摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Highly polarized electron donors, acceptors and donor–acceptor compounds for organic conductors Yoshiro Yamashita*† and Masaaki Tomura Institute forMolecular Science,Myodaiji, Okazaki 444-8585, Japan New types of p-electron donors and acceptors with high polarizability are reviewed. TTF derivatives with polarizable substituents, p-extended TTF analogues fused with nitrogen-containing heterocycles and non-TTF donors (i.e.electron donors) and chalcogen atom-containing quinones, TCNQ and DCNQI analogues (i.e. electron acceptors) are described. Donor–acceptor compounds showing electrochemically amphoteric properties are also discussed. Introduction Recently much attention has been focused on molecular-based organic conductors and superconductors.1 Development of new electron donors and acceptors aVording them is particularly important in order to make progress in this field.For this purpose many kinds of electron donors and acceptors have been designed and synthesized.2–4 Among them highly polarized molecules containing heteroatoms have attracted considerable attention. Polarized molecules can be produced by introduction of polarizable substituents into donor or acceptor skeletons, or linkage of electron-donating units and accepting ones.In the latter compounds donor or acceptor properties appear depending on the strength of their electrondonating or -accepting abilities. The highly polarized molecules have several advantages as components for aVording organic conductors.First, intermolecular interactions can be enhanced S S S S X X X X S S S S Y X S S S S S S RTe RTe S S S S Te S S S S R R Te S S S S Te Te S S 1 2 3 4 5 TeR TeR S S R R by electrostatic interactions, which increase dimensionality in the complexes and suppress metal–insulator transitions. Such derivatives 1, halogen-halogen contacts are expected to lead interactions may lead to the formation of unique molecular to strong intermolecular interactions in the crystals. However, networks which have special functions such as inclusion attachment of halogen atoms reduces electron-donating ability properties.Second, in the donor–acceptor systems intermolecuand the tetra-substituted derivatives 1 (X=Cl, Br) do not lar charge-transfer (CT) interactions can be expected, which aVord conducting materials.5 On the other hand, the tetraiodo may be useful for crystal engineering.Third, p-extended systems derivative 1 (X=I) aVords cation radical salts. Although the have reduced on-site Coulomb repulsion which is one of the I3- salt is an insulator due to its 151 stoichiometry, strong important requirements for the molecular design of organic I,I interactions are observed, which connect the stacks of conductors.Fourth, unstable strong electron donors can be cation radicals in a three-dimensional structure.6 stabilized by introduction of electron-withdrawing groups to Unsymmetrical TTF derivatives 2 have been prepared to give polarized electron donors. Similarly unstable strong elecincrease donating abilities as well as to use intermolecular tron acceptors can be stabilized by electron-donating groups.interactions involving sulfur atoms. The donor 2 (X=I, Y= Finally, highly polarized molecules have low excitation ener- H) aVords a complex with Pd(dmit)2 which shows metallic gies, and interesting optical properties such as non-linear properties down to 4.2 K.7 X-Ray analysis reveals strong and optical ones are expected.When the electron-donating and directional I,S donor–acceptor interactions. The donor 2 -accepting abilities are balanced in the donor–acceptor com- (X=Y=I) aVords metallic cation radical salts 24M(CN)4(M= pounds, they show amphoteric properties. Such compounds Ni, Pd, Pt).8 In the crystals there are two kinds of short I,NC are of special interest as candidates for single-component contacts resulting in unusual three-dimensional structures.In conductors showing intrinsic conductivities. We will highlight the CT complex of 2 (X=Y=Cl) with TCNQ, Cl,Cl short here highly polarized donors, acceptors and donor–acceptor contacts are observed, which control the packing in the crystal.9 compounds. In spite of the mixed stacking structure, its conductivity (s=1 S cm-1) is high.Polarized electron donors Compounds containing highly polarizable tellurium atoms tend to form molecular networks through the heteroatom TTF derivatives contacts. This property can be used to increase the intermolecu- The derivatives of TTF such as BEDT-TTF have played a lar interactions between TTF molecules. In the tetrakis(alkylmajor role in the development of organic conductors and telluro)-TTFs 3, a two dimensional network based on Te,Te superconductors.2 The synthetic methods of TTF derivatives contacts is observed, which leads to an unusually high conduchave been improved and several derivatives containing polariz- tivity (ca. 10-5 S cm-1) as a single component.10 The crystal able substituents have been prepared.In the halogenated TTF of tetrakis(phenyltelluro)-TTF 3 (R=Ph) has Te,S interstack and Te,Te intrastack interactions, resulting in a single-component conductivity of 10-6 S cm-1.11 TTF-di-substituted ditellu- †E-mail: yoshiro@ims.ac.jp J. Mater. Chem., 1998, 8(9), 1933–1944 1933ride 4 prepared by Becker et al. also shows a high single- On the other hand, the diselenium analogue 8 forms 251 salts showing metallic behaviour down to 20 K.16 In the crystal component conductivity (5×10-5 S cm-1).12 The crystal has a stacking structure where several short Te,S contacts are a two dimensional network of S,N interactions in addition to S,S(Se) ones is formed.Since there are no distinctive observed. TTF derivative 5 has been obtained by a one-pot reaction of 2,3-dimethyl-TTF with bis(phenylacetylenyl) tellu- donor columns, the thiadiazole network is considered to aVord a conduction path.ride.13 The TTF units are significantly bent and both Te,S and Te,Te contacts are observed. Papavassiliou et al. have prepared the pyrazine derivative 9. This molecule aVords a cation radical salt 93I3 which is Introduction of heterocycles containing CNN bonds on the TTF units is of interest since such heterocycles are electron- considered to be a three-dimensional conductor due to the S,N intermolecular interactions.17 Pyrazine-fused TTF withdrawing and so polarized structures leading to strong intermolecular interactions are expected.In addition, the derivatives 10 and 11 have been prepared by nucleophilic reaction of dianion 12 with tetrachloropyrazine.18 The bis- extended p-conjugation decreases on-site Coulomb repulsion.In this context, 1,2,5-thiadiazole- and pyrazine-fused TTF TTF derivative 10 (R=hexyl) exhibits four reversible sequential one-electron oxidation steps with potentials of 0.49, 0.71, derivatives have been synthesized. Bis[1,2,5]thiadiazolo-TTF 1.24 and 1.50 V vs.SCE. TTF analogues with quinoid structures Such TTF analogues are of interest due to their high electrondonating abilities and reduced on-site Coulomb repulsion.4 S S S S S S S S N X N R R R R N X N O O 14 X = S 15 X = Se 16 X = CH=CH 13 17 Although a TTF analogue 13 has been prepared with this aim, it is unstable in solution due to its extremely low oxidation N S N S S S S N S N N S N S S S S N S N S S S S S S S S S S R R N N S S S S R R R R N N S S S S S S R R N N Cl Cl S S S– S– R R 6 7 8 9 10 11 12 Se Se S S potential.19 Introduction of electron-withdrawing heterocycles is expected to enhance stability as well as intermolecular 6 has been prepared independently by Underhill et al.and interaction and p-delocalization. us.14 In the crystal structure of the neutral molecule a unique TTF analogues fused with 1,2,5-thiadiazole 14, selenadiazole molecular network is formed by short S,N and S,S contacts. 15 and pyrazines 16 have been prepared by a Wittig–Horner However, this molecule has not aVorded CT complexes with reaction of the corresponding carbanions derived from phoselectron acceptors due to its poor electron-donating ability. In phonate esters with diones 17 followed by a retro-Diels–Alder order to enhance this ability, unsymmetrical molecules 7 have reaction.20 They are air-stable violet solids.Their absorption been synthesized.15 The donor 7 (R=SC2CH2S) aVords radical maxima are observed at 482–522 nm due to intramolecular cation salts with 151 stoichiometries. The conductivities are charge transfer from the 1,3-dithiole units to the nitrogenfairly good in spite of their being 151 salts.X-Ray analysis containing heterocycle. The oxidation potentials are lower reveals short S,N and S,S contacts (Fig. 1). than those of TTF, as shown in Table 1, indicating that they are stronger donors than TTF although they contain electron- Table 1 Oxidation potentials of donorsa E/V vs. SCE compound E1 E2 DE ref.TTFa 0.46 0.87 0.41 20(a) 13b -0.11 -0.04 0.07 19 14 (R=H)a 0.36 0.53 0.17 20(a) 15 (R=H)a 0.37 0.55 0.18 20(b) 16 (R=H)a 0.23 0.43 0.20 20(c) 18 (R=H)a 0.63 0.01 22 19a 0.76 0.85 0.09 23 20a 0.68 0.05 26 25 (X=S)b 0.54 1.30 0.76 31 26b 0.15 0.95 0.80 32 28 (R=H)a 0.22 0.66 0.44 33 29 (R=Me)c 0.59 0.75 0.16 34 30 (R=Me)a 0.27 0.57 0.30 35 Fig. 1 Two-dimensional network in the crystal of (7) (ClO4) (reprinted with permission from ref. 15) aIn PhCN.bIn MeCN. cIn CH2Cl2. 1934 J. Mater. Chem., 1998, 8(9), 1933–1944withdrawing heterocycles. The diVerences between the first and second oxidation potentials are smaller than that for TTF. This fact shows that on-site Coulomb repulsion is decreased in these molecules due to the extended p-conjugation. The donors 14 (R=Me) and 15 (R=Me) aVord cation radical salts showing metallic behavior down to 100 K.The structures are similar to those of superconducting TMTSF salts.21 Organic conductors are usually obtained from planar molecules; the planar geometry is considered to be necessary for good p-overlapping via stacking. However, it seems possible to use non-planar molecules to increase dimensionality via multi-dimensional interactions in the crystal.In this context, Fig. 2 Crystal structure of (19)2(PF6) (THF) (reprinted with permission from ref. 23) a composition of 202(PF6)(solvent) where THF, DHF and DO are used as solvent.26 Although their conductivities at room temperature are all 10 S cm-1, the properties at low temperature are dependent on the solvent molecules.Thus, the THF salt is metallic down to 180 K, while the DHF and DO ones show semiconducting behavior at room temperature. No distinctive diVerence in crystal structure between the metallic and semiconducting salts is observed. The diVerence of the ordering of the solvent is considered to appear in the salts of 20 at room temperature. S S S S N S N R R R R S S S S N S N S S S S R R S S S S R R R R 18 19 R, R = benzo 20 R = Me 21 Martý�n et al.have recently reported larger p-extended compounds 21, where a charge transfer from the 1,3-dithiol-2- bis(1,3-dithiole) compounds 18, which are non-planar owing ylidene moieties to the fused anthracene is observed.27 to steric interactions caused by the peri hydrogens, are interesting. 22 These molecules are butterfly-shaped and there are Non-TTF donors intramolecular S,N interactions between the heterocycles, which suppress a large conformational change upon oxidation. Most molecular conductors are formed from donor molecules based on TTF and its derivatives.However, it is important to They undergo one-stage two-electron oxidation, indicating that the cation radicals are thermodynamically unstable. explore new classes of electron donors which do not contain such skeletons to extend the range of conductors.We describe However, some compounds aVord the cation radical salts as stable single crystals. The donor molecule 18 (R, R=benzo) here non-TTF donors with high polarizability. Dithiapyrene derivative 22 forms metallic CT complexes aVords 151 salts where butterfly-shaped molecules are uniformly stacked.22 The overlap of molecules is an alternative with TCNQ, chloranil and bromanil.28 On the other hand, the more polarized nitrogen-containing molecule 23 also aVords mode due to the intermolecular CT interaction. Although the conductivities are fairly high (8.3 to 10-2 S cm-1), they show metallic cation radical salts where several S,S contacts are observed.29 semiconducting behavior due to completely charged salts.On the other hand, a naphthalene-fused derivative 19 aVords 5,10-Dimethyl-5,10-dihydrophenazine (DMPH) is a strong electron donor due to its 16p-electron ring system and aVords cation radical salts incorporating solvent molecules whose composition is 25151 (donor5cation5solvent).23 The PF6 salt conducting CT complexes with acceptors.30 On the other hand, 1,4-dimethyl-1,4-dihydroquinoxaline 24 with a 12p-electron involving THF shows metallic behavior down to 3 K.24 The crystal structure is shown in Fig. 2 where good overlap between ring system is unstable and has not been isolated. We have replaced the benzene rings with 1,2,5-thiadiazole or -selenadia- the butterfly-shaped molecules is observed.The physical properties are strikingly dependent on the incorporated solvent zole rings to give 25 and 26 in order to enhance their stability as well as polarizability, leading to strong intermolecular molecules. Thus, in contrast to the THF incorporated salt, 2,5- dihydrofuran (DHF) and 1,3-dioxolane (DO) incorporated interactions. The absorption maxima of 25 are red-shifted by ca. 50 nm compared with those of DMPH due to the polariz- ones undergo a metal–insulator transition at 150 and 100 K, respectively.25 Since their crystal structures are almost the ation eVect.31 Their electron-donating abilities are comparable to that of dibenzo-TTF and they give cation radical salts and same, the diVerence is attributed to the ordering of the incorporated solvent molecules.The dimethyl derivative 20 also CT complexes. The 12p-electron ring system 26 is a stronger electron donor than TTF in spite of the presence of the gives cation radical salts incorporating solvent molecules with J. Mater. Chem., 1998, 8(9), 1933–1944 1935S S SMe SMe S N S N SMe SMe N N Me Me N N Me Me N N Me Me N X N N N Me Me N S N 22 23 DMPH 24 25 X = S, Se 26 N Me N S N N Me S S S S R R N S N N N S N N R R R R N N R R N S N N S N N N S N OSO2CF3 R 27 28 29 + –OSO2CF3 30 R = Me, Et 31 electron-withdrawing heterocycle (Table 1).32 In the crystal the molecules are connected by S,N contacts (3.05 A° ) to form a coplanar dyad which is held together by hydrogen bonding between the olefinic hydrogens and nitrogens to form a sheetlike network (Fig. 3). The sheet is stacked to give a threedimensional structure. The strong S,N interactions are attributed to electrostatic eVects caused by the polarized structure. This molecule aVords conducting complexes with acceptors containing a 1,2,5-thiadiazole or -selenadiazole ring although no complex with TCNQ is formed. On the other hand, 4-(1,3-dithiol-2-ylidene)-1-methyl-1,4- dihydropyridine 27 has not been prepared because of the instability of the ring system.We have replaced one amino group of 26 by a 1,3-dithiol-2-ylidene unit to give 28.33 These molecules are stronger electron donors than TTF. Their absorption maxima are observed around 515 nm, and are Fig. 4 Crystal structure of 28; broken lines: S,S contacts (reprinted regarded as intramolecular CT bands.The crystal structure of with permission from ref. 33) 28 (R=H) is constructed of four columns which interact with each other via S,S contacts (Fig. 4). These donors form highly Introduction of fused 1,2,5-chalcogenadiazole units to conducting CT complexes with TCNQ. benzidines which are known as strong Wurster type electron donors leads to new polarized donors 29.34 They have absorption maxima in the 500–550 nm region assignable to intramolecular CT bands.X-Ray analyses reveal that the twisted geometry of the neutral molecule 29 becomes planar upon one-electron oxidation, and a coplanar tape-like network is formed by S,N contacts in the crystal of the PF6 salt. 1,1-Dihydro-4,4¾-bi(pyridylidene) has also been used as a strong electron-donating skeleton. Highly polarized electron donors 30 containing this unit have been synthesized by reductive coupling of pyridinium compounds 31.35he absorption maxima are observed at 623 (R=Me) and 627 nm (R= Et) in CH2Cl2.They are stronger donors than TTF (Table 1) and aVord conductive CT complexes with TCNQ and cation radical salts. In the crystal structure of the PF6 salt the donor molecules form two-dimensional columnar stacks and a tapelike network is also formed by short S,N contacts (3.05 A ° ) (Fig. 5). Polarized electron acceptors Quinones Quinones bearing sulfur-containing substituents are considered to be polarized electron acceptors. 2,355,6-Bis(ethylenedithio)- 1,4-benzoquinone 32 is obtained, by reaction of chloranil and ethane-1,2-dithiol, as green crystals.36 Quinones 33 bearing a 1,4-dithiine unit have been prepared by an unusual reaction Fig. 3 Sheet-like network of 26; broken line: S,N interactions and hydrogen bonding (reprinted with permission from ref. 32) of the corresponding dichloro-substituted quinones with phos- 1936 J. Mater. Chem., 1998, 8(9), 1933–1944The quinones 35 have a large contribution of resonance structure 35¾ which leads to high electron aYnities.39 These quinones aVord highly conducting CT complexes with some electron donors.S S O O X X O O MeS MeS SMe SMe S S O O O O S S O– O O O 38 X = Br, Cl 39 + 40 40' Quinones 38 with extended p-conjugation also have strong electron-accepting abilities and aVord conducting complexes with tetrathiatetracene.40 4,4¾-Biphenoquinone 39 has recently Fig. 5 Crystal structure of (30)PF6 (reprinted with permission from been prepared using oxidative coupling of the corresponding ref. 35) phenol.41 This molecule shows the longest absorption maximum at 562 nm. Bis(p-benzoquinone) 40 has a resonance form 40¾ and is deep violet with an absorption maximum of 508 nm.42 This quinone can be reduced to a stable tetraanion species.Heterocyclic analogues of TCNQ Chalcogen atom-containing TCNQ analogues are highly polarized and are expected to have strong intermolecular interactions via heteroatom contacts. In this context, thio- S NC CN S NC S CN CN S S CN CN NC NC S NC CN S NC CN N N R R S NC S CN CN N N N N R R R R NC NC NC NC CN CN CN NC 41 42 43 44 45 46 S S S S O O R' R' S S O O R R S S H PO(OMe)2 R R S S S S X O O X S S S S X– O O X– S S S S O O CN CN NC NC O O S S NC CN CN NC 32 33 34 + + 35 X = C(CN)2, S, O 35' 36 37 phene-TCNQ 41,43 its p-extended analogue 42,44 and fused heteroquinonoid compound 4345 have been prepared.These phonate esters 34.37 The absorption maximum of 33 (R=H, R¾, R¾=benzo) is observed at 546 nm due to intramolecular and related molecules, which are called hetero-TCNQs, have recently been reviewed by Ogura and Otsubo.46 Therefore, we charge transfer.These quinones are weaker electron acceptors and have not been used as components for conducting will focus our attention on the TCNQ analogues containing fused heterocycles. materials. On the other hand, quinones 35–37 fused with electron- Benzothiophene-TCNQ 44 has been prepared to extend the p-conjugation of thiophene-TCNQ.47 However, this molecule accepting sulfur-containing heterocycles are stronger electron acceptors than chloranil.38 The first reduction potential of 35 is a weak electron acceptor due to the fused benzene ring.In order to enhance its acceptor ability, we have replaced the [X=C(CN)2, 0.30 V vs. SCE] is comparable to that of TCNQ. J. Mater.Chem., 1998, 8(9), 1933–1944 1937Table 2 Reduction potentials of acceptors benzene ring with an electron-withdrawing pyrazine ring to give 45.48 These molecules have been prepared together with E/V vs. SCE 46 by reaction of 5,7-dibromothieno[3,4-b]pyrazine with tetracyanoethylene oxide (TCNEO), although the yields are low. compound E1 E2 DE ref. In the crystal of 45 (R=H) there exist three kinds of short TCNQ 0.18 -0.36 0.54 49 S,N contacts between the S atom of the thiophene ring and 35 [X=C(CN)2] 0.30 -0.44 0.74 38 the N atoms of the pyrazine ring, leading to an interesting 35 (X=S) 0.14 -0.60 0.74 38 molecular network with a helical structure (Fig. 6). This 35 (X=O) 0.05 -0.70 0.75 38 acceptor aVords two kinds of CT complexes with TTF, in 36 0.27 -0.39 0.66 38 which unique structures are constructed by S,N interactions 37 0.04 -0.66 0.70 38 between the S atoms of TTF and the N atoms of the CN group. 47 -0.02 -0.49 0.47 49 48 -0.12 -0.55 0.43 52 TCNQ analogues fused with sulfur-containing heterocycles 49 -0.23 -0.55 0.32 52 are also polarized electron acceptors. Bis(thiadiazolo)-TCNQ 50 (R=H) -0.01 -0.46 0.45 53 51 0.12 -0.38 0.50 54 52 0.04 -0.43 0.47 56 53 0.22 -0.29 0.51 56 complex of BTDA 47 with tetraselenotetracene shows metallic behavior down to 1.5 K without undergoing a Peierls transition. 50 The acceptor forms a sheet-like network by S,N interactions in the crystal (Fig. 7).51 The selenium analogues 48 and 49 form similar crystal structures. These molecular N Y N N X N NC CN NC CN N S N NC CN NC CN N N R R 47 X = Y = S (BTDA) 48 X = S, Y = Se (TSDA) 49 X = Y = Se (BSDA) 50 networks are used to selectively incorporate aromatic hydrocarbons such as xylene.52 47 (BTDA) has been prepared from the corresponding dione Thiadiazolopyrazino-TCNQs 50 have also been prepared by reaction with malononitrile in the presence of TiCl4.49 The from the corresponding dione.53 Their reduction potentials are reduction potentials of BTDA are lower than those of TCNQ, a little higher than those of BTDA 47 (Table 2).They undergo as shown in Table 2. The diVerence between the first and reversible four-stage one-electron reduction and give conducsecond reduction potentials is smaller than that of TCNQ, tive CT complexes with electron donors. The X-ray analysis indicating that on-site Coulomb repulsion is reduced.The CT of 50 (R=H) reveals that the coplanar sheet-like network is formed by S,N heteroatom contacts and hydrogen bonding. Thiadiazolo-TCNQ 51 has been prepared by reaction of dibromo compound 54 with malononitrile anion in the presence of a Pd catalyst followed by oxidation with PbO2.54 Compound 51 is a stronger acceptor than BTDA judging from the reduction potentials (Table 2).This acceptor aVords highly conducting complexes with TTF in contrast to BTDA, which forms a mixed stacking with TTF due to its inclusion properties. 55 The selenium and oxygen analogues 52 and 53 have been similarly prepared.56 They also aVord highly conducting complexes with electron donors such as TMTTF. Tetracyanodiphenoquinodimethane 55 fused with four thiadiazole rings has recently been prepared by Suzuki et al.57 Fig. 7 Sheet-like network of BTDA 47 (reprinted with permission Fig. 6 Crystal structure of 45 (reprinted with permission from ref. 48) from ref. 51) 1938 J. Mater. Chem., 1998, 8(9), 1933–1944N X N NC CN NC CN N S N Br Br N S N N S N NC CN NC CN N S N N S N 51 X = S 52 X = Se 53 X = O 54 55 This molecule exhibits reversible mechano- and thermo-chromism involving the folded (yellow) and twisted (violet) conformers.The folded conformer forms a sheet-like network via electrostatic S,N interactions. Kobayashi et al. have prepared a series of thiophene-fused S S N N NC CN R R S S N N CN NC S S N N CN NC N N CN CN S R R R R O O X CN NC N N N S N N S N R R R R 61 62 63 64 65 X = S, Se, Te R = But 66 Heterocyclic compound 66 containing a hypervalent sulfur atom has been prepared by reaction of 5,6-diaminothiadiazolopyrazine with thionyl chloride.64 Its first reduction potential (0.10 V) is higher than that of the corresponding TCNQ S S CN NC NC CN S S CN NC NC CN S S CN NC NC CN S S CN NC NC CN S CN NC NC CN R R 56 57 58 59 60 R = H, Cl analogue 47 (0.02 V), indicating that 66 is a stronger electron acceptor than 47 although it has no electron-withdrawing TCNQs 56–59.58 They are non-planar due to the steric intersubstituents.This molecule has a unique three-dimensional actions caused by the peri hydrogens. The molecules 56 and network in the crystal, which is formed by S,N interactions. 57 keeping the TCNQ skeleton are stronger acceptors than the others, and aVord conductive CT complexes with TTF.Benzo derivative 60 is butterfly-shaped and a weaker Donor–acceptor compounds acceptor.59 There has been no report on their complexes Compounds containing both electron donor and acceptor units with donors. are highly polarized due to intramolecular charge transfer. These compounds are expected to have small HOMO–LUMO DCNQI derivatives gaps which lead to interesting properties such as absorption in the near-infrared region, nonlinear optical properties, and Hu�nig has reported N,N¾-dicyanoquinone diimine (DCNQI) analogues 61 with a thieno[3,2-b]thiophene skeleton.60 These single component conductivity.We discuss here unusual donor–acceptor systems with electrochemically amphoteric compounds are obtained by reaction of the corresponding quinone with bis(trimethylsilyl )carobodiimide and TiCl4.The properties. acceptor 61 (R=Br) aVords highly conductive complexes with electron donors. Donor–p–acceptor systems Thiophene-fused DCNQI analogues 62 and 63 have been We have prepared bis[1,2,5]thiadiazolo-p-quinobis(1,3-dithiprepared by a similar method. They aVord metallic complexes ole) (BTQBT) 67 where the 1,3-dithiole units are the electronwith CuI.An interesting structure composed of polynuclear donating parts and the thiadiazole ones are the electron CuI chains bridged by bidentate ligands 62 or 63 has been accepting parts.65 The conductivity of BTQBT is good (10-3 suggested.61 Martý�n et al. have prepared benzo-fused derivato 10-5 S cm-1) as a single component.A high hall mobility tives 64 which are planar molecules in contrast to the correfor the single crystal has been observed.66 In the crystal the sponding butterfly-shaped TCNQ analogues 60.62 The presence planar molecule forms a sheet-like network via interactions of four fluorine atoms in 64 increases its accepting ability. between the S atoms of the 1,3-dithiole rings (Fig. 8). The S,S contact distance of 3.26 A ° is much shorter than the sum Others of the van derWaals radii (3.70 A° ). The molecules are uniformly stacked with a distance of 3.46 A ° between the molecular planes. [3]Radialenes containing a sulfur, selenium or tellurium atom 65 are highly polarized acceptors with strong electron accepting The ratio of the conductivity along the stacking direction to that along the intercolumnar direction is only 2. The selenadia- abilities.They aVord metallic CT complexes with some donors in spite of the presence of the bulky tert-butyl groups.63 zole analogues 68 and 69 have also been prepared.65b Their J. Mater. Chem., 1998, 8(9), 1933–1944 1939S S S S N Y N N X N O S S N S N N S N 70 67 X = Y = S (BTQBT) 68 X = S, Y = Se 69 X = Y = Se S S CN CN S S CN CN CN NC NMe2 Me2N N C16H33 NC CN CN 74 75 76 77 Push-pull types of conjugated p-electronic systems become highly-coloured chromphores.71 Quinoid compound 74 con- Fig. 8 Sheet-like network of BTQBT 67 [reprinted with permission taining a dicyanomethylene group shows a strong intramolecufrom ref. 65(a)] lar CT band with a longest absorption maximum of 644 nm.72 The CT absorption of diphenoquinoid compound 75 recently prepared is red-shifted to the near-infrared region (957, conductivities are a little higher than that of BTQBT due to 1045 nm).73 Dicyanodiaryl-p-quinodimethane 76 shows its the stronger intermolecular interactions caused by the sellongest absorption maximum at 698 nm.74 enium atom.Metzger et al. have used highly polarized molecule 77 for We have replaced one of the 1,3-dithiol-2-ylidene units of electrical rectification.75 Compound 77 shows an absorption 67 by a carbonyl group to give 70 which is more highly maximum at 884 nm in dichloromethane. It shows both polarized than 67.67 This molecule shows both oxidation and oxidation (Ep=0.49 V) and reduction potentials (E1/2= reduction potentials.In the crystal the planar molecules form -0.513 V).Electrical rectification has been observed in a uniform stack, where short S,S and S,N contacts are Langmuir–Blodget multilayers and even a monolayer of 77. observed between the columns. Compound 70 shows unusual This property is related to a transition of the zwitterionic non-ohmic behavior on measurement of the resistivity.ground-state to the neutral excited state. Martý�n and co-workers have prepared TCNQ analogues Non-classical 1,2,5-thiadiazole containing a hypervalent sulfur atom has a high electron aYnity. Introduction of electron- donating groups to a skeleton containing this unit aVords S N N X X N S N S S S N S N N S N N S N R R 78 79 X = S 80 X = NH 81 X S X S N N NC CN NC CN NC NC S S Me Me Me Me CN CN CN CN 71 X = S, O 72 X = S, O 73 new donor–acceptor systems.Terthiophene derivative 78 has an absorption maximum at 618 nm in dichloromethane.76 This 7168 and DCNQI ones 7269 and 7370 containing a fused dithiine skeleton. These molecules show amphoteric properties absorption is red-shifted to 990 nm in 79 where a pyrazine ring is inserted. Replacement of the thiophene rings in 79 by and have intramolecular CT absorptions.For example, 72 (X=S) shows an absorption maximum at 628 nm, an oxidation more electron-donating pyrrole rings results in a further redshift. 77 The heterocycle 80 has an absorption maximum at potential at 1.51 V, and reduction potentials at 0.05 and -0.36 V.68 1345 nm and the end-absorption reaches to 2100 nm (0.6 eV).78 1940 J.Mater. Chem., 1998, 8(9), 1933–1944The electronic spectra of 79 and 80 are shown in Fig. 9. This low energy gap is attributed to the low LUMO of the nonclassical heterocyclic part and the high HOMO of the pyrrole ring. These heterocycles 78–80 aVord narrow bandgap polymers by electrochemical oxidation.76–78 The bandgaps are dependent on the HOMO–LUMO gaps of the monomers.The electrochemical bandgaps of poly-78, -79 and -80 determined from their cyclic voltammograms (Fig. 10) are ca. 0.9, 0.3 and 0 eV, respectively. Benzobis(thiadiazole)s 81 with electron-donating substituents also have small HOMO–LUMO gaps.79 For example, the morpholine-substituted derivative shows an absorption maximum at 764 nm. X-Ray structure analysis of the diphenyl derivative reveals the formation of a tape-like network via short S,N contacts (Fig. 11). The derivative containing 2- thienyl groups has an absorption maximum at 702 nm and Fig. 11 Tape-like network of 81 (R=Ph) [reprinted with permission from ref. 79(b)] aVords a narrow bandgap polymer with a bandgap below 0.5 eV by electrochemical oxidation.80 TTF derivatives containing electron-accepting parts are Fig. 9 Electronic spectra of 79 and 80 S S O O S S O O S S S S O O S S S S SR SR RS RS O S O R R R R S S R R S S R' 82 83 84 R = H, Me, But 85 R' = CHO, CH=C(CN)2 promising donor–acceptor systems. The derivative 82 containing fused naphthoquinone units, which is slightly soluble in solvent, shows a reduction potential at -0.32 V, but no oxidation potential is reported.81 On the other hand, bis-TTF molecules 83 show amphoteric properties.82 Thus, 83 (R=nbutyl ) shows oxidation potentials at 0.67 (2e) and 1.02 eV, and a reduction potential at -0.27 V.They show intramolecular CT absorptions in solution and in the solid state. The absorption maximum of 83 (R=n-hexyl) is observed at 819 nm in toluene and 1042 nm in the solid state. Quinonoid compounds 84 also show both oxidation and reduction potentials [84 (R=H); Ered=-0.12,-0.22 V, Eox= +1.46 V].83 Their absorption maxima are observed at 531–558 nm.Very recently push-pull TTF derivatives 85 have been reported to exhibit high second-order NLO activities.84 This report shows that highly polarized TTF molecules are promising candidates for materials showing NLO properties. Donor–s–acceptor systems TCNQ derivative 86 has been designed to control the degree Fig. 10 Cyclic voltammograms of (a) poly-79 and (b) poly-80 (reprinted with permission from ref. 78) of charge-transfer.85 However, the donor phenyl groups are J. Mater. Chem., 1998, 8(9), 1933–1944 1941the donor and acceptor units is so small that no CT interaction was observed. Conclusion Highly polarized molecules showing electron-donating, electron-accepting and amphoteric properties have been highlighted in this article.TTF derivatives containing polarizable tuents such as iodine or tellurium atoms aVord conductors with unique three-dimensional crystal structures. Such substituents can be used to increase dimensionality so as to suppress metal–insulator transitions as well as to strengthen intermolecular interactions leading to wide bandwidths. Nitrogen-containing heterocycles such as 1,2,5-thiadiazole or pyrazines are used as electron-accepting units in highly polarized donors.Negatively charged atoms electrostatically interact with positively charged chalcogen atoms to form interesting molecular assemblies. Interactions between the nitrogen and sulfur atoms of thiadiazole rings result in the formation of a molecular tape.Intermolecular CT interactions sometimes lead to alternative stacking of molecules. These facts suggest that such interactions can be used for crystal engineering. Chalcogen atom-containing quinones, and TCNQ and DCNQI analogues are highly polarized acceptors. Short atom contacts between the chalcogen atoms and negatively charged heteroatoms are often observed in the crystals.They are promising electron acceptors to aVord multi-dimensional conductors. Donor–acceptor compounds showing amphoteric properties have small HOMO–LUMO gaps. This leads to interesting optical properties such as absorptions in the near-infrared region and non-linear optical properties. They are also candidates for single-component conductors.Some compounds show semiconducting behavior without doping. However, the synthesis of single-component organic metals has not been accomplished yet and the preparation of such molecules remains as a challenging theme. References 1 Handbook of Organic Conductive Molecules and Polymers, ed. H. S. Nalwa, Wiley, vol. 1, 1997. 2 M. R. Bryce, J.Mater.Chem., 1995, 5, 1481. 3 N. Martý�n, J. L. Segura and C. Seoane, J. Mater. Chem., 1997, 7, 1661. NC CN NC CN NC CN NC CN NC CN NC CN O S S S S O S S O O S S S S SHex SHex S S S S SCH3 H3CS S S N N (CH2) n (CH2) n 86 87 88 89 + + 2PF6 – 90 n = 4, 5 4 J. Roncali, J.Mater. Chem., 1997, 7, 2307. 5 M. Jorgensen and K. Bechgaard, Synthesis, 1989, 207; M. R. Bryce and G. Cooke, Synthesis, 1991, 263. 6 R. Gompper, J. Hock, K. Polborn, E. Dormann and H. Winter, orthogonal to the TCNQ moiety due to steric interactions. Adv.Mater., 1995, 7, 41. TCNQ derivative 87 and the N-cyanoimine analogue bearing 7 T. Imakubo, H. Sawa and R. Kato, J. Chem. Soc., Chem. Commun., 1995, 1097. a longer methylene chain have similar stacking motifs and the 8 T. Imakubo, H. Sawa and R. Kato, J.Chem. Soc., Chem. Commun., formation of segregated stacks of donor and acceptor moieties 1995, 1667. has not been accomplished.86 9 M. Iyoda, H. Suzuki, S. Sasaki, H. Yoshino, K. Kikuchi, K. Saito, Bryce has developed the novel donor–s–acceptor compound I. Ikemoto, H. Matsuyama and T. Mori, J. Mater. Chem., 1996, 88 containing tetracyanoanthraquinodimethane (TCNAQ) and 6, 501.TTF moieties.87 Compound 88 shows a very weak, low-energy 10 N. Okada, H. Yamochi, F. Shinozaki, K. Oshima and G. Saito, Chem. L ett., 1986, 1861. intramolecular CT band at 420–680 nm. In this molecule clean 11 V. Y. Khodorkovsky, C. Wang, J. Y. Becker, A. Ellern, L. Shapiro oxidation of the TTF and reduction of the TCNAQ take place. and J. Bernstein, Adv.Mater., 1994, 6, 656. TTF derivative 89 bearing a benzoquinone unit has also been 12 J.Y. Becker, J. Bernstein, M. Dayan and L. Shahal, J. Chem. Soc., prepared.88 The interaction between the units is small and Chem. Commun., 1992, 1048. each component preserves its identity. 13 C. Wang, A. Ellern, V. Khodorkovsky, J. Y. Becker and In compound 90 a TTF-derived donor is covalently tethered J. Bernstein, J.Chem. Soc., Chem. Commun., 1994, 2115. 14 A. E. Underhill, I. Hawkins, S. Edge, S. B. Wilkes, K. S. Varma, to a bipyrdinium acceptor.89 The CT absorption maxima are A. Kobayashi and H. Kobayashi, Synth. Met., 1991, 42, 2093; observed at 673 (n=4) and 650 nm (n=5). The redox potentials M. Tomura, S. Tanaka and Y. Yamashita, Heterocycles, 1993, 35, of 90 (n=4) are observed at 0.72, 0.58, -0.36 and -0.85 V. 69. Donor–s–acceptor systems composed of TTF and C60 15 M. Tomura and Y. Yamashita, Synth. Met., 1997, 86, 1871. moieties have recently been prepared.90 They show oxidation 16 T. Naito, A. Kobayashi, H. Kobayashi and A. E. Underhill, Chem. potentials derived from the TTF unit and reduction potentials Commun., 1996, 521. 17 G. C. Papavassiliou, S. Y. Yiannopoulos and J.S. Zambounis, derived from the C60 unit. However, the interaction between 1942 J. Mater. Chem., 1998, 8(9), 1933–1944J. Chem. Soc., Chem. Commun., 1986, 820; V. Psycharis, 54 T. Suzuki, Y. Yamashita, C. Kabuto and T.Miyashi, J. Chem. Soc., Chem. Commun., 1989, 1102. A. Hountas, A. Terzis and G. Papavassiliou, Acta Crystallogr., 55 T. Suzuki, C. Kabuto, Y. Yamashita and T.Mukai, Bull. Chem. 1988, C44, 125. Soc. Jpn., 1987, 60, 2111. 18 K. Lahlil, A. Moradpour, C. Merienne and C. Bowlas, J. Org. 56 T. Suzuki, Y. Yamashita, T. Fukushima and T. Miyashi, Mol. Crst. Chem., 1994, 59, 8030. L iq. Cryst., 1997, 296, 165. 19 Y. Yamashita, Y. Kobayashi and T. Miyashi, Angew. Chem., Int. 57 T. Suzuki, T. Fukushima, T. Miyashi and T. Tsuji, Angew. Chem., Ed.Engl., 1989, 28, 1052. Int. Ed. Engl., 1997, 36, 2495. 20 (a) Y. Yamashita, S. Tanaka, K. Imaeda, H. Inokuchi and M. Sano, 58 K. Kobayashi and C. L. Gajurel, J. Chem. Soc., Chem. Commun., Chem. L ett., 1992, 419; (b) Y. Yamashita, S. Tanaka and 1986, 1779; K. Kobayashi, C. L. Gajurel, K. Umemoto and M. Tomura, Phosphorus Sulfur Silicon, 1992, 67, 327; Y. Mazaki, Bull. Chem. Soc.Jpn., 1992, 65, 2168; F. Iwasaki, (c) Y. Yamashita, S. Tanaka, K. Imaeda, H. Inokuchi and M. Sano, N. Toyoda, M. Hirota, N. Yamazaki, M. Yasui and K. Kobayashi, J. Chem. Soc., Chem. Commun., 1991, 1132. Bull. Chem. Soc. Jpn., 1992, 65, 2173; F. Iwasaki, S. Hironaka, 21 Y. Yamashita and S. Tanaka, Chem. L ett., 1993, 73. N. Yamazaki and K. Kobayashi, Bull. Chem. Soc. Jpn., 1992, 65, 22 Y.Yamashita, K. Ono, S. Tanaka, K. Imaeda and H. Inokuchi, 2187. Adv. Mater., 1994, 6, 295. 59 P. Cruz, N. Martý�n, F. Miguel, C. Seoane, A. Albert, F. H. Cano, 23 Y. Yamashita, S. Tanaka and K. Imaeda, Synth. Met., 1995, 71, A. Gonza�lez and J. M. Pingarro� n, J. Org. Chem., 1992, 57, 6192. 1965. 60 E. Gu� nther, S. Hu� nig, K. Peters, H. Rieder, H. G. von Schnering, 24 K.Imaeda, Y. Yamashita, S. Tanaka and H. Inokuchi, Synth.Met., J.-V. von Schu� tz, S. So�derholm, H.-P. Werner and H. C. Wolf, 1995, 73, 107. Angew. Chem., Int. Ed. Engl., 1990, 29, 204. 25 Y. Yamashita, M. Tomura, S. Tanaka and K. Imaeda, Synth. Met., 61 K. Kobayashi, Y. Mazaki, H. Namba, K. Kikuchi, K. Saito, 1997, 86, 1795; K. Imaeda, Jro� ber, C. Nakano, M. Tomura, I. Ikemoto, S.Hino and N. Kosugi, J.Mater. Chem., 1995, 5, 1625. S. Tanaka, Y. Yamashita, H. Kobayashi, H. Inokuchi and 62 N. Martý�n, P. de Miguel, C. Seoane, A. Albert and F. H. Cano, A. Kobayashi, Mol. Cryst. L iq. Cryst., 1997, 296, 205. J.Mater. Chem., 1997, 7, 25. 26 Y. Yamashita, M. Tomura and K. Imaeda, Chem. Commun., 1996, 63 K. Takahashi and S. Tarutani, J. Chem. Soc., Chem. Commun., 2021. 1994, 519; K. Takahashi and S. Tarutani, Adv. Mater., 1995, 7, 639; 27 N. Martý�n, L. Sa�nchez, C. Seoane, E. Ortý�, P. M. Viruela and S. Tarutani, T. Mori, H. Mori, S. Tanaka and K. Takahashi, Chem. R. Viruela, J. Org. Chem., 1998, 63, 1268. L ett., 1997, 627. 28 K. Nakasuji, M. Sasaki, T. Kotani, I. Murata, T. Enoki, K. Imaeda, 64 Y. Yamashita, K. Saito, T. Suzuki, C. Kabuto, T.Mukai and H. Inokuchi, A. Kawamoto and J. Tanaka, J. Am. Chem. Soc., T. Miyashi, Angew. Chem., Int. Ed. Engl., 1988, 27, 434. 1987, 109, 6970. 65 (a) Y. Yamashita, S. Tanaka, K. Imaeda and H. Inokuchi, Chem. 29 J. S. Zambounis, E. Christen, J. PfeiVer and G. Rihs, J. Am. Chem. L ett., 1991, 1213; (b) Y. Yamashita, S. Tanaka, K. Imaeda, Soc., 1994, 116, 925. H. Inokuchi and M.Sano, J. Org. Chem., 1992, 57, 5517. 30 K. Dietz, H. J. Keller, D. No�the and D. Wehe, J. Am. Chem. Soc., 66 K. Imaeda, Y. Yamashita, Y. Li, T. Mori, H. Inokuchi and 1982, 104, 7581. M. Sano, 5; K. Imaeda, Y. Li, 31 Y. Yamashita, K. Saito, T. Mukai and T. Miyashi, T etrahedron Y. Yamashita, H. Inokuchi and M. Sano, J. Mater. Chem., 1995, L ett., 1989, 30, 7071. 5, 861. 32 Y. Yamashita, J. Eguchi, T. Suzuki, C. Kabuto, T. Miyashi and 67 Y. Yamashita, K. Ono, S. Tanaka, K. Imaeda, H. Inokuchi and S. Tanaka, Angew. Chem., Int. Ed. Engl., 1990, 29, 643. H. Sano, J. Chem. Soc., Chem. Commun., 1993, 1803. 33 K. Ono, S. Tanaka, K. Imaeda and Y. Yamashita, J. Chem. Soc., 68 P. Bando, N. Martý�n, J. L. Segura, C. Seoane, E. Ortý�, Chem. Commun., 1994, 899.P. M. Viruela, R. Viruela, A. Albert and F. H. Cano, J. Org. Chem., 34 T. Suzuki, T. Okubo, A. Okada, Y. Yamashita and T. Miyashi, 1994, 59, 4618. Heterocycles, 1993, 35, 395. 69 N. Martý�n, J. L. Segura, C. Seoane, E. Ortý�, P. M. Viruela, 35 Y. Yamashita, K. Ono, M. Tomura and K. Imaeda, Chem. R. Viruela, A. Albert, F. H. Cano, J. V.-Gancedo, C. Rovira and Commun., 1997, 1851.J. Veciana, J. Org. Chem., 1996, 61, 3041. 36 T. Otsubo, Y. Nobuhara, K. Kanefuji, Y. Aso and F. Ogura, 70 H. Gonza�lez, P. Miguel, N. Martý�n, J. L. Segura, C. Seoane, E. Ortý�, J. Phys. Org. Chem., 1988, 1, 275. R. Viruela and P. M. Viruela, Adv. Mater., 1994, 6, 765. 37 N. Martý�n, L. Sa�nchez, C. Seoane, J. Garý�n and J. Orduna, 71 R. Gompper and H.-U. Wagner, Angew.Chem., Int. Ed., Engl., T etrahedron L ett., 1995, 36, 7153. 1988, 27, 1437. 38 Y. Yamashita, T. Suzuki, G. Saito and T. Mukai, J. Chem. Soc., 72 R. Gompper, H.-U. Wagner and E. Kutter, Chem. Ber., 1968, 101, Chem. Commun., 1986, 1489. 4123. 39 T. Suzuki, C. Kabuto, Y. Yamashita and T. Mukai, Bull. Chem. 73 S. Inoue, Y. Aso and T. Otsubo, Chem. Commun., 1997, 1105. Soc. Jpn., 1987, 60, 3459. 74 T. Kawase, M. Wakabayashi, C. Takahashi and M. Oda, Chem. 40 M. Nakatsuka, K. Nakasuji, I. Murata, I. Watanabe, G. Saito, L ett., 1997, 1055. T. Enoki and H. Inokuchi, Chem. L ett., 1983, 905. 75 R. M. Metzger, B. Chen, U. Ho� pfner, M. V. Lakshmikantham, 41 Y. Morita, A. Kashiwagi and K. Nakasuji, J. Org. Chem, 1997, D. Vuillaume, T. Kawai, X. Wu, H. Tachibana, T. V. Hughes, 62, 7464.H. Sakurai, J. W. Baldwin, C. Hosch, M. P. Cava, L. Brehmer and 42 S. Hu� nig, K. Sinzger, R. Bau, T. Metzenthin and J. Salbeck, Chem. G. J. Ashwell, J. Am. Chem. Soc., 1997, 119, 10455. Ber., 1993, 126, 465. 76 S. Tanaka and Y. Yamashita, Synth.Met., 1993, 55–57, 1251. 43 S. Gronowitz and B. Uppstro�m, Acta Chem. Scand., 1974, B28, 981. 77 S. Tanaka and Y. Yamashita, Synth.Met., 1995, 69, 599. 44 K. Yui, Y. Aso, T. Otsubo and F. Ogura, J. Chem. Soc., Chem. 78 S. Tanaka and Y. Yamashita, Synth.Met., 1997, 84, 229. Commun., 1987, 1816. 79 (a) K. Ono, S. Tanaka and Y. Yamashita, Angew. Chem., Int. Ed., 45 K. Yui, H. Ishida, Y. Aso, T. Otsubo, F. Ogura, A. Kawamoto and Engl., 1994, 33, 1977; (b) Y. Yamashita, K. Ono, M. Tomura and J. Tanaka, Bull. Chem.Soc. Jpn., 1989, 62, 1547. S. Tanaka, T etrahedron, 1997, 53, 10169. 46 F. Ogura and T. Otsubo, Conductive Hetero-TCNQs, in Handbook 80 M. Karikomi, C. Kitamura, S. Tanaka and Y. Yamashita, J. Am. of Organic Conductive Molecules and Polymers, ed. H. S. Nalwa, Chem. Soc.,1995, 117, 6791. Wiley, 1997, vol. 1. 81 W. H. Watson, E. E. Eduok, R. P. Kashyap and M. Krawiec, 47 D. Lorcy, K. D. Robinson, Y. Okuda, J. L. Atwood and T etrahedron, 1993, 49, 3035. M. P. Cava, J. Chem. Soc., Chem. Commun., 1993, 345. 82 S. Frenzel and K. Mu� llen, Synth.Met., 1996, 80, 175. 48 K. Suzuki, M. Tomura and Y. Yamashita, J. Mater. Chem., 1998, 83 K. Takahashi, T. Suzuki, K. Akiyama, Y. Ikegami and 8, 1117. Y. Fukazawa, J. Am. Chem. Soc., 1991, 113, 4576. 49 Y. Yamashita, T. Suzuki, T. Mukai and G. Saito, J. Chem. Soc., 84 A. I. de Lucas, N. Martý�n, L. Sa�nchez, C. Seoane, R. Andreu, Chem. Commun., 1985, 1044. J. Garý�n, J. Orduna, R. Alcala� and B. Villacampa, T etrahedron, 50 A. Ugawa, K. Iwasaki, A. Kawamoto, K. Yakushi, Y. Yamashita 1998, 54, 4655. and T. Suzuki, Phys. Rev. B, 1991, 43, 14718. 85 J. Y. Becker, J. Bernstein, S. Bittner, N. Levi and S. S. Shaik, J. Am. 51 C. Kabuto, T. Suzuki, Y. Yamashita and T. Mukai, Chem. L ett., Chem. Soc., 1983, 105, 4468; J. Y. Becker, J. Bernstein, S. Bittner, 1986, 1433. N. Levi and S. S. Shaik, J. Org. Chem., 1988, 53, 1689. 52 T. Suzuki, H. Fujii, Y. Yamashita, C. Kabuto, S. Tanaka, 86 N. Martý�n, J. L. Segura, C. Seoane, A. Albert and F. H. Cano, M. Harasawa, T. Mukai and T. Miyashi, J. Am. Chem. Soc., 1992, J. Chem. Soc., Perkin T rans. 1, 1993, 2363. 114, 3034. 87 P. de Miguel, M. R. Bryce, L. M. Goldenberg, A. Beeby, 53 Y. Tsubata, T. Suzuki, Y. Yamashita, T. Mukai and T. Miyashi, V. Khodorkovsky, L. Shapiro, A. Niemz, A. O. Cuello and V. Rotello, J. Mater. Chem., 1998, 8, 71. Heterocycles, 1992, 33, 337. J. Mater. Chem., 1998, 8(9), 1933–1944 194388 J. L. Segura, N. Martý�n, C. Seoane and M. Hanack, T etrahedron 9391; C. Boulle, J. M. Rabreau, P. Hudhomme, M. Cariou, M. Jubault, A. Gorgues, J. Orduna and J. Garý�n, T etrahedron L ett., L ett., 1996, 37, 2503. 89 K. B. Simonsen, K. Zong, R. D. Rogers, M. P. Cava and J. Becher, 1997, 38, 3909; J. Llacay, M. Mas, E. Molins, J. Veciana, D. Powell and C. Rovira, Chem. Commun., 1997, 659. J. Org. Chem., 1997, 62, 679. 90 N. Martý�n, L. Sa�nchez, C. Seoane, R. Andreu, J. Garý�n and J. Orduna, T etraheron L ett., 1996, 37, 5979; A. I. de Lucas, Paper 8/03151G; Received 27th April, 1998 N. Martý�n, L. Sa�nchez and C. Seoane, T etrahedron L ett., 1996, 3

