The reaction of tertiary phosphines with (Ph2Se 2I2)2 - The influence of steric and electronic effects

Article abstract of DOI:10.1039/b608453b

The reaction of (Ph₂Se₂I₂)₂ with a wide variety of tertiary phosphines possessing different steric and electronic properties has been studied, leading in most cases to R ₃PSe(Ph)I adducts; [R₃

Full text of DOI:10.1039/b608453b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
The reaction of tertiary phosphines with (Ph Se I ) —the influence of steric
2
2 2 2
and electronic effects
Nicholas A. Barnes, Stephen M. Godfrey,* Ruth T. A. Halton, Imrana Mushtaq and Robin G. Pritchard
Received 15th June 2006, Accepted 17th July 2006
First published as an Advance Article on the web 3rd August 2006
DOI: 10.1039/b608453b
The reaction of (Ph
electronic properties has been studied, leading in most cases to R
2
Se
2
I
2
)
2
with a wide variety of tertiary phosphines possessing different steric and
PSe(Ph)I adducts; [R
P =
MeP (6), Me PhP (7), Me
P (9)]. All of the products formed were characterised by elemental analysis, Raman and
3
3
(
p-CH
3
C
6
H
4
)
3
P (1), (m-CH
3
C
6
H
4
)
3
P (2), (o-OCH
3
C
6
H
4
)
3
P (4), Ph
2
2
3
P (8),
Cy
3
multinuclear NMR spectroscopy. Both steric and electronic factors are important in determining the
structural motif (CT vs. ionic) observed in the solid-state. In general, highly basic phosphines result in a
lengthening of the Se–I interaction, and a preference for an ionic structure. The reaction with
(
o-CH
The unusually long P–I bond, [2.5523(12) A], and short I–I bond, [3.0724(4) A], exhibited by 3 is a
result of the high steric requirements of this phosphine. The similarly bulky (o-SCH P yields a
mixture of (o-SCH PSe(Ph)I (5a) and [(o-SCH PSePh]I (5b). The crystal structures of
m-CH PSe(Ph)I, 2, (o-CH PI , 3, [(o-OCH PSePh]I.CH Cl , 4,
(o-SCH PSePh]I , 5b, two pseudo-polymorphs of Ph MePSe(Ph)I, 6a/6b, and
Me PSe(Ph)I] Cl , 8, are reported. The R PSe(Ph)I adducts formed exhibit one of four types of
·CH
3
C
6
H
4
)
3
P does not yield a stable R
3
PSe(Ph)I adduct, and instead (o-CH
3
C
6
H
4
)
3
PI (3) is formed.
2
˚
˚
3
C
6
H
)
4 3
3
C
6
H
4
)
3
3
C
6
H
4
)
3
3
(
[
[
3
C
6
H
4
)
3
3
C
6
H
4
)
3
2
3
C
6
H
4
)
3
2
2
3
C
6
H
4
)
3
3
2
3
2
2
2
3
behaviour. Type I products, (such as 2) are CT in the solid-state and display fluxionality in solution.
Type II products (such as 6a/6b) lie close to the CT/ionic structural borderline, displaying long Se–I
−
bonds, and are more appropriately classified as [R
complexes ionise in solution to [R
3
PSePh] (acceptor)/I (donor) CT complexes. Type II
3
PSePh]I. Type III products, such as 8, are ionic in solution, but
frequently show cation–anion, or cation–solvent interactions in the solid-state, although these
interactions are weak and the linear P–Se–I motif is lost. Type IV products (such as 4) are ionic and
feature bulky phosphines. They display no short cation–anion interactions in the solid-state.
−
Introduction
and I the donor. Evidence for the latter interpretation has been
provided by the short P–I and long I–I distances observed in
The ability of group 15 or 16 donor molecules to form charge
transfer (CT) adducts with dihalogens or interhalogens has long
been recognised, with considerable emphasis being placed upon
these compounds, and the absence of a m(I–I) band in the Raman
5
spectra. In fact, the I–I distances vary considerably, ranging
i
˚
from 3.021(1) A for (Mecarb) Pr
2
PI
2
, [Mecarb = 1-(2-Me-1,2-
studies involving dihalogen adducts of tertiary phosphines, R
3
PX
2
,
6
˚
4
C
2
B
10
H
10)], to 3.408(2) A for PhMe PI
2
2
, suggesting that the type
in view of their widespread use as reagents in synthetic organic
of CT compound formed depends on the R groups. Extensive
1
2
3
chemistry. Ph
3
PBr
2
and Ph
3
PI
2
both display a four coordinate
31
1
solution phase P{ H} NMR studies on Ph
the molecular CT structure may persist in solution provided the
Ph P : I ratio is 1 : 1. The presence of an excess of Ph P results in
averaged signals, and is indicative of exchange between free Ph
and Ph PI . Where an excess of diiodine is present, polyiodide
species such as [Ph
3
PI
2
have shown that
CT ‘spoke’ structure Ph
3
P–X–X in the solid-state, with a linear P–
X–X arrangement. The nature of the adduct formed when tertiary
phosphines are reacted with dihalogens is greatly influenced both
by the halogen, and by the steric and electronic properties of the R
groups bound to the phosphine. For example, the CT complex
3
2
3
3
P
3
2
5,7
3
PI]I
3
and [(Ph
3
PI)
2
I
3
]I are observed.
3
˚
PhMe
shorter P–I bond [2.410(2) A], than that observed for Ph
2
PI displays a much longer I–I bond, [3.408(2) A], but
2
Studies of the adducts formed between phosphines and pseudo-
halogens are rarer, and the products formed are of interest since
they may exhibit different solid-state structures from the dihalogen
adducts, with steric factors likely to be more important. For
˚
3
PI , [I–I:
2
˚
3
.161(2), P–I: 2.481(4) A], illustrating the effect that an increase
4
in the basicity has upon the strength of the CT complex formed.
The interpretation of the CT structure of R PI
been the subject of some controversy, since these compounds may
be considered either as CT adducts with R P as the donor and
as the acceptor, or alternatively, with [R
3
2
compounds has
example, ionic complexes of the type [R PCN]X are formed when
3
8
R
3
P are reacted with XCN (X = Br, I); these complexes display
3
no cation–anion interactions in the solid-state. No evidence of
a CT motif was observed for any of these adducts, consistent
+
I
2
3
PI] as the acceptor
with structural data for R
more electronegative halogens (or in this case CN ).
recently, work has been undertaken studying the reactions of
3
PX compounds involving the harder,
2
−
9–11
More
School of Chemistry, The University of Manchester (North Campus),
Manchester, PO Box 88, UK M60 1QD. E-mail: Stephen.godfrey@
manchester.ac.uk; Fax: +44 (0)161 200 4559; Tel: +44 (0)161 200 4525
12–14
R
3
P with PhSeX (X = Br, I),
a system involving the softer
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4795–4804 | 4795

View Article Online
−
SePh pseudohalogen moiety. The solid-state structures of PhSeBr
consistent with a weakening of the Se–Br bond when the much
and PhSeI are considerably different, the former existing as a
more strongly donating Me
3
P group is present. It therefore appears
14
“
square” Ph
latter is a centrosymmetric dimer, (Ph
that adducts of formula R PSe(Ph)X are readily formed when four
equivalents of R P are reacted with either Ph Se Br or (Ph Se
in anhydrous diethyl ether. These adducts have been shown
to possess structural isomerism in the solid-state, adopting either
a charge-transfer donor/acceptor motif, R PSe(Ph)X, Fig. 1(a),
or an ionic [R PSePh]X phosphonium salt structure, Fig. 1(b).
4
Se
4
Br
4
, linked by weak Se–Se bonds, whilst the
that some CT R PSe(Ph)X adducts with elongated Se–X bonds
3
15
2
Se
2
I
2
)
2
.
We have shown
do ionise in solution, and thus display isomerism between their
solid-state and solution structures, making structural assignment
by spectroscopic data alone unreliable.
3
3
4
4
4
2
2
I
)
2 2
12–14
Clearly, the structures adopted by R PSe(Ph)X compounds are
3
heavily influenced by the donor power of the tertiary phosphine.
Such structural dependency upon the identity of the R groups
necessitated a thorough investigation into the products obtained
3
3
In the CT compounds the geometry at selenium is T-shaped,
and may be considered a 10-Se-3 system, with a linear P–Se–X
angle. This structural motif has been observed for the adducts
from the reaction of (Ph
2
Se
2
I
2
) with a series of tertiary phosphines,
2
in order to compare the effects of the steric and electronic
properties of these phosphines. Additionally, these systems may
then be compared with the data obtained for the analogous
bromide adducts, in order to ascertain the effects of changing
the halogen in these systems. Herein we report the results of this
study.
12,14
Ph
3
PSe(Ph)X (X = Br, I).
An identical CT motif was observed
PTe(Ph)I, prepared by reaction of
These CT complexes may be thought of as
PX
(X = Br, I) systems, with the “PhSe” moi-
ety behaving as a pseudohalogen. The long Se–I bond [3.2564(5)
for the tellurium analogue, Ph
3
16
Ph
3
P with Ph
4
Te
4
4
I .
analogous to the Ph
3
2
˚
A] in Ph
3
PSe(Ph)I suggests that the CT system is best considered as
Results and discussion
−
the donation of electrons from the I to the r* (SeP) antibonding
orbital. However, Ph
3
PSe(Ph)I does not ionise in solution as a
The reaction of (Ph
possessing different steric and electronic properties was pursued.
