Organometallic complexes of platinum-group metals incorporating substituted guanidine dianion (triazatrimethylenemethane) ligands

Article abstract of DOI:10.1016/S0022-328X(97)00654-2

Reactions of the platinum-group metal halide complexes [PtCl₂(COD)] (COD=1,5-cyclo-octadiene), [Cp*RhCl₂(PPh₃)] (Cp=η⁵-C₅Me₅), [Cp*IrCl₂(PPh₃)], [(p-cymene)RuCl₂(PPh₃)] and [(p-cymene)OsCl₂(PPh₃)] with symmetrically trisubstituted (acetyl or phenyl) guanidines, mediated by silver(I) oxide, give complexes formally containing the triazatrimethylenemethane ligand. A full X-ray crystal structure determination is reported for the N,N′,N″-triphenylguanidine dianion complex [Pt{NPhC(=NPh)NPh}(COD)] 4a which shows the presence of a planar Pt-NR-C(=NR)-NR four-membered platinacycle. At room temperature (r.t.), the ¹H- and ¹³C{¹H}-NMR spectra of 4a yield a single set of COD CH and CH₂ resonances. At 240 K however, two sets of resonances are observed, interpreted in terms of fluxionality of the C=N-Ph moiety. Attempted synthesis of the analogous platinum triacetylguanidine complex yields the new ureylene complex [Pt{NAcC(=O)NAc}(COD)], via a hydrolysis reaction. Starting with the osmium compound, both the guanidine complex [(p-cymene)Os{NAcC(=NAc)NAc}(PPh₃)] 9 and the ureylene complex [(p-cymene)Os{NAcC(=O)NAc}(PPh₃)] 10 were formed; similar results were obtained for the iridium system.

Full text of DOI:10.1016/S0022-328X(97)00654-2

Journal of Organometallic Chemistry 556 (1998) 75–88
Organometallic complexes of platinum-group metals incorporating
substituted guanidine dianion (triazatrimethylenemethane) ligands
Maarten B. Dinger, William Henderson *, Brian K. Nicholson
Department of Chemistry, Uni6ersity of Waikato, Pri6ate Bag 3105, Hamilton, New Zealand
Received 26 August 1997
Abstract
Reactions of the platinum-group metal halide complexes [PtCl2(COD)] (COD=1,5-cyclo-octadiene), [Cp*RhCl2(PPh3)] (Cp*=
p5-C5Me5), [Cp*IrCl2(PPh3)], [(p-cymene)RuCl2(PPh3)] and [(p-cymene)OsCl2(PPh3)] with symmetrically trisubstituted (acetyl or
phenyl) guanidines, mediated by silver(I) oxide, give complexes formally containing the triazatrimethylenemethane ligand. A full
X-ray
crystal
structure
determination
is
reported
for
the
N,N%,N¦-triphenylguanidine
dianion
complex
¸¹¹¹¹¹¹¹¹¹º
[Pt{NPhC(ꢀNPh)NPh}(COD)] 4a which shows the presence of a planar Pt-NR-C(ꢀNR)-NR four-membered platinacycle. At
1
room temperature (r.t.), the H- and 13C{1H}-NMR spectra of 4a yield a single set of COD CH and CH2 resonances. At 240 K
however, two sets of resonances are observed, interpreted in terms of fluxionality of the CꢀN-Ph moiety. Attempted synthesis of
¸¹¹¹¹¹¹¹º
the analogous platinum triacetylguanidine complex yields the new ureylene complex [Pt{NAcC(ꢀO)NAc}(COD)], via a hydrolysis
¸¹¹¹¹¹¹¹¹¹º
reaction. Starting with the osmium compound, both the guanidine complex [(p−cymene)Os{NAcC(ꢀNAc)NAc}(PPh3)] 9 and the
¸¹¹¹¹¹¹¹º
ureylene complex [(p−cymene)Os{NAcC(ꢀO)NAc}(PPh3)] 10 were formed; similar results were obtained for the iridium system.
© 1998 Elsevier Science S.A. All rights reserved.
Keywords: Guanidine; Triazatrimethylenemethane; Metallacycle; Silver oxide; Platinum group metal
[1]. Bailey et al. have recently reported some ruthenium
[3] and dimolybdenum [4] complexes of the N,N%,N¦-
triphenylguanidine monoanion. However, transition-
metal complexes containing the guanidine dianion
ligand [C(NH)3]2−, or substituted derivatives, are much
less common. To the best of our knowledge only one
transition-metal complex containing a guanidine dian-
ion has been previously reported, the dinuclear iron
carbonyl complex 2, formed by the reaction of dicyclo-
1. Introduction
Guanidines, with the general structure 1, are versatile
ligands and are capable of bonding to metal centres in
a variety of coordination modes, the most common
being as neutral donor ligands [1,2] or as monoanions
* Corresponding
author.
Fax:
+64-7838-4219;
e-mail:
w.henderson@waikato.ac.nz.
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(97)00654-2

76
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
hexylcarbodiimide and Fe(CO)5 ([5]a); a very recent
Sb example also contains formally dianionic
a
guanidine ligand ([5]b). The guanidine dianion ligand
is of interest, since it is formally isoelectronic with the
trimethylenemethane (TMM) ligand [C(CH2)3]2−, for
which there are numerous reports of complexes [6,7].
Complexes of monoazatrimethylenemethane have re-
cently been described [8]. The recent synthesis of the
dilithio-salt of N,N%,N¦-triphenylguanidine [9] and its
reaction to form a heterobimetallic cadmium–lithium
complex [10] suggests that a diverse range of metal
complexes containing guanidine dianion ligands
should be accessible from this reagent. Reactions of
1,3-dianions with metal halide complexes have been
shown to be versatile, general synthetic routes to a
wide range of metallacyclic complexes [11]. Complexes
studies into the synthesis of organometallic platinu-
m(II), rhodium(III), iridium(III), ruthenium(II) and os-
mium(II) complexes containing triphenyl- and
triacetyl-guanidine dianion ligands.
containing
oxodimethylenemethane
[12]
and
trimethylenemethane [7] ligands have also been pre-
pared using such types of reagent.
2. Results and discussion
We have adopted a complementary strategy for the
synthesis of platinum group metal complexes of
guanidine dianion ligands, using the reagent silver(I)
oxide. This reagent simultaneously acts as a halide-
abstracting reagent and a base, and a variety of com-
plexes containing metal–oxygen, –nitrogen, –carbon
and –sulfur bonds have been synthesised under mild
reaction conditions [13]. We have recently reported
the use of silver(I) oxide in the synthesis of a range
of platinum(II) ureylene complexes 3 [14] which are
2.1. Syntheses
The ready availability of N,N%,N¦-triphenyl- and
N,N%,N¦-triacetylguanidines suggested that their reac-
tion with platinum-group metal–dihalide complexes
might form a general synthetic route into the synthesis
of guanidine dianion (triazatrimethylenemethane) com-
plexes. Thus, the complex [PtCl2(COD)] readily reacts
with N,N%,N¦-triphenylguanidine in refluxing dichloro-
methane in the presence of excess silver(I) oxide to give
the guanidine dianion complex 4a in 94% yield [17]. In
marked contrast however, the attempted synthesis of
the analogous N,N%,N¦-triacetylguanidine platinum
complex 4b was unsuccessful. In this case, the N,N%-di-
acetylureylene complex [¸Pt¹{¹N¹A¹cC¹¹(ꢀ¹O¹)NºAc}(COD)] 5
derived from urea dianion ligands [RNC(O)NR]2−
,
formally isoelectronic with the guanidine dianion
[RNC(NR)NR]2−. Complexes of the urea-derived lig-
and are more widely reported in the literature [14–
16].
The use of silver(I) oxide as a general synthetic
reagent is limited to moisture-stable complexes (in par-
ticular those of the platinum group metals described
herein), since water is a by-product of the reaction. Its
advantage lies in the ease of product isolation. Using
the silver oxide method, we recently reported the syn-
thesis of the first mononuclear complex of a guanidine
dianion, this being the cyclo-octadiene platinum com-
plex 4a [17]. In this paper, we report our detailed
was isolated and characterised in 87% yield as a white
crystalline solid. The silver(I) oxide-mediated reaction
appears to be only applicable to the synthesis of plat-
inum complexes of phenyl-substituted guanidines, since
the attempted synthesis of analogous platinum com-
plexes with alkyl-substituted guanidines led either to
significant decomposition of the starting [PtCl2(COD)]
(with methyldiphenylguanidine and 1-adamantyl-
diphenylguanidine) or to recovery of unreacted

