An exploratory study of regiocontrol iii the heck type reaction. Influence of solvent polarity and bisphosphine ligands

Article abstract of DOI:10.1021/om9803265

The regiochemistry of the addition of arylpalladium species to styrene and propene has been studied. It was found that for the reaction of P2Pd(Ph)X the counterion X, the polarity of the solvent, and the structure of the bisphosphine ligand ?2 all have an influence. Using bis(diphenylphosphino)ethane as ligand, triflate as counterion, and a moderately polar solvent mixture, 98% selectivity for addition of the phenyl group to the /?-carbon of styrene could be obtained. Finally, using low-temperature NMR, some of the palladium intermediates in the addition could be observed, specifically those where the palladium species is stabilized by ?/3-allylic interaction with a phenyl group.

Full text of DOI:10.1021/om9803265

970
Organometallics 1999, 18, 970-975
An Exp lor a tor y Stu d y of Regiocon tr ol in th e Heck Typ e
Rea ction . In flu en ce of Solven t P ola r ity a n d
Bisp h osp h in e Liga n d s
Maik Ludwig,* Staffan Stro¨mberg, Mats Svensson, and Bjo¨rn Åkermark*,†
Department of Chemistry, Organic Chemistry, The Royal Institute of Technology,
S-10044 Stockholm, Sweden
Received April 27, 1998
The regiochemistry of the addition of arylpalladium species to styrene and propene has
been studied. It was found that for the reaction of P2Pd(Ph)X the counterion X, the polarity
of the solvent, and the structure of the bisphosphine ligand P2 all have an influence. Using
bis(diphenylphosphino)ethane as ligand, triflate as counterion, and a moderately polar solvent
mixture, 98% selectivity for addition of the phenyl group to the â-carbon of styrene could be
obtained. Finally, using low-temperature NMR, some of the palladium intermediates in the
addition could be observed, specifically those where the palladium species is stabilized by
η3-allylic interaction with a phenyl group.
Sch em e 1. Gen er a l Mech a n ism of th e Heck
Rea ction
In tr od u ction
The Heck reaction is one of the most important
methods for the preparation of aryl-functionalized alk-
enes, and it is used both in synthetic chemistry and in
the pharmaceutical industry. The scope and the limita-
tions of the reaction have been reviewed in a number
of recent publications.1 The reaction is often referred
to as a selective reaction, and a large number of
synthetically optimized Heck reactions are found in the
literature. However, general and good rationales for the
observed selectivity are rare. For example, it has been
observed that regioselectivity is strongly affected by the
substitution pattern of the alkene and by the use of
mono vs bidentate ligands.1b The selectivity is rational-
ized mainly on the basis of empirical results. It therefore
seems likely that the development of the Heck reaction
into a more general procedure which tolerates different
starting materials and is highly regioselective requires
a deeper understanding of the mechanism, including
probable intermediates and transition states.
Resu lts a n d Discu ssion
As a first attempt toward fine-tuning the catalyst for
higher stereo- and regioselectivity, we set out to identify
intermediates in the reaction between arylpalladium
species and the alkenes styrene and propene using
diphosphines as ligands for the catalyst. Styrene seemed
particularly interesting as substrate because it is known
to react in a nonregioselective fashion in the presence
of a Pd(OAc)2/dppp catalyst (dppp ) bis(diphenylphos-
phino)propane).2 To improve the understanding of the
factors which determine regioselectivty in Heck type
reactions, we have varied the steric and electronic
factors by varing solvent, the counterions, and the
bisphosphine ligands.
In the now-classic Heck cycle (Scheme 1),3 the regio-
chemistry in the reaction of a monosubstituted alkene
with an arylpalladium species is determined by the
balance between addition of the aryl group at the
terminal or internal position of the alkene. The cycle is
completed by â-elimination, generating a palladium
hydride species.
The classic Heck cycle has been modified by Cabri et
al.1b and Hayashi et al.,3d who have shown that the
counterion (X in Scheme 1)4 plays an important role.
(3) (a) Cabri, W.; Candiani, I.; Bedeschi, A.; Penco, S. J . Org. Chem.
1992, 57, 1481. (b) J utand, A.; Mosleh, A. Organometallics 1995, 14,
1810. (c) Kawataka, F.; Shimizu, I.; Yamamoto, A. Bull. Chem. Soc.
J pn. 1995, 68, 654. (d) Hayashi, T.; Kubo, A.; Ozawa, F. Pure Appl.
Chem. 1992, 64, 421.
(4) While this work was in progress, a related and complementary
study on the addition of arylpalladium triflates to methyl acrylate was
published: Brown, J . M.; Hii, K. K. Angew. Chem. 1996, 108, 679;
Angew. Chem., Int. Ed. Engl. 1996, 35, 657.
† E-mail: bear@orgchem.kth.se.
(1) Recent reviews: (a) de Meijere, A.; Meyer, F. E. Angew. Chem.
1994, 106, 2473; Angew. Chem., Int. Ed. Engl. 1994, 33, 2379. (b) Cabri,
W.; Candiani, I. Acc. Chem. Res. 1995, 28, 2.
(2) Cabri, W.; Candiani, I.; Bedeschi, A. J . Org. Chem. 1992, 57,
3558.
10.1021/om9803265 CCC: $18.00 © 1999 American Chemical Society
Publication on Web 02/12/1999

Regiocontrol in the Heck Type Reaction
Organometallics, Vol. 18, No. 6, 1999 971
Sch em e 2. P r ep a r a tion of th e Com p lexes P 2P d (P h )X
Ta ble 1. 31P NMR Da ta of th e P h en ylp a lla d iu m
Com p ou n d s P 2P d (P h )X (1-4) a t 0 °C in DMF
u n less Oth er w ise Sta ted
When X ) halide, it probably remains coordinated
during the cycle and one of the ligands L has to
dissociate prior to coordination of the alkene. In con-
trast, when X is less strongly coordinating, e.g. X )
triflate (OTf), it dissociates to give a cationic palladium
species. The alkene then coordinates and the ligands L
remain coordinated throughout the catalytic cycle.1b,3d
The intermediate palladium hydride species are known
to isomerize alkenes, presumably by readdition and
renewed â-elimination. This may preclude detection of
the primary intermediates in a catalytic reaction. We
have therefore chosen to perform the reactions under
stoichiometric conditions, using well-defined arylpalla-
dium complexes as reactants.
