Experimental and DFT study of the tautomeric behavior of cobalt-containing secondary phosphine oxides

Article abstract of DOI:10.1002/chem.200601051

New cobalt-containing secondary phosphine oxides [(μ-PPh ₂CH₂PPh₂)Co₂(CO)₄{μ, η-PhC≡C-P(=O)(H)(R)}] (8a: R = tBu; 8b: R = Ph) were prepared by reaction of secondary phosphine oxides PhC≡CP-(=O)(H)(R)

Full text of DOI:10.1002/chem.200601051

FULL PAPER
DOI: 10.1002/chem.200601051
Experimental and DFT Study ofthe Tautomeric Behavior ofCobalt-
Containing Secondary Phosphine Oxides
Chu-Hung Wei, Cheng-En Wu, Yi-Luen Huang, Roman G. Kultyshev, and
Fung-E. Hong*[a]
ꢀ
ꢀ
Abstract: New cobalt-containing secon-
dary phosphine oxides [(m-
PPh2CH2PPh2)Co2(CO)4{m,h-PhC C-
P(=O)(H)(R)}] (8a: R=tBu; 8b: R=
PhC CP(OH)(Ph)}] 8b’ indeed takes
place, but it is unlikely between 8a and
PPh2CH2PPh2)Co2(CO)4{m,h-PhC CP-
(NEt2)(tBu)}] (7a), and its oxidation
G
ACHTREUNG
ꢀ
ꢀ
[(m-PPh2CH2PPh2)Co2(CO)4{m,h-PhC
CP(OH)
product [(m-PPh2CH2PPh2)Co2(CO)4-
ꢀ
E
A
A
(NEt2)
A
7a’
Ph) were prepared by reaction of sec-
demonstrated that reasonable activa-
tion energies for the tautomeric con-
versions can be achieved only via a bi-
molecular pathway. Since a tBu group
is much larger than a Ph group, the
conversion is presumably only feasible
in the case of 8bÐ8b’, but not in the
case of 8aÐ8a’. Another cobalt-con-
were prepared from the reaction of
ꢀ
ꢀ
ondary phosphine oxides PhC CP-
PhC CP
(NEt2)
(tBu) (5a) with 2. The
(=O)(H)(R) (6a: R=tBu; 6b: R=Ph)
with dppm-bridged dicobalt complex
[(m-PPh2CH2PPh2)Co2(CO)6] (2). The
molecular structures of 8a and 8b were
determined by single-crystal X-ray dif-
fraction. Although palladium-catalyzed
Heck reactions employing 8b as ligand
gave satisfying results, 8a performed
poorly in the same reaction. Judging
from these results, a tautomeric equi-
librium between 8b and its isomeric
form [(m-PPh2CH2PPh2)Co2(CO)4{m,h-
molecular structures of 7a and 7a’
were determined by single-crystal X-
ray diffraction. The phosphorus atom is
surrounded by substituents in a tetra-
A P N single
bond (1.676(3) Š) is observed in the
À
hedral environment.
taining
phosphine,
namely,
[(m-
molecular structure of 7a. Heck reac-
tions employing 7a/Pd
lyst system exhibited efficiency compa-
rable to that of 8a/Pd(OAc)2.
ACHTRE(UGN OAc)2 as cata-
Keywords: density functional calcu-
lations · Heck reaction · P ligands ·
palladium · tautomerism
AHCTREUNG
Introduction
phosphines in various phosphine-assisted catalytic reactions,
has received a great deal of attention.[3] Probably the most
widely used and best known metal-containing bidentate
phosphine ligands are bis(diphenylphosphino)ferrocene
(dppf, 1) and its derivatives.[4] In the past few years, we have
been developing another class of cobalt-containing (mono-
or bidentate) phosphines, namely, 3. Each of these com-
pounds was prepared by reaction of dppm-bridged dicobalt
complex [(m-PPh2CH2PPh2)Co2(CO)6] (2) with the corre-
sponding alkynyl phosphine. This reaction allows straightfor-
ward access to a new class of phosphines having bulky char-
acter. Interestingly, besides the commonly observed palladi-
um–phosphorus bonds, unusual palladium–cobalt bonds
were formed during the complexation of 3 with palladium
sources.[5] Therefore, the dicobalt fragment present in metal-
containing phosphines 3 does not act merely as an innocent
spectator, as in the conventional organic phosphine/metal
catalyst system. The unique character of 3 indeed opens a
new territory for the use of metal-containing phosphines as
ligands in cross-coupling reactions. The roles played by
Trialkyl and triaryl phosphines (R3P, Ar3P) are among the
most versatile and commonly employed ligands in reactions
catalyzed by transition-metal complexes.[1] Recently, N-het-
erocyclic carbenes (NHC) have been introduced as poten-
tially effective ligands for coupling reactions.[2] Nevertheless,
these types of ligands are normally either air-/moisture-sen-
sitive or expensive, and this significantly limits on their syn-
thetic applications. Lately, the concept of using metal-con-
taining phosphines, rather than the conventional organic
[a] C.-H. Wei, C.-E. Wu, Y.-L. Huang, Dr. R. G. Kultyshev,
Prof. F.-E. Hong
Department of Chemistry
National Chung Hsing University
Taichung 40227 (Taiwan)
Fax : (+886)4-2286-2547
E-mail: fehong@dragon.nchu.edu.tw
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2007, 13, 1583 – 1593
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1583

these newly made compounds as assisting ligands in reac-
tions catalyzed by palladium complexes were thoroughly ex-
plored.[6] In most cases, the phosphorus atoms of these li-
gands are electron-rich and, unfortunately, prone to oxida-
tion. Oxidized ligands lose their coordination capabilities
toward low-valent transition metals such as PdII or Pd0, and
this results in greatly reduced catalytic performance.
Results and Discussion
ꢀ
Preparation ofPhC CP
Ph) and PhC CP(=O)(H)(R) (6a: R=tBu; 6b: R=Ph):
A
ꢀ
ꢀ
New alkynyl amino phosphine PhC CP
A
U
was prepared by
a
modified literature procedure
(Scheme 2).[9] The 31P NMR spectrum of 5a displays a sin-
glet at d=50.8 ppm, which is shifted considerably upfield
compared with d=161.6 ppm for 4a. Conversion of 5a to
ꢀ
secondary alkynyl phosphine oxide PhC CP(=O)(H)ACHTREUNG(tBu)
(6a) was achieved in acidic medium according to the ingeni-
ous preparative procedure described by Denis et al.[9] The
31P NMR spectrum of 6a displays a singlet at d=18.7 ppm.
ꢀ
A
similar preparation was carried out for PhC CP-
AHCTREUNG
(NEt2)(Ph) (5b, Scheme 2). The 31P NMR spectrum of 5b
exhibits a singlet at d=35.3 ppm. The conversion of 5b to
ꢀ
The tautomeric equilibrium between a secondary phos-
phine oxide RR’HP=O (I) and the less stable phosphinous
acid RR’POH (I’) is shifted to the left (Scheme 1).[7] Gener-
PhC CP(=O)(H)(Ph) (6b) was achieved readily by column
chromatography on silica.[9] In principle, a tautomeric equi-
librium between 6 and the less stable secondary phosphi-
nous acid 6’ could take place in solution (Scheme 2).
ꢀ
Preparation of[( m-PPh2CH2PPh2)Co2(CO)4{m,h-PhC CP-
(=O)
A
ligand having both soft (e.g., P) and hard (e.g., N) donor
atoms normally exhibits interesting coordinating capacity
Scheme 1.
ally, I’ exhibits a good coordi-
nating capability toward low-
valent transition metals, but I
does not. Satisfactory catalytic
performance for palladium-cat-
alyzed cross-coupling reactions
employing I’-like phosphine li-
gands has been reported.[8]
To the best of our knowledge,
an extensive study on the tauto-
merism of a metal-containing
phosphine oxide and the corre-
sponding phosphinous acid has
never been reported before.
