Synthesis and destannylation of η3-1-stannylallylpalladium(II) complexes

Article abstract of DOI:10.1016/S0022-328X(00)00801-9

Reaction of 1-tributylstannyl-3-chloropropene with a Pd(PPh₃) species, generated in situ from Pd₂(dba)₃ and two equivalents of PPh₃, afforded Pd(η³-Bu₃SnCHCHCH₂)Cl(PPh₃) (1). Complex 1 underwent PPh₃ promoted protodestannylation with acetic acid or diethyl malonate to give an unsubstituted η³-allylpalladium moiety as either a detectable product or a reaction intermediate. The reaction of 1 with PPh₃ and 0.5 equivalent of PdCl₂(PhCN)₂ afforded the dinuclear complex (μ-1-3-η³:4-6-η³-CH₂CHCHCHCHCH ₂){PdCl(PPh₃)}₂ (4) containing a hexatriene ligand, a formal vinylcarbene dimer.

Full text of DOI:10.1016/S0022-328X(00)00801-9

Journal of Organometallic Chemistry 625 (2001) 54–57
www.elsevier.nl/locate/jorganchem
Synthesis and destannylation of h3-1-stannylallylpalladium(II)
complexes
Takuma Nishida, Saisuke Watanabe, Tomohiro Yoshida, Sensuke Ogoshi,
Tetsuro Murahashi, Hideo Kurosawa *
Department of Applied Chemistry, Faculty of Engineering, Osaka Uni6ersity, Suita, Osaka 565-0871, Japan
Received 2 October 2000; received in revised form 25 October 2000; accepted 26 October 2000
Abstract
Reaction of 1-tributylstannyl-3-chloropropene with a Pd(PPh3) species, generated in situ from Pd2(dba)3 and two equivalents of
PPh3, afforded Pd(h3-Bu3SnCHCHCH2)Cl(PPh3) (1). Complex 1 underwent PPh3 promoted protodestannylation with acetic acid
or diethyl malonate to give an unsubstituted h3-allylpalladium moiety as either a detectable product or a reaction intermediate.
The reaction of 1 with PPh3 and 0.5 equivalent of PdCl2(PhCN)2 afforded the dinuclear complex (m-1-3-h3:4-6-h3-
CH2CHCHCHCHCH2){PdCl(PPh3)}2 (4) containing a hexatriene ligand, a formal vinylcarbene dimer. © 2001 Elsevier Science
B.V. All rights reserved.
Keywords: Destannylation; h3-1-Stannylallylpalladium(II) complexes
1. Introduction
2. Results and discussion
Organometallic complexes bearing readily removable
functional groups on organic ligands have attracted
attention as potential precursors of metal bound reac-
tive intermediates, as exemplified by 1-haloalkyl or
1-alkoxyalkyl metal complexes serving as carbene–
metal sources [1,2]. We have been interested in h3-allyl-
palladium complexes of the type Pd{h3-CH2C(CH2Y)-
CH2}(X)(L) as a possible precursor of the trimethylen-
emethane–palladium intermediate where Y represents
an electrophilic (e.g. R3Sn) [3] or nucleophilic (e.g. Cl,
OS(O)Ph) [4] leaving group. Stimulated by a proposal
[5], without proof, on the generation of vinylcarbene–
palladium species from h3-1-silylallylpalladium interme-
diate, we undertook studies of synthesis and reactivities
of 1-silylallyl and 1-stannylallyl complexes of palla-
dium.
Initial attempts were made to accomplish desilylation
of the known complex {Pd(h3-Me3SiCHCHCH2)Cl}2
[6] via, e.g. treatment with excess PPh3 and/or AgOAc,
but no well-characterizable results were obtained. We
then turned our attention to stannyl analogs in view of
the expected weaker SnꢀC bond than the SiꢀC bond.
Indeed, we had found earlier that the stannylmethyl-
substituted complex [Pd{h3-CH2C(CH2SnMe3)CH2}-
(PPh3)2]Cl is a much better trimethylenemethane pre-
cursor than the silyl analog [3]. Treatment of 1-tributyl-
stannyl-3-chloropropene with Pd2(dba)3 and PPh3
(Pd/PPh3=1:1) in CH2Cl2 at room temperature af-
forded pale yellow solids of the composition Pd(h3-
Bu3SnCHCHCH2)Cl(PPh3) (1) in 21% isolated yield.
