Volume 2 Chapter 11 Text Menu *DISORDER

Back to Table of Contents

*ERROR

Function

*ERROR is used to test for a comment describing the correction made to a published error.

CSD Contents

Since textual comments in the CSD are composed in "free-language" it is not recommended that such searches are conducted. Rather, when a hit is encountered the textual comment can be displayed to provide details of the entry's history.

An example of an error correction comment might be:

x(N5) is 0.4321, not 0.3421.
It is appropriate to describe here the bond length analysis which is conducted for each new entry.

The published bond lengths, although not stored in the CSD, are keyboarded and used for comparison against values calculated at the CCDC.

Errors in atomic coordinates, unit cell dimensions, space group (and published bond lengths) can be detected by the discovery of :

A bond length is declared to be short if the calculated value is less than the sum of the elemental radii minus a tolerance (the tolerance is normally 0.4 Å).

Suppose a published bond length is Dpub and the calculated value is Dcalc

Two levels of discrepancy are recognised :

Many errors are capable of correction by CCDC staff and when this is done an error correction comment is introduced into the entry indicating the correct value for the parameter.

The check process is iterated until all errors that can be corrected at the CCDC have been eliminated.

At this stage the entry is either error-free or an error set.

The entry is declared to be an error set if it contains at least :

Note that, for earlier entries in the CSD, a discrepancy at the * level did not constitute an error situation.

Each error set is flagged as such and a comment (see *REMARKS) is introduced into the entry suggesting the possible source of error. The check output is sent to the author with a request for the correct value(s). When the corrected information is received from the author the entry is extracted from the master database, edited, re-checked and re-archived (see also *ADATE, *MDATE in the NUMERIC sub-menu, chapter 7).

If the entry is error-free with respect to bond length analysis then the next stage of the check process is carried out ie. connectivity matching (see *MATFLAG in the NUMERIC sub-menu, chapter 7).

Graphics QUEST3D Procedure

Select *ERROR from the TEXT sub-menu and respond to the standard text prompts.

Related Bit Screens

The following database screens are set as a result of the bond length analysis :

37
entry contains valence errors
41
valence errors have been over-ridden (a common example would be oxygen with a valence of 3 in a water molecule coordinated to a metal atom)

38
entry contains short bond(s)

39
bond length discrepancies are at 0.02-0.05Å level (*)
40
bond length discrepancies are greater than or equal to 0.05Å (**)

32
entry is error-free at the 0.02Å level
33
entry is error-free at the 0.05Å level
34
entry contains one or more uncorrected errors
42
all error checks have been over-ridden ie. entry contains errors
43
at least one problem referred to author
44
entry contains correction(s) supplied by author

If you wish to exclude error sets and mis-matched entries from the hits generated by a search then :

SCRE -34 85

will register hits for all entries containing no uncorrected errors and which are perfectly matched.

Related Keywords

*REMARKS

Back to Table of Contents

Volume 2 Chapter 11 Text Menu *FORMULA.