Therefore, we were interested
in exploring the tautomerism of
a
metal-containing phosphine
oxide using both experimental
and DFT methods. Herein we
report the preparation of sever-
al cobalt-containing secondary
phosphine oxides and the ex-
ploration of their catalytic per-
formance in Heck coupling re-
actions. In addition, results of
Scheme 2.
the DFT studies on the pathways of tautomeric rearrange-
ment available for these ligands are presented.
toward low-valent transition metals.[10] The relative ease of
breaking/formation of the nitrogen–metal bond makes it
useful in various catalytic reactions.[11] Accordingly, the coor-
dination of P,N bidentate ligands to transition metals has re-
cently attracted a lot of attention.[12]
1584
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1583 – 1593

Cobalt-Containing Secondary Phosphine Oxides
FULL PAPER
Table 1. Crystal data of 7a, 7a’, 8a, and 8b.
Compound
7a
7a’
8a
8b
formula
formula weight
cryst syst
space group
a [Š]
b [Š]
c [Š]
a [8]
b [8]
C45H46Co2N1O4P3
875.60
C45H46Co2N1O5P3
891.60
C41H36Co2O5P3
822.49
C43H33Co2O5P3·CH2Cl2
925.39
triclinic
monoclinic
P2(1)/n
19.3785(13)
21.1861(14)
22.5262(15)
–
107.9150(10)
–
8799.8(10)
4
monoclinic
P2(1)/n
11.4373(9)
20.8937(17)
16.8703(13)
–
103.473(2)
–
3920.5(5)
4
triclinic
¯
¯
P1
P1
17.6981(9)
12.8417(6)
19.3480(10)
90.0000(10)
97.7970(10)
90.0000(10)
4356.6(4)
2
12.4124(14)
13.0915(15)
15.1746(18)
83.627(2)
89.614(2)
62.164(2)
2164.2(4)
2
g [8]
V [Š3]
Z
1calcd [MgmÀ3
]
1.335
1.346
1.393
1.420
l
A
0.71073
0.913
0.71073
0.907
0.71073
1.011
0.71073
1.044
)
1.16–26.01
16917
991
0.0417
0.0904
1.55–26.04
17302
1009
0.0393
0.0818
1.95–26.10
7728
460
0.0775
0.1937
1.35–26.02
8393
509
0.0655
0.1695
observed reflections (F>4s(F))
no. of refined parameters
R1[a] for significant reflections
wR2[b] for significant reflections
GoF[c]
0.904
0.888
1.149
0.952
[a] R1=jꢀ(jFojÀjFcj)/ꢀjFo jj. [ b]wR2=[ꢀw(F2oÀF2c)2/ꢀw(F2o)2]1/2. [c] GoF=[ꢀw(F2oÀFc2)2/(NreflectionsÀNparameters)]1/2
.
Cobalt-containing
phosphine
[(m-PPh2CH2PPh2)-
ꢀ
A
G
ACHTREUNG
ꢀ
product
(NEt2)
U
G
(tBu)}] (7a’) were obtained from the reaction of
ꢀ
PhC CPACHTREUNG(NEt2)ACHTREUNG(tBu) (5a) with 2 (Scheme 2). Presumably,
the latter was formed during chromatography. Compounds
7a and 7a’ were identified by spectroscopic means and
single-crystal X-ray diffraction (Table 1). The 31P NMR spec-
trum of 7a displays three sets of signals at d=36.7, 37.8
(both dppm), and 86.5 ppm [P
those of 7a, the corresponding signals of 7a’ at d=34.7, 35.3
(both dppm), and 49.7 ppm [P(=O)(tBu)(NEt2)] were ob-
ACHRTU(NEG tBu)ACHTREU(GN NEt2)]. Compared with
A
ACHTREUNG
served upfield. The ORTEP diagrams of 7a and 7a’ are de-
À
picted in Figures 1 and 2, respectively. A P N single bond,
À
(7a: 1.676(3) Š; 7a’: 1.652(3) Š), and a P C
A
bond (7a: 1.878(4) Š; 7a’: 1.837(4) Š) were observed for
each compound. In addition,
a
P=O double bond
(1.484(2) Š) was observed for 7a’. The oxidized phosphorus
atom of 7a’ is located in a nearly tetrahedral surrounding.
The configuration of the rest of 7a (or 7a’) is similar to that
of its counterpart 3e. Attempts to prepare 7b by a similar
procedure always led to the formation of [(m-
ꢀ
Figure 1. ORTEP drawing of 7a. Hydrogen atoms are omitted for clarity.
À
À
Selected bond lengths [Š] and angles [8]: Co(1) C(8) 1.949(3), Co(1)
À
À
À
C(7) 2.002(3), Co(1) P(2) 2.2270(9), Co(1) Co(2) 2.4870(6), Co(2) C(7)
À
À
À
1.966(3), Co(2) C(8) 1.975(3), Co(2) P(3) 2.2196(9), P(1) N(1) 1.676(3),
P(1) C(7) 1.808(3), P(1) C(9) 1.878(4), P(2) C(29) 1.833(3), P(3) C(29)
PPh2CH2PPh2)Co2(CO)4{m,h-PhC CP(=O)(H)(Ph)}]
instead. Presumably, the polarity of the Ph group of 7b facil-
itates hydrolysis of 7b to yield 8b.
(8b)
À
À
À
À
À
À
1.838(3), C(1) C(8) 1.466(4), C(7) C(8) 1.366(4); C(8)-Co(1)-C(7)
40.44(12), C(7)-Co(2)-C(8) 40.57(12), C(7)-Co(2)-Co(1) 51.83(9), N(1)-
P(1)-C(7) 108.69(15), N(1)-P(1)-C(9) 105.85(17), C(7)-P(1)-C(9)
103.98(15), C(8)-C(7)-P(1) 133.6(3), Co(2)-C(7)-Co(1) 77.62(11), C(7)-
C(8)-C(1) 136.0(3), Co(1)-C(8)-Co(2) 78.67(12), P(2)-C(29)-P(3)
111.84(16).
ꢀ
Preparation of[( m-PPh2CH2PPh2)Co2(CO)4{m,h-PhC CP-
(=O)(H)(R)}] (8a: R=tBu; 8b: R=Ph): By similar proce-
dures, cobalt-containing secondary phosphine oxides [(m-
ꢀ
PPh2CH2PPh2)Co2(CO)4{m,h-PhC CP(=O)(H)(R)}]
(8a:
ꢀ
R=tBu; 8b: R=Ph) were prepared by reaction of PhC
CP(=O)(H)(R) (6a: R=tBu; 6b: R=Ph) with 2. In addi-
tion, 8a can be prepared from 7a by a procedure analogous
to that described by Denis et al. (Scheme 2).[9] Both 8a and
8b were identified by NMR spectroscopy and single-crystal
X-ray diffraction (Table 1). The 31P NMR spectrum of 8a
shows three sets of signals at d=35.8, 33.8 (br, both dppm),
À
and 47.2 ppm (d, P(=O)ACHTRE(UGN tBu)(H)). The presence of a P H
Chem. Eur. J. 2007, 13, 1583 – 1593
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1585

F.-E. Hong et al.
Figure 3. ORTEP drawing of 8a. Hydrogen atoms are omitted for clarity.
Figure 2. ORTEP drawing of 7a’. Hydrogen atoms are omitted for clarity.
