(π-Allyl) palladium complexes with N,N′-diphenylbispidinone derivatives as a new type of chelating nitrogen ligand: Complexation studies, spectroscopic properties, and an X-ray structure of (3,7-diphenyl-1,5-dimethylbispidinone)[(1,3-η 3-propenyl)-palladium] trifluoromethanesulfonate

Article abstract of DOI:10.1021/om9609527

A series of 3,7-diazabicyclo[3.3.1]nonane (bispidine) derivatives have been synthesized, and their properties as bidentate nitrogen ligands for (π-allyl)palladium complexes have been investigated. Complexes of these ligands and of N,N′-diphenylpiperazine (7) and N,N′-diphenyl-1,5-diazacyclooctane (8) with (1,3-η³-propenyl)palladium are described, in particular their effects on the proton chemical shifts of the π-allyl ligand. Ligand dynamics of the complexes is discussed. The structure of [(3,7-diphenyl-1,5-dimethylbispidinone)(1,3-η ³-propenyl)Pd]CF₃SO₃ (15) has been determined by X-ray crystallography. N,N′-Diphenyl-bispidine derivatives show an unusually large steric interaction with the π-allyl ligand, indicated by a tilt of the π-allyl plane toward the N-Pd-N plane by 122.8(8)°. Chemical shift changes of the π-allyl protons due to the aromatic ring current are related to the geometry of the complexes. The ligands are tested on the larger 2-methylene-6,6-dimethylbicyclo[3.1.1]hept-2,3,10-η³-enyl ligand, demonstrating their potential as chemical shift reagents.

Full text of DOI:10.1021/om9609527

Organometallics 1997, 16, 1167-1178
1167
(π-Allyl)p a lla d iu m Com p lexes w ith
N,N′-Dip h en ylbisp id in on e Der iva tives a s a New Typ e of
Ch ela tin g Nitr ogen Liga n d : Com p lexa tion Stu d ies,
Sp ectr oscop ic P r op er ties, a n d a n X-r a y Str u ctu r e of
(3,7-Dip h en yl-1,5-d im eth ylbisp id in on e)[(1,3-η3-p r op en yl)-
p a lla d iu m ] Tr iflu or om eth a n esu lfon a te†
Adolf Gogoll,* Helena Grennberg, and Andreas Axe´n
Department of Organic Chemistry, University of Uppsala, Box 531, 751 21 Uppsala, Sweden
Received November 4, 1996X
A series of 3,7-diazabicyclo[3.3.1]nonane (bispidine) derivatives have been synthesized,
and their properties as bidentate nitrogen ligands for (π-allyl)palladium complexes have
been investigated. Complexes of these ligands and of N,N′-diphenylpiperazine (7) and N,N′-
diphenyl-1,5-diazacyclooctane (8) with (1,3-η3-propenyl)palladium are described, in particular
their effects on the proton chemical shifts of the π-allyl ligand. Ligand dynamics of the
complexes is discussed. The structure of [(3,7-diphenyl-1,5-dimethylbispidinone)(1,3-η3-
propenyl)Pd]CF3SO3 (15) has been determined by X-ray crystallography. N,N′-Diphenyl-
bispidine derivatives show an unusually large steric interaction with the π-allyl ligand,
indicated by a tilt of the π-allyl plane toward the N-Pd-N plane by 122.8(8)°. Chemical
shift changes of the π-allyl protons due to the aromatic ring current are related to the
geometry of the complexes. The ligands are tested on the larger 2-methylene-6,6-
dimethylbicyclo[3.1.1]hept-2,3,10-η3-enyl ligand, demonstrating their potential as chemical
shift reagents.
In tr od u ction
Recently we have investigated the suitability of bpy-
type nitrogen ligands for structural analysis of pal-
ladium complexes with acyclic π-allyl ligands.6 Here,
interligand contacts are not sufficiently strong to restrict
the conformational flexibility of the π-allyl ligand,
making the method less powerful. It is therefore of
great interest to find new types of ligands that give
stronger interligand contacts.