δ(Pa) δ(Pb) J (Pa,Pb) δ(P)a
X (compd no.) (ppm) (ppm)
(Hz)
(ppm)
(dppp)Pd(Ph)X I (1a )b
I (1a )
13.1
12.9
25.8
26.2
26.2
19.4
50.7
56.4
67.0
75.1
56.8
58.8
-8.0
-8.4
-1.7
-1.3
-1.3
-3.7
35.3
39.0
61.7
66.8
49.3
52.8
52.8
53.2
48.9
50.4
50.4
48.5
27.7
25.0
18.8
15.8
14.9
d
-17.3
OTf (1b)
PF6 (1c)c
BF4 (1d )
OAc (1e)
(dppe)Pd(Ph)X I (2a )b
-12.5
0.5
OTf (2b)
(dcpe)Pd(Ph)X I (3a )b
OTf (3b)
(dppet)Pd(Ph)X I (4a )b
OTf (4b)
-22.7
The complexes P2Pd(Ph)I5 (1a -4a ) were prepared and
then converted into the corresponding complexes P2Pd-
(Ph)X (Scheme 2). The conversions were monitored by
31P NMR: the two doublets of the compounds 1a -4a
were progressively replaced by two doublets at lower
field. In all compounds, the high-field shifts were
assigned to the phosphorus atom Pb in a position trans
to the organo group on the palladium by using P-H
COSY experiments on selected compounds (1a ,b, 2a ,b,
3a ). In the compounds P2Pd(Ph)X (1-3) a strong influ-
ence of the counterion X on the 31P chemical shift δ(31P)
was observed. Among the compounds (dppp)Pd(Ph)X (1)
the δ(31P) increased in the following order: I < OAc <
OTf ≈ PF6, BF4. As might be expected, the influence of
X on δ(31P) is larger for the phosphorus atom situated
in a trans position relative to the exchanged counterion
(Pa, Table 1). The shifts at Pa on going from the strongly
coordinating I- to the noncoordinating OTf- were 12.9
ppm for complex 1, 5.7 ppm for 2, 8.1 ppm for 3, and
2.0 ppm for 4 (Table 1). It is not clear if this means that
the dppp ligand is the most polarizable, but it is evident
from Table 2 that the regiochemistry of the reaction of
the dppp complexes 1 with styrene is strongly affected
by the counterion. Using DMF as the solvent, the
selectivity for 8 was reduced by weakly coordinating
anions, probably because of the increased cationic
character of the phenylpalladium complex. Thus, with
the complex 1 and styrene the selectivity for 8 was
decreased from ca. 80% using iodide or acetate as the
counterion X to ca. 55% (X ) PF6, BF4) (Table 2).
a
Free ligand. b In CDCl3. c δ(PF6) -142.9 ppm, 1J (P-F) ) 709.5
d
Hz. Broad signals.
Ta ble 2. In flu en ce of th e Cou n ter ion X on th e
Regioselectivity of th e P h en yla tion of Styr en e
w ith [(d p p p )P d (P h )]+X- (1a -e) in DMF .
X
7/8
cis/trans for 8
1a
1b
1c
1d
1e
I
20/80
35/65
43/57
42/58
18/82
2/98
OTf
PF6
BF4
OAc
<1/99
<1/99
<1/99
<1/99
The importance of cationic character is also illustrated
by the solvent dependence of the regiochemistry of the
reaction between [(dppp)Pd(Ph)]+OTf- (1b) and styrene.
Using the complex (dppp)Pd(Ph)OTf (1b) in DMF or
acetonitrile solution, there is a balance between addition
of the aryl group at the terminal position (65% or 67%
of 8) and at the internal position (35% or 33% of 7). This
is approximately the same result as in catalytic reac-
tions reported earlier by Cabri et al.2 The selectivity of
65% for 8 that was observed in DMF solution was
lowered to 58% when the polarity was increased by
addition of water (DMF/water 9:1). Likewise, decreasing
the solvent polarity increased the selectivity for 8 to 78%
with DMF/CH2Cl2 (1:1) and further to 92% and 95%
with THF and DMF/CH2Cl2 (1:9), respectively, as
solvent (Table 3). This is clearly in accordance with a
decreased cationic character in the complex 1b as the
polarity of the solvent is decreased.
Also, the bisphosphine ligand P2 had a strong influ-
ence on the regioselectivity of the phenylation of styrene
(Table 4). In a comparison of the reactions of the triflate
complexes in DMF solution, it was found that the dppp
complex 1b gave by far the lowest selectivity for the
product 8 from addition of the phenyl group to the
2-position of styrene, 65%. With the complexes 3b and
(5) (a) Herrmann, W. A.; Thiel, W. R.; Brossmer, C.; O¨ fele, K.;
Priermeier, T.; Scherer, W. J . Organomet. Chem. 1993, 461, 51. (b)
Herrmann, W. A.; Brossmer, C.; Priermeier, T.; O¨ fele, K. J . Organomet.
Chem. 1994, 481, 97. (c) Markies, B. A.; Canty, A. J .; de Graf, W.;
Boersma, J .; J anssen, M. D.; Hoderheide, M. P.; Smets, W. J . J .; Spek,
A. L.; van Koten, G. J . Organomet. Chem. 1994, 482, 191. (d) de Graf,
W.; van Wegen, J .; Boersma, J .; Spek, A. L.; van Koten, G. Recl. Trav.
