(Reçu le 2 février 1988, révisé le 22 novembre 1988, accepté le 22 novembre 1988). Résumé. 2014 Neuf termes de la série des dérivés tétra-n-alcanoyloxy de ...
J.
Phys.
France 50
(1989)
539-547
ler
MARS
539
1989,
Classification
Physics Abstracts 61.30
-
v
Ellagic acid derivatives : a new mesogenic series exhibiting a thermotropic cubic mesophase J. Billard
(1),
H. Zimmermann
(2),
(1)
Laboratoire de
(2)
Max-Planck-Institut für medizinische
(3)
The Weizmann Institute of
(Reçu
le 2
Physique
février 1988,
R.
Poupko (3)
de la Matière Condensée
and Z. Luz
(*), Collège
(3)
de France,
Paris, France
Forschung Jahnstrasse 29, D-6900 Heidelberg I,
F.R.G.
Science, Rehovot 76100, Israel
révisé le 22 novembre 1988,
accepté le
22 novembre
1988)
Résumé. 2014 Neuf termes de la série des dérivés tétra-n-alcanoyloxy de l’acide ellagique (de l’octanoyle à l’hexadécanoyle) sont synthétisés et étudiés par calorimétrie et microscopie optique. Les dérivés octanoyloxy et nonanoyloxy ne sont pas mésogènes, les autres termes présentent une mésophase thermodynamiquement stable, biréfringente et fortement organisée. Les cinq substances qui ont les plus longues chaînes latérales présentent, en plus, une mésophase métastable optiquement isotrope, donc cubique. D’après les observations de miscibilités cette phase cubique, que nous appelons CD, est différente des autres mésophases cubiques antérieurement identifiées dans les mésogènes thermotropes achiraux. Cette mésophase CD peut être conservée pendant plusieurs semaines à la température ambiante.
Nine members of the
tetra-n-alkanoyloxy ellagic acid series, ranging from the octanoyl hexadecanoyl were synthesized and studied by calorimetry and optical microscopy. The octanoyloxy and nonanoyloxy derivatives are not mesogenic, while the other homologues exhibit a highly organized birefringent enantiotropic mesophase. Of these the five members with the longest side chains exhibit in addition a monotropic optically isotropic, and consequently, a cubic mesophase. From miscibility studies it appears that this cubic phase, which we name CD, is different from other cubic mesophases previously identified in non-chiral thermotropic mesogens. This monotropic CD mesophase can be maintained at room temperature in a Abstract.
2014
to
metastable state for several weeks.
Introduction.
Liquid crystalline mesophases with cubic symmetry have been identified both in lyotropic [1] and thermotropic systems. In the latter category the best known examples are the various blue phases which are found in many chiral thermotropic mesogens [2]. Besides the plastic crystals
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01989005005053900
540
only three distinct non-chiral thermotropic cubic mesophases have been reported so far [3]. These include the cubic CA (smectic D) found [4-9] in members of series 1, with X N02 or CN, and m = 16 or 18, the cubic mesophase CB found [10,11] in members of series 2, with m 8, 9, 10, and the cubic mesophase Cc recently identified [12] in one of the inositol 3. Another compound, 4, exhibiting an optically isotropic derivatives, scyllo has not yet been subjected to miscibility studies has recently been however which mesophase, described [13]. As part of our study on the relation between molecular structure and mesomorphic properties [14] we have investigated the polymorphic behaviour of tetraesters of ellagic acid, 5, and discovered a new non-chiral cubic mesophase which appears to be different from those exhibited by the above compounds. In the present paper we present the phase diagrams of these ellagic acid derivatives and discuss the optical properties of their mesophases. =
=
541
Experimental. A. SYNTHESIS.
Tetraesters of
ellagic acid (2, 3, 7, 8-tetraalkanoyloxy-1-benzopyrano(5, 4, 3-cde)-1-benzopyran-5, 10-dione) were prepared by refluxing dry ellagic acid (2gr) in excess of the n-alkanoylchloride for about 8 hours. The residual acid chloride was removed by vacuum distillation and the residue recrystallized several times from chloroform. Yield : 65 %. Elemental analysis gave the correct carbon and hydrogen content and high resolution 1H NMR spectra were as expected with no impurity signals. -
-
B. DIFFERENTIAL SCANNING CALORIMETRY. - Transition températures and transition enthalpy changes were measured using a Mettler, T.A. 3000 differential scanning calorimeter. The results reported are for increasing temperature (1 to 5 °C/min).
