Kiralj R., Ferreira M. M. C., “Use of the Cambridge Structural Database in Study of Single and Partial Double C-X (X=C,N,O) Bonds in Organic Molecules in Crystalline State”. São Carlos, SP, 16-19/03/2003: XVI Reunião da Sociedade Brasileira de Cristalografia [16th Meeting of Brazilian Crystallographic Society], Livro de Resumos [Book of Abstracts], (2003) 32. Oral, poster OPQM02.
OPQM02. Use of the Cambridge
Structural Database in Study of Single and
Partial Double C-X
(X = C, N, O) Bonds in Organic Molecules in Crystalline
State
Kiralj R. and Ferreira M. M.
C
Instituto deQuímica,UniversidadeEstadualde
Campinas.E-mails:rudolf@iqm.unicamp.br,
marcia@iqm.unicamp.br
The Cambridge Structural Database
(CSD) [1] is an organic/organometallic
crystal
structures database with various possibilities
of visualization and analysis of the hits. The
current CSD version (November 2002) contains
272066 crystal structures determined by
diffraction methods (X-ray, neutron, synchrotron)
[2]. This is an encouragement to study
multiple C-X bonds, where X is usually C, N or O
in typical organic compounds. Another
reason to study these bonds is their presence
in bioactive molecules as drugs and natural
biomolecules, and in
environmental pollutants as polycyclic aromatic
hydrocarbons
(PAHs) and their derivatives. Moreover, C-C
bonds were among the first ones exhibiting
structure correlation as bond length-bond order relationships
[3]. Today, when the number
of determined crystal structures with high-level
precision is increasing, it is possible to
update such relationships, generalize to other
bond types and molecular classes, and by
careful analysis qualify and even quantify delocalization,
substitution and crystal packing
effects on bond lenghts. Furthermore,
in structural, organic and theoretical chemistry,
simple equations for calculation of bond lengths
from some simple variables, are welcome.
The aim of this work is
to present new proper findings on bond
length-bond order
relationships taking into account
those phenomena affecting the bond length.
Current
findings on extremely long C-C
bond length around 3 Å [4] put traditional teaching
on
chemical bonding to be reexamined
and more generalized, what is also the aim of
this
work. This presentation discusses accurate
CSD searches appropriate for chemometric
analysis, and generation of bond variables
as bond orders, topological indices,
crytal
packing effect parameters. Finally,
regression equations for bond length prediction
are
established, validated and interpreted. The compounds
under study are PAHs, nucleobases,
aza-PAHs and carbon allotropic modifications and carbon
clusters, all obtained from CSD.
[1] F. H. Allen, Acta Cryst., B58 (2002) 380.
[2] The Cambridge Structural Database. http://www.ccdc.cam.ac.uk/support/
[3] L. Pauling, L. O. Brockway, J. Y. Beach, J. Am. Chem. Soc., 57 (1935) 2705.
[4] J. J. Novoa et al., Angew. Chem. Int. Ed., 40 (2001) 2540.
The authors acknowledge FAPESP for the support.
32