ISSN:0959-9428    

DOI:10.1039/a803151g

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 Bis(ethylenedioxy)diselenadithiafulvalene (BEDO-STF) Tatsuro Imakubo*† and Keiji Kobayashi Department of Chemistry, Graduate School of Arts and Sciences, T he University of T okyo, Komaba 3-8-1, Meguro-ku, T okyo 153-8902, Japan As a new promising p-donor for organic metals and superconductors, the title compound has been synthesized without the use of highly toxic CSe2 or H2Se gas; the preparation of stable organic metals including a new k-type salt is also mentioned.Bis(ethylenedioxy)tetrathiafulvalene (BO)1 is one of the most successful p-donors derived from bis(ethylenedithio)tetrathiafulvalene (BEDT-TTF or ET), the flagship molecule of the organic p-donors.2 BO has supplied a large number of metallic cation radical salts with various inorganic or organic counter anions3 and two superconducting salts are also known.4,5 The most distinguished feature of the BO molecule compared with other TTF derivatives is the existence of the ethylenedioxy group, and the eVect of the oxygen substitution on the electronic state and crystal structure has been of interest.However, X-ray structure analysis of the BO salt is usually diYcult due to the low quality and small size of the single crystals obtained, and further research on the physical properties has been prevented.From the viewpoint of the improvement of the crystal quality and fine tuning of the electronic state of the cation radical salts, we have tried to introduce selenium atoms into the inner frame of the parent BO molecule. Several modifications of the BO skeleton have been reported by our group and others;6,7 however, all these donors contain only two oxygen atoms on the skeleton. The origin of the characteristic features of the BO molecule, e.g.CH,O type hydrogen bonding and high Scheme 1 Reagents and conditions: i, BuLi (1 equiv.), 0 °C; ii, Se, THF, solubility, must be due to the existence of two ethylenedioxy -35 °C; iii, morpholine-4-thiocarbonyl chloride, -78 °C (68%); iv, groups on the edges of the skeleton, and we have decided Br2, CH2 Cl2; v, 110 °C; vi, Se, NaBH4, AcOH–EtOH (11%); vii, P(OEt)3, benzene, reflux (12%) therefore to fix the outer part of the BO framework and modify the inner TTF moiety. The selenium substitution method is well-known as the one of the most eVective methods of ate.After cooling to-78 °C, morpholine-4-thiocarbonyl chlor- stabilizing metallic and/or superconducting states at low temide was added in one portion and the key intermediate 2 was perature, and the disadvantage of the selenium substitution, obtained in 68% isolated yield as a cream–white powder. lowering of the solubility, should be reduced in the present Synthesis of selone 3 from 2 was achieved via the same case by the two ethylenedioxy groups. Our goal is the synthesis procedure as for the synthesis of 4,5-ethylenedioxy-1,3-dithiole- of the all selenium-substituted p-donor bis(ethylenedioxy)tetra- 2-seleone,1 except for the selenocarbonyl synthesis, in which selenafulvalene (BEDO-TSeF); however, it is usually diYcult we used the NaSeH–AcOH method9 to avoid the use of H2Se to synthesize TSeF analogues without using the highly toxic gas.Treatment of the selone 3 with triethyl phosphite in and foul-smelling reagents, CSe2 or H2Se gas. As the first step refluxing benzene gave the title compound 1 as deep-red of the project, we planned to substitute half of the TTF sulfur microcrystals‡ which had good solubility towards the usual atoms of BO with selenium atoms without using the above organic solvents.troublesome reagents. Here we report the synthesis and proper- The doubling of the all 13C NMR signals of 1 demonstrates ties of a novel BO analogue, bis(ethylenedioxy)diselenadithiathe existence of cis- and trans-isomers. It was impossible to fulvalene 1 (BEDO-STF).The preparation, conductivity and separate these isomers by the usual chromatography methods; crystal structures of metallic cation radical salts are also however, isolation of the neutral isomer is not always necessary mentioned. for the preparation of conducting salts, as has been shown for The synthesis of 1 is outlined in Scheme 1. The starting other mixed chalcogenofulvalenes.10 material is commercially available 2,3-dihydro-1,4-dioxine and Cyclic voltammetry measurements were taken at room the key intermediate of the present route is thiocarbonyl selenoester 2.Highly viscous 5-lithio-2,3-dihydro-1,4-dioxine ‡New compounds were characterized by elemental and/or spectral was obtained by treatment of neat 2,3-dihydro-1,4-dioxine with analyses. Selected data for 1: deep-red microcrystals; 1H NMR (CDCl3, 1 equiv.of BuLi at 0 °C.8 To a THF solution of 5-lithio-2,3- 500 MHz): d 4.267 (m, 4H); 13C NMR (CDCl3, 126 MHz): d (minor dihydro-1,4-dioxine was added powdered selenium at -35 °C isomer) 66.15, 66.54, 102.06, 121.62, 126.57; d (major isomer) 66.18, to give an orange solution of 2,3-dihydro-1,4-dioxine-5-selenol- 66.48, 101.34, 122.19, 125.74; MS (EI, 70 eV): 416 (100%, M+ for C10H8O4S280Se2), 388 (M+-C2H4, 27%), 300 (M+-C4H4O2S, 38%), 252 (M+-C4H4O2Se, 69%); Calc.for C10H8O4S2Se2: C, 29.00; H, 1.95. Found: C, 28.94; H, 2.03%. †E-mail: cimax@komaba.ecc.u-tokyo.ac.jp J. Mater. Chem., 1998, 8(9), 1945–1947 1945Table 1 Cyclic voltammetric data for 1 and related p-donors. Values are given in V vs.the ferrocene/ferrocenium couple donor E1/21/V E1/22/V DE/V ET 0.06 0.38 0.32 BO -0.05 0.27 0.32 BEDO-STF 1 0.01 0.30 0.29 Fig. 2 Crystal structures of k-(1)2GaCl4: top, donor layer arrangement Fig. 1 Temperature dependence of the resistivity for the AuBr2 and viewed along the a axis and bottom, crystal packing viewed along the GaCl4 salts of 1 c axis temperature using a glassy-carbon working electrode in benzo- two-dimensional electronic state (Fig. 2). The calculated Fermi nitrile at 100 mV s-1 with 0.1 M Bu4NBF4 and measured vs. a surface is a closed two-dimensional one and is also in accord- 0.01 M Ag/AgNO3 reference. Table 1 summarizes the half-wave ance with the stable metallic feature. potentials of 1 and related p-donors vs. the potential of the In conclusion, we have synthesized the first seleniumferrocene/ ferrocenium couple.The donor 1 showed two revers- substituted BO analogue using a newly developed CSe2- and ible redox waves. As expected from the selenium substitution H2Se-free synthetic method and obtained a large single crystal of the TTF skeleton, the first redox potential (E1/21) is at a of a new k-type salt suitable for X-ray structure analysis. slightly higher voltage than that of the parent BO; however, Examination of the other metallic and superconducting salts the diVerence between the first and second redox potentials is in progress.(DE=E1/22-E1/21), which reflects the on-site coulombic repulsion, is 0.03 V smaller than those of BO and ET. The decrease We thank Professor R. Kato for the use of his conductivity in DE must be due to the larger size of the 4p atomic orbital measurement apparatus, and the Material Design and on the inner selenium atoms and is promising for the prep- Characterization Laboratory of ISSP (University of Tokyo) aration of stable metallic salts.for the use of their X-ray diVractometer. This work was To confirm the potential ability of the p-donor 1, cation partially supported by Grant-in-Aids for Scientific Research radical salts were prepared via galvanostatic oxidation.Two from the Ministry of Education, Science, Sports and Culture, conducting salts (AuBr2 and GaCl4 salts) were obtained and Japan. both salts are fundamentally metallic down to 4.2 K (Fig. 1). Large thick plate crystals of the GaCl4 salt were harvested and References X-ray structure analysis was performed on a single crystal.§ The donor–anion ratio is 251 and disorder of the selenium 1 T.Suzuki, H. Yamochi, G. Srdanov, K. Hinkelmann and F. Wudl, J. Am. Chem. Soc., 1989, 111, 3108. and sulfur atoms has been observed in all four positions. It is 2 For recent progress in molecular conductors, see Synth.Met., 1997, known that the parent BO molecule tends to construct similar 86 (the latest proceedings of the international conference on sci- types of donor arrangements in most cases,3 however the ence and technology of synthetic metals, ICSM’96).present GaCl4 salt forms a so-called k-type arrangement, which 3 (a) M. A. Beno, H. H. Wang, K. D. Carlson, A. M. Kini, is well-known as the most promising for construction of a G.M. Frankenbach, J. R. Ferraro, N. Larson, G. D. McCabe, J. Thompson, C. Purnama, M. Vashon, J. M. Williams, D. Jung and M.-H. Whangbo, Mol. Cryst. L iq. Cryst., 1990, 181, 145; §X-Ray diVraction data were collected on a Rigaku AFC6S automatic four-circle diVractometer with monochromated Mo-Ka (l=0.71069 A° ) (b) M. Fettouhi, L. Ouahab, D. Serhani, J.-M.Fabre, L. Ducasse, J. Amiell, R. Canet and P. Delhaes, J. Mater. Chem., 1993, 3, 1101; radiation up to 2h=60.2°. The structure was solved by direct methods and refined with full-matrix least-squares methods using reflections (c) S. Horiuchi, H. Yamochi, G. Saito, K. Sakaguchi and M. Kusunoki, J. Am. Chem. Soc., 1996, 118, 8604; with I3s(I). The data were corrected for Lorentz and polarization eVects.Anisotropic thermal parameters were used for non-hydrogen (d) E. I. Zhilyaeva, R. N. Lyubovskaya, S. A. Torunova, S. V. Konovalikhin, O. A. Dyachenko and R. B. Lyubovskii, Synth. atoms except for disordered sulfur atoms. All calculations were performed using the teXsan program package of MSC. Crystal data for Met., 1996, 80, 91. 4 S. Kahlich, D. Schweitzer, I. Heinen, S.E. Lan, B. Nuber, (1)2GaCl4: (C10H8O4S2Se2)2(GaCl4), M=1039.95, monoclinic, space group C2/c (#15), a=36.14(1), b=10.532(5), c=8.073(3) A ° , b= H. J. Keller, K. Winzer and H. W. Helberg, Solid State Commun., 1991, 80, 191. 94.46(3)°, V=3063(1) A ° 3, Z=4, m=63.18 cm-1, Dc=2.255 g cm-3, F(000)=1996.0, R=0.074, Rw=0.046, GOF=3.00 for 2420 observed 5 M. A. Beno, H. H. Wang, A. M.Kini, K. D. Carlson, U. Geiser, W. K. Kwok, J. E. Thompson, J. M. Williams, J. Ren and M.- reflections out of 4735 unique reflections. Full crystallographic details, excluding structure factors, have been H. Whangbo, Inorg. Chem., 1990, 29, 1599. 6 (a) A. M. Kini, T. Mori, U. Geiser, S. M. Budz and J. M. Williams, deposited at the Cambridge Crystallographic Data Centre (CCDC).See Information for Authors, J. Mater. Chem., 1998, Issue 1. Any J. Chem. Soc., Chem. Commun., 1990, 647; (b) J.-M. Fabre, D. Serhani, K. Saoud, S. Chakroune and M. Hoch, Synth. Met., request to the CCDC for this material should quote the full literature citation and the reference number 1145/107. 1993, 60, 295; (c) J. Hellberg, M. Moge, D. Bauer and J-U. von 1946 J. Mater.Chem., 1998, 8(9), 1945–1947Schutz, J. Chem. Soc., Chem. Commun., 1994, 817; (d) T. Imakubo, (b) M. Fetizon, I. Hanna and J. Rens, T etrahedron L ett., 1985, 26, 3453. H. Sawa and R. Kato, J. Chem. Soc., Chem. Commun., 1995, 2493. 7 (a) A.M. Kini, U. Geiser, H. H.Wang, K. R. Lykke, J. M. Williams 9 F. Wudl, E. Aharon-Shalom and S. H. Bertz, J. Org. Chem., 1981, 46, 4612. and C. F. Campana, J. Mater. Chem., 1995, 5, 1647; (b) J. Hellberg, K. Balodis, M. Moge, P. Korall and J-U. von Shu� tz, J. Mater. 10 (a) S. Etemad, T. Penney, E. M. Engler, B. A. Scott and P. E. Seiden, Phys. Rev. L ett., 1975, 34, 741; (b) K. Takimiya, A. Morikami, Chem., 1997, 7, 31; (c) J. Yamada, S. Tanaka, H. Anzai, T. Sato, H. Nishikawa, I. Ikemoto and K. Kikuchi, J. Mater. Chem., 1997, Y. Aso and T. Otsubo, Chem. Commun., 1997, 1925. 7, 1311. 8 (a) R.W. Saylor and J. F. Sebastian, Synth. Commun., 1982, 12, 579; Paper 8/04723E; Received 22nd June, 1998 J. Mater. Chem., 1998, 8(9), 1945–1947 19

ISSN:0959-9428    

DOI:10.1039/a804723e

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 A new low temperature one-step route to metal chalcogenide semiconductors: PbE, Bi2E3 (E=S, Se, Te) Shu-Hong Yu,a,b Jian Yang,a Yong-ShengWu,a Zhao-Hui Han,a Jun Lu,a Yi Xiea and Yi- Tai Qiana,b*† aDepartment of Chemistry, University of Science and T echnology of China, Hefei, Anhui 230026, People’s Republic of China bStructure Research L aboratory, University of Science and T echnology of China, Hefei, Anhui 230026, People’s Republic of China diamine (en) and pyridine (py).The reactions can be expressed by eqn. (1) and (2): A solvothermal reaction of metal oxalates such as PbC2O4, Bi2(C2O4)3 with E (E=S, Se, Te) in organic solvents at relatively low temperature (120–160 °C) produces crystalline PbC2O4+E CA solvent PbE+2CO2( (1) PbS, PbSe, PbTe, Bi2S3, Bi2Se3, and Bi2Te3: Bi2(C2O4)3+3E CA solvent Bi2E3+6CO2( (2) PbC2O4+E CA solvent PbE+2CO2( In a typical procedure, 0.01 mol analytical S (or Se, Te) and Bi2(C2O4)3+3E CA solvent Bi2E3+6CO2( 0.01 mol analytical PbC2O4 were put into a Teflon-lined autoclave of 100 ml capacity, which was filled with py or en to 80% of the total volume.The autoclave was sealed and maintained at 140 °C for 6–12 h and then allowed to cool to room temperature.The dark grey precipitate was filtered and Recently, the synthesis of binary metal sulfides, selenides and washed with ethanol, distilled water, dilute HNO3 solution, tellurides of groups 12, 14 and 15 has been the focus of and absolute ethanol several times to remove the impurities.attention because of their important physical and chemical The product was dried in vacuum at 70 °C for 4 h. properties,1–13 good commercial applications in semi- Reaction of PbC2O4 with E (E=S, Se, Te) produced crystalconductors, pigments, luminescence devices,14 solar cells, IR line PbS, PbSe, and PbTe. All the products were characterized detectors, and optical fiber communications,15 and modern by X-ray powder diVraction.‡ XRD patterns show that the asthermoelectric coolers.6,10,11 Especially, with the planned prepared PbS, PbSe and PbTe powders were pure cubic phases.phase-out of chlorofluorocarbon (CFC) refrigerants within this The X-ray powder diVraction data for the synthesized metal decade, the synthesis of Bi2Te3, which is the most eYcient chalcogenides are summarized in Table 1.The cell parameters semiconductor material for modern thermoelectric cooling are in good agreement with the reported data. The TEM devices, has been the focus of recent research.5–13 image§ in Fig. 1(a) shows that the PbSe powders synthesized Conventionally, metal chalcogenides are synthesized by the in en consist of uniform square particles with an average size reaction of the elements at elevated temperature, typically of 80 nm, which is in good agreement with the average size 500–600 °C, in evacuated tubes,1,16–18 by a solid state metath- calculated by the Scherrer equation.TEM observation indiesis reaction of anhydrous metal halide with Na2(S,Se) or cates that the PbTe powders produced consist of agglomerates Li2(S,Se,Te)19 at high temperature, typically 500 °C, or involved of particles with irregular shape.Elemental analysis¶ show the use of complex and expensive organometallic precur- that the compositions of the products are PbS, PbSe, and sors,3,8,9,11–13,20,21 or by reaction of aqueous metal salt solu- PbTe, which are consistent with the calculated results by XPS. tions with the highly toxic and malodorous gaseous H2S, H2Se, The reaction of bismuth oxalate with sulfur, selenium and H2Te.22,23 Moreover, H2Te is unstable at room temperature, tellurium proceeded similarly to that of lead oxalate.The decomposing into hydrogen and metallic tellurium, which produced Bi2S3, Bi2 Se3, and Bi2Te3 powders display the results in impurities in the final product.7 expected colours of the metal chalcogenides.The Bi2S3, Bi2Se3 Recently, Parkin et al.1,2 reported a new method for synthes- powders produced are the pure orthorhombic phase and izing metal chalcogenides by a reaction between metal and E hexagonal phase, respectively, the cell parameters of which are in liquid ammonia at room temperature. Crystalline PbS and also in excellent agreement with the reported data as shown PbSe powders were obtained at room temperature.1 However, in Table 1.The TEM image in Fig. 1(b) indicates that the the reaction of lead with tellurium in liquid ammonia did not Bi2Se3 powders display flake-like morphology. The selected form PbTe.2 Ritter et al.10 reported a new two-step process for the preparation of polycrystalline Bi2Te3 by reduction of a ‡X-Ray diVraction analysis was performed using a Japan Rigaku complex precursor (Bi2O3 3TeO2 xH2O) with hydrogen at D/Max-cA X-ray diVractometer equipped with graphite monochroma- 275 °C.Our group successfully prepared Bi2S3 nanorods by a tized Cu-Ka radiation (l=1.54178 A ° ), employing a scanning rate of solvothermal reaction between bismuth trichloride and thio- 0.02 ° s-1 in the 2h range from 10° to 60°.urea in ethanol at 140 °C.24 In addition, nanocrystalline b- §TEM images and selected area electron diVraction (ED) patterns were taken with a Hitachi Model H-800 transmission electron micro- In2S3 was also prepared in organic madia in our laboratory.25 scope, using an accelerating voltage of 200 kV. Here we report a novel low temperature one-step route to ¶Elemental analysis by atomic absorption on a Perkin-Elmer 1100B crystalline PbE and Bi2E3 by a solvothermal reaction between atomic absorption spectrophotometer.X-Ray photoelectron spectra metal oxalates and E in organic solvents such as ethylene- (XPS) were recorded on a VGESCALAB MKII X-ray photoelectron spectrometer, using non-monochromatized Mg-Ka X-rays as the excitation source.†E-mail: yqian@mail.ach.ustc.edu.cn J. Mater. Chem., 1998, 8(9), 1949–1951 1949Table 1 X-Ray crystallographic data for metal chalcogenides synthesized by the present route crystalline reference lattice reagents phase detected26 lattice parameters/nm parameters/nm26 PbC2O4+S PbS a=0.59340 a=0.59362 PbC2O4+Se PbSe a=0.6126 a=0.6124 PbC2O4+Te PbTe a=0.6447 a=0.6443 Bi2(C2O4)3+S Bi2 S3 a=1.1150, b=1.1306, c=0.3980 a=1.1149, b=1.1304, c=0.3981 Bi2(C2O4)3+Se Bi2Se3 a=0.4141, c=2.875 a=0.4133, c=2.862 Bi2(C2O4)3+Te Bi2Te3+Te a=0.43852, c=3.0484 a=0.43772, c=3.0483 area electronic diVraction (ED) pattern [Fig. 1(c)] for the sample in Fig. 1( b) demonstrates that the Bi2Se3 powders are single crystalline. Elemental analysis¶ and the XPS analysis show that the compositions of the products are Bi2S3, Bi2Se3.The produced Bi2Te3 sample can be indexed as a hexagonal phase. However, a small amount of Te was detected by XRD in the Bi2Te3 synthesized by the present process. The eVects of solvents on the synthesis of crystalline PbE and Bi2E3 were investigated.We found that crystalline powders of PbE can be prepared both in py and in en at 120–160 °C; however, X-ray diVraction spectra indicated the presence of some Se, Te in the samples PbSe and PbTe synthesized in py.Similar phenomena are observed in the synthesis of crystalline Bi2Se3 and Bi2Te3. These results indicated that solvents en and py have diVerent influences on the completeness of the solvent thermal reaction, and the reaction in en is more complete.The influences of temperature and time on the synthesis of crystalline PbE and Bi2E3 were also studied. If the temperature is lower than 120 °C, the reaction is very incomplete and sometimes can not be initiated. The reaction is also incomplete if the time is shorter than 6 h. It was found that some metallic Bi will be present in the final product Bi2E3 when the temperature exceeds 160 °C, even though Bi2(C2O4)3 is not in excess.We believe that this results from the decomposition of Bi2(C2O4)3 in the system at higher temperature. However, no metallic Pb was detected by XRD even at temperatures up to 180 °C or in the presence of an excess of PbC2O4. These results revealed that the optimum conditions for the synthesis of PbE and Bi2E3 are at 120–160 °C for 6–12 h.In conclusion, a new low temperature one-step route to metal chalcogenides was successfully developed. This simple route eVectively avoids the use of expensive organometallic precursors and malodorous H2S, H2Se, H2Te, and prevents the release of toxic gases. We are beginning to extend this synthetic route to the preparation of other important metal chalcogenides.Financial support from the Chinese National Foundation of Natural Science Research and Anhui Provincial Foundation of Natural Science Research is gratefully acknowledged. This work is also supported by a grant for a key research project from the National Climbing Program. References 1 G. Henshaw, I. P. Parkin and G. Shaw, Chem. Commun., 1996, 1095. 2 G. Henshaw, I. P. Parkin and G.Shaw, J. Mater. Sci. L ett., 1996, 15, 1741. 3 C. B. Murray, D. J. Norris and M. G. Bawendi, J. Am. Chem. Soc., 1993, 115, 8706. 4 C. B. Murray, C. R. Kagan and M. G. Bawendi, Science, 1995, 270, 1335. 5 W. S. Sheldrick and M. Wachhold, Angew. Chem., Int. Ed. Engl., 1997, 36, 206. 6 T. Chivers, J. Chem. Soc., Dalton T rans., 1996, 1185. Fig. 1 Transmission electron micrographs of PbSe and Bi2Se3 powders: 7 C.J. Warren, R. C. Haushalter and B. Bocarsly, J. Alloys Compd., (a) PbSe powders synthesized in en at 160 °C for 12 h; (b) Bi2Se3 1995, 229, 175. powders synthesized in en at 140 °C for 12 h; (c) the selected area electronic diVraction (ED) pattern for (b) 8 A. L. Seligson and J. Arnold, J. Am. Chem. Soc., 1993, 115, 8214. 1950 J. Mater. Chem., 1998, 8(9), 1949–19519 M.Bochmann, X. J. Song, M. B. Hursthouse and A. Karaulor, 19 P. R. Bonneau, R. F. Jarvis and R. B. Kaner, Nature, 1991, 349, 510; J. C. Fitzmaurice, A. Hector and I. P. Parkin, Main Group J. Chem. Soc., Dalton T rans., 1995, 1649. 10 J. J. Ritter, Inorg. Chem., 1994, 33, 6419. Met. Chem., 1994, 17, 537. 20 A. C. Jones, Chem. Soc. Rev., 1997, 101. 11 J.J. Ritter and M. Pichai, Inorg. Chem., 1995, 34, 4278. 12 J. T. Groshens, R. W. Gedridge and C. K. Lowe-Ma, Chem.Mater., 21 A. K. Verma, T. B. Rauchfuss and S. R. Wilson, Inorg. Chem., 1995, 34, 3072; S. Dev, E. Ramli, T. B. Rauchfuss and S. R. Wilson, J. Am. 1994, 6, 727. 13 H. J. Breunig, K. H. Ebert, R. E. Schulz, M. Wieber and I. Sauer, Chem. Soc., 1993, 115, 3316. 22 L. C.Roof and J. W. Kolis, Chem. Rev., 1993, 93, 1037. Z. Naturforsch., T eil B, 1995, 50, 735. 14 N. N. Greenwood and E. A. Earnshaw, Chemistry of the Elements, 23 R. B. King, Encyclopedia of inorganic compounds, Wiley, Chichester, 1994, p. 4113. Pergamon, Oxford, 1990, p. 1403; G. Q. Yeo, H. S. Shen, E. D. Honig, R. Kershaw, K. Dwight and A. Word, Solid State 24 S. H. Yu, Y. T. Qian, L. Shu, Y. Xie, L. Yang and C. S. Wang, Mater. L ett., 1998, 35, 116. Ionics, 1987, 24, 249. 15 A. J. Strausse, Phys. Rev. L ett., 1966, 16, 1193. 25 S. H. Yu, L. Shu, Y. S. Wu, Y. Xie, Y. T. Qian and L. Yang, J. Am. Ceram. Soc., in press. 16 D. Arivuoli, F. D. Gnanam and P. Ramasamy, J. Mater. Sci. L ett., 1988, 7, 711. 26 PDF-2 database, 1990, International center for diVraction data, Swarthmore, PA 19081, 1990. 17 C. Kaito, Y. Saito and K. Fujita, J. Cryst. Growth, 1989, 94, 967. 18 F. D. Rosi, B. Ables and R. V. Jensen, J. Phys. Chem. Solids, 1959, 10, 191. Communication 8/04105I; Received 1st June, 1998 J. Mater. Chem., 1998, 8(9), 1949–1951 1951

