摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Liquid crystalline surfactant phases in chemical applications Thomas Engels andWolfgang von Rybinski Henkel KGaA, Henkelstrasse 67, D-40191 Du�sseldorf, Germany The importance of lyotropic liquid crystalline structures is shown for two selected surfactant applications, i.e. cosmetics and detergency. The properties of lyotropic liquid crystals are demonstrated for binary surfactant–water systems, ternary surfactant–oil–water systems and multicomponent systems.Detailed knowledge of the phase behavior is crucial for tailor-made product development. Furthermore cosmetic products with definite viscosities are 1 Introduction able to incorporate additives which at much lower viscosities Liquid crystalline structures have received a good deal of would segregate out of the product, i.e.pearlescent agents in attention in recent years. Besides being used in the scientific hair shampoos. study of cooperative phenomena and complex fluid phases In ternary systems composed of water, surfactants and oils these structures have found applications in electrooptic dis- or hydrocarbons the presence of liquid crystals has a major plays, sensors, optical switches and shutters, and thermogra- impact on system properties such as microscopic structure, phy.The liquid crystalline structures formed by amphiphilic viscosity, stability and foaming performance. molecules form the basis for emulsions and have been studied thoroughly by researchers in the food, drug, oil and chemical Stabilization of foams industries.The foaming ability of, for example, oil continuous solutions Two diVerent types of structures have been distinguished: is dependent on the phase behavior of the system.1 On the thermotropic liquid crystals and lyotropic liquid crystals. surfactant and hydrocarbon rich side of the phase diagram Although lyotropic phases have not been given the same (Fig. 1, black area) two diVerent phases can be distinguished. prominence as thermotropic phases, their importance should In the L-phase (w/o emulsion) as well as the LC phase not be underestimated. The phases are crucial in the manufac- ( liquid crystal) foam lifetimes were only of the order of seconds. ture and mode of operation of detergents and have an import- On the other hand, in the two-phase region (hatched area) ant role in cosmetics.In lyotropic systems, the transition from foam lifetimes increased up to several hours indicating that one phase to another can occur owing to the change of the presence of a liquid crystal enhances and stabilizes the concentration. Of course temperature can also cause phase foam. This stabilization eVect can be attributed to the influence transitions in these systems, so this aspect of thermotropic of the liquid crystal on the foam drainage.Microscopic pictures liquid crystals is shared by lyotropics. The real distinctiveness using polarized light2 reveal that the liquid crystals concentrate of lyotropic liquid crystals is the fact that at least two diVerent in the plateau borders of the foam where its high viscosity species of molecules (e.g.solute and solvent) must be present reduces the drainage of the foam considerably. For this reason for these structures to form. the plateau borders and the radii of its curvature remain large The following review will summarize basic principles of causing the Laplace pressures and the thinning rates of the lyotropic liquid crystalline phases using specific applications foam lamellae to be small.in cosmetics and detergency. Of course the results can be In addition, the surfactants in the liquid crystalline phase transferred to other applications. result in lower surface tensions and hence higher surface pressures in the foam lamellae compared to the situation where only surfactants are present. In this respect the liquid crystals 2 Cosmetic applications In the field of cosmetics, surfactants are widely used as wetting agents, emulsifiers and/or stabilizers.Cosmetic products can be regarded as belonging to either one of the two following categories: surfactant+solvent (water or oil ) or surfactant+ oil+solvent. In most of the cases cosmetic products are based on water being the solvent and lyotropic liquid crystals as the predominating structures providing the specific properties of the product.Typical cosmetic products of the first category are hair and body shampoos or shower gels, facial cleansers and toothpastes. All these products have in common the fact that the presence of liquid-crystalline surfactant phases is induced in order to provide appropriate rheological properties.In particular, lamellar liquid crystals are preferred since their bulk viscosities are not as high as those of hexagonal or cubic liquid crystals and the shear thinning eVect in these systems is convenient during application. Of course other means to adjust product Fig. 1 Phase diagram of water–hydrocarbon–oil-soluble surfactant rheology are common, e.g.addition of polymers or salt-induced forming foams only in the two-phase area L+LC (reproduced with permission from ref. 1; for explanation of abbreviations see text) micelle-to-rod transitions for low concentrated formulations. J. Mater. Chem., 1998, 8(6), 1313–1320 1313in the plateau border serve as a surfactant reservoir of optimum composition for the stabilization of the foam films.Stabilization of emulsions and dispersions The application of lamellar liquid crystalline phases for the stabilization of emulsions has been reported by Friberg et al.3 Emulsions or dispersions can be further stabilized by crystallization of lamellar liquid crystalline phases located at the water/oil interface resulting in the formation of a so-called gel phase.4–6 We have demonstrated6 that changes in the type and concentration of the co-emulsifier caused the formation of a lamellar gel phase surrounding the oil droplets inside an oil-in-water (o/w) emulsion which increased the stability of the emulsion from 5 days (emulsion 1) to more than 1 month (emulsion 2; see Table 1 and Fig. 2).Great care was taken to make sure that the physico-chemical properties of the system remained unaVected by the change of the co-emulsifier.In Table 1 the viscosities and the phase inversion temperatures (PIT) are compiled together with the particle size distributions of the two emulsions. The data suggest that only the presence of the lamellar liquid crystalline phase causes the observed increase in emulsion stability. The viscous lamellar film surrounding the emulsion droplets may be several layers thick and reduces the attraction potential between the droplets [see Fig. 2(c)].7,8 As a result, the lamellar layer acts as a barrier against coalescence. Friberg9 distinguished between emulsions or two-phase systems being stabilized by a monomolecular surfactant layer at the interface between oil and water and three-phase emulsions with regular structures of multimolecular layers that can be regarded as a distinct phase which can exist independently from the emulsion, e.g.after separation by centrifugation. Emulsions containing a third phase or multimolecular layers of lyotropic liquid crystals are often found in cosmetic products (Fig. 3) and is therefore widely used by the cosmetics industry to adjust or optimize specific properties of the product such as viscosity or consistency, storage stability or application convenience.Fig. 3 shows a conventional and a polarization microscope picture of a commercial hand lotion with a liquid crystal present at the interface between the oil droplets and the continuous water phase. The liquid crystal is assumed to consist of multiple layers or shells stabilizing the dispersed phase (see Fig. 4). These kinds of three-phase emulsions can be found in cosmetic products ranging from lotions to creamy emulsions Fig. 2 Influence of co-emulsifier type on emulsion stability. Ultrasonic density scans of the vessel containing emulsions 1 (a) and 2 (b); (c) TEM Table 1 PIT emulsion systems (% by mass of active substance) image of emulsion 2 containing liquid-crystalline gel phases stabilizing the oil droplets (reproducewith permission from ref. 6). emulsion final concentrate formulation and are believed also to exist in food products in which lecithin component/property 1 2 1 2 and monoglycerides are commonly used as emulsifiers.10,11 C16/18E12 5.7 8.0 2.7 3.8 Networks formed by liquid crystals glyceryl monostearate — 5.0 — 2.4 C12/14E4 a 7.3 — 3.5 — Similar to the eVect of polymer additives, liquid crystalline dicapryl ether 42.0 42.0 20.0 20.0 phases are capable of forming three-dimensional networks water 45.0 45.0 73.8 73.8 extending through the continuous phase of the system.Like PIT/°C 73.4 74.0 — — polymer thickeners the network reduces the Brownian motion viscosity/mPa 300 300 30 36 of the dispersed particles or droplets and thus add to the mean particle sizeb/nm stability of the system (see Fig. 5). fresh — — 350 305 In the field of pharmacology rigid network structures made aged — — 3500 450 up of liquid crystalline structures may dissolve substances that mean particle sizec/mm otherwise show only limited solubility. Wahlgren et al.13 demfresh — — <1 <1 onstrated that the solubility of hydrocortisone in isotropic aged — — 2.7 <1 solvents is small, ca. 1.5% in ethylene glycol, whereas in a lamellar liquid crystalline phase of lecithin and water the aNarrow range ethoxylate. bZetasizer, Malvern. cBy light microscopy (from ref. 6). solubility exceeds 4%. 1314 J. Mater. Chem., 1998, 8(6), 1313–1320Fig. 3 Liquid crystals in a hand lotion visualized using polarized light microscopy (reproduced with permission from ref. 9) Fig. 5 Three-dimensional network structure of oil/liquid crystal/water emulsions (reproduced with permission from ref. 12) the dispersed and continuous phase which requires the energy Ac (internal surface A×interfacial tension co/w) and thus renders the emulsion thermodynamically metastable or even unstable.The addition of surfactants which reduce the interfacial tension co/w, or the presence of liquid crystals which prevent the phase separation, i.e. the decrease of the total energy of the system towards the thermodynamic equilibrium, by modifying the rheological properties of the system, are the two main concepts to stabilize emulsions usually of either oil-in-water (o/w) or water-in-oil (w/o) types.However, for cosmetic and technical applications, microemulsions with even higher interfacial areas, but being thermodynamically stable, Fig. 4 Electron microscope image of freeze-fracture of liquid crystalline and multiphase emulsions, e.g. of the w/o/w type, are quite layers in an emulsion (reproduced with permission from ref. 9) common too.In the case of microemulsions the considerable amount of positive entropy of mixing over-compensates the Liposomes amount of interfacial energy required to form the microemulsion which on the other hand has been minimized because Liposomes are supramolecular structures or vesicles of mono- (uni-) or multilamellar surfactant bilayers with a hollow core of an extreme reduction of the interfacial tension due to an optimal match of emulsifier and oil properties.which can be regarded as an intermediate structure between micelles (surfactant monolayer on the surface of an imaginary The stabilization eVects of surfactants on emulsions are due to their amphiphilic molecular structure with hydrophilic and sphere) and lamellar liquid crystals (plane bilayers). In a way liposomes can also be regarded as lamellar liquid bilayers lipophilic moieties.The tendency to accumulate at interfaces resulting in the reduction of interfacial energies and to aggre- which have transformed by bending and have fused to a closed spherical shell. They are usually discussed in terms of controlled gate in aqueous solutions forming thermodynamically stable micellar structures or lyotropic liquid crystals oVers possibil- release and specific drug targeting in the fields of pharmacology, cosmetics, food industry and agrochemicals.14–16 ities for changing the macroscopic appearance and properties of emulsions.With respect to controlled drug release a delayed delivery because of low diVusion coeYcients for the solubilized drugs The presence of liquid crystalline structures during the formation or production of technical or cosmetic emulsions is via vesicles or liposomes incorporating, for example, insulin is discussed.17,18 of paramount importance for the PIT, as well as the gel-phase emulsification.Specific drug targeting is aimed at by designing vesicles binding to cells with high selectivities and aYnity; e.g. immunoliposomes bearing ligands (e.g.antibodies) that are PIT- emulsification recognized by specific cell receptors.19,20 The above mentioned optimized balance between the amphiphilic properties of the emulsifier system with the hydro- Liquid crystals in emulsification processes phobicity of the oil phase and the hydrophilicity of the water phase, being modified by the presence of salts or water soluble The mixing of at least two mutually insoluble liquid phases is usually referred to as emulsification or dispersion.21 The pro- components of the oil phase, represents the essential part of the PIT emulsification process resulting in the extreme fine cess itself involves the creation of large interfacial areas between J.Mater. Chem., 1998, 8(6), 1313–1320 1315dispersity of the emulsion which is necessary, for example, for the required long-term stability during storage.The phase-inversion temperature (PIT) emulsification has been developed for emulsions based on ethoxylated fatty alcohol nonionic surfactants and requires a heating–cooling cycle22,23 (see Fig. 6). Starting at ambient temperatures with a coarsely dispersed o/w emulsion upon temperature increase the system passes through a microemulsion or bicontinuous lamellar phase at the PIT resulting in the formation of a w/o emulsion at temperatures exceeding the PIT-range.The temperature induced inversion eVect is related to the clouding phenomenon of the ethoxylated nonionic surfactants which change their solubilization properties with increasing temperature.The clouding phenomenon finally leads to a separation of the system into two phases, one of which is a dilute surfactant phase whereas the other phase consists of a concentrated micellar solution in thermodynamic equilibrium with the dilute phase. The eVect is ascribed to a temperature dependent change of the hydration of the ethylene oxide groups of the hydrophilic headgroup of the nonionic surfactant.At higher temperatures the intermolecular interactions between water and the surfac- Fig. 7 Principle of the gel-phase emulsification (reproduced with per- tant head group are less energetically favorable than the mission from ref. 25): 1, formation of an o/lc gel emulsion containing interaction between water or the nonionic molecules oil (o), monoarginine hexyl decyl phosphate (R6R10MP-1Arg), glycerol themselves.and a small amount of water (w) ( lc=lamellar phase); 2, dilution of During the PIT-process upon reduction of the temperature the o/lc emulsion with cold water the system re-inverts from the w/o emulsion to an o/w emulsion causing a simultaneous break-up of the planar or bicontinuous (monoarginine hexyldecyl phosphate), polyol (glycerol) and a intermediate phase into a multitude of spherical oily droplets small amount of water.Oil phase (o) and lamellar phase ( lc) of extreme fine dispersity with only minimal mechanical energy together form a transparent o/lc gel phase. In contrast to the required. The extreme fine dispersity is the reason why PITPIT- process the gel-phase emulsification does not require a emulsions exhibit much longer shelflives than emulsions with definite type of surfactant.The surfactant may be anionic more coarsely distributed particle sizes. or nonionic. In the second step the o/lc gel is diluted with a definite Gel-phase emulsification amount of water to form the final o/lc/w emulsion. In this emulsion the presence of the lamellar liquid crystalline phase Another way to produce finely dispersed emulsions associated protects the oil droplets against coalescence ( like in ref. 6). with a change in the hydrophilic/lipophilic balance is called The particle size distribution of the dispersed oil droplets is gel-phase emulsification (Fig. 7). In this two-stage process a primarily dependent on the surfactant/oil ratio and can be microemulsion or a lamellar gel phase is induced through adjusted to cover the whole range from less than 100 nm to an additional component which modifies the hydrophilic/ even more than 1000 nm.Thus together with the PIT-process lipophilic properties of the solvent, e.g. addition of a polyol to the gel-phase emulsification oVers the opportunity to produce water.24–26 By proper choice of this additive and the surfactant finely dispersed o/w emulsions for the use in the cosmetics as well as optimization of their relative ratio the gel-phase industry in a simple and cost-saving way which satisfies the emulsification can be adapted to diVerent oils.The whole fundamental requirements of long shelflives and stability process can be operated at a definite temperature, e.g.at room against coalescence. temperature. In the first step the oil phase is dispersed in a lamellar phase. Structure of the epidermis In this particular case the lamellar phase consists of surfactant So far liquid crystalline structures have been discussed merely in the sense that they aVect the structure and stability of oil and/or water-based products. But apart from that they are an essential part of the human epidermis (Fig. 8).27–29 During the cell diVerentiation from the lower parts of the epidermis (stratum basale) to the upper parts (stratum corneum) the chemical composition of the lipids of the skin change. The stratum basale is mainly composed of phospholipids and small amounts of cholesterol, ceramides and fatty acids (malpighian cells).In the granular cells of the intermediate stratum granulosum nearly equal amounts of phospholipids, cholesterol, ceramides and fatty acids can be found. The formation of lamellar liquid crystalline structures in the human skin requires a minimum amount of fatty acids. Together with the skin lipids these lamellar structures are stored in the granules of the stratum granulosum.Between the horny cells of the stratum corneum these granules fuse to form lamellar lipid membranes Fig. 6 Principle of the PIT method: an o/w emulsion changes into a and thus constitute an epidermal barrier against water loss. w/o emulsion through increasing temperature. In the phase inversion The skin lipids of the stratum corneum are mainly composed range, a microemulsion or lamellar bicontinuous phase develops which of fatty acids, cholesterol and ceramides which together deter- becomes a blueish o/w emulsion after cooling (reproduced with permission from ref. 21).mine the properties of this water barrier. Knowledge of the 1316 J. Mater. Chem., 1998, 8(6), 1313–1320Fig. 9 Phase diagram of the binary system water–pentaoxyethylene n- Fig. 8 Epidermal structure of the human skin.Schematic represen- dodecanol (C12E5) (reproduced with permission from ref. 33) tation of the horny layer (a) and the lipid barrier (b, c). Lipids are formed during cell diVerentiation and stored in granula as disks, which this boundary with a mesophase region may result, as depicted are released into the intracellular space during transfer to the horny in Fig. 9. At low surfactant concentrations in such systems, layer (b). There they fuse into ‘endless’ lipid membranes (c) which are several two-phase areas are observed in addition to the single- water impermeable. (Reproduced with permission from ref. 30.) phase isotropic L1 range: two coexisting liquid phases (W+L1), a dispersion of liquid crystals (W+La) and a two-phase region structural and chemical composition of the human epidermis of water and a surfactant liquid phase (W+L2).can be taken advantage of in cosmetic products. The repair of In ternary mixtures of water, surfactant and oil, three phases the lamellar structures in the epidermis or the supply of may coexist in equilibrium. These systems are also referred to essential fatty acids or ceramides should result in a well as three-phase microemulsions. When these three phases are developed skin barrier.As a consequence the reduction in formed, extremely low interfacial tensions between two phases trans-epidermal water loss increases the water content of the are observed. Because the interfacial tension is generally the upper skin and wrinkle formation should be significantly restraining force, with respect to the removal of liquid soil in reduced.the washing and cleaning process, it should be as low as possible for optimal soil removal. Other quantities such as the wetting energy and the contact angle on polyester, as well as 3 Lyotropic liquid crystals in detergency the emulsifying ability of e.g. olive oil, also show optima at Detergent performance the same mixing ratio at which the minimum interfacial tension is observed.39 Nonionic surfactants of the alkylpolyglycol ether type are key Fig. 10 (right) represents the three-phase temperature interingredients of detergent formulations because of their good vals for C12E4 and C12E5 vs. the number (n) of carbon atoms detergency properties.31 The interfacial and colloidal properof n-alkanes.40 Both parts of Fig. 10 indicate that the maximum ties of alkylpolyglycol ethers have been the subject of numeroil removal is in the three-phase interval of the oil used ous publications. In particular, the phase behavior of (n-hexadecane). This means that not only the solubilization binary mixtures of water and nonionic surfactants has capacity of the concentrated surfactant phase, but probably been intensively investigated.32,33 Besides molecular, micellar also the minimum interfacial tension existing in the range of and inverse-micellar solutions, single and two-phase liquid the three-phase body are responsible for the maximum oil crystal regions, as well as ‘anomalous’ phases, have been removal.observed. Apart from these binary mixtures, ternary systems of the water/surfactant/oil type have also been studied.These mixtures may form three-phase microemulsions that are of interest with regard to special applications, since extremely low interfacial tensions exist between the single phases.34 Generally, interfacial tension is the restraining force regarding removal of liquid oil from a solid surface35 and in the displacement of oil from narrow capillaries, as is the case with enhanced oil recovery.At the surfactant solution/oil or the surfactant solution/fat interface, liquid crystals may be formed by penetration of surfactant molecules. Model experiments have shown that the formation of mesophases strongly aVects detergency.36–38 Beside this eVect liquid crystalline phases are of equal importance for the formulation of liquid products.The phase behavior of nonionic surfactants with a low Fig. 10 Phase behavior of the polyoxyethylene alcohols C12E4 and degree of ethoxylation is very complex. As the lower cosolute C12E5, and detergency:40 three-phase temperatures versus number n of boundary is shifted to lower temperatures with decreasing EO carbon atoms (right) and detergency (R) as a function of temperature ( left) (reproduced with permission from ref. 42) (ethylene oxide) number of the molecule, an overlapping of J. Mater. Chem., 1998, 8(6), 1313–1320 1317Tests with pure low ethoxylated surfactants have shown that a discontinuity is observed with respect to oil removal versus temperature in cases of the existence of dispersions of liquid crystals in the binary system water/surfactant.Fig. 11 shows that the detergency values for mineral oil and olive oil, i.e. two oils with significantly diVerent polarities, are at diVerent levels. It also shows that in both cases a similar reflectance vs. temperature curve exists. In the region of the liquid crystal dispersion, i.e. between 20 °C and 40 °C, the oil removal increases significantly.Above the phase transition W+La � W+L3, between 40 °C and 70 °C, no further increase in oil removal takes place. For olive oil, a small decrease in detergent performance is served. The interfacial tensions between aqueous solutions of C12E3 and mineral oil lie at about 5 mNm-1 at 30 °C and 50 °C. These relatively high values indicate that in this system the interfacial activity is not the decisive factor in oil removal from fabrics.Fig. 12 Phase behavior of the polyoxyethylene alcohols C12/18E4 and In the case of C10E4, which is substantially more hydrophilic C12/18E5 and detergency. The line above the number of the oxyethylene than C12E3 (Fig. 11), a phase transition of the two co-existing groups indicates that it is a mean value (reproduced with permission liquid phases into a liquid crystal dispersion takes place in the from ref. 42). temperature range investigated. Here, too, the reflectance greatly increases above the phase transition temperature in the liquid phases (W+L1) and for C12–18E49 it is in the range of region where the liquid crystal dispersion exists. Whereas an the surfactant liquid phase (W+L2). An unusually strong interfacial tension of 0.3 mN m-1 occurs at 40 °C, a value increase of oil removal with increasing temperature occurs in approximately ten times higher is observed at 60 °C.Thus, the the region of the liquid crystal dispersion (W+La). At 30 °C interfacial tension increases with increasing temperature in this and 50 °C, the interfacial tensions between aqueous surfactant ternary mixture.The strong increase of reflectance cannot, solutions and mineral oil and the contact angle on glass and therefore, be attributed to the increase of the amount of polyester were determined for C12–18E49 . Whereas the values of surfactants adsorbed, which would manifest itself in a decrease the interfacial tensions are practically identical (approximately in the interfacial tension. 10-1 mN m-1 after 15 min), the contact angles on both In both the examples, neither the position of the cloud point substrates are slightly less advantageous at higher temperature. nor the existence of a three-phase body are responsible for the Hence, the increased oil removal between 30 °C and 50 °C strong temperature dependence of oil removal. Rather, the cannot be attributed to an increase in the adsorbed amounts macroscopic properties of the liquid crystal dispersion seem to of surfactants. Rather, in both cases, the decisive part is be responsible for the strong temperature dependence.It can probably played by the macroscopic properties of the liquid be assumed that fragments of liquid crystals are adsorbed onto crystal dispersion and their temperature dependence.fabric and oily soil in the W+La range during washing. The local surfactant concentration is therefore substantially higher Flow behavior in comparison to the monomolecular surfactant layer that forms when surfactant monomers adsorb. As the viscosity of Besides detergency the liquid crystalline phases of surfactant liquid crystals in the single-phase range is strongly temperature systems at higher concentrations are of crucial importance for dependent,41 it can be assumed that the viscosity of a fragment the processing of concentrated surfactant systems and the of a liquid crystal deposited on a fabric also significantly formulation and application of liquid products.In such cases decreases with increasing temperature. Thus, the penetration not only the phase behavior but also the rheological properties of surfactant into the oil phase and removal of oily soil are are of interest for the user.Characteristic properties of promoted. surfactant systems will be shown in the following examples for Apart from pure nonionic surfactants, technical grade surfac- diVerent types of nonionic surfactants. tants are of specific interest for applications.As in the case of Alkyl polyglycosides form a new class of surfactant, which pure nonionic surfactants, definite ranges exist in which there show favourable physical chemical properties.43 The phase is only a slight dependence of oil removal on the temperature behavior of a technical C8/10-alkyl polyglycoside (C8/10-APG) (Fig. 12). For C12–18E5 9 , this is in the range of the two co-existing is illustrated in Fig. 13.44 At temperatures above 20 °C, the C8/10-APG is present up to very high concentrations in an isotropic phase, the viscosity of which increases considerably. Fig. 11 Phase behavior of the polyoxyethylene alcohol C12E3 and Fig. 13 Phase diagram of the C8/10-APG/water system (2W=two- detergency (2 g l-1 surfactant) (reproduced with permission from ref. 42) phase region) (reproduced with permission from ref. 44) 1318 J. Mater. Chem., 1998, 8(6), 1313–1320A birefringent lyotropic phase of nematic texture is formed at the measurements are repeated, the yield value on the ascending curve corresponds to that on the first descending curve. This around 95% by mass, which changes at around 98% by mass into a cloudy two-phase region of liquid and solid alkyl suggests that an orientation of the anisotropic phases took place in the shear field when they were first subjected to shear polyglycoside.At relatively low temperatures, a lamellar liquid crystalline phase is additionally observed between 75 and 85% stress. The lower yield value obtained when the measurements were repeated corresponds to that of the aligned sample.by mass. The phase diagram of the C12/14-alkyl polyglycoside (C12/14- Fig. 16 shows the yield value given by the ascending and descending curves for C10E4 as a function of the surfactant APG)/water system (Fig. 14) diVers clearly from that of the short-chain APG. At low temperatures, a region resembling a concentration. When first subjected to shear stress, the samples gave yield values up to 80 Pa; when the measurements were solid/liquid system below the KraVt point is formed over a wide concentration range. With an increase in temperature, repeated, these values sank below 20 Pa.The yield value increases with the surfactant content of the lamellar phase, the system changes into an isotropic liquid phase. Since crystallization is kinetically retarded to a considerable extent, only to decrease rapidly in the two-phase region La+L2.Before the first measurements are taken, the orientation of the this phase boundary changes position with the storage time. At low concentrations, the isotropic liquid phase changes anisotropic phases is not defined. Therefore, their rheological properties vary distinctly.This is not the case with the oriented above 35 °C into a two-phase region of two liquid phases, as is normally observed with nonionic surfactants.45 At concen- samples; there is scarcely any recognizable concentration dependence; the yield values are around 10 Pa. trations above 60% by mass, a sequence of liquid crystalline phases is formed at all temperatures examined.It is important Similar results are obtained for the shear viscosity as a function of concentration. For the freshly produced samples to mention that, in the isotropic single-phase region, a distinct streaming birefringence can be observed at concentrations just not previously subjected to shear stress, the viscosity increases systematically as the concentration of the surfactant is below the lyotropic phases, disappearing again rapidly on completion of the shearing process.However, no multiphase increased. After being subjected to shear stress, the samples show distinctly smaller variations in viscosity when the regions separating this region from the L1 phase could be found. In the dilute L1 phase, there is another region with measurements are repeated.This also suggests that the viscous flow of the oriented samples occurs by sliding of layers whose weaker streaming birefringence which is situated near the minimum of the liquid/liquid miscibility gap. The rheological properties are barely aVected by the concentration. Fig. 17 shows the yield values of samples of the same properties of highly concentrated alkyl polyglycoside systems are summarized in ref. 42. composition (10 mol%), obtained from the ascending curve, as a function of EO and alkyl chain length. As the EO number Paasch et al. studied the rheological properties of lamellar liquid crystalline phases of alkyl polyethylene glycol ethers.46 increases, the yield value increases by about 10 Pa in the case of C12 ethoxylates and about 5 Pa in the case of C10 surfactants.The rheological measurements of lamellar liquid crystals indicate that the phases studied are, without exception, plastic The value for the C10-polyethylene glycol ethers is higher than systems. As an example, flow curves obtained from measurements of C12E4, 60% by mass, are reproduced in Fig. 15. The first measurement on the ascending curve indicates a clearly higher yield value than is given by the descending curve.When Fig. 16 Yield values of the polyoxyethylene alcohol (C104EO)–water (w) system (up=ascending curve, down=descending curve) (reproduced with permission from ref. 46) Fig. 14 Phase diagram of the C12/14-APG/water system (a, b and c indicate crystalline precipitates or liquid crystalline phases) (reproduced with permission from ref. 44) Fig. 17 Yield values given by the ascending curves as a function of Fig. 15 Flow curves of the polyoxyethylene alcohol C12E4 (60 the length of the alkyl chain and the number of oxyethylene groups n (EO) (concentration=10 mol%) (reproduced with permission from mass%)–water system: (a) first measurement; (b) repeat measurement (reproduced with permission from ref. 46) ref. 46) J. Mater. Chem., 1998, 8(6), 1313–1320 131914 R. R. C. New, L iposomes—a practical approach, Oxford University the values of the C12-ethoxylates; this is probably a result of Press, New York, 1990. the higher proportion of EO chains in the molecule. The same 15 J. R. Philippot and F. Schuber, L iposomes as tools in basic research relationship holds true at diVerent concentrations. and industry, CRC Press, Boca Raton, FL, 1995.In contrast, the yield values of the oriented samples at a 16 D. R. Karsa and R. A. Stephenson, Encapsulation and controlled surfactant concentration of 10 mol% show no dependence on release, Royal Society of Chemistry, Cambridge, 1993. 17 S. Ng and S. Frank, J. Dispersion Sci. T echnol., 1982, 3, 217. the type of surfactant. The liquid crystals formed from the 18 D.Kavaliunas and S. Frank, J. Colloid Interface Sci., 1978, 66, 586. various surfactants all have roughly the same value of approxi- 19 P. Machy and L. Leserman, L iposomes in cell biology and pharma- mately 10 Pa. This is surprising, since the lengths of the cology, John Libbey, London, 1987. polyethylene glycol and alkyl chains apparently have no 20 G. Gregoriadis, L iposome T echnology, CRC Press, Boca Raton, influence on the plastic properties of the oriented liquid FL, 1992, 2nd edn., vol.III. crystals. These aspects must be considered for the handling of 21 T. Fo� rster, W. von Rybinski and A. Wadle, Adv. Colloid Interface Sci., 1995, 58, 119. such fluids in technological processes and consumer products. 22 T. Fo� rster, F.Schambil and H. Tesmann, Int. J. Cosmetic Sci., 1990, 12, 217. 4 Outlook 23 T. Fo� rster, F. Schambil and W. von Rybinski, J. Dispersion Sci. T echnol., 1992, 13, 183. The importance of liquid crystalline structures has been shown 24 H. Sagitani, J. Dispersion Sci. T echnol., 1988, 9, 115. for two selected surfactant applications, i.e. cosmetics and 25 T. Suzuki, H. Takei and S.Yamazaki, J. Colloid Interface Sci., 1989, detergency. The demonstrated basic principles of liquid crystals 129, 491. 26 T. Suzuki, M. Nakamura, H. Sumida and A. Shigeta, J. Soc. which are needed for the optimization of product properties Cosmet. Chem., 1992, 43, 21. are of course valid in other application fields as well. For each 27 S. Friberg, J. Soc. Cosmet. Chem., 1990, 41, 155.application where surfactants are part of a product detailed 28 S. Friberg and L. L. Rhein, J. Dispersion Sci. T echnol., 1988, 9, 371. knowledge has to be gathered about how to cope with the 29 S. Friberg and D. W. Osborne, J. Dispersion Sci. T echnol., 1985, special properties of liquid crystals. 6, 485. The generation, incorporation or modification of liquid 30 U.Zeidler, Skin Care Forum, 1992, 3. 31 H. Andree and P. Krings, in Waschmittelchemie, Hu� thig, crystalline structures are the daily challenges of product devel- Heidelberg, 1976, p. 84. opment in order to optimize the products. Further eVorts are 32 J. C. Lang and R. D. Morgan, J. Chem. Phys., 1980, 73, 5849. still needed to transfer and apply fundamental knowledge to 33 D. J. Mitchell, G.J. T. Tiddy, L. Warring, T. Bostock and more application oriented problems concerning the tailoring M. P. McDonald, J. Chem. Soc., Faraday T rans. 1, 1983, 79, 975. of liquid crystal-based products to specific market or customer 34 M. Kahlweit and R. Strey, Angew. Chem., Int. Ed. Engl., 1985, demands. 24, 654. 35 W. Kling and H. Lange, J. Am. Oil Chem. Soc., 1960, 37, 30. 36 H. S. Kielman and P. J. F. van Steen, Surface Active Agents, References London Society of Chemical Industries, 1979, p. 191. 37 C. A. Miller and K. H. Raney, Colloids Surf. A: Physicochem. Eng. 1 S. Friberg, Advances in L iquid Crystals, Academic Press, New Aspects, 1993, 74, 169. York, 1978, vol. 3, p. 149. 38 J. C. van de Pas, C. J. Buytenhek and L. F. Brouwn, Rec. T rav. 2 H. Saito and S.Friberg, in L iquid Crystals, ed. S. Chandrasekhar, Chim. Pays-Bas, 1994, 113, 231. Indiana Academy of Science, 1975. 39 M. J. Schwuger, in Structure/Performance Relationships in 3 S. Friberg, L. Mandell and M. Larsson, J. Colloid Interface Sci., Surfactants, ACS Symp. Ser. 253, ed. M. J. Rosen,Washington DC, 1969, 29, 155. 1984, p. 3. 4 K. Larsson, Prog. Chem. Fats L ipids, 1978, 16, 163. 40 M. Kahlweit and R. Strey, in Proceeding of Vth International 5 K. Shinoda and S. Friberg, Emulsions and solubilization, Wiley, Conference on Surface and Colloid Science, ed. H. L. Rosano, New York, 1986. Potsdam, New York, 1985. 6 T. Engels, T. Fo� rster and W. von Rybinski, Colloids Surf. A: 41 J. Munoz, C. Gallegos and V. Flores, J. Dispersion Sci. T echnol., Physicochem. Eng. Aspects, 1995, 99, 141. 1986, 7, 453. 7 S. Friberg and M. A. El-Nokaly, in Surfactants in Cosmetics, 42 F. Schambil and M. J. Schwuger, Colloid Polymer Sci., 1987, 165, ed. M. M. Rieger, Marcel Dekker, New York, 1985, p. 55. 1009. 8 S. Friberg and C. Solans, L angmuir, 1986, 2, 121. 43 K. Hill, W. von Rybinski and G. Stoll, Alkyl Polyglycosides— 9 S. Friberg, J. Soc. Cosmet. Chem., 1979, 30, 309. T echnology, Properties and Applications, VCH,Weinheim, 1997. 10 N. Krog and J. B. Lauridsen, in Food Emulsions, ed. S. Friberg, 44 G. Platz, J. Po� licke, Chr. Thunig, R. Hofmann, D. Nickel and W. Marcel Dekker, New York, 1976, p. 67. von Rybinski, L angmuir, 1995, 11, 4250. 11 S. Friberg, P. O. Jansson and E. Cederberg, J. Colloid Interface 45 R. A. Mackay, in Nonionic Surfactants; Physical Chemistry, ed. Sci., 1976, 55, 614. M. J. Schick, Marcel Dekker, New York, 1987, p. 297. 12 T. Suzuki, H. Tsutsumi and A. Ishida, J. Dispersion Sci. T echnol., 46 S. Paasch, F. Schambil and M. J. Schwuger, L angmuir, 1989, 5, 1984, 5, 119; T. Forster, in Surfactants in Cosmetics, ed. 1344. M. M. Rieger and L. D. Rhein, M. Dekker, New York, 1997, p. 105. 13 S. Wahlgren, A. L. Lindstrom and S. Friberg, J. Pharm. Sci., 1984, 73, 1484. Peper 7/06141B; Received 21st August, 1997 1320 J. Mater. Chem., 1998, 8(6), 131

