Synthesis of lanthanide(II)-imine complexes and their use in carbon-carbon and carbon-nitrogen unsaturated bond transformation

Article abstract of DOI:10.1016/j.tet.2003.06.003

Ytterbium and samarium metals reduced aromatic ketimines to give directly divalent azalanthanacyclopropane complexes 1 quantitatively, the structure of which was characterized by X-ray analysis. The imine complexes 1 catalyzed dehydrogenative silylation of terminal alkynes, hydrosilylation of imines and alkenes, and intermolecular hydrophosphination of alkynes. Moreover, dehydrogenative double silylation of conjugated dienes was achieved with 1.

Full text of DOI:10.1016/j.tet.2003.06.003

Tetrahedron 59 (2003) 10381–10395
Synthesis of lanthanide(II)–imine complexes and their use in
carbon–carbon and carbon–nitrogen unsaturated bond
transformation
*
Ken Takaki, Kimihiro Komeyama and Katsuomi Takehira
Department of Chemistry and Chemical Engineering, Graduate School of Engineering, Hiroshima University, Kagamiyama, Higashi-
Hiroshima 739-8527, Japan
Received 19 April 2003; accepted 23 June 2003
Abstract—Ytterbium and samarium metals reduced aromatic ketimines to give directly divalent azalanthanacyclopropane complexes 1
quantitatively, the structure of which was characterized by X-ray analysis. The imine complexes 1 catalyzed dehydrogenative silylation of
terminal alkynes, hydrosilylation of imines and alkenes, and intermolecular hydrophosphination of alkynes. Moreover, dehydrogenative
double silylation of conjugated dienes was achieved with 1.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
temperature for 4 h, the metal was mostly consumed, and a
reddish-black solution was formed. On quenching the
mixture with D2O, deuterated amine 2-d2 was formed
quantitatively (Eq. (1)). In contrast, aromatic aldimines
were changed to pinacol-type products, 1,2-diaminoethanes,
by treatment with Yb.7 Similar results were obtained by
using Sm metal. Deuteration of 2 suggested the formation of
dianion species 1 through two-electron reduction. Fortu-
nately, the Yb(II)–imine complex 1a (Ln¼Yb, Ar¼Ph,
n¼3) was isolated as red–black needles in 45% yield, which
is not only the first azalanthanacyclopropane, but also the
first example of direct synthesis of azametallacycles by the
reduction of imines with low-valent metallic reductants.
Rich chemistry has been explored in organic synthesis by
using lanthanides in the last two decades.1 Of the many
lanthanide reagents, we were interested in the reducing
ability of low-valent lanthanides, and studied various
coupling reactions promoted by ytterbium and samarium
metals,2 and selective C–O bond cleavage of allylic,
propargylic, and vinylic ethers with divalent samarocenes
leading to the corresponding reactive intermediates.3 More-
over, we found that aromatic ketones and thioketones were
readily reduced to a dianion species that reacted with many
electrophiles at the carbonyl-4 and thiocarbonyl-carbon,5
respectively. Aromatic ketimines were similarly reduced by
two-electron transfer from the metals, but the resulting
Ln(II)–imine complexes exhibited very different behavior
from the ketone and thioketone complexes. In this article,
we would like to disclose the synthesis and chemical
properties of the imine complexes and their catalyst
activities in some transformations of carbon–carbon and
carbon–nitrogen unsaturated bonds.
ð1Þ
Figure 1 shows the structure of 1a obtained by the X-ray
crystallographic study.8 The complex is five-coordinate and
monomeric in contrast to the corresponding dimeric
ytterbium(II)–benzophenone complex 3.4 The sum of the
angles formed by N(10), C(2), and C(8) around C(1) is over
3598, which is larger than that of 3, suggesting that C(1) is
still sp2-hybridized. The sp2 character of C(1) is consistent
with the Yb–C(1)–N(10) angle (87.8(10)8), which is nearly
orthogonal to the C(1)–N(10) bond. The C(1)–N(10) bond
2. Synthesis, structure and chemical properties of
Ln(II)–imine complexes.6
When aromatic ketimines were treated with equimolar
amounts of ytterbium metal in THF–HMPA solvent at room
˚
distance of the imine moiety in the complex 1a (1.43(3)A)
C
Keywords: azalanthanacyclopropane; lanthanide catalyst; hydrosilylation;
dehydrogenative silylation; hydrophosphination.
can be considered as a C–N single bond. In the 1H and 13
NMR spectra of 1a, all signals of the three phenyl groups
and the imine carbon are shifted to upfield and appear
*
Corresponding author. Tel.: þ81-824-24-7738; fax: þ81-824-24-5494;
e-mail: ktakaki@hiroshima-u.ac.jp
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.06.003

10382
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
corresponding benzophenone and thiobenzophenone com-
plexes 3 and 4 in spite of the similar dianion complexes: less
nucleophilic and more basic.
ð2Þ
ð3Þ
ð4Þ
ð5Þ
Based on these properties of the imine complexes 1, we
intended to use them in catalytic reactions (Scheme 1). The
basic character of 1 could abstract acidic protons of the
substrate H–X like alkyl ligands of the well-documented
Figure 1. ORTEP drawing of 1a.
trivalent lanthanocene Cp*Ln–R (Cp*¼C5Me5) to generate
2
without broadening, as expected for a divalent species.9 In
addition, rotation of the phenyl ring on the nitrogen should
be ceased or very slow even in the solution, because their
five protons and six carbons are independently observed.
the ring-opened active intermediate A having an amide
ligand. Further reaction of A with the substrate as well as
disproportionation of A would produce homoleptic inter-
mediate B. Once the [Yb]–X species A and B are generated,
various catalytic reactions could be expected. For example,
A or B would react with H–Y to give the coupling product
X–Y and [Yb]–H, and the catalyst [Yb]–X could be
regenerated with hydrogen evolution (cycle A). If insertion
of olefin to A or B is possible, catalytic addition of H–X to
olefin would be realized (cycle B).
Chemical properties of the Yb–imine complexes 1 were
compared with those of the corresponding Yb–benzo-
phenone 34 and Yb–thiobenzophenone complexes 4.5
These complexes 1, 3, and 4 are extremely air- and
moisture-sensitive. Exposure of 1 and 3 to air resulted in
the nearly quantitative recovery of the imine and benzo-
phenone, respectively. As mentioned above, the complexes
3 and 4 were umpoled from electrophilic to nucleophilic
and, thus, they reacted with many electrophiles: for
example, the reaction with acetone gave the addition
products in good yields (Eq. (2)). In contrast, the imine
complex 1a abstracted a-D of acetone-d6 to yield deuterated
amine, Ph2CDNHPh, after quenching with H2O and no
addition product was obtained. Exceptionally, the imine
complexes 1 reacted with CO2 and isocyanates to afford
a-amino acid10 and a-aminoacetamide derivatives11 in high
yields, respectively (Eq. (3)). The complexes 1 served as
unique base catalysts in the isomerization of 1-alkynes to
2-alkynes, wherein no by-products such as allenes and
dienes were contaminated, in contrast to conventional bases
(Eq. (4)).12 The basic feature of the imine complex was also
observed on treatment of p-tolualdehyde with 10 mol% of
1f (Ar¼2,6-iPr2C6H3, n¼0) in toluene to give the
Tishchenko product in 56% yield along with pinacol (6%)
and alcohol (10%) (Eq. (5)). The stoichiometric reaction
gave selectively the pinacol in 80% yield, indicating the
moderate reducing ability of 1f. Moreover, while the
benzophenone and thiobenzophenone complexes 3 and 4
reacted with 2,6-xylyl isocyanide to yield the coupling
products through insertion of the isocyanide to the C–Yb
bond of 3 and 4,13 no reaction occurred in the case of 1.
Thus, it can be concluded that chemical properties of the
imine complexes 1 are very different from those of the
Scheme 1.
Recently, lanthanocene complexes have attracted much
attention in organic synthesis because of their efficient and
powerful catalyst activities in alkyne and alkene transform-
ations such as hydrosilylation, hydroamination, hydro-
phosphination, and dimerization.1 Compared with late
transition metal catalysts, the synthesis, stocking, and
handling of the lanthanocenes are, however, very difficult
and, thus, their use seems to be limited actually. The imine
complexes 1 can be easily and almost quantitatively

