GENERAL DISCUSSIONt Dr R. Zentel (University of Mainz, West Germany) said: I would like to make two comments on the paper presented by Dr Noel. First, we have carried out dielectric relaxation measurements on liquid-crystalline side-group polymers’ and were able to determine up to five different relaxation processes, depending on the molecular structure. One of these is the so-called ?‘-relaxation, a relaxation of the alkyl-group spacer between the polymer chain and the mesogenic groups. For spacers of six methylene groups it is active and comparable to your y-relaxation. However, if one shortens the spacers to only two methylene groups, it is no longer possible. Secondly I would like to comment OII the fact that Dr Noel was able to supercool the isotropic-liquid-crystalline transition for her polymers.This is very unusual in low-molar-mass liquid crystals, because just below the observed first-order transition, isotropic-nematic or isotropic-smectic A or C, there is a hypothetical second-order transition to the same phases. Thus at the phase transition temperatures, there are still long-range correlations and the fluctuations are strong.2 Therefore one does not have problems with nucleation. Now if one can supercool the transition for liquid-crystal main-chain polymers, there may be two reasons. First the viscosity is higher for polymers. However, more importantly, Maret and Blumstein found3 that for main-chain polymers the hypothetical second-order phase transition temperature is far below the observed first-order transition.Therefore the fluctuations are still small at the transition temperature and one has similar problems with nucleation as for a crystallization. Does Dr Noel think that this may be a correct explanation? ’ R. Zentel, G. Strobl and H. Ringsdorf, Macromolecules, 1985, 18, 960; H. Ringsdorf, G. Strobl and R. Zentel, 29th Symp. on Macromolecules, Bucharest-Romania, September, 1983, Abstracts, section IV, p. 27. ’ See for example: P. G. de Gennes, The Physics ofLiquid Crystals (Clarendon Press, Oxford, 1975). A. Blumstein, Polym. J., 1985, 17, 277. Dr C . Noel (ESPCI, Paris, France) replied: Dr Zentel has made an interesting comment on the intramolecular motional processes occurring in thermotropic liquid- crystal polymers in the glassy state. Dr Lauprctre and I have also investigated molecular motions in side-chain liquid-crystal polymers by an e.s.r.spin-probe technique. ly2 Different relaxation processes have been observed depending on the molecular structure. In the glassy state, one relaxation process is found in all samples investigated and can be assigned to internal motions of mesogenic groups. Another depends on,the length of the flexible spacers and must arise from a local motion of thefCH,), units. Support for these assignments has been obtained for the polymer f CH- CH, t 0’ ‘O+CH,)t,--CO-O 0 0 CN using 13C solid-state n.m.r. techniques.2 From proton-decoupled 13C n.m.r. lineshape analysis of non-spinning samples, it appears that the spectrum of the carboxy carbon I C w t Plates 1-3 face p. 228. 225226 GENERAL DISCUSSION adjacent to the main chain is that of the rigid limit in the glassy state.In contrast, the reduction of the chemical-shift anisotropy of the protonated aromatic carbons indicates fast oscillations of the phenyl rings about their symmetry axis. The increase in magnetization as a function of the contact duration in a spin-lock cross-polariz- ation experiment using magic-angle spinning shows that the main-chain carbons are frozen within the timescale of the experiments ( lo5 Hz). In contrast, oscillations on the valence cone of ca. 20" about one equilibrium conformation are observed for the three central methylene carbons of the flexible spacers. This suggests that jumps between two equilibrium conformations as observed for the y-relaxation of the main-chain polyester TO1 1 require flexible spacers containing more than five CH2 groups.' P. Le Barny, J. C. Dubois, C. Friedrich, F. Lauprgtre, C. Noel and L. Monnerie, IUPAC Int. * G. Decobert, J. C. Dubois, S. Lukovic, C. Noel and L. Monnerie, IUPAC Int. Symp. Non-crystalline Symp. Non-crystalline Order in Polymers, Naples, May, 1985. Order in Polymers, Naples, May, 1985. Prof. G. Williams (University College of Wales, Aberystwyth) said: Dr Zentel has indicated that multiple dielectric loss processes are observed for liquid-crystalline side-chain polymers. In his assignment of the mechanisms for the individual process he has suggested that the process observed at the highest temperatures (say in a plot of loss factor against temperature at a fixed frequency) is due to the end-over-end motion of the side chain with respect to the main chain.It seems likely that the topological constraints of having the flexible alkyl spacer attached to the main chain and having local angular correlations between side-chain