author checklist

The following is a summary of the conventions used by the journal for the recording and presentation of data and text. It concentrates on those data items which may prove most useful to authors.

Absolute structure – see Anomalous scattering effects

Absorption corrections

checkCIF may indicate rescaling of transmission factors.

If absorption corrections are not made, then _exptl_absorpt_correction_T_min and _exptl_absorpt_correction_T_max should be set as ‘?” and not numerical values. The experimentally determined T_min and T_max values should be reported and not those estimated theoretically from the crystal size.


If the compound name is not included in the title, it should be given.

The same applies to the chemical formula, either as a sum formula or (if appropriate) divided up into ions/molecules, including separate solvent molecules; in metal complexes, it may be divided up into separate ligands.

The most important aspects and results of the study should be summarized. An Abstract that says only that a compound has been prepared and its crystal structure determined is of little value. Avoid saying that a certain aspect will be discussed – summarize the finding instead.

Any crystallographic symmetry of molecules or ions should be stated.

There should be no crystal data (cell parameters, space group, etc.) in the Abstract unless there is good reason.

Avoid specific atom numbers in the Abstract, which should be capable of being understood without reference to the rest of the paper.

References should be avoided in the Abstract. An exception would be when the paper reports a redetermination of a structure (see References for example citation).

Only one paragraph is allowed in the Abstract. See also Keywords.


Acknowledgement should be given for any assistance provided to the study. If diffraction data collection was not carried out by one of the authors, or in the laboratory of one of the authors, details of who collected the data and where the data collection was carried out should be provided.

If there are no acknowledgements, you will be asked whether this is intentional, in case they have been omitted accidentally.

Anomalous scattering effects

Absolute structure is relevant in any non-centrosymmetric space group. Authors should be aware of the difference between absolute structure and absolute configuration. To improve the precision of the absolute structure parameter, the use of a large fraction of the complete set of Friedel pairs in the data set is strongly recommended.

If the space group is non-centrosymmetric and there are significant anomalous scattering effects (usually elements > Si with Mo radiation, several O atoms or heavier with Cu radiation and good data), the absolute structure parameter (Flack parameter or equivalent) should be refined and reported, together with its s.u.; the number of Friedel pairs in the data set should also be stated (even if there are none, this refinement should still work, though it will be less good).

Unless the Flack parameter is close to zero with a small s.u., there will probably need to be some discussion of its meaning, possibly with reference to the standard Flack & Bernardinelli papers. There should be a statement of what has been determined (absolute configuration, crystal polarity, partial inversion twinning, etc.).

These points need to be addressed, even if the “absolute structure” is not important, because having the structure inverted by mistake can introduce other errors or bias.

For non-centrosymmetric structures with no significant anomalous scattering effects and the precision of the absolute structure parameter renders the analysis inconclusive, if authors merge Friedel-pair reflections before final refinement, they should mention that fact in the Refinement section and not report a value for the Flack parameter in the CIF. If the molecule is chiral, the merging of Friedel pairs is not recommended; it has been argued that there may be information in the data that could be extracted on the basis of future developments in the subject.

If the absolute configuration of a chiral compound has been assumed from the synthesis, or assigned arbitrarily, this should be stated.

For all structures of chiral molecules, there should be a CIF entry for _chemical_absolute_configuration, stating the method or assumption regarding the determination of the absolute configuration. This term is irrelevant for structures of racemates (not just centrosymmetric structures, but any containing mirror or glide planes, or improper rotation axes of any kind).

The chemical scheme and the molecular structure as shown (usually in Fig. 1) should be of the same absolute configuration. The name of the compound should also be correct in terms of R and S designations. For a racemic structure in which both enantiomers are present, the correct notation uses RS and SR designations, and for a single enantiomer for which the absolution configuration is not known, R* and S* are used.

Note that in _refine_ls_abs_structure_details it is sufficient to cite the method, for example, ‘Flack & Bernardinelli (1999, 2000)’. The full reference is recorded in _publ_section_references and not here.

If, in the case of a non-centrosymmetric space group, the compound contains no element heavier than Si and Friedel pairs have been merged then, in Refinement section, there should be the statement “In the absence of significant anomalous scattering effects Friedel pairs have been merged.” or similar.

Atom and geometry tables

Atom names (including H atoms) should be sensible and simple. Chemical and crystallographic numbering should be in agreement wherever possible.

The usual convention is for no parentheses in atom names. Atom labelling needs to be consistent in all tables, text sections, and figures.