M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
77
[PtCl2(COD)], in the case of the reaction with
trimethylguanidine.
guanidine mediated by silver(I) oxide gives the
monodentate guanidine intermediate X, which can
then react by two different pathways, A and B. Path-
When N,N%,N¦-triacetylguanidine was reacted under
the same conditions (as for the platinum complex 5)
with the metal complexes [Cp*RhCl2(PPh3)] (Cp*=
way
A involves simple cyclisation to give the
guanidine dianion complex. It is suggested that the
greater lability of the second-row metals (Ru and Rh)
over their third-row counterparts results in pathway
A being rapid for these metals, resulting in the isola-
tion of solely the guanidine dianion complexes. For
the metals Pt, Ir and Os however, pathway B be-
comes competitive with pathway A (entirely so for
Pt). These third-row metals are less labile, cyclisation
p5-C5Me5),
[Cp*IrCl2(PPh3)],
and
[(p-cymene)-
RuCl2(PPh3)] the corresponding complexes 6a, 7 and
8, containing the triacetylguanidine dianion ligand
could be isolated. However, the attempted synthesis
of the osmium complex 9 was less successful, with the
31P-NMR of the crude reaction mixture showing two
products in almost equal amounts. These were char-
acterised as being the desired product, [(p−cymene)
¸¹¹¹¹¹¹¹¹¹º
Os{NAcC(ꢀNAc)NAc}(PPh3)]
9 and the ureylene
¸¹¹¹¹¹¹¹¹º
complex, [(p−cymene)Os{NAcC(ꢀO)NAc}(PPh3)] 10.
When the complex [Cp*IrCl2(PPh3)] was reacted
with N,N%,N¦-triacetylguanidine under the same stan-
dard conditions, there was NMR and electrospray
mass spectrometric evidence for the corresponding
ureylene complex 11 being formed. Pure samples of
the osmium and iridium ureylene complexes (10 and
11, respectively) were subsequently obtained by the
silver(I) oxide-mediated reactions of the dihalide com-
plexes with N,N%-diacetylurea in refluxing CH2Cl2.
This subsequently facilitated unambiguous NMR
spectroscopic characterisation of the osmium complex
9, which was not obtained in a pure state.
The formation of the diacetylureylene complexes
clearly arises as a result of hydrolysis of the triacetyl-
guanidine ligand. Additionally, the hydrolysis reaction
occurs to variable extent with the different metal cen-
tres. For example, rhodium gives solely the guanidine
complex 6a, while iridium gives ca. 25% ureylene
complex 11 in the crude product 7. Similarly, a
higher degree of hydrolysis occurs for the osmium
complex 9 when compared to the ruthenium analogue
8. Platinum gives solely the ureylene complex 5. The
relative amounts of 9 and 10 remained constant when
the reaction was carried out by refluxing the triacetyl-
guanidine and Ag2O for 18 h, with and without
added water, before adding a stoichiometric amount
of [(p-cymene)OsCl2(PPh3)]. However, if the reaction
mixture containing [(p-cymene)OsCl2(PPh3)], silver ox-
ide and the guanidine was refluxed for 20 h, a 42:58
distribution of 9:10 was obtained. When the reaction
time was decreased to 2 h, a ca. 50:50 mixture was
obtained. These observations suggests that the free
guanidine and the final guanidine dianion product 9
essentially resist hydrolysis under the reaction condi-
tions, and so conversion to the ureylene ligand re-
quires coordination to the metal in an intermediate
species. A possible mechanism explaining the forma-
tion of differing amounts of ureylene and guanidine
dianion products is shown in Scheme 1. Initial reac-
tion of the metal–dihalide complex with triacetyl-

78
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
Scheme 1. A possible mechanism explaining the formation of differing amounts of ureylene and guanidine dianion products.
by pathway A is slower, and thus metal-promoted
hydrolysis of the coordinated guanidine monoanion
to the coordinated urea monoanion, intermediate Y
in Scheme 1, may occur. The formation of the ure-
ylene complexes then simply proceeds by silver(I)
oxide-mediated cyclisation of intermediate Y.
The preparations of the rhodium and ruthenium
complexes (6 and 8, respectively) had to be carried
out under a nitrogen atmosphere, while the platinum
complexes 4b and 5 could also be prepared success-
fully in air. All the complexes are air- and moisture-
stable once isolated, and are typically bright yellow
to orange in colour. Freshly-prepared solutions of
the ruthenium complex 8 are initially orange in
colour, but rapidly turn dark green upon standing
in air. Since NMR spectra essentially only show res-
onances due to 8, it was concluded that the crude
product is contaminated with a small quantity of a
highly coloured impurity. After recrystallisation by
diffusion of diethyl ether and pentane into
dichloromethane solution, stable orange crystals were
obtained.
The synthesis of the triphenylphosphine complex
4c was also attempted. Reaction of cis-[PtCl2(PPh3)2]
with N,N%,N¦-triphenylguanidine and silver(I) oxide
led to very little product formation after 24 h reac-
tion, while the attempted ligand substitution reaction
of the COD ligand of 4a with PPh3 surprisingly led
to recovery of unreacted starting material.
a
2.1.1. X-ray crystal structure determinations
A single-crystal X-ray analysis was carried out in
order to ascertain the nature of the bonding of the
ligand to the metal centre in the platinum complex 4a.
The molecular structure and atom numbering scheme
are shown in Fig. 1, while Table 1 gives atomic

M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
79
¸¹¹¹¹¹¹¹¹º
Fig. 1. Molecular structure of [Pt{NPhC(ꢀNPh)NPh}(COD)] 4a, showing the atom numbering scheme. The dichloromethane and disordered
diethyl ether of crystallisation have been omitted for clarity.
coordinates for the structure, and Table 2 gives
selected bond lengths and angles.
substituents on nitrogens N(1) and N(3), which
permits a closer approach of N(2) to the platinum.
The orientation of the phenyl substituent on the
imino group presumably precludes a fully symmetrical
binding of the ligand to the platinum centre. The
phenyl substituents are tipped out of the metallacyclic
plane, again presumably to minimise steric
interactions. Thus, carbon atoms C(11), C(21) and
The complex contains an p2-triazatrimethylene-
methane (guanidine dianion) ligand, coordinated to
platinum via two nitrogen atoms, forming an
essentially planar Pt–N–C–N metallacycle. No atom
deviates from the least-squares plane of this
˚
˚
metallacycle by more than 0.054(12) A, for N(2). The
˚
˚
C(31) are 0.84(1) A above, 0.62(1) A below and
C(1)–N(3) bond distance, at 1.30(1) A, is significantly
˚
0.34(1) A below the metallacyclic plane, respectively,
shorter than the C(1)–N(1) and C(1)–N(2) bond
˚
as depicted in the orientation of Fig. 1.
distances, which are both 1.40(1) A. This indicates
The triphenylguanidine dianion complex 4a overall
bears a very strong resemblance to formally isoelec-
tronic carbonato complexes. Thus, a number of platin-
um(II)–phosphine complexes, e.g. [Pt(CO3)(PPh3)2],
the presence of localised single [C(1)–N(1) and
C(1)–N(2)] and double [C(1)–N(3)] bonds in the
complex. By comparison, the average C–N bond
distance to the central carbon in the delocalised
˚
[Pt(CO3)(PMe3)2]·2H2O
and
[Pt(CO3)(Ph2P-
dianion of Li2[C(NPh)3] is 1.36(1) A [9]. The X-ray
CH2CH2CH2PPh2)] also contain essentially planar
structure of N,N%,N¦-triphenylguanidine has also been
¸¹¹¹¹¹¹¹¹¹º
reported [18] and the C–N bond lengths in this
Pt−O−C(O)−O met¸al¹la¹cy¹c¹le¹s¹[1¹9¹].¹Mº etal–ureylene
complexes, containing M−N−C(O)−N ring systems
(of the type 3) also have planar ring systems [14,16]. In
marked contrast, the palladium p3-trimethylen-
˚
structure are equal at ca. 1.34 A. The difference
between the Pt–N(1) and Pt–N(2) bond distances in
˚
4a, 2.034(8) and 2.002(7) A, respectively, can be
attributed to steric interactions between the phenyl
emethane
complex
[Pd{p3-CH2C{C(CO2Me)2}-

