Preparation of a range of copper complexes of diphenylsulfimide: X-ray crystal structures of [Cu(Ph2SNH)4]Cl2 and [Cu4(μ4-O)(μ-Cl)6(Ph2SNH) 4]

Article abstract of DOI:10.1016/S0277-5387(99)00237-5

Reaction of Ph₂SNH 1 with CuCl₂ (molar ratio 2:1) in acetonitrile gives trans-[CuCl₂(Ph₂SNH)₂] 2a which may be isolated as blue crystals (square planar form) by crystallisation from hot MeCN or as green needles (pseudo-tetrahedral form) by slow evaporation of a CH₂Cl₂/petroleum ether solution. In contrast, [CuBr₂(Ph₂SNH)₂] 2b, most efficiently prepared by reaction of [Cu(Ph₂SNH)₄]Br₂ with CuBr₂, appears to only exist in a dark green, pseudo-tetrahedral form. When 1.5 equivalents of 1 are reacted with CuX₂ in air, [Cu₄(μ₄-O)(μ-X)₆(Ph₂SNH) ₄] (X=Cl 3a or Br 3b) forms as orange crystals. X-ray crystallography reveals 3a to have the expected tetrahedral arrangement of Cu atoms within an oxo-centered Cu₄Cl₆O core. Four equivalents of 1 react with CuX₂ to give purple [Cu(Ph₂SNH)₄]X₂ (X=Cl 4a or Br 4b); the X-ray crystal structure of 4a reveals a square planar structure exhibiting strong hydrogen bonding interactions between the ligands and the counterions. Even in the presence of excess 1, no more than four of the ligands may be added to a copper centre. Elsevier Science Ltd.

Full text of DOI:10.1016/S0277-5387(99)00237-5

Polyhedron 18 (1999) 3173–3179
www.elsevier.nl/locate/poly
Preparation of a range of copper complexes of diphenylsulfimide: X-ray
crystal structures of [Cu(Ph2SNH)4]Cl2 and [Cu4(m4-O)(m-
Cl)6(Ph2SNH)4]
*
Paul F. Kelly , Sit-Ming Man, Alexandra M.Z. Slawin, Kevin W. Waring
Department of Chemistry, Loughborough University, Loughborough, Leics. LE11 3TU, UK
Received 30 June 1999; accepted 13 August 1999
Abstract
Reaction of Ph2SNH 1 with CuCl2 (molar ratio 2:1) in acetonitrile gives trans-[CuCl2(Ph2SNH)2] 2a which may be isolated as blue
crystals (square planar form) by crystallisation from hot MeCN or as green needles (pseudo-tetrahedral form) by slow evaporation of a
CH2Cl2 /petroleum ether solution. In contrast, [CuBr2(Ph2SNH)2] 2b, most efficiently prepared by reaction of [Cu(Ph2SNH)4]Br2 with
CuBr2, appears to only exist in a dark green, pseudo-tetrahedral form. When 1.5 equivalents of 1 are reacted with CuX2 in air,
[Cu4(m4-O)(m-X)6(Ph2SNH)4] (X5Cl 3a or Br 3b) forms as orange crystals. X-ray crystallography reveals 3a to have the expected
tetrahedral arrangement of Cu atoms within an oxo-centered Cu4Cl6O core. Four equivalents of 1 react with CuX2 to give purple
[Cu(Ph2SNH)4]X2 (X5Cl 4a or Br 4b); the X-ray crystal structure of 4a reveals a square planar structure exhibiting strong hydrogen
bonding interactions between the ligands and the counterions. Even in the presence of excess 1, no more than four of the ligands may be
added to a copper centre.
1999 Elsevier Science Ltd. All rights reserved.
Keywords: Complexes; Copper; Sulfimides; Isomerism; Hydrogen bonding
1. Introduction
with CuCl2, proved to be unusual in its ability to crys-
tallise in either square-planar or pseudo-tetrahedral
geometries [5]. Such isomerism appears to be unique for a
neutral Cu(II) species, again indicating that the ligand
properties of 1 are not as simple as one might imagine.