ISSN:0959-9428    

DOI:10.1039/a706141b

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 The selenium analogue of DOET and its conducting salts Hiroyuki Nishikawa,*a† Hanako Ishikawa,a Tatsuo Sato,a Takeshi Kodama,a Isao Ikemoto, a Koichi Kikuchi,*a Satoru Tanaka,b Hiroyuki Anzaib and Jun-ichi Yamada*b‡ aDepartment of Chemistry, T okyo Metropolitan University, Hachioji, T okyo 192-0397, Japan bDepartment of Material Science, Faculty of Science, Himeji Institute of T echnology, 1475-2 Kanaji, Kamigori-cho, Ako-gun, Hyogo 678-1297, Japan Table 1 Conducting behaviour of the DOES salts Synthesis and electrochemical properties of the title donor anion solvent D5Aa srt/S cm-1 b (DOES), electrical conductivities of its radical-cation salts, and crystal structure of (DOES)2(AuI2)0.75 are described.AsF6- PhCl 552 5.1 (Ea=52 meV) PF6- PhCl 251 0.62 (Ea=74 meV) ClO4- PhCl 251 2.3 (Ea=3.7 meV) BF4- PhCl 352 5.0 (Ea=34 meV) I3- PhCl 251 60(TM–I d=250 K) AuI2- PhCl 250.75c 1.4 (TM–I d=55 K) We have already disclosed the crystal structure of the metallic (DOET)2BF4 [DOET=(1,4-dioxanediyl-2,3-dithio)ethylene- aDetermined by elemental analysis.bRoom temperature conductivity dithiotetrathiafulvalene] salt.1 Despite the fact that the 1,4- measured by a four-probe technique on a single crystal.cDetermined dioxane ring of DOET is condensed by cis fusion,2 the DOET by the X-ray analysis. dTemperature of a metal-to-insulator transition. molecules in this salt form the so-called b-type packing arrangement, and the band structure calculation of this salt reveals nearly two-dimensional character.One interesting modification of the DOET molecule, for inducing appreciable change in the crystal structure, might be to replace the outer sulfur atoms with selenium atoms. In this communication, we report the synthesis and electrochemical properties of the selenium analogue of DOET, viz. DOES [(1,4-dioxane-2,3-diyldithio)ethylenediselenotetrathiafulvalene], and the conducting behaviour of its radical-cation salts.Furthermore, the crystal structure of a metallic (DOES)2(AuI2)0.75 salt is described. Synthesis of DOES§ was carried out by cross-coupling between ketone 13 and 2 equiv. of thione 2¶ in the presence of (MeO)3P in benzene under reflux for 1 h (27% yield). Fig. 1 Temperature dependence of the resistivity of the AuI2 and I3 salts of DOES DOES formed a 151 charge-transfer complex with TCNQ (tetracyanoquinodimethane), the room temperature conductivity of which was rather low (srt<10-6 S cm-1).Thus, we examined the preparation of the radical-cation salts by controlled- current electrochemical oxidation4 in chlorobenzene containing the corresponding tetrabutylammonium salt. As summarized in Table 1, the salts with octahedral and tetra- O O S S S S X X S S O O S S S S O S Se Se S S 1 2 DOET; X = S DOES; X = Se hedral anions (AsF6-, PF6 -, ClO4-, and BF4-) exhibited semiconducting behaviour with activation energies ranging The electrochemical properties of DOES were investigated from 3.7 meV to 74 meV.On the other hand, the salts with by cyclic voltammetry in benzonitrile (0.1 M Bun4NClO4, Pt linear anions (I3- and AuI2-) displayed metallic behaviour.electrode, room temp., scan rate 50 mV s-1). DOES showed The temperature dependence of their resistivities is shown in two pairs of reversible redox waves, and its E1 and E2 values Fig. 1. The resistivity of the AuI2 salt shows metallic tempera- (+0.54, +0.84 V vs. SCE) are slightly lower than those of ture dependence to 83 K, and gradually increases until 55 K DOET (E1=+0.58 V, E2=+0.88 V).where a metal-to-insulator transition occurs.|| While the resistivity of the I3 salt increases slightly with decreasing tempera- †E-mail: hiron@comp.metro-u.ac.jp. ture around room temperature, its temperature dependence is ‡E-mail: yamada@sci.himeji-tech.ac.jp. §Physical and spectroscopic data for DOES: orange cryst.; mp very small and a clear transition to insulator takes place 168–171 °C (decomp.); MS (m/z) 538 (M++2), 536 (M+); 1H NMR at 250 K.|| (CDCl3) d 3.25 (s, 2 H), 3.40 (m, 2 H), 3.63 (m, 2 H), 3.99 (m, 2 H), The X-ray crystal analysis of the AuI2- salt of DOES** 5.31 (s, 2 H).The preliminary X-ray study of DOES confirmed that revealed that this salt has a b-type crystal structure, like the dioxane ring was condensed by cis fusion like that of the (DOET)2BF4 (Fig. 2). The anion AuI2- is aligned almost DOET molecule. ¶Thione 2 was prepared by reaction of (Bu4N)2[Zn(dsit)2] (dsit=4,5- diselenolato-1,3-dithiole-2-thione) with 1,2-dibromoethane in THF ||The transition temperature of this salt was estimated by the derivative of the Arrhenius plot of its resistivity.under reflux in 97% yield.5 J. Mater. Chem., 1998, 8(6), 1321–1322 1321Fig. 3 Energy band structure and Fermi surface of (DOES)2(AuI2)0.75 occupied by Au and I atoms with 3/4 occupancy to make the array such as ,-I-Au-I-V-I-Au-I-V-, (V=vacancy) along the a axis and there is no correlation between these arrays. Fig. 2( b) shows the donor arrangement of (DOES)2(AuI2)0.75.The DOES molecules are stacked along the [110] direction with some dimerization, so as to avoid the steric hindrance of the bulky dioxane rings. Although the intradimer and interdimer face-to-face distances (intradimer: 3.47 A ° ; interdimer: 3.78 A ° ) are larger than those in (DOET)2BF4 (intradimer: 3.50 A° ; interdimer: 3.77 A° ), the donor stack in (DOES)2(AuI2)0.75 is much more uniform because the ratio of overlap integral p1/p2 is close to 1.0 compared with the value of (DOET)2BF4 [0.94 for (DOES)2(AuI2)0.75, 1.6 for (DOET)2BF4].The pattern of chalcogen contacts of (DOES)2(AuI2)0.75 is similar to that of (DOET)2BF4, hence (DOES)2(AuI2)0.75 has the 2D energy dispersion relation and closed Fermi surface shown in Fig. 3. Because the band filling of this salt increased by 8.3% compared with that of (DOET)2BF4, the area of the Fermi surface is smaller than in Fig. 2 Crystal structure showing: (a) intermolecular face-to-face disthe case of (DOET)2BF4. tances: d1=3.47 A ° , d2=3.78 A ° ; (b) intermolecular overlap integrals In conclusion, when comparing the crystal structure of (×10-3) p1, p2, q1, q2 and a are 21.31, 22.72, 7.51, 8.51 and -3.15, (DOES)2(AuI2)0.75 with that of (DOET)2BF4, the substitution respectively of sulfur atoms of DOET with selenium atoms has a minor influence on the donor arrangement.However, this modifi- parallel to the a axis and the length of the anion (4.492 A ° ) is cation yields a metallic salt of diVerent stoichiometry, which nearly 2/3a. In the oscillation photographs the diVuse lines can realize a diVerent band filling.The next aim of our ongoing were also observed at 3/4, 3/2, 9/4 and 3a*. These facts indicate the 4/3a periodicity of the anion without the three- investigation is further development of the metallic DOET and dimensional order. Namely, three sites for the anion are DOES salts and clarification of their crystal structures.**Crystal data for (DOES)2(AuI2)0.75: (C24H20O4S12Se4(AuI2)0.75, M=1411.06, triclinic, space group P19, a=6.814(3), b=9.197(5), c= References 16.491(8) A ° , a=94.49(4), b=96.45(4), c=109.71(5)°, V=959.3(9) A ° 3, Z=1, Dc=2.442 g cm-3, m=98.395 cm-1, F(000)=662.75. The data 1 J. Yamada, S. Tanaka, H. Anzai, T. Sato, H. Nishikawa, I. Ikemoto were collected on a Mac Science MXC18 diVractometer equipped and K.Kikuchi, J.Mater. Chem., 1997, 7, 1311. with graphite monochromated Mo-Ka radiation (l=0.71073 A ° ) using 2 For other TTF donors substituted with oxygen-containing rings, the v–2h scan technique to a maximum 2h of 60°. The structure was see: A. M. Kini, U. Geiser, H. H.Wang, K. R. Lykke, J. M. Williams solved by direct methods using CRYSTAN (MacScience, Japan), and and C.F. Campana, J. Mater. Chem., 1995, 5, 1647; J. Hellberg, refined by full-matrix least-squares analysis (anisotropic for S and Se K. Balodis, M. Moge, P. Korall and J.-U. von Schu� tz, J. Mater. atoms) to R=0.082 and Rw=0.093 for 2501 independent reflections Chem., 1997, 7, 31. [I2s(I)]. The refinement was carried out by assung that the 3 J. Yamada, Y.Nishimoto, S. Tanaka, R. Nakanishi, K. Hagiya and specific position (0, 0.5, 0.5) and the general ones for the anion are H. Anzai, T etrahedron L ett., 1995, 36, 9509. occupied by Au and I atoms, respectively. The occupancies for these 4 H. Anzai, J. M. Delrieu, S. Takasaki, S. Nakatsuji and J. Yamada, sites were fixed to 0.586 (weight is 0.293) and 0.873, considering that J. Cryst. Growth, 1995, 154, 145. every site is occupied by both Au (1/4 occupancy) and I (1/2 5 For the synthesis of thione 2, see: J. Garý�n, J. Orduna, M. Saviro�n, occupancy) atoms. Full crystallographic details, excluding structure M. R. Bryce, A. J. Moore and V. Morisson, T etrahedron, 1996, factors, have been deposited at the Cambridge Crystallographic Data 52, 11 063. Centre (CCDC). See Information for Authors, J. Mater. Chem., 1998, Issue 1. Any request to the CCDC for this material should quote the Communication 8/02556H; Received 3rd April, 1998 full literature citation and the reference number 1145/95. 1322 J. Mater. Chem., 1998, 8(6), 1321–

ISSN:0959-9428    

DOI:10.1039/a802556h

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 Formation of carbon nanocapsules with SiC nanoparticles prepared by polymer pyrolysis Takeo Oku,*† Koichi Niihara and Katsuaki Suganuma Institute of Scientific and Industrial Research, Osaka University,Mihogaoka 8-1, Ibaraki, Osaka 567-0047, Japan performed with a 200 kV electron microscope (JEM-2010) having a point-to-point resolution of 0.194 nm.The electron Carbon nanocapsules with SiC nanoparticles were produced by thermal decomposition of polyvinyl alcohol with SiC microscope is equipped with a TEM-IP system (PIXsysTEM), and imaging plates (IP) with the advantage of a large detection clusters at 500 °C in Ar gas atmosphere. High-resolution electron microscopy also showed the formation of carbon area of digital data were used to record the observed images.The detection area of the IP is 102×77 mm with a pixel size hollow structures such as nanoparticles, polyhedra, and clusters. The present work indicates that the pyrolysis of of 25×25 mm and an image depth of 0–16383 gray scale. The digital data were saved in digital data storage (DDS) by Digital polymer materials with clusters is a useful fabrication method for the mass-production of carbon nanocapsules at low Micro-Luminography (Fuji Film Co.Ltd.). For image processing and analysis of the observed HREM images, Image temperatures compared to the ordinary arc discharge method. Gauge, L Process (Fuji Film Co. Ltd.), Digital Micrograph (Gatan Inc.) and Adobe Photoshop software were used. A typical HREM image of as-prepared SiC nanoparticles with polyvinyl alcohol is shown in Fig. 1. The SiC particle size is in the range of 10–50 nm. An amorphous layer with a Carbon has various hollow-cage structures such as C60, giant thickness of 1–5 nm is observed at the surface of the SiC fullerenes, nanocapsules, bucky onions, nanopolyhedra, cones nanoparticles as indicated by the arrows. This layer must be and nanotubes.1–5 These structures show diVerent physical amorphous carbon, formed by decomposition of polyvinyl properties, and oVer the potential to study low-dimensional alcohol.materials in isolated environments. Nanoclusters encapsulated A HREM image of SiC nanoparticles prepared with poly- within these carbon hollow-cage structures are intriguing for vinyl alcohol annealed at 500 °C in Ar gas atmosphere is both scientific research and future device applications such as shown in Fig. 2(a). The particle size is 10–30 nm, and all cluster protection, nano-ballbearings, nano-optical–magnetic particles are surrounded by graphite sheets as indicated by devices, catalysis, and biotechnology.6–9 The arc discharge arrows A. Graphite sheets without any nanoparticles are also method10 is an ordinary method for the formation of hollowobserved as indicated by arrows B.An enlarged HREM image cage structures. However, it is hard to separate these cages of a nanoparticle is shown in Fig. 2(b). Lattice fringes with a from carbon soot, and understanding of the formation process distance of 0.25 nm which corresponds to the distance of {111} is diYcult because of the coexistence of various carbon byplanes of b-SiC are observed in the clusters.The {002} planes products and because of the high-temperature annealing. of graphite grow epitaxially on the {111} planes of b-SiC, The purpose of the present work is to prepare carbon which divides the SiC nanoparticle. nanocapsules with nanoclusters at low temperatures without Fig. 3(a) is a HREM image of double graphite-layered using the arc discharge method. SiC nanoparticles with polynanoparticles with a size of ca. 10 nm. These nanoparticles vinyl alcohol were selected for nanocapsule formation. were formed by the decomposition of polyvinyl alcohol which Polyvinyl alcohol decomposes readily at elevated temperatures, consists of C, H, and O atoms. The polyvinyl alcohol has the and produces carbon-based materials.SiC is a hard material general formula [CH2CH(OH)]m[CH2CH(OCOCH3)]n, and with a hardness of ca. 30 GPa. If the SiC nanoparticles are combined with slippery graphite, nano-ballbearings and solidstate lubricants can be obtained, which are similar in properties to the metal dichalcogenides produced by Rapoport et al.8 To understand the formation mechanism of the nanocapsules, high-resolution electron microscopy (HREM) was carried out for microstructure analysis.These studies will provide guidelines for the design and synthesis of the carbon nanocapsules, which are expected as future nanoscale devices. b-SiC nanoparticles (Sumitomo Cement Co. Ltd.) were dispersed in de-ionized water with polyvinyl alcohol (PVA- 706, Kuraray Co., Ltd.)9 at 60 °C.This polymer is a copolymer of polyvinyl alcohol and polyvinyl acetate. The solution with SiC and polyvinyl alcohol was dried in a drying oven prior to loading into the vacuum chamber. The annealing chamber was first evacuated to 1×10-6 Pa, and the samples were annealed at 500 °C for 30 min in an atmosphere of 0.12 MPa Ar gas. Samples for HREM observation were prepared by dispersing the materials on a holey carbon grid.HREM observation was Fig. 1 HREM image of as-prepared SiC nanoparticles with polyvinyl alcohol †E-mail: Oku@sanken.osaka-u.ac.jp J. Mater. Chem., 1998, 8(6), 1323–1325 1323Fig. 3 HREM images of carbon hollow cages. (a) Double wall nanocages. (b) Single wall polyhedra. (c) Carbon cluster at nanocapsule surface.{002} planes is also observed at the graphite surface near the Fig. 2 (a) HREM image of annealed SiC nanoparticles with polyvinyl SiC nanoparticle as shown in Fig. 3(b). Weak contrast is alcohol. (b) Enlarged HREM image of the nanocapsule. observed inside the polyhedron, which might be due to the existence of amorphous carbon. Hollow carbon clusters with diameter in the range it decomposes into H2O and CO2 at 120–170 °C in air. In the present work, only the carbon atoms remained in the sample 0.7–1.0 nm are often observed as indicated by asterisks in Fig. 2( b). An enlarged HREM image of a carbon cluster is after the pyrolysis of polyvinyl alcohol into H2O and CO2 in Ar atmosphere. A single wall polyhedron with flat graphite shown in Fig. 3(c). These are giant fullerenes derived from the 1324 J. Mater. Chem., 1998, 8(6), 1323–1325C60 series structures, which consist of 12 pentagons and is a useful method for the formation of carbon nanocapsules at very low temperatures compared to the conventional arbitrary numbers of hexagons with 60+ carbon atoms.6 In methods. Since the polyvinyl alcohol has a pyrolysis tempera- Fig. 3(c), the carbon cluster is connected at the step edge of ture of ca. 120 °C, formation of carbon nanocapsules would be the graphite sheets. Dark contrast corresponding to a carbon also expected at low temperatures below 100 °C. This kind of layer are observed, and the number of atoms is ca. 120, which chemical process is also useful for large-scale production of is estimated from the size of the clusters (ca. 1 nm). In nanocapsules with nanoclusters (semiconductors, metals, cer- addition, dark contrast is observed inside the cluster, which amics, etc.) from solution with colloids and metal complexes, indicates the existence of several atoms inside the carbon which is expected to be applicable to future nanoscale devices. cluster. Formation of the graphite layer around nanoparticles is The authors would like to acknowledge Prof.Y. Hirotsu and useful for cluster protection against grain growth of nano- Dr. T. Ohkubo for allowing us to use the electron microscope. particles. For various nanostructured materials, nanograins are The authors would like to thank Mr. M. Ueshima for experi- needed to obtain various required properties such as mechanmental help.This work is supported by The Mitsubishi ical, electronic, and magnetic properties. Similar formation of Foundation. graphite layers had been reported in the SiC ceramics prepared from polysilastyrene at 1800 °C in Ar gas atmosphere11 and References by liquid phase sintering of SiC at 1700 °C in vacuum.12 In addition, formation of carbon nanotubes had also been 1 H.W. Kroto, J. R. Heath, S. C. O’Brien and R. E. Smalley, Nature reported upon decomposition of SiC at 1700 °C in vacuum.13 (L ondon), 1985, 318, 162. In these works, {002} planes of graphite sheets grow epitaxially 2 S. Iijima, Nature (L ondon), 1991, 354, 56. 3 D. Ugarte, Nature (L ondon), 1992, 359, 707. on the {111} planes of b-SiC, which agrees well with the 4 Y. Saito, T. Yoshikawa, M.Inagaki, M. Tomita and T. Hayashi, present results. The graphite structure was also produced by Chem. Phys. L ett., 1993, 204, 277. pyrolysis of a polyvinylidene chloride polymer at temperatures 5 A. Krishnan, E. Dujardin, M. M. J. Treacy, J. Hugdahl, S. Lynum in the range 2100–2600 °C in an inert atmosphere.14 In the and T. W. Ebbesen, Nature (L ondon), 1997, 388, 451.present work, we have succeeded in producing carbon nanocap- 6 Y. Saito, T. Yoshikawa, M. Okuda, N. Fujimoto, K. Sumiyama, sules at ‘low’ temperature, 500 °C, in Ar gas atmosphere. K. Suzuki, A. Kasuya and Y. Nisina, J. Phys. Chem. Solids, 1993, 54, 1849. Although carbon nanocapsules have previously been produced 7 J. Sloan, J. Cook, M. L. H. Green, J. L. Hutchinson and R. Tenne, with the aid of catalytic eVects of metal elements such as Fe, J.Mater.Chem., 1997, 7, 1089. Co, and Ni,6,9 these catalytic elements are not needed in the 8 L. Rapoport, Y. Bilik, Y. Feldman, M. Homyonfer, S. R. Cohen present work. By the ordinary arc discharge method, it is and R. Tenne, Nature (L ondon), 1997, 387, 791. diYcult to control the formation of nanocapsules. In the 9 Kuraray Co., Ltd., Osaka, Japan. 10 B. R. Elliott, J. J. Host, V. P. Dravid, M. H. Teng and J.-H. Hwang, present work, the number of the graphite sheets is in the range J.Mater. Res., 1997, 12, 3328. of 1–3; this number can be controlled by altering the solution 11 W. Kra�tschmer, L. D. Lamb, K. Fostiropoulos and D. R. HuVman, concentration of the polyvinyl alcohol. Nature (L ondon), 1990, 347, 354.In conclusion, carbon nanocapsules with SiC nanoparticles 12 K. Suganuma, G. Sasaki, T. Fujita, M. Okumura and K. Niihara, were produced by thermal decomposition of polyvinyl alcohol J.Mater. Sci., 1993, 28, 1175. 13 C. M. Wang, M. Mitomo and H. Emoto, J. Mater. Res., 1997, with SiC clusters at the low temperature of 500 °C in Ar gas 12, 3266. atmosphere. Formation of carbon nanoparticles, polyhedra, 14 M. Kusunoki, M. Rokkaku and T. Suzuki, Appl. Phys. L ett., 1997, and clusters were also observed by HREM. A HREM image 71, 2620. of the nanocapsules showed that the graphite {002} planes 15 P. J. F. Harris and S. C. Tsang, Philos.Mag. A, 1997, 76, 667. were epitaxially grown on the {111} surface of the b-SiC. The present work indicates that the pyrolysis of polymer materials Communication 8/01912F; Received 9th March, 1998 J. Mater. Chem., 1998, 8(6), 1323–1325 13