80
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
Table 2
CH2}(PPh3)2] (also formally isoelectronic with the
guanidine dianion platinum complex) shows a highly
non-planar metal–ligand arrangement [20].
˚
Selected
bond
lengths
(A)
and
angles
(°)
for
¸¹¹¹¹¹¹¹¹º
[Pt{NPhC(ꢀNPh)NPh}(COD)] 4a, with estimated standard deviations
in parentheses
An X-ray structure determination of the ruthenium
complex 8 was also attempted. A reasonable data set
was obtained and solved, but the refinement was bedev-
iled by extensive disorder of the guanidine dianion
ligand and possibly by unresolved twinning, so that an
R1 factor of only 0.145 was obtained. The analysis can
therefore only be taken to confirm the atom connectiv-
ity and the overall conformation (as shown in Fig. 2)
and so no bond parameters are given or discussed here.
The molecule does contain the expected triacetyl-
guanidine dianion ligand attached to the ruthenium via
two nitrogen atoms to give the expected four-membered
ring complex.
Pt(1)–N(2)
Pt(1)–C(41)
Pt(1)–C(45)
N(1)–C(1)
N(3)–C(1)
N(2)–C(21)
Pt(1)· · ·C(1)
2.002(7) Pt(1)–N(1)
2.173(9) Pt(1)–C(44)
2.155(9) Pt(1)–C(48)
1.398(11) N(2)–C(1)
1.298(11) N(1)–C(11)
1.406(12) N(3)–C(31)
2.561(9)
2.034(8)
2.188(9)
2.183(9)
1.402(11)
1.402(12)
1.406(11)
N(2)–Pt(1)–N(1)
C(1)–N(2)–Pt(1)
65.9(3)
96.0(5)
C(1)–N(1)–Pt(1)
94.7(5)
129.2(6)
103.3(7)
123.7(8)
125.1(8)
174.7(7)
C(11)–N(1)–Pt(1)
N(1)–C(1)–N(2)
N(3)–C(1)–N(2)
C(1)–N(1)–C(11)
N(3)–C(1)–Pt(1)
C(21)–N(2)–Pt(1) 133.5(6)
N(3)–C(1)–N(1) 133.0(8)
C(1)–N(3)–C(31) 122.1(8)
C(1)–N(2)–C(21) 124.8(7)
N(2)–N(1)–C(11) 36.8(7)
N(1)–N(2)–C(21) 26.2(5)
2.1.2. Characterisation by NMR spectroscopy and
electrospray mass spectrometry (ESMS)
Unambiguous NMR assignments were made using a
1
combination of NOE, H-13C COSY, and long range
BIRDTRAP
experiments.
(1J
suppression)
1H-13C
COSY
Table 1
Final positional and equivalent thermal parameters for
Since most of the starting materials contained the
triphenylphosphine ligand, 31P-NMR was conveniently
used to monitor reaction progress and outcomes. De-
tailed NMR assignments are given in the experimental
section, while significant and diagnostic NMR features
of the starting materials and prepared complexes are
summarised in Table 3.
A number of general conclusions can be drawn. In
the 31P-NMR, a downfield shift of the PPh3 (relative to
the starting material) is observed on coordination of the
guanidine dianion ligand, consistent with a decreased
availability of electron density about the metal (and
hence the phosphorus) centre. A downfield shift (rela-
tive to the starting material) of the 13C-NMR resonance
of the central guanidine carbon is produced on coordi-
nation, largely independent of metal.
¸¹¹¹¹¹¹¹¹º
[Pt{NPhC(ꢀNPh)NPh}(COD)] 4a
Atom
x
y
z
Ueq
Pt(1)
N(1)
N(2)
N(3)
0.2226(1)
0.3042(8)
0.2096(8)
0.3009(8)
0.2777(9)
0.3004(9)
0.2362(10)
0.2323(10)
0.2901(9)
0.3517(10)
0.3595(9)
0.2012(10)
0.3017(10)
0.2855(10)
0.1725(12)
0.0725(10)
0.0859(10)
0.3928(9)
0.3548(11)
0.4452(12)
0.5767(11)
0.6125(10)
0.5218(10)
0.1865(9)
0.0552(9)
0.0229(10)
0.0805(9)
0.2101(10)
0.3161(10)
0.3341(10)
0.3155(9)
0.0701(1)
0.0871(1)
0.1128(4)
0.014(1)
0.019(2)
0.017(2)
0.019(2)
0.016(2)
0.018(2)
0.025(2)
0.023(2)
0.026(2)
0.027(2)
0.021(2)
0.021(2)
0.021(2)
0.027(2)
0.025(2)
0.024(2)
0.022(2)
0.015(2)
0.027(2)
0.030(2)
0.032(3)
0.025(2)
0.025(2)
0.021(2)
0.017(2)
0.024(2)
0.018(2)
0.019(2)
0.025(2)
0.021(2)
0.019(2)
0.210(19)
−0.0580(5)
−0.0111(5)
−0.1638(5)
−0.0871(6)
−0.1149(7)
−0.1996(7)
−0.2533(7)
−0.2221(8)
−0.1338(8)
−0.0832(7)
0.0065(6)
−0.0247(7)
−0.0051(7)
0.0446(6)
0.0764(7)
0.0558(7)
−0.2312(6)
−0.3239(7)
−0.3934(7)
−0.3718(8)
−0.2820(7)
−0.2112(8)
0.1157(7)
0.1707(6)
0.2263(7)
0.1847(6)
0.2043(6)
−0.0071(4)
0.0011(4)
0.0343(5)
0.1773(6)
0.1710(6)
0.2362(6)
0.3109(6)
0.3179(6)
0.2534(5)
−0.0873(6)
−0.1273(6)
−0.2067(6)
−0.2474(6)
−0.2087(5)
−0.1301(6)
0.0412(5)
0.0254(6)
0.0574(6)
0.1060(6)
0.1201(5)
0.0875(6)
0.1996(5)
0.1887(5)
0.1136(6)
0.0461(5)
0.0328(5)
0.0828(6)
0.1712(6)
0.1914(5)
0.0000
C(1)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(31)
C(32)
C(33)
C(34)
C(35)
C(36)
C(41)
C(42)
C(43)
C(44)
C(45)
C(46)
C(47)
C(48)
In the triacetyl- and triphenyl-guanidine complexes
the 13C and H spectra indicate that the substituents on
1
the metal-coordinated N atoms are equivalent, despite
the expected inequivalence arising from the syn or
anti-arrangement of the substituents on the third N
atom. A variable-temperature H- and 13C{1H}-NMR
1
study was therefore undertaken on the COD complex
4a. The H-NMR spectra of 4a at four temperatures
1
ranging from 300–240 K are shown in Fig. 3. At r.t., a
single slightly broadened COD CHꢀCH resonance is
observed at l 4.92, which broadens further at 280 K,
before forming two CH resonances at 260 K, which
sharpen at 240 K. The 13C{1H}-NMR spectra of 4a at
the same temperatures show analogous behaviour, with
two CH and two CH2 resonances being observed at 240
K. These observations are interpreted as a r.t. fluxional
process (Scheme 2), which serves to interchange the
0.2674(7)
0.2513(7)
0.1509(7)
0.5000
O(111) 1.0000
C(111) 0.9406(46)
C(112) 0.9605(22)
0.5278(27)
0.6002(12)
0.0581(19) 0.249(25)
0.0909(10) 0.084(6)

