Synthesis and investigation of new macrocyclic diphosphine - Palladium(0) complexes based on the barbiturate binding receptor

Article abstract of DOI:10.1021/om020364v

A series of diphosphine ligands possessing a barbiturate-binding receptor have been synthesized with the goal of preparing new palladium(0) complexes for the Heck reaction, which position the alkene with respect to the metal center and thereby control the regioselectivity of the insertion step. Some of the diphosphines prepared were found to efficiently form macrocyclic bisphosphine palladium(0) complexes even though a 26-membered cycle is produced. A significant solvent effect for the oxidative addition of the Pd⁰ complexes with phenyl iodide was noted in the case of one of the diphosphine ligands. This descrepancy may well be accounted for by the ability of the ligand when complexed to Pd⁰ to possess different conformational preferences in the solvents tested, which influences the bite angle to the metal center. The receptors possessing an isophthaloyl connector bind barbital with affinities corresponding to those of the previously reported open receptors. However, upon complexation with Pd(dba)₂, none of the bidentate ligands revealed a capacity to bind barbital, reflecting again the conformational changes that occur upon coordination to Pd⁰. The new palladium(0) complexes were tested for their ability to promote the Heck reaction between aryl halides and n-butyl acrylate. Whereas all the ligand:Pd⁰ complexes were found to catalyze this reaction with reactivities similar to triphenylphosphine, in one case a higher reactivity was noted.

Full text of DOI:10.1021/om020364v

Organometallics 2002, 21, 5243-5253
5243
Syn th esis a n d In vestiga tion of New Ma cr ocyclic
Dip h osp h in e-P a lla d iu m (0) Com p lexes Ba sed on th e
Ba r bitu r a te Bin d in g Recep tor
Hanne S. Sørensen, J ens Larsen, Brian S. Rasmussen, Bolette Laursen,
Signe G. Hansen, and Troels Skrydstrup*
Department of Chemistry, University of Aarhus, Langelandsgade 140,
8000 Aarhus C, Denmark
Christian Amatore* and Anny J utand*
Department de Chimie, Ecole Normale Superieure, CNRS UMR 8640, 24 Rue Lhomond,
75231 Paris Cedex 5, France
Received May 6, 2002
A series of diphosphine ligands possessing a barbiturate-binding receptor have been
synthesized with the goal of preparing new palladium(0) complexes for the Heck reaction,
which position the alkene with respect to the metal center and thereby control the
regioselectivity of the insertion step. Some of the diphosphines prepared were found to
efficiently form macrocyclic bisphosphine palladium(0) complexes even though a 26-membered
cycle is produced. A significant solvent effect for the oxidative addition of the Pd0 complexes
with phenyl iodide was noted in the case of one of the diphosphine ligands. This descrepancy
may well be accounted for by the ability of the ligand when complexed to Pd0 to possess
different conformational preferences in the solvents tested, which influences the bite angle
to the metal center. The receptors possessing an isophthaloyl connector bind barbital with
affinities corresponding to those of the previously reported open receptors. However, upon
complexation with Pd(dba)2, none of the bidentate ligands revealed a capacity to bind barbital,
reflecting again the conformational changes that occur upon coordination to Pd0. The new
palladium(0) complexes were tested for their ability to promote the Heck reaction between
aryl halides and n-butyl acrylate. Whereas all the ligand:Pd0 complexes were found to catalyze
this reaction with reactivities similar to triphenylphosphine, in one case a higher reactivity
was noted.
In tr od u ction
play including the ligand type, solvent polarity, and
substitution pattern of the alkene.1,3,4 Fine-tuning of the
catalyst and reaction conditions is therefore required
The palladium(0)-catalyzed coupling of aryl iodides,
bromides, or triflates with an olefin substrate, known
as the Heck reaction, represents an important synthetic
transformation for the preparation of arylalkenes on a
laboratory and an industrial scale.1 Traditionally, Pd0-
complexes with tertiary phosphine ligands have been
studied and applied to the Heck coupling; however, the
past few years have witnessed intensive efforts devoted
to the design of new and more electron-donating ligands
necessary for promoting the initial oxidative addition
step in the catalytic cyclic with the cheaper but less
reactive aryl chlorides.2
(2) For some recent examples, see: (a) Herrmann, W. A.; Reisinger,
C.-P.; Spiegler, M. J . Organomet. Chem. 1998, 557, 93. (b) Littke, A.
F.; Fu, G. C. J . Org. Chem. 1999, 64, 10. (c) Shaughnessy, K. H.; Kim,
P.; Hartwig, J . F. J . Am. Chem. Soc. 1999, 121, 2123. (d) Herrmann,
W. A.; Bohm, V. P. W. J . Organomet. Chem. 1999, 572, 141. (e) Bohm,
V. P. W.; Herrmann, W. A. Chem. Eur. J . 2000, 6, 1017. (f) McGuin-
ness, D. S.; Cavell, K. J . Organometallics 2000, 19, 741. (g) Iyer, S.;
Ramesh, C. Tetrahedron Lett. 2000, 41, 8981. (h) Peris, E.; Loch, J .
A.; Mata, J .; Crabtree, R. H. Chem. Commun. 2000, 201. (i) Calo, V.;
Nacci, A.; Lopez, L.; Mannarini, N. Tetrahedron Lett. 2000, 41, 8973.
(j) Gruber, A. S.; Zim, D.; Ebeling, G.; Monteiro, A. L.; Dupont, J . Org.
Lett. 2000, 2, 1287. (k) Morales-Morales, D.; Redon, R.; Yung, C.;
J ensen, C. M. Chem. Commun. 2000, 1619. (l) Ehrentraut, A.; Zapf,
A.; Beller, M. Synlett 2000, 11, 1589. (m) Feuerstein, M.; Doucet, H.;
Santelli, M. J . Org. Chem. 2001, 66, 5923. (n) Littke, A. F.; Fu, G. C.
J . Am. Chem. Soc. 2001, 123, 6989, and references therein.
(3) For a few recent examples of regiocontrolled arylation of alkenes
bearing an internal coordinating group, see: (a) Itami, K.; Nokami,
T.; Ishimura, Y.; Mitsudo, K.; Kamei, T.; Yoshida, J .-i. J . Am. Chem.
Soc. 2001, 123, 11577. (b) Olofsson, K.; Sahlin, H.; Larhed, M.;
Hallberg, A. J . Org. Chem. 2001, 66, 544.
(4) For papers on the topic of insertion and regioselectivity, see: (a)
Ozawa, F.; Kubo, A.; Hayashi, T. J . Am. Chem. Soc. 1991, 113, 1417.
(b) Cabri, W.; Candiani, I. Acc. Chem. Soc. 1995, 28, 2. (c) Ludwig, M.;
Stromberg, S.; Svensson, M.; Akermark, B. Organometallics 1999, 18,
970. (d) For a recent review on the topic of insertion and regioselec-
tivity, see: Larhed, M.; Hallberg, A. In Handbook of Organopalladium
Chemistry for Organic Synthesis; Negishi, E., Ed.; J ohn Wiley & Sons:
New York, in press.
Less attention has been concentrated on providing a
general method for controlling the regiochemical out-
come of these reactions, where many factors come into
(1) For recent reviews, see: (a) Beletskaya, I.; Cheprakov, A. V.
Chem. Rev. 2000, 100, 3009. (b) Crisp, G. T. Chem. Soc. Rev. 1998, 27,
427. (c) Reetz, M. T. In Transition Metal Catalyzed Reactions; Davis,
S. G., Murahashi, S.-I., Eds.; Blackwell Science: Oxford, 1999. (d) Link,
J . T.; Overman, L. E. In Metal-Catalyzed Cross-Coupling Reactions;
Diederich F., Stang, P. J ., Eds.; Wiley-VCH: New York, 1998. (e) Brase,
S.; DeMeijere, A. In Metal-Catalyzed Cross-Coupling Reactions;
Diederich F., Stang, P. J ., Eds.; Wiley-VCH: New York, 1998. (f)
Gibson, S. E.; Middleton, R. J . Contemp. Org. Synth. 1996, 3, 447.
10.1021/om020364v CCC: $22.00 © 2002 American Chemical Society
Publication on Web 10/31/2002

5244 Organometallics, Vol. 21, No. 24, 2002
Sørensen et al.
for many cases in order to obtain higher regioselectivi-
ties. Whereas the Heck reaction works well for the
arylation at the terminal position of electron-deficient
olefins with monodentate ligands by the neutral path-
way, there are no procedures for the preparation of the
corresponding branched olefins. In contrast, protocols
for the preparation of branched electron-rich alkenes are
available employing bidentate ligands via the cationic
pathway. For the obtention of linear products with
electron-rich olefins, the most successful approach
requires the incorporation of a chelating auxiliary. More
problematic are the unfunctionalized 1-alkenes, where
efficient methodology for their regioselective arylation
is lacking.
We have recently commenced a program directed to
the design of new Pd0-ligand complexes that possess a
receptor domain capable of binding and subsequently
organizing the alkene substrate with respect to the
metal center. Through the optimization of both the size
of the receptor and positioning of the palladium(0)
center it may therefore be possible to control the
regioselectivity of the C-C bond forming step and
thereby override any intrinsic directive properties of the
olefin substrate. In this way, the regiochemistry may
either be controlled for difficult alkenes, which are not
regiospecific in the arylation step, or even potentially
inverted compared to the normal products observed in
the absence of such a receptor ligand.