À
À
Selected bond lengths [Š] and angles [8]: Co(1) C(1) 1.940(5), Co(1)
À
À
Selected bond lengths [Š] and angles [8]: Co(1) C(8) 1.951(3), Co(1)
À
À
À
C(2) 1.981(5), Co(1) P(3) 2.2346(13), Co(1) Co(2) 2.4779(10), Co(2)
À
À
À
C(7) 1.982(3), Co(1) P(2) 2.2311(9), Co(1) Co(2) 2.4729(6), P(1) O(1)
À
À
À
C(1) 1.957(5), Co(2) C(2) 1.963(5), Co(2) P(2) 2.2286(14), P(1) O(1)
1.379(6), P(1) C(39) 1.741(7), P(1) C(1) 1.775(5), P(2) C(34) 1.838(5),
À
À
À
1.484(2), P(1) N(1) 1.652(3), P(1) C(7) 1.783(3), P(1) C(9) 1.837(4),
À
À
À
À
À
À
À
Co(2) C(7) 1.962(3), Co(2) C(8) 1.979(3), Co(2) P(3) 2.2315(9), P(2)
À
À
À
P(3) C(34) 1.831(5), C(1) C(2) 1.354(7), C(2) C(3) 1.484(6); C(1)-
Co(1)-C(2) 40.4(2), C(1)-Co(2)-C(2) 40.40(19), O(1)-P(1)-C(39) 122.3(4),
O(1)-P(1)-C(1) 118.4(3), C(39)-P(1)-C(1) 112.9(3), C(2)-C(1)-P(1)
144.4(4), Co(1)-C(1)-Co(2) 78.96(19), C(1)-C(2)-C(3) 139.7(4), Co(2)-
C(2)-Co(1) 77.86(17), P(2)-C(34)-P(3) 106.5(2).
À
À
À
C(29) 1.833(3), P(3) C(29) 1.834(3), C(6) C(8) 1.478(4), C(7) C(8)
1.361(4); C(8)-Co(1)-C(7) 40.48(11), O(1)-P(1)-N(1) 110.30(14), O(1)-
P(1)-C(7) 110.97(13), N(1)-P(1)-C(7) 109.78(15), O(1)-P(1)-C(9)
109.89(15), N(1)-P(1)-C(9) 107.98(16), C(7)-P(1)-C(9) 107.83(15), C(42)-
N(1)-C(44) 116.3(3), C(7)-Co(2)-C(8) 40.40(11), C(8)-C(7)-P(1) 132.2(2),
C(7)-C(8)-C(6) 139.2(3), P(2)-C(29)-P(3) 108.44(15).
bond in 8a is evidenced by a large coupling constant of the
last-named signal (JP, H =457.8 Hz). Similarly, three sets of
signals at d=36.8, 35.7 (br, both dppm), and 21.5 ppm (d,
J
P, H =483.4 Hz, P(=O)(Ph)(H)) were observed in the
31P NMR spectrum of 8b. The ORTEP diagrams of 8a and
8b are depicted in Figures 3 and 4, respectively. In both
compounds the oxidized phosphorus atoms are situated in
almost tetrahedral environments. The P=O bond legths for
8a and 8b are 1.379(6) and 1.470(4) Š, respectively. In solu-
tion, a tautomeric equilibrium between secondary phosphine
oxide 8 and its phosphinous acid form 8’ is expected.
Heck reactions employing cobalt-containing secondary
phosphine oxides 8a and 8b: In principle, 8’ could act as a
ligand toward a palladium complex. Therefore, the coordi-
nation of 8’ to palladium in basic media would yield a cata-
lyst precursor ready for coupling reactions (Scheme 3).
Palladium-catalyzed Heck reactions of bromobenzene
with styrene were carried out by employing 8a, 8b, and re-
lated secondary phosphine oxides as ligands (Scheme 4). In
the general procedure for the catalytic reactions, a Schlenk
tube was charged with 1.0 mmol of bromobenzene, 1.2 mmol
of styrene, 2.0 mL of solvent, 1.4 mmol of a base, 1.0 mol%
Figure 4. ORTEP drawing of 8b. Hydrogen atoms are omitted for clarity.
À
À
Selected bond lengths [Š] and angles [8]: Co(1) C(1) 1.944(5), Co(1)
À
À
À
C(2) 1.965(5), Co(1) P(2) 2.2608(16), Co(2) C(2) 1.965(5), Co(2) C(1)
À
À
À
1.968(5), Co(2) P(3) 2.2243(15), P(1) O(5) 1.470(4), P(1) C(1) 1.760(5),
P(1) C(9) 1.799(6), P(2) C(35) 1.838(5), P(3) C(35) 1.835(5), C(1) C(2)
1.366(7), C(2) C(3) 1.454(7); C(1)-Co(1)-C(2) 40.9(2), P(2)-Co(1)-Co(2)
À
À
À
À
À
97.65(5), C(2)-Co(2)-C(1) 40.6(2), O(5)-P(1)-C(1) 116.0(2), O(5)-P(1)-
C(9) 112.6(3), C(1)-P(1)-C(9) 107.3(2), C(2)-C(1)-P(1) 142.1(4), C(2)-
C(1)-Co(1) 70.4(3), C(2)-C(1)-Co(2) 69.6(3), Co(1)-C(1)-Co(2) 78.96(19),
C(1)-C(2)-C(3) 143.3(5), Co(2)-C(2)-Co(1) 78.5(2).
of PdACHTREUNG(OAc)2, and 2.0 mol% of a secondary phosphine oxide
ligand. The mixture was stirred at the designated reaction
temperature and time. It was then worked up.
1586
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1583 – 1593

Cobalt-Containing Secondary Phosphine Oxides
FULL PAPER
played by the secondary phos-
phine oxide ligand in the reac-
tion can be clearly seen. Low
conversion was observed in the
absence
entry 2).
of
8b
(Table 4,
PdACHTRE(UNG OAc)2 is probably the
most frequent chosen palladium
source in coupling reactions
due to its outstanding perfor-
mance in most cases. In the fol-
lowing runs, the influence of
various palladium sources on
Scheme 3.
the reaction yield was evaluated
(Table 5). Heck reactions be-
tween bromobenzene and styrene employing 8b/8b and var-
ious Pd sources in the K2CO3/DMF system were carried out.
Again, the best performance was achieved with PdACHTRE(UNG OAc)2 as
palladium source (Table 5, entry 1). Interestingly, a rather
Scheme 4. Palladium-catalyzed Heck reactions with assistance by secon-
dary phosphine oxide.
Table 3. Heck reactions employing 8b/Pd(OAc)2 or 8a/Pd(OAc)2 and
K2CO3 as base in various solvents.[a]
The performance of a palladium-catalyzed carbon–carbon
cross-coupling reaction is governed by a number of fac-
tors.[13] Among them, a well-chosen base is crucial to the
success of the reaction.[14] The influence of the base used in
these coupling reactions catalyzed by secondary phosphine
oxide was examined (Table 2). The best performance was
Entry
Solvent
Yield [%] (L=8b)[b,d]
Yield [%] (L=8a)[c,d]
1
2
3
4
DMF
75.1
45.6
39.4
2.1
60.5
21.6
48.3
8.4
dioxane
DMA[e]
toluene
[a] Reaction conditions as in the footnote of Table 2. [b] 1108C, 7 h.
[c] 1308C, 13 h. [d] Yield of isolated product; average of two runs.
[e] DMA=N,N-dimethylacetamide.
