Is it possible to get glass transition increases, while elastic properties decrease, optical energy gap decrease and FTIR result show increment of non bridging oxygen when adding some modifier into glass former?
by definition, the modifiers create non-bridging oxygens (NBO) in the glass former network, hence the increment of NBO you observe is absolutely normal and expected. The creation of NBO and the fact that the modifier-oxygen bonds have (in most cases) lower bond strength than the glass former-oxygen bonds, usually leads to lower elastic properties, as you observed. However, the increase in NBO is usually accompanied by a decrease in the Tg of the glass, as the glass network becomes less "rigid" (you would also expect lower viscosities for the glass containing the modifier for instance). But the value of Tg measured can vary as function of the thermal history of the glass (melting temperature, quenching rate, annealing...) and the heating ramp used during the measurement. Thus, variations in one (or several) of these parameters could explain the higher Tg measured, while a lower Tg would be expected for a glass with more NBO prepared in the exact same conditions.