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
10383
generated in situ from imines and Yb or Sm metal by stirring
at room temperature. In addition, non-Cp* complexes 1 may
provide different chemistry from that of the lanthanocenes,
because the Cp* ligand has been known to control the
selectivities of the reaction by the steric and electronic
environment around the metal center. On the other hand,
catalyst activity of 1 may be decreased, since coordinative
unsaturation of the lanthanide metals, the origin of high
activities in the case of the lanthanocene, would be
diminished by the HMPA ligand and THF solvent that are
necessary for solvation and stabilization of 1. With such
consideration in mind, we investigated the catalytic
activities of the imine complexes 1.14
of alkynide.19 The imine complexes 1 were found to serve as
good catalysts for the present reaction.
When triphenylsilane and 1-hexyne were successively
added to a THF solution of 10 mol% of 1a, 1-triphenyl-
silyl-1-hexyne (5f) was formed in 82% yield along with
hydrogen evolution, and the starting silane and alkyne were
recovered in 14 and 12% yields, respectively. But reduced
products like hexene and silylhexenes were not produced.
Moreover, the isomerization of 1-alkyne to 2-hexyne was
suppressed in the presence of the silane.12 The reactivity of
various monohydrosilanes was tested under the same
conditions (Table 1). Trialkylsilanes and alkoxysilane
gave poor yields of the products, and no alkynylsilane was
obtained in the reaction with trichlorosilane, which led to an
untractable mixture instead. Diphenylmethylsilane and
triphenylsilane gave satisfactory results (runs 5 and 6).
3. Dehydrogenative silylation of terminal alkynes.15
Although alkynylsilanes are readily obtained from metal
alkynides and silyl halides conventionally, the process via
catalytic dehydrogenative silylation of terminal alkynes is
expected to be advantageous if a methodology can be
designed that avoids the stoichiometric use of a base under
halogen-free conditions, particularly for the preparation of
polyalkynylsilyl compounds.16 If lanthanocene complexes
are used in this process, considerable problems would arise
from the side reactions such as oligomerization of hydro-
silanes17 and terminal alkynes,18 and the coupling reaction
The activity of the imine complexes, [Yb(h2-Ph2-
CNR)(hmpa)4] (1), was greatly influenced by the substituent
groups on the imine nitrogen. Generally, electron-with-
drawing groups increased the activity, and, in contrast,
electron-donating and bulky substituents decreased it. For
example, the reaction of 1-hexyne with dimethylphenyl-
silane showed the following order of decreasing yield of the
product 5d: R¼4-FC6H4 (1c) (51%).Ph (1a) (31%).4-
MeOC6H4 (1d), 2,6-Me2C6H3 (1e) (23%).2,6-iPr2C6H3
(1f) (17%).H (1g) (7%).SiMe3 (1h) (4%). In addition, the
silylation did not take place with the imine complexes 1
prepared in the absence of HMPA.
Table 1. Silylation of 1-hexyne with monohydrosilanes
Silylation of 1-hexyne with di- and trihydrosilanes took
place similarly (Table 2). In the reaction with dihydro-
silanes, product distribution was mainly determined by the
ratio of 1-hexyne to the silanes. Thus, when the ratio was
1/2, monohexynylsilanes 6a and 6c were obtained in about
five-fold excess of dihexynylsilanes 6b and 6d, respectively
(runs 1 and 4). On the contrary, use of 2 equiv. of 1-hexyne
resulted in the predominant formation of 6b and 6d (runs 3
and 6). In the case of phenylsilane, reaction of an equimolar
mixture of the two substrates gave the alkynylsilanes 6e–g
in low yields (run 7), which was probably due to
Run
Hydrosilane
Product
Yield (%)a
1
2
3
4
5
6
7
Et3SiH
5a
5b
5c
5d
5e
5f
7
tBuMe2SiH
(EtO)3SiH
PhMe2SiH
Ph2MeSiH
Ph3SiH
Trace
38
31 (21)
76
82 (81)
b
Cl3SiH
–
a
GC yield (isolated yield).
No hexynylsilane was obtained.
b
Table 2. Silylation of 1-hexyne with di- and trihydrosilanesa
Run
Hydrosilane
Ph2SiH2
1-Hexyne/hydrosilane
Product and yield (%)b
C6H9SiHPh2 6a
54
42
9
C6H9SiHMePh 6c
(C6H9)2SiPh2 6b
11
25
84
(C6H9)2SiMePh 6d
12
26
69
1
2
3c
1/2
1/1
2/1
PhMeSiH2
PhSiH3
4
1/2
1/1
2/1
77
46
5
6c
Trace
C6H9SiH2Ph
6e
5
(C6H9)2SiHPh
65
4
Trace
(C6H9)3SiPh
6g
13
60
7
8c
1/1
3/1
0
a
10 mol% of 1a, based on the silane, was used.
GC yield based on the silane, except for runs 1 and 4 (based on 1-hexyne).
10 mol% of 1a, based on 1-hexyne, was used.
b
c

10384
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
dehydrogenative oligomerization of the silane.17 However,
trialkynylsilane 6g was selectively formed in 60% yield
with 3 equiv. of 1-hexyne (run 8).
yield together with 1,7-octadiyne (9%). Interestingly, when
9b was treated with 1a in the absence of the silane, a mixture
of 1,7-octadiyne, 9b, and 10b in a ratio of 1/2/1 was formed
(Eq. (8)). A mixture in the same ratio was also obtained
from the diyne and 10b. These results clearly indicate that
the equilibrium is attained through a redistribution of the
silyl group of alkynylsilanes catalyzed by the imine
complexes 1.
Silylation of 1-hexyne with 1,4-bis(dimethylsilyl)benzene
(2/1) gave monohexynyl- and dihexynylsilanes 7a and 7b in
34 and 64% yields, respectively (Eq. (6)). However, the
reaction with 1,1,2,2-tetraphenyldisilane did not yield any
hexynyldisilanes. Instead, hexynylmonosilanes 6a and 6b
were obtained in 84 and 54% yields, respectively, together
with diphenylsilane (Eq. (7)). The catalyst 1a was proved to
cleave the Si–Si bonds of the starting disilane and the
expected product, 1,2-di(1-hexynyl)tetraphenyldisilane (8).
Thus, on treatment with 1a (10 mol%, rt, 17 h), the disilane
gave a mixture of diphenylsilane (43%), tetraphenyldisilane
(11%), and oligomers, and 8 prepared separately afforded a
mixture of 6a (11%), 6b (2%), and 8 (71%).
ð8Þ
Co-oligomerization of 1,7-octadiyne and diphenylsilane
with 1c (10 mol%) took place in good conversion yields,
from which oligomer 11 (MW¼2907, MW/MN¼2.1, nø10)
was isolated in 12% yield as the higher molecular weight
fraction (Eq. (9)).
ð6Þ
ð9Þ
The present reaction would proceed as depicted in cycle A
of Scheme 1 (H–X¼terminal alkyne, H–Y¼hydrosilane).
Of the individual steps, silanolysis of the lanthanide–
alkynide bond has not been reported, and it seems to be
rather difficult compared to that of lanthanide-sp3 and sp2
carbon bonds, which has been known as a key process for
the hydrosilylation of alkenes and internal alkynes.20
Nevertheless, the catalytic cycle would be driven by the
facile elimination of hydrogen. Thus, the mono-, di-, and tri-
hydrosilanes can be converted to the desired alkynylsilanes
selectively by using 1, including the co-oligomers of a,v-
diynes and dihydrosilanes. Dehydrogenative silylation of
ð7Þ
Next, bissilylation of 1,7-octadiyne with monohydrosilanes
was investigated (Table 3). While monosilyl and bissilyl
products 9a and 10a were obtained in 26 and 59% yields,
respectively, by the reaction with 2 equiv. of triphenylsilane
in the presence of 1a (run 1), the reaction with diphenyl-
methylsilane gave the bissilyl product 10b in low yield (run
3). Although use of a more active complex 1c0 (n¼6) in lieu
of 1a and excess silanes (4 equiv.) improved the yield of
10b up to 62% (run 4), the low yield would be attributed to
slow silylation of the monosilyl compound 9b. In fact,
treatment of 9b with an equimolar amount of diphenyl-
methylsilane and 1a (10 mol%) afforded 10b only in 14%
1-octyne with triphenylsilane was not promoted by Cp*-
2
Sm(thf)2 in hexane, which afforded a head-to-tail dimer of
the alkyne, 2-hexyldec-1-en-3-yne, instead. Therefore, the
success of the present silylation is attributed to the
characteristics of the imine complexes 1.
Table 3. Silylation of 1,7-octadiyne with monohydrosilanes
4. Dehydrogenative silylation of amines and
hydrosilylation of imines.21
In spite of many useful reactions of alkynes and alkenes
catalyzed by organolanthanide complexes, the catalytic
reactions of carbonyl and imine functions have been rarely
explored, because the catalysts are difficult to regenerate
due to strong oxygen– and nitrogen–lanthanide bonds.
Marks et al. reported the first example of lanthanide-
catalyzed imine hydrogenation, in which acyclic imines
were reduced under more severe conditions than olefin
hydrogenation.22 We studied the catalyst activity of the
imine complexes 1 for the hydrosilylation of imines, and
found that the reaction took place efficiently.
Run Hydrosilane Equiv. of silane Catalysta
Product and yield
(%)b
9
10
1
Ph3SiH
2
4
2
4
1a
1c0
1a
1c0
9a 26 10a 59
31 65
9b 33 10b 14
26 62
2
3
4c
Ph2MeSiH
a
20 mol% of 1, based on the diyne, was used. 1c0¼[Yb(h2-Ph2CNC6H4F-4)-
As in olefin hydrosilylation with organolathanides,20 imine
hydrosilylation would involve also two key steps: insertion
of imine to the [Ln]–H bond and silanolysis of the resulting
(hmpa)6].
Isolated yield.
The reaction was carried out without THF.
b
c

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
10385
[Ln]–N bond (Scheme 2). The second step is very rare and
could be difficult, compared to the reaction of olefin. Thus,
we studied at first a catalytic dehydrogenative silylation of
amines with 1 to confirm the second step.
Scheme 2.
As a model reaction, dehydrogenative silylation of amyl-
amine (12a) was tested with various hydrosilanes and the
imine complexes 1 (Table 4). When triphenylsilane and
amylamine (1/1) were successively added to a solution of 1a
(3 mol%) in THF, N-triphenylsilylamylamine (13a) was
produced in 61% yield along with vigorous hydrogen
evolution (run 1). The complexes 1 with electron-with-
drawing groups generally gave better yields of 13a than
those with electron-donating ones, as observed in the
dehydrogenative silylation of terminal alkynes. However,
2,6-xylyl catalyst 1e showed the highest activity among the
imine complexes tested, suggesting that a steric factor can
play also an important role in the reaction (run 4). With
respect to the reactivity of hydrosilanes, those with a phenyl
substituent gave the aminosilanes 13 in over 90% yields
(runs 4, 7, and 8), whereas trialkyl and trialkoxysilanes
decreased the yield to about 20% (runs 9 and 10).
Table 4. Dehydrogenative silylation of amylamine
Run
Imine complex (Ar)
Hydrosilanea
Ph3SiH
Product
Yield (%)b
1
2
3
4
5
6
7
8
1a
1c
1d
1e
1f
Ph
13a
13a
13a
13a
13a
13a
13b
13c
13d
13e
61
52
28
93 (92)
9
Trace
(90)
(93)
(21)
27
4-FC6H4
4-MeOC6H4
2,6-Me2C6H3
2,6-1Pr2C6H3
H
1g
1e
1e
1e
1e
Ph2MeSiH
PhMe2SiH
Et3SiH
9c
10c
Dehydrogenative silylation of various primary and second-
ary amines with Ph3SiH was studied by using the imine
complex 1e (Table 5). Of the primary amines, n- and sec-
alkylamines 12a and 12b were readily converted to 13a and
13f in 93 and 97% yields, respectively (runs 1 and 2).
Although reaction of tert-alkylamine 12c was very slow
(run 3), the product 13g was formed in 81% yield by
(EtO)3SiH
a
12a/R3SiH¼1/1.
b
c
GC yield (isolated yield).
Carried out for 15 h.
Table 5. Dehydrogenative silylation of various amines with Ph3SiHa
Run
1
Amine
1e (mol%)
Time (h)
2
Product
Yield (%)b
93
12a
12b
3
13a
2
3
3
3
2
13f
97
16
12c
20
13g
4
5
6
10
10
10
192
20
5
13g
13h
13i
81
56
95
12d
12e
PhNH2
7
8
12f
12g
BnNHMe
PhNHMe
10
10
2
2
13j
13k
89
79
9
12h
10
20
No reaction
a
12/Ph3SiH¼1/1.
b
GC yield.