Table 2. Heck reactions employing 8b and various bases.[a]
Entry
Base
Yield [%][b]
Table 4. Dependence of the Heck reaction yield on the Pd(OAc)2/8b
ratio.[a]
1
2
3
4
5
6
7
8
K2CO3
KF
CsF
75.1
65.1
59.1
51.8
51.3
43.8
35.5
8.5
Entry
Pd/L
Yield [%][b]
1
2
3
4
5
6
7
8
0/1
1/0
1/1
1/1.5
1/2
1/3
1.5/1
2/1
0.0
16.1
47.3
68.1
75.1
62.5
49.1
59.6
NEt3
K3PO4
DABCO
Cs2CO3
NaOtBu
[a] Reaction conditions: 1.0 mmol of bromobenzene, 1.2 equivalents of
styrene, 1.4 equivalents of base, 2 mL DMF, 1.0 mol% of Pd(OAc)2,
2.0 mol% of 8b, 1108C, 7 h. [b] Yield of isolated product; average of two
runs.
[a] With the exception of Pd(OAc)2/8b ratio, reaction conditions as in
the footnote of Table 2. [b] Yield of isolated product; average of two
runs.
observed with K2CO3 (Table 2, entry 1). Unexpectedly, the
yield was low when a strong base such as NaOtBu was em-
ployed (Table 2, entry 8).
Table 5. Dependence of the Heck reaction yield on the Pd source.[a]
Subsequently, the impact of various high-boiling solvents
on the reaction was evaluated. As shown in Table 3, the re-
action is greatly affected by the nature of the solvent used.
For instance, the coupling reaction in toluene was ineffective
(Table 3, entry 4). However, the yield was greatly improved
when DMF was employed as solvent (Table 3, entry 1).
Carrying out the Heck reaction in DMF in the presence
Entry
Pd source
Yield [%]
Yield [%]
(L=8b)[b,c]
(L=8a)[c,d]
1
2
3
4
5
6
Pd(OAc)2
[Pd(cod)Cl2]
75.1
60.5
59.2
58.3
54.6
15.7
60.5
57.6
49.3
56.2
54.4
10.7
[
2(dba)3] Pd
[Pd(CH3CN)2Cl2]
[
h3-C3H5)PdC{(l}2][e]
PdCl2
of K2CO3 at various Pd
red ratio (Table 4). The optimum yield was achieved with
Pd(OAc)2/8b=1/2 (Table 4, entry 5). The positive role
ACHTRE(UNG OAc)2/8b ratios revealed the prefer-
[a] Reaction conditions as in the footnote of Table 2. [b] 1108C, 7 h.
[c] Yield of isolated product; average of two runs. [d] 1308C, 13 h.
[e] dba=trans,trans-dibenzylideneacetone.
ACHTREUNG
Chem. Eur. J. 2007, 13, 1583 – 1593
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1587

F.-E. Hong et al.
Table 8. Dependence of the Heck reaction yield on Pd(OAc)2 concentra-
tion.[a]
low yield was obtained with PdCl2, which is another fre-
quently chosen palladium source (Table 5, entry 6).
Entry
[Pd] [mol%]
Yield [%][b]
Reaction temperature is another critical factor for Heck
reactions. In general, Heck reactions require much high re-
action temperatures than other coupling reactions. Poor per-
formance was observed when the reaction temperature was
below 1108C (Table 6, entries 1 and 2). However, excellent
1
2
1.0
0.1
93.1
51.2
[a] With the exception of temperature (1208C), the reaction conditions as
in the footnote of Table 2. [b] Yield of isolated product; average of two
runs.
Table 6. Dependence of the Heck reaction yield on temperature.[a]
Entry
T [8C]
Yield [%][b]
ing electron-donating groups (Table 9, entries 5 and 6).[17]
Poor performance was observed with o-bromotoluene due
to severe steric hindrance (Table 9, entry 7). As expected
from the generally accepted mechanism, trans-1,2-diaryl-
ethylene B rather than 1,1-diarylethylene A was obtained as
the major product of the reaction (Scheme 5).
1
2
3
4
5
6
90
100
110
120
130
140
7.8
9.2
75.1
90.1
>99.0
>99.0
[a] With the exception of temperature, reaction conditions as in the foot-
note of Table 2. [b] Yield of isolated product; average of two runs.
Table 9. Dependence of the yield of Heck reaction with Pd(OAc)2/8b or
Pd(OAc)2/8a on substrate.[a]
L=8b[b, d]
Yield [%]
L=8a[c,d]
Yield [%]
yields were obtained when the reaction temperature was
Entry
Substrate
A/B
A/B
higher than 1208C (Table 6, entries 5 and 6).
In a PdACHTREUNG(OAc)2-catalyzed coupling reaction an induction
1
2
3
4
5
6
7
4-NO2
4-COCH3
4-COH
H
4-CH3
4-OCH3
2-CH3
83.6
84.4
70.1
3/97
2/98
52.4
54.1
40.3
60.5
62.3
63.7
14.1
1/99
1/99
3/97
0/100
2/98
5/95
0/100
period due to reduction of PdII to Pd0 active species is ob-
served.[15] This is particularly true of the Heck reaction,
which requires more severe reaction conditions. A satisfac-
tory result was obtained when the reaction was carried out
at 1208C for 5 h (Table 7, entry 3). On the contrary, the
yield dropped to 68.6% or even 20.5% (Table 7, entries 2
and 1) when the reaction time
10/90
0/100
1/99
19/81
0/100
93.1
>99.0
>99.0
21.5
[a] Reaction conditions as in the footnote of Table 2. [b] 1208C, 5 h.
[c] 1308C, 13 h. [d] Yield of isolated product; average of two runs.
was shortened to 3 or 1 h. Pro-
longing the reaction time might
lead to the decomposition of
the active catalyst and eventual-
ly diminish the reactivity. Side
reactions may occur as well
(Table 7, entry 4).
Scheme 5.
For the purpose of comparison, the same Heck reactions
employing various phosphine ligands were carried out in
DMF at 1208C for 7 h. A poor yield was observed in the ab-
sence of ligand (Table 10, entry 1). The yields were greatly
improved on addition of 7a, 8b, or 3e (Table 10, entries 2,
4, and 5). After 7 h of reaction, almost quantitative yield
was observed with 7a as ligand (Table 10, entry 2).
Table 7. Dependence of the Heck reaction yield on reaction time.[a]
Entry
Time [h]
Yield [%][b]
1
2
3
4
1
3
5
7
20.5
68.6
93.1
90.1
[a] With the exception of reaction time, reaction conditions as in the foot-
note of Table 2. [b] Yield of isolated product; average of two runs.
Table 10. Heck reactions with various ligands.[a]
Entry
Ligand
Time [h]
Yield [%][b]
A/B
While the Heck coupling reaction with 1.0 mol% of Pd-
ACHTREUNG(OAc)2 was effective (Table 8, entry 1), 0.1 mol% of catalyst
1
2
3
4
5
6
7
–
7
5(7)
30.2[c]
80.6(>99.0)[c]
45.7[c]
7/93
7a
7a’
8b
3e
3b
3 f
1/99(1/99)
11/89
0/100
0/100
2/98
resulted in a much poorer performance (Table 8, entry 2).
It has been commonly observed that in a palladium-cata-
lyzed coupling reaction a better conversion is achieved with
aryl halides bearing electron-withdrawing rather than elec-
tron-donating substituents.[16] Nevertheless, the reverse trend
7
5
5
5
5
90.1[c]
>99.0[d]
63.7[d]
52.6[d]
1/99
was observed in the Pd
ACHTRE(UNG OAc)2/8a-catalyzed Heck reaction,
[a] Reaction conditions as in the footnote of Table 2. [b] Yield of isolated
where the best yields were obtained with aryl halides bear-
product; average of two runs. [c] Base=K2CO3. [d] Base=K3PO4.
