checkCIF procedure

PROC-NAME: ABSTM02 Type_3

PURPOSE: To check that _exptl_absorpt_correction_T_min and _exptl_absorpt_correction_T_max are appropriate to the _exptl_crystal_size_ and _exptl_absorpt_coefficient_mu values.

PROCEDURE: DEFINE

mu is the value _exptl_absorpt_coefficient_mu
tmax is the distance _exptl_crystal_size_max
tmid is the distance _exptl_crystal_size_mid
tmin is the distance _exptl_crystal_size_min
trad is the radius _exptl_crystal_size_rad

Tmax(rep) is reported _exptl_absorpt_correction_T_max
Tmin(rep) is reported _exptl_absorpt_correction_T_min

TEST

IF _exptl_absorpt_correction_type ~ "psi-scan"
OR "empirical"
OR "multi-scan"
OR "refdelf"

IF mu * tmid > 3.0 exit and issue ALERT A
   "Alert A Crystal and compound unsuitable for non-numerical corrections. Product of mu and tmid > 3.0 "
When the product of the linear absorption coefficient (mu) and the median crystal dimension (_exptl_crystal_size_mid) is greater than 3.0, the application of a numerical or analytical absorption correction based on the indexing of the crystal faces is considered to be compulsory. When choosing crystals from samples that are likely to fall into this category, one should try as far as possible to select a crystal whose faces are clearly defined and can readily be indexed. If, despite all efforts, a regularly shaped crystal cannot be obtained, or the crystal is immersed in oil, and the faces cannot be indexed adequately, empirical or multi-scan absorption corrections may be acceptable, PROVIDED the experimental and predicted values of Tmin and Tmax do not differ significantly.
IF crystal dimensions are in the wrong order, issue General ALERT
   "Crystal dimensions are not in the correct order. Correcting and continuing."
IF crystal given as circular/spherical, issue General ALERT
   "Crystal is circular/spherical. Crystal dimensions set as twice the given radius."
IF _exptl_absorpt_correction_type is note set as none, bit Tmin/Tmax not given, issue ALERTA
   "Alert A Tmin and Tmax not reported, even though the correction is not set as 'none'"
If no absorption correction was applied, set _exptl_absorpt_correction_type to 'none', otherwise give the appropriate values for _exptl_absorpt_correction_T_min and _exptl_absorpt_correction_T_max

IF _exptl_absorpt_correction_type ~ "psi-scan"
OR "empirical"
OR "multi-scan"
OR "refdelf"
OR "analytical"
OR "numerical"
OR "gaussian"
OR "none"

CALCULATE

Tmax(shape) = EXP(-tmin * mu)
Tmin(shape) = EXP(-temx * mu)
Tmin(shape)'= EXP(-tmax * mu)

where temx = MAX{tmid, MIN[(1.2*tmid), tmax]}

Tmax(angle) = 1.0
Tmin(angle) = TAB(_diffrn_reflns_theta_max, mu*tmid)
TAB=Lookup value for angular factor

mu*tmidMoK\aCuK\a
0.01.001.00
0.51.011.05
1.01.041.15
1.51.091.25
2.01.151.5
2.51.251.9
3.01.402.5

Tmax(exp) = Tmax(shape) * Tmax(angle)
Tmin(exp) = Tmin(shape) * Tmin(angle)
Tmin(exp)'= Tmin(shape)'/ Tmin(angle)

RT(rep) = Tmax(rep) / Tmin(rep)
RT(exp) = Tmax(exp) / Tmin(exp)
RT(exp)'= Tmax(exp) / Tmin(exp)'

RR = RT(exp) / RT(rep)
RR'= RT(exp)'/ RT(rep)

TEST

IF _exptl_absorpt_correction_type ~ "refdelf"
OR "empirical"
OR "psi-scan"
OR "multi-scan"
THEN
IF RR' < 0.50 OR RR > 2.00 issue ALERT A (1)
   "Alert A The ratio of expected to reported Tmax/Tmin(RR') is < 0.50
Please check that your absorption correction is appropriate."
OR
   "Alert A The ratio of expected to reported Tmax/Tmin(RR) is > 2.00
Please check that your absorption correction is appropriate."

IF RR' < 0.75 OR RR > 1.50 issue ALERT B (1)
   "Alert B The ratio of expected to reported Tmax/Tmin(RR') is < 0.75
Please check that your absorption correction is appropriate."
OR
   "Alert B The ratio of expected to reported Tmax/Tmin(RR) is > 1.50
Please check that your absorption correction is appropriate."

IF RR' < 0.90 OR RR > 1.10 issue ALERT C (1)
   "Alert C The ratio of expected to reported Tmax/Tmin(RR') is < 0.90
Please check that your absorption correction is appropriate."
OR
   "Alert C The ratio of expected to reported Tmax/Tmin(RR) is > 1.10
Please check that your absorption correction is appropriate."


ELSE IF _exptl_absorpt_correction_type ~ "numerical"
OR "analytical"
OR "gaussian"
THEN
IF RR' < 0.50 OR RR > 2.00 issue General ALERT
   "The ratio of expected to reported Tmax/Tmin(RR') is < 0.50
Please check that your absorption correction is appropriate."
OR
   "The ratio of expected to reported Tmax/Tmin(RR) is > 2.00
Please check that your absorption correction is appropriate."