ISSN:0959-9428    

DOI:10.1039/a804105i

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 Determination of surface free energy components for heterogeneous solids by means of inverse gas chromatography at finite concentrations V. I. Bogillo,a V. P. Shkilev and A. Voelkelb*† aInstitute of Surface Chemistry of the Ukrainian National Academy of Sciences, Pr. Nauki 31, 252022 Kiev, Ukraine bInstitute of Chemical T echnology and Engineering, Poznan� University of T echnology, Pl.M. Sk�odowskiej-Curie 2, 60–965 Poznan�, Poland The approach for calculating the adsorption free energy and adsorption energy distributions for a heterogeneous solid surface as the sum of two uniform functions directly from co-ordinates of the tail of the probe’s chromatographic peak is presented and discussed. Distribution functions are derived from data collected for the series of test adsorbates (n-alkanes C5—C10 and five polar organic compounds) by means of inverse gas chromatography at finite concentrations.Average dispersive components of surface free energy and the donor/acceptor components of the adsorption energy in the monolayer region for the parent and mixed Si and Al pyrogenic oxides are also determined and discussed. face adhesive ability in the monolayer region should be distri- Introduction butions on the dispersive and donor/acceptor components of The surfaces of most solids used as adsorbents, polymer fillers, the surface free energy and their initial moments, rather than catalysts and their supports are chemically and structurally the parameters related to the Henry region.heterogeneous. Surface properties are generally diVerent to The aim of the present paper is the development of an those of a bulk solid. Depending on the structures of the initial approach for the determination of the components of the reactants, the preparation method and the pretreatment tem- surface free energy, characterising the ability of a heterogeneous perature, diVerent types of adsorption sites may exist on a solid surface to take part in donor–acceptor and dispersive given surface.For example, several Bronsted and Lewis interactions. This approach is based on data collected by acid/base sites are the main types of active sites on the silica, means of inverse gas chromatography at finite concentrations titania, alumina and mixed (Si/Al) oxide surfaces.The proper- for a series of organic probes possessing diVerent polarizability ties of these sites depend strongly on their geometry, and and acid/base properties. The proposed procedure is applied display a distribution of the surface characterising parameters to characterization of the parent and mixed Si and Al pyrogenic and its adsorption capacity.1,2 oxides surfaces in their monolayer region.Inverse gas chromatography is one of the most convenient methods for the determination of the surface properties of Experimental powders.3 This method allows the examination of a given material in terms of the surface free energy and acid/base Materials characteristics of its surface.4 When non-polar probes (n- The following HPLC grade compounds (Aldrich) were used alkanes) are used only London interactions exist between the as test adsorbates: non-polar compounds: n-pentane, n-hexane, adsorbate and the solid surface.The dispersive component of n-heptane, n-octane, n-nonane, n-decane; polar compounds: the surface free energy of any solid in the Henry region may acetonitrile, ethyl acetate, chloroform, ethyl alcohol, and iso- be determined by measuring adsorption free energies for npropyl alcohol.The characteristics of the test adsorbates used alkanes. This method has been applied in the examination of in the calculations of dispersive components of the surface free several inorganic oxides, polymers and fibres.5 However, this energy of the solids and the donor/acceptor components of approach is valid only for systems having a linear adsorption their adsorption energy are listed in Table 1.isotherm. In the case a of heterogeneous solid surface, the The pyrogenic parent and mixed Si and Al oxides were recorded chromatographic peaks are strongly asymmetrical, even for the low volumes of the injected liquid probe. They are typical for non-ideal, non-linear chromatography and are Table 1 Molar deformation polarization (PD), donor (DN) and not suitable for measurements of the dispersive increment of acceptor (AN and AN*) numbers of test compounds the surface free energy and its donor/acceptor characteristics (i.e.acid/base properties). During investigation of the adsorp- DN/ AN*/ tion equilibrium at low coverage of a heterogeneous surface, compound PD/cm3 kcal mol-1 AN kcal mol-1 the most active sites will be covered first.These sites correspond n-pentane 249.0 0 0 0 only to the initial part of the adsorption isotherm and they n-hexane 297.4 0 0 0 are not representative of all of the active sites. Therefore, the n-heptane 245.5 0 0 0 components of the surface free energy determined by the use n-octane 391.6 0 0 0 of extremely low volumes of the test probes are related only n-nonane 438.2 0 0 0 to those strongest (and highly energetic) adsorption sites, which n-decane 483.1 0 0 0 acetonitrile 110.5 14.1 18.9 41.7 form only a small fraction of all active sites.Owing to the ethyl acetate 221.0 17.1 9.3 1.5 surface energetic heterogeneity, the correct parameters of surchloroform 212.5 4 25.1 5.4 ethyl alcohol 127.4 20 37.9 10.3 isopropyl alcohol 184.3 29 33.0 3.1 †E-mail: voelkel@fct.put.poznan.pl J.Mater. Chem., 1998, 8(9), 1953–1961 1953synthesized by Chlorovinyl Co. (Kalush, Ukraine) in joint follows: flame hydrolysis of their chlorides. These oxides were used as chromatographic supports in the present study: silica sample q=k D mP0 h (t-t0)dh (1) (Aerosil 175) with a specific adsorption area measured by the BET method (low-temperature nitrogen adsorption) SA= where q is the adsorbed amount per gram of adsorbent, D is 170±15 m2 g-1, alumina sample (SA=140±12 m2 g-1) and the corrected flow of gas through the column, m is the weight mixed alumina–silica containing 30 wt.% Al in a silica matrix of adsorbent in the column, t0 and t are the retention times with SA=170±16 m2 g-1. corresponding to the air peak and to the maximum of the solute peak, respectively, k is a proportionality coeYcient between the height of the peak, h, and the corresponding IGC experiments concentration, c, of the solute in the gas phase (k=c/h).The The gas chromatographic measurements were carried out with concentration c is given by: the use of an LHM-80 gas chromatograph (Russia), equipped with a katharometer detector.The analog output from the c= hqm DSpeak (2) detector was digitalized and recorded on an IBM PC 386 microcomputer controlled by original software (Turbo Pascal where Speak is the area of the chromatographic peak and qm is 7.0). Helium was used as the carrier gas. Air, as a nonthe injected amount of the test solute.interacting marker, was used to measure the dead volume of Eqn. (1) is valid if the temperature and the carrier gas/probe the column. Injection of the test compounds was repeated at volumetric rate are constant in the whole column. The con- least three times. Flow rate was measured at the end of the ditions of the one-peak method are fulfilled when the tails of column with a bubble flow meter and its value was maintained the peaks from several injections (at increasing probe concen- at 20 cm3 min-1. Pressure measured at the inlet and outlet of tration) superimpose.Therefore it is possible to use a single the column was used to calculate the net retention volume by peak corresponding to the maximum concentration of the test the usual procedure.6 The molecular probes were injected solute in the detector. This single peak should be cut into i manually with a Hamilton microsyringe (Hamilton microliter slices corresponding to i pressures and amounts of the test 700 and 7000 series syringe).The volume of injected liquid solute. Eqn. (1) and (2) may then be transformed into eqn. (3) probe varied from 0.5–10 ml. and (4): The examined solids were agglomerated, crushed and sieved to give particles 200–320 mmze and placed in the chromaqi= k D m P0 h (ti-t0)dhi (3) tographic column (stainless steel, 40 cm long, 4 mm i.d.).The columns were conditioned under helium at 200 °C for 12 h before their use. The chromatographic measurements were ci= hiqm DSpeak (4) carried out at temperature varying from 120–170 °C (isothermal conditions) with an increment of 10 °C.The where ti is the retention time for the point localized on the temperature of the detector and sampler was 200 °C. right side of the i-slice at its height hi, i.e. on the peak’s tail. The integration of eqn. (3) is performed from 0 to the maximum Calculation of primary data height, h, of the chromatographic peak. Fig. 1 and 2 present the chromatographic peak of n-hexane Adsorption free energy distribution (or adsorption energy on the alumina/silica surface at 130 °C and the adsorption distribution) parameters were calculated immediately from the isotherms obtained from sequential integration of the n-hexane profile of the tail of the chromatographic peak at constant peaks measured at three diVerent temperatures, respectively.temperature and adsorbate amount (5 ml ). Calculation of This peak is strongly asymmetrical and it does not relate to adsorbate pressure in the gas phase and saturated vapour the adsorption in the Henry region. The n-hexane adsorption pressure at the temperature of the experiment was performed isotherms are convex relative to the adsorbate pressure axis. with the use of Antoine’s equation.6 The cross-sectional area This isotherm shape corresponds to a stronger adsorbate/ of adsorbate molecules on the flat surface was estimated using surface interaction in comparison with attractive interactions their liquid densities at 298 K, assuming a spherical molecular between adsorbate molecules on the surface.Usually, the shape in a hexagonal close-packed configuration, or by using critical volumes of the test adsorbates.7 Surface coverage was determined directly from the injected amount, the height of the peak, the area of the chromatographic peak, the crosssectional area of the probe, the specific adsorption area of the chromatographic support and the amount of probe in the column.6 Results and Discussion Determination of the adsorption free energy distribution directly from the parameters of the chromatographic peak The ‘multiple injection’ and ‘one-peak’ methods are commonly applied to the analysis of the chromatographic peak dependence on the known amount of the liquid probe injected into the chromatographic column, and for the following determination of the adsorption isotherm on the surface of the chromatographic support.6 In the first method a given amount of solute is injected quickly, inducing a peak.Adsorption or desorption may be followed by a mathematical examination of the front and tail of the recorded peak. The amount of the Fig. 1 Profile of the chromatographic peak of n-hexane on the alumina/silica surface at 403 K solute adsorbed onto the solid surface can be calculated as 1954 J. Mater.Chem., 1998, 8(9), 1953–1961The distribution function r(x) may be presented as a stepped function including the sum of two uniform distribution functions r1(x) and r2(x) with joint first initial moment: r(x)=a r1(x)+(1-a)r2(x) (10) where 0a1 and r1(x), r2(x) are defined as follows: r1(x)=G 1 x2-x1 at x1<x<x2 0 atx<x1; x>x2 and r1(x)=G 1 x4-x3 at x3<x<x4 0 atx<x3; x>x4 (11a) The limits of these distributions are related by the following inequality: x1<x3<x4<x2 (11b) The analytical solutions of eqn.(8) and for the derivative of Fig. 2 Adsorption isotherms of n-hexane on the alumina/silica surface the adsorption isotherm in terms of eqn. (10) and (11) may be derived from the peak profiles recorded at three diVerent temperatures written as follows: H(P, T )= a x2-x1 ln G1+P exp (x2) 1+P exp (x1)H determination of the adsorption energy distribution from the chromatographic peak parameters requires the approximation + 1-a x4-x3 ln G1+P exp (x4) 1+P exp (x3)H (12) of its co-ordinates by suitable polynomials or by spline functions and subsequent sequential integration of the peak’s slices.and This leads to the accumulation of numerical errors and to an increase of the distribution function uncertainty. Therefore we dH(P, T ) dP = a x2-x1 G exp (x2) 1+P exp (x2) - exp (x1) 1+P exp (x1)H propose a simple direct analytical procedure for the calculation of the adsorption free energy distribution directly from the co-ordinates of the peak’s tail.+ 1-a x4-x3 G exp (x4) 1+P exp (x4) - exp (x3) 1+P exp (x3)H(13) As shown in Fig. 1, the following relationship can be evaluated immediately from the co-ordinates of the tail of the All these values x1, x2, x3, x4 and a may be calculated by peak recorded in the chromatographic experiment the least-mean-square method from an adsorption isotherm using eqn. (12), or from the co-ordinates of a chromatographic dH dP =f (P) (5)peak by using eqn. (13).The next step in the procedure is the determination of the distribution cumulants, i.e. the average value (xav) and its where H and P are the relative surface coverage and the variance (sx2) from values of the initial moments of the adsorbate vapour pressure in the gas phase respectively. The distribution on x: common integral equation of the adsorption on the heterogeneous solid surface at isothermal conditions, assuming the Langmuir isotherm for the local surface coverage, may be xav=M1= a(x2+x1) 2 + (1-a) (x4+x3) 2 (14a) presented as follows:8 M2= a(x23-x13) 3(x2-x1) + (1-a) (x43-x33) 3(x4-x3) (14b) H(P, T)=P0 2 (K0)-1 exp (E/RT ) [1+(K0)-1 exp (E/RT )] x(E)dE (6) sx2=M2-M12 (14c) where H(P, T) is the overall surface coverage at temperature T and adsorbate vapour pressure P, x(E) is a normalized where M2 and M2 are the first and second initial moments of diVerential distribution function of the surface of adsorption the distribution function on variable x.energy E, K0 is the Langmuir constant, R is the universal gas The xav values are related to the limits of the distribution constant.The Langmuir constant may be estimated by using (x1, x2, x3, x4) by simple expressions: the Jaroniec equation:9 x1=xav-d1 (15a) K0=PS exp (DHvap/RT) (7) x2=xav+d1 (15b) where DHvap is the heat of evaporation of the adsorbate and x3=xav-d2 (15c) PS is its saturated vapour pressure at temperature T. Let us substitute in eqn. (6) the distribution function of E x4=xav+d2 (15d) [x(E)] by a distribution function [r(x)] of the following where d1 and d2 are the half-widths of the outer and inner variable x=E/RT-ln(K0).This equation is then transformed uniform distribution functions r1(x) and r2(x), respectively. to: The average (Eav) and the variance (sE2) of the adsorption energy distribution [x(E] are related to the above cumulants H(P, T)=P0 2 P exp (x) 1+exp (x) r(x)dx (8) of the distribution function on x, as follows: Eav=RT xav+ln(K0) (16a) These functions are related by the simple expression: x(E)=RTr(x) (9) sE2=(RT)2sx2 (16b) J.Mater. Chem., 1998, 8(9), 1953–1961 1955Using the above expressions one can calculate the limits of the sum of two uniform distribution functions r1(x) and r2(x) [or x1(E) and x2(E)] with equivalent areas.It should be stated two distribution functions on E in eqn. (17a), similar to eqn. (10): that the adsorption energy distribution, as the sum of two uniform distribution functions, was chosen in the present study x(E)=ax1(E)+(1-a)x2(E) (17a) to represent the possibility of the analytical solution of eqn. (8). Only a few analytical solutions are obtained for common integral equation of adsorption on the heterogeneous solid x1(E)=G 1 E2-E1 at E1<E<E2 0 atE<E1; E>E2 surface, containing simple local adsorption isotherms (Langmuir, Henry, Jovanovic) by using c, exponential, Gaussian and two discrete distribution functions of the adsorption energy.10–13 However, any of these functions may reflect and x2(E)=G 1 E4-E3 at E3<E<E4 0 atE<E3; E>E4 (17b) the actual properties of the heterogeneous surface under investigation.From a mathematical point of view, eqn. (6) and (8) where E1<E3<E4<E2. are Fredholm integral first-order equations. The solution of The limits of the function x(E) (Ei, i=1–4) are related to the this equation with respect to r(x) or x(E) functions is a limits of the function r(x) (xi, i=1–4) by a simple relation: numerically ill-posed problem, i.e., small changes in the measured adsorption H(P, T) caused by experimental errors may Ei=RT xi+ln(K0) (18b) significantly distort the sought function.One of most suitable It is clear that the half-widths of the outer and inner uniform numerical methods for solving this ill-posed problem is the functions x1(E) (h1) and x2(E) (h2) coincide with those for regularization procedure.14 The number and areas of the peaks functions r1(x) and r2(x) multiplied by RT: h1=RT d1 and in the distribution curve obtained using this method can, in h2=RT d2.principle, reflect the number of diVerent types of adsorption Then the adsorption energy distribution x(E) may be calcu- sites, their actual concentrations and adsorption potentials on lated from eqn.(12)–(18) and eqn. (7)–(9). The Langmuir the heterogeneous surface. For example, two peaks are exhibconstant K0 in the present study was calculated given the ited in the adsorption energy distribution curves for diethyl assumption that this value is independent of the type of solids ether and n-pentane on hydroxylated and trimethylsilylated and of the temperature within the temperature range used in silica surfaces.15 Two peaks are also observed in the adsorption our experiments (120–170 °C).Parameters for the adsorption energy distribution curves for methanol and dichloromethane free energy distribution function [DGA=RT ln(K0)-E], its on hydroxylated and octadecyldimethylsilylated silica suraverage (DGAav) and variance (sDGA) were calculated directly faces.16 The position and intensity of the peaks depend on the from eqn.(13)–(16) using the following expressions: r(DGA)= structure of the adsorbates and the solid surface under investix( E); DGAav=-RT xav=RT ln(K0)-Eav; sDGA=sE2. gation. The above distribution curves are calculated using For example, the adsorption energy distribution calculated diVerent modifications of the regularization procedure from from the parameters of the n-hexane peaks’ tail on the alumina/ the adsorption isotherms measured by the classical volumetric silica surface, at three diVerent temperatures, as the sum of method15 or by inverse gas chromatography.16 The number of two uniform functions, is shown in Fig. 3.This distribution is peaks is in contradiction with the possible diVerence in the equal to the reference one calculated from eqn. (12) with the number of site types on the hydrophilic and hydrophobic silica use of adsorption isotherms. surfaces for adsorption of non-polar and polar organic As the exact computation of the five parameters (x1, x2, x3, molecules.17 x4 and a) by the least-mean-squares method requires highly The optimization of the regularization parameter is crucial accurate experimental points for the peaks’ tail, the a value is for the sought-for adsorption energy distribution.This paramassumed to be 0.5 in all our calculations. In such a case, the eter is dependent on the relative experimental error. A very distributions r(x) or x(E) represent stepped functions including low value of the regularization parameter gives rise to spurious peaks, while too high a value over-smooths the distribution function.It was shown that two nearby peaks may be recovered well by this method only for simulated isotherms without experimental error. Even an error as low as 0.1% flattens sharp peaks and gives only a shoulder. Wide peaks may be recovered by this method almost independently of the number of experimental points and the error inherent in the input data.For example, two to three peaks in the adsorption energy distribution curve are evaluated from low-temperature nitrogen adsorption on the surfaces of diVerent activated carbons which accounts for a relative experimental error of 0.01%. Characteristically, these distribution curves at a relative error of 1% present only one peak with a long tail.18 As a rule, the relative experimental error for points on the adsorption isotherms which are determined by means of gas chromatography exceeds 1%.This means that only wide peaks may be well represented on the adsorption energy distribution curve from these data using the regularization procedure.Hence, any analytical method including such a wide distribution may be used in these calculations. It was shown that analytical solution of the Fredholm integral equations of the first order, including two uniform distributions of both radiative and non-radiative rate constants for bimolecular photoreaction kinetics on the heterogeneous surface, gives the apparent rate constant distri- Fig. 3 The adsorption energy distribution function of n-hexane on the bution, which is close to that calculated by the regularization alumina/silica surface as the sum of two uniform functions calculated method.19 Therefore, as proposed in the present study, the from parameters of the peaks tail recorded at three diVerent temperatures adsorption energy distribution as the sum of two uniform 1956 J.Mater. Chem., 1998, 8(9), 1953–1961distribution functions is suitable for the description of adsorp- topography of the sites with a diVerent correlation length between their adsorption energies.8 Therefore, one can expect tion data evaluated from chromatographic experiments with intermediate accuracy. Moreover, as the following calculations that the sDGA(CH2)2 value is equal to zero in the case of a patchwise surface topography and an increase in the case of a of dispersive and donor/acceptor components of the surface free energy for the heterogeneous solids require only knowledge random site topography.of the first and second initial moments of the adsorption free energy distribution, the choice of such a simple distribution Determination of donor/acceptor components of adsorption energy in the monolayer region of a heterogeneous solid surface function possessing well defined limits is valid.Several attempts were made to evaluate the specific interaction Determination of the dispersive component of the surface free parameters of surface free energy from the adsorption data of energy in the monolayer region of heterogeneous solid test compounds by means of IGC experiments at infinite dilution.4,21,22 It was proposed to calculate the dispersive component of the By analogy with the method of calculation of specific surface free energy of a solid (cSd) in the Henry region values interaction components of the adsorption heats of polar test of the adsorption free energy (DGA) for n-alkanes series4 adsorbates,22 the average values of these components in the obtained from gas chromatographic data (IGC) finite concentration region may be calculated by using the average values of the adsorption energy distributions for test cSd= 1 4cCH2 CDGA (CH2) NaCH2 D2 (19) nonpolar and polar probes.23 The average values of the adsorption energy distributions where cCH2 is the surface energy of methylene group for polar test compounds may be presented as the sum of (35.6 mJ m-2 at 293 K).The variation of cCH2 with temperature average values for the components of specific (donor–acceptor is given by cCH2=35.6+0.058(293-T) (in mJ m-2).20 aCH2 is or hydrogen bonds) and non-specific (dispersive) interactions the area occupied by a methylene group (aCH2=0.06 nm2).N between the surface and adsorbates is the Avogadro number and DG(CH2) is the adsorption free energy of a methylene group. Eav=Eavsp+Eavnsp (25) Eqn. (19) may be transformed into the following relationship The energy of dispersive interactions between the large between the adsorption free energy at infinite dilution for the organic compound and the surface sites is proportional to the n-alkanes series and their carbon number (n) compound’s molecular induced polarizability22 or its molar -DGA=n2NaCH2 (cSd)1/2 (cCH2)1/2+b (20) deformation polarization (PD)1 where b is a constant depending on the surface area of the Eavnsp=(KP)avPD+f (26) solid in the column and on the choice of a reference state of where the (KP)av coeYcient is proportional to the average the adsorbed solutes.When we have to apply the linear polarizability of the surface adsorption sites and f is a constant eqn. (20) for the description of the adsorption equilibrium on characteristic for a given solid surface. The donor (DN) and the heterogeneous surface, we can use cumulants of the first acceptor (AN) numbers24 describe the ability of the polar test (DGAav) and second order (sDGA 2) for the adsorption free energy solute to act as an electron donor and electron acceptor, distribution [r(DGA)] in this expression.respectively, in interactions with the surface active sites As the term 2NaCH2 (cCH2)1/2=z does not depend on the (acid/base parameters) surface coverage, the average value and variance of distribution on the [DGA(CH2)]2 value may be calculated from the relation- Eavsp=AN (KD)av 100 +DN (KA)av 100 (27) ships between the average adsorption free energy of n-alkanes (DGAav) or its variance (sDGA2) and the number of carbon atoms where (KA)av and (KD)av coeYcients denote the average ability (n).These parameters may be determined from the following of the surface sites to act as acceptor (electron acceptor or relationships Lewis acid) and donor (electron donor or Lewis base) at -DGAav=nz-1[DGA(CH2)]av+bav (21) interaction with the polar adsorbates.The quantities (KA)av and (KD)av are equal to the average acceptor and donor and numbers of the surface sites. sDGA 2=n2z-2/sDGA(CH2)+sb2 (22) The (KD)av and f parameters may be determined by using values of the average adsorption energies for the n-alkanes where bav and sb2 are the average and variance of the series on the examined solid surfaces.The coeYcients of constant b. eqn. (27) are easily determined by the least-squares method The average value for DGA(CH2), given as [DGA(CH2)]av is from the linear relationship for the series of polar adsorbates related to the average value for [DGA(CH2)]2, given as characterized by diVerent DN and AN values {[DGA(CH2)]2}av by the simple relationship {[DGA(CH2)]2}av={[DGA(CH2)]av}2+sDGA(CH2)2 (23) 100Eavsp AN = DN AN (KA)av+(KD)av (28) The average value for the dispersive component of surface free energy of a heterogeneous solid can be calculated by using However, diVerent DN and AN scales for the test adsorbates eqn.(24) are usually used for the evaluation of the donor–acceptor characteristics for solid surfaces from IGC data.For example, (cSd)av=z-2{[DGA(CH2)]2}av (24) extended DN and AN scales calculated from the plot of the original DN Gutmann values vs. parameter B* (shift of the It should be mentioned that the theory of multisite adsorption on the heterogeneous solid surface predicts the valence vibration band of the OD group of deuteromethanol in the liquid under investigation) in accordance with the dependence of variance on the adsorption energy distribution for the n-alkanes series on the type of surface topography.8 In relation DN=-6.36+0.19B* and from the plot of the original AN Gutmann values vs. ET parameters (energy of the low the case of random topography of the surface sites this variance (sE2 or sDGA2) should increase with the increase in the number electron transition of some betaine zwitterions) in accordance with the equation AN=-40.52+1.29ET,25 were proposed in of carbon atoms in the n-alkane, whereas this is independent of the n value in the case of a patchwise surface topography.ref. 26. Riddle and Fowkes have shown that the 31P NMR spectrum of triethylphosphine oxide, used in determining the As a rule, real heterogeneous surfaces possess intermediate J.Mater. Chem., 1998, 8(9), 1953–1961 1957Table 3 Half widths of two uniform distribution functions (h1 and h2), AN values for polar compounds, is appreciably shifted downthe average adsorption energy (Eav) and its mean-square deviation field due to van der Waals interactions with the solvents.27 (sE) (in kJ mol-1) at a=0.05, calculated from adsorption of the test Hence, the AN values were corrected for the van der Waals solutes on the pyrogenic alumina surface contribution to the chemical shift on the basis of the determination of the cd values from the measurements of the surface adsorbate h1 h2 Eav sE and interfacial tensions of the test liquids.In many cases, this n-pentane 24.2 19.1 67.0 12.6 correction is quite substantial. The corrected AN values for n-hexane 40.9 7.0 70.6 16.7 polar test adsorbates are designated as AN*. The corrected n-heptane 13.0 0.001 75.0 5.3 (KD*)av values were determined from eqn. (29) by analogy to n-octane 12.2 0.0013 79.4 5.0 eqn. (28) n-nonane 12.8 0.0013 82.7 5.2 n-decane 14.4 2.53 84.9 6.0 acetonitrile 87.7 2.66 84.4 35.8 100Eavsp AN* = DN AN* (KA*)av+(KD*)av (29) ethyl acetate 21.0 11.7 79.6 9.8 chloroform 12.9 0.0018 74.1 5.3 Usually, one can also take into account the variances in ethyl alcohol 28.6 20.8 91.1 12.0 donor (sKD 2) and acceptor (sKA 2) numbers of the heterogeneous isopropyl alcohol 17.4 8.8 90.4 7.8 surface.In accordance with the theorem of probability theory,21 the common expression for these variances may be written as Table 4 Half widths of two uniform distribution functions (h1 and h2), follows the average adsorption energy (Eav) and its mean-square deviation (sE) (in kJ mol-1) at a=0.05, calculated from adsorption of the test sE2=PD2sKP2+AAN 100B2 sKD2+ADN 100B2 sKA2+sf2 (30) solutes on the pyrogenic alumina/silica surface adsorbate h1 h2 Eav sE where sKP 2 and sf2 are the variances of distributions on the KP parameter of the surface and on the constant f, respectively.n-pentane 21 2.6 59.0 13.6 n-hexane 29.8 25.0 62.6 16.0 Components of the surface free energy and adsorption energy of n-heptane 25.1 18.0 68.0 12.8 Si and Al pyrogenic oxides n-octane 15.1 4.5 72.4 6.4 n-nonane 10.8 0.022 76.7 4.4 The above approach was applied to the calculation of the n-decane 7.32 0.0005 80.9 3.0 adsorption energy distributions for n-alkanes series and diVer- acetonitrile 14.3 6.2 81.4 6.4 ent polar organic compounds and to estimate the average ethyl acetate 21.0 9.4 81.6 9.5 chloroform 18.4 10.8 66.1 8.7 values of distribution functions and on the dispersive compoethyl alcohol 22.7 17.5 92.1 11.7 nent of surface free energy as well as those for the donor/ isopropyl alcohol 27.0 18.1 88.4 14.2 acceptor components of the adsorption energy distributions in the monolayer region. These parameters were determined for pyrogenic parent and mixed Si and Al oxide surfaces.region were determined by using eqn. (23) from the squares of The half widths of two uniform distribution functions, average values [DGA(CH2)]av. The average adsorption free average adsorption energies of the test solutes and their meanenergy for n-alkanes series on the examined oxides are square deviations at a=0.05 on the surfaces of the examined presented in Table 5. oxides are presented in Tables 2–4.The linear relationships between the average adsorption free Variance of the adsorption free energy distribution for the energy of n-alkanes and their carbon numbers for the oxide studied adsorbate/surface systems changes from 0.5 to 14 kJ2 surfaces are shown in Fig. 4. We calculated the average mol-2. These high values of the variance are due to the increment of the adsorption free energy of the methylene considerable energetic heterogeneity of the oxide surfaces.No group, dispersive component of the surface free energy, specific significant change in the variance of adsorption free energy for components of the adsorption energy of polar adsorbates n-alkanes series on their carbon number was found. Therefore, (Table 6) as well as average acceptor and donor numbers for it was assumed that sDGA(CH2)2#0 in the eqn.(23) for all studied the adsorption sites of these oxides. These data are presented inorganic oxides. This is in accordance with predictions of the in Table 7. The plots of average specific components of the multisite adsorption theory on the heterogeneous surface.8 The adsorption energy for test polar compounds on the oxide above value indicates in our case patchwise rather than random surfaces vs.their DN/AN or DN/AN* ratios in the co-ordinates topography of the surface sites of the examined inorganic of eqn. (28) and (29) are shown in Fig. 5(a) and (b). oxides. The values of the average dispersive component of The average dispersive components of the surface free energy surface free energy of these surface sites in the monolayer of pyrogenic inorganic oxides (Table 7) decrease in the following order: alumina>alumina/silica>silica. The dispersive component of the surface free energy is proportional to the overall Table 2 Half widths of two uniform distribution functions (h1 and h2), the average adsorption energy (Eav) and its mean-square deviation polarizability of the surface sites, their ionization energy and (sE) (in kJ mol-1) at a=0.05, calculated from adsorption of the test to maximum partial charges on the atoms of these sites.The solutes on the pyrogenic silica surface above sequence may be explained by the presence of very adsorbate h1 h2 Eav sE Table 5 Average values of the adsorption free energy (DGAav, in kJ mol-1) at 403 K of n-alkanes on the surfaces of pyrogenic parent n-pentane 39.5 23.1 64.0 18.7 n-hexane 27.4 20.6 65.6 14.0 and mixed Si and Al oxides, calculated at a=0.05 n-heptane 19.1 7.5 67.0 8.3 n-octane 20.9 12.9 69.4 10.0 adsorbate SiO2 Al2O3 Al2O3/SiO2 n-nonane 19.4 11.1 70.7 9.1 n-decane 27.1 18.7 71.9 13.4 n-pentane 13.0 5.0 6.7 n-hexane 12.6 1.5 4.5 acetonitrile 31.3 30.0 71.4 17.7 ethyl acetate 25.3 18.3 72.6 12.8 n-heptane 10.0 -1.8 1.1 n-octane 5.4 -4.4 0.35 chloroform 17.0 0.0006 66.1 6.9 ethyl alcohol 51.2 27.7 84.1 23.4 n-nonane 3.1 -7.0 -3.1 n-decane 2.3 -9.5 -3.8 isopropyl alcohol 21.2 12.1 84.4 10.0 1958 J.Mater. Chem., 1998, 8(9), 1953–1961adsorption sites on the silica surface. Increase of temperature and surface coverage causes a decrease in the cSd value. For example, estimates of cSd values at 293 K when the temperature coeYcient of cSd from ref. 29 was taken into account and our (cdS)av data at 403 K lead to cSd=60–80 mJ m-2. Hence, the average dispersive component of the surface free energy of silica (28.6 mJ m-2 at 403 K) found in the present study may be treated as the true value of this parameter characterizing the overall popularity of the surface sites. The dispersive component of the surface free energy for alumina samples depends on the modification (a- or c-phases, boehmite), temperature of the pretreatment of the surface and on the content of silica and other oxides in their matrix.It was reported that c-alumina pretreated at 473 K has higher cSd=115 mJ m-2 at 373 K, in comparison with Aerosil 200 (cSd=65 mJ m-2) under the same conditions.31 The boehmite (Al2O3, H2O or AlOOH) is characterized by cSd=172 mJ m-2 at 353 K.31 The cSd value for alumina samples decreases from Fig. 4 Relationships between the average values of the adsorption free 100 mJ m-2 to 65 mJ m-2 and to 42 mJ m-2 at 373 K with an energy and the carbon number of n-alkanes on the pyrogenic parent increase in silica content from 45–630 ppm and to 1060 ppm and mixed Si and Al oxide surfaces at 403 K in the alumina matrix.32 Additionally, it was reported that cSd increased from 48 to 71 mJ m-2 for aluminas after their dry Table 6 Average specific interaction components of the adsorption energy (Eavsp, in kJmol-1) of polar test compounds on the surfaces of grinding.33 Therefore, despite the large diVerences in the parent and mixed Si and Al pyrogenic oxides reported cSd values for silica, alumina and alumina/silica, one may conclude that the cSd value increases during the transition compound SiO2 Al2O3 Al2O3/SiO2 from silica to alumina samples and may decrease due to the increase of silica content in the alumina matrix.This conclusion acetonitrile 12.3 28.2 36.0 fits well with our data for variation of (cSd)av values for ethyl acetate 9.6 14.7 25.6 chloroform 3.4 9.8 10.9 pyrogenic silica, alumina and alumina/silica samples.ethyl alcohol 24.4 33.6 45.0 No clear relationship was found between the variances for isopropyl alcohol 22.7 28.4 35.9 adsorption energy distribution of polar probes and squares of their donor or acceptor numbers (correlation coeYcients were found to be below 0.5). Table 7 Average values of the increment of adsorption free energy for The electron donor ability or basicity of adsorption sites the methylene group ([DGA(CH2)]av, in kJmol-1), dispersive component of the surface free energy at 403 K [(cSd)av, in mJm-2], acidities of the examined oxides in accordance with the change of [(KA)av and (KA*)av], basicities [(KD)av, in kcal mol-1 and (KD*)av], their (KD)av and (KD*)av values increases in the order: relative acidities (KA/KD, mol kcal-1 and KA*/KD*) of the surfaces of silica<alumina/silica<alumina and their electron acceptor pyrogenic parent and mixed Si and Al oxides, and correlation ability or acidity, as it follows from their (KA)av and (KA*)av coeYcients (R) of the relationships (21), (28) and (29) values, varies as: silica<alumina<alumina/silica.The ability of the inorganic oxide surfaces to interact with parameter SiO2 Al2O3 Al2O3/SiO2 organic compounds as acid or as base can be estimated from [DGA(CH2)]av 2.15 2.89 2.29 the comparison of the mean electronegativities of these solids. (cSd)av 28.6 51.8 32.5 The mean orbital electronegativity of a solid (xS) may be R for eqn.(21) 0.996 0.996 0.999 calculated from the following relationships34 (KD*)av 21.8±7.8 46.7±21.0 34.4±25.9 (KA*)av 11.7±1.4 16.8±3.8 32.9±4.7 KA*/KD* 0.54 0.36 0.96 xS= S(nixi) Sni (31) R for eqn. (29) 0.979 0.931 0.971 (KD)av 5.8±2.8 12.4±6.6 9.6±8.0 and (KA)av 11.2±2.8 15.0±6.6 30.9±8.0 KA/KD 1.93 1.21 3.21 xi=IPi+EAi (32) R for eqn. (28) 0.916 0.793 0.911 where xi, IPi and EAi are the Mulliken orbital electronegativity, *In calculating these parameters the average specific component of adsorption energy was expressed in kcal mol-1 while DN and AN first ionization potential and electron aYnity of the ith atom values were taken from ref. 24. in the inorganic molecule, respectively, while ni is the number of these atoms in the molecule. The estimated xS values for alumina and silica are 5.82 eV and 6.61 eV, respectively.This polar Brønsted acid sites and highly polarizable Lewis acid/base sites on the surfaces of alumina and mixed oxides means that acidic properties increase during the transition from alumina to silica, whereas the basic properties increase containing Al in the silica matrix, in comparison to the parent silica. in the opposite direction.It is well known that bulk alumina exhibits basic properties during interaction with typical Some diYculties exist when the cSd values from the present study are compared with those evaluated from IGC data at Brønsted acids, whereas silica typically displays acidic properties. This conclusion coincides with that observed in the present infinite dilution. It was reported that the cSd value for pyrogenic silica (Aerosil 130 from Degussa) equals 46.5 mJ m-2 at 383 K,4 study of an increase of (KD)av during the transition from silica to alumina.The surface of a-Al2O3 is characterized by a higher or 40±4 mJm-2 at 353 K.28 These cSd values were obtained in Henry’s region. It is known that cSd for initial and modified donor number (KD=21 kcal mol-1) in comparison with its acceptor number (KA=15)35 (data from inverse chromatogra- silicas decreases with the increasing temperature by 0.3–0.5 mJ m-2 °C-1.29 Much higher cSd values were reported phy at infinity dilution and zetametry measurements).This agrees with the increase of basic properties, determined on the for Aerosil 300 (cSd=76 mJ m-2 at 333 K, 67 mJ m-2 at 363 K and 68 mJ m-2 at 393 K)30 and for Spherosil XOB 75 (cSd= basis of its (KD)av and (KA)av values in the present study, on changing from silica to alumina samples. 80 mJ m-2 at 293 K).29 These values are related to the strongest J. Mater. Chem., 1998, 8(9), 1953–1961 1959Fig. 5 Plots of the average specific components of the adsorption energy for test polar compounds on the parent and mixed Si and Al pyrogenic oxide surfaces vs.(a) their DN/AN ratio [eqn. (28)] and (b) their DN/AN* ratio [eqn. (29)] It is known that OH groups of the alumina surfaces display Conclusions a wide range of Brønsted acidic and basic properties.2 The The approach was proposed for the calculation of the single OH groups of the surfaces exhibit mainly typical basic adsorption free energy distribution and the parameters describ- properties, whereas the bridged OH groups bonded to trigonal ing donor–acceptor and dispersive properties for hetero- aluminium atoms behave as typical Brønsted acid sites.36 geneous solid surfaces.The calculations were carried out with It should be mentioned that the behaviour of such main the use of chromatographic data for a peak profile measured active sites as OH groups at the oxide surfaces depends at finite adsorbate concentrations.Moreover, the average strongly on the composition of the oxide and the local chemical dispersive and donor–acceptor components of surface free environment. On partially dehydroxylated surfaces they have energy for a heterogeneous solid in the monolayer region may varying acid/base character and interact with adsorbates be obtained by the use of the proposed procedure.These according to acid/base characteristics, such as ionization potenparameters seem to be the most appropriate characteristics of tial, electron aYnity, proton aYnity and partial charges on the adhesion ability for a heterogeneous solid surface. The calcuatoms in the surface cluster. The comparison of the positive lations of such parameters were performed for adsorption sites charges on the OH groups and lengths of the OH bonds in on the surfaces of parent and mixed Si and Al pyrogenic the surface clusters of possible OH groups on the silica, oxides.It was found that the surface adsorption sites of mixed alumina and alumina/silica surfaces leads to the conclusion alumina–silica exhibit a highly acidic character in comparison that the acidity of these groups decreases in the transition to the parent oxides.from alumina/silica to alumina and to silica.36 Recent quantum chemical computations have shown that the main Brønsted This work was partially supported by PUT grant DS 32/265/97. acidic sites of silica are single and geminal OH groups, whereas those of binary alumina/silicas are bridged OH groups and water molecules coordinated on a trigonal aluminium References atom.36,37 The observed enhancement of acidic characteristics 1 V.I. Bogillo, A. S. Semenyuk and E. V. Utlenko, Ukr. Khim. Zh. for the mixed alumina/silica compared with the parent Al and (Russ. Ed.), 1994, 60, 398. Si oxides may be explained in terms of the formation of strong 2 H.Knozinger and P. Ratnasamy, Catal. Rev. Sci. Eng., 1978, 17, 31. Lewis acidic sites and by enhancement of the acidity of OH 3 Inverse Gas Chromatography. Characterization of Polymers and groups which are bound to the aluminium cations on the Other Materials, ed. D. R. Lloyd, T. C. Ward and H. P. Schreiber, ACS Symp. Ser. No. 391, American Chemical Society,Washington boundaries between Al2O3 patches and the SiO2 lattice.38 DC, 1989.The average acidity of the oxide surfaces, which can be 4 C. Saint Flour and E. Papirer, Ind. Eng. Chem. Prod. Res. Dev., estimated as the KA/KD ratio, depends on the choice of 1982, 21, 666. acceptor number (AN or AN*) (Table 7). When the AN values 5 A. Voelkel, Crit. Rev. Anal. Chem., 1991, 22, 411.are used, i.e. acceptor properties of the test adsorbates are 6 R. C. Reid, J. M. Prausnitz and T. K. Sherwood, T he Properties of Gases and L iquids, McGraw-Hill, New York, 1977. overestimated, the relative acidity of all examined oxides is 7 J. R. Conder and C. L. Young, Physicochemical Measurements by less than one, i.e. they exhibit basic rather than acidic proper- Gas Chromatography,Wiley, New York, 1979. ties.Their relative acidity decreases in the following order: 8 W. Rudzinski and D. H. Everett, Adsorption of Gases on alumina/silica>silica>alumina. This order is not changed Heterogeneous Surfaces, Academic Press, New York, 1991. when the AN* values for polar adsorbates accounting for the 9 M. Jaroniec, Surf. Sci., 1975, 50, 553. 10 S. Ross, Adsorption T echnol., 1971, 67, 1.contribution of van der Waals interactions are used. However, 11 J. Roles and G. Guiochon, J. Chromatogr., 1992, 591, 345. in this case (Table 7) the relative acidity of the oxides is more 12 J. A. Lum Wan and L. R. White, J. Chem. Soc., Faraday T rans., than one, i.e. they exhibit acidic rather than basic properties. 1991, 87, 3051. Therefore, application of the AN* values in the calculations 13 M.Jaroniec, X. Lu and R. Madey, J. Phys. Chem., 1990, 94, 5917. of the acid/base properties of the solid surfaces results in the 14 A. M. Puziy, V. V. Volcov, O. I. Poznayeva, V. I. Bogillo and V. P. Shkilev, L angmuir, 1997, 13, 1303. increase of their relative acidity. 1960 J. Mater. Chem., 1998, 8(9), 1953–196115 V. I. Bogillo, V. P. Shkilev and G. R. Yurchenko, Ukr. Khim. Zh., 27 F. L. Riddle, Jr. and F. M. Fowkes, J. Am. Chem. Soc., 1990, 112, 3259. 1995, 61, 318. 28 H. Haidar, H. Balard and E. Papirer, Colloids Surf. A, 1995, 99, 45. 16 M. Pyda and G. Guiochon, L angmuir, 1997, 13, 1020. 29 W. J. Wang, S. WolV and J. B. Donnet, Rubber Chem. T echnol., 17 R. K. Iler, T he Chemistry of Silica, Wiley Interscience, New York, 1991, 64, 559. 1979. 30 E. Papirer, H. Balard, Y. Rahmani, A. P. Legrand, L. Facchini and 18 R. Leboda, J. Skubiszewska-Zieba and V. I. Bogillo, L angmuir, H. Hommel, Chromatographia, 1987, 23, 639. 1997, 13, 1211. 31 E. Papirer, G. Ligner, H. Balard, A. Vidal and F. Mauss, in Proc. 19 V. I. Bogillo, V. P. Shkilev and V. V. Osipov, L angmuir, 1997, Chemically Modified Surface Symp., Midland, Michigan, June 13, 945. 28–30, 1989, ed. D. E. Leyden and W. T. Collins, Gordon and 20 G. M. Dorris and D. G. Gray, J. Colloid Interface Sci., 1980, 77, Breach, New York, 1989, pp. 15–26. 353. 32 E. Papirer, J. M. Perrin, B. SiVert and G. Philipponneau, Prog. 21 J. Schultz, L. Lavielle and C. Martin, J. Adhesion, 1987, 23, 45. Colloid Polym. Sci., 1991, 84, 252. 22 J. B. Donnet, S. J. Park and H. Balard, Chromatographia, 1991, 33 E. Papirer, J. M. Perrin, B. SiVert, G. Philipponneau and 31, 434. J. M. Lamerant, J. Colloid Interface Sci., 1993, 156, 104. 23 G. A. Korn and T. M. Korn,Mathematical Handbook for Scientists 34 V. I. Bogillo, Composite Interfaces, submitted. and Engineers, Nauka, Moscow, 1968. 35 B. SiVert, J. Eleli-Letsango, A. Jada and E. Papirer, Colloids Surf., 24 V. Gutmann, T he Donor-Acceptor Approach to Molecular 1994, 92, 107. Interactions, Plenum Press, New York, 1978. 36 V. I. Bogillo and V. M. Gun’ko, L angmuir, 1996, 12, 153. 25 R. Schmidt and V. N. Sapunov, Non-Formal Kinetics, Verlag 37 G. M. Zhidomirov and V. B. Kazansky, Adv. Catal., 1986, 34, 131. Chemie,Weinheim, 1982. 38 J. Sauer, J.Mol. Catal., 1989, 54, 312. 26 A. Vidal, E. Papirer, W. M. Jiao and J. Donnet, Chromatographia, 1987, 23, 121. Paper 8/01703D; Received 2ndMarch, 1998 J. Mater. Chem., 1998, 8(9), 1953–1961 1961