1588
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1583 – 1593

Cobalt-Containing Secondary Phosphine Oxides
FULL PAPER
Table 11 shows the results obtained for Heck reactions as-
sisted by 7a. The trend previously observed in Table 9 is
also apparent here: the catalytic performance was reduced
theory (DFT) has proved itself repeatedly as a useful tool in
providing reliable results in studies on transition-metal-
mediated catalytic reactions.[18] This method at the B3LYP
level was utilized to examine the validity of the proposed
routes for transformation of 8 to 8’ and that of its organic
ꢀ
Table 11. Dependence of the yield of Heck reaction with 7a as ligand on
substrate.[a]
counterpart PhC CP(=O)(H)ACHTER(UNG tBu) (6) to 6’ (Scheme 2). To
make the computations feasible, smaller model compounds
replaced the real reaction participants. Thus, the tBu and Ph
groups were replaced by Me and H groups, respectively.
Two probable routes for the transformation of secondary
phosphine oxides to phosphinous acids were proposed and
examined. The first pathway, referred to as the monomer
route, is based on intramolecular hydrogen migration. The
second pathway, or dimer route, is based on intermolecular
hydrogen transfer.
Entry
Substrate (R)
Yield [%][b]
A/B
1
2
3
4
5
6
7
4-NO2
4-COCH3
4-COH
H
4-CH3
4-OCH3
2-CH3
93.5
86.2
75.1
>99.0
97.1
>99.0
96.8
2/98
4/96
8/92
1/99
1/99
10/90
2/98
[a] Reaction conditions as in the footnote of Table 2. The reaction was
carried out in DMF at 1208C for 7 h. [b] Yield of isolated product; aver-
age of two runs.
DFT studies on the monomer route: In the monomer route,
conversion of the phosphine oxide to the phosphinous acid
takes place via an intramolecular hydrogen migration
(Scheme 6).
when electron-withdrawing groups were present in the sub-
strate Table 11 (Table 11, entries 1–3). On the contrary,
almost quantitative yields were observed with electron-do-
nating groups (Table 11, entries 5 and 6) even in the case of
o-bromotoluene (Table 11, entry 7). This indicates that the
steric effect is much more important in the case of 8a than
in the case of 7a.
Scheme 6. Conversion of a phosphine oxide to a phosphinous acid via the
monomer route (Case 1A: R=CH3; Case 1B: R=CF3).
For further comparison, the Heck reaction between bro-
motoluene and styrene employing in-situ-prepared Pd-
ACHTREUNG
A
In the case of the monomer route for an organic phosphine
oxide, the geometry-optimized molecular structures, includ-
ing the transition state, on the way from 1PO to 1POH
(Case 1A) are shown in Figure 5. The outward appearance
of 1PO on the B3LYP potential energy surface is a distorted
tetrahedron (Td) with the phosphorus atom at its center.
The configuration of 1POH is a triangular pyramid with the
phosphorus atom situated on top. On the 1PO!1TS!
1POH pathway, H1 (from P1 of 1PO) is transferred to O1
Table 12. Dependence of the yield of palladium-catalyzed Heck reaction
with 8a as ligand on base, reaction time and temperature.[a]
Entry
Base
Time [h]
T [8C]
Yield [%][b]
1
2
3
4
5
6
7
8
9
K2CO3
KF
K3PO4
CsF
Cs2CO3
K2CO3
K2CO3
K2CO3
K2CO3
7
7
7
7
7
5
7
7
13
110
110
110
110
110
120
120
130
130
42.7
33.2
11.9
trace
trace
29.4
47.8
59.8
60.5
À
and a new O1 H1 bond is generated. Except for H1, the
relative positions of all other atoms remain almost un-
changed. The O1-P1-H1 bond angle is 115.18 in 1PO. In
1POH the P1-O1-H1 bond angle is 108.68. The two methyl
carbon atoms, P1, and H1 are almost coplanar in 1TS.
Similar geometry-optimized molecular structures, includ-
ing the transition state on the way from FPO to FPOH,
[a] Reaction conditions as in the footnote of Table 2. [b] Yield of isolated
product; average of two runs.
The best catalytic performance
was obtained with PdACHTRE(UNG OAc)2/
8a=1/2, DMF as solvent, and
K2CO3 as base. In general,
lower efficiencies were ob-
served than with in-situ-pre-
pared PdACHTREUNG(OAc)2/8b.
Computational studies on the
tautomerism of6 and 6 ’ as well
Figure 5. Optimized structures of stationary points pertinent to Case 1A showing bond lengths [Š] and
as 8 and 8’: Density functional angles [8].
Chem. Eur. J. 2007, 13, 1583 – 1593
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1589

F.-E. Hong et al.
Figure 6. Optimized structures of stationary points pertinent to Case 2 showing bond lengths [Š] and angles [8].
were observed for Case 1B, where both methyl groups are
replaced by CF3 substituents.
bond angle is 109.68. There is little geometric change on the
side of the metal moiety during the conversion.
In the case of the monomer route for a metal-containing
phosphine oxide (Case 2), the geometry-optimized molecu-
lar structures, including the transition states on the way
from MPO to MPOH, are shown in Figure 6. Results were
similar to those obtained for Case 1. The phosphorus atom
is located in the center of a distorted tetrahedron in MPO
on the B3LYP potential energy surface. An analogous trian-
gular pyramid was observed for MPOH with the phosphorus
atom located at the top. On the MPO!MTS!MPOH
pathway, H4 (from P4 of MPO) is transferred to O4 of
DFT studies on the dimer route: As shown in Scheme 7, the
transformation of phosphine oxide into phosphinous acid is
possible via intermolecular hydrogen transfer.
In the case of the dimer route for an organic phosphine
oxide (Case 3), the geometry-optimized molecular struc-
tures, including transition states on the way from 2PO to
2POH, are depicted in Figure 7. First, two phosphine oxide
molecules approach each other and adopt the proper orien-
tation. Then, two hydrogen atoms are transferred synchro-
nously to the oxygen atoms of the opposite molecules. Final-
ly, the two phosphinous acid molecules formed move away
from each other. In 2PO six
À
4POH with creation of a new O4 H4 bond. The O4-P4-H4
bond angle is 115.08 in MPO. In MPOH, the P4-O4-H4
atoms, H2, P2, O2, H3, P3, O3,
are arranged in a chair form.
They lie almost in the same
plane in 2TS and 2POH.
Scheme 7. Conversion of a phosphine oxide to a phosphinous acid via the dimer route (R=CH3)
Figure 7. Optimized structures of stationary points pertinent to Case 3 showing bond lengths [Š] and angles [8].
1590
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1583 – 1593

Cobalt-Containing Secondary Phosphine Oxides
FULL PAPER
Figure 8. Free energies showing bond lengths [Š] and angles [kcalmolÀ1] of states involved in Case 1A (diamond), Case 1B (circle), Case 2 (square), and
Case 3 (triangle), calculated at the B3LYP level of theory.
The potential energy surfaces of the proposed transforma-
tions of Cases 1A, 1B, 2, and 3 are depicted in Figure 8.
First, in the intramolecular transfer process the activation
energy for the metal-containing species is lower than that
for its organic counterpart. It is 55.8 kcalmolÀ1 for Case 2
versus 59.6 kcalmolÀ1 and 57.7 kcalmolÀ1 for Case 1A and
Case 1B, respectively. Obviously, the presence of the metal
fragment assists in lowering the activation energy. The re-
placement of CH3 groups by the strongly electron-withdraw-
ing CF3 groups also reduces the activation energy. Bis(tri-
fluoromethyl)phosphinous acid is also more stable than the
corresponding phosphine oxide and dimethylphosphinous
acid. In fact, it has been shown experimentally that bis(tri-
fluoromethyl)phosphinous acid is not prone to rearrange to
the oxide.[19] Second, the activation energy is significantly
lower for the intermolecular hydrogen transfer process
(Case 3). The severe strain caused by formation of a three-
membered ring during the intramolecular hydrogen migra-
tion processes (Cases 1 and 2) is avoided in the intermolecu-
lar processes (Case 3).