M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
81
¸¹¹¹¹¹¹¹¹¹º
Fig. 2. Molecular structure of [(p−cymene)Ru{NAcC(ꢀNAc)NAc}(PPh3)] 8. Only one component of the disordered triacetylguanidine dianion
ligand is shown.
lone pair and phenyl substituent on the CꢀN nitrogen
N(3). At 240 K, this process is frozen out, leading to
the structure also observed in the solid state, with the
phenyl on N(3) rendering the two sides of the COD
ligand inequivalent. Unfortunately, it was not possible
to resolve J(PtC) or J(PtH) couplings in the frozen-
out structure, though small differences would be ex-
pected for CH groups trans to N(1) and N(3).
plexes [22]. It is also worth noting that the
triacetylguanidine dianion complexes typically showed
some loss of MeCN in their ES spectra and, since
no ureylene complexes were formed in the cases of
rhodium and ruthenium, this loss is interpreted in terms
1
2
of a hydrolysis reaction occurring in the mass
spectrometer.
Electrospray mass spectrometry (ESMS) can be suc-
cessfully used as an additional characterisation tech-
nique for the new complexes reported herein. ESMS is
finding increased use in the characterisation of a range
of organometallic and coordination complexes, and
generally leads to observation of strong parent ions
[21]. As an example, the platinum complex 4a gives the
parent ion [MH]+ as the base peak at a cone voltage of
50 V when recorded in MeCN–H2O solution; the other
guanidine dianion complexes show similar behaviour.
However, for the rhodium, iridium, ruthenium and
2.1.3. Conclusions
A number of platinum-group metal complexes con-
taining guanidine dianion (triazatrimethylenemethane)
ligands have been successfully synthesised using the
silver(I) oxide method. X-ray crystallography reveals
that the ligand bonds in a planar p2-type arrangement,
reminiscent of ureylene and carbonate complexes.
3. Experimental details
osmium complexes, which contain
a
coordinated
1H- and 13C{1H}-NMR spectra were recorded on a
Bruker AC300P spectrometer, at 300.13 and 75.47
MHz, respectively, in CDCl3, with SiMe4 (l 0.0) as the
external reference. 31P{1H}-NMR spectra were record-
ed, unless stated otherwise, in CDCl3 solution at 121.49
MHz on a Bruker AC300P spectrometer with 85%
H3PO4 (l 0.0) as the external reference. The 31P-NMR
triphenylphosphine ligand, ions of the type [MH−
PPh3]+ and [MH−PPh3+MeCN]+ were typically ob-
served. Presumably the sterically crowded environment
around the metal labilises the phosphine ligand in these
complexes. We have previously reported analogous be-
haviour for a series of Cp* rhodium(III) oxolene com-

82
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
Table 3
A comparison of significant NMR spectroscopic properties of the guanidine dianion complexes
Compound
13CꢀN (ppm)
Product PPh3 (l/ppm), (1JP−M/Hz)
Starting material PPh3 (l/ppm), (1JP−M/Hz)
PhNHC(NPh)NHPh
145.2
–
–
4a
4c
6b
166.4
Not observed
154.6
–
–
8.1 (3344)
35.6 (158.7)
14.0 (3750)
30.0 (144.0)
AcNHC(NAc)NHAc
7
6a
8
9
150.3
166.0
164.8
161.9
164.0
–
15.4
40.9 (150.2)
44.2
11.6 (301.5)
–
2.2
30.0 (144.0)
24.9
−12.4 (281.8)
spectrum of the rhodium complex 6b was recorded on a
JEOL FX90 spectrometer at 36.23 MHz. IR spectra
were recorded as KBr disks on a BioRad FTS-40
spectrometer; only major peaks in the region 1700–
1500 cm−1 are reported as distinctive fingerprints.
Melting points were recorded on a Reichert Hotstage
apparatus and are uncorrected.
Electrospray mass spectra were recorded in positive-
ion mode on a VG Platform II instrument, using
MeCN–H2O (1:1 v/v) as the mobile phase. Fragmenta-
tion was investigated by varying the skimmer cone
voltage, typically from 10 to 80 V. Isotope patterns for
major species were recorded and compared to calcu-
lated patterns obtained using the Isotope simulation
program [23].
N,N%,N¦-triphenylguanidine was prepared by reac-
tion of aniline with diphenylcarbodiimide [24]. ESMS
data for triphenylguanidine: cone voltage+10 V, m/z
288 (MH+, 100%). Cone voltage 100 V, m/z 288
(MH+, 3%), 196 ([PhNꢀCꢀNPh]H+, 100%)]. N,N%,N¦-
triacetylguanidine [25] was prepared by the literature
procedure, while N,N%-diacetylurea was prepared by
reaction of urea with acetyl chloride, by modification of
the literature procedure for propionylurea [26]. The
complex [PtCl2(COD)] was prepared via a minor mod-
ification of the literature procedure [27] and the com-
plex cis-[PtCl2(PPh3)2] was prepared from this complex
by displacement of the labile COD ligand by reaction
with two mole equivalents of PPh3 in CH2Cl2. The
complexes [Cp*RhCl2(PPh3)] (Cp*=p5-C5Me5) [28],
[Cp*IrCl2(PPh3)] [28], [(p-cymene)RuCl2(PPh3)] [29]
and [(p-cymene)OsCl2(PPh3)] [30] were prepared by the
appropriate literature procedures and their purities
checked by 31P{1H}-NMR.
dichloromethane (20 ml) for 3 h. Filtration to remove
the silver salts gave a bright yellow solution. The
solvent was removed by evaporation and subsequent
recrystallisation of the residue from dichloromethane–
diethyl ether gave bright yellow crystals of 4a, (0.074 g,
94%). This complex was further characterised by an
X-ray structure analysis (see below). M.p. 175°C (de-
comp.). Found: C, 54.4; H, 5.1; N, 6.8%.
C27H27N3Pt·Et2O·0.5CH2Cl2 requires: C, 54.3; H, 5.1;
N, 6.5%. ESMS: (Cone voltage+50 V): m/z 589
(MH+, 100%). IR: wmax 1609(s), 1590(s), 1574(s) and
1565(vs) cm−1. 1H-NMR: l 7.05 (4H, t, br, 3J3%,2%=7.33
Hz, H-3%,5%), 6.97 (4H, s, br, H-2%,6%), 6.86–6.73 (6H, m,
3
4
H–Ar), 6.51 (1H, tt, J4%,3%=7.01 Hz, J4%,2%=1.48 Hz
2
H-4¦), 4.92 (4H, (s, br), (d, JH,Pt=57.3 Hz), CHꢀCH),
2.59 (4H, m, CH–CH2), 2.21 (4H, m, CH–CH2). 13C-
NMR: l 166.4 (s, CꢀN), 148.0 (s, 1¦), 147.7 (s, C-1%),
129.4 (d, C-3%,5%), 128.3 (d, C-3¦,5¦), 123.2 (d, C-2%,6%),
123.0 (d, C-2¦,6¦), 122.0 (s, C-4%), 119.9 (s, C-4¦), 93.8
(d, (d, 1JC,Pt=140.5 Hz), CHꢀCH), 30.1 (t, CH–
CH2).The atom numbering scheme is shown for the r.t.
1H- and 13C-NMR spectra in Scheme 2.
3.¸2.¹A¹t¹te¹m¹p¹te¹d¹p¹reºparation of
[Pt{NPhC(ꢀNPH)NPh}(PPh3)2] 4c
A mixture of [PtCl2(COD)] (0.052, 0.139 mmol),
triphenylphosphine (0.073 g, 0.278 mmol), N,N%,N¦-
triphenylguanidine (0.082 g, 0.139 mmol), and silver(I)
oxide (0.987 g, excess) was refluxed in dichloromethane
(20 ml) for 4 h. Filtration to remove the silver salts gave
a dark orange solution. The solvent was removed by
evaporation, to give a solid residue that failed to crys-
tallise. NMR studies revealed the reaction does not
proceed cleanly, with considerable quantities of an un-
known impurity present. ESMS: (Cone voltage+10 V):
m/z 1019 (unidentified, 39%), 1005 (MH+, 88%), 771
([(Ph3P)2Pt(OH)2]NH4+, 100%), 754 ([(Ph3P)2Pt(OH)+
+H2O], 20%). 31P-NMR: (36.23 MHz) (CDCl3) l 8.14
¸¹¹¹¹¹¹¹¹¹º
3.1. Preparation of [Pt{NPhC(ꢀNPh)NPh}(COD)] 4a
[PtCl2(COD)] (0.050 g, 0.134 mmol), N,N%,N¦-
triphenylguanidine (0.039 g, 0.136 mmol) and silver(I)
1
1
(s, (d, JP,Pt=3344 Hz), Ph3P), 6.06 (s, (d, JP,Pt=3709
oxide
(0.103
g,
excess)
were
refluxed
in
Hz), impurity). 1H-NMR: l 7.42–7.07 (45H, m, H–Ar).

M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
83
¸¹¹¹¹¹¹¹¹º
Fig. 3. 1H-NMR spectra of the complex [Pt{NPhC(ꢀNPh)NPh}(COD)] 4a at temperatures of 300, 280, 260 and 240 K, in CDCl3 solution. The
COD alkene protons (l 4.92 at 300 K) become inequivalent at the lowest temperature due to freezing out of the fluxional process involving the
CꢀNPh moiety, see Scheme 2. The peaks marked * are due to CHCl3 (l 7.27) and CH2Cl2 (l 5.32), respectively.
¸¹¹¹¹¹¹¹¹º
An alternative synthesis of 4c, by ligand substitution
of the COD complex 4a was also attempted. Complex
4a (0.035 g, 0.059 mmol) and triphenylphosphine (0.032
g, 0.122 mmol) were refluxed in dichloromethane (25
ml) for 3 h. No colour change was noted and upon
removing the solvent under reduced pressure, the dis-
3.3. Preparation of [Pt{NAcC(ꢀO)NAc}(COD)] 5
The complex [PtCl2(COD)] (0.050 g, 0.134 mmol),
N,N%,N¦-triacetylguanidine (0.025 g, 0.135 mmol) and
silver(I) oxide (0.08 g, excess) were refluxed in
dichloromethane (20 ml) for 2 h. Filtration to remove
the silver salts gave a colourless solution. The solvent
was removed by evaporation and subsequent recrys-
tallisation of the residue from dichloromethane–diethyl
1
tinctive smell of free COD was not detected. H- and
31P-NMR showed only starting materials, indicating no
reaction had taken place.

