There are a lot of possible sources of error: differences in the true temperatures, inappropriately assessed wave lengths, instrumental abberations. However, giving the absolute change of the of the lattice parameter is relatively meaningless. You actually have to refer it to the abslute of the lattice parameter. If your lattice parameter (a cubic translation lattice???) is 1000 Angstrom, your error is small, but if your lattice parameter is 2.8 Angstrom, your error is large. The decisive quantitivy is Delta_a/a and not Delta_a.

Have you included an off-axis shift parameter in the refinement?

Best

Arman

There are a lot of possible sources of error: differences in the true temperatures, inappropriately assessed wave lengths, instrumental abberations. However, giving the absolute change of the of the lattice parameter is relatively meaningless. You actually have to refer it to the abslute of the lattice parameter. If your lattice parameter (a cubic translation lattice???) is 1000 Angstrom, your error is small, but if your lattice parameter is 2.8 Angstrom, your error is large. The decisive quantitivy is Delta_a/a and not Delta_a.

Cell parameters of one sample at same temp(300K) shall be the same irrespective of diffraction tool. The precision and accuracy with which they should match (within experiment limitations) are decided by instrumentation (lab or synchrotron) including analysis tool (XRD peak profile chosen in fitting e.g. pV or TCH etc) and X-ray source parameters. Generally, these parameter do not lead to 2nd decimal inaccuracy as you mentioned. In such situation, it is better to cross check instrument calibration and determine the accuracy and precision (0.1 or 0.01 or 0.001 angstrom or better) with which one is able to reproduce NIST SRM lattice constants at both Kek facility(wavelength 0.77 A) and Rigaku(1.54 A) . In this process instrumental aberrations need to be considered, and same procedure to be adopted for analysing sample/s.