10386
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
increasing the catalyst amount (10 mol%) and prolonging
reaction time (run 4). Secondary amines 12e–12g were
silylated similarly in good yields (runs 6–8). In contrast,
aniline was converted to the aminosilane 13h in lower yield
(56%) than alkylamines (run 5), and no reaction occurred
with N-methylaniline (run 9).
most active catalyst in the dehydrogenative silylation of
amines, did not promote any reaction. In addition,
triphenylsilane and diphenylsilane could not be used for
the present hydrosilylation, probably because a hydride
species was not generated effectively in the reaction
mixture.
Dehydrogenative silylation of amines could be performed
with di- and trihydrolsilanes, giving rise to aminosilanes and
diaminosilanes (Eqs. (10) and (11)). In the reaction of
amylamine (12a) with diphenylsilane, when the ratio of
12a/Ph2SiH2 was unity, aminosilane 14a and diaminosilane
15a were obtained in 18 and 35% yields, respectively. Use
of 2 equiv. of 12a resulted in the selective formation of 15a.
N-Benzylamylamine (12i) was silylated with phenylsilane
(3 equiv.) in the presence of the imine complex 1a to afford
aminosilane 17b in 80% yield together with a minor
diaminosilane 18b.
ð12Þ
a
Table 6. Hydrosilylation of imines with PhSiH3
Run
1
Imine
Product and yield (%)b
17
18
ð10Þ
16a
16b
17a
17b
2
18a
95
35
2
52
18b
ð11Þ
3
4
16c
16d
17c
17d
15
3
18c
18d
47
89
Since it was confirmed that silanolysis of lanthanide amide,
i.e. the second step of imine hydrosilylation (Scheme 2),
took place readily, we tried to generate a hydride species
from the imine complex 1 for the first step. Fortunately,
phenylsilane was found to oligomerize with hydrogen
evolution in the presence of 1.23 It has been reported that
dehydrogenative oligomerization of the silane catalyzed by
lanthanocene complexes proceeds through a hydride and
silyl species.17 Therefore, similar intermediates could be
generated in the present system as shown in Scheme 3, and
if the oligomerization is suppressed, the hydride species C
can be available for the imine hydrosilylation.
5
6
7
16e
16f
16g
17e
17f
17g
21
28
56
18e
18f
18g
64
43
0
8
9
16h
16i
17h
3
18h
17
Scheme 3.
Reaction of benzylidenemethylamine (16a) with an equi-
molar amount of phenylsilane in the presence of 1a
(5 mol%) gave aminosilane 17a and diaminosilane 18a in
7 and 65% yields based on 16a (7 and 32% yields based on
the silane), respectively, along with a very small amount of
the oligomer (Eq. (12)). Of course, no reaction took place in
the absence of 1a. Substitution of 1a with 1c caused an
increase of the oligomer, though the products 17a and 18a
were obtained in similar yields. The imine complex 1e, the
No reaction
a
16/PhSiH3¼1/1. 1a0¼[Yb(h2-Ph2CNPh)(hmpa)6].
GC yield based on 16.
b

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
10387
Hydrosilylation of aldimines and ketimines with phenyl-
silane was investigated by using the imine complex 1a0
(n¼6) (Table 6). Aldimines 16a–16f were converted to
aminosilanes 17 and diaminosilanes 18 in good yields (runs
1–6). The diaminosilanes 18 were selectively obtained from
N-methylimines (runs 1 and 4), whereas the ratio of 17/18
was increased in the reaction of N-pentylimines (runs 2, 5,
and 6). Ketimines 16g–16i were less reactive than
aldimines (runs 7–9). While hydrosilylation of ketimine
16g produced aminosilane 17g in 56% yield as the sole
product, little or no product was obtained in the reaction of
16h and 16i.
a
Table 7. Dehydrogenative silylation of amines with PhSiH3
For comparison of hydrosilylation of imines with dehy-
drogenative silylation of amines, amines corresponding to
the imines listed in Table 6 were treated with equimolar
amounts of phenylsilane in the presence of the imine
complex 1a0 (5 mol%), and these results are summarized in
Table 7. As can be seen in Tables 6 and 7, the two reactions
are in good agreement as regards product selectivities and
yields, except for runs 2 and 6. These results indicate that
the two reactions proceed through the same intermediates.
Run
1
Amine
Product and yield (%)b
17 18
12f
12i
17a
17b
1
18a
93
Dehydrogenative silylation of amine can be explained as
shown in cycle A of Scheme 1 (H–X¼HNR2, H–
Y¼hydrosilane), wherein the heteroleptic diamides A
would be an active species and silanolysis takes place
only at one amide moiety derived from the substrate amine.
If homoleptic diamide B participates in the reaction, all
imine complexes 1 bearing a different imine ligand are
likely to exhibit similar catalyst activities, contrary to the
results in Table 4. Moreover, no aminosilanes derived from
the imine ligand were detected in any of the reactions. A
possible catalytic cycle for the imine hydrosilylation is
outlined in Scheme 4. The cycle is essentially similar to the
dehydrogenative silylation of amines in that they both
include hydride E and diamide F as common intermediates.
The only difference between the two is the process to form
diamide F. In the hydrosilylation, insertion of imine to
hydride species E, which is generated in situ by the reaction
of 1 with phenylsilane, affords diamide F. Subsequent
silanolysis of one Yb–N bond of F with phenylsilane yields
the product and E.
2
2
18b
86
3
4
5
12j
12k
12l
17c
17d
17e
15
18c
18d
18e
47
83
57
Trace
34
6
7
12m
12n
17f
87
38
18f
Trace
The aminosilanes, particularly those derived from the less
crowded imine, would be more reactive than phenylsilane in
the second step, probably because metal coordination at
their nitrogen atom facilitates this process. Therefore,
17g
18g
0
8
9
12o
12p
17h
7
18h
11
No reaction
a
Conditions: 12/PhSiH3¼1/1, THF, 20 h, 2358C (runs 1–5) or rt (runs
6–9).
GC yield based on 12.
b
Scheme 4.

10388
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
Table 8. Hydrosilylation of olefins
diaminosilanes could be obtained as the major product in
most of the reactions. Decreased reactivity of ketimines
would be attributed to the lower efficiency in the silanolysis
step, since substituents on the amide moieties are secondary
hydrocarbons. In fact, dehydrogenative silylation of 12n–p
gave lower yields of the products than that of the amines
corresponding to aldimines (runs 7–9 versus 1–6, Table 7).
Although it has been known that the stoichiometric reaction
of imine with lanthanocene hydrides proceeds via C–H
activation rather than imine insertion to the Ln–H bond,22,24
the first lanthanide-catalyzed hydrosilylation of imine is
realized with the imine complexes 1.
Run
1
Olefin
Product and yield (%)a
Method Ab
85d
Method Bc
19a
95
2
3
19be
19c
76
71
90
94
5. Hydrosilylation of olefins and dehydrogenative double
silylation of conjugated dienes.25
Lanthanocene-catalyzed hydrosilylation has been proved to
proceed via insertion of olefins to lanthanide hydride,
followed by s-bond metathesis as depicted in Scheme 5
(path a).26 In contrast to late transition metal catalysts, there
is no precedent for the process initiated by a lanthanide silyl
species (path b). However, if the silyl species once added to
olefins, this process may lead to double silylation, instead of
hydrosilylation, owing to the polarity of hydrosilane in the
s-bond metathesis (path c). For synthetic application of the
lanthanide–imine complexes 1, we studied their catalyst
activities in the hydrosilylation of olefins and conjugated
dienes.
4
5
6
19d
19e
19f
10
0
88
35
95
26f
7
8
19g
19h
0t
0t
95
79g
Scheme 5.
a
GC yield.
b
c
d
e
f
Catalyst: 1at (n¼6), silane/olefin¼1, 20 h.
Catalyst: 21a, silane/olefin¼2, 3 h.
Carried out with 1a (5 mol%), at 2358C.
Ph2SiH2 was used instead of PhSiH3.
508C.
When styrene was treated with an equimolar amount of
PhSiH3 in the presence of 1a in THF at 2358C for 20 h, 2-
(phenylsilyl)ethylbenzene (19a) was formed in 85% yield
together with polyphenylsilane after evolution of hydrogen,
but the other regioisomer 20a, a major product by
lanthanocene catalysts,20 was not detected (Eq. (13)).
g
Carried out with 21b at 508C for 48 h.
The limitation was significantly improved by addition of
equimolar amounts of primary and secondary amines,
particularly diphenylamine, to 1a. Thus, the reaction of
styrene with phenylsilane by using 1a and Ph2NH (catalyst
21a) resulted in reduction of the reaction time and increase
of the yield of 19a to 95%. As can be seen in Table 8
(Method B), all styrene derivatives, including a- and
b-methylstyrene, were converted to the expected products
19 in good yields, except for 4-methoxystyrene (run 5). In
the reaction of 1-decene, the hydrosilylation did not occur
with 21a, but it could be conducted with [Sm(h2-Ph2-
CNPh)(hmpa)0] (1b) and Ph2NH (1/1, catalyst 21b) to
afford the product 19h in 79% yield (run 8).
ð13Þ
The reaction at room temperature gave 19a in lower yield
(33%) probably because of the predominant oligomerization
of the silane. No noticeable reaction was observed with
SmI2 (10 mol%) and without the catalyst 1a even under
refluxing conditions. Results on the hydrosilylation by using
1a0 (n¼6) are summarized in Table 8 (Method A).
Phenylsilane could be substituted by diphenylsilane to
afford 2-(diphenylsilyl)ethylbenzene (19b) in 76% yield
(run 2), but triphenylsilane was inactive. In contrast to the
reaction of styrene and 4-fluorostyrene (runs 1–3), this
method was not applicable to electron-rich and substituted
styrenes nor to a-olefins (runs 4–8).
Various features of the reaction were elucidated through the
improvement mentioned above. With respect to the catalyst
activity of 21, Yb was superior to Sm for the hydrosilylation
of styrene, but the order was reversed for 1-decene. The