84
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
Scheme 2. Fluxional process interconverting the Pt–NR groups, and COD CH and CH2 groups.The NMR numbering scheme for the phenyl
substituents is also shown for the r.t. spectrum with identical Pt–NPh moieties.
ether gave white crystals of 5 (0.052 g, 87%). M.p.
264°C (decomp.). Found: C, 35.1; H, 3.9; N, 6.2%.
C15H21N3O3Pt requires: C, 35.1; H, 4.1; N, 6.3%.
ESMS: (Cone voltage+50 V): m/z 446 (MH+, 100%),
360 (CODPt(OH)+ +CH3CN, 28%). IR: wmax 1726(vs),
3J3%,2%=8.88, C-3%, 5%), 7.43–7.35 (9H, m, br, C-2%,4%,6%),
1.97 (6H, s, RhNC(O)CH3), 1.83 (3H, s,
CꢀNC(O)CH3), 1.56 (15H, d, 4JH,P=3.17 Hz,
Cp-CH3). 13C-NMR: l 177.8 (s, RhNC(O)CH3), 174.1
(s, CꢀNC(O)CH3), 164.8 (s, CꢀN), 134.7 (d, (d,
3JC,P=11.17 Hz), C-3%,5%), 131.3 (d, C-4%), 130.1 (s,
1
1641(vs, br) cm−1. H-NMR: l 6.17 (4H, (s, br), (d, br,
2JH,Pt=64.0 Hz), CHꢀCH), 2.50 (4H, m, CH-CH2),
2.39 (6H, s, CH3) 2.31 (4H, m, CH-CH2). 13C-NMR: l
174.6 (s, CH3CꢀO), 165.3 (s, NC(O)N), 96.7 (d, (d,
1JC,Pt=132.7 Hz), CHꢀCH), 30.4 (t, CH-CH2), 26.7 (q,
1JC,P=45.35 Hz, C-1%), 128.2 (d, (d, JC,P=10.49 Hz),
2
C-2%,6%), 99.9 (s, Cp), 26.7 (q, RhNC(O)CH3), 26.1 (q,
CꢀNC(O)CH3), 8.8 (q, Cp–CH3).
3
(d, JC,Pt=36.5 Hz), CH3).
3.5. Preparation of
¸¹¹¹¹¹¹¹¹¹¹º
[CpꢀRh{NPhC(ꢀNPh)NPh}(PPh3)] 6b
¸¹¹¹¹¹¹¹¹¹¹º
3.4. Preparation of [CpꢀRh{NAcC(ꢀNAc)NAc}(PPh3)]
6a
The complex [Cp*RhCl2(PPh3)] (0.050 g, 0.088
mmol), N,N%,N¦-triphenylguanidine (0.026 g, 0.091
mmol) and silver(I) oxide (0.06 g, excess) were added to
dichloromethane (25 ml), which had previously been
degassed and flushed with nitrogen. The mixture was
refluxed under nitrogen for 1 h, during which time no
colour change was noted. At completion, with no fur-
ther efforts at excluding air, the silver salts were filtered
off and the solvent removed under reduced pressure, to
give a bright yellow oil which did not crystallise. 31P-
NMR revealed product with B50% purity, with a very
broad singlet (\100 Hz) indicating substantial decom-
position. ESMS: (Cone= +20 V) 978 (unidentified,
8%), 937 (unidentified, 10%), 834 (unidentified, 20%),
787 (MH+, 100%), 566 ([MH−PPh3+CH3CN]+,
The complex [Cp*RhCl2(PPh3)] (0.051 g, 0.089
mmol), N,N%,N¦-triacetylguanidine (0.017 g, 0.092
mmol) and silver(I) oxide (0.05 g, excess) were added to
dichloromethane (25 ml), which had previously been
degassed and flushed with nitrogen. The mixture was
refluxed under nitrogen for 1 h, during which time a
marked colour change from orange to bright yellow
was observed. An inert atmosphere appears to be
essential, with attempts at carrying out the reaction in
air proving unsuccessful. At completion, with no
further efforts at excluding air, the silver salts were
filtered off and the solvent removed under reduced
pressure. Recrystallisation from dichloromethane–
pentane gave 6a as bright yellow crystals (0.051 g,
85%). M.p. 153–155°C. Found: C, 59.7; H, 5.6; N,
5.7%. C35H38N3O3PRh·0.5CH2Cl2 requires: C, 58.8; H,
5.4; N, 5.0%. IR: wmax 1669(w), 1610(m) cm−1. ESMS:
(Cone voltage+15 V): m/z 1368 (2MH+, 3%), 684
(MH+, 100%). (Cone voltage+50 V): m/z 684 (MH+,
1
32%). 31P-NMR: (36.23 MHz) l 35.6 (d, JP,Rh=158.7
1
Hz, PPh3), 9.6 (s, br, impurity). H-NMR: l 7.48–6.64
(m, Ar–H), 1.61 (15H, s, Cp–CH3). 13C-NMR: l 154.6
(s, CꢀN), 144.9 (s, C-1¦), 138.3 (s, C-1%¦), 134.1 (d, (d,
1
3JC,P=16.45 Hz), C-3%,5%), 132.0 (s, JC,P=29.51 Hz,
2
C-1%), 130.3 (d, C-4%), 129.3 (d, Ar), 128.8 (d, JC,P
=
90%),
([MH−PPh3]+, 100%). 31P-NMR:
1JP,Rh=150.2 Hz, PPh3). 1H-NMR: l 7.53 (6H, t,
643
([MH−CH3CN]+,
30%),
422
10.04 Hz, C-2%,6%), 128.6 (d, C-3¦,5¦), 128.2 (d, Ar),
l
40.9 (d,
124.2 (d, C-2¦,6¦), 122.3 (d, Ar), 120.9 (d, Ar), 99.9 (s,
2
(d, JC,P=8.53 Hz), Cp), 9.2 (q, Cp–CH3).

