As is shown in Ref.1, the trans isomer is much more stable, at least at molecular level. Based on the crystal structural data from Ref.2 and atomic radii of O and N from wikipedia and some little calculation, it seems that steric effects can be neglected. What else factors can contribute to this phenomenon? Or can it be rationalized with LFT, CFT, or QMOT?
The Journal of Physical Chemistry B, 2004, 108(6), 2098–2102. doi:10.1021/jp0364497
Australian Journal of Chemistry, 2002, 55(5), 331-341. doi:10.1071/ch01091