Determination of the Glass Transition Temperature of Difficult Samples by Flash DSC

Crystalline pharmaceutical substances often decompose immediately before or during melting. To determine the glass transition temperature, the substance must be melted and then cooled as rapidly as possible so that decomposition and crystallization do not occur. In many cases, the heating and cooling rates of conventional DSCs are not high enough for this purpose. The METTLER TOLEDO Flash DSC however offers new possibilities. This is illustrated in this article using prednisolone as an example.

Introduction

Amorphous forms of active pharmaceutical ingredients (APIs) are often preferred in order to obtain the highest level of bioavailablity. Amorphous APIs are only stable in the glassy state below the glass transition temperature (Tg). Above the Tg, they can crystallize. This can have a large effect on bioavailablity.

APIs are often crystallized from solution in order to obtain the purest form. Conversion to the amorphous state can be achieved by grinding the crystals, whereby the temperature of the material being ground must not rise above the Tg during grinding [1, 2]. Knowledge of the glass transition temperature of amorphous APIs is therefore important both from the point of view of storage and from process engineering.

To determine the glass transition temperature of a (crystalline) starting material, the material must be melted and then cooled as rapidly as possible so that no decomposition or crystallization occurs. In many cases, the heating and cooling rates of conventional DSCs are inadequate for this purpose.

The Flash DSC offers new possibilities with heating and cooling rates of up to 40,000 K/s (heating) and 4,000 K/s (cooling). This allows a substance to be heated up to several hundred degrees and then cooled within a few milliseconds [3, 4]. In this short time interval, practically no decomposition can occur. This makes it possible to determine the glass transition temperatures of crystalline substances that would have decomposed in a conventional DSC during the comparatively slow melting process [5].

In this article, we illustrate the procedure using prednisolone as an example.

Prednisolone is a synthetic corticoid used to treat inflammation. Prednisolone is available as the anhydride in two stable polymorphic forms and as a sesquihydrate [6]. The expected solid-solid transition between the two anhydrides should occur between 120 and 130 °C. This has however previously not been observed [6].

The form stable at room temperature (Form I) melts between 236.5 and 239 °C [6]. The high-temperature form (Form II) melts between 224 and 228 °C [6].

The melt is thermally unstable and decomposes [7]. Form I is however stable at room temperature and is commercially available.

Conclusions

The amount of information that can be gained from DMA measurements can be greatly increased when the normal temperature scan measurements are complemented by frequency scan measurements and measurements in which the displacement amplitude is varied.

Thanks to the high heating rates that can be achieved with the Flash DSC 1, it is possible to separate the melting and decomposition of substances whose decomposition begins together with or immediately before melting. The glass transition temperature of such substances can be determined after rapid cooling of the non-contaminated melt.

n the case of prednisolone, a glass transition temperature of 118 °C was measured at a heating rate of 100 K/s. Prednisolone can therefore easily be made amorphous by grinding at room temperature. The temperature of the prednisolone during grinding must not however exceed about 112 °C (onset of Tg). The melting point of the form of prednisolone stable at room temperature (Form I) was also be determined (257.2 °C). This is appreciably higher than the temperature of 236.5 to 239 °C published in the literature [6]).

Determination of the glass transition temperature of difficult samples by Flash DSC | Thermal Analysis Application No. UC 453 | Application published in METTLER TOLEDO Thermal Analysis UserCom 45