First of all, please notice that this server is in development.
Your comments and bug reports will be greatly appreciated and we will
try to respond quickly with an updated version. We recommend that
you critically inspect the resulting ADPs using a visualization
program, e.g. Ortep, Platon, Mercury or Peanut. The procedure and
underlying program has been thoroughly tested, but the server setup
(including the use of cif files) may contain bugs.
Bug reports, requests for improvements and all other sorts of comments
are greatly appreciated and should be sent to
Anders Ø. Madsen
you wish to get a notice about updates of the software you can send
a mail to Anders Østergaard Madsen.
Requirements / Limitations
The cif file should
contain a description of one molecule in a crystal. The ADPs of
the non-hydrogen atoms will be analysed using the THMA11 program
(Trueblood and Schomaker, 1968) as a rigid body using the TLS
formalism. This rigid body motion will be imposed on the hydrogen
atoms along with a contribution from internal motion. In the
present implementation this internal motion is taken from a
database of displacements derived from accurate neutron
diffraction structures. If a H atom does not belong to a fragment
present in the database, we use a mean square displacement of
0.005 Å² in the bond direction and 0.02 Å²
in the directions perpendicular to the bond.
Atoms will be
recognized by the THMA11 program on the basis of their name - so
all H atom names should start with an H, e.g. "H5",
"HO1", all carbon atoms with a C, and so on.
If your crystal
contains more than one molecular entity, you have to split the cif file into several files containing each entity.
Only rigid fragments
with at least 4 non-hydrogen atoms can be analyzed. The rigid body
analysis is a procedure where the rigid-body translations (T) and
librations (L) and their correlation (the S matrix) are fitted
against the ADPs of the non-hydrogen atoms. There must be a
sufficient amount of data for the fitting procedure. Each
non-hydrogen atom have 6 anisotropic displacement parameters. If
the fragments are smaller than 4 non-hydrogen atoms, the server
will revert to a 'riding motion' procedure, where the overall TLS
motion is replaced by the motion of the atom that the H atom is
At present, the
program does not take internal molecular symmetry into account.
Thus, a full ADP matrix should be included for each and every atom
in the molecule.
more than 280 atoms can not be analyzed, because of a limitation
in the THMA11 program.
Please cite the following paper
when you use SHADE for your research
SHADE web server for estimation of hydrogen anisotropic displacement
Anders Østergaard Madsen.
J. Appl. Cryst. 39, 757---758
A simple approach to estimate isotropic displacement parameters for hydrogen atoms.
Anders Ø. Madsen and Anna Agnieszka Hoser.
Acta Crystallographica. Section A: Foundations of Crystallography (2015). 71, 2, 169-174.
SHADE3 server : a streamlined approach to estimate H-atom anisotropic displacement parameters using
periodic ab initio calculations or experimental information.
Anders Ø. Madsen and Anna Agnieszka Hoser.
Journal of Applied Crystallography (2014). 47, 6, 2100-2104.
Estimated H-atom anisotropic displacement parameters: a comparison between
different methods and with neutron diffraction results [ reprint ].
Parthapratim Munshi, Anders Ø. Madsen, Mark A. Spackman, Sine Larsen and Riccardo Destro.
Acta Cryst. (2008). A64, 465-475.
A neutron diffraction study of Xylitol: derivation of mean square internal vibrations for H atoms from a rigid body description.
Anders Østergaard Madsen, Sax Mason and Sine Larsen. Acta Cryst. (2003)
Section B59, 653-663.
The modeling of hydrogen atoms in charge density analysis [reprint].
Anders Østergaard Madsen, Henning Osholm Sørensen, Robert F. Stewart, Claus
Flensburg and Sine Larsen. Acta Cryst. (2004) Section A60,
Charge density study of naphthalene based on X-ray diffraction data at
four different temperatures and theoretical calculations.
Jette Oddershede and Sine Larsen. J. Phys. Chem. A 2004,
About the program
The shade server is written in python
and uses the PyCifRW
parser to read and write cif files.
