menuadvice on CIFs containing completed validation response forms (VRFs)
Validation Co-editors will receive papers containing one or more A alerts, along with the corresponding Validation Response Form(s). If the Co-editor determines that the alerts are sufficiently serious to prevent the paper progressing, the Co-editor may have to inform the authors that their paper has been rejected, or that significant remedial action is required. Conversely, the Co-editor may determine that although critical information is incorrect or missing this can be readily supplied. In all the above cases, authors should be contacted as soon as possible, but in other circumstances the Co-editor may decide that the Alerts can be best dealt with as part of the overall refereeing process.
A further step is to determine whether the authors have supplied a scientifically valid explanation for the alert: in some cases there may be no acceptable explanation. Authors may occasionally offer irrelevant, invalid or circular responses.
The A alerts may be of any type but a large proportion of them are accounted for by problems concerned with (a) data resolution, completeness and quality, (b) various aspects of refinement, including disorder, and (c) the treatment of hydrogen atoms. Problems with absorption corrections now occur much less frequently than previously. The frequent occurrence of problems with data resolution and completeness raises a number of issues, given that the required remedial action may well extend to further data collection, which may not always be feasible. Nevertheless, the data quality criteria are not demanding and are easily met or exceeded daily in laboratories around the world, and adherence to these should only be relaxed in exceptional circumstances, and then only slightly.
FAQ list for Co-editors who receive CIFs containing completed validation response Forms (VRFs)
The suggested actions broadly cover cases where
(a) the authors are contacted immediately because they should be able to supply critical missing information quickly;
(b) the authors are contacted immediately because of serious problems that are insurmountable or which require action before further review can be undertaken;
(c) the alerts will be assessed as part of the overall review process;
(d) no serious problem exists but additional explanatory text will be requested; and
(e) no action is necessary.
Co-editors should adopt a fairly strict stance. When an author gives a
long pleading argument about how important this structure is (in their
eyes?), despite its deficiencies, and that the overall structure is
clear and chemically unambiguous, it is easy to start to feel a little
soft-hearted. It is important to maintain a fairly level standard,
otherwise one starts down a slippery slope. Authors will be quick to
draw our attention to accepted papers with low standards if they can
find any and argue that their paper ought to be accepted for this
reason, even if their work is obviously shoddy, while the example they
use might have had quite reasonable grounds why it did not meet the
usual criteria.
A good rule of thumb: if it does not feel adequate, then it isn't. In
cases of doubt, contact the Validation Editor currently on duty;
Validation Editors are listed in the Editors pages of Section C
and Section E.
In general, author responses in the VRF should be reasonable scientific justifications, and not emotional, illogical or pleading-for-forgiveness type responses. If the argument is irrational, ask for a better one. A common argument is that the parameter is "near the acceptable limit", but they forget this is the limit for achieving only a B alert, not the limit for a completely acceptable value. If they are already in the A alert region, then things are far from normally acceptable values, never simply borderline.
Arguments that the data are old and met the acceptable standards at the time of collection are also poor ones. This argument usually appears when a theta(max) of 23 degrees or lower was used. The data would have to be at least prior to 1996 to be in this class, but even back then it was an undesirable trend. Old data may actually become less publishable with time as modern expectations move ahead, so the argument that publishing old data is OK is not really acceptable - otherwise it allows the scenario that one could publish any mediocre result if one waited long enough.
Theta(max) below 25 degrees (for Mo) is normally inadequate. In exceptional cases, 24 degrees may be OK, but 23 degrees or lower is insufficient for Acta C papers.
Arguments of weak diffraction, therefore theta(max) was kept low, need to be treated cautiously, as this is an easy-to-use excuse. A quick test that this is a bad argument is if the _gt reflections is a high percentage of the unique reflections, which suggests that the crystal was actually diffracting quite well. Low theta(max) should not normally come from a CCD or IP experiment nowadays, as almost by default one gets data to higher angles, even if they are weak. Low theta(max) from a serial diffractometer suggests a deliberate choice of this limit, perhaps even before it was known how strong the diffraction really was, and suggests a poor choice of experimental conditions, perhaps for expediency, rather than true consideration of the data actually required.