ISSN:0959-9428    

DOI:10.1039/a801703d

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 A phenomenological approach to the inversion of the helical twist sense in the chiral nematic phase Marcus J. Watson,a Mark K. Horsburgh,b John W. Goodby,*c† Kohki Takatoh,d Andrew J. Slaney,e Jay S. Patelf and Peter Styringc aCavendish L aboratory, T he University of Cambridge,Madingley Road, Cambridge, UK CB3 0HE bDepartment of AppliedMathematics and T heoretical Physics, T he University of Cambridge, Silver Street, Cambridge, UK CB3 9EW cDepartment of Chemistry, T he University of Hull, Hull, UK HU6 7RX dT oshiba Corporation, 33, Shin-Isogo-Cho, Isogo-Ku, Yokohama, 235, Japan eDRA, St Andrew’s Road,Malvern, UK WR14 3PS fDepartment of Physics, 215 Davey L aboratory, University Park, PA 16802-6301, USA Optically active materials that show a temperature dependent inversion of the helical twist sense in the chiral smectic C* and chiral nematic phases have been known for many years. However, it has only recently been found that inversions can occur in compounds which have single chiral centres.It was found previously that the temperature range and the magnitude and sign of the helical twist in the chiral nematic phase are related to the concentration of the optically active material(s) dissolved in a nematic host.In a similar way, we propose to describe the inversion of the helical twist sense in the chiral nematic phase of pure materials containing a single chiral centre. Additionally, we seek to verify the possible validity of the latter model by means of molecular modelling on appropriate compounds, and by deriving a suitable mathematical expression to allow the direct use of experimental data.Introduction Addition of a chiral dopant to a nematogen will induce a helical distortion in the nematic structure. The same helical distortion is also found in pure chiral nematogens. Locally, a chiral nematic mesophase is very similar to its achiral equivalent, where there is no long-range positional ordering and the preferred molecular orientation is defined by a director n� .However, the preferred orientations of the molecules are such that the director n� is not unidirectional in space but is helical with one of two possible helical twist directions, dextro (D) and laevo (L), as represented in Fig. 1. Other chiral mesophases, such as the smectic C* phase, exhibit helical distortions similar to the one described above.1 Optically active compounds that show a temperature dependent inversion of the helical twist sense in the chiral smectic C* and chiral nematic phases have been discovered over a period of many years.2–11 In particular, Slaney et al.showed that an inversion could occur for single chiral centre systems6 and, more recently, a compound has been reported to exhibit a temperature dependent unwinding of the helix in both the chiral smectic C* and chiral nematic phases.4 It was hypothetically suggested12 that the temperature range of the chiral nematic phase and, more importantly, the magnitude and sign of the helical twist are related to the concentration of the single optically active material dissolved in a Fig. 1 Schematic representation of the arrangement of molecules in nematic host. Within the dilute limit, the helical pitch length the chiral nematic mesophase. The successive planes have been drawn p is inversely proportional to the concentration.13 Additionally, as a guide for the eye but do not have any specific physical meaning for concentrated solutions of cholesterol derivatives, it was empirically found that the total twist q: of a mixture is the helical twist, defined as the helical wavevector p-1, is an approximately equal to the average of the component twists, additive property of the various components of a chiral nematic qi.12 The various component twists may be positive or negative. system.The observation of a temperature dependent pitch Hence, a suitable mixture of two components with opposing inversion in the chiral nematic phase of a pure material twists can be nematic, when, q:=0, i.e.it was suggested that containing only one chiral centre6 can also be seen to be due to a competition between component entities. In this case, the component species with opposing helical twists can be thought †E-mail: j.w.goodby@chem.hull.ac.uk J.Mater. Chem., 1998, 8(9), 1963–1969 1963of as being molecular species, say conformational isomers, and microscopy and DSC, are summarised in Table 1. In the first that the relative concentration of the species varies with instance, compound 4 does not appear to exhibit a twist temperature. Thus, the additivity model can be extended to inversion.On the other hand, there is a N*–N2 * transition single molecular systems where the qi’s are the twists of the which leads us to believe that the temperature range of the competing species of the system. Firstly, as for the dilute lower temperature chiral nematic phase is very short and was model, it is assumed that there is a linear dependence between not detected.reciprocal pitch and concentration, even at very large concen- The pitch length was determined as a function of temperature trations of component species, and secondly, it is assumed that for all compounds represented. For compounds 1–6, the pitch the concentration of the component species will change with was determined as a function of temperature by measuring the temperature. Hence, a pure system with opposite component distance between the dechiralisation lines in the fingerprint twists may show a temperature dependence of the helical texture of the phase using a calibrated Filar eyepiece attached twist sense.as the ocular of a Zeiss polarizing light microscope. The pitch The purpose of this study is to provide a phenomenological of compound 3 was determined as a function of temperature model for the inversion of the helical twist sense in helical using the Grandjean wedge method (also known as the Cano liquid-crystalline phases.Additionally, molecular modelling of wedge method).4 The wedge cells were obtained from EHC materials which exhibit a helical twist inversion in the chiral (Japan) and calibrated by interferometry.The experimental nematic phase was performed and a microscopic origin to the errors for all pitch measurements were estimated to be ca. phenomenon is proposed. 10% of the experimental pitch. Molecular modelling studies were performed on a Silicon Graphics workstation (Indigo XS24, 4000) using QUANTA Experimental and CHARMm. Within CHARMm, the Adopted Basis The compounds employed for the purpose of this study are Newton-Raphson (ABNR) algorithm was used to locate the given by structures 1–6 in Fig. 2, and were selected because molecular conformation with the lowest potential energy. The some exhibit a progression in chemical structure (1–4) and the minimisation calculations were performed until the root mean others, in relation, show a diversity in molecular structures (5, square (RMS) force reached 4.184 kJ mol-1 A ° -1, which is 6).Materials 1, 2 and 4 were prepared by Dr K. Takatoh, close to the resolution limit. The RMS force is a direct measure compound 3 was synthesised by J. D. Vuijk, and compounds of the tolerance applied to the energy gradient (i.e. the rate of 5 and 6 were synthesised by Drs A. J. Slaney and C. Loubser, change of potential energy with step number) during each cycle respectively.The synthesis and optical purity of terphenyl of minimisation. If the average energy gradient was less than epoxides 1–4,4,8 the phenyl propiolate 56 and the difluorobi- the specified value, the calculation was terminated. phenyl 67 have been reported previously. The results of the molecular mechanics calculations were The latter two compounds clearly have single chiral centres.generated using the programs QUANTA ver. 4.0 and However, the first four have epoxide chiral ring structures that CHARMm ver. 22.2. The programs were developed by can, in principle, act as single chiral entities. Finally, for Molecular Simulations Inc. The modelling packages assume unequivocal comparison, all materials were selected on the the molecules to be a collection of particles held together criterion that they exhibit a helix inversion in the chiral nematic by elastic forces, in the gas phase, at absolute zero, in an ideal phase.The phase sequences, as determined by optical motionless state, and the force fields used are those described in CHARMm ver. 22.2. Results Model The additivity model described earlier expresses the total twist q: as a function of the average of the component twists qi of a mixture.As suggested, it is proposed to extend this model to pure systems where the competing entities are molecular species which could be, for example, conformational isomers or rotomers, and the qi’s are the molar twists of the system’s contributing species.Firstly, in a similar way to the model describing a single component in dilute solution,13 it is assumed that there is a linear dependence between reciprocal pitch and concentration, even at very large concentrations of the single component. Secondly, it is assumed that the helical twist, defined as the inverse pitch or helical wavevector p-1, is an additive property of the various components of a chiral nematic system.For simplicity, it is proposed that the inversion of the helical screw sense in the chiral nematic phase is brought about by the action of two competing species within the phase, set a with a positive twist giving rise to a right-handed (D) helix and set b with an opposite twist favouring a left-handed (L) helix. These two groups are presumed to be two molecular species weighted around two energy minima corresponding to two static species.In fact, these are presumed to be in dynamic flux and are constantly interconverting from one species to another. The model therefore only describes an average picture of the structure of the mesophase. ROCH2 O C3H7 O C9H19O O O O O C8H17O F F O O O F O C6H13 F Me Cl F * * * * 1 R = CH3 2 R = C2H5 3 R = C3H7 4 R = C4H9 5 6 4''- n-alkoxymethyl-2'-fluoro-4-[(2 S, 3 S)-3- n-propyloxiran- 2-ylmethoxy][1,1':4',1'']terphenyl ( S)-2-chloropropyl 4'-(4- n-nonyloxyphenylpropioloyloxy)- biphenyl-4-carboxylate 4- n-octyloxy-2,3-difluorobiphenyl-4'-yl 3-fluoro-4-[( S)-2-fluorooctanoyloxy]benzoate Fig. 2 Compounds 1–6 In essence, these assumptions lead to the expression in 1964 J.Mater. Chem., 1998, 8(9), 1963–1969Table 1 Transition temperatures (°C) for materials 1–6 as determined by polarized light optical microscopy and diVerential scanning calorimetry material K SmC* SmA* N* N2 * N* BPI BPII I 1a · ——————— 86.3 · 90.7 · 90.7 · —————— 176.3 · 2a · 52.9 · ——— 100.6 · 106.4 · 109.6 · 159.5 · 159.5 · 159.5 · 3b · <46.2 · ——— 103.3 · 106.3 · 112.1 · 158.8 · 162.9 · 164.4 · 4a · 32.6 · 112.4 · ——— 114.1 · 116.1 · 138.8 · 138.8 · 139.5 · 5c · ——— 62.0 · 137.0 · ~141 · ~142 · —————— 166.0 · 6d · 89.7 · ——— 139.3 · ~140 · ~140 · —————— 149.6 · aRef. 8. bRef. 4. cRef. 6. dRef. 7. eqn. (1), 1 p = [a] pa + [b] pb (1) where p is the total pitch of the system, [a] and [b] are the concentrations, and pa and pb are the pitch coeYcients, of the rival species, respectively.Thus, the reciprocal of pa and pb are a direct measure of the twisting power of the species considered. Let the energy diVerence between the two species be DE. Their relative concentration will therefore be temperature dependent and will follow a Boltzmann distribution [eqn. (2)], [a] [b] =e -DE RT (2) where R is the fundamental constant 8.314 J mol-1 K-1 and T is the absolute temperature.For simplicity, let [a]+[b]=1, which yields eqns. (3) and (4). Fig. 4 Temperature dependence of the reciprocal experimental and theoretical helical pitch of compound 5 [a]= 1 (1+eDE/RT (3) Though such an apparent relationship has also been observed [b]= eDE/RT (1+eDE/RT (4) in other systems,9 eqn. (5) clearly expresses inverse pitch as a non-linear function of temperature.However, in the tempera- Therefore, the total inverse helical pitch is given by eqn. (5). ture range under study, the relationship may appear to be linear. 1 p = 1 (1+eDE/RT) A1 pa + eDE/RT pb B (5) Theoretical curves were fitted to the helical pitch and reciprocal helical pitch of compound 5 according to the model Taking compound 5 as an example, the pitch length deter- described by eqn.(5). These are shown concurrently with the mined experimentally as a function of temperature is shown experimental data in Fig. 3 and 4, respectively. in Fig. 3. Examination of Fig. 3 shows that at the inversion For all materials studied, the unknown parameters T 2, pa, temperature T 2 the pitch is infinite. This corresponds to the pb and DE were ascertained using a least-squares method and formation of a nematic phase which exhibits schlieren and are summarised in Table 2.homeotropic defect textures when observed optically under It has to be noted that the theory will give good fits only if crossed polars. Thus, at T 2, the reciprocal of the pitch is equal data points ca. 5 °C below the clearing temperature T Cl are to zero.Furthermore, there appears to be a linear relationship taken. Experimental pitch data between T Cl and T Cl-5 °C between temperature and reciprocal pitch,6 as shown in Fig. 4. tend to be distorted, possibly due to the eVects of molecular fluctuations and the unpinning of defects at the surfaces of the cell. For materials 1–6, all pitch length data points were taken below T Cl-5 °C.Molecular modelling Molecular modelling studies are described using compound 5 as an example. A conformational search around the OMC(1)MC(2)*MCl dihedral angle, shown in Fig. 5, was performed on the geometrically optimised structure. The search was performed using a 360° grid scan. A 5° step size was Table 2 Values for T 2, DE, pa and pb derived from eqn.(1) material T 2/K DE/kJ mol-1 pa/mm pb/mm 1 361.6 6.480 -0.0297 0.2564 2 381.2 8.329 0.0257 -0.3559 3 389.7 6.757 0.031 -0.2495 4 377.3 7.065 -0.0389 0.37 5 414.6 8.261 0.0612 -0.6724 Fig. 3 Temperature dependence of the experimental and theoretical 6 408.1 5.408 -0.0118 0.0581 helical pitch of material 5 J. Mater. Chem., 1998, 8(9), 1963–1969 1965Table 3 Comparison of the energy diVerence between the two lowest minimum energy static conformers DEc, estimated by computer modelling, and the energy diVerence between two competing species, DE, evaluated from eqn.(5) computer modelling theoretical model material DEc/kJ mol-1 DE/kJ mol-1 C9H19O O O O O Cl Me * C(1) C(2) Fig. 5 Description of the OMC(1)MC(2)*MCl dihedral angle in 5 1 2.3 6.480 2 2.0 8.329 selected because, in this model, we are considering sets of 3 2.0 6.757 4 1.7 7.065 conformers about energy wells.At each step, the grid torsion 5 8.7 8.261 was artificially fixed to prevent the structure from returning to 6 8.1 5.408 the initial geometrically optimised structure and the resulting conformation was minimised using a Steepest Descents algorithm; 200 iterations were suYcient because complete geometri- 8.7 kJ mol-1.The energy diVerences between the two lowest cal optimisation had already been performed. The search gives energy conformers of compounds 1–4 were found to be very the variation of the relative torsional energy as a function of similar at ca. 2 kJmol-1. This might have been expected since torsion angle h for rotations around the C(1)MC(2)* bond, all the modifications in this molecular progression occur at as shown in Fig. 6. Three energy minima corresponding to the the alkoxymethylene group and the group containing the chiral three minimum energy conformers of compound 5 were centre remains unchanged throughout the chemical observed. The computer generated models of these conforprogression. mations are shown in Fig. 7. The same study was conducted on the other materials where the bond containing the chiral centre was allowed to rotate in Discussion a stepwise manner with respect to the molecular core, and the Eqn. (5) was used to qualify and phenomenologically quantify resulting conformations were minimised at each step in the changes in the helix of chiral nematic phases of thermotropic rotation. Table 3 lists the energy diVerence DEc between the liquid crystals.However, pitch inversion in other mesophases two lowest minimum energy conformers thus generated for can be modelled in the same way. These phases might include materials 1–6, and from Fig. 6, it can be seen that DEc for 5 is chiral smectic C*, chiral lyotropic phases or even mixtures of chiral materials.It can be seen from Fig. 3 that eqn. (5) gives good theoretical fits to the experimental data with correlation coeYcients of the order of 0.99 for all materials studied. Furthermore, it can be seen from Table 2 that the theory predicts realistic values for T 2 and DE. In the first instance, the estimated temperatures at which inversions occur (1/p=0) agree with those determined by optical microscopy. In the molecular modelling studies, the rotation around the bond containing the chiral centre was not selected arbitrarily.Indeed, the emergence of spontaneous polarization in chiral smectic C* phases is a chiral property, and it was found that its magnitude depends very strongly on the dipole at the chiral centre and the amount of freedom that the chiral centre has to rotate.14 Likewise, helicity in liquid-crystalline phases is a chiral attribute and it is therefore reasonable to assume that the helical nature of a phase depends primarily on the chemical and geometrical environment around the chiral centre. This is confirmed by Kuball and co-workers, who showed that the Fig. 6 Representation of the torsion energies of 5 for the contribution of a ‘chiral area’, induced around an asymmetric OMC(1)MC(2)*MCl dihedral angle plotted as a function of the centre, to the helical twisting power (HTP)15 depends on its torsion angle h along with the Newman projections of the three orientation with respect to the director.16,17 Ultimately, each energetically preferred conformers (a), (b) and (c) conformer contributes diVerently to the HTP of a compound where a variation of the orientation of the ‘‘chiral region’’ could cause a change of sign of the HTP.In the preceding molecular simulations, the orientation of the ‘chiral region’ with respect to the director may be described as a function of the torsional angle for the rotation around the bond containing the chiral centre, i.e. the C(1)MC(2)* bond shown in Fig. 5 for compound 5. As pictured in Fig. 6, the energy diVerence DEc between the lowest energy conformer (c) and the next conformational minimum, that of conformer (a) (8.7 kJ mol-1), is of the same order of magnitude as the energy diVerence between the two hypothetical species described eqn. (5) (8.3 kJ mol-1). This suggests that the competing molecular species in the pure component system are likely to be conformational isomers related to the structure about the chiral moiety, and that changing the temperature will alter their relative concentrations.Thus, the higher energy conformations, although in Fig. 7 Computer generated models of the three minimum energy conformers of 5; left to right: conformers (a), (b) and (c) lower concentration, dictate the twist sense in the high tempera- 1966 J.Mater. Chem., 1998, 8(9), 1963–1969ture region of the chiral nematic phase. On the other hand, as assume that pa and pb are independent of temperature. On the other hand, it would be interesting to search for materials the temperature is lowered below the inversion point, the increased concentration of the lower energy conformations is where the conformations around the chiral atom have similar energies in view of providing, perhaps, constant pitch chiral such that the said conformations now dominate and dictate the helical twist sense of the phase.As indicated in the nematic phases. The third special case is where the species have degenerate description of the model, the two conformational groups are not static conformers but rather two groups weighted around energy levels and pa=-pb.Here, throughout the temperature range of the chiral nematic phase, and though the system may two potential energy minima and are undergoing dynamic interconversion. be entirely composed of chiral molecules, the macroscopic structure would be that of an achiral nematogen. This case is In this idealised case, conformations around the two lowest energy minima for rotations around the C(1)MC(2)* bond highly unlikely, however it does stress that the chiral character of a mesophase is not necessarily linked to the strength of were selected on the criterion that they are most likely to be the more densely populated conformational groups and there- chirality of a molecule.More so, it may depend on the constructive vs.destructive eVects of opposing chiral species, fore are most likely to have the greatest eVect on the mesophase structure. However, in a liquid-crystalline system at elevated whatever their nature. An interesting point to note is that, according to the model temperatures, species weighted around higher energy conformational minima may also have to be considered because the and with parameters outside the boundary conditions given in the special cases above, there will always be an inversion in conformers weighted around the lowest energy minima may not necessarily have opposing twists and will therefore not single chiral centre systems where the constituent species have opposite helical twists.Experimentally, however, it would often generate a temperature dependent twist inversion.This, in turn, may lead to diVerent values for the energy diVerence not be possible to observe an inversion point since it is likely to occur outside the thermal range of the chiral nematic phase between the minimum energy conformers of interest (see Fig. 6). This could be one of the many reasons why there seems to be as a hypothetical inversion point.Indeed, the model describes an infinite temperature range available for inversion to occur, a larger discrepancy in the energy diVerences for materials 1–4, where the energy diVerence between the two lowest minimum whereas experimentally, we are confined to small temperature ranges, at best of the order of 100 °C. energy conformations is ca. 3–4 times smaller than the energy diVerence between the two competing hypothetical species, as calculated from eqn. (5). Number of species involved in the inversion phenomenon Obviously, the fact that DEc and DE are of the same order Competition between species with opposing helical twists is of magnitude does not disprove the perception that the competnot limited to bi-component systems and experimental evidence ing species could be entities other than conformational isomers, indicates that many species compete, but it may happen that such as rotomers or molecular pairs.However, it does aYrm only a small number, and often only one, dominates. In fact, that conformational interconversion is a reasonable mechanism Vill et al. recently reported unusual changes in the chiral by which twist inversion in helical phases can be brought about.nematic helical pitch.18 A liquid-crystalline trioxadecalin derivative shows a temperature dependent inversion of the Special cases in the model chiral nematic helix at lower temperatures. At higher tempera- For all materials studied, the parameters pa and pb are shown tures, the pitch reaches a minimum then increases, tending to to have opposite signs.This is wholly expected as it is one of a second inversion point just above the clearing point. This the assumptions for the inversion model. On the other hand, infers that at least three species contribute to the overall twist, pa and pb are approximately an order of magnitude diVerent albeit in varying degrees. Using the same assumptions as for in their absolute values.One may infer that for a helical twist eqn. (5), it is very easy to extend the model to take into to occur at reasonably low temperatures in a pure system account a number N of species [eqn. (7)], where the contributing species follow a Boltzmann distribution, the latter must have opposite HTP’s of diVerent magnitude. Indeed, consider a system composed of two species with 1 p = .N n=1 CeEn/RT Pn D . N n=1 eEn/RT (7) opposite HTP, where pa=-pb. If the species are not degenerate (DE>0), then an inversion point will only be reached at infinite temperature. This is apparent from the notion that with a where En is the energy level of the nth species. As an example Boltzmann distribution, the two species will be equally popuof the flexibility of the model, a qualitative plot of a double lated only at infinite temperature.At this temperature, hypopitch inversion as a function of temperature is represented in thetically, one would observe a nematic phase. This is Fig. 8. Three contributing species a, b and c were chosen with confirmed by looking at eqn. (5) where, under these conditions, appropriate values for the parameters DEab=5 and DEac= inversion occurs if DE/RT=0 or T is infinite.At all other 10 kJ mol-1, and pa=0.1, pb=-0.2489 and pc=2.507 mm. The temperatures, a chiral nematic phase is observed. two inversion temperatures will therefore be T 21=350 and Let us now consider the case where competing species have T 22=400 K. Hence Vill’s observations can be explained degenerate energy levels, where DE=0.In this case, it is quite using eqn. (7). clear that a chiral nematic phase would result for pa<-pb, with the inverse pitch expressed in eqn. (6). Limitations of the model 1 p = 1 2 A1 pa + 1 pbB (6) The assumptions described by eqn. (5), even though viable within the phases under study, are nevertheless limited. They do not take account of phase transitions, i.e.it is assumed that The implication from eqn. (6) is that a system where all competing species have degenerate energy levels will have a the phase under study will exist at any temperature. To compensate for this, and therefore estimate the pitch divergence pitch in the cholesteric phase which is independent of temperature. Practically, this means that the smaller the energy diVer- near phase transitions, other terms are required, however this is outside the scope of this study.Also, the model described is ence between the species, the smaller the changes in pitch. This oversimplified view does not take into account such things as phenomenological and gives no insight into the microscopic origin of the phenomenon of twist inversion.Furthermore, it fluctuations near transitions and relies on the assumption that the phase will exist at all temperatures. Also, one is forced to is seen from Table 1 that the inversion of the helical twist sense J. Mater. Chem., 1998, 8(9), 1963–1969 1967Additionally, Saito et al. described inversions in spontaneous polarization in the chiral smectic C* phase as being driven by the competition between the dipolar and quadrupolar coupling of molecules,19 i.e.it was proposed that dipolar and quadrupolar coupling produce polarizations with opposite signs and that, at a critical temperature, the two can be exactly compensated and give rise to an inversion in the direction of the polarization. Therefore, it is not unreasonable to envisage that twist inversion in helical mesophases could also be explained by the competition between dipolar and quadrupolar ordering, though molecular correlations are much weaker in the chiral nematic phase than they are in the smectic phase where there is some positional order. This notion does not contradict the phenomenological model which is macroscopic and, moreover, changes in dipolar and quadrupolar coupling may be driven by changes in conformational distribution.Conclusions Fig. 8 Representation of the temperature dependence of pitch in a system where three species aVect the sense of the helix of the chiral There is much open debate as to the nature of inversion of nematic phase in such a way that a double inversion of the helical helical structures in liquid crystal systems.Unfortunately, the twist sense occurs (DEab=5 and DEac=10 kJ mol-1, pa=0.1, pb= debate is clouded by the fact that many systems studied are -0.2489 and pc=2.507 mm. Hence T 21=350 and T22=400 K) multi-component ones comprising a variety of materials or single component systems with multiple chiral centres. The study of single component, single chiral centre systems negates in the chiral nematic phase of material 4 is close to the SmA this problem and we are now allowed to examine what happens phase.Therefore, we might expect strong smectic fluctuations on the molecular scale more critically. Clearly, the mechanisms to adversely aVect the quality of the fit between the phenomenoinvolved in this process are diVerent from inversions in multi- logical model and the experimental pitch data.However, this component systems. The understanding of single component, was not found to be the case, implying that the model allows single chiral centre systems should however shed light on for good fits even when there are presumed substantial devimulti- component systems. At this present moment, no simple ations from an ideal system.This may therefore yield erroneous experimental techniques will allow us to measure the concen- values for the parameters pa, pb and DE. trations of conformers. Hence, the only way of rationalising Even though values for DE can be estimated to a certain this phenomenon is through predictive modelling. degree via computer simulations, unfortunately, it is diYcult Although this study gives no real insight into the mechanisms to use the model in a predictive fashion because the quantities of pitch inversion, the experimental results observed in a pa and pb cannot be estimated a priori.However, a range of variety of materials have been explained using a phenomeno- values for the quantities pa and pb can be approximated from logical model where twisting power is an additive property.the results reported in Table 2, e.g. pa#0.06 and pb#-0.6 mm. Moreover, through molecular modelling, it is hoped that This might then assist in determining whether inversion is information has been added to the debate in such a way that likely to occur within the temperature range of a chiral helical conformational interconversion is a viable mechanism by which phase.However, one ought to emphasise that, at the present helical inversion can be eVected. stage of computational or even experimental evaluation, the It is also hoped that the present article may spur researchers predictive nature of the model is rather limited. in the field to hunt for new materials with the view of providing, perhaps, chiral nematic phases whose pitch is largely indepen- Inversion phenomena in related systems dent of temperature.Describing the behaviour of a system by summing up its components’ properties is not a new concept. Component We gratefully acknowledge the EPSRC and the Defence property averaging was used to describe the inversion of the Research Agency (Malvern) for support of an EPSRC/CASE spontaneous polarization in chiral smectic C* phases3,5,10–11 studentship (M.J.W.).Finally, we would also like to thank Mr of single chiral centre systems. The model assumed that the A.T. Rendell for his assistance with photography. relative magnitude and direction of the dipole for all molecular species present and their rotational distribution about the References director have to be taken into account in the generation of the polarization. It was proposed that the species would fall into 1 J.W. Goodby, in Ferroelectric L iquid Crystals: Principles, two groups, one where the average dipole lies in a direction Properties and Applications, ed. G. W. Taylor, Gordon and Breach f (+n4) relative to the c-director and another where the average Science Publishers, Amsterdam, 1991. 2 P. Martino�t-Lagarde, R. Duke and G. Durand, Mol. Cryst. L iq. dipole lies in an opposing direction, f (-n4). In this way, the Cryst., 1981, 75, 249. two species are in competitn. Thus, for two such species a 3 J. S. Patel and J. W. Goodby, Philos.Mag. L ett., 1987, 55, 283. heuristic approach was used in order to describe the tempera- 4 P. Styring, J. D. Vuijk, I. Nishiyama, A. J. Slaney and ture dependence of the polarization.3,5,10 If the polarization J.W. Goodby, J. Mater. Chem., 1993, 3, 399. directions of species a and b are opposite, then the polarization 5 J. S. Patel and J. W. Goodby, J. Phys. Chem., 1987, 91, 5838. can fall to a zero value with decreasing temperature5,10 when 6 A. J. Slaney, I. Nishiyama, P. Styring and J. W. Goodby, J. Mater. Chem., 1992, 2, 805.xaPa=xbPb, where xa and xb are the temperature dependent 7 C. Loubser, P. L. Wessels, P. Styring and J. W. Goodby, J. Mater. concentrations of species a and b. Additionally, the energy Chem., 1994, 4, 71. diVerence between the two rival species, which can be deter- 8 K. Takatoh, M. J. Watson, P. Styring, M. Hird and J. W. Goodby, mined from fitting polarization data, was found to be similar in preparation. in magnitude to the torsional energy for the rotation around 9 I. Dierking, F. Gießelmann, P. Zugenmaier, W. Kuczynski, a CMC bond. This observation led the authors to suggest that S. T. Lagerwall and B. Stebler, L iq. Cryst., 1993, 13, 45. 10 J. W. Goodby and J. S. Patel, Opt. Eng., 1987, 26, 373. the competing species were likely to be conformational isomers. 1968 J. Mater. Chem., 1998, 8(9), 1963–196911 J. W. Goodby, E. Chin, J. M. Geary, J. S. Patel and P. L. Finn, 15 H. Finkelmann and H. Stegemeyer, Ber. Bunsenges. Phys. Chem., 1978, 82, 1302. J. Chem. Soc., Faraday T rans., 1987, 83, 3429. 12 H. Baessler and M. Labes, J. Chem. Phys., 1970, 52, 631; J. Adams 16 H. G. Kuball, H. Bruning, T. Muller, O. Turk and A. Schonhofer, J.Mater. Chem., 1995, 5, 2167. and W. Haas, Mol. Cryst. L iq. Cryst., 1971, 15, 27; In the case of cholesterol derivatives, it was found that pitch determination is 17 H. G. Kuball, T. Muller, H. Brunning and A. Schonhofer, Mol. Cryst. L iq. Cryst., 1995, 261, 205. more linear when analysed according to weight rather than molar concentration. 18 V. Vill, H. W. Tunger, K. Hensen, H. Stegemeyer and K. Diekmann, L iq. Cryst., 1996, 20, 449. 13 R. Cano and P. Chatelain, C.R. Acad. Sci., 1964, B259, 252; J. Adams, W. Haas and J.Wysocki, in L iquid Crystals and Ordered 19 S. Saito, K. Murashiro, M. Kikuchi, T. Inukai, D. Demus, M. Neundorf and S. Diele, Ferroelectrics, 1993, 147, 367. Fluids, Plenum press, New York, 1970, p. 463. 14 J. W. Goodby, J. S. Patel and E. Chin, J. Phys. Chem., 1987, 91, 5151. Paper 8/04413I; Received 10th June, 1998 J. Mater. Chem., 1998, 8(9), 1963–1969 1969

