Figure 4.2: A typical DSC thermogram
sensitivity of 0.3meal/ sec for full scale deflection and is normally used with mil ligram quantities of samples. The operational temperature range is from 5°C to 600°C.
It is known [102] that the slope of the leading edge of a melting endotherm for a pure sample is given by (1 / R 0)(dT/dt), where (dT/dt) is the scanning rate. Variation in 1 / R 0 affect the peak shape, but the peak area is unchanged. For reproducible and accurate enthalpy data fewer conditions need to be imposed than given above since the area of integration of a plot of dH/ dt versus time is less dependent on the variation of R a than the transition temperature value.
4 .8 .1 E v a lu a tio n o f tr a n s it io n t e m p e r a t u r e s an d e n th a lp ie s
In order to determ ine the transition tem p eratu re the n atu re of th e transition has to be first established. For an isotherm al transition (e.g. eutectoid and peritec- toid reactions in binary systems [103]) the transition tem p eratu re is given as an intercept of th e interpolated base line and the slope of the leading edge of the endotherm (tem p eratu re T\ in Figure 4.2). This value is given as the onset by th e TADS DSC S tandard Program .
Transitions corresponding to m onovariant equilibrium in binary systems are non-isotherm al. In this case, as suggested by Grabielle-M adelm ont and Perron [104], th e tem peratures corresponding to the onset and completion (T2 and T'2in Figure 4.2) on th e heating curves are to be considered.
T he enthalpies of transition are determ ined from the areas under transition endotherm al peaks using th e TADS DSC S tandard Program . The procedure used is presented graphically in Figure 4.2.
Typical calorim etric heating curves in a binary system are shown in Figure 4.3. T he peaks corresponding to isotherm al transitions are relatively sharp and n ar row w ith the m axim um am plitude at the eutectoid com position, whereas for the nonisotherm al transitions the observed peaks are broad.
4 .8 .2 S p e c ific h e a t m e a s u r e m e n t s
T he heat flow rate (dH/ dt) which is measured in differential scanning calorim etry versus te m p eratu re is proportional to the instantaneous specific heat cp of the sam ple [105]
(4.9)
A ♦ D
t T t ^
C 0 M P 0 S T I 0 H
Figure 4.3: Correlation of calorimetric heating curves with the features of a binary phase diagram (after Grabielle-Madelmont and Perron [104]).
Figure 4.4: Method of the specific heat measurement (after O’Neill [105]). accuracy of the specific heat measurements with DSC (of the order of 0.3%, as demonstrated by O’Neill [105] and Fransson and Bäckström [106]) the elimination of the ordinate calibration and the heating rate in equation 4.9 is required. The method used in this work is illustrated in Figure 4.4. In order to measure the specific heat of a sample, three consecutive runs are programmed as follows:
• Empty aluminium pans are placed in the sample holders and the heat flow is measured in the required temperature range using the appropriate heating rate. The thermal cycle begins and ends isothermally.
• The same measurement is repeated with the material of known specific heat placed in the same aluminium pans as used for the base line.
• The sample measurement (programmed the same way) is carried out using the same pans.
T he isotherm al com ponents during the m easurem ents are identical for all three runs. T he ordinate deflections are however different, owing to th e different values of specific heat and masses of the samples. In order to calculate the specific heat th e ord in ate deflections from the base line are com pared at the same tem perature. From equation 4.9 one obtains for each run:
C
-L = ^ (4.10)
c'p m y'
T he Perkin-Elm er Specific Heat Software allows the calculation of the specific heat versus tem p eratu re assuming long term energy-calibration stability of DSC. T he calibration constant required for the program was obtained using specific heat stan d ard s and was checked before and after the series of the m easurem ents. T he DSC-4 calorim eter used in this work had stability b e tte r then 0.1% w ithin several days of operation.
P u re indium m etal and Laurie acid were used for the tem p eratu re and en thalpy calibration [107]. Benzoic acid, alum inium oxide and distilled water were used as specific heat stan d ard s [108]. Sample preparation m ethods are described in Section 4.1. Portions of samples weighting about 10mg were transferred from th e am poules to alum inium pans (supplied by Perkin-Elm er) and sealed. Samples of th e tern ary system surfactant /w ater/cyclohexane were sealed in a satu rated atm osphere of cyclohexane to prevent the change of the sample composition. Samples were weighed before and after calorim etric m easurem ents in order to test th e quality of seal. Only d a ta taken using perfectly sealed pans were pro cessed.