Here we present the results of further investigations into
the reactivity of 1 towards Cu(II) centres.
As part of our continuing interest in the coordination
chemistry of sulfur–nitrogen species [1] we have recently
reported on the reactivity of SS9-diphenylsulfimide,
Ph2SNH 1, towards a number of metal centres. For
example, we showed that homoleptic [Co(Ph2SNH)6]Cl2
readily forms from 1 and CoCl2 and exhibits strong,
concerted, directed hydrogen bonding between both sets of
triple N–H units on either side of the coordination
octahedron and the chloride counterions [2]. We have also
shown that simple complexes of Pt and Pd are isolable [3],
though in the former case reactions can be complicated by
an unexpected activation of nitrile leaving groups towards
sulfimide addition [4]. Another intriguing aspect of the
coordination chemistry of 1 came with the observation that
trans-[CuCl2(Ph2SNH)2], the product of the reaction of 1
2. Experimental
Except where otherwise stated all reactions were per-
formed under an atmosphere of dry nitrogen using dried,
degassed solvents (MeCN and CH2Cl2 from calcium
hydride, 60/80 petroleum ether from sodium). Copper
halides were dried by heating in vacuo and then stored and
handled under nitrogen; as noted below, 1 was prepared by
a modification of the literature method [6] (the latter, and
indeed other published methods, proving to be unreliable
in our hands). % Yields, IR and micro-analytical data for
2–4 are given in Table 1.
*Corresponding author. Tel.: 144-1509-222-578; fax: 144-1509-223-
925.
E-mail address: p.f.kelly@lboro.ac.uk (P.F. Kelly)
0277-5387/99/$ – see front matter
PII: S0277-5387(99)00237-5
1999 Elsevier Science Ltd. All rights reserved.

3174
P.F. Kelly et al. / Polyhedron 18 (1999) 3173–3179
Table 1
Analytical and selected data for complexes 2–4
Yield
(%)
IR spectra, cm21
[n(N–H), n(S–N)]
Analysis (calculated) (%)
C
H
N
CuCl2(Ph2SNH)2 2a
Blue
Green
76
73a
70
50
61
89
87
3292, 913
3202, 918
3286, 888
3283b, 947
3286b, 943
3081, 940
3126, 936
53.6 (53.7)
53.5 (53.7)
45.7 (46.1)
44.3 (44.8)
36.5 (37.1)
61.2 (61.4)
55.9 (56.1)
3.8 (4.1)
4.2 (4.1)
3.7 (3.5)
3.5 (3.4)
2.9 (2.9)
4.8 (4.7)
4.3 (4.3)
5.1 (5.2)
5.2 (5.2)
4.8 (4.4)
5.1 (4.4)
3.6 (3.6)
5.7 (6.0)
5.4 (5.5)
CuBr2(Ph2SNH)2 2b
[Cu4OCl6(Ph2SNH)4] 3a
[Cu4OBr6(Ph2SNH)4] 3b
[Cu(Ph2SNH)4]Cl2 4a
[Cu(Ph2SNH)4]Br2 4b
a Upon crystallisation from the blue form.
b Additional N–H stretch at 3319 cm21
.
3. Preparations
above was dissolved in CH2Cl2 (15 ml) and 60/80
petroleum ether (15 ml) added. The resulting green
mixture was treated with just enough CH2Cl2 to redissolve
the precipitate that had formed and was then allowed to
slowly evaporate in an open round bottom flask overnight,
during which time fine well formed green needles were
produced together with a small crop of the heavier, more
three-dimensional crystals of the blue form. The former
were decanted from the latter, filtered and dried in vacuo.
Yield: 55 mg. In order to confirm the polymorphism of the
system some of the latter green material (12 mg) was
redissolved in hot MeCN (2 ml) and the solution cooled
over a number of days in a freezer. This resulted in the
crystallisation of well formed blocks of the blue isomer
(11 mg).
