Moisture sorption/desorption

When dealing with hydrates, or indeed any crystal form, it can be interesting to investigate its stability at different relative humidities (RH). Especially when the hydrated crystal form will be processed, it is necessary to know its range of stability, i.e. below which RH level it will release the incorporated water or above which RH level it will transform to a higher hydrated form or even liquefy. Therefore, moisture sorption/desorption analyses are carried out, involving the measurement of equilibrium water content as a function of RH at a specific temperature (moisture sorption/desorption isotherms). The easiest but also most tedious way to gather this information is to store a sample in a desiccator over a saturated salt solution. Salt solutions can generate a distinct humidity level. The RH levels for selected salts over a wide temperature range are listed here (http://www.omega.com/temperature/z/pdf/z103.pdf). The water content can be measured by weighing the sample, as the only mass gained is due to water. More sophisticated approaches are conducted using dynamic vapour sorption (DVS) instruments, which consist of a balance in an air-conditioned chamber. By purging the chamber with air or nitrogen of known RH level, the weight of the sample can be monitored over time and the point of equilibrium easier and quicker measured. Is equilibrium reached, the instrument changes to the next programmed RH level. Having the possibility to continuously record the weight of the sample, a sorption isotherm for each step can be recorded, from which the kinetics of this step can be extracted.

From the measured sorption/desorption isotherms, characteristics of the hydrated sample can be extracted, which is not accessible by any other technique. For example, it can be determined whether the hydrate shows stoichiometric or non-stoichiometric behaviour, or even a mixture thereof. Stoichiometric hydrates will show a distinct RH level, below which they release all or parts of the incorporated water in a single step. This single step is accompanied by a phase transition (see Definition). If the same sample is subsequently submitted to a sorption cycle, the sorption isotherm will most likely show single step rehydration behaviour. Between the dehydration and the hydration, however, can be a more or less emphasised hysteresis, a kinetically controlled delay of the two events, which thermodynamically should occur at the same RH level. Non-stoichiometric hydrates however have different water contents at different RH levels and thus show a smoother water release in the desorption isotherm. As these hydrates do not undergo phase change upon dehydration, a following sorption step will follow the same slope as the desorption step. There is no hysteresis between the desorption and the sorption isotherm. Hydrates can show one or more of the described steps and indeed have stoichiometric and non-stoichiometric characteristics depending on their hydration state.

Unsolvated crystal forms can show a water uptake at higher RH levels, or indeed can be unaffected by change of humidity. Especially in the pharmaceutical sciences the latter case is preferred, as a change in crystal form from an anhydrous to a hydrated modification can cause a changed efficiency and safety of the formulation and have an impact on the consistency and integrity of the application form (e.g. capping or disintegration of tablets).

If solvates are subjected to a sorption/desorption analysis, several different behaviours can occur. The solvate can stay stable up to a certain RH level, above which it releases the incorporated solvent in a single step and either transforms to a unsolvated crystal form or a corresponding hydrate. Another behaviour would be that the solvate takes up some water and then releases the incorporated solvent in a single step. This is due to the destabilisation of the absorbed water, while the “dry” solvate can be reasonably stable. The solvent loss can sometimes take place over two or more cycles of increasing and decreasing the humidity. In addition, solvents with low vapour pressure can stay in the sample pan, i.e. they do not evaporate into the gas phase, which leads to a re-incorporation and reformation of the solvate when the RH levels are sufficiently decreased.

Amorphous samples normally show a rather high hygroscopicity, as moisture can dissolve in the glass easier than being absorbed into a crystalline material. However, moisture can induce crystallisation of amorphous samples. This is detectable by a rapid weight gain in the sorption isotherm followed by an abrupt weight loss, when the phase change from amorphous to crystalline material reduces the hygroscopicity of the sample. The resulting crystalline phase can be a hydrate but might also be an anhydrous material or a mixture of crystalline phases, depending on the kinetics of the crystallisation process.