Dear Geoffrey, Thanks for your response. The structure was optimized at MP2/aTZ level. I use XYZ-format because it is the main format in MolPro QM program. The problem with C=N bond was met in several compounds listed bellow. Methanimine EthanImine VinylImine Benzenemethanimine and Imidazoline
The definition of sp2 nitrogen via just angle seems to be little oversimplified. There are cases when angles are significantly distorted, especially in ring compounds. See for example, the C-N-CH2 angle in Imidazoline is only 103. 11 Imidazoline.xyz C -0.88262 -0.68699 0.03537 N -0.92913 0.64788 -0.34690 C 0.38783 1.18838 0.01773 C 1.22818 -0.09818 -0.01329 N 0.27522 -1.19770 0.25046 H -1.80129 -1.25731 0.09109 H -1.74753 1.17193 -0.07458 H 0.36896 1.62120 1.02219 H 0.72296 1.93780 -0.69462 H 2.03179 -0.10284 0.71867 H 1.66075 -0.25683 -1.00408 Thank you though for pointing the reason. Igor > > I came across a problem of converting XYZ to HIN format for imine > > compounds. In more detail, the double C=N bond in imine series is > > determined as a single bond. (As a result, atom typification for Merck > > force field is completely screwed up.) > > The problem here is that the H-N-C bond angle is 110 degrees. It's not > anywhere close to "looking like" an sp2 nitrogen. So the PerceiveBondOrders > is left assuming there's some missing valence somewhere. > > It might be worth having some sort of special case for indicating that "all > hydrogens are included, so funny angles are OK" but I'm a little hesitant > because this doesn't come up very often. > > Out of curiosity, where did this file come from? Some calculation? Why did > you use XYZ? I'm wondering why the angle is so off and whether a different > format would preserve the bond information for you. > > If others have bad experiences with imines, I'd like to know. > > Thanks, > -Geoff > > ------------------------------------------------------------------------------ November Webinars for C, C++, Fortran Developers Accelerate application performance with scalable programming models. Explore techniques for threading, error checking, porting, and tuning. Get the most from the latest Intel processors and coprocessors. See abstracts and register http://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk _______________________________________________ OpenBabel-discuss mailing list OpenBabel-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