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
10389
HMPA ligand showed the promotive and inhibitory effect in
the reaction of styrene and 1-decene, respectively. More-
over, regiochemistry in the reaction of styrene depended on
the ligand: linear silane 19a with HMPA and branched
silane 20a without the ligand were formed exclusively. For
example, 20a was obtained in 72% yield as a single
regioisomer with 1a (n¼0) and PhMeNH (1/1).
(E)-isomers. In addition, similar or somewhat better results
were obtained with 21a.
These results clearly indicate that the reaction mode is
completely different between olefins and conjugated dienes.
As mentioned above, lanthanide hydride C and silyl species
D would be generated in the reaction mixture by partial
oligomerization of the silanes (Scheme 3). In the case of
olefins, it is likely that the reaction was initiated by olefin
insertion to the hydride species C, and followed by
silylation to give the products 19 and 20 (path a in
Scheme 5). This mechanism can explain the regiocontrol by
the HMPA ligand observed for the reaction of styrene. That
is, the [Yb] metal was introduced to the less crowded
b-carbon to yield the linear products 19 in the presence of
HMPA, and, in contrast, the metal was delivered to the
a-carbon for p-coordination of the Ph ring in the absence of
the ligand, leading to the blanched products 20. If styrene
inserted to the silyl species D, a reverse ligand effect should
be observed.
Next, hydrosilylation of conjugated dienes was investigated
(Table 9). The reaction of isoprene with phenylsilane
(3 equiv.) in the presence of 1a at 08C for 20 h in THF gave
1,4-bis(phenylsilyl)-2-methyl-2-butene (22a) and 1-phenyl-
3-methyl-3-silacyclopentene (23a) in 61 and 13% yields,
respectively (run 1). The normally expected 1,4- and 1,2-
hydrosilylation products were not detected at all. Use of
equimolar or less amounts of the silane only caused a
decrease of 22a (ca. 20%) with similar yields of 23a. The
reaction with diphenylsilane also produced the products 22b
and 23b in 57 and 16% yields, respectively (run 2).
Similarly, the dehydrogenative double silylation took place
in the reaction of 2-octyl- and 2,3-dimethylbutadiene (runs 3
and 4), wherein the silacyclopentene 23c was formed as a
major product in the former case. Unfortunately, alkyl
substituents on the terminal position seemed to significantly
decrease the reaction efficiency as shown by 1,3-pentadiene
(run 5). All products 22 were preferentially obtained as
On the other hand, the dehydrogenative double silylation
would be initiated by the reaction of diene with the silyl
species D to yield allylic lanthanide G (Scheme 6).
Silylation of the syn isomer of G gave the product 22 and
the hydride C, and the latter was converted finally to D by
the reaction with another molecule of the silane with
hydrogen evolution. Similarly, intramolecular silylation of
the anti isomer of G would yield the silacyclopentene 23.
The reason for the change in the reaction mode by the two
substrates still remains unclear. Alternatively, the two
reactions may be explained by the addition of a silyl
radical, followed by H abstraction from Si–H for olefin and
Si abstraction for diene. However, this process seems less
likely, because a radical reaction of dienes with hydrosilane
was reported to yield hydrosilylation products,27 and in fact,
22 and 23 were not formed by the reaction with AIBN.
Table 9. Dehydrogenative double silylation of conjugated dienes
Run
1
Diene
Hydrosilanea
PhSiH3
Product, yield (%)b and E/Z ratio
22
23
22a
61 (93/7)
23a
13
2c
3
Ph2SiH2
PhSiH3
22b
22c
57 (93/7)
18 (97/3)
23b
23c
16
41
4
PhSiH3
22d
22e
60 (96/4)
23d
23e
9
0
Scheme 6.
5
PhSiH3
19 (71/29)
Although some ambiguous points still remain regarding the
reaction mechanism, double silylation of conjugated dienes
with hydrosilane, not with disilane, has been uniquely
catalyzed by the imine complexes 1, in contrast to the
reaction with lanthanocenes that were reported to produce
1,4-hydrosilylation products.28
a
Silane/diene¼3.
b
c
GC yield based on the diene.
Carried out at rt.

10390
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
6. Intermolecular hydrophosphination of alkynes and
related carbon–carbon multiple bonds.29
26 by the reaction with Ph2P(O)H was unsuccessful, wherein
the catalyst 1 was immediately oxidized with the phosphine
oxide to give some trivalent lanthanide species and Ph2PH.
Hydrophosphorylation and hydrophosphinylation of
alkynes via P–H bond activation of pentavalent phos-
phorous compounds with Pt and Pd catalysts have been
extensively investigated in the last decade.30 However, the
reaction of trivalent phosphines has been known to be
difficult generally, though a successful example was
reported very recently.31 Besides the group 10 catalysts,
trivalent lanthanocenes were reported to exhibit high
catalyst activities in the intramolecular hydrophosphination
of phosphino-alkynes and alkenes.32 We have inde-
pendently found that the intermolecular hydrophosphination
of alkynes is effectively catalyzed by the lanthanide–imine
complexes 1.
It has been reported that Ph2PH reacts with alkynes under
radical conditions to yield normally (E)-diphenylphos-
phinoalkenes as the primary products, which isomerize to
(Z)-isomers finally.33 Thus, the reaction with AIBN
(10 mol%, THF or neat, 80–858C, without oxidative
workup) was performed for comparison with the imine
complex 1. The radical reaction gave the alkenylphosphine
250 as single regioisomers in lower yields than the latter;
particularly, no or minimal products were formed in the
1
reaction of 24b and 24g. On monitoring by H and 31P
NMR, (Z)-isomers were preferentially formed irrespective
of aromatic and aliphatic alkynes 24, even at an early stage
of the reaction, and this selectivity was not reversed, though
decreased, after longer reaction period. These results clearly
indicate that the hydrophosphination with 1 is different from
the radical reaction in regard to regio- and stereochemistry.
When diphenylphosphine and equimolar amounts of
1-phenyl-1-propyne (24c) were successively added to a
THF solution of 1a0 (n¼6, 5 mol%) at room temperature,
1-phenyl-2-diphenylphosphino-1-propene (25c0) was
quantitatively formed within 5 min (E/Z¼80/20). The
alkenylphosphine 25c0 was isolated in decreased yield
(80%) because of a partial oxidation to the corresponding
alkenylphosphine oxide 25c during the usual workup and
column chromatography. Thus, it was convenient to isolate
the reaction product as the phosphine oxide 25c after
oxidation with H2O2. No reaction took place with Yb(OiPr)3
and SmI2 (10 mol%, rt, 12 h) nor without the catalyst.
Moreover, many products, including polymeric materials,
The present hydrophosphination was applied to other
carbon–carbon multiple bonds to examine its scope and
limitations (Table 11). The reaction of 7,9-hexadecadiyne
with Ph2PH (2 equiv.) gave bis(diphenylphosphinyl)diene
27a in 51% yield as a single regio- and stereoisomer after
the oxidative workup. In contrast, 2,2,7,7-tetramethyl-3,5-
octadiyne was converted to a mixture of diene 27b and
allene 28b in 12 and 71% yields, respectively, under similar
conditions. An attempt to get monophosphinylated products
with equimolar amounts of the phosphine was unsuccessful
because of rapid oligomerization of the diynes even at
2788C. Although the reaction of aliphatic alkenes did not
occur, a- and b-methylstyrene afforded 29a and 29b in good
yields. In the reaction of isoprene, the Ph2P group was
selectively delivered to the less hindered side of the two
terminal carbons to give 1,4- and 1,2-addition products 30
and 31 in 64 and 35% yields, respectively. Two olefinic
phosphine oxides 32 and 33 were formed in 79 and 13%
yields, respectively, by the reaction of cyclohexylallene, but
1-cyclohexyl-3-diphenylphosphinyl-1-propene was not
detected in contrast to the radical reaction.33
n
were formed with BuLi (10 mol%, rt, 30 min).
Results on the reaction of various alkynes 24 are
summarized in Table 10. The expected products 25 and
26 were obtained in high yields under mild conditions,
except for aliphatic internal alkynes 24e–f, in which the
reaction should be heated at 808C without the solvent (runs 5
and 6). With respect to regiochemistry, a Ph2P group was
introduced into the opposite side to the aryl substituents of
the aromatic alkynes exclusively (runs 2–4), and into the
less hindered side of the aliphatic alkynes (runs 6–8).
Stereochemistry of the products 25 and 26 was dependent on
the substrates, i.e. (E)- and (Z)-isomers were preferentially
formed from the aromatic and aliphatic alkynes, respect-
ively. In contrast to these results, direct synthesis of 25 and
The reaction would proceed as shown in cycle B of
Scheme 1 (H–X¼HPPh2), in which the active species could
Table 10. Hydrophosphination of alkynes with diphenylphosphine
Run
Alkyne
R1
Ph
Ph
Ph
Ph
nPr
nPen
tBu
nHex
R2
Ph
SiMe3
Me
H
nPr
Me
H
Conditions
25
Yielda (%)
E/Z
26
Yielda (%)
1
2
3
4
5
6
7
8
24a
24b
24c
24d
24e
24f
THF, rt, 5 min
THF, rt, 4 h
25a
25b
25c
25d
25e
25f
Quant
Quant
Quant
Quant
95
61
62
52
100/0
100/0
80/20
76/24
0/100
0/100
0/100
27/73
–
26b
26c
26d
–
26fc
26g
26h
0
0
0
THF, rt, 5 min
THF, rt, 5 min
Neat, 808C, 6 hb
Neat, 808C, 6 hb
THF, rt, 3 hb
28
10
34
24g
24h
25g
25h
H
THF, rt, 5 min
a
GC yield.
10 mol% of 1a0 was used.
E/Z¼21/79.
b
c