C. OPTICAL MICROSCOPY. Thin samples were observed between two untreated coverslips, made of ordinary glass. The textures as well as the miscibilities were studied using a polarizing microscope (Leitz, Panphot or Zeiss, Universal) equipped with a hot stage (Mettler, F.P. 52). Binary phase diagrams, at atmospheric pressure, were constructed by observations of contact preparations [15], and the solubilities of the solids calculated using the Le Chatelier-Schrôder equation [16, 17]. Results. A. THE PHASES OF -5-m AND THE PHASE DIAGRAMS OF THEIR BINARY MIXTURES. - The transition temperatures and molar enthalpy changes of the ellagic acid alkanoyloxy derivatives studied in the present work, are summarized in the table I. The compounds are referred to as 52m where m is the number of methylen and methyl carbon atoms in the side chains. All these compounds are stable in air in the temperature ranges of the investigation. At room temperature they appear as opaque fine crystalline needles and all of them exhibit a solid-solid transition below the melting point. Thè 5-7 and 5-8 homologues melt directly from the crystalline phases to the fluid (isotropic) liquids, and by rapid cooling they crystallize into fragile birefringent rectilinear needles. On melting of 5-9 or 5-10 an enantiotropic mesophase designated Ml, is obtained as manifested by slight rounding of the edges of the crystalline needles. Pressing over the cover slips shows deformation and a highly confused birefringence characteristic of fluid mesophases. Further heating eventually results in a sharp transition to the liquid fluid phase. Cooling back to Mi gives a birefringent network of long needles with uniform extinctions. By pressing the cover-slips at a temperature near the clearing point these preparations behave as fluid. From these results and the high values of the clearing molar enthalpy changes we conclude that MI in the 5-9 and 5-10 compounds corresponds to a highly organized mesophase. Similar observations on the mesophase Mi were made by heating the homologues 5-11, to 5-15. However in these cases rapid cooling of the liquids resulted in significant supercooling, followed by the formation of an optically transparent and isotropic, viscous mesophase, which we refer to as M2. Under microscopic observation a well defined boundary is seen between the normal liquid and M2, indicating that the latter is not a glass. By pressing over the cover-slips this boundary deforms, hence M2 is not a solid. The , growth of M2 on cooling is manifested in the formation of finger-like contours or fine, either rectilinear or curved needles. Slow heating (1 °C/min) of the M2 mesophase results in transformation to the normal liquid, but at temperatures slightly below the clearing points of MI observed on first heating of the crystals (see Tab. I). Consequently Mi is enantiotropic while M2 is monotropic (Fig. 1). We were unable to obtain Mi by cooling. The monotropic M2 mesophase
542
could be conserved at room temperature for many weeks. No maltese crosses were observed in this phase with convergent light. The binary phase diagrams of mixtures of neighbouring homologues of séries 5 (Fig. 2) indicate that all mesogenic compounds exhibit the same enantiotropic Ml and monotropic M2 mesophases. To obtain the Mi mesophase in the compounds, which also exhibit M2, the following procedure was used : the compound with the higher clearing point was inserted into the heating stage at a temperature below the clearing temperature and the sample pressed, and removed from the hot stage to room temperature. The second component was added and the preparation reinserted into the hot stage at a temperature where the first component is in Mi and the second a liquid, thus establishing a contact boundary between the two phases. By lowering the temperature the continuous growth of the mesophase Ml into the liquid phase was observed. This procedure is also convenient for the mixtures of 5-11 and 5-12 where the equilibrium curves of MI and L exhibit a minimum.
(in degrees centigrade) and, in parentheses, enthalpy changes (in kJ/mole) for 2, 3, 7, 8-tetra-n-alkanoyloxy derivatives of ellagic acid (5) Table I.
-
Transition temperatures
(’) In the table K stands for crystalline phases, M for mesophases, and L for liquids. The data for monotropic phases are given in parentheses. m is the number of methylene and methyl carbons in the side chains. (b) Sum of
melting
and
clearing enthalpies.
diagram of the free energy versus temperature for the 5-m compounds exhibiting monotropic M2 mesophase (labels in parenthesis indicate metastable states). The indicated temperatures correspond to the 5-11 homologue.
Fig.
the
1. - A schematic
543
2. Phase diagrams of binary mixtures of neighbouring mesogenic compounds of the 5-m series. The dashed lines represent the coexistence curves of the monotropic M2 mesophase and the
Fig.