Chim. Pays-Bas 1989, 108, 275.

972 Organometallics, Vol. 18, No. 6, 1999
Ludwig et al.
F igu r e 1. Reaction products of the phenylation of propene.
Ta ble 3. In flu en ce of th e Solven t on th e
Regioselectivity of th e P h en yla tion of Styr en e
w ith [(d p p p )P d (P h )]+OTf-
solvent
7/8
cis/trans for 8
DMF/H2O (9/1)
DMF
acetonitrile
DMF/CH2Cl2 (1/1)
DMF/CH2Cl2 (1/9)
THF
42/58
35/65
33/67
22/78
5/95
<1/99
<1/99
<1/99
<1/99
<1/99
4/96
8/92
Ta ble 4. In flu en ce of th e Bisp h osp h in e Liga n d P 2
on th e Regioselectivity of th e P h en yla tion of
Styr en e w ith [P 2P d (P h )]+OTf- in DMF
[P2Pd(Ph)]+OTf-
7/8
cis/trans for 8
F igu r e 2. Reaction of (dppp)Pd(Ph)OTf (1b) with 10 equiv
of styrene at -20 °C.
1b
2b
3b
4b
(dppp)Pd(Ph)]+OTf-
(dppe)Pd(Ph)]+OTf-
(dcpe)Pd(Ph)]+OTf-
(dppet)Pd(Ph)]+OTf-
35/65
5/95
16/84
14/86
<1/99
<1/99
2/98
Ta ble 5. Regioselectivity of th e P h en yla tion of
P r op en e w ith P 2P d (P h )X
<1/99
P2Pd(Ph)
X
solvent
9/10
cis/trans for 10
4b ca. 85% selectivity was obtained and the dppe
complex 2b gave very high selectivity for 8, 95%. The
reasons for this great difference in the effects of fairly
closely related ligands are not clear at present. However,
it is interesting to note that these differences correspond
roughly to the changes in the 31P shifts as the P2 ligands
are bound to palladium. Thus, the shift for Pa of the
dppp ligand is ca. 40 ppm to lower field on going to the
triflate complex, while it is 70-80 ppm for the other
three ligands (Table 1). Also, the phosphorus couplings
are very different, ca. 50 Hz for 1b and 25-15 Hz for
2b-4b. These effects might be related to the “bite
angle”, which should be greater for the dppp ligand than
for dppe and the related ligands, as demonstrated by
the values 90.6° for (dppp)PdCl2 and 85.8° for (dppe)-
PdCl2.6 Irrespective of the exact nature of the observed
effects, it is evident that they can be used for controlling
the regiochemistry of the addition of arylpalladium
complexes to styrenes. In fact, by using the dppe
complex 1b and DMF/CH2Cl2 (1:9) as solvent, the
selectivity for the product 8 over the isomer 7 can be
increased to 98%.
The strong influence of alkene structure on the
regiochemistry has been demonstrated in a large num-
ber of earlier studies.1b,2,3a,10 It is also nicely demon-
strated by the reaction between 1b and the fairly
electron rich propene, which in DMF solution gave
1-methylstyrene (9) from addition at the internal posi-
tion as the main product (>85%; Figure 1). In analogy
with observations with styrene as reactant, the amount
of terminally substituted product trans-â-methylstyrene
(10) was increased slightly, from 12 to 19%, when the
(dppp)Pd(Ph) OTf DMF/H2O (9/1) 87/13
1/99
3/97
1/99
12/88
6/94
6/94
OTf DMF
OTf acetonitrile
OTf THF
88/12
88/12
85/15
88/12
81/19
OAc DMF
(dppe)Pd(Ph) OTf DMF
complex 2b (P2 ) dppe) was used in place of 1b (P2
dppp). The influence of the solvent polarity on the
regioselectivity was even smaller (Table 5).
)
To get more information on the mechanism, we have
finally tried to detect intermediates in the phenylation
of styrene. To ensure fast reaction at low temperature,
we have worked with the triflate complex3,4 [(dppp)-
Pd(Ph)]+OTf- (1b). The progress of the reaction and the
appearance of intermediates were monitored by 31P
NMR, and the intermediates were identified by 1D and
2D NMR techniques.
On addition of 2-10 equiv of styrene to a solution of
the complex 1b, in DMF at -20 °C, four new doublets
appeared in the 31P NMR spectrum. These signals
correspond to two new palladium species, one minor (5b)
and one major (6b) (Scheme 3). The kinetic results
(Figure 2) indicate that 5b is the primary product which
is converted to 6b as the reaction proceeds. The struc-
tures of the compounds could be determined by NMR:
for the compound 6b, the 13C DEPT NMR showed three
high-field peaks from the CH2 groups of the dppp ligand
and a double doublet (C1) at 67.4 ppm corresponding to
a CH group directly bounded to palladium. Further-
more, there was a doublet at 13.3 ppm (C2), correspond-
ing to a CH3 group bounded to the carbon C1. The
assignments were confirmed by 1H NMR and H-H and
C-H COSY experiments.
(6) (a) Steffen, W. L.; Palenik, G. J . Inorg. Chem. 1976, 15, 2432.
(b) Hofmann, P.; Heiss, H.; Mu¨ller, G. Z. Naturforsch. 1987, 42B, 395.
(7) We used the hybrid density functional method (DFT) B3LYP with
a double-ú-quality basis set as implemented in GAUSSIAN 94.
(8) Brookhart, M.; Green, M. L. H. J . Organomet. Chem. 1983, 250,
395.