These included substituted aryl, mixed aryl/alkyl, and trialkyl
2
Se
2
I
2
) with a range of tertiary phosphines
2
3
1
1
broad resonance is observed in the P{ H} NMR spectrum, and
no selenium satellites can be resolved. This suggests exchange
between Ph
3
P/PhSeX and Ph
3
PSe(Ph)X occurs in solution.
phosphines. Four molar equivalents of R
equivalent of (Ph Se in diethyl ether solution, affording mostly
yellow solids, see scheme 1.
3
P were reacted with one
2
2
I
)
2 2
3
Fig. 1 Structural isomers for R PSe(Ph)I compounds, (a) CT motif, (b)
ionic selenophosphonium salt.
3 2 2 2 2
Scheme 1 Synthesis of R PSe(Ph)I compounds from (Ph Se I ) .
The second structural motif adopted by R
pounds is an ionic [R PSePh]X selenophosphonium salt for-
mulation, which has been observed for [(Me
3
PSe(Ph)X com-
The compounds formed were characterised by elemental anal-
ysis, Raman and multinuclear NMR spectroscopy, and where
possible by single crystal X-ray diffraction. In most cases the
3
13
2
N)
3
PSePh]I, and
14
[Cy
3
PSePh]Br. Ionic compounds tend to be formed with
products formed analysed as R PSe(Ph)I, but the products
3
more strongly donating phosphines, resulting in elongation,
and cleavage of the Se–X bond, although in the structure of
obtained from the reactions with substituted aryl phosphines
(1 to 5) were often rich in iodine. This was especially true
13
˚
[
(Me
2
N)
3
1
PSePh]I a long Se · · · I contact is observed, 3.825(1) A.
for the reactions of (Ph
SCH P, with the former analysing to (o-CH
and the latter to a mixture of products (see below). With the
exceptions of (o-CH PI , 3, and (o-OCH PSe(Ph)I,
4, all the substituted aryl products produce varying amounts
of [R PSePh]I adducts, accounting for the high iodine content.
This was confirmed by Raman spectroscopy with bands typically
2
Se
2
I
2
)
2
with (o-CH
3
C
6
H
4
)
3
P and (o-
3
1
The P{ H} NMR spectra of the ionic compounds display well
3
C
6
H
4
)
3
3
C
6
H
4
)
3
PI
2
,
resolved resonances with attendant selenium satellites. A series of
R
3
PSe(Ph)Br adducts, [where R
3
= Me
2
Ph, Me
3
, Cy
3
, and (R
2
N)
3
3
C
6
H
4
)
3
2
3
C
6
H
)
4 3
n
(
R = Me, Et, Pr )], all displayed similar NMR features, with
1
J(SeP) couplings varying between 412 and 497 Hz, and on this
3
3
14
basis were all assigned ionic structures. However, assignment
by NMR spectroscopic data alone must be treated with some
−
1
observed at ca. 110 cm . Peaks attributable to m(P–Se) were
caution, as exemplifiedby Me
3
PSe(Ph)Br. This compound displays
observed for most of the complexes, typically between 450 and
−
1
all the features of an ionic adduct in solution, viz. a sharp peak
490 cm . The m(P–Se) bands in the R PSe(Ph)I compounds are
3
3
1
1
in the P{ H} NMR spectrum at 37.3 ppm with coupling to
found at lower wavenumbers than m(P=Se) bands for tertiary
1
17
selenium, J(SeP) = 412 Hz. Yet the structure observed in the
phosphine selenides, and also at lower wavenumbers than the
14
solid-state is CT, with a linear P–Se–Br angle. The Se–Br bond
m(PSe) bands in halogen adducts of phosphine selenides, R
where the P–Se bonds contain significant double bond character.
3
PSeI
2
,
18
˚
of 3.327(3) A is considerably weaker than in the Ph
3
P analogue,
4
796 | Dalton Trans., 2006, 4795–4804
This journal is © The Royal Society of Chemistry 2006

View Article Online
Reactions with substituted aryl phosphines
A series of tri-tolylphosphines, (p-CH
angle away from a linear environment. The tolyl rings in 2 do
not adopt a regular propeller orientation, as one of the rings
3
C
6
H
4
)
3
P, (m-CH
3
C
6
H
4
)
3
P
is orientated in the opposite direction. The three Se–P–C
i
–C
o
and (o-CH P, were reacted with (Ph
to yellow adducts 1 to 3, respectively. The P{ H} NMR spectra of
and 2 displayed broad resonances with no discernible selenium
satellites, suggesting that these adducts do not ionise in solution
3
C
6
H
4
)
3
2
Se
2
1
I
2
)
2
, yielding orange
torsion angles differ considerably from each other; Se(1)–P(1)–
3
1
◦
◦
C(7)–C(12): 127.4(2) , Se(1)–P(1)–C(14)–C(15): −17.3(3) , Se(1)–
1
◦
P(1)–C(21)–C(26): −69.48(19) .
The reaction between the bulkier (o-CH
3
C
6
H ) P and
4
3
1
2
and may exhibit a CT structure similar to Ph
reaction of (m-CH P with (Ph Se afforded pale yellow
crystals over time, which were suitable for X-ray crystallography.
m-CH PSe(Ph)I, 2, crystallises in the triclinic space group
3
PSe(Ph)I. The
(
Ph
orange solid, 3, after ca. one week. The orange solid formed did
not analyse to a stoichiometry of R PSe(Ph)I, but was rich in
iodine, and microanalytical values suggested a formulation of (o-
CH PI . Confirmation of this assignment was sought by
2
Se
2
I
2
)
2
proceeded more slowly with formation of a bright
3
C
6
H
4
)
3
2
2
I
)
2 2
3
(
3
C
6
H
)
4 3
P-1, and the molecular structure is shown in Fig. 2 below, along
with selected bond lengths and angles.
3
C
6
H
4
)
3
2
X-ray crystallography and orange crystals of 3 were grown. The
structure obtained was indeed the diiodide adduct (triclinic, P-1),
as shown in Fig. 3, along with selected bond lengths and angles.
3 6 4 3
Fig. 2 Molecular structure of (m-CH C H ) PSe(Ph)I, 2. Thermal ellip-
soids are shown at the 30% probability level. Selected bond lengths [ A˚ ]
and angles [ ]: P(1)–Se(1): 2.2612(8), Se(1)–I(1): 3.2972(8), Se(1)–C(1):
Fig. 3 Molecular structure of (o-CH
shown at the 30% probability level. Selected bond lengths [A
C
H
)
PI
, 3. Thermal ellipsoids are
3
6
4
3
2
◦
◦
˚
] and angles [ ]:
1
.932(3), P(1)–C(7): 1.801(3), P(1)–C(14): 1.797(2), P(1)–C(21): 1.798(3),
P(1)–I(1): 2.5523(12), I(1)–I(2): 3.0727(4), P(1)–C(1): 1.817(4), P(1)–C(8):
1.810(4), P(1)–C(16): 1.815(4), P(1)–I(1)–I(2): 174.32(3), I(1)–P(1)–C(1):
107.50(15), I(1)–P(1)–C(8): 112.60(15), I(1)–P(1)–C(16): 111.19(15),
C(1)–P(1)–C(8): 109.4(2), C(1)–P(1)–C(16): 107.9(2), C(8)–P(1)–C(16):
108.1(2).
P(1)–Se(1)–I(1): 172.05(3), P(1)–Se(1)–C(1): 96.55(8), C(1)–Se(1)–I(1):
9
1.32(7).
12
The structure of 2 is analogous to that of Ph PSe(Ph)I, and
3
3
1
1
is consistent with the CT assignment suggested by the P{ H}
◦
NMR spectroscopic data, although the P–Se–I angle of 172.05(3)
(o-CH
3
C
6
H
4
)
3
PI
2
displays the molecular “spoke” structure
PI compounds. The P–I bond in 3
is somewhat distorted from an ideal linear geometry. The Se–
I bond in 2, 3.2972(8) A, is slightly longer than observed for
observed previously for R
is 2.5523(12) A, which is considerably longer than observed for
3
2
˚
˚
3
˚
˚
˚
Ph
m-CH
.2612(8) A, and is similar in magnitude to both Ph
3
PSe(Ph)I, 3.2564(5) A, reflecting the increased basicity of
P compared with Ph P. The P–Se bond in 2 is
PSe(Ph)I,
P–Se: 2.2585(9) A], and other reported data for P–Se single
Ph
3
PI
2
, [P–I: 2.481(4) A], whilst the I–I bond is 3.0727(4) A,
3
˚
(
2
[
3
C
6
H
˚
4
)
3
3
compared to 3.161(2) A for Ph
shortest observed for an R PI
appears that the long P–I and short I–I bonds are a consequence of
the high steric demand of (o-CH P, as an argument based
3
PI
2
. The I–I bond is one of the
3
3
2
spoke compound, see Table 1. It
12
˚
19
˚
bonds, such as in P
4
Se
3
(average P–Se: 2.24(1) A), and a number
3
C
6
H
)
4 3
of 5-coordinate arylseleno-phosphoranes, where the P–Se bonds
on the electronic properties of the phosphine, [pK
a
of 2.73 for
20
23
˚
vary between 2.233(2) and 2.2432(13) A. The lengthening of
the Se–I bond in 2 compared to Ph PSe(Ph)I is contrary to
Ph
3
P vs. 3.08 for (o-CH
3
C
6
H
4
)
3
P)], would predict a shorter P–I
3
and longer I–I bond in 3 compared to Ph
3
PI
2
, due to the higher
expectations as the introduction of a more basic phosphine would
be expected to weaken the Se–I bond, and strengthen the P–
Se bond. This suggests that steric factors are important in 2,
basicity of (o-CH P. The phosphine adopts an almost
3
C
6
H
4 3
)
perfect propeller conformation about the P(1)–I(1) bond, with
◦
I(1)–P(1)–C
Typically, in structures containing (o-CH
rings adopt one of two conformations, denoted exo
exo orientation involves all three methyl groups pointing in the
i
–Co(Me) torsion angles of 54.6(4), 53.6(4) and 50.0(4) .