ISSN:0959-9428    

DOI:10.1039/a801912f

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 Assembly of alternating TiO2/CdS nanoparticle composite films Encai Hao,a Bai Yang,*a Junhu Zhang,a Xi Zhang,a,b Junqi Suna,b and Jiacong Shena,b aDepartment of Chemistry, and bKey L ab. of Supramolecular Structure and Spectroscopy, Jilin University, Changchun, 130023, P.R. China A composite film comprising TiO2 and CdS nanoparticles was fabricated based on an alternating deposition method; this was confirmed by UV–VIS spectroscopy, TEM, and photocurrent measurements.Recently, the organization of semiconductor nanoparticles into layered structures has received increasing attention.1,2 Several diVerent strategies for the fabrication of nanoparticle composite films have been reported, including Langmuir–Blodgett (LB), self-assembly and casting techniques.Self-assembly is a rapid and experimentally very simple way to produce complex Fig. 1 Schematic drawing for the build-up of an alternating film of layered structures with precise control of layer composition TiO2/CdS. S is the abbreviation for surface charged substrate. and thickness. Over the past few years, many nanoparticles Procedures I and II represent the adsorption of TiO2 and CdS (such as PbI2, TiO2, CdS, and Au) have been assembled nanoparticles.successfully into layered systems using the self-assembly method.3,4 repeating the above procedure using the TiO2 and CdS col- ‘Coupled’ nanoparticles consist of semiconductor particles loidal solutions. having a large bandgap and an energetically low-lying conduc- Since the isoelectric point of titanium dioxide is at pH= tion band, combined with particles having a small bandgap 5–7, the surface of TiO2 particles will therefore carry net and an energetically high-lying conduction band.5,6 Charge positive charges at pH<4.Therefore we envisaged the substiinjection from one semiconductor into another can lead to tution of TiO2 particles for polycationic compounds. Here we eYcient and longer charge separation by minimizing the first obtained a TiO2 monolayer covered substrate based on unwanted electron–hole recombination pathway.Therefore electrostatic interaction, after which the substrate was dipped ‘coupled’ type nanoparticles are anticipated to have potential into CdS colloid stabilized by mercaptoacetic acid.It is well applications in photocatalysis and solar energy conversion. known that carboxyl groups bind strongly to the surface of Here, by utilizing TiO2 and CdS colloids, we demonstrate the TiO2.6 Therefore, it is possible to transfer the mercaptocarformation of a composite film comprising diVerent nanopart- boxylic acid stabilized CdS nanoparticles to the surface of the icles based on an alternating deposition method.UV–VIS TiO2 monolayer, and a heteromultilayer would be fabricated spectra illustrated that TiO2 and CdS nanoparticles can be by repeating the TiO2 and CdS units. deposited uniformly onto multilayers. Photoelectrochemical UV–VIS spectra, obtained using a Shimadzu 3100 measurements demonstrated that eVective charge separation UV–VIS–NIR recording spectrophotometer, were used to occurred in the nanocomposite film.monitor the self-assembly of the repeating TiO2 and CdS units The cationic TiO2 colloidal solution was prepared by the (see Fig. 2). The absorption spectra of CdS and TiO2 nanoforced hydrolysis technique.6 A stable CdS colloid was pre- particles exhibit typical features around 450 nm and below pared upon the addition of mercaptoacetic acid as a stabilizing 350 nm respectively, details of which can be found elsewhere.7,9 agent.7 The carboxylic groups from the mercaptoacetic acid The spontaneous assembly of the cationic TiO2 nanoparticles modified on the surface of the particles resulted in anionic onto the anionic polyelectrolyte (PSS) is based on the ionic CdS particle surfaces.In a typical preparation of acidic CdS attraction developed between the oppositely charged species.10 colloid, a 100 ml solution of 1.0×10-3 mol l-1 CdCl2 and From curve 1, the obvious absorption below 350 nm illustrated 2.0×10-3 mol l-1 mercaptoacetic acid was prepared, Na2S that TiO2 nanoparticles were absorbed onto the ionic substrate.(10 ml, 1.0×10-3 mol l-1) was then injected under vigorous By comparing curve 1 with curve 2, it was found that an stirring.The pH of the resulting yellow colloid was adjusted obvious absorption emerged around 450 nm when the CdS to 3.0 by dropwise addition of 0.01 mol dm-3 HCl. nanoparticles were adsorbed onto the film, an eVect caused by Multilayer ultrathin films comprising TiO2 and CdS have CdS particles. With increasing the number of TiO2 units, there been fabricated by an alternating deposition process (Fig. 1). is an obvious increase of the absorption below 350 nm, which The first step is to modify the corresponding substrate to decreases only slightly at longer wavelengths owing to slight create a charged surface according to the literature.8 The desorption. Furthermore, the absorbance around 450 nm was resulting charged substrate was first dipped into a TiO2 found to increase linearly with increasing numbers of CdS colloidal solution for 40 min to absorb one layer of TiO2 units.Another linear relationship between the absorbance at nanoparticles. After washing with deionized water and 250 nm for TiO2 and the number of layers of TiO2 deposited drying, the slide was transferred to the acidic CdS colloidal also appears, if the absorbance at 250 nm for CdS is omitted.solution for another 40 min in order to adsorb one layer of These indicate that both TiO2 and CdS can be deposited uniformly onto multilayers. However, the development of the CdS nanoparticles. The multilayer films were obtained through J. Mater. Chem., 1998, 8(6), 1327–1328 1327process, 40 min immersions still result in well packed TiO2/CdS alternating assemblies according to TEM measurements.Fig. 3 shows the photocurrent response for a 15-layer TiO2/CdS composite film and a TiO2/PSS composite film modified on ITO electrodes. Photocurrents were measured by illuminating the self-assembled, film-carrying ITO electrode by a defocused light (300 W xenon lamp).The photocurrent was registered at potentiostatic conditions, DV=0.6 V (Ag/AgCl standard). The standard electrolyte consisted of 0.4 mol l-1 Na2S and 0.1 mol l-1 Na2SO3. Comparison of curves b and a shows that the intensity of the former sample was enhanced significantly relative to that of the latter, which indicates that eYcient charge transfer occurred in the composite films.In addition, the photovoltage of TiO2/CdS composite films with 83 mW cm-2 in standard electrolyte versus the Ag/AgCl rest potential was -50.8 mV. CdS has a small bandgap and an energetically high-lying conduction band, while TiO2 has a large bandgap and an energetically low-lying conduction band. Under irradiation, electrons generated in the CdS layers would transfer to the conduction band of TiO2.On the other hand, holes remain in CdS. It was found that the photocurrent Fig. 2 UV–VIS absorption spectra of alternating film deposition of response was greater when a sulfide electrolyte which consists TiO2/CdS with diVerent numbers of layers on a quartz slide. From of a redox couple such as S2-/SO32- was selected instead of the lower to the upper curves, the number of TiO2 and CdS layers is 1, 3, 5, 7 and 2, 4, 6, 8 respectively.Inset: the relationship of absorbance Na2SO4. From earlier studies of Q-particle-sensitized nanoat 420 and 360 nm vs. the number of CdS layers, and 250 nm vs. the porous electrodes,5 we therefore considered that holes but not number of TiO2 (the absorbance at 250 nm for CdS was omitted). electrons would react with the electrolyte when the electrolyte consists of the redox couple, which would reduce the electron losses in the TiO2 layers after the charge separation and thus alternating TiO2/CdS component structure is shown in a very lead to a greater photocurrent response.idealized manner in Fig. 1. From the transmission electron In conclusion, a new kind of TiO2/CdS alternating film micrographs, we found TiO2 particles in the first layer (S/TiO2) comprising diVerent nanoparticles was fabricated based on an were well dispersed over a large area, the particles in the alternating deposition method.UV–VIS spectra illustrated that second layer (S/TiO2/CdS) were close-packed with a high TiO2 and CdS could be deposited uniformly onto multilayers. surface coverage, while the images of the multilayer (the third Significant enhancement was found in the photocurrent layer) showed obvious spherical agglomerates.Although the response of a TiO2/CdS composite film modified ITO electrode absorption kinetics experiments illustrated a rapid adsorption compared with a TiO2/PSS composite film modified ITO electrode, indicating that eYcient charge transfer occurred in the former system.This work was supported by the National Natural Science Foundation of China. The authors would like to thank Dr. Dongshe Zhang, and Dr. Yuhong Sun (Institute of Photographic Chemistry, Academic Sinica, Beijing) for their help in carrying out the photocurrent measurements. References 1 J. H. Fendler and F. C. Meldrum, Adv. Mater., 1995, 7, 607. 2 G. Decher, Science, 1997, 277, 1232. 3 M. Y. Gao, M. L. Gao, X. Zhang, Y. Yang, B. Yang and J. C. Shen, J. Chem. Soc., Chem. Commun., 1994, 2777. 4 J. H. Fendler, Chem. Mater., 1996, 8, 1616. 5 R. Vogel, P. Hoyer and H. Weller, J. Phys. Chem., 1994, 98, 3183. 6 D. Lawless, S. Kapoor and D. Meisel, J. Phys. Chem., 1995, 99, 10 329 and references therein. 7 V. L. Colvin, A. N. Goldstein and A. P. Alivisatos, J. Am. Chem. Soc., 1992, 114, 5221. 8 Y. Sun, E. Hao, X. Zhang, B. Yang, M. Gao and J. Shen, Chem. Commun., 1996, 2381. 9 K. R. Gopidas, M. Bohorquez and P. V. Kamat, J. Phys. Chem., 1990, 94, 6435. 10 Y. Liu, A. Wang and R. Clus, J. Phys. Chem., 1997, 101, 1385. Fig. 3 Photocurrent response of ITO electrodes modified with nanoparticle multilayers: (a) 15 layers of TiO2/PSS, (b) 15 layers of Communication 8/02655F; Received 7th April, 1998 TiO2/CdS 1328 J. Mater. Chem., 1998, 8(6), 1327–1328

ISSN:0959-9428    

DOI:10.1039/a802655f

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 function for anionic clays: selective transition metal cation uptake by diadochy Sridhar Komarneni,* Naofumi Kozai and Rustum Roy Materials Research L aboratory and Department of Agronomy, T he Pennsylvania State University, University Park, PA 16802, USA carbonate anion containing material was supplied by the Aluminum Company of America (ALCOA) while the nitrate Here we report for the first time, the extremely high and selective uptake of transition metal cations such as Cu2+, anion containing material was synthesized hydrothermally at 200 °C for 4 h under saturated steam pressure in our laboratory Ni2+, Co2+ and Zn2+ by two anionic clays of nominal composition, [Mg6Al2(OH)16]2+[CO3·4H2O]2- and according to the method described previously.24 The carbonate and nitrate anion containing materials have surface areas of 9 [Mg2Al(OH)6]+[NO3·2H2O]-.We postulate that the principal mechanism of this selective cation uptake is by and 54 m2 g-1, respectively. We determined selective transition metal (M2+) cation (M2+=Cu2+, Co 2 +, Ni 2 + and Zn2+) substitution in the anionic clay structure for mainly Mg through a process known as diadochy.This newly discovered uptake by using solutions containing 5×10-5, 5×10-4, 5×10-3 or 1×10-2 M M2+ in the presence of 0.5 M NaCl. The function for anionic clays is useful for the decontamination and immobilization of the above transition metals at room pH of the starting solutions except for 1×10-2 M CuCl2 and 5×10-3 M CuCl2 was adjusted to pH 5 with HCl and NaOH temperature and may find applications in the remediation of metal contaminated soils, filtration of drinking water as well as solutions. Because Cu precipitates at pH 5.0 in 10-2 M and 5×10-3 M solutions, the pH of these solutions was adjusted decontamination of waste waters.to 4.0 or used at their unadjusted pH (4.7). Each anionic clay (50 mg) was equilibrated by shaking with 25 ml of solution for 1 day.After equilibration the solid and solution phases were separated by centrifugation and the pH of the solutions were measured. The solutions were analyzed by atomic emission This paper deals with transition metal cation uptake in anionic spectroscopy (AES) and the solid phases were characterized clays by a process known as diadochy.Copper, cobalt, nickel by powder X-ray diVraction (XRD) and scanning electron and zinc are some of the regulated metals which are of microscopy (SEM). Kinetics of uptake of Co and Ni were environmental concern because they are hazardous to humans investigated using 5×10-4 M concentrations of these ions in and/or to other forms of life depending upon their concen- 0.5 M NaCl at pH 5.The solutions and solids were separated tration in drinking or irrigation water.1 Remediation of the and analyzed as above. All of the data points represent the environment or decontamination of drinking water through average of triplicate runs. filtration is possible with the use of a highly selective material Both anionic clays selectively removed 100% of all the for these metals.Inorganic exchange materials such as zeolites transition metal (M2+) cations from 5×10-4 and 5×10-5 M and cationic clays are frequently used in remediation. Minerals M2+Cl2 solutions in the presence of 0.5 M NaCl after equilibelonging to the anionic clay group have been reported by bration for 24 h. When the concentration of metal cations was several researchers since the 1920s,2–5 and CO32- or NO3- increased to 5×10-3 or 1×10-2 M the uptake of cations hydrotalcites were first synthesized hydrothermally in these changed with the type of transition metal cation as well as the laboratories in 1952.6 The term ‘anionic clay’ refers to natural nature of the anionic clay, i.e., whether the anionic clay or synthetic layered or lamellar double hydroxides with intercontained carbonate anions or nitrate anions (Table 1).The layers containing anionic species to balance the positive charge commercial sample with lower charge density containing on the layers while the term ‘cationic clay’ refers to natural or synthetic layered structures whose interlayers contain cations to balance the negative charge on the layers7 and these cations Table 1 Transition metal cation uptake by two anionic clays a are exchangeable with other cations, i.e., cation exchange.The uptake (%) process of cation exchange in natural and synthetic clays is well known.8,9 Anionic clays are rare in nature while cationic carbonate nitrate clays are ubiquitous in all soils and sediments. Hydrotalcite, nominal anion anion manasseite, pyroaurite, takovite, stichtite, honessite, meixnerite transition starting containing containing and sjo� grenite, are some examples of minerals of the anionic metal cation concentrationb/M material material clay family.7 Anionic clays are a multifunctional group of Co2+ 5×10-3 22.9±2.8 22.9±1.6 materials with much research exploring their structural, chemi- Co2+ 1×10-2 10.4±3.0 8.1±1.0 cal, ionic, catalytic, optical and electronic properties.7,10–19 Ni2+ 5×10-3 29.3±3.8 21.2±0.8 Anionic exchange, i.e., exchange of interlayer anions with Ni2+ 1×10-2 14.6±2.2 9.3±1.6 other anions in anionic clays has been studied previously.20–23 Zn2+ 5×10-3 60±3.6 50.6±0.8 However, to the best of our knowledge, no cation uptake Zn2+ 1×10-2 35.6±4.8 24.9±0.8 Cu2+ (pH 4.67) 5×10-3 100±0.05 78.3±1.8 studies by anionic clays have been reported.Here we report Cu2+ (pH 3.99) 5×10-3 100±0.05 79.4±0.5 the extremely high and selective uptake of transition metal Cu2+ (pH 4.68) 1×10-2 95.1±1.1 38.9±1.5 cations such as Cu2+, Ni 2 +, Co 2 + and Zn2+ by two anionic Cu2+ (pH 4.00) 1×10-2 93.4±0.7 37.1±1.5 clays of nominal composition, [Mg6Al2(OH)16]2+- [CO3·4H2O]2- and [Mg2Al(OH)6]+[NO3·2H2O]-. aComplete uptake occurred from 5×10-4 and 5×10-5 M concen- Two synthetic Mg–Al double hydroxide materials of the trations of transition metals from 0.5 M NaCl by both anionic clays.bThe transition metals are in the presence of 0.5 M NaCl. nominal composition given above were used in this study. The J. Mater. Chem., 1998, 8(6), 1329–1331 1329Fig. 2 Kinetics of uptake of Co and Ni from 0.5 M NaCl containing 5×10-4 M Co2+ or 5×10-4 M Ni2+ Fig 1 Powder XRD patterns of carbonate form of anionic clay: (A) untreated and (B) treated with 5×10-3 M Ni2+ CO32- as the charge balancing ion reacted to take up higher percentages of all the transition metals compared to the hydrothermally synthesized higher charge density anionic clay with nitrate ions as the interlayer species even though the former sample has a larger particle size than the latter (SEM results not shown; see surface areas above).Powder XRD and scanning electron microscopy results showed that there were no changes in the patterns (Fig. 1) or morphology, respectively when the anionic clays were treated with 5×10-5 or 5×10-4 M Co2+, Ni 2 +, Zn 2 + or Cu2+ and with 5×10-3 or 1×10-2 M Co2+ or Ni2+.When the anionic clays were treated with 5×10-3 or 1×10-2 M Zn2+, a small unidentified peak at ca. 2.70 A ° appeared while the remainder of the pattern remained unchanged. For 5×10-3 or 1×10-2 M Cu2+ solution, the formation of botallackite, Cu2+(OH)3Cl as an additional phase was observed. Solution analyses indicated more or less stoichio- Fig. 3 Schematic of M2+(=Cu, Ni, Co and Zn) substitution for Mg2+ in anionic clay structure through the interlayer space metric release of Mg2+ into the solution by the uptake of all the transition metals (Table 2) except at concentrations of 5×10-3 and 1×10-2 M for Cu2+ where precipitation of botallackite was observed. We have also tested calumite, Table 3 Theoretical capacities for metal cation uptake of some clays [Ca2Al(OH)6]+[NO3·xH2O which is an analogue of hydro- and zeolites in comparison to those of anionic clays talcite for Cd2+ and Pb2+ uptake using the same concentheoretical capacity trations of these ions as the above transition metals in 0.5 M for metal cation NaCl.The uptakes of Cd and Pb by calumite were determined material uptake/mequiv.(100 g)-1a to be very high but the amounts of Ca released into solution were in excess of stoichiometric exchange with Cd or Pb. The cationic clays mechanism of uptake of these ions on calumite, however, is montmorillonite ~100 vermiculite ~180 not clear. Because the equilibrium pH with calumite is very zeolites high, precipitation along with other mechanisms are possible clinoptilolite ~220 (detailed results on calumite will be reported elsewhere).zeolite, 4A ~540 The above results of transition metal cation uptake on anionic clays Mg–Al anionic clays and the rapid kinetics involved (Fig. 2) [Mg6Al2(OH)16]2+[CO3·4H2O]2- ~1990 suggest that the mechanism of transition metal ion uptake is [Mg2Al(OH)6]+[NO3·2H2O]- ~1450 by diadochy, i.e., substitution for Mg2+ in the anionic clay aCalculated from chemical formula. crystal structure facilitated by the easy access for the cations through the interlayer spaces (Fig. 3). Diadochy is a well Table 2 Uptake of transition metal (M2+) cations from 0.5 M NaCl containing 5×10-3 M M2+ with concurrent stoichiometric (within experimental errors) release of Mg2+ from anionic clays carbonate anion nitrate anion containing material containing material metal pH after ion equilibration M2+ uptake/mM Mg2+ release/mM M2+ uptake/mM Mg2+ release/mM Co2+ 7.1 1.17 0.83 1.17 1.18 Ni2+ 7.0 1.45 1.40 1.03 1.28 Zn2+ 6.3 2.77 2.52 2.33 2.32 Cu2+ 5.4 — — 3.45 2.57 1330 J.Mater. Chem., 1998, 8(6), 1329–13315 C. Frondel, Am. Mineral., 1941, 26, 295. known reaction in calcite1 and hydroxyapatites.25 This mechan- 6 D.M. Roy, R. Roy and E. F. Osborn, Am. J. Sci., 1953, 251, 337. ism is very likely because anionic clays with all these transition 7 A. de Roy, C. Forano, K. El Malki and J. P. Besse, in Expanded metals have been prepared by many researchers.7 The kinetics Clays and Other Microporous Solids, ed. M. L. Occelli and of Co2+ and Ni2+ ion uptake are rapid as shown in Fig. 2 H. E. Robson, Van Nostrand Reinhold, New York, 1992, p. 108. which is a very important consideration in certain applications 8 W. J. Lay, Ind. Eng. Chem., 1954, 46, 1061; J. M. Kerr, Bull. Am. Ceram. Soc., 1954, 38, 374. such as waste water treatment or drinking water filtration. The 9 S. Komarneni and R. Roy, Science, 1988, 239, 1286; W. J. Paulus, new function discovered here for anionic clays is expected to S.Komarneni and R. Roy, Nature (L ondon), 1992, 357, 571. have applications in remediation of metal contaminated soils 10 R. Allman, Acta Crystallogr., Sect. B, 1968, 24, 972. and groundwater (for example, at superfund sites) and clean 11 H. F. W. Taylor,Miner. Mag., 1969, 37, 338. up of drinking as well as waste waters because of their very 12 K.Hashi, S. Kikkawa and M. Koizumi, Clays Clay Miner., 1983, high theoretical capacities for transition metals compared to 31, 152. 13 A. de Roy, J. P. Besse and P. Bondot, Mater. Res. Bull., 1985, the existing cation exchange materials (Table 3). The anionic 20, 1091. clay materials not only selectively remove transition metal 14 K. El Malki, A. de Roy and J. P. Besse, Eur.J. Solid State Inorg. cations but also lock them up in their structure leading to Chem., 1989, 26, 339. their immobilization. An additional advantage of these mate- 15 A. de Roy and J. P. Besse, Solid State Ionics, 1991, 46, 95. rials compared to synthetic zeolites and clays is that they can 16 M. Lal and A. T. Howe, J. Solid State Chem., 1981, 39, 368. 17 S. Miyata and T. Kumura, Chem.L ett., 1973, 843. be inexpensively prepared at room temperature.7 18 S. Miyata, Clays ClayMiner., 1975, 23, 369. 19 J. G. Nunan, P. B. Himelfarb, R. G. Herman, K. Kleir, This research was supported by the Materials Research C. E. Bogdan and G. W. Simmons, Inorg. Chem., 1989, 28, 3868. Laboratory’s Consortium on Chemically Bonded Ceramics. 20 H. P. Boehm, J. Steinle and C. Vieweger, Angew. Chem., Int. Ed. Engl., 1977, 16, 265. 21 D. L. Bish, Bull.Mineral., 1980, 103, 170. 22 S. Miyata, Clays ClayMiner., 1983, 31, 305. References 23 T. Kwon, G. A. Tsigdinos and T. J. Pinnavaia, J. Am. Chem. Soc., 1 J. R. Conner, Chemical Fixation and Solidification of Hazardous 1988, 110, 3653. 24 G. Fetter, F. Herna�ndez, A. M. Maubert, V. H. Lara and P. Bosch, Wastes, Van Nostrand Reinhold, New York, 1990, p. 692. J. PorousMater., 1997, 4, 27. 2 N. S. Kurnakov and V. V. Chernykh, Zap. Vserass. Mineral Ova., 25 T. Suzuki, K. Ishigaki and M. Miyake, J. Chem. Soc., Faraday 1926, 55, 118. T rans. 1, 1984, 80, 3157. 3 G. AminoV and B. Broome�, Kungl. Svenska. Vetensckaps Handel, 1930, 9, 23. 4 H. H. Read and B. E. Dixon,Miner.Mag., 1933, 23, 309. Communication 8/01631C; Received 26th February, 1998 J. Mater. Chem., 1998, 8(6), 1329–1331

