DOI 10.1140/epje/i2010-10632-1

Regular Article

Eur. Phys. J. E 32, 243–247 (2010) THE EUROPEANPHYSICAL JOURNAL E

Calorimetric study of the Paranematic-to-Nematic transition ofpolydomain side-chain liquid-crystalline elastomers with differentmesogen composition

G. Cordoyiannis1,2,a, B. Rozic1, H. Finkelmann3, S. Zumer1,4, and Z. Kutnjak1,b

1 Condensed Matter Physics Department, Jozef Stefan Institute, Jamova 39, Ljubljana 1000, Slovenia2 EN FIST Centre of Excellence, Dunajska 156, Ljubljana 1000, Slovenia3 Institut fur Makromoleculate Chemie, 056 Albert-Ludwigs-Universitat Freiburg, Freiburg 79104, Germany4 Department of Physics, University of Ljubljana, Jadranska 19, Ljubljana 1000, Slovenia

Received 15 February 2010 and Received in final form 7 May 2010Published online: 8 July 2010 – c© EDP Sciences / Societa Italiana di Fisica / Springer-Verlag 2010

Abstract. The phase transition behaviour of various nematic side-chain liquid-crystalline elastomers withdifferent mesogen composition has been explored by means of high-resolution ac calorimetry. Polydomainsamples of the same crosslinking density and different type of mesogens have been investigated. Theresults show a strong dependence of the phase transition features upon the composition of the mesogen.The distance from the critical point, reflected in the sharpness of the heat capacity anomalies, increaseswhen adding a shorter-length mesogen. The results provide new insight for the impact of mesogens onthe thermodynamic behaviour and, thus, on the thermomechanical response of nematic liquid-crystallineelastomers.

1 Introduction

Liquid-crystalline elastomers are materials that simulta-neously possess the orientational order of liquid crystalsand the rubber elasticity of polymer networks [1–3]. Sev-eral of their optical, thermomechanical and ferroelectricproperties have endowed the elastomers with a high po-tential for applications and technological leaps [4–10]. Inthe last decades, these materials have been extensivelystudied in terms of theory [11–16], experiments [17–20]and simulations [21,22].

The precise knowledge of the phase transition be-haviour of elastomers is very useful for tuning their ther-momechanical response. The temperature range at whichmost of the enthalpy change is observed corresponds tothe regime where most of thermomechanical expansion (orcontraction) occurs. Hence, the gradual or abrupt changeof their shape in the vicinity of a phase transition dependson the thermodynamic behaviour. The Paranematic-to-Nematic (PN-N) phase transition has been recently ex-plored for monodomain side-chain and main-chain elas-tomers, revealing a strong dependence on the crosslink-ers density [23,24]. Note that the term Paranematic isused (instead of Isotropic), due to the non-zero order pa-rameter observed in elastomers by nuclear magnetic reso-

a e-mail: george.cordoyiannis@ijs.sib e-mail: zdravko.kutnjak@ijs.si

nance above the Isotropic-to-Nematic transition tempera-ture [23]. By solely changing this parameter (i.e. crosslink-ers density) one may drive the PN-N transition from su-percritical to first order, via a critical point of liquid-vapor type. The role of the crosslinkers topology was alsoexplored in [23], where the same densities of differentcrosslinkers had different impact on the order of the tran-sition. The role of the mechanical load and the crosslink-ing temperature has been also demonstrated in [24,25].In the present work the impact of the mesogens type onthe thermodynamic behaviour is explored for polydomainside-chain liquid-crystalline elastomers by means of high-resolution ac calorimetry.

2 Samples and apparatus

The building blocks of a side-chain liquid-crystalline elas-tomer are polymer backbones, crosslinkers and mesogens.The latter are attached to the monomer units of thepolymer backbones via the functional alkene chain. Inthe present study, different polydomain (i.e. non-aligned)side-chain liquid-crystalline elastomers have been chosen.Regarding the building blocks of the samples, the polymerbackbone consists of methyl-siloxane units. Tri-functionalcrosslinkers of point-like geometry have been used, namely1,3,5-tris(undec-10-enyloxy)cyclohexane (V3). The con-centration of the crosslinkers has been kept the same for

244 The European Physical Journal E

Fig. 1. The chemical formulae of the building blocks forthe studied polydomain liquid-crystalline elastomers are givenhere: the mesogens MC3 and MC4 with a biphenyl core anddifferent carbon chains, the methyl-siloxane polymer backboneand the tri-functional (point-like) crosslinkers V3.