Palladium-catalyzed reactions are among the most
important tools of current preparative organic chemis-
try,1 and various types of ligands have been developed
to increase their versatility and selectivity.2 More
recently, the importance of interligand steric interac-
tions for the stereoselectivity of these reactions have
been addressed,2e,3 in particular regarding enantiose-
lective palladium-catalyzed reactions. Such steric in-
teractions have also been utilized for structure analysis
of (π-allyl)palladium complexes. The method is based
on measurement of interligand NOEs in complexes with
chelating nitrogen ligands based on 2,2′-bipyridyl (bpy)
(1) and its congeners.4 For both synthetic work and
analytical applications it is important to identify the
dynamic processes in these complexes.5
However, increased steric interaction is not necessar-
ily obtained by increasing the ligand size. A larger bpy-
type ligand tends to minimize steric contacts by sliding
beneath the π-allyl ligand plane.7 We therefore need a
ligand which (i) is very rigid in order to prevent
conformational changes, (ii) has substituents which
occupy space both above and below the metal coordina-
tion plane, thus minimizing decrease of steric interac-
tion by distortion of the square planar metal coordina-
tion, and (iii) has a high binding constant for the metal,
thus preventing release of sterical strain by dissociation.
To be useful as an analytical tool, the ligand would also
need to have a high degree of symmetry to produce
reasonably simple 1H NMR spectra. As promising
† Dedicated to Prof. Hans-J . Scha¨fer on the occasion of his 60th
birthday.
X Abstract published in Advance ACS Abstracts, February 15, 1997.
(1) (a) Tsuji, J . Palladium Reagents and Catalysts: Innovations in
Organic Chemistry; Wiley: Chichester, U.K., 1995. (b) Collman, J . P.;
Hegedus, L. S.; Norton, J . R.; Finke, R. G. Principles and Applications
of Organotransition Metal Chemistry; University Science Books: Mill
Valley, CA, 1987.
(2) Examples for ligand types: (a) Trost, B. M.; Van Vranken, D.
L.; Bingel, C. J . Am. Chem. Soc. 1992, 114, 9327. (b) Trost, B. M.; van
Vranken, D. L. Chem. Rev. 1996, 96, 395. (c) Albinati, A.; Pregosin, P.
S.; Wick, K. Organometallics 1996, 15, 2419. (d) Trost, B. M.; Bunt, R.
C. Angew. Chem., Int. Ed. Engl. 1996, 35, 99. (e) Togni, A.; Burckhardt,
U.; Gramlich, V.; Pregosin, P. S.; Salzmann, R. J . Am. Chem. Soc. 1996,
118, 1031. (f) Frost, C. G.; Williams, J . M. J . Tetrahedron Lett. 1993,
34, 2015. (g) Sprinz, J .; Helmchen, G. Tetrahedron Lett. 1993, 34, 1769.
(h) Knu¨hl, G.; Sennhenn, P.; Helmchen, G. J . Chem. Soc., Chem.
Commun. 1995, 1845. (i) Leutenegger, U.; Umbricht, G,; Fahrni, C.;
von Matt, P.; Pfaltz, A. Tetrahedron 1992, 48, 2143. (j) van Asselt, R.;
Elsevier, C. J . Organometallics 1994, 13, 1972.
(4) (a) Albinati, A.; Ammann, C. J .; Pregosin, P. S.; Ru¨egger, H.
Organometallics 1990, 9, 1826. (b) Albinati, A.; Kunz, R. W.; Ammann,
C. W.; Pregosin, P. S. Organometallics 1991, 10, 1800. (c) Ba¨ckvall, J .
E.; Granberg, K. L.; Andersson, P. G.; Gatti, R.; Gogoll, A. J . Org.
Chem. 1993, 58, 5445. (d) Bookham, J . L.; McFarlane, W. J . Chem.
Soc., Chem. Commun. 1993, 1352.