ISSN:0959-9428    

DOI:10.1039/a801631c

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 Molecular modelling of interactions at the composite interface between surface-treated carbon fibre and polymer matrices: the influence of surface functional groups Ian Hamerton, John N. Hay, Brendan J. Howlin, John R. Jones and Shui-Yu Lu* Department of Chemistry, University of Surrey, Guildford, UK GU2 5XH To simulate the results of surface treatments, commonly encountered functional groups were introduced onto the surface of the carbon fibre model.The carbon fibre model used in this study is based on the single layer diagonal graphitic plane, comprising 52 six-membered rings, in a 4×13 configuration, and of 150 carbon atoms. Surface treatment was represented by the introduction of functional groups (MR): each time, a CMC bond was broken along the edge of the plane, and a pair ofMR groups was added to the graphitic plane.The total number of functional groups (n) was six. The eVect of these functional groups on the non-covalent bonding interactions at the composite interface between carbon fibre and epoxy resin was investigated using a previously established BLENDS method. The compatibility of the resin and fibre in this model, indicated by the interaction parameter x(T ), was dependent upon two factors: steric bulk and electrostatic interactions.The halogen substituents show a decrease in x(T ), as one descends the group. Maximum interaction tends to be a function of steric bulk and polarizability in this group. The alkyl (CnH2n+1) and phenyl substituents also show a decreasing trend in x(T ) with increasing size, although the interaction parameter with methyl is anomalously low in all cases.Reinforcement in composites such as those containing epoxy The eVect of other functional groups, apart from the most common ones such as MOH, MCOOH and MNH2 on the resin matrices (e.g. the diglycidyl ether of bisphenol A, DGEBA) interfacial properties is of great interest to both chemists and incorporating carbon fibres, is achieved by suYcient stress materials scientists alike.Current surface treatment techniques transfer between fibres and matrix. The stress transfers can be limit the number and type of functional groups that can be realised by mechanical interlocking, physical adhesion and introduced onto the carbon fibre surfaces, hence preclude chemical bonding.1 The technology to improve stress transfer investigation by experimental methods.15 Molecular simulation within composites, as indicated by the improved interlaminar provides a useful tool to tackle these problems by introducing shear strength (ILSS), consists of surface treatment methods ‘artificial’ surface functional groups.As a result, our under- such as (a) treating fibres in oxidative acids,2 (b) catalytic standing of the complex nature of the fibre-reinforced com- oxidation in air,3 (c) anodic treatment in aqueous solutions,4 posite at interfaces can be increased.In this paper, we report (d) plasma surface treatments,5 and (e) metal oxide coating.6 our investigation into the eVect of various surface functional Despite its importance, the nature of the interfaces/interphase groups on the interactions between carbon fibres and polymer region in fibre reinforced polymers remains largely unresolved.matrices, using the BLENDS method. Wright suggested7 that carbon fibre surface treatments lead to increased fibre surface area, removal of a weak surface layer, and modification of the surface chemistry.All three phenomena Calculation serve to improve resin wetting and bonding, and of these the A Silicon Graphics Indy workstation (MIPS R4000) running available evidence suggests that the change in surface area is the computer program CERIUS2 v.1.6 (Molecular Simulations not a significant parameter, although polar surface free energy Inc.) was used to generate models and to calculate the non- was found to increase on treatment. bonding interaction.The Dreiding 2.21 force field16 was used Increasingly, there is a need to understand the properties in this work. The charge calculation method, charge equili- and features of polymeric materials at the molecular level. bration (Qeq),17 was used to assign, edit and calculate point Recent advances in molecular modelling have led to new charges.methods able to predict the properties of structural, electromagnetic and optical materials.8–14 Simulation based on simpli- Models fied models has been carried out using ab initio, semi-empirical and force-field methods in the area of composite interfacial The monomer models (Fig. 1) include the diglycidyl ether of properties.12,13 Our previous work14 demonstrates that the bisphenol A (DGEBA), bis(4-aminophenyl) sulfone (DDS), molecular modelling method, BLENDS, can be used to simu- bis(4-aminophenyl)methane (DDM) and ethylene.The polylate the non-covalent bonding interaction between electrolyti- mer models include the linear DGEBA homopolymer chain, cally surface treated carbon fibre and various amine cured and the amine-cured epoxy resins, both alternating (alt-co) epoxy polymer models. The carbon fibre was modelled as and random (ran-co) copolymers of DGEBA–DDM and graphitic layer(s) with MOH and MCOOH functional groups DGEBA–DDS.The homopolymer, random copolymers (amiattached to one edge. The polymers were modelled as an ne5epoxy monomers in the ratio 258), and alternating copolyhomopolymer and as both alternating and random copolymers mers represent diVerent degrees of chain extension of the epoxy to represent diVerent degrees of cure of the epoxy resin.The resin. The carbon fibre models are a series of graphitic planes. change in the free energy of mixing (DmixG) was used to indicate The basic considerations for model construction and details of the magnitude of the interaction and, hence, the interfacial the polymer models have been described elsewhere.14 adhesion.The results show a trend, in relation to the level of surface treatment, in agreement with literature data of com- Single-layered graphitic models for carbon fibres. The carbon fibre model used in this study is based on the single layer posite interfacial strength.J. Mater. Chem., 1998, 8(6), 1333–1337 1333h t S O O N H h N t H N t H N H h O t O OH h OH O t O h OH ( a ) ( b ) ( c ) ( d ) ( e ) Fig. 1 Structural repeat units for (a) the diglycidyl ether of bisphenol A (DGEBA) in homopolymer and random copolymers, (b) bis(4- aminophenyl) sulfone (DDS), (c) bis(4-aminophenyl)methane (DDM), (d) DGEBA in alternating copolymers and (e) polyethylene (PE).h=head linkage, t=tail linkage for the polymer builder. graphitic sheet. A diagonal graphitic plane, comprising 52 six- Fig. 2 Structure of the single-layered graphitic plane as an example of membered rings, in a 4×13 configuration, and of 150 carbon the surface-treated carbon fibre model, in which R=But and n=6 atoms, was built as the principal, non-surface treated carbon fibre model.In this structure the carbon atoms have sp2 hybridisation, and form partial double bonds (the perfect graphitic structure is a flat plane). Those carbons at the outermost edge allows one to obtain more representative interaction energies of form CH (in closed ring) or CH2 (open end). A further 18 ij pairs, Eij values. Graphitic structures in carbon fibres are variants of the model were built based on this structure.Surface highly oriented, therefore models are aligned along the principal treatment was represented by the introduction of functional axes (the allowed range of orientation was 10°). On the other groups (MR): each time, a CMC bond was broken along the hand, the epoxy and polyethylene polymer models have an edge of the plane, and a pair of MR groups was added to the isotropic (random) packing with no restrictions at all.graphitic plane (Fig. 2). The total number of functional groups The interaction parameter x(T ) is defined as eqn. (1), (n) was six. The functional groups were arranged in such a way as to be evenly distributed along the edge of the plane. These x(T )= Emix(T ) RT = (Z12E12+Z21E21-Z11E11-Z22E22) 2RT (1) structures were then relaxed by molecular mechanics energy minimisation, using conjugate gradients and charges derived where Eij is the interaction energy for a pair of ij .BLENDS from the Qeq method, until energy convergence (at 0.01 uses Monte Carlo atomistic simulations both to generate kcal mol-1; 1 cal=4.184 J) was achieved. thousands of diVerent molecular orientations and to calculate For R=CH3, Bu t , CF3, models with n=12 were constructed their pair-interaction energies (this method results in four using the same procedure described above.Boltzmann averaged Eij values); Zij, the coordination number, is the number of molecules of type j that can be packed around Molecular simulation method a single molecule of type i. A single coordination number has a definite physical significance only when two components of The property prediction method, BLENDS,18 was used to investigate the interactions between two macromolecules.The the binary mixture have a similar volume or surface area. Z is explicitly calculated for each of the possible molecular pairs module combines a modified Flory–Huggins model19 and molecular simulation techniques to calculate the compatibility using a molecular simulation method called nearest-neighbours packing.It involves generating clusters in which nearest neigh- of binary mixtures. The theoretical and computational considerations were developed by Blanco and co-workers.20 In bours are packed around the central molecule until no more will fit. The van der Waals surface is used to represent the implementing the Flory–Huggins lattice model for polymers, BLENDS requires that the lattice sites be occupied by polymer shape of the molecules.We matched the single layer graphitic model (150 atoms) with polymer models comprising ten repeat segments. BLENDS is also an oV-lattice calculation, meaning that molecules are not arranged in a regular lattice as in the units (about 200 atoms).BLENDS analysis options can be used both to calculate original Flory–Huggins theory. In practice, each of the graphitic models and the polymer models occupy one lattice site. thermodynamic functions (entropy, enthalpy and free energy of mixing) for a binary system, and to create plots of these The degrees of polymerisation X1 and X2 were both set to 1, and the total molecular weight of the polymers was ignored.functions versus composition at a specific temperature. The plots generated reflect the current choice of the interaction BLENDS also provides options that place restrictions on both molecule alignment and atom contact during packing, and thus parameter model x(T ) and the degree of polymerisation of the 1334 J.Mater. Chem., 1998, 8(6), 1333–1337two components in eqn. (2), atom (from sp2 to sp3) and the requirement for bond angles of 109.5° rather the 120° to which the group is attached. The electrostatic repulsion between functional groups on each layer DmixG RT = w1 X1 ln w1+ w2 X2 ln w2+xw1w2 (2) also contributed to the deformation in these carbon fibre models. The models in all cases are relaxed to the same energy where DmixG is the free energy of mixing, w is the volume convergence criterion (0.001 kcal mol-1) in order to be directly fraction of each component, X1 and X2 are the degrees of comparable.This, of necessity, also requires the use of the polymerisation (or chain length) of each component. The plots same force field (in this case Dreiding 2.21).of thermodynamic isotherms showed that DmixG is sensitive to the change of volume fraction, but not to temperature in the range 300–400 K. When other parameters such as w1=w2= The nature of the surface functional groups 0.5, and T=300 K are equal, DmixG is proportional to x(T ). It is important to remember that the nature of the interfacial Therefore x(T ) is used to characterise the interaction between region between bulk resin and fibres is the subject of this study the carbon fibre and polymer models in this work.and the results will be discussed with regard to the interaction between these two components. The compatibility of the resin Results and Discussion and fibre in this model depends upon two factors: steric bulk and electrostatic interactions.The halogen substituents (the This work is part of an ongoing program involving both second group in Table 1) show a decrease in the interaction experimental and simulation studies of fibre–resin interactions parameter x(T ), as one descends the group. Fluorine has the covering XPS, SIMS analysis,4 surface radioisotope and stable greatest interaction parameter and iodine the least.Therefore, isotope labelling, surface chemical reaction,21 and inverse gas the interaction with iodine would be the most favourable. chromatography (IGC).8g The object of this paper was to look Maximum interaction tends to be a function of steric bulk and at the specific interactions between functionalised carbon fibre polarizability from this group. The group of substituents which surfaces and matrix resins.The model only deals with the comprises the alkyl homologues and the phenyl ring (the third improvement in the compatibility between the fibre and the group in Table 1) also shows a general trend with increasing resin owing to the functionalisation. It does not deal with size, although the interaction parameter with methyl is anomal- mechanical keying, pores or clefts, etc.It is currently impossible ously low in all cases. Interestingly, in most cases the substi- to model at the atomistic level a whole fibre and this may not tution of fluorine atoms for the hydrogen atoms (i.e. in changing be productive. Such a model would contain too many variables the substituent from MCH3 to MCF3) causes a decrease in to allow for the systematic control of each.The models used the interaction parameter and hence an increase in the compati- simulate, to the best of our current knowledge, a representative bility. This does not follow for the case when polyethylene is section of the fibre surface, in eVect a close up of a section of the polymer matrix. the surface. These models could be extended in three dimen- Considering the functional groups which actually exist on the sions to produce the entire surface of a fibre but would contain carbon fibre surface as a result of the electrolytic oxidation (e.g.no extra information than that used. MOH, MCOOH and MCHO, the seventh group in Table 1), The dependence on size of the models used has been dealt there is a general decreasing trend from MOH to MCOOH to with in our previous paper, where it was demonstrated that MCHO for the epoxy and polyethylene homopolymers and the both the multi-layer graphitic models and the single layer alternating epoxy–amine copolymers.With the random epoxy- models gave quantitatively similar results.14 The size of the amine copolymers there is no obvious trend. The small steric polymer needs to be commensurate with that of the graphitic bulk of the MOH group appears to predominate over its hydro- model.Obviously, as with any simulation technique, the results gen bonding potential. Conversely, in the case of the MCOOH obtained depend on the force field used. We have employed group its large steric bulk reduces its ability to form hydrogen- the Dreiding 2.21 force field for these simulations. This force bonds.Therefore, the MCHO group, being of intermediate size field has been validated by calculating the physical and with hydrogen-bonding potential gives the strongest interaction mechanical properties of structural polymers in a number of with all of the polymer matrices. These results, in line with previous papers.8 Naturally, in a study of this kind the electroexperimental data,7 do suggest that mild oxidative conditions static model is important and the magnitude of the point might facilitate optimal interfacial adhesion. Further complication charges can vary depending on the calculation used to derive might arise from the potential of the MCOOH group to form them.The ideal method would be to use ab initio calculations covalent bonds, leading to increased stress at the interface between to derive these but this method is limited to structures containthe matrix resin and the carbon fibre surface.Having said that ing about 40 atoms or less, and is not feasible with the current there is also evidence to show that the ILSS does not necessarily models. The force field chosen is balanced (i.e.the charges are decrease with a decrease in the oxygen–carbon ratio when the scaled up or down) so that the point charges do not dominate functional groups introduced by oxidative treatment are elimin- the covalent interactions and any alternative calculation would ated.22 The result highlights the importance of physical or also need to have the charges scaled. A further proof of the mechanical eVects in determining fibre–matrix adhesion.validity of the models and force field employed can be found in a study which showed that simulation of the electrostatic eVects compared very well with IGC experimental results.8g Interactions between single-layered graphitic models and polymers The geometry of the graphitic plane A linear free energy treatment23 was attempted for the entire Table 1 by plotting sI+sR versus the interaction parameter All structures of the carbon fibre model with functional groups attached deviated from the perfectly flat graphitic plane after x(T ).However, as expected this problem cannot be treated simply in terms of inductive eVects, and a poor correlation energy minimisation. In the oxidised portion of the models (where functional groups were added), the graphitic planes was obtained.Furthermore, a multiple linear regression analysis of x(T ) against CMX distances (which were obtained from curved when energy minimisation was performed (Fig. 2). The bulkier groups resulted in more severe distortion. Similar the centre of the carbon to the centre of the outermost atom for each functional group after optimising the structure on distortion was also observed when the concentration of the groups increased.The curvature induced in the surface is a CERIUS2), sI+sR (the Hammett substituent constants23) and DHG (the sum of bond enthalpy24) gave an r2 value of 29% direct result of substitution with the functional groups, i.e. it represents a change in the hybridisation state of the carbon with a constant of 211 indicating that most of the variation J.Mater. Chem., 1998, 8(6), 1333–1337 1335Table 1 Interaction parameter x(T ) at 300 K between polymer models and single-layered carbon fibre models with diVerent surface functional groups DGEBA copolymers functional group DGEBA DDS DDM DDS DDM n=6 homopolymer alt-co alt-co ran-co ran-co polyethylene H 306.2 279.9 355.2 312.3 307.1 226.4 F 159.7 137.7 214.3 165.5 164.5 95.4 Cl 129.5 61.1 111.1 96.6 109.7 68.7 Br 113.6 67.6 102.9 78.2 96.5 41.1 I 57.3 77.3 63.1 22.0 56.5 2.8 OMe 138.8 87.7 144.8 107.2 109.7 91.2 SMe 145.3 145.7 155.5 118.9 150.2 99.5 CF3 105.5 67.8 82.4 138.9 71.6 81.6 Me 157.6 154.2 185.2 149.3 124.8 74.4 Et 179.4 181.0 163.4 163.7 196.0 144.6 Pri 153.6 164.7 145.3 135.7 171.4 88.3 But 149.8 163.9 152.8 174.8 164.4 72.3 Ph 108.8 95.2 96.6 93.9 118.2 22.9 CN 142.2 118.8 148.9 137.0 154.8 89.5 NO2 98.2 74.0 74.5 75.7 72.0 67.9 NH2 148.9 124.7 175.4 150.5 139.5 58.4 OH 169.9 136.5 216.8 188.3 149.3 145.5 COOH 142.0 125.9 131.3 137.9 162.6 107.3 CHO 118.5 92.0 115.6 143.0 121.3 72.0 was not in the parameters chosen.Perhaps the situation is just too complex for these simple treatments.In order to find a simple parameter to correlate, the steric substituent constant (Es),23 which is a function of the steric bulk for spherical substituents, was plotted against the interaction parameter x(T ). The plot for the halogens is shown in Fig. 3 where an approximately linear decrease with Es is found as one descends the group.This provides some support for the idea that steric bulk favours interaction between the polymer matrix and the carbon fibre surface where the groups are essentially spherical and inherently polarizable. Similarly, for the alkyl and phenyl residues a plot of x(T ) against Es is given in Fig. 4. This shows a non-linear increasing trend with increasing Es with the most favourable interaction being formed with Fig. 4 Plots of x(T ) vs. steric substituent constant (Es) representing interaction of alkyl and phenyl substituted carbon fibre models with (+) DDM–DGEBA alternating copolymer, (*) DGEBA homopolymer, (#) DDM-DGEBA random copolymer, ($) DDS–DGEBA alternating copolymer, (&) DDS–DGEBA random copolymer and (×) polyethylene the phenyl substituent.The hydrogen substituent is the least favourable and this supports the idea that both steric bulk and polarizability are the important parameters to consider. Es contains information pertaining to steric bulk and it is interesting to note that this parameter gives a better correlation than that relating to polarizability. EVect of changing surface group concentration Table 2 lists the interactions between a DGEBA homopolymer and multi-layered carbon fibre models at diVerent values of n Fig. 3 Plots of x(T ) vs. steric substituent constant (Es) representing (number of functional groups on the carbon fibre model’s interaction of halogen substituted carbon fibre models with (+) DDM– surface). An increase of methyl group concentration on the DGEBA alternating copolymer, (*) DGEBA homopolymer, (#) DDMcarbon fibre model does not result in much change (+10%) DGEBA random copolymer, ($) DDS–DGEBA alternating copolymer, (&) DDS–DGEBA random copolymer and (×) polyethylene in x(T ).An increase of tert-butyl group concentration on the 1336 J. Mater. Chem., 1998, 8(6), 1333–1337Table 2 Interaction parameter x(T ) at 300 K between DGEBA homo- polar in nature and is more compatible with the non-polar polymer and carbon fibre models at diVerent number of surface polyethylene chains.functional groups (n) We wish to thank the EPSRC for funding (Grant GR/H95891) functionality n=6 n=12 a postdoctoral research fellowship (S.Y.L.). The published Me 157.6 172.8 results were generated using the program CERIUS2 (developed But 149.8 104.7 by Molecular Simulations Inc.).CF3 105.5 200.0 References carbon fibre model results in a decrease (-30%) in the value 1 J. D. H. Hughes, Composites Sci. T echnol., 1991, 41, 13. of x(T ). An increase in the concentration of MCF3 groups on 2 Z.Wu, C. U. Pittman Jr. and S. D. Gardner, Carbon, 1995, 33 597. the carbon fibre model results in an increase (+90%) in the 3 D.W. McKee and V. J. Mimeault, in Chemistry and Physics of value of x(T ). The substitution of hydrogen for methyl and Carbon, ed. P. L. Walker Jr and P. A. Thrower, Marcel Dekker, New York, 1973, vol. 8. then fluorine in low surface concentrations improves the 4 I. Hamerton, J. N. Hay, B. J. Howlin, J. R. Jones, S. Y. Lu, G. A. interaction, but at higher concentrations the more bulky tert- Webb, M.G. Bader, A. M. Brown and J. F. Watts, Chem. Mater., butyl group is preferred. 1997, 9, 1972. 5 Y. Da, D. Wang, M. Sun, C. Chen and J. Yue, Composites Sci. EVect of changing the polymer matrix T echnol., 1987, 30, 119. 6 Y. Xie and P. M. A. Sherwood, Carbon, 1994, 6, 650. Similar trends in interfacial adhesion were observed for poly- 7 W. W.Wright, Composite Polym., 1990, 3, 231 (part I); 1990, 3, 258 ethylene.However, the absence of polar groups in the polymer (part II). chain tends to decrease the interaction parameter x(T ). This 8 (a) J.M. Barton, G. J. Buist, A. S. Deazle, I. Hamerton, B. J. Howlin and J. R. Jones, Polymer, 1994, 35, 4326; (b) I. P. Aspin, J. M. is probably because being a graphitic structure the bulk of the Barton, G. J.Buist, A. S. Deazle, I. Hamerton, B. J. Howlin and carbon fibre surface is non-polar in nature and the surface J. R. Jones, J. Mater. Chem., 1994, 4, 385; (c) A. S. Deazle, C. R. groups are present at low concentration (<10% surface cover- Heald, B. J. Howlin, I. Hamerton and J. M. Barton, Polymer age),14 therefore making the carbon fibres more compatible Preprints, Japan (Engl.Edn.), 1995, 44, E11; (d) I. Hamerton, C. R. with the non-polar polyethylene chains. The thermodynamic Heald and B. J. Howlin, Macromol. T heory Simul., 1996, 5, 305; (e) work of adhesion can be expressed by eqn. (3),25 A. S. Deazle, I. Hamerton, C. R. Heald and B. J. Howlin, Polym. Int., 1996, 41, 151; ( f ) I. Hamerton, C. R. Heald and B. J. Howlin, Wa=cSV+cLV-cSL (3) Modell.Simul. Mater. Sci. Eng., 1996, 4, 151; (g) R. D. Allington, D. Attwood, I. Hamerton, J. N. Hay and B. J. Howlin, Composites where Wa is the thermodynamic work of adhesion, cSV is the Sci. T echnol., 1998, in the press. surface energy of the solid–vapour interface, cLV is the surface 9 J. W. Holubka, R. A. Dickie and J. C. Cassatta, J. Adhesion Sci. energy of the liquid–vapour interface, cSL is the surface energy T echnol., 1992, 6, 243.of the solid–liquid interface. The surface energy of the fibre 10 L. H. Lee, J. Adhesion, 1994, 46, 15. 11 A. R. Tiller and B. Gorella, Polymer, 1994, 35 (15), 3251. (cSV) should be greater than that of the matrix (cLV) for proper 12 D. Attwood and P. I. Marshall, Composites, Part A, 1996, 27, 775. wetting to occur.Better wetting can enhance the adhesive 13 A. Calderone, V. Parente and J. L. Bredas, Synth. Met., 1994, bond strength by increasing the thermodynamic work of 67, 151. adhesion Wa which is directly proportional to the fracture 14 I. Hamerton, J. N. Hay, B. J. Howlin, J. R. Jones, S. Y. Lu, G. A. energy of the adhesion bond.26 Thus, carbon fibres (typically Webb and M. G. Bader, J.Mater.Chem., 1997, 7, 169. with surface energy of 40–50 mJ m-2) can be wetted readily 15 B. Z. Jang, in Advanced Polymer Composites, ASM International, Ohio, 1994. by polymer matrices such as epoxy (with surface energy of 16 S. L. Mayo, B. B. Olafson and W. A. Goddard III, J. Phys. Chem., 43 mJ m-2) and polyethylene (with surface energy of 1990, 94, 8897. 31 mJ m-2).26 These analyses are consistent with our simu- 17 A.K. Rappe� and W. A. Goddard III, J. Phys. Chem., 1991, 95, 3358. lation results above. 18 Computational Instruments Property Prediction User’s Reference, CERIUS2, v.1.6, Molecular Simulations Inc., 1994. 19 P. J. Flory, Principles of Polymer Chemistry, Cornell University Conclusions Press, Ithaca, New York, 1953. 20 (a) M. Blanco, J. Comput. Chem., 1991, 12, 237; (b) C. F. Fan, B. D. The interfacial interaction between carbon fibres and polymer Olafson, M. Blanco and S. L. Hsu,Macromolecules, 1992, 25, 3667. matrices can be modelled in terms of steric bulk and hydrogen- 21 I. Hamerton, J. N. Hay, B. J. Howlin, J. R. Jones, S. Y. Lu, G. A. bonding potential of the substituent. There is a trade-oV Webb and M. G. Bader, High Perform. Polym., 1997, 9, 281. between these two parameters and, in practice, mild oxidation 22 K. Morita, K. Murata, A. Ishitani, K. Muragama, T.Ono and is the preferred route to enhance interfacial adhesion. More A. Nakajima, Pure Appl. Chem., 1986, 58, 455. 23 J. Shorter, in Advances in L inear Free Energy Ralationships, ed. polar groups (such as MCF3, MI and MNO2) give more N. B. Chapman and J. Shorter, Plenum Press, London, 1972, p. 71. favourable interactions, but we accept that these are not easily 24 P. W. Atkins, Physical Chemistry, Oxford University Press, accessible nor in some cases desirable with current technologies Oxford, 1991. for surface treatment. Whilst similar trends in interfacial 25 D. Hull, An Introduction to Composite Materials, Cambridge adhesion were observed for epoxy polymers and polyethylene, University Press, Cambridge, 1981, p. 38. the presence of polar groups in the polymer backbone tends 26 F. Hoecker and J. Karger-Kocsis, J. Appl. Polym. Sci., 1996, 59, 139. to reduce the interaction. This is probably because being a graphitic structure the bulk of the carbon fibre surface is non- Paper 7/09161C; Received 22nd December 1997 J. Mater. Chem., 1998, 8(6), 1333–1337 13