M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
85
decomposed without melting. Found: C, 48.3; H, 4.5; N,
3.4%. C33H36N2O3PIr·CHCl3 requires: C, 48.0; H, 4.4;
N, 3.3%. IR: wmax 1694(s), 1616(s), 1600(s) cm−1. ESMS:
(Cone voltage+20 V) 733 (MH+, 100%). (Cone
voltage+50 V) 731 (MH+, 100%), 471 ([MH−PPh3]+,
22%), 263 ([PPh3H]+, 12%). (Cone voltage+80 V) 733
(MH+, 20%), 493 ([MNa−PPh3]+, 22%), 471
([MH−PPh3]+, 100%), 427 ([M−CH3C(O)–PPh3]+,
12%), 386 (unidentified, 80%), 263 ([PPh3H]+, 38%).
31P-NMR: l 17.6 (s, PPh3). 1H-NMR: l 7.61 (6H, m, br,
Ar–H), 7.41–7.37 (9H, m, Ar–H), 2.16 (6H, s,
IrNC(O)CH3), 1.56 (15H, s, (d, 4JP,H=1.51 Hz),
Cp–CH3).13C-NMR: l 175.9 (s, IrNC(O)CH3), 165.0 (s,
3
CꢀO), 134.9 (d, (d, JC,P=10.79 Hz), C-3%,5%), 130.7 (d,
1
C-4%), 130.0 (s, (d, JC,P=54.72 Hz), C-1%), 127.9 (d, (d,
¸¹¹¹¹¹¹¹¹¹º
3.6. Prepar¸at¹io¹n¹¹of¹¹[C¹p¹ꢀºIr{NAcC(ꢀNAc)NAc}(PPh3)]
7 and [Cpꢀ Ir{NAcC(ꢀO)N Ac}(PPh3)] 11
2JC,P=10.49 Hz), C-2%,6%), 92.9 (s, Cp), 27.5 (q,
IrNC(O)CH3), 9.3 (q, Cp–CH3).
3.8. Preparation of
The complex [Cp*IrCl2(PPh3)] (0.050 g, 0.076 mmol),
N,N%,N¦-triacetylguanidine (0.014 g, 0.076 mmol) and
silver(I) oxide (0.05 g, excess) were added to
dichloromethane (20 ml), which had previously been
degassed and flushed with nitrogen, and the mixture was
refluxed under nitrogen for 6 h. At completion, with no
further efforts at excluding air, the silver salts were
filtered, and the solvent removed from the filtrate under
reduced pressure to give a yellow oil, shown by NMR to
be a mixture of 7 and 11 in a 3:1 ratio. Recrystallisation
from dichloromethane and pentane gave 7 as bright
yellow crystals (0.036 g, 61%). M.p. 224–226°C. Found:
C, 52.3; H, 4.9; N, 5.2%. C35H39N3O3PIr·0.5CH2Cl2
requires: C, 54.3; H, 5.1; N, 5.4%. IR: wmax 1616(m),
1590(w), 1525(s, br) cm−1. ESMS: (Cone voltage+15
V): m/z 774 (MH+, 100%), 732 ([M−CH3CN]H+,
10%). (Cone voltage+50 V): m/z 774 (MH+, 100%),
732 ([11+H]+, 5%), 512 ([MH−PPh3]+, 20%).
[(p−cymene)¸R¹u{¹N¹A¹c¹C¹(ꢀ¹N¹A¹c¹)NºAc}(PPh3)] 8
To a Schlenk flask containing dichloromethane (20
ml) (which had previously been degassed and flushed
with nitrogen) was added [(p-cymene)RuCl2]2 (0.0.30 g,
0.049 mmol) and triphenylphosphine (0.026 g, 0.099
mmol). The solution was refluxed for 15 min. To this
was added N,N%,N¦-triacetylguanidine (0.018 g, 0.097
mmol) and silver(I) oxide (0.07 g, excess) and the
mixture was refluxed under nitrogen for a further 2 h,
during which time a marked colour change from bright
orange to pale yellow was observed. Without exclusion
of air, the silver salts were filtered off and solvent was
evaporated under reduced pressure to give a yellow oil.
Solutions of the compound became dark green, al-
though NMR revealed no decomposition, implying the
colouration is probably due to trace impurities. Recrys-
tallisation by vapour diffusion of pentane and ether
into dichloromethane over 3 months gave bright orange
blocks of 8 (0.040 g, 60%). M.p. 188–189°C. Found: C,
58.8; H, 5.6; N, 5.9%. C35H38N3O3PRu·0.5CH2Cl2 re-
quires: C, 59.0; H, 5.4; N, 5.8%. IR: wmax 1673(m),
1651(m), 1590(m), 1577(m), 1569(m). ESMS: (Cone
voltage+20 V): m/z 682 (MH+, 100%). (Cone
voltage+50 V): m/z 682 (MH+, 70%), 419 ([MH−
PPh3]+, 100%). (Cone voltage+80 V): m/z 419
1
31P-NMR: l 15.4 (s, PPh3). H-NMR: l 7.51 (6H, m,
Ar–H), 7.43–7.26 (9H, m, Ar–H), 1.96 (6H, s,
IrNC(O)CH3), 1.83 (3H, s, CꢀNC(O)CH3), 1.56 (15H, s,
4
(d, JP,H=1.51 Hz), Cp–CH3). 13C-NMR: l 176.3 (s,
IrNC(O)CH3), 174.4 (s, CꢀNC(O)CH3), 166.0 (s, CꢀN),
3
134.9 (d, (d, JC,P= 10.79 Hz), C-3%,5%), 130.8 (d, C-4%),
130.0 (s, (d, 1JC,P=54.72 Hz), C-1%), 128.1 (d, (d,
2JC,P=10.49 Hz), C-2%,6%), 93.6 (s, Cp), 26.3 (q,
CꢀNC(O)CH3), 26.1 (q, IrNC(O)CH3), 9.3 (q,
Cp–CH3). Atom numbering scheme as for complex 6b.
([MH−PPh3]+,
22%),
335
([(p-cymene)Ru+
2CH3CN+H2O]+, 100%), 294 ([(p-cymene)Ru+
CH3CN+H2O]+, 80%). 31P-NMR: l 44.2 (s, PPh3).
¸¹¹¹¹¹¹¹¹º
3.7. Preparation of [Cpꢀ Ir{NPhC(ꢀO)N Ac}(PPh3)] 11
1H-NMR: l 7.56 (6H, t, 3J3¦,2¦=8.63 Hz, C-3¦,5¦),
3
The complex [Cp*PPh3IrCl2] (0.051 g, 0.077 mmol),
N,N%-diacetylurea (0.011 g, 0.076 mmol) and silver(I)
oxide (0.08 g, excess) were refluxed in dichloromethane
(25 ml) under nitrogen for 2 h. Workup gave a yellow
oil that readily crystallised by vapour diffusion of
pentane into a saturated chloroform solution at r.t., to
give large yellow blocks of 11 (0.047 g, 84%). M.p.
7.43–7.35 (9H, m, C-2¦, 4¦, 6¦), 5.95 (2H, d, J2¦,3¦
=
3
6.12 Hz, C-2%,6%), 5.05 (2H, d, J3¦,2¦=5.98 Hz, C-3%,5%),
2.74 (1H, q, 3J4,1=6.87 Hz, C-4), 1.93 (6H, s,
RuNC(O)CH3), 1.82 (3H, s, CꢀNC(O)CH3), 1.68 (3H,
s, C-3), 1.16 (6H, d, J1,4=6.90 Hz, C-1, 2). 13C-NMR:
3
l177.9 (s, RuNC(O)CH3), 174.4 (s, CꢀNC(O)CH3),
161.9 (s, CꢀNC(O)CH3), 134.6 (d, (d, 3JC,P=10.94 Hz),