IF RR' < 0.75 OR RR > 1.50 issue General ALERT
   "The ratio of expected to reported Tmax/Tmin(RR') is < 0.75
Please check that your absorption correction is appropriate."
OR
   "The ratio of expected to reported Tmax/Tmin(RR) is > 1.50
Please check that your absorption correction is appropriate."

IF RR' < 0.90 OR RR > 1.10 issue General ALERT
   "The ratio of expected to reported Tmax/Tmin(RR') is < 0.90
Please check that your absorption correction is appropriate."
OR
   "The ratio of expected to reported Tmax/Tmin(RR) is > 1.10
Please check that your absorption correction is appropriate."

Alerts concerning the ratio of expected to reported Tmax/Tmin(RR') indicate that the range of transmission factors given in the CIF is either smaller or larger than that predicted from the given crystal dimensions. This means that some reflection intensities will be over or under-corrected for absorption effects.

Confusion often arises here because the experimental and predicted values of Tmax are given on a different scale. This of itself is unimportant. If one normalises the experimental Tmax so that it is set equal to the predicted Tmax, then the discrepancy will be highlighted by the significant difference between the experimental and predicted values of Tmin.

Alerts of this nature need to be addressed carefully. First check that the crystal dimensions given in the CIF do represent the actual crystal dimensions as closely as possible. Inaccuracies here can lead to a poor prediction of Tmin & Tmax and give rise to these alerts. It should normally be possible to estimate the crystal dimensions to 2 decimal places. Rough estimates to only 1 decimal place may be too inaccurate to provide reliable estimates of Tmin & Tmax.

If the crystal dimensions are appropriate, then the correctness of the absorption correction itself needs to be checked. If an analytical or numerical correction has been applied, have the crystal faces been indexed as accurately and fully as possible and have distances of these faces from a central point been measured as accurately as possible?

If a psi-scan correction has been applied, were sufficient reflections used to adequately define the absorption surface? One needs to ensure that the chosen psi-scan reflections were not also suffering from secondary extinction or the Renninger effect. If discrepancies in the spread of the Tmin & Tmax values remain, consider the alternative of applying a numerical or analytical absorption correction. Even if the crystal is only weakly absorbing, numerical and analytical corrections are still the most reliable and realistic method.

If the multi-scan method (e.g. SADABS) has been employed and there are discrepancies in the spread of the Tmin & Tmax values, one needs to be sceptical of the correction, particularly if the experimental Tmin is much smaller than predicted as this implies that an overly extreme correction is being applied to some reflections. Absorption effects due to the glue or glass fibre (which should never be too thick) are usually negligible and should not give rise to excessive ranges of Tmin & Tmax.

Corrections of the DIFABS type should normally be avoided if at all possible. Where such corrections are employed, discrepancies between the calculated and experimental values of Tmin & Tmax which give rise to alerts of this kind will not be accepted unless they can be precisely justified.


ELSE IF _exptl_absorpt_correction_type ~ "none"

THEN
IF RT(exp) > 1.30 issue ALERT A (2)
   "Alert A The ratio of Tmax/Tmin expected RT(exp) is > 1.30 An absorption correction should be applied."
IF RT(exp) > 1.20 issue ALERT B (2)
   "Alert B The ratio of Tmax/Tmin expected RT(exp) is > 1.30 An absorption correction should be applied."
IF RT(exp) > 1.10 issue ALERT C (2)
   "Alert C The ratio of Tmax/Tmin expected RT(exp) is > 1.30 An absorption correction should be applied."
You have indicated that an absorption correction has not been applied (exptl_absorpt_correction_type 'none'). However, the predicted values of Tmin & Tmax, based on the crystal dimensions given in the CIF, are sufficently unequal that absorption effects appear to be significant. Therefore, the application of a suitable absorption correction would appear to be required. Also check that the crystal dimensions given in the CIF do represent the actual crystal dimensions as closely as possible. Inaccuracies here can lead to a poor prediction of Tmin & Tmax and give rise to these alerts. It should normally be possible to estimate the crystal dimensions to 2 decimal places. Rough estimates to only 1 decimal place may be too inaccurate to provide reliable estimates of Tmin & Tmax.

SCALE

IF _exptl_absorpt_correction_type ~ "refdelf"
OR "empirical"
OR "psi-scan"
OR "multi-scan"

AND IF Tmax(exp)/Tmax(rep) < 0.95 AND > 1.05

THEN
Tmax(scl) = Tmax(exp)
Tmin(scl) = Tmax(exp) / RT(rep)

[[ For printed publication the reported T values will be replaced by the scaled T values. Since the ratio of scaled T's is identical to the ratio of reported T values, the scaling does not imply a change to the absorption corrections used in the study. It simply places the published T values on an absolute scale with respect to the crystal dimensions and the crystal mu value. In this way the T values conform to their definition, and the reader is able to easily deduce the relative importance of absorption corrections to the study.]]

Follow IUCr Journals
Sign up for e-alerts
Follow IUCr on Twitter
Follow us on facebook
Sign up for RSS feeds