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
Table 11. Hydrophosphination of various C–C multiple bonds with diphenyphosphinea
10391
Substrate
Conditions
Product and yieldb (%)
2358C, 3 hc
rt, 4 h
08C, 15 hd
rt, 30 min
a
All reactions were carried out with 10 mol% of 1a0 (n¼6) in THF and the products were isolated after oxidation with H2O2.
b
c
d
GC yield (isolated yield).
2 equiv. of Ph2PH were used.
2 equiv. of isoprene were used.
be diphosphido complex B. In fact, Yb(PPh2)2(hmpa)n was
and samarium metals to give azalanthanacyclopropanes,
[Ln(h2-Ph2CNAr)(hmpa)n] (1) in quantitative yields,
wherein the metals changed to divalent. Their structures
were characterized by X-ray and NMR analyses. These
dianion complexes 1 are less nucleophilic and more basic
than the corresponding Ln–benzophenone and thiobenzo-
phenone analogs. They can abstract acidic protons of the
substrates H–X such as terminal alkynes, amines, and
diphenylphosphine through ring-opening to afford reactive
species [Ln]–X, which initiate various catalytic reactions.
Thus, in the presence of hydrosilane, dehydrogenative
silylation of terminal alkynes and amines has been
accomplished. Intermolecular hydrophosphination of
alkynes also proceeded catalytically via lanthanide diphos-
phido intermediate. Moreover, with active hydrosilane like
PhSiH3, the imine complexes 1 exhibited good catalyst
activities for hydrosilylation of olefins and aldimines. In the
reaction of conjugated dienes, unusual dehydrogenative
double silylation occurred to yield 1,4-bissilyl-2-butene
derivatives. These results are very different from those with
lanthanocenes, demonstrating the unique reactivities of 1.
isolated by treatment of 1a with 2 equiv. of Ph2PH, and the
reaction by using this complex gave similar results to 1a.
Stereochemistry of the hydrophosphination would not be
determined by isomerization of the products, because their
E/Z ratio was nearly independent of the reaction time.
Accordingly, (Z)-selectivity for the reaction of aliphatic
alkynes seems to be somewhat curious in the light of syn
addition, but this point is obscure at present. Thus, a new
catalytic intermolecular hydrophosphination of alkynes has
been achieved with the Yb–imine complexes 1, giving rise
to alkenylphosphines or phosphine oxides after oxidative
workup in high yields under mild conditions. This method is
also applicable to a wide range of carbon–carbon multiple
bonds such as conjugated diynes and dienes, allenes, and
styrene derivatives.
7. Conclusion
Aromatic ketimines were readily reduced with ytterbium

10392
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
Therefore, combined with simple preparation, the
Ln–imine complexes 1 have potential utility in organic
synthesis.
25.3, 117.3. Anal. calcd for C17H22Si2: C, 72.27; H, 7.85.
Found: C, 72.31; H, 7.93.
8.2.2. 1,4-Bis(diphenylsilyl)-2-methyl-2-butene (22b).
Isolated as a mixture of (E)- and (Z)-isomer (93/7); colorless
oil; IR (neat) 1157, 1145, 802 cm21; MS m/z 434 (Mþ), 252
(Mþ2Ph2Si), 183 (Ph2SiHþ); 1H NMR (CDCl3) d (E)-
isomer (major) 1.46 (3H, s), 1.98 (2H, dd, J¼8.2, 3.5 Hz),
2.07 (2H, d, J¼3.5 Hz), 4.74 (1H, t, J¼3.5 Hz), 4.84 (1H, t,
J¼3.5 Hz), 5.20 (1H, t, J¼8.2 Hz), 7.30–7.58 (20H, m);
(Z)-isomer (minor) (clearly assignable peaks) 1.58 (3H, s),
1.83 (2H, dd, J¼8.0, 3.6 Hz), 4.77 (1H, t, J¼3.6 Hz), 4.89
(1H, t, J¼3.6 Hz), 5.14 (1H, t, J¼8.0 Hz); 13C NMR
(CDCl3) d (E)-isomer (major) 14.4, 18.3, 25.5, 118.4, 127.9,
129.5, 130.7, 134.3, 134.5, 135.1, 135.2; (Z)-isomer (minor)
(clearly assignable peaks) 13.8, 18.6, 25.7; HRMS calcd for
C29H30Si2 (Mþ) 434.1887, found 434.1841.
8. Experimental
8.1. Synthesis and isolation of the Yb(II)–imine complex
(1a)
Ytterbium metal (40 mesh, 346 mg, 2 mmol) and
N-diphenylmethylideneaniline (515 mg, 2 mmol) were
placed in a 20-mL Schlenk tube under argon. THF
(8 mL), HMPA (2 mL), and methyl iodide (2 mL, activator
of the metal) were successively added into the tube, and the
mixture was stirred for 4 h at room temperature to give a
red–black suspension. The suspension was moved into
another Schlenk tube to remove a small amount of metallic
residue. A homogeneous solution was obtained by heating
the suspension under reflux, which was allowed to stand at
room temperature to precipitate red–black needles. The
complex 1a (860 mg, 45%) was isolated by removing the
solvent, followed by drying in vacuo. Crystal data of 1a
have been previously reported,6 and its selected bond
lengths, angles, and NMR data are shown in Refs. 8 and 9,
respectively.
8.2.3. 1-Phenylsilyl-3-phenylsilylmethyl-2-undecene
(22c). Isolated as a mixture of (E)- and (Z)-isomer (97/3);
colorless oil; IR (neat) 1115, 876, 841 cm21; MS m/z 380
(Mþ), 107 (PhSiH2þ); 1H NMR (CDCl3) d (E)-isomer
(major) 0.87 (3H, t, J¼7.1 Hz), 1.15–1.44 (12H, m), 1.78
(2H, dt, J¼8.2, 3.6 Hz), 1.84 (2H, t, J¼3.6 Hz), 1.94 (2H, t,
J¼7.6 Hz), 4.22 (2H, t, J¼3.6 Hz), 4.28 (2H, t, J¼3.6 Hz),
5.14 (1H, br t, J¼8.2 Hz), 7.31–7.39 (6H, m), 7.51–7.57
(4H, m); (Z)-isomer (minor) (clearly assignable peaks) 1.72
(2H, dt, J¼8.0, 3.7 Hz), 4.24 (2H, t, J¼3.7 Hz); 13C NMR
(CDCl3) d (E)-isomer (major) 12.1, 14.1, 20.2, 22.7, 28.0,
29.3, 29.5, 29.7, 31.4, 31.9, 118.7, 127.9, 129.5, 129.6,
132.4, 132.7, 135.16, 135.22, 135.5.
8.2. General procedure for the reactions catalyzed by the
Yb–imine complexes 1
Generally, the imine complex 1 was generated in situ from
Yb metal (0.2 mmol), imine (0.2 mmol), and HMPA
(0.8 mmol) in THF (1 mL) as above. Then, hydrosilane
(2 mmol) or diphenylphosphine (2 mmol) and the substrate
(2 mmol) such as alkyne, alkene, diene, amine, and imine
were added successively to the mixture. In the reaction with
phenylsilane, a mixture of the silane and the substrate in
THF (0.5 mL) was added to the Yb–imine catalyst 1,
because its oligomerization commenced immediately. The
reaction was carried out under appropriate conditions
indicated in the Tables with monitoring by GC. After
completion of the reaction, an internal standard like
tridecane was added to the mixture and the reaction was
quenched with 2 M HCl and water. The organic layer was
extracted with ether, washed with brine, dried over MgSO4,
and concentrated in vacuo. The products were purified by
column chromatography on silica gel. The unstable products
such as aminosilanes were isolated by direct column
chromatography without an aqueous workup.
8.2.4. 1,4-Bis(phenylsilyl)-2,3-dimethyl-2-butene (22d).
Isolated as a mixture of (E)- and (Z)-isomer (96/4); colorless
oil; IR (neat) 1169, 1115, 868, 845 cm21; MS m/z 296 (Mþ),
107 (PhSiHþ2 ); 1H NMR (CDCl3) d (E)-isomer (major) 1.56
(6H, s), 1.87 (4H, t, J¼3.9 Hz), 4.28 (4H, t, J¼3.9 Hz),
7.31–7.40 (6H, m), 7.55–7.56 (4H, m); (Z)-isomer (minor)
(clearly assignable peaks) 1.54 (6H, s), 1.80 (4H, t,
J¼3.9 Hz), 4.24 (4H, J¼3.9 Hz); 13C NMR (CDCl3) d
(E)-isomer (major) 19.4, 20.4, 122.7, 127.9, 129.6, 132.9,
135.2. Anal. calcd for C18H24Si2: C, 72.90; H, 8.16. Found:
C, 72.89; H, 8.23.
8.2.5. 1,4-Bis(phenylsilyl)-2-pentene (22e). Isolated as a
mixture of (E)- and (Z)-isomer (71/29); colorless oil; IR
(neat) 1119, 934, 837 cm21; MS m/z 282 (Mþ), 107
(PhSiHþ2 ); 1H NMR (CDCl3) d (E)-isomer (major) 1.13
(3H, d, J¼7.3 Hz), 1.83 (2H, m), 2.00 (1H, m), 4.17 (2H, d,
J¼2.9 Hz), 4.26 (2H, t, J¼3.8 Hz), 5.34 (1H, dt, J¼15.6,
7.7 Hz), 5.45 (1H, dd, J¼15.6, 7.7 Hz), 7.31–7.40 (6H, m),
7.50–7.57 (4H, m); (Z)-isomer (minor) (clearly assignable
peaks) 1.06 (3H, d, J¼7.3 Hz), 1.79 (2H, m), 2.23 (1H, m),
4.19 (2H, d, J¼2.9 Hz), 4.30 (2H, t, J¼3.8 Hz); 13C NMR
(CDCl3) d (E)-isomer (major) 15.4, 15.9, 22.0, 122.6, 127.9,
129.6, 132.5, 135.3, 135.7; HRMS calcd for C17H22Si2
(Mþ) 282.1261, found 282.1304.
8.2.1. 1,4-Bis(phenylsilyl)-2-methyl-2-butene (22a). Iso-
lated as a mixture of (E)- and (Z)-isomer (93/7); colorless
oil; IR (neat) 1653, 1157, 1117, 876, 841 cm21; MS m/z 282
1
(Mþ), 175 (Mþ2PhSiH2), 107 (PhSiHþ2 ), 105 (PhSiþ); H
NMR (CDCl3) d (E)-isomer (major) 1.57 (3H, s), 1.79 (2H,
dt, J¼7.8, 3.6 Hz), 1.87 (2H, t, J¼3.9 Hz), 4.24 (2H, t,
J¼3.6 Hz), 4.31 (2H, t, J¼3.9 Hz), 5.17 (1H, t, J¼7.8 Hz),
7.31–7.40 (6H, m), 7.52–7.57 (4H, m); (Z)-isomer (minor)
(clearly assignable peaks) 1.52 (3H, s), 1.69 (2H, m), 1.82
(2H, t, J¼4.0 Hz); 13C NMR (CDCl3) d (E)-isomer (major)
12.3, 17.7, 23.1, 118.4, 127.9, 129.6, 130.9, 132.4, 135.2;
(Z)-isomer (minor) (clearly assignable peaks) 12.6, 16.5,
8.2.6. 1-Phenyl-3-methyl-3-silacyclopentene (23a). Color-
less oil; IR (neat) 1638, 1155, 1117, 866, 804 cm21; MS m/z
174 (Mþ), 105 (PhSiþ); 1H NMR (CDCl3) d 1.51 (1H, dm,
J¼18.8 Hz), 1.59 (1H, dm, J¼18.8 Hz), 1.70 (1H, dm,
J¼18.8 Hz), 1.76 (1H, dm, J¼18.8 Hz), 1.82 (3H, s), 4.74

