Chemistry and ligating properties of the 1,2,4-thiadiphosphole P2SC2Bu2t

Article abstract of DOI:10.1016/S0022-328X(02)01397-9

New synthetic routes to the 1,2,4- and 1,3,4-thiadiphospholes, P₂SC₂Bu₂,^t, are presented. Both η¹-W(CO)₅ and η⁵-M(CO)₃ complexes (M = Mo, W) of the 1,2,4-thiadiphosphole ring are described and the molecular structure of the former determined by a single crystal X-ray study is in good agreement with theoretical calculations. The thermal conversion of the η¹-M(CO)₅ complex to the η⁵-M(CO)₃ complex is attributable to the entropy change in the reaction. A sequence of [4 + 2] and [2 + 2 +2] cyclo-addition reactions of the 1,2,4-thiadiphosphole, P₂SC₂Bu₂^t_, with the phosphaalkyne, P≡CBu^t, affords the tetracyclic cage compound P₄SC₄Bu₄^t which has also been structurally characterised.

Full text of DOI:10.1016/S0022-328X(02)01397-9

Journal of Organometallic Chemistry 655 (2002) 7Á15
/
www.elsevier.com/locate/jorganchem
Chemistry and ligating properties of the 1,2,4-thiadiphosphole
P2SC2But2
Sarah E. d’Arbeloff-Wilson a, Peter B. Hitchcock a, John F. Nixon a,*, La´szlo´ Nyula´szi b
a School of Chemistry, Physics and Environmental Sciences, University of Sussex, Brighton BN1 9QJ, Sussex, UK
b Department of Inorganic Chemistry, Technical University of Budapest, H-1521 Budapest Gelle´rt te´r 4, Hungary
Received 2 January 2002; accepted 5 March 2002
Abstract
New synthetic routes to the 1,2,4- and 1,3,4-thiadiphospholes, P2SC2But2, are presented. Both h1-W(CO)5 and h5-M(CO)3
complexes (Mꢀ
single crystal X-ray study is in good agreement with theoretical calculations. The thermal conversion of the h1-M(CO)5 complex to
the h5-M(CO)3 complex is attributable to the entropy change in the reaction. A sequence of [4ꢁ
2] and [2ꢁ2ꢁ2] cyclo-addition
CBut, affords the tetracyclic cage compound
/Mo, W) of the 1,2,4-thiadiphosphole ring are described and the molecular structure of the former determined by a
/
/
/
reactions of the 1,2,4-thiadiphosphole, P2SC2But2, with the phosphaalkyne, Pꢀ
/
P4SC4But4 which has also been structurally characterised. # 2002 Published by Elsevier Science B.V.
Keywords: Thiadiphosphole; Phosphaalkyne; Carbonyl complexes
1. Introduction
(2JP(A)P(B)) 59 Hz). Markl et al. [2] found that the
¨
1,2,4-thiadiphosphole P2SC2Ph2 (2) was formed, albeit
in low yield, among a mixture of other cyclic products
from the reaction between a 1,3,4-oxa-thia-azole and the
Only very few 1,2,4- and 1,3,4-thiadiphospholes,
P2SC2R2, have been reported and there is only a single
chlorophosphaalkene Ph(SiMe3)Cꢁ
/
PCl.
report of a 1,2,5-thiadiphosphole (RꢀCF3). The first
/
The oily 3,5-di-tert-butyl-1,2, 4-thiadiphosphole (3)
was first reported by Lindner et al. [3] via an unusual
known 1,2,4-thiadiphosphole P2S(CSiMe3)C(SSiMe3)
(1) reported by Appel et al. [1] was synthesised by the
reaction of carbon disulphide with lithium bis-trimethyl-
silyl phosphide and chlorotrimethylsilane (Fig. 1).
synthetic route involving insertion of Pꢀ
Pꢁ
S bond of the h2-thiophosphinito manganese com-
plex [Mn(CO)4(PPh3)(PCy2S)]. The resulting intermedi-
ate is kinetically unstable and undergoes a [2ꢁ2] cyclo-
/
CBut into the
/
/
addition with a second molecule of phosphaalkyne to
give 3 with elimination of [Mn(CO)4(PPh3)(PCy2)]. The
orientation of the inserted phosphaalkyne is both
electronically and sterically controlled [4], thereby
resulting specifically in the 1,2,4-isomer. The same
authors [5] subsequently reported that 3 could also be
synthesised using [Co(PEt2S)(CO)2(PPh3)] and were able
to structurally characterise the crystalline 1,2,4-thiadi-
Fig. 1.
Characterisation of 1 was solely by elemental analysis
and NMR spectroscopic data, the 31P{1H}-NMR spec-
trum showing an [AB] system consisting of two doublets
at 309 (PA) and 304 (PB) ppm in the unsaturated region
with a characteristic two bond coupling constant
phosphole, P2SC2R2 (4) (Rꢀ
/
adamantyl) and establish
that the ring is planar [6].
In seeking alternative synthetic approaches to this
type of ring system, we took advantage of the reactivity
of the Pꢀ
its remarkable similarity with the Cꢀ
Phospha-alkyne cycloaddition reactions afford a variety
/
C triple bond in cycloaddition reactions [7] and
/
C triple bond [8].
* Corresponding author. Fax: ꢁ44-1273-677196.
E-mail address: j.nixon@sussex.ac.uk (J.F. Nixon).
/
0022-328X/02/$ - see front matter # 2002 Published by Elsevier Science B.V.
PII: S 0 0 2 2 - 3 2 8 X ( 0 2 ) 0 1 3 9 7 - 9

8
S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á15
/
of five-membered heterophospholes [7,9] and thiadi-
phospholes have been considered as possible [3ꢁ2]
products from cycloadditions between Pꢀ
CBut and a
PC(H)S intermediate [10].