1
The H- and 31P-NMR spectra of 1 showed the pres-
ence of almost one isomer (Eq. (1)). The resonance of
the allylic hydrogen geminal to Sn group (l 5.06)
showed a strong NOE correlation (15%) with the hy-
drogen at the allylic center (l 6.27), suggesting the
anti-Sn configuration. This preferred anti configuration
is unusual for a h3-allyl palladium complex, but we
cannot give any reason for this [7].
* Corresponding author. Fax: +81-66-68797394.
E-mail
address:
kurosawa@ap.chem.eng.osaka-u.ac.jp
(H.
Kurosawa).
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00801-9

T. Nishida et al. / Journal of Organometallic Chemistry 625 (2001) 54–57
55
ization at different temperatures was difficult owing to
the broadness of the peaks caused by the dynamic
movement.
(1)
In view of the previous assumption [5] that the
1-silylallyl–palladium moiety underwent protodesilyla-
tion with malonate esters, complex 1 was treated with
diethyl malonate in CH2Cl2 at 25°C, but no reaction
took place. However, addition of excess PPh3 (three
equivalents) to a mixture of 1 and diethyl malonate
under the same condition caused the coupling reaction
to give diethyl 2-allylmalonate (40%, 46 h) (Eq. (2)).
(3)
As expected, addition of CD3OD to a CD2Cl2 solu-
tion of the mixture of 1 and PPh3 (CD3OD–CD2Cl2=
1:6) led to predominant formation of the cationic
complex 2 even at room temperature.
It is to be noted that no protodestannylation oc-
curred in this mixed solvent solution. In contrast, the
trimethylenemethane precursor, [Pd{h3-CH2C(CH2-
(2)
SnMe3)CH2}(PPh3)2]Cl
underwent
very
facile
In order to clarify the role of PPh3 in promoting the
protodestannylation under the same conditions to give
a deuterated h3-2-methylallyl complex [3]. Addition of
acetic acid, instead of methanol, to the mixture of 1 and
PPh3 resulted in protodestannylation to give a quantita-
tive yield (40 h) of Pd(h3-CH2CHCH2)Cl(PPh3).
above coupling, a CD2Cl2 solution of 1 and PPh3 (1:1)
1
was examined by H- and 31P-NMR spectroscopy. At
room temperature, both 1H- and 31P-NMR spectra
showed very broad peaks with small chemical shift
changes compared to those of 1 alone, suggesting the
occurrence of dynamic movements and/or equilibrium
on the NMR time scale. On the other hand, at −60°C,
the new allyl proton resonances predominated at l 2.47
(br t, J=12 Hz, 1H), 4.25 (br, 1H), 4.30 (br d, J=8
Hz, 1H), 6.58 (br, 1H). These chemical shifts are consis-
tent with a h3-allyl–Pd, but not a h1-allyl–Pd form; in
the latter case the CH2 protons would have appeared
For the PPh3-promoted destannylation of 1 with
diethyl malonate or acetic acid, we suggest a plausible
reaction path depicted in Scheme 1 where the Sn atom
of the cationic complex 2 is susceptible to the attack of
Cl−, affording the vinylcarbene–palladium intermedi-
ate (3) which would abstract a proton from the mal-
onate or acetic acid [10]. It should be pointed out here
that in the absence of the proton source, the reaction
mixture of 1 and PPh3 underwent a very slow uncharac-
terizable decomposition path where no evidence for the
vinylcarbenepalladium moiety was obtained. An alter-
native reaction sequence for Eq. (1) involving initial
CꢀC coupling between 2 and CH2(COOEt)2 to give
either as
a singlet for equivalent two protons
(Bu3SnCHꢁCHCH2ꢀPd linkage) or as part of terminal
olefin proton multiplets (CH2ꢁCHCH(SnBu3)ꢀPd link-
age) [8]. Moreover, these 1H spectral changes were
accompanied by the new 31P aspect showing two dou-
blets (l 21.59, 24.09, JPP=38 Hz), consistent with the
cis configuration of two PPh3 groups in h3-allyl com-
plexes [9]. Almost identical spectral data were observed
for a sample prepared by the treatment of the mixture
of 1 and PPh3 with AgBF4 at room temperature. As to
the mixture of 1 and PPh3, we propose the occurrence
of the equilibrium (Eq. (3)) where the neutral
monophosphine complex 1 dominates at room tempera-
ture and the cationic bisphosphine counterpart [Pd(h3-
Bu3SnCHCHCH2)(PPh3)2]Cl (2) dominates at the lower
temperature. Accurate estimation of the degree of ion-
CH2ꢁCHCH(SnBu3)CH(COOEt)2
or
Bu3SnCHꢁ
CHCH2CH(COOEt)2 followed by their protonolysis
seems unlikely, since [Pd(h3-allyl)(PPh3)2]Cl does not
react with neutral malonate esters.