(9) (a) Stevens, R. R.; Shier, G. D. J . Organomet. Chem. 1970, 21,
495. (b) de Felice, V.; Cucciolito, M. E.; De Renzi, A.; Ruffo, F.; Tesauro,
D. J . Organomet. Chem. 1995, 493, 1.
When the reaction between styrene and 1b was
performed in a 9:1 mixture of CD2Cl2 and DMF-d7 at
-20 °C, the conversion of 5b to 6b could be slowed and
5b was formed exclusively in ca. 90% yield. Its 13C NMR
DEPT spectrum showed the three high-field peaks of
the methylene chain of the dppp ligand. In addition,
there was a double doublet at 72.3 ppm corresponding
(10) Crisp, G. T.; Gebauer, M. G. Tetrahedron 1996, 52, 12465.

Regiocontrol in the Heck Type Reaction
Organometallics, Vol. 18, No. 6, 1999 973
F igu r e 3. Conformers with a â-agostic or η3-allylic inter-
action.
to a CH group bonded to palladium (C1) and a doublet
at 34.3 ppm corresponding to a benzylic CH2 group (C2).
The connection between C1 and C2 was again estab-
lished by H-H and C-H COSY NMR.
When the reaction between styrene and 1b was
performed in DMF and allowed to go to complete
conversion to 6b, the organic products 1,1-diphenyleth-
ylene (7) and trans-stilbene (8) were formed in a ratio
of ca. 2:3. The product 8 is clearly derived from the
intermediate 5b, in which the phenyl group has been
added to the terminal position of styrene. Similarly, the
product 7 must stem from addition of the phenyl group
to the internal position of styrene via the intermediate
5b′, which was not observed. Thus, although the two
pathways via 5b and 5b′ are of comparable efficiency
at -20 °C, only 5b could be detected. Perhaps the reason
is a relative stabilization of the intermediate 5b by
coordination of the phenyl group to give a system with
η3-allyl character, which is indicated by theoretical
calculations.7 The difference in the energies of the two
low-energy conformers, the â-agostic8 and the η3-allylic,9
was calculated to be 8 kcal mol-1 in favor of the latter
form (Figure 3). This stabilization is probably also
responsible for the detection of 6b as the final product
in the stoichiometric reaction described in Figure 2. The
regiochemistry of the Heck reaction is determined by
the insertion step, and it is therefore possible that the
energy of the transition state leading to 8 is also affected
by the η3-allylic stabilization.
Con clu sion
A number of factors clearly affect the regiochemistry
of of Heck type reactions. As shown by addition of
cationic phenylpalladium species, the reaction with
propene, an electron-rich alkene, gave mainly pheny-
lation at the more substituted position of the double
bond. In contrast, an electron-poor alkene such as
methyl acrylate gave predominant addition of the phen-
yl group to the less substituted position.4 Styrene
appears to occupy a position intermediate between
propene and acrylate and gave initially approximately
equal addition to both termini of the alkene. The
decrease of the cationic character of the phenylpalla-
dium species, either by an increase in the coordinating
power of the counterion or by a decrease in the polarity
of the solvent, led to an increase in the relative pheny-
lation at the less substituted position of styrene, and in
CH2Cl2/DMF (9:1) stilbene (8) was the major product
(Table 2). With propene, however, no such effect was

974 Organometallics, Vol. 18, No. 6, 1999
Ludwig et al.
3
4J (Pb,C) ) 9.1 Hz, J (Pa,C) ) 1.5 Hz, Cm), 121.6 (Cp); dppp, δ
observed (Table 3). A reasonable explanation for the
result with styrene is to assume that η3-allylic stabiliza-
tion of the transition state will increase in a less polar
solvent. Since such stabilization is only possible in the
transition state leading to 5b, this product would
become favored over 5b′ as the solvent polarity goes
down and as less coordinating counterions are added.
Thus, decreasing the cationic character of the inter-
mediate arylpalladium species is an excellent way of
controlling regiochemistry in the Heck type of addition
to aryl-substituted alkenes. The effect on the addition
to simple alkenes such as propene appears to be small.
This is also true for changes in the structure of the
bisphosphine ligands, which strongly affect the addtion
to styrene. In summary, changes of solvent polarity and
ligand structure have a moderate effect on the addition
of arylpalladium species to unactivated alkenes but
seem to offer a general way of controlling the regio-
chemistry of the addition to arylalkenes. The exact
nature of the ligand effects is not clear, but we are
presently trying to understand them, using quantum
chemical and molecular mechanics calculations in com-
bination with experiments.
28.5 (dd, 1J (Pa,C) ) 25.0 Hz, 3J (Pb,C) ) 7.6 Hz, Pa-CH2), 27.0
1
3
(dd, J (Pb,C) ) 19.0 Hz, J (Pa,C) ) 3.8 Hz, Pb-CH2), 19.0 (d,
2J (Pa,C) ) 3.8 Hz, CH2), 131.0 (d, 1J (Pa,C) ) 50.8 Hz, Ci,a),
132.9 (d, 1J (Pb,C) ) 35.6 Hz, Ci,b), 133.7 (d, 2J (Pa,C) ) 10.6
2
3
Hz, Co,a), 133.1 (d, J (Pb,C) ) 10.6 Hz, Co,b), 128.3 (d, J (Pa,C)
3
) 7.6 Hz, Cm,a), 128.1 (d, J (Pb,C) ) 10.6 Hz, Cm,b), 131.1 (d,
4J (Pa,C) ) 2.3 Hz, Cp,a), 131.1 (d, J (Pb,C) ) 2.3 Hz, Cp,b).
