A new oxapalladacycle generated via ortho C-H activation of phenylphosphinic acid: An efficient catalyst for Markovnikov-type additions of E-H bonds to alkynes

Article abstract of DOI:10.1039/c0cc03436c

A new oxapalladacycle 3 can be conveniently prepared via direct ortho palladation of diphenylphosphinic acid with palladium acetate. Catalysts derived from 3 can efficiently catalyze Markovnikov-type additions of E-H bonds (P(O)-H, S-H and spC-H) to alkyn

Full text of DOI:10.1039/c0cc03436c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 2333–2335
www.rsc.org/chemcomm
COMMUNICATION
A new oxapalladacycle generated via ortho C–H activation of
phenylphosphinic acid: an efficient catalyst for Markovnikov-type
additions of E–H bonds to alkynesw
Qing Xu,ab Ruwei Shen,a Yutaka Ono,a Ritsuko Nagahata,a Shigeru Shimada,a Midori Gotoa
and Li-Biao Han*a
Received 24th August 2010, Accepted 26th November 2010
DOI: 10.1039/c0cc03436c
A new oxapalladacycle 3 can be conveniently prepared via direct
ortho palladation of diphenylphosphinic acid with palladium
acetate. Catalysts derived from 3 can efficiently catalyze
Markovnikov-type additions of E–H bonds (P(O)–H, S–H and
spC–H) to alkynes via a unique catalytic cycle.
Herein we disclose the first direct ortho C–H activation of
phenylphosphinic acid Ph2P(O)OH with Pd(OAc)2 to produce
high yields of a stable oxapalladacycle 3 (eqn (2)).4 In addition,
we also found that complexes of 3 are efficient catalysts
superior to common ones for catalytic additions of several
E–H bonds to alkynes (vide infra). To our knowledge, 3
represents a rare example of stable oxapalladacycles without
bearing external ligands prepared via direct ortho palladation
processes.1–3 Although oxapalladacycles 2a (3’s analogue)
were proposed to be intermediates in Pd-catalyzed
functionalizations of aryl carboxylic acids,2 these complexes
could not be generated by a similar ortho palladation
of the corresponding carboxylic acids.5 In addition, only
ligand-stabilized 2a could be isolated.2,3
Palladacycles are novel catalysts in homogeneous catalysis,
active intermediates in C–H activations and cascade
transformations.1 They have also attracted attention in
medicinal chemistry.1a Oxapalladacycles (C,O-palladacycles)
2 bearing a C–Pd and an O–Pd bond are intermediates
related to catalytic C–O coupling reactions and Pd-catalyzed
functionalizations of aryl carboxylic acids.1,2 Although
C-palladacycles 20 are easily generated by direct ortho C–H
activation of benzene rings directed by the chelating functional
group L,1 C,O-palladacycles 2 are usually prepared through
tedious procedures.3 Preparation of 2 via a direct C–H
activation of 1 seems to have not been achieved (eqn (1)).2,3
Heating a mixture of Pd(OAc)2 (1.123 g, 5 mmol) and
Ph2P(O)OH (1.64 g, 7.5 mmol) in THF (50 mL) at 60 1C
resulted in rapid formation of 3 as grey precipitates. Collecting
the precipitates by filtration, washing with THF and drying
under vacuum gave 1.38 g of 3 (86% yield). Recrystallization
of the solid from dichloromethane gave crystals suitable for
X-ray analysis which leads to the determination of its structure
unambiguously (Fig. 1).6
As to the structure of complex 3, we found that in the solid
state, six 3 aggregate to form a novel hollow cylinder-like
shape bearing approximate D3 symmetry.4 Three monomer
ð1Þ
ð2Þ
a National Institute ofAdvanced Industrial Science and Technology
(AIST), Tsukuba, Ibaraki 305-8565, Japan.