(5) (a) Vrieze, K.; Volger, H. C.; van Leeuwen, P. W. N. M. Inorg.
Chim. Acta Rev. 1969, 109. (b) Vrieze, K. in Dynamic Nuclear Magnetic
Resonance Spectroscopy; J ackman, L. M., Cotton, F. A., Eds.; Academic
Press: New York, 1975. (c) Ref 2b.
(6) Gogoll, A.; Gomes, J .; Bergkvist, M.; Grennberg, H. Organome-
tallics 1995, 14, 1354.
(7) (a) Deeming, A. J .; Rothwell, I. P. J . Chem. Soc., Chem. Commun.
1979, 670. (b) Fanizzi, F. P.; Intini, F. P.; Maresca, L.; Natile, G.;
Lanfranchi, M.; Tiripicchio, A. J . Chem. Soc., Dalton Trans. 1991, 1007.
(3) (a) Sjo¨gren, M.; Hansson, S.; Norrby, P.-O.; A° kermark, B.
Organometallics 1992, 11, 3954. (b) Brown, J . M.; Hulmes, D. I.; Guiry,
P. J . Tetrahedron 1994, 50, 4493. (c) Pfaltz, A. Acta Chem. Scand. 1996,
50, 189.
S0276-7333(96)00952-1 CCC: $14.00 © 1997 American Chemical Society

1168 Organometallics, Vol. 16, No. 6, 1997
Gogoll et al.
Ch a r t 1
Sch em e 1
candidates emerged bispidine (bpd) derivatives (Chart
1). Related ligands have recently been used as strong
complexing agents for a variety of metal cations.8
In the present paper we report on the synthesis and
properties of a series of N,N′-diphenylbispidinone de-
rivatives and their interaction with (1,3-η3-propenyl)-
palladium complexes. The analytical potential of these
ligands is shown on the [2-methylene-6,6-dimethylbicyclo-
[3.1.1]hept-2,3,10-η3-enyl]Pd complex derived from
â-pinene.
Sch em e 2
Resu lts a n d Discu ssion .
Syn th eses. The most commonly used synthetic route
to bispidine derivatives involves Mannich condensation
between a ketone or a piperidone, an amine and
formaldehyde.9 N,N′-diphenylbispidinone derivatives
cannot be obtained directly from aniline in this way.10
However, 1,5-dicarbomethoxybispidinone derivative 2a
could be prepared by a modified procedure (Scheme 1),
representing the first direct synthesis of a bispidine
derivative from an aromatic amine. An alternative
although longer pathway via the N,N′-dibenzyl deriva-
tive 3b was found to be more suitable for the 1,5-
dimethylbispidinone derivative 2b (Scheme 1). The
synthesis of 3b by Mannich reaction from pentane-3-
one was improved to 18% as compared to previously
reported 5%.8c Two reference ligands, N,N′-diphe-
nylpiperazine (7) and N,N′-diphenyl-1,5-diazacyclooc-
tane (8), were prepared by phenylation of the corre-
sponding amines (Scheme 2). Recently introduced
methods for phenylation of monoamines11 all failed with
the diamines described here, probably due to interfer-
ence of the strongly chelating starting materials or small
amounts of products with the catalytic amounts of
metals. Complexes with (1,3-η3-propenyl)palladium
were prepared by known procedures.4,6
Upon debenzylation of 3a the diazaadamantane de-
rivative 6 was obtained, rendering this route impractical
for the synthesis of 2a . Obviously under the acidic
reaction conditions part of the starting material under-
goes a retro-Mannich reaction, generating free formal-
dehyde which traps the debenzylation product to pro-
duce 6.
Ch a r a cter iza tion of Liga n d s. Bispidine deriva-
tives are known to undergo conformational equilibria
in solution which are shifted far to the side of the chair,
chair conformer (CC, eq 1) even with large substituents
(8) (a) Hosken, G. D.; Hancock, R. D. J . Chem. Soc., Chem. Commun.