with the crystallographic cone angle of the (m-CH
in 2 calculated to be 165 , significantly larger than the cone
3
C
6
H
4
)
3
P group
3
C
6
H
4
)
3
P groups the aryl
or exo . An
◦
3
2
◦
angle of Ph
3
P in Ph
3
PSe(Ph)I, calculated as 149 . This steric
3
effect may also be responsible for the distortion of the P–Se–I
same direction, (where an exo substituent is defined as a group
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4795–4804 | 4797

View Article Online
Table 1 Comparison of P–I and I–I bonds in R
3
PI
2
molecular spoke
Ref.
structures
Compound
P–I/A˚
I–I/A˚
Ph
3
PI
PhPI
PI
PI
2,4,6-(MeO)
2
2.481(4)
2.410(2)
2.461(2)
3.161(2)
3.408(2)
3.326(1)
3
4
21
Me
2
2
t
Bu
Pr
3
2
i
3
2
2.409(2)/2.420(2) 3.383(1)/3.372(1) 22
2.482(1)
2.5978(14)
2.5523(12)
[
(
(
3
C
6
H
2
]
3
PI
2
3.3394(5)
3.021(1)
3.0727(4)
8
6
i
Mecarb) Pr
o-CH
2
PI
2
3
C
6
H
4
)
3
PI
2
, 3
This paper
(
2 10 10
Mecarb) = 1-(2-Me-1,2-C B H )
pointing towards the apex of a pyramid consisting of the three
para hydrogens at the base and the atom bound to phosphorus as
the apex). This is the conformation observed for the rings in 3. The
alternative is an exo
2
conformation with one ring twisted so that
the ortho substituent is in an endo orientation, pointing towards
the base of the pyramid. The E–P–C
i
–C torsion angle to this ring
o
◦
is close to ± 180 , the remaining two torsion angles being typically
◦
24–27
±
50–70 .
The adoption of the more open exo
3
conformation
Fig.
4
3 6 4 3 2 2
Molecular structure of [(o-OCH C H ) PSePh]I.CH Cl , 4.
in 3 suggests that the I
2
spoke provides minimal steric hindrance to
Thermal ellipsoids are shown at the 30% probability level and the
the ligand. The crystallographic cone angle of (o-CH
is large at 192 . The short I–I bond is similar to that of 3.021(1) A
3
C
6
H ) P in 3
4
3
◦
molecule of solvent is omitted for clarity. Selected bond lengths
˚
◦
[A]
˚
and angles [ ]: Se(1)–P(1): 2.241(5), Se(1)–C(22): 1.898(17),
i
observed for (Mecarb) Pr
correspondingly long P–I bonds. It would therefore appear that
these two R PI adducts are best considered as CT complexes
of the R P (donor)/I (acceptor) type. Furthermore, 3 does not
ionise to [(o-CH PI]I in solution, (in contrast to most other
PI
compounds), as the P{ H} NMR spectrum of the solid
dissolved in CDCl
shows a single resonance at −21.5 ppm, the
low frequency of which is not consistent with an [R PI] cation.
2
PI
2
, and both compounds also display
P(1)–C(1): 1.783(19), P(1)–C(8): 1.753(15), P(1)–C(15): 1.776(16), shortest
Se(1) · · · I(1) contact: 5.438, P(1)–Se(1)–C(22): 98.8(5), Se(1)–P(1)–C(1):
3
2
1
09.4(5), Se(1)–P(1)–C(8): 108.8(6), Se(1)–P(1)–C(15): 107.6(5).
3
2
3
C
6
H
)
4 3
3
1
1
13,14,30
˚
R
3
2
2.3027(9) A.
The aryl rings adopt an exo
2
conformation with
◦
3
torsion angles Se(1)–P(1)–C(1)–C(2): −50.7(13) , Se(1)–P(1)–
+
4,28
◦
◦
3
C(8)–C(9): −63.3(14) , Se(1)–P(1)–C(15)–C(16): −178.6(12) .
The two o-anisyl rings in exo conformations are orientated such
that short intramolecular interactions are set up between the anisyl
It therefore appears that the molecular CT structure of 3 is retained
in solution. This provides further evidence that 3 should not be
+
−
˚
considered a CT adduct of the [R
In view of the unusual formation of the diiodide adduct
from the reaction of (o-CH P and (Ph Se , we in-
vestigated the analogous reactions with (o-OCH
3
PI] (acceptor), I (donor) type.
oxygen atoms and the selenium atom; Se(1) · · · O(1): 3.002(11) A,
˚
Se(1) · · · O(2): 3.305(10) A, which are shorter than the sum of the
˚
3
C
6
H
4
)
3
2
2
I
2
)
2
van der Waals radii for selenium and oxygen, 3.42 A. The other
3
C
6
H
4
)
4 3
)
3
P and
P and
o-anisyl ring has the OMe group twisted away from selenium,
although there is a short intramolecular Se · · · H interaction,
(
(
o-SCH
Ph Se
3
C
6
H
4
)
3
P. The reaction between (o-OCH
3
C
6
H
˚
2
1
2
I
2
)
2
yielded a yellow powder, 4, after several days, the
Se(1) · · · H(20): 2.7920 A (sum of the van der Waals radii of
3
1
˚
P{ H} NMR spectrum of which displayed a peak at 35.2 ppm
selenium and hydrogen: 3.10 A). These interactions may help to
1
with attendant selenium satellites, J(SeP) = 474 Hz. The pres-
favour the exo
angle is again very large (190 ).
2
conformation, although the crystallographic cone
◦
ence of selenium satellites suggests that [(o-OCH
3
C
6
H
4
)
3
PSePh]I
1
is formed, the J(SeP) coupling being consistent with values
The reaction between (Ph
2
Se
2
I
2
)
2
and (o-SCH
3
C
6
H ) P is slow,
4
3
13,14
observed for other ionic [R
crystals were grown from diethyl ether/dichloromethane, and
confirmed as [(o-OCH PSePh]I.CH Cl , 4, (triclinic, P-
), which incorporates one molecule of dichloromethane co-
3
PSePh]X compounds.
Yellow
with formation of an orange–brown solid, 5, after several days. The
microanalysis again revealed the compound to be rich in iodine,
but the elemental composition matched neither a formulation of
3
C
6
H
4
)
3
2
2
3
1
1
1
R
3
PSe(Ph)I or R
3
PI
2
. The P{ H} NMR spectrum showed that
crystallised in the unit cell. The molecular structure is shown in
Fig. 4, along with selected bond lengths and angles.
a mixture of products had been formed, with peaks observed
at 52.1 (broad), 39.1 (sharp) and −30.6 ppm (broad), the latter
[(o-OCH
3
C
6
H
4
)
3
PSePh]I.CH
2
Cl
2
,
4, is the second ionic
of which corresponds to free (o-SCH
3
C
6
H ) P. A sample of 5
4
3
[
R
3
PSePh]I compound to be crystallographically characterised,
was dissolved in diethyl ether/dichloromethane and a crop of
red crystals were grown which were suitable for single crystal X-
13
but unlike [(Me
2
N) PSePh]I there are no cation–anion in-
3
teractions present in 4, the shortest Se · · · I separation being
ray studies. The structure obtained was [(o-SCH
5b, which crystallises in the monoclinic space group P2
3
C
6
H
4
)
3
PSePh]I
3
,
˚
5
.438 A, presumably as a result of the high steric bulk of (o-
1
/n. The
˚
OCH
3
C
6
H
4
)
3
P. The P–Se bond length in 4 is 2.241(5) A which is
formation of 5b is consistent with the high iodine content of the
bulk material 5, and also the Raman spectrum, which contains
29
˚
PSeMe]ClO , 2.205 A,
slightly longer than the P–Se bond in [Ph
3
4
−
1
but consistent with the P–Se distances observed for phenylse-
lenophosphonium cations, which range between 2.2260(16) and
strong peaks at 141 and 110 cm , consistent with the presence of
−
an I
3
ion. The species 5b is likely to be the sharp peak observed
4
798 | Dalton Trans., 2006, 4795–4804
This journal is © The Royal Society of Chemistry 2006

View Article Online
3
1
1
at 39.1 ppm in the P{ H} NMR spectrum, as the chemical shift
interaction, and the adduct is essentially ionic. The aryl rings
+
is consistent with an [R
3
PSePh] cation, although no selenium
again adopt an exo
2
conformation, with Se · · · S interactions to
˚
satellites could be resolved. It therefore seems likely that the broad
peak at 52.1 ppm is (o-SCH PSe(Ph)I, 5a, which is probably
fluxional in solution, hence the additional resonance for free (o-
SCH P is also observed. The molecular structure of 5b is
shown in Fig. 5 below, along with selected bond lengths and angles.
the two exo SMe groups, Se(1) · · · S(2): 3.314(2) A (shorter than
˚
C
6
H
4
)
3
the sum of the van der Waals radii of selenium and sulfur, 3.70 A),
3
˚
and one interaction slightly longer, Se(1) · · · S(3): 3.756(2) A. The
C
6
H
4
)
3
endo orientated ring again displays a short Se · · · H interaction,
3
˚
Se(1) · · · H(12): 2.7692 A.
