This hadn't really occurred to me until an experimentalist with no experience in computational chemistry asked me "to how many significant figures is Gaussian accurate when measuring distance?"
I think Gaussian's internal algorithms compute in Bohr radius but outputs in angstroms. In GView until now I was using the exact numeric-figure when measuring distance., e.g., 0.95708 angstroms, which is 1.8086190 Bohr radius. However, (I am not a physicist) what is 0.8, 0.00, 0.008, 0.0006, 0.00001 etc..., of a Bohr radius? Do these significant figures make any sense? Therefore, how many significant figures in an angstrom make sense.
I fear that this question might get a bit philosophical. However, I just wish to know to what signifiant figure ought I publish using distances derived in Gaussian.
Thanks
Anthony