Attempts were next made to substitute another Pd
unit for Sn of 1 by using PdCl2(PhCN)2 in the hope of
obtaining m-h3-vinylcarbene dinuclear complexes, since
vinylcarbene ligands were known to be stabilized when
bridging over the PdꢀPd bond [11]. When the 1:1
mixture of 1 and PPh3 was treated with an equimolar
Scheme 1.

56
T. Nishida et al. / Journal of Organometallic Chemistry 625 (2001) 54–57
1
amount of PdCl2(PhCN)2, the H-NMR analysis indi-
cated the appearance of resonances for some products
among which a set of resonances ascribable to a C3H4
unit are discerned; l 2.98 (l, J=12 Hz, 1H), 3.11 (l,
J=7 Hz, 1H), 4.69 (m, 1H), 5.78 (m, 1H). These peaks
became dominant (92%, 40 h) when the amount of
PdCl2(PhCN)2 was decreased to 0.5 equivalent with
respect to 1. The reaction was not so clean in the
absence of PPh3. After several attempts had been made
to identify the major product, its structure was deter-
mined as 4 (Eq. (4)) having the 1-3-h3:4-6-h3-hexadi-
enediyl ligand by synthesizing an authentic sample via
an alternative route. This route involved substitution of
hexatriene for m-butadiene in Pd2Cl(PPh3)(m-Cl)(m-buta-
diene) [12], followed by addition of one equimolar
Scheme 2.
3.2. Preparation of 1-tributylstannyl-3-hydroxypropene
[14]
1
amount of PPh3. The H-NMR data described above
are consistent with 4 indicated. Strong NOE correla-
tions were observed between the peaks at l 4.69 (at C3)
and 2.98 (anti-H at C1) (3%) as well as those at l 4.69
and 5.78 (at C2%) (9%) [13].
Into a stirred THF suspension (350 ml) of LiAlH4
(4.00 g, 105 mmol) cooled at 0°C was added propargyl
alcohol (11.7 g, 209 mmol) drop by drop. Stirring was
(4)
A possible route to 4 is shown in Scheme 2. The
electrophilic replacement of the stannyl group in 2 by
Pd(II), possibly via the intermediate 3, would give a
trinuclear intermediate (5) containing two m-h3-vinyl-
carbene ligands. This would then undergo coupling of
two vinylcarbene ligands on the central Pd. The CꢀC
coupling between the vinylcarbene ligand and Ph within
the PdꢀPd framework has been known [11b]. An alter-
continued for 17 h at room temperature (r.t.). The
mixture was cooled to −60°C, and an Et2O solution
(70 ml) of Bu3SnOTf (25.9 g, 61.5 mmol) was added.
The mixture was stirred further at the same tempera-
ture for 4 h. Gaseous ammonia was bubbled through
the solution, and 55 ml of methanol, 35 ml of ammo-
nia-saturated aqueous ammonium chloride, and 70 ml
of hexane were added successively. After filtration with
Celite, the organic solvents were evaporated. The
residue was extracted with hexane (50 ml×3), and
subsequent drying (MgSO4) and evaporation gave 13.7
g (67%) of Bu3SnCHꢁCHCH2OH.
native path may be reductive elimination of
a
diorganopalladium complex (6) formed by the attack of
Cl− at 5.