The rigid body analysis is performed by the THMA11 program.
program is released under the GNU
general public license
Author: Anders Østergaard Madsen.
Citations and links
On the Rigid-Body Motion of Molecules in Crystals.
Schomaker, Verner and Trueblood, K. N.. Acta Cryst. (1968)
Section B24, 63-76.
of internal torsional motion with overall molecular motion in
crystals. Schomaker, Verner and Trueblood, Kenneth N. Acta
Cryst. (1998) Section B54, 507-514.
Computer graphics program to represent atomic displacement
parameters. Journal of Molecular Graphics (1990), 8, 214-220.
15 01 2017
The server has moved to a new machine. Some security features have been enabled. This should not change functionality or performance.
3 09 2011
New feature: if you submit ADPs without associated standard uncertainties, you can use the following option to use unit weights in the TLS analysis: Input the following entry in the cif file:
Now SHADE gives input in the correct format for the MoPro program.
13 08 2011
Version 2.1: More liberal atom labeling: Parentheses and longer atom names are allowed.This feature is accomplished via a new internal labeling for the thma11 program. This labeling is output in the cif files, in order to ease the interpretation of the thma input and output.
Reprint of new publication added to homepage.
SHADE2 made available.
16 08 2007
The thma11 program, and thus SHADE, has been dimensioned to analyze up to 280 atoms.
13 08 2007
Fixed a bug related
to the use of _atom_site_U_iso_or_equiv values of zero in the
submitted cif file. Thanks to Parthapratim Munshi for reporting
05 07 2006
to PyCifRW 3.0. This fixes the incompatibility problem with
PLATON. Thus, PLATON is now able to read the CIF files produced by
Bug related to metal-organic structures have been fixed.
I discovered that analysis of structures of more than 150
atoms is presently not possible due to limitations in the THMA11
has been updated in several ways. A publication describing the
details and comparisons with neutron-diffraction data and other
methods of estimating H ADPs has been published:
H-atom anisotropic displacement parameters: a comparison between
different methods and with neutron diffraction results [reprint].
Munshi, Anders Ø. Madsen, Mark A. Spackman, Sine Larsen and
SHADE library of internal displacements has been altered to give an
overall better fit with ADPs from neutron diffraction experiments.
Furthermore, it is now possible to perform a segmented rigid
body approach with shade. This approach uses the 'attached rigid
groups' method implemented in the THMA11 program (here is the
To instruct SHADE to use attached rigid groups, a non-standard
cif-loop should be put in the cif-file.
cif entry ---
'C26 C25' 'N3 N4 N5 N6 HN6'
'C6 C5' 'C7 C8 C9 S1 C10 O3 O2 C11 H9 H8 H11A H11B H11C'
C17' 'C21 C22 C23 C24 C26 C25 N3 N4 N5 N6 HN6 H24 H23 H22 H21'
'C14 C13' 'C19 C18 C15 C16 C17 H16 H15 H19 H18'
'N2 C5 C2
N2 N2' 'C12 C2 N1 C1 C4 C3 O1 C27 C28 C29 C30 C13 H12A H12B H12C
H30B H30A H30C H29A H29B H28A H28B H27A H27B'
--- end of cif
input is closely tied to the THMA input. THMA allows at most
attached rigid groups, and this limitation of-course persists.
The _segmented_tls_axis_defining_atoms entry should
either two or five atom names in a list enclosed by
each atom name separated by white space. If
two atoms are present,
they define the libration axis (similar
to THMA entries LBAT1 and LBAT2). If five atoms
this signals another way of constructing the libration
This corresponds to a negative NAFA entry in THMA. The five
then corresponds to LBAT1, LBAT2, LBAT3, LBAT4 and LBAT5, as
defined in the THMA
_segmented_tls_atoms_in_segment entry should be a quotation-
mark-enclosed white-space delimited list of atom names
corresponding to the atoms affected by the libration of the
including H-atoms, though they are not part of the TLS
This could probably be explained better, so don't
hesitate to ask
it is unclear!