As a suitable receptor ligand, our attention was first
drawn to the barbiturate-recognizing hosts originally
reported by Hamilton in 1988,5 due to the well-
documented exploitation of this host:guest system for
numerous applications.6 The host is built up of two
appropriately separated 2,6-diaminopyridine units form-
ing a cavity for the barbiturates, which ultimately binds
by a six hydrogen bonding network, leading in general
to strong complex formation in nonparticipating sol-
vents.7 With this host as the basis structure, we have
therefore been interested in preparation of receptors
incorporating two terminal triarylphosphine units ca-
pable of complexing palladium(0) and resulting in the
formation of a metal-containing macrocycle, as sche-
matically illustrated in Figure 1. Eventually, complex-
ation of a suitable metal-host with an olefin covalently
linked to a barbiturate unit via a temporary connector
F igu r e 1. Macrocyclic bisphosphine palladium(0) complex
possessing a barbiturate-recognizing moiety.
would subsequently position the alkene with respect to
the metal center in much the same way as a chelating
auxiliary and thereby potentially control the regio-
selectivity of the insertion step.8
In this paper, we report our preliminary studies in
this direction, revealing both the synthesis of several
of such receptors, whereby phenyl groups of varying
substitution patterns represent the linker, and their
propensity to form macrocyclic complexes upon addition
of Pd0(dba)2 (dba ) trans,trans-dibenzylideneacetone),
measured by 31P NMR spectroscopy and cyclic voltam-
metry.9 In addition, we disclose our investigation on the
ability of these complexes to bind barbital and on their
comparative reactivity in the oxidative addition with
phenyl iodide, representing the first step in the catalytic
cycle of the Heck reaction. Finally, the aptitude of these
new complexes to promote the Heck reaction was
explored, being an important requisite for the success
of this project.
Resu lts a n d Discu ssion
I. Syn th esis of th e Dip h osp h in e Recep tor s. The
bidentate phosphine ligands 1-3 first studied are
illustrated in Figure 2. All contain the central barbitu-
rate binding motif though modified at the terminal
position with two aryl diphenylphosphine units of
varying substitution pattern. The preparation of these
compounds was easily performed by an initial amide-
forming step between the three iodobenzoyl chlorides
and the known diamine 76d in the presence of triethyl-
amine (Scheme 1). A subsequent iodine to phosphorus
exchange with diphenylphosphine promoted by pal-
ladium acetate in refluxing acetonitrile afforded the
crystalline diphosphines 1 (mp 184-186 °C) and 2 (mp
174-176 °C) in good yields.10 The rather unexceptional
(5) Chang, S.-K.; Hamilton, A. D. J . Am. Chem. Soc. 1988, 110, 1318.
(6) (a) Tecilla, P.; Dixon, R. P.; Slobodkin, G.; Alavi, D. S.; Waldeck,
D. H.; Hamilton, A. D. J . Am. Chem. Soc. 1990, 112, 9408. (b) Tecilla,
P.; Chang, S.-K.; Hamilton, A. D. J . Am. Chem. Soc. 1990, 112, 9586.
(c) Tecilla, P.; Hamilton, A. D. J . Chem. Soc., Chem. Commun. 1990,
1232. (d) Chang, S.-K.; Fan, E.; Van Engen, D.; Hamilton, A. D. J .
Am. Chem. Soc. 1991, 113, 7640. (e) Aoki, I.; Harada, T.; Sakaki, T.;
Kawahara, Y.; Shinkai, S. J . Chem. Soc., Chem. Commun. 1992, 1341.
(f) Aoki, I.; Kawahara, Y.; Sakaki, T.; Harada, T.; Shinkai, S. Bull.
Chem. Soc. J pn. 1993, 66, 927. (g) Motesharei, K.; Myles, D. C. J . Am.
Chem. Soc. 1994, 116, 7413. (h) Tecilla, P.; J ubian, V.; Hamilton, A.
D. Tetrahedron 1995, 51, 435. (i) Chin, T.; Gao, Z.; Lelouche, I.; Shin,
Y. K.; Purandare, A.; Knapp, S.; Isied, S. S. J . Am. Chem. Soc. 1997,
119, 12849. (j) Salameh, A. S.; Ghaddar, T.; Isied, S. S. J . Phys. Org.
Chem. 1999, 12, 247. (k) Nabeshima, T.; Takahashi, T.; Hanami, T.;
Kikuchi, A.; Kawabe, T.; Yano, Y. J . Org. Chem. 1998, 63, 3802. (l)
Inoue, K.; Ono, Y.; Kanekiyo, Y.; Ishi-i, T.; Yoshihara, K.; Shinkai, S.
J . Org. Chem. 1999, 64, 2933. (m) Ghaddar, T. H.; Castner, E. W.;
Isied, S. S. J . Am. Chem. Soc. 2000, 122, 1233. (n) Rasmussen, B. S.;
Elezcano, U.; Skrydstrup, T. J . Chem. Soc., Perkin Trans. 1 2002, 1723.
(7) For studies using synthetic receptors capable of binding ligands
with high association constants via multiple hydrogen bonding, see:
(a) Lehn, J .-M. Supramolecular Chemistry-Concepts and Perspectives;
VCH: Weinheim, 1995. (b) Schneider, H.-J .; Yatsimirsky, A. Principles
and Methods in Supramolecular Chemistry; J ohn Wiley and Sons: New
York, 2000.
(8) For
a few examples on the use of synthetic receptors for
synthesis: (a) Rowe, H. L.; Spencer, N.; Philip, D. Tetrahedron Lett.
2000, 41, 4475. (b) Bassani, D. M.; Darcos, V.; Mahony, S.; Desvergne,
J .-P. J . Am. Chem. Soc. 2000, 122, 8795. (c) Bach, T.; Bergmann, H.
J . Am. Chem. Soc. 2000, 122, 11525.
(9) (a) Amatore, C.; Broeker, G.; J utand, A.; Khalil, F. J . Am. Chem.
Soc. 1997, 119, 5176. (b) Amatore, C.; J utand, A.; Meyer, G. Inorg.
Chim. Acta 1998, 273, 76. (c) For a review, see: Amatore, C.; J utand,
A. Coord. Chem. Rev. 1998, 178-180, 511.

Diphosphine-Palladium(0) Complexes
Organometallics, Vol. 21, No. 24, 2002 5245
Sch em e 2
ligand 4, the pyridine rings have been replaced by
benzene, and for ligands 5 and 6, a ferrocene entity
bridges the two 2,6-diamidopyridine units.
F igu r e 2. Receptor-containing diphosphines used for this
study.
Sch em e 1
Adaptation of the same synthetic sequence to 4 as for
ligands 1-3 proved less rewarding, as the monoacyla-
tion of 1,3-diaminobenzene with isophthaloyl chloride
as reported for the effective preparation of 7 afforded
only a 9% yield of the corresponding diamine. However,
by reversing the synthetic sequence as indicated in
Scheme 2, access to 4 was more satisfying. Hence,
monoacylation of 1,3-diaminobenzene with m-iodoben-
zoyl chloride produced the amide 11 in 88% yield, which
was coupled twice to isophthaloyl chloride. In constrast
to the diphosphine 2, ligand 4 proved exceptionally
insoluble in both chloroform and THF, which compli-
cated its purification by column chromatography. How-
ever, the crude product obtained after introduction of
diphenylphosphine proved to be reasonably pure, as
1
assessed by both H and 31P NMR spectroscopy.
Synthesis of the diphosphines 5 and 6 started with
the double carboxylation of ferrocene via the dianion
intermediate (Scheme 3). Conversion to its diacid chlo-
ride and amide formation with diaminopyridine afforded
14 in 66% yield. Subsequent acylation and introduction
of diphenylphosphine afforded 5 and 6.
II. In vestiga tion of th e P a lla d iu m (0) Com p lexes
F or m ed in Situ Betw een P d (d ba )2 a n d th e Bid en -
ta te P h osp h in e Liga n d s by 31P NMR Sp ectr oscop y
a n d /or Cyclic Volt a m m et r y. (a ) P d (d b a )2 a n d
Liga n d 2. The ability of the synthesized diphosphines
to form macrocyclic complexes with Pd0 was investigated
by 31P NMR with solutions of Pd(dba)2 (14 mM) and 1
equiv of the ligand (Table 1).9 In the event of bidentate
complexation, the 31P NMR spectrum will feature two
doublets, which is indicative of two nonequivalent
yield obtained for the para-substituted ligand 3 could
be the result of the particularly low solubility of the
diiodide 10 in the reaction conditions employed com-
pared to its counterparts 8 and 9.
In the subsequent investigation on the applicability
of these ligands for promoting the Heck reaction (see
section V), a series of similar ligands with small struc-
turial variations were made, to study their effect on the
reactivity of the Pd0-complexes. For example, with
(10) (a) Herd, O.; Hessler, A.; Hingst, M.; Tepper, M.; Stelzer, O. J .
Organomet. Chem. 1996, 522, 69. (b) Bergbreiter, D. A.; Liu, Y.-S.;
Furyk, S.; Case, B. L. Tetrahedron Lett. 1998, 39, 8799.

5246 Organometallics, Vol. 21, No. 24, 2002
Sørensen et al.