E-mail: libiao-han@aist.go.jp; Fax: +81 298-61-6344;
Tel: + 81 298-61-4855
Fig. 1 Molecular structure of 3 in solid state (hydrogen atoms
omitted for clarity. Right: view from top). Selected bond lengths (A)
and angles (1): Pd1–O1 = 2.071(5), Pd1–O3 = 2.169(6), Pd1–O6 =
2.047(5), Pd1–C1 = 1.947(8), P1–O1 = 1.537(6), P1–O2 = 1.514(6),
P1–C6 = 1.782(9), O1–Pd1–C1 = 87.4(3), Pd1–O1–P1 = 115.6(3),
Pd1–C1–C6 = 115.9(6), P1–C6–C1 = 116.7(7).
b College of Chemistry and Materials Engineering,
Wenzhou University, Wenzhou 325035, China
w Electronic supplementary information (ESI) available: Experimental
procedures and characterization of new compounds. CCDC 773371
and 773372. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/c0cc03436c
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 2333–2335 2333

View Article Online
Scheme 1 A proposed reaction mechanism.
generate 6 followed by a protonolysis to give the adduct and
regenerate 4 (Scheme 1).
Fig. 2 Molecular structure of 5a (hydrogen atoms omitted for
clarity). Selected bond lengths (A) and angles (1): Pd–P1 = 2.3320(5),
Pd–P2 = 2.3129(5), Pd–P3 = 2.3201(5), Pd–C = 2.0609(17),
P3–O1 = 1.5530(14), P4–O2 = 1.5423(14), P4–O3 = 1.4921(14),
P1–Pd–P2 = 84.044(17), P2–Pd–P3 = 172.173(17), P3–Pd–P1 =
103.349(17).
ð4Þ
units are crystallographically independent and bonded with
each other through three Pd–O bonds using the Pd atoms and
PQO oxygen atoms forming the upper circle. The upper and
lower circles are connected through six Pd–O bonds using the
Pd atoms and P–O oxygen atoms. The distances between
Pd1–Pd2, Pd1–Pd3, and Pd2–Pd3 are 4.37, 4.39, and 4.07 A,
respectively.
The phenomenon that the C–Pd bond of 3 is retained during
the catalytic cycle is noticeable considering it is well recognized
that in palladacycle-involved catalytic reactions so far
reported, palladacycle complexes are usually not the real
catalysts but precursors to zerovalent palladium generated
by the breakdown of the C–Pd bonds of the complexes.1
Complexes derived from oxapalladacycle
3 are more
As shown in eqn (3), mixing 3 and 1 equiv. of a
bidentate phosphine dppe (Ph2P(CH2)2PPh2) or dmpe
(Me2P(CH2)2PMe2) quantitatively gave the corresponding
phosphine-chelating oxapalladacycles 4. Interestingly,
complexes 4 reacted smoothly at room temperature with
Ph2P(O)H to give complex 5 via the cleavage of the
Pd–O bond, which was unambiguously confirmed by
X-ray analysis (Fig. 2).7 The bond length of O2–P4 (POH) is
substantially longer than that of O3–P4 (PQO), but is similar
to that of O1–P3, reflecting a hydrogen bond between O1
and O2 (O2–H–O1), which was also observed in similar
complexes.8
efficient and selective catalysts for the additions of a few
E–H bonds to alkynes than common palladium complexes
used so far (Table 1).9 Thus, 3 could catalyze the addition of
Ph2P(O)H to not only linear 1-octyne but also to the bulky
tert-butylacetylene, which only gave poor results so far by
other palladium catalysts,8,10 to give the Markovnikov
adducts in high yields and high regioselectivities (runs 1 and 2).