In conclusion, we succeeded in preparing 1-stannyl-
substituted allyl–palladium complexes which served as
a formal source of nucleophilic vinylcarbene–palladium
intermediate.
3.3. Preparation of 1-tributylstannyl-3-chloropropene
A THF solution (30 ml) of the 1-tributylstannyl-hy-
droxypropene (13.7 g, 39.5 mmol), PPh3 (14.5 g, 55.3
mmol) and 83 ml of CCl4 was heated at 70°C for 2 h.
Pentane (450 ml) was added and the mixture filtered.
The solvents were evaporated under vacuum. After
addition of another pentane (200 ml), filtration, and
evaporation, the residue was dissolved in hexane.
Column chromatography (Wako C-200) of the hexane
solution gave 9.5 g (65%) of the stannyl–chloro-
propene. 1H-NMR (CDCl3): l 0.84–1.57 (m, 27H),
3.97 (d, J=7.3, JSn=7.0 Hz, 2H), 6.16 (d, J=12.2,
3. Experimental
3.1. General procedures and measurements
Most of the commercially available reagents were
used without further purification. All reactions and
manipulations of air- and moisture-sensitive com-
pounds were carried out under an atmosphere of dry
Ar by use of standard vacuum line techniques. NMR
spectra were obtained on JEOL GSX-270 and JEOL
GSX-400 spectrometers. Chemical shifts are given in
ppm using TMS or H3PO4 as standard.
JSn=57.3 Hz, 1H), 6.64 (dt, J=12.2, 7.3, JSn=72.4
Hz, 1H). Anal. Calc. for C15H31ClSn: C, 49.29; H, 8.55.
Found: C, 49.17; H, 8.37%.

T. Nishida et al. / Journal of Organometallic Chemistry 625 (2001) 54–57
57
3.4. Preparation of Pd(p3-Bu3SnCHCHCH2)(Cl)(PPh3)
(1)
50 mg (0.083 mmol) of Pd2Cl(PPh3)(m-Cl)(m-C4H8) [12],
and then 22 mg of PPh3 (0.083 mmol) added. Filtration
and evaporation of the solvent gave dark yellow solids
(30 mg, 41%). Recrystallization from CH2Cl2–hexane
gave pale yellow solids of (CH2CHCHCH-
CHCH2){PdCl(PPh3)}2·1/2CH2Cl2. 31P-NMR (CDCl3):
l 24.02 (s). Anal. Calc. for C42.5H39Cl3P2Pd2: C, 54.84;
H, 4.22. Found: C, 54.38; H, 4.29%.
To a CH2Cl2 solution (5 ml) of Pd2(dba)3 (414 mg,
0.399 mmol) and PPh3 (212 mg, 0.808 mmol) was
added Bu3SnCHꢁCHCH2Cl (300 mg, 0.822 mmol) at
r.t. The solution color became deep green after 10 min
stirring. The mixture was purified by column chro-
matography (alumina, CH2Cl2) to give a yellow eluent.
Evaporation of the solvent, washing with pentane, and
filtration gave a yellow solution. Recrystallization from
CH2Cl2–pentane gave crystalline materials of 1 (125
mg, 21%). 1H-NMR (CD2Cl2): l 0.8–1.6 (m, 27H), 2.63
(bd, J=10.8 Hz, 1H), 3.05 (dd, J=7.5, 2.6 Hz, 1H),
5.06 (m, 1H), 6.27 (m, 1H), 7.45 (m, 9H), 7.6 (m, 6H).
31P-NMR (CD2Cl2): l 21.68 (s, JSn=44 Hz). Anal.
Calc. for C35H46ClPPdSn: C, 53.98; H, 6.31. Found: C,
54.08; H, 6.28%.
Acknowledgements
Partial support of this work through Grants-in-Aid
for Scientific Research from the Ministry of Education,
Science and Culture, Japan is gratefully acknowledged.
References
[1] J.P. Collman, L.S. Hegedus, J.R. Norton, R.G. Finke, Principles
and Applications of Organotransition Metal Chemistry, Univer-
sity Science Books, Mill Valley, CA, 1987, p. 127.
[2] M.A. Gallop, W.R. Roper, Adv. Organomet. Chem. 25 (1986)
121.