all samples at the value x = 0.105, where x denotes thecoverage of active chain groups of the polymer backbone.Two types of biphenyl mesogens have been used, namely 4-butyl-3-enyl-benzoic acid 4-methoxy-phenyl ester (MC4)and 3-propyl-2-enyl-benzoic acid 4-methoxy-phenyl ester(MC3). The difference between the two molecules liessolely at the length of the carbon chain, with MC4 be-ing longer by one carbon atom. Three different elastomersamples were prepared differing only on the type of meso-gens used in each one. The first sample contained onlyMC3 mesogens, while the other two had a mixture ofMC3 and MC4 with mass ratio mMC3/mMC4 equal to 1.00and 0.44, respectively. The crosslinking of all samples hasbeen carried out in the N phase. The chemical formulaefor all the components (methyl-siloxane backbone units,V3 crosslinkers, mesogens MC3 and MC4) of our liquid-crystalline elastomers samples are given in fig. 1.

The experimental apparatus used for the present mea-surements is a computerised home-made high-resolutionac calorimeter. It operates in the classical ac mode andalso in a relaxation mode [26–28]. The former detects onlythe continuous changes of the enthalpy, while the latter de-tects the total change of enthalpy, including the continu-

ous and discontinuous part (i.e., latent heat L). A detaileddescription of ac calorimetry has been given in [26,27].

Sample quantities of 40mg were placed in hand-madecells made from high-purity silver. A very thin layer ofteflon was carefully placed between the elastomer and thecell, in order to prevent the sticking of the sample to thewalls and, hence, the induction of any unwanted extrastress. The samples and cell sizes were appropriately cho-sen to allow for an unhampered expansion of the formerinside the latter. Since the liquid-crystalline elastomerstypically exhibit broad phase transitions, the heat capac-ity temperature profiles Cp(T ) were obtained on heatingthe elastomers over a large temperature range of 50K.This way we have scanned from deep inside the N phaseto much higher than the completion of the PN-N tran-sition. The scanning rate was in the order of 0.85Kh−1

for all ac runs. A relaxation run with a step of 0.7K wasadditionally performed for one of the samples. The heatcapacity of the empty cell as well as the one of teflon wassubtracted in order to obtain the net Cp of the elastomersample in each experiment.

3 Results and discussion

In a previous calorimetric study on side-chain elas-tomers [23], no discontinuity was observed in the PN-Nphase transition of monodomain liquid-crystalline elas-tomer samples with V3 crosslinkers of density x = 0.105,siloxane-based polymer backbone and MC4 mesogens.Due to the wide Cp anomaly, the uncertainty in the de-termination of latent heat was in the order of 0.05 J/g−1.Nevertheless, the continuous character of the transitionwas corroborated by the phase shift φ of the ac oscilla-tions. In case of a second order (continuous) transitionor supercritical evolution, Cp has only a real part andφ evolves inversely proportional to Cp. On the contrary,in case of a first-order (discontinuous) transition, Cp hasreal and imaginary parts and φ exhibits an anomalousbehaviour (e.g. positive spikes) in the temperature rangewhere latent heat is released [29]. In fig. 2 the Cp(T ) pro-file of the 0.105 monodomain sample is shown at the toplayer (a) and the φ(T ) profile at the bottom layer (b),respectively. It is clearly demonstrated that the evolutionof φ in the vicinity of the phase transition is inverselyproportional to Cp. This rules out the possibility of anydiscontinuity and, thus, for samples of the above crosslink-ing density composed of MC4 mesogens the transition iscertainly in the supercritical regime.

In order to see how the mesogen composition affectsor not the distance from the critical point, the crosslink-ers density has been kept the same and the shape of theanomalies is probed for various compositions. Since thesamples of this study are polydomains, the transitions areexpected to be broader with respect to the ones observedfor the monodomains. Nevertheless, the relative change ofthe Cp anomalies between samples with different meso-gen composition can reliably probe the distance from thecritical point.

G. Cordoyiannis et al.: Effect of mesogens on the paranematic-to-nematic transition elastomers 245

Fig. 2. The excess specific heat capacity ΔCp of ac mode (solidcircles) is plotted at the top layer (a) for the monodomainelastomer composed of MC4 mesogens. The phase shift φ ofthe temperature oscillations (open circles) is plotted at thebottom layer (b). The dotted line serves as a guide to the eyeto demonstrate the inversely proportional evolution of φ withrespect to ΔCp.

Fig. 3. The excess specific heat capacity ΔCp of ac mode(solid circles) and the effective ΔCp,eff of the relaxation mode(open circles) for the sample containing only MC3 mesogensare plotted. Within our experimental resolution no latent heatis detected, identifying a smeared, supercritical-like transition.

