(
)
M.P. Murphy et al.rJournal of Organometallic Chemistry 550 1998 151–164
159
this is observed quite commonly in borane-containing
compounds of palladium, platinum, rhodium and similar
elements when the transition-element has two exo-clus-
ter ligands. Hence, the common approach to electron
counting in these compounds often needs to be modified
to be in accord with the observed cluster geometry.
However, the alternative explanation of the unexpected
structurerelectron-count of compound 4 which is based
on the suggestion that two extra cluster electrons may
be provided from two C–H . . . Rh agostic interactions
sulphur Õs. boron binding. Thus, choice of metal unit
and reaction conditions will permit some structural
‘tailoring’ of these types of systems in the future.
3. Experimental
General methods for the syntheses and spectroscopic
characterisation of compounds have been described in
w x
previous Parts in this series 1 . The compound
w
x
10,11 is worth consideration. The agostic interactions
could possibly arise from the triphenylphosphine CH
arachno-4,6-S2 B H was prepared according to the
7 w 9 x
w
Ž
.3x
literature method 23 . The complexes, RhCl PPh3
w
x
w
x w
Ž
.
x w
x
w
Ä
Ž
.
424
w
x
units if they were located in suitable positions 10,11 .
The two shortest C–H . . . Rh distances in the three
24 , Pt PPh3
25 and PtCl2 POMe3
26 were
w
4
synthesized as described in the literature from RhCl3 P
w
x
w
x
crystallographically independent molecules of 8-dppe-
xH2O and K2 PtCl4 , respectively. NMR chemical
shifts d B , d P and d H are given in ppm to
Ž11
.
Ž31
.
Ž1
.
x
w
x
8,7-RhSB9 H10 5a 10,11 which were observed in two
dichloromethane solvates of 5a, i.e. 5a, 2CH2Cl2 and
Ž
.
high frequency low field of BF3OEt2 , 85% H3PO4
and SiMe4, respectively.
˚
.
Ž
.
Ž
5a. 1r2 CH2Cl2 were 3.254 11 and 2.742 12 A,
˚
.
˚
Ž
.
Ž
3.119 14 and 2.849 10 A, and 3.174 and 2.809 A,
[
(
)
]
respectively. In compound 4, the two closest C–H . . .
3.1. 9,9- PPh3 2-9,6,8-PtS2 B7 H7 1a
˚
Ž
.
Rh 10 distances are 3.189 and 3.394 A and involve
Ž
.
Ž
.
w
Ž
.
4 x
Ž
.
H 336 and H 316 , respectively. However, for the
A solution of Pt PPh3
0.166 g, 0.134 mmol in
3
Ž
.
rhodium atom with higher coordination, the related
dichloromethane 15 cm was added to a solution of
Ž .
Ž
.
C–H . . . Rh 8 distances fall between the C–H
arachno-4,6-S2 B7 H9 0.018 g, 0.134 mmol in
3
˚
Ž
.
Ž
.
. . . Rh 10 distances at 3.251 and 3.260 A and involve
dichloromethane 10 cm . The yellow reaction mixture
was stirred at room temperature for 72 h. The solution
was concentrated under reduced pressure and subjected
Ž
.
Ž
.
H 112 and H 226 , respectively. It seems unlikely that
both rhodium atoms have gained two extra electrons
from two, one-electron, agostic C–H . . . Rh bonds. If
this were so, it would lead to an arachno electron
count. Furthermore, it is noteworthy that some of the
B–H . . . Rh distances are shorter than the C–H . . . Rh
ones above. The two shortest B–H . . . Rh distances to
Ž
. Ž
.
to preparative tlc CH2Cl2–heptane 8:2 . The single
major yellow band was extracted into dichloromethane.
Recrystallisation from CH2Cl2–hexane afforded pale
yellow block crystals of 9,9- PPh3 2-9,6,8-PtS2 B7 H7
1a 0.062 g, 53.1% . Found: C, 49.9; H, 4.3.
w
Ž
.
x
Ž
.