Ta ble 1. 31P NMR Ch em ica l Sh iftsa of P d 0-Com p lexes F or m ed fr om th e Mixtu r e of P d (d ba )2 + P ,P a n d of
th e Com p lexes P h P d I(P ,P ) Gen er a ted fr om th e Oxid a tive Ad d ition w ith P h I
P,P
δ0
Pd(dba)2 + 1P,P f Pd(dba)(P,P)
PhPdI(P,P)
δ4
P,P
solvent
CDCl3
δ1
δ2
δ3
2
-4.78
23.76 (J PP ) 12 Hz)
(J PP ) 10 Hz)
24.34 (J PP ) 12 Hz)
(J PP ) 10 Hz)
24.71 (J PP ) 12 Hz)
(J PP ) 10 Hz)
24.73 (J PP ) 12 Hz)
(J PP ) 10 Hz)
24.15
2
2
THF
-4.20
24.37
DMF
-4.35
25.40
26.08
27.82b
24.03
1
1
CDCl3
THF
-8.97
-8.40
22.94 (J PP ) 459 Hz)
44.71 (J PP ) 459 Hz)
n.m.c
n.m.c
-8.69
-9.29
26.70b
29.84b
1
DMF
-9.31
27.50b
31.16b
21.09
a
b
Determined vs H3PO4 at 20 °C. Broad singlet. c n.m. ) not measured.
Sch em e 3
F igu r e 3. Partial 31P NMR spectrum (162 MHz) in
CDCl3: (a) Pd(dba)2 (14 mM) + diphosphine 2 after 10 min,
(b) Pd(dba)2 (14 mM) + diphosphine 1 after 24 h, (c) Pd-
(dba)2 (14 mM) + diphosphine 3 after 24 h.
spray mass spectroscopy, where a peak corresponding
to Pd0(2) + Na+ with loss of the dba ligand was
observed.
Slightly more complicated was the 31P NMR spectrum
from an identical experiment in DMF where in addition
to two broad singlets at δ1 ) 25.40 and δ2 ) 26.08 ppm
of equal intensity representing the cis-macrocyclic
Pd0(dba)(2) species, a third signal was observed by a
singlet at δ3 ) 27.82 ppm, implying the presence of a
second complex displaying two equivalent phosphorus.
Addition of excess dba (10 equiv) did not perturb the
spectrum, thereby excluding the possibility that this
peak arises from the formation of Pd0(2), whereby dba
has dissociated from the metal center. Exposure of the
solution to oxygen or hydrogen peroxide did not produce
signals that were comparable to the chemical shift of
δ3.
Addition of 100 equiv of PhI to the three solutions
led to the disappearance of the above signals and the
formation of a single signal at δ4 ∼24 ppm. Its appear-
ance as a singlet suggests that the formed complex
PhPdI(2) possesses a trans-configuration, as generally
observed for monodentate ligands, but which is in
contrast to the cis-preference previously reported for
other bidentate ligands such as dppm, dppe, dppp, dppb,
dppf, DIOP, and BINAP. The large ring size and thereby
greater flexibility in the macrocyclic complex, PhPdI-
phosphorus compatible with the structure Pd0(dba)(P,P)
in which the dba ligand remains ligated to the Pd0
center by only one CdC bond, as classically observed
with other bidentate P,P ligands.9a,c The 31P NMR
spectrum of a mixture of a slight excess of ligand 2 with
Pd(dba)2 in CDCl3 or THF exhibited after 10 min two
double doublets of equal magnitude with coupling
constants of approximately 12 and 10 Hz, the magnitude
of which is indicative of a cis-geometry, which appeared
stable even after 11 days (Figure 3a).11 Macrocyclization
by complexation is therefore both fast and efficient
considering the formation of a 26-membered cycle.
Further evidence for this ring size and not that of a
dimeric or trimeric structure was gained from electro-
(11) It is difficult to explain the appearance of two double doublets,
considering the nature of the Pd0-complexes. We suggest the possible
formation of two closely related dba-containing cis-macrocycles, the
structures of which are elaborated upon in section IV.

Diphosphine-Palladium(0) Complexes
Organometallics, Vol. 21, No. 24, 2002 5247
Sch em e 4
Sch em e 5
(2), compared to similar complexes with other bidentate
ligands, could be the main reason for the generation of
the thermodynamically more favored trans-relationship
between the two phosphorus. In DMF, a small byprod-
uct (approximately 5%) was also detected by the two
singlets observed at δ ) -4.99 and 24.77 ppm, possibly
attributed to the complex PhPdI(2)2, in which only one
of the phosphorus atoms of the ligand is bound to
palladium. 31P NMR spectroscopy performed on a 1:1
mixture of PhPdI(PPh3)2 and 2 indicated the same major
product and byproduct as well as free PPh3 (δ ) -5.19
ppm).
Pd0(2) excluded the monitoring of the kinetics of the
oxidative addition by amperometry.9a UV spectroscopy
was the best adapted and selected technique (vide infra)
for this purpose.9b
In THF, the mechanism of the formation of Pd0(dba)-
(2) looked very similar (oxidation potentials of O1 and
O2 at +0.34 and +0.83 V). However, the oxidation peak
of Pd0(2) was not detected. This suggests that the
equilibrium between Pd0(dba)(2), Pd0(2), and dba (last
reaction of Scheme 4) lies less in favor of Pd0(2) in THF
than in DMF, or is less labile in THF than in DMF with
some kinetic consequences for the oxidative addition
(vide infra).
The kinetic evolution of the mixture Pd(dba)2 + 2 in
THF and DMF was monitored by cyclic voltammetry.
When ligand 2 was added to a solution of Pd(dba)2 (2
mM) in DMF (containing n-Bu4NBF4 0.3 M), two oxida-
tion peaks of almost equal peak current were detected
at +0.27 (O1) and +0.78 V (O2). After 10 min, the
oxidation peak O1 shifted to a slightly more positive
potential of +0.30 V (O′1), and the oxidation peak
current of O2 was approximately 4 times that of O′1. This
ratio was then constant with time. The oxidation peak
O′1 was plateau-shaped, suggesting the presence of two
Pd0 complexes involved in a fast equilibrium (CE
mechanism).9c When 10 equiv of dba was added to the
solution, only the oxidation peak O2 was detected,
indicating that the two Pd0 complexes detected at O2
and O′1 were involved in an equilibrium with dba.
(b) P d (d ba )2 a n d Liga n d 1. The ortho-substituted
ligand 1 produced a macrocyclic trans-Pd0(dba)(1) com-
plex in CDCl3 after 24 h of mixing with Pd(dba)2, as
observed by the formation of two doublets at δ1 ) 22.94
and δ2 ) 44.71 ppm with a large geminal trans-coupling
constant of 459 Hz (Table 1, Figure 3b).12 However, in
DMF, three signals were detected, 1 h after mixing
Pd(dba)2 and 1 equiv of 1 and remained unchanged from
1 to 24 h. Two singlets of equal integration were
observed at -9.21 ppm (δ1) and 31.16 ppm (δ3) (Table
1), and a third singlet was seen at 27.5 ppm (δ2). A
similar spectrum was obtained in THF (Table 1). Both
the close vicinity of δ1 when compared to the signal of
the free ligand (Table 1) and the same intregration for
δ1 and δ3 suggest that these two signals characterize a
Pd0 complex ligated by only one phosphorus of the
ligand 1, the second one being free. This complex did
not exhibit any absorption band around 390 nm, as is
always observed for the Pd0(η2-dba)(P,P) complex.9c
Consequently, a structure such as A is proposed (Scheme
5) with a bis-ligation of dba, leading to a 16-electron
complex, compatible with the 31P NMR and UV spec-
troscopy data. The singlet at δ2 characterizes a second
complex displaying two equivalent phosphorus ligated
to the Pd0 center (Scheme 5) as in Pd0(1). The coordina-
tion of the Pd0 center by DMF (or THF) would enhance
the stability of such a complex.
Pd0(dba)(2) formed at longer time (characterized by 31
P
NMR spectroscopy, Table 1, and by the oxidation peak
O2) is then involved in an equilibrium with Pd0(2)
(oxidation peak O′1) and dba (Scheme 4, last reaction).
The complex Pd0(2) could not be characterized by 31P
NMR spectroscopy (vide supra), suggesting that its
thermodynamic concentration in the equilibrium with
dba and Pd0(dba)(2) (Scheme 4, last reaction) was very
low. However, Pd0(2) was detected by cyclic voltamme-
try. Indeed, its oxidation at O′1 resulted in a shift of
the equilibrium toward the formation of Pd0(2) due to
its consumption in the diffusion layer (CE mechanism).9c
The oxidation current of Pd0(2) at O′1 reflected then its
dynamic concentration. A 14-electron Pd0 complex li-
gated to a bidentate P,P ligand9a as in Pd0(2) is then
detected and characterized for the first time by cyclic
voltammetry performed in DMF.
Complexes A and Pd0(1) are involved in an equilib-
rium with dba (Scheme 5) since the ratio A/Pd0(1)
(determined by the integration of the corresponding 31
P
NMR signals describe above) increased upon addition
of dba to the 1:1 mixture of Pd(dba)2 and 1 in DMF.
Upon addition of 100 equiv of PhI to the DMF
solution, all three signals disappeared with the evolu-
tion of a new singlet at δ4 ) 21.09 ppm ascribed to trans-
PhPdI(1), as was done for the diphosphine 2. This is
also in favor of an equilibrium between complexes A and
Pd0(1) drawn by the oxidative addition of PhI to one (or
to both) Pd0 complex(es). A single macrocyclic Ph-PdII
As reported for other bidentate ligands,9a the forma-
tion of Pd0(dba)(2) proceeds first via the fast formation
of Pd0(2)2 generated at short times (t < 10 min, oxidation
peak O1, not detectable in 31P NMR spectroscopy)
(Scheme 4).