Very recently we showed in a collaboration with the
Regitz group, [11] that the phosphalkyne, Pꢀ
CBut,
readily reacts with CS2 (or its complexes) to produce
an isomeric mixture of the 1,2,4- and 1,3,4-thiadiphosp-
hole rings 3 and 5 (Fig. 2) and provided a detailed
theoretical explanation of the reaction pathways. We
also structurally characterised some h1-Pt(II) complexes
of both ring systems.
/
/
/
Fig. 3.
Other possible synthetic routes were investigated to
see if alternative sources of reactive sulphur could be
found to combine with the two phosphaalkyne units.
One of the more simple and convenient routes involved
the UV irradiation of 2-methyl-thiacyclopropane in the
CBut, which readily led to a mixture of
both thiadiphospholes 3 and 5 in the ratio of 8:1, albeit
in overall poor yield, with propene presumed to be the
other product (Fig. 4).
Ruf [12] and Grobe et al. [13] have independently
CBut with CS2 but
presence of Pꢀ
/
described a similar reaction of Pꢀ
/
only identified 3. Regitz and coworkers [14], in unpub-
lished work, have very recently developed an alternative
synthetic route to 3 utilising Pꢀ
/
CBut and TaSCl3.We
Interestingly, Lawesson’s reagent (PhP(ꢁ
/
S)S)2 did not
present here new studies on other preparative routes to
the 1,2,4- and 1,3,4-thiadiphospholes 3 and 5 and the
reactivity and ligating behaviour of the former.
prove to be a source of sulphur in its reaction with Pꢀ
/
CBut, although there are reports of its use as a 1,3-
dipolarophile [16]. However, the more reactive dimeric
ferrocene derivative [Fe2(C5H4Me)2(C5H3MeP(S)S)2],
treated with Pꢀ
/
CBut (Fig. 5) did prove to be a mild
method of synthesis of the rings 3 and 5 in the ratio of
6:1
Unexpectedly, the direct reaction between Pꢀ
and elemental sulphur was unsuccessful, unlike the
analogous reactions with the other chalcogenides Se
and Te, which afford mixtures of the selena- and tellura-
diphospholes [17]. Likewise, treatment of Pꢀ
the phosphine sulphide, Ph3Pꢁ
ful, leading instead to the formation of several unchar-
acterised products.
2. Results and discussion
/
CBut
2.1. Alternative syntheses of the thiadiphospholes
P2SC2But2
/
CBut with
In the light of our previous results, in which one S
atom is effectively removed from CS2, in its reaction
with two phosphaalkynes, it was decided to extend this
chemistry to include other possible sources of sulphur. It
was anticipated that reactions of other heterocumulenes
/
S, also proved unsuccess-
such as isothiocyanides, RNꢁ
offer an alternative route to that using CS2 with the
elimination of the isocyanide CꢀNR and these expecta-
tions were fully borne out. Thus, treatment of SꢁCꢁ
NBut, with an excess of PCꢀBut, under gentle heat
/
Cꢁ
/
S, with Pꢀ
/
CBut might
2.2. Ligating properties of the 1,2,4-thiadiphosphole 3
/
2.2.1. Synthesis and characterisation of [W(CO)5h1-
P2SC2But2] (7)
/
/
/
Treatment of the 1,2,4-thiadiphosphole
3 with
for half an hour, led to the expected formation of 3 and
5 (ratio 4:1) as evidenced by their characteristic 31P{1H}-
[W(CO)5THF] in THF at ambient temperature afforded
the yellow complex [W(CO)5(h1-P2SC2Bu2t )] (7) which
was found to involve h1-coordination of the ring
phosphorus adjacent to sulphur (Fig. 6).
The 31P{1H}-NMR spectrum of a solution of 7
showed an [AB] spin system, with chemical shifts (d
259.2 P(B), 210.1 P(A) ppm). Since the corresponding
values for uncoordinated 3 are (d 266.1 P(B), 254.4 P(A)
ppm), the data strongly suggest that P(A) is the donor
NMR spectra, with simultaneous elimination of Cꢀ
/
NBut (Fig. 3).
The formation of the isonitrile was confirmed by its
subsequent treatment with [Mo(norbornadiene)(CO)4]
to afford the red crystalline product cis-[Mo(CN-
But)2(CO)4] (6) which was structurally characterised by
a single crystal X-ray diffraction study to be published
elsewhere [15].
Fig. 2.
Fig. 4.

S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á
/15
9
Fig. 7. Molecular structure of [W(CO)5h1-P2SC2Bu2t ] (7) together with
the atomic numbering scheme and some selected bond lengths (A) and
˚
bond angles (8) P(1)ꢂ
1.690(11), P(2)ꢂC(7)ꢀ
2.483(3), WꢂC(1)ꢀ2.014(11), P(1)ꢂ
C7)ꢀ101.8(5), P(2)ꢂC(7)ꢂP(1)ꢀ117.8(6), Sꢂ
P(1)ꢂWꢂC(1)ꢀ173.2(3).
/
Sꢀ
1.763(11), P(1)ꢂ
Sꢂ
/
2.045(4), Sꢂ
/
C(6)ꢀ
C(7)ꢀ
C(6)ꢀ
/
1.726(10), C(6)ꢂ
1.695(11), P(1)ꢂ
100.3(4), C(6)ꢂP(2)ꢂ
P(1)ꢂWꢀ113.04(14),
/
P(2)ꢀ
/
/
/
/
/
/
Wꢀ
/
/
/
/
/
/
/
/
Fig. 5.
/
/
/
/
/
/
/
/
/
/
those at P or S are closer to 1008. The two Pꢁ
/
C bond
distances (1.695(11), 1.690(11) A) lie in the expected
range for PꢁC double bonds and the structure is in good
˚
/
agreement with that obtained from theoretical calcula-
tions [23], on the complex 7 and on the free ring 3 (vide
infra). The B3LYP/LANL2DZ* calculated geometry of
7 is in reasonably good agreement with that of the X-ray
structure and also with that of the free ring (with the real
tert-Bu substituents) calculated at the same level of the
theory (See Table 1). Structural data are also similar to
those observed previously [6] for the thiadiphosphole
P2SC2Ad2 4.