ISSN:0959-9428    

DOI:10.1039/a804413i

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 Disk-like liquid crystals of transition metal complexes Part 20.‡—Pursuit of chemistry to directly visualize van derWaals interactions Kazuchika Ohta,*† Mayumi Ikejima, Mitsuo Moriya, Hiroshi Hasebe and Iwao Yamamoto Department of Functional Polymer Science, Faculty of T extile Science and T echnology, Shinshu University, 386 Ueda, Japan This work reports on mesomorphic thermochromism caused by metal centres leading to visualization of van derWaals interactions.We succeeded in obtaining columnar liquid crystals exhibiting very interesting thermochromism by introduction of long chains into disk-like bis(glyoximato) d8 metal(II ) complexes.These long-chain-substituted NiII, PdII and PtII complexes, (CnO)8-M, containing one-dimensional metal chain change color according to the metal–metal stacking distance. With increasing temperature, the metal–metal distance increases leading to an enlargement of the band gap between the filled ndz2 valence band and the empty (n+1)pz conduction band of the central metal.Generally speaking, when mesomorphic materials are heated, the peripheral alkyl chains melt, whereas the central aromatic part remains rigid.Adjacent peripheral long chains tend to lie within the van derWaals radius at lower temperatures, which results in a shortening of the distance between the central d8 metals in the present (CnO)8-M systems. This ‘fastener eVect’ of the van derWaals interaction of peripheral long chains is weakened with increasing temperature, leading to visible color changes due to the ndz2–(n+1)pz interaction in the central metal chain.The ‘apparent pressures’ of the long chains in the (CnO)8-M complexes have been estimated from the shift coeYcient of the d–p band vs. pressure for the corresponding non-substituted metal complexes. 1 Introduction Is it possible to directly visualize van der Waals interactions? It is well known that the main intermolecular force in molecular condensed phases is van der Waals interaction.If a molecular solid is gradually heated, the condensation force by this van der Waals interaction gradually becomes weaker. It is especially interesting that this force weakening causes the phase changes as solid(crystal)Aliquid crystalAliquid. These changes are attributable to the molecular structure of liquid crystalline materials.Many liquid crystals are, generally, composed of aromatic groups at the molecular center and long alkyl chains at the surroundings. On gradually heating liquid crystalline materials, the peripheral alkyl chains begin to melt, whereas the central aromatic part remains essentially rigid. This is the reason N R O H R R H O N R N R R R M O O R N 4 : (C12)8-Ni 1 : (C nO)8-Ni (ref. 2,3) R = C12H25 This work ( n = 4–14) M=Pt ( n = 1–12) M=Pd ( n = 4, 8, 12) M=Ni M=Ni R = OC nH2 n+1 R = OC nH2 n+1 R = OC nH2 n+1 2 : (C nO)8-Pd (ref. 4) 3 : (C nO)8-Pt why liquid crystalline phases appear before completely melting Fig. 1 Structural formula of long chain-substituted bis(diphenylglyoxi- into the liquid phase. In this context, can one directly visualize mato)metal(II ) complexes van der Waals interactions of alkyl chains changing with temperature? This is our starting viewpoint for the present work.Recently, we investigated a novel phenomenon of ‘mesomorphic thermochromism due to metal centres’.2–4 This alkoxy groups into core bis(diphenylglyoximato)metal(II ) moieties which exhibit piezochromism as reported by Shirotani phenomenon is largely concerned with an attempt to directly visualize van der Waals interactions.Recently, we succeeded et al.5–9 They reported that the non-substituted bis(diphenylglyoximato) platinum(II ) complex is red–brown under atmos- in obtaining discotic columnar liquid crystals exhibiting very interesting thermochromism by introduction of long chains pheric pressure but green under 0.69 GPa.Suprisingly, the present alkoxy-chain-substituted bis(diphenylglyoximato)plat- into a disk-like molecule, bis(diphenylglyoximato)M (M=Ni, Pd or Pt). We found that these long-chain-substituted NiII, inum(II ) complexes already show a green color even under atmospheric pressure as if 0.69 GPa of external high pressure PdII and PtII complexes having a one-dimensional columnar structure show temperature dependent color changes, i.e., were applied.This ‘apparent pressure’ is attributed to an eVect of self-pressing by van der Waals interaction of the peripheral thermochromism. With increasing temperature, the metal–metal distance increase leads to an enlargement of band gap chains, i.e., a ‘fastener eVect’.10–15 The neighboring peripheral alkyl chains tend lie within the van der Waals radius at lower between the filled ndz2 valence band and the empty (n+1)pz conduction band of the metal.This results in a blue-shift of temperatures, which results in shortening the distance between the central d8 metals in the one-dimensional columnar struc- the electronic transition spectrum which is directly connected with d–p interactions between the upper and lower metals in ture. The fastener eVect of the van der Waals interaction is weakened with increasing temperature, which leads to a color the one-dimensional metal chains.The present d8-metal complexes (Fig. 1) were synthesized by introduction of eight long change due to an alteration of the d–p interaction in the metal chain. Hence, this color change can be directly visualized.For example, the present alkoxy-chain-substituted bis(diphenylglyoximato) platinum(II ) complexes are green at r.t., and turn †E-mail: ko52517@giptc.shinshu-u.ac.jp ‡Part 19: ref. 1. red in the mesophase with increasing temperature. J. Mater. Chem., 1998, 8(9), 1971–1977 1971hydroxylamine hydrochloride (12.9 g, 185 mmol) and 85% 2 Experimental potassium hydroxide (12.9 g, 185 mmol) were added and the 2.1 Synthesis mixture was stirred vigorously for ca. 1 h and filtered to remove precipitates. 3,3¾,4,4¾-Tetradecyloxybenzil 5 (3.00 g, The synthetic routes for (CnO)8-Pt (n=4–14) 3 and (C12)8-Ni 2.83 mmol) was added to the filtrate and under a nitrogen 4 are shown in Scheme 1. Detailed procedures of the prepatmosphere the mixture was refluxed with stirring for 12.5 h.aration of precursors 5 and 6 have described previously.16 The To the hot reaction mixture was added potassium tetrachloro- (CnO)8-Pt (n=4–14) and (C12)8-Ni complexes were prepared platinate(II ) (0.64 g, 1.54 mmol) dissolved in a small amount by the reaction of the corresponding a-diketones, 5 and 6, with of ethane-1,2-diol.Immediately the solution was neutralized hydroxylamine hydrochloride in the reaction solvent, followed with glacial acetic acid as confirmed by pH test paper. After by addition of an ethane-1,2-diol or propane-1,3-diol solution neutralization, reflux was continued for more than 6 h during of the metal salt and then neutralization with glacial acetic which dark green precipitates gradually appeared. After reflux, acid.Detailed procedures are presented only for the representathe hot reaction mixture was filtered and the dark green tive (C14O)8-Pt complex. Table 1 summarizes reaction solvents, precipitate collected on filter paper. The product was dissolved solvents for the metal salt, elemental analysis data, yields, and in chloroform and the solvent evaporated to give dark green colors of the complexes.liquid crystals. Purification was carried out by column chromatography [silica gel, chloroform–benzene (151 v/v), Rf=0.70] Bis[1,2-bis(3,4-di-n-tetradecyloxyphenyl )ethanedione and reprecipitation carried out by adding acetone to a hot dioximato]platinum(II) 3, (C14O)8-Pt. To 313 ml ethanol, solution of the product in chloroform to give dark green liquid crystals (0.37 g, yield: 10%). 2.2 Measurements The products were characterized by elemental analyses using a Perkin Elmer 240B Elemental Analyzer. The phase transition behavior of the compounds was observed by a polarizing microscope equipped with a heating plate controlled by a thermoregulator (Mettler FP80 and FP82), and heat changes measured with a Rigaku Thermoflex DSC-10A or a Shimadzu DSC-50 diVerential scanning calorimeter.To establish the mesophases, powder X-ray patterns were measured with Cu- Ka radiation using a Rigaku Geigerflex instrument equipped with a hand-made heating plate controlled by a thermoregulator. 17 Temperature-dependent electronic spectra were recorded with a Hitachi 330 spectrophotometer equipped with a hand-made heating plate controlled by a thermoregulator (CHINO DB-125).17 For thermochromism measurements, thin films of the Ni and Pt complexes were prepared by casting from a solution of the products in chloroform on glass.Glass Techno, P-102 glass was used, since it is transparent in the wavelength range 250–2100 nm. Measurements were carried out in air as reference. 3 Results and Discussion 3.1 Mesomorphic properties We have already reported that each of the (CnO)8-Ni 1 and -Pd 2 complexes (Fig. 1) has a discotic hexagonal disordered columnar mesophase (Colhd), and that the mesomorphic structure has a linear chain of metals which originates the ndz2–(n+1)pz transition of the metal ion.2,4 For the present complexes, (CnO)8-Pt (n=4–14) 3 and (C12)8-Ni 4, these mesophases are identified by DSC measurements, polarizing microscopy observations and temperature-dependent X-ray diVraction measurements. 3.1.1 Mesomorphic properties of the (CnO)8-Pt (n=4–14) complexes, 3. As is shown in Table 2, each of the (CnO)8-Pt (n=4–14) complexes shows an enantiotropic columnar mesophase. The (CnO)8-Pt complexes for n7 are in a viscous liquid crystalline state at r.t., whereas the (C4O)8-Pt complex is in a supercooled mesomorphic state and the (CnO)8-Pt (n= 5, 6) complexes are in a crystalline state at r.t.A fan-shaped natural texture was observed on cooling for each of the (CnO)8- Pt complexes for n6. Generally this texture is characteristic of a Colhd mesophase. X-Ray diVraction of each of the (CnO)8- Pt complexes for n=4–14 gave spacings in the ratios, R R R O O N O H N R R H O N R N R Ni O O R R R R R N RO O H RO OR RO H O N OR N RO OR OR Pt O O OR RO RO OR O O N 4 : (C12)8-Ni 6 1)NH2OH•HCl/KOH 2)NiCl2•6H2O/EG R = C12H25 3 : (C nO)8-Pt R = C nH2 n+1 ( n = 4–14) 2)K2PtCl4/EG or PG 5 1)NH2OH•HCl/KOH 151/Ó351/251/Ó75,, which correspond to a two-dimensional Scheme 1 Synthetic routes to long chain-substituted bis(diphenylhexagonal lattice, and a diVuse band at 2h#25o corresponding glyoximato)metal(II) complexes; EG=ethane-1,2-diol, PG=propane- 1,3-diol to fluctuation of the one-dimensional molecular stack.Hence, 1972 J. Mater. Chem., 1998, 8(9), 1971–1977Table 1 Reaction solvents, solvents for the metal salt, elemental analysis data, yields, colors and pristine states of the (CnO)8-Pt (n=4–14) complexes, 3, and (C12)8-Ni complexes, 4 elemental analysis (%) found (calc.) reaction solvent for yielda pristine n solvent (v/v) the metal salt C H N (%) color stateb 4 MeOH–EtOH(151) EG 57.38(57.63) 7.00(6.93) 4.39(4.48) 4 dark green supercooled Colhd 5 EtOH EG 59.97(59.94) 7.53(7.55) 4.00(4.11) 21 dark green K1 6 EtOH EG 62.21(61.89) 8.08(8.80) 3.68(3.80) 7 dark green K 7 EtOH EG 63.91(63.57) 8.47(8.51) 3.48(3.53) 9 dark green Colhd 8 MeOH–EtOH(151) EG 65.31(65.03) 9.02(8.90) 3.18(3.30) 8 dark green Colhd 9 EtOH EG 66.38(66.30) 9.19(9.24) 3.01(3.09) 8 dark green Colhd 10 EtOH EG 67.53(67.43) 9.54(9.54) 2.79(2.91) 8 dark green Colhd 11 EtOH EG 68.72(68.43) 9.85(9.80) 2.68(2.72) 4 dark green Colhd 12 EtOH PG 69.20(69.33) 10.05(10.04) 2.56(2.61) 8 dark green Colhd 13 EtOH EG 70.34(70.14) 10.18(10.26) 2.41(2.48) 9 dark green Colhd 14 EtOH EG 70.56(70.87) 10.49(10.45) 2.28(2.36) 10 dark green Colhd (C12)8-Ni EtOH EG 79.05(77.78) 11.45(11.21) 2.97(2.64) 6 red Colrh aRecrystallization solvent: CHCl3–acetone.bK=crystal, Colhd=hexagonal disordered columnar mesophase, Colro=rectangular ordered columnar mesophase. Table 2 Phase transition temperatures (T/°C) and enthalpy changes (DH/kJ mol-1 in parentheses) for the (CnO)8-Pt (n=4–14) complexes, 3a n phase transitions supercooled Colhd 4 @ K1 CCA 206.9 (3.07) K2 CCA 272.5 (4.42) Colhd CCA 312.5 IL (Decomp.) 5 K1 CCA 219.7 (1.01) K2 CCA 252.9 (1.99) Colhd CCA 303.2 IL (Decomp.) 6 Kcbe 224.4 (1.25) Colhd cbe 285.4 (20.0) IL Fig. 2 Phase transition temperatures vs.the number of carbon atoms (n) in the alkoxy chains of the (CnO)8-Pt (n=4–14) complexes, 3 7 Colhd cbe 281.5 (20.1) IL Table 3 Comparison of phase transition temperatures (°C) enthalpy changes (DH/kJ mol-1 in parentheses) and X-ray data of the 8 Colhd cbe 268.8 (17.8) IL (C12O)8-M (M=Ni, Pd, Pt) and (C12)8-Ni complexes 1–4 complex phase transitions 9 Colhd cbe 263.3 (18.2) IL 10 Colhd cbe 255.5 (17.7) IL (C12O)8-Ni 1a Colhd cbbbe 211(36.9) IL (a=34.4 A ° at r.t.) 11 Colhd cbe 247.3 (16.2) IL (C12O)8-Pd 2b Colhd cbbbe 229(23.5) IL (a=35.4 A ° at r.t.) 12 Colhd cbe 239.3 (16.4) IL (C12O)8-Pt 3 Colhd cbbbe 239.3(16.4) IL (a=34.2 A ° at 150°C) 13 Colhd cbe 232.8 (15.9) IL (C12)8-Ni 4 Colro(P2/a) cbbbe 32.1(46.6) (a=61.7, b=46.6, h=3.37 A° at 22 °C) Colhd cbbbe 82.3(23.1) IL (a=33.7 A ° at 60 °C) 14 Colhd cbe 227.8 (15.6) IL aRef. 2 and 3. bRef. 4. aK=crystal, Colhd=hexagonal disordered columnar mesophase, IL= isotropic liquid, Decomp.=decomposition,@=relaxation. order of (C12O)8-Ni<(C12O)8-Pd<(C12O)8-Pt, so that the temperature region of the mesophase also becomes larger in each of these mesophases can be assigned to a Colhd mesophase, as well as those of the (CnO)8-Ni and Pd complexes.The this order. assignment is also consistent with the fan-shaped texture. Phase transition temperatures of the (CnO)8-Pt complexes for 3.1.2 EVects of long chains on the liquid crystalline (DPG)2Ni derivatives, 1 and 4. We have already reported the influence of n=4–14 are plotted vs. the carbon number (n) in the alkoxy chain in Fig. 2. This figure clearly shows that the melting point long chains on substituted b-diketonato copper complexes18 and dithiolene nickel complexes.16,19 Interestingly, these com- of the Pt complexes becomes lower upon increasing the chain length and the temperature region of the mesophase conse- plexes have mesomorphic properties when the chain is an octyloxy group, whereas they do not show mesophases when quently becomes larger.As Table 3 shows, the clearing points of (C12O)8-M complexes for M=Ni, Pd, Pt are higher in the the chain is an octyl group. Hence, we wished to investigate J. Mater. Chem., 1998, 8(9), 1971–1977 1973the influence of diVerent long-chain-substituents for the core bis(diphenylglyoximato)metal(II) derivatives, (DPG)2M, on their mesomorphic properties.We investigated the Ni derivatives, since the syntheses of the dodecyl-substituted Pd and Pt derivatives were not successful whereas the dodecyl-substituted Ni derivative was obtained (Fig. 1). It has been already reported in a previous paper17 that (C12O)8-Ni 1, shows a Colhd mesophase for which the lattice constant a is 34.4 A ° at r.t. Table 3 summarizes the phase transition temperatures, enthalpy changes and X-ray data of (C12O)8-Ni 1 and (C12)8- Ni 4.X-Ray diVraction of the lower temperature mesophase of (C12)8-Ni 4, at 22°C gave reflections from a two-dimensional rectangular lattice with a=61.7 A ° and b=46.6 A ° . Since Z could be estimated to be ca. 4 in the unit cell and since the Miller indices (0k) were absent when k was odd, it could be concluded that the rectangular lattice has P2/a symmetry.In addition, a reflection appeared at 2h#25° which corresponds to a one-dimensional stack of the molecules with an interdisk separation h of 3.37 A° . Hence, the lower temperature mesophase could be identified as a rectangular ordered columnar mesophase, Colro (P2/a). X-Ray diVraction of the higher temperature mesophase at 60 °C gave a diVuse band at 2h#25° corresponding to fluctuation of the molecular stack and reflections from a two-dimensional hexagonal lattice having lattice constant a=33.7 A ° .Therefore, the higher temperature mesophase in the region from 31.2–82.3 °C could be identified as a Colhd mesophase, as for (C12O)8-Ni. A fan-shaped texture was observed. A schematic representation of the mesomorphic structural change from the Colro mesophase to the Colhd mesophase is shown in Fig. 3. The molecular planes tilt to the Fig. 4 Photomicrographs of the (C12O)8-Pt complex at various rectangular plane in the lower temperature mesophase. On the temperatures: (a) r.t., (b) 200 °C, (c) 235 °C and (d) 238 °C other hand, the molecules begin to rotate by thermal movement in the higher temperature mesophase region, so that they are packed in a hexagonal lattice.As can be seen in Fig. 3, the discotic mesophase was observed as yellow and fan like when two-dimensional lattice expands to 67.4/61.7=1.09 times in the yellow isotropic liquid(IL) at >239.3 °C was cooled to the a-axis direction and 58.4/46.6=1.25 times in the b-axis 238 °C [Fig. 4(d)]. Shirotani et al.8 reported that the direction. 5dz2–6pz transition band of Pt2+ which is consistent with a one-dimensional platinum stack is located at 18 200 cm-1 at 3.2 ‘Visible fastener eVect’ on mesomorphic thermochromism atmospheric pressure, and that the band shifts toward lower 3.2.1 ‘Apparent pressure’ by surrounding long chians. We energy by -3000 cm-1 GPa-1. By using this shift coeYcient have already reported that the alkoxy-chain-substituted bis(di- the ‘apparent pressure’ on the present long-chain-substituted phenylglyoximato)nickel(II)2,3 and palladium(II )4 complexes, Pt complex could be estimated from eqn.(1) (CnO)8-Ni 1 and (CnO)8-Pd 2 (Fig. 1), exhibit color changes with increasing temperature. The Ni and Pd complexes gradu- Pressure GPa= (n�obs-18 200) cm-1 -3000 cm-1 GPa-1 (1) ally turn from red to yellow and from orange to yellow, respectively.On the other hand, the present alkoxy chainwhere n� obs is the observed wavenumber of the d–p band. The substituted bis(diphenylglyoximato)platinum(II ) complexes, d–p band of the (C12O)8-Pt complex is located at 15 600cm-1 (CnO)8-Pt 3, synthesized here exhibit sharp color changes from (641 nm) at 35 °C so the ‘apparent pressure’ can be estimated green at r.t.to red, orange, and yellow with increasing temperato be 0.87 GPa at atmospheric pressure. ture, as is shown in Fig. 4 and 5. The natural texture of the 3.2.2 Influences of metals on the ‘fastener eVect’. Fig. 5(a)–(c) show electronic spectra at various temperatures for the (C12O)8- Ni, Pd and Pt complexes, respectively.Comparison with the detailed assignment of electronic spectra for the (C12O)8-Pd complex in a previous paper,4 the A and B bands can be assigned to the ndz2–(n+1)pz transition of M2+ and to a metal to ligand charge transfer transition (MLCT), respectively; band C has not been assigned.4 As can be seen from this figure, the A band of the (C12O)8-Pt complex at 641 nm at 35 °C shifts to higher energy ( blue-shift) with increasing temperature more than the A bands of the Ni and Pd complexes (located at 550 and 467 nm at 35 °C, respectively).This blue-shift can be reasonably explained by an increase of the intermolecular distance within the columns. The d–p band wavenumbers of each of the metal complexes are plotted vs. temperature in Fig. 6(a). A linear relationship between the d–p band wavenumber and temperature is observed. Table 4 lists the rates of Fig. 3 Schematic representation of the mesomorphic structural change from Colro to Colhd for the (C12)8-Ni complex, 4 blue-shift as determined by least-squares. It is clear that the 1974 J. Mater. Chem., 1998, 8(9), 1971–1977Fig. 6 (a) Wavenumbers of the d–p bands and (b) apparent pressures at various temperatures for the (C12O)8-M (M=Ni, Pd, Pt) complexes Table 4 Rates of blue-shift of the d–p bands for the (C12O)8-M (M= Ni, Pd, Pt) complexes with increasing temperature M rate of blue-shift (cm-1 °C-1) Ni 6.07 Pd 10.1 Fig. 5 Electronic absorption spectra of a thin film of the (C12O)8-M Pt 15.6 complexes at various temperatures: (a) M=Ni, ( b) Pd, and (c) Pt.According to the assignment in the literature,19 A band= ndz2–(n+1)pz transition(B1u) and B band=metal-to-ligand charge Table 5 Shifts of pressure and values of apparent pressure at 35 °C for transfer transition (MLCT5B2u or B3u). The C band has not been the (C12O)8-M (M=Ni, Pd, Pt) complexes reported previously. pressure/GPa M Xa/ cm-1 Y a/cm-1 GPa-1 at 35 °C Pt complex exhibits the largest blue-shift-rate, the rates being in the order (C12O)8-Ni<(C12O)8-Pd<(C12O)8-Pt. By using Ni 19610 -1300 0.83 these shift-rates and the wavenumber of the d–p bands at Pd 22940 -1940 0.79 various temperature, ‘apparent pressures’ were calculated from Pt 18200 -3000 0.87 eqn.(2) in the same manner as for eqn. (1) aRef. 8. Pressure GPa= n� obs-X Y (2) molecules at the clearing point.The expansion of the interdisk distance with increasing temperature corresponds to an expan- where X cm-1 and Y cm-1 GPa-1 are the wavenumber of the d–p transition band under atmospheric pressure and the blue- sion of metal–metal distance in the one-dimensional metal chain which is the origin of the blue-shift of the d–p band shift-rate of the non-substituted core (DPG)2M complex, respectively. X and Y values and the ‘apparent pressures’ of leading to thermochromism. Thus, measurements of the temperature- dependent electronic spectra enable us to reasonably (C12O)8-Ni, Pd and Pt complexes at 35 °C are summarized in Table 5.As can be seen from Fig. 6( b), all complexes have a elucidate the relationship between the ‘fastener eVect’ and the mechanism of thermochromism.similar linear relationship between temperature and apparent pressure with slopes of ca. -5×10-3 GPa °C-1. This means The diVerent rates of blue-shift (Table 4) can be explained by the diVerent size of the metal ions. Shirotani et al.8 reported that the ‘apparent pressure’ on the central core complex imparted by the fastener eVect of the surrounding long chains the rates of red-shift of the d–p bands vs.pressure for bis(1,2-cyclohexanedione dioximato)metal(II) complexes is released by ca. -5×10-3 GPa °C-1 with increasing temperature, irrespective of the central metal. [(NIOX)2M], bis(dimethylglyoximato)metal(II ) complexes [(DMG)2M] and bis(diphenylglyoximato)metal(II ) complexes Thus, under atmospheric pressure long-chain-substituted bis(diphenylglyoximato)metal(II) complexes are self-pressed by [(DPG)2M]; the rates of red-shift increase in the order (NIOX)2M<(DMG)2M<(DPG)2M for a given central metal.van derWaals interaction between the surrounding long chains, as if they were subject to an external high pressure. With This means that the (DPG)2M complex having the bulkiest ligand shows the largest rate of red-shift; on the other hand, increasing temperature the apparent pressure is gradually weakened and the interdisk distance within the column extends rates are in the order of Ni<Pd<Pt complex for a given ligand.Drickamer et al. also reported a similar order for the until finally the columnar structure disintegrates into discrete J. Mater. Chem., 1998, 8(9), 1971–1977 1975metals.20,21 Since the radii are in the order of Ni2+>Pd2+>Pt2+, platinum has the smallest ion radius among these metals.When the complex has both the bulkiest ligand (DPG) and the smallest metal ion (Pt2+), it has the loosest metal–metal distance in the one-dimensional metal chain to originate the biggest rate of red-shift of the d–p band with pressure.8 The present Colhd mesophases of the long chain- substituted (C12O)8-M complexes have a columnar structure containing one-dimensional metal–metal stacks.The (C12O)8-Pt complex has the loosest metal–metal stack among the (C12O)8-M complexes (M=Ni, Pd, Pt). Such a structure leads to the largest blue-shift of the d–p band with temperature, leading to the most drastic ‘mesomorphic thermochromism’.Hence, it can beuded that the compressibility of the one-dimensional metal–metal stacking distance and the resulting delocalization of d-electrons depends on the bulkiness Fig. 8 Electronic absorption spectra of a thin film of the (C12)8-Ni complex at various temperatures: (a) M=Ni, (b) Pd, and (c) Pt. of the ligands and the size of the metal ions, and that they can According to the assignment in the literature,19 A band= be sensitively varied by the apparent pressure of the fastener ndz2–(n+1)pz transition (B1u) and B band=metal-to-ligand charge eVect of the surrounding long chains.transfer transition (MLCT5B2u or B3u). 3.2.3 Influence of the length of long chains on fastener eVect. Fig. 8 shows electronic spectra of a thin film of the (C12)8- As mentioned above, the nature of the metal crucially influences Ni complex at various temperatures.The A and B bands of the thermochromism of the alkoxy-chain-substituted comthis (C12)8-Ni complex show diVerent behavior from those of plexes, (CnO)8-M 1–3. Fig. 7 shows the variation of the wavethe (C12O)8-M complexes for M=Ni, Pd and Pt, (Fig. 5). The number of the d–p bands and the ‘apparent pressures’ at r.t.A band of the (C12)8-Ni complex shows a blue-shift with plotted vs. the number (n) of carbon atoms in the alkoxy chains increasing temperature, as does the (C12O)8-Ni complex, how- of the (CnO)8-Pt complexes, respectively. As can be seen in ever its intensity increases with increasing temperature in direct Fig. 7, the d–p band shows a discontinuity between n=5 and contrast to the (C12O)8-Ni complex.The intensity of the B 6; the reason for this behaviour is not clear at the present time. band of the (C12)8-Ni complex increases with increasing tem- As can also be seen in Fig. 7, it seems that ‘apparent pressure’ perature, similarly to the (C12O)8-Ni complex, but shows no via the ‘fastener eVect’ attains a plateau at ca. 0.9 GPa at blue-shift in contrast to the (C12O)8-Ni complex. Irrespective n12. In other words, the ‘fastener eVect’ may not work of these diVerent behaviors, the A and B bands of the (C12)8- eVectively for highly extended alkoxy chains. Ni complex can be assigned to the 3dz2–4pz transition Ni2+ and the metal–ligand charge transfer (MLCT) band, respect- 3.2.4 Influence of the nature of long chains on the fastener eVect.We next investigated the influence of the nature of the surrounding long chains on the ‘fastener eVect’ and thermochromism; the dodecyloxy chains were replaced by dodecylchains to give the (C12)8-Ni complexes. Fig. 9 (a) Wavenumbers of the d–p bands and (b) apparent pressures Fig. 7 (a) Wavenumbers of the d–p bands and (b) apparent pressures vs.the number of carbon atoms (n) in the alkoxy chains of the (CnO)8- at various temperares for Colhd mesophases of the (C12O)8-Ni and (C12)8-Ni (n=4–14) complexes, 1 and 4 (temperature 35 °C) Pt (n=4–14) complexes, 3, at r.t. 1976 J. Mater. Chem., 1998, 8(9), 1971–1977ively, since their band positions are approximately the same van der Waals interactions visually.This work may also represent a new frontier in applications of liquid crystalline as those of the (C12O)8-Ni complex. These assignments are also supported by the disappearance of the A band in a organic metal complexes (metallomesogens). In addition, we can expect to obtain high electronic conductivities by such a solution of the (C12)8-Ni complex in chloroform, similarly to the (C12O)8-Ni complex.2,3 Fig. 9 shows the d–p bands and fastener eVect under atmospheric pressure even for the compounds exhibiting high conductivities only under high pressure. the ‘apparent pressures’ at r.t. vs. temperatures, respectively, for the (C12O)8-Ni and (C12)8-Ni complexes. The ‘apparent pressure’ decreases with increasing temperature at a rate of ca. References -1.1×10-2 GPa °C-1 for the (C12)8-Ni complex.This rate is about twice that of (C12O)8-Ni (ca. -5×10-3 GPa °C-1). 1 K. Ohta, S. Azumane, T.Watanabe, S. Tsukada and I. Yamamoto, Appl. Organomet. Chem., 1996, 10, 623. The diVerence of the rates between these complexes may be 2 K. Ohta, H. Hasebe, M. Moriya, T. Fujimoto and I. Yamamoto, due to the presence of oxygen atoms, since the dodecyloxy Mol.Cryst. L iq. Cryst., 1991, 208, 43. group is more electron-donating than the dodecyl group, the 3 K. Ohta, H. Hasebe, M. Moriya, T. Fujimoto and I. Yamamoto, core complex part of the (C12O)8-Ni complex becomes very J.Mater. Chem., 1991, 1, 831. electron rich so that the intermolecular interactions of the core 4 K. Ohta, M. Moriya, M. Ikejima, H. Hasebe, T. Fujimoto and I.Yamamoto, Bull. Chem. Soc. Jpn., 1993, 66, 3553, 3559. (C12O)8-Ni complex are stronger than that of (C12)8-Ni. Hence, 5 I. Shirotani, K. Suzuki, T. Suzuki, T. Yagi and M. Tanaka, Bull. the (C12O)8-Ni molecules more strongly adhere with each other Chem. Soc. Jpn., 1992, 65, 1078. at the core complex part than do (C12)8-Ni molecules. When 6 I. Shirotani, Petrotech, 1993, 16, 781.they are heated, the ‘apparent pressure’ of the surrounding 7 I. Shirotani, Gendai Kagaku, 1987, 42, frontispiece, 30. long chains is reduced so that their core complex parts should 8 I. Shirotani, Y. Inagaki, W. Utsumi and T. Yagi, J. Mater. Chem., be released from the ‘fastener eVect’. However, the strong 1991, 1, 1041. 9 I. Shirotani, K. Suzuki and T. Yagi, Proc. Jpn. Acad., Ser.B, 1992, interaction between the core complex parts of the (C12O)8-Ni 68, 57. complex may persist at higher temperatures compared with 10 H. Inokuchi, G. Saito, P. Wu, K. Seki, T. B. Tang, T. Mori, the (C12)8-Ni complex. K. Imaeda, T. Enoki, Y. Higuchi, K. Inaka and N. Yasuoka, Chem. L ett., 1986,1263. 11 H. Inokuchi, K. Imaeda, T. Enoki, T. Mori, Y. Maruyama, 4 Conclusion G.Saito, N. Okada, H. Yamochi, K. Seki, Y. Higuchi and N. Yasuoka, Nature (L ondon), 1987, 329, 39. We have described the influence of central metals and the 12 Z. Shi, T. Enoki, K. Imaeda, K. Seki, P. Wu, H. Inokuchi and surrounding long chains on thermochromism of long-chain- G. Saito, J. Phys. Chem., 1988, 92, 5044. substituted bis(diphenylglyoximato)metal(II) complexes. Novel 13 P.Wang, T. Enoki, K. Imaeda, N. Iwasawa, H. Yamochi, complexes (CnO)8-Pt (n=4–14) and (C12)8-Ni were prepared. H. Urayama, G. Saito and H. Inokuchi, J. Phys. Chem., 1989, It was revealed that the long-chain-substituted bis(diphenyl- 93, 5947. glyoximato)metal(II ) complexes exhibit not only mesomorphic 14 K. Imaeda, T. Mitani, C. Nakano, H. Inokuchi and G. Saito, Chem. Phys. L ett., 1990, 173, 298. properties but also thermochromism caused by the ‘fastener 15 J. K. Jeszuka, T. Enoki, Z. Shi, K. Imaeda, H. Inokuchi, eVect’ of the van der Waals interaction of the peripheral long N. Iwasawa, H. Yamochi and G. Saito, Mol. Cryst. L iq. Cryst., chains. The length and nature of the long chains influence not 1991, 196, 167. only on the mesomorphic properties but also the thermochro- 16 K. Ohta, H. Hasebe, H. Ema, M. Moriya, T. Fujimoto and mism. The one-dimensional metal stacking structure is central I. Yamanoto, Mol. Cryst. L iq. Cryst., 1991, 208, 21. to this thermochromism. The value of ‘apparent pressure’ was 17 H. Ema, Master Thesis, Shinshu University, Ueda, 1988, ch. 7; H. Hasebe, Master Thesis, Shinshu University, Ueda, 1991, ch. 5. estimated to be 0.87 GPa for the (C12O)8-Pt complex at 35 °C. 18 K. Ohta, H. Ema, H. Muroki, I. Yamamoto and K. Matsuzaki, The apparent pressure by the surrounding long chains reaches Mol. Cryst. L iq. Cryst., 1987, 147, 61. a saturated plateau around 0.9 GPa for n12 for the (CnO)8- 19 K. Ohta, H. Hasebe, M. Moriya, T. Fujimoto and I. Yamamoto, Pt complexes. The (C12O)8-Pt complex which has the bulkiest Mol. Cryst. L iq. Cryst., 1991, 208, 33. ligand and the smallest metal ion exhibits the most drastic 20 J. C. Zahner and H. G. Drickamer, J. Chem. Phys., 1960, 33, 1625. thermochromism among the (C12O)8-M complexes for M= 21 M. Tkacz and H. G. Drickamer, J. Chem. Phys., 1986, 85, 1184. Ni, Pd, Pt. Such a visible fastener eVect so far as we know is unprecedented. The present thermochromism enables to see Paper 8/00894IK; Received 2nd February, 1998 J. Mater. Chem., 1998, 8(9), 1971–1977 1977