The oxidation peaks of Pd0(dba)(2) and Pd0(2) disap-
peared upon addition of 5 equiv of PhI, and a new
oxidation peak appeared at +0.51 V, which was as-
signed to PhPdI(2), characterized by 31P NMR spectros-
copy (vide supra, Table 1). The location of the oxidation
peak of PhPdI(2) between those of Pd0(dba)(2) and
(12) (a) Fahey, D. R. J . Organomet. Chem. 1973, 57, 385. (b) Ogilvie,
F. B.; J enkins, J . M.; Verkade, J . G. J . Am. Chem. Soc. 1970, 92, 1916.

5248 Organometallics, Vol. 21, No. 24, 2002
Sørensen et al.
complex is indeed generated. Again, the trans-relation-
ship between the two phosphorus is attributed to the
large flexibility of the macrocyclic ring in comparison
to other well-known bidentate ligands.
The two Pd0 complexes described in Scheme 5 have
been characterized by 31P NMR and UV spectroscopy
at long times, i.e., 1 h after mixing Pd(dba)2 and 1 equiv
of 1 (vide supra). The solution could be characterized
at shorter times by UV spectroscopy. An absorption
band was observed at 395 nm just after mixing, which
usually characterizes a Pd0 complex monoligated by dba,
such as in Pd0(dba)(1). However, such a complex was
not stable since the absorbance at 395 nm continuously
decreased with time. This suggests that the formation
of complex A detected after 1 h reaction (Scheme 5)
proceeds via the initial formation of Pd0(dba)(1).
F igu r e 4. Kinetics of the oxidative addition of PhI to the
Pd0 generated from Pd0(dba)2 (1 mM) and 2 (1 mM) in DMF
at 25 °C as monitored by UV spectroscopy at 392 nm. Plot
of kobs versus PhI concentration: (9) in the absence of added
(c) P d (d ba )2 a n d Liga n d s 3-6. Subjecting the para-
isomer 3 to Pd(dba)2 in CDCl3 afforded a more compli-
cated mixture upon complexation to Pd0 according to
the 31P NMR spectrum, even when left standing over
longer periods of time (Figure 3c). Hence, further work
with this ligand was abandoned. On the other hand,
ligands 4 and 6 produced more consistent results.
Although these diphosphines proved to be sparingly
soluble in either THF or CDCl3, both afforded macro-
cyclic cis-complexes corresponding to Pd0(dba)(4) or
Pd0(dba)(6) upon mixing with Pd(dba)2 in DMF, as
revealed by a set of small doublets at δ ) 25.50 and
26.40 ppm and δ ) 27.24 and 28.96 ppm, respectively
A less well-defined system arose with the ligand 5,
which upon mixing with Pd(dba)2 in CDCl3 produced
two signals as a sharp and broad singlet at δ ) 24.30
and 31.00 ppm, respectively, the origins of which were
not pursued. In any event, the macrocyclization was
supported by mass spectroscopy (ES), where a peak for
Pd0(5) + Na+ was obtained. It is interesting to note the
contrast in complexation between the meta- and para-
substituted pairs of ligands, 2 and 3, with 5 and 6, which
most likely reflect the differences in geometry and cavity
size of the receptor complexes.
dba; (b) in the presence of added dba (10 equiv). kobs
Kk[PhI]/[dba], slope ) 0.685, r ) 0.998.
)
Sch em e 6
dba. The observed rate constants kobs of the overall
reaction are displayed in Figure 4. This deaccelerating
effect of dba establishes that Pd0(2) is the more reactive
species. The reaction order in PhI was determined in
the presence of added dba (10 equiv) by plotting the
variation of kobs versus the PhI concentration. A straight
line was obtained, which went through origin (Figure
4), in contrast to what was observed with other P,P
ligands.9a,c This established that the reaction order in
PhI is +1 and that Pd0(2) is the unique reactive species
(Scheme 6). kobs ) Kk[PhI]/[dba]. Kk ) 0.008 s-1
.
This oxidative addition was considerably slower in
THF than in DMF and consequently was not quantified.
This means that the concentration of Pd0(2) was con-
siderably higher in DMF than in THF. This is consistent
with the observation of Pd0(2) by cyclic voltammetry in
DMF but not in THF (vide supra). Interestingly, this
solvent effect was not observed for the similar reaction
with triphenylphosphine as the ligand.13
The effect of barbital on the kinetics of the oxidative
addition was tested. The rate constant was not signifi-
cantly affected upon addition of barbital (1 equiv),
suggesting that this molecule was not incorporated
inside the ligand of the Pd0-complex (vide infra).
IV. Det er m in a t ion of Bin d in g Con st a n t s of
Liga n d s 1-3, 5, a n d 6 w ith Ba r bita l a n d Bin d in g
Stu d ies w ith th e P d 0(d ba )(2) Com p lex. Titration
experiments were performed in CDCl3 where complex-
ation was expected to be greatest for the three solvents
examined, measuring the typical large downfield shifts
of the host’s amide NH’s (approximately 1.5 ppm) as a
function of the barbiturate concentration. From the
titration curve, the association constant (Ka) can there-
III. Ra te Stu d ies on th e Oxid a tive Ad d ition of
P h I w it h t h e P a lla d iu m (0) Com p lex F or m ed in
Situ betw een P d (d ba )2 a n d th e Bid en ta te P h os-
p h in e Liga n d 2. A previous investigation of the
reactivity of Pd0 complexes generated from Pd0(dba)2
and a bidentate P,P ligand (1 equiv) in the oxidative
addition of PhI in THF has established that both Pd0-
(dba)(P,P) and Pd0(P,P) were reactive but that Pd0(P,P)
was considerably more reactive than Pd0(dba)(P,P).9a,c
The overall reaction was slower in the presence of dba,
by decreasing the concentration of the more reactive
Pd0(P,P).
The UV spectrum of a solution in DMF of Pd0(dba)-
(2) formed in situ from Pd0(dba)2 (1 mM) and 2 (1 mM)
exhibited an absorption band at 392 nm, characteristic
of the dba monoligated to the Pd0 center in Pd0(dba)-
(2).9b,c The absorbance remained unchanged after ad-
dition of dba (10 equiv). Addition of PhI in excess to the
UV cell at 25 °C resulted in the decay of the absorbance
of Pd0(dba)(2) versus time. The kinetics of the oxidative
addition was then monitored by UV spectroscopy. Two
reactions were first investigated with PhI (10 equiv) in
the absence and then in the persence of added dba (10
equiv). The reaction was slower in the presence of added
(13) Amatore, C.; J utand, A.; Khalil, F.; M’Barki, M. A.; Mottier, L.
Organometallics 1993, 12, 3168.

Diphosphine-Palladium(0) Complexes
Organometallics, Vol. 21, No. 24, 2002 5249
Ta ble 2. Bin d in g Con sta n ts Deter m in ed for
Ba r bita l w ith Liga n d s 1-3, 5, a n d 6
a
receptor
Ka (M-1
)
1
2
3
5
6
7900
5100
7100
700
290
fore be extracted according to that previously reported.
Initial experiments with ligands 1-3 and barbital
afforded association constants in the range expected for
open receptors of this type (Table 2). Less exceptional
were the ligands 5 and 6, displaying Ka’s of less than
1000 M-1, being indicative of barbital binding to only a
single diaminopyridine unit of the receptor. This was
somewhat unexpected, as simple molecular modeling
revealed that a particular conformer of these hosts could
be located with a sufficient cavity size for the binding
of the barbiturate guests.
F igu r e 5. Proposed structure of Pd(dba)(2) in CDCl3 and
THF.
between 2 and Pd(dba)2. The existence of the structures
17a and 17b provides a plausible explanation for this
observation.14
Of all the ligands tested, only the diphosphine 2
appeared suitable for further binding studies due to both
its fast complexation with Pd0(dba)2, being independent
of the three solvents examined, and its tendancy for cis-
complex formation, which is a necessary requirement
in the following Heck reactions. Whereas the Pd0(dba)-
(2) complex was initially studied by 31P NMR, quite
Interestingly, when the 1H NMR spectrum of the
Pd0(6)(dba) complex was examined, no significant down-
field shifts were noted for the four amide protons, which
may be consistent with the smaller cavity for this
receptor and hence being more appropriate for the
complexation to Pd0.
1
surprisingly, its H NMR spectrum in CDCl3 displayed
In any event, it appears that for all cases examined
the barbiturate-binding site in the receptor is signifi-
cantly perturbed upon complexation to palladium(0),
thereby excluding the possibility for recognition of the
barbiturate substrates. Diphosphines with larger and/
or less rigid linkers are required in order that a
particular conformer of the macrocyclic palladium(0)
complex may be adapted to bind the barbiturate guest.
V. Heck Rea ction Stu d ies w ith th e Liga n d s 1, 2,
a n d 4-6. Finally, a small study was undertaken to
examine whether this class of macrocyclic diphosphine-
palladium(0) complexes could indeed promote the Heck
reaction. These reactions were carried out with various
aryl halides and n-butyl acrylate with 1:1 mixtures of
the synthesized diphosphines and Pd(dba)2 using tri-
ethylamine as the stoichiometric base. As the rates of
oxidative addition were found to be highest in DMF, all
reactions were performed in this solvent. The results
were compared to those employing PPh3 and are shown
Table 3. Whereas PPh3 and ligand 1 could only advance
the reaction with phenyl iodide at 60 °C (entries 1-4),
it was interesting to observe the aptitude of ligand 2 to
promote the same reaction at room temperature (entry
6). More importantly, the Pd(dba)2:2 system exhibited
sufficient catalytic activity for the same coupling with
both an activated aryl bromide at 60 °C (entry 7) and
an unactivated aryl bromide such as phenyl bromide at
120 °C (entry 8), providing in the latter case n-butyl
cinnamate in an 85% isolated yield. With an activated
aryl chloride the reaction proved also possible but
required longer reaction times (entry 9). A product yield
of 13% was obtained after 42 h at 120 °C along with
recovered starting material.
some interesting hydrogen-bonding features in com-
parison to the uncomplexed 2. For example, in 2 alone,
the four amide protons are observed as two broad
singlets at δ ) 8.52 (2H) and 8.33 (2H) ppm. Upon
treatment with Pd(dba)2, these NH signals display a
large downfield shift in addition to being split up into
four sharp singlets representing one proton each (δ )
11.20, 10.34, 9.75, and 9.65 ppm), suggesting the
involvement of some form of hydrogen-bonding event
with these protons. Evidence for this process being
intramolecular was gained by performing dilution ex-
periments, resulting in the same large chemical shifts.