Fig. 6.
centre and this was confirmed by the magnitude of the
one bond 183Wꢂ
/
31P coupling constant, measured from
the satellites around the resonance of P(A), (1J(P(A)W)
250 Hz) which lies in the expected range for tertiary
phosphine complexes containing an h1-ligated [W(CO)5]
fragment.
2.2.3. Synthesis and characterisation of [Mo(CO)3(h5-
P2SC2But2)] (8) and [W(CO)3(h5-P2SC2But2)] (9)
In view of the known exclusive h5-ligating properties
of the isoelectronic aromatic 1,2,4-triphosphole
1
As expected, the H-NMR spectrum of 7 shows two
distinct resonances corresponding to the different types
of But groups. Further evidence supporting the char-
acterisation of 3 results from the mass spectrum which
exhibits the expected parent ion at m/z 556 and peaks
corresponding to the stepwise loss of all five CO groups
(m/z 528, 500, 472, 444, 416, respectively) and a single
crystal X-ray structure determination.
P3RC2But2 (Rꢀ
transitions metal, [19Á
/
(Me3Si)2CHꢂ
/
) towards a variety of
/
22], reactions were also carried
out with the 1,2,4-thiadiphosphole 3 to see if it would
also prove possible to synthesise analogous h5-metal
complexes. Theoretical calculations concerning the ‘ar-
omaticity’ of 3 suggested that it ought to behave as a 6e-
2.2.2. Crystal and molecular structure of [W(CO)5h1-
P2SC2But2] (7)
Table 1
Experimental and calculated bond length data (A) for 3 and 7
˚
Suitable single crystals of 7 were grown from a hexane
solution and a single crystal X-ray diffraction study was
performed. The molecular structure which is shown in
Fig. 7, together with selected bond lengths and angles,
confirms the h1-ligation mode via P(A) and is similar to
those we published previously [18] for the analogous
1,2,4-selena and tellura-diphosphole complexes. A no-
teworthy structural feature of 7 is the planarity of the
ring The ring angles at carbon lie close to 1208 while
7: X-ray
7: Calculated
3: Calculated
P1S
2.045(4)
1.695(11)
1.736(11)
1.690(11)
1.726(10)
2.483(3)
2.07(2)
1.69(3)
1.76(5)
1.71(1)
1.72(4)
2.31(1)
2.08(9)
1.71(7)
1.78(0)
1.72(0)
1.73(7)
P1C7
P2C7
P2C6
SC6
P1W

10
S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á15
/
donor in view of the comparability of the magnitude of
the BDSHRT, BI and NICS parameters (53, 56 and
ꢂ
/
12.4, respectively) with those for thiophene (57, 68
and ꢂ13.2, respectively) [24].
/
Thus, treatment of the 1,2,4-thiadiphosphole 3 with
[Mo(cycloheptatriene)(CO)3] in THF readily afforded
[Mo(CO)3(h5-P2SC2Bu2t )] (8) in good yield (Fig. 8).
It was also possible to thermally convert the h1-
complex [{W(CO)5}(h1-P2SC2But2)] (7) on heating in
boiling THF, to the h5 product [W(CO)3(h5-P2SC2But2)]
(9) in moderate yield, with evolution of CO (Fig. 9).
Unfortunately, despite many attempts, it did not
prove possible to obtain crystals of either 8 or 9 suitable
for an X-ray diffraction study; however, the correct
formulation can be readily deduced from spectroscopic
data. The mass spectra of both h5-metal tricarbonyl
complexes were very similar, showing the expected
parent ions (m/z 412 (8), 500 (9)) and peaks correspond-
ing to stepwise loss of the three CO groups. The
31P{1H}-NMR spectra for 8 and 9 are also both
distinctly different from those of the corresponding h1-
pentacarbonyl complex 7 consisting of two resonances,
each of which appears as a doublet resulting from the
Fig. 9.
Table 2
31P{1H}-NMR data for 3, 7, 8 and 9 (d in ppm, J in Hz)
Compound
d P(A)
d P(B)
2J(P(A)P(B))
3
7
8
9
266.1
259.2
79.2
254.4
210.1
48.5
49.8
62.5
32.6
33.2
51.7
34.3
2
two bond J(P(A)P(B)) coupling (32 Hz). The resonances
for both P(A) and P(B) in 8 and 9 have shifted
significantly to high field compared with those in both
the free ring 3 and the complex [W(CO)5(h1-P2SC2Bu2t )]
Fig. 10.
2
(7) (Table 2). It can also be seen that the J(P(A)P(B))
coupling constant increases in the h1-complex 7 and
significantly decreases in the h5-complexes 8 and 9, in
comparison to the J(PP) found in the free thiadiphosp-
Table 3
Relative energies of 10, 11 and 12ꢁ2CO in terms of ZPE corrected
total energy (DEꢁZPE), enthalpy (DH) and Gibbs free energy (DG)
2
hole 3 reflecting the strong interaction of the whole of
the ring system with the [M(CO)3] fragments.
DEꢁZPE
DH298 K
DG298 K
DG500 K
10
11
0.0
2.0
0.0
2.7
0.7
0.0
3.0
0.0
4.6
12ꢁ2CO
35.0
36.3
16.2
3. Theoretical aspects of 1,2,4-thiadiphosphole
complexation
All data are in kcal molꢂ1 at the B3LYP/6-311ꢁG* level of the
theory.
The relative stabilities of the different complexation
modes of the 1,2,4-thiadiphosphole 3 were studied
computationally [23]. As model compounds, the com-
putationally amenable chromium complexes of the
parent 1,2,4-thia-diphosphole ring (C2H2P2S) have
been investigated. The three possible coordination
compounds 10, 11 and 12 are shown in Fig. 10.
energy are presented in Table 3. The most stable
structure according to the calculations is 10, for which
the analogous complexation mode has not been ob-
served for the 1,2,4-thiadiphosphole ring having the real
(tBu) substituents.