3.1. Ph2SNH 1
A modification of the route of Vlasova et al. was used
[6]; thus Ph2SNSO2Ph (7 g, 0.02 mol) was added in a
number of portions to well stirred, degassed 95% H2SO4
(25 ml) under a blanket of N2. The resulting mixture was
then heated at 808C for 15 min before being cooled and
cautiously poured onto ice (200 ml). After warming, the
resulting mixture was taken to pH 11 by addition of 5 M
NaOH; the response of the mixture to this appears to be
very variable – on some occasions solid 1 precipitated and
was then filtered and recrystallised from Et2O, on other
occasions only a pale orange oil was apparent. In the latter
case, the entire mixture was extracted into warm Et2O
(typically 2350 ml), the solvent reduced in volume and
the resulting colourless crystalline mass filtered. In either,
case typical yield: 1.8 g, 45%; mp 69–708C (lit. 698C).
3.4. [CuBr2(Ph2SNH)2 ] 2b
A suspension of [Cu(Ph2SNH)4]Br2 4b (0.1 g, 0.097
mmol), prepared as detailed below, in MeCN (10 ml) was
treated with CuBr2 (22 mg, 0.099 mmol) added with
vigorous stirring. The solid immediately started to dissolve
and the mixture turned green; after stirring for 20 min the
mixture was heated until all the precipitate had redis-
solved, lagged to promote slow cooling to ambient tem-
perature and then further cooled in a freezer overnight.
This yielded a large crop of well-formed dark green
crystals. Yield: 85 mg.
3.2. [CuCl2(Ph2SNH)2 ] 2a blue (square planar) form
A solution of 1 (75 mg, 0.38 mmol) in acetonitrile (20
ml) was slowly added to a solution of CuCl2 (25 mg, 0.19
mmol) in the same solvent (20 ml) giving a green solution
which started to precipitate a blue solid after some 15 min
of stirring. At this point the solution was heated until all
the solid had dissolved, then lagged and slowly cooled,
first to ambient temperatures then overnight in a freezer,
yielding well formed blue crystals of the product.
3.5. [Cu4(m4-O)(m-Cl)6(Ph2SNH)4 ] 3a
The same material can also be prepared by the reaction
of [Cu(Ph2SNH)4]Cl2 4a (see below) with CuCl2; thus
addition of solid CuCl2 (15 mg, 0.11 mmol) to a stirred
suspension of 4a (0.1 g, 0.11 mmol) in MeCN (20 ml)
resulted in rapid formation of a green solution which in
turn yielded a blue precipitate after 15 min stirring.
Crystallisation as before yielded blue crystals of the
product. Yield: 87 mg.
A solution of 1 (90 mg, 0.45 mmol) in undried MeCN
(10 ml) in air was added to a solution of CuCl2 (40 mg,
0.3 mmol) in the same volume of the same solvent and the
mixture stirred then the reaction vessel sealed and allowed
to stand. After 1 day a mass of brown/orange crystals was
apparent; after 4 days the solution was decanted from this
solid which was then washed with a small amount of cold
MeCN and dried in vacuo. Yield: 48 mg.
3.3. [CuCl2(Ph2SNH)2 ] 2a green (pseudo-tetrahedral)
form
3.6. [Cu4(m4-O)(m-Br)6(Ph2SNH)4 ] 3b
A sample of crystalline blue isomer (75 mg) isolated as
A solution of 1 (90 mg, 0.45 mmol) in undried MeCN

P.F. Kelly et al. / Polyhedron 18 (1999) 3173–3179
3175
˚
(10 ml) in air was added to a solution of CuBr2 (66 mg,
0.3 mmol) in the same volume of the same solvent and the
mixture stirred. The dark mixture began to produce a
precipitate within 2–3 min; after a few hours the latter was
filtered from the solution, washed with a small amount of
cold MeCN and dried in vacuo. Yield: 70 mg.
0.2030.20 mm, m (Cu Ka)53.78 mm21, l51.54178 A,
F(000)5974.