86
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
1
C-3¦,5¦), 131.8 (s, (d, JC,P=45.58 Hz), C-1¦), 130.5 (d,
silver(I) oxide (0.13 g, excess) were refluxed in
dichloromethane (25 ml) under nitrogen for 2 h. Workup
gave a very pale yellow oil that resisted crystallisation.
Evacuation of all solvent, however, gave 10 as a pale
yellow solid of sufficient purity (0.055 g, 95%). M.p.
235–238°C. Found: C, 53.6; H, 5.0; N, 4.0%.
C33H35N2O3POs requires: C, 54.4; H, 4.8; N, 3.8%. IR:
wmax 1692(s), 1616(s), 1599(s) cm−1. ESMS: (Cone
voltage+20 V) 731 (MH+, 100%). (Cone= +50 V) 731
(MH+, 100%), 510 ([MH−PPh3+CH3CN]+, 3%), 469
([MH−PPh3]+, 10%), 263 ([PPh3H]+, 22%). (Cone
voltage+80 V) 731 (MH+, 20%), 510 ([MH−PPh3+
CH3CN]+, 22%), 469 ([MH−PPh3]+, 48%), 425 ([M−
CH3C(O)–PPh3]+, 48%), 384 (unidentified, 80%), 263
([PPh3H]+, 100%). 31P-NMR: l 13.9 (s, (d, 1JP,Os=301.1
Hz), PPh3). 1H-NMR: l 7.62–7.55 (6H, m, Ar–H),
7.42–7.33 (9H, m, Ar–H), 6.03 (2H, d, 3J2%,3%=5.77 Hz,
C-2%,6%), 5.15 (2H, d, 3J3¦,2¦=5.73 Hz, C-3%,5%), 2.53 (1H,
quintet, 3J4,1=6.90 Hz, C-4), 2.10 (6H, s, OsNC(O)CH3),
2
C-4¦), 128.2 (d, (d, JC,P=9.96 Hz), C-2¦,6¦), 112.4 (s,
(d, 2JC,P=8.75 Hz), C-1%), 102.6 (s, C-4%), 88.4 (d, C-2%,6%),
86.8 (d, C-3%,5%), 30.9 (d, C-4), 26.4 (q, CꢀNC(O)CH3),
25.8 (q, RuNC(O)CH3), 22.5 (q, C-1,2), 18.7 (q, C-3).
3¸.9¹. ¹P¹re¹p¹ar¹a¹ti¹on¹oºf [(p−cymene)-
Os{NAcC(ꢀN¸A¹c)¹N¹A¹c¹}(¹P¹P¹h3º)] 9 and
[(p−cymene)Os{NAcC(ꢀO)NAc}(PPh3)] 10
3
1.80 (3H, s, C-3), 1.04 (3H, d, J1,4=6.93 Hz, C-1, 2).
13C-NMR: l 176.7 (s, OsNC(O)CH3), 164.6 (s, CꢀO),
To a Schlenk flask containing dichloromethane (20 ml)
(which had previously been degassed and flushed with
nitrogen) was added [(p-cymene)OsCl2(PPh3)] (0.050 g,
0.076 mmol), N, N%, N¦-triacetylguanidine (0.014 g, 0.076
mmol) and silver(I) oxide (0.10 g, excess) and the mixture
was refluxed under nitrogen for 2 h, during which time
the colour changed from bright yellow to almost colour-
less. The silver salts were filtered off and solvent was
evaporated under reduced pressure to give a pale yellow
oil which, despite repeated efforts, did not crystallise.
NMR spectroscopy revealed the material to be a ca.
50:50 mixture of 9 and 10. ESMS: (Cone voltage+15 V)
m/z 772 ([9+H]+, 100%), 731 ([10+H]+, 28%).
3
134.9 (d, (d, JC,P=10.64 Hz), C-3¦,5¦), 131.3 (s, (d,
1JC,P=52.15 Hz), C-1¦), 130.5 (d, C-4¦), 128.4 (d, (d,
2
2JC,P=9.58 Hz), C-2¦,6¦), 104.0 (s, (d, JC,P=9.89 Hz),
C-1%), 94.2 (s, C-4%), 79.8 (d, C-2%,6%), 76.7 (d, C-3%,5%), 31.2
(d, C-4), 26.6 (q, OsNC(O)CH3), 22.6 (q, C-1, 2), 18.4
(q, C-3). Atom numbering scheme as for 8 above.
3.10. X-ray st¸ru¹c¹tu¹re¹d¹e¹te¹rm¹¹inºation of
[(p−cymene)Pt{NPhC(ꢀNPh)NPh}(COD)]
4a ·CH2Cl2 ·0.5Et2O
Yellow rectangular blocks were obtained on crystalli-
sation by vapour diffusion of diethyl ether into a
saturated dichloromethane solution of 4a at 4°C. Accu-
rate cell parameters and intensity data were collected on
a Nicolet R3 diffractometer, using a crystal of dimen-
sions 0.80×0.24×0.22 mm, and Mo–Kh radiation
3.9.1. [(p−cymene)¸O¹s{¹N¹A¹c¹C¹(ꢀ¹N¹A¹c¹)NºAc}(PPh3)] 9
31P-NMR: l 11.6 (s, (d, 1JP,Os=301.5 Hz), PPh3).
1H-NMR: l 7.60 (6H, m, Ar–H), 7.38–7.33 (9H, m,
3
Ar–H), 5.99 (2H, d, J2%,3%=5.58 Hz, C-2%,6%), 5.13 (2H,
˚
d, 3J3%,2%=5.58 Hz, C-3%,5%), 2.61 (1H, quintet, 3J4, 1=6.87
Hz, C-4), 1.92 (6H, s, OsNC(O)CH3), 1.84 (3H, s, C-3),
1.81 (3H, s, CꢀNC(O)CH3), 1.12 (3H, d, 3J1,4=6.89 Hz,
C-1,2). 13C-NMR: l 177.0 (s, OsNC(O)CH3), 174.6 (s,
CꢀNC(O)CH3), 164.0 (s, CꢀNC(O)CH3), 134.9 (d, (d,
3JC,P=10.72 Hz), C-3¦,5¦), 131.4 (s, (d, 1JC,P=52.45
Hz), C-1¦), 130.6 (d, C-4¦), 128.1 (d, (d, 2JC,P=9.58 Hz),
(u=0.71073
A).
Crystal
data:
C27H27N3Pt·
CH2Cl2 ·0.5Et2O, M=625.67, monoclinic, space group
˚
P21/c with a=9.862(1), b=14.497(3), c=17.436(3) A,
i=103.02(1)°, U=2428.7(6) A , Dcalc. =1.711 g cm−3
,
3
˚
Z=4, v(Mo–Kh)=5.80 mm−1, F(000)=1236. A total
˚
of 4291 reflections in the range 2BqB25 A were
collected at 130(2) K, of which 4270 were unique. These
were subsequently corrected for Lorentz and polarisation
effects, and for linear absorption by a C scan method
(Tmax, min=0.49, 0.18). The structure was solved by
Patterson interpretation [31] and developed routinely. A
penultimate difference map revealed electron density
which was attributed to a disordered diethyl ether
molecule lying across an inversion centre in the lattice.
In the final cycle of full-matrix least-squares refinement
based on F2 using SHELXL-93 [32] all non-H atoms were
assigned anisotropic temperature factors, with all H-
2
C-2¦,6¦), 105.5 (s, (d, JC,P=8.98 Hz), C-1%), 94.7 (s,
C-4%), 80.0 (d, C-2%,6%), 77.8 (d, C-3%,5%), 31.0 (d, C-4), 25.2
(q, OsNC(O)CH3), 22.6 (q, C-1,2), 22.1 (q,
CꢀNC(O)CH3), 18.4 (q, C-3). Atom numbering scheme
as for 8 above.
3.9.2. Preparation of
¸¹¹¹¹¹¹¹¹º
[(p−cymene)Os{NAcC(ꢀO)NAc}(PPh3)] 10
The complex [(p-cymene)OsCl2(PPh3)] (0.052 g, 0.079
mmol), N,N%-diacetylurea (0.012 g, 0.083 mmol) and