ISSN:0959-9428    

DOI:10.1039/a709161c

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 Liquid crystal trimers. The synthesis and characterisation of the 4,4¾-bis[v-(4-cyanobiphenyl-4¾-yloxy)alkoxy]biphenyls Corrie T. Imrie*a† and GeoVrey R. Luckhurstb aDepartment of Chemistry, University of Aberdeen, Meston Walk, Old Aberdeen, Scotland, UK AB24 3UE bDepartment of Chemistry, University of Southampton, Southampton, England, UK SO17 1BJ The synthesis and characterisation of a homologous series of liquid crystal trimers, the 4,4¾-bis[v-(4-cyanobiphenyl-4¾- yloxy)alkoxy]biphenyls, is reported in which the length of the flexible spacers is varied from 3 to 12 methylene units.All the members of the series exhibit enantiotropic nematic behaviour and, in addition, monotropic smectic A behaviour is observed for compounds with spacers containing from 4 to 11 methylene units.The formation of a smectic phase in these compounds is attributed to a specific interaction between the unlike mesogenic groups, namely, the central biphenyl and the terminal cyanobiphenyl units. The nematic–isotropic transition temperatures and the associated entropy changes exhibit a dramatic odd–even eVect as the length and parity of the spacers is varied, in which the even members exhibit the higher values. This behaviour is interpreted in terms of the geometry as well as the flexibility of the spacers and how these control the average molecular shape.A comparison of the magnitudes of these odd–even eVects with those for the analogous dimeric series suggests that in the nematic phase the mesogenic units in the trimers are correlated to the same extent as in the dimers but that their orientational ordering is significantly higher.Semi-flexible main-chain liquid-crystalline polymers are composed of mesogenic units linked via flexible alkyl spacers.1,2 The transitional properties, in particular, the clearing tempera- NC OCnH2 n+1 ture and the associated entropy change, exhibited by such a nOCB polymer are strongly dependent on the length and parity of the alkyl spacers.3 In contrast, the transitional properties of the corresponding dimers, the a,v-bis(4-cyanobiphenyl-4¾- conventional low molar mass mesogens, which are composed yloxy)alkanes, have been extensively investigated.8,21–23 The of molecules possessing a single semi-rigid core, attached to which are one or two alkyl chains, are modified to a much smaller extent by changes in the length and parity of the alkyl chain.4 If just two mesogenic units, however, are linked through a flexible spacer, yielding the so-called liquid crystal dimers,5,6 NC O(CH2) nO CN BCBOn then the transitional properties are found to be critically dependent on the length and parity of the alkyl spacer in a acronym used to refer to the monomers is nOCB where n manner reminiscent of the behaviour observed for polymeric denotes the number of methylene units in the terminal chain systems.7,8 In order to investigate how these properties evolve while the dimers are described by BCBOn in which n now from the dimers to the polymers a small number of higher refers to the number of methylene units in the flexible spacer. oligomers have been prepared and characterised.The majority The liquid crystalline behaviour of several members of the of these are described as liquid crystal trimers as they contain TCBOn9,15 series has been reported in the literature but to the molecules comprising three mesogenic groups and two flexible best of our knowledge this is the first report describing a spacers9–15 although liquid crystal tetramers containing four homologous series of trimers in which as many as ten homolmesogenic groups and three spacers have also been ogues have been characterised.reported.16,17 In these examples the mesogenic units are linked in terminal positions but a range of other molecular architec- Experimental tures are also possible.18–20 To extend such studies we have prepared and studied the Synthesis transitional behaviour of a homologous series of trimeric mesogens, the 4,4¾-bis[v-(4-cyanobiphenyl-4¾-yloxy)alkoxy]bi- The trimers were prepared using the synthetic route shown in Scheme 1.The synthesis of the a-bromo-v-(4-cyanobiphenyl- phenyls, and use the acronym TCBOn to describe them; n NC O(CH2) nO O(CH2) nO CN TCBOn refers to the number of methylene units in the flexible alkyl 4¾-yloxy)alkanes, 1, has been described in detail elsewhere.24 The coupling of 1 with 4,4¾-dihydroxybiphenyl was performed spacers. This particular series was selected because both the analogous monomers, the 4-cyano-4¾-alkoxybiphenyls, and using potassium carbonate as base in dimethylformamide. The reaction mixture was refluxed for 12 h, allowed to cool and then poured into water.The resulting precipitate was filtered †E-mail: c.t.imrie@abdn.ac.uk J. Mater. Chem., 1998, 8(6), 1339–1343 1339NC OH NC O(CH2) nBr HO OH NC O(CH2) nO O(CH2) nO CN + Br(CH2) nBr K2CO3 Acetone K2CO3 DMF 1 Scheme 1 oV and dried. The crude trimer was recrystallized at least twice point singularities when viewed through the polarising microscope, while for the smectic A phase coexisting regions of focal with hot filtration from large volumes of either toluene or ethyl acetate.The trimers were too insoluble to allow for conic fan and homeotropic textures were observed. The monotropic nature of the smectic phase precluded the possibility of structural characterisation using NMR spectroscopy, but their IR spectra were consistent with the proposed structures.X-ray diVraction studies in order to determine the layer spacings. The nematic–isotropic temperatures listed in Table 1 Specifically, the IR spectra in all cases contained a band corresponding to the cyano stretch while there was no evidence are in good agreement with those given in the literature.9,15 The smectic A phase exhibited by the TCBOn series has been to suggest the presence of unreacted hydroxy groups.In addition mass spectroscopy identified the molecular ion. The overlooked in the previous studies9,15 and a plausible explanation for this is the monotropic nature of the phase. It is high level of purity of the trimers is also reflected in the narrow temperature ranges observed for the melting transitions and noteworthy, however, that the interpretation of the electrooptical properties of TCBO10 invoked smectic fluctuations also in the good agreement with the data reported for members of this series in the literature.9,15 within the nematic phase.15 The dependence of the transition temperatures on the number of methylene units, n, in the flexible alkyl spacers for Thermal characterisation.The thermal properties of the trimers were investigated by diVerential scanning calorimetry the TCBOn series is shown in Fig. 1. It is immediately apparent that the melting points, the nematic–isotropic temperatures using a Perkin-Elmer DSC-2C diVerential scanning calorimeter interfaced to an Opus PC II microcomputer.The optical and the smectic A–nematic temperatures all depend critically on the length and parity of the flexible spacers. Specifically, textures of the mesophases were examined by polarised light microscopy using a Nikon polarising microscope equipped the melting points exhibit a pronounced alternation which does not attenuate on increasing n. The nematic–isotropic and with a Linkam hot stage.The transition temperatures quoted for the trimers were extracted from the calorimetric data with smectic A–nematic transition temperatures also show a dramatic odd–even eVect although in both cases it is attenuated the exception of the smectic A–nematic transition temperatures which were measured using the optical microscope because on increasing n. For even members the stability of the smectic A phase decreases with increasing n while for odd members the samples crystallised during the calorimetric analysis.the thermal stability of the phase passes through a weak maximum. Thus, increasing the spacer length does not strongly Results and Discussion The transitional properties of the trimers are listed in Table 1. All ten members exhibit an enantiotropic nematic phase and also, with the exception of the propyl and dodecyl members, a monotropic smectic A phase.The nematic phases have a schlieren optical texture containing both two and four brush Table 1 The transitional properties of the TCBOn homologous series n TCN/°C TSmAN/°C TNI/°C DSCN/R DSNI/R 3 179 202 10.5 0.81 4 226 (196) 297 18.4 3.51 5 173 (124) 215 18.7 1.15 6 224 (190) 262 23.9 3.93 7 157 (130) 206 19.9 1.40 8 198 (175) 231 20.2 4.08 9 152 (133) 196 21.9 1.86 Fig. 1 The dependence of the transition temperatures on the number 10 191 (155) 213 23.9 4.57 of methylene units (n) in the flexible alkyl spacers for the TCBOn 11 144 (127) 184 20.2 2.25 series: (#) melting points, ($) nematic–isotropic transitions and 12 183 194 26.0 4.64 (%) monotropic smectic A–nematic transitions 1340 J.Mater. Chem., 1998, 8(6), 1339–1343promote smectic behaviour. This is similar to the behaviour observed for symmetric liquid crystal dimers25 but quite unlike that of semi-flexible main chain liquid crystal polymers for which increasing spacer length enhances smectic behaviour.1,2 Thus the driving force for smectic phase formation must diVer between the dimers and trimers, and the polymers.Presumably for the polymers this driving force must be an entropic one in order to disentangle the polymer chains. In considering possible molecular factors for the observation of smectic behaviour for the TCBOn series, it is interesting to note that the analogous dimers, the BCBOn series, are exclusively nematics.8 Indeed, the smectic A–nematic temperatures exhibited by the trimers are generally higher than the temperatures to which the nematic phases of the dimers can be supercooled.This implies that smectic phase formation is more favourable for the trimers. The trimers contain two electron Fig. 2 (#) The dependence of the entropy change associated with the nematic–isotropic transition on the number of methylene units (n) in deficient moieties, the cyanobiphenyl groups, and a central the flexible alkyl spacers for the TCBOn series.Also shown are the electron rich biphenyl unit. Smectic phase formation in nonnematic –isotropic entropies for (%) the BCBOn series8 and (6) the symmetric liquid crystal dimers containing electron rich and nOCB series.29 Filled symbols denote smectic A–isotropic transitions electron poor moieties is attributed to the specific interaction for the monomers.between the unlike mesogenic units.26,27 The precise nature of this interaction is unclear and most recently it has been suggested that it is electrostatic quadrupolar interactions then considered to be more compatible with the molecular organisation found in the nematic phase than for the odd- between groups with quadrupole moments which are opposite in sign.28 A similar explanation would serve to rationalise the membered dimers and it is this greater compatibility which results in, for example, the higher nematic–isotropic entropies observation of smectic behaviour for the trimers and not for the dimers.For the trimers, therefore, the driving force respon- found for the even-membered dimers.Such an argument neglects the flexibility of the spacer and a more realistic sible for smectic behaviour may be the specific interaction between the electron rich biphenyl unit and the cyanobiphenyl interpretation of the dependence of the transitional properties on the parity of the spacer certainly includes a wide range of moieties of high electron aYnity. This view is supported by the transitional properties of four homologues of a closely conformations and not solely the all-trans conformation.5 In the isotropic phase approximately half the conformers of an related trimeric series to the TCBOn series,12 which exhibit NC O(CH2) nO N=N O(CH2) nO CN Trimeric series exclusively smectic behaviour.The central azobenzene unit is even-membered dimer are essentially linear whereas for an odd membered dimer just 10% are linear.There exists a synergy more electron rich than the biphenyl unit in the TCBOn series and it is reasonable to assume that the specific interaction between conformational and orientational order and hence at the transition to the nematic phase for even-membered dimers between the quadrupole moments of the unlike groups will be more favourable.Thus it would be expected, and indeed many of the bent conformers are converted to a linear form. This enhances the orientational order of the nematic phase observed, that smectic behaviour would be enhanced for these trimers relative to the corresponding members of the TCBOn resulting in a larger nematic–isotropic entropy than would be expected for a monomer.For odd-membered dimers, however, series. By analogy to non-symmetric dimers, this driving force for smectic phase formation should result in the formation of the diVerence in free energy between the bent and linear conformers is such that the orientational order of the nematic intercalated phases26 but unfortunately, and as we noted earlier, it was not possible to investigate the structure of the phase is insuYcient to convert bent into linear conformers. Hence, the orientational order is not enhanced and a smaller smectic A phase using X-ray diVraction.Hence, it would be unwise to speculate further on the local structure within nematic–isotropic entropy would be expected. Models have been developed based on a molecular field to describe this the phase. The entropy change associated with the nematic–isotropic synergy between conformational and orientational order which are able to predict most of the properties of nematic liquid transition for the TCBOn series also exhibits a dramatic alternation as n is increased (Fig. 2). The values for the even crystal dimers.5 It is reasonable to assume that a similar treatment of liquid crystal trimers would also successfully members are typically 2–3 times larger than those of the odd members.This odd–even eVect does not attenuate on increasing predict their transitional behaviour and such calculations will be reported elsewhere. n, although it may be argued that it is attenuated in a relative sense since the values of DSNI/R increase with increasing n but The dependence of the melting points on n for the monomeric, dimeric and trimeric series is shown in Fig. 3. The the diVerence between the n and (n+1) homologues appears to be approximately constant. trimers exhibit the highest melting points with the single exception of the propane homologues for which the melting Thus we have seen that the transitional properties of the trimers depend critically on the length and parity of the spacer point of BCBO3 is just higher than that of TCBO3. The strong alternation in the melting points of the dimers and trimers is in a manner reminiscent of both dimers and the semi-flexible main chain polymers. For the dimers this behaviour is most not evident in those of the monomers; it is interesting to note that this pronounced alternation in the melting points is only often attributed to the dependence of the molecular shape on the parity of the spacer considered in the all-trans confor- found for nematic dimers and is not observed for smectic dimers.25 It was suggested that for nematic dimers this may mation.Specifically, in an even-membered dimer the mesogenic groups are antiparallel whereas in an odd-membered dimer indicate that the change in the conformation statistical weights of the spacer on melting is small for even-membered spacers they are inclined.This structure for even-membered dimers is J. Mater. Chem., 1998, 8(6), 1339–1343 1341dimeric and trimeric series. For any given value of n, the trimer has the highest clearing point, then the dimer, while the monomer invariably exhibits the lowest clearing temperature. This trend is to be expected given the enhanced shape anisotropy on passing from the monomeric to the dimeric and trimeric structures.Fig. 2 shows the dependence of the entropy change associated with the liquid crystal–isotropic transition for the trimers, dimers8 and monomers.29 Both odd and even members of the trimeric TCBOn series exhibit higher, typically twice as large, values of DSNI/R than observed for the analogous dimer.The size of the alternation exhibited by DSNI/R on increasing n for the trimers is also greater than for the dimers. The values of DSNI/R exhibited by the monomers are considerably smaller than either the dimers or the trimers and in addition, do not exhibit the dramatic alternation seen for the BCBOn and Fig. 3 The dependence of the melting points on the number of TCBOn series. methylene units (n) in the alkyl chains for (#) the TCBOn series, (%) The question arises, however, when we compare the entropies the BCBOn series8 and (6) the nOCB29 series. Filled symbols denote crystal–smectic A transitions. and nematic–isotropic temperatures for the monomers, dimers and trimers how or indeed whether we should normalise the results to allow for the diVerent number of mesogenic units.but large for odd-membered compounds.8 This would also be In order to address this issue, Fig. 6 shows the dependence of a reasonable assumption to make for the trimeric compounds. the ratios of the TNIs for the dimers to those of the monomers An alternative explanation, however, considers enthalpic and the trimers to those of the dimers on the number of carbon eVects.Thus at the root of this odd–even eVect is possibly the atoms in the terminal chain or the flexible spacers. The diYculty that the odd membered compounds, with their bent dimer/monomer ratio exhibits a pronounced odd–even eVect conformations, experience in packing eYciently into a crystalwhich reflects the much larger alternation exhibited by the line structure as compared with the more elongated evendimers when compared with the monomers.In contrast, the membered trimers, see Fig. 4. ratios for the trimer/dimer transition temperatures exhibit a Fig. 5 compares the clearing temperatures of the monomeric, weak odd–even eVect and this is a result of the strong alternations exhibited by both sets of compounds cancelling each other out.A far more dramatic increase in TNI is evident on passing from the monomer to the dimer as compared with going from dimer to trimer, as is clear from the higher values of the ratios of the transition temperatures, see Fig. 6. Fig. 7 compares the ratios of the entropies associated with the nematic–isotropic transition and again a dramatic odd–even eVect is observed for the dimer/monomer ratios whereas the trimer/dimer values exhibit a very weak dependence on the length of the flexible spacers.It should be noted that the dimer/monomer ratio for n=3 is probably too large, resulting from an underestimation of DSNI/R for 3OCB.29 We note, however, that the clearing temperatures and entropies of the monomers do not show any significant dependence on n on the scale of that seen for the dimers and trimers and hence, the dimer/monomer ratios essentially reflect the behaviour of the dimers.A comparison of Fig. 6 and 7 reveals that the change in DSNI/R on going from the dimer to the trimer is Fig. 4 Schematic representation of the all-trans conformation of (a) an larger than that seen in TNI.The ratios of TNI all lie in the even and (b) an odd membered trimer range 1.05–1.09 whereas for the entropies the corresponding Fig. 5 The dependence of the clearing points on the number of Fig. 6 The dependence of the ratio of the nematic–isotropic transition methylene units (n) in the alkyl chains for (#) the TCBOn series, (%) the BCBOn series8 and (6) the nOCB29 series. Open symbols denote temperatures for ($) the BCBOn series to those of the nOCB series,29 and for (#) the TCBOn series to those of the BCBOn series,8 on the nematic–isotropic transitions and filled symbols smectic A–isotropic transitions for the monomers.number of carbon atoms (n) in the alkyl chains 1342 J.Mater. Chem., 1998, 8(6), 1339–1343vol. 2B, ed. D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess and V. Vill, Wiley-VCH,Weinheim, 1998, ch. 10, pp. 801–834. 6 G. R. Luckhurst, Macromol. Symp., 1995, 96, 1. 7 A. C. GriYn and T. R. Britt, J. Am. Chem. Soc., 1981, 103, 4957. 8 J. W. Emsley, G. R. Luckhurst, G. N. Shilstone and I. Sage, Mol. Cryst. L iq. Cryst. L ett., 1984, 102, 223. 9 H.Furuya, K. Asahi and A. Abe, Polym. J., 1986, 18, 779. 10 G. S. Attard and C. T. Imrie, L iq. Cryst., 1989, 6, 387. 11 R. Centore, A. Roviello and A. Sirigu,Mol. Cryst. L iq. Cryst., 1990, 182B, 233. 12 T. Ikeda, T. Miyamoto, S. Kurihara, M. Tsukada and S. Tazuke, Mol. Cryst. L iq. Cryst., 1990, 182B, 357. 13 A. T. Marcelis, A. Koudijs and E. J. R. Sudho� lterr, L iq.Cryst., 1996, 21, 87. 14 A. T. Marcelis, A. Koudijs and E. J. R. Sudho� lterr, L iq. Cryst., 1995, 18, 851. 15 N. V. Tsvetkov, V. V. Zuev and V. N. Tsvetkov, L iq. Cryst., 1997, 22, 245. Fig. 7 The dependence of the ratio of the nematic–isotropic entropies 16 A. C. GriYn, S. L. Sullivan and W. E. Hughes, L iq. Cryst., 1989, for ($) the BCBOn series to those of the nOCB series, and for (#) 4, 677.the TCBOn series to those of the BCBOn series, on the number of 17 C. T. Imrie, D. Stewart, C. Remy, D. W. Christie, I. W. Hamley, carbon atoms (n) in the alkyl chains R. Harding and J. Pople, unpublished work. 18 D. Demus, L iq. Crys., 1989, 5, 75. 19 K. Zab, D. Joachimi, E. Novotna, S. Diele and C. Tschierske, L iq. range is 1.5–2.4. This small increase in TNI suggests that the Cryst., 1995, 18, 631.mesogenic units in the trimers are correlated to the same 20 V. Percec and M. Kawasumi, J.Mater. Chem., 1993, 3, 725. extent as in the dimers whereas the larger increase in DSNI/R 21 J. W. Emsley, G. R. Luckhurst and G. N. Shilstone, Mol. Phys., suggests a significant increase in the orientational order of the 1984, 53, 1023. mesogenic groups on passing from the dimer to trimer. 22 J.W. Emsley, G. R. Luckhurst and B. A. Timimi, Chem. Phys. Alternatively, if the results were scaled according to the number L ett., 1985, 114, 19. 23 D. A. Dunmur and M. R. Wilson, J. Chem. Soc., Faraday T rans. 2, of mesogenic units, the value 3/2 might be expected. These 1988, 84, 961. observations, however, are based on just a single set of 24 G. S. Attard, C. T. Imrie and F. E. Karasz, Chem. Mater., 1992, materials and much research is now required to establish the 4, 1246. generality or otherwise of their behaviour. 25 R. W. Date, C. T. Imrie, G. R. Luckhurst and J. M. Seddon, L iq. Cryst., 1992, 12, 203. 26 J. L. Hogan, C. T. Imrie and G. R. Luckhurst, L iq. Cryst., 1988, References 3, 645. 27 G. S. Attard, S. Garnett, C. G. Hickman, C. T. Imrie and L. Taylor, 1 H. Finkelmann, in T hermotropic L iquid Crystals, ed. G. W. Gray, L iq. Cryst., 1990, 7, 495. Wiley, Chichester, 1987, ch. 6. 28 A. E. Blatch, I. D. Fletcher and G. R. Luckhurst, L iq. Cryst., 1995, 2 C. K. Ober, J.-I. Jin and R. W. Lenz, Adv. Polym. Sci., 1984, 59, 103. 18, 801. 3 A. Blumstein and O. Thomas, Macromolecules, 1982, 15, 1264. 29 C. T. Imrie, F. E. Karasz and G. S. Attard, Macromolecules, 1993, 4 G. W. Gray, in T he Molecular Physics of L iquid Crystals, ed. 26, 3803. G. R. Luckhurst and G. W. Gray, Academic Press, London, 1979, ch. 1. 5 C. T. Imrie and G. R. Luckhurst, in Handbook of L iquid Crystals, Paper 8/01128A; Received 9th February, 1998 J. Mater. Chem., 1998, 8(6), 1339–1343