Thus, whilst (p-CH
3
C
6
H
4
)
3
P and (m-CH
3
C
6
H
4
)
3
P react with
(
Ph
2
Se
2
I
2
)
2
to yield CT adducts of the type R
3
PSe(Ph)I, (with
−
small amounts of I
3
adducts detected by Raman spectroscopy),
the bulky tris-ortho substituted aryl phosphines behave somewhat
differently. The products obtained appear to be related to the
relative steric bulk of these phosphines, with small changes in
the ortho-substituents resulting in the formation of different
products. These observations can be rationalised by a comparison
of the known steric properties of the phosphines, and available
crystallographic data, both in the literature, and for the com-
pounds reported here. The steric properties of the phosphine
fragments in 2, 3, 4 and 5b have been calculated using both
the traditional crystallographic cone angle method, and also the
ꢀ
symmetric deformation coordinate, S
4
. This alternative measure
of steric characteristics is defined as the sum of the three E–
P–C angles (in the compounds here E = Se or I), minus the
32
ꢀ
Fig. 5 Molecular structure of [(o-SCH
3
C
6
H
4
)
3
PSePh]I
3
, 5b. Thermal
sum of the three C–P–C angles, thus a smaller S value will
4
ellipsoids are shown at the 30% probability level. Selected bond
lengths [ A˚ ] and angles [ ]: Se(1)–P(1): 2.2686(18), Se(1)–C(1): 1.939(6),
result from more sterically hindered phosphines. Table 2 shows
◦
ꢀ
the cone angles and S
4
values obtained from compounds 2 to 5b,
Se(1) · · · I(1_4): 3.8250(9), I(1)–I(2): 2.9264(7), I(2)–I(3): 2.8915(8),
with relevant details included from the structure of Ph
for comparison. It can be clearly seen that the cone angle
3
PSe(Ph)I
P(1)–C(7): 1.815(7), P(1)–C(14): 1.806(7), P(1)–C(21): 1.811(7), S(1)–C(8):
12
1
.760(7), S(1)–C(13): 1.802(7), S(2)–C(15): 1.766(7), S(2)–C(20): 1.799(8),
of (m-CH
3
C
6
H ) P is significantly smaller in these compounds
4
3
S(3)–C(22): 1.758(7), S(3)–C(27): 1.795(8), P(1)–Se(1)–C(1): 99.37(19),
P(1)–Se(1)–I(1#4): 175.32(5), I(1)–I(2)–I(3): 176.56(3).
than tris-ortho substituted aryl phosphines. The crystallographic
cone angles for the three tris-ortho substituted aryl phosphines
˚
in the compounds studied here are very similar, and the very
The P–Se bond in 5b is 2.2686(18) A, rather longer than
ꢀ
observed for 4. The asymmetric unit cell shows no close links
between the selenophosphonium cation and the I
ever, there is a very weak contact from the selenium atom to the
low magnitude of the S
4
values corroborates this. It appears
−
3
anion, how-
that the exo
distortion of the angles at phosphorus, as the S
2
conformations observed for 4 and 5b result in
ꢀ
4
values are
−
I
˚
3
anion of a neighbouring molecule, Se(1) · · · I(1#4): 3.8250(9)
negative, with larger C–P–C angles than Se–P–C. This further
suggests that the interactions between the Se atom and the EMe
groups (E = O, S) help to stabilise these compounds. In the case
A, which is slightly shorter than the sum of the van der Waals
˚
radii of selenium and iodine (3.88 A). The P(1)–Se(1) · · · I(1#4)
angle is close to linear at 175.32(5) . The I anion in 5b is
unsymmetrical, with the I(1)–I(2) distance, 2.9264(7) A, somewhat
◦
−
of (o-CH
3
C
6
H ) P there are no heteroatoms present on the ortho
4
3
3
˚
groups, suggesting that the high steric bulk of the phosphine either
disfavours formation of the R PSe(Ph)I adduct in the first place,
˚
longer than the I(2)–I(3) distance of 2.8915(8) A. The elongated I–
3
I bond contains the iodine atom which is weakly interacting with
or decomposition occurs, producing Ph
which may then react with (o-CH
presence of the stabilising interactions for (o-OCH
2
Se
P to produce 3. The
P, plus
P, vs. 3.08 for
2
and I
2
, the latter of
−
selenium. Such asymmetry in I
3
anions has been observed in
3
C
6
H
4 3
)
other systems where strong electrostatic interactions are present,
for example in extended iodine adducts of tertiary phosphines
3
C
6
H
)
4 3
33
(
a higher basicity (pK
a
of 4.47 for (o-OCH
3
C
6
H
4 3
)
7
,31
23
and phosphites), although in 5b this asymmetry is quite small
(o-CH
3
C
6
H
4
)
3
P), helps to stabilise 4, whilst the reaction with (o-
˚
SCH C H ) P results in a mixture of (o-SCH C H ) PSe(Ph)I, 5a
[0.0349(8) A], considerably less than in [(PhO)
3
PI]I
3
, 0.2493(7)
3
6
4
3
3
6
4 3
31
7
˚
˚
and [(o-SCH C H ) PSePh]I , 5b; the latter formed via the reaction
A, and [(Ph
3
PI)
2
I
3
]I
3
, 0.254(4) A. This suggests that the Se · · · I
3
6
4
3
3
contact in 5b is best regarded as a very weak electrostatic
of 5a with the excess iodine present in the system.
ꢀ
Table 2 Comparison of cone angles, S
4
values, and torsion angles in Ar
3
PSe(Ph)I compounds
◦
ꢀ
◦
◦
Compound
Crystallographic cone angle/
S
4
/
i o
Se–P–C –C torsion angles/
1
2
Ph
m-CH
(o-OCH
(o-SCH
o-CH
3
PSe(Ph)I
149
165
190
191
192
0.86
1.19
−5.20
−2.80
5.89
−117.0(2), −125.9(2), −138.1(2)
−17.3(3), −69.48(19), 127.4(2)
−50.7(13), −63.3(14), −178.6(12)
−52.0(5), −61.0(6), 171.6(5)
50.0(4), 53.6(4), 54.6(4)
(
[
[
3
C
3
6
H
4
)
H
3
PSe(Ph)I, 2
C
6
4
)
3
PSePh]I, 4
, 5b
3
C
6
H
4
)
3
PSePh]I
3
(
3
C
6
H
4
)
3
PI , 3
2
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4795–4804 | 4799

View Article Online
◦
Reactions with mixed aryl/alkyl phosphines
The reactions of (Ph Se with the mixed aryl/alkyl phosphines
anion interaction. The P–Se–I angle in 6a is 176.13(6) , close to
the expected linear geometry for a CT motif.
2
2
I
)
2 2
An alternative pseudo-polymorph of Ph MePSe(Ph)I, 6b, was
obtained when crystals were grown from a 1 : 1 mixture of
dichloromethane : diethyl ether at room temperature. The crystal
2
Ph MeP, and Me PhP have also been performed, with resultant
formation of the yellow compounds Ph MePSe(Ph)I, 6 and
2
2
2
3
1
1
Me
2
PhPSe(Ph)I, 7, respectively. The P{ H} NMR spectrum
structure was solved in the monoclinic space group P2 /c with two
1
1
of 7 displays a signal at 33.8 ppm with a J(SeP) coupling of
independent molecules in the unit cell, along with two molecules of
dichloromethane co-crystallised (one disordered). The molecular
structure is shown in Fig. 7 below, (solvents of crystallisation
omitted), along with selected bond lengths and angles.
4
27 Hz, data which is identical to that observed for the bromide
14
analogue [Me PhPSePh]Br, suggesting that 7 adopts the same
2
ionic structure in solution. We have previously assigned a CT
structure to the compound Ph MeSe(Ph)Br on the basis of NMR
spectroscopic data, as a broad peak was observed at 35.8 ppm
2
3
1
1
in the P{ H} NMR spectrum with no observable selenium
satellites. In contrast, the iodide analogue 6 displays a sharp
1
peak at 34.4 ppm with a J(SeP) coupling of 445 Hz. This
suggests that 6 does ionise in solution to [Ph MePSePh]I, and thus
2
behaves differently to the bromide adduct. Crystals of 6 have been
grown in two different solvents, diethyl ether and dichloromethane
and we have obtained two pseudo-polymorphs, (one containing
a solvent of crystallisation), from the different conditions. The
◦
orthorhombic form (6a) was obtained from diethyl ether at 0 C,
and crystallises in the space group Pbca. The molecular structure
of 6a is shown in Fig. 6, along with selected bond lengths and
angles.
2 2 2
Fig. 7 Molecular structure of Ph MePSe(Ph)I.CH Cl , 6b. Thermal
ellipsoids are shown at the 30% probability level and the molecule of solvent
◦
is omitted for clarity. Selected bond lengths [ A˚ ] and angles [ ]: Se(1)–P(1):
2
3
1
.2244(15), Se(2)–P(2): 2.2189(15), Se(1)–I(1): 3.4885(8), Se(2)–I(2):
.5235(8), Se(1)–C(14): 1.933(5), Se(2)–C(34): 1.925(6), P(1)–C(1):
.790(5), P(1)–C(7): 1.788(6), P(1)–C(13): 1.783(6), P(2)–C(21): 1.800(6),
P(2)–C(27): 1.803(5), P(2)–C(33): 1.785(6), P(1)–Se(1)–I(1): 173.94(6),
P(1)–Se(1)–C(14): 94.97(16), P(2)–Se(2)–I(2): 165.42(5), P(2)–Se(2)–C(34):
9
4.30(17).