Fig. 4. At the top layer (a), the ac mode ΔCp(T ) profilesare given for polydomain samples with only MC3 mesogens(cross), and with mMC3/mMC4 ratio 1.00 (open circles) and0.44 (solid circles), respectively. The difference in the shape ofthe anomaly is significant when increasing the concentration ofMC4, though the width of the anomaly stays rather constant.At the bottom layer (b) the ac mode ΔCp(T ) profile for a mon-odomain sample with MC4 mesogens is given for comparison(from [23]).

No anomalous behaviour of φ was noticed for any ofthe presently investigated samples, indicating a continu-ous PN-N transition in all cases. Particularly for the sam-ple with MC3 mesogens an additional relaxation run wasperformed with a step (linear variation of the tempera-ture) of 0.7K. The excess specific heat capacity ΔCp(T )and ΔCp,eff(T ) of the ac and relaxation run, respectively,for this sample are plotted in fig. 3. The use of ΔCp in-stead of Cp denotes that the background heat capacity hasbeen subtracted (i.e. the background has been shifted tozero). No difference can be detected between the anomaliesobtained by the two modes. It is noteworthy that the re-laxation run is more noisy, because it is performed in smallsteps (0.7K) and somewhat faster than ac run. More-over, the reader may take into account that the particularCp anomaly is essentially small, since the heat capacitychanges for only 5% of its background value in the phasetransition regime.

A common plot of the ac runs ΔCp(T ) anomalies forall three samples, with MC3 mesogens only and with ra-tio of mMC3/mMC4 1.00 and 0.44, respectively, is shown infig. 4(a). Since the transition temperature TPN-N is chang-

246 The European Physical Journal E

ing with the addition of MC4, the heat capacity profilesare plotted as a function of T −TPN-N for clarity. This wayone can directly compare the shape (broadness and steep-ness) of the anomaly for different samples. Substantial dif-ferences are observed between the heat capacity profilesof the polydomain samples. The increase of MC3 (i.e. de-crease of MC4) mass in the elastomer composition yieldsmore broadened and suppressed anomalies. This type ofanomalies is typically observed along a Widom line [30]deep in the supercritical region. Thus, the use of MC3enforces more supercritical behaviour, i.e. by shorteningthe mesogen length the critical point is shifted to lowercrosslinking densities.

The results on the monodomain liquid-crystalline elas-tomer (taken from [23]), with x = 0.105 of V3 crosslinkersand MC4 mesogens are plotted in fig. 4(b) for compari-son. The ΔCp(T ) of the monodomain (i.e. oriented) MC4elastomer is not only steeper (as expected from the trendby adding MC4), but also less broad than the polydomainones. This will be demonstrated further down by plottingthe full-width at half-maximum (FWHM) of all heat ca-pacity peaks.

The PN-N phase transition of polydomain samples ofthe present work exhibits a broad and smeared (especiallyfor MC3 mesogens) anomaly of supercritical-like type. Insome other very recent study sharp continuous behaviourhas been observed by nuclear magnetic resonance tech-nique for polydomain samples of similar crosslinkers den-sity (0.10). The critical exponent β of the order param-eter was found to be 0.22 [31,32]. At present it is notclear weather that polydomain sample was not acciden-tally close to the critical point, which may be at differentvalue of crosslinking density than for our samples, due todifferent sample preparation and treatment procedures. Inparticular, in the present work tri-functional crosslinkershave been used and the samples were crosslinked in the Nphase. On the contrary, in [31,32] bi-functional crosslink-ers were used and the crosslinking was performed in theIsotropic (I) phase.

A substantial increase of the total transition enthalpyas a function of the MC4 mass in the composition isalso observed. The total enthalpy changes are ΔH =0.71±0.02 J g−1 for the MC3 sample, 1.17±0.02 J g−1 forthe sample with mass ratio of 1.00 and 1.39±0.02 J g−1 forthe sample of 0.44, respectively. The enthalpy change forthe monodomain MC4 sample (taken from [23] and givenhere only for comparison) is ΔH = 1.50 ± 0.02 J g−1. Inspite of the monotonous evolution in the enthalpy changeswith MC4 concentration, the transition remains broad andcontinuous (within resolution) for all polydomain samples,in agreement with the monodomain sample of the samecrosslinking density. The change of mesogen length, byonly one carbon atom, exhibits a considerable effect on theshape, especially on the steepness of the anomaly. The evo-lution of the total transition enthalpy ΔH is graphicallyrepresented in fig. 5, as a function of the MC3 mass per-centage. For liquid-crystalline elastomer containing solelyMC4 mesogens the data of a monodomain sample [23]are plotted with a different symbol. ΔH significantly de-

