Synthesis and crystal structure of a zwitterionic nickel(II) complex

Article abstract of DOI:10.1016/j.inoche.2004.12.027

The crystal structure of complex 3, a Ni(II)(dppe) ligated in a square planar arrangement by a novel S, N-ligand, has been determined at 123 K. The neutral and air stable complex is of interest as it shows an apparently zwitterionic structure, one counter-anion to the Ni(II) being delocalised into the backbone of the ligand. Support for this conclusion comes from an analysis of bond lengths in the complex 3, the crystal structure of which is presented and the result of a density functional calculation performed on the complex. A discussion of the mechanism leading to the observed product is presented.

Full text of DOI:10.1016/j.inoche.2004.12.027

Inorganic Chemistry Communications 8 (2005) 293–296
www.elsevier.com/locate/inoche
Synthesis and crystal structure of a zwitterionic nickel(II) complex
*
Andrew Dinsmore , Priscilla Ikhide, David G. Billing, Manuel A. Fernandes,
Fabrizio Marsicano, Letlhogonolo Selemela
School of Chemistry, University of the Witwatersrand, PO WITS, Johannesburg 2050, South Africa
Received 13 August 2004; accepted 30 December 2004
Available online 23 January 2005
Abstract
The crystal structure of complex 3, a Ni(II)(dppe) ligated in a square planar arrangement by a novel S, N-ligand, has been deter-
mined at 123 K. The neutral and air stable complex is of interest as it shows an apparently zwitterionic structure, one counter-anion
to the Ni(II) being delocalised into the backbone of the ligand. Support for this conclusion comes from an analysis of bond lengths
in the complex 3, the crystal structure of which is presented and the result of a density functional calculation performed on the com-
plex. A discussion of the mechanism leading to the observed product is presented.
Ó 2005 Published by Elsevier B.V.
Keywords: Nickel; dppe; Aza-thiol ligand; Zwitterion; Density functional calculation
The molybdenum dependent oxygen-transfer en-
zymes are essential to all life [1]. Enzymes isolated and
characterised from this diverse family to date share a
structurally conserved cofactor. This cofactor has been
called the molybdenum cofactor and the general struc-
ture, as determined by protein crystallography, is
shown. Additional ligation on the metal is not shown,
but may be provided by a second pterin–dithiolene or
serine, cysteine or selenocysteine, while the free hydroxyl
group may be replaced by a phosphate–dinucleotide, in
the case of bacterial sources of the co-factor.
We are engaged in research directed at elucidation of
the mechanism by which the molybdenum cofactor al-
lows electron transfer between the active site (molybde-
num) via the reduced tricyclic pterin to one of a number
of different biological redox centres, this being depen-
dant on the class of oxygen-transfer enzyme. An interest
in modelling essential features of this process led us to
attempt to prepare the nickel dithiolene complex 1
(Scheme 1).
We had expected dithiolene complex 1 to be readily
formed by hydrolysis of the precursor ligand 2c,
followed by transmetallation to Ni(II)dppe (Ph2PCH2
CH2PPh2). Though none of the dithiolene 1 was identi-
fied, a novel nickel complex 3 was isolated. This had
apparently resulted from incomplete hydrolysis of the li-
gand 2c under the conditions employed, prior to the
complexation with NiCl2 Æ dppe. The structure of this
product 3 was elucidated by X-ray crystallographic
analysis and is presented here. A comparison of the
structure with that of the parent ligand 2c is made to-
gether with a consideration of the novel zwitterionic
structure and a discussion of the likely mechanism lead-
ing to the formation of complex 3.
O
Mo
S
O
H
N
H
H
S
HN
H2N
OH
N
N
H
O
Molybdenum cofactor
*
Corresponding author. Tel.: +27117176765; fax: 27117176749.
E-mail address: andy@aurum.wits.ac.za (A. Dinsmore).
1387-7003/$ - see front matter Ó 2005 Published by Elsevier B.V.
doi:10.1016/j.inoche.2004.12.027