ISSN:0959-9428    

DOI:10.1039/a800894i

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 Disk-like liquid crystals of transition metal complexes Part 21.‡—Critical molecular structure change from columnar to lamellar liquid crystal in bis(diphenylglyoximato)nickel(II )-based complexes Kazuchika Ohta,*a† Ryuji Higashi,ab Mayumi Ikejima,a Iwao Yamamotoa and Nagao Kobayashib aDepartment of Functional Polymer Science, Faculty of T extile Science & T echnology, Shinshu University, Ueda, 386, Japan bDepartment of Chemistry, Graduate School of Science, T ohoku University, Sendai, 980-77, Japan The critical molecular structure change from columnar to lamellar mesophases in bis(diphenylglyoximato)nickel(II)-based complexes has been investigated. When the number of the long chains surrounding the core complex part was reduced from eight to four, the mesophase changes its structure from columnar Colhd to novel disk-like lamellar DL.rec(P2121).When the length of the four chains at the p-positions was fixed and the remaining four chains at m-positions were gradually made shorter, each of the complexes shows a columnar Colho mesophase. Surprisingly, even when the substituents at m-positions are methoxy or methyl groups, a columnar Colho mesophase is still observed.When the substituents at m-positions were OH groups, another novel disklike lamellar DL.rec(P211) mesophase is observed. Hence, it can be shown that the critical molecular structure change from the columnar to lamellar mesophase occurs between methoxy and hydroxy groups at the m-position. In both of the two novel DL.rec mesophases in the present study, the core complex parts are parallel to the layers and the long chains are normal to both the layers and the core complex planes. Such a unique mesophase structure has not been previously observed.The synthesis of these complexes and the temperature-dependent X-ray structural analyses of the unique mesophases are reported. substituted tetraphenylporphyrins6 and tetrakis(alkyldithiola- 1 Introduction to)dinickel(II) complexes.7 Two types of disk-like lamellar Many types of organic transition metal complexes which are mesophases(DL1, DL2) have been found so far.5 Here, we peripherally substituted by long chains aVording columnar focused on our interest to prepare a new disk-like lamellar liquid crystalline properties have been reported.These columnar mesogen for bis(diphenylglyoximato)nickel(II)-based comliquid crystals can self-organize to show much higher plexes. As described above, an eight-chain dodecyloxy groupfunctionality than the unsubstituted non-mesogenic core com- substituted bis(diphenylglyoximato)nickel(II) complex [1: plexes. For example, we reported that bis(diphenylglyoximato)- Ni(12,12) in Fig. 2] shows a hexagonal disordered columnar metal(II) (M=Ni, Pd, Pt) complexes1 substituted with eight (Colhd) mesophase. If the number of chains is reduced from long alkoxy chains exhibit columnar mesophases and show eight to four, a new disk-like lamellar mesogen is expected to other various properties, e.g., thermochromism, solvatochro- appear. Hence, initially we synthesized a four-chain dodecymism and gelation.These properties originate not from the loxy-substituted nickel complex [5: Ni(12) in Fig. 2]. As individual molecules but from their supramolecular structure. expected, Ni(12) shows a lamellar type of mesophase. Furthermore, liquid crystalline compounds are also well known Moreover, it unexpectedly has a novel disk-like lamellar recto dramatically change their supramolecular structure upon tangular (DL.rec) mesophase, whose structure is quite diVerent slightly changing the molecular structure.For example, long- from those of the previous DL1 and DL2 mesophases. chain-substituted bis(b-diketonato)copper(II) complexes Nevertheless, we can state that when the number of the reported by us2 and long-chain-substituted perylene derivatives dodecyloxy chains surrounding the core Ni(DPG)2 complex reported by Spieß and co-workers3 dramatically change their is reduced from eight to four, the mesophase changes its liquid crystalline phase from calamitic to columnar upon slightly structure from columnar Colhd to disk-like lamellar DL.rec.We changing the number and length of the peripheral chains.focused on our aim to investigate critical molecular structure Praefcke and co-workers also investigated the influence of the changing from columnar to lamellar phase structure when the number and length of the peripheral chains on columnar length of the four chains is fixed and the remaining four chains mesomorphism.4 The functionalities of materials depends on are gradually made shorter.Thus we synthesized complexes 2 their supramolecular structures. Therefore, liquid crystalline and 3 (Fig. 2) which have four dodecyloxy chains at p-positions compounds whose supramolecular structures can be controlled and four shorter chains at the m-positions. In the event, each will play an important role in the development of functional of the complexes shows a columnar (Colho) mesophase.It is materials. very surprising that even if the substituent at the m-position As shown in Fig. 1, disk-like molecules containing more is a methoxy or methyl group, a columnar Colho mesophase than six side-chains generally pile up one-dimensionally to is still obtained. When chains at the m-positions were finally give columnar mesophases. On the other hand, four-chainremoved, i.e.in complex 4 (Fig. 2) which has OH groups at substituted disk-like molecules tend to show disk-like lamellar the m-positions, a disk-like lamellar (DL.rec) mesophase was (DL) mesophases.5–7 Although many columnar liquid crystaleventually observed. Thus the critical molecular structure line compounds have been reported, a very few disk-like changing from columnar to lamellar mesophase occurs between lamellar liquid crystals have been reported only for long chainn= 0 and 1 for alkoxy groups (CnH2n+1O) at m-positions. substituted bis(b-diketonato)copper(II) complexes,5 long chain- Furthermore, it is also seen that DL.rec in 4 has a lower symmetry than DL.rec in 5, and that the two-dimensional rectangular lattice shrinks owing to a hydrogen-bonding net- †E-mail: ko52517@giptc.shinshu-u.ac.jp ‡Part 20: ref. 1(f ). work (HBN) in the layer compared to 5. Surprisingly, in both J. Mater. Chem., 1998, 8(9), 1979–1991 1979Fig. 1 Relationship between the molecular structures and the liquid-crystalline phases Colhd 1: Ni (12,12) X X X X N R O H N R H O N R N R Ni O O H O H O N Ni N OC12H25 OC12H25 N N O H25C12O O H25C12O OC12H25 OC12H25 OC12H25 OC12H25 N H25C12O O H N H25C12O H O N OC12H25 N OC12H25 Ni O O DL ? ? This Work 3 : Ni (12, ñ) X = C ñH2 ñ+1 ( ñ=1-3) 2 : Ni (12, n) X =OC nH2 n+1 ( n=1-6) 4 : Ni (12,0) X = OH � 5 : Ni (12) R=OC12H25 6 : Ni (12) R=C12H25 Fig. 2 Our strategy for pursuit of critical molecular structure changing from columnar to lamellar liquid crystal in the bis(diphenylglyoximato) nickel(II)-based complexes of these two novel DL.rec mesophases, the core complex parts conventional columnar mesophase whereas the Ni(12,0) complex exhibits a unique DL.rec mesophase.are parallel to the layers and the long chains are normal to both the layers and the core planes. Such a unique mesophase We synthesized complex 6 containing four dodecyl chains only at the p-positions, in order to investigate the influence of structure has not been previously observed.The long chains in conventional smectic, columnar and disk-like lamellar meso- oxygen atoms on the mesophase formation. Complex 6 showed only tetragonal columnar (Coltet) mesophases attributable to phases lie parallel to the core planes. Hence, one can change the long chains from parallel to normal to the core planes by dimerization which apparently leads to a disk unit composed of the dimer having eight long chains.changing the number (n) of carbon atoms at the m-position from n=1 to n=0, since the Ni(12,1) complex shows a Thus, we found the critical molecular structure changing 1980 J. Mater. Chem., 1998, 8(9), 1979–1991from columnar to lamellar mesophase and two novel unique 6.82–7.59(m, aryl H, 6.IR (Nujol, cm-1) 3360(OH), 1670(CNO). DL.rec mesophases. 4,4¾-Didodecyloxy-3,3¾-dipropyloxybenzil 9 2 Experimental 4,4¾-Didodecyloxy-3,3¾-dihydroxybenzil, 8 (0.90 g, 1.5 mmol), Fig. 3 illustrates the structural formulae of long-chainpotassium carbonate (0.43 g, 3.1 mmol) and n-propyl bromide substituted bis(diphenylglyoximato)nickel(II ) complexes 1–6 of (0.39 g, 3.2 mmol) were added to 70 ml of N,N-¾dimethylacewhich 2–6 were synthesized in this work. The abbreviation of toamide and the mixture stirred for 15 h at 90 °C under a these complexes is composed of the central metal and the nitrogen atmosphere. The reaction mixture was then extracted length of long chains in the two directions as M(xy); x and y with chloroform and washed with water and the organic layer being the numbers of carbon atoms along the x and y axis, dried over anhydrous sodium sulfate.The solvent was removed respectively. Numbers n and n� represent the numbers of carbon under reduced pressure using a rotary evaporator and the atoms in the alkoxy and alkyl chains, respectively. crude product was purified by column chromatography using silica gel and chloroform and then recrystallized from ethanol Synthesis to give a white powder.Yield 0.98 g (96%), m.p. 94.5 °C. 1H NMR(CDCl3, TMS) d 0.85–1.53(m, CH3 and CH2, 56H), Synthetic routes to complexes 2–6 are shown in Scheme 1. 4.02(t, OCH2, 8H), 6.65–8.09(m, aryl H, 6H). IR (Nujol, cm-1) Detailed procedures of preparations of the precursors 7, 13 1660(CNO).and 14 have described in a previous paper.8 a-Diketone 8 which has only two dodecyloxy chains at p- 2-Methyldodecyloxybenzene 11 position was prepared by reaction of the compound 78 and dodecylbromide in the presence of potassium carbonate in 2-Methyl phenol, 10 (2.00 g, 18.5 mmol), potassium carbonate N,N-dimethylacetoamide under a nitrogen atmosphere. The (1.30 g, 9.41 mmol) and n-dodecyl bromide (4.66 g, 18.7 mmol) further m-alkylated a-diketone 9 was prepared by the same were added to 250 ml of N,N-dimethylacetoamide and the method for a-diketone 8. 2-Alkyl dodecyloxybenzenes 11 were mixture stirred for 15 h at 100 °C under a nitrogen atmosphere. prepared by reaction of the commercially available compounds The reaction mixture was then extracted with chloroform and 10 and dodecyl bromide in the presence of potassium carbonate washed with water and the organic layer dried over anhydrous in N,N-dimethylacetoamide under a nitrogen atmosphere.sodium sulfate. The solvents were removed under reduced Compounds 12 which possess two dodecyloxy groups and two pressure using a rotary evaporator and the crude product was short alkyl chains (methyl, ethyl or propyl), on the p- and mpurified by column chromatography using silica gel and chloropositions, respectively were synthesized by Mohr’s method.9 form to give a white syrup.Yield 4.31 g (80.8%). 1H The Ni(12,n):2, Ni (12,n� ):3, Ni(n,0):4, Ni(n):5 and Ni(n� ):6 com- NMR(CDCl3, TMS) d 0.68(t, CH2CH3, 3H), 1.08–1.80(m, plexes were synthesized by use of our previous method.2 CH2, 20H), 2.01(s, aryl-CH3, 3H), 3.58(t, OCH2, 2H), Detailed procedures are described below only for representative 6.29–6.75(m, aryl, 4H). compounds.Table 1 lists the yields and elemental analysis data of all the complexes. 4,4¾-Didodecyloxy-3,3¾-dimethylbenzil 12 To a mechanically stirred suspension of 2-methyldodecyloxy- 4,4¾-Didodecyloxy-3,3¾-dihydroxybenzil 8 benzene 11 (2.00 g, 7.23 mmol) and aluminium chloride (1.06 g, 3,3¾,4,4¾-Tetrahydroxybenzil 7(2.1 g, 7.7 mmol), potassium 7.95 mmol) in 20 ml of carbon disulfide at 0 °C, was added a carbonate (1.1 g, 8.0 mmol) and n-dodecyl bromide (4.35 g, solution of oxalyl chloride (0.55 g, 4.3 mmol) in 10 ml of carbon 18.2 mmol) were added to 120 ml of N,N-dimethylacetoamide disulfide over 1 h under a nitrogen atmosphere, and then the and the mixture stirred for 8 h at 70 °C under a nitrogen mixture was stirred for 5 h.After reaction the solvent was atmosphere. The reaction mixture was then extracted with removed under reduced pressure using a rotary evaporator, diethyl ether and washed with water and the organic layer the residue was extracted with dichloromethane and washed dried over anhydrous sodium sulfate.The solvents were with water. The organic layer was dried over anhydrous removed under reduced pressure using a rotary evaporator sodium sulfate and the solvent removed under reduced and the crude product was purified by column chromatography pressure. The crude product was purified by column chromausing silica gel and chloroform and then by recrystallization tography using silica gel and chloroform and then by recrysfrom ethanol to give a white powder.Yield, 2.45 g (52%), m.p. tallization from ethanol to give a white powder. Yield 1.03 g 98 °C. 1H NMR(CDCl3, TMS) d 0.88(t, CH3, 6H), (47.0%), m.p. 85.5 °C. 1H NMR(CDCl3, TMS) 0.88(t, 6H, 1.35–1.91(m, CH2, 40H), 4.13(t, OCH2, 4H), 5.74(s, OH, 2H), CH3), 1.34–1.88(m, 40H, CH2), 2.25(s, 6H, aryl-CH3), 4.04(t, 4H, OCH2), 6.80–7.83(m, 6H, aryl H).IR (KBr, cm-1) 1655(CNO). Bis[1,2-bis(4-n-dodecyloxy-3-butoxyphenyl )ethane dioximato]nickel(II ); Ni(12,4): 2 Hydroxylamine hydrochloride (3.41 g, 49.1 mmol) and 85% potassium hydroxide (3.24 g, 49.1 mmol) were added to 100 ml of ethanol and the mixture vigorously stirred for ca. 1 h, and then filtered to remove the resulting precipitate of potassium chloride. 4,4¾-Didodecyloxy-3,3¾-dibutoxybenzil (0.59 g, 0.82 mmol) was added to the filtrate and the mixture refluxed with stirring under a nitrogen atmosphere for 15 h. Nickel(II) chloride hexahydrate (0.16 g, 0.67 mmol) dissolved in a small amount of ethane-1,2-diol was added to the hot reaction mixture and the resulting solution was then immediately neutralized with glacial acetic acid.After further refluxing for N N O O H H O O N N Ni R R� R R R R R R� 1 2 3 45 6 ' ' R = R� = OC nH2 n+1 : Ni( n, n) n = 4,8,12 a R = OC12H25, R� = OC nH2 n+1 : Ni(12, n) n = 1-6 R = OC12H25, R� = C ñH2 ñ+1 : Ni(12, ñ) ñ = 1-3 R = OC nH2 n+1, R� = OH R = OC nH2 n+1, R� = H : Ni( n) n = 10,12,14 R = C ñH2 ñ+1, R� = H : Ni( ñ) ñ = 12 : Ni( n,0) n = 8,12,16 This work ~ ~ ~ 4 h, the reaction mixture was cooled to room temperature.Fig. 3 Fomulae of the long chain-substituted bis(diphenylglyoximato)- nickel(II ) complexes 1–6. aRef. 1(a) The reaction solvents were removed under reduced pressure J. Mater. Chem., 1998, 8(9), 1979–1991 1981O O HO OH OH HO O O RO OR OH HO O O RO OR OR� RO N N O O H H O O N N Ni OR OR¢ OR OR¢ OR¢ RO RO OR¢ N N O O H H O O N N Ni OR OH OR OH OH RO RO OH HO RO O O RO OR R¢ N N O O H H O O N N Ni OR R¢ OR RO RO R¢ R' (a) 7 8 9 2 : Ni(12, n) 4 : Ni( n,0) 3 : Ni(12, ñ) (b) 10 11 12 i ii i v iv , R' R' R' R' iii iv iii iv iii O O R R N N O O H H O O N N Ni R R R R 13 : R=OC nH2 n+1 14 : R=C nH2 n+1 6 : Ni( ñ) 5 : Ni( n) (C) iii iv Scheme 1 Synthetic routes of the long chain-substituted bis(diphenylglyoximato)nickel(II ) complexes 2–6.Reagents: i, RBr, K2CO3/N,Ndimethylacetoamide; ii, R¾Br, K2CO3/N,N-dimethylacetoamide; iii, NH2OH·HCl, KOH; iv, NiCl2·6H2O/ethane-1,2-diol; v, AlCl3, oxalyl chloride/CS2. Table 1 Yield and elemental analysis data for the long-chain-substituted bis(diphenylglyoximato)nickel(II) complexes, 2–6 elemental analysis (%) found (calc.) yield molecular molecular complex n(n� ) (%) formula weight C H N 2: Ni (12,n) 1 18 C80H126N4O12Ni 1394.61 69.06(68.89) 9.17(9.11) 4.05(4.02) 2 5 C84H134N4O12Ni 1450.17 69.34(69.54) 9.38(9.31) 3.72(3.86) 3 43 C88H142N4O12Ni 1506.83 69.99(70.14) 9.60(9.50) 3.79(3.72) 4 24 C92H150N4O12Ni 1562.93 70.74(70.69) 9.72(9.67) 3.59(3.58) 5 25 C96H158N4O12Ni 1619.04 71.35(71.21) 10.03(9.84) 3.37(3.46) 6 14 C100H166N4O12Ni 1675.15 71.75(71.69) 9.97(9.99) 3.05(3.34) 3: Ni (12,n� ) 1� 5 C80H126N4O8Ni 1330.61 72.02(72.20) 9.59(9.54) 4.30(.72 72.68(72.74) 9.76(9.74) 3.98(4.04) 3� 11 C88H142N4O8Ni 1441.96 73.43(73.25) 10.06(9.92) 3.85(3.88) 4: Ni(n,0) 8 9 C60H86N4O12Ni 1114.07 64.97(64.69) 8.00(7.78) 4.74(5.03) 12 19 C76H118N4O12Ni 1338.51 68.05(68.20) 8.89(8.89) 4.10(4.19) 16 10 C92H150N4O12Ni 1562.94 70.65(70.70) 9.52(9.67) 3.52(3.58) 5: Ni(n) 10 24 C68H102N4O8Ni 1162.29 69.63(70.27) 9.03(8.85) 4.92(4.82) 12 22 C76H118N4O8Ni 1274.51 71.48(71.62) 9.31(9.33) 4.38(4.40) 14 16 C84H134N4O8Ni 1386.73 72.94(72.76) 9.71(9.74) 4.05(4.04) 6: Ni(n� ) 12~ 17 C76H118N4O4Ni 1210.51 75.65(75.41) 9.52(9.83) 4.64(4.63) using a rotary evaporator and the residue extracted with then by reprecipitation by adding acetone to the chloroform solution to give red liquid crystals. Yield, 0.14 g (24%). 1H chloroform and washed with water. The organic layer was dried over anhydrous sodium sulfate, the solvent removed NMR(CDCl3, TMS) d 0.65–0.97(m, CH3, 24H), 1.33–1.76(m, CH2, 96H), 3.65(t, OCH2, 8H), 3.92(t, OCH2, 8H), under reduced pressure and the crude product purified by column chromatography using silica gel and chloroform and 6.59–6.86(m, aryl H, 12H). 1982 J.Mater. Chem., 1998, 8(9), 1979–1991Bis[1,2-bis(4-n-dodecyloxy-3- crude product was purified by column chromatography using silica gel and chloroform (Rf=0.80) and then by recrystalliz- methylphenyl )ethanedioximato]nickel( II ); Ni(12,1�): 3 ation from acetone–chloroform to give orange crystals. Yield, Hydroxylamine hydrochloride (5.04 g, 72.6 mmol) and 85% 0.13g (22%).UV–VIS (CHCl3, 4×10-5 mol dm-3) lmax/nm= potassium hydroxide (4.80 g, 72.6 mmol) were added to 125 ml 381, 424, 461(sh). 1H NMR(CDCl3, TMS) d 0.88(t, CH3, 12H), of ethanol and the mixture vigorously stirred for ca. 1 h, and 1.26–1.76(m, CH3, 80H), 3.90(t, CH2, 8H), 6.68–7.11(m, aryl then filtered to remove the resulting precipitate of potassium H, 16H).chloride. To the filtrate 4,4¾-didodecyloxy-3,3¾-dimethylbenzil 12 (0.70 g, 1.15 mmol) was added and the mixture was refluxed Bis[1,2-bis(4-n-dodecylphenyl )ethanedioximato]nickel(II ); with stirring under a nitrogen atmosphere for 15 h. Nickel(II ) Ni(12 ~ ): 6 chloride hexahydrate (0.29 g, 1.22 mmol) dissolved in a small amount of ethane-1,2-diol was added to the hot reaction Hydroxylamine hydrochloride (3.83 g, 55.1 mmol) and 85% mixture and the solution was immediately neutralized with potassium hydroxide (3.83 g, 58.0 mmol) were added to 91 ml glacial acetic acid.After further refluxing for 4 h, the reaction of ethanol and the mixture stirred vigorously for ca. 1 h, and mixture was cooled to room temperature and the reaction then filtered to remove the resulting precipitate of potassium solvents were removed under reduced pressure using a rotary chloride. 4,4¾-Didodecylbenzil (0.50 g, 0.86 mmol) was added evaporator. The residue was extracted with chloroform and to the filtrate and the mixture was refluxed with stirring under washed with water and the organic layer was dried over a nitrogen atmosphere for 17 h.Nickel(II) chloride hexahydrate anhydrous sodium sulfate. The solvent was removed under (0.43 g, 1.8 mmol) which was dissolved in a small amount of reduced pressure and the crude product was purified by column ethane-1,2-diol was added to the hot reaction mixture which chromatography using silica gel and chloroform and then by was immediately neutralized with glacial acetic acid. After reprecipitation by adding methanol to the chloroform solution further refluxing for 4 h, the reaction mixture was cooled to to give red liquid crystals. Yield, 0.040 g (5.0%). 1H NMR room temperature and the resulting precipitate collected by (CDCl3, TMS) d 0.87(t, CH2CH3, 12H), 1.26–1.78(m, CH2, filtration and dissolved in chloroform.The organic layer was 80H), 2.08(s, aryl- CH3, 12H), 3.89(t, OCH2, 8H), 6.52–7.09(m, dried over anhydrous sodium sulfate and the solvent removed aryl H, 12H). under reduced pressure. The crude product was purified by column chromatography using silica gel and chloroform (Rf= 0.91) and then by recrystallization from acetone to give orange Bis[1,2-bis(4-n-dodecyloxy-3- crystals.Yield, 0.090 g (17%). 1H NMR(CDCl3, TMS) d 0.88(t, hydroxyphenyl )ethanedioximato]nickel(II ); Ni(12,0): 4 CH2CH3, 12H), 1.25–1.55(m, CH2, 80H), 2.54(t, aryl- CH2, Hydroxylamine hydrochloride (14.4 g, 207 mmol) and 85% 8H), 7.01–7.08(m, aryl H, 16H). potassium hydroxide (14.4 g, 218 mmol) were added to 100 ml of ethanol and the mixture stirred vigorously for ca. 1 h, and Measurements filtered to remove the resulting precipitate of potassium chloride. 4,4¾-Didodecyloxy-3,3¾-dihydroxybenzil 8 (2.00 g, The products were identified by elemental analyses using a 3.27 mmol) was added to the filtrate and the mixture was Perkin Elmer 240B Elemental Analyzer. The phase transition refluxed under an N2 atmosphere with stirring for 15 h.behaviors of these compounds were observed by a polarizing Nickel(II) chloride hexahydrate (0.69 g, 2.90 mmol) dissolved microscope equipped with a heating plate controlled by a in a small amount of ethane-1,2-diol was added to the hot thermoregulator (Mettler FP80 and FP82), and measured with reaction mixture and the solution was immediately neutralized a Shimadzu DSC-50 diVerential scanning calorimeter.To with glacial acetic acid. After further refluxing for 4 h, the establish the mesophases, powder X-ray patterns were measreaction mixture was cooled to room temperature and the ured with Cu-Ka radiation using a Rigaku Geigerflex instrureaction solvents removed under reduced pressure using a ment equipped with a hand-made heating plate controlled by rotary evaporator.The residue was extracted with chloroform a thermoregulator. and washed with water and the organic layer was dried over anhydrous sodium sulfate. The solvent was removed under 3 Results and Discussion reduced pressure and the crude product purified by column chromatography using silica gel and n-hexane–ethyl acetate 3.1 Mesomorphic properties of Ni(12,n):2 and Ni(12,n� ):3 (151) and then by recrystallization from ethanol to give an We have already reported that the eight long-chain-substituted orange powder.Yield, 0.41g (19%). 1H NMR(CDCl3, TMS) d complexes Ni(n,n):1 show a hexagonal disordered columnar 0.89(t, CH3, 12H), 1.37–1.75(m, CH2, 80H), 4.01(t, OCH2, (Colhd) mesophase.2a,b In order to pursue the critical molecular 8H), 5.55(s, OH, 4H), 6.72–6.87(m, aryl H, 12H).structure changing from the columnar to the lamellar mesophase, we have synthesized the Ni(12,n):2 complexes in which Bis[1,2-bis(4-n-dodecyloxyphenyl )ethanedioximato]nickel(II ); the alkoxy group at the p-positions is fixed as a dodecyloxy Ni(12): 5 (C12H25O–) chain and the number of carbon atoms in the alkoxy group (CnH2n+1O) at the m-position is gradually Hydroxylamine hydrochloride (3.61 g, 51.9 mmol) and 85% potassium hydroxide (3.43 g, 51.9 mmol) were added to 86 ml reduced from n=6 to 1; i.e., they gradually approach from eight long-chain-substituted complexes to four long-chain- of ethanol and the mixture stirred vigorously for ca. 1 h, and then filtered to remove the resulting precipitate of potassium substituted complexes.Then, we aimed to investigate the influence of the oxygen atom in the alkoxy group at the m- chloride. 4,4¾-Didodecyloxybenzil (0.50 g, 0.91 mmol) was added to the filtrate and the mixture was refluxed with stirring positions on the mesomorphic properties, and replaced alkoxy groups by alkyl groups (Cn� H2n� +1: n� =1, 2, 3) to obtain the under an N2 atmosphere for 17 h.Nickel(II) chloride hexahydrate (large excess; 2.04 g, 8.58 mmol) dissolved in a small Ni(12,n� ) complexes. The mesomorphism of these Ni(12,n) and Ni(12,n� ) complexes were studied. amount of ethane-1,2-diol was added to the hot reaction mixture and the solution was immediately neutralized with The phase transition sequences, transition temperatures, enthalpy changes and lattice constants for mesophases of the glacial acetic acid.After further refluxing for 1 h, the reaction mixture was cooled to room temperature and the resulNi(12,n� ) complexes are listed in Table 2. The Ni(12,n) (n=1–6) and Ni(12,3�) complexes are dark red liquid precipitate collected by filtration and dissolved in chloroform. The organic layer was dried over anhydrous sodium sulfate crystals at room temperature.All of the Ni(12,n) and Ni(12,n� ) complexes show a considerable temperature region of a hexag- and the solvent was removed under reduced pressure. The J. Mater. Chem., 1998, 8(9), 1979–1991 1983Table 2 Phase transition sequences (T /°C) and enthalpy changes onal ordered columnar (Colho) mesophase. Interestingly, (DH/kJ mol-1 in parentheses) of the Ni(12,n) and Ni(12,n� ) complexes Ni(12,3) shows two types of Colho mesophases; Colho1 2 and 3a (r.t.–105.6 °C) and Colho2 (105.6–219.6 °C).However, the diVerence between them, as yet, is not clear. The lattice complex phase transitions constants of the Colho2 phase at 150 °C are almost the same as those of the Colho phase of Ni(12,1) and Ni(12,2) at 150 °C. 2: Ni (12,1) Colho CCCCCCCCCCCCCA 297 IL (decomp.) It is most surprising that even the shortest methoxy groupsubstituted complex, Ni(12,1), shows a columnar Colho meso- Ga=28.1A ° h=3.33A ° Hat r.t. Ga=28.1A ° h=3.41A ° Hat 150 °C phase. This might be attributable to the oxygen atom in the methoxy group which would make the rectangular cores rotate Ni(12,2) Colho ,bbbbbbbbbbbb) 245.9(13.6) IL readily so that they would appear as round disks.Hence, we then synthesized the Ni(12,n� ) complexes containing alkyl Ga=28.8A ° h=3.39A ° Hat r.t. Ga=29.1A ° h=3.47A ° Hat 150 °C chains but no oxygen. The Ni(12,1�) and Ni(12,2� ) complexes give a crystalline phase at room temperature and a Colho phase at higher temperatures. This means that free rotation is likely Ni(12,3) Colho1 ,bbbb) 105.6(4.43) Colho2 ,bbbb) 219.6(15.9) IL to be attributable to the influence of carbon rather than oxygen Ga=28.7A ° h=3.38A° Hat r.t.Ga=29.1A ° h=3.46A° Hat 150 °C at the m-position. Table 3 lists X-ray diVraction data of the Ni(12,1) and Ni(12,1� ) complexes; each showed (100), (110), (200), [(210) only for Ni(12,1)] reflection lines in the low angle Ni(12,4) Colho ,bbbbbbbbbbbb) 206.5(25.0) IL region corresponding to the two-dimensional hexagonal lattice, Ga=29.2A° h=3.36A ° Hat r.t.a sharp (001) reflection line in the medium angle region corresponding to the ordered stacking distance between the disks in the column, and a very broad halo corresponding to Ni(12,5) Colho ,bbbbbbbbbbbb) 209.0(33.9) IL the melting of the chains.Ga=29.7A ° h=3.38A ° Hat r.t. Thus, we found that even if the substituents at the mpositions are methoxy or methyl groups, the carbon atom in the substituent makes them form a Colho mesophase. We also Ni(12,6) Colho ,bbbbbbbbbbbb) 187.5(35.3) IL found no influence of oxygen atoms at m-positions on mesomorphism. Ga=28.0A ° h=3.34A ° Hat r.t. 3.2 Mesomorphic properties of Ni(n,0):4 and Ni(n):5 3: Ni (12,1�) K,bbbb) 42.1(30.1) Colho ,bbbb) 228.7(9.24) IL We also synthesized the Ni(n,0):4 and Ni(n):5 complexes in Ga=28.2A ° h=3.35A ° Hat 170 °C which carbon atoms in the m-positions were absent and investigated their mesomorphism.Table 4 lists the phase transition sequences, transition temperatures, enthalpy changes and Ni(12,2�) K,bbbb) 32.6(4.57)b Colho ,bbbb) 256.2(17.2) IL lattice constants for mesophases of the Ni(n,0), Ni(n) and Ga=28.2A ° h=3.35A ° Hat 170 °C Ni(12) complexes.As can be seen from this table, Ni(8,0) is not mesogenic but both Ni(12,0) and Ni(16,0) show a disklike lamellar mesophase.8 Moreover, each of the Ni(n) com- Ni(12,3�) Colho ,bbbb) 202.5(14.0) IL plexes shows a disk-like lamellar phase similar to Ni(12,0) and Ni(16,0).The structures of these disk-like lamellar mesophases Ga=28.9A ° h=3.50A ° Hat 170 °C have been established by temperature-dependent X-ray diVraction. X-Ray diVraction data of representative Ni(12,0) and aPhase nomenclature: K=crystal, Colho=hexagonal ordered columnar Ni(12) complexes are listed in Table 5. mesophase and IL=isotropic liquid. bThis small enthalpy change might be caused by a mixture of the K and Colho phases in the 3.2.1 Mesophase structure of Ni(n):5.As shown in Table 5, pristine sample. the X-ray diVraction pattern of the mesophase of Ni(12) at 120 °C showed (001), (002) and (003) reflections in the low angle region corresponding to a lamellar structure, and (110), Table 3 X-Ray diVraction data of the Ni(12,1) and Ni(12,1� ) complexes 2 and 3 spacing/A° mesophase Miller indices complex (temperature) lattice constants/A ° dobs dcalc (hkl) 2: Ni (12,1) Colho a=28.1 24.3 24.3 (100) (r.t.) h=3.33 14.1 14.1 (110) 12.5 12.2 (200) ca. 4.8 — —a 3.33 — (001) 3: Ni (12,1�) Colho a=28.4 24.7 24.6 (100) (170 °C) h=3.36 14.2 14.2 (110) 12.3 12.3 (200) 9.24 9.31 (210) ca. 4.8 — —a 3.36 — (001) aMelt of the alkoxy chains. 1984 J. Mater.Chem., 1998, 8(9), 1979–1991Table 4 Phase transition sequences (T/°C) and enthalpy changes (DH/kJ mol-1 in parentheses) of the Ni(n,0), Ni(n) and Ni(n� ) complexes 4, 5 and 6a complex phase transitions 4: Ni (8,0) K1 103.0 ,bbb) (16.6) K2 176.2 ,bbb) (32.2) IL Ni(12,0) Kb 105.4 ,bbb) (1.89) DL.rec.(P211) 170.1 ,bbb) (53.4) IL Ga=18.0A° b=14.0A ° c=34.4A ° Hat 150 °C Ni(16,0) K1 40.6 ,bbb) (6.02) K2 56–75(broad)c ,bbbb) (6.87) DL.rec.(P211) 164.5 ,bbb) (56.3) IL Ga=18.0A ° b=14.0A ° c=42.2A ° Hat 150 °C 5: Ni(10) K1 69.8 ,bbb) (39.7) DL.rec.(P2121) 159.8 ,bbb) (40.0) IL 85.4 (81.8) K2 Ga=22.2A ° b=15.7A ° c=31.1A ° Hat 120 °C Ni(12) K1 69.0 ,bbb) (33.0) DL.rec.(P2121) 148.8 ,bbb) (45.6) IL 87.3 (140.8) K2 Ga=22.2A ° b=15.3A ° c=34.5A° Hat 120 °C Ni(14) K1 27.5 ,bbb) (4.60) K2 49.2 ,bbb) (1.17) K3 77.0 ,bbb) (49.4) DL.rec (P2121) 141.2 ,bbb) (50.0) IL Ga=22.6A ° b=15.9A ° c=38.6A ° Hat 120 °C 85.4 CCCCA Coltet1 CCCA Ga=24.9A ° a¾=10.2A ° b¾=7.01A° Hat r.t.IL fast K1 115.3 ,bbb) (58.2) ColX(Coltet2) 120.5(33.1) ,bbbbbbbbbD 121.2 K2 ,bbbbbb) IL very slow cooling 6: Ni(12 ~ ) / rapid cooling slow aPhase nomenclature: K=crystal, DL.rec.=disk-like lamellar rectangular mesophase, Coltet=tetragonal columnar mesophase, ColX=unidentified columnar mesophase and IL=isotropic liquid.bSeveral unidentified transitions were observed before the melting. cThis peak was very broad. @ @ values of Ni(10), Ni (12) and Ni(14) are 31.1, 34.5 and 38.6 A ° , respectively (Table 4). Furthermore, the a and b values are surprisingly the same as those of a single crystal of the non- CCCCCA CCCCCA (200), (210), (220) and (400) reflections in the medium angle region corresponding to a two-dimensional rectangular lattice; a very broad halo due to the melting of the alkyl chains was also located in the medium angle region.These reflections substituted core complex, Ni(DPG)2, (orthorhombic, space group Iba2, Z=4, a=15.1936, b=22.3500, c=7.109 A ° ), for were also observed for mesophases in the Ni(10) and Ni(14) complexes.Table 4 lists lattice constants for all the Ni(n) which X-ray analysis was carried out by Konno et al.11 Thus, the a and b values of the long-chain-substituted Ni(n) com- complexes. As revealed from the extinction rules for twodimensional rectangular lattices10 the present lattices have plexes are completely independent of the chain length and equal to the a and b values of the non-substituted core complex.P2121(=P21/a) symmetry. Lattice constant c values can be calculated from the (00l ) This means that the long alkoxy chains are normal to both the core complex planes and the layers whereas the core reflections in the low angle region and correspond to the distance between neighbouring layers.The lattice constant c complex moieties are parallel to the layers. J. Mater. Chem., 1998, 8(9), 1979–1991 1985Table 5 X-Ray diVraction data of the Ni(12,0) and Ni(12) complexes 4 and 5 spacing/A ° mesophase Miller indices complex (temperature/°C) lattice constants/: Ni (12,0) DL.rec.(P211) a=18.1 34.7 34.4 (001) (150) b=14.0 17.2 17.2 (002) c=34.4 14.0 14.0 (010)b 11.5 11.5 (003) 11.1 11.1 (110) 9.04 9.04 (200) 7.63 7.59 (210) 6.94 6.98 (020) 6.48 6.51 (120) ca. 4.5 — —a 4.48 4.51, 4.52 (130),(400) 5: Ni(12) DL.rec.(P2121) a=22.0 34.3 34.5 (001) (120) b=15.3 17.3 17.3 (002) c=34.5 12.3 12.5 (110) 11.6 11.5 (003) 11.0 11.0 (200) 9.08 8.92 (210) 6.42 6.27 (220) 5.51 5.49 (400) ca. 4.6 — —a aMelt of the alkoxy chains. bThis assignment is not in accordance with the extinction rule of P2121(=P21/a) symmetry. The number (Z) of molecules per unit cell can be calculated mesophase of Ni(12). The top view shows the two-dimensional rectangular lattice with P2121 symmetry in the layer. The side to be ca. 4 by using the measured a, b and c values whereas Z is generally 2 for conventional two-dimensional rectangular view illustrates the bilayer structure and the alkoxy chains normal to the layers. The complete three-dimensional view lattices with P2121 symmetry.Furthermore, although the core complex moieties are parallel to the layers, the lattice constant illustrates the mesophase structure, combining the top and side views. This mesophase has a quite novel structure com- c value is equal to twice the alkoxy chain length as mentioned above.From these facts it can be concluded that this mesophase pletely diVerent from any type of mesophases reported so far. We denote this phase ‘disk-like lamellar rectangular (P2121) in the Ni(n):5 complexes should have a bilayer structure as does the non-substituted core complex.11 mesophase’ [abbreviated as DL.rec(P2121)], from its structural characteristics. Evidence for this phase structure could in principle come from the electronic absorption spectra of a thin film of the mesophase.It is well known that molecules of the core 3.2.2 Mesophase structure of Ni(n,0):4. As listed in Table 5, the X-ray diVraction pattern of the mesophase of Ni(12,0) Ni(DPG)2 complex align one-dimensionally to form metal chains.The NiMNi intermetal distance in these one-dimen- showed (001), (002) and (003) reflections in the low angle region corresponding to a lamellar structure, and (010), (110), sional chains sensitively determines the band-shift of the dz2–pz electronic transition.2d However, this mesophase does (200), (210), (020), (120) and (130)/(400) reflections in the low to medium angle regions corresponding to a two-dimensional not show such a dz2–pz electronic transition band indicating that the NiMNi intermetal distance is too wide to lead to such rectangular lattice.Almost the same reflections were also observed for the mesophase in the Ni(16,0) complex. The a transition. This is compatible with the results of X-ray diVraction studies.lattice constant c values for Ni(12,0) and Ni(16,0) are almost equal to twice the alkoxy chain length as is the case for the Fig. 4 illustrates the structural model of the unique bilayer P2121 Symmetry c = 34.5Å a = 22.0Å b = 15.3Å b = 15.3Å c = 34.5Å a = 22.0Å A whole view A top view A side view =(the length of the dodecyloxy chain) 2 or Fig. 4 Structural model of the novel lamellar mesophase of DL.rec.(P2121) of the Ni(12) complex 5 1986 J.Mater. Chem., 1998, 8(9), 1979–1991Ni(n) complexes. The lattice constant a and b values also do two-dimensional rectangular lattices C2/m and P2m symmetries, they are also excluded. Therefore, this two-dimensional not change with change of alkoxy chain length (Table 4), as for the Ni(n) complexes.However, the a and b values are rectangular lattice seems to have a novel symmetry.We inferred that the molecular directions might be regularly controlled smaller than for the Ni(n) complexes (Table 4). This shrinkage can be attributed to formation of a hydrogen-bonding network because the mesophase has a hydrogen-bonding network in the layers. On the other hand, since the lattice constant c (HBN) of the OH groups at the m-positions of the Ni(n,0) complexes in the layers, as illustrated in Fig. 5. values of Ni(12,0) and Ni(12) are the same, the long chains of Ni(12,0) should be normal to the core complex plane as for It was revealed from the temperature-dependent X-ray analysis (Table 5) that this mesophase has a unique structure Ni(12). Accordingly, as illustrated in Fig. 6, two of the long alkoxy chains may extend upward and the other two chains that we have not encountered previously. The symmetry of this lattice was considered from the extinction rules for Miller downward,12 so that the molecular direction can be assigned. In this figure, the direction from up to down is marked by an indices. As described above, the symmetry of the two-dimensional rectangular lattice of the mesophase in the Ni(n) com- arrow and taking the resulting molecular direction into the symmetry considerations, the P2121 symmetry is reduced to plexes is P2121(=P21/a).If the present mesophase in the Ni(n,0) complexes has the same P2121 symmetry, the (010) P121 or P211, as shown in Fig. 7. For P121 or P211 symmetry, a two-fold spiral axis exists in the b- and a-axis directions, reflection should disappear by the corresponding extinction rule of this symmetry [h0: h=2n+1, 0k: k=2n+1] (Table 5); respectively.In order to clearly distinguish these, these symmetries are denoted P121 and P211 following international hence, the Ni(n,0) mesophases do not have P2121 symmetry. Furthermore, if this lattice had P2/a symmetry, one of the four crystallographic notation. As summarized in Table 6, the extinction rules are 0k: k=2n+1 for P121 and h0: h=2n+1 symmetries for two-dimensional rectangular lattices reported so far,10 two molecules must align per unit cell in the a-axis for P211.Since this mesophase showed a (010) reflection, it can be concluded from these extinction rules that it has P211 direction.If this was so, the lattice constant a should be much larger than that in P2121 symmetry. By contrast, the a value symmetry. Therefore, we denote this phase ‘disk-like lamellar rectangular (P211) mesophase’ [abbreviated as DL.rec(P211)] of the present Ni(12,0) complex is smaller than that of the Ni(n) complexes. Moreover, from the extinction rule of P2/a from its structural characteristics.The control of the molecular orientations by the hydrogen- symmetry [0k: k=2n+1], the (010) reflection should disappear, hence, this mesophase does not have P2/a symmetry. bonding network can be also proven by considerations from statistical thermodynamics. First from the measured enthalpy After similar considerations from the extinction rules of two N N N N O H O O H O Ni RO OR OR O O O H H N N N N O H O O H O Ni RO RO OR OR HO O O O H H N N N N O H O O H O Ni RO RO OR OR O O OH H H O O H H N N N N O H O O H O Ni RO RO OR OR HO O O H H N N N N O H O O H O Ni RO RO OR OR O HO O H H O O H H N N N N O H O O H O Ni RO RO OR OR HO O O O H H N N N N O H O O H O Ni RO OR OR HO O O O H H OR O H RO O H OR O H RO O H RO RO a = 18.1Å b = 14.0Å H H H H Fig. 5 Possible model of the lattice shrinking by the hydrogen-bonding network in the layer of the DL.rec. mesophase for the Ni(12,0) complex 4 J. Mater. Chem., 1998, 8(9), 1979–1991 1987bilayer P211 Symmetry c = 34.4Å a = 18.1Å b = 14.0Å b = 14.0Å c = 34.4Å a = 18.1Å A whole view A top view A side view =(the length of the dodecyloxy chain) x2 Fig. 6 Structural model of the novel lamellar mesophase of DL.rec.(P211) of the Ni(12,0) complex 4.For this structure, the (010) reflection can be observed. See Table 5. changes DH listed in Table 4, the experimental enthropy diVer- the number of states W for Avogadro’s number N of the molecules is 4N. ence DSexp between the entropy change DSNi(12) for the phase transition from DL.rec(P2121) to isotropic liquid (IL) in Ni(12) hW=4N and the entropy change DSNi(12,0) for the phase transition from DL.rec(P211) to IL in Ni(12.0) can be calculated.On the other hand, the molecular directions in the mesophase having P211 symmetry are fixed as shown in Fig. 8(B). DSNi(12,0)=DH/T=53.4/(170.1+273.15)=0.1205 kJ mol-1 K-1 Therefore, the entropy diVerence DStheor between these =120.5 J mol-1 K-1 mesophases can be calculated as; DSNi(12)=DH/T=45.6/(148.8+273.15)=0.1081 kJ mol-1 K-1 DStheor=k ln W=k ln 4N=kN ln 4 =108.1 J mol-1 K-1 where k is Boltzman’s constant. hDSexp=DSNi(12,0)-DSNi(12)=120.5-108.1 hDStheor=1.38×10-23×6.022×1023×ln 4 =12.4$12 J mol-1 K-1 =11.5$12 J mol-1 K-1 This experimental entropy diVerence DSexp corresponds to the Therefore, diVerence of molecular orientational order in the layers between P211 and P2121 symmetries.Then, we theoretically DStheor=DSexp calculate the entropy diVerence DStheor by statistical thermodynamics. As illustrated in Fig. 8(A), the molecular directions This means that the control of the molecular orientations by the hydrogen-bonding network in the DL.rec(P211) mesophase in the mesophase having P2121 symmetry are not fixed but random. In this case, a molecular direction has four possibil- can be also supported by statistical thermodynamics.Fig. 6 illustrates the structural model of the unique ities; upper up, upper down, lower up, or lower down. Hence, (10) (20) (10) (01) (02) (01) A B +a +b a b A A A (10) (20) (10) (01) (02) (01) A B -b-p +a +b a b A A A (10) (20) (10) (01) (02) (01) A B +a+p +b a b A A A or (1) P2121 (2) P121 (3) P211 DL.rec.in Ni(12) DL.rec. in Ni(12,0) -a -b ( h0) line (0 k) line -a -a -b X-ray X-ray Fig. 7 Structures of the rectangular lattices having P2121, P121 and P211 symmetries 1988 J. Mater. Chem., 1998, 8(9), 1979–1991Table 6 Extinction rules for the two-dimensional rectangular lattice shaving P2121, P121 and P211 symmetries symmetry P2121 P121 P211 type of equivalent molecule molecule A molecule B molecule A molecule B molecule A molecule B +b -b +b -b-p +b -b tilt angle of the molecule from the (h0) line hfA=fB=f hfAfB hfA=fB=f -a +a -a +a -a +a+p tilt angle of the molecule from the (0k) line hfA=fB=f hfA=fB=f hfAfB two-dimensional liquid crystalline F=fA+fB(-1)(h+k) F=fA+fB(-1)(h+k) F=fA+fB(-1)(h+k) structure factor: F extinction rule [(h,k) for F=0] h0:h=2n+1 h0:h=2n+1 0k:k=2n+1 0k:k=2n+1 or A: DL.rec.( P2121) B: DL.rec.( P211) DS Fig. 8 DiVerence between the entropy changes of clearing for the DL.rec.(P2121) and DL.rec.(P211) mesophases mesophase DL.rec(P211) in Ni(12,0). Although this mesophase structure is basically the same as that of DL.rec(P2121) in Ni(12), the molecular orientations are fixed by a hydrogenbonding network and the two-dimensional rectangular lattice shrinks compared to that of DL.rec(P2121) in Ni(12).Thus, this DL.rec(P211) mesophase has a novel structure completely diVerent from any kind of mesophases reported to date. Fig. 10 Photomicrographs of the natural textures for the Coltet1 and Coltet2 mesophases of the Ni(12 ~ ) complex 6. (a) The Coltet2 mesophase at 119.3 °C.The angle in this texture is 90 °. ( b) The Coltet2 mesophase at 120.6 °C. A four-fold axis of symmetry could be observed in this texture. (c) Fingerprint texture of the Coltet1 mesophase at r.t. 3.3 Mesomorphic properties of Ni(12 ~ ):6 The pristine state of the Ni(12 ~ ) complex, 6, is a light orange needle-like crystalline phase (K1) at r.t.When it was heated to 115.3 °C, the K1 phase transformed into an unidentified columnar mesophase, ColX, as indicated in Table 4. On further heating, the ColX mesophase clears into an isotropic liquid phase, IL, at 120.5 °C. When the ColX mesophase is slowly heated, it slowly relaxes into another plate-like crystal (K2). The K2 crystals melt into the IL at 121.2 °C. A red tetragonal columnar mesophase (Coltet1) can be also obtained at r.t.when the IL at >122 °C is cooled rapidly. This Coltet1 mesophase melts into the IL at 85.4 °C in the second heating stage. This IL immediately crystallizes into the pristine state of the K1 phase. When the IL at >122 °C is cooled very slowly, the K2 Coltet1 K1 K2 Colx(Coltet2) IL 85.4 115.3 121.2 120.5 G Temperature/°C crystalline phase can be obtained.On the other hand, when Fig. 9 Schematic free energy vs. temperature (G–T) diagram of the the IL is cooled at a moderate rate, the ColX mesophase can Ni(12 ~ ) complex 6. #: transition temperature; [, A: heating; @: relaxation. be obtained. This complicated phase transition behavior can J. Mater. Chem., 1998, 8(9), 1979–1991 1989N N N N Ni O O H H O O OC12H25 OC12H25 C12H25O C12H25O OC12H25 OC12H25 OC12H25 OC12H25 X X OC12H25 OC12H25 H O O N Ni N N O H O C12H25O C12H25O X X N N N N Ni O O H H O O OC12H25 OC12H25 C12H25O C12H25O N X X X X Colhd Colho Lamellar phase DL.rec.Columnar phase 1 : Ni (12,12) 2 : Ni (12,1) 3 : Ni (12,1) 4 : Ni (12,0) 5 : Ni (12) X = OCH3 X = CH3 X = OH X = H ~ Fig. 11 The relationship between the molecular structure and the resulting mesophase structure for the bis(diphenylglyoximato)nickel(II)-based complexes.Critical change from the columnar mesophase to the lamellar mesophase occurs between 3 and 4. sponds to the intersection of the K1 line and the ColX line. The ColX mesophase slowly relaxes into the K2 phase. When the ColX mesophase does not relax completely into the K2 phase, the residue of the ColX mesophase clears into the IL at 120.5 °C which corresponds to the intersection of the ColX line and the IL line.This IL relaxes and crystallizes into the K2 phase. The K2 phase melts into the IL at 121.2 °C which corresponds to the intersection of the K2 line and IL line. Thus, the complicated phase transition and relaxation behavior can be rationally explained by using the G–T diagram.The natural texture of the ColX mesophase could be obtained by moderate cooling as shown in Fig. 10. Fig. 10(a) shows a texture growing in four directions, the angle between each direction being 90°. Fig. 10(b) clearly shows a four-fold axis symmetry in this texture. This is very similar to the texture of the Coltet mesophase of a tetrapyrazinoporphyrazine deriva- Conventional mesophases S Col DL Novel mesophase DL.rec. 1 : Ni ( n, n) 2 : Ni (12, n) 3 : Ni (12, ñ) 6 : Ni ( ñ) 4 : Ni (12,0) 5 : Ni (12) tive.13 Therefore, the present ColX mesophase may also be Fig. 12 Structural change from conventional mesophases (S, Col, DL) identified as a Coltet mesophase from these textures. However, to the novel mesophase(DL.rec) it was impossible to carry out an X-ray diVraction study for the ColX mesophase because of its narrow temperature region be rationally explained by a schematic free energy versus and the relaxation from the DX mesophase into the K2 phase.temperature (G–T) diagram. For example, the heating process Considering these microscopic observations and the G–T diaof the Coltet1 mesophase is illustrated in Fig. 9. When the gram mentioned above, it can be deduced that the ColX Coltet1 mesophase is heated from r.t., it clears into the IL at mesophase is another diVerent Coltet mesophase from the 85.4 °C which corresponds to the intersection of the Coltet1 Coltet1 mesophase in the lower temperature region. Hence, we, line and the IL line. The IL at 85.4 °C immediately relaxes and tentatively denote the ColX mesophase Coltet2.crystallizes into the metastable K1 phase. On heating, this K1 The natural texture of the Coltet1 mesophase obtained by rapidly cooling the IL to r.t. is a fingerprint-like texture, as is phase melts into the ColX mesophase at 115.3 °C which corre- 1990 J. Mater. Chem., 1998, 8(9), 1979–1991shown in Fig. 10(c). The X-ray diVraction powder pattern of kind of mesophase reported so far.It was also revealed from the extinction rules that the DL.rec mesophases of Ni(12,0) and the Coltet1 mesophase gave twenty-three reflections corresponding to a two-dimensional tetragonal lattice having lattice Ni(12) have P211 and P2121 symmetries, respectively. The molecular orientations in the DL.rec(P211) mesophase are fixed constant a=24.9 A ° (Table 4).The pattern also gave three additional reflections corresponding to a rectangular lattice as by a hydrogen-bonding network but not in the DL.rec(P2121) mesophase. The two-dimensional rectangular lattice of the the sub-lattice (a¾=10.2, b¾=7.01 A ° ). This sub-lattice may be ascribed to packing of the alkyl chains. The fact that so many DL.rec(P211) mesophase shrinks as a consequence of the hydrogen- bonding network in the layer compared to that of the reflections appear is attributed to the absence of any extinctions for a two-dimensional tetragonal symmetry.Since a diVuse DL.rec(P2121) mesophase. Thus, we have observed two kinds of novel DL.rec mesophases band could be observed at 2h#20°, the packing of the long chains is somewhat disordered, although the presence of the and determined the critical molecular structure changing from lamellar to columnar liquid crystals.sub-lattice suggests a degree of ordering of the packing of the long chains. As will be described elsewhere, the electronic spectra of a thin film of this Coltet1 mesophase exhibits a d–p References transition band which implies that the central metals align in 1 M=Ni: (a) K. Ohta, H.Hasebe, M. Moriya, T. Fujimoto and a one-dimensional manner. Columnar mesohphases are thus I. Yamamoto, Mol. Cryst. L iq. Cryst., 1991, 208, 43; (b) K. Ohta, observed although the Ni(12) complex has only four long H. Hasebe, M. Moriya, T. Fujimoto and I. Yamamoto, J. Mater. chains. This behavior may be attributable to dimerization by Chem., 1991, 1, 831; M=Pd: (c) K.Ohta, M. Moriya, M. Ikejima, which a disk unit can apparently possess eight long chains. H. Hasebe, T. Fujimoto and I. Yamamoto, Bull. Chem. Soc., Jpn., 1993, 66, 3553; (d) K. Ohta, M. Moriya, M. Ikejima, H. Hasebe, T. Fujimoto and I. Yamamoto, Bull. Chem. Soc., Jpn., 1993, 66, 4 Conclusion 3559; M=Pt: (e) M. Ikejima, M. Moriya, H.Hasebe, T. Fujimoto, I. Yamamoto and K. Ohta, in Chemistry of Functional Dyes, ed. As shown in Fig. 11, we continuously shortened the alkoxy Z. Yoshida and Y. Shirota, Mita Press, Tokyo, 1993, vol. 2, p. 801; chains at the m-position of the Ni(12,12):1 complex exhibiting ( f ) K. Ohta, M. Ikejima, M. Moriya, H. Hasebe and I. Yamamoto, a columnar Colhd mesophase to finally synthesize the methoxy- J.Mater. Chem., 1998, 8, preceding paper. substituted Ni(12,1):2 complex. It was found that this complex 2 K. Ohta, H. Akimoto, T. Fujimoto and I. Yamamoto, J. Mater. Chem., 1994, 4, 61. shows a columnar Colho mesophase. Furthermore, we synthe- 3 C. Go� ltner, D. Pressner, K. Mu� llen and H. W. Spieß, Angew. sized the methyl-substituted Ni(12,1� ):3 complex and found Chem., Int. Ed. Engl., 1993, 32, 1660. that it also shows a columnar Colho mesophase. It is very 4 B. Kohne and K. Praefcke, Chem. Ztg., 1985, 109, 121; B. Heinrich, surprising that such short group-substituted Ni(12,1) and K. Praefcke and D. Guillon, J.Mater. Chem., 1997, 7, 1363. Ni(12,1�) complexes also show a columnar mesophase. Finally, 5 K. Ohta, H. Muroki, A. Takagi, K. Hatada, H. Ema, I. Yamamoto we removed the carbon atom at the m-position to synthesize and K. Matsuzaki,Mol. Cryst. L iq. Cryst., 1986, 140, 131. 6 Y. Shimizu, M. Miya, A. Nagata, K. Ohta, A. Matsumura, the OH group m-substituted Ni(12,0):4 and non-substituted I. Yamamoto and S. Kusabayashi, Chem. L ett., 1991, 25. Ni(12):5 complexes. 7 K. Ohta, Y. Morizumi, H. Ema, T. Fujimoto and I. Yamamoto, These Ni(12,0) and Ni(12) complexes show novel lamellar Mol. Cryst. L iq. Cryt., 1991, 208, 55. (DL.rec) mesophases. Therefore, we found that the critical 8 K. Ohta, H. Hasebe, H. Ema, M. Moriya, T. Fujimoto and change from columnar to lamellar mesophase occurs between I. Yamamoto, Mol. Cryst. L iq. Cryst., 1991, 208, 21. n=0 and 1 for alkoxy (OCnH2n+1) groups at m-positions. 9 B.Mohr, V. Enkelman and G.Wegner, J. Org. Chem., 1994, 59, 635. 10 T. Komatsu, K. Ohta, T. Watanabe, H. Ikemoto, T. Fujimoto and Furthermore, we revealed the detailed mesophase structures I. Yamamoto, J.Mater. Chem., 1994, 4, 537. of the DL.rec phases by temperature-dependent X-ray diVraction 11 M. Konno, A. Kashima and I. Shirotani, Z. Kristallogr., in the studies. The long chains in these DL.rec phases are normal to press. the core complex plane, whereas the long chains in conven- 12 I. Chambrier, M. J. Cook, M. Helliwell and A. K. Powell, J. Chem. tional smectic(S), columnar(Col) and disk-like lamellar(DL) Soc., Chem. Commun., 1992, 444. mesophases are parallel to the core plane (Fig. 12). Such a unique mesophase structure is completely diVerent from any Paper 8/00897C; Received 2nd February, 1998 J. Mater. Chem., 1998, 8(9), 1979–1