Fig. 5. The total enthalpy change ΔH (left y-axis) and thefull-width at half-maximum FWHM (right y-axis) of the PN-NCp anomaly are plotted versus the MC3 mass percentage (i.e.samples containing only MC3 and MC4 correspond to 1 and0, respectively). The solid circle and open rhombus symbolsare for ΔH and FWHM of the polydomain elastomers of thiswork, while the star and cross correspond to the monodomainelastomer of [23]. The dotted lines serve as guides to the eye.

creases when adding MC3 mesogens in the composition ofthe elastomer sample. In addition, in fig. 5 the FWHMof the Cp anomalies is plotted versus the MC3 mass per-centage. The broadness of the transition strongly increaseswith increasing the MC3 content. Such decrease of en-thalpy, suppression and broadening of Cp anomaly is typi-cally observed when the system is shifted deeper inside thesupercritical region along the Widom line. The substan-tial change in the shape of the anomaly evinces that forthe same crosslinking density the samples composed of theshorter MC3 mesogen are more supercritical. This meansthat lower densities of crosslinkers are required to drivethe elastomers composed of MC3 in the below-critical (i.e.first-order) regime compared to MC4 ones. Equivalently,the critical point shifts to lower crosslinking densities forelastomers composed of MC3 mesogens.

The observed difference in the shape and the enthalpyinvolved in the PN-N transition for MC3 and MC4 meso-gens could be related to changes in the conformation ofthe elastomer [33] and, particularly, in the orientation ofthe molecules with respect to the backbone. As reportedin ref. [33] the shorter MC3 molecule favours a locallyoblate (i.e. perpendicular to the backbone) conformation,while the longer and more flexible MC4 molecule favorsthe prolate (i.e. parallel to the backbone) one. For in-termediate compositions a frustration is possible betweenthe oblate (due to MC3) and prolate (due to MC4) whichcan result to stronger internal random fields. Nevertheless,the changes of both the enthalpy and the shape of the Cp

anomaly are continuous as a function of the mesogen com-position and they do not support significant random fieldsoriginating from such frustration.

G. Cordoyiannis et al.: Effect of mesogens on the paranematic-to-nematic transition elastomers 247

Furthermore, the polydomain samples have wider tran-sitions than the monodomain (see fig. 5) as well as othermonodomain liquid-crystalline elastomers [23,24]. Suchincrease in the width of heat capacity anomaly with re-spect to monodomains is typically observed in varioussolid and soft systems [34] and it is attributed to theincrease of random mechanical fields in the polydomainstructure.

Additional experimental studies are needed to fully re-solve the relative role of all the crucial parameters that af-fect the critical behaviour of the nematic liquid-crystallineelastomers. So far, it has been clearly demonstrated thatthe critical behaviour depends strongly on the crosslink-ers density and geometry [23], the applied mechanicalload [24], as well as the temperature of crosslinking [25].Nevertheless, the critical-behaviour dependence on thetype of mesogens was not adequately studied. The im-portant finding of this work is that the distance from thecritical point is certainly affected by the mesogen composi-tion (even when one uses mesogens with slightly differentcarbon chains) in the supercritical regime. This impliesthat similar effects must be expected in the below-criticalregime where first-order transitions are observed. We hopethat this work will stimulate additional experimental stud-ies of the phase transition behaviour of elastomers withdifferent mesogen composition.

4 Conclusions

The effects of the mesogen composition on the nature ofPN-N phase transition have been investigated for variouspolydomain side-chain liquid-crystalline elastomers, withthe same concentration of tri-functional crosslinkers V3.The shape of the heat capacity anomaly exhibits a strongdependence on the type of mesogen used; the longer themesogenic molecule, the steeper the transition and thebigger the enthalpy change involved in it. For the partic-ular crosslinking density of the current study, the orderof the transition remains unaffected by the mesogen type.However, the critical point apparently shifts towards lowervalues of crosslinking density when using shorter mesogensas derived from the heat capacity temperature profiles.

This work was performed with the financial support of EU(Project No. HPRN-CT-2002-00169) and Slovenian Office ofScience (Projects No. J1-9368, P1-0125). G.C. acknowledgesthe financial support of the CO EN FIST Centre of Excellence.The authors are grateful to Elke Stibal-Fischer for help in thesynthesis of the elastomer samples.