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
10393
(1H, m), 5.60 (1H, br s), 7.35–7.39 (3H, m), 7.57–7.59 (2H,
m); 13C NMR (CDCl3) d 16.0, 20.2, 22.3, 124.7, 128.0,
129.6, 134.4, 135.7, 140.2; HRMS calcd for C11H14Si (Mþ)
174.0865, found 174.0843.
NMR (CDCl3) d 15.0 (d, J¼10.7 Hz), 128.3 (d, J¼4.0 Hz),
128.5, 128.6, 129.4, 129.6 (d, J¼40.6 Hz), 131.2 (d,
J¼102.5 Hz), 131.9 (d, J¼2.5 Hz), 132.1 (d, J¼9.8 Hz),
135.8 (d, J¼18.9 Hz), 142.7 (d, J¼11.5 Hz). (Z)-Isomer
(minor): white solid; mp 120–1238C; IR (Nujol)
1173 cm21; MS m/z 318 (Mþ), 303 (Mþ2Me), 241
8.2.7. 1,1-Diphenyl-3-methyl-3-silacyclopentene (23b).
Colorless oil; IR (neat) 1639, 1157, 941, 845, 806 cm21
;
(Mþ2Ph), 201 (Ph2POþ); H NMR (CDCl3) d 1.90 (3H,
1
MS m/z 250 (Mþ), 172 (Mþ2PhH), 105 (PhSiþ); 1H NMR
(CDCl3) d 1.77 (2H, br s), 1.84 (5H, br s), 5.64 (1H, t,
J¼1.6 Hz), 7.33–7.42 (6H, m), 7.54–7.59 (4H, m); 13C
NMR (CDCl3) d 17.5, 21.7, 22.6, 124.7, 127.9, 129.4, 134.7,
136.2, 140.1; HRMS calcd for C17H18Si (Mþ) 250.1178,
found 250.1176.
dd, J¼12.3, 1.5 Hz), 6.89–6.96 (3H, m), 7.18–7.31 (8H,
m), 7.36 (1H, dm, J¼37.1, 1.5 Hz), 7.53–7.62 (4H, m); 13C
NMR (CDCl3) d 24.7 (d, J¼13.1 Hz), 127.5, 127.7, 128.2
(d, J¼12.3 Hz), 129.3 (d, J¼1.6 Hz), 129.4 (d, J¼92.7 Hz),
131.2 (d, J¼3.3 Hz), 131.3 (d, J¼9.0 Hz), 132.9 (d,
J¼102.5 Hz), 135.7 (d, J¼6.6 Hz), 145.7 (d, J¼7.4 Hz).
8.2.8. 1-Phenyl-3-octyl-3-silacyclopentene (23c). Color-
less oil; IR (neat) 1636, 1153, 1115, 864, 818 cm21; MS m/z
8.2.13. 2-Diphenylphosphinylstyrene (25d) [3582-82-9
and 78045-10-0]. Isolated as a mixture of (E)- and (Z)-
isomer (76/24); white solid; mp 87–898C; IR (Nujol)
1176 cm21; MS m/z 304 (Mþ), 227 (Mþ2Ph), 201
(Ph2POþ), 124 (PhPOþ); 1H NMR (CDCl3) d 6.23
(0.24H, dd, J¼19.6, 14.0 Hz, Z-isomer), 6.76 (0.76H, dd,
J¼22.2, 17.4 Hz, E-isomer), 7.07–7.70 (16H, m).
1
272 (Mþ), 105 (PhSiþ); H NMR (CDCl3) d 0.87 (3H, t,
J¼7.3 Hz), 1.27–1.54 (13H, m), 1.59 (1H, br d,
J¼18.2 Hz), 1.68 (1H, br d, J¼18.2 Hz), 1.76 (1H, br d,
J¼18.2 Hz), 2.13 (2H, t, J¼7.3 Hz), 4.74 (1H, quin,
J¼2.9 Hz), 5.62 (1H, br s), 7.33–7.41 (3H, m), 7.54–7.59
(2H, m); 13C NMR (CDCl3) d 14.1, 15.6, 17.8, 22.9, 27.6,
29.3, 29.4, 29.5, 31.9, 36.3, 123.8, 127.9, 129.5, 134.4,
135.8, 144.5.
8.2.14. (Z)-4-Diphenylphosphinyl-4-octene (25e)
[195148-52-8]. Colorless oil; IR (neat) 1188 cm21; MS
m/z 312 (Mþ), 283 (Mþ2Et), 201 (Ph2POþ), 185 (Ph2Pþ);
1H NMR (CDCl3) d 0.65 (3H, t, J¼7.3 Hz), 0.66 (3H, t,
J¼7.3 Hz), 1.16–1.24 (4H, m), 1.96 (2H, dt, J¼13.8,
6.9 Hz), 2.16 (2H, dq, J¼3.0, 7.3 Hz), 6.29 (1H, dt, J¼37.9,
7.3 Hz), 7.34–7.46 (6H, m), 7.60–7.72 (4H, m); 13C NMR
(CDCl3) d 13.5, 13.6, 22.4 (d, J¼1.6 Hz), 22.9 (d,
J¼3.3 Hz), 32.6 (d, J¼7.4 Hz), 37.7 (d, J¼13.1 Hz), 128.3
(d, J¼12.3 Hz), 131.5, 131.6 (d, J¼11.8 Hz), 131.9 (d,
J¼96.8 Hz), 133.8 (d, J¼100.0 Hz), 148.6 (d, J¼8.2 Hz).
8.2.9. 1-Phenyl-3,4-dimethyl-3-silacyclopentene (23d).
Colorless oil; IR (neat) 1173, 1115, 895 cm21; MS m/z
188 (Mþ), 105 (PhSiþ); 1H NMR (CDCl3) d 1.60 (2H, br d,
J¼18.7 Hz), 1.76 (6H, s), 1.78 (2H, br d, J¼18.7 Hz), 4.61
(1H, m), 7.33–7.39 (3H, m), 7.56–7.59 (2H, m); 13C NMR
(CDCl3) d 19.1, 22.6, 127.9, 129.5, 130.8, 134.4, 135.9.
Anal. calcd for C12H16Si: C, 76.53; H, 8.56. Found: C,
76.39; H, 8.70.
8.2.10. (E)-1-Diphenylphosphinyl-1,2-diphenylethylene
(25a) [14447-40-6]. White solid; mp 156–1578C; IR
8.2.15. (Z)-2-Diphenylphosphinyl-2-octene (25f) [187471-
88-1]. Colorless oil; IR (neat) 1188 cm21; MS m/z 312
(Mþ), 269 (Mþ2Pr), 201 (Ph2POþ), 185 (Ph2Pþ); 1H NMR
(CDCl3) d 0.79 (3H, t, J¼6.9 Hz), 1.11–1.31 (6H, m), 1.74
(3H, d, J¼12.3 Hz), 2.29–2.38 (2H, m), 6.41 (1H, dt,
J¼37.5, 6.3 Hz), 7.46–7.54 (6H, m), 7.66–7.72 (4H, m);
13C NMR (CDCl3) d 13.9, 22.3, 23.8 (d, J¼14.0 Hz), 28.8,
30.4 (d, J¼6.6 Hz), 31.3, 126.9 (d, J¼95.2 Hz), 128.4 (d,
J¼12.3 Hz), 131.48 (d, J¼4.9 Hz), 131.54 (d, J¼2.5 Hz),
133.4 (d, J¼101.7 Hz), 149.8 (d, J¼8.2 Hz).
1
(Nujol) 1177 cm21; H NMR (CDCl3) d 6.92–7.53 (16H,
m), 7.58 (1H, d, J¼21.0 Hz, olefinic), 7.64–7.70 (4H, m);
13C NMR (CDCl3) d 127.7 (d, J¼1.5 Hz), 128.16, 128.22
(d, J¼12.3 Hz), 128.7 (d, J¼1.6 Hz), 128.9, 129.9 (d,
J¼4.1 Hz), 130.2 (d, J¼1.6 Hz), 131.0 (d, J¼103.4 Hz),
131.8 (d, J¼2.5 Hz), 132.3 (d, J¼9.0 Hz), 134.9 (d,
J¼2.4 Hz), 135.2 (d, J¼73.0 Hz), 135.7 (d, J¼26.3 Hz),
143.0 (d, J¼9.0 Hz).
8.2.11. (E)-2-Diphenylphosphinyl-2-trimethylsilyl-
styrene (25b). White solid; mp 122–1238C; IR (Nujol)
1177 cm21; 1H NMR (CDCl3) d 20.03 (3H, d, J¼1.0 Hz),
0.0 (6H, s), 7.21–7.82 (16H, m); 13C NMR (CDCl3) d 1.1
(d, J¼7.4 Hz), 1.5 (d, J¼2.5 Hz), 127.2, 127.8, 127.9 (d,
J¼1.4 Hz), 128.0 (d, J¼1.6 Hz), 128.4 (d, J¼3.3 Hz),
128.5, 128.8, 131.4 (d, J¼3.6 Hz), 131.9 (d, J¼9.9 Hz),
133.4 (d, J¼99.3 Hz), 134.6 (d, J¼20.5 Hz), 139.0 (d,
J¼41.1 Hz), 139.4 (d, J¼77.9 Hz), 141.0 (d, J¼6.6 Hz).
Anal. calcd for C23H25OPSi: C, 73.37; H, 6.69. Found: C,
73.04; H, 6.57.
8.2.16. (Z)-1-Diphenylphosphinyl-3,3-dimethyl-1-butene
(25g). White solid; mp 116–1178C; IR (Nujol) 1180 cm21
;
MS m/z 284 (Mþ), 269 (Mþ2Me), 201 (Ph2POþ), 124
1
(PhPOþ); H NMR (CDCl3) d 1.21 (9H, s), 5.97 (1H, dd,
J¼22.4, 14.6 Hz), 6.68 (1H, dd, J¼43.5, 14.6 Hz), 7.41–
7.50 (6H, m), 7.71–7.77 (4H, m); 13C NMR (CDCl3) d 30.3
(d J¼9.8 Hz), 35.4 (d, J¼5.7 Hz), 119.2 (d, J¼97.5 Hz),
128.4 (d, J¼12.3 Hz), 130.8 (d, J¼9.8 Hz), 131.3, 135.7 (d,
J¼105.0 Hz), 164.5. Anal. calcd for C18H21OP: C, 76.04; H,
7.44. Found: C, 76.14; H, 7.32.
8.2.17. 1-Diphenylphosphinyl-1-octene (25h) [195148-53-
9 and 178943-30-1]. (Z)-Isomer (major): white solid; MS
8.2.12. 1-Phenyl-2-diphenylphosphinyl-1-propene (25c)
[62556-17-6]. (E)-Isomer (major): white solid; mp 129–
1308C; IR (Nujol) 1177 cm21; MS m/z 318 (Mþ), 303
(Mþ2Me), 241 (Mþ2Ph), 201 (Ph2POþ); 1H NMR
(CDCl3) d 2.03 (3H, d, J¼14.0 Hz), 7.13 (1H, d,
J¼22.2 Hz), 7.19–7.50 (11H, m), 7.66–7.71 (4H, m); 13C
1
m/z 312 (Mþ), 255 (Mþ2Bu), 241 (Mþ2Pen); H NMR
(CDCl3) d 0.83 (3H, t, J¼7.1 Hz), 1.10–1.28 (6H, m), 1.34
(2H, quin, J¼7.1 Hz), 2.54 (2H, qm, J¼7.1 Hz), 6.11 (1H,
ddt, J¼25.6, 12.8, 3.0 Hz), 6.69 (1H, ddt, J¼40.5, 12.8,
7.1 Hz), 7.42–7.51 (6H, m), 7.71–7.77 (4H, m); 13C NMR