ISSN:0959-9428    

DOI:10.1039/a801128a

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 Thermotropic properties of monosubstituted ferrocene derivatives bearing bidentate N-benzoyl-N¾-arylthiourea ligands—novel building blocks for heterometallic liquid crystal systems† Tarimala Seshadri* and Hans-Ju�rgen Haupt Department of Inorganic and Analytical Chemistry, University of Paderborn, 33098 Paderborn, Germany Ferrocene-based derivatives such as 4-{3-[4-(octyloxy)benzoyl]thioureido}phenyl 4-ferrocenylbenzoate and other higher homologues (n=12,16 and 18; n=length of alkoxy chain) were prepared by reacting 4-alkoxybenzoyl isothiocyanates with the corresponding amines containing the ferrocenyl moiety.Their mesomorphic properties were investigated by means of polarized optical microscopy and diVerential scanning calorimetry. All the compounds exhibit enantiotropic nematic phases and the nematic range increases with increasing terminal alkyl chain length.On cooling, the nematic phase persists below 0 °C in the first three compounds and in the case of n=18, a phase transformation, possibly to the SC phase, around 72 °C during cooling was observed. In all cases, a glass transition was observed around Tg=18–35 °C, which is remarkable for low molecular mass calamitic metallomesogen systems.Studies focussed on the synthesis and mesomorphic properties have been investigated and found to exhibit enantiotropic SC and SA phases. A few of them are monotropic. The results on of ferrocene-based metallomesogens have gained increasing importance in recent years because of their novel thermal, these will be communicated separately.We report in this paper the synthesis and liquid crystalline optical and magnetic properties and they are well documented. 1,2 Ferrocene, because of its aromatic character, facili- properties of monosubstituted ferrocene derivatives bearing BATU ligands which form the building blocks for heterometal- tates several substitution reactions whereby a range of low molecular mass calamitic (rod-like) systems can be prepared lic liquid crystal systems.Initially, we prepared ferrocene derivatives similar to BATU by monosubstitution, or 1,1-, 1,2-, 1,3-disubstitution or 1,1,3- trisubstitution of the ferrocene nucleus.3,4 Little work has been with two benzene rings in the mesogenic core. The fact that they did not exhibit any liquid crystalline properties clearly reported on the monosubstituted-ferrocene derivatives in recent years.This is attributed partly to their unfavourable indicates the negative influence of the bulky metallocene in the system. Hence, we planned addition of a third ring into the molecular shape (L-shaped geometry) and also to the repulsive steric eVects of the ferrocene unit reducing the ability of the core.In this connection we evaluated the diVerent methods (see Fig. 2) of attaching mesogenic units to the ferrocene molecules to be placed in layers thus mostly favouring the formation of nematic and smectic A phases. Fig. 1 shows some moiety. Imrie7 observed that the nature of groups immediately adjacent to the ferrocenyl moiety is important, especially the of the monosubstituted ferrocene derivatives reported in recent years.electron withdrawing ability of the carbonyloxy group in (a), which may weaken greatly the chemical and thermal stability However, ferrocene derivatives with additional chelating groups used to attain homo- and hetero-metallic complexes of the resulting compounds. To circumvent this problem, we have synthesized only the have not been reported, except in the work of Galyametdinov5 (see Fig. 1); such systems would allow investigations of the last two series (b) and (c) of monosubstituted ferrrocene derivatives bearing BATU chelating groups. We report in this metal–metal interactions in the vicinity and through the p ligand system to be made. The ferrocene subunit in such paper only the series (b) shown in Fig. 2, where the phenyl group is placed between the ferrocenyl moiety and the car- systems might function as a redox switch by further complexation with suitable metal ions.We have selected N-benzoyl- bonyloxy link of the mesogenic core. The synthetic route to prepare these compounds is shown in Scheme 1. The prep- N¾-arylthiourea (BATU) ligands for this purpose.aration of the final compounds is very simple; they are abbreviated hereafter as Fc-BATU-n, where n is the number of carbon atoms in the flexible alkyl chain. C O NH C S NH N-benzoyl-N¢-arylthiourea Experimental These ligands possess strong donor groups, namely the General amide and the thiourea group. The BATU ligands react with The solvents acetone and dichloromethane were distilled over transition metal ions mostly in monoanionic and bidentate CaCl2 and P2O5 prior to use.Benzene was distilled over form by deprotonation, resulting in neutral complexes with sodium. All the reactions were carried out under argon. We S,O-coordination. The complex forming properties of these used for column chromatography (CC) silica gel 60 (70–230 ligand systems have been intensively investigated.6 We have mesh) and thick-layer chromatography (TPC), silica gel plates synthesized several 4-alkloxy-substituted BATU ligands with prepared in our laboratory.Transition temperatures (onset increasing alkyl chain length and their mesomorphic properties point of endotherm or exotherm) and enthalpies were determined with a diVerential scanning calorimeter (Mettler DSC- †Presented at the International Symposium on metallomesogens, 30) connected to a Mettler TA 4000 processor, rate 10 °C min-1 3–6 June 1997, University of Neuchatel, Neuchatel, Switzerland.*E-mail: ts@chemie.uni-paderborn.de under nitrogen; treatment of data used Mettler TA 72.2/.5 J. Mater. Chem., 1998, 8(6), 1345–1350 1345SO3 – H2 n+1CnO O O C H N O O C O O (CH2)6 O O (CH2)6 O C CH O CH2 CO2 CO2 OC10H21 O O O O OC8 H17 F F C 152 N 252 I (dec.) Loubser and Imrie, ref. 3( c ) G 37 SA83 I Deschenaux, ref. 3( i ) Malthete, ref. 3( b) N HC HO O2C OC10H21 N C HO O2C OC10H21 H CuO O C N H CO2 C10H21O Fe C 221 N 230 I Galyametdinov, ref. 5( b ) C 140 N 184 I Galyametdinov, ref. 5( b ) Fe n = 8 C 153 N 167 I n = 10 C 143 N 159 I Fe + Fe C 43 SA 51 I Tanaka, ref. 3( h ) Fe Fe Fe Fig. 1 with ferrocene and subsequent hydrolysis. A mixture of the acid (2.85 g, 9.31 mmol), 4-nitrophenol (1.29 g, 9.31 mmol), N,N¾-dicyclohexylcarbodiimide (1.96 g, 9.31 mmol), and 4-pyrrolidinopyridine (0.1 g) in 100 ml of dry dichloromethane was stirred at room temperature for 48 h. The reaction mixture was filtered oV and the solvent was removed in vacuo; the solid obtained was recrystallized from ethanol.Yield: 3.54 g (89%). dH (CDCl3) 8.35 (d, 2H, Ar), 8.12 (d, 2H, Ar), 7.65 (d, 2H, Ar), 7.40 (d, 2H, Ar), 4.78 (t, 2H, C5H4), 4.46 (t, 2H, C5H4), 4.07 (s, C5H5). n/cm-1 (FTIR: 1720 (CNO). 4-Aminophenyl 4-ferrocenylbenzoate 4 A mixture of the nitro derivative obtained above (4.10 g, 10 mmol), 10% Pd/C (0.5 g) and dioxane (50 ml ) was stirred under H2 in an autoclave for several hours till no more H2 was consumed and then filtered through Celite.Removal of the solvent resulted in a solid residue which was recrystallized from ethanol. Yield: 3.30 g (90%). n/cm-1 (FTIR) 3405 (s), nas OR C O O C O O OR (CH2)11 OR O C O O (c) (b) (a) MU Fe MU Fe MU Fe (NH2), 3330 (sb), nsy (NH2), 1726 [n(CNO)]. dH (CDCl3) 8.09 Fig. 2 DiVerent ways of attaching mesogenic units to a ferrocenyl (d, 2HAr), 7.57 (d, 2H, Ar), 7.03 (d, 2H, Ar), 6.74 (d, 2H, Ar), moeity. L-shape: 1-substitution where MU=mesogenic unit; R= 4.75 t, 2H, C5H4), 4.42 (t, 2H, C5H4), 4.06 (s, 5H, C5H5), 3.75 terminal alkylchain. ( br, 2H, NH2). GRAPHWARE. Optical studies were conducted using a Zeiss- Acioscop polarizing microscope equipped with a Linkam- Preparation of methyl 4-hexadecyloxybenzoate THMS-600 variable temperature stage under nitrogen. 1H A mixture of 1-bromohexadecane(0.11 mol), methyl 4- NMR spectra were recorded on a Bruker AMX 300 spechydroxybenzoate (0.1 mol), potassium carbonate (0.2 mol) and trometer and IR spectra on a Nicolet P 510 FTIR spectrometer. a few mg of potassium iodide in 250 ml of acetone was heat A Perkin Elmer Microanalyser PE 2400 was used for the under reflux for 72 h and allowed to cool to room temperature.elemental analyses. The solvent was removed under reduced pressure and the residue was taken up in 200 ml of water, extracted with CH2Cl2 Synthesis and dried over MgSO4. The crude product obtained after removal of solvent was recrystallized from ethanol.Yield: 80%. 4-Nitrophenyl 4-ferrocenylbenzoate 3 dH (CDCl) 8.0 (d, 2H, Ar), 6.9 (d, 2H, Ar), 4.05 (t, 2H, OCH2), 3.88 (s, 3H, OCH3), 1.85 (m, 2H, CH2), 1.30 (m, 28H), 0.9 (t, The starting compound, 4-ferrocenylbenzoic acid,8 was prepared from ethyl p-aminobenzoate by diazotization, coupling 3H, CH3). 1346 J. Mater. Chem., 1998, 8(6), 1345–1350C O O NHC NH C OCn H2 n+1 Fe Fe CO2Et CIN2 CO2Et Fe CO2H KOH HCI 1 2 Fe C 3 NO2 HO DCC O O NO2 Fe C 4 O O NH2 Pd(10%)C H2 C H2 n+1CnO NCS O 5 Fe O 6 ( n = 8,12,16,18) Fc-BATU- n S Scheme 1 4-Hexadecyloxybenzoic acid 1.20–155 (m, 26H), 0.9 (t, 3H, CH3).n/cm-1 (FTIR) 1988, [n(NNCNS)] 1685, [n(CNO)]. The ester obtained above (5.6 g, 15 mmol) was dissolved in The other 4-alkyloxybenzoyl isothiocyanates used here were 60 ml of dioxane and KOH (1.7 g, 30 mmol) in 2 ml of water prepared in 40–50% yields by following the same method and the mixture was heated under reflux for 8 h.The solvent mentioned above using the corresponding acid chlorides. was removed under reduced pressure and the residue was taken up in water and acidified with 6 M HCl. The resulting General procedure for preparing 4-ferrocenylbenzoate precipitate was filtered oV and washed with water and dried derivatives (Fc-BATU-n) 6 in air. Yield: 4.30 g (80%).To a benzene solution (5 ml) containing 256 mg (0.5 mmol) of 4-Hexadecyloxybenzoyl chloride 4-aminophenyl 4-ferrocenylbenzoate was added dropwise the corresponding 4-alkyloxybenzoyl isothiocyanate(0.5 mmol) in To 4-hexadecyloxybenzoic acid (3.6 g, 10 mmol) in 50 ml of 10 ml of benzene and the reaction mixture was stirred at room CH2Cl2 was added SOCl2 (2.38 g, 20 mmol) and a few drops temperature for 2 h.Addition of methanol facilitated the of dimethylformamide. The reaction mixture was refluxed precipitation of the desired compound which was then filtered overnight and the excess thionyl chloride was removed under oV.The residue was recrystallized from CH2Cl2–heptane and vacuum. dH (CDCl3) 8.10 (d, 2H, Ar), 6.9 (d, 2H, Ar), 4.05 (t, dried over P2O5. Yields: 80–85%. 2H, OCH2), 1.85 (m, 2H, CH2), 1.20–1.60 (m, 30H), 0.9 (t, 3H, CH3). 4-{3-[4-(Octyloxy)benzoyl]thioureido}phenyl 4-ferrocenylbenzoate (Fc-BATU-8). Elemental analysis: Calc. for 4-Hexadecyloxybenzoyl isothiocyanate 5 C39H40N2O4SFe: C, 68.02; H, 5.85 and N, 4.07%; Found: C, 67.92; H, 5.80 and N, 4.12%.dH (CDCl3) 0.89 (t, 3H, CH3), The crude acid chloride which is free from SOCl2 obtained 1.30–1.60 (m, 26H), 1.83 (m, 2H, CH2), 4.05 (t, 2H, OCH2), above was suspended in dry acetone (100 ml) and to it was 4.07 (s, 5H, C5H5), 4.44 (s, 2H, C5H4), 4.76 (s, 2H, C5H4), 7.10 added potassium thiocyanate (1.94 g, 20 mmol) and refluxed (d, 2H, Ar), 7.27 (d, 2H, Ar), 7.58 (d, 2H, Ar), 7.79–7.88 (dd, for 4 h.After cooling, the reaction mixture was filtered and the 4H, Ar), 8.13 (d, 2H, Ar), 9.04 (s, 1H, MNHMCNO), 12.74 (s, solvent was removed. The residue was treated with benzene 1H, NH-aryl). and filtered again to remove the solid formed. The benzene solution was evaporated under vacuum and the oily residue was chromatographed (silica gel, CH2Cl2) to give a pure light- 4-{3-[4-(Dodecyloxy)benzoyl]thioureido}phenyl 4-ferrocenylbenzoate (Fc-BATU-12). Elemental analysis: Calc.for yellow compound. Yield: 1.75 g (40%). dH (CDCl3) 8.05 (d, 2H, Ar), 6.95 (d, 2H, Ar), 4.05 (t, 2H, OCH2), 1.85 (m, 2H, CH2), C43H48O4N2SFe: C, 69.36; H, 6.50 and N, 3.76%; Found C, J. Mater. Chem., 1998, 8(6), 1345–1350 134769.73, H, 6.45 and N, 3.81%.dH (CDCl) 0.89 (t, CH3), 1.28–1.63 (m, 18H), 1.85 (m, 2H, CH2), 4.05 (t, 2H, OCH2), 4.07 (s, 5H, C5H4), 4.43 (s, 2H, C5H4), 4.76 (s, 2H, C5H4), 7.0 (d, 2H, Ar), 7.32 (d, 2H, Ar), 7.61 (d, 2H, Ar), 7.82–7.88 (dd, 4H, Ar), 8.12 (d, 2H, Ar), 9.05 (s, 1H, MNHMCNO), 12.75 (s, 1H, MNHMaryl). 4-{3-[4-(Hexadecyloxy)benzoyl]thioureido}phenyl 4-ferrocenylbenzoate (Fc-BATU-16).Elemental analysis: Calc. for C47H56N2O4SFe: C, 70.47, H: 7.05 and N, 3.50%; Found C, 70.37; H, 7.04 and N, 3.58%. dH (CDCl3) 0.91 (t, 3H, CH3), 1.30–1.65 (m, 26H), 1.83 (m, 2H, CH2), 4.03 (t, 2H, OCH2), 4.07 (s, 5H, C5H5), 4.44 (s, 2H, C5H4), 4.76 (s, 2H,C5H4), 7.02 (d, 2H, Ar), 7.27 (d, 2H, Ar), 7.61 (d, 2H, Ar), 7.79–7.88 (dd, 4H, Ar), 8.13 (d, 2H, Ar), 9.07 (s, 1H, MNHMCNO), 12.75 (s, Fig. 3 EVect of increasing chain length, n 1H, NHMaryl). 4-{3-[4-(Octadecyloxy)benzoyl]thioureido}phenyl 4-ferrocenylbenzoate (Fc-BATU-18). Elemental analysis: Calc. for C49H60N2O4SFe: C, 71; H, 7.29 and N, 3.38%; Found: C, 71.05; H, 7.30; N, 3.38%. dH (CDCl3) 0.89 (t, 3H, CH3), 1.27–1.61 (m, 30H), 1.83 (m, 2H, CH2), 4.03 (t, 2H, OCH2), 4.07 (s, 5H, C5H5), 4.44 (s, 2H, C5H4), 4.76 (s, 2H, C5H4), 7.02 (d, 2H, Ar), 7.29 (d, 2H, Ar), 7.59 (d, 2H, Ar), 7.83–7.88 (dd, 4H, Ar), 8.12 (d, 2H, Ar), 9.05 (s, MNHMCNO), 12.74 (s, 1H, MNHMaryl).Results and Discussion These compounds can be readily prepared by reacting the 4- alkoxybenzoyl isothiocyanates with the corresponding aminecontaining ferrocenyl moiety. The completion of the reaction can be easily monitored by FTIR by the disappearance of stretching and deformation bands of the primary amine as well as the isothiocyanate band around 2000 cm-1.Further, Fig. 4 Representative thermal polarized optical micrograph of the nematic schlieren texture displayed by Fc-BATU-12 on cooling from these derivatives are most likely to undergo intramolecular the isotropic liquid to 28 °C hydrogen bond formation between the H atom of the NH group in position 3 and the O atom of the carbonyl group as Table 1 Phase-transition temperaturesa and enthalpy changes of ferro- shown or the molecules are linked into dimers by NMH,S cene derivatives intermolecular hydrogen bonds in addition to intramolecular hydrogen bonding.The NH between 4¾-aryl and thiocarbonyl ferrocene transition T /°C DH/kJ mol-1 groups is at d ca. 12.70 (lower field) and the NH between the carbonyl and thiocarbonyl groups is at d ca. 9.05 (higher field) Fc-BATU-8 C–N 160 47.0 which may be ascribed to the deshielding eVect of the intramol- N–I 176 2.0 I–N 170 2.1 ecular hydrogen bond. Such hydrogen bond formation was Tg 21 — observed in the case of 1-aroyl-3-arylthioureas as well as 1-(pchlorobenzoyl)- 3-phenylthiourea by Dago et al.9 Fc-BATU-12 C–N 144 52.0 N–I 157 1.6 I–N 154 2.2 Tg 18 — Fc-BATU-16 C1–C2 114 49.0 C2–N 126 13.8 N–I 146 1.8 I–N 144 2.2 C O O N H O C C N S H O CnH2 n+1 Fe Tg 26 — All the compounds Fc-BATU-n where n=8, 12, 16 and 18 Fc-BATU-18 C–C1 94 20.8 exhibit purely nematic phases on heating which are charac- C1–C2 102 18.5 C2–N 116 16.1 terized by the formation of a nematic schlieren texture and the N–I 139 2.1 appearence of droplets immediately below the clearing point.I–N 134 1.2 The nematic range increases with an increase in the terminal Tg 31 — alkyl chain length (see Fig. 3). On cooling, the nematic phase reappears and persists to below 0 °C in the first three derivatives aC: crystal, N: nematic; I: isotropic, Tg: glass transition. (see Fig. 4) whereas in the case of Fc-BATU-18 there is a phase transition from nematic to (possibly) SC at about 72 °C which Fig. 5 shows the DSC curve of Fc-BATU-8. On the first remains unchanged below -20 °C. heating, the compound exhibited an endothermic peak at 159.9 °C which corresponds to the melting point followed by Mesomorphic properties the formation of a nematic schlieren texture and the appearence of droplets below the clearing point reached at 177.7 °C.On The thermal and liquid crystal properties of Fc-BATU-n where n=8, 12, 16 and 18 were investigated by a combination of first cooling from the isotropic liquid, however, the enantiotropic nematic phase persisted below -20 °C with a baseline diVerential scanning calorimetry (DSC) and polarized optical microscopy (POM).The data are collected in Table 1. shift at 21 °C. The cooling curve A (5 mW) is a magnified 1348 J. Mater. Chem., 1998, 8(6), 1345–1350Fig. 7 DSC curves of Fc-BATU-16. Scanning rate: 10 °C min-1. nematic schlieren texture. The first cooling of this compound shows a small exothermic peak at 142 °C corresponding to an isotropic–nematic transition.The nematic texture remains Fig. 5 DSC curves of Fc-BATU-8. Scanning rate: 10 °C min-1. unchanged below -20 °C and a baseline shift is observed at 31 °C. Again, the shape of the curve clearly resembles that of version of curve B (20 mW) and resembles a typical glass the typical glass transition point Tg of a DSC curve. transition Tg curve.On second heating from the glassy state, In the second heating, Fc-BATU-16 underwent a glass two exothermic peaks at 94.9° and 119 °C and a sharp endotransition at 37 °C followed by a complex melting and crysthermic peak at 157 °C were observed; the former peaks are tallization process. A broad exothermic peak was seen at due to cold crystallization and the latter corresponds to a 111 °C corresponding to a crystal–nematic transition again crystal–nematic transition followed by an isotropic melt at followed by an isotropic melt at 146 °C. 170.6 °C. The second cooling cycle gave the same results as the first On first heating compound Fc-BATU-12 (see Fig. 6), two cooling. The thermal behaviour is reversible which suggests exothermic peaks were observed at 113 and 144 °C which that the liquid crystal structures are frozen unchanged even correspond to a crystal–crystal transition (C1–C2) and a crysbelow the glass transtion temperatures.tal–nematic transition. The isotropic melt was found at 157 °C. The first heating of compound Fc-BATU-18 (see Fig. 8) On cooling, the nematic phase reappears and persists below shows two sharp endothermic peaks at 94 and 102 °C due to 0 °C with a baseline shift at 18 °C which resembles the typical a crystal–crystal (C–C1) transition followed by an exothermic shape of a glass transition curve and correponds to a glass peak at 107 °C corresponding to the the crystallization of the transition point Tg.On a second heating from the glassy state, metastable crystal C1, which melts immediately at 116 °C.At a metastable crystal (C1) at onset peak temperature 73.2 °C is this temperature, formation of the nematic phase was observed found which melted at 82.3 °C followed by crystallization. A followed by an isotropic melt at 139 °C. On first cooling, the sharp endothermic peak was seen at 131 °C which corresponds nematic phase reappears and a change of texture was observed to a crystal–nematic transition (C2–N) followed by an isotropic melt at 155 °C.This phenomenon is typical of double melting behaviour as was observed by Nakamura et al.3a for their ferrocene compounds as well as by Ohta et al.10 The second cooling proceeds in a similar manner to the first. The first heating of compound Fc-BATU-16 (see Fig. 7), shows a sharp endothermic peak at 113.6 °C, which is the melting point, followed by an exothermic peak at 120 °C corresponding to crystallization of the metastable crystal C1 which melts immediately at 128.4 °C. At this temperature, formation of the mesophase was observed, which was identified by the appearence of droplets and the formation of a typical Fig. 8 DSC curves of Fc-BATU-18. Scanning rate: 10 °C min-1.Fig. 6 DSC curves of Fc-BATU-12. Scanning rate: 10 °C min-1. J. Mater. Chem., 1998, 8(6), 1345–1350 1349D. W. Bruce, InorganicMaterials, ed. D. W. Bruce and D. O. Hare, at 72 °C under the polarizing optical microscope to a texture Wiley, Chichester, 2nd edn., 1996. resembling the schlieren texture of the SC phase. This is also 3 (a) N. Nakamura, T. Hanasaki and H. Onoi, Mol.Cryst. L iq. evidenced from the DSC curves of both the heating and cooling Cryst., 1993, 225, 269; N. Nakamura, H. Onoi, T. Oida and cycles of this compound. Curve B is a magnified version T. Hanasaki, Mol. Cryst. L iq Cryst., 1994, 257, 43; (b) J. Malthete (2 mW) of curve A (10 mW). One can clearly see in this curve and J. Billard, Mol. Cryst. L iq. Cryst., 1976, 34, 177; (c) C.Imrie and C Loubser, J. Chem. Soc., Chem. Commun., 1994, 2159; a peak at 72 °C indicating a change of phase. This SC phase C. Loubser and C. Imrie, J. Chem. Soc., Perkin T rans. 2, 1997, 399; continues to persist below -20 °C with a baseline shift at (d) N. J. Thompson, J. W. Goodby and K. J. Toyne, L iq. Cryst., 31 °C which corresponds to a glass transition Tg. 1993, 13, 381; (e) J.Bhatt, B. M. Fung, K. M. Nicholas and C.- The absence of first order peaks during the second heating D. Poon, J. Chem. Soc., Chem. Commun., 1988, 1439; ( f) K. P. from the glassy state is presumably due to delayed crystalliz- Reddy and T. L. Brown, L iq. Cryst., 1992, 12, 369; (g) ation of the alkyl chains. R. Deschenaux, I. Kosztics, J. L. Marendaz and H. Stoeckli-Evans, Chimia, 1993, 47, 206; (h) H.Tanaka and T. Hongo, Makromol. Rapid Commun., 1996, 17, 91; (i) R. Deschenaux, M. Schweissguth Conclusion and A.-M. Levelut, Chem. Commun., 1996, 1275. 4 R. Deschenaux and J. Santiago, D. Guillon and B. Heinrich, The ferrocene derivatives bearing BATU ligands reported here J. Mater. Chem., 1994, 4, 679; R. Deschenaux and J. L. Marendaz, represent the first steps to achieving the desired heterometallic J.Chem. Soc., Chem. Commun., 1991, 909; R. Deschenaux and complexes. Work is in progress on their preparation, as well J. Santiago, T etrahedron L ett., 1994, 35, 2169. as studying their mesomorphic properties. 5 (a) Yu. G. Galyametdinov and O. V. Ovchinnikov, Izv. Akad. Nauk., Ser. Khim. (Russia), 1990, 10, 2462; (b) Yu. G.The persistence of the nematic phase below the glass trans- Galyametdinov, O. N. Kadkin and I. V. Ovchinnikov, Izv. Akad. ition (Tg) temperatures in Fc-BATU-n compounds (remarkable Nauk., Ser. Khim. (Russia) 1992, 2, 402. for low molecular mass calamitic sytems) suggests possible 6 L. Beyer and E. Hoyer, Z. Chem., 1981, 21, 81; K. C. Satpathy, applications in display devices. The synthetic route reported H.P. Misra, A. K. Panda, A. K. Sathpathy and A. Tripathy, here provides several opportunities for structural variations J. Indian Chem, Soc., 1979, 56, 761; S. N. Jigalur and A. S. R. and we are exploring these and investigating their mesogenic Murthy, Curr. Sci., 1979, 48, 942, Chem. Abstr., 92: 51185w; L. Beyer, R. Scheibe, S. Behrendt and P. Scheibler, Z. Chem., 1978, properties in detail. Work is also in progress on the synthesis 18, 74. of selenium analogues such as N-benzoyl-N¾-arylselenourea 7 C. Imrie, Appl. Organomet. Chem., 1995, 9, 75 and references derivatives containing ferrocene. therein. 8 W. F. Little, C. N. Reilley, J. D. Johnson, K. N. Lynn and We thank Professor R. Deschenaux for helpful discussions and A. P. Sanders, J. Am. Chem. Soc., 1964, 86, 1377. Dr B. Donnio for his help in DSC as well as OPM studies in 9 A. Dago, M. A. Simonov, E. A. Pobedimskaya, A. Macias and the initial stages of this work. A. Martin, Kristallografia, 1987, 32, 1024; A. Dago, M. A. Simonov, E. A. Pobedimskaya, A. Macias and A. Martin, Kristallografia, 1988, 33, 1021. References 10 K. Ohta, M. Yokayama, S. Kusaba and H. Mikawayashi, Mol. Cryst. L iq. Cryst., 1981, 69, 131–142. 1 R. Deschenaux and J. W. Goodby, in Ferrocenes, ed. A. Togni and T. Hayashi, VCH,Weinheim, 1995, ch. 9. 2 J. L. Serrano, Metallomesogens, VCH, Weinheim, 1996; Paper 7/08738Ad; Received 4th December, 1997 1350 J. Mater. Chem., 1998, 8(6), 1345–1350