References

1. R. Zentel, Angew. Chem. Int. Edn Engl. 28, 1407 (1989).2. S.M. Kelly, Liq. Cryst. 24, 71 (1998).3. M. Warner, E.M. Terentjev, Liquid Crystal Elastomers

(Oxford University Press, New York, 2003).

4. W. Lehmann, H. Skupin, C. Tolksdorf, E. Gebbhard, R.Zentel, P. Kruger, M. Losche, F. Kremer, Nature 410, 447(2001).

5. D.K. Shenoy, D.L. Thomsen, A. Srinivasan, P. Keller, B.R.Ratna, Sensor Actuat. A 96, 184 (2002).

6. M. Camacho-Lopez, H. Finkelmann, P. Palffy-Muhoray,M. Shelley, Nat. Mater. 3, 307 (2004).

7. M.H. Godinho, A.C. Trindade, J.L. Figuerinhas, L.V.Melo, P. Broguiera, A.M. Deus, P.I.C. Teixeira, Eur. Phys.J. E 21, 319 (2006).

8. A. Sanchez-Ferrer, T. Fischl, M. Stubenrauch, H. Wurmus,M. Hoffmann, H. Finkelmann, Macromol. Chem. Phys.210, 1671 (2009).

9. P. Palffy-Muhoray, Nat. Mater. 8, 614 (2009).10. G.L. van Oosten, C.W.M. Mastiaansen, D.J. Broer, Nat.

Mater. 8, 677 (2009).11. G.C. Verway, M. Warner, Macromolecules 30, 4196 (1997).12. O. Stenull, T.C. Lubensky, Phys. Rev. E 69, 021807

(2004).13. J.V. Selinger, B.R. Ratna, Phys. Rev. E 70, 041707 (2004).14. R. Ennis, L.C. Malacarne, P. Palffy-Muhoray, M. Shelley,

Phys. Rev. E 74, 061802 (2006).15. L. Petridis, E.M. Terentjev, Phys. Rev. E 74, 051707

(2006).16. A.M. Menzel, H. Pleiner, H.R. Brand, J. Appl. Phys. 105,

013503 (2009).17. S. Disch, C. Schmidt, H. Finkelmann, Makromol. Rapid.

Commun. 15, 303 (1994).18. S.M. Clarke, A. Hotta, A.R. Tajbakhsh, E.M. Terentjev,

Phys. Rev. E 64, 061702 (2001).19. J.V. Selinger, H.G. Jeon, B.R. Ratna, Phys. Rev. Lett. 89,

225701 (2002).20. A. Lebar, Z. Kutnjak, S. Zumer, H. Finkelmann, A.

Sanchez-Ferrer, B. Zalar, Phys. Rev. Lett. 94, 197801(2005).

21. P. Pasini, G. Skacej, C. Zannoni, Chem. Phys. Lett. 413,463 (2005).

22. J. Liu, S.Z. Wu, D.P. Cao, L.Q. Zhang, J. Chem. Phys.129, 154905 (2008).

23. G. Cordoyiannis, A. Lebar, B. Zalar, S. Zumer, H. Finkel-mann, Z. Kutnjak, Phys. Rev. Lett. 99, 197801 (2007).

24. G. Cordoyiannis, A. Lebar, B. Rozic, B. Zalar, Z. Kutn-jak, S. Zumer, S. Krause, F. Brommel, H. Finkelmann, Z.Kutnjak, Macromolecules 42, 2069 (2009).

25. B. Rozic, S. Krause, H. Finkelmann, G. Cordoyiannis, Z.Kutnjak, Appl. Phys. Lett. 96, 111901 (2010).

26. H. Yao, K. Ema, C.W. Garland, Rev. Sci. Instrum. 69,172 (1998).

27. Z. Kutnjak, S. Kralj, G. Lahajnar, S. Zumer, Phys. Rev.E 68, 021705 (2003).

28. Z. Kutnjak, J. Petzelt, R. Blinc, Nature 441, 956 (2006).29. H. Yao, T. Chan, C.W. Garland, Phys. Rev. E 51, 4585

(1998).30. B. Widom, J.S. Rowlinson, J. Chem. Phys. 52, 1670

(1970).31. G. Feio, J.L. Figuerinhas, A.R. Tajbakhsh, E.M. Terentjev,

Phys. Rev. B 78, 020201(R) (2008).32. G. Feio, J.L. Figuerinhas, A.R. Tajbakhsh, E.M. Terentjev,

J. Chem. Phys. 131, 074903 (2009).33. A. Greve, H. Finkelmann, Macromol. Chem. Phys. 202,

2926 (2001).34. Z. Kutnjak, Phys. Rev. E 70, 061704 (2004).