ISSN:0959-9428    

DOI:10.1039/a800897c

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 EVect of alkyl sulfate anion on the mesomorphism of 3,4- dialkoxystilbazole complexes of silver(I ) Bertrand Donnio† and Duncan W. Bruce* School of Chemistry, University of Exeter, Stocker Road, Exeter, UK EX4 4QD. E-mail: d.bruce@exeter.ac.uk 3,4-Dialkoxystilbazoles have been complexed to silver(I ) alkyl sulfates in which the alkyl sulfate chain length has been varied.The mesomorphism of the materials is dominated by the formation of cubic and columnar phases, and the alkyl sulfate chain length is shown to aVect the mesophase range and type. The mesomorphism of these, and related, stilbazole complexes of silver alkyl sulfates is discussed in terms of the eVect of aromatic/paraYnic interfacial curvature and the eVects of ionic interactions.Introduction In the last 10–15 years, the literature on metallomesogens has increased hugely and has been well reviewed.1 Since initial studies on largely flat and linear complexes, the range of materials found to be capable of showing liquid crystal properties has grown substantially and numerous diVerent metals have now been incorporated successfully into mesogenic systems.In these studies, it is of interest both to define an increasing number of metal/ligand possibilities for the realisation of metallomesogens, and also to undertake systematic N OCnH2 n+1 N H2 n+1CnO Ag Y Z Y Z X + X = BF4, NO3, CF3SO3, CmH2m+1OSO3 Y, Z = H or F 1 X = CmH2m+1OSO3 Y = OC nH2 n+1 Z = H 2 investigations of the structure/property relationships so that Fig. 1 Structure of stilbazole complexes of silver(I) understanding, and ultimately prediction, of mesomorphic behaviour will arise. We have worked extensively with silver(I) complexes of appropriate bis(3,4-dialkoxy-4-stilbazole)silver triflate using monoalkoxystilbazoles (1; Fig. 12–7), and more recently on sodium hexyl- or octyl-sulfate. The metathesis reaction was mesomorphic, tetracatenar complexes based on dialkoxystilba- performed at reflux in acetone for 12 h in the dark after which zoles (2; Fig. 1).8 In this work, we have started to accumulate the products precipitated out of solution as pale yellow a substantial amount of data relating structure to property, powders (Fig. 2b). particularly in relation to cubic phases. As an extension of this The ligands and complexes were characterised by 1H and work, we now report the results of a study of the eVect of alkyl 13C NMR spectroscopy.All the stilbazoles were trans about sulfate chain length on the mesomorphism of silver(I) com- the ethylenic double bond as evidenced by proton NMR, in plexes of 3,4-dialkoxystilbazoles (Fig. 1; 2). which the coupling constant of the AB system was ca. 16.0–16.5 Hz. The complexes had a similar spectra to the free ligands, although some diVerences indicated that the com- Results and Discussion plexation had been achieved. For example, the protons on the Synthesis and characterisation of the ligands and their carbons ortho to the nitrogen of the pyridine ring, and the complexes vinylic protons were shifted downfield by a few tenths of ppm, relative to their positions in the spectra of the free ligands, The ligands were synthesised as described previously,8 and will be abbreviated St(n-3,4), where n represents the number of carbon atoms in the alkoxy chains.Sodium alkyl sulfates were obtained by reaction of chlorosulfonic acid with the appropriate terminal alcohol in ether, followed by treatment with sodium hydroxide to give the sodium salt in good yield.Silver alkyl sulfates were then prepared by precipitation after mixing aqueous solutions of silver nitrate and the corresponding sodium alkyl sulfate, although silver octyl sulfate and hexyl sulfate could not be prepared in this way, being water soluble. The bis(3,4-dialkoxy-4¾-stilbazole)silver(I) alkyl sulfate complexes were then prepared by stirring two equivalents of the stilbazole with one equivalent of silver alkyl sulfate in dichloromethane at room temperature in the dark for 4 h (Fig. 2a). The products so obtained remained stable for a prolonged period of time when stored at room temperature in the dark. The complexes of silver hexyl- and octyl-sulfate could not be prepared this way, and so were obtained via metathesis of the †Present address: Institut fu� r Makromolekulare Chemie, Universita� t [AgL2][CF3SO3] N OCnH2 n+1 N CnH2 n+1O Ag CnH2 n+1O OCnH2 n+1 CmH2m+1OSO3 – AgCF3SO3 + 2 L Acetone rt AgOSO3CmH2m+1 + 2 L Dichloromethane rt NaOSO3CmH2m+1 Acetone heat a b + Fig. 2 Synthesis of the complexes under study Frieburg, Sonnenstrasse 5, D-79104 Freiburg, Germany. J. Mater. Chem., 1998, 8(9), 1993–1997 1993while the carbons ortho and para to the ring nitrogen were shifted downfield by 2–4 ppm.The complexes were further characterised by the presence of an additional triplet at around 4 ppm due to the OCH2 protons of the alkyl sulfate anion in addition to the triplets from the alkoxy groups bound to the stilbazole. The purity of the ligands and the complexes was confirmed by elemental analysis.EVect of alkyl sulfate chain length in complexes of 3,4- dialkoxystilbazoles, [Ag{St(n-3,4)}2][CmH2m+1OSO3] We have already shown that the alkyl sulfate chain length influences dramatically the mesomorphic properties of silver complexes of this general type. For example, silver complexes of 4-alkoxy-4¾-stilbazoles with dodecyl- and tetradecyl-sulfate anions showed nematic, SA, SC, and cubic phases, while the Fig. 4 Representation of the phase behaviour of [Ag{St(4-3,4)}2]- [CmH2m+1OSO3] analogous complexes with decyl-and octyl-sulfate showed only nematic, SA, SC phases, with the cubic phase being absent. We rationalised this behaviour in relation to the position of the anion chain relatively to the paraYnic/aromatic interface.6 Thus, when the anion chain length was short so that it did not extend beyond the aromatic core (octyl- and decyl-sulfate),3 the cubic phase was absent, but when it extended beyond the interface (dodecyl- and tetra-decylsulfate), the cubic phase was observed.In this present study, therefore, it is of interest to see the eVect of the anion chain length on a system where columnar and cubic phases predominate.It is, therefore, first appropriate to recall the mesomorphism of the complexes [Ag{St(n-3,4)}2][C12H25OSO3] which can act as a reference point (Fig. 3). Thus, for 4n10, a cubic phase was seen while for 6n12, a columnar hexagonal phase was observed. Melting points were relatively constant at between 55–70 °C, while clearing points varied from 100–170 °C.8 In these complexes, as in the complexes described Fig. 5 Representation of the phase behaviour of [Ag{St(6-3,4)}2]- [CmH2m+1OSO3] below, the mesophases were identified initially by optical microscopy, with both the cubic and columnar mesophases giving characteristic textures. In the latter case, these textures were consistent with the columnar phase having hexagonal symmetry. The mesomorphism in these complexes was finally confirmed using both X-ray diVraction and freeze-fracture electron microscopy.In order to ‘sample’ the eVects of the alkyl sulfate chain length on mesomorphism, the chain length was varied for three diVerent 3,4-dialkoxystilbazoles, namely the butyloxy, hexyloxy and dodecyloxy derivatives. Up to six diVerent alkyl sulfates were used with m=6, 8, 10, 12, 14 and 16.The mesomorphism of the three series with n=4, 6 and 12, as a function of the alkyl sulfate carbon chain length, m, is shown in Fig. 4–6 respectively, the phase diagrams being drawn at the same scale. Thermal data are collected in Table 1. A brief overview of the data shows that in general, the longer chain alkyl sulfates led to lower melting points, while Fig. 6 Representation of the phase behaviour of [Ag{St(12-3,4)}2]- longer chains on the stilbazoles led to stabilisation of the [CmH2m+1OSO3] mesophases.Thus, there was a huge increase in the domains of stability of the mesophases with both increasing m and n. The mesophases were characterised by a combination of optical microscopy and X-ray diVraction. The cubic phases were seen as viscoisotropic phases which could be observed growing as characteristic square features from either the columnar or isotropic phase.The columnar phases showed a characteristic fan-like texture which was indicative of a columnar hexagonal structure. The X-ray data for the columnar phases confirmed the assignment as hexagonal, and structural data are included in Table 2. X-Ray data were also consistent with the observation of a cubic phase, and allowed the phase to be identified as having Ia39d symmetry; some data are also included in Table 2. The detailed X-ray analysis will be published separately as part of a much larger structural study of related Fig. 3 Phase diagram for the complexes [Ag{St(n-3,4)}2]- [C12H25OSO3] complexes.9 1994 J. Mater. Chem., 1998, 8(9), 1993–1997Table 1 Thermal behavior of the complexes formation of a columnar hexagonal phase with a wide (#100 °C) range for m=12 and 14, the often-observed cubic m transition T/ °C DH/kJ mol-1 DS/J K-1 mol-1 phase having completely disappeared. However, on heating the complexes with m=8 and 10, they first formed a phase 3,4-dibutyloxystilbazole which appeared optically isotropic and which gave many 10 Crys–I 98 29.0 78 reflections in the X-ray diVraction pattern, suggesting a crystal- 12 Crys–Cub 90 27.9 77 Cub–I 102 2.1 6 line cubic phase.However, on annealing for some 20 min (e.g. 14 Crys–Cub 84 30.1 84 at 100 °C for m=10), a birefringent texture was found to be Cub–I 102 1.0 3 growing in which resembled that of a columnar phase; the 16 Crys–Crys¾ 53 12.6 39 columnar phase subsequently cleared at a higher temperature.Crys¾–Cub 69 20.9 61 These observations will be discussed subsequently in a paper Cub–I 108 1.0 3 detailing the structural studies on these materials. 3,4-dihexyloxystilbazole In general, the mesomorphism in 3,4-disubstituted poly- 6 Crys–I 109 25.2 66 catenar systems can be regarded as depending to a very large 10 Crys–Cub 64 29.6 88 degree on the curvature at the aromatic/paraYnic interface,10 Cub–I 115 2.6 7 by analogy with lyotropic systems for which the interfacial 12 Crys–Cub 61 29.3 88 Cub–Colh 118 2.6 7 curvature is defined by the ratio between the volume of the Colh–I 153 1.0 2 lipophilic moiety and that of the hydrophilic moiety.11 Thus, 14 Crys–Cub 50 46.7 144 at short chain lengths, nematic and smectic C phases can be Cub–Colh 113 2.5 6 seen as there is little or no curvature at the interface, while at Colh–I 145 0.4 1 much longer chain lengths, a significant curvature is introduced 16 Crys–Crys¾ 34 2.3 8 which tends to promote the formation of columnar phases Crys¾–Cub 44 13.8 43 formed from repeating units apparently containing 3–4 mol- Cub–Colh 113 2.0 5 Colh–I 150 0.8 2 ecules.In certain cases, the change from lamellar to columnar mesomorphism happens on moving from one homologue to 3,4-didodecyloxystilbazole the next,12 although it is more common to see some homol- 8 Crys–Crys¾ 80 17.2 49 Crys¾–Crysa 86 45.6 127 ogues exhibiting, for example, both a smectic C and a columnar Crys–Colh 101 2.6 7 phase. However, it is also common to see that at some Colh–I 170 — — intermediate value of the curvature, a cubic phase is inserted 10 Crys–Crys¾ 37 12.8 41 into the phase sequence between the lamellar and columnar Crys¾–Crys 62 10.4 31 phase.This is beautifully illustrated by the phase diagram of Crys–Crys¾a 83 39.4 111 some tetracatenar bipyridines.13 The curvature therefore arises Crys¾–Colh 97 2.6 7 as a mismatch between the volume of the molecular core and Colh–I 172 3.3 8 12 Crys–Crys¾ 71 54.8 159 that of the chains, and so in the silver complexes it is necessary Crys¾–Colh 83 10.6 30 to take into account the contributions of the rigid core, the Colh–I 172 3.7 8 chains and the anion.Against this background, we can then 14 Crys–Crys¾ 60 18.9 57 rationalise the mesomorphism of the complexes in this study.Crys¾–Colh 73 91.5 275 Transitions between phases in individual homologues are then Colh–I 173 1.8 4 generally accounted for by the increased volume of the chains aThe Crys phase (m=8) and the Crys¾ phase (m=10) are the which results from greater motion at higher temperature. isotropic crystal phases referred to in the text. Our previous work with tetracatenar silver complexes with m=12 showed8 that at shorter chain lengths, the cubic phase predominated, while our much earlier work with simple, two- Table 2 Structural parameters for the columnar phase of the complexes chained, calamitic silver complexes showed that for cubic m T/°C phase d211/d01 a phases to be seen, longer stilbazole chain lengths were needed and the anion chain had to extend out beyond the rigid, 3,4-dihexyloxystilbazole aromatic core of the molecule, which happened only for m12. 12 95 Cub 31.5 In this case, when the alkyl sulfate chain crosses the aromatic/ 14 95 Cub 31.8 paraYnic interface, it contributes to the increase in the curva- 14 130 Colh 29.8 ture in such a way that a cubic phase is formed, although over 16 95 Cub 31.8 16 130 Colh 30.4 a shorter temperature range than in the case of the tetracatenar silver complexes.In this two-chained system, the cubic phase 3,4-didodecyloxystilbazole can be seen as a frustrated structure, owing to the presence of 8 130 Colh 35.5 10 130 Colh 35.3 a SA phase above the cubic, and a SC phase below it. However, 12 130 Colh 35.7 when the alkyl sulfate is confined in the aromatic part of the 14 130 Colh 35.8 molecule, there is no contribution to the curvature and as a result, only smectic and nematic phases are seen.These ad211 for cubic phase; d10 for columnar phase. complexes behave as simple calamitic mesogens. In the case of the 3,4-dibutyloxystilbazoles, when the anion The phase behaviour of the 3,4-dibutyloxystilbazole extends beyond the core and as such contributes to chain complexes as a function of the alkyl sulfate chain length (Fig. 4) volume and, to a lesser extent, the curvature, then a cubic showed that for m=10, the complex was non-mesomorphic, phase appears whose stability increases with increasing alkyl while for m12, a cubic phase was seen whose phase range sulfate chain length. However, neither the density of chains, increased from 12 °C (m=12) to 38 °C (m=16), due both to nor the degree of curvature, is suYcient to stabilise a columnar destabilisation of the crystal phase and stabilisation of the phase.However, for m=10, the anion cannot contribute to the cubic phase. For the 3,4-dihexyloxystilbazole complexes chain volume and the two butyloxy chains are too short to (Fig. 5), it was found that for m=6, the complex was non- create the curvature (Fig. 7) necessary for the formation of a mesomorphic, while for m=12, 14 and 16, both a cubic and a cubic phase, so no cubic phase is seen and the complex simply columnar phase were observed while the complex with m=10 melts to isotropic on heating. Interestingly, it is also clear that showed only a cubic phase. The mesomorphism of the 3,4- what curvature is induced is suYcient to suppress the formation of lamellar or nematic phases.dodecyloxystilbazole complexes (Fig. 6) was dominated by the J. Mater. Chem., 1998, 8(9), 1993–1997 1995bution to the interfacial curvature is greatly enhanced as it begins to ‘interfere’ with the terminal alkoxy chains. For twochained systems, this additional curvature leads to the formation of cubic phases, while for tetracatenar systems, both cubic and columnar phases are observed.It may then be that in the case of the 3,4-dihexyloxystilbazole complexes of silver decyl sulfate where a cubic phase is rather unexpectedly observed, that ionic interactions are playing a key ro� le is stabilising the cubic phase. One way of attempting to evaluate such an idea is to study extensively the mesomorphism of complexes of palladium carboxylates bearing polycatenar stilbazoles; these studieare Fig. 7 Schematic diagram to show the relative steric eVects of lateral alkyl sulfate chains currently underway and the results will be reported in due course. A similar explanation accounts for the lack of mesomorphism in the 3,4-dihexyloxystilbazole complexes with a hexyl Experimental sulfate anion, in that the stilbazole chains alone are insuYcient to give rise to a curvature that stabilises the cubic phase in Dichloromethane was distilled from calcium hydride, methanol the absence of a contribution from the anion chain.However, from magnesium and iodine, and acetone from potassium the presence of the cubic and columnar phase for m12 is permanganate. Diethyl ether was dried and stored over sodium consistent with the general mesomorphism of tetracatenar wire.All other chemicals were used as supplied. 3,4- materials. The slightly unexpected result is that for m=10, a Dialkoxystilbazoles were synthesised as described previously.8 cubic phase is seen when the anion chain does not extend Proton and carbon NMR spectra were recorded on a Bru�ker beyond the rigid, aromatic core and this is diYcult to rational- ACL250 spectrometer and referenced to external tetramethylise in the light of the absence of mesomorphism for the same silane.In order to measure with accuracy the diVerent coupling stilbazole when m=6, particularly as in this latter case, the constants (2JHH, 3JHH and 4JHH), a program Window NMR core volume (including the anion) is smaller.(1D-Win NMR, Bru� ker, MS Windows) was used. The assign- The mesomorphism of the 3,4-didodecyloxystilbazole ment of all the carbon peaks was possible using one-bond complexes is entirely consistent with that now expected and (1JCH) and multiple-bond (2JCH, 3JCH and 4JCH) C–H correthe volume occupied by the ligand chains is the dominant lation15 as well as the estimation of 13C and 1H chemical shifts parameter, giving rise to an interfacial curvature allowing in substituted benzenes.16 Microanalysis was performed by the stabilisation of the columnar phase only.University of SheYeld micro-analytical service. The study of We can now begin to obtain some understanding of the the thermal behaviour of the complexes was achieved by DSC general ro� le of the intermolecular ionic interactions which we analysis, carried out using a Perkin-Elmer DSC7 instrument proposed previously6 as being significant in stabilisation of the using various heating rates (2, 5 and 10 K min-1) and the cubic phases in these compounds.Recall once more that in mesomorphism was studied by hot-stage, polarising the simple two-chained silver complexes with alkyl sulfate microscopy using a Zeiss Labpol microscope equipped with a anions, N, SA, SC and cubic phases are seen.However, in Linkam TH600 hot-stage and PR600 temperature controller. analogous, non-ionic complexes of palladium(II) (Fig. 8) which possess a lateral substituent, only nematic phases are seen and indeed, such behaviour is typical of mesogens substituted by Sodium alkylsulfates lateral alkyl chains.14 The procedure is illustrated for sodium decyl sulfate and That a nematic phase is seen at all in any of the silver follows a general method described by Gilbert.17 All other complexes shows that the molecules can exist as discrete (ionalkyl sulfate salts were similarly prepared. paired) entities, at least when the chain length is short.Chlorosulfonic acid (1.5 g, 12.6 mmol) was added dropwise However, as the stilbazole chain length is increased, the to a flask containing diethyl ether (50 cm3), and the mixture resulting microphase separation tends to promote a layering. was left stirred for a few min. Decan-1-ol (2 g, 12.6 mmol) in In the absence of ionic interactions, such microphase separation diethyl ether (20 cm3) was then added to the solution, and the would not occur in a laterally alkylated system, but in these stirring was maintained for 30 min.The solvent was evaporated cases, the layering must be stabilised by the additional ionic in vacuo and the residue was dissolved in ethanol (100 cm3, interactions possible.Thus, the microphase separation and 95% v/v). Sodium hydroxide (0.5 g, 12.6 mmol) was added, ionic interactions act synergically to stabilise lamellar phases, and the solution was stirred for 5 h. The precipitate was and the volume of the molten lateral chain cannot lead to the formation of a nematic phase, rather exerting a steric influence filtered, washed thoroughly with absolute ethanol, and then generating some degree of curvature at the aromatic/aliphatic with diethyl ether. Sodium decyl sulfate was obtained in good interface.So long as the chain is confined within the extent of yield (70%, 2.3 g) after being air-dried. the rigid core the steric influence is clearly not large which implies that the ionic interactions are quite strong, but once Silver(I ) alkyl sulfates the chain extends beyond the core (at m=12), then its contri- The procedure is illustrated for silver decyl sulfate.Silver salts of longer-chain alkyl sulfates were similarly prepared, but the hexyl- and octyl-sulfate salt could not be so obtained as they were too soluble in water. Sodium decyl sulfate (2.3 g, 8.8 mmol) in warm water (10 cm3) was added to an aqueous solution of silver nitrate (1.5 g, 8.8 mmol) and the resulting mixture stirred for 2 h, in the vessel protected from light.The precipitate was filtered, washed with cold water and dried under high vacuum. The N CnH2 n+1O N OCnH2 n+1 Pd O O O O m m Fig. 8 Structure of laterally substituted palladium stilbazole complexes yield was 80% (2.4 g). 1996 J. Mater. Chem., 1998, 8(9), 1993–1997Table 3 Analytical data for the complexes Bis[(3,4-dialkoxy)-4¾-stilbazole]silver(I ) alkyl sulfates The complexes with an alkyl sulfate with more than 10 carbons found (calc.) (%) were directly prepared as described below, while the hexylm yield (90) C H N S and octyl-sulfate complexes were obtained in two steps as described later. 3,4-dibutyloxystilbazole A solution of 3,4-diheptoxy-4¾-stilbazole (500 mg, 1.22 10 76 62.3 (62.7) 7.5 (7.6) 3.2 (2.8) 3.6 (3.2) mmol) in dichloromethane (10 cm3) was added dropwise to a 12 78 63.0 (63.3) 7.5 (7.8) 2.5 (2.7) 3.0 (3.1) stirred suspension of AgC12H25OSO3 (250 mg, 0.67 mmol) in 14 67 63.6 (63.9) 7.8 (7.9) 2.6 (2.7) 2.9 (3.0) 16 30 63.9 (64.5) 8.0 (8.1) 2.6 (2.6) 3.1 (3.0) dichloromethane (10 cm3) and stirred (4 h, room temp.) with the vessel protected from the light.The mixture was then 3,4-dihexyloxystilbazole 6 58 63.0 (63.9) 8.0 (7.9) 2.7 (2.7) 3.5 (3.0) filtered through Celite and the solvent removed under reduced 10 52 64.7 (65.0) 8.3 (8.3) 2.8 (2.5) 2.8 (2.9) pressure. The yellow solid was crystallised from hot acetone 12 78 65.3 (65.5) 8.1 (8.4) 2.5 (2.5) 3.0 (2.8) and then recrystallised from methanol.The product was col- 14 61 65.8 (66.0) 8.6 (8.6) 2.5 (2.4) 2.6 (2.7) lected as a pale yellow solid in 67% yield (488 mg). 16 56 66.6 (66.5) 8.7 (8.7) 2.4 (2.3) 2.5 (2.7) All the other complexes were similarly prepared and 3,4-didodecyloxystilbazole obtained in similar yields. The analytical data are collected in 8 65 68.8 (69.4) 9.3 (9.6) 2.4 (2.0) 2.6 (2.3) Table 3; 1H and 13C NMR data are given below. 10 74 69.2 (69.8) 9.8 (9.7) 2.1 (1.9) 2.2 (2.2) 12 83 69.7 (70.1) 9.6 (9.8) 1.8 (1.9) 2.3 (2.2) 14 62 70.5 (70.4) 9.8 (9.9) 1.9 (1.9) 2.3 (2.1) Typical preparation of the hexyl- and octyl-sulfate silver(I ) complexes 113.0 ( j), 121.4 (n), 121.5 (i), 122.3 ( l ), 128.2 (h), 135.6 (k), The preparation of these complexes involved the formation of 147.2 (m), 149.2 (e), 150.7 (f ), 152.2 (o).the triflate complex followed by a metathesis reaction with the relevant sodium alkyl sulfate. B. D. would like to thank the European Union for the award 3¾,4¾-Didodecyloxy-4-stilbazole (250 mg, 4.6×10-4 mol) was of a Human Capital and Mobility Category 20 fellowship. The stirred in acetone (5 cm3) with silver trifluoromethanesulfonate authors thank Drs Daniel Guillon and Benoý�t Heinrich (60 mg, 2.3×10-4 mol) at room temperature in the dark for (IPCMS, Strasbourg) for access to X-ray data. 4 h. The mixture was then cooled in ice and the precipimed was filtered and washed with cold acetone to yield the References pure complex (94%, 291 mg) as a pale yellow solid of analyt- 1 A.-M. Giroud-Godquin and P.M. Maitlis, Angew. Chem., Int. Ed. ical purity. Engl., 1991, 375; D. W. Bruce, in Inorganic Materials, ed A solution of the complex previously prepared (150 mg, D. W. Bruce and D. O’Hare, Wiley, Chichester, 2nd edn., 1996; 1.1×10-4 mol) in acetone (20 cm3) was added to stirred Metallomesogens, ed. J. L. Serrano, VCH, Weinheim, 1996; suspension of sodium octyl sulfate (77 mg, 3.3×10-4 mol) in D.W. Bruce, J. Chem. Soc., Dalton T rans., 1993, 2983; P. Espinet, the same solvent (10 cm3) and was heated at reflux in the dark J. L. Serrano, L. A. Oro and M. A. Esteruelas, Coord. Chem. Rev., 1992, 117, 215; A. P. Polishchuk and T. V. Timofeeva, Russ. Chem. for 12 h. When cooled, the solution was filtered through Celite, Rev., 1993, 291. to eliminate the excess of sodium octyl sulfate, and the solvent 2 D.W. Bruce, D. A. Dunmur, S. A. Hudson, P. M. Maitlis and evaporated. The solid was then crystallised once from meth- P. Styring, Adv.Mater. Opt. Electron., 1992, 1, 37. anol, and twice from acetone. The complex was recovered by 3 H. Adams, N. A. Bailey, D. W. Bruce, S. C. Davis, D. A. Dunmur, filtration, washed with acetone and dried under vacuum. Yield P.D. Hempstead, S. A. Hudson and S. Thorpe, J. Mater. Chem., 1992, 2, 395. 65% (102 mg). 4 D. W. Bruce, B. Donnio and C. Fernihough, unpublished work. 5 D. W. Bruce and S. A. Hudson, J. Mater. Chem., 1994, 4, 479; D. W. Bruce, D. A. Dunmur, S. A. Hudson, E. Lalinde, [Ag{St(6-3,4)}2][GoH21OSO3] Spectroscopic data P. M. Maitlis, M. P. McDonald, R. Orr, P.Styring, A. S. Cherodian, R. M. Richardson, J. L. Feijoo and G. Ungar, Mol. Cryst. L iq. Cryst., 1991, 206, 79. 6 D. W. Bruce, B. Donnio, S. A. Hudson, A.-M. Levelut, S. Megtert, D. Petermann and M. Veber, J. Phys. II, 1995, 5, 289. 7 J. Bell, S. A. Hudson and D. W. Bruce, unpublished work. 8 D.W. Bruce, B. Donnio, D. Guillon, B. Heinrich and M. Ibn-Elhaj, L iq. Cryst., 1995, 19, 537; B.Donnio, D. W. Bruce, B. Heinrich, D. Guillon, H. Delacroix and T. Gulik-Krzywicki, Chem. Mater., 1997, 9, 2951. OCdH2-CcH2-(CbH2)3-CaH3 OCd'H2-Cc'H2-(Cb'H2)3-Ca'H3 i j g N Ag h e f k l m n o + CsH3-(CrH2)7-CqH2-CpH2-OSO3 – 2 9 B. Heinrich and D. Guillon, work in progress. 10 A. Skoulios and D. Guillon,Mol. Cryst., L iq. Cryst., 1988, 165, 317; The 1H and 13C NMR data for [Ag{St(6-3,4)}2]- A.-M.Levelut and M. Clerc, L iq. Cryst., 1998, 24, 105. [C10H21OSO3] are given in detail. The spectra for all other 11 J. M. Seddon and R. H. Templer, Philos. T rans. R. Soc. L ondon complexes were eVectively identical save for the number of Sect. A, 1993, 344, 377. 12 B. Donnio and D. W. Bruce, J. Chem. Soc., Dalton T rans., 1997, protons/carbons associated with the various alkyl chains. 2745. dH(250.13 MHz, CDCl3): 0.85 (s, t, 3J=6.7 Hz, 3 H), 0.89 (a, 13 K. E. Rowe and D. W. Bruce, J.Mater. Chem., 1998, 8, 331. a¾, t, 3J=6.6 Hz, 12 H), 1.38 (b, b¾, r, m, 38 H), 1.54 (q, m, 2 14 J. P. Rourke, F. P. Fanizzi, N. J. S. Salt, D. W. Bruce, H), 1.80 (c, c¾, m, 8 H), 3.95, 3.98 (d, d¾, 2 t, 3J=6.6 Hz, 8 H), D. A. Dunmur and P. M. Maitus, J. Chem. Soc., Chem. Commun., 4.16 (p, t, 3J=6.7 Hz, 2 H), 6.66 ( l, AB, 3Jlk=16.6 Hz, 2 H), 1990, 229. 15 R. J. Abraham, J. Fisher and P. Loftus, Introduction to NMR 6.70 ( j, d, 3Jji=7.9 Hz, 2 H), 6.86 (i, dd, 4Jig=2.0 Hz, 3Jij= spectroscopy, Wiley and Sons, 3rd edn., 1993. 7.9 Hz, 2 H), 6.95 (g, d, 4Jgi=2.0 Hz, 2 H), 7.17 (k, AB, 3Jkl= 16 D. H. Williams and I. Fleming, Spectroscopic Methods in Organic 16.6 Hz, 2 H), 7.26 (n, AA¾XX¾, |Jno+Jno¾|=6.5 Hz, 4 H), 8.68 Chemistry, McGraw-Hill Publishers, London, 4th edn., 1989. (o, AA¾XX¾, |Jon+Jon¾|=6.5 Hz, 4 H). dC(62.9 MHz, CDCl3): 17 E. E. Gilbert, Synthesis, 1969, 1. 13.8 (a, a¾), 14.1 (s), 22.6, 25.3, 26.0 29.1, 29.2, 29.4, 29.7, 31.4, 31.6 (b, b¾, c, c¾, r), 32.0 (q), 67.9 (p), 69.2, 69.4 (d, d¾), 111.8 (g), Paper 8/03964J; Received 27th May, 1998 J. Mater. Chem., 1998, 8(9), 1993–1997 1997

ISSN:0959-9428    

DOI:10.1039/a803964j

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 Novel metal–chelate emitting materials based on polycyclic aromatic ligands for electroluminescent devices Hiromitsu Tanaka,* Shizuo Tokito, Yasunori Taga and Akane Okada T oyota Central R&D laboratories, Inc., Nagakute, Aichi, 480-11, Japan We have designed and synthesized novel metal–chelate complexes based on polycyclic aromatic ligands for electroluminescent devices.These complexes exhibited strong luminescence with blue and green colors. EL properties of devices using these complexes for an emitting layer have been studied. Several good emitting materials were obtained and the EL properties were found to strongly depend on the ligand structure. Introduction Organic electroluminescent (EL) devices are of great interest because of attractive applications such as large-area light emitting displays which are operative at a low drive voltage.1 Organic EL devices can be easily tuned by choosing a suitable emitting material.2 The most representative emitting material is Alq3 [tris (8-quinolinolato)aluminium] which was originally reported by Tang et al.3 Alq3 has high luminous eYciency, high electron mobility and also good chemical stability.The chemical structure of Alq3 consists of one aluminium ion and three ligands [8-quinolinolate (quin)] based on a fused ring system with imine and phenolate functionalities as coordination sites. Recently, metal–chelate materials based on polycyclic aromatic ligands which chelate through a phenolic oxygen and doubly bound nitrogen [NN(o-phenol)] have been reported.These compounds are attractive for tuning emission colors, because such metal–chelate materials with polycyclic aromatic ligands emit various luminescence colors depending on the molecular structure. For example, it has been reported that Zn(BOX)2 {bis[2-(2-hydroxyphenyl)benzoxazolate]zinc}4 emits blue light and Zn(BTZ)2 {bis[2-(2-hydroxyphenyl) benzothiazolate]zinc}5 emits whitish light.However, the relationship between the ligand structure and the PL spectrum is, as yet not clear. The hetero atoms (O,S,N) in the aromatic ligands are presumed to aVect the photoluminescence (PL) and EL properties by changing the energy level of the molecular orbitals. Here, we report novel metal–chelate complexes based on polycyclic aromatic ligands containing an NN(o-phenol) moiety, in which electron donating (N,S,O) groups are introduced at the 1 position of the imino ring and/or electron withdrawing (MNN) groups are introduced at the 4 position of the imino ring (Fig. 1). The PL and EL properties of the complexes have been studied. Especially, the influence of the N N O N N S Zn O N O N N N O N N Ph O O O N N N N O O Zn S O Zn N N Ph O Zn N N O O O Al O N N O Al(ODZ)3 Zn(ODZ)2 Zn(TDZ)2 Zn(BIZ)2 Zn(PhPy)2 hetero group in the polycyclic ligands and metal center in the Fig. 1 Structures of the complex molecules with polycyclic ligands complexes upon the PL spectrum has been studied systematically. The EL spectra of the EL devices using these materials coordination site. ODZ and TDZ have one donating group were almost identical to the PL spectra, and several good blue (O, S) and one electron withdrawing group in the imino ring.and green emitting materials were obtained in this study. BIZ, BTZ and BOX has one electron donating group (N, S, O) in the coordinated imino ring while PhPy contains no hetero group. Results and Discussions ODZ and TDZ were synthesized by cyclization of the corresponding 1,2-diacylhydrazine. Cyclization with thionyl The novel metal–chelate complexes synthesized are shown in Fig. 1 and the ligands are shown in Fig. 2. ODZ, TDZ, chloride yielded ODZ, and cyclization with phosphorus pentasulfide yielded TDZ. BIZ was obtained by condensation of o- BIZ and PhPy are novel ligands while BTZ, BOX and 8- quinolinol have been reported.4,5 The novel ligands (Fig. 2) anisic acid and N-phenyl-1,2-phenylenediamine. The methoxy group of BIZ intermediate was demethylated in the course of were synthesized according to Scheme 1. Each synthesized ligand molecule has anNN(o-phenol) type the condensation reaction. PhPy was synthesized by Grignard J. Mater. Chem., 1998, 8(9), 1999–2003 1999N N O N N S O N S N N N Ph HO HO HO HO HO N HO N N HO HO X Quinolinol BTZ BOX PhPy BIZ TDZ ODZ 1 2 3 imino ring o -phenol (X=N,O,S,C) 4 Fig. 2 Structures of the polycyclic ligand molecules O Cl O NHNH2 OMe O NH OMe NH O N N S MeO N N O MeO NH2 NHPh OMe CO2H N Br BrMg MeO MeO N TDZ ODZ PhPy BIZ 1 2 3 4 + + Scheme 1 Synthetic schemes for the target ligands coupling of o-methoxybromobenzene with nickel phosphine spectral halfwidth (Dl) of Zn(BIZ)2 was less than those of the Zn(BOX)2 and Zn(BTZ)2.Broadening of the PL spectrum catalyst6 followed by demethylation with BBr3. EL devices were fabricated by vacuum deposition of the appears to be suppressed when the nitrogen group is introduced as a donating group into the ring compared to oxygen and organic materials and must be stable to heating during the deposition process, hence thermal durability is important in sulfur.The PL peak red-shifted as the electron-donating character of the heteroatoms increased [ionization potential these materials. The complexes synthesized were thermally stable up to 300 °C. No phase transitions of the complexes of heteroatom/eV:8 14.53 for N, 13.62 for O and 10.36 for S, PL peak wavelength/nm: 440 for Zn(BIZ)2, 478 for were observed by DSC measurements between room temperature and 300 °C.Zn(BOX)24 and 487 for Zn(BTZ)25 with large spectral expansion towards longer wavelengths]. All of the complexes exhibited strong fluorescence. PL spectra of the metal complexes are summarized in Table 1. A similar relationship between the PL peak wavelength and electron-donating character was found in Zn(ODZ)2 and These emitting materials showed blue or green fluorescence.Both Zn(PhPy)2 and Zn(quin)27 coordinate to zinc via imino Zn(TDZ)2. These complexes contain 3,4-diazole moieties with one donating group (O or S) at the 1 position. The PL peak nitrogen and hydoxyl oxygen. Zn(PhPy)2 consists of a polycyclic ligand while Zn(quin)2 consists of a fused ring ligand. of Zn(TDZ)2 is more red shifted than that of Zn(ODZ)2 by ca. 50 nm as the electron-donating character of the heteroatom The PL peaks of Zn(PhPy)2 and Zn(quin)2 occur at 455 and 542 nm, respectively.7 The PL peak of the polycyclic ligand increased. In both cases the PL spectrum was aVected strongly by the introduction of a sulfur atom with a broadening of the complex was at a shorter wavelength than that of the fused ring ligand complex.spectrum for Zn(BTZ)2 and a large red shift for Zn(BTZ)2. The nature of the metal center (Zn or Al ) did not aVect the PL spectra of Zn(BIZ)2, Zn(BOX)24 and Zn(BTZ)25, which contain an electron dontating group in each fused imino ring PL spectrum with the spectra of Zn(ODZ)2 and Al(ODZ)3 being the same. On the other hand, in chelete complexes were compared.Zn(BIZ)2 exhibited a sharp PL spectrum. The 2000 J. Mater. Chem., 1998, 8(9), 1999–2003Table 1 Photoluminescence (PL) and electroluminescence (EL) results for the complexes PL ELa max. luminance/cd m-2 emitting material PL, EL color lmax/nm Dl/nm lmax/nm Dl/nm (applied voltage/V) luminous eYciencyb/lm W-1 Zn(ODZ)2 blue 457 73 460 66 3000 (17.6) 0.45 Zn(TDZ)2 green 505 73 505 100 1000 (18) 0.13 Zn(PhPy)2 blue 455 56 483 91 5300 (14) 0.45 Zn(BIZ)2 blue 438 51 450 73 800 (19) 0.15 Al(ODZ)3 blue 455 73 460 100 400 (22) 0.08 Zn(BOX)2 c blue 478 112 Zn(BTZ)2 d whitish green 487 157 aEL device structure: [ITO/TPD (70 nm/emitting layer (70 nm)/Mg5Ag (180 nm)].bValue at luminance 100 cd m-2. cRef. 4. dRef. 5. having fused ring ligands (8-quinolinolate),7 the PL peak photonics PMA-11) upon irradiating with a UV lamp (l= 350 nm).The EL devices were fabricated by the vacuum wavelength is aVected by the coordinated metal [542 nm for Zn(quin)2 and 519 nm for Al(quin)3)]. deposition method. A 70 nm layer of TPD was deposited on a substrate consisting of glass slides precoated with indium tin The synthesized metal complexes were used as emitting materials of EL devices after conventional vacuum-vapor oxide (with sheet resistance of 15 V/%-1), by thermal evaporation under a vacuum of 10-4 Pa.A 70 nm layer of emitting deposition. The device structures were [ITO/hole transport layer (70 nm)/emitting layer(70 nm)/Mg5Ag(1051, 180 nm)]. material was then deposited on the TPD layer. A top electrode was subsequently deposited by coevaporation of Mg and Ag The hole transport layer was composed of TPD (N,N¾- dimethyl-N,N¾-m-ditolylbenzidine),9 and the emitting layer was (1051 atomic ratio) under a vacuum of 10-5 Pa.The luminance of the device was measured with a Minolta photometer composed of the synthesized metal complexes. These complexes formed uniform, smooth and clear films.No crystallization (nt-1/3° P). EL measurements were carried out under a nitrogen atmosphere. was found in the films over several months. The EL peak wavelengths lmax and half-widths (Dl) of the complexes are listed in Table 1. The PL and EL spectra were almost identical, Syntheses which implies that the EL emission originated from the metal 1-Benzoyl-2-(2-methoxybenzoyl )hydrazine 1.To an ice- complex layer in the EL device, and that no exiplex is formed cooled suspension of 15.34 g (0.11 mol) of benzoylhydrazide between the hole transport layer and the emitting layer to and 11.96 g (0.11 mol) of sodium carbonate in 184 g of dioxane reduce EL eYciency. The EL color of the devices was blue for was added 19.16 g (0.11 mol) of o-methoxybenzoyl chloride Zn(ODZ)2, Zn(PhPy)2, Zn(BIZ)2 and Al(ODZ)3 and green dropwise over 10 min with stirring under a nitrogen atmos- for Zn(TDZ)2. A blue color was characteristic for complexes phere.The reaction mixture was stirred for 1 h at room based on polycyclic ligand systems while complexes based on temperature (r.t.) then for 1 h at 90 °C. After cooling to r.t., to fused ring systems emit green light in most cases.7,10 The the mixture was added 400 ml of water.The reaction mixture maximum luminance and luminous eYciency of these devices was then filtered and washed with water. The white solid was are given in Table 1. The EL devices fabricated with Zn(ODZ)2, dried in vacuo to give 21.0 g (71%) of 1: mp 134 °C. 1H NMR Zn(TDZ)2 and Zn(PhPy)2 exhibited a high luminance (90 MHz,CDCl3) d 4.10 (s,3H), 7.18–6.96 (m,2H), 7.62–7.40 >1000 cd m-2.Zn(PhPy)2 showed the highest luminance (m,4H), 7.98–7.80 (m,2H), 8.19 (d,1H,J=8.4), 9.79 (br,2H). value of 5300 cd m-2. In luminance–current density measure- IR(KBr) 3260, 3050, 1630, 1475, 1290, 1240, 1100, 755, 705, ments of the EL devices, the luminance was found to be 595 cm-1. MS for C15H14N2O3: calc. 270(M), found 270. proportional to the injection current. The high luminance and charge injection eYciency of the EL devices imply electron 2-(2-Methoxyphenyl )-5-phenyl-1,3,4-oxadiazole 2. A mixture transporting ability of the complexes.7 of 5.5 g (20 mmol) of 1, 20 g of thionyl chloride and 0.3 g of In summary, we have synthesized metal–chelate complexes pyridine was stirred at 80 °C under a nitrogen atmosphere.based on polycyclic aromatic novel ligands containing After 4 h, the reaction mixture was cooled and poured onto NN(o-phenol) moieties as coordination sites. The EL and PL 200 g of crushed ice. The reaction mixture was then filtered wavelength of the polycyclic system was shorter than those of and washed with water. The white solid was dried in vacuo corresponding fused ring systems.The peak wavelengths of to give 5.2 g (quantitative) of 2: mp 91°C. 1H NMR the EL and PL spectra were red shifted by the introduction of (90 MHz,CDCl3) d 3.99 (s,3H), 7.00–7.18 (m,2H), 7.40–7.63 electron donating atoms into the imino ring of the ligands. (m,4H), 7.94–8.21 (m,3H). IR(KBr) 3050, 3000, 2950, 2850, Especially, sulfur was found to induce a large red shift.The 1605, 1535, 1475, 1270, 1020, 745, 710, 685 cm-1. MS for nature of the coordinated metal in the complexes based on C15H12N2O2: calc. 252(M), found 252. polycyclic systems did not aVect the PL spectrum. In EL devices these materials exhibited an electron transporting ability and emitted blue or green light. 2-(2-Hydroxyphenyl )-5-phenyl-1,3,4-oxadiazole (ODZ).To a solution of 5.13 g (20 mmol) of 2 in 50 ml of CH2Cl2 at -78 °C was added 3 ml (32 mmol) of BBr3 in 20 ml of CH2Cl2 Experimental dropwise over 10 min. The reaction mixture was allowed to warm slowly to room temperature. Water (50 ml ) was added All metal–chelate materials were purified by the train sublimation method reported by Wanger et al.11 Yields of to the reaction mixture carefully, then the reaction mixture was extracted twice with 30 ml of CH2Cl2.The organic layer 70–80% were obtained for the metal–chelate complexes. A Perkin-Elmer DSC7 instrument was used for the DSC was dried over Na2SO4. The solvent was removed by evaporation to yield a white solid. Recrystallization of the residue measurements and 1H NMR spectra were recorded at 90 MHz on a JEOL FX90Q spectrometer. Chemical shifts refer to from 200 ml of ethanol gave ODZ as a white crystalline solid: 3.89 g (82%): mp 165 °C. 1H NMR (90 MHz,CDCl3) d Me4Si as internal standard. IR spectra were recorded on a JASCO FT/IR-5M spectrophotometer. EL and PL spectra 6.95–7.23 (m,2H), 7.35–7.78 (m,4H), 7.87 (d,1H,J=8), 8.16 (m,2H), 10.19 (s,1H). IR(KBr) 3200, 3160, 3060, 1625, 1590, were measured with a multichannel analyzer (Hamamatsu J.Mater. Chem., 1998, 8(9), 1999–2003 20011540, 1490, 1410, 1240, 1070, 750, 710, 690 cm-1. MS for product was chromatographed (SiO2, 50% CHCl3–hexane) to yield BIZ as needles, 0.37g (10%): mp 119 °C. 1H NMR C14H10N2O2: calc. 238(M), found 238. (90MHz,CDCl3) d 6.55 (t,1H,J=8), 6.89 (d,1H,J=8), 7.01–7.50 (m,7H), 7.52–7.73 (m,3H), 7.75–7.91 (m,1H), 13.53 (br,1H). 2-(2-Methoxyphenyl )-5-phenyl-1,3,4-oxathiazole 3. A mixture of 5.5 g (20 mmol) of 1, 5.5 g (12 mmol) of P4S10 and IR(KBr) 3050, 2600, 1580, 1480, 1380, 1280, 1255, 810, 740, 700 cm-1. MS for C19H14N2O: calc. 286(M), found 286. 50 ml of xylene was stirred at 140 °C for 4 h. After cooling, 140 ml of water was added to the reaction mixture. This was heated at 80 °C for 1 h.Sodium hydrogen carbonate was added Bis[2-(2-hydroxyphenyl )-5-phenyl-1,3,4-oxadiazolato]zinc to the water layer until a pH of 6 was obtained. The xylene Zn(ODZ)2. To a suspension of 500 mg (2.1 mmol) of ODZ layer was washed with water and dried over Na2SO4. The and 5 g of methanol was added chloroform until the suspension solvent was removed by evaporation to yield a white solid.was clear. To this solution was added a solution of 230 mg The residue was dried in vacuo. The crude product was (1.0 mmol) of zinc acetate in 1.55 g of methanol and the chromatographed (SiO2, CHCl3) to yield 3 as a pale yellow mixture was heated at 70 °C for 1.5 h. The solvent was removed solid, 3.5 g (65%): mp 137 °C. 1H NMR (90 MHz,CDCl3) d by evaporation.The residue was washed with water and dried 4.02 (s,3H), 6.98–7.22 (m,2H), 7.35–7.59 (m,4H), 7.92–8.20 in vacuo to give a pale yellow solid. The crude materials were (m,2H), 8.52 (d,1H,J=8Hz). IR(KBr)3050, 3000, 2940, 2830, purified by the train sublimation method to give Zn(ODZ)2. 2700, 2530, 2560, 1595, 1450, 1410, 1300, 1260, 1015, 725, Anal. Calc. for C28H18N4O4Zn: C,62.30; H,3.36; N,10.38. 680 cm-1. MS for C15H12N2OS: calc. 268(M), found 268. Found: C,62.04; H,3.33; N,10.14%. 2-(2-Hydroxyphenyl )-5-phenyl-1,3,4-oxathiazole (TDZ). To Bis[2-(2-hydroxyphenyl )-5-phenyl-1,3,4-oxathiazolato]zinc a solution of 3.5 g (13 mmol) of 3 in 40 ml of CH2Cl2 at Zn(TDZ)2. To a suspension of 597 mg (2.3 mmol) of TDZ, -78 °C was added dropwise 17.5 ml of a solution of 1.0 M 204 mg (2.3 mmol) of morpholine and 5 g of methanol was BBr3 in CH2Cl2.The reaction mixture was allowed to warm added a solution of 258 mg (1.2 mmol) of zinc acetate and slowly to room temperature overnight. Water (50 ml ) was 1.7 g of methanol with stirring. The mixture was heated at added to the reaction mixture carefully, then the reaction 70 °C for 1 h and the solvent was removed.The residue was mixture was extracted twice with 30 ml of CH2Cl2. The organic washed with water and dried in vacuo. The crude materials layer was dried over Na2SO4. The solvent was removed by were purified by the train sublimation method to give evaporation to yield TDZ as a white solid: 1.75 g (53%): mp Zn(TDZ)2. Anal. Calc. for C28H18N4O2S2Zn: C,58.80; H,3.17; 128 °C. 1H NMR (90 MHz,CDCl3) d 6.85–7.43 (m,3H), N,9.80. Found: C,58.63; H,3.13; N,9.56%. 7.43–7.68 (m,4H), 7.91–8.08 (m,2H), 11.48 (s,1H). IR(KBr) 3060, 3030, 2920, 2730, 2630, 2570, 1600, 1455, 1430, 1310, Bis[2-(2-hydroxyphenyl )pyridinato]zinc Zn(PhPy)2. To a 1260, 1100, 1000, 750, 680, 600 cm-1. MS for C14H10N2OS: solution of 816 mg (4.8 mmol) of PhPy, 419 mg (4.9 mmol) of calc. 254(M), found 254.piperidine, 3 g of methanol and 10 g of chloroform was added a solution of 523 mg (2.4 mmol) of zinc acetate in 3.5 g of 2-(2-Methoxyphenyl )pyridine 4. To a mixture of 1.28 g methanol. The solvent was evaporated slowly for 5 days. The (53 mmol ) of Mg and 1 ml of THF was added a solution of residue was washed with water and dried in vacuo to give a 10 g (53 mmol ) of o-methoxybromobenzene in 15 ml of dry pale yellow solid.The crude materials were purified by the THF. The reaction mixture was refluxed for 30 min. This train sublimation method to give Zn(PhPy)2. Anal. Calc. for Grignard solution was added dropwise to a mixture of 7.5 g C22H16N2O2Zn: C,65.12; H,3.97; N,6.90. Found: C,64.70; (47 mmol ) of 2-bromopyridine, 0.25 g of 1,3-bis(diphenylphos- H,3.97; N,6.70%.phinopropane)nickel(II) chloride and 15 ml of dry THF over 10 min at 0 °C. The reaction mixture was allowed to warm to Bis[2-(2-hydroxyphenyl )-1-phenylbenzimidazolato]zinc room temperature and stirred overnight. To the reaction mixture Zn(BIZ)2. To a solution of 365 mg (1.3 mmol) of BIZ, 114 mg was added 30 ml of water and the solvent was then evaporated.(1.3 mmol) of piperidine and 2 g of methanol was added a The residue was partitioned between 70 ml of water and 50 ml solution of 140 mg (0.64 mmol) of zinc acetate in 0.92 g of of CHCl3. The organic layer was dried over Na2SO4 and methanol. The reaction mixture was stirred for 6 h. The evaporated to yield 4 as a light brown oil: 8.84 g (quantative). precipitates were filtered oV and washed with water and dried 1H NMR (90MHz,CDCl3) d 3.82 (s,3H), 6.91–7.50 (m,4H), in vacuo.The crude materials were purified by the train 7.55–7.90 (m,3H), 8.7 (d,1H,J=6Hz). IR(KBr) 3060, 3000, 2950, sublimation method to give Zn(BIZ)2. Anal. Calc. for 2840, 1600, 1585, 1500, 1460, 1425, 1260, 1240, 1025, 755 cm-1. C38H26N4O2Zn: C,71.76; H,4.12; N,8.81. Found: C,71.39; MS for C12H11NO: calc. 185(M), found 185.H,4.13; N,8.66%. 2-(2-Hydroxyphenyl )pyridine (PhPy). To a solution of 8.84 g Tris[2-(2-hydroxyphenyl )-5-phenyl-1,3,4-oxadiazolato]- (48 mmol) of 4 in 50 ml of CH2Cl2 at -78 °C was added aluminium Al(ODZ)3. To a solution of 1.00 g (4.2 mmol) of dropwise 49.5 ml of a solution of 1.0 M BBr3 in CH2Cl2. The ODZ and 0.75 g (8.6 mmol) of morpholine in 190 g of ethanol reaction mixture was allowed to warm slowly to room temperawas added a solution of 0.187 g (1.4 mmol) of aluminium ture overnight.Water (60 ml ) was added to the reaction chloride in 3.15 g of ethanol. The reaction mixture was heated mixture carefully, then the reaction mixture was extracted at 80 °C for 4 h. The precipitates were filtered oV and washed twice with 100 ml of CH2Cl2.The organic layer was dried over with water and dried in vacuo. The crude materials were Na2SO4. The solvent was removed by evaporation to yield purified by the train sublimation method to give Al(ODZ)3. PhPy as a light brown oil: 3.35g (41%). 1H NMR Anal. Calc. for C42H27N6O6Al: C,68.29; H,3.68; N,11.38. Found: (90 MHz,CDCl3) d 6.79–7.35 (m,4H), 7.65–7.95 (m,3H), 8.52 C,68.32; H,4.08; N,10.96%. (d,1H,J=6), 14.35 (br,1H).IR(KBr) 3400, 3060, 2950, 2840, 1600, 1505, 1495, 1260, 1025, 830, 760 cm-1.MS for C11H9NO: calc. 171(M), found 171. One of the authors would like to thank Mr. Masao Tsuji (Toyota Central R&D Labs. Inc.) for the elemental analysis, Mrs. Atsuko Takagi (Toyota Central R&D Labs. Inc.) for the 2-(2-Hydroxyphenyl )-1-phenylbenzimidazole (BIZ).A mixture of 2.0 g (13 mmol ) of o-anisic acid, 2.42 g (13 mmol) of N- DSC measurements, Mr. Tadao Ogawa (Toyota Central R&D Labs. Inc.) for the MS measurements and Mr. Koji Noda phenyl-1,2-phenylenediamine and 5 ml of o-dichlorobenzene was heated at 170 °C for 10 h, then at 220 °C for 30 h. The crude (Toyota Central R&D Labs. Inc.) for the device fabrication. 2002 J. Mater. Chem., 1998, 8(9), 1999–20034 N. Nakamura, S. Wakabayashi, K. Miyairi and T. Fujii, Chem. References L ett., 1994, 1741. 1 J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, 5 Y. Hamada, T. Sano, T. Fujii, Y. Nishio, H. Takahashi and K. M. Makay, R. H. Friend, P. L. Burns and A. B. Holmes, Nature, K. Shibata, Jpn. J. Appl. Phys., 1996, 35, L1339. 1990, 347, 539; G. Gustafsson, Y. Cao, G. M. Treacy, F. Klavettefi 6 K. Tamao, S. Kodama, I. Nakajima and M. Kumada, T etrahedron, N. Colaneri and A. J. Heeger, Nature, 1992, 357, 477; C. Adachi, 1982, 38, 3347. T. Tsutsui and S. Saito, Appl. Phys. L ett., 1989, 55, 1489; J. Kido, 7 P. E. Burrows, L. S. Sapochak, D. M. McCarty, S. R. Forrest and M. Kohda, K. Okuyama and K. Nagai, Appl. Phys. L ett., 1992, M. E. Thompson, Appl. Phys. L ett., 1994, 64, 2718. 61, 761. 8 CRC Handbook of Chemistry and Physics, ed. R. C. West, CRC 2 C. W. Tang, S. A. Van Slyke and C. H. Chen, J. Appl. Phys., 1989, Press, Ohio, 57th edn., 1976, p. E-68. 65, 3610; J. Kido, M. Kohda, K. Hongawa, K. Okuyama and 9 C. Adachi, T. Tsutsui and S. Saito, Appl. Phys. L ett., 1989, 55, 1489. K. Nagai, Mol. Cryst. L iq. Cryst., 1993, 227, 277; C. Hosokawa, 10 Y. Hamada, T. Sano, M. Fujita, T. Fujii, Y. Nishio and K. Shibata, N. Kawasaki, S. Sakamoto and T. Kusumoto, Appl. Phys. L ett., Chem. L ett., 1993, 905. 1992, 61, 2503; C. Adachi, T. Tsutsui and S. Saito, Appl. Phys. L ett., 11 J. Wanger, R. O. Loufty and C. K. Hiao, J. Mater. Sci., 1982, 1990, 56, 799. 17, 2781. 3 C. W. Tang and S. A. Van Slyke, Appl. Phys. L ett., 1987, 51, 913. Paper 8/03308K; Received 1st May, 1998 J. Mater. Chem., 1998, 8(9), 1999–2003 2003

