You can try with a thermo gravimetric analysis. In general, the adsorbed water is released at low temperature (about 100°C), while the lattice water at higher temperature, if you are lucky you can separate the two thermal events.

You can try with a thermo gravimetric analysis. In general, the adsorbed water is released at low temperature (about 100°C), while the lattice water at higher temperature, if you are lucky you can separate the two thermal events.

yes you can try thermo gravimetric analysis method as Marcella Bini has mentioned.

Thank you for your suggestions. Unfortunately, as mentioned in the post, both types of water leave the sample simultaneously upon heating. 

hi Michal! just a question, can't you determine the crystal structure of your sample?

and have you also tried with DSC? does it occur the same as with TGA? 

IR spectroscopy. Adsorbed and structural (please do not use the term "lattice" that means nothing else than a regular array of points) should have different vibrational frequencies.

You may also try solid-state NMR. There it should be possible to see differences to the local environment.

Do you know the crystal structure? Is this a channel like structure where water can enter and leave easily? Is the DVS analysis giving you a hysteresis upon dehydration?

Dear Michal,

                                   I agree with Marcella and Paolo point of view. Nevertheless, I would like recall that more precise information are needed, according to the relationships of water with your crystalline sample: 

a)- water can be ad-sorbed on the sample surfaces

b)- water can be incorporated in the crystal bulk (during growth) in three ways:

b1)  as molecular H2O, building the lattice, as in the case of gypsum CaSO4.2H2O

b2) as -OH group, building the lattice, as in the case of PbCl(OH)

b3) as liquid water, captured into the crystal bulk during growth in the shape of solvent 3D-inclusion

In case a),  water is highly unstable and escapes from the crystal surfaces "continuously" with temperature, as can be seen through DSC and TGA curves

In cases b1 ) and b2 ),  water is more or less strictly bound to the lattice. In both cases, it leaves the crystal with a sharp phase transition (which is revealed by sharp peaks in DSC and TGA). Moreover, IR  absorption frequencies are well defined,  for both stretching and bending vibrations, and typical for each crystal (or amorphous) species.

In case b3) IR spectra are those of liquid water (a large absorption band in the near IR). Further, there are sharp peaks in DSC at the temperature of "decrepitation" where the crystal breaks when water inclusions explose. 

All the best !

Dino

Could you attach your TG measurement.. The temperature at which   the water releases from the sample has some relation  with how strong is a bonding between water and sample. If you have a patience try to use lower heating rate in TG measurement in order to decouple both process.  Use also derivative of TG signal, start at 50 deg  . I will be surprised if both process occur at one and same temperature.

Thank you very much for all your valuable suggestions. We will most probably start with TGA measurements at different heating rates to see if there is any usable differences in water evolution temperature ranges. IR was also considered, but the substance is typically crystallized as a mixture of 8 isomers, which also highlights possible complications for SC-XRD approach.

You can also try Variable temperature PXRD at low rate of heating. This can differentiate both events based on change in powder pattern. Since loss in lattice water may lead to different crystal structure.