On the addition of excess barbital, no change was
observed in the NH chemical shifts in either the Pd0-
(dba)(2) complex or barbital, implying that complexation
cannot be made. Furthermore, addition of Pd(dba)2 to
a solution of the barbital:2 system led to quick formation
of Pd0(dba)(2) and free barbital, indicating that com-
plexation of Pd0 to the diphosphine 2 is stronger than
the six-point hydrogen-bonding complex between the
host and guest. Hence, it appears that the binding of 2
to Pd0 possibly results in either a considerable contrac-
tion of the cavity of the ligand or a pertubation from
planarity required for the inclusion of the guest. The
above chemical shifts observed for the amide protons
upon complexation indicate some form of internal
hydrogen bonding. We speculate as to whether the
origins of these shifts are due to the tautomerization of
one of the two terminal amides leading to the presence
of tautomers 17a and 17b, as illustrated in Figure 5,
where hydrogen bonding is favored to the pyridyl
nitrogen. The driving force for this tautomerization
could be the obtainment of more optimal Pd-P bond
lengths and the P-Pd-P angle for the complexation of
2 to Pd(dba)2. As discussed in section IIa, two sets of
signals appear in the 31P NMR spectrum, suggesting the
possibility of two closely related cis-complexes formed
Last, a few structural modifications were made on
ligand 2 in order to examine what effect such changes
(14) We have so far not been successful in crystallizing the Pd0-
complexes with the diphosphine ligands 1-6.

5250 Organometallics, Vol. 21, No. 24, 2002
Sørensen et al.
Ta ble 3. P d -Ca ta lyzed Heck Rea ction of Ar yl
(solvent), in the more ligating solvents such as DMF
could also explain this effect.
Ha lid es w ith n -Bu tyl Acr yla tea
Whereas the receptors possessing an isophthaloyl
connector bind barbital with affinities corresponding to
those of the previously reported open receptors, there
are considerable deviations from the preferred con-
former of the receptor for binding barbital upon com-
plexation with Pd(dba)2. More flexible linkers in the
receptor are therefore required in order for these Pd0
complexes to bind barbiturate guests. Such work is
currently ongoing and will be reported in due course.
Finally, these complexes were found to promote the
Heck reaction with reactivities similar to triphenylphos-
phine. In one case, with Pd0(dba)(2), a higher reactivity
was noted.
entry
ArX
ligand temp (°C) rxn time (h) yield (%)b
c
1
2
3
4
5
6
7
8
PhI
PhI
PhI
PhI
PhI
PhI
PPh3
PPh3
60
20
60
20
60
20
60
120
120
20
24
48
24
48
3
30
24
24
42
30
24
30
7
88
d
89
d
87
82
48e
85
13e
d
85
d
87
10
47
c
1
1
2
2
2
2
2
4
5
5
6
6
6
(p-Ac)PhBr
PhBr
(p-Ac)PhCl
PhI
9
10
11
12
13
14
15
PhI
PhI
PhI
PhI
60
20
60
20
Exp er im en ta l Section
44
24
Gen er a l Con sid er a tion s. Unless otherwise stated, all
reactions were carried out under argon. Dichloromethane was
freshly distilled over P2O5, THF and diethyl ether from sodium/
benzophenone, and acetonitrile from calcium hydride, whereas
DMF was distilled under reduced pressure and stored over 4
Å molecular sieves. Reactions were monitored by thin-layer
chromatography (TLC) analysis on Kieselgel 60 F254 (Merck).
NMR spectra (1H at 200 MHz and 13C at 50 MHz) were
recorded on a Varian Gemini 2000 spectrometer. Chemical
shifts (δ in ppm) are given relative to those for Me4Si. 31P NMR
spectra were recorded on a Bruker or Varian spectrometer (162
MHz) using H3PO4 as an external reference. ES mass spectra
were recorded with a Micromass LC-TOF instrument. UV
spectroscopy was recorded on a Beckman DU 7400 spectrom-
eter. The electrochemical setup and electrochemical procedure
for cyclic voltammetry and the UV studies were performed as
previously published.9 The complexation studies and monitor-
ing by 31P NMR were carried out as previously reported.9 The
1H NMR binding studies were performed employing standard
methods.7 The diamine 7 and Pd(dba)2 were prepared accord-
ing to published procedures.5,9a
N,N′-Bis[6-(2-iodoben zoylam in o)pyr idin -2-yl]isoph th al-
a m id e (8). Gen er a l P r oced u r e for th e Ben zoyla tion of
th e Dia m in es 7 a n d 14. Oxalyl chloride (1.75 mL, 20.2 mmol)
and a catalytic amount of DMF (15 µL) were added to a
solution of 2-iodobenzoic acid (2.12 g, 8.54 mmol) in CH2Cl2
(30 mL). After stirring for 1 h at 20 °C, the solvent was
removed under reduced pressure, affording the crude benzoyl
chloride, which was redissolved in THF (10 mL). To this
solution were added the diamine 7 (686 mg, 1.97 mmol) and
triethylamine (1.1 mL, 7.88 mmol) in THF (50 mL). An
additional 20 mL of THF was added, and the reaction mixture
was stirred at 20 °C for 18 h. After removal of the solvent
under reduced pressure, ethyl acetate was added and the
organic phase was washed with water. The aqueous phase was
extracted once with ethyl acetate, and the combined organic
phases were dried (Na2SO4) and evaporated to dryness under
vacuum. Column chromatography (ethyl acetate/CH2Cl2, 1:9)
afforded 8 as light yellow crystals (1.35 g, 85% yield): mp 231-
235 °C; 1H NMR (DMSO-d6) δ 10.72 (s, 2H, CONH), 10.60 (s,
2H, CONH), 8.54 (s, 1H, Ar), 8.15 (dd, 2H, J ) 7.8, 1.4 Hz,
Ar), 7.92-7.89 (m, 8H, Ar), 7.66 (t, 1H, J ) 7.8 Hz, Ar), 7.47-
7.42 (m, 4 H, Ar), 7.25-7.17 (m, 2H, Ar); 13C NMR (DMSO-
d6) δ 169.1 (2C), 166.3 (2C), 151.3 (2C), 151.1 (2C), 143.4 (2C),
141.0 (2C), 139.8 (2C), 134.9 (2C), 132.3 (2C), 131.9 (2C), 129.6,
128.9 (4C), 128.3, 112.1 (2C), 111.8 (2C), 95.3 (2C); IR (KBr) ν
3264, 1677, 1583, 1505, 1446, 1300 cm-1; MS (ES) calcd for
PhBr
120
a
Reaction conditions: aryl halide (2.5 mmol), n-butyl acrylate
b
(2.5 mmol), NEt3 (2.75 mmol). Isolated yield after column chro-
matography. c 1 mol % phosphine used. No product detected
d
according to TLC analysis. e 1H NMR spectrum of the crude
product mixture revealed the remaining material to be unreacted
starting material.
would evoke on the activity of the palladium catalyst.
However, replacement of the pyridyl rings with phenyl,
as in the diphosphine 4, or the isophthaloyl group with
a ferrocene unit (diphosphine 5) led only to ligands
revealing a outcome comparable to PPh3 in the same
Heck reaction with phenyl iodide (Table 3, entries 10-
12). The para-ligand 6 revealed only a slight improve-
ment but still not quite as effective as the ligand 2
(entries 13-15).
These results establish that a Heck reaction per-
formed from an aryl halide and an alkene containing
an electron-withdrawing group may be efficiently cata-
lyzed by a Pd ligated to a bibentate bisphosphine ligand
(P,P), which are usually considered to be inappropri-
ate1b,4b for such purposes due to the difficult deligation
of one P in cis-ArPdX(P,P), required for the coordination
of the olefin. However, in the present case, a trans-
ArPdX(2) is formed in the oxidative addition as moni-
tored by 31P NMR. This suggests that the two phospho-
rus are less close to each other than in a cis-complex
and that the deligation of one P should then be easier
in this trans-complex than in a cis-complex.
Con clu sion
In conclusion, we have prepared a series of diphos-
phine ligands possessing a barbiturate-binding receptor,
where some were found to efficiently form macrocyclic
bisphosphine palladium(0) complexes even though a 26-
membered cycle is created. In one case, with ligand 2,
a significant solvent effect was observed for the oxida-
tive addition of the Pd0 complexes with phenyl iodide.
Although the exact cause for this effect has not been
deduced, it could be related to distinct conformational
or tautomer preferences of the metal-containing mac-
rocycle in the solvents tested, thereby giving rise to
potentially different bite angles of the bidentate ligand
to the metal center. Higher concentrations of the species
responsible for the oxidative addition, namely, Pd(P,P)-
C
32H22I2N6O2 831.0 (M + Na), found 831.1.