10394
K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
297–310. (b) Fujiwara, Y.; Takaki, K.; Taniguchi, Y. J. Alloys
Compd. 1993, 192, 200–204. (c) Taniguchi, Y.; Takaki, K.;
Fujiwara, Y. Rev. Heteroat. Chem. 1995, 12, 163–178.
(d) Makioka, Y.; Nakagawa, I.; Taniguchi, Y.; Takaki, K.;
Fujiwara, Y. J. Org. Chem. 1993, 58, 4771–4774. (e) Takaki,
K.; Itono, Y.; Nagafuji, A.; Naito, Y.; Shishido, T.; Takehira,
K.; Makioka, Y.; Taniguchi, Y.; Fujiwara, Y. J. Org. Chem.
2000, 65, 475–481.
(CDCl3) d 14.0, 22.5, 28.76 (d, J¼3.3 Hz), 28.78, 30.9 (d,
J¼8.2 Hz), 31.5, 121.2 (d, J¼100.9 Hz), 128.5 (d,
J¼12.2 Hz), 130.9 (d, J¼9.8 Hz), 131.4 (d, J¼2.5 Hz),
134.6 (d, J¼103.4 Hz), 155.2. (E)-Isomer (minor): white
1
solid; H NMR (CDCl3) d 0.87 (3H, t, J¼7.0 Hz), 1.17–
1.34 (6H, m), 1.48 (2H, quin, J¼7.0 Hz), 2.29 (2H, qm,
J¼7.0 Hz), 6.22 (1H, ddm, J¼24.5, 17.0 Hz), 6.73 (1H, ddt,
J¼19.6, 17.0, 7.0 Hz), 7.27–7.53 (6H, m), 7.67–7.77 (4H,
m); 13C NMR (CDCl3) d 14.0, 22.5, 27.8, 28.8, 31.5, 34.5
(d, J¼16.4 Hz), 121.6 (d, J¼103.4 Hz), 128.5 (d,
J¼12.3 Hz), 131.3 (d, J¼9.9 Hz), 131.6, 133.2 (d,
J¼104.2 Hz), 152.9.
3. (a) Takaki, K.; Kusudo, T.; Uebori, S.; Makioka, Y.;
Taniguchi, Y.; Fujiwara, Y. Tetrahedron Lett. 1995, 36,
1505–1508. (b) Makioka, Y.; Koyama, K.; Nishiyama, T.;
Takaki, K.; Taniguchi, Y.; Fujiwara, Y. Tetrahedron Lett.
1995, 36, 6283–6286. (c) Takaki, K.; Maruo, M.; Kamata, T.;
Makioka, Y.; Fujiwara, Y. J. Org. Chem. 1996, 61,
8332–8334. (d) Takaki, K.; Kusudo, T.; Uebori, S.;
Nishiyama, T.; Kamata, T.; Yokoyama, M.; Takehira, K.;
Makioka, Y.; Fujiwara, Y. J. Org. Chem. 1998, 63,
4299–4304.
8.2.18. 3-Diphenylphosphinyl-2-octene (26f). Isolated as a
mixture of (E)- and (Z)-isomer (21/79); yellow oil; MS m/z
1
312 (Mþ), 201 (Ph2POþ); H NMR (CDCl3) (Z)-isomer
(major) d 0.75 (3H, t, J¼7.5 Hz), 1.03–1.31 (6H, m), 1.85
(3H, d, J¼7.0 Hz), 1.96–2.06 (2H, m), 6.51 (1H, dq,
J¼37.9, 7.0 Hz), 7.43–7.53 (6H, m), 7.67–7.81 (4H, m);
(E)-isomer (minor) (clearly assignable peaks) 0.77 (3H, t,
J¼7.0 Hz), 1.74 (3H, d, J¼6.8 Hz), 6.25 (1H, dq, J¼21.3,
6.8 Hz).
4. (a) Hou, Z.; Takamine, K.; Aoki, O.; Shiraishi, H.; Fujiwara,
Y.; Taniguchi, H. J. Org. Chem. 1988, 53, 6077–6084.
(b) Hou, Z.; Yamazaki, H.; Kobayashi, K.; Fujiwara, Y.;
Taniguchi, H. J. Chem. Soc., Chem. Commun. 1992, 722–723.
(c) Hou, Z.; Yamazaki, H.; Fujiwara, Y.; Taniguchi, H.
Organometallics 1992, 11, 2711–2714.
8.2.19. 2-Diphenylphosphinyl-3,3-dimethyl-1-butene
(26g). White solid; mp 133–1358C; IR (Nujol)
1180 cm21; MS m/z 284 (Mþ), 227 (Mþ2tBu), 201
5. (a) Makioka, Y.; Uebori, S.; Tsuno, M.; Taniguchi, Y.; Takaki,
K.; Fujiwara, Y. Chem. Lett. 1994, 611–614. (b) Makioka, Y.;
Uebori, S.; Tsuno, M.; Taniguchi, Y.; Takaki, K.; Fujiwara, Y.
J. Org. Chem. 1996, 61, 372–375.
1
(Ph2POþ), 124 (PhPOþ); H NMR (CDCl3) d 1.27 (9H,
s), 5.20 (1H, d, J¼22.2 Hz), 5.99 (1H, d, J¼45.6 Hz), 7.36–
7.53 (6H, m), 7.64–7.76 (4H, m); 13C NMR (CDCl3) d 30.5
(d, J¼8.2 Hz), 38.0 (d, J¼9.0 Hz), 127.6 (d, J¼9.0 Hz),
128.3 (d, J¼11.5 Hz), 131.4, 131.6 (d, J¼9.8 Hz), 133.6 (d,
J¼105.0 Hz), 152.7 (d, J¼87.8 Hz). Anal. calcd for
C18H21OP: C, 76.04; H, 7.44. Found: C, 76.09; H, 7.50.
6. Makioka, Y.; Taniguchi, Y.; Fujiwara, Y.; Takaki, K.; Hou, Z.;
Wakatsuki, Y. Organometallics 1996, 15, 5476–5478.
7. (a) Takaki, K.; Tsubaki, Y.; Tanaka, S.; Beppu, F.; Fujiwara,
Y. Chem. Lett. 1990, 203–204. (b) Jin, W.; Makioka, Y.;
Taniguchi, Y.; Kitamura, T.; Fujiwara, Y. J. Chem. Soc.,
Chem. Commun. 1998, 1101–1102. (c) Jin, W.; Makioka, Y.;
Kitamura, T.; Fujiwara, Y. J. Org. Chem. 2001, 66, 514–520.
8.2.20. 2-Diphenylphosphinyl-1-octene (26h). White
solid; mp 140–1428C; IR (Nujol) 1188 cm21; MS m/z 312
(Mþ), 297 (Mþ2Me), 269 (Mþ2Pr), 241 (Mþ2Pen), 201
˚
8. Selected bond lengths (A) and angles (deg): Yb–C(1), 2.679
(18), Yb–N(10), 2.318 (15), C(1)–N(10), 1.43 (3), Yb–O(1),
2.320 (14), Yb–O(2), 2.275 (12), Yb–O(3), 2.298 (12),
Yb–C(1)–N(10), 87.8 (10), Yb–N(10)–C(1), 59.9 (9), C(1)–
Yb–N(10), 32.3 (5), C(2)–C(1)–C(8), 125.6 (15),
C(2)–C(1)–N(10), 118.2 (15), C(8)–C(1)–N(10), 115.3
(14), C(1)–N(10)–C(14), 117.5 (15), O(1)–Yb–O(2), 107.1
(5), O(2)–Yb–O(3), 99.7 (5), O(3)–Yb–O(1), 95.6 (5),
C(1)–Yb–O(1), 102.7 (6), C(1)–Yb–O(2), 128.2 (6), C(1)–
Yb–O(3), 118.6 (5).
1
(Ph2POþ); H NMR (CDCl3) d 0.83 (3H, t, J¼6.8 Hz),
1.13–1.30 (6H, m), 1.42–1.52 (2H, m), 2.30 (2H, q,
J¼8.5 Hz), 5.62 (1H, d, J¼21.0 Hz), 5.94 (1H, d,
J¼43.0 Hz), 7.45–7.56 (6H, m), 7.68–7.73 (4H, m); 13C
NMR (CDCl3) d 14.0, 22.5, 28.0 (d, J¼5.8 Hz), 28.9, 31.5,
31.6 (d, J¼9.0 Hz), 128.5 (d, J¼12.3 Hz), 128.6 (d,
J¼10.7 Hz), 131.5 (d, J¼101.7 Hz), 131.8, 131.9 (d,
J¼9.8 Hz), 144.1 (d, J¼91.9 Hz). Anal. calcd for
C20H25OP: C, 76.90; H, 8.07. Found: C, 76.67; H, 8.13.
9. 1H NMR (THF-d8/HMPA) (5.52 (1H, t, J¼6.8 Hz, p-PhN),
5.58 (2H, t, J¼7.7 Hz, p-PhC), 5.82 (1H, d, J¼6.8 Hz, o-PhN),
6.30 (1H, t, J¼6.8 Hz, m-PhN), 6.35 (1H, d, J¼6.8 Hz,
o-PhN), 6.42 (4H, t, J¼7.7 Hz, m-PhC), 6.56 (1H, t, J¼6.8 Hz,
m-PhN), 6.95 (4H, d, J¼7.7 Hz, o-PhC). 13C NMR (THF-d8/
HMPA) (94.2 (CN), 105.1 (p-PhN), 107.3 (p-PhC), 112.8
(o-PhN), 116.3 (o-PhC), 117.9 (m-PhN), 127.4 (m-PhN), 128.6
(m-PhC), 128.9 (o-PhN), 143.4 (ipso-PhC), 162.0 (ipso-PhN).
10. Takaki, K.; Tanaka, S.; Fujiwara, Y. Chem. Lett. 1991,
493–494.
Characterization of other products such as alkynylsilanes
and aminosilanes is cited in Refs. 15 and 21, respectively.
References
1. For reviews see: (a) Imamoto, T. Lanthanides in Organic
Synthesis; Academic: London, 1994. (b) Edelmann, F. T.
Comprehensive Organometallic Chemistry II; Lappert, M. F.,
Ed.; Pergamon: Oxford, 1995; Vol. 4. Chapter 2. (c) In Topics
in Current Chemistry 179; Herrmann, W. A., Ed.; Springer:
Berlin, 1996. (d) In Lanthanides: Chemistry and Use in
Organic Synthesis; Kobayashi, S., Ed.; Springer: Berlin, 1999.
(e) Chem. Rev.; Kagan, H. B., Ed.;, 2002; 102, pp 1805–2476.
2. (a) Takaki, K.; Fujiwara, Y. Appl. Organomet. Chem. 1990, 4,
11. Ueno, R.; Yano, K.; Makioka, Y.; Fujiwara, Y.; Kitamura, T.
Chem. Lett. 2002, 790–791.
12. (a) Makioka, Y.; Saiki, A.; Takaki, K.; Taniguchi, Y.;
Kitamura, T.; Fujiwara, Y. Chem. Lett. 1997, 27–28.
(b) Makioka, Y.; Taniguchi, Y.; Kitamura, T.; Fujiwara, Y.;
Saiki, A.; Takaki, K. Bull. Soc. Chim. Fr. 1997, 134, 349–355.
13. Makioka, Y.; Tsuno, M.; Takaki, K.; Taniguchi, Y.; Fujiwara,
Y. Chem. Lett. 1995, 821–822.