Summary
We have demonstrated that novel cobalt-containing secon-
dary phosphine oxides 8a and 8b could be prepared by the
reactions of corresponding secondary phosphine oxides 6a
and 6b with dppm-bridged dicobalt complex 2. Using 8a as
ligand in palladium-catalyzed Heck reactions led to poor re-
sults, probably due to the extremely low quantity of its tau-
tomeric form 8a’ in solution. Nevertheless, reactions em-
ploying 8b under the same conditions are very efficient.
Density functional calculations on the transformation 8!8’
supports the experimental results. The dimer route has been
shown to have a much lower activation energy than the mo-
nomer route for the transformation 8!8’. In the dimer
route, two reacting molecules must approach each other and
arrange themselves in the proper orientation. Since the tBu
group of 8a is much larger than the Ph group of 8b, it is
much more difficult for two 8a molecules to engage appro-
priately. Therefore, the catalytic reaction relying on the
transformation 8a!8a’ is much less effective.
It has been shown that in the monomer route the transfor-
mation of the transition metal-containing compound pro-
ceeds with a lower activation energy. It has also been dem-
onstrated that the conversion via the dimer route is much
more feasible than that via the monomer route. However,
due to the severe steric hindrance caused by the bulkiness
of the metal-containing compound, the dimer route is prob-
ably not feasible for a compound with a bulky substituent,
such as 8a.[20] Therefore, the 8a!8a’ conversion is less
likely. However, our experimental observation that 8b is an
efficient ligand in the Pd-catalyzed Heck reaction between
bromobenzene and styrene suggests that the transformation
8b!8b’ indeed occurs to some extent in solution.
Experimental Section
General information: All manipulations were carried out under a dry ni-
trogen atmosphere. Solvents and deuterated solvents were purified
before use. Products were mostly separated by centrifugal thin layer
chromatography (CTLC, Chromatotron, Harrison model 8924). The 1H
and 31P NMR spectra were recorded on a Varian-400 spectrometer at
400.44 and 162.10 MHz, respectively; 13C NMR spectra were recorded on
a Varian VXR-300S spectrometer at 75.43 MHz. Chemical shifts are re-
ported in parts per million relative to the residual proton signals of
CDCl3 or CD2Cl2. Mass spectra were recorded on a JOEL JMS-SX/SX
102 A GC/MS/MS spectrometer. Elemental analyses were obtained on a
Heraeus CHN-O-S-Rapid instrument.
7a and 7a’: A solution of [Co2(CO)6ACHTER(UNG dppm)] (2, 1.340 g, 2.000 mmol) and
ꢀ
PhC CP
G
(NEt2) (5a, 0.523 g, 2.000 mmol) in THF (15 mL) was stir-
red at 708C for 16 h. Then the solvent was removed under reduced pres-
Chem. Eur. J. 2007, 13, 1583 – 1593
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1591

F.-E. Hong et al.
sure, and the resulting dark red residue was separated by CTLC. A dark
red band was eluted with hexanes/CH2Cl2 (1:1). The product was identi-
fied as 7a (1.365 g, 1.560 mmol, 78.0% yield). A similar procedure was
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
Computational methods: All calculations were carried out using the
Gaussian03 package[23] with the tight criterion (10À8 Hartree) as default
for the SCF convergence. The molecular geometries were fully optimized
with the hybrid B3LYP-DFT method under C1 symmetry with the Becke
three-parameter exchange functional[24] and the Lee–Yang–Parr correla-
tion functional.[25] LANL2DZ including the double-z basis sets for the va-
lence and outermost core orbitals combined with pseudopotential was
used for the preparation of 7a’: A solution of [Co2(CO)6ACHTREU(NG dppm)] (2,
ꢀ
1.340 g, 2.000 mmol) and PhC CP(=O)
G
(NEt2) (0.555 g, 2.000 mmol)
in THF (15 mL) was stirred at 708C for 16 h and then worked up. A dark
red band was eluted with ethyl acetate/CH2Cl2 (1:1). The product was
identified as 7a’ (1.568 g, 1.760 mmol, 88.0% yield).
7a: 1H NMR (CDCl3): d=7.39–6.75 (m, 24H; arene), 3.48 (m, 1H;
dppm), 3.27 (m, 4H; NEt2), 3.11 (m, 1H; dppm), 1.45 (t, 6H; NEt2),
1.10–1.19 ppm (d, 9H; PtBu); 13C NMR (CDCl3, d/ppm): 133.07–127.86
(arene), 30.43 (dppm); 31P NMR (CDCl3, d/ppm): 37.80–36.72 (d, dppm),
86.52 ppm (s, PNEt2); MS (FAB): m/z: 876.5 [M+1]+; elemental analysis
(%) calcd for 7a: N 1.60, C 61.71, H 5.26; found: N 1.11, C 61.20, H
54.88.
used for Co,[26,27] and 6-31G
ACHTRE(UNG d,p) basis sets were employed for the other
atoms, which has proven successful in describing the cobalt-mediated
Pauson–Khand reaction.[28] All the stationary points found were charac-
terized via harmonic vibrational frequency analysis as minima (number
of imaginary frequencies Nimag =0) and transition states (Nimag =1). For
the determination of transition-state geometry, intrinsic reaction coordi-
nate (IRC)[29] analyses followed geometry optimization to ensure the
transition structures are smoothly connected by two proximal minima
along the reaction coordinate. Thermodynamic quantities, calculated
electronic energies, enthalpies of reaction DHR, and free activation ener-
gies DG°, were obtained and corrected at constant pressure and 298 K.
Stability analysis[30] was performed to determine whwether the Kohn–
Sham (KS) solutions are stable with respect to variations which break
spin and spatial symmetry. Tables listing coordinates and energies of the
optimized stationary points are available as Supporting Information
7a’: 1H NMR (CDCl3): d=7.81–6.71 (m, 24H; arene), 3.25 (m, 1H;
dppm), 3.41 (m, 1H; dppm), 2.72 (m, 4H; NEt2), 1.69 (t, 6H; NEt2),
1.15 ppm (d, 9H; PtBu); 13C NMR (CDCl3): d=133.07–127.86 (arene),
30.43 ppm (dppm); 31P NMR (CDCl3): d=49.28 (s, P(=O)
ACHTREU(NG NEt2)),
35.27 ppm (d, dppm); MS (FAB): m/z: 892.6 [M+1]+; elemental analysis
(%) calcd for 7a’: N 1.57, C 60.51, H 5.16; found: N 1.31, C 59.51, H
5.23.
ꢀ
8a and 8b:
2.000 mmol) in THF (5 mL) was added to [Co2(CO)
A
solution of PhC CP(=O)(H)
G
N
2.000 mmol) and THF (10 mL). The mixture was stirred for 24 h at room
temperature. The solvent was removed under reduced pressure, and the
resulting dark red residue purified by CTLC. A dark red band was eluted
with CH2Cl2/ethyl acetate (10:1). It was identified as 8a (1.246 g,
1.520 mmol, 76.0% yield). A similar procedure was used for the prepara-
Acknowledgements
We thank the National Science Council of the R.O.C. (Grant NSC 94-
2113M-005-011) for financial support.
tion of 8b: A solution of [Co2(CO)6ACHTER(UNG dppm)] (2, 1.340 g, 2.000 mmol) and
ꢀ
PhC CP(Ph)
A
red at 508C for 8 h and then worked up. The solvent was removed under
reduced pressure, and the resulting dark red residue was purified by
CTLC. A dark red band was eluted with ethyl acetate/CH2Cl2 (1:1). The
product was identified as 8b (1.546 g, 1.840 mmol, 92.0% yield). Appa-
rently, the phosphorus atom was oxidized during the chromatographic
[1] a) J. Hassan, M. SØvingnon, C. Gozzi, E. Schulz, M. Lemaire, Chem.