3.9.3. Data collection and processing
In the case of 3a data were collected (using a Bruker
SMART diffractometer with graphite monochromated
Mo Ka radiation) using small slices: 26828 data collected,
8892 data used (Rint50.0571). Data were corrected for
Lorentz and polarisation effects and an empirical absorp-
tion correction was applied resulting in transmission
factors ranging from 0.698 to 1.00.
In the case of 4a data were collected (using a Rigaku
AFC7S diffractometer with graphite monochromated
Cu Ka radiation) by the v–2u scan technique to a
maximum 2u value of 120.28. Of 3757 measured reflec-
tions, 3604 were unique. Data were corrected for Lorentz
and polarisation effects and an empirical absorption correc-
tion was applied resulting in transmission factors ranging
from 0.45 to 1.00
3.7. [Cu(Ph2SNH)4 ]Cl2 4a
A solution of 1 (120 mg, 0.6 mmol) in MeCN (10 ml)
was treated with solid CuCl2 (20 mg, 0.15 mmol). After
vigorous stirring for 20 min the mixture consisted of a pale
purple solid and a pale green solution. At this point further
1 (10 mg) was added which resulted in an intensification
of the purple colour and a concomitant reduction in the
colour of the solution. The mixture was then heated until
all the solid had dissolved, lagged to promote slow cooling
and placed in the refrigerator. Overnight a crop of well-
formed purple crystals was obtained; these were filtered
and dried in vacuo. Yield: 124 mg.
3.9.4. Structure analysis and refinement
The structures were solved by direct methods and
refined by full-matrix least-squares against F2 (3a) or F
(4a).
In the case of 3a two acetonitrile molecules disordered
into four locations were found; the full and half weight
solvents were refined anisotropically whereas the two
quarter weight occupancies were refined isotropically. The
N–H protons were located from a DF map and refined
isotropically subject to a distance constraint. Refinement
against F2 lead to R150.0512, wR250.1182 with I .
3.8. [Cu(Ph2SNH)4 ]Br2 4b
A solution of 1 (120 mg, 0.6 mmol) in MeCN (10 ml)
was treated with solid CuBr2 (33 mg, 0.15 mmol). After
vigorous stirring for 30 min the mixture consisted of a very
pale purple solid and a quite intense green solution. At this
point further 1 (12 mg, 0.06 mmol) was added which
resulted in an increase in the amount of precipitate and a
reduction in the colour of the solution; addition of further 1
(10 mg, 0.05 mmol) completed the precipitation of the
intensely purple product and left only a pale green colour
to the solution. The mixture was then heated until all the
solid had dissolved, lagged to promote slow cooling and
placed in the refrigerator. Overnight a crop of well-formed
purple crystals was obtained; these were filtered and dried
in vacuo. Yield: 132 mg.
2s(I). The maximum/minimum residual electron densities
23
˚
in the final DF map were 0.715 and 20.318 e A
;
calculations were performed using SHELXTL [7].
For 4a non-hydrogen atoms were refined anisotropically;
hydrogen atoms were included but not refined. Refinement
against F lead to R50.062 [R 5 o(uFou 2 uFcu)/ouFou],
Rw50.057. The maximum/minimum residual electron
23
˚
densities in the final DF map were 0.86 and 20.6 e A
;
calculations were performed using the teXsan crystallo-
graphic software package of Molecular Structure Corpora-
tion [8].
3.9. Crystallography
3.9.1. Crystal data for 3a
C52H50Cl6Cu4N6OS4 M51370.08; monoclinic, space
˚
group I2/a; a524.238(1), b528.949(1), c517.640(1) A;
4. Results and discussion
3
˚
b594.78(1)8; U512334.0(2) A , Z58, calculated density
1.476 g cm23. Orange needle, dimensions 0.0130.0330.2
4.1. Preparation of [CuX2(Ph2SNH)2 ]
mm, m (Mo Ka)51.796 mm21, l50.71073 A, F(000)5
˚
5552.