If
the data are indeed weak, then the next thing is to ensure that the
experiment was indeed conducted optimally for that particular case. For
example, were any or all of the following tried: best efforts at getting
the optimum possible crystals; recollecting data more slowly; low
temperature; Cu radiation? Sometimes it has been found that people
collect data far too quickly for weakly diffracting samples and a slower
data collection would help significantly. We can also probably
reasonably expect, at least in the Western world, that a modern properly
equipped laboratory has LT apparatus, so the absence of LT measurements
from such laboratories when needed is perhaps not acceptable. If they
really have no LT equipment, what about collaborating with someone who
does?
Incomplete data sets are often explained as being due to weak
diffraction. This argument seems irrelevant, as one should have obtained
all unique reflections to the chosen theta(max), or at least to 25
degrees (with Mo), regardless of whether or not many of them were
unobserved. However, DENZO may
throw out some reflections in the outer regions if there really is no
diffraction except at quite low angles (but then perhaps a better data
set is needed anyway).
Problems and actions
In the following, common problems are outlined. In cases where the
authors are asked to resubmit, they should be directed to do this at
the following web addresses:Section C: http://journals.iucr.org/services/submitbdyc.html
Section E: http://journals.iucr.org/services/submitbdye.html
and should also be asked to provide the Co-editor code for the paper when resubmitting.
Symmetry and space group
SY1. PROBLEM: Wrong space group/genuine missed symmetry (and consequent problems in refinement).
ACTION: Immediately referred back to author for re-refinement in correct space group. Resubmit.
SY2. PROBLEM: Possible missed symmetry but inconclusive.
ACTION: Can be assessed as part of the overall refereeing process.
SY3. PROBLEM: Strong pseudo-symmetry present.
ACTION: Can be assessed as part of the overall refereeing process, but author should discuss this in the Comment section.
SY4. PROBLEM: Possible spurious lattice doubling; Z' = 2.
ACTION: Can be assessed as part of the overall refereeing process.
SY5. PROBLEM: Triclinic unit cell not in reduced form.
ACTION: Immediately referred back to author for correction. Re-refine and resubmit.
SY6. PROBLEM: Monoclinic unit cell with beta < 90.
ACTION: Immediately referred back to author for correction. Re-refine and resubmit.
SY7. PROBLEM: Unit cell has s.u.'s on constrained values, e.g. 90.00(3).
ACTION: Immediately referred back to author for correction. Re-refine and resubmit.
Data completeness, resolution and quality
Theta(max) below 25 degrees (for Mo) is normally inadequate. In exceptional cases, 24 degrees may be OK, but 23 degrees or lower is insufficient for Acta C papers.
DC1. PROBLEM: Data completeness
is only 89% to theta(max) of 29.55 degrees.
ACTION: Author immediately asked to calculate and supply completeness to appropriate lower resolution, i.e. fill in correctly the lines
_diffrn_reflns_theta_full?
_diffrn_measured_fraction_theta_full ?
DC2. PROBLEM: Major gaps in data completeness throughout dataset.
ACTION: Author contacted immediately, but if cannot supply a satisfactory explanation, reject.
DC3. PROBLEM: Dataset is only 0.69 complete to 2theta 50 degrees on Mo radiation
ACTION: Author contacted immediately and replied that a filter was applied in SAINT; author was immediately asked to reprocess data without this filter, re-refine the structure and resubmit the paper.
DC4. PROBLEM: Synchrotron dataset with high redundancy but only 0.72 complete (after beam dump).
ACTION: Completeness is not satisfactory: rejected.
DC5. PROBLEM: About half the data were skipped or suppressed.
ACTION: Author could not provide a valid explanation: paper rejected.
DC6. PROBLEM: Very low r/p ratios (2.5, 1.9); extensive disorder.