Rev. 2002, 102, 1359–1469; b) W. Tang, X. Zhang, Chem. Rev. 2003,
103, 3029–3070; c) N. G. Andersen, B. A. Keay, Chem. Rev. 2001,
101, 997–1030; d) C. A. Bessel, P. Aggarwal, A. C. Marschilok, K. J.
Takeuchi, Chem. Rev. 2001, 101, 1031–1066; e) T. E. Barder, S. D.
Walker, J. R. Martinelli, S. L. Buchwald, J. Am. Chem. Soc. 2005,
127, 4685–4696; f) S. D. Walker, T. E. Barder, J. R. Martinelli, S. L.
Buchwald, Angew. Chem. 2004, 116, 1907–1912; Angew. Chem. Int.
Ed. 2004, 43, 1871–1876; g) J. F. Hartwig, Angew. Chem. 1998, 110,
2154–2177; Angew. Chem. Int. Ed. 1998, 37, 2046–2067; h) G.
Mann, J. F. Hartwig, J. Am. Chem. Soc. 1996, 118, 13109–13110;
i) N. Kataoka, Q. Shelby, J. P. Stambuli, J. F. Hartwig, J. Org. Chem.
2002, 67, 5553–5566; j) A. F. Littke, C. Dai, G. C. Fu, J. Am. Chem.
Soc. 2000, 122, 4020–4028; k) J. G. Planas, J. A. Gladysz, Inorg.
Chem. 2002, 41, 6947–6949; l) Q.-S. Hu, Y. Lu, Z.-Y. Tang, H.-B.
Yu, J. Am. Chem. Soc. 2003, 125, 2856–2857; m) T. E. Pickett, F. X.
Roca, C. J. Richards, J. Org. Chem. 2003, 68, 2592–2599; n) Z.-Y.
Tang, Y. Lu, Q.-S. Hu, Org. Lett. 2003, 5, 297–300.
process.
1
8a: 1H NMR (CDCl3): d=7.42–7.01 (m, 25H; arene), 7.32 (d, JP,H
=
457.8 Hz, 1H; PH), 3.33 (m, 1H; dppm), 3.00 (m, 1H; dppm), 0.883 ppm
(br, 9H; tBu); 31P NMR (CDCl3): d=47.2 (1JP, H =457.8 Hz, 1P, P(=O)-
ACHTREUNG(tBu)(H)) 35.8 (d, 1P, dppm), 33.8 ppm (d, 1P, dppm); MS (FAB): m/z:
821.4 [M+1]+; elemental analysis (%) calcd for 8a: C 60.09, H 4.56;
found: C 60.23, H 4.06.
8b: 1H NMR (CDCl3): d=8.92 (d, 1JP, H =483.4 Hz, 1H; PH), 7.69–6.98
(m, 29H; arene), 4.12 (m, 1H; dppm), 3.29 ppm (m, 1H; dppm);
13C NMR (CDCl3): d=133.07–127.86 (arene), 30.43 ppm (dppm);
31P NMR (CDCl3): d=36.8 (d, 1P; dppm), 35.7 (d, 1P; dppm), 21.5 ppm
(1JP, H =483.4 Hz, 1P; P(=O)(Ph)(H)); MS (FAB): m/z: 841.4 [M+1]+; el-
emental analysis (%) calcd for 8b: C 61.43, H 3.93; found: C 60.82, H
3.41.
[2] a) O. Navarro, R. A. Kelly III, S. P. Nolan, J. Am. Chem. Soc. 2003,
125, 16194–16195; b) M. S. Viciu, R. A. Kelly III, E. D. Stevens, F.
Naud, M. Studer, S. P. Nolan, Org. Lett. 2003, 5, 1479–1482; c) J.
Yin, M. P. Rainka, X. X. Zhang, S. L. Buchwald, J. Am. Chem. Soc.
2002, 124, 1162–1163; d) J. P. Stambuli, R. Kuwano, J. F. Hartwig,
Angew. Chem. 2002, 114, 4940–4942; Angew. Chem. Int. Ed. 2002,
41, 4746–4748.
[3] a) S. Eichenseher, O. Delacroix, K. Kromm, F. Hampel, J. A. Gla-
dysz, Organometallics 2005, 24, 245–255; b) O. Delacroix, J. A. Gla-
dysz, Chem. Commun. 2003, 665–675; c) K. Kromm, F. Hampel,
J. A. Gladysz, Organometallics 2002, 21, 4264–4274.
X-ray crystallographic studies: Suitable crystals of 7a, 7a’, 8a, and 8b
were sealed in thin-walled glass capillaries under nitrogen atmosphere
and mounted on a Bruker AXS SMART 1000 diffractometer. Intensity
data were collected in 1350 frames with increasing w (width of 0.38 per
frame). The absorption correction was based on symmetry-equivalent re-
flections using the SADABS program. The space group determination
was based on a check of the Laue symmetry and systematic absences,
and was confirmed by means of the structure solution. The structure was
solved by direct methods using a SHELXTL package.[21] All non-H
atoms were located from successive Fourier maps and hydrogen atoms
were refined by using a riding model. Anisotropic thermal parameters
were used for all non-H atoms, and fixed isotropic parameters were used
for H atoms.[22] Crystallographic data of 7a, 7a’, 8a, and 8b are summar-
ized in Table 1.
[4] a) T. J. Colacot, Chem. Rev. 2003, 103, 3101–3118; b) R. C. J. Atkin-
son, V. C. Gibson, N. J. Long, Chem. Soc. Rev. 2004, 33, 313–328.
[5] F.-E Hong, C.-P. Chang, Y.-C. Chang, Dalton Trans. 2003, 3892–
3897.
CCDC-615123 (7a), CCDC-296107 (7a’), CCDC-615124 (8a) and
CCDC-615125 (8b) contain the supplementary crystallographic data for
this paper. These data can be obtained free of charge from the Cam-
[6] a) F.-E Hong, Y.-C. Chang, R.-E Chang, S.-C. Chen, B.-T. Ko, Orga-
nometallics 2002, 21, 961–967; b) F.-E Hong, C.-P. Chang, Y.-C.
Chang, Dalton Trans. 2003, 3892–3897; c) F.-E Hong, Y.-C. Lai, Y.-J.
1592
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1583 – 1593

Cobalt-Containing Secondary Phosphine Oxides
FULL PAPER
Ho, Y.-C. Chang, J. Organomet. Chem. 2003, 688, 161–167; d) F.-E
Hong, Y.-C. Chang, C.-P. Chang, Y.-L. Huang, Dalton Trans. 2004,
157–165; e) F.-E Hong, Y.-J. Ho, Y.-C. Chang, Y.-C. Lai, Tetrahe-
dron 2004, 60, 2639–2645; f) F.-E Hong, Y.-J. Ho, Y.-C. Chang, Y.-L.
Huang, J. Organomet. Chem. 2005, 690, 1249–1257; g) C.-P. Chang,
Y.-L. Huang, F.-E Hong, Tetrahedron 2005, 61, 3835–3839; h) J.-C.
Lee, M.-G. Wang, F.-E Hong, Eur. J. Inorg. Chem. 2005, 5011–5017;
i) Y.-C. Chang, F.-E Hong, Organometallics 2005, 24, 5686–5695.