ISSN:0959-9428    

DOI:10.1039/a708738a

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 Metallomesogens: synthesis and properties Emerson Meyer, Ce�sar Zucco and Hugo Gallardo* Department of Chemistry, Universidade Federal de Santa Catarina, Cep 8800.040–900 Floriano�polis, SC Brazil The synthesis, characterization and mesogenic behaviour of the copper(II) and oxovanadium(IV) complexes derived from phenyltetrazole and benzothiazole and their corresponding ligands are reported.The ligands did not exhibit mesomorphism, whereas the complexes form monotropic smectic A and smectic C mesophases. The mesophases were identified according to their textures by optical microscopy. Introduction Liquid crystals have been known for more than 100 years, but the number of structural types and classes of chemical compounds capable of exhibiting mesomorphic properties has increased significantly only during the last 15–20 years, for example, metal-containing liquid crystals, which are termed ‘metallomesogens’, that combine the variety and range of metal-based coordination chemistry with the extraordinary physical characteristics of liquid crystals.1,2 Coordination compounds, usually complexes of mesogenic ligands, salts of organic acids, and certain organometallic and organoelement compounds with representative metals of s-, OR N O H O H N N N N R OR OR O H N S OR (I) (II) (III) p-, d- and even f-block elements, have been made, and fall into Fig. 1 Schematic of the similarity between the coordination sites of two broad classes: the calamitic, where the metal atoms are salicylaldehydes and five-membered heterocyclic rings containing bound to long thin ligands, giving complexes which are long, nitrogen thin, and rod-like, and the discotic, where the metal is generally coordinated in the center of a flat disc-like organic ligand system.The ligands play a key role in determining the mesomorphic behavior, since they usually compose the periphery of the molecule, and hence play a role in controlling the shape.SchiV bases derived from substituted salicyladehyde (I) are very versatile ligands which form (N–O) chelates with many metals.3 Due to the diversity of substituents that can be introduced, a great variety of these mesogenic complexes has been reported.4,5 In this work we study the potentiality of tetrazole (II) and benzothiazole (III) derivatives in the generation of mesomorphic behavior.The design of these materials is based on the similarity of these ligands with SchiV ’s base imines derived from salicylaldehyde (Fig. 1). In order to verify whether the structural correlation approach is suitable in the case of the heterocyclic ligands tetrazole and benzothiazole, we have studied their complexation behaviour and investigated the thermal behaviour of both ligands and complexes, with the general structure shown in Fig. 2. Synthesis The synthetic route used to prepare the tetrazole ligands and the copper(II) complexes is shown in Scheme 1. The heterocyclic ring was synthesized in several steps, starting from commercial 2,4-dihydroxybenzaldehyde 1, aldoxime formation Fig. 2 Representation of the metal complexes of tetrazole and benzo- 2 and dehydration with acetic anhydride forming the 2,4- thioazole derivatives diacetoxybenzaldehyde 3.The 2,4-dihydroxyphenyltetrazole 4 was obtained by treating 3 with sodium azide. Alkylation The 5-(2-hydroxy-4-alkoxyphenyl)-2-alkyltetrazole 5 was reaction with the appropriate alkyl halide furnished the 5-(2- prepared by O-alkylation and N-alkylations in one-pot synhydroxy- 4-alkoxyphenyl)-2-alkyltetrazoles 5.Treatment of 5 thesis from 5-(2,4-dihydroxyphenyl)tetrazole 4. The tetrazolate with copper(II) acetate in ethanol gives the corresponding anion is an ambidentate system in which alkylation reactions can occur at the N-1 or N-2 positions, the relative proportions copper complexes. J. Mater. Chem., 1998, 8(6), 1351–1354 1351CHO OH OC10H21 NH2 SH + SH N OC10H21 HO a S N HO OC10H21 10 7 8 9 b N S O H21C10O N S O M OC10H21 c M = CuII M = VO 11 12 O N N N N Cu N O N N N OR R RO R HO CHO OH 1 HO OH 2 NOH AcO CN OAc 3 a b HO OH 4 N N N N H RO OH 5 N N N N R c d 6 e Scheme 2 a, Pyridine; b, FeCl3 6H2O, ethanol; c, Cu(OAc)2 H2O or Scheme 1 a, NH2OH HCl, H2O; b, Ac2O; c, NaN3, NH4Cl, DMF; d, VOSO4 5H2O, ethanol RBr, K2CO3, cyclohexanone; e, Cu(OAc)2, EtOH gen bonding between the 2-hydroxyl group and the N-4 position of the tetrazole ring.This hydrogen bond was clearly observed both in the solid state by an X-ray crystallographic N N N N S N N N N S – – 1 2 3 4 5 study of 5-(2,4-dihydroxyphenyl)tetrazole8,9 4 and in solution Fig. 3 Schematic diagram of canonical tetrazolate anions by 1H and IR spectroscopy of 4 and its alkylated products 5.Complexation was carried out with copper(II ) diacetate yielding the mesogenic structures 6. Rotamer complexes are considered forbidden, a priori, because of steric hindrance of the aliphatic groups which would impede the ligand bonding (Fig. 4). The tetrazole ligands and the intermediate compounds were investigated by a variety of techniques, including IR, 1H NMR, 13C NMR, X-ray crystallography and elemental analyses.The analyses showed that the structures of all of the materials were consistent with those expected. The general reaction pathway used to the target ligand 2- (4-decyloxy-2-hydroxyphenyl)benzothiazole and its complexes OH O N N N N HO O N N N N ( ) n ( ) n ( ) n ( ) n is shown in Scheme 2.Fig. 4 Schematic diagram of the steric hindrance between rotamers of The ligand was synthesized using well known literature the tetrazole ligands (n=0,1,2,3…) methods in two steps: first, reaction of 2-hydroxy-4-decyloxybenzaldehyde 7 with 2-aminothiophenol 8: secondly, cyclization of which depend upon the conditions of the alkylation and the of the obtained SchiV base 9 with FeCl3 6H2O.The chelates influence of the 5-substituent (Fig. 3). 11 were prepared by treatment of 2-(4-decyloxy-2-hydroxyphe- However, in our case, spectroscopic evidence suggests that nyl )benzothiazole 10 with CuII and VOII salts. The elemental the N-2 anion in the tetrazole ring is the nucleophile in the and spectral analyses of the complexes and the intermediates alkylation step.The alkyl substituents at the 1- and 2-positions were consistent with their proposed structures. can be readily distinguished by the 1H and 13C chemical shifts of the the N-alkyl group. Alkyl groups bonded to N-1 are Results and Discussion more shielded by ca. 0.15–0.35 ppm in the 1H spectra and by ca. 2–6 ppm in the 13C spectra to their corresponding N-2 Tetrazole system isomers.6 The regioselectivity in the alkylation was corroborated by the analysis of the 13C NMR chemical shifts of the The optical and thermal data of the ligands and complexes are gathered in Table 1.carbon atom C5, 164.8 ppm, in accordance with Butler and Garvin,7 that in similar systems there is a perfect distinction between the isomeric forms N-1 and N-2, in view of the Ligands. All the free ligands synthesized were not mesomorphic.This fact is due in part to a loss of linearity in the sensitive diVerence in the position of such peaks in the spectrum. This regioselectivity is due to the large steric hindrance molecule due to the presence of the tetrazole ring, which is unable to make co-linear disubstitution bondings. Another at the N-1 position considering the bulky and large alkylating agents used.contribution to the loss of liquid crystallinity is the strong intermolecular dipolar repulsions brought about by the pres- The O-alkylation is also regioselective at the 4-hydroxyl group, because of the existence of strong intramolecular hydro- ence of the lateral hydroxide groups. 1352 J. Mater. Chem., 1998, 8(6), 1351–1354Table 1 Optical and thermal data (in °C) of the tetrazolic ligands and complexes of CuII ligand mp complex Cryst SmA SmC Iso n=10 40–42 n=10 117.0 (106.7) (67.4) n=12 50–52 n=12 110.0 (103.4) — — n=14 60–62 n=14 116.0 (92.7) — — Complexes.The CNN stretching vibration of the ligands is polarizability of transition metals), suYciently strong to cause three diional order and consequently the absence of a located in the 1630–1634 cm-1 region and is shifted to lower wavenumbers (aproximately 20 cm-1) upon chelation, indicat- mesophase. The addition of a long aliphatic chain to the rigid nucleus ing that the tetrazolic N atom is involved in metal–nitrogen bond formation.furnishes both the anisotropy and irregular packing needed to promote mesomorphism.As can be observed, all the complexes exhibit mesogenic behavior. The mesophases were identified according to their Based on these considerations the absence of mesomorphic behavior in the free ligand 10 and the complexes 11 and 12 textures which were observed by optical microscopy.10–12 All the complexes n=10–14 show monotropic SmA meso- can be attributed to the small number of aliphatic substituents permitting interactions that favor only three dimensional order.phases. Complex n=10, in addition, gives a monotropic SmC phase. On cooling the isotropic liquid, the SmA phase appears, via bato�nnets, forming a focal-conic textures. Experimental The diVerent mesomorphic behaviour when compared to the SchiV bases complexes is unsurprising, given that one The transition temperatures for all compounds were deterwould expect them to be not eVectively isostructural, but in mined by optical microscopy using a Leitz Ortholux polarizing the coordination geometry only small diVerences in metal– microscope in conjunction with a Mettler FP-52 heating stage.ligand bond lengths and angles can be expected. The purity of the compounds was evaluated by thin layer The diVerence in mesomorphic behavior compared to the chromatography and elemental analysis. The IR spectra were SchiV bases is not surprising considering the very diVerent recorded using the KBr disc method with a Perkin-Elmer electronic and steric character of the tetrazole heteroaromatic model 283 spectrometer, and the 1H NMR and 13C NMR system.spectra were recorded at 80 MHz (Bruker WP-80) or 270 MHz The low thermal stability of the mesophases can be inter- (Bruker HX-270). preted as being due to the reduced anisotropy of the complexes because of the deviation from linearity of the aliphatic chains Materials caused by the heterocyclic ring (Fig. 5). 2,4-Dihydroxybenzaldehyde (Aldrich), 2-hydroxybenzaldehyde (Merck), hydroxylamine hydrochloride (Aldrich), sodium azide Benzothiazole system (Aldrich), 1-bromodecane (Aldrich), 1-bromododecane The thermal data of the ligand and complexes are gathered (Aldrich), 1-bromotetradecane (Aldrich), 2-aminothiophenol in Table 2.(Aldrich) and resorcinol (Aldrich) were used as received. The IR spectra show a shift of the ligand CNN stretching Anhydrous sodium sulfate (Aldrich), potassium carbonate band (1632 cm-1) to lower wavenumbers (1608 cm-1, cop- (Merck), copper acetate (Aldrich), iron(III ) chloride hexahyper( II ) complex; 1605 cm-1, oxovanadium(IV) complex), which drate (Merck) and vanadyl sulfate pentahydrate (Fluka) were indicates that the N atom of benzothiazole participates in the used without purification.Acetic anhydride (Merck), N,Nformation of the complex.For the oxovanadium(IV) complex dimethylformamide (Merck), ethanol (Merck), methanol the stretching band characteristic of the VNO group was (Merck), acetic acid (Merck), acetone (Aldrich), cyclohexanone observed at 976 cm-1. (Carlo Erba) and pyridine (Merck) were purified by distillation While in classic organic structures, the basic question is, prior to use.frequently, to increase intermolecular contact to induce mesomorphism, in coordination compounds the problem is to avoid 2,4-Dihydroxybenzaldoxime 2. To a solution of hydroxylamintermolecular contacts (whenever a coordination site is access- ine hydrochloride (0.05 mol) and sodium acetate (0.05 mol) in ible), and strong dipolar interactions (associated with the high water (30 ml ), at 40 °C, 2,4-dihydroxybenzaldehyde (0.05 mol) was added and the mixture was stirred for 10 min.The mixture was cooled and filtered. The solid was collected and recrystallized from water (92% yield). Mp 194–196 °C; IR (KBr): 3359, 1624, 1525 cm-1. 2,4-Diacetoxybenzonitrile 3. 2,4-Dihydroxybenzaldoxime (0.05 mol) was added slowly to acetic anhydride (30 ml ) at room temperature. After an additional 30 min at room temperature (rt) the reaction mixture was warmed slowly to 100 °C over a 2 h period.Removal of the solvent gave a solid which was recrystallized from methanol–water (351) (75% yield). Mp 72–73 °C; IR (KBr): 2229, 1780, 1186 cm-1. Fig. 5 Sketch of geometry of copper complexes 5-(2,4-Dihydroxyphenyl)tetrazole 4.A suspension of 2,4- Table 2 Transition temperatures (°C) diacetoxybenzonitrile (0.05 mol), sodium azide (0.15 mol) and structure C I ammonium chloride (0.15 mol) in dimethylformamide (50 ml ) was heated to 150 °C over a 6 h period. The reaction mixture ligand 97 was allowed to come to rt with stirring over 1 h. The reaction copper complex 216 mixture was poured into ice cold water (100 ml), and the oxovanadium complex 194 resulting material was filtered and washed with cold water.J. Mater. Chem., 1998, 8(6), 1351–1354 1353The crude product was recrystallized from ethanol–water (351) temperature. The pyridine was then removed by evaporation in vacuo, and the resulting residue was added to a hot ethanolic to give 5-(2,4-dihydroxyphenyl)tetrazole as colorless crystals (85% yield).Mp 304 °C; IR (KBr): 3348, 1612, 1490 cm-1; solution of FeCl3 6H2O. The mixture was then allowed to cool and filtered. The solid was collected and recrystallized Anal. Calc. for C7H6N4O2: C, 47.19; H, 3.37; N, 31.46. Found: C, 47.20; H, 3.20; N, 31.92. from ethanol (82% yield). Mp 97 °C; IR (KBr): 3400, 2919, 1632 cm-1. 1H NMR (200 MHz, CDCl3): 12.71 (s, 1H), 7.96–6.52 (m, 7H), 4.02 (t, J=6.4 Hz, 2H), 1.82 (t, J=6.75 Hz, General procedure for preparation of 5-(2-hydroxy-4-alkoxy- 2H), 1.30 (m, 14H), 0.90 (t, 3H) ppm. 13C NMR (50 MHz, phenyl)-2-alkyltetrazoles 5. To a suspension of 5-(2,4- CDCl3): 169.28, 163.03, 159.87, 151.87, 131.87, 129.51, 126.44, dihydroxyphenyl)tetrazole (0.03 mol) and anhydrous potass- 124.89, 121.58, 121.32, 110.15, 108.05, 101.77, 68.23, ium carbonate (0.06 mol), in cyclohexanone (50 ml ), was added 31.82–14.04 ppm.Anal. Calc. for C23H29NO2S: C, 71.96; H, the appropriate alkyl halide (0.065 mol). The reaction mixture 7.56; N, 3.65. Found: C, 71.66; H, 7.55; N, 4.00. was stirred for 72 h at 150 °C. The mixture was then allowed to cool and filtered. The filtrate was poured into ice cold water Bis[2-(2-hydroxy-4-decyloxyphenyl)benzothiazole] (100 ml) and the residue formed was recrystallized from copper(II) 11. An ethanolic solution (10 ml ) of 2-(2-hydroxy-4- ethanol.decyloxyphenyl)benzothiazole (2 mmol) was added to a hot homologue melting point/°C yield (%) ethanolic solution (10 ml ) of copper(II ) acetate (1 mmol). The solution was refluxed for 20 min and then cooled.The precipi- C10H21 40–42 66 tate was collected by filtration and recrystallized from cyclohex- C12H25 50–52 70 ane–ethanol (153) (68% yield). The brown copper complex C14H29 60–62 70 was characterized by IR spectroscopy and elemental analysis. The 1H NMR spectrum of the complex exhibits large shifts relative to the free ligand, and all signals are broadened as a For the homologue C14H29: IR (KBr): 3314, 2954, 1634, result of the paramagnetic metallic ion.Mp 216 °C; IR (KBr): 1472 cm-1. 1H NMR (200 MHz, CDCl3): 9.81 (s, 1H), 2922, 1608, 1468 cm-1. Anal. Calc. for C46H56N2O4S2Cu: C, 7.90–6.57 (m, 3H), 4.64 (t, J=7.1 Hz, 2H), 3.98 (t, J=6.5 Hz, 66.71; H, 6.76; N, 3.38. Found: C, 66.76; H, 6.67; N, 3.48. 2H), 2.05 (q, 2H), 1.79 (q, 2H), 1.44–1.19 (m, 44H), 0.87 (t, 6H) ppm. 13C NMR (50 MHz, CDCl3): 164.18, 162.50, 158.01, Bis[2-(2-hydroxy-4-decyloxyphenyl)benzothiazole]oxo- 128.31, 108.16, 104.18, 102.11, 68.21, 58.30, 31.95–14.15 ppm. vanadium(IV ) 12. An ethanolic solution (10 ml ) of 2-(2- Anal. Calc. for C35H62N4O2: C, 73.56; H, 10.86; N, 9.80. Found: hydroxy-4-decyloxyphenyl)benzothiazole (2 mmol) was added C, 73.58; H, 105; N, 9.52.to a hot ethanolic solution (10 ml ) of vanadyl sulfate pentahydrate (1 mmol). The solution was refluxed for 20 min and then General procedure for preparation of bis[5-(2-hydroxy-4- cooled. The precipitate was collected by filtration and recrys- alkoxyphenyl)-2-alkyltetrazole]copper(II) 6. An ethanolic solutallized from cyclohexane–ethanol (153) (65% yield).The tion (10 ml ) of the appropriate tetrazole (2 mmol) was added green oxovanadium complex was characterized by IR spec- to a hot ethanolic solution (10 ml ) of copper(II) acetate troscopy and elemental analysis. The 1H NMR spectrum of (1 mmol). The solution was refluxed for 20 min and then the complex exhibits large shifts relative to the free ligand, and cooled. The precipitate was collected by filtration and recrysall signals are broadened as a result of the paramagnetic tallized from chloroform–ethanol (153).The green copper metallic ion. Mp 194 °C; IR (KBr): 2920, 1605, 1469, 976 cm-1. complexes were characterized by IR spectroscopy and elemen- Anal. Calc. for C46H56N2O5S2V: C, 66.44; H, 6.73; N, 3.37. tal analysis. The 1H NMR spectra of the complexes exhibit Found: C, 66.43; H, 6.58; N, 3.44.large shifts relative to the free ligands, and all signals are broadened as a result of the paramagnetic metallic ion. The authors gratefully acknowledge the financial support of CNPq, CAPES and PRONEX. homologue yield (%) C10H21 62 References C12H25 65 1 A. M. Giroud-Godquin and P. M. Maitlis, Angew. Chem., Int. Ed. C14H29 60 Engl. 1991, 30, 375. 2 P. Espinet, M. A. Estruelas, L. A. Oro and J. L. Serrano, Coord. Chem. Rev., 1992, 117, 215. For the homologue C14H29: IR (KBr): 2956, 1610, 1466 cm-1. 3 S. A. Hudson and P. M. Maitlis, Chem. Rev., 1993, 93, 861. Anal. Calc. for C70H122N8O4Cu: C, 69.80; H, 10.14; N, 9.31. 4 M. J. Baena, J. Barbera�, P. Espinet, A. Ezcurra, M. B. Ros and J. L. Found: C, 69.66; H, 10.18; N, 9.55.Serrano, J. Am. Chem. Soc., 1994, 116, 1899. 5 E. Campillos, M. Marcos and J. L. Serrano, J.Mater. Chem., 1993, 2-Hydroxy-4-decyloxybenzaldehyde 7. To a suspension of 3, 1049. 6 R. N. Butler and V. C. Garvin, J. Chem. Soc., Perkin T rans. 1, 2,4-dihydroxybenzaldehyde (0.03 mol) and anhydrous potass- 1981, 390. ium carbonate (0.03 mol) in acetone (50 ml ) was added 1- 7 R. N. Butler, T. M. Mcevoy, F. C. Scott and J. C. Tobin, Can. bromodecane (0.035 mol). The reaction mixture was refluxed J. Chem., 1977, 55, 1564. for 56 h. The mixture was then allowed to cool and filtered. 8 H. Gallardo, E. Meyer and I. Vencato, Acta Crystallogr., Sect. C, The solvent was removed and the residue was distilled at 1995, 51, 2430. reduced pressure giving the product as a clear liquid (63% 9 H. Gallardo, I. M. Begnini and I. Vencato, Acta Crystallogr., Sect. C, 1997, 51, 2430. yield): bp 185 °C (0.5 mm); IR (thin film): 3300, 2920, 10 D. Demus and H. Zaschke, Flu�ssige Kristalle in T abellen II, 1650 cm-1. Springer Verlag, Leipzig, 1984. 11 H. Sackmann and D. Demus, Mol. Cryst. L iq. Cryst., 1966, 2, 81. 2-(2-Hydroxy-4-decyloxyphenyl)benzothiazole 10. 2- 12 G. W. Gray and J. W. Goodby, Smectic L iquid Crystals: T extures Hydroxy-4-decyloxybenzaldehyde (0.02 mol) was dissolved in and Structure, Heyden and Son Inc., 1984, pp. 47–48. 10 ml of dry pyridine containing 2-aminothiophenol (0.02 mol) under an atmosphere of nitrogen and was stirred 5 h at room Paper 8/01159A; Received 9th February, 1998 1354 J. Mater. Chem., 1998, 8(6), 1351–

