Deposition of crystal structure data sets in the ICSD

written by Felix C. Goerigk, Institute of Inorganic Chemistry at the University of Stuttgart


There is sometimes confusion about depositing crystal structures in the frequently used ICSD, especially after the union of the ICSD with the CambridgeStructuralDatabase. In the past, a finished data set consisting of a .cif file and a .hkl file could simply be sent by e-mail to the Fachinformationszentrum (FIZ) Karlsruhe. Since the fusion, this is now done online via an assistant. Although this is somewhat more complicated, it has the advantage of highlighting any errors in the datasets that might otherwise be overlooked. Since the ICSD is one of the most important structure databases for inorganic structures, the following is about the common procedure of structure deposition.


To deposit a new crystal structure, the following requirements must be fulfilled:

  • The crystal structure data set is finished, consisting of .cif- , .hkl- and .fcf-file. The .cif and the .fcf file is usually output when the crystal structure is refined by the respective software. The .hkl file should be the file generated by the diffractometer during integration, which has subsequently undergone absorption correction, e.g. with STOE X-SHAPE, Habitus or the Laue Analyser (LANA). The latter programs are commercially available from Stoe Darmstadt in the software package Stoe X-Area and are common for data processing of Stoe diffractometers [1].
  • An account at the Cambridge Crystallographic Data Centre (CCDC) has already been created. This can be set up free of charge.
  • The EnCIFer software is installed (not absolutely necessary, but makes work much easier). This can also be obtained free of charge from the CCDC. Alternatively, the online service can be used.

Step-by-step procedure

Checking the .cif file

First, the finished .cif file should be loaded using the EnCIFer programme. This tool knows the permitted commands of the .cif nomenclature and immediately issues errors if incorrect entries were made during manual editing. Of course, any text editor can be used, but then errors can easily be overlooked, as there is no feedback on entries. The knowledge of basic handling of .cif files is required in the following paragraphs.

The following table gives an overview of some entries that should be paid special attention to, as manual changes might be necessary. In addition to the values given below, the residual values, number of reflections, measuring range, calculated density, etc. should also be checked for consistency. Depending on the hardware and software used and the measurement geometry, some entries may not be necessary.

Name of the entry Meaning What to do
Sum formula of the compound with alphabetical sorting of the elements Check for correctness
Crystal system

as one word, e.g.

space group type in Hermann Mauguin symbolism

enter, e.g.



Cell parameters check. All three axes and angles must always be indicated, regardless of the symmetry!
Minimum and maximum value of θ (not 2θ) for the cell determination Determine and enter from the measurement report of the diffractometer
Habitus and colour of the crystal


needle, bar, plate, 
cube, square prims,
hexagonal, prismatic,
tetrahedral, octahedral,
blocky, ...

and for colour

pale yellow, 
colourless, blue,
maximum, average and minimum edge length of the crystal Enter values in millimetres. It makes sense to determine these from light or, better, scanning electron microscope images of the crystal.
Type of absorption correction

add, e.g.



Minimum and maximum transmission coefficient is usually calculated by the software of the absorption correction. Allowed values go from 0.0 to 1.0
Absorption correction programmes used Enter programme, version and publisher, put all values in inverted commas (‘apostrophe’).
Diffractometer model Enter manufacturer and device, all in inverted commas
Geometry and detector of the diffractometer

enter, e.g.

'four-circle, ccd-detector'

all in inverted commas

Measurement, refinement and data reduction software Enter, all in inverted commas
Editor and Version oft the software mostly automatically set, check for correctness, in inverted commas, e.g.
'SHELXL-97 (Sheldrick, 1997)
Most important cif keywords and their meaning.

Then click on the control symbol in EnCIFer (yellow warning triangle with two arrows). Ideally, “Errors – none”, “Warnings – none” and “remarks – none” is displayed below. In Figure 1, on the other hand, an error has been deliberately built in. The correct space group from the symmetry operations would be P21/c, EnCIFer highlights the incorrectly entered space group P2/c (at “Warnings”).

The EnCIFer programme can be used as an comfortable editor for cif files and displays entered, incorrect values directly.