ACTION: Make various suggestions to author including refining on F^2^ with all data, constraining H atoms wherever possible or recollecting data more slowly, at low temperature and/or with Cu rather than Mo radiation. If no progress, reject.
DC7. PROBLEM: The r/p ratio appears OK but only 20% of data have I > 2 sigma(I).
ACTION: Same as previous question.
DC8. PROBLEM: Data completeness only 0.90 on a one-circle IP system.
ACTION: This can be reasonable if the missing data are in a cusp but the author will be asked to provide an explanation for inclusion in _publ_section_exptl_refinement. Triclinic is more problematic with such diffractometers than are higher symmetry space groups. If higher symmetry space groups have low completeness on a one-circle IP system, ask if the crystal was mounted with an axis almost along the spindle axis. This is a bad idea and completeness will normally improve if the crystal is mounted skew to the spindle axis.
DC9. PROBLEM: High value of R~int~ (0.18) for a weak dataset.
ACTION: The author will be asked to provide an explanation for inclusion in _publ_section_exptl_refinement.
DC10. PROBLEM: Low resolution dataset with theta(max) on Mo radiation of 23 degrees
ACTION: The author was informed that the resolution of the dataset was well below the minimum of about 25 degrees.
Absorption corrections
AB1. PROBLEM: Significant mu*tmid value (0.7) but no absorption correction.
ACTION: Author immediately asked to apply correction, re-refine and resubmit.
AB2. PROBLEM: High mu*tmid value (> 3.0) but multi-scan absorption correction.
ACTION: Author immediately asked to apply numerical correction, re-refine and resubmit.
AB3. PROBLEM: Extreme ADPs due to poor absorption correction.
ACTION: Author immediately asked to apply a better correction, re-refine and resubmit.
AB4. PROBLEM: Extreme Tmin or Tmax.
ACTION: If SADABS used, may indicate a problem with the correction. Works well only if highly redundant data present. Author immediately asked to check correction and improve if possible. If a light-atom structure, maybe no correction is better. If appropriate, re-refine and resubmit.
AB5. PROBLEM: Tmin and Tmax match calculated values.
ACTION: Values derived from the SIZE instruction in SHELXL. Author immediately asked to provide true experimental values (this often occurs when SADABS is used).
Refinement and disorder
RE1. PROBLEM: High U values, unrealistic geometry - due to unmodelled disorder.
ACTION: Author immediately asked to refine a suitable disorder model and resubmit.
RE2. PROBLEM: Poor convergence - maximum shift/s.u. > 1.5.
ACTION: Author immediately asked to identify the problem (Flack parameter? extinction parameter? H atoms? disorder?), re-refine and resubmit.
RE3. PROBLEM: Central heavy atom has a high U value compared with immediate neighbours - wrong element?
ACTION: Author immediately asked to investigate this possibility.
RE4. PROBLEM: Indicators of element mis-assignment (Hirschfeld test; U mismatch with neighbours; N-H or O-H without acceptor, but C...O or C...N contact present).
ACTION: Author immediately asked to investigate this possibility and, if correct, re-refine and resubmit.
RE5. PROBLEM: Significant voids in solvent regions of the structure explained as unable to model disordered solvent, but field of peaks observed.
ACTION: Author immediately asked to consider trying SQUEEZE and, if appropriate, re-refine and resubmit.
RE6. PROBLEM: Significant voids in solvent regions of the structure - no explanation.
ACTION: As above initially. At the very least author should provide explanation for _publ_section_exptl_refinement.
RE7. PROBLEM: Solvent molecules have problems with some geometric and U values.
ACTION: Authors will be asked to check their modelling of these molecules, improve using restraints, or maybe try SQUEEZE and, if appropriate, re-refine and resubmit.
RE8. PROBLEM: Poor geometry and extreme U values for disordered CF~3~ group.
ACTION: Author immediately asked to re-evaluate disorder model and, if appropriate, re-refine and resubmit.
RE9. PROBLEM: Suspected disorder; chemically equivalent C-C bonds differ by 0.11 Angstroms.