294
A. Dinsmore et al. / Inorganic Chemistry Communications 8 (2005) 293–296
O
C6H4CH3
O
Ph
P
i) NaOH / MeOH
ii) NiCl .dppe
Ph
O
S
2
Ph
P
N
N
S
N
Ph
S
S
Ni
HN
S
P
SO2
Ni
X
S
Ph
Ph
P
(2a); X = H, H
(2b), X = O
Ph
Ph
(3)
S
OMe
(2c), X = NNHSO C H CH
2
6
4
3
(1)
Scheme 1.
O
The aldehyde 2b was prepared in 70% yield from the
This air stable red crystalline solid was found to be sol-
uble in organic solvents. X-ray crystallographic analysis
at 123 K showed the compound to have the structure 3
dithiolene 2a by oxidation employing selenium dioxide
in refluxing dioxane. Aldehyde 2b has previously been
made, in a 45% yield, by a di-iso-butylaluminium hy-
dride (DiBAl) mediated reduction of the corresponding
methylcarboxylate substituted dithiolene [2]. This route
is, however, accompanied by over-reduction, which
gives the corresponding alcohol. Thus, the route re-
ported here, in one step from the readily available meth-
yldithiolene 2a, represents an improved synthesis of the
aldehyde 2b. Reaction of the aldehyde 2b with toluene-
sulfonylhydrazide in ethanol gave the sulfonylhydrazone
1
shown in Fig. 2 and both H NMR and mass spectral
data were consistent with this analysis. Interesting fea-
tures of the structure shown in Fig. 2 are the planarity
of the ligand arranged around the square planar Ni(II)
and the absence of both a counter-ion in the lattice
and a proton at the N-2 position of the structure. Thus,
the ligand is a dianion, balancing the charge on the nickel
and resulting in a neutral complex.
Various canonical forms can be drawn to account for
the distribution of charge in this structure (Scheme 2).
Significantly, no canonical form allows the de-local-
ised charge in the ligand backbone to cancel the devel-
opment of positive charge at N1. Support for the
likely contribution of each of the canonical forms shown
comes from an examination of the bond lengths.
The bond lengths in the complex 3 are consistent with
contributions from each of the canonical forms (I)–(III)
shown (Fig. 2). Thus, the bonds C1–C2, C3–N1 and the
two S3–O bonds are longer, while the bonds C2–C3,
1
2c in 93% yield. H, 13C NMR and mass spectral data
were consistent with the proposed structure and for
the purposes of comparison with the complex 3 an X-
ray crystallographic analysis was performed, the result
of which is shown in Fig. 1.
Hydrolysis of the sulfonylhydrazone 2c employing
two equivalents of sodium methoxide in methanol was
followed to completion (disappearance of starting mate-
rial, as judged by TLC analysis of the reaction mixture).
Upon addition of the NiCl2 Æ dppe, a red colour formed
immediately. Chromatographic purification of the crude
product, on silica gel, yielded one principal product.
Fig. 2. ORTEP diagram for complex 3 showing the atom-labelling
scheme and 50% probability ellipsoids. H atoms have been omitted
˚
from the dppe ligand for clarity. Selected bond lengths (A): C1–C2
Fig. 1. ORTEP diagram for pro-ligand 2c showing the atom-labelling
˚
scheme and 50% probability ellipsoids. Selected bond lengths (A): C1–
C2 1.332(4), C2–C3 1.450(3), C3–N1 1.274(3), N1–N2 1.391(3), N2–S3
1.650(2), S3–O2 1.430(2), S3–O3 1.426(2).
1.352(3), C2–C3 1.434(3), C3–N1 1.310(3), N1–N2 1.386(2), N2–S3
1.614(18), S3–O2 1.448(16), S3–O3 1.455(17).