The two molecules in 6b show similar features to 6a, with short
˚
Se–P bonds of 2.2244(15) and 2.2189(15) A, and long Se–I bonds
˚
of 3.4885(8) and 3.5235(8) A. The P–Se–I angles observed for
the two molecules deviate quite considerably from linearity, [P(1)–
◦
◦
Se(1)–I(1): 173.94(6) , P(2)–Se(2)–I(2): 165.42(5) ]. This deviation
is quite marked, especially in the second molecule, and reflects
the borderline nature of the structure of Ph MePSe(Ph)I. This
2
distortion of the linear P–Se–I motif is suggestive that this
compound is close to exhibiting an ionic structure with a bent
P–Se–I interaction, although the Se–I distances observed in both
6
a and 6b are considerably shorter than seen for the ionic
13
˚
[(Me
2
N)
3
PSePh]I, where the Se–I interaction was 3.825(1) A, and
Fig. 6 Molecular structure of Ph
2
MePSe(Ph)I, 6a. Thermal ellipsoids
◦
the P–Se–I angle measured 114.5(2) . In view of the considerable
are shown at the 30% probability level. Selected bond lengths [ A˚ ]
elongation of the Se–I bonds, it is not surprising that 6 ionises
◦
and angles [ ]: Se(1)–P(1): 2.229(2), Se(1)–C(14): 1.933(7), Se(1)–I(1):
31
1
in solution to [Ph
2
MePSePh]I, as suggested by the P{ H} NMR
3
.5313(14), P(1)–C(1): 1.777(7), P(1)–C(2): 1.785(7), P(1)–C(8): 1.795(7),
data, and it appears that 6 exhibits different structures in the solid-
P(1)–Se(1)–I(1): 176.13(6), C(14)–Se(1)–I(1): 83.6(2), P(1)–Se(1)–C(14):
1
4
state and solution, as observed for Me
3
PSe(Ph)Br. The structure
9
3.8(2).
The Se–P bond of 2.229(2) A in 6a is shorter than that
of Me PhPSe(Ph)I, 7, would be of interest as a further increase in
2
the basicity of the phosphine is likely to result in an adduct which
displays further elongation of the Se–I bond, but unfortunately we
were unable to obtain suitable crystals.
˚
12
observed for Ph
3
PSe(Ph)I, consistent with the increased basicity
P. The increased donating power of
the phosphine also results in a considerable elongation of the Se–I
of Ph
2
MeP compared to Ph
3
Reactions with trialkyl phosphines
˚
bond, which in 6a is 3.5313(14) A. Whilst this is still well within
the sum of the van der Waals radii of selenium and iodine (3.88
The reaction of (Ph
phines, Me P and Cy
adducts. The analogous bromide complexes have previously been
reported by us, with [Cy PSePh]Br adopting an ionic structure
with no cation–anion interactions, whilst Me PSe(Ph)Br behaved
2
Se
2
I
2
)
2
with two highly basic trialkyl phos-
˚
A), it is considerably longer than observed for other R
3
PSe(Ph)I
3
3
P, resulted in the formation of R
3
PSe(Ph)I
compounds assigned a CT structure [Se–I: 3.256(3)–3.2972(8)
˚
A]. Therefore, 6a can either be described as a very weak CT
3
compound, or an ionic compound with a considerable cation–
3
4
800 | Dalton Trans., 2006, 4795–4804
This journal is © The Royal Society of Chemistry 2006

View Article Online
unusually as it was shown to be a weak CT compound in the
solid-state, but ionised to [Me PSePh]Br in solution, (d : 37.3,
may also be influenced by a number of weak hydrogen bonding
interactions from methyl protons (Fig. 9).
3
P
1
14
J(SeP) = 412 Hz). It was therefore of interest, (especially for the
Me
Me
3
P system), to study the analogous iodide compounds. Both
3
1
1
77
1
3
PSe(Ph)I 8, and Cy
3
PSe(Ph)I 9, display P{ H} and Se{ H}
NMR spectra identical to their bromide analogues, and are thus
assigned ionic structures in solution. We were able to grow crystals
of 8 from a refluxing 1 : 1 dichloromethane : diethyl ether solution
of the compound. [Me
3
PSe(Ph)I]
2
.CH
2
2
Cl 8, crystallises in the
orthorhombic space group Aba2, and incorporates half a molecule
of dichloromethane (disordered), see Fig. 8.
Fig. 9 Secondary I · · · H interactions in the extended structure of
[
Me
3
PSe(Ph)I]
2
.CH
2
Cl
2
. Contact distances [ A˚ ]: I(1) · · · H(9b): 3.02(5),
I(1) · · · H(7b): 3.16(4), I(1) · · · H(8c): 3.06(6).
It would therefore appear that a very weak CT array may be
viable in R PSe(Ph)X systems, even for electron-rich phosphines,
3
providing the steric demand of the phosphine is low, although 8 is
probably best described as an ionic compound in the solid-state,
with significant cation–anion and cation–solvent interactions. The
solid-state structure can clearly be influenced by secondary hydro-
gen bonding or solvent interactions, which appear to contribute
to the presence of a linear P–Se–X motif in 8.
3 2 2 2
Fig. 8 Molecular structure of [Me PSe(Ph)I] .CH Cl , 8. Thermal ellip-
soids are shown at the 30% probability level. Solvent disorder omitted for
◦
clarity. Selected bond lengths [ A˚ ] and angles [ ]: Se(1)–P(1): 2.2220(9),
Se(1)–I(1): 3.7636(4), Se(1)–Cl(1a): 3.556(6), Se(1)–C(1): 1.939(3),
P(1)–Se(1)–Cl(1a): 174.94(9), P(1)–Se(1)–C(1): 96.91(11), C(1)–Se(1)–I(1):
Conclusions
1
68.71(11).
The reaction of (Ph
2
Se
2
I
2
2
) with a wide variety of tertiary
phosphines possessing different steric and electronic properties
has now been studied. Whilst the electron donating character of
the phosphine is chiefly responsible for the structural motif (CT vs.
ionic) observed in the solid-state, we have now shown that steric
factors are also important, since bulky aryl phosphines such as
Despite the fact that both 8 and Me
3
PSe(Ph)Br were recrys-
, a different structure
tallised from the same solvent, CH
2
Cl
2
is observed for 8, with the solvent of crystallisation having
an important impact. A weak Se · · · I interaction is present,
˚
Se(1) · · · I(1): 3.7636(4) A, which is considerably longer than
(o-CH
3
C
6
H
4
)
3
P do not form R
3
PSe(Ph)I adducts.
PSe(Ph)I adducts fall in a fairly
that observed in either of the Ph
2
MePSe(Ph)I structures, 6a
The P–Se bond lengths in R
3
˚
˚
and 6b, although the Se–P bond length is 2.2220(9) A, which
is similar in magnitude to 6a/6b. Additionally, a short contact
from the selenium atom to the solvent chlorine atom is observed,
compact range [2.189(15)–2.2686(18) A], irrespective of the struc-
1
tural motif observed. For the ionic complexes, J(SeP) coupling is
observed [412 to 474 Hz], but there is no evidence of a correlation
between the P–Se distance and the magnitude of the coupling. In
contrast, the Se–I distances observed in the CT complexes show
great sensitivity to the basicity of the tertiary phosphine, and there
˚
Se(1) · · · Cl(1a): 3.556(6) A, (compared to the sum of the van
˚
der Waals radii of selenium and chlorine of 3.65 A). Unusually,
the same molecule of dichloromethane bridges between two
Me
3
PSe(Ph)I units via these interactions, and the P(1)–Se(1)–
appears to be a direct linear correlation between the pK
the starting tertiary phosphine and the Se–I bond length observed
in the R PSe(Ph)I adduct. If one also considers the purely ionic 4,
where there is no Se · · · I interaction, then R PSe(Ph)I compounds
may be said to display one of four types of behaviour. Type I
adducts include (m-CH PSe(Ph)I, 2, and the previously
PSe(Ph)I. These adducts can be considered as 10-
Se-3 CT compounds with a linear P–Se–I arrangement, and Se–I
a
value of
◦
Cl(1a) angle is close to linear, 174.94(9) .