ACTION: Authors immediately asked to consider a disorder model and, if appropriate, re-refine and resubmit.
RE10. PROBLEM: Absolute structure determination [Z(max) = 8, Mo radiation] claimed but Friedel equivalents have been merged.
ACTION: Author will be asked to remove this claim before acceptance.
RE11. PROBLEM: As above but with Z(max) = 35, Mo radiation.
ACTION: Author immediately asked to re-refine without merging Friedels and resubmit.
RE12. PROBLEM: Flack parameter is 1.05(5) - suspect wrong absolute structure.
ACTION: Author immediately asked to invert the structure, re-refine and resubmit.
RE13. PROBLEM: Flack parameter has no s.u. (means TWIN used without BASF)
ACTION: Author immediately asked to repeat refinement with BASF and resubmit.
RE14. PROBLEM: Flack parameter is exactly 0.00(x). (means TWIN used with the matrix defined explicitly, and while the BASF parameter is correct, the Flack parameter may be wrong).
ACTION: Author immediately asked to repeat refinement with just TWIN (without any numbers following) or simply provide the BASF parameter and its s.u. and resubmit.
RE15. PROBLEM: Friedels not merged for light-atom structure, Mo radiation.
ACTION: Author immediately asked to merge Friedels, re-refine and resubmit.
RE16. PROBLEM: Pervasive disorder and unrealistic geometry.
ACTION: Author immediately asked to re-assess the structure, including the space group assignment and, if appropriate, re-refine and resubmit.
RE17. PROBLEM: Maximum difference electron-density peak is very high given the elements present.
ACTION: Authors immediately asked to re-assess their structure (heavier element present? space group correct? absorption correction adequate? twinning? solvent?). If appropriate, re-refine and resubmit.
RE18. PROBLEM: Riding non-H atoms.
ACTION: Usually an AFIX 0 is missing after an H atom. Author immediately asked to rectify, re-refine and resubmit. Will be valid if atom is in a rigid group, e.g. rigid benzene ring.
RE19. PROBLEM: Most non-H atoms are isotropic.
ACTION: Author immediately asked for explanation (simple mistake? do anisotropic atoms go NPD? restraints to ADPs?). Otherwise, suggest recollecting data more slowly, at low temperature and/or with Cu rather than Mo radiation. If appropriate, re-refine and resubmit.
RE20. PROBLEM: Very large structure has 10 non-H atoms isotropic.
ACTION: Author will be asked to provide explanation for _publ_section_exptl_refinement.
RE21. PROBLEM: Disordered C atoms (4) refined with isotropic displacement parameters.
ACTION: Author will be asked to provide explanation for _publ_section_exptl_refinement.
RE22. PROBLEM: Central heavy atom has a very low U value compared with its neighbours.
ACTION: Author will be asked to look at this and, if appropriate, provide some text.
RE23. PROBLEM: The mean U3/U1 value (3.3) is a rather high (ellipsoids like trees in a storm).
ACTION: Author will be asked to look at this and provide some text for _publ_section_exptl_refinement.
RE24. PROBLEM: Solvent molecules have problems with some geometric and U values.
ACTION: Authors will be asked to check their modelling of these molecules and, if appropriate, re-refine and resubmit.
RE25. PROBLEM: High residual electron density near Rb centre.
ACTION: Authors will be asked to check their model and their data processing; if no improvement, they should supply appropriate explanatory text.
RE26. PROBLEM: Extreme perchlorate O ADPs.
ACTION: Can be assessed as part of the overall refereeing process.
RE27. PROBLEM: Several rigid-bond alerts involving central metal atom.
ACTION: Author will be asked to look at this and, if appropriate, provide some text.
RE28. PROBLEM: Unmerged data.
ACTION: OK if a twin and e.g. HKLF 5 data used.
Hydrogen atoms
HY1. PROBLEM: Wide range of H-atom U values; very close H...H contacts.