A. Dinsmore et al. / Inorganic Chemistry Communications 8 (2005) 293–296
295
O3 tol
tol
O
O
O
tol
O
3
S
O 2
S
S
2N
N
N
N1
N
N
C3
C2
(3)
Ni
Ni
Ni
S
S
S
C1
1
S2 OMe
S
OMe
S
OMe
(III)
(I)
(II)
O
O
O
Scheme 2. Canonical forms accounting for electron distribution in complex 3.
N1–N2 and N2–S3 are shorter than the corresponding
bonds in the ligand precursor 2c, though these differ-
ences are small and it is evident that both structures 3
and 2c show bond lengths intermediate between single
and double bonds. The canonical form (I) presumably
benefits in stability from charge localisation onto the
more electronegative oxygen, while forms (II) and (III)
from the known ability of sulphur to stabilise adjacent
carbanions [3].
which to yield the observed regioisomer is directed by
the Lewis acidity of the sodium cation, itself coordinated
to the hydrazide (Scheme 3). It is conceivable that the di-
anionic nature of the hydrolysis product 6 then slows
attack of the second methoxide at the terminal thiocar-
bonate, the product from which would have been the
initially desired dithiolate 1. We have previously ob-
served the hydrolysis of similar protected dithiolenes
lacking potentially acidic side chains to be a facile [5].
In conclusion, a neutral nickel(II) complex has been
isolated, wherein one counterion is delocalised in the li-
gand backbone and not on a ligating atom.
This charge distribution is supported by a density
functional calculation of the atomic charges, selected
values from which are shown in Table 1 [4].
An interesting question concerning the course of the
reaction during the hydrolysis and subsequent transmet-
allation is why the sole product identified is the five
membered chelate ring 3. None of the isomeric six mem-
bered chelate ring 4 (Scheme 3) was detected. We sug-
gest two possible explanations. The six membered
chelate 4 is thermodynamically less stable and, if
formed, spontaneously rearranges, with acyl transfer
to the observed product 3. Another explanation is that
the side chain on the protected ligand directs the hydro-
lysis to give the one regioisomer. Nucleophilic attack of
methoxide on the deprotonated ligand presumably re-
sults in a tetrahedral intermediate 5, the collapse of
Acknowledgement
This material is based on work supported by the Na-
tional Research Foundation under Grant No. 2053652.
Financial support from the University of the Witwaters-
rand is also gratefully acknowledged.
Appendix A. Supplementary data
Crystallographic data have been deposited with the
Cambridge Crystallographic data centre, CCDC No.
221256 for pro-ligand 2c and CCDC No. 221257 for
nickel complex 3. Copies of this information may be ob-
tained free of charge from the Director, CCDC, 12 Un-
ion Road, Cambridge CB2 1EZ, UK or http://
www.ccdc.cam.ac.uk. Full experimental details, charac-
terisation for all compounds described in this paper,
and crystallographic tables are contained in the
Table 1
Selected partial atomic charges calculated for complex 3
C1
C2
C3
N1
À0.141
0.173
N2
O2
O3
Ni
À0.542
À0.563
À0.567
À0.234
P1
S3
S2
S1
0.146
1.162
À0.37
0.308
À0.183
À0.122
O
S
O
tol
O
tol
O
S
Ph
HN
N
OMe
Ph
N
N
O
NiCl .dppe
2
P
Na
O
Ni
(3)
S
S
SO2tol
Na
S
P
S
N N
S
OMe
Ph Ph
S
OMe
(6)
(5)
(4)
O
Scheme 3.

296
A. Dinsmore et al. / Inorganic Chemistry Communications 8 (2005) 293–296
[2] T. Nozdryn, J. Cousseau, A. Gourges, M. Jubault, J. Orduna, S.
Uriel, J. Garin, J. Chem. Soc., Perkin Trans. 1 (1993) 1711.
[3] F. Bernardi, A. Bottoni, A. Venturini, A. Mangini, J. Am. Chem.
Chem. Soc. 108 (1986) 8171.
supplementary data file. Supplementary data associated
with this article can be found, in the online version, at
doi:10.1016/j.inoche.2004.12.027.
[4] Atomic charges can be calculated in various ways, leading to
different numerical values. The atomic charges given in Table 1
were calculated by the program Spartan Pro to produce the
electrostatic potentials around the molecule. Some other molecular
modelling packages calculate Mulliken atomic charges. Calculated
atomic charges are sensitive to the basis set used and to the method
of calculation..
References
[1] (a) D. Collison, C.D. Garner, J.A. Joule, Chem. Soc. Rev. (1996)
25;
(b) R. Hille, Chem. Rev. 96 (1996) 2757;
(c) J.H. Enemark, J.J.A. Cooney, Jun-Jieh Wang, R.H. Holm,
Chem. Rev. 104 (2004) 1175.
[5] S. Davies, R.L. Beddoes, D. Collison, A. Dinsmore, A. Docrat,
J.A. Joule, C.R. Wilson, C.D. Garner, J. Chem. Soc., Dalton
Trans. (1997) 3985.