The iodine atom is unusually situated so that the C(1)–Se(1)–
I(1) angle of 168.71(11) , is approaching linearity, and it seems that
in 8a a situation intermediate between the CT and ionic motifs is
observed where the solvent molecule is more effective in competing
for the available electron density than the iodide anion, although
the latter may still be considered to be participating in a very weak
CT system. The site of the iodine atom in the crystal packing
3
◦
3
3
C
6
H
)
4 3
12
reported Ph
3
˚
bonds varying between 3.2564(5) and 3.2972(8) A. Whilst Se–I
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4795–4804 | 4801

View Article Online
bonds of this length suggest that these compounds ought to be
−
considered as CT adducts of the [R
3
PSePh] (acceptor)/I (donor)
type, they do not ionise in solution, but instead display fluxional
3
1
1
behaviour. From the P{ H} NMR spectroscopic data it is likely
that (p-CH PSe(Ph)I, 1, is also an adduct of this type. Type
II adducts lie close to the CT/ionic structural borderline, and are
exemplified by the two pseudo-polymorphs of Ph MePSe(Ph)I,
a and 6b. Here, the Se–I bond lengths are very long [3.4885(8)–
3
C
6
H
)
4 3
2
6
3
˚
.5313(14) A], but the P–Se–I angle displays only small deviations
from linearity. Ionisation does occur in solution to produce
[
[
Ph
2
MePSePh]I, and therefore 6a and 6b are best considered as
−
R
3
PSePh] (acceptor)/I (donor) CT adducts. Type III adducts are
ionic in solution, but if the steric size of the phosphine permits,
as for example in [Me PSe(Ph)I] .CH Cl 8, then both cation–
3
2
2
2
anion, and cation–solvent interactions may be observed in the
solid-state. Even in a compound with a highly basic phosphine,
1
3
such as (Me
2
N)
3
PSe(Ph)I (pK
a
= 9.73), there is a weak Se · · · I
interaction, but the linear nature of the P–Se–I angle is lost. The
lack of bulky aromatic rings in these systems may be significant,
with “soft–soft” interactions being the most favourable packing
force. There is also a fourth grouping where the steric effect of the
phosphine allows no close Se · · · I contact in the crystal packing,
as shown in the structure of [(o-OCH
Whilst (o-OCH P does form an R
CH P reacts with (Ph Se to yield (o-CH
suggesting that (o-CH PSe(Ph)I is too sterically hindered
adduct is formed in preference.
3
C
6
H
4
)
3
PSePh]I.CH
PSe(Ph)I adduct, (o-
PI , 3,
2
Cl
2
, 4.
3
C
6
H
4
)
3
3
3
C
6
H
4
)
3
2
2
I
2
)
2
3
C
6
H
4
)
3
2
3
C
6
H
)
4 3
to be stable, and an R PI
3
2
Experimental
Reagents and physical measurements
Most of the compounds synthesised here are air and moisture
sensitive so strict anhydrous and anaerobic conditions were used
throughout their synthesis. Diethyl ether was purchased commer-
cially (BDH) and freshly distilled from sodium/benzophenone
ketyl before use. All the tertiary phosphines used were purchased
commercially (Aldrich) except (o-SCH
3
C
6
H
4
)
3
P, which was syn-
was synthesised
34
thesised as described in the literature. (Ph
2
Se
2
I
)
2 2
15
as described previously. Elemental analyses were performed
by the UMIST Chemistry Department Microanalytical Service.
Raman spectra were recorded as solid samples on a Nicolet-Nexus
1
13
1
combined FT-IR/FT-Raman spectrometer. H and C{ H}
NMR spectra were recorded on a Bruker DPX400 spectrometer
3
1
1
operating at 399.9 and 100.6 MHz, respectively. P{ H} and
7
7
1
Se{ H} NMR spectra were recorded on a Bruker DPX200
spectrometer operating at 81.8 and 38.2 MHz, respectively. Peak
1
13
positions are quoted relative to external TMS ( H/ C), Me
2
Se
7
7
31
(
Se), and 85% H
3
PO
4
( P), using the high frequency positive
convention throughout.
Crystallographic details
Details of the structural analysis for compounds 2, 3, 4, 5b,
a, 6b and 8 are summarised in Table 3. Diffraction data were
recorded on a Nonius MAC3 CAD4 diffractometer (2, 6a and
b), or a Nonius j-CCD four-circle diffractometer (3, 4, 5b and 8)
using graphite-monochromated Mo-Ka radiation (k = 0.71073
6
6
˚
A), at 203(2) K (2, 6a and 6b), 150(2) K (4, 5b and 8), or
1
33(2) K (3). Structural data were solved by direct methods, with
4
802 | Dalton Trans., 2006, 4795–4804
This journal is © The Royal Society of Chemistry 2006

View Article Online
2
2
3
full-matrix least squares refinement on F using the SHELX-97
[s, SePh], 131.0 [d, C
o
, J(PC) = 13.8 Hz], 133.8 [d, C
m
, J(PC) =
35
31
1
program. Absorption corrections for compounds 2, 6a and 6b
were applied via the psi-scan method, for compound 4 via the
sphere method, and for compounds 3, 5b and 8 by the multiscan
method, using the SORTAV program. Non-hydrogen atoms were
refined with anisotropic thermal parameters. All hydrogen atoms
in 2 were refined isotropically, whilst those in compounds 4, 5b,
10.2 Hz], 136.5 [s, SePh], 146.2 [s, C
37.0 [s, br]. Raman (cm ): 3051, 2917, 1595, 1096, 805, 456, 355,
294, 161, 113.
p
]. P{ H} NMR (CDCl
3
):
−
1
(
m-CH
3
◦
C
6
H
4
)
3
PSe(Ph)I, 2. Yield = 78.4%. Yellow solid, mp
1
12–116 C. Calculated for C27
H
26PSeI: C, 55.2; H, 4.4; I, 21.6:
1
P, 5.3; Found: C, 53.8; H, 4.3; I, 24.7; P, 5.4. H NMR (CDCl
3
):
6
a and 6b were allowed for as riding atoms. In 3 all phenyl
1
3
1
2.37 [s, CH
3
], 7.18–7.30 [m, Ar], 7.51–7.56 [m, Ar]. C{ H} NMR
hydrogen atoms were allowed for as riding atoms, whilst the methyl
hydrogens were refined isotropically. In 8 all hydrogen atoms were
refined isotropically, except for atoms H10A and H10B which
were allowed for as riding atoms. All thermal ellipsoid plots were
1
(CDCl
3
): 22.1 [s, CH
3
], 119.8 [s, SePh], 120.2 [d, C
i
, J(PC) = 61.0
Hz], 129.5 [s, SePh], 130.7 [s, SePh], 131.0 [s, C
m
], 132.0 [s, C
, J(PC) = 15.8], 137.6 [s, SePh], 138.5 [s, C ], 141.1 [d,
]. P{ H} NMR (CDCl ): 47.9 [s, br]. Raman (cm ): 3049,
p
]
2
1
34.1 [d, C
o
o
3
1
1
−1
36
br, C
m
3
generated using ORTEP-3 for Windows.
2
4
914, 1591, 1576, 1221, 1086, 997, 854, 696, 673, 546, 488, 453,
15, 361, 222, 187, 152, 114.
CCDC reference numbers 279361 and 611257–611262.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b608453b
Reaction of (o-CH
3
C
6
H
4
)
3
P and (Ph
2
Se
2
I
2
)
2
.
0.231 g (7.6 ×
−
4
1
0
0
moles) of (o-CH
3
C
6
H
)
4 3
P was added to a solution containing
−
4
3
.215 g (1.9 × 10 moles) of (Ph
2
Se
2
I
2
) dissolved in 40 cm
2
Synthetic details
of dry diethyl ether. After ca. one week a bright orange solid,
34
Synthesis of tris-o-thioanisyl phosphine . 10.210 g (0.050
moles) of 2-bromothioanisole were dissolved in 100 cm of freshly
distilled dry diethyl ether. The solution was cooled in ice and was
treated dropwise over one hour with 20.10 cm of a 2.5 M BuLi
solution. The solution was left to stir at 0 C for a further hour,
after which time a solution containing 1.461 cm (2.300 g, 0.017
in 150 cm of dry diethyl ether was added dropwise
via a pressure equalising dropping funnel and the mixture was
stirred overnight. The solution was quenched with 150 cm of a
.2 M HCl solution and the white precipitate formed was filtered
off and washed with water, ethanol and diethyl ether. The solid was
then dried in vacuo to yield 13.166 g (65.5%) of tris-o-thioanisyl
(
o-CH
3
C
6
H
4
)
3
PI , 3 was formed. Yield = 52.3%. Calculated for
: C, 45.2; H, 3.8; I, 45.5; Found: C, 44.9; H, 3.7; I, 45.1.
): 2.30 [s, CH ], 7.28–7.45 [m, Ar], 7.54–7.63 [m,
2
3
C H21PI
21
2
1
H NMR (CDCl
3
3
3
n
13
1
3
Ar]. C{ H} NMR (CDCl
3
): 23.5 [d, CH
3
3
, J(PC) = 5.8 Hz], 120.4
◦
1
[
[
d, C
d, C
i
, J(PC) = 50.9 Hz], 127.6 [d, C
m
, J(PC) = 12.4 Hz], 133.8
3
2
3
o
, J(PC) = 10.9 Hz], 134.4 [s, C
p
], 135.0 [d, C
m
, J(PC) =
3
moles) of PCl
3
2
31
1
9.5 Hz], 143.7 [d, C
o
, J(PC) = 9.5 Hz]. P{ H} NMR (CDCl
3
):
−1
−
21.5 [s]. Raman (cm ): 3057, 3006, 2922, 1588, 1564, 1282, 1201,
3
1
132, 1047, 800, 661, 557, 515, 465, 399, 175, 137.