Extensive numerical details in the text instead of a Table are undesirable, and repeating significant amounts of the contents of a Table in the text should also be avoided.

Lists of geometry should be in a sensible order.

All geometry involving H atoms should preferably be included in the CIF (but not flagged for publication); this is required if H atoms have been refined rather than constrained, and in such cases, the range of bond lengths to H atoms should be stated in the Refinement section.

Check the consistency in the number of atoms and the chemical formula (checkCIF).

Even though space restrictions are not important for electronic publishing, and full geometry is available to readers, sensible selection for the html/pdf versions should still be made. Avoid symmetry-equivalent items in the selected list.

There should be no s.u.'s on H-bond geometry involving constrained H atoms (D---H, H...A and D---H...A).

Coordinates of constrained H atoms should not have more decimal places than they would if the H atoms were refined; similarly for geometry parameters (usually 3 decimal places for distances and whole degrees for angles).

Numerical results, including torsion and dihedral angles, quoted in text sections should usually have s.u.'s.

Values calculated by including atoms from other asymmetric units must have the symmetry code included with the atom label and the associated symmetry operator defined (or a pointer given to a table or caption where it is defined). Symmetry codes should be consistent within the text, tables and figures for any given structure (i.e. the code numbers for a specific symmetry operator in the text and the corresponding table or figure must be the same).

Authors and addresses

One full forename for each author is preferred.

If there are more than five authors, there will be a checkCIF A-level alert; you will need to specify the role of each author by including a Validation Response Form in the CIF.

Addresses a, b, c etc. should all be different. Sometimes a small typing error in one author’s address makes it appear different from the others when they should actually be the same.

Chemical scheme – see Graphics

Compound identification

Compound synthesis

Unless a published method is used (and included as a reference), synthetic details should normally include quantities of reagents and solvents, reaction times, recrystallization procedure, melting points, optical rotation etc. Other characterization data (chemical analysis, spectroscopy etc.) may be included,and lists of spectroscopic data should include assignments. More extensive information can be included in the various _special_details sections of the CIF or as Supporting information.

Crystal colour

Crystal density

If there is no measured density, _exptl_crystal_density_method should be ‘not measured’ and _exptl_crystal_density_meas should be ‘?’

Crystal habit

Very often the crystal is declared to be a needle or plate, but the dimensions do not support this. As a reasonable guideline, a needle should have one dimension at least double the other two, and a plate should have a thickness no more than half the other dimensions.

Decimal places

Adjustment of the number of decimal places is carried out in the production of the final publication so it is not necessary to make these changes in the CIF.

Formula mass 2
Calculated density 3
Wavelength 5 (or 4)
Theta ranges (cell and data collection) 1
mu (in mm-1) 2
Transmission coefficients 3
Crystal dimensions 1 or 2 (usually, depending upon apparent precision)
R values 3
Goodness of fit 2
Intensity decay not more than 1
Shift/su 3
Electron density extremes 2
Cell measurement temperature not more than 1


The manufacturer and model name of the diffractometer should be reported in _diffrn_measurement_device_type.

Dihedral angles

Discussion of dihedral angles involving groups that are not approximately planar is meaningless. One often occurring example is the calculation of a dihedral angle between a benzene ring (sensibly planar) and a cyclohexane ring (not planar).

Dihedral angles between groups that are in completely different parts of a molecule are often pointless; such information should not be used to extend discussions with little content.


If the structure is disordered, then details, including site occupancy factors (with standard uncertainties if they are refined and explanation of the values chosen if not) and restraints or constraints used, should be reported in the Refinement section.

In any relevant Figure, the caption should indicate whether all disordered components are shown or, for example, only the major component.

Extinction coefficients

Extinction coefficients with values <3sigma from zero are insignificant and should not be given or used in the refinement.

Extra tables (non-standard)

Information on adding extra non-standard tables to a CIF can be found at the online help page 'Extra tables in CIF'.

Geometry tables – see Atom and geometry tables


Check the scheme (chemical formula diagram) for agreement with the overall formula. All entities in the structure should be present, including counter-ions and solvent in the appropriate stoichiometric ratio. The correct stereochemistry must be included and be consistent with the structure model.

Avoid unnecessary detail, e.g. use PPh3 instead of fully drawn phenyl groups for a triphenylphosphane ligand.

Charges should be given where appropriate. Check also for correct bonding representations, especially in delocalized groups.