4
(d p p e)P d (P h )I (2a ). 1H NMR: Pd-Ph, δ 7.05 (2H, t,
1J (Ho,Hm) ) 7.7 Hz, Ho), 6.72 (2H, m, Hm), 6.63 (1H, dt,
1J (Hp,Hm) ) 7.2 Hz, Hp; dppe, δ 2.36 (2H, m, Pa-CH2), 2.20
(2H, m, Pb-CH2), 7.33 (8H, m, Ho,a + Hm,a), 7.88 (4H, m, Ho,b),
7.45 (8H, m, Hm,a + Hp,a + Hp,a). 31P{1H} NMR: δ 50.7 (d,
J (Pa,Pb) ) 27.7 Hz, Pa), 35.3 (d, Pb). 13C{1H} NMR: Pd-Ph, δ
154.6 (d, 2J (Pb,C) ) 130.5 Hz, Ci), 137.4 (Co), 127.1 (d, 4J (Pb,C)
) 9.2 Hz, Cm), 122.2 (Cp); dppe, δ 29.4 (dd, 1J (Pa,C) ) 29.8 Hz,
3J (Pb,C) ) 22.1 Hz, Pa-CH2), 24.5 (dd, 1J (Pb,C) ) 25.2 Hz,
3J (Pa,C) ) 13.0 Hz, Pb-CH2), 129.2 (d, 1J (Pa,C) ) 49.6 Hz, Ci,a),
131.1 (d, 1J (Pb,C) ) 33.6 Hz, Ci,a), 133.1 (d, 2J (Pa,C) ) 11.4
2
3
Hz, Co,a), 133.6 (d, J (Pb,C) ) 11.4 Hz, Co,a), 128.6 (d, J (Pa,C)
) 10.7 Hz, Cm,a), 128.7 (d, 3J (Pb,C) ) 9.2 Hz, Cm,a), 130.8 (Cp,a),
131.1 (Cp,a).
(d cp e)P d (P h )I (3a ). A degassed solution of (tmeda)Pd(Ph)I
(324 mg, 0.759 mmol) in chloroform (10 mL) was cooled to 0
°C, and a cold solution (0 °C) of 1,2-bis(dicyclohexylphosphino)-
ethane (dcpe; 324 mg, 0.776 mmol) in degassed chloroform (10
mL) was added under nitrogen. The resulting clear solution
was stirred for 10 min at 0 °C, evaporated to dryness, and dried
for 6 h in vacuo to give a white solid. Column chromatography
(silica gel, chloroform) yielded 289 mg (52%) of 3a .
1H NMR: Pd-Ph, δ 7.39 (2H, t, 1J (Ho,Hm) ) 7.3 Hz, Ho),
7.01 (2H, m, Hm), 6.83 (1H, t, J (Hp,Hm) ) 7.3 Hz, Hp); dppe,
δ 1.88 (m, Pa-CH2), 1.67 (m, Pb-CH2), 1.9 (m, Hi,a), 2.2 (m,
Exp er im en ta l Section
The compound (tmeda)Pd(Ph)I was prepared by a literature
procedure.5c,d DMF was dried over molecular sieves (4 Å). All
other chemicals were used as received. 1,2-Bis(dicyclohexy-
lphosphino)ethane (dcpe) and cis-1,2-bis(diphenylphosphino)-
ethylene (dppet) were purchased from Strem and Aldrich,
respectively. The compounds 1a -4a were handled in air but
stored under nitrogen.
GC measurements were recorded on a Varian 3700 or on a
Varian 3400 GC. GC-MS measurements were recorded on a
Finnigan SSQ 7000 GC-MS system, including a Varian 3400
GC.
1
H
i,a), 1.75 (m), 1.65 (m), 1.45 (m), 1.1-1.4 (m), 0.7-0.8 (m).
31P{1H} NMR: δ 67.0 (d, J (Pa,Pb) ) 18.8 Hz, Pa), 61.7 (d, Pb).
13C{1H} NMR: Pd-Ph, δ 155.6 (d, 2J (Pb,C) ) 129.5 Hz, Ci),
138.3 (Co), 126.7 (d, 4J (Pb,C) ) 8.5 Hz, Cm), 122.5 (Cp); dppe, δ
25.3 (dd, 1J (Pa,C) ) 23.1 Hz, 3J (Pb,C) ) 23.1 Hz, Pa-CH2), 20.8
1
3
(dd, J (Pb,C) ) 18.8 Hz, J (Pa,C) ) 12.0 Hz, Pb-CH2), 34.2 (d,
1J (Pa,C) ) 25.4 Hz, Ci,a), 34.7 (d, 1J (Pb,C) ) 17.8 Hz, Ci,a), 29.5
(d, 2J ) 3.9 Hz, Co), 28.7 (Co), 28.2 (Co), 27.7 (Co), 27.1 (d, 3J )
13.1 Hz, Cm), 27.0 (d, J ) 12.3 Hz, Cm), 26.8 (d, J ) 9.2 Hz,
Cm), 26.6 (d, J ) 10.0 Hz, Cm), 25.9 (Cp), 25.7 (Cp).
(d p p et)P d (P h )I (4a ). To a solution of (tmeda)Pd(Ph)I (500
mg, 0.977 mmol) in chloroform (5 mL) was added cis-1,2-bis-
(diphenylphosphino)ethylene (dppet; 400 mg, 1.01 mmol), and
the solution was stirred for 5 min at room temperature. After
fast evaporation of the solvent, the resulting solid was dried
overnight in vacuo. Due to the formation of unspecified dppet
oligomers, an additional portion of dppet (110 mg) was added
to a chloroform solution of the product mixture. The solvent
was evaporated, and the resulting solid was dried in vacuo
for 4 h and purified by column chromatography (silica gel,
methylene chloride) to give 94 mg (14%) of pure 4a and 69 mg
of a mixture of (tmeda)Pd(Ph)I and 4a .
1H, 13C{1H}, and 31P{1H} NMR spectra were recorded at 400,
100, and 162 MHz on a Bruker AMX400 spectrometer or at
500, 126, and 202 MHz on a Bruker DMX500 spectrometer,
respectively. The NMR measurements were run in CDCl3 at
0 °C unless otherwise stated. The assignments of the peaks
in the 1D NMR spectra were clarified using 2D NMR spectra
(H-H COSY, H-C COSY, and H-P COSY) which were
recorded on a Bruker DMX500 spectrometer using gradient
3
3
3
1
pulse techniques. H and 13C chemical shifts are reported in
ppm downfield from SiMe4 and are referenced to the solvent
peaks (δ(CHCl3) 7.25 ppm, δ(CDHCl2) 5.31 ppm, δ(Me2CONH)
8.02 ppm, δ(CDCl3) 77.0 ppm, δ(CD2Cl2) 53.8 ppm, δ(Me2COND)
162.7 ppm). 31P chemical shifts are in ppm downfield from 85%
H3PO4 (external).
