In my opinion it depends in what temperature region the transition are occuring, intensity and the even the molecular distribution of the neat DMA, is it monodisperse ? It could be beta-transition (dimethyl amide). It could be some efect relative to the cast, curing process, since "thermal history" is present in DMA, DTMA if you did not do a aneling process, in that case maybe a DSC analisys of second heating process could help you to elucidate such questions.

A DSC of individual components could help you elucidate the contribution of each component. shoulders and shifting in Tg around know values could be due to mixture thermodanamic's effects (phase separation, blending) or be a effect of cross-linkage degree. I sugest you to analise the polymer and the crosslinker separate, if you already didn't it it, maybe it could spots some new answers.

Dear all, more likely this due to heterogenity in crosslinks density. Different responses from highly crosslinks density outer phase and low crosslinks density inner phase. Decay of UV radiation intensity while penetrating the matrix. My Regards

Agree with Abdelkader: crosslinked amorphous phase has lower segmental mobility than uncrosslinked phase, thus requiring higher temperature for glass relaxation to occur and showing higher glass transition peak temperature

Dear all, thank you for your answers. The difference in segmental mobility could be the reason for the two different Tgs. I will have a look again and might come back to you, if I have further questions.