In a recent communication we noted that reaction of
Ph2SNH with CuCl2 in MeCN in the ratio 2:1 resulted in
the formation of trans-[CuCl2(Ph2SNH)2]. Intriguingly we
found that the latter could be crystallised in two forms:
blue (which proved to have square-planar geometry) and
green (pseudo-tetrahedral) [5]. Such isomerism is well-
known for ionic Cu(II) species but is much rarer for
neutral complexes – indeed we have yet to confirm the
3.9.2. Crystal data for 4a
C48H44Cl2CuN4S4 M5939.59; monoclinic, space group
P21 /n; a514.960(2), b510.372(1), c515.185(2) A; b5
105.663(8)8; U52268.7(4) A , Z52 (molecule located
˚
3
˚
about a crystallographic centre of symmetry), calculated
density 1.375 g cm23. Purple prism of dimensions 0.23

3176
P.F. Kelly et al. / Polyhedron 18 (1999) 3173–3179
existence of any other example. To build upon these initial
results, and in order to gain a more complete picture of the
reactivity of 1 towards copper centres, we have now
developed methods to specifically isolate reasonable quan-
tities of either isomer (in our preliminary work we noted
that appearance of the green form seemed to be unpredict-
able and in very low yield) and have investigated the effect
that changing the nature of the halogen has on the system;
in addition we have determined the ability of copper to
coordinate more than two units of 1 and have isolated the
first examples of oxo cluster complexes of this ligand.
Simple recrystallisation of crude 2a from hot MeCN
results in the formation of the blue isomer, which may be
isolated as very well-formed, three-dimensional blue crys-
tals if the solution is cooled slowly. Conversely, if the
original material is dissolved in CH2Cl2, 60/80 petroleum
ether added and the mixture allowed to slowly evaporate
then the faster loss of CH2Cl2 results in crystallisation of
2a from the mixture as very fine, long needles of the green
form. In the latter case a small amount of the blue material
sometimes forms, but the difference in size and density of
the two types of crystal makes it easy to decant the needles
of the green isomer off from their blue counterparts.
Although this method does not lend itself to scaling up too
far, we have found that it can certainly be used on amounts
in the order of 100 mg or so, allowing isolation of
reasonable amounts of this isomer for the first time.
If samples of the green isomer of 2a thus formed are
redissolved in a small volume of hot MeCN and the
solutions cooled in a freezer, the material crystallises out in
the blue form, thus confirming that a true polymorphism is
present. This begs the question as to the nature of the
driving force behind this unique isomerism, and some
insight into this problem may be gained by the results
obtained from the analogous reaction with CuBr2. The
copper bromide reaction proceeds in much the same
manner as the chloride, giving a green mixture, although in
this case the oxo cluster 3b (see below) tends to form as
substantial impurity. The latter can be avoided by first
forming pure [Cu(Ph2SNH)4]Br2 4b as detailed below and
then treating this with one equivalent of CuBr2 as in
Scheme 1. The [CuBr2(Ph2SNH)2] thus formed may be
isolated as dark green crystals from MeCN. The results of
an X-ray study on these crystals is not formally presented
here as the data1 resulted in a poor solution due to disorder
at one of the sulfurs; however, the results are good enough
to confirm that the geometry at the metal centre is pseudo-
tetrahedral (with, for example a N–Cu–N angle of 149.98),
Scheme 1. The formation of 2, 3 and 4 and the two isomers of 2a.
a conclusion that also fits in with the colour of the
material. Crucially, we have found no evidence for the
formation of any other isomer (i.e. square-planar) of this
material – crystallisation from hot MeCN, from petroleum
ether/CH2Cl2 or by slow diffusion of Et2O into CH2Cl2
only results in the formation of the green material. Thus it
seems that the act of changing from the chloride to
bromide ligands mitigates against the formation of a
square-planar isomer. What, if anything, this tells us about
the driving forces behind the isomerism in 2a is not
necessarily clear, though it possibly indicates that in-
tramolecular interaction (or, in solution before crystallisa-
tion, intermolecular interaction) between the N–H units
and the halide ligands could play a part. If this were the
case then changing the electronegativity of the halide
would clearly be expected to have a significant effect. In
proposing this model, however, we should clearly add the
caveat that we have no definitive proof of its veracity; one
observation which would add weight to the suggestion
would come with the structure of analogues of 2 in which
the N–H groups had been replaced with N–R, where R
was alkyl or aryl. In such cases there would be minimal
interaction between the R group and the chloride. Un-
fortunately, attempts to prepare analogous complexes of,
for example, Ph2SNnBu have so far only resulted in dark
red/brown coloured products from which we have yet to
isolate crystalline material. Work in this area is, however,
still ongoing.