1994, 1363. (b) Black, D. St. C.; Horsham, M. A.; Rose, M. Tetrahedron
1995, 51, 4819. (c) Black, D. St. C.; Deacon, G. B.; Rose, M. Tetrahedron
1995, 51, 2055. (d) Hancock, R. D.; Pattrick, G.; Wade, P. W.; Hosken,
G. D. Pure Appl. Chem. 1993, 65, 473. (e) Black, D. St. C.; Craig, D.
C.; Horsham, M. A.; Rose, M. J . Chem. Soc., Chem. Commun. 1996,
2093.
(9) (a) J eyaraman, R.; Avila, S. Chem. Rev. 1981, 81, 149. (b) Zefirov,
N. S.; Palyulin, V. A. In Topics in Stereochemistry; Eliel, E. L., Wilen,
S. H., Eds.; Wiley: Chichester, 1991; Vol. 20, p 171.
(10) The only N,N′-diphenylbispidine derivative which could be
located in the literature was reported in an X-ray crystallography
investigation without any synthetic data: Levina, O. I.; Potekhin, K.
A.; Kurkutova, E. N.; Struchkov, Y. T.; Palyulin, V. A.; Zefirov, N. S.
Dokl. Akad. Nauk SSSR 1984, 277, 367.
(11) (a) Driver, M. S.; Hartwig, J . F. J . Am. Chem. Soc. 1996, 118,
7217. (b) Louie, J .; Hartwig, J . F. Tetrahedron Lett. 1995, 36, 3609. (c)
Wolfe, J . P.; Buchwald, S. L. J . Org. Chem. 1996, 61, 1133. (d) Guram,
A. S.; Rennels, R. A.; Buchwald, S. L. Angew. Chem., Int. Ed. Eng.
1995, 34, 1348.
on the nitrogen.12,13 For the present derivatives 2 and
3, the presence of CB conformers, which might be
indicated by the appearance of additional signals in the
(12) (a) Douglass, J . E.; Ratliff, T. B. J . Org. Chem. 1968, 33, 355.
(b) Zefirov, N. S. Russ. Chem. Rev. 1975, 44, 196.
(13) CB conformers might be present if the nitrogens have endo
substituents: (a) McCabe, P. H.; Milne, N. J .; Sim, G. A. J . Chem.
Soc., Chem. Commun. 1985, 625. (b) References 8b and 9b. (c) Quast,
H.; Mu¨ller, B.; Peters, E.-M.; Peters, K.; von Schnering, H. G. Chem.
Ber. 1982, 115, 3631. (d) J ackman, L. M.; Dunne, T. S.; Mu¨ller, B.;
Quast, H. Chem. Ber. 1982, 115, 2872. (e) Zefirov, N.; Palyulin, V. A.;
Chemodanova, S. V.; Samoshin, V. V. Mendeleew Commun. 1992, 162.
(f) the ∆Gq for the exchange CB h BC was 40.6 kJ /mol at -63 °C
(observed by 13C NMR): Takeuchi, Y.; Scheiber, P.; Takada, K. J .