-
liquid phase. mixtures of 5-11 to 5-15, after heating to above the clearing points the lowered to the monotropic M2 mesophase and the coexistence curves of the latter and the liquid phase established. These curves (dashed lines in Fig. 2) indicate that M2 in all five compounds is the same monotropic mesophase. In all
binary
temperature
was
B. BINARY MIXTURES OF 5-m WITH OTHER COMPOUNDS EXHIBITING CUBIC MESOPHASES. To compare the mesophase M2 with other previously described optically isotropic thermotropic non-chiral mesophases, binary phase diagrams of 5-13 with three reference compounds exhibiting the cubic mesophases [3] CA, CB and Cc were constructed by examination of contact preparations. The reference compounds used were 1-18 (with X N02), 2-10 and 3 [5, 7, 11, 12]. The resulting phase diagrams provide of course only the coexistence curves (dashed lines), and not the equilibrium curves, since M2 is monotropic. These curves could not be obtained using the polarizing properties, since the phases in contact are isotropic. Rather the observations were made with the analyzer removed and viewing the Becke lines [18]. Such lines appear in the contact boundary of two isotropic phases having different refractive indices. They are detected as shadow or bright lines by defocusing the microscope above and below the plane of the sample. If at a particular temperature the refractive indices of the two phases become identical the Becke line disappears and the boundary is not visible (see below the discussion of 5-13 and 3). In mixtures of 5-13 and 1-18 (Fig. 3) the mesophase M2 of 5-13 coexists with the CA and Sc mesophases. Since M2 and CA are immiscible they most likely have different symmetries or at least different structures. From the slopes of the coexisting curves between CA, Sc and M2 it follows that in the binary mixtures under discussion M2 exists at temperatures below those of CA and Sc. In the phase diagram of the mixtures of 5-13 and 2-10 (Fig. 4) the mesophase M2 coexists with the cubic mesophase, CB, of the reference compound. To observe the boundary between these two phases it is necessary to cool the liquid mixture very slowly (0.2 °C/min or slower) in order to avoid the formation of fine needles of M2. Observations of the Becke lines at temperatures higher but near the triple point liquid-M2-CB (151 °C) indicate that the differences between the refractive indeces of the mesophases and the liquid are of opposite signs. Thus the two isotropic mesophases, with essentially the same composition, have different refractive indices. On cooling the liquid on both sides of the triple point the growth rate of the mesophases is not accelerated as the 151 °C point is approached. After the disappearance of the liquid phase, below 151 °C, a Becke line subsists between the two mesophases. Thus there is no minimum in the coexisting mesophase-liquid curves but rather a =
544
Fig.
3. - Phase
diagram
of
binary
mixtures of 5 -13
(right)
Fig.
4. - Phase
diagram
of
binary
mixtures of 5-13
(left)
and
1-18 (left).
and 2-10
(right).
triple point suggesting that M2 is different from the cubic mesophase CBI of 2-10. From the slope of the coexisting curves of M2 and CB it follows that M2 exists at lower temperatures than CB. To perform the contact preparations for the binary phase diagram (Fig. 5) of the mixtures of 5-13 and 3 some special precautions were necessary in order to obtain [12] the Cc mesophase of 3 : crystals of 3 were inserted between two cover-slips and heated to the liquid phase. The temperature was rapidly lowered to 207 °C, maintained at that point until the growth of Cc was completed, and the system cooled to room temperature. Crystals of 5-13 were then placed near the edge of the upper cover-slip and heated, using a Kofler heating block, to its clearing temperature. As a result the liquid of 5-13 entered by capillarity between the cover-slips to form a contact with 3. The preparation was then inserted into the hot stage at a temperature above 175 °C. At 185 °C the Becke line at the boundary between Cc and the coexisting liquid disappeared indicating that the two phases have the same refractive index at this temperature ; at lower and higher temperatures the Becke line
545
Fig.
5.
-
Phase
diagram
of
binary
mixtures of 5-13
(right)
and 3
(left).
réappears. Further observations on binary mixtures of 5-13 and 3 gave similar results to those of the mixtures with 2-10 (opposite signs of the differences of the refractive indices, no acceleration of the growth rate, Becke line between the two mesophases) indicating the existence of a triple point. The temperature dependence of the coexistence lines for Cc and M2 indicates that the latter exists at higher temperatures than Cc.
Summary and discussion. The transition temperatures of the 5-m homologues studied in the present work are summarized in figure 6. It may be seen that the solid-solid transition temperatures form a zigzag pattern often encountered in homologous series, while the melting temperatures decrease monotonically with the length of the side chains. The homologues with sufficiently long side chains are mesogenic, exhibiting an enantiotropic birefringent phase M1, and for even longer
Fig.
6.
-
Plots of transition temperatures of the neat 5-m
compounds
studied in the present work.
