Why does Mercury not generate an ESD for a particular bond length, angle, torsion or other measurement in my structure?

Mercury provides the tabulation of measured and standard molecular geometric parameters, along with estimated standard deviations (ESDs) for those values, in various tabs of the More Info dialogue box (such as Bonds, Distances and All Angles). The calculation of ESDs for any of these measurements within crystal structures relies on the underlying atomic coordinates having associated ESDs. Mercury will only calculate an ESD for a measurement if ALL the associated atoms have full atomic coordinate ESDs.

An example where this is highly relevant is a standard organic crystal structure where the atomic coordinates of the non-hydrogen atoms within the model have been freely refined, but the hydrogens have been placed at geometric positions and either set to be ‘fixed’ or ‘riding’. In this example the non-hydrogen atomic coordinates would have ESDs, but the hydrogen positions would not. Mercury will tabulate the measured values of all bond lengths, angles and torsions, but those geometric parameters involving hydrogen will not have ESDs as they would not make any sense scientifically – essentially those parameters have been fixed, not refined.