Chem. Soc., Chem. Commun. 1980, 403. (g) ∆Gq ) 38.1 kJ /mol was
reported for a macrocyclic derivative, where the CB isomer had lower
energy than CC.8e

(π-Allyl)palladium Complexes
Organometallics, Vol. 16, No. 6, 1997 1169
Ta ble 1. p Ka Va lu es for Va r iou s Am in es a n d
Ta ble 2. Ch em ica l Sh ifts of th e π-Allyl P r oton s in
Va r iou s (1,3-η3-P r op en yl)p a lla d iu m Com p lexesa
Nitr ogen Ch ela tin g Liga n d sa
compd
pKa
9.2
Hsyn
∆δar
Hanti
∆δar
H-2
∆δar
triethylamine
complex ligand
δ
δ
δ
1
4.2
5.2b
4.4
4.5
5.3
10
13
14
18
15
16
17
19
20
b
1
4.11
4.29
3.04
3.56
5.46
6.01
pyridine
2a
2b
3a
3b
4
2a
5
2b
3a
3b
7
2.21 -1.67 2.98 -0.77 5.67 -0.54
2.11 -1.77 2.96 -0.79 5.52 -0.69
2.22 -1.65 3.07 -0.71 5.60 -0.61
3.88
3.87
3.98
7.7
7.0
3.75
3.78
6.21
6.21
N,N′-dimethyl-1,5-diphenylbispidinone
N,N′-dimethylbispidine
8.1c
11.8812a
0.11 3.05 -0.73 5.58 -0.63
8
2.28 -1.59 2.86 -0.92 5.57 -0.64
a
Dimethyl sulfoxide solution, c ) 0.02 M, 25 °C, titrated with
a
CDCl3 solutions, counterion CF3SO3-. Chemical shift differ-
b
0.1 M p-TosOH, referring to the first protonation step. Solvent
ences ∆δar refer to the the corresponding complex with 3b or with
3a (for 2a and 5). Chloro dimer 10.
H2O. c Solvent 2-methoxyethanol/H2O.18b
b
low-temperature 13C NMR spectra,13a,f,8e could not be
detected. This is favorable for complex formation.14
Compounds 2 and 3 have reasonably simple H NMR
spectra in the aliphatic chemical shift region, namely a
singlet for the methoxy or carbomethoxy protons and
the benzyl protons for 3 and a pair of doublets for the
methylene protons. The doublet at higher chemical
shift belongs to the equatorial protons in the 1,5-
dimethyl derivatives 3 but to the axial protons in the
1,5-dicarbomethoxy derivatives 2.15
Binding capabilities, i.e. donor properties of nitrogen
ligands may be correlated to their pKa, but few data are
available in the literature.16 Binding behavior may also
be discussed in view of σ-donor and π-acceptor proper-
ties of the ligands, but these parameters are not directly
available.17 In the present investigation, the pKa values
were determined for solutions in dimethyl sulfoxide18
by titration with p-toluenesulfonic acid and are sum-
marized in Table 1. The N,N′-dibenzyl derivative 3b
shows a pKa comparable to aliphatic amines, whereas
the corresponding N,N′-diphenyl derivative 2b has a
value close to heteroaromatic compounds. We notice
also that the keto function has a strong lowering effect
on the pKa, which changes by ca. 4 orders of magnitude
from N,N′-dimethylbispidine (pKa ) 11.88)12a to 3b (pKa
) 7.7) and 4 (pKa ) 7.0). Such an effect has been related
to interactions between the nitrogen lone pairs and the
keto function through σ bonds.18c,19 An approximately
equally large further decrease by 3.2 pKa units is
observed upon introduction of a phenyl substituent, i.e.
comparing 3b with 2b. A smaller decrease by only 0.9
units is observed for 3a and 2a , since the pKa of 3a is
already low. In a similar way, the carbomethoxy
substituent lowers the pKa, i.e. comparing 2a with 2b
and 3a with 3b. In conclusion, due to similar pKa
values, the N-phenyl-substituted bispidinone derivatives
are expected to have σ-donor properties comparable to
the bipyridyl type ligands but might bind weaker since
they are not π-accepors.
Com p lexa tion . All bispidinone ligands 2 and 3 form
stable complexes with (1,3-η3-propenyl)Pd. Complex
formation is indicated by interligand NOEs, which for
the N-phenylated ligands are observed between the
ortho and meta protons and the terminal π-allyl protons.
In addition, substantial chemical shift changes of the
terminal π-allyl protons are observed for complexes
involving ligands 2a ,b (Table 2). These result from
electronic effects and from the magnetic anisotropy of
the aromatic ring (vide infra).