Deposit with the help of the online assistant

After the .cif file is error-free, the structure can be deposited. To deposit it in the ICSD, it must first be deposited in the CCDC. Sounds strange, but this is due to the fusion. To do this, create a profile if you do not already have one, log in to the CCDC, click on “Profile” and then on “Deposit” in the menu on the left. The assistant is relatively well explained. First, the personal data should be checked, then the .cif, . hkl and . fcf files are uploaded under “Select Files”. At the bottom, the option “I wish to run the IUCr checkCIF/PLATON service on my data” can be selected; this checks the database for existing, identical metrics and for overseen errors in the .cif file. If “View Report” lights up orange or red, errors have obviously occurred and should be displayed and corrected. These are categorised into different “Alert Levels”, where Level A and B usually mean serious problems (typos in the .cif file?) and e.g. Level G are just general comments. The temperature at which the measurement was performed is often shown as a Level C error. This note can be ignored after checking.

Once all problems have been solved, you can go to the next page “Add Publication”. The next step is to enter the reference in which the link was first published. However, since one normally first deposits the dataset(s) and then publishes the corresponding article, only the names and address data of all planned authors are temporarily entered here. The journal and the exact page number are then simply added later after publication.

On the following pages, “Enhance Data” and “Review”, the data entered is summarised again. Cataloguing entries can also be entered here to facilitate faster searching and finding for other users. These can be ticked on the right and should be answered as precisely as possible. For example, in the case of a compound with luminescent properties produced by solid-state chemistry from a molten salt, the properties “from the melt” and “luminescent” should be selected. Further, self-explanatory options, e.g. on stability, can also be selected. After one is satisfied, the data set can be submitted with “Submit”. Within a few minutes, an automatic email with the CSD number should arrive, and the data set can also be examined under “Profile”.

Further publication and linking to the ICSD

Under “My structures” in “Profile” you will find a tabular list of all deposited data records. If these are not yet linked to any literature, an embargo date can be found as shown in Figure 2. This is usually one year after the date of the deposition. A literature reference should be given by this date, otherwise the link can no longer be directly associated with the publication. To link, click on “Details” in the line of the corresponding link, then on the next page at the bottom click on “Add Publication”, where the corresponding details are entered. After checking the entries, the status changes to “Published”.

The deposited crystal structures and their status can be displayed in the CCDC’s own profile.

You might think you are now finished and everything is in order. Unfortunately, one last step is necessary, which is not explained: Namely, the connection to the ICSD must be made. Unfortunately, this does not work automatically. In order to include the finished structure in the ICSD, a mail is written manually from the personal mail address (or the address entered in the CCDC profile) to, which, in addition to naming the account and the dataset (CSD number), also contains a .pdf file of the published paper (finished version including page numbers, no galley proof etc.). Then, the receipt of the reference is confirmed by mail and the dataset from the CCDC is inserted into the ICSD at the next update.

Why you must manually send an email with the paper to the FIZ is unclear to me. The online assistant at the CCDC works extremely well; there only seems to be a communication problem between the CCDC and the FIZ, since the FIZ, according to its own statement, does not receive any information about the publication details from the CCDC and therefore does not automatically see when a dataset has actually been properly published. This last step should be improved and automated. If nothing else is done, the connection remains “Published” in the CCDC, but is not included in the ICSD.

I hope that this article will help some of you to successfully deposit your own structures!

With greetings from Stuttgart,

Felix C. Goerigk

Felix Goerigk did his PhD in the research group of Prof. Thomas Schleid at the Institute for Inorganic Chemistry at the University of Stuttgart. His research focuses on the structural elucidation of new solid-state compounds using single crystal and powder X-ray diffraction and related methods such as SEM-EDS and -WDS.

[1] W. Herrendorf, H. Bärnighausen, HABITUS: Programm zur Optimierung der Kristallgestalt für die numerische Absorptionskorrektur als Version X-SHAPE (Version 1.06, Fa. Stoe, Darmstadt 1999). Karlsruhe, Gießen, 1996.