4.2. Formation of oxo clusters
1Monoclinic, a521.637(1), b512.468(1), c518.364(1) A, b5
As we have seen, formation of 2a can be achieved by
reaction of 2 equivalents of 1 with CuCl2 in dry MeCN
under a nitrogen atmosphere; if 1.5 equivalents of 1 are
employed instead, and the reaction performed in air using
undried MeCN, a different product may be isolated. In this
case the solution gradually deposits an orange crystalline
˚
100.29(1)8. The structure exhibits disorder in one of the ligands resulting
in two occupancies for the associated sulfur atom of 70:30%. However,
the associated phenyl groups could not be resolved into separate
occupancies and were refined in one full weight position. This model
worked fairly well and gave unambiguous results for the geometry at the
Cu as no significant disorder is noted for either nitrogen.

P.F. Kelly et al. / Polyhedron 18 (1999) 3173–3179
3177
is bubbled through, or thoroughly degassed water added to
such a solution then a minimal yield of 3a is obtained (¯7
and 3%, respectively). Interestingly, if both degassed water
and dried O2 are added to the original anaerobic solution
then a larger yield is obtained (23%); the highest yield of
all, 50%, is obtained when the reaction is performed in the
presence of water and oxygen from the start. The latter
result is not surprising and, by analogy with other systems
investigated, probably suggests that initial solutions of
CuCl2
in
MeCN
are
in
equilibrium
with
[Cu4OCl6(MeCN)4] making the formation of 3a a simple
substitution reaction. Quite why the presence of both O2
and water would be required for the formation of 3a from
the anhydrous starting system is not obvious. The structure
of 3a consists of a Cu4OCl6 core in which the copper
atoms are tetrahedrally displaced about the centre oxygen
˚
at an average Cu–O bond length of 1.90 A, a distance
which is typical for such clusters (Table 2). While the
˚
average S–N bond distance (1.583 A) within the sulfimide
ligands is similar to that of other complexes of 1, the
average angle at the nitrogen (121.98) is significantly
smaller than in any complex of 1 we have yet prepared
(including 4a, Table 3). The latter observation may well
indicate that there is a weaker H . . . Cl interaction between
the N–H hydrogens and the bridging chlorides in 3a than
is found in either the terminal chloride complexes such as
2a or the ionic chlorides such as 4a.
The formation of the analogous bromine complex
[Cu4(m4-O)(m-Br)6(Ph2SNH)4] 3b, which we have char-
acterised by IR and microanalysis, is much more facile and
indeed is hard to avoid; even traces of water in the system
appear to take the reaction mixture through to this product.
It is for this reason that we prefer to prepare 2b from 4b
rather than directly from 1 plus CuBr2, as the latter method
invariably generates significant amounts of 3b contami-
Fig. 1. The molecular structure of [Cu4(m4-O)(m-Cl)6(Ph2SNH)4], 3a.
material which X-ray crystallography reveals as the cluster
compound [Cu4(m4-O)(m-Cl)6(Ph2SNH)4] 3a (Fig. 1).
This is an example of a well known class of oxo cluster
species of the general type [Cu4OCl6L4], which have been
reported with a range of ligands L including pyridine,
triphenylphosphine etc. [9]. Such products commonly form
when copper halide reactions are performed using the
hydrated material or when products are refluxed in air.