M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
87
[2] See for example (a) J. Pickardt, B. Kuhn, Z. Naturforsch. 51b
(1996) 1701. (b) J. Pickardt, B. Kuhn, Z. Naturforsch. 51b (1996)
1469.
[3] P.J. Bailey, L.A. Mitchell, S. Parsons, J. Chem. Soc. Dalton
Trans. (1996) 2839.
[4] P.J. Bailey, S.F. Bone, L.A. Mitchell, S. Parsons, K.J. Taylor,
L.J. Yellowlees, Inorg. Chem. 36 (1997) 867.
[5] (a) N.J. Bremer, A.B. Cutcliffe, M.F. Farona, W.G. Kofron, J.
Chem. Soc. A (1971) 3264. (b) P.J. Bailey, R.O. Gould, C.N.
Harmer, S. Pace, A. Steiner, D.S. Wright, Chem. Commun.
(1997) 1161.
atoms in calculated positions. The refinement converged
with R1=0.0456 for 3197 data with I52|(I), 0.0612 for
all data; wR2=0.1147, and GOF=0.933. The largest
parameter shift was 0.4| (for the disordered diethyl ether
solvent molecule) in the final cycle, and in the final
difference map the largest features were +2.58 and
−3
˚
−1.97 e A
near the platinum atom.
Full tables of atomic coordinates, bond lengths and
angles, and thermal parameters have been deposited with
the Cambridge Crystallographic Data Centre, and can be
obtained from the authors on request.
[6] For a review see: (a) M.D. Jones, R.D.W. Kemmitt, Adv.
Organomet. Chem. 27 (1987) 279. For selected recent references
see: (b) V. Plantevin, P.W. Blosser, J.C. Gallucci, A. Wojcicki,
Organometallics 13 (1994) 3651. (c) S. Watanabe, S. Ogoshi, K.
Kakiuchi, H. Kurosawa, J. Organomet. Chem. 481 (1994) 19. (d)
I.-Y. Wu, M.-C. Cheng, Y.- C. Lin, Y. Wang, Organometallics
12 (1993) 1686. (e) G.E. Herberich, T.P. Spaniol, J. Chem. Soc.
Dalton Trans. (1993) 2471. (f) L. Girard, J.A. MacNeil, A.
Mansour, A.C. Chiverton, J.A. Page, S. Fortier, M.C. Baird,
Organometallics 10 (1991) 3114.
[7] (a) G. Rodriguez, G.C. Bazan, J. Am. Chem. Soc. 119 (1997)
343. (b) G. Rodriguez, G.C. Bazan, J. Am. Chem. Soc. 117
(1995) 10155. (c) G.C. Bazan, G. Rodriguez, B.P. Cleary, J. Am.
Chem. Soc. 116 (1994) 2177. (d) G.E. Herberich, T.P. Spaniol, J.
Chem. Soc. Chem. Commun. (1991) 1457. (e) G.E. Herberich, C.
Kreuder, U. Englert, Angew. Chem. Int. Ed. Engl. 33 (1994)
2465. (f) G.E. Herberich, U. Englert, L. Wesemann, P. Hof-
mann, Angew. Chem. Int. Ed. Engl. 30 (1991) 313.
[8] (a) J.-T. Chen, Y.-K. Chen, J.-B. Chu, G.-H. Lee, Y. Wang,
Organometallics 16 (1997) 1476. (b) M.W. Baize, V. Plantevin,
J.C. Gallucci, A. Wojcicki, Inorg. Chim. Acta 235 (1995) 1. (c)
T.-M. Huang, R.-H. Hsu, C.-S. Yang, J.-T. Chen, G.-H. Lee, Y.
Wang, Organometallics 13 (1994) 3657. (d) K. Ohe, H. Matsuda,
T. Morimoto, S. Ogoshi, N. Chatani, S. Murai, J. Am. Chem.
Soc. 116 (1994) 4125. (e) J.-T. Chen, T.-M. Huang, M.-C.
Cheng, Y.-C. Lin, Y. Wang, Organometallics 11 (1992) 1761.
[9] P.J. Bailey, A.J. Blake, M. Kryszczuk, S. Parsons, D. Reed, J.
Chem. Soc. Chem. Commun. (1995) 1647.
[10] P.J. Bailey, L.A. Mitchell, P.R. Raithby, M.-A. Rennie, K.
Verhorevoort, D.S. Wright, Chem. Commun. (1996) 1351.
[11] (a) D. Seyferth, T. Wang and W.M. Davis, Organometallics 13
(1994) 4134. (b) H.O. Frohlich, R. Wyrwa, H. Gorls, J.
Organomet. Chem. 456 (1993) 7. (c) S.D. Chappell, D.J. Cole-
Hamilton, Polyhedron 1 (1982) 739.
3.11. X-Ray structure determination of
[(p−cymene)¸R¹u{¹N¹A¹c¹C¹(ꢀ¹N¹A¹c¹)NºAc}(PPh3)] 8
An orange rosette of 8 was obtained on crystallisation
by vapour diffusion of diethyl ether and pentane into a
saturated dichloromethane solution at 4°C, from which
a rectangular block (dimensions 0.88×0.34×0.14 mm)
suitable for X-ray analysis was cleaved. Preliminary
precession photography indicated orthorhombic symme-
try, with space group Pbcn. Crystal data:
C35H38N3O3PRu, M=682.58, orthorhombic, space
group Pbcn, a=25.174(4), b=14.847(2), c=16.913(9)
3
A, U=6322(4) A , Dcalc. =1.435 g cm−3, Z=8, v(Mo–
Kh)=5.86 mm−1, F(000)=2832. A total of 5445 reflec-
tions in the range 2.41BqB22.50° were collected at
130(2)K, of which 4131 were unique (Rint=0.056). These
were corrected for Lorentz and polarisation effects, and
˚
˚
for linear absorption by a C scan method (Tmax, min
=
0.35, 0.32). Solution (direct methods, SHELXS-86) [31]
gave the positions of the Ru and P atoms. Subsequent
difference maps slowly revealed the rest of the structure,
though with extensive disorder of the acetyl substituents
on the metallacyclic ring. No satisfactory refinement
model could be developed, the best giving R1 0.1455 (2|
data), Rw2 0.2896 (all data). Although the overall connec-
tivity is unambiguous, detailed parameters are unreliable
so are not reported in this paper.
[12] (a) D. Seyferth, T. Wang, W.M. Davis, Organometallics 13
(1994) 4134. (b) D. Seyferth, T. Wang, R.L. Ostrander, A.L.
Rheingold, Organometallics 14 (1995) 2136. (c) G.E. Herberich,
T.P. Spaniol, J. Chem. Soc. Chem. Commun. (1991) 1457. (d)
K.W. Chiu, W. Henderson, R.D.W. Kemmitt, L.J.S. Prouse,
D.R. Russell, J. Chem. Soc. Dalton Trans. (1988) 427.
[13] (a) W. Henderson, J. Fawcett, R.D.W. Kemmitt, C. Proctor,
D.R. Russell, J. Chem. Soc. Dalton Trans. (1994) 3085 and refs.
therein. (b) A.D. Burrows, D.M.P. Mingos, A.J.P. White, D.J.
Williams, J. Chem. Soc. Dalton Trans. (1996) 149. (c) A.D.
Burrows, D.M.P. Mingos, A.J.P. White, D.J. Williams, J. Chem.
Soc. Dalton Trans. (1996) 3805.
Acknowledgements
We thank the University of Waikato for financial
support, the New Zealand Lottery Grants Board for a
grant-in-aid towards a mass spectrometer and Prof Ward
Robinson (University of Canterbury) for collection of the
X-ray data sets. We also thank Johnson Matthey for a
generousloanofplatinumandM.B.Dingeracknowledges
the William Georgetti Foundation for a scholarship.
[14] M.B. Dinger, W. Henderson, B.K. Nicholson, A.L. Wilkins, J.
Organomet. Chem. 526 (1996) 303.
[15] (a) P. Braunstein, D. Nobel, Chem. Rev. 89 (1989) 1927. (b)
A.K. Burrell, A.J. Steedman, Organometallics 16 (1997)1203.
[16] (a) A.J. Blake, P. Mountford, G.I. Nikonov, D. Swallow, Chem.
Commun. (1996) 1835. (b) F. Paul, J. Fischer, P. Ochsenbein,
J.A. Osborn, Angew. Chem. Int. Ed. Engl. 32 (1993) 1638. (c)
W.-H. Leung, G. Wilkinson, B. Hussain-Bates, M.B. Hurst-
house, J. Chem. Soc. Dalton Trans. (1991) 2791.
References
[1] R.C. Mehrotra, in: G. Wilkinson (Ed.), Comprehensive Coordi-
nation Chemistry, Pregamon, Oxford, pp. 281-285.
[17] M.B. Dinger, W. Henderson, Chem. Commun. (1996) 211.

88
M.B. Dinger et al. / Journal of Organometallic Chemistry 556 (1998) 75–88
[18] A. Kemme, M. Rutkis, J. Eiduss, Latv. PSR Zinat. Akad. Vestis
Khim. Ser. (1988) 595.
rations, vol. II, Academic Press, New York, 1971, p. 217.
[25] R. Greenhalgh, A.A.B. Bannard, Can. J. Chem. 37 (1959)
1810.
[26] G. Hilgetag, A. Martini (Eds.), Preparative Organic Chemistry,
Wiley, Toronto, 1972, p. 471.
[19] (a) M.R. Gregg, J. Powell, J.F. Sawyer, Acta Crystallogr. Sect. C
44 (1988) 43. (b) M.A. Andrews, G.L. Gould, W.T. Klooster,
K.S. Koenig, E.J. Voss, Inorg. Chem. 35 (1996) 5478. (c) T.K.
Miyamoto, Y. Suzuki, H. Ichida, Bull. Chem. Soc. Japan 65
(1992) 3386.
[27] J.X. McDermott, J.F. White, G.M. Whitesides, J. Am. Chem.
Soc. 98 (1976) 6521.
[20] C.- C Su, J.-T. Chen, G.-H. Lee, Y. Wang, J. Am. Chem. Soc.
116 (1994) 4999.
[28] J.W. Kang, K. Moseley, P.M. Maitlis, J. Am. Chem. Soc. 91
(1969) 5970.
[21] (a) C.E.C.A. Hop, R. Bakhtiar, J. Chem. Educ. 73 (1996) A162.
(b) R. Colton, A. D’Agostino, J.C. Traeger, Mass Spectrom.
Rev. 14 (1995) 79.
[22] W Henderson, J. Fawcett, R.D.W. Kemmitt, D.R. Russell, J.
Chem. Soc. Dalton Trans. (1995) 3007.
[29] A. Demonceau, A.F. Noels, E. Saive, A.J. Hubert, J. Mol. Catal.
76 (1992) 123.
[30] H. Werner, K. Zenkert, J. Organomet. Chem. 345 (1988) 151.
[31] G.M. Sheldrick, SHELXS-86, Program for solving crystal struc-
tures, University of Gottingen, 1986.
[23] L.J. Arnold, J. Chem. Educ. 69 (1992) 811.
[24] S.R. Sandler, W. Karo (Eds.), Organic Functional Group Prepa-
[32] G.M. Sheldrick, SHELXL-93, Program for refining crystal struc-
tures, University of Gottingen, 1993.
.