N,N′-Bis[6-(2-{d ip h en ylp h osh a n yl}ben zoyla m in o)p yr -
id in -2-yl]isop h th a la m id e (1). Gen er a l P r oced u r e for th e
In tr od u ction of Dip h en ylp h osp h in e. Triethylamine (733

Diphosphine-Palladium(0) Complexes
Organometallics, Vol. 21, No. 24, 2002 5251
µL, 5.26 mmol), Ph2PH (322 µL, 1.85 mmol), and Pd(OAc)2 (1.0
mL from a solution of 10.5 mg of Pd(OAc)2 in 5 mL of CH3CN,
9.4 µmol) were added to a solution of the diiodide 8 (709 mg,
0.88 mmol) in CH3CN (15 mL). The reaction mixture was
warmed to 85 °C in a closed tube and stirred for 2 h. Upon
completion of the reaction, a clear red-colored solution is
obtained, which was cooled and then poured into a separatory
flask with CH2Cl2. The organic phase was washed with a
saturated solution of NH4Cl, dried (Na2SO4), and then evapo-
rated to dryness under vacuum. Column chromatography
(ethyl acetate/CH2Cl2, 1:19) afforded 1 as light yellow crystals
(811 mg, 83% yield): mp 184-186 °C; 1H NMR (CDCl3) δ 8.48
(s, 1H, Ar), 8.42 (br s, 2H, CONH), 8.24 (br s, 2H, CONH),
8.14 (br d, 2H, J ) 8.4 Hz, Ar), 8.05 (d, 2 H, J ) 8.4 Hz, Ar),
7.88 (d, 2H, J ) 8.4 Hz, Ar), 7.74-7.66 (m, 5H, Ar), 7.42-7.26
(m, 24H, Ar), 7.05-7.02 (m, 2H, Ar); 13C NMR (CDCl3) δ 167.7
(2C), 164.9 (2C), 149.8 (2C) 149.7 (2C), 140.8 (d, 2C, J CP ) 24.5
Hz), 140.5 (2C), 137.0 (d, 2C, J CP ) 20.1 Hz), 136.6 (d, 4C,
J CP ) 9.0 Hz), 134.6 (2C), 134.5 (d, 2C, J CP ) 5.7 Hz), 134.1
(d, 8C, J CP ) 20.1 Hz), 131.8 (2C), 130.9 (2C), 129.3 (2C), 129.2
(4C), 128.8 (d, 8C, J CP ) 7.2 Hz), 128.1 (d, 2C, J CP ) 4.7 Hz),
125.8 (2C), 110.8 (2C) 110.5 (2C); 31P NMR (CDCl3) δ -8.97;
IR (KBr) ν 3394, 1685, 1585, 1501, 1446, 1302 cm-1; MS (ES)
calcd for C56H42N6O4P2 947.3 (M + Na), found 947.6.
N,N′-Bis[6-(3-iodoben zoylam in o)pyr idin -2-yl]isoph th al-
a m id e (9). The diiodide 9 was prepared according to the
general procedure outlined for 8, with the following quanti-
ties: 3-iodobenzoic acid (1.25 g, 5.04 mmol) in CH2Cl2 (25 mL),
oxalyl chloride (1.75 mL, 20.2 mmol), diamine 7 (729 mg, 2.09
mmol), and triethylamine (1.2 mL, 8.36 mmol) in THF (40 mL).
Column chromatography (ethyl acetate/CH2Cl2, 1:9) afforded
9 as colorless crystals (1.18 g, 70% yield): mp 272-275 °C; 1H
NMR (DMSO-d6) δ 10.57 (s, 4H, CONH), 8.51 (s, 1H, Ar), 8.24
(dd, 2H, J ) 1.6, 1.6 Hz, Ar), 8.11 (dd, 2 H, J ) 7.8 Hz, 1.4 Hz,
Ar), 7.92-7.73 (m, 10 H, Ar), 7.62 (t, 1H, J ) 7.8 Hz, Ar), 7.23
(t, 2 H, J ) 7.8 Hz, Ar); 13C NMR (DMSO-d6) δ 166.1 (2C),
165.2 (2C), 151.2, 151.1 (4C), 141.3 (2C), 141.0 (2C), 137.2 (2C),
136.9 (2C), 134.9 (2C), 132.3 (2C), 131.4 (2C), 128.1 (4C), 112.5
(2C) 112.1 (2C), 95.55 (2C); IR (KBr) ν 3286, 1652, 1588, 1516,
1447, 1306 cm-1; MS (ES) calcd for C32H22I2N6O2 831.0 (M +
Na), found 831.1.
combining the organic phases, a precipitate was obtained. This
was filtered and washed with cold ethyl acetate, affording
crude 10 as a light brown powder (1.23 g, 95%): mp 282-284
°C; 1H NMR (DMSO-d6) δ 10.56 (s, 2H, CONH), 10.49 (s, 2H,
CONH), 8.50 (s, 1H, Ar), 8.10 (dd, 2H, J ) 7.8, 1.4 Hz, Ar),
7.87-7.58 (m, Ar); 13C NMR (DMSO-d6) δ 166.1 (4C), 151.2
(2C) 151.1 (2C), 141.0 (2C), 138.4 (2C), 138.2 (2C), 134.9 (2C),
134.3 (4C), 131.9 (4C), 130.6 (2C), 112.4 (2C), 112.0 (2C), 100.8
(2C); IR (KBr) ν 3293, 1782, 1652, 1584, 1517, 1456, 1307 cm-1
;
MS (ES) calcd for C32H22I2N6O2 831.0 (M + Na), found 831.0.
N,N′-Bis[6-(4-{d ip h en ylp h osh a n yl}ben zoyla m in o)p yr -
id in -2-yl]isop h th a la m id e (3). The bisphosphine 3 was pre-
pared according to the general procedure outlined for 1, with
the following quantities: diiodide 10 (294 mg, 0.36 mmol) in
CH3CN (1.5 mL), triethylamine (304 µL, 2.18 mmol), Ph2PH
(132 µL, 0.76 mmol), and Pd(OAc)2 (499 µL from a solution of
8.2 mg of Pd(OAc)2 in 5 mL of CH3CN, 3.6 µmol). A reaction
time of 20 h was allowed due to the low solubility of the
diiodide. Column chromatography (ethyl acetate/CH2Cl2, 1:19)
afforded 3 as colorless crystals (118 mg, 35% yield): mp 190-
1
192 °C; H NMR (CDCl3) δ 8.50 (br s, 2H, CONH), 8.43 (m,
3H, Ar + CONH), 8.08 (d, 2H, J ) 8.2 Hz, Ar), 8.03 (d, 4H,
J ) 7.6 Hz, Ar), 7.81 (d, 4H, J ) 7.6 Hz, Ar), 7.75 (t, 2H, J )
8.2 Hz, Ar), 7.55 (t, 1H, J ) 8.2 Hz, Ar), 7.38-7.30 (m, 24H,
Ar); 13C NMR (CDCl3) δ 165.6 (2C), 164.7 (2C), 150.0 (2C),
149.6 (4C), 143.8 (d, 2C, J CP ) 14.3 Hz), 141.0 (2C), 136.2 (d,
4C, J CP ) 10.4 Hz), 134.6 (2C), 134.1 (d, 8C, J CP ) 20.1 Hz),
133.9 (2C), 133.7 (d, 4C, J CP ) 19 Hz), 131.1 (2C), 129.4 (4C),
128.9 (d, 8C, J CP ) 7.2 Hz), 127.3 (d, 4C, J CP ) 6.1 Hz), 126.2
(2C), 110.7 (2C), 110.3 (4C); 31P NMR (CDCl3) δ -4.75; IR
(KBr) ν 3411, 1686, 1586, 1508, 1446, 1301 cm-1; MS (ES) calcd
for C56H42N6O4P2 947.3 (M + Na), found 947.6.
N-(3-Am in op h en yl)-3-iod oben za m id e (11). Oxalyl chlo-
ride (0.40 mL, 4.59 mmol) and a catalytic amount of DMF (15
µL) were added to a solution of 3-iodobenzoic acid (250 mg,
1.02 mmol) in CH2Cl2 (20 mL). After stirring for 1 h at 20 °C,
the solvent was removed under reduced pressure, affording
the crude benzoyl chloride, which was redissolved in THF (10
mL). To this solution was added the m-diaminobenzene (170
mg, 0.79 mmol) and triethylamine (0.29 mL, 2.62 mmol) in
THF (20 mL). The reaction mixture was stirred at 20 °C for
70 h. After removal of the solvent under reduced pressure,
ethyl acetate was added and the organic phase was washed
with water. The aqueous phase was extracted once with ethyl
acetate, and the combined organic phases were dried (Na2SO4)
and evaporated to dryness under vacuum. Column chroma-
tography (ethyl acetate/pentane, 1:2) afforded 11 as a colorless
solid (240 mg, 88% yield): 1H NMR (DMSO-d6) δ 10.03 (s, 1H,
CONH), 8.25 (t, 1H, J ) 1,6 Hz, Ar), 7.92 (dd, 2H, J ) 1.8, 7.8
Hz, Ar), 7.31 (t, 1H, J ) 7,8 Hz, Ar), 7.08 (t, 1H, J ) 1,8 Hz,
Ar), 6.97 (t, 1H, J ) 7,8 Hz, Ar), 6.86 (dt, 1H, J ) 8.0, 1.2 Hz,
Ar), 6.32 (dt, 1H, J ) 7.8, 1.2 Hz, Ar), 5.10 (s, 2H, NH2); 13C
NMR (DMSO-d6) δ 164.5, 149.8, 140.7, 140.3, 138.1, 136.7,
131.3, 129.7, 127.9, 110.8, 109.2, 106.9, 95.5; IR (KBr) ν 3448,
3354, 3239, 1648, 1610, 1546, 1494, 1441, 1327 cm-1; HR-MS
(ES) calcd for C13H11N2OI 360.9814 (M + Na), found 360.9815.