[7] a) K. R. Dixon, A. D. Rattray, Can. J. Chem. 1971, 49, 3997–4004;
b) G. O. Doak, L. D. Freedman, Chem. Rev. 1961, 61, 31–44; c) B.
Silver, Z. Luz, J. Am. Chem. Soc. 1962, 84, 1091–1905; d) Z. Luz, B.
Silver, J. Am. Chem. Soc. 1961, 83, 4518–4521; e) Z. Luz, B. Silver,
J. Am. Chem. Soc. 1962, 84, 1095–1098.
[8] a) G. Y. Li, Angew. Chem. 2001, 113, 1561–1564; Angew. Chem. Int.
Ed. 2001, 40, 1513–1516; b) G. Y. Li, G. Zheng, A. F. Noonan, J.
Org. Chem. 2001, 66, 8677–8681; c) L. Ackermann, R. Born, J. H.
Spatz, D. Meyer, Angew. Chem. 2005, 117, 7382–7386; Angew.
Chem. Int. Ed. 2005, 44, 7216–7219; d) L. Ackermann, R. Born,
Angew. Chem. 2005, 117, 2497–2500; Angew. Chem. Int. Ed. 2005,
44, 2444–2447; e) I. Pryjomska, H. B. Bechowski, Z. Ciunik, A. M.
Trzeciak, J. J. Ziółkowski, Dalton Trans. 2006, 213–220.
751–760; e) P. Fristrup, S. L. Quement, D. Tanner, P.-O. Norrby, Or-
ganometallics 2004, 23, 6160–6165.
[16] N. Kataoka, Q. Shelby, J. P. Stambuli, J. F. Hartwig, J. Org. Chem.
2002, 67, 5553–5566.
[17] a) Z. Xiong, N. Wang, M. Dai, A. Li, J. Chen, Z. Yang, Org. Lett.
2004, 6, 95–98; b) R. Singh, M. S. Viciu, N. Kramareva, O. Navarro,
S. P. Nolan, Org. Lett. 2005, 7, 1829–1832; c) C. Dai, G. C. Fu, J.
Am. Chem. Soc. 2001, 123, 2719–2724; d) M. E. Limmert, A. H.
Roy, J. F. Hartwig, J. Org. Chem. 2005, 70, 9364–9370.
[18] a) J. A. Pople, Angew. Chem. 1999, 111, 2014–2023; Angew. Chem.
Int. Ed. 1999, 38, 1894–1902; b) W. J. Hehre, L. Radom, P. von R.
Schleyer, J. A. Pople, Ab Initio Molecular Orbital Theory, Wiley-In-
terscience, New York, 1986; c) W. Koch, M. C. Holthausen, A Chem-
istꢀs Guide to Density Functional Theory, Wiley-VCH, Weinheim,
2000; d) T. Ziegler, Chem. Rev. 1991, 91, 651–667; e) E. R. David-
son, Chem. Rev. 2002, 102, 351–352.
[19] J. E. Griffiths, A. B. Burg, J. Am. Chem. Soc. 1962, 84, 3442–3450.
[20] D. G. Ho, R. Gao, J. Celaje, H.-Y. Chung, M. Selke, Science 2003,
302, 259–262.
[21] G. M. Sheldrick, SHELXTL PLUS Userꢁs Manual, Revision 4.1,
Nicolet XRD Corporation, Madison, Wisconsin, 1991.
[9] a) G. Baba, J.-F. Pilard, K. Tantaoul, A.-C. Gaumont, J.-M. Denis,
Tetrahedron Lett. 1995, 36, 4421–4424; b) J.-F. Pilard, G. Baba, A.-
C. Gaumont, J.-M. Denis, Synlet 1995, 1168–1170.
[10] a) R. G. Pearson, J. Am. Chem. Soc. 1963, 85, 3533–3539; b) Hard
and Soft Acids and Bases (Ed.: R. G. Pearson), Dowden, Hutchin-
son, and Ross, Stroudsburg, PA, 1973; c) R. G. Pearson, Chemical
hardness - A historical introduction, Struct. Bonding (Berlin) 1993,
80, 1–10.
[11] a) J. P. Farr, M. M. Olmstead, F. E. Wood, A. L. Balch, J. Am. Chem.
Soc. 1983, 105, 792–798; b) E. Linder, H. Rauleder, P. Wegner, Z.
Naturforsch. B 1984, 39, 1224–1229; c) K. Kurtev, D. Ribola, R. A.
Jones, D. J. Cole-Hamilton, G. Wilkinson, J. Chem. Soc. Dalton
Trans. 1980, 55–58.
[12] a) Z.-Z Zhang, H. Cheng, Coord. Chem. Rev. 1996, 147, 1–39;
b) Z.-Z. Zhang, H.-K. Wang, Z. Xi, X.-K. Yao, R.-J. Wang, J. Orga-
nomet. Chem. 1989, 376, 123–131; c) Z.-Z. Zhang, H.-K. Wang, H.-
G. Wang, R.-J. Wang, J. Organomet. Chem. 1986, 314, 357–367;
d) J. P. Farr, M. M. Olmstead, N. M. Rutherford, F. E. Wood, A. L.
Balch, Organometallics 1983, 2, 1758–1762; e) J. P. Farr, F. E. Wood,
A. L. Balch, Inorg. Chem. 1983, 22, 1229–1235; f) J. P. Farr, M. M.
Olmstead, A. L. Balch, Inorg. Chem. 1983, 22, 3387–3393; g) R. J.
Puddephatt, Chem. Soc. Rev. 1983, 12, 99–127.
[22] The hydrogen atoms rode on carbon or oxygen atoms in their ideal-
À
ized positions and were held fixed with C H distances of 0.96 Š.
[23] Gaussian03, RevisionC.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomer-
y, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyen-
gar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussian, Inc., Walling-
ford, CT, 2004.
[24] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[25] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[26] T. H. Dunning, Jr., P. J. Hay, Modern Theoretical Chemistry, Plenum,
New York, 1976, p. 1.
[27] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299–310.
[28] M. Yamanaka, E. Nakamura, J. Am. Chem. Soc. 2001, 123, 1703–
1708.
[13] a) E. R. Goldman, I. L. Medintz, J. L. Whitley, A. Hayhurst, A. R.
Clapp, H. T. Uyeda, J. R. Deschamps, M. E. Lassman, H. Mattoussi,
J. Am. Chem. Soc. 2005, 127, 6744–6751; b) A. A. C. Braga, N. H.
Morgon, G. Ujaque, F. Maseras, J. Am. Chem. Soc. 2005, 127, 9298–
9307.
[14] a) N. Miyaura, K. Yamada, H. Suginome, A. Suzuki, J. Am. Chem.
Soc. 1985, 107, 972–980; b) G. A. Grasa, A. C. Hillier, S. P. Nolan,
Org. Lett. 2001, 3, 1077–1080; c) S. R. Dubbaka, P. Vogel, Org. Lett.
2004, 6, 95–98.
[29] a) C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154–2161;
b) C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523–5527.
[30] a) R. Seeger, J. A. Pople, J. Chem. Phys. 1977, 66, 3045–3050; b) R.
Bauernschmitt, R. Ahlrichs, J. Chem. Phys. 1996, 104, 9047–9052.
[15] a) R. F. Heck, J. P. Nolley, Jr., J. Org. Chem. 1972, 37, 2320–2322;
b) I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100, 3009–
3066; c) Q. Yao, E. P. Kinney, Z. Yang, J. Org. Chem. 2003, 68,
7528–7531; d) J. Mo, L. Xu, J. Xiao, J. Am. Chem. Soc. 2005, 127,
Received: July 20, 2006
Published online: November 8, 2006
Chem. Eur. J. 2007, 13, 1583 – 1593
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1593