The phenylated products 7-10 were identified using 1H
NMR and GC-MS.
(d p p p )P d (P h )I (1a ) a n d (d p p e)P d (P h )I (2a ). To a solu-
tion of (tmeda)Pd(Ph)I (417 mg, 1.17 mmol) in chloroform (5
mL) was added 1 equiv of the chelating phosphine (dppp, 483
mg, dppe, 467 mg). After the mixture was stirred for 5-10
min at room temperature, the solvent was evaporated and the
resulting yellow solid was powdered and dried in vacuo (at
least 4 h). Recrystallization (CH2Cl2/n-pentane) afforded yellow
microcrystals (882 mg of 1a (96%), 723 mg of 2a (87%).
(d p p p )P d (P h )I (1a ). 1H NMR: Pd-Ph, δ 6.90 (2H, m, Ho),
1H NMR: Pd-Ph: δ 7.01 (2H, t, J (Ho,Hm) ) 8.0 Hz, Ho),
6.76 (2H, m, Hm), 6.70 (1H, m, Hp); dppet, δ 7.02 (1H, ddd,
1
3J (Pb,Ha) ) 54.1 Hz, J (Pa,Ha) ) 9.9 Hz, J (Ha,Hb) ) 9.9 Hz,
Pa-CH), 7.39 (u¨, Pb-CH), 7.25 (4H, m, Ho,a), 7.77 (4H, m, Ho,a),
7.30 (4H, m, Hm,a), 7.42 (8H, m, Hm,a + Hp,a + Hp,a). 31P{1H}
NMR: δ 56.8 (d, J (Pa,Pb) ) 14.9 Hz, Pa), 49.3 (d, Pb). 13C{1H}
NMR: Pd-Ph, δ 153.4 (d, 2J (Pb,C) ) 133.3 Hz, Ci), 138.0 (Co),
126.9 (d, 4J (Pb,C) ) 9.3 Hz, Cm), 122.7 (Cp); dppet, δ 146.7 (dd,
1J (Pa,C) ) 40.8 Hz, 3J (Pb,C) ) 40.8 Hz, Pa-CH), 146.0 (dd,
1J (Pb,C) ) 34.7 Hz, 3J (Pa,C) ) 33.1 Hz, Pb-CH), 128.2 (d,
1J (Pa,C) ) 53.2 Hz, Ci,a), 130.9 (d, 1J (Pb,C) ) 37.8 Hz, Ci,a),
132.9 (d, 2J (Pa,C) ) 11.6 Hz, Co,a), 133.5 (d, 2J (Pb,C) ) 12.3
2
2
1
6.55 (2H, m, Hm), 6.45 (1H, t, J (Hp,Hm) ) 7.5 Hz, Hp); dppp,
δ 2.54 (2H, m, Pa-CH2), 2.42 (2H, m, Pb-CH2), 1.89 (2H, m,
CH2), 7.81 (4H, m, Ho,a), 7.44 (6H, m, Hm,a + Hp,a), 7.29 (6H,
m, Ho,b + Hp,b), 7.14 (4H, m, Hm,b). 31P{1H} NMR: δ 13.1 (d,
J (Pa,Pb) ) 52.8 Hz, Pa), -8.0 (d, Pb). 13C{1H} NMR: Pd-Ph, δ
158.6 (dd, 2J (Pb,C) ) 126.6 Hz, 2J (Pa,C) ) 3.8 Hz, Ci), 137.0
(dd, 3J (Pb,C) ) 4.6 Hz, 3J (Pa,C) ) 2.3 Hz, Co), 127.0 (dd,
3
3
Hz, Co,a), 128.7 (d, J (Pa,C) ) 10.8 Hz, Cm,a), 128.8 (d, J (Pb,C)
) 10.8 Hz, Cm,a), 130.9 (d, 4J (Pa,C) ) 1.5 Hz, Cp,a), 131.2 (d,
4J (Pb,C) ) 2.3 Hz, Cp,a).

Regiocontrol in the Heck Type Reaction
Organometallics, Vol. 18, No. 6, 1999 975
[P 2P d (P h )]+OTf- (1b-4b). In a typical procedure AgOTf
(28.6 mg, 0.111 mmol) was added to a precooled (-20 °C)
solution of (P2)Pd(Ph)I (1a -4a ; 0.117 mmol) in 0.87 mL of
DMF-d7. After 2-4 min the precipitate was removed (cen-
trifugation followed by filtration through 0.45 µm PTFE),
resulting in a clear, colorless, or slightly reddish solution (135
mmol of Pd/mL of solvent), which was immediately used for
NMR experiments or stored at -196 °C.
complexes 1b-4b using a DMF/DMF-d7 mixture (9/1) and the
silver salts AgPF6, AgBF4, and AgOAc instead of triflate.
[(d p p p )P d (P h )]+P F 6 (1c). 31P{1H} NMR: δ 26.2 (d,
-
1
J (Pa,Pb) ) 50.4 Hz, Pa), -1.3 (d, Pb), -142.9 (sept, J (P-F) )
709.5 Hz, PF6).
[(d p p p )P d (P h )]+BF 4 (1d ). 31P{1H} NMR: δ 26.2 (d,
-
J (Pa,Pb) ) 50.4 Hz, Pa), -1.3 (d, Pb).
[(d p p p )P d (P h )]+OAc- (1e). 31P{1H} NMR: δ 19.4 (d,
J (Pa,Pb) ) 48.5 Hz, Pa), -3.7 (d, Pb).