In terms of Gibbs free energy, however, 11 is slightly
more stable than 10.
The relative stabilities of 10, 11 and 12ꢁ2 CO in
/
terms of total energy, enthalpy (298 K) and Gibbs free
The complexation mode analogous to 10 has been
determined computationally [18] to be the most stable
one for the 1,2,4-selenadiphosphole (C2H2P2Se) (by 1.6
kcal molꢂ1 B3LYP/3-21G(*)), while the observed
coordination site for the tert-butyl substituted ring
was also at the 4-position. Apparently, the bulky tert-
butyl groups at the 3- and 5-position destabilise the
complex at the 4-position. Indeed, calculating the
relative stabilities of the two possible h1-complexes,
Fig. 8.

S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á
/15
11
the 2-[Cr(CO)5C2But2P2Se] compound, turned out to be
more stable by 6.5 kcal molꢂ1 (B3LYP/3-21G*) than
the corresponding 4-[Cr(CO)5C2But2P2Se] complex [18].
We, therefore, considered the relative stabilities of the
complexes (MꢀCr, Mo, W) [21]. An increase of the
/
bond angle sum about the tri-coordinate phosphorus
was also observed on complexation, indicating that the
lone pair of the tri-coordinate phosphorus is somewhat
more involved in the formation of the delocalised p-
system than in the parent uncomplexed system.
h1-complex 11 and the h5-complex 12 (ꢁ
/2CO). The
latter is far less stable than 11 and cannot be stabilised
by kinetic factors. To understand the formation of 12
from 11 on heating, the entropy factor should be
considered. Whereas the enthalpy function is nearly
unchanged by varying the temperature (not shown in
Table 3), the relative Gibbs free energies show a strong
variance, as expected since three particles are formed
4. Cyclo-addition reactions
4.1. Synthesis and characterisation of P4SC4But4 (13)
from one in the 110
/
12ꢁ
/
2CO reaction. At about 570 K
Although the above discussion indicates that the
1,2,4-thiadiphosphole 3 exhibits some aromatic charac-
the calculated Gibbs free energies of 11 and 12ꢁ
/2CO
are equal. Considering that the calculations refer to the
unsubstituted systems and the tert-Bu substituent must
have a larger steric repulsive effect on 11 (since both the
tert-Bu and Cr(CO)5 groups are in plane) than on 12, it
is understandable that 7 rearranges to 9 at 66 8C (i.e. at
much lower temperature than the calculated change in
the Gibbs free energies of the model compounds 11 and
ter, it was also found that it would undergo [4ꢁ
cycloaddition reactions with Pꢀ
CBut and in this respect
it differs from the isoelectronic triphosphole P3RC2Bu2t
(RꢀHC(SiMe3)2) which shows no such behaviour.
Thus, treatment of a solution of 3 in toluene with
two equivalents of Pꢀ
CBut and heating at 80 8C for 8
h, led to the formation of the tetracyclic product,
2,3,6,9-tetrat-butyl-5-thia-1,4,7, 8-tetraphosphatetracy-
clo[4.3.0.02,403,7]non-8-ene, P4SC4But4 (13) (see Scheme
1 below).
/
2]
/
/
/
12). Further to the understanding of the 7Á9 conversion,
/
the effect of the gaseous CO should be considered,
which on leaving the reaction mixture, will shift the
equilibrium further towards 12.
The calculated structural parameters (Table 4) of the
1,2,4-thiadiphosphole ring upon complexation are also
of interest. Whereas h1-complexation hardly effects the
geometry of the ring, as shown above, h5-coordination
results in significant equalisation of the bond lengths,
indicating an increased delocalisation and aromaticity.
It is also worthwhile mentioning that similar structural
changes were observed when comparing the molecular
structure of 1-(bis-trimethylsilyl-methyl)-2,4-di-tert-bu-
tyl-1,2,4-triphosphole [19] with those of its [M(CO)3]
Compound 13 was fully characterised by 31P{1H}-,
1H-NMR spectroscopy, mass spectroscopy, elemental
analysis and its molecular structure established by a
single crystal X-ray crystallographic determination. The
31P{1H}-NMR spectrum for 13 clearly shows the
addition of two equivalents of Pꢀ
/
CBut to the thiadi-
Table 4
˚
Calculated (B3LYP/6-311ꢁG*) bond lengths (A) of 10, 11 and 12
phosphole 3, as four signals are clearly visible. The high
field resonance at dꢀ
/
ꢂ98 ppm (P1) consists of a triplet
/
(2J(P(1)P(2)) 7 Hz), resulting from the equal two bond
The numbering of the ring atoms is identical to those used for the X-
ray structure.
Scheme 1.

12
S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á15
/
couplings to P2 and P3, and lies in the region typical for
a phosphirane system [25]. The low field doublet of
phosphaalkene unit in the structure is characterised by a
˚
typical Pꢁ
/
C bond length of 1.692(3) A, the P2ꢂ
/
P4 bond
carbon bond
˚
doublets resonance (dꢀ
the presence of a phosphaalkene fragment. This signal
shows direct bonding to P2 (dꢀ125 ppm) by a large
1J(P(4)P(2)) coupling constant of 286 Hz and is further
/406 ppm (P4)) clearly indicates
length is 2.2133(11) A and the phosphorusꢂ
/
˚
lengths lie between 1.857(3) and 1.905(3) A and are thus
in the usual range for phosphorusꢂcarbon single bonds.