ACTION: Author immediately asked to review the placement and refinement of these H atoms, re-refine the structure and resubmit.
HY2. PROBLEM: Isolated H atoms (symmetry transformations do not bring them close to their parent atoms).
ACTION: Author immediately asked whether these atoms have really been located reliably. If appropriate, re-refine and resubmit.
HY3. PROBLEM: Unsatisfactory H-atom geometry.
ACTION: Authors immediately asked to review their treatment of these H atoms, re-refine the structure and resubmit.
HY4. PROBLEM: C-bound H atoms refined in presence of heavy atom (Br).
ACTION: Author immediately asked whether this was intentional; recommended the application of appropriate constraints (e.g. SHELXL HFIX) and/or restraints, re-refinement and resubmission.
HY5. PROBLEM: Missing water H atoms - no explanation.
ACTION: Author will be asked to confirm that these could not be located and, if this is the case, to supply appropriate text. If H atoms can be included, re-refine and resubmit.
HY6. PROBLEM: Extreme values for freely refined H-atom U values - ratio 9:1.
ACTION: Author immediately contacted to recommend the application of appropriate U~iso~ constraints, re-refinement and resubmission.
HY7. PROBLEM: Wrong geometric constraints used for H atoms of methyl group on benzene ring.
ACTION: Author immediately asked to apply the correct constraints, re-refine and resubmit.
HY8. PROBLEM: O-H hydrogen-bond donor without a corresponding acceptor.
ACTION: Author will be asked to check whether this is correct. If not, re-refine and resubmit.
HY9. PROBLEM: H-atom bonds/angles missing.
ACTION: Author asked immediately to include this info, even if all H-atom positions calculated (as sometimes this reveals mis-calculation). If your own checks show that the H-atom geometry is OK, this could be left until a further assessment of the paper has been done, in case there are other issues requiring a new refinement.
Consistency
CO1. PROBLEM: Molecular geometry parameters do not correspond to fractional coordinates.
ACTION: Author immediately asked to resolve the problem and, if appropriate, re-refine and resubmit.
CO2. PROBLEM: Author has manually edited atom labels, but not consistently.
ACTION: Author immediately asked to resolve the problem and, if appropriate, re-refine and resubmit.
CO3. PROBLEM: Centre of gravity of moiety outside base unit cell.
ACTION: Assessed as part of the overall refereeing process.
CO4. PROBLEM: Atoms do not make a properly connected moiety.
ACTION: Author immediately asked to resolve the problem (may be a spurious alert for extended or inorganic structures which you need to check carefully, but the test is generally reliable) and, if appropriate, re-refine and resubmit.
Other
OT1. PROBLEM: Crystal has long needle axis (1.25 mm) w.r.t. beam diameter.
ACTION: Author will be asked to supply appropriate text.
OT2. PROBLEM: Many alerts resulting from missing data items, but in a Rietveld powder study.
ACTION: Assessed as part of the overall refereeing process.
OT3. PROBLEM: Many alerts resulting from a neutron study.
ACTION: Assessed as part of the overall refereeing process if the alerts are clearly a result of experimental restrictions, e.g. wavelength, restricted access to reciprocal space.
OT4. PROBLEM: Long N-N bonds, but plausible on steric grounds.
ACTION: No specific action required.
OT5. PROBLEM: Missing standard reflections for area-detector data set.
ACTION: Alert treated as spurious - note that setting _diffrn_standards_number to zero is valid and cures this problem.
OT6. PROBLEM: Missing data item _chemical_absolute_configuration.
ACTION: Author immediately contacted and asked to supply missing item.
OT7. PROBLEM: Refined extinction parameter within range.
ACTION: Author immediately contacted and asked to re-refine without extinction and resubmit.
OT8. PROBLEM: Validation report has very long lists of mostly trivial things.
ACTION: Author immediately contacted and asked to tidy up the CIF and resolve as many items as possible, if necessary by re-refinement. Resubmit. May also be indicative of generally poor structure, so you need to decide if it is good enough to be acceptable at all.