For organic and metal-organic papers, Fig. 1 is usually a view of the molecular structure. Sufficient atom labels (with no parentheses and with H atoms not labelled) should be included on the diagram to allow all atoms mentioned in the text to be identified. The labels should be consistent throughout the article. If H atoms have been omitted, this should be stated. It helps if the orientation is not completely different from that of the scheme.

The Figure will usually contain more than one molecule if Z’>1 in the asymmetric unit.

For molecules with crystallographic symmetry, the whole molecule should normally be shown, but it is not usually necessary to label more than a few symmetry-related atoms; the caption should give the symmetry operation relating them to the corresponding labelled atoms or refer to a Table for this.

The caption should also state if any or all H atoms, or any other parts of the structure (e.g. disorder components) have been omitted.

Unless there are very good reasons, this Figure should show displacement ellipsoids, for which the probability level (%) should be stated in the caption. Atom labels and any other text should not intersect any atoms or bonds, and should be large enough to be clearly legible.

For packing diagrams, the caption should state the view direction or orientation and the unit cell outline should be shown, with the origin and axes labeled, or the axes directions drawn at the side of the diagram. The caption should state what any dashed/dotted lines indicate (most commonly for hydrogen bonds); this applies for all Figures. Packing diagrams and other Figures that are not discussed in any detail should be removed; “Fig. 2 shows a packing diagram” is not a significant discussion.

Stereoviews rarely give any advantage over a well produced single view of comparable overall size. If one is used, any labels should appear, correctly positioned, in both views, not just in one.

Figures should be numbered, and referred to sequentially in the text sections. There should be a caption for each Figure.

hkl ranges – see Index ranges

Hydrogen bonding

Extensive numerical details in the text instead of a Table are undesirable, and repeating significant amounts of the contents of a Table in the text should also be avoided. The same applies to molecular geometry (bond lengths, bond angles and torsion angles).

Hydrogen-atom treatment

If the refinement of H atoms has involved constraints or restraints, then numerical details should be reported in the Refinement section. It is not sufficient to reference software defaults since not all readers use the same software.

For the commonly used procedure of riding H atoms, the assumed distances and information about the treatment of U values should be given.

For freely refined H atoms, the range of distances obtained should be stated (with standard uncertainties).

Very often

_refine_ls_hydrogen_treatment mixed

should be replaced by

_refine_ls_hydrogen_treatment constr

In many cases, this results from a default of the refinement program.

Independent reflections

If the number of independent reflections is different from the number used in the refinement, then this should be clarified by the author and the reason stated in the text.

Index ranges

For area-detector diffractometers the full measured ranges of h, k, l should be reported, not just the unique data set after merging of equivalents. There should be no published entries concerning standard reflections.

If _diffrn_reflns_av_R_equivalents is > 0.10 and markedly higher than the final R factor, some comment should preferably be included in the Refinement section.

Intensity decay

If this is reported as negative, it should be checked. Negative decay means the standard reflection intensities have increased, which is unusual.


At least preferred keywords describing the areas covered by the paper should be included. In a CIF, these will come under the data name _publ_section_keywords. Examples might include “crystal structure”, “pharmaceuticals”, “metal-organic frameworks”, “coordination polymers”, “water clusters”, “powder diffraction” etc.


Chemical compounds should where possible be given their full name according to IUPAC rules. Although this is checked by the editorial staff, and created for the author if necessary, correct naming of a compound is the author's responsibility.

All components, in the correct proportions, should also be shown in the chemical scheme (not required for inorganic network compounds).

The scheme should show charges for ionic compounds.

The word “phenyl” should be used only for the unsubstituted phenyl ring, C6H5; otherwise use “benzene” to refer to substituted rings.

Correct bracket notation for planes (), directions [] and forms {} should be used.

Unusual abbreviations should be explained in full when first used. Authors will be asked to supply definitions of any nomenclature or usage which is not common in scientific journals.

The word “moiety” is frequently misused. It means “half” or “one of two approximately equal parts” and should not be used to refer to relatively small groups of atoms.

A very useful website for chemical terminology is the so-called IUPAC Gold Book at


In the Abstract (_publ_section_abstract), if, for example, there is a reference to an earlier publication of the structure then the full reference should be cited, e.g. Smith, Jones & Anderson [Chem. Commun. (1967), pp. 408—409].

In the text sections of the paper, for two authors, the format is, for example, (Smith & Jones, 1988) for more than two authors, the format is, for example, (Allen et al., 1987), with no comma after the first author name.

In _publ_section_references the ordering of references should be :

(1) First, alphabetically by surname of first author, then second author, etc.

(2) Second, numerically by year for the same set of authors.

