Gypsum converts into hemihydrate at around 100-140°C, and into anhydrite at around 140-150°C. Note that these ranges can vary depending on the sample weight, heating rate, open/closed vessel, etc.
Besides, by coupling TGA and XRF you can verify the presence of contaminants such as dolomite (see the MgO content and weight loss at around 600-800°C).
I totally agree with Paulo Ricardo de Matos , TGA + XRF are very good options indeed.
By using closed vessel in TGA with an adequate heating rate (10 or 20°C/min) you can estimate the amount of gypsum and hemihydrate in your sample. For more details regarding this, please refer to the book "A Practical Guide to Microstructural Analysis of Cementitious Materials", Ed. Karen Scrivener, Ruben Snellings and Barbara Lothenbach.
As mentioned by Paulo Ricardo de Matos , gypsum (CaSO4.2H2O) loses 1.5 H2O around 100-140°C and the other 0.5 H2O around 140-150°C. Therefore, a simple idea is to measure the weight loss up to 150°/200°C. You can do this with an oven, for example. The theoretical weight loss of gypsum is 20.91%. Therefore, if your gypsum is 100% pure, it should lose 20.91% of mass up to 150/200°C.
Obviously, this method is not accurate, but it can have you an idea of how pure is your material. To minimize the influence of moisture in this analysis you can first dry your sample at 40-50°C and then compare the weight at 50°C and at 150/200°C.
You didn't gave any information about your sample. If there is no overlapping between calcium sulfates in DTG curve. The TGA will be good option. Or you can also try FT-IR with TGA.
Dear Muhammad Rashid . I agree with Paulo Ricardo de Matos and José da Silva Andrade Neto . With XRF you can get a good idea of the purity of your gypsum. With the oxide configuration, almost all of the material should be CaO and SO3, with an approximate ratio of 40% CaO and 60% SO3. This is obtained by summing the atomic weights of the elements that make up gypsum in its anhydrous form (CaSO4).
Additionally with TGA you can verify that the dehydration temperature does not vary for both Hemihydrate and Anhydrite and that the total mass loss of water is 20.91%.
With this you can get an idea of how pure your material is.