The addition of Ph2P(O)H to diphenylacetylene also gave the
Table 1 Complexes of palladacycle 3-catalyzed highly selective
addition of E–H to alkynesa
Temp./1C, Yield (%),
time/h
Run E–H
Alkyne
sele
1b
Ph2P(O)H
1-Octyne
t-BuCRCH 70, 10
PhCRCPh
1-Octyne
70, 3
99, 98/2
97, 99/1
99, —
2c
ð3Þ
3b
70, 20
110, 5
4b
Me2P(O)H
99, 98/2
96, 95/5
97, 96/4
96, 98/2
89, 96/4
95, 98/2
96, 97/3
97, 99/1
95, 94/6
95, 94/6
96, 95/5
98, 97/3
5c
t-BuCRCH 110, 25
1-Octyne 110, 3
t-BuCRCH 110, 17
1-Octyne 110, 2
t-BuCRCH 110, 25
(Pinacolato)P(O)H 1-Octyne 70, 3
t-BuCRCH 70, 19
Both complexes 4b and 5b can be clearly observed in the
catalytic addition of Ph2P(O)H to 1-octyne. Thus, a mixture
of 4b (0.015 mmol), Ph2P(O)H (0.1 mmol) and 1-octyne
(0.11 mmol) at room temperature afforded 5b quantitatively
(eqn (4)). Allowing the solution to stand overnight at room
temperature, a trace amount of the adduct was obtained. The
possible alkenylpalladium intermediates 6 were not observed,
perhaps due to a quick collapse of this possible intermediate
under current conditions. Upon heating the mixture at 70 1C
overnight resulted in a complete consumption of Ph2P(O)H to
give the adduct in 99% yield with 99/1 selectivity, while 4b was
completely recovered (0.015 mmol). Therefore, the reaction
could be tentatively rationalized via addition of 5 to alkyne to
6b
Ph(EtO)P(O)H
(MeO)2P(O)H
7c
8b
9c
10b
11c
12d
13b
14b
15c
(Pinacolato)2PH
PhSH
1-Octyne
1-Octyne
1-Octyne
1-Octyne
70, 6
70, 16
110, 25
t-BuCRCH
t-BuCRCH 70, 19
a
E–H, alkyne (1.1 equiv.), 5 mol% 3 and a phosphine ligand
b
in toluene were heated in a glass tube. L = Ph2P(CH2)3PPh2.
c
d
L = Ph2P(CH2)2PPh2. In the absence of a phosphine ligand.
e
Yield based on E–H. Regioselectivity was determined by NMR.
c
2334 Chem. Commun., 2011, 47, 2333–2335
This journal is The Royal Society of Chemistry 2011

View Article Online
adduct in high yield (run 3). As thoroughly investigated, 3 is a
rather general catalyst for the addition of a variety of
H-heteroatom compounds to alkynes to produce the
Markovnikov adducts highly selectively.10–15 Thus, all other
P(V)H compounds, a H-phosphinate (runs 6 and 7),10d,11
H-phosphonates (runs 8–11),10a,12 and a H-spirophosphorane
(run 12)13 all added to terminal alkynes to give the
Markovnikov adducts selectively in high yields. In addition,
3 also catalyzed the addition of PhSH to 1-octyne (run 13) to
give the Markovnikov adduct in 95% yield with 94%
selectivity.14 Interestingly, 3 can also mediate the dimerization
of terminal alkynes such as 1-octyne and tert-butylacetylene to
give the homoaddition products in 96% and 98% yields,
respectively, with regioselectivities over 95% (runs 14 and 15).15
As noted above, as a catalyst, a complex of 3 is superior to
the common palladium complexes for the additions. For
example, the additions of Me2P(O)H, (MeO)2P(O)H and
(pinacolato)P(O)H to tert-butylacetylene using Pd(OAc)2/
phosphine as catalysts did not or only sluggishly took place
to give the corresponding adducts in poor selectivities.10
In summary, a new stable oxapalladacycle 3 was prepared
for the first time via a simple direct ortho palladation process
of phenylphosphinic acid with palladium acetate. Complexes
derived from 3 are new catalysts for the selective addition of a
variety of H-heteroatom compounds to alkynes. A remarkable
feature of 3 is that its C–Pd bond retains during the catalytic
process which is rare for palladacycle catalysts. Further
applications of 3 as a catalyst and catalytic transformation
of phosphinic acid based on C–H activation are in progress.