The three-membered ring contains two almost identical
/
/
split by a typical 2J(P(4)P(3)) coupling of 18 Hz to P3 (dꢀ
/
bond lengths (1.861(3) A P1/C1, 1.857(3) A P1/C2) while
the internal angle at P1 of 49.938 is similar to those
found in the literature [25] for these types of com-
pounds. It should be mentioned here that analogous
products have recently been reported by Regitz and
˚
˚
3
116 ppm) and also shows a small J(P(4)P(1)) coupling
constant (6 Hz) to P1. Both chemical shifts of P2 and P3
are in good agreement with those of l3s3-phosphorus
atoms in comparable compounds [26,27]. The P3 signal
2
has a multiplet structure resulting from J(P(3)P(1))
coworkers, [17,26Á28], from the reaction between the
/
couplings to P1 and P4 (16 Hz each) as well as to P2
(33 Hz). The 1H-NMR spectrum of 9 shows the presence
of four different But groups and the mass spectrum
exhibits the expected parent ion (m/z 432). Although no
intermediates can be detected in the 31P{1H}-NMR
spectrum, the mechanism of formation is assumed to
1,2,4-oxa and selena-diphospholes and several phos-
phaalkynes which led to the formation of both possible
regioisomers.
5. Experimental
involve an initial [4ꢁ
3 and one equivalent of Pꢀ
mediate shown in Scheme 1, which then undergoes a
homo DielsÁAlder reaction with a further equivalent of
Pꢀ
CBut to give 13.
/
2] cycloaddition reaction between
/
CBut to afford an inter-
All manipulations were carried out using conventional
high vacuum and Schlenk line techniques, under an
atmosphere of dry argon, or dinitrogen in an MBraun or
/
/
MillerÁHowe glove box. Solvents were refluxed over
/
suitable drying agents, and distilled and degassed prior
to use. Toluene and C6H6 were refluxed over sodium,
4.2. Crystal and molecular structure of P4SC4But4 (13)
THF was refluxed over potassium. Petroleum ether (40Á
/
60 8C b.p.) and pentane were refluxed over sodiumÁ
/
A single crystal X-ray crystallographic analysis con-
firmed the structure of 13 which is shown in Fig. 11,
together with selected bond lengths and angles. The
potassium alloy. Calcium hydride was used as a drying
agent for MeCN. NMR solvents were dried over
potassium (d6-benzene), sodium (d8-toluene) or CaH2
(d2-dichloromethane), then vacuum transferred into
ampoules and stored under dinitrogen prior to use.
NMR spectra were recorded on Bruker DPX 300 or
AMX 500 spectrometers. Chemical shifts reported in
ppm (d) are relative to the residual proton chemical shift
of the deuterated solvent (1H), the carbon chemical shift
of the deuterated solvent (13C) and external H3PO4
(31P). Mass spectra were recorded on a VG autospec
Fisons instrument (Electron ionisation at 70 eV) by Dr.
A. Abdul-Sada. Elemental analyses were performed by
Micro Analytisches Labor Paschen (Germany) or A.T.
Stones of the Department of Chemistry, University
College, London.
Both Pꢀ
/
CBut and [Mo(cycloheptatriene)(CO)3] were
synthesised according to standard literature methods
[29,30] and [W(CO)5THF] was made by UV irradiation
of [W(CO)6] in THF for 6 h. CS2, ButNCS and
CH3CH2CHS were purchased from Sigma Aldrich and
used without further purification. The ferrocene
[Fe2(C5H4Me)2(C5H3MeP(S)S)2] was a gift from Pro-
fessor D.J. Woollins, University of St. Andrews, UK.
Fig. 11. Molecular structure of P4SC4But4 (13) together with the
˚
atomic numbering scheme and some selected bond lengths (A) and
bond angles (8). Sꢂ
1.857(3), P(1)ꢂC(1)ꢀ
1.875(3), P(2)ꢂP(4)ꢀ2.2133(11), P(3)ꢂ
1.877(3), P(3)ꢂC(1)ꢀ1.905(3), P(4)ꢂC(4)ꢀ
1.569(4), C(3)ꢂSꢂP(1)ꢀ95.71(9), C(2)ꢂP(1)ꢂC(1)ꢀ
P(1)ꢂSꢀ100.12(9), C(3)ꢂP(2)ꢂC(2)ꢀ93.34(12), C(3)ꢂ
96.77(9), C(4)ꢂP(3)ꢂC(3)ꢀ97.78(12), C(4)ꢂP(4)ꢂP(2)ꢀ
P(4)ꢂC(4)ꢂP(3)ꢀ117.63(15).
/
C(3)ꢀ
/
1.836(3), Sꢂ
/
P(1)ꢀ
C(3)ꢀ
C(4)ꢀ
1.692(3), C(1)ꢂ
49.93(11), C(2)ꢂ
P(2)ꢂP(4)ꢀ
98.16(10),
/
2.0939(13), P(1)ꢂ
1.866(3), P(2)ꢂ
1.851(3), P(3)ꢂ
/
C(2)ꢀ
C(2)ꢀ
C(3)ꢀ
C(2)ꢀ
/
/
/
1.861(3), P(2)ꢂ
/
/
/
/
/
/
/
/
/
/
5.1. Reaction of ButNꢁ
CBut (2.608 g, 26.08 mmol) was added to a
/
Cꢁ
/
S with Pꢀ
/
CBut
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
Pꢀ
/
/
/
/
/
/
/
/
/
/
solution of ButNCS (1.5 g, 13.04 mmol) in ether (10

S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á
/15
13
3
PCC(CH3)3P, J(PC) 8.9 Hz, J(PC) 7.9 Hz], 31.2 [d, 3C,
3
ml) and the reaction mixture was heated to 35 8C for 1/
2 h and then stirred for a further 3 h. The solvents were
removed in vacuo and the resulting yellow solution was
¯
3
PCC(CH3)3S, J(PC) 8.8 Hz] ppm.
¯
EIMS m/z (%):556 (36) [M], 528 (42) [Mꢂ
/
CO], 500
3(CO)], 444 (52) [Mꢂ
5(CO)] 232 (100) [Mꢂ5(CO)ꢂW].