K. Takaki et al. / Tetrahedron 59 (2003) 10381–10395
10395
14. Because the accurate number of the coordinated HMPA of 1
prepared in situ is unknown, equivalents of this ligand used for
generation of 1 are tentatively shown as a coordination number
(n) in such a way that [Yb(imine)(hmpa)n]. Activities of 1
increased generally with increasing of the number (n) up to 4,
and thus the complex 1 (n¼4) was used unless otherwise noted
for the catalytic reactions described hereafter.
produced polyphenylsilanes with vigorous hydrogen
evolution, from which oligomer (MW¼2966, MW/MN¼1.1)
was isolated in 29% yield by aqueous workup, followed by
short-column chromatography on silica gel using hexane–
ethyl acetate (1/1).
24. Radu, N. S.; Buchwald, S. L.; Scott, B.; Burns, C. J.
Organometallics 1996, 15, 3913–3915.
15. Takaki, K.; Kurioka, M.; Kamata, T.; Takehira, K.; Makioka,
Y.; Fujiwara, Y. J. Org. Chem. 1998, 63, 9265–9269.
16. (a) Liu, H. Q.; Harrod, J. F. Can. J. Chem. 1990, 68,
1100–1105. (b) Itoh, M.; Mitsuzuka, M.; Iwata, K.; Inoue, K.
Macromolecules 1994, 27, 7917–7919, and references cited
therein.
25. Takaki, K.; Sonoda, K.; Kousaka, T.; Koshoji, G.; Shishido,
T.; Takehira, K. Tetrahedron Lett. 2001, 42, 9211–9214.
26. Fu, P.-F.; Brard, L.; Li, Y.; Marks, T. J. J. Am. Chem. Soc.
1995, 117, 7157–7168.
27. Stacey, F. W.; Harris, J. F., Jr. Organic Reactions; Adams, R,
Blatt, A. H., Boekelheide, V., Cairns, T. L., Cope, A. C.,
Curtin, D. Y., Niemann, C., Eds.; Wiley: New York, 1963;
Vol. 13, pp 150–376.
17. Forsyth, C. M.; Nolan, S. P.; Marks, T. J. Organometallics
1991, 10, 2543–2545.
18. Heeres, H. J.; Teuben, J. H. Organometallics 1991, 10,
1980–1986.
28. Onozawa, S.; Sakakura, T.; Tanaka, M. Tetrahedron Lett.
1994, 35, 8177–8180.
19. (a) Heeres, H. J.; Nijhoff, J.; Teuben, J. H. Organometallics
1993, 12, 2609–2617. (b) Evans, W. J.; Keyer, R. A.; Ziller,
J. W. Organometallics 1993, 12, 2618–2633. (c) Forsyth,
C. M.; Nolan, S. P.; Stern, C. L.; Marks, T. J. Organometallics
1993, 12, 3618–3623.
29. (a) Takaki, K.; Takeda, M.; Koshoji, G.; Shishido, T.;
Takehira, K. Tetrahedron Lett. 2001, 42, 6357–6360.
(b) Takaki, K.; Koshoji, G.; Komeyama, K.; Takeda, M.;
Shishido, T.; Kitani, A.; Takehira, K. J. Org. Chem. 2003, 68,
6554–6565.
20. For reviews see: (a) Molander, G. A.; Dowdy, E. D. In
Lanthanides: Chemistry and Use in Organic Synthesis;
Kobayashi, S., Ed.; Springer: Berlin, 1999; pp 119–154.
(b) Molander, G. A.; Romero, J. A. Chem. Rev. 2002, 102,
2161–2185.
30. (a) Han, L.-B.; Tanaka, M. J. Am. Chem. Soc. 1996, 118,
1571–1572. (b) Han, L.-B.; Choi, N.; Tanaka, M. Organo-
metallics 1996, 15, 3259–3261.
31. Kazankova, M. A.; Efimova, I. V.; Kochetkov, A. N.;
Afanasev, V. V.; Beletskaya, I. P.; Dixneuf, P. H. Synlett
2001, 497–500.
21. Takaki, K.; Kamata, T.; Miura, Y.; Shishido, T.; Takehira, K.
J. Org. Chem. 1999, 64, 3891–3895.
22. Obora, Y.; Ohta, T.; Stern, C. L.; Marks, T. J. J. Am. Chem.
32. Douglass, M. R.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc.
2001, 123, 10221–10238.
Soc. 1997, 119, 3745–3755.
23. For example, treatment of PhSiH3 with 1c0 ((10 mol%, rt, 17 h)
33. Mitchell, T. N.; Heesche, K. J. Organomet. Chem. 1991, 409,
163–170.