This work was partly supported by New Energy and
Industrial Technology Development Organization (NEDO)
of Japan (Industrial Technology Research Grant Program in
2004). Q.X. thanks the National Natural Science Foundation
of China (No. 20902070).
5 Under similar conditions, benzoic acid did not react with Pd(OAc)2
to give 2a. No C–H activation took place either with Ph2PO2Et or
Ph3P(O), indicating that the PO2H functional is essential for this
ortho palladation.
6 CCDC 773371 contains the supplementary crystallographic
data for compound 3. Crystal data for 3: C72H54O12P6Pd6,
M
=
1935.45, orthorhombic,
a
=
20.7526(17) A,
b = 17.8092(14) A, c = 18.6170(15) A, a = 90.00001, b =
90.00001, g = 90.00001, V = 6880.6(10) A3, T = 153(2) K,
space group Pccn, Z = 4, m(Mo Ka) = 1.738 mmꢁ1, 45 474
reflections measured, 8562 independent reflections (Rint = 0.0618).
The final R1 values were 0.0611 (I
wR(F2) values were 0.1505 (all data). The goodness of fit on F2
was 1.258.
> 2s(I)). The final
7 CCDC 773372 contains the supplementary crystallographic data
for compound 5a. Crystal data for 5a: C30H36O3P4PdꢀCH2Cl2,
M = 759.79, monoclinic, a = 14.8490(9) A, b = 11.5995(7) A,
c = 19.2145(12) A, a = 90.001, b = 93.2570(10)1, g = 90.001,
V = 3304.2(3) A3, T = 183(2) K, space group P2(1)/c, Z = 4,
m(MoKa) = 0.948 mmꢁ1, 20 039 reflections measured, 7503
independent reflections (Rint = 0.0194). The final R1 values were
0.0266 (I
>
2s(I)). The final wR(F2) values were 0.0690
(I > 2s(I)). The final R1 values were 0.0292 (all data). The final
wR(F2) values were 0.0707 (all data). The goodness of fit on F2 was
1.035.
8 L.-B. Han, N. Choi and M. Tanaka, Organometallics, 1996, 15,
3259.
9 For reviews: (a) L.-B. Han and M. Tanaka, Chem. Commun., 1999,
395; (b) O. Delacroix and A. C. Gaumont, Curr. Org. Chem., 2005,
9, 1851; (c) S. Greenberg and D. W. Stephan, Chem. Soc. Rev.,
2008, 37, 1482; (d) I. P. Beletskaya and M. M. Kabachnik,
Mendeleev Commun., 2008, 18, 113; (e) L. Coundray and
J.-L. Montchamp, Eur. J. Org. Chem., 2008, 3601;
(f) I. P. Beletskaya and C. Moberg, Chem. Rev., 2006, 106, 2320;
(g) Q. Xu and L.-B. Han, J. Organomet.Chem., 2011, 696, 130.
10 The catalyst Pd(OAc)2/phosphine did not catalyze the addition of
Me2P(O)H to alkynes. Whereas the so far reported metal-catalyzed
P(O)–H addition reactions to alkynes generating the anti-
Markovnikov regioisomer are general and the selectivity is
excellent, reactions giving the Markovnikov adducts are usually
accompanied by the anti-Markovnikov regioisomers which can
even become the major product in the case of bulky substrates.
Thus, the PdMe2(PPh2Me)2 catalyzed addition of (RO)2P(O)H to
tert-butylacetylene only proceeded sluggishly to give 42% yield of
the adducts with 36% regioselectivity to the Markovnikov product
(ref. 10a).
Ph3P/CF3CO2H (ref. 10b) too.
A
poor selectivity was obtained with Pd2(dba)3/
similar phenomenon was
Notes and references
A
1 (a) J. Dupont and M. Pfeffer, Palladacycles: Synthesis,
Characterization and Applications, Wiley-VCH, Weinheim, 2008;
(b) I. P. Beletskaya and A. V. Cheprakov, J. Organomet. Chem.,
2004, 689, 4055; (c) J. Dupont, C. S. Consorti and J. Spencer,
Chem. Rev., 2005, 105, 2527; (d) M. P. Munoz, B. Martin-Matute,
C. Fernandez-Rivas, D. Cardenas and A. M. Echavarren, Adv.
Synth. Catal., 2001, 343, 338.