Crystal data for 7: C15H18P2SW, Mꢀ556.1, mono-
clinic P21/n (No. 14), aꢀ6.345(2), bꢀ30.209(4), cꢀ
92.07(2)8, Vꢀ2040.2(14) A , Tꢀ
4, Dcalc 1.81 mg mꢂ3, mꢀ5.94 mmꢂ1
0.71073 A, F(000) 1072, crystal size 0.3ꢃ0.1ꢃ0.1
mm3 3939 measured reflections, 3606 independent
reflections (Rint 0.0925), 2759 reflections with I ꢀ
2s(I), Final indices R1ꢀ0.047, wR2ꢀ0.111 for I ꢀ
2s(I), R1ꢀ0.072, wR2ꢀ0.131 for all data. Data
collection: EnrafÁNonius CAD Program package
purified by column chromatography (kieselgelÁ
/hexane)
(38) [Mꢂ
/
2(CO)], 472 (71) [Mꢂ
/
/
to give a yellow oil (yield 1.138 g, 37.6%) which was
found to be the products 3 and 5 in the approximate
ratio of 4:1.
4(CO)], 416 (29) [Mꢂ
/
/
/
/
/
/
/
3
˚
˚
10.651(2) A, bꢀ
/
/
/
5.2. Reaction of CH3CH2CHS with Pꢀ
/
CBut
173(2) K, Zꢀ
/
/
,
˚
lꢀ
/
/
/
Pꢀ
/
CBut (0.81 g, 8.1 mmol) was added to a solution of
ꢀ
/
/
CH3CH2CHS (0.4 g, 5.4 mmol) in toluene (5 ml) and the
reaction mixture was irradiated with a UV lamp for 4 h.
The solvents were removed in vacuo and the resulting
yellow solution was purified by column chromatography
/
/
/
/
/
/
4
WINGX-97. Structure was refined using full-matrix
least-squares on F2 with SHELXL-97.
(kieselgelÁhexane) to give a yellow oil (yield 0.355 g,
/
28.4%) which was found to be a roughly 8:1 mixture of 3
and 5.
5.5. Synthesis of [Mo(CO)3h5-P2SC2Bu2t ] (8)
5.3. Reaction of [Fe2(C5H4Me)2(C5H3MeP(S)S)2]
CBut
Similarly [Mo(C7H8)(CO)3] (0.15 g, 0.54 mmol) was
added as a solid to a solution of P2SC2But2 3 (0.1 g, 0.43
mmol) in THF (8 ml) and the resulting mixture stirred
for 48 h at room temperature (r.t.). Removal of the
solvent in vacuo produced an oily red residue which was
with Pꢀ
/
Pꢀ
/
CBut (0.048 g, 0.485 mmol) was added to a
solution of the ferrocene [(Fe2(C5H4Me)2(C5H3-
MeP(S)S)2] (0.1 g, 0.16 mmol) in toluene (5 ml) and
the dark brown reaction mixture was stirred for 48 h.
The solvents were removed in vacuo and the resulting
brown solution was purified by column chromatography
purified by column chromatography (kieselgelÁhexane)
/
to give
a
red solid which was identified as
[{Mo(CO)3}h5-P2SC2Bu2t ] (7) (yield 0.121 g, 68.4%).
31P{1H}-NMR data (d6-benzene, 295 K): d 79.2 (d,
(kieselgelÁhexane) to give a yellow oil (yield 0.011 g,
/
2
2
P(A), JP(A)P(B) 32.6 Hz) 48.5 (d, P(B), JP(A)P(B) 32.1
Hz) ppm. 1H-NMR data (d6-benzene, 295 K): d 1.42 (d,
29.3%) which was found to be a 6:1 mixture of 3 and 5.
4
4
9H, C(CH3)3 J(HP) 1.6 Hz) 1.2 (d, 9H, C(CH3)3 J(HP)
1.1 Hz) ppm.
5.4. Synthesis of [W(CO)5h1-P2SC2Bu2t ] (7)
13C{1H}-NMR data (d6-benzene, 295 K): d 183.4 [d,
1C, PCP, 1J(PC) 100.3 Hz], 159.2 [d, 1C, PCS, 1J(PC) 94.3
A solution of [W(CO)5THF] (0.35 g, 0.86 mmol) in
THF (5 ml) was added to a solution of P2SC2But2 3 (0.1
g, 0.43 mmol) (containing about 10% of the isomeric 5)
in THF (5 ml) and the yellow resulting mixture stirred
for 48 h. Removal of the solvent in vacuo produced an
oily yellow residue which was purified by column
¯
¯
2
Hz], 32.6 [d, 1C, PCC(CH3)3P, J(PC) 15.2 Hz], 38.1 [d,
¯
2
1C, PCC(CH3)3S, J(PC) 12.9 Hz], 36.9 [dd, 3C,
¯
3
PCC(CH3)3P, J(PC) 7.6 Hz, J(PC) 7.1 Hz], 30.2 [d, 3C,
3
¯
3
PCC(CH3)3S, J(PC) 11.3 Hz] ppm.
¯
EIMS m/z (%): 412 (30) [M], 384 (32) [Mꢂ
/CO], 356
chromatography (kieselgelÁhexane) to give a yellow
/
(38) [Mꢂ
/
2(CO)], 328 (51) [Mꢂ
/
3(CO)], 232 (100) [Mꢂ
/
solid identified as [W(CO)5h1-P2SC2Bu2t ] (yield 0.146
g, 61.3%). Found: C, 32.3; H, 3.2; C15H18P2SO5W
requires: C, 32.37; H, 3.23%. Crystals suitable for an
X-ray diffraction study were grown from a cooled and
3(CO)ꢂMo].
/
5.6. Synthesis of [W(CO)3h5-P2SC2Bu2t ] (9)
concentrated hexane solution (ꢂ
/
38 8C) 31P{1H}-NMR
2
data (d6-benzene, 295 K): d 259.2 (d, P(A), J(P(A)P(B))
62.5 Hz, 1JP(A)W 250.5 Hz) 210.1 (d, P(B), 2JP(A)P(B) 62.4
Hz) ppm.
