The method of differential scanning calorimetry (DSC) allows the measurement of the energy flow to and from a sample during a temperature controlled program. The sample is prepared in DSC capsules, which can be either sealed or open, and then submitted to a temperature program. The actual measurement is performed against a control, which normally is an empty DSC capsule, and the energy required to keep both sample and control at the same temperature is plotted vs. the temperature. The resulting curve is called a thermogram. Generally, endothermic events (i.e. events that require energy flow to the sample) are plotted up, while exothermic events (i.e. events that require energy flowing from the sample) are plotted down.

DSC thermogram with definitions of onset temperature (Tonset), peak temperature (Tpeak) and offset temperature (Toffset).

DSC thermogram with definitions of onset temperature (Tonset), peak temperature (Tpeak) and offset temperature (Toffset).


Any peak in the thermogram is characterised by a peak temperature, the onset and the offset temperature. The peak temperature is obviously the temperature at the maximum/minimum of the thermal event. This temperature is highly dependent on the sample crystallinity, crystal size, sample preparation and heating rate, which makes this value unreliable for comparison. The onset and the offset temperatures are defined as the intersection of the tangents of the peak with the extrapolated baseline. The onset temperature should stay unchanged when the peak temperature shifts due to changing heating rate or sample preparation. Thus, this value should be stated and used for comparison of different thermal events.

Figure 2 The area under the curve of a thermal event gives its enthalpy

The area under the curve of any event in the thermogram is proportional to the enthalpy of this event. After suitable calibration with standard materials with known thermal events and their related enthalpy, it is possible to utilise the integral of a thermal event and obtain very accurate information about the energetics of thermal events. Direct integration will give this enthalpy as energy per sample mass (Joule per gram, J g-1). For better comparison, this value is transformed to give energy per moles by multiplication with the molar mass. This value is then given as kJ mol-1 and allows the comparison of enthalpies for corresponding events detached from the actual measured sample. The slope of the thermogram’s baseline gives the heat capacity of the sample. As this characteristic will change with temperature, the baseline normally increases with increasing temperature. Most published thermograms, however, are baseline-corrected and thus show an unchanged baseline slope.

Several thermal events can be detected during the heating/cooling of a sample. Starting from a crystalline sample, the most likely event to observe is the melting point. This event requires energy flowing to the sample and is thus endothermic. Normally, the melting point is the biggest endotherm of a DSC thermogram (by height as well as by area), as the enthalpy required to melt the sample is bigger than any enthalpy connected with e.g. a phase transition.

In general, the slope of the baseline will change during the melting event, as the resulting melt has a considerably different heat capacity than the incident crystal form. In the special case of solvates/hydrates, the detection of the melting point can pose a non-trivial problem, as the desolvation/dehydration prior to the melting has to be avoided. For this purpose, the sample is placed in a sealed capsule to keep possible desolvation/dehydration products in close vicinity of the sample, thus saturating the atmosphere with the volatile component and increasing the activation energy of further desolvation/dehydration. In the case of very stable solvates/hydrates, which show a melting point well above the boiling point of the incorporated solvent/water, the sample can be placed in high pressure capsules, which will hold tight even at pressures of up to 150 bar.

Another frequent thermal event is the phase transition. For a polymorphic phase transition, the thermal event can be either endothermic or exothermic depending on the nature of the transition. Two types of phase transitions can be distinguished: the enantiotropic and the monotropic transition. Enantiotropic transitions are reversible and are normally observable as endothermic events in the heating cycle and exothermic events in the cooling cycle. Monotropic transitions are irreversible and can be observed as exothermic events.

Desolvation/dehydration events can normally be detected as endotherms, as the sample takes up energy to evaporate the released solvent. These events are usually broad and well distinguished from polymorphic phase transitions or melting processes. This is due to the requirement in energy to evaporate the released solvent from the sample. Nonetheless, desolvation events can become sharp for very stable solvates which release the solvent at a temperature where its vapour pressure is sufficiently high to evaporate the released solvent immediately. If the desolvation/dehydration event occurs very much above the boiling point of the solvent/water, the evaporation endotherm is no longer detectable and the thermogram will only show the recrystallisation/reorientation event of the desolvated/dehydrated crystal form as an exothermic peak. Desolvation/dehydration can overlap with the melting endotherm of the solvate/hydrate, which manifest as a shoulder on the higher temperature side of the desolvation/dehydration endotherm.

Amorphous samples show a range of characteristic events. First of all, the glass transition temperature can be detected as a localised change in the baseline. During this event, the super-cooled melt or glass undergoes a second order phase transition and becomes a liquid. The glass transition temperature, however, is strongly influenced by the thermal treatment of the amorphous sample and can additionally be shifted to up to 100 °C by the inclusion of 2 – 4% of water.

The crystallisation of an amorphous sample or the melt can be detected as an exothermic event, in which the enthalpy of crystallisation is released from the melt. This event is highly kinetically controlled and can vary widely in onset temperatures. The heat of crystallisation is comparable with the heat of fusion and can be used to identify the crystallisation of single phases. In the case of a mixture crystallising, however, the heat of crystallisation will be a mixture of the two individual heats of crystallisation and a deconvolution of the two will not be easily achieved.

More often than not, samples will decompose at higher temperatures. This can be detected as strong exotherms followed by an unstable baseline, normally following directly above the melting endotherm. Since the sample itself is chemically altered, the enthalpies measured for these events are meaningless unless it can be unambiguously established which compounds are present and in which relative rations.