ISSN:0959-9428    

DOI:10.1039/a803308k

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 Synthesis and properties of highly soluble third-order optically nonlinear chromophores and methacrylate monomer based on distyrylbenzene M. S. Wong,*a,b†M. Samoc,a A. Samoc,a B. Luther-Daviesa and M. G. Humphreyb aAustralian Photonics Cooperative Research Centre, L aser Physics Centre, Research School of Physical Sciences and Engineering, T he Australian National University, Canberra, ACT 0200, Australia bDepartment of Chemistry, T he Australian National University, Canberra, ACT 0200, Australia A novel series of highly soluble distyrylbenzenes bearing poly(alkyleneoxy) and/or alkylsulfonyl solubilizing substituents has been synthesized using the stereoselective Wadsworth–Emmons reaction as a key step.The corresponding poly(alkyleneoxy) alkylsulfonyl 4,4¾-disubstituted distyrylbenzene has been covalently incorporated into an acrylic monomer which has also been successfully co-polymerized with methyl methacrylate (MMA).All the 4,4¾-disubstituted distyrylbenzenes and the polyacrylate based co-polymer show excellent solubility, processibility and thermal stability. Importantly, 2-(2-butoxyethoxy)ethoxyhexylsulfonyl 4,4¾-disubstituted distyrylbenzene exhibits only a minor bathochromic shift of the absorption maximum compared to those of the corresponding symmetrically 4,4¾-disubstituted distyrylbenzenes; however, its third-order optical nonlinearity derived from Z-scan measurements at 800 nm is approximately twice as large as those of its symmetrically 4,4¾-disubstituted counterparts.Poly(phenylenevinylene) (PPV) is one of the widely (non)linear optical and thermal properties of PPV-based materials are greatly dependent upon the degree of p-conjugation, investigated p-conjugated polymeric materials exhibiting technologically useful multi-functional properties such as a large the co-planarity of the conjugated system and the electronic properties of the substituents.Towards the goal of understand- third-order optical nonlinearity,1 an eYcient electroluminescent response,2 a high conductivity upon doping and a laser ing and establishing the structure–property relationship of OPVs, we report here our initial studies on the synthesis of a emission.4 Unsubstituted PPV is insoluble in organic solvents and thin-films of this polymer are traditionally prepared by novel series of highly soluble three-phenyl-ring OPVs (distyrylbenzenes) bearing poly(alkyleneoxy) electron-donor(s) and/or spin-casting of a water soluble polysulfonium precursor followed by pyrolysis.5,6 However, the polydispersity and the hexylsulfonyl electron-acceptor(s).In addition, the influence of substitution on various physical properties including the poor structural perfection that are incurred in this preparation procedure may aVect the optical and electrochemical properties absorption spectroscopic properties, the third-order nonlinear optical properties and the thermal behaviour will be discussed. of the polymeric thin-film.Over the last few years, various soluble PPV derivatives and copolymers have been synthesized For practical applications, the active chromophores or OPVs are required to be developed in the form of either side- to enhance the solubility and processibility as well as to tune the desirable properties.7–9 With the solubilizing substituents chain or main-chain polymers in order to obtain better miscibility, a larger chromophore density and a higher thermal influencing the electronic properties of the PPV chain, e.g. via induction and steric eVects, and with the various compositions stability. Towards this end, we also report the synthesis of the acrylic monomer and polyacrylate bearing the corresponding of the copolymers, the physical properties of the PPV-based materials are greatly varied.As a result, the exploration and donor–acceptor disubstituted distyrylbenzene unit.development of eYcient and useful PPV-based materials for a specific application still present a great challenge for the scientific communities. Results and Discussion It has recently been shown that highly p-conjugated PPV In general, any solubilizing substituent grafted on the lateral thin-films possess a large eVective nonlinear refractive index side of the main chain of an oligomer (including a vinylic and a desirable two-photon merit factor for all-optical switchlinkage) will greatly distort the co-planarity of the p-conjugated ing.10 This suggests that PPV-based materials may be potential system, leading to a decrease in conjugation and in desirable candidates for applications in all-optical signal processing. It physical properties.Therefore, in this study, the solubilizing has also been shown that the third-order nonlinear optical substituents will only be placed at the ends of the main chain properties of PPV originate from the relatively short pof an oligomer. Although several symmetrically disubstituted conjugated chain segments of the polymer.11 Therefore, it is of distyrylbenzenes were previously synthesized for spectroscopic interest to understand and establish the structure–property studies we have found that their solubilities were often too relationship of the well-defined oligo-phenylenevinylenes low to perform any reliable nonlinear optical measurement in (OPV) in order to further explore and optimize the PPV-based solution [e.g.the solubility of 1,4-bis(-4-bromostyryl)benzene materials for photonic applications.Unfortunately, there are in chloroform is less than 0.1% w/w). To overcome this, the still no reliable guidelines for optimizing the second molecular poly(alkyleneoxy) and alkylsulfonyl moieties were incorporated hyperpolarizability, c, of an organic molecule, in contrast to into the ends of the main chain to act as solubilizing as well the first molecular hyperpolarizability, b.In addition, the as electron-donating and electron-withdrawing substituents, respectively. The stereoselective Wadsworth–Emmons reaction was used †Present address: Department of Chemistry, Hong Kong Baptist University, Hong Kong. to construct the trans carbon–carbon double bonds. The J. Mater. Chem., 1998, 8(9), 2005–2009 2005BrCH2 CH2Br + P(OC2H5)3 CH 2 CH 2 P O (C2H5O)2 P O (OC2H5)2 3 F CHO R H R CHO 4 R = C6H13S 5 R = C4H9O(C2H4O)2 6 R¢ CHO 1 2 + 7 R = C6H13S 8 R = C4H9O(C2H4O)2 9 R¢ = C6H13SO2 3 + 8 (OC2H4)2OC4H9 C4H9O(C2H4O)2 10 3 + 9 SO2C6H13 C6H13O2S 11 CH 2 C4H9O(C2H4O)2 P O (OC2H5)2 + 9 12 SO2C6H13 C4H9O(C2H4O)2 i ii iii iv iv iv 13 Scheme 1 Reagents and conditions: i, 150 °C; ii, K2CO3, DMSO, 150 °C; iii, MCPBA, CH2Cl2, 0°C; iv, NaH, DME, room temp.general scheme for the synthesis of symmetrically and unsym- with methacrylic anhydride in the presence of dimethylaminopyridine (DMAP) as catalyst and triethylamine as base metrically 4,4¾-disubstituted distyrylbenzenes is summarized in Scheme 1. The bis-phosphonate 3 was prepared by the aVorded the methacrylate monomer 19.To demonstrate the ability of this monomer to be polymerized, thermal radical Michaelis–Arbuzov reaction using a,a¾-dibromo-p-xylene 1 and triethyl phosphite 2. The solubilizing hexylsulfanyl and 2-(2- polymerization conditions were employed to co-polymerize the methacrylate monomer 19 with MMA using the procedure butoxyethoxy)ethoxy substituents, were introduced by aromatic nucleophilic substitution of 4-fluorobenzaldehyde 6 described by Robello.13 With an 8.3% w/w dye content in the feed composition, the estimated weight average of copolymer aVording the corresponding aldehydes 7 and 8, respectively.The Wadsworth–Emmons reaction of the bis-phosphonate 3 derived from gel permeation chromatography (GPC) is 35 000 g mol-1 with Tg=124 °C.More detailed characteriz- with 2 equiv. of aldehyde 8 yielded the corresponding symmetrical distyrylbenzene 10. Surprisingly, the yellow product ation and the physical properties of the copolymer will be reported elsewhere. obtained from the Wadsworth–Emmons reaction of 3 and 2 equiv. of 7 was not soluble enough to be characterized analyti- The results of the electronic absorption measurements, the third-order nonlinearities determined by Z-scan measurements cally by the conventional spectroscopic techniques.However, the oxidation of this yellow solid with 3-chloroperoxybenzoic using 800 nm irradiation wavelength and the thermal properties are summarized in Table 1. As seen from the electronic acid (MCPBA) aVorded the highly soluble distyrylbenzene 11. Alternatively, the hexylsulfanyl functionality of 7 could be absorption spectra (Fig. 1), all the compounds show strong and intense low-lying absorption bands which indicate a highly converted into the hexylsulfonyl functionality by means of MCPBA oxidation. Subsequent double Wadsworth–Emmons p-conjugated system. Unlike the donor–acceptor 4,4¾-disubstituted distyrylbenzene 13, the low-lying absorption bands of reactions also gave the desired distyrylbenzene 11.To synthesize the donor–acceptor disubstituted distyrylbenzene 13, the the symmetrically 4,4¾-disubstituted distyrylbenzenes 10 and 11 are apparently composed of several vibronic electronic precursor mono-phosphonate 12 was first prepared from the reaction of the aldehyde 8 and excess of the bis-phosphonate transitions.14 There are also significant bathochromic shifts of the absorption maxima lmax of the 4,4¾-disubstituted distyryl- 3, followed by Wadsworth–Emmons reaction with aldehyde 9.It was found that reversing the reaction sequence did not benzenes compared to that of the unsubstituted distyrylbenzene (Dl=12–19 nm),12 in spite of the moderate donating and aVord the desired product, as the corresponding monophosphonate could not be prepared directly from this method.withdrawing strength of the 2-(2-butoxyethoxy)ethoxy and hexylsulfonyl functionalities, respectively. Such a red shift is The synthetic route for the preparation of the methacrylate monomer bearing the poly(alkyleneoxy) alkylsulfonyl 4,4¾- consistent with the fact that the donating and withdrawing substituents at the para-positions of distyrylbenzene enhance disubstituted distyrylbenzene is outlined in Scheme 2.A similar synthetic approach as used above was adopted to prepare the the p-electron delocalization along the entire unsaturated system. On the other hand, the donor–acceptor 4,4¾-disubsti- functionalized poly(alkyleneoxy) alkylsulfonyl 4,4¾-disubstituted distyrylbenzene 18. The reaction of the chromophore 18 tuted distyrylbenzene 13 exhibits only a small bathochromic 2006 J.Mater. Chem., 1998, 8(9), 2005–2009SO2C10H21 (C2H4O)2 18 i C10H21S CHO 14 + 3 CH 2 C10H21S P O (OC2H5)2 15 CH 2 C10H21SO2 P O (OC2H5)2 16 + (C2H4O)2 CHO HO 17 HO SO2C10H21 (C2H4O)2 19 O O ii i iii Scheme 2 Reagents and conditions: i, NaH, DME, room temp.; ii, MCPBA, CH2Cl2; iii, DMAP, Et3N, methacrylic anhydride, room temp.shift of lmax compared to those of the symmetrically 4,4¾- disubstituted analogues (Dl=5–7 nm). This is presumably due Table 1 Summary of the measured physical properties of 10, 11, and 13 to the strong inductive nature of the hexylsulfonyl group. lmax/nm transition To evaluate the third-order molecular nonlinearities, the (e/10-4 M-1 creal/ cimag/ temperature/ Z-scan technique15 was employed as it can determine the sign, Compound cm-1) 10-36 esu 10-36 esu °C the real part of the molecular nonlinearity creal (which relates to the refractive nonlinearity) and the imaginary part of the 10 368 (7.78) 200 54 46, 225 molecular nonlinearity cimag (due to the multi-photon absorp- 11 370 (7.62) 250 95 231, 259 13 375 (5.97) 420 200 216 tion).All the 4,4¾-disubstituted distyrylbenzenes exhibit positive creal values. In spite of the very diVerent electronic nature of the 2-(2-butoxyethoxy)ethoxy and hexylsulfonyl functionalities, both of the symmetrically 4,4¾-disubstituted distyrylbenzenes 10 and 11 show comparable creal values. On the other hand, the donor–acceptor 4,4¾-disubstituted distyrylbenzene 13 exhibits an enhanced creal compared to those of its symmetrically disubstituted counterparts 10 and 11.This confirms the advantage of using conjugated donor and acceptor substituents to enhance c,16 in particular the poly(alkyleneoxy) alkylsulfonyl pair which have been shown to provide an excellent transparency –nonlinearity trade-oV. In contrast to their creal values, bis(hexylsulfonylstyryl )benzene 11 shows a substantially higher cimag than that of bis[2-(2-butoxyethoxy)ethoxystyryl]- benzene 10.The cimag value of the donor–acceptor 4,4¾-disubstituted distyrylbenzene 13 is also relatively large, presumably due to the relatively strong two-photon absorption at irradiation wavelength. With the incorporation of the solubilizing substituents, the distyrylbenzene derivatives and the corresponding methacrylate monomer show enhanced solubility in various solvents; in particular, the donor–acceptor 4,4¾-disubstituted distyrylbenzene 13 has a solubility of more than 6% w/w in chloroform.The solubilizing substituents also induce liquid crystalline phases of the symmetrically 4,4¾-disubstituted distyrybenzenes (10 and 11).Experimental All the new compounds were fully characterized with standard Fig. 1 Electronic absorption spectra of (+) 10, (#) 11 and (6) 13 measured in chloroform spectroscopic techniques. All the physical measurements were J. Mater. Chem., 1998, 8(9), 2005–2009 2007performed in CHCl3. 1H NMR spectra were recorded using a 170 mg, 0.34 mmol) in anhydrous DME was slowly added 1.2 equiv.of NaH (15 mg, 0.42 mmol) at 0 °C. After stirring for Varian Gemini-300 FT NMR spectrometer and are referenced to the residual CHCl3 (7.24 ppm). Infrared spectra were 0.5 h at 0 °C, the reaction mixture was slowly warmed up to room temp. After stirring for 2 h at room temp., the solution recorded using a Perkin-Elmer System 2000 FT-IR spectrometer. Electronic absorption (UV–VIS) spectra were was quenched with water.The crude product was either collected by suction filtration or extracted twice with CH2Cl2, recorded using a Shimadzu UV-3101PC Spectrophotometer. Thermal properties were determined by diVerential scanning dried over anhydrous MgSO4 and evaporated to dryness. The crude product was then purified by silica gel chromatography calorimetry (DSC) using a Shimadzu Thermal Analysis System TA-50ASI with a heating rate of 10 °Cmin-1.The reported using the gradient elution technique with CH2Cl2–ethyl acetate as eluent [aVording 110 g (54%) of 13]. For the double temperatures were the peak temperature of the traces obtained from the rerun. Molecular weight of the co-polymer was Wadsworth–Emmons reactions, a 251 mixture of aldehyde (i.e. 9: 350 mg, 1.37 mmol) and phosphonate ester (i.e. 3: 260 mg, estimated by gel permeation chromatography (GPC) using a Spectra Physics HPLC instrument equipped with a Jordi 0.69 mmol) was used [aVording 170 mg (43%) of 11]. Mixed-Bed GPC column. THF was used as the eluent with toluene as an internal standard. The polystyrene standards 1,4-Bis{4-[2-(2-butoxyethoxy)ethoxy]styryl}benzene 10. 1H NMR (300 MHz, CDCl3) d 7.44 (s, 4H), 7.42 (d, J 8.91 Hz, were used for the calibration. Z-Scan measurements were performed with a system 4H), 7.05 (d, J 16.20 Hz, 2H), 6.94 (d, J 16.30 Hz, 2H), 6.89 (d, J 8.79 Hz, 4H), 4.15 (t, J 4.88 Hz, 4H), 3.86 (t, J 4.88 Hz, 4H), consisting of a Coherent Mira Ar-pumped Ti-sapphire laser generating a mode-locked train of approximately 100 fs 800 nm 3.71 (m, 4H), 3.60 (m, 4H), 3.46 (t, J 6.72 Hz, 4H), 1.55 (m, 4H), 1.35 (m, 4H), 0.90 (t, J 7.28 Hz, 6H).MS (EI) m/z 602.3 pulses and a Ti-sapphire regenerative amplifier pumped by a Q-switched pulsed YAG laser at 30 Hz. The open- and closed- (M+). HRMS (EI) C38H50O6: calc. 602.3607 found, 602.3628. nmax(CH2Cl2)/cm-1 3029, 2934, 2875, 1605, 1516, 1457, 1250, aperture Z-scans were recorded at two or three concentrations for each compound, and the real and imaginary part of the 1176, 1111, 1065.Crystal–mesophase=46 °C, mesophase– isotropic=225 °C. Found: C, 75.60; H, 8.59. C38H50O6 requires nonlinear phase shift was determined by numerical fitting. The real and imaginary parts of the hyperpolarizability of the C, 75.72; H, 8.34%.solute were calculated by assuming a linear concentration dependence of the solution susceptibility. The nonlinearities 1,4-Bis[4-(hexylsulfonyl )styryl]benzene 11. 1H NMR (300 MHz, CDCl3) d 7.87 (d, J 8.46 Hz, 4H), 7.66 (d, J 8.52 Hz, and light intensities were calibrated using measurements of a 1 mm thick silica plate for which the nonlinear refractive index 4H), 7.55 (s, 4H), 7.25 (d, J 16.32 Hz, 2H), 7.15 (d, J 16.26 Hz, 2H), 3.08 (t, J 8.1 Hz, 4H), 1.70 (m, 4H), 1.34 (m, 4H), 1.24 n2=3×10-16 cm2 W-1 was assumed.(m, 8H), 0.84 (t, J 6.84 Hz, 6H). MS (EI) m/z 578.2 (M+). General procedure for the aromatic nucleophic substitution HRMS (EI) C34H42O4S2: calc. 578.2525, found 578.2523. nmax(CH2Cl2)/cm-1 3018, 2930, 2859, 1592, 1511, 1466, To an equimolar solution of 4-fluorobenzaldehyde 6 (2 ml, 1310, 1142, 1089.Crystal–mesophase=231 °C, mesophase– 18.6 mmol) and the corresponding solubilizing substituent (e.g. isotropic=259 °C. Found: C, 70.47; H, 7.17; S, 10.94. 4: 2.6 ml, 18.6 mmol) in DMSO was added 2 equiv. of Na2CO3 C34H42O4S2 requires C, 70.55; H, 7.31; S, 11.08%. (i.e. 5.2 g, 37.3 mmol). The mixture was heated at 150 °C for 24 h under N2.After cooling to room temp., the reaction Diethyl 4-{4-[2-(2-butoxyethoxy)ethoxy]styryl}benzylphosmixture was poured into water and extracted twice with phonate 12. 1H NMR (300 MHz, CDCl3) d 7.40 (d, J 8.79 Hz, CH2Cl2, dried over anhydrous MgSO4 and evaporated to 4H), 7.24 (dd, J 8.19 Hz, J 2.40 Hz, 2H), 7.01 (d, J 16.47 Hz, dryness. The crude product was then purified by silica gel 2H), 6.89 (d, J 16.41 Hz, 1H), 6.88 (d, J 8.70 Hz, 2H), 4.12 (t, chromatography using the gradient elution technique with J 5.22 Hz, 2H), 3.99 (m, 4H), 3.84 (t, J 4.98 Hz, 2H), 3.70 (m, CH2Cl2–ethyl acetate as eluent [aVording 3.79 g (91%) of 7]. 2H), 3.58 (m, 4H), 3.58 (m, 2H), 3.44 (t, J 6.72 Hz, 2H), 3.12 (d, J 21.69 Hz, 1H), 1.55 (m, 2H), 1.32 (m, 2H), 1.22 (t, J 4-Hexylsulfanylbenzaldehyde 7. 1H NMR (300 MHz, CDCl3) 6.99 Hz, 6H), 0.89 (t, J 7.41 Hz, 3H). d 9.89 (s, 1H), 7.73 (d, J 8.55 Hz, 2H), 7.31 (d, J 8.31 Hz, 2H), 2.97 (t, J 7.37 Hz, 2H), 1.68 (m, 2H), 1.29 (m, 2H), 1.28 (m, 1-{4-[2-(2-Butoxyethoxy)ethoxy]styryl}-4-(4-hexylsulfonyls- 4H), 0.86 (t, J 7.11 Hz, 3H). tyryl )benzene 13. 1H NMR (300 MHz, CDCl3) d 7.85 (d, J 8.49 Hz, 2H), 7.64 (d, J 8.52 Hz, 2H), 7.49 (s, 4H), 7.43 (d, J 4-[2-(2-Butoxyethoxy)ethoxy]benzaldehyde 8. 1H NMR 8.79 Hz, 2H), 7.23 (d, J 16.38 Hz, 1H), 7.10 (d, J 16.41 Hz, (300 MHz, CDCl3) d 9.85 (s, 1H), 7.79 (d, J 8.64 Hz, 2H), 6.99 1H), 7.09 (d, J 16.41 Hz, 1H), 6.95 (d, J 16.26 Hz, 1H), 6.90 (d, (d, J 8.82 Hz, 2H), 4.19 (t, J 4.79 Hz, 2H), 3.86 (t, J 4.79 Hz, J 8.79 Hz, 2H), 4.14 (t, J 4.92 Hz, 2H), 3.86 (t, J 4.94 Hz, 2H), 2H), 3.70 (m, 2H), 3.58 (m, 2H), 3.43 (t, J 6.72 Hz, 2H), 1.53 3.71 (m, 2H), 3.59 (m, 2H), 3.46 (t, J 6.72 Hz, 2H), 3.07 (t, J (m, 2H), 1.32 (m, 2H), 0.87 (t, J 7.80 Hz, 3H). 8.10 Hz, 2H), 1.70 (m, 2H), 1.56 (m, 2H), 1.33 (m, 4H), 1.24 (m, 4H), 0.90 (t, J 7.35 Hz, 3H), 0.84 (t, J 6.87 Hz, 3H).MS 4-Hexylsulfonylbenzaldehyde, 9 (EI) m/z 590.2 (M+).HRMS (EI) C36H46O5S: calc. 590.3066, To a stirred solution of 7 (400 mg, 1.8 mmol) in CH2Cl2 at found 590.3077. nmax(CH2Cl2)/cm-1 3029, 2960, 2933, 1591, 0 °C was slowly added MCPBA (650 mg, 3.6 mmol). After 1514, 1459, 1307, 1250, 1176, 1143, 1111, 1089. Mp=216 °C. stirring for 1 h, the white suspension was filtered oV and the Found: C, 73.20; H, 8.12; S, 5.16. C36H46O5S requires C, 73.19; filtrate was washed with Na2CO3 solution, dried over anhy- H, 7.85; S, 5.43%.drous MgSO4 and evaporated to dryness. The crude product was then purified by silica gel chromatography using the Diethyl 4-(4-decylsulfanylstyryl )benzylphosphonate 15. 1H gradient elution technique with CH2Cl2–ethyl acetate as eluent NMR (300 MHz, CDCl3) d 7.43 (d, J 8.79 Hz, 2H), 7.40 (d, J aVording 350 g of 9 in 76% yield. 1H NMR (300 MHz, CDCl3) 8.61 Hz, 2H), 7.27 (d, J 8.37 Hz, 2H), 7.02 (s, 2H), 4.00 (m, d 10.12 (s, 1H), 8.07 (s, 4H), 3.10 (t, J 8.09 Hz, 2H), 1.69 (m, 4H), 3.14 (d, J 21.72 Hz, 2H), 2.91 (t, J 7.38 Hz, 2H), 1.62 (m, 2H), 1.34 (m, 2H), 1.23 (m, 4H), 0.83 (t, J 6.78 Hz, 3H). 2H), 1.40 (m, 2H), 1.23 (m, 20H), 0.86 (d, J 6.74 Hz, 3H). MS (EI) m/z 502.1 (M+).General procedure for theWadsworth–Emmons reaction Diethyl 4-(4-decylsulfonylstyryl )benzylphosphonate 16. 1H To an equimolar solution of an aldehyde (i.e. 9: 89mg, 0.34 mmol) and the corresponding phosphonate ester (i.e. 12: NMR (300 MHz, CDCl3) d 7.85 (d, J 8.46 Hz, 2H), 7.63 (d, J 2008 J. Mater. Chem., 1998, 8(9), 2005–20098.55 Hz, 2H), 7.47 (d, J 7.92 Hz, 2H), 7.30 (dd, J 8.31 Hz, J developed.The thermal free-radical copolymerization of distyrylbenzene- derived methacrylate monomer withMMA has been 2.46 Hz, 2H), 7.21 (d, J 16.47 Hz, 1H), 7.08 (d, J 16.26 Hz, 1H), 4.01 (m, 4H), 3.14 (d, J 21.96 Hz, 2H), 3.06 (t, J shown to be achievable. All distyrylbenzene derivatives show excellent solubility, processibility and thermal stability. In 8.01 Hz, 2H), 1.69 (m, 2H), 1.23 (m, 22H), 0.86 (d, J 6.75 Hz, 3H).MS (EI) m/z 534.1 (M+). addition to enhanced third-order nonlinearity, the 2-(2-butoxyethoxy) ethoxyhexylsulfonyl 4,4¾-disubstituted distyrylbenzene exhibits an excellent transparency–nonlinearity trade-oV 1-{4-[2-(2-Hydroxyethoxy)ethoxy]styryl}-4-(4-decylcompared to those of the symmetrically 4,4¾-disubstituted sulfonylstyryl )benzene 18. 1H NMR (300 MHz, CDCl3) d 7.85 distyrylbenzenes. (d, J 8.46 Hz, 2H), 7.65 (d, J 8.46 Hz, 2H), 7.50 (s, 4H), 7.45 (d, J 8.64 Hz, 2H), 7.23 (d, J 16.24 Hz, 1H), 7.11 (d, J 16.20 Hz, We gratefully acknowledge the Australian Photonics 1H), 7.09 (d, J 16.26 Hz, 1H), 6.96 (d, J 16.41 Hz, 1H), 6.91 (d, Cooperative Research Centre for financial support and J 8.82 Hz, 2H), 4.15 (s, 2H), 3.87 (s, 2H), 3.76 (m, 2H), 3.67 N.T. Lucas for the GPC analysis. M.G.H. is an ARC (m, 2H), 3.07 (t, J 7.80 Hz, 2H), 1.70 (m, 2H), 1.21 (m, 16H), Australian Research Fellow. 0.85 (t, J 6.69 Hz, 3H). MS (FAB) m/z 590.1 (M+). Synthesis of 19 References To a stirred solution of 18 (240 mg, 0.41 mmol), freshly distilled 1 B. Luther-Davies and M. Samoc, Curr. Opin. Solid State Phys., methacrylic anhydride (75 mg, 0.49 mmol) and 3-tert-butyl-4- 1997, 2, 213.hydroxy-5-methylphenyl sulfide (15 mg, 0.04 mmol) in anhy- 2 J. Burroughs, D. D.Bradley, A. R. Brown, R. N. Marks, K. Mackey, drous CH2Cl2 at room temp. was slowly added a solution of R. H. Friend, P. L. Burns and A. B. Holmes, Nature, 1990, 347, 539. 3 M. Hirooka, I. Murase, T. Ohnishi and T. Noguchi, in Frontiers of DMAP (6 mg, 0.04 mmol) and Et3N (49 mg, 0.49 mmol) in Macromolecular Science, ed.T. Saegusa, Blackwell Scientific CH2Cl2. The reaction was monitored by TLC. After complete Publications, Oxford, UK, 1989, p. 425. disappearance of 18 via TLC, the reaction mixture was 4 F. Hide, M. A. Diaz-Garcia, B. J. Schwartz and A. I. Heeger, Acc. quenched with water and extracted twice with CH2Cl2, dried Chem. Res., 1997, 30, 430.over anhydrous MgSO4 and evaporated to dryness. The crude 5 D. R. Gagnon, J. D. Capistran, F. E. Karasz and R. W. Lenz, product was purified by silica gel chromatography using the Polym. Bull., 1984, 12, 293. 6 S. Antoun, F. E. Karasz and R. W. Lenz, J. Polym. Sci., Part A, gradient elution technique with CH2Cl2–ethyl acetate as eluent, 1988, 26, 1809.aVording 125 mg (47%). 1H NMR (300 MHz, CDCl3) d 7.85 7 Z. Bao, Y. Chen, R. Cai and L. Yu,Macromolecules, 1993, 26, 5281. (d, J 8.46 Hz, 2H), 7.65 (d, J 8.55 Hz, 2H), 7.50 (s, 4H), 7.44 8 R. M. Gurge, A. Sarker, P. M. Lahti, B. Hu and F. E. Karasz, (d, J 8.79 Hz, 2H), 7.23 (d, J 16.38, 1H), 7.11 (d, J 16.47, 1H), Macromolecules, 1996, 29, 4287. 7.09 (d, J 16.41, 1H), 6.96 (d, J 16.35 Hz, 1H), 6.90 (d, J 9 T.Maddux, W.Li and L. Yu, J. Am. Chem. Soc., 1997, 119, 844. 8.82 Hz, 2H), 6.12 (m, 1H), 5.56 (m, 1H), 4.33 (t, J 4.80 Hz, 10 A. Samoc, M. Samoc, M. WoodruV and B. Luther-Davies, Opt. L ett., 1995, 20, 1241. 2H), 4.14 (t, J 4.59 Hz, 2H), 3.87 (t, J 4.77 Hz, 2H), 3.82 (t, J 11 C. Bubeck, in Organic T hin Films for Waveguiding Nonlinear 4.86 Hz, 2H), 3.07 (t, J 7.95 Hz, 2H), 1.93 (m, 3H), 1.70 (m, Optics, ed. F. Kajzar and J. D. Swalen, Gordon and Breach 2H), 1.21 (m, 16H), 0.85 (t, J 6.42 Hz, 3H). MS (EI) m/z 658.3 Publishers, Amsterdam, The Netherlands, 1996, p. 137. (M+). HRMS (EI) C60H50O6S: calc. 658.3328, found 658.3336. 12 S. Nakatsuji, K. Matsuda, Y. Uesugi, K. Nakashima, S. Akiyama, nmax(CH2Cl2)/cm-1 3029, 2929, 2857, 1718, 1591, 1514, 1299, G. Katzer and W. Fabian, J. Chem. Soc., Perkin T rans. 2, 1991, 861. 1250, 1176, 1141, 1089. Mp=213 °C; decomp. temp.=290 °C. 13 D. R. Robello, J. Polym. Sci., Part A: Polym. Chem., 1990, 28, 1. 14 N. N. Barashkov, D. J. Guerrero, H. J. Olivos and J. P. Ferrais, Synth.Met., 1995, 75, 153. Conclusions 15 M. Samoc, A. Samoc, B. Luther-Davies, Z. Bao, L. Yu, B. Hsieh and U. Scherf, J. Opt. Soc. Am. B, 1998, 15, 817. A novel series of poly(alkyleneoxy) and/or alkylsulfonyl 4,4¾- 16 C. Bosshard, R. Spreiter, P. Gu� nter, R. R. Tykwinski, M. Schreiber disubstituted distyrylbenzenes has been synthesized. and F. Diederich, Adv.Mater., 1996, 8, 231. Furthermore, the distyrylbenzene-derived methacrylate monomer has been synthesized by adopting the synthetic route Paper 8/03229G; Received 29th April, 1998 J. Mater. Chem., 1998, 8(9), 2005–2009 20

ISSN:0959-9428    

DOI:10.1039/a803229g

出版商:RSC    

年代:1998

数据来源: RSC