[W(CO)5(h1-P2SC2Bu2t )] (0.15 g, 0.26 mmol) was
heated in THF (10 ml) at 66 8C for 2 h. Removal of
the solvent in vacuo produced an oily yellow residue
which was purified by column chromatography
1H-NMR data (d6-benzene, 295 K): d 1.53 (d, 9H,
4
4
C(CH3)3 JHP 1.7 Hz) 1.1 (d, 9H, C(CH3)3 JHP 1.3 Hz)
ppm.
(kieselgelÁhexane) to give a yellow solid which was
/
identified as [W(CO)3h5-P2SC2Bu2t ] (yield 0.073 g,
54.2%).
13C{1H}-NMR data (d6-benzene, 295 K): d 222.4 [d,
1C, PCP, 1J(PC) 79 Hz], 216.2 [d, 1C, PCS, 1J(PC) 76 Hz],
31P{1H}-NMR data (d6-benzene, 295 K): d 51.7 (d,
¯
¯
2
41.6 [d, 1C, PCC(CH3)3P, J(PC) 18.2 Hz], 40.1 [d, 1C,
2
P(A), JP(A)P(B) 33.2 Hz) 34.3 (d, P(B), JP(A)P(B) 33.2
2
¯
PCC(CH3)3S, 2J(PC) 17.9 Hz], 36.9 [dd, 3C,
Hz) ppm.
¯

14
S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á15
/
1H-NMR data (d6-benzene, 295 K): d 1.33 (d, 9H,
Acknowledgements
4
C(CH3)3 J(HP) 1.8 Hz). 1.09 (d, 9H, C(CH3)3 J(HP) 1.2
4
Hz) ppm.
L. N. and J.F. N. wish to thank the Royal Society and
FKFP-0029/2000 for financial support for part of this
work.
EIMS m/z (%): 500 (31) [M], 472 (42) [Mꢂ
(24) [Mꢂ2(CO)], 416 (40) [Mꢂ3(CO)], 232 (100) [Mꢂ
3(CO)ꢂW].
/
CO], 444
/
/
/
/
References
5.7. Synthesis of P4SC4But4 (13)
[1] R. Appel, M. Moors, Angew. Chem. Int. Ed. Engl. 25 (1986) 567.
[2] G. Markl, W. Holzl, A. Kallmunzer, Tetrahedron Lett. 33 (1992)
4421.
¨
¨
¨
P2SC2But2 3 (0.2 g, 0.86 mmol) (containing ca. 10% of
5) was dissolved in toluene (10 ml) and to this was added
[3] E. Lindner, C. Haase, H.A. Mayer, M. Kemmler, R. Fawzi, M.
Steimann, Angew. Chem. Int. Ed. Engl. 32 (1993) 1424.
[4] E. Lindner, Adv. Heterocycl. Chem. 39 (1986) 237.
[5] E. Lindner, T. Schlenker, C. Haase, J. Organomet. Chem. 464
(1989) C31.
Pꢀ
/
CBut (0.17 g, 1.7 mmol). The pale yellow reaction
mixture was stirred and heated to 80 8C for 8 h, during
which time the solution turned deep yellow in colour.
The solvents were then removed in vacuo and the
remaining yellow product was purified by column
[6] E. Lindner, E. Bosch, C. Maichle-Mossmer, U. Abram, J.
¨
Organomet. Chem. 524 (1996) 67.
chromatography (kieselgelÁ
/
hexane) to give a yellow
[7] M. Regitz, in: M. Regitzand, O.J. Scherer (Eds.), Multiple Bonds
and Low Coordination in Phosphorus Chemistry, G. Thieme,
Stuttgart, 1990.
solid, which was identified as P4SC4But4 (yield 0.35 g).
Found: C, 55.6; H, 8.1; C20H36P4S requires: C, 55.55; H,
8.13%. Crystals suitable for an X-ray diffraction study
[8] K.B. Dillon, F. Mathey, J.F. Nixon, Phosphorus: The Carbon
Copy, Wiley, Chichester, 1998.
were grown from a cooled hexane solution (ꢂ38 8C).
/
[9] (a) M. Regitz, P. Binger, Angew. Chem. Int. Ed. Engl. 27 (1988)
1485;
(b) M. Regitz, Chem. Rev. 90 (1990) 191.
31P{1H}-NMR data (d6-benzene, 295 K): d 406.2
(ddd, P(4), 1JP(4)P(2) 286.3 Hz, 2JP(4)P(3) 18.1 Hz, 3JP(4)P(1)
1
6.1 Hz), 125.2 (ddd, P(2), JP(2)P(4) 285.3 Hz, JP(2)P(3)
2
[10] B. Burghardt, S. Krill, Y. Okano, W. Ando, M. Regitz, Synlett
(1991) 5356.
2
15.2 Hz, JP(2)P(1) 17.4 Hz), 116.4 (ddd, P(3), JP(3)P(1)
2
[11] S.E. d’Arbeloff-Wilson, P.B. Hitchcock, S. Krill, J.F. Nixon, L.
Nyula`szi, M. Regitz, J. Am. Chem. Soc. 122 (2000) 4557.
[12] S.G. Ruf, unpublished results, Ph.D. Thesis, Universitat Kaiser-
slautern, 2000.
2
2
16.3 Hz, JP(3)P(4) 16.1 Hz, JP(3)P(2) 15.2 Hz), ꢂ98.5 (t,
/
P(1), JP(1)P(2)ꢀ2
/
JP(1)P(3)
ꢀ17.1 Hz) ppm.
/
2
1H-NMR data (d6-benzene, 295 K): d 1.57 (s, 9H,
C(CH3)3), 1.34 (s, 9H, C(CH3)3), 1.32 (d, 9H, C(CH3)3
4J(HP) 2.1 Hz) 1.27 (s, 9H, C(CH3)3) ppm.
[13] J. Grobe, D. le Van, T. Pohlmeyer, F. Immel, H. Pucknat, B.
Krebs, J. Kuchinke, M. Lage, Tetrahedron 56 (2000) 27.
[14] J. Dietz, T. Schmidt, J. Renner, U. Bergstrasser, F. Tabellion, F.