ISSN:0959-9428    

DOI:10.1039/a801159a

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 Discotic metallomesogens: mesophase crossover of columnar rectangular to hexagonal arrangements in bis(hydrazinato)nickel(II ) complexes Chung K. Lai,* Chun-Hsien Tsai and Yung-Shyen Pang Department of Chemistry, National Central University, Chung-L i, T aiwan, ROC Three series of N-(3,4-dialkoxybenzylidene)-N¾-(3¾,4¾,5¾-trialkoxybenzoyl)hydrazine and their nickel(II ) complexes, bis[N-(3,4- dialkoxybenzylidene)-N¾-(3¾,4¾,5¾-trialkoxybenzoyl)hydrazinato]nickel(II), were prepared and characterized. The mesomorphic properties of these disc-like compounds were also studied in terms of liquid crystallinity.The orange nickel(II) complexes with ten soft alkoxy side chains exhibited columnar disordered mesophases, which were characterized based on DSC analysis and optical polarized microscopy.Nickel complexes with shorter side chains (n=5–8) displayed a monotropic columnar rectangular disordered (Colrd) phase; however, complexes with longer side chains showed monotropic (n=10) and enantiotropic (n=12, 14) columnar hexagonal disordered (Colhd) phases. Similar nickel complexes (n=14) with eight side chains exhibited a columnar hexagonal phase, whereas complexes (n=14) with six side chains formed crystalline phases.The structures of these columnar disordered phases were confirmed by X-ray powder diVraction. XRD data indicated that mesophase crossover from columnar rectangular (Colr) to hexagonal arrangements (Colh) was observed with increasing side chain lengths. Recently, many disc-like coordination complexes with a central geometry of either square planar (Cu, Ni, Pd) or square pyramidal (VO) structure as the rigid core have been prepared, and these metal complexes have been found to exhibit novel mesomorphic properties.1 In general this type of metallomesogenic structure was generated by incorporation of single or multiple metal centers2 into traditional organic molecules.This synthetic strategy has been widely used to design metallomesogenic materials with a variety of geometrical structures and molecular shapes.1e The rich electronic and magnetic properties originating from metal centers makes such materials potential candidates1e,3 for molecular based applications. Columnar hexagonal and rectangular arrangements are among the two common phases observed among discotic2e,4,5 N N O N N O RO OR OR X Y RO X Y OR OR Ni R = (CH2) nH 1 X = Y = OR 2 X = H, Y = OR 3 X = Y = H molecules.Moreover, certain molecules, which are not themselves disc-shaped,6 may also exhibit similar columnar phases. A typical monomesomorphic transition7 of the crystal-tocolumnar- to-isotropic (KAColAI) has often been observed Results and Discussion for this type of columnar mesogen; however, discotic molecules Synthesis with polymesomorphic properties are relatively rare.4 Among them mesophase crossover from columnar rectangular to The synthetic procedures for this type of monodentate hydrahexagonal arrangements identified by high-resolution X-ray zone SchiV base10 are quite straightforward, and summarized diVraction (XRD) with increasing side chain lengths has been in Scheme 1.Reaction of methyl 3,4,5-trialkoxybenzoate esters observed in several columnar systems.4,8 This type of phase with hydrazine hydrate in refluxing absolute ethanol gave transition9 has generally been attributed to the fact that shorter 3,4,5-trialkoxybenzoylhydrazine as a white solid. The SchiV side chain complexes favored the greater core interaction bases of N-(3,4-dialkoxybenzylidene)-N¾-(3¾,4¾,5¾-trialkoxybenneeded for the formation of Colrd phases.A similar preference zoyl)hydrazine were obtained by the condensation reactions of Colr over the Colh phase for larger metal centers, such as of these hydrazine derivatives with 3,4-dialkoxybenzaldehyde M=Pd over the Cu analogue8a also occurs in bis(b-diketonate) in refluxing ethanol.The SchiV bases were isolated by metal complexes. recrystallization from ethanol as light yellow solids. In this paper, we report the preparation, characterization Aroylhydrazone derivatives11 exist in two diVerent keto–enol and mesomorphic properties of a series of nickel(II) com- tautomeric forms (Scheme 2). The conjugating ability12 of the plexes 1–3 of N-(3,4-dialkoxybenzylidene)-N¾-(3¾,4¾,5¾-trialkox- substituents on the ketone moiety was found to influence the ybenzoyl)hydrazine. These disc-like molecules exhibited col- ratio of these two isomers. The increasing conjugating ability umnar mesophases as expected.The mesophase crossover of of aryl substituents preferred the enolic tautomer form. rectangular to hexagonal phase with increasing side chain Two diVerent structural types10 of nickel(II) complexes could length was also observed in these nickel(II ) complexes.In be prepared and isolated depending on the nickel(II) salts used. addition, the eVect of side chain density on the formation Reaction of SchiV bases with nickel(II) acetate produced the and/or stability of columnar mesophases was also studied square planar bis(aroylhydrazinato)nickel(II) complexes; however, the reaction with nickel(II ) chloride gave the octahedral in detail.J. Mater. Chem., 1998, 8(6), 1355–1360 1355Scheme 1 Reagents and conditions: (a) RBr (3.0 equiv.), K2CO3 (7.0 equiv.), Kl (catalyst), refluxing in MeCOMe, 72 h, 73–94%. (b) NH2NH2 (1.5 equiv.), refluxing in absolute EtOH, 12 h, 85–90%.(c) RBr (2.0 equiv.), K2CO3 (3.0 equiv.), Kl, refluxing in MeCOMe, 24 h, 79–88%. (d) Acetic acid (3 drops), refluxing in EtOH, 12 h, 70–75%. (e) Ni(OAc)2 (1.1 equiv.), refluxing in absolute C2H5OH, 24 h 73–80%. that the formation2d,e,6a and stability of the columnar phases by disc-like molecules were strongly dependent on the numbers of flexible side chains, i.e.side chain density. Increasing the number of side chains often improved the liquid crystalline behavior in this type of metallomesogenic system. Complexes with ten side chains These nickel complexes all exhibited the liquid crystalline behavior of columnar discotics and were characterized based on DSC analysis and polarized optical microscopy. The phasetransition temperatures and enthalpies of nickel complexes 1–3 are given in Table 1.Complexes with shorter carbon chain length, n=5–8, exhibited monotropic columnar phase behavior. On heating the crystals melt directly to isotropic liquids above 203.0–218.0 °C, whereas the transition of mesophases at ca. 187.0–219.0 °C took place on cooling from the isotropic phase. The temperature range of the mesophases was quite wide depending on the carbon chain length during the N NH O OR OR RO RO OR N N HO OR OR RO RO OR Keto Enol cooling process.The columnar rectangular phases4,8a were Scheme 2 easily identified by observation of the mosaic textures displaying prominent wedge-shaped defect patterns (Fig. 1) under dichlorobis(aroylhydrazone)nickel(II) complexes. Dehaloprothe polarized microscope.All these complexes were also studied tonation with alcoholic potassium hydroxide converted the at various temperatures by XRD to examine their mesophase octahedral nickel(II ) to planar nickel(II) complexes. The prepstructures. Rectangular4,8 columnar phases often displayed two aration of octahedral complexes was not attempted in this intense peaks in the low angle region, which were indexed as work.The square planar nickel(II ) complexes studied in this (200) and (110) reflections. The XRD data for the nickel work were prepared by reaction of N-(3,4-dialkoxybenzylidcomplexes 1 and 2 are summarized in Table 2. For example, ene)-N¾-(3¾,4¾,5¾-trialkoxybenzoyl)hydrazine with Ni(OAc)2 in complexes n=5 gave a diVraction of two strong reflections at refluxing ethanol.The nickel complexes were then isolated as d 25.94 and 21.70 A ° , a weaker peak at d 8.57 A ° and also a orange crystals by recrystallization from absolute ethanol in a broad diVuse peak at d 4.75 A ° at 130 °C. This diVraction high yield of 76–86%. All the nickel(II) complexes are diamagpattern (Fig. 2) corresponded to rectangular lattice constants: netic, and all display sharp peaks in their 1H NMR spectra. a=51.88, b=23.89 A ° .Three diVerent columnar Colrd phases4 Elemental analyses also confirmed the identity of the have now been identified: P(21/a), P(2/a) and C(2/m) complexes. depending on the point symmetry of the molecules. In rectangular phases the average molecular plane is not perpendicular to Mesomorphic properties the column axis, and the tilt angle can be determined only by known molecular symmetry.Mesomorphic studies in columnar systems of metal bis(bdiketonates) with various numbers of side chains indicated Interestingly, the nickel complexes with longer side chains, 1356 J. Mater. Chem., 1998, 8(6), 1355–1360Table 1 Phase behaviora of nickel(II ) complexes 1–3 I K2 K1 n = 5 25.2 (46.6) 78.2 (34.9) 187.2 (4.45) 189.9 (4.58) 68.8 (12.0) 44.8 (41.6) 78.4 (11.3) 170.2 (2.18) 173.9 (2.31) 39.6 (24.7) 131.9 (0.76) 154.8 (0.80) 158.0 (0.13) 56.9 (3.91) 76.1 (5.71) 92.6 (55.0) 116.3 (52.1) 72.4 (29.8) Colrd 216.6 (5.46) 35.7 (22.8) 211.4 (6.05) 79.9 (6.59) Colrd 224.5 (7.39) 42.4 (36.7) 218.6 (7.35) 74.8 (37.7) Colrd 217.8 (4.87) 42.2 (2.77) 208.8 (4.83) 27.2 (2.77) Colrd 217.8 (4.99) 17.2 (29.8) 214.5 (4.83) 63.5 (28.7) Colhd 203.5 (4.91) 21.1 (32.9) 200.9 (4.75) 73.6 (28.2) Colhd Colhd Colhd K2 1 75.7 (24.2) 6 7 8 10 12 14 14 14 2 3 K1 K1 K1 K1 K1 K1 K1 K2 K2 K2 K2 K2 K2 K2 K1 I I I I I I I I an represents the number of carbon atoms in the alkoxy chain.K1, K2=crystal phases; Colhd=columnar hexagonal disordered phase; Colrd=columnar retangular disordered phase; I=isotropic.The transition temperatures (°C) and enthalpies (in parentheses, kJ mol-1) were determined by DSC at a scan rate of 10.0 °C min-1. n=10, 12 and 14, showed a typical columnar hexagonal phase. Two transitions10 of crystal-to-columnar-to-isotropic (KA ColAI), and additional crystal-to-crystal (K1AK2) transitions were also observed for these complexes.From DSC analysis a typical larger enthalpy for the crystal-to-liquid crystal transition at lower temperatures (ca. 78.0 °C) and a lower enthalpy for the liquid crystal-to-isotropic transition at higher temperatures (174.0–190.0 °C) were observed. This lower value of the transition enthalpy indicated that the mesophases were relatively highly disordered. The temperature range of columnar mesophases for nickel complexes is quite wide (between 125–162 °C) in the cooling process and slightly sensitive to the carbon number of the side chains.The melting and clearing points both decrease as the side chain length increases. Upon heating, orange complexes melt to give birefringent fluid phases with columnar superstructures. When cooled from their isotropic phases, they displayed an optical texture (Fig. 3) which was a mixture of pseudo focal-conics and mosaic regions with linear birefringent defects, suggesting hexagonal columnar4,6,8 structures. The existence of a large area of homeotropic domains may lead to the conclusion of preferred uniaxial character4a in this phase. A summary of the diVraction peaks and lattice constants for these nickel complexes is also given in Table 1.For example, nickel complex 1 (n=14) displays a diVraction pattern of a two-dimensional hexagonal lattice (Fig. 4) with an intense peak and two weaker peaks of 36.65, 21.77 and 18.86 A ° at 155 °C. These are characteristic of the columnar hexagonal columnar (Colhd) phase with a d-spacing ratio of 1, (1/3)1/2 and (1/4)1/2, corresponding to Miller indices (100), (110) and (200), respectively.This reflection pattern corresponds to an intercolumnar distance (a parameter of the hexagonal lattice) of 43.20 A ° . In the hexagonal lattice the column axes are located at the nodes of the two-dimensional structure and are oriented along the c axis. However, a broad diVuse band, which is likely Fig. 1 Optical texture (100×) observed for complex 1 (n=8): Colrd phase at (a) 207 °C (thick sample) and (b) 195 °C (thin sample) due to molten alkyl chains, occurred at 4.73–4.82 A ° in the J.Mater. Chem., 1998, 8(6), 1355–1360 1357Table 2 Variable-temperature X-ray diVraction data for nickel(II ) complexes 1 and 2 lattice d-spacing/A ° Miller complex mesophase spacing/A ° obs. (calcd.) indices 1 n=5 Colrd at 100 °C a=52.28 26.14 (26.14) (200) b=24.10 21.89 (21.89) (110) 8.57 (8.86) (420) 4.80 (br) Colrd at 130 °C a=51.88 25.94 (25.94) (200) b=23.89 21.70 (21.70) (110) 8.57 (8.79) (420) 4.75 (br) 6 Colrd at 200 °C a=54.44 27.22 (27.22) (200) b=24.92 22.66 (22.66) (110) 12.58 (12.46) (020) 4.91 (br) 7 Colrd at 150 °C a=58.08 29.04 (29.04) (200) b=28.62 25.67 (25.67) (110) 16.30 (16.03) (310) 14.65 (14.31) (020) 4.75 (br) 8 Colrd at 200 °C a=56.84 28.42 (28.42) (200) b=28.23 25.28 (25.28) (110) 5.18 (br) 10 Colhd at 100 °C a=37.49 32.40 (32.40) (100) 18.60 (18.70) (110) 16.38 (16.20) (200) 4.73 (br) Colhd at 190 °C a=37.03 32.07 (32.07) (100) 18.39 (18.51) (110) 16.08 (16.03) (200) 4.73 (br) 12 Colhd at 120 °C a=40.64 35.63 (35.63) (100) 21.17 (20.57) (110) 17.58 (17.81) (200) Fig. 2 X-Ray diVraction data for complex 1 (n=5): (a) Colrd phase at 4.79 (br) 130 °C and (b) K phase at room temperature 14 Colhd at 155 °C a=43.20 36.65 (36.65) (100) 21.77 (21.60) (110) 18.86 (18.32) (200) 4.82 (br) 2 14 Colhd at 120 °C a=43.92 38.04 (38.04) (100) 21.85 (21.96) (110) 19.62 (19.62) (200) 4.78 (br) wide-angle region. The absence of distinct peaks at a high angle rules out the existence of periodicity along the columns, and is also consistent with the DSC analysis showing low enthalpies for the columnar-to-isotropic transition.This positively indicates a highly disordered mesophase; i.e. there is no long-range order along the columns. The temperature dependence of the lattice parameters in these complexes was also studied. We find that the low-angle reflections of complexes generally shift to a larger d-spacing with decreasing temperatures (i.e.d=37.03 A ° at 190 °C and d=37.49 A ° at 100 °C for complex 1; n=10). The hexagonal lattices are also correlated well with increasing side chain length. The mesophase change of columnar rectangular to hexagonal arrangements with increasing side chain length was observed in this system; with increasing carbon number the hexagonal phase arrangement predominates over the rectangular phase.The mesophase crossover from Colrd to Colhd has been observed in similar metallomesogenic systems, and was generally attributed to the greater core interaction necessary for the formation of the Colrd phases. The tilted Colrd phase reduced the interactions between the bulky side chains and allowed closer contacts between the cores. Similar mesophase crossover was also observed in systems4,8a,b,d with the larger Pd replacing Cu in the Cu analogue.8a This pseudohexagonal lattice constant, i.e.a rectangular lattice with an axial ratio of b/a from the ideal hexagonal of 31/2 (Scheme 3), for this series Fig. 3 Optical texture (100×) observed for complex 1 (n=12): (a) Colhd phase at 180 °C and (b) K phase at room temperature of nickel complexes was in the range of 2.01–2.18 (from 1358 J.Mater. Chem., 1998, 8(6), 1355–1360Conclusion The mesomorphic studies of the title nickel complexes indicated that the formation of columnar phases was found to be strongly dependent on the number of side chains around the core group. The complexes with a total of ten and eight side chains exhibited columnar phases, however, complexes with six and four side chains exhibited crystalline phases. Therefore, the number of side chains for this type of disc-like nickel complexes studied must be at least eight to form a stable mesogen.Experimental section All chemicals and solvents were reagent grade from Aldrich Chemical Co., and used without further purification. 1H and 13C NMR spectra were measured on a Bruker DRS-200. J Values are in Hz. Infrared spectra were recorded on a Bio- Rad FTS-155 using polystyrene as a standard. DSC thermographs were recorded on a Perkin-Elmer DSC-7 and calibrated with a pure indium sample. All phase behavior was determined at a scan rate of 10.0 °C min-1 unless otherwise noted. Optical polarized microscopy was carried out on a Nikkon MICROPHOT-FXA with a Mettler FP90/FP82HT hot stage system.X-Ray powder diVraction (XRD) studies were conducted on an INEL MPD-diVractometer with a 2 kW Cu-Ka Xray source equipped with an INEL CPS-120 position sensitive detector and a variable temperature capillary furnace with an accuracy of ±0.10 °C in the vicinity of the capillary tube.The detector was calibrated using mica and silicon standards. The powder samples were charged in Lindemann capillary tubes (80 mm long and 0.01 mm thick) from Charles Supper Co. Fig. 4 X-Ray diVraction data for complex 1 (n=14): (a) Colhd phase with a diameter of 0.1 or 0.2 mm, and the end was flame at 155 °C and (b) K phase at room temperature sealed.The samples were heated above the isotropic temperatures and left for ca. 10 min. The samples were then cooled at a rate of 5.0 °C min-1 to the desired temperature and the diVraction data collected. The methyl 3,4,5-trialkoxybenzoate esters, methyl 4-alkoxybenzoate esters, ethyl 3,4-dialkoxybenzoate esters, and 3,4-dialkoxybenzaldehyde were prepared by literature2a–d,6a procedures.Methyl 4-tetradecyloxybenzoate White solid, yield 90%. dH(CDCl3): 0.82 (t, J 6.73, MCH3, 3H), 1.24–1.84 [m, M(CH2)12, 24H], 3.85 (s, MOCH3, 3H), b = 31/2 a a a b 4.27 (t, J 7.31, MOCH2, 2H), 6.84 (d, J 8.82, MC6H4, 2H), Scheme 3 7.92 (d, J 8.82, MC6H4, 2H). dC(CDCl3): 14.76, 23.26, 26.58, 29.79, 29.98, 30.05, 30.30, 30.33, 30.36, 32.53, 52.54, 68.83, 114.62 (C2,6), 122.91 (C1), 132.13 (C3,5), 163.45 (C4), 167.53 n=5–10). This value indicated that the structural departure (CNO).from the ideal hexagonal lattice was about 20% in this system. Complexes 2 and 3 with eight and six side chains Ethyl 3,4-bis(tetradecyloxy)benzoate Nickel complexes 2 and 3 (n=14) were also prepared to study White solid, yield 89%. dH(CDCl3): 0.87 (t, MCH3, 9H), the eVect of side chain density on the formation of mesophases. 1.24–1.83 (m, MCH2, 44H), 3.97 (t, J 8.19, OCH2, 2H), 4.34 DSC analysis and XRD diVraction showed that complex 2 (tt,MOCH2, 2H), 6.80 (d, J 8.44,MC6H3, 1H), 7.51 (s,MC6H3, with eight side chains exhibited a similar columnar hexagonal 1H), 7.57 (d, J 6.42, MC6H3, 1H).dC(CDCl3): 13.91, 14.21, disordered phase (Colhd). The isotropic point for the complex 22.51, 25.84, 28.93, 29.21, 31.74, 60.48, 68.84, 69.13, 111.7 (C6), was ca. 20°C lower, and the temperature of the mesophase 114.2 (C2), 122.6 (C5), 123.2 (C1), 148.3 (C4), 152.9 (C3), was narrower than for the similar complex 1 (n=14). This 166.3 (CNO). result indicated that higher side chain density tended to easily stabilize the mesogenic core, as is often observed in this type of columnar system.However, complex 3 (n=14) with six side Methyl 3,4,5-tris(dodecyloxy)benzoate chains showed only a transition of the crystal to isotropic White solid, yield 90%. dH(CDCl3): 0.86 (t, MCH3, 9H), phase at 116.3 °C. Similar nickel complexes (n=12) of disc-like 1.26–1.84 (m, CH2, 60H), 3.88 (s, MOCH3, 3H), 4.03 (tt, molecules have also been prepared, in which nickel complexes10 MOCH2, 6H), 7.25 (s, C6H2, 2H). dC(CDCl3): 14.18, 22.75, with two or four dodecyloxy chains were found to be non- 26.14, 29.36, 29.45, 29.62, 29.71, 29.75, 29.39, 30.39, 32.00, 52.13, mesogenic. The broadening10 of the molecular width and the 69.22, 75.53, 105.02 (C2,6), 124.69 (C1), 142.01 (C3,5), 152.83 non-coplanarity of the benzylidene ring were attributed to this non-mesomorphic behavior.(C4), 166.89 (CNO). J. Mater. Chem., 1998, 8(6), 1355–1360 13593,4-Bis(dodecyloxy)benzaldehyde References 1 (a) S. A. Hudson and P. M. Maitlis, Chem. Rev., 1993, 93, 861; White crystals, yield 85%. dH(CDCl3): 0.85 (t, J 6.79, CH3, (b) P. Espinet, M. A. Esteruelas, L. A. Oro, J. L. Serrano and 6H), 1.23–1.43 (m, CH2, 36H), 1.74–1.86 (m, CH2, 4H), 4.03 E.Sola, Coord. Chem. Rev., 1992, 117, 215; (c) D. W. Bruce and (tt, MOCH2, 4H), 6.96 (d, J 8.63, C6H3, 1H), 7.36 (s, C6H3, D. O’Hare, Inorganic Materials, Wiley, Chichester, 1992, 1H), 7.48 (d, J 1.81, C6H3, 1H), 9.79 (s, CHO, 1H). dC(CDCl3): pp. 407–490; (d) P. Maitlis and A. M. Giroud-Godquin, Angew. 14.03, 22.63, 25.94, 28.98, 29.06, 29.31, 29.57, 31.88, 69.10, 111.1 Chem., Int.Ed. Engl., 1991, 30, 375; (e) J. L. Serrano, in (C6), 111.8 (C2), 126.4 (C5), 129.9(C1), 149.5 (C4), 154.7 (C3), Metallomesogens; Synthesis, Properties, and Applications, VCH, New York, 1996. 190.8 (CNO). 2 (a) C. K. Lai, M.-Y. Lu and F.-J. Lin, L iq. Cryst., 1997, 23, 313; (b) Y.-F. Leu and C. K. Lai, J. Chin. Chem. Soc., 1997, 44, 89; 4-Tetradecyloxybenzoylhydrazine (c) C.K. Lai and F.-J. Lin, J. Chem. Soc., Dalton T rans., 1997, 17; (d) A. G. Serrette, C. K. Lai, T. M. Swager, Chem. Mater., 1994, 6, White solid, yield 90%. dH(CDCl3): 0.84 (t, J 6.30, MCH3, 2252; (e) J. Barbera�, A. Elduque, R. Gime�nez, L. A. Oro and 3H), 1.23–1.83 (m, MCH2, 24H), 3.28 ( br, MOCNHNH2, 2H), J. L. Serrano, Angew. Chem., Int. Ed. Engl., 1996, 35, 2832; 3.92 (t, J 6.47, MOCH2, 2H), 6.79 (d, J 8.76, MC6H4, 2H), ( f ) S.Eguchia, T. Nozaki, H. Miyasaka, N. Matsumoto, 7.73 (d, MC6H4, 2H), 8.53 (br, MCONH, 1H). dC(CDCl3): H. Okawa, S. Kohata and N. Hoshino-Miyajima, J. Chem. Soc., Dalton T rans., 1996, 1761; (g) L. Bonnet, F. D. Cukiernik, 13.94, 16.26, 22.51, 25.37, 25.81, 28.94, 29.18, 29.46, 30.74, 31.74, P. Maldivi, A.-M. Giroud-Godquin, J.-C.Marchon, M. Ibn-Elhaj, 68.05, 114.14 (C2,6), 125.24 (C1), 129.04 (C3,5), 161.89 (C4), D. Guillon and A. Skoulios, Chem. Mater., 1994, 6, 31; 168.74 (CNO). (h) R. H. Cayton, M. H. Chisholm and F. D. Darrington, Angew. Chem., Int. Ed. Engl., 1990, 29, 1481; (i) M. Ghedini, D. Pucci, 3,4-Bis(tetradecyloxy)benzoylhydrazine G. D. Munno, D. Viterbo, F. Neveal and S. Armentano, Chem.Mater., 1991, 3, 65; ( j) W. Pyzuk, A. Krowczynski, L. Chen, White solid, 78%. dH(CDCl3): 0.85 (t, J 6.69, MCH3, 6H), E. Gorecka and I. Bickczantaev, L iq. Cryst., 1995, 19, 675; 1.23–1.84 (m, MCH2, 48H), 3.72 ( br, MOCNHNH2, 2H), 4.05 (k) M. Ghedini, M. Longeri and R. Bartolino, Mol. Cryst. L iq. Cryst., 1982, 84, 207; (l )M. Ghedini, S. Licoccia, S. Armentano and (tt,MOCH2, 4H), 6.83 (d, J 8.47,MC6H3, 1H), 7.51 (s,MC6H3, R.Bartolino, Mol. Cryst. L iq. Cryst., 1984, 108, 269. 1H), 7.58 ( br, MCONH, 1H), 7.62 (d, J 8.39, MC6H3, 1H). 3 (a) N. Boden, R. J. Bushby and J. Clements, J. Chem. Phys., 1993, dC(CDCl3): 13.93, 14.22, 22.53, 25.84, 28.92, 29.04, 29.22, 31.77, 98, 5920; (b) D. Adam, P. Schuhmacher, J. Simmerer, L. Haussling, 68.81, 69.10, 111.71 (C3), 114.13 (C4), 122.58 (C1), 123.28 (C5), K.Siemensmeyer, K. H. Etzbach, H. Ringsdorf and D. Haarer, 148.31 (C2), 152.95 (C6), 168.12 (CNO). Nature, 1994, 371, 141. 4 (a) S. Chandrasekhar, Adv. L iq. Cryst., 1982, 5, 47; (b) C. Destrade, P. Foucher, H. Gasparoux, H. T. Nguyen, A. M. Levelut and 3,4,5-Tris(dodecyloxy)benzoylhydrazine J. Malthete, Mol. Cryst. L iq. Cryst., 1984, 106, 121. 5 (a) C. Destrade, P. Foucher, H. Gasparoux, H. T. Nguyen, White solid, yield 80%. dH(CDCl3): 0.82 (t, J 7.46, MCH3, A. M. Levelut and J. Malthete, Mol. Cryst. L iq. Cryst., 1984, 106, 3H), 1.23–1.81 (m, MCH2, 60H), 3.33 ( br, MCONHNH2, 2H), 121; (b) S. Chandrasekhar and G. S. Ranganath, Rep. Prog. Phys., 3.96 (tt, J 6.16, MOCH2, 6H), 6.90 (s, MC6H2, 2H), 7.34 ( br, 1990, 53, 57.MCONH, 1H). dC(CDCl3): 14.09, 22.67, 26.05, 29.29, 29.34, 6 (a) H.-X. Zheng C. K. Lai and T. M. Swager, Chem. Mater., 1994, 29.37, 29.55, 30.28, 31.90, 69.27, 73.50, 105.41, 127.38 (C2,6), 6, 101; (b) J. Barbera�, C. Cativiela, J. L. Serrano and 141.33 (C1), 153.16 (C3,5), 168.62 (CNO). M. M. Zurbano, Adv. Mater. 1991, 3, 602; (c) L. Bonnet, F. D. Cukiernik, P. Maldivi, A.-M.Giroud-Godquin, J.- C. Marchon, M. Ibn-Elhaj, D. Guillon and A. Skoulios, Chem. N-(3,4-Dialkoxybenzylidene)-N¾-(3¾,4¾,5¾-trialkoxybenzoyl ) Mater., 1994, 6, 31; (d) W. Pyzuk, A. Krowczynski, L. Chen, hydrazine E. Gorecka and I. Bickczantaev, L iq. Cryst. 1995, 19, 67. M. Collard and C. P. Lillya, J. Am. Chem. Soc., 1991, 113, 8577. n/cm-1 (thin film): 3421, 2923, 2847, 1640, 1626, 1599, 1576, 8 (a) H.-X.Zheng, C. K. Lai and T. M. Swager, Chem. Mater., 1995, 1514, 1470, 1437, 1387, 1337, 1264, 1233, 1177, 1144, 1069, 1069. 7, 2067; (b) A. Takada, N. Ide, T. Fukuda, T. Miyamoto, K. Yamagata and J. Watanabe, L iq. Cryst., 1995, 19, 441; (c) C. Destrade, M. C. Monodon and J. Malthete, J. Phys., 1979, General procedures for preparation of nickel complexes. 40-C3, 17; (d) C. Destrade, M. C. Monodon-Bernaud and Bis[N-(3,4-diheptyloxybenzylidene)-N¾-(3¾,4¾,5¾- H. T. Nguyen, Mol. Cryst. L iq. Cryst. L ett., 1979, 49, 16; triheptyloxybenzoyl )hydrazinato]nickel(II ) (e) J. Billard, J. C. Dubois, C. Vaucher and A. M. Levelut, Mol. Cryst. L iq. Cryst., 1981, 66, 115. A suspension of N-(3,4-diheptyloxybenzylidene)-N¾-(3¾,4¾,5¾- 9 (a) C. R. Safinya, K. S. Liang, W. A. Varady, N. A. Clark and triheptyloxybenzoyl)hydrazine (0.31 g, 0.342 mmol) and nickel G. Anderson, Phys. Rev. L ett., 1984, 54, 1172; (b) C. R. Safinya, acetate (0.090 g, 0.362 mmol) in 100 ml of absolute ethanol N. A. Clark, K. S. Liang, W. A. Varady and L. Y. Chiang, Mol. was heated under reflux overnight. The orange solid was Cryst. L iq. Cryst., 1985, 123, 205. filtered oV, and then washed with hot ethanol (3×20 ml). The 10 (a) M. N. Abser, M. Bellwood, C. Buckley, M. C. Holmes and R. W. McCabe, J. Mater. Chem., 1994, 4, 1173; (b) M. N. Abser, product was isolated as orange crystals after twice recrystalliz- M. Bellwood, M. C. Holmes and R. W. McCabe, J. Chem. Soc., ation from THF–ethanol. Yield 73%. n/cm-1 (thin film): 2929, Chem. Commun., 1993, 4, 1062. 2855, 1597, 1522, 1502, 1466, 1473, 1397, 1362, 1273, 1221, 11 (a) L. Sacconi, J. Am. Chem. Soc., 1952, 74, 4503; (b) H. Ohta, Bull. 1180, 1137, 1069, 1038. (Calc. for C138H242N4O12Ni: C 71.46, Chem. Soc. Jpn., 1958, 31, 1056; 1960, 33, 202. H 9.91, N 3.40. Found C 71.17, H 10.06, N 3.44%). 12 (a) L. Sacconi, J. Am. Chem. Soc., 1954, 76, 3400; (b) R. A. Morton, A. Hassan and T. C. Calloway, J. Chem. Soc., 1934, 883. We thank the National Science Council of Taiwan, ROC for generous funding (NSC-87-2113-M008-007) of this work. Paper 8/00657A; Received 23rd January, 1998 1360 J. Mater. Chem., 1998, 8(6), 1355–1360

ISSN:0959-9428    

DOI:10.1039/a800657a

出版商:RSC    

年代:1998

数据来源: RSC