Studies have shown that the oxygen in the product comes
from water in the system; in the case of 3a results go some
way to confirming this though the mechanism is not
straightforward. If the reaction is performed in scrupulous-
ly dried and degassed MeCN no 3a forms; if dried oxygen
Table 2
˚
Selected bond distances (A) and angles (8) in complex 3a
Cu(1)–O(1)
Cu(1)–Cl(4)
Cu(2)–O(1)
Cu(2)–Cl(6)
Cu(3)–O(1)
Cu(3)–Cl(5)
Cu(4)–O(1)
Cu(4)–Cl(6)
1.904(4)
2.413(2)
1.906(4)
2.385(2)
1.898(4)
2.407(2)
1.909(4)
2.430(2)
Cu(1)–Cl(1)
Cu(1)–N(1)
Cu(2)–Cl(1)
Cu(2)–N(2)
Cu(3)–Cl(2)
Cu(3)–N(3)
Cu(4)–Cl(3)
Cu(4)–N(4)
2.444(2)
1.932(5)
2.458(2)
1.930(5)
2.410(2)
1.929(5)
2.372(2)
1.928(5)
Cu(1)–Cl(5)
N(1)–S(1)
Cu(2)–Cl(2)
N(2)–S(2)
Cu(3)–Cl(3)
N(3)–S(3)
Cu(4)–Cl(4)
N(4)–S(4)
2.476(2)
1.586(5)
2.489(2)
1.576(5)
2.499(2)
1.584(6)
2.517(2)
1.586(6)
Cu(1)–O(1)–Cu(2)
Cu(1)–N(1)–S(1)
Cu(2)–N(2)–S(2)
Cu(4)–N(4)–S(4)
110.0(2)
Cu(1)–O(1)–Cu(3)
Cu(2)–O(1)–Cu(3)
Cu(3)–O(1)–Cu(4)
108.9(2)
110.8(2)
109.9(2)
Cu(1)–O(1)–Cu(4)
Cu(2)–O(1)–Cu(4)
Cu(3)–N(3)–S(3)
110.5(2)
106.7(2)
122.1(3)
122.3(3)
123.2(3)
119.8(3)
Table 3
˚
Selected bond distances (A) and angles (8) in complex 4a
Cu–N(1)
N(2)–S(2)
Cu–N(2)–S(2)
1.991(4)
1.591(4)
124.2(3)
Cu–N(2)
N(1)–Cu–N(2)
N(2)–Cu–N(1A)
1.977(4)
87.7(2)
92.3(2)
N(1)–S(1)
Cu–N(1)–S(1)
1.592(4)
126.3(3)

3178
P.F. Kelly et al. / Polyhedron 18 (1999) 3173–3179
nant. The similarity in the IR spectra of 3a and 3b suggests
that structural differences between the two are probably
minimal.
positions (and thus making up an overall octahedral
coordination environment via long Cl–Cu bonds) as might
be expected, the chlorides actually lie substantially nearer
the CuL4 coordination plane. The side-on view shown in
Fig. 2 highlights this; the Cu–N–Cl angle is 468 indicating
just how strong this N–H . . . Cl interaction is. The strength
of the latter is, of course, undoubtedly heightened by the
fact that there are two such hydrogen bonds acting
cooperatively. In addition, it should be noted that there are
no significant long range interactions between the Cu axial
position and chlorides from other units; indeed the longest
4.3. Preparation of [Cu(Ph2SNH)4 ]X2
Reaction of 3b with excess 1 results in the breaking up
of the cluster structure and formation of the homoleptic
[Cu(Ph2SNH)4]21 cation; the latter is is best prepared,
however, in a more direct manner by addition of four
equivalents of 1 to CuCl2 or CuBr2. In both cases the
products can be isolated as purple crystalline material from
MeCN. Efficient formation of [Cu(Ph2SNH)4]Cl2 4a and,
especially, [Cu(Ph2SNH)4]Br2 4b can only be achieved by
addition of a slight excess of 1 (¯15%) above the precise
stoichiometric amount. Thus when exactly four equivalents
of 1 are added to CuBr2 the solution retains a quite intense
green colour and there is only a pale precipitate; it is only
upon addition of the extra 1 that the colour of the solution
is discharged and an intensely purple precipitate produced.