[(d p p p )P d (P h )]+OTf- (1b; 1/9 DMF -d 7/CD2Cl2, -80 °C).
1H NMR: Pd-Ph, δ 6.90 (2H, m, Ho), 6.57 (3H, m, Hm + Hp);
dppp, δ 2.67 (2H, br, Pa-CH2), 2.52 (2H, br, Pb-CH2), 1.69
(2H, m, CH2), 7.58 (4H, m, Ho,a), 7.45 (6H, m, Hp,a + Hm,a),
7.27 (6H, m, Ho,a + Hp,a), 7.10 (4H, m, Hm,a). 31P{1H} NMR: δ
25.5 (d, J (Pa,Pb) ) 49.6 Hz, Pa), -3.3 (d, Pb). 13C{1H} NMR:
Pd-Ph, δ 162 (d, one satellite overlapped with the solvent
peak, Ci), 133.5 (Co), 127.4 (d, 4J (Pb,C) ) 6.9 Hz, Cm), 123.3
[(dppp)P d(CHP h CH2P h )]+OTf- (5b). A solution of [(dppp)-
Pd(Ph)]+OTf- (1b) (135 mmol of Pd/L) in 9/1 CD2Cl2/DMF-d7
was prepared in a way similar to that described above and
cooled to -80 °C. After addition of 1.2 equiv of styrene the
sample was inserted into the NMR (-20 °C) and the reaction
was followed via 31P NMR spectroscopy. After ca. 20 min 1b
was converted mainly to 5b, whose composition (80-90% 5b)
did not change during the 1H, 13C, DEPT-135, H-C COSY,
and H-H COSY experiments (10 h). The signals of the
performed NMR experiments (-20 °C) could be partially
assigned as follows.
1
3
(Cp); dppp, δ 27.2 (dd, J (Pa,C) ) 33.6 Hz, J (Pb,C) ) 7.6 Hz,
Pa-CH2), 23.8 (d, 1J (Pb,C) ) 22.1 Hz, Pb-CH2), 17.6 (CH2),
129.1 (d, 1J (Pa,C) ) 59.5 Hz, Ci,a), 129.0 (d, 1J (Pb,C) ) 35.9
2
2
Hz, Ci,a), 132.7 (d, J (Pa,C) ) 10.7 Hz, Co,a), 132.5 (d, J (Pb,C)
3
) 11.4 Hz, Co,a), 127.4 (d, J (Pa,C) ) 10.7 Hz, Cm,a), 128.7 (d,
1H NMR: Pd-R, δ 3.64 (1H, m, H1), 2.09 (1H, m, H2), 2.93
(1H, m, H′2); dppp, δ 2.80 (1H, m, Pa-CH2), 2.95 (1H, m, Pa-
CH′2), 2.57 (2H, m, Pb-CH2), 1.8 (2H, br, CH2). 31P{1H} NMR:
δ 18.9 (d, J (Pa,Pb) ) 74.6 Hz, Pa), 6.2 (d, Pb). 13C{1H} NMR:
3J (Pb,C) ) 9.2 Hz, Cm,a), 130.4 (Cp,a), 130.4 (Cp,a), CF3: 120.2
1
(q, J (C,F) ) 320.4 Hz).
[(d p p e)P d (P h )]+OTf- (2b; DMF -d 7, -60 °C). 1H NMR:
Pd-Ph, δ 7.12 (2H, br, Ho), 6.77 (3H, br, Hm + Hp); dppe, δ
3.05 (2H, d, 2J (Pa,H) ) 29.3 Hz, Pa-CH2), 2.53 (2H, d, 2J (Pb,H)
) 26.4 Hz, Pb-CH2), 7.70 (4H, br, Ho,a), 7.96 (4H, br, Ho,a),
7.47 (4H, br, Hm,a), 7.62 (8H, br, Hm,a + Hp,a + Hp,a). 31P{1H}
NMR: δ 57.2 (d, J (Pa,Pb) ) 24.8 Hz, Pa), 38.3 (d, Pb). 13C{1H}
NMR: Pd-Ph, δ 162 (d, one satellite overlapped with the
2
2
Pd-R, δ 72.3 (dd, J (Pb,C) ) 43.2 Hz, J (Pa,C) ) 10.6 Hz, C1),
34.3 (d, 3J (Pb,C) ) 4.5 Hz, C2), dppp, δ 26.7 (dd, 1J (Pa,C) )
1
1
27.5 Hz, J (Pb,C) ) 3.8 Hz, Pa-CH2), 25.3 (d, J (Pb,C) ) 23.7
Hz, Pb-CH2), 18.3 (CH2); CF3, δ 120.2 (q, 1J (C,F) ) 320.4 Hz).
[(d p p p )P d (CHMeP h )]+OTf- (6b). To a solution of [(dppp)-
Pd(Ph)]+OTf- (1b) (135 mmol of Pd/L) in DMF-d7 was added
10 equiv of styrene at -20 °C. After 12 h at -20 °C the
conversion of 1b via the intermediate 5b into 6b was com-
pleted by warming to room temperature (5 min) to give a
mixture containing 6b and the Heck products trans-stilbene
and 1,1-diphenylethylene. The NMR signals at -20 °C could
partially be assigned as follows.
4
solvent peak, Ci), 135.9 (Co), 128.1 (d, J (Pb,C) ) 7.7 Hz, Cm),
124.6 (Cp), dppe: δ 30.6 (dd, 1J (Pa,C) ) 33.9 Hz, 3J (Pb,C) )
20.2 Hz, Pa-CH2), 20.9 (d, 1J (Pb,C) ) 29.3 Hz, Pb-CH2), 129.9
(d, 1J (Pa,C) ) 57.8 Hz, Ci,a), 131.4 (d, 1J (Pb,C) ) 33.9 Hz, Ci,a),
134.2 (d, 2J (Pa,C) ) 11.6 Hz, Co,a), 133.6 (d, 2J (Pb,C) ) 13.1
3
3
Hz, Co,a), 129.5 (d, J (Pa,C) ) 10.8 Hz, Cm,a), 130.0 (d, J (Pb,C)
) 10.0 Hz, Cm,a), 132.4 (Cp,a), 132.0 (Cp,a), CF3: 121.6 (q, 1J (C,F)
) 322.1 Hz).