For identical lists of authors and the same year, the a, b, etc. notation is used.

Examples are:

Lehmler, H. J., Robertson, L. W. & Parkin, S. (2005). Acta Cryst. E61, o3025—o3026.

Lehmler, H. J., Robertson, L. W., Parkin, S. & Brock, C. P. (2002). Acta Cryst. E61, o123—o124.

Drew, M. G., Hall, R. S. & Long, F. I. (2004a). Organometallics, 23, 456—459.

Drew, M. G., Hall, R. S. & Long, F. I. (2004b). J. Med. Chem. 47, 579—586.

Consider the following two references:

Smith, A. B., Jones, C. D., Wolf, E. F., Baird, M. N. & Hunt, G. H. (1994). Organometallics, 13, 123—128.

Smith, A. B., Jones, C. D., Lyle, R. S., Hunt, G. H. & Baird, M. N. (1994). Chem. Rev. 94, 345—380.

In the text of the paper, these should be distinguished as follows:

Smith, Jones, Wolf et al. (1994) and Smith, Jones, Lyle et al. (1994).

Each reference is separated from the next by a blank line.

Page numbers should be inclusive (first–last).

Check consistency with references in the body of paper (same authors, same years, no references missing or extra in list). If there are problems, you will be asked to supply missing/corrected items, or a complete new reference list if several changes are needed. The program publCIF (available from will check whether references in the reference list are cited in the text and vice versa. It can also be used to order the reference list correctly. A list of standard and commonly used references is included in publCIF under the “tools” menu.

For software references, addresses should be city and country, but not full address.

Refinement section

Routine structure solution and refinement procedures are not needed and should be omitted; only include details of non-routine procedures or properties of the structure.

Comment should be provided here on the treatment of disorder and on non-standard data collection or refinement, etc., including use of constraints and restraints. This is particularly useful if there are checkCIF alerts to be explained.

Details of twinning or the use of SQUEEZE should be included here – see SQUEEZE and Twinning.

If the number of independent reflections is different from the number used in the refinement, then you will be asked to clarify this.

Refinement instruction files

A refinement instruction file should be included for each data block in the CIF using the data name _iucr_refine_instructions_details. If the refinement was carried out using SHELXL-2013 or later, then the refinement instructions will be present in the CIF under the data name _shelx_res_file and should be left as such along with its checksum. Also, do not remove any reflection data appended to the CIF, e.g. _shelx_hkl_file and its associated checksum.

The advantages of upgrading from SHELXL-97 to SHELXL-2014/7 or later versions are described here.

Reflection observation threshold

The reflection observation threshold is often given as ‘>2sigma(I)' instead of ‘I>2\s(I)'. Note the need for I before > and correct use of the Greek symbol.

Residual electron density

If a validation alert of any level is present for _refine_diff_density_max or _refine_diff_density_min, then the location of the highest peak/deepest hole should be recorded in the Refinement section.

Scheme – see Graphics

Software references

All software citations:


are needed. For each program/package, there should be an entry in the reference list. Check for consistency of these, including the dates.

Some examples of very commonly used programs and their recommended standard references are:
PLATON: Spek, A. L. (2009). Acta Cryst. D65, 148–155.
SHELXL: Sheldrick, G. M. (2008). Acta Cryst. A64, 112--122.
publCIF: Westrip, S. P. (2010). J. Appl. Cryst. 43, 920-925.
ORTEP-3 for Windows: Farrugia, L. J. (2012). J. Appl. Cryst. 45, 849--854.

References are required also for absorption, extinction and absolute structure procedures, if used.

References are not needed for scattering factors from International Tables.

A full set of standard references is available from the online help page 'Standard software references'.

Space-group symbols

Use the correct CIF-defined format, not SHELX style, e.g. P2(1)/c should be ‘P 21/c’ and Fdd2 should be ‘F d d 2’.


If the SQUEEZE routine has been used to handle disordered solvent (use with caution!) the contents of the .sqf file generated (if PLATON is used) should be included in the CIF. Details of the number and volume of voids per unit cell, estimated electron count per void and the estimated solvent content of the structure deduced therefrom should be included in the Refinement section. It is recommended that the reported formula, formula weight, F(000), density and linear absorption coefficient included the true estimated solvent content of the crystal. The correct value of the linear absorption coefficient is important if numerical or analytical absorption corrections are applied. Such absorption corrections usually need to be repeated once the true content of the crystal has been established.

Standard reflections

For area-detector data collections when decay is not monitored, the CIF can be filled out as follows to avoid validation alerts.

