202
M. Rodriguez i Zubiri et al. / Inorganic Chemistry Communications 7 (2004) 201–203
we also have an interest in the ease of distortion of
planar aromatic systems [8].
and dichloromethane and air- and moisture-stable for
short periods in the solid state but oxidises readily in
solution in air.
H
N
N
1 was oxidised to the P(V) diselenide 2 by stirring at
reflux in toluene with selenium. Alternatively, reaction of
1 with PtCl2(cod) gives cis-[PtCl2{Ph2PN(C6H4) (C6H4)
NPPh2}] (3). The 31P–{1H} NMR spectrum of 3 is a
singlet with satellites due to coupling to 195Pt (d(P) 43.3
ppm, 1J{195Pt–31P} 3855 Hz); it has the expected parent-
ion (m/z 818,816) and fragmentation patterns consistent
with the loss of one chloride [m/z 781 [M–Clꢀ]). Ele-
mental analysis data are in agreement with calculated
values and the IR spectrum shows bands due to m(P–N)
(949 cmꢀ) and m(PtCl) (311 cmꢀ1).
Na2S2O4/H2O
O2
ð1Þ
N
N
H
Phenazine
5,10-Dihydrophenazine
5,10-Dihydrophenazine (Eq. 1) was reacted with two
equivalents of Ph2PCl, in the presence of n-BuLi at
4
)78 °C, in thf to give 1 in 91% yield (Eq. 2).
PPh2
H
N
N
thf, -78oC
Ph2PCl
+
n-BuLi
The molecular structure of 3 confirms that the ligand
is bidentate and in a cis geometry and that the chloride
ligands are mutually cis. The complex contains two
seven-membered Pt–P–N–C–C–N–P rings and the metal
adopts the expected square planar geometry. The
square-planar platinum centre is distorted and the bite
angle of the ligand is larger than the ideal 90° with P(1)–
Pt(1)–P(6) 99.90(3)° whilst Cl–Pt–Cl is 90.37(3)°. The
backbone of the ligand forms an Ôumbrella-likeÕ struc-
ture about the platinum centre with the C12N2 tricyclic
ring adopting a butterfly geometry with the aryl rings
being inclined by 122° to each other and the N(2) and
N
N
H
PPh2
ð2Þ
The 31P–{1H}NMR spectrum of 1 consists of a sin-
glet at d(P) 61.0 ppm – the large downfield shift when
compared to its P–C relative [9,10] (PAnP, d(P) ¼ )21.9
ppm) reflecting the substitution of the carbon centres in
PAnP by nitrogen atoms in 1 (it should be noted that
PanP is the aromatic analogue of 1 there is no direct
carbon analogue).
ꢀ
N(5) atoms lying 0.45 and 0.50 A above the C(3)–C(4)–
PPh2
C(11)–C(16) mean plane, and C–N–C angles of ca. 109°
as appropriate for sp3 hybridised nitrogen atoms.
The P–N bond lengths in 3 are at the upper end of the
range we have observed in P–N containing phosphines
[11].
(PAnP)
PPh2
FABþ mass spectrometry confirms the proposed
identity of the ligand 1 with the expected parent-ion
peaks (m/z 550 [M], 551 [M + Hþ] and 573 [M + Naþ])
whilst its IR spectrum shows a band that can be assigned
to m(P–N) (951 cmꢀ1) and elemental analysis data are in
agreement with calculated values. 1 is soluble in both thf
We tested the usefulness of 1 for the catalytic hydro-
formylation of oct-1-ene using a batch autoclave and
similar conditions to previously [6] – it proved to be fairly
active with a similar initial rate of reactivity to the ho-
mopiperazine ligand Ph2PN(C3H6)(C2H4)NPPh2 (4)
(2:45 ꢁ 10ꢀ3 of CO uptake in 1 versus 2:98 ꢁ 10ꢀ3 mol
dmꢀ3 in the piperazine) and a rather better lin-
ear:branched ratio of the aldehyde produce (2.5:1 in 1
versus 1.7:1 in 4) which suggest that 1 may have some
potential in catalytic applications – and certainly that it is
robust enough to behave as a supporting ligand under
hydroformylation conditions. According to simple mo-
lecular modelling studies although 1 and 4 have different
natural bite angles [ca. 90° and 110°, respectively) they
have similar flexibility ranges in their bite angles and
this may be the most significant feature in relating their
similarity in catalytic behaviour. Further work is in
progress.
4
Catalytic studies (1) The ligand (0.03 mmol, 0.005 mol dmꢀ3),
[Rh(CO)2(acac)] (5.16 ꢁ 10ꢀ3 g, 0.02 mmol, 0.004 mol dmꢀ3) and
toluene (4 ml) were charged into a Schlenk tube and stirred until
complexation of the rhodium complex with the phosphine species
(total dissolution). The phosphine/rhodium ratio used was 3/2. This
solution was placed in a glass lined steel autoclave, CO/H2 added until
20 bar pressure, the autoclave was heated to 100 °C for the appropriate
reaction time and stirred magnetically.
Single crystal X-ray diffraction studies on crystals were performed
using a Bruker SMART diffractometer with graphite-monochromated
ꢀ
Mo Ka radiation (k ¼ 0:71073 A). The structure was solved by direct
methods, the non-hydrogen atoms were refined with anisotropic dis-
placement parameters; hydrogen atoms were fixed. Structural refine-
ments were by the full-matrix least-squares method on F2 using
SHELXTL [12].
(3) C36H28N2P2Cl2Pt, M ¼ 816:5, monoclinic, a ¼ 10:7076ð17Þ, b ¼
Supplementary data
ꢀ
ꢀ3
15:392ð3Þ, c ¼ 19:078ð3Þ A, b ¼ 92:036ð3Þ°. U ¼ 1138 A , T ¼ 125 K,
space group P2ð1Þ=n, Z ¼ 4, l(Mo Ka) ¼ 4.77 mmꢀ1. Of 13,327 mea-
sured data, 4479 were unique (Rint ¼ 0:0308), to give R1 ½I > 2rðIÞꢂ ¼
0:0201.
Crystallographic data for the structural analyses has
been deposited with the Cambridge Crystallographic