3.5. Reaction of 1 with PPh3 and diethyl malonate
To a CD2Cl2 solution (0.6 ml) of 1 (10 mg, 0.014
mmol) and PPh3 (10 mg, 0.039 mmol) in an NMR tube
was added diethyl malonate (2.2 mg, 0.014 mmol). The
[3] S. Watanabe, S. Ogoshi, K. Kakiuchi, H. Kurosawa, J.
Organomet. Chem. 481 (1994) 19.
1
mixture was kept at r.t. for 46 h. H-NMR measure-
[4] (a) S. Watanabe, H. Kurosawa, Organometallics 16 (1997) 3601.
(b) S. Watanabe, H. Kurosawa, Organometallics 17 (1998) 479.
[5] B.M. Trost, C.R. Self, J. Am. Chem. Soc. 105 (1983) 5942.
[6] S. Ogoshi, W. Yoshida, K. Ohe, S. Murai, Organometallics 12
(1993) 578.
ments indicated formation of diethyl 2-allylmalonate in
40% yield.
3.6. Reaction of 1 with PPh3 and acetic acid
[7] The
silylallylpalladium
complexes
[PdCl{h3-Me3Si(Me)-
CCHCH2}]2 and PdCl{h3-Me3Si(Me)CCHCH2}(PPh3) showed
the anti-Me3Si preference in solution (18/82 and 13/87, respec-
tively). T. Ohta, T. Hosokawa, S. Murahashi, Organometallics 4
(1985) 2080.
To a CD2Cl2 solution (0.6 ml) of 1 (9.0 mg, 0.013
mmol) and PPh3 (3.4 mg, 0.013 mmol) in an NMR tube
was added 1 ml of acetic acid. The mixture was kept
at r.t. for 40 h. 1H-NMR measurements confirmed
almost quantitative formation of Pd(h3-CH2CHCH2)ꢀ
(Cl)(PPh3).
[8] H. Kurosawa, A. Urabe, K. Miki, N. Kasai, Organometallics 5
(1986) 2002.
[9] JPP values for [Pd(h3-allyl)(dppe)]+ were reported to be ca. 40
Hz. R. Malet, M.M. Menas, T. Parella, R. Pleixats,
Organometallics 14 (1995) 2463.
3.7. Reaction of 1 with PPh3 and PdCl2(PhCN)2
[10] It is also feasible that the destannylation occurs synchronously
with the attack of proton at certain bisphosphine coordinated
intermediates such as 2, or shorter-lived h1-Bu3SnCHꢁ
CHCH2ꢀPd or h1-CH2ꢁCHCH(SnBu3)ꢀPd bonded complexes.
[11] (a) S. Ogoshi, K. Tsutsumi, M. Ooi, H. Kurosawa, J. Am.
Chem. Soc. 117 (1995) 10415. (b) S. Ogoshi, K. Tsutsumi, T.
Shinagawa, K. Kakiuchi, H. Kurosawa, Chem. Lett. (1999) 123.
[12] T. Murahashi, N. Kanehisa, Y. Kai, H. Kurosawa, J. Chem.
Soc. Chem. Commun. (1996) 825.
To a CD2Cl2 solution (0.6 ml) of 1 (9.0 mg, 0.013
mmol) and PPh3 (3.5 mg, 0.013 mmol) in an NMR tube
was added PdCl2(PhCN)2 (2.7 mg, 0.007 mmol). The
solution color changed from pale yellow to orange, and
1
then dark orange. H-NMR measurements after 40 h
showed a set of four resonances as the major product at
l 2.98 (br d, J=12 Hz, 1H), 3.11 (br d, J=7 Hz, 1H),
4.69 (m, 1H), 5.78 (m, 1H). These data were identical
with those of a sample prepared separately. Thus, hexa-
triene (9 ml) was added to a CH2Cl2 solution (10 ml) of
[13] The observed weak NOE correlation between the peaks at l 4.69
and 3.11 (syn-H at C1) (B1%) would have been caused by
moderately fast exchange between anti- and syn-H at C1 via
h1-allyl intermediate(s).
[14] E.J. Corey, T.M. Echrich, Tetrahedron Lett. 25 (1984) 2419.
.