¨
EIMS m/z (%): 432 (100) [M], 417 (12) [Mꢂ
(32) [MꢂCH3ꢂ
P2C2But2CC2H6] 200 (35) [[Mꢂ
P2C2But2CC2H6ꢂ
S].
Crystal data for 13: C20H36P4S, Mꢀ
clinic P21/c (No. 14), aꢀ16.417(9), bꢀ
118.08(2)8, Vꢀ2250.8(14) A , Tꢀ
4, Dcalc 1.28 mg mꢂ3, mꢀ0.43 mmꢂ1
0.71073 A, F(000) 928, crystal size 0.2ꢃ0.2ꢃ0.05
mm3 4090 measured reflections, 3950 independent
reflections (Rint 0.029), 3160 reflections with I ꢀ
2s(I), Final indices R1ꢀ0.040, wR2ꢀ0.090 for I ꢀ
2s(I), R1ꢀ0.057, wR2ꢀ0.092 for all data, Data
collection: EnrafÁNonius CAD Program package
/
CH3], 231
Preuss, P. Binger, H. Heydt, M. Regitz, Eur. J. Org. Chem. (2002)
1664; H. Heydt, personal communication to JFN.
[15] S.E. d’Arbeloff-Wilson, P.B. Hitchcock, J.F. Nixon, manuscript
in preparation.
/
/
/
CH3ꢂ
/
/
/
432.43, mono-
9.762(3), cꢀ
[16] N. Dubau-Assibat, A. Baceiredo, G. Bertrand, J. Org. Chem. 60
(1995) 3904.
[17] S.F.M. Asmus, U. Bergstrasser, M. Regitz, Synthesis 9 (1999)
¨
/
/
/
3
˚
˚
15.917(4) A, bꢀ
173(2) K, Zꢀ
/
/
/
/
/
,
1642.
˚
[18] M.D. Francis, D.E. Hibbs, P.B. Hitchcock, M.B. Hursthouse, C.
Jones, T. Mackewitz, J.F. Nixon, L. Nyulaszi, M. Regitz, N.
Sakarya, J. Organomet. Chem. 580 (1999) 156.
lꢀ
/
/
/
ꢀ
/
/
[19] V. Caliman, P.B. Hitchcock, J.F. Nixon, J. Chem. Soc. Chem.
Commun. (1995) 1661.
/
/
/
/
/
[20] P.B. Hitchcock, J.F. Nixon, N. Sakarya, J. Chem. Soc. Chem.
Commun. (1996) 2751.
/
4
[21] V. Caliman, P.B. Hitchcock, J.F. Nixon, L. Nyulaszi, N. Sakarya,
J. Chem. Soc. Chem. Commun. (1997) 1305.
WINGX-97. Structure was solved by direct methods and
refined using full-matrix least-squares on F2 with
SHELXL-97.
[22] G.K.B. Clentsmith, P.B. Hitchcock, J.F. Nixon, N. Sakarya, J.
Organomet. Chem. 584 (1999) 58.
[23] Calculations were carried out by the GAUSSIAN 98 program
package. (GAUSSIAN 98, Revision A.5, M.J. Frisch, G.W. Trucks,
H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G.
Zakrzewski, J.A. Montgomery, Jr., R.E. Stratmann, J.C. Burant,
S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A.
Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V.
Ortiz, B.B. Stefanov, G. Liu, Liashenko, P. Piskorz, I. Komar-
6. Supplementary material
Crystallographic data for the structural analysis has
been deposited with the Cambridge Crystallographic
Data Centre.

S.E. d’Arbeloff-Wilson et al. / Journal of Organometallic Chemistry 655 (2002) 7Á
/15
15
omi, R. Gomperts, R.L. Martin, J. Fox, T. Keith, M.A. Al-
Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challa-
combe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L.
Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A.
Pople, Gaussian, Inc., Pittsburgh, PA, 1998) at the B3LYP/6-
therefore, they are unlikely to change the qualitative conclusions.
The structure of 7 and also that of the uncomplexed 3,5-di-tBu-
1,2,4-thiadiphosphole was calculated at the B3LYP level using the
LANL2DZ pseudopotential with polarisation functions on the
heavy atoms as implemented in the GAUSSIAN 98 package.
[24] F.G.N. Cloke, P.B. Hitchcock, J.F. Nixon, D.J. Wilson, F.
Tabellion, U. Fishbeck, F. Preuss, M. Regitz, L. Nyulaszi,
J.C.S. Chem. Commun. (1999) 2363.
311ꢁG* level of the density functional theory. Preliminary
/
calculations were also carried out at the B3LYP/3-21G* level,
resulting in minor changes between relative energies and struc-
tures (see below). Optimisations were followed by the calculation
of the second derivatives which proved that real minima were
obtained. (It should be noted that the B3LYP/3-21G* optimised
minimum of 10 and 11 has turned out to be a first order saddle
point with respect to the rotation of the Cr(CO)5 group. The real
minima were just slightly lower (0.1 kcal molꢂ1) in energy in
[25] F. Mathey, Chem. Rev. 90 (1990) 997.
[26] S.G. Ruf, U. Bergstrasser, M. Regitz, Eur. J. Org. Chem. (2000)
¨
2225.
[27] A. Mack, U. Bergstrasser, G.J. Reiss, M. Regitz, Eur. J. Org.
¨
Chem. (1999) 587.
[28] M. Regitz, S. Krill, Phosphorus Sulfur Silicon 115 (1996) 99.
[29] G. Becker, G. Gresser, W. Uhl, Inorg. Synth. 27 (1990) 249.
[30] E.W. Abel, M.A. Bennett, R. Burton, G. Wilkinson, J. Chem.
Soc. (1958) 4559.
agreement with the low imaginary frequencies (10Á
20 cmꢂ1) of
/
the transition structures. The calculation of the thermodynamic
data were uncorrected for the internal rotations, such effects are
considered to be similar for the structures investigated here,