mesogenic groups would make it difficult for such a 'flip-flop' motion to occur. Indeed, it is possible that in small-molecule nematics (e.g. alkylcyanobiphenyls) that the well defined principal dielectric relaxation process in the MHz range, usually assigned as a flip-flop motion in the P2 nematic potential, is not of that origin. Several alternative mechanisms may be proposed which would accommodate the main dielectric process in both monomeric and polymeric (side-chain) nematics. One possibility is that in both systems the dipolar mesogenic group is able to move in an effective 'cone' prescribed by the neighbouring mesogenic groups. Such a motion may be modelled by the Warchol-Vaughan- Wang-Pecora model of rotational diffusion, as we have applied it to lyotropic-nematic rigid-rod polymer systems,' but would give a reduced relaxa- tion magnitude compared with that for isotropic motion.Experimentally it is found for alkylcyanobiphenyls2 that the apparent Kirkwood g-factor is ca. 0.66, which might be interpreted as being due to motions of mesogenic groups limited to an effective cone. The total relaxation magnitude would then result from a combination of limited one-body motions and a Kirkwood factor representing the preferred angular correlations of the dipole vectors in the nematic phase.J. K. Moscicki and G. Williams, J. Polym Sci., Polym Phys. Ed., 1983, 21, 197, 213. 2, 1976, 72, 1447. * M. Davies, R. Moutron, A. H. Price, M. S. Beevers and G . Williams, J. Chem. SOC., Furuday Trans. Prof. A. Blumstein (University of LoweEZ, U.S.A.) said: By analogy with the observation by Noel et al. of two distinct glass-transition temperatures ( Tg) for the smectic and isotropic domains of the polyester TO1 1, I would like to mention that the polyester DDA-9 (see our paper) also displays two distinct 7'' values. Both can be observed on samples of DDA-9 only after quenching the isotropic phase in liquidGENERAL DISCUSSION 227 nitrogen. The Tg values observed are at -3 to -7 "C and 13-15 "C and are attributed to the isotropic and nematic domains, respectively.They are the result of quenching of the isotropic and nematic phases in DDA-9. The simultaneous quenching of an isotropic and a nematic phase is a rather rare occurrence, and may be due to the cybotactic nature of the nematic phase of DDA-9. Indeed, we have not been able, under similar conditions, to quench the isotropic phase of MA-9 (again see our paper), which displays a normal nematic mesophase. This difference in the super- cooling behaviour in the homologous series of polyesters between the 'even' and 'odd' specimens can be understood from table 1 of our paper. In that table, high values of the order parameter S are related to high values of supercooling (samples with n even) and vice versa. One can also remark that for DDA-9 ( n = 10) the N/I transition is more first order than for MA-9 ( n = 7 ) , with values of T,- T* = 27 "C for DDA-9 as compared with 14.5 "C for MA-9 and only 5°C for p-azoxyanisole.' ( T , is the clearing temperature and T* is the virtual second-order transition temperature.) These results are in agreement with our proposed model of two distinct molecular nematic organizations for the even and the odd polyesters containing the 4,4'-dioxy- 2,2'-dimethylazoxy mesogen.2 ' G.Maret, Am. Chem. SOC., PoZym. Prepr., 1983, 24, 2, 249. ' A. Blumstein, R. B. Blumstein, M. M. Gauthier, 0. Thomas and J. Asrar, Mol. Cryst. Liq. Cryst. Lett., 1983, 92, 87. Dr C . Noel (ESPCI, Paris, France) said: I would like to make a further comment. We completely agree with the points raised by Prof.Blumstein and Dr Zentel: the simultaneous quenching of an isotropic and a liquid-crystalline phase is a rather rare occurrence. First, we would like to recall that Frosini et al.' were the first to detect two glass transitions in some thermotropic polyesters based on a,o-bis(4- hydroxybenzoy1oxy)alkanes and terephthalic acid. In agreement with our data, the thermal and dynamic viscoelastic behaviour of these polymers suggests that the lower of the two glass transitions is associated with the amorphous phase while the upper glass transition 'is connected with the unfreezing of the super-cooled mesophase and is observed in samples quenched from the liquid-crystalline state. Secondly, we would like to point out that although polyester TO1 1 has a relatively high molecular weight and hence a viscosity which is much higher than that of a conventional liquid crystal, we have not been able to quench the isotropic phase of polyester TO1 1 in liquid nitrogen under the usual conditions.Only ultraquenching in isopentane cooled in liquid nitrogen has resulted in the preparation of sample with a marked TgL. Thus, the thermal behaviour of polyester Toll is not very different from that of a conventional liquid crystal. The remarks that the difference between the N/I transition