Ž
˚
Ž
.
Ž .
Ž .
Ž
.
Ž .
.
Rh 10 are 2.77 4 and 2.83 4 A to H 11 and H 6 ,
respectively, and those to Rh 8 are 2.91 4 and 3.02 4
A to H 3 and H 4 respectively. A similar observation
about the proximity of B–H . . . Rh atoms has been made
C36 H37 B7 P2 PtS2 requires C, 49.85; H, 4.65% . IR:
nmax KBr 2555 m , 2530 s , 2490 s BH cm
.
11B
y1
Ž . Ž . Ž .
Ž
.
Ž .
Ž .
Ž . Ž
.
1
˚
Ž .
Ž .
Ž
.
and H NMR data CD2Cl2 230 K ordered as assign-
ment d B d H of directly attached proton : BH 4
Ž11 Ž1
.
w
.
x
Ž .
w
w
x
Ž .
w
x
Ž
.
elsewhere in discussion of the structure of m-9,10-
q21.1 q3.09 , BH 5 q0.2 q2.94 , BH 10 y3.0
4
Ž
.
Ž
.
x w
x
w
x
Ž . Ž .
w
x
w
x
SMe -8- h -C5Me5H -nido-8,7-RhSB9 H9 22 . At this
q1.98 , BH 7 y6.0 q2.72 , BH 1 y18.7 q1.42 ,
Ž . Ž .
w
x
w
x
stage, we are unconvinced of the agostic bonding expla-
nation of the structures of 4, 5 and 5a and we intend to
address this problem in more detail in later publications.
The origins of the M–S–M–S sequence in 4 are of
interest. The formation of compound 4 must involve a
cluster skeletal rearrangement which may be related to
the 2 to 3 conversion mentioned above. Thus the reac-
BH 2 y23.4 q1.94 , BH 3 y33.8 q1.18 ; and at
Ž11
.
297 K ordered as assignment d B , multiplicity,
w
w
w
w
w
Ž1
.
x
Ž
.
d H of directly attached proton : BH 4,5 q10.2, 2,
x
Ž
.
w
x
Ž .
q3.13 , BH 10 y2.5, 1, q2.16 , BH 7 y3.6, 1,
x
Ž .
w
P
x
Ž
.
q2.82 , BH 1 y16.3, 1, q1.56 , BH 2,3 y27.5, 2,
x
Ž31 . Ž
CD2Cl2 , 294 K q33.1
1
.
31
q1.65 . Additionally, d
Ž195
.
x
w
Ž195
.
Pt– P 2695
1
31
J
Pt– P 4284 Hz and q25.4
J
2
31
x
Ž31
.
Ž
Ž
tion possibly involves the initial formation of 2 sche-
Hz with J P– P 2112 Hz. For X-ray Diffraction
.
.
matic structure IV followed by a vertex ‘flip’ to yield
analysis of 1a, see Section 3.6 and Table 3 .
the desired precursor geometry V, i.e. that proposed for
compound 3, before the addition of the second rhodium
unit. Notwithstanding the details of the intimate
mechanism, it is apparent that with these types of metal
centres there is a clear driving force for facile metal–
sulphur grouping and thence M2 S2 subcluster self-
assembly on the boron hydride framework. It is likely
that this depends critically on the particular metal unit
being added and the relative affinities of metal units for
[
{
(
)
}
]
3.2. 9,9- P OMe 3 2-9,6,8-PtS2 B7 H7 1c
Ž
A solution of arachno-4,6-S2 B7 H9 0.024 g, 0.160
mmol and tetramethylnaphthalenediamine tmnd; 0.092
g, 0.430 mmol in dichloromethane 20 cm was stirred
for 10 min and then PtCl2 POMe3
mmol added. Stirring was continued for 24 h. The
mixture was then filtered through silica tlc-grade, ca. 3
.
Ž
3
Ä
4
.
Ž
.
.
w
Ä
Ž
42
x
Ž
0.100 g, 0.195
.
Ž
Murphy, Michael P.