This presumably indicates that there is appreciable dis-
sociation of the complexes in solution. The crystal struc-
ture of 4a confirms the presence of the four sulfimide
ligands in a square-planar arrangement about the copper
centre (Fig. 2) with all the ligands bound by their nitrogen
donors. The most interesting aspect of the structure arises
from the relative orientations of the four ligands. They
arrange themselves in such a way that two pairs of cis
sulfurs extend on to opposite sides of the CuN4 plane. An
important consequence arising from the resulting orienta-
tion of the N–H groups is that each pair can cooperate in
significant interactions with the chloride counterion. One
result of this strong interaction (average H–Cl distance
˚
such interaction is as far out as 9.3 A. Four appears to be
the maximum number of diphenylsulfimide ligands that
copper will accommodate. Thus if either six equivalents of
1 are used in the synthesis from the start, or if solutions of
4 are heated in the presence of excess 1, 4 is the only
isolable product.
As Table 1 indicates, the IR spectra of all the complexes
prepared in this work show both N–H and S–N stretches
which vary significantly from complex to complex (most
other bands due to the ligands that are common to all vary
only minimally in wavenumber). The S–N stretches occur
in the range 888–947 cm21 and so can be quite diagnostic;
the precise pattern behind the variations in position is,
however, difficult to glean. The most dramatic of contrasts
within the spectra come with the N–H stretches. In the
case of neutral species, such as 2, the latter are weak; in
contrast the cationic species 4 exhibit strong, broad bands
that are comparable in magnitude to any other ligand bands
in the spectra. This difference undoubtedly originates from
the strong interactions between the N–H groups and the
counterions noted in the crystal structure of 4a.
To conclude it is apparent that 1 shows excellent ligand
properties towards copper centres. The isomerism shown
by 2a appears to be unique for a neutral Cu(II) species and
the methods outlined here allow isolation of pure samples
of each isomer in good yield. The lack of analogous
behaviour for 2b indicates that the isomerism in such
systems is significantly dependent upon the halide present.
Strong interactions between N–H groups and halide coun-
terions are a prominent feature of 4 (as they are for the
related cobalt species [Co(Ph2SNH)6]Cl2) and suggest
that, in the presence of an appropriate anion, it may well
be possible to generate extended arrays in which metal
centres are linked by such interactions. We are currently
investigating this possibility.
˚
2.20 A) is that rather than occupying the expected axial
Supplementary data
Supplementary data are available from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ on request quoting the deposition numbers
CCDC 127562 (3a) and CCDC 127563 (4a).
Fig. 2. The molecular structure of [Cu(Ph2SNH)4]Cl2 4a.

P.F. Kelly et al. / Polyhedron 18 (1999) 3173–3179
3179
[6] O.G. Vlasova, O.A. Rakitin, L.I. Khmelnitski, Org. Prep. Proc. Int.
26 (1994) 331.
References
[7] Siemens, SHELXTL, revision 5.03, Siemens Analytical X-ray Inc,
Madison, WI, 1995.
[8] teXsan, Crystal Structure Analysis Package, Molecular Structure
Corporation (1985 and 1992).
[9] R.E. Norman, N.J. Rose, R.E. Stenkamp, Acta Crystallogr., Sect. C
45 (1989) 1707.
[1] P.F. Kelly, A.M.Z. Slawin, A. Soriano-Rama, J. Chem. Soc., Dalton
Trans. (1996) 53 and references therein.
[2] P.F. Kelly, A.M.Z. Slawin, K.W. Waring, Inorg. Chem. Commun. 1
(1998) 249.
[3] P.F. Kelly, A.C. Macklin, A.M.Z. Slawin, K.W. Waring, R. Yates,
manuscript in preparation.
[4] P.F. Kelly, A.M.Z. Slawin, Chem. Commun. (1999) 1081.
[5] P.F. Kelly, A.M.Z. Slawin, K.W. Waring, J. Chem. Soc., Dalton
Trans. (1997) 2853.