temperature and the virtual second-order transition temperature may explain the simultaneous quenching of an isotropic and a nematic phase in DDA-9 are correct. This suggestion, however, has not been substantiated for other liquid-crystal polymers, and much study is needed in this area.Systematic research directed toward characterization of the structure and the properties of thermotropic liquid-crystal polymers as a function of quenching technique, quench temperature, melt temperature from which quenched and subsequent annealing occurs and a careful analysis in certain homologous series should suffice to establish certain laws. V. Frosini, S. de Petris, E. Chiellini, G. Galli and R. W. Lenz, MoZ. Cryst. Liq. Cryst., 1983,98,223.228 GENERAL DISCUSSION Dr C. Viney (University of Cambridge) said: Dr Coles' paper refers to molecular orientation effects induced by electric and (occasionally) magnetic fields in ther- motropic side-chain polymers. Such an effect can also be identified when a magnetic field acts on a thermotropic main-chain polymer.Relevant experimental conditions, and some implications of thus being able to modify molecular correlations by means of a magnetic field, are given below. The particular polymer investigated was that designated as B-ET in our paper at this Discussion. Specimens having a reproducible starting texture were prepared by shearing onto an aluminium substrate at 300°C and quenching to room tem- perature. Subsequently, specimens were annealed at 280 "C for 2 h, in a magnetic field of approximate strength 0.5 T; they were then quenched to room temperature so that their textures could be observed microscopically between crossed polars. The magnetic field was oriented either perpendicular or parallel to the specimen surface. [Specimens annealed on aluminium must of course be removed from the substrate before they can be viewed in transmitted light.The aluminium was dissolved in a 1 mol dmP3 solution of NaOH in water, to which a few drops of HgC12 solution (0.01 mol dm-3) had been added; specimens were then washed twice in distilled water and allowed to dry.] It is expected that the magnetic field will encourage particular rotational correla- tions of molecules about their chain axes. The planes of phenyl groups will tend to lie parallel to flux lines, since this gives the least distortion of the field and therefore the lowest energy.' The molecular arrangement actually obtained may also reflect surface interactions between polymer and substrate. The initial texture resulting from shear is typified by plate 1.It is shown together with the corresponding conoscopic image, obtained by using a Bertrand lens. The effect of annealing in the magnetic field is shown in plate 2 (field normal to specimen surface) and plate 3 (field parallel to specimen surface). Clearly, the texture obtained depends markedly on the orientation of the magnetic field relative to the specimen. This effect is less evident if glass substrates, rather than aluminium ones, are used; this may be because glass-polymer surface interactions are stronger. The above observations have some significant implications. ( 1 ) Long-range molecular correlations in thermotropic main-chain liquid-crystal polymers can be influenced by magnetic fields. The necessary field strength is relatively small. (2) The possibility exists for producing 'single crystal' domains of uniform molecular correlation, of lateral extent large enough for structural hnalysis based on techniques such as X-ray diffraction and optical conoscopic imaging.This would allow direct identification of uniaxial and biaxial ordering in polymers such as those referred to in our paper. (3) If fibres are spun from these materials, a magnetic field could be used to control the molecular correlations which involve rotation about the chain axes. The fibres could thus be made to have anisotropic transverse properties. P. G. de Gennes, in 7'he Physics ofLiquid Crystals (Oxford University Press, Oxford, 1979), p. 80.Plate 1. Texture typical of B-ET sheared on an aluminium substrate at 300 "C and quenched to room temperature. The corresponding conoscopic image is also shown. The specimen was photographed between crossed polars, with the polarizer parallel to the shear direction (horizontal). [facing page 228Plate 2. Sheared B-ET specimen annealed for 2 h at 280 "C in a magnetic field applied normal to the specimen surface. The specimen was photographed between crossed polars with the polarizer parallel to the original shear direction (horizontal). The corresponding conoscopic image is also shown.Plate 3. Sheared B-ET specimen annealed for 2 h at 280 "C in a magnetic field applied parallel to the specimen surface and parallel to the original shear direction (horizontal). The cono- scopic image corresponding to the upper micrograph is also shown. (p=polariser; a = analyser. )