Thermochimica Acta 492 (2009) 51–54 Contents lists available at ScienceDirect Thermochimica Acta journal homepage: www.elsevier.com/locate/tca Glass transition in vapor deposited thin films of toluene E. León-Gutierrez a , G. Garcia a , M.T. Clavaguera-Mora a , J. Rodríguez-Viejo a,b,∗ a Grup de Nanomaterials i Microsistemes, Departament de Física, Universitat Autònoma de Barcelona, Campus UAB, 08193 Bellaterra, Spain b MATGAS Research Centre, Campus UAB, 08193 Bellaterra, Spain article info Article history: Available online 3 June 2009 Keywords: Nanocalorimetry Glass transition Toluene Relaxation time Thin films abstract We report on nanocalorimetric measurements of 50 nm thick toluene films at heating rates spanning 600 to 8 × 10 4 K/s. The films are grown from the vapour at 90K directly onto the nanocalorimetric cell. The kinetic and thermodynamic stability of as-deposited films is higher than the stability of films cooled from the supercooled liquid at 2000 K/s. We also show that at those heating rates, the calorimetric T g does not correlate with the relaxation time obtained by dielectric spectroscopy. © 2009 Elsevier B.V. All rights reserved. 1. Introduction Many reports in the literature have dealt with the glass transition associated to organic glasses [1–5 and references therein]. While there are many organic liquids that may form a glass upon cooling, only few are simple enough that their intramolecular degrees of freedom can be ignored [6]. One of such simple molecules is toluene in which the only low-frequency intramolecular vibration is that of methyl rotation. Toluene also exhibits extremely high fragility with a fragility index m = 105 [7]. The super-Arrhenius depen- dence of the relaxation time when approaching the glass transition in supercooled toluene is well documented by several experi- ments including dielectric spectroscopy (DS) [7], deuteron-spin lattice relaxation [8] and microwave spectroscopy [9]. An ultra- fast microcalorimetric study performed on micron thick toluene films obtained from the vapour has also been recently reported by Chonde et al. [10]. The heat capacity showed an unusually broad glass transition which may not be related to the fast rates but to the existence of temperature gradients within the sample, however the effect on the onset temperature was below the ±5 K tempera- ture uncertainty of the measurements. Interestingly, these authors found that the use of a relaxation time of 100 s at T g for a heating rate, q, of 10K/min did not provide a good comparison with the DS data of Döss et al. Instead they use the relation  =1 K/q to find a good correlation of their data with the relaxation time obtained by DS. ∗ Corresponding author at: Grup de Nanomaterials i Microsistemes, Departament de Física, Universitat Autònoma de Barcelona, Campus UAB, 08193 Bellaterra, Spain. E-mail address: javier.rodriguez@uab.es (J. Rodríguez-Viejo). Recent measurements by Wang and Richter [11] provide data on the comparison between calorimetric and dielectric relaxation measurements on non-Debye and Debye-type liquids. For non- Debye liquids at moderate heating/cooling rates of 20 K/min the difference between the onset of the glass transition measured by calorimetry with respect to the kinetic glass transition (determined as the temperature at which the dielectric relaxation time equals 100 s) was between 1 and 3 K depending on the fragility of the liq- uid. Based on this small difference, the authors indicate that the definition of the kinetic glass transition temperature at  = 100 s, although arbitrary, provides a satisfactory correlation with calori- metric data. Johari and Aji [12] comment on the proper use of relaxation times determined using different techniques to com- pare ultraviscous liquid dynamics since they may be associated to different molecular/diffusive/rotational processes. They also point out that the use of a relaxation time of 100 s for the onset of the calorimetric glass transition associated to a heating rate of 20 K/min may not be a universal value to apply for the glass transition of all materials [12,13]. Nanocalorimetric measurements using microfabricated chips permit a good control of temperature differences within the sam- ple region, a negligible thermal lag and a differential scheme that enables an extremely high resolution [14–17]. Previous works by Efremov et al. [18,19] have shown the suitability of the technique to analyze the glass transition on ultrathin polymer films. Glassy materials prepared from its vapour are usually con- sidered to be more unstable than similar glasses obtained by quenching [20]. In addition, organic glassy films grown at low tem- peratures through the vapour phase have rarely been the object of analysis due to the difficulties inherent to in situ calorimet- ric measurements. The pioneering work of Swallen et al. [21] has shown the extraordinary stability of glasses of 1,3-bis-(1-naphthyl)- 0040-6031/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.tca.2009.05.016