546
side chains, also a monotropic, optically isotropic mesophase, M2. The clearing temperatures of Ml and M2 decrease regularly with the chain length. The characteristic properties of Ml, i.e. the growth habit as manifested in the rounding of the crystalline edges and the deformations observed under pressure, indicate that it is a liquid crystalline mesophase. However its high viscosity and its high clearing enthalpy (much larger than for the melting) suggest that it is highly ordered. We have performed some preliminary powder X-ray measurements on the Ml mesophase in the two homologues, 5-12 and 5-14. The low angle diffraction region (d > 15 À) always showed peaks which also appear in the solid phases, but in addition three new weak lines, which in 5-14 could be indexed on a lamellar lattice (with spacing - 57 À). In 5-12 the new peaks are much broader and cannot readily be indexed. The high angle diffraction pattern of the Ml mesophase in both compounds is similar to that of the corresponding solids, except that region around 4.7 Â becomes diffused. This most likely reflects packing disorder and chain melting upon the solid-MI transition, but no quantitative analysis of these results was attempted. Thus although the X-ray measurements do not provide conclusive identification of Ml, the optical microscopy results and in particular the fact that there is complete miscibility of the Ml phases over the range of homologues from 5-9 to 5-15 provide strong evidence for Ml being a mesophase. We note that unlike the Ml mesophases, the crystalline phases of the ellagic acid homologues show no complete miscibility and in fact there are no cases known of complete miscibility of solid phases over such a wide range of homologue as in figure 2. As for Ml the clearing enthalpies of M2 are also larger than those corresponding to melting. Miscibility studies suggest that M2 is different from the cubic mesophases CA of 1-18 (Fig.3), CB of 2-10 (Fig.4) and from Cc of 3 (Fig.5). We note that the CA mesophase lies above smectic C in series 1, while in series 2 the cubic mesophase, CB, lies below smectic C. From figure 4 it follows that the temperature range of existence of M2 is lower than the cubic phase CB of 2-10. Although all these observations do not rule out conclusively the possibility that M2 has a symmetry similar to CA, CB or Cc, it most likely has a different structure and we propose to name it CD. Recently we have synthesized several derivatives of corruleoellagic acid, 6 (hexaalkanoyloxy[1]-benzo-pyrano-[5.4.3-cde]-[1]-benzopyran-5.10 dione). These hexa-substituted compounds are similar in structure to the discotic mesogens and apparently also exhibit columnar mesophases as evidenced by optical microscopy. One of these mesophases appears to be optically isotropic [19].
Acknowledgments. We thank Professors D. Demus and K. Praefcke for the loans of the samples used as reference compounds, and Dr. E. Wachtel for performing the X-ray measurements. This research was supported by grants from the German-Israeli Foundation (GIF) for Scientific Research and Development, the National Council for Research and Development, Israel and the KFA Jülich, West Germany and the Israel Academy of Sciences. One of us (J.B.) thanks the France-Israel Scientific Exchange Program for visiting fellowship at the Weizmann Institute of Science.
547
References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
CHARVOLIN J., J. Phys. Colloq. France 46 (1985) C3-173-90. STEGEMEYER H., BLÜMEL Th., HILTROP K., ONUSSEIT H. and PORSCH F., Liq. Cryst. 1 (1986) 328. BILLARD J., C.R. Acad. Sci. Paris 305II (1987) 843-5. GRAY G. W. , JONES B. and MARSON J., J. Chem. Soc. (1957) 393-5. DEMUS D., KUNICKE G., NEELSEN J. and SACKMANN H., Z. Naturforsch. 23a (1968) 84-90. TARDIEU A. and BILLARD J., J. Phys. Colloq. France 37 (1976) C3-79-81. DEMUS D., MARZOTKO D., SHARMA N. K. and WIEGELEBEN A., Kristall Techn. 15 (1980) 331-9. GRAY G. W. and GOODBY J. W., Smectic liquid crystals textures and structures (Heyden, Philadelphia) 1984. ETHERINGTON G., LEADBETTER A. J. , WANG X. J., GRAY G. W. and TAJBAKHASH A., Liq. Cryst. 1 (1986) 209-14. SCHUBERT H., HAUSCHILD J., DEMUS D. and HOFFMANN S., Z. Chem. 18 (1978) 256. DEMUS D., GLOZA A., HARTUNG H. , HAUSER A., RAPTHEL I. and WIEGELEBEN A., Cryst. Res. Techn. 16 (1981) 1445-51. KOHNE B., PRAEFCKE K. and BILLARD J., Z. Naturforsch. 41b (1986) 1036-1044. NGUYEN HUU TINH, DESTRADE C., LEVELUT A. M. and MALTHETE J.,J. Phys. France 47 (1986) 553-7. ZIMMERMANN H., POUPKO R., LUZ Z. and BILLARD J.,Z. Naturforsch. 40a (1985) 149-60 ; 41a (1986) 1137-40 ; Liq. Cryst. 3 (1988) 759-770. KOFLER L. and KOFLER A., Thermomikromethoden (Verlag Chemie, Weinheim) 1958. LE CHATELIER H., C. R. Acad. Sci. Paris 100 (1885) 50-2. SCHRÖDER I., Z. Phys. Chem. 11 (1893) 449-65. BECKE F., Sitz. Math. Naturwiss. Cl. Kais. Akad. Wiss. 102I (1893) 358-76. ZIMMERMANN, H., POUPKO R., BILLARD J. and LUZ Z. (to be published).