N,N′-Bis[6-(3-{d ip h en ylp h osh a n yl}ben zoyla m in o)p yr -
id in -2-yl]isop h th a la m id e (2). The bisphosphine 2 was pre-
pared according to the general procedure outlined for 1, with
the following quantities: diiodide 9 (150 mg, 0.19 mmol) in
CH3CN (3.5 mL), triethylamine (155 µL, 1.11 mmol), Ph2PH
(68 µL, 0.39 mmol), and Pd(OAc)2 (204 µL from a solution of
10.2 mg of Pd(OAc)2 dissolved in 5 mL of CH3CN, 1.9 µmol).
Column chromatography (ethyl acetate/CH2Cl2, 1:19) afforded
2 as colorless crystals (148 mg, 86% yield): mp 174-176 °C;
1H NMR (CDCl3) δ 8.52 (br s, 2H, CONH), 8.46 (br s, 1H, Ar),
8.33 (br s, 2H, CONH), 8.09-8.04 (m, 6H, Ar), 7.93 (br d, 2H,
J ) 8.4 Hz, Ar), 7.84-7.82 (m, 2H, Ar), 7.76 (t, 2H, J ) 8.2
Hz, Ar), 7.59 (t, 1H, J ) 7.8 Hz, Ar), 7.43-7.38 (m, 4H, Ar),
7.34-7.26 (m, 20H, Ar); 13C NMR (CDCl3) δ 165.7 (2C), 164.9
(2C), 149.9 (2C) 149.6 (2C), 140.9 (2C), 139.2 (d, 2C, J CP ) 13.6
Hz), 137.2 (d, 2C, J CP ) 13.6 Hz), 136.3 (d, 4C, J CP ) 10.4 Hz),
N,N′-Bis[3-(3-iod ob en zoyla m in o)p h en yl)isop h t h a la -
m id e (12). Triethylamine (50 µL, 0.36 mmol) and isophthaloyl
chloride (24 mg, 0.12 mmol) dissolved in THF (5 mL) were
added to a solution of the amine 11 (91 mg, 0.27 mmol) in THF
(25 mL). After stirring for 20 h at 20 °C, the solvent was
removed under reduced pressure. Water was added, and the
solid obtained was filtered and washed with water and then
ethyl acetate. This afforded the crude diiodide 12 as a colorless
powder (68 mg, 70% yield), which was sufficiently pure for the
next step: 1H NMR (DMSO-d6) δ 10.50 (s, 2H, CONH), 10.40
(s, 2H, CONH), 8.54 (s, 1H, Ar), 8.35 (t, 2H, J ) 1.4 Hz, Ar),
8.31 (t, 2H, J ) 1.4 Hz, Ar), 8.16 (dd, 2H, J ) 7.6, 1.6 Hz, Ar),
7.97 (dt, 4H, J ) 8.0, 1.6 Hz, Ar), 7.70 (dt, 1H, J ) 7.6 Hz,
Ar), 7.53 (t, 4H, J ) 7.4 Hz, Ar), 7.34 (t, 4H, J ) 8.0 Hz, Ar);
134.6 (2C), 134.4 (d, 2C, J CP ) 7.6 Hz), 134.0 (d, 8C, J CP
)
19.8 Hz), 132.9 (d, 2C, J CP ) 19.8 Hz), 131.3 (2C), 129.4 (d,
2C, J CP ) 19.8 Hz), 129.3 (4C), 128.9 (d, 8C, J CP ) 7.2 Hz),
127.7 (3C), 126.1, 110.7 (2C) 110.4 (4C); 31P NMR (CDCl3): δ
-4.78; IR (KBr) ν 3277, 1685, 1586, 1507, 1447, 1299 cm-1
;
MS (ES) calcd for C56H42N6O4P2 947.3 (M + Na), found 947.7.
N,N′-Bis[6-(4-iodoben zoylam in o)pyr idin -2-yl]isoph th al-
a m id e (10). The diiodide 10 was prepared according to the
general procedure outlined for 8, with the following quanti-
ties: 4-iodobenzoic acid (1.19 g, 4.80 mmol) in CH2Cl2 (25 mL),
oxalyl chloride (1.70 mL, 19.2 mmol), diamine 7 (528 mg, 1.52
mmol), and triethylamine (846 µL, 6.07 mmol) in THF (50 mL).
The product was only slightly soluble in ethyl acetate; hence
after extraction several times with ethyl acetate followed by

5252 Organometallics, Vol. 21, No. 24, 2002
Sørensen et al.
IR (KBr) ν 3304, 1648, 1607, 1538, 1489, 1428, 1327 cm-1; HR-
MS (ES) calcd for C34H24I2N4O4 828.9785 (M + Na), found
828.9774.
(m, 10H, Ar), 7.31 (br s, 2H, Ar), 5.11 (br s, 4H, Cp), 4.51 (br
s, 4H, Cp); 13C NMR (DMSO-d6) δ 167.3 (2C), 164.0 (2C), 150.2
(2C), 149.8 (2C), 140.2 (2C), 139.6 (2C), 136.1 (2C), 135.9 (2C),
130.4 (2C), 127.1 (2C), 110.8 (2C), 110.4 (2C), 94.6 (2C), 77.1
(2C), 72.4 (2C), 70.0 (2C); IR (KBr) ν 3410, 1676, 1586, 1508,
1444 cm-1; HR-MS (ES) calcd for C36H26O4N6I2Fe (M + Na)
938.9352, found 938.9321.
N,N′-Bis[6-(3-{d ip h en ylp h osh a n yl}ben zoyla m in o)p yr -
id in -2-yl]-1,1′-fer r ocen ed ica r boxylic Am id e (5). The bis-
phosphine 5 was prepared according to the general procedure
outlined for 1, with the following quantities: diiodide 15 (173
mg, 0.189 mmol) in CH3CN (4 mL), triethylamine (158 µL, 1.13
mmol), Ph2PH (69 µL, 0.396 mmol), and Pd(OAc)2 (123 µL from
a solution of 10.3 mg of Pd(OAc)2 in 5 mL of CH3CN, 1.9 µmol).
Column chromatography (pentane/ethyl acetate, 2:1) afforded
5 as a yellowish brown powder (165 mg, 85% yield): 1H NMR
(CDCl3) δ 8.33 (br s, 2H, CONH), 8.15 (br s, 2H, CONH), 7.98
(dd, 2H, J ) 8.0, 0.7 Hz, Ar), 7.88 (dd, 2H, J ) 8.0, 0.7 Hz,
Ar), 7.82 (d, 4H, J ) 8.6 Hz, Ar), 7.63 (t, 2H, J ) 8.0 Hz, Ar),
7.41-7.30 (m, 24H, Ar), 4.84 (t, 4H, J ) 2.0 Hz, Cp), 4.54 (t,
4H, J ) 2.0 Hz, Cp); 31P NMR (CDCl3) δ -4.7; IR (KBr) ν 3409,
1677, 1585, 1507, 1444 cm-1; HR-MS (ES) calcd for C60H46O4-
N6P2Fe (M + Na) 1055.2303 found 1055.2307.
N,N′-Bis[3-(3-{diph en ylph osh an yl}ben zoylam in o)ph en -
yl]isop h th a la m id e (4). The bisphosphine 4 was prepared
according to the general procedure outlined for 1, with the
following quantities: diiodide 12 (150 mg, 0.44 mmol) in CH3-
CN (2 mL), triethylamine (250 µL, 1.79 mmol), Ph2PH (160
µL, 0.92 mmol), and Pd(OAc)2 (900 µL from a solution of 12
mg of Pd(OAc)2 in 10 mL of CH3CN, 4.8 µmol). A reaction time
of 24 h was allowed due to the low solubility of the diiodide.
During the course of the reaction a preciptitate was formed,
which was filtered and then washed with water and dichlo-
romethane. Drying afforded 4 as a light red powder (110 mg,
28% yield), which was sparingly soluble in THF, CH2Cl2, ethyl
acetate, methanol, and ethanol, but soluble in DMF and
DMSO: 1H NMR (DMSO-d6) δ 10.47 (s, 2H, CONH), 10.39 (s,
2H, CONH), 8.53 (s, 1H, Ar), 8.32 (m, 4H, Ar), 8.14 (dd, 2H,
J ) 7.6, 1.6 Hz, Ar), 8.01 (dt, 2H, J ) 8.2, 1.4 Hz, Ar), 7.89
(m, 2H, J ) 7.8 Hz, Ar), 7.73-7.24 (m, 29 Ar); 31P NMR
(DMSO-d6) δ -5.83; IR (KBr) ν 3218, 1644, 1605, 1527, 1435,
1316 cm-1; MS (ES) calcd for C58H44N4O4P2 977.9 (M + Na),
found 977.8.