1H NMR: Pd-R, δ 3.28 (1H, m, H1), 0.86 (3H, m, H2); dppp,
δ 2.74 (2H, m, Pa-CH2), 2.55 (2H, m, Pb-CH2), 1.40 (1H, m,
CH2), 1.88 (1H, m, CH′2). 31P{1H} NMR: δ 18.3 ppm (d, J (Pa,Pb)
) 76.0 Hz, Pa), 6.3 (d, Pb). 13C{1H} NMR: Pd-R, δ 67.4 (dd,
2J (Pb,C) ) 42.7 Hz, 2J (Pa,C) ) 11.9 Hz, C1), 13.3 (d, 3J (Pb,C) )
5.4 Hz, C2); dppp, δ 26.6 (d, 1J (Pa,C) ) 28.5 Hz, Pa-CH2), 24.7
[(d cp e)P d (P h )]+OTf- (3b; DMF -d 7, -60 °C). 1H NMR:
Pd-Ph, δ 7.38 (2H, m, Ho), 7.13 (2H, m, Hm), 7.02 (1H, t,
1J (Hp,Hm) ) 7.3 Hz, Hp); dcpe, δ 2.3 (m, P-CH2), 2.04 (m),
1.85-2.15 (m), 1.75 (br), 1.55 (br), 1.05-1.50, 0.95 (br). 31P-
{1H} NMR: δ 75.1 (d, J (Pa,Pb) ) 15.8 Hz, Pa), 66.8 (d, Pb), 13C-
{1H} NMR: Pd-Ph, δ 160.7 (d, 2J (Pb,C) ) 116.4 Hz, Ci), 136.1
(Co), 127.5 (Cm), 124.3 (Cp); dcpe, δ 16.8 (d, 1J (Pb,C) ) 22.4
1
(d, J (Pb,C) ) 24.7 Hz, Pb-CH2), 17.9 (CH2); CF3, δ 120.2 (q,
1J (C,F) ) 320.4 Hz).
Gen er a l P r oced u r e for th e Rea ction w ith Styr en e a n d
P r op en e. To a solution of [P2Pd(Ph)]+OTf- (0.5 mL of 1a -
1d , 2b, 3b, and 4b) containing 45 mmol of Pd/L, prepared as
described above, was added styrene (2-10 equiv) or propylene
(excess) were added at 0 °C, and the solutions were stirred at
room temperature overnight. The resulting dark mixture was
partitioned between diethyl ether (2 × 2 mL) and water (2 ×
2 mL). The ethereal layers were dried over sodium sulfate and
concentrated to 1 mL by rotary evaporation before analyses
by GC.
1
Hz, Pb-CH2), 32.6 (d, J (Pb,C) ) 17.0 Hz, Ci,a), 28.1 (Co), 27.7
(Co), 24.4 (Co), 26.1-26.5 (Cm), 25.5 (Cp), 25.3 (Cp), CF3: 121.0
1
(q, J (C,F) ) 322.1 Hz).
[(d p p et)P d (P h )]+OTf- (4b; DMF -d 7, -60 °C). 1H NMR:
Pd-Ph, δ 7.09 (2H, br, Ho), 6.84 (2H, br, Hm), 6.87 (1H, br,
Hp), dppet, δ 7.0 (1H, d, one satellite is overlapped, Pa-CH),
3
8.33 (1H, d, J (Pa,Hb) ) 54.5 Hz, Pb-CH), 7.55 (8H, br, Ho,a
+
H
m,a), 7.90 (4H, br, Ho,a), 7.65 (8H, br, Hm,a + Hp,a + Hp,a). 31P-
{1H} NMR: δ 58.8 (br, Pa), 52.8 (br, Pb). 13C{1H} NMR: Pd-
Ph, δ 160.8 (d, 2J (Pb,C) ) 120.2 Hz, Ci), 136.3 (Co), 128.2 (Cm),
1
3
125.2 (Cp), dppet, δ 149.9 (dd, J (Pa,C) ) 43.2 Hz, J (Pb,C) )
40.8 Hz, Pa-CH2), 143.4 (dd, 1J (Pb,C) ) 38.5 Hz, 3J (Pa,C) )
16.6 Hz, Pb-CH2), 128.0 (d, 1J (Pa,C) ) 53.1 Hz, Ci,a), 131.0 (d,
1J (Pb,C) ) 37.8 Hz, Ci,a), 134.1 (d, 2J (Pa,C) ) 10.8 Hz, Co,a),
133.5 (d, 2J (Pb,C) ) 12.2 Hz, Co,a), 129.8 (d, 3J (Pa,C) ) 10.8
Ack n ow led gm en t. This work was supported by the
Deutsche Forschungsgemeinschaft (postdoctoral fellow-
ship for M.L.), by the Swedish Research Council for
Engineering Sciences, and by the Swedish Natural
Science Research Council. We are also grateful to the
Parallelldatorcentrum (PDC) at KTH for providing the
computer facilities.
3
Hz, Cm,a), 130.2 (d, J (Pb,C) ) 7.8 Hz, Cm,a), 132.8 (Cp,a), 132.3
1
(Cp,a); CF3, δ 121.7 (q, J (C,F) ) 322.1 Hz).
[(d p p p )P d (P h )]+X- (X ) P F 6 (1c), BF 4 (1d ), OAc (1e)).
These complexes were prepared similarly to the triflate
OM9803265