0
(
o-OCH
3
C
6
H
4
)
3
PSe(Ph)I, 4. Yield = 82.1%. Yellow solid, mp
PSeI: C, 51.0; H, 4.1; I, 20.0;
◦
1
55–157 C. Calculated for C27
H
26
O
3
1
1
Found: C, 50.9; H, 3.9; I, 20.3. H NMR (CDCl ): 3.65 [s, OCH ],
phosphine. Spectroscopic data is as follows: H NMR (CDCl
3
):
3
3
1
3
1
6
.95–7.17 [m, Ar], 7.33–7.76 [m, Ar]. C{ H} NMR (CDCl
3
):
2
.48 [s, SCH
3
], 6.73–6.78 [m, Ar], 7.06–7.12 [m, Ar], 7.35–7.39 [m,
1
1
3
1
4
57.2 [s, OCH ], 106.6 [d, C anisyl, J(PC) = 85.7 Hz], 113.3 [d,
Ar]. C{ H} NMR (CDCl
25.8 [s, C ], 127.1 [d, C
s, C ], 135.8 [d, C
3
2
): 17.7 [d, SCH
3
, J(PC) = 1.65 Hz],
3
i
3
C anisyl, J(PC) = 6.5 Hz], 119.3 [s, SePh], 122.2 [d, C anisyl,
2
m
m
o
1
[
m
o
, J(PC) = 4.8 Hz], 129.9 [s, C
p
], 133.9
1
2
J(PC) = 14.5 Hz], 129.8 [s, SePh], 130.8 [s, SePh], 135.7 [d, C
m
i
, J(PC) = 9.7 Hz], 144.6 [s, C
o
, J(PC) = 29.0
3
3
1
1
anisyl, J(PC) = 8.7 Hz], 137.1 [s, SePh], 137.9 [s, C anisyl], 161.1
Hz]. P{ H} NMR (CDCl ): −30.6 [s].
3
p
3
1
1
1
[
s, C
o
anisyl]. P{ H} NMR (CDCl
3
): 35.2 [s, J(SeP) = 474 Hz].
−
1
Raman (cm ): 3066, 2981, 2941, 2837, 1587, 1574, 1475, 1460,
General method of synthesis of R
3
PSe(Ph)I adducts
1
279, 1251, 1167, 1142, 1044, 791, 699, 176, 111.
To a dry diethyl ether solution containing one equivalent of
Reaction of (o-SCH
3
C
6
H
4
)
)
3
P and (Ph
P was added to a solution containing
2
Se
2
I
2
)
2
.
0.268 g (6.7 ×
(
Ph Se was added four equivalents of the tertiary phosphine
with vigorous stirring. The burgundy solution was rapidly de-
colourised with precipitation of the R PSe(Ph)I adduct (typically
2
2
I
2
)
2
−4
1
0
0
moles) of (o-SCH
3
C
6
H
4 3
−4
3
.190 g (1.7 × 10 moles) of (Ph
2
Se
2
I
2
) dissolved in 40 cm of
2
3
dry diethyl ether. After ca. one week an orange–brown solid,
which consists of a mixture of (o-SCH PSe(Ph)I, 5a,
and [(o-SCH PSePh]I , 5b, was formed. Calculated for
PSeI: C, 47.4; H, 3.8; I, 18.6; S, 14.1 Found: C, 46.1;
yellow in colour) when trialkyl or mixed alkyl/aryl phosphines
were used (compounds 6 to 9 below). The reactions with substi-
tuted aryl phosphines (compounds 1 to 5 below) took several days
3
C
6
H
4 3
)
3
C
6
H
4
)
3
3
C
27
H
26
S
3
to proceed to completion. The reaction with (o-SCH
yielded a mixture of products, (o-SCH PSe(Ph)I, 5a, and
P yielded (o-
, 3. All solids formed were isolated using standard
3
C
6
H
4
)
3
P
31
1
H, 3.8; I, 23.5; S, 15.7. P{ H} NMR (CDCl
3
): −30.6 [s, br, free
3
C
6
H
4
)
3
−1
(
2
4
o-SCH
3
C
6
H
4
)
3
P], 39.1 [s, 5b], 52.1 [s, br, 5a]. Raman (cm ): 3049,
[
(o-SCH
3
C
6
H
4
2
)
3
PSePh]I
3
, 5b, whilst (o-CH
3
C
6
H
)
4 3
987, 2914, 1572, 1425, 1267, 1167, 1105, 1036, 708, 650, 480, 449,
22, 403, 349, 229, 141, 110.
CH
3
C
6
H ) PI
4
3
Schlenk techniques. Characterising data for compounds 1 to 9 is
given below.
Ph
2
MePSe(Ph)I, 6. Yield = 52.6%. Yellow solid, mp 104–
◦
1
05 C. Calculated for C19
C, 46.8; H, 3.7; I, 26.0. H NMR (CDCl
12.8 Hz], 7.18 [m, Ar], 7.35 [m, Ar], 7.44–7.62 [m, Ar], 7.65–7.83
H
18PSeI: C, 47.2; H, 3.7; I, 26.3; Found:
1
2
(
p-CH
3
C
6
H
4
)
H
3
PSe(Ph)I, 1. Yield = 30.7%. Yellow solid. Cal-
3
): 3.01 [d, CH
3
, J(PH) =
culated for C27
26PSeI: C, 55.2; H, 4.4; I, 21.6; Found: C, 54.9; H,
1
13
1
3
7
.9; I, 24.2. H NMR (CDCl
3
): 2.46 [s, CH
3
], 7.30–7.46 [m, Ar],
[m, Ar], 7.90–7.92 [m, Ar]. C{ H} NMR (CDCl
3
): 13.8 [d, PCH ,
3
1
3
1
1
1
.68–7.76 [m, Ar]. C{ H} NMR (CDCl
3
): 22.2 [s, CH ], 119.7
3
J(PC) = 47.2 Hz], 119.4 [s, SePh], 120.2 [d, PPh
2
, J(PC) = 82.3
1
2
[
d, C
i
, J(PC) = 66.8 Hz], 120.8 [s, SePh], 128.8 [s, SePh], 129.9
Hz], 130.7 [d, PPh
2
, J(PC) = 13.1 Hz], 131.0 [s, SePh], 131.8
Dalton Trans., 2006, 4795–4804 | 4803
This journal is © The Royal Society of Chemistry 2006

View Article Online
3
[
s, SePh], 133.4 [d, PPh
2
, J(PC) = 10.9 Hz], 135.6 [s, PPh
2
], 138.0
2 N. Bricklebank, S. M. Godfrey, A. G. Mackie, C. A. McAuliffe and
3
1
1
1
R. G. Pritchard, J. Chem. Soc., Chem. Commun., 1992, 355.
[
s, SePh]. P{ H} NMR (CDCl
3
): 34.4 [s, J(SeP) = 445 Hz].
3
S. M. Godfrey, D. G. Kelly, C. A. McAuliffe, A. G. Mackie, R. G.
Pritchard and S. M. Watson, J. Chem. Soc., Chem. Commun., 1991,
1163.
7
7
1
1
Se{ H} NMR (CDCl
3
): 303.0 [d, J(SeP) = 445 Hz]. Raman
−
1
(
cm ): 3053, 2960, 2883, 1583, 1190, 1155, 1101, 1024, 997, 874,
685, 615, 511, 484, 415, 380, 307, 256, 229, 206, 129.
4 N. Bricklebank, S. M. Godfrey, H. P. Lane, C. A. McAuliffe, R. G.
Pritchard and J.-M. Moreno, J. Chem. Soc., Dalton Trans., 1995, 2421.
5
6
P. Deplano, S. M. Godfrey, F. Isaia, C. A. McAuliffe, M. L. Mercuri
and E. Trogu, Chem. Ber., 1997, 130, 299.
F. Teixidor, R. N u´ n˜ ez, C. Vi n˜ as, R. Sillanp a¨ a¨ and R. Kivek a¨ s, Angew.
Chem., Int. Ed., 2000, 39, 4290.
Me
for C14
2
PhPSe(Ph)I, 7. Yield = 64.3%. Yellow solid. Calculated
H
16PSeI: C, 39.9; H, 3.8; I, 30.2; Se, 18.8; P, 7.3; Found: C,
1
4
0.2; H, 3.9; I, 27.8; Se, 19.8; P, 7.1. H NMR (CDCl
3
): 2.57 [d,
2
CH
3
, J(PH) = 13.9 Hz], 7.08 [m, Ar], 7.22–7.24 [m, Ar], 7.33–
): 13.6 [d,
, J(PC) = 45.8 Hz], 119.4 [d, SePh, J(PC) = 6.5 Hz], 121.1
7 F. A. Cotton and P. A. Kibala, J. Am. Chem. Soc., 1987, 109, 3308.
8 S. M. Godfrey, C. A. McAuliffe, R. G. Pritchard and J. M. Sheffield,
J. Chem. Soc., Dalton Trans., 1998, 1919.
1
3
1
7
.37 [m, Ar], 7.41–7.46 [m, Ar]. C{ H} NMR (CDCl
3
1
2
PCH
3
9
W. S. Sheldrick, Acta Crystallogr., Sect. B, 1975, B31, 1776; F. Weller, D.
Nuszh a¨ r, K. Dehnicke, F. Gingl and J. Str a¨ hle, Z. Anorg. Allg. Chem.,
1991, 602, 7; K. M. Doxsee, E. R. Hanawalt and T. J. R. Wreakly, Acta
Crystallogr., Sect. C, 1992, C48, 1288; M. R. Kopp and B. Neum u¨ ller,
Z. Anorg. Allg. Chem., 1999, 625, 363.
1
2
[
1
1
d, PPh, J(PC) = 75.6 Hz], 130.2 [d, PPh, J(PC) = 13.8 Hz],
3
30.9 [s, SePh], 131.6 [s, SePh], 132.0 [d, PPh, J(PC) = 11.6 Hz],
4
35.2 [d, PPh, J(PC) = 2.9 Hz], 137.8 [d, SePh, J(PC) = 2.2
3
1
1
1
77
1
Hz]. P{ H} NMR (CDCl
3
): 33.8 [s, J(SeP) = 427 Hz]. Se{ H}
1
0 S. M. Godfrey, C. A. McAuliffe, R. G. Pritchard and J. M. Sheffield,
Chem. Commun., 1996, 2521; S. M. Godfrey, C. A. McAuliffe, R. G.