2 (a) M. Miura, T. Tsuda, T. Satoh, S. Pivsa-Art and M. Nomura,
J. Org. Chem., 1998, 63, 5211; (b) R. Giri and J.-Q. Yu, J. Am.
Chem. Soc., 2008, 130, 14082.
3 (a) K. Nagayama, F. Kawataka, M. Sakamoto, I. Shimizu and
A. Yamamoto, Bull. Chem. Soc. Jpn., 1999, 72, 573;
(b) C. Fernandez-Rivas, D. J. Cardenas, B. Martin-Matute,
A. Monge, E. Gutierrez-Puebla and A. M. Echavarren,
Organometallics, 2001, 20, 2998; (c) G. Lu and
H. C. Malinakova, J. Org. Chem., 2004, 69, 8266; (d) R. Giri
and J.-Q. Yu, J. Am. Chem. Soc., 2008, 130, 14082.
4 To our knowledge, 3 is also the first isolated palladacycle directly
generated from a simple organoheteroatom acid Ph2P(O)OH via a
mild and efficient ortho-cyclometallation. No decomposition of 3 in
CD2Cl2 was observed at 80 1C for 5 days. Only one 31P signal was
observed in CD2Cl2 (d 69.6) and in CDCl3 with 1% DMSO
(d 47.4). In CD2Cl2, ESI-TOF MS detected a hexamer of 3 as
shown in Fig. 1, while a monomeric 3 (3ꢀ2DMSO) was detected as
the major species in CDCl3 with 1% DMSO.
observed for the addition of Ph2P(O)H, that the addition of
Ph2P(O)H to tert-butylacetylene only gave the Markovnikov
adduct in 60% selectivity (ref. 10c). (a) L.-B. Han and
M. Tanaka, J. Am. Chem. Soc., 1996, 118, 1571;
(b) V. P. Ananikov, L. L. Khemchyan and I. P. Beletskaya, Synlett,
2009, 2375; (c) L.-B. Han, R. Hua and M. Tanaka, Angew. Chem.,
Int. Ed., 1998, 37, 94; (d) L.-B. Han, C.-Q. Zhao, S. Onozawa,
M. Goto and M. Tanaka, J. Am. Chem. Soc., 2002, 124, 3842;
(e) N. Dobashi, K. Fuse, T. Hoshino, J. Kanada, T. Kashiwabara,
C. Kobata, S. K. Nune and M. Tanaka, Tetrahedron Lett., 2007,
48, 4669.
11 (a) L.-B. Han, C.-Q. Zhao and M. Tanaka, JP Patent, 3877151,
2006; (b) S. K. Nune and M. Tanaka, Chem. Commun., 2007, 2858.
12 (a) L.-B. Han, 225th National Meeting of the American Chemical
Society, Mar. 23–27, 2003, New Orleans, Louisiana; Abstracts of
Papers of The American Chemical Society, 225 (2003) U148;
(b) M. Tanaka and L.-B. Han, JP Patent, 3041396, 2000;
(c) M. Tanaka and L.-B. Han, JP Patent, 3007984, 2000.
13 L.-B. Han, Y. Ono, Q. Xu and S. Shimada, Bull. Chem. Soc. Jpn.,
2010, 83, 1086.
14 (a) A. Ogawa, J. Organomet. Chem., 2000, 611, 463;
(b) H. Kuniyasu, A. Ogawa, K. Sato, I. Ryu, N. Kambe and
N. Sonoda, J. Am. Chem. Soc., 1992, 114, 5902.
15 B. M. Trost, M. T. Sorum, C. Chan, A. E. Harms and G. Ruhter,
¨
J. Am. Chem. Soc., 1997, 119, 698.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 2333–2335 2335