_diffrn_standards_number 0
_diffrn_standards_interval_count .
_diffrn_standards_interval_time .
_diffrn_standards_decay_% ?

Standard uncertainties (denoted s.u.)

These are normally required for all interatomic distances (not just bond lengths), bond angles, torsion angles, deviations from mean planes, dihedral angles, ring puckering parameters, etc.

Of course, these are not available if, for example, H-atom parameters have been constrained.

In some other cases, they may also not be provided by software (especially some ring-puckering parameters), and they may not be available for numerical results quoted from other publications or database entries.

They may be essentially meaningless for some averages, and for the sum of angles around an atom (because of ignored covariance terms).

Structural databases

Identification of individual structures in the article by use of database reference (identification) codes should be accompanied by a full citation of the original literature in the reference list. Citations in supporting information should also appear in the main body of the article.

Databases need an appropriate reference, and the database version should be stated, e.g. Cambridge Structural Database, Version 5.24 (Allen, 2002).

In the reference list, database entries can be included as:
Robertson, L. (2014). Private communication (refcode JIPFAQ). CCDC, Cambridge, England.

Text formatting

The length of lines in the CIF is unimportant, except they should be <80 characters. Be careful not to generate longer lines when editing. Avoid hyphenation of words across line breaks.

Spaces are needed after punctuation (period, comma, semi-colon and colon) and before \%A, but no spaces before punctuation, after ([ and before )], and before \%.

Paragraph breaks in text sections are created by inserting blank lines.

Entries in the reference list should be separated by blank lines.

Ensure that there are no comment lines beginning with # inside semi-colon delimited text blocks; they are treated as normal text !

Common errors include the following:

  • / instead of \ when defining special characters
  • A instead of \%A for the Angstrom symbol
  • ^o^ instead of \% for the degrees symbol
  • Greek letters spelt out instead of coded, e.g. sigma for \s, eta for \h
  • Chemical formulae in text blocks without subscripts

Spelling of the English language should be either all British or all American. A common error is to use “colorless” in the text and “colourless” in Crystal data, or vice versa. Other words to watch out for include: neighbouring/neighboring, analogue/analog, labelling/labeling, centre/center.


The title should usually be an interesting short description of the content of the paper. Descriptive text (e.g. redetermination of; a neutron study; at high pressure; a strained polycyclic molecule; the first example of) is welcomed. The title could also include the compound name(s) and some informative brief text referring to the nature of the compound. Long systematic names are undesirable and can be added to the Abstract.

Avoid titles consisting only of a chemical formula.

The first letter of the title should be capital (except for prefixes like ortho-, trans- or DL-); no capital letters are used for the start of other words.

Transmission factors – see Absorption corrections



Details of the handling of any twinning, including the type of twinning, the transformation matrix relating the twin components, the integration procedure (were reflections belonging to all or only one of the lattice orientations integrated) and the major twin component fraction should be reported in the Refinement section. The numbers of non-overlapping reflections from each twin component and the number of overlapping reflections should be given.

Unit-cell parameters: precision

If any cell parameters are reported with a final zero digit and an s.u. of (10) this should be checked; more often than not, it is a spurious addition by the refinement program! In such cases the s.u. is (1) and the trailing zero on the parameter should be deleted (for example, the COLLECT software on a KappaCCD was known for this).

It is recognized that diffractometer software packages currently produce unit cell parameters with far greater precision than is realistic. However, it is recommended that authors do not apply some random factor to bring the precision to more realistic levels (at least without stating the fact in the Refinement section), because a reader cannot tell if the precision has been adjusted post-integration or if it is the true diffractometer result and a sign that the crystal quality is very poor.

Validation Response Form (VRF)

In some cases, a validation response form (VRF) will be supplied by checkCIF. If the related validation issue cannot be resolved, this form should be completed and included in the CIF, preferably with the addition of appropriate explanatory text in the published experimental section of the article. A completed VRF will look something like:

# start Validation Reply Form
PROBLEM: The ratio of given/expected molecular weight as calculated from the _atom_site* data lies outside...
RESPONSE: SQUEEZE used to remove disordered diethyl ether solvent molecule, but the reported formula includes the solvent. See _refine_special_details.
# end Validation Reply Form

Weighting scheme

An actual formula should be provided instead of expressions such as ‘based on sigmas’ or ‘counting statistics’.

Remove meaningless terms such as +0.0000P and trailing zeros on all numbers.

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

[# https x2 cm 20170801 %]