Pritchard and J. M. Sheffield, Chem. Commun., 1998, 921.
1 N. A. Barnes, S. M. Godfrey, R. T. A. Halton, S. Law and R. G.
Pritchard, Angew. Chem., Int. Ed., 2006, 45, 1272.
2 P. D. Boyle, S. M. Godfrey, C. A. McAuliffe, R. G. Pritchard and J. M.
Sheffield, Chem. Commun., 1999, 2159.
13 S. M. Godfrey, R. T. A. Ollerenshaw, R. G. Pritchard and C. L.
Richards, J. Chem. Soc., Dalton Trans., 2001, 508.
14 N. A. Barnes, S. M. Godfrey, R. T. A. Halton, I. Mushtaq, R. G.
Pritchard and S. Sarwar, Dalton Trans., 2006, 1517.
5 S. Kubiniok, W. W. du Mont, S. Pohl and W. Saak, Angew. Chem., Int.
Ed. Engl., 1988, 27, 431.
6 P. D. Boyle, W. I. Cross, S. M. Godfrey, C. A. McAuliffe, R. G. Pritchard,
S. Sarwar and J. M. Sheffield, Angew. Chem., Int. Ed., 2000, 39, 1796.
7 See for example: R. A. Zingaro, Inorg. Chem., 1963, 2, 192; R. P. Pinnell,
C. A. Mergerle, S. L. Manatt and P. A. Kroon, J. Am. Chem. Soc., 1973,
95, 977.
1
−1
NMR (CDCl
3
): 320.2 [d, J(SeP) = 427 Hz]. Raman (cm ): 3064,
3
8
041, 2972, 2952, 2883, 1587, 1572, 1398, 1155, 1113, 1032, 997,
77, 773, 685, 665, 615, 472, 442, 364, 303, 256, 233, 191, 125.
1
1
Me
3
PSe(Ph)I, 8. Yield = 59.8%. Yellow solid. Calculated for
C
9
H
14PSeI: C, 30.1; H, 3.9; I, 35.4; Se, 22.0; P, 8.6; Found: C, 30.7;
1
H, 3.9; I, 34.9; Se, 22.4; P, 8.2. H NMR (CDCl
3
): 2.49 [d, CH ,
3
2
J(PH) = 14.0 Hz], 7.37–7.45 [m, Ar], 7.48–7.57 [m, Ar], 7.60–
1
3
1
1
7
4
2
.67 [m, Ar]. C{ H} NMR (CDCl
3
): 14.3 [d, PCH
3
, J(PC) =
1
1
1
2
5.8 Hz], 118.9 [d, Ph, J(PC) = 8.0 Hz], 131.3 [d, Ph, J(PC) =
.1 Hz], 132.1 [d, Ph, J(PC) = 2.9 Hz], 138.0 [d, Ph, J(PC) = 2.9
3
1
1
1
77
1
Hz]. P{ H} NMR (CDCl
3
): 37.3 [s, J(SeP) = 412 Hz]. Se{ H}
1
−1
NMR (CDCl
3
): 295.8 [d, J(SeP) = 412 Hz]. Raman (cm ): 3049,
2
4
915, 1591, 1572, 1380, 1220, 1174, 1088, 998, 850, 692, 555, 524,
84, 468, 420, 360, 307, 248, 207, 193, 111.
1
8 S. M. Godfrey, S. L. Jackson, C. A. McAuliffe and R. G. Pritchard,
J. Chem. Soc., Dalton Trans., 1997, 4499.
1
2
9 E. Keulen and A. Vos, Acta Crystallogr., 1959, 12, 323.
0 V. A. Pinchuk, C. M u¨ ller, H. Th o¨ nnessen, P. G. Jones, R. Schmutzler
and Y. G. Shermolovich, Z. Anorg. Allg. Chem., 1996, 622, 1250.
Cy
3
PSe(Ph)I, 9. Yield = 70.3%. Cream solid. Calculated for
C
24
H
38PSeI: C, 51.2; H, 6.7; I, 22.6; Found: C, 51.6; H, 6.1;
1
I, 22.8. H NMR (CDCl
3
): 0.99–1.33 [m, CH
2
], 1.46–1.50 [m,
21 W. W. du Mont, M. B a¨ tcher, S. Pohl and W. Saak, Angew. Chem., Int.
Ed. Engl., 1987, 26, 912.
2
2
CH
2
], 1.63 [m, CH
2
], 1.76 [m, CH
2
], 2.47 [m, PCH, J(PH) =
2 F. Ruthe, P. G. Jones, W. W. du Mont, P. Deplano and M. L. Mercuri,
Z. Anorg. Allg. Chem., 2000, 626, 1105.
3
1
1.3 Hz, J(HH) = 10.2 Hz], 7.25 [m, Ar], 7.34 [m, Ar], 7.47 [m,
13 1
3 2 2 2
Ar]. C{ H} NMR (CDCl
): 25.3 [s, PCHCH
, J(PC) = 12.4 Hz], 28.1 [d, PCHCH
CH
CH
], 26.4 [d,
23 Md. M. Rahman, H. Y. Liu, A. Prock and W. P. Giering,
Organometallics, 1987, 6, 650.
24 T. S. Cameron and B. Dahl e` n, J. Chem. Soc., Perkin Trans. 2, 1975,
2
3
PCHCH
2
2
CH
2
, J(PC) =
1
2
4
7
(
1
2
8
.4 Hz], 34.3 [d, PCH, J(PC) = 24.7 Hz], 117.9 [d, Ph, J(PC) =
1
737.
3
1
1
.3 Hz], 131.1 [s, Ph], 132.1 [s, Ph], 138.2 [s, Ph]. P{ H} NMR
):
2
2
2
5 J. A. S. Howell, M. G. Palin, P. C. Yates, P. McArdle, D. Cunningham,
Z. Goldschmidt, H. E. Gottlieb and D. Hezroni-Langerman, J. Chem.
Soc., Perkin Trans. 2, 1992, 1769.
6 J. A. S. Howell, N. Fey, J. D. Lovatt, P. C. Yates, P. McArdle,
D. Cunningham, E. Sadeh, H. E. Gottlieb, Z. Goldschmidt, M. B.
Hursthouse and M. E. Light, J. Chem. Soc., Dalton Trans., 1999, 3015.
7 R. A. Baber, A. G. Orpen, P. G. Pringle, M. J. Wilkinson and R. L.
Wingad, Dalton Trans., 2005, 659.
1
77
1
CDCl
3
): 62.2 [s, J(SeP) = 438 Hz]. Se{ H} NMR (CDCl
3
1
−1
46.7 [d, J(SeP) = 438 Hz]. Raman (cm ): 3045, 2948, 2929,
853, 1569, 1443, 1358, 1293, 1204, 1176, 1048, 1032, 1014, 1000,
50, 813, 705, 667, 450, 300, 227, 204, 124.
Acknowledgements
28 N. Bricklebank, S. M. Godfrey, A. G. Mackie, C. A. McAuliffe, R. G.
Pritchard and P. J. Kobryn, J. Chem. Soc., Dalton Trans., 1993, 101.
29 P. G. Jones and C. Th o¨ ne, Z. Kristallogr., 1994, 209, 78.
We are grateful to the Engineering and Physical Sciences Research
Council (EPSRC) for a research studentship to RTAH, and also for
support of the UMIST FT IR-Raman (Grant No: GR/M30135),
NMR (Grant No: GR/L52246), and X-ray (Research Initiative
Grant) facilities.
3
3
3
3
3
0 J. M u¨ nchenberg, O. B o¨ ge, H. Th o¨ nnessen, P. G. Jones and R.
Schmutzler, Z. Anorg. Allg. Chem., 1998, 624, 655.
1 S. M. Godfrey, C. A. McAuliffe, A. T. Peaker and R. G. Pritchard,
J. Chem. Soc., Dalton Trans., 2000, 1287.
2 B. J. Dunne, R. B. Morris and A. G. Orpen, J. Chem. Soc., Dalton
Trans., 1991, 653.
3 Y. Yamashoji, T. Matsushita, M. Wada and T. Shono, Chem. Lett.,
1
988, 43.
References
4 G. Dyer and D. W. Meek, Inorg. Chem., 1965, 4, 1398; D. W. Meek, G.
Dyer and M. O. Workman, Inorg. Synth., 1976, 16, 168.
1
See for example: G. A. Wiley, R. L. Hershkowitz, B. M. Rein and
B. C. Chung, J. Am. Chem. Soc., 1964, 86, 964; A. G. Anderson and
F. J. Freenor, J. Am. Chem. Soc., 1964, 86, 5037; G. A. Wiley, B. M.
Rein and R. L. Hershkowitz, Tetrahedron Lett., 1964, 5, 2509; A. G.
Anderson and F. J. Freenor, J. Org. Chem., 1972, 37, 626.
35 G. M. Sheldrick, SHELX-97, University of G o¨ ttingen, G o¨ ttingen,
Germany, 1998.
36 ORTEP 3 for Windows: L. J. Farugia, J. Appl. Crystallogr., 1997, 30,
565.
4
804 | Dalton Trans., 2006, 4795–4804
This journal is © The Royal Society of Chemistry 2006