The hydroxyl group content in nine different types of silicas was determined by thermogravimetry (TG). The TG measurements were performed by heating the silica samples (15–30mg) at a rate of 5°Cmin−1 from room temperature to 1250°C in an argon flow. The dehydration and dehydroxylation phenomena (two mass loss steps) were distinguished with the help of the differential TG curve (DTG). The temperature where physisorbed water was assumed to be totally released was dependent on the silica and was found to range from 100 to 130°C. The OH group content was calculated from the second mass loss step by assuming that there were only siloxane bridges left in silica at 1250°C. The total hydroxyl group content was calculated from the entire mass loss of the second step. The remaining amount of OH groups at any temperature below 1250°C could also be calculated and was done in this study at 550 and 1000°C. When comparing the total OH group content and the OH group content at 550°C determined by the TG method with the results of calcined samples measured by 1H MAS NMR, they were found to be in a good agreement with each other. However, a significant difference was observed between the methods regarding the remaining OH group content at 1000°C.

Surface reactions between bis(cyclopentadienyl)zirconium dichloride (Cp2ZrCl2) and porous SiO2 surface were studied by preparing series of samples by saturated gas–solid reactions in an atomic layer epitaxy (ALE) reactor. Partially dehydroxylated silica supports were obtained by preheating. Cp2ZrCl2 reacted mainly with the isolated OH groups. The availability of OH groups influenced the bonding mode of the zirconocene complex which was mainly singly bonded on silica preheated at 600°C, while preheating at 350°C resulted in multiple bonding as revealed by NMR and elemental analyses. There was no evidence for surface alkylation or chlorination.

A method based on diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was developed for the rapid determination of the hydroxyl group content in calcined silica. Thermogravimetry (TG) was used as a reference method to determine the hydroxyl concentrations. The data pretreatment was performed with orthogonal signal correction (OSC) and it was followed by subsequent partial least squares (PLS) modelling. The calibration model was validated by predicting the hydroxyl content of samples not used in the construction of the models. The OSC‐PLS model was able, with sufficient accuracy, to predict the hydroxyl content of several calcined silicas with different physical properties. Copyright © 2000 John Wiley & Sons, Ltd.