Vacuum setup with the heating/cooling and the deposition system, left. The right picture shows the copper block with the cooling tube and the mechanical shutter which can be used to shield the sample sensor before deposition and during the measurement. The reference sensor is located opposite of the sample sensor but is not seen because it is protected by the copper.
Photographs of the calorimetric chip sensor XI39390 at three different magnifications. Left: chip mounted on TO5 housing, middle: the chip, and right: the central part, including the active area.
Scheme of the electronic setup of the differential ac calorimeter. The oscillating voltage of the lock-in amplifier drops over the two inner heaters of the chip sensors and the known resistor R, which are connected in series. The voltage drop at the inner heater of the sample sensor as well as the voltage drop at the reference resistor is amplified by instrumentation amplifiers (amp) and measured by an ADC. From these values the heater resistance and the applied power can be determined. The differential thermopile voltage is directly measured by the digital lock-in.
The temperature amplitude derived from the measured differential thermopile voltage and the phase angle (a) and heater resistance and rms power (b) for a temperature scan experiment of an ordinary toluene glass with the experimental parameters given in the figure. The heater resistance was fitted (bold black line) and then used to calculate the temperature using a calibration function (see Figure 5).
Temperature calibration of the heater resistance. The black dots show the calibration points determined from the three transitions of cyclopentane. The black line is the resulting calibration function. The open diamonds and stars correspond to independent calibrations with the frequency dependence of the dynamic glass transition temperature of toluene and ethylbenzene, using VFT equations taken from Refs. 49 and 50, respectively. The inset magnifies the range between 121 K and 130 K.
Scheme of the central part of the membrane including the active area (a). (A) indicates the white aluminum electrical leads, (B) the thermocouple stripes, and (C) the free standing silicon nitride membrane. The diagonal black line indicates the line of symmetry. The aluminum interconnections, heaters and thermopiles as used in the model are shown in (b) to (d) in more detail. The thermopiles and the heater stripes are made from 300 nm thick doped polysilicon. They are in green in (a) and in blue in (d). The inner heater, providing the power, is marked as the dark area in (d).
Steady state comparison between the experimental data (open circles) and the model data (dots). The power was applied to the inner heater and in the experiment the temperature was recorded by the heater resistivity. The silicon frame temperature was 110 K. The insets show, for illustration only, the normalized temperature distributions from the FEM calculations for a dc-heated chip sensor in vacuum. The scaling is as follows: X and Y directions span 0 to 900 μm and Z is normalized temperature between 0 and 1.
Comparison of simulated (circle, square) and experimental (triangle, diamond) data points for an empty sensor and a sensor loaded with a 390 nm toluene sample. Temperature amplitude is shown in (a) and the phase angle between temperature and power in (b).
Differential temperature amplitude (a) and phase angle between temperature and power for the differential signal (b) for a 390 nm toluene film. Experimental differential data are shown as circles and the modeled data are shown as triangles.
Experimental and modeled data for three films of slowly cooled toluene glasses of different thickness. The differential amplitude (a) and the phase angle between power and temperature (b). Experimental data are black, light gray, and gray curves and the corresponding modeled data are squares, diamonds, and dots, respectively. In (b) the phase data are shifted by 3° (open symbols) to compensate for additional time constants from the electronics.
The black curve corresponds to the measured differential temperature amplitude during deposition of toluene and the points show the modeled data with different thicknesses. The modeling and experiment were done at 106 K and a thermal frequency of 20 Hz. The deposition rate was held constant at about 1.5 nm s−1. The inset shows the temporal evolution of the sample thickness.
Measured differential temperature amplitudes (dashed, right axes) and specific heat capacities (solid, left axes) for a slowly cooled toluene film from ac calorimetry (solid line without dots) and a bulk sample by adiabatic calorimetry62 (solid line with dots). The calculation of the specific heat capacity from differential temperature amplitude was performed by the FEM assuming a constant mass per unit area corresponding to a film thickness of 390 nm at 106 K. The inset shows the ratio between specific heat capacity and differential temperature amplitude. The AC-calorimetric temperature scans were performed with a heating rate of 0.7 K min−1 and a thermal frequency of 20 Hz.
Specific heat capacities for an as-deposited thin toluene film (solid line) and the reheating of the ordinary glass (dashed line). The recalculation from differential temperature amplitude to specific heat capacity was performed by the FEM assuming a constant mass per unit area corresponding to a film thickness of 390 nm at 106 K. Results from adiabatic calorimetry are shown for amorphous62 (squares) and crystalline66 (diamonds) toluene. The film was deposited at a substrate temperature of 106 K and a deposition rate of 1.5 nm s−1. The ac-calorimetric temperature scans were performed with a heating rate of 0.7 K min−1 and a thermal frequency of 20 Hz. The vertical arrow at 120 K indicates the isothermal transformation from the stable glass towards the supercooled liquid above the static T g as shown in Figure 14.
Transformation of a stable glass into the SCL as indicated by the arrow at 120 K in Figure 13. Inset: ΦSG is the portion of the stable glass remaining at any point in time. ΦSG = 1 indicates the sample that is completely stable glass whereas ΦSG = 0 indicates the sample that is fully transformed into the SCL.
Material properties of the membrane and the other elements on the chip sensor. Toluene was used as the sample.
Article metrics loading...
Full text loading...