N,N′-Bis[6-(4-iod oben zoyla m in o)p yr id in -2-yl]-1,1′-fer -
r ocen ed ica r boxylic Am id e (16). The diiodide 16 was pre-
pared according to the general procedure outlined for 8 with
the following quantities: 4-iodobenzoic acid (455 mg, 1.83
mmol) in CH2Cl2 (10 mL), oxalyl chloride (1.57 mL, 18.0 mmol),
and a catalytic amount of DMF (15 µL), diamine 14 (205 mg,
0.44 mmol), and triethylamine (256 µL, 1.84 mmol) in THF
(13 mL). Column chromatography (ethyl acetate/pentane, 1:1)
afforded 16 as a reddish brown powder (344 mg, 86% yield):
1H NMR (DMSO-d6) δ 10.39 (br s, 2H, CONH), 9.74 (br s, 2H,
CONH), 8.9 (m, 14H, Ar), 5.11 (br s, 4H, Cp), 4.51 (br s, 4H,
Cp); 13C NMR (DMSO-d6) δ 167.4 (2C), 165.0 (2C), 150.4 (2C),
150.0 (2C), 139.7 (2C), 137.3 (4C), 133.4 (2C), 129.7 (4C), 110.8
(2C), 110.5 (2C), 100.0 (2C), 77.3 (2C), 72.5 (4C), 70.1 (4C); IR
(KBr) ν 3407, 2924, 1676, 1585, 1508, 1443 cm-1; HR-MS (ES)
calcd for C36H26O4N6I2Fe (M + Na) 938.9352, found 938.9352.
N,N′-Bis[6-(4-{d ip h en ylp h osh a n yl}ben zoyla m in o)p yr -
id in -2-yl]-1,1′-fer r ocen ed ica r boxylic Am id e (6). The bis-
phosphine 6 was prepared according to the general procedure
outlined for 1, with the following quantities: diiodide 16 (760
mg, 0.83 mmol) in CH3CN (8 mL), triethylamine (0.70 mL, 5.0
mmol), Ph2PH (304 µL, 1.74 mmol), and Pd(OAc)2 (518 µL from
a solution of 18 mg of Pd(OAc)2 in 5 mL of CH3CN, 8.3 mmol).
Column chromatography (CH2Cl2/MeOH, 60:1) afforded 6 as
a brown powder (473 mg, 55% yield): 1H NMR (CDCl3) δ 8.30
(br s, 2H, CONH), 8.13 (br s, 2H, CONH), 7.98 (d, 2H, J ) 8.0
Hz, Ar), 7.88 (d, 2H, J ) 8.0 Hz, Ar), 7.81 (m, 4H, Ar), 7.63 (t,
2H, J ) 8.0 Hz, Ar), 7.37 (m, 24H, Ar), 4.84 (t, 4H, J ) 1.9
Hz, Cp), 4.54 (t, 4H, J ) 1.9 Hz, Cp); 31P NMR (CDCl3) δ -4.74;
IR (KBr) ν 3409, 1677, 1585, 1507, 1444 cm-1; HR-MS (ES)
calcd for C60H46O4N6P2Fe (M + Na) 1055.2303, found 1055.2307.
Gen er a l P r oced u r e for th e Heck Cou p lin g betw een
th e Ar yl Ha lid es a n d n -Bu tyl Acr yla te. A Schlenk tube
containing Pd(dba)2 (7.2 mg, 12.5 µmol, 0.5 mol %) and the
phosphine, either PPh3 (6.6 mg, 25 µmol, 1.0 mol %) or the
bidentate phosphines (11.6 mg, 12.5 µmol, 0.5 mol %), was
purged with argon. DMF (800 µL) was added followed by
triethylamine (383 µL, 2.75 mmol), n-butyl acrylate (358 µL,
2.5 mmol), and phenyl halide (2.5 mmol). The reaction mixture
was stirred at 20, 60, or 120 °C over a period of time, as
indicated in Table 3. Diethyl ether and water were added, and
the aqueous phase was extracted three times with ether. The
combined organic phases were dried (MgSO4) and evaporated
to dryness under vacuum. Column chromatography (ethyl
acetate/pentane, 1:199) afforded n-butyl cinnamate as a color-
less oil: 1H NMR (CDCl3) δ 7.69 (d, 1H, J ) 16.0 Hz, CHd
CH), 7.56-7.51 (m, 2H, Ar), 7.41-7.36 (m, 2H, Ar), 6.44 (d,
1H, J ) 16 Hz, CHdCH), 4.21 (t, 2H, J ) 6.4 Hz, CH2O), 1.77-
1,1′-F er r ocen ed ica r boxylic Acid (13). n-BuLi (64.0 mL,
0.102 mol) was added dropwise to a solution of TMEDA (15.7
mL, 0.103 mol) and ferrocene (7.87 g, 42.2 mmol) in diethyl
ether (260 mL) at 20 °C. After stirring overnight, CO2 was
bubbled through the orange slurry for 2 h. Water was added,
and the mixture was extracted three times with CH2Cl2 and
then acidified with concentrated aqueous HCl. The precipitate
was filtered and dried, affording 13 as a yellow powder (7.07
g, 68% yield), which was sufficiently pure for the next step:
1H NMR (DMSO-d6) δ 4.69 (t, 4H, J ) 2.0 Hz, Cp), 4.45 (t,
4H, J ) 2.0 Hz, Cp); 13C NMR (DMSO-d6) δ 171.1 (2C), 73.5
(2C), 72.6 (4C), 71.3 (4C); IR (KBr) ν 2923, 1677, 1492, 1405,
1300 cm-1; HR-MS (ES) calcd for C12H10O4Fe 296.9826 (M +
Na), found 296.9841.
Bis(6-a m in op yr id in -2-yl)-1,1′-fer r ocen ed ica r b oxylic
Am id e (14). Oxalyl chloride (0.6 mL, 6.9 mmol) and a catalytic
amount of DMF (15 µL) were added to a solution of diacid 13
(188 mg, 0.69 mmol) in CH2Cl2 (20 mL). After stirring for 1 h
at 20 °C, the solvent was removed under reduced pressure,
affording the crude benzoyl chloride, which was redissolved
in THF (11 mL). To this solution were added the 2,6-diamino-
pyridine (445 mg, 4.14 mmol) and triethylamine (0.29 mL, 2.07
mmol) in THF (20 mL), and the reaction mixture was stirred
at 20 °C for 16 h. After removal of the solvent under reduced
pressure, ethyl acetate was added and the organic phase was
washed with water. The aqueous phase was extracted once
with ethyl acetate, and the combined organic phases were
dried (Na2SO4) and evaporated to dryness under vacuum,
affording 14 as a brown powder (208 mg, 66% yield): 1H NMR
(DMSO-d6) δ 9.43 (br s, 2H, CONH), 7.36 (t, 2H, J ) 7.8 Hz,
Pyr), 7.25 (dd, 2H, J ) 7.8, 0.8 Hz, Pyr), 6.21 (dd, 2H, J ) 7.8,
0.8 Hz, Pyr), 5.75 (br s, 4H, R-NH2), 5.05 (t, 4H, J ) 2.0 Hz,
Cp), 4.41 (t, 4H, J ) 2.0 Hz, Cp); 13C NMR (DMSO-d6) δ 167.6
(2C), 158.5 (2C), 150.4 (2C), 138.9 (2C), 103.6 (2C), 102.2 (2C),
77.6 (2C), 72.6 (2C), 70.1 (2C); IR (KBr) ν 3434, 3340, 3226,
1641, 1625, 1464 cm-1; HR-MS (ES) calcd for C22H20O2N6Fe
(M + Na) 479.0895, found 479.0893.
N,N′-Bis[6-(3-iod oben zoyla m in o)p yr id in -2-yl]-1,1′-fer -
r ocen ed ica r boxylic Am id e (15). The diiodide 15 was pre-
pared according to the general procedure outlined for 8 with
the following quantities: 3-iodobenzoic acid (2.10 g, 8.47 mmol)
in CH2Cl2 (40 mL), oxalyl chloride (7.3 mL, 83.7 mmol), and a
catalytic amount of DMF, diamine 14 (1.01 g, 2.21 mmol), and
triethylamine (1.2 mL, 8.60 mmol) in THF (60 mL). Column
chromatography (CH2Cl2/MeOH, 50:1) afforded 15 as a reddish
brown powder (1.56 g, 77%): 1H NMR (DMSO-d6) δ 10.48 (br
s, 2H, CONH), 9.73 (br s, 2H, CONH), 8.30 (br s, 2H, Ar), 7.86

Diphosphine-Palladium(0) Complexes
Organometallics, Vol. 21, No. 24, 2002 5253
1.63 (m, 2H, CH2), 1.50-1.35 (m, 2H, CH2), 0.97 (t, 3H, J )
7.4 Hz, CH3); 13C NMR (CDCl3) δ 167.2, 144.7, 134.6, 130.4,
129.0 (2C), 128.2 (2C), 118.5, 64.6, 31.0, 19.4, 14.0; HR-MS
(ES) calcd for C13H16O2 227.1048 (M + Na), found 227.1053.
Note Ad d ed in P r oof: We have recently prepared
a homologue of the ligand 2, which both complexes to
palladium(0) and binds the guest, barbital. The ability
of this complex to catalyze the Heck reaction and control
regioselectivities is currently under investigation and
will be reported in a separate paper.
Ack n ow led gm en t. We are indebted to the Danish
National Science Foundation, The University of Aarhus,
the Carlsberg Foundation, the Centre National de la
Recherche Scientifique, the Ministere de la Recherche,
Ecole Normale Supe´rieure, and The University Paris
VI for generous financial support.
1
Su p p or tin g In for m a tion Ava ila ble: Copies of H NMR
spectra for compounds 1-6. This material is available free of
charge via the Internet at http//pubs.acs.org.
OM020364V