1
X-r a y Cr ysta llogr a p h y. Because of the consider-
able difference between the N,N′-diphenylbispidinone
ligands and previously applied chelating ligands, it was
important to obtain accurate structural information for
their (π-allyl)palladium complexes. Thus, we have
investigated the structure of complex 15 by X-ray
crystallography (Tables 3 and 4). An ORTEP presenta-
tion of the structure is shown in Figure 1.20 The
geometry of the complex shows considerable deviation
from other (π-allyl)palladium complexes with chelating
nitrogen ligands. The main difference is the unusually
large inclination of the π-allyl plane relative to the
N-Pd-N plane, which is 122.8(8)°. For comparison,
this angle is 111.5° in the chloro dimer, i.e. bis[(1,3-η3-
propenyl)palladium chloride],21 and 109.4° in (2,2′-
bipyridyl)(2-methyl-1,3-η3-propenyl)palladium
tri-
(14) Stronger complexation can be expected if the conformations of
free and complexed ligand are similar (i.e., close to CC), which can be
achieved by preorganizing the ligand through rigid structural fea-
tures: Hancock, R. D.; Ngwenya, M. P.; Wade, P. W.; Boeyens, J . C.
A.; Dobson, S. M. Inorg. Chem. 1990, 29, 264; see also ref 8a,d.
(15) Gogoll, A.; Axe´n, A.; Grennberg, H. Magn. Reson. Chem. 1997,
35, 13. This order is reversed to δH,ax < δH,eq in the palladium
complexes.
(16) (a) For various chelating nitrogen ligands it was shown that
ligand preorganization affects ligand basicity, i.e. pKa, and the equi-
librium constant for binding of Cu(II) in the same way. Incorporation
of the rigid bispidine skeleton into the ligand structure resulted in a
further increase of both the pKa and the equilibrium constant for Cu-
(II) binding.8a (b) Togni, A.; Venanzi, L. M. Angew. Chem., Int. Ed.
Engl. 1994, 33, 497.
fluoromethanesulfonate.4b We find the trifluoromethane-
sulfonate counterion in the apical position above the
metal facing C-2 of the π-allyl ligand, which is less
obscured due to the tilt of the π-allyl group. Further-
more, the distances between the palladium and the
nitrogen atoms are larger (2.169(4) and 2.184(4) Å) than
in the (2,2′-bipyridyl)(2-methyl-1,3-η3-propenyl)palla-
dium complex (2.089(2) and 2.085(3) Å)4b and also
slightly larger than in a reported (spartein)(1,3-η3-
(17) Moberg, C.; Adolfsson, H.; Wa¨rnmark, K. Acta Chem. Scand.
1996, 50, 195.
(20) The structure determination was complicated by the presence
of an equimolar amount of solvent in the crystals. These solvent
molecules are oriented randomly at room temperature; i.e., they are
not complexed to the metal. An example of a palladium complex with
coordinated solvent as well as intercalated solvent molecules has been
reported: Hansson, S.; Norrby, P.-O.; Sjo¨gren, M. P. T.; A° kermark,
B.; Cucciolito, M. E.; Giordano, F.; Vitagliano, A. Organometallics 1993,
12, 4940.
(18) (a) Meurling, L. Chem. Scr. 1975, 7, 23. The pKa values in water
and dimethyl sulfoxide for various amines were different by ca. (1.5
units. (b) Settimj, G.; del Guidice, M. R.; D’Angelo, S.; di Simone, L.
Ann. Chim. (Rome) 1974, 64, 281. (c) Settimj, G.; del Guidice, M. R.;
D’Angelo, S.; di Simone, L. Gazz. Chim. Ital. 1979, 109, 345.
(19) Sasaki, T.; Eguchi, S.; Kiriyama, T.; Sakito, Y. J . Org. Chem.
1973, 38, 1648.
(21) Smith, A. E. Acta Crystallogr. 1965, 18, 331.

1170 Organometallics, Vol. 16, No. 6, 1997
Gogoll et al.
have been developed for this purpose,23 but practical
applications to small molecules are not common,24 with
the exception of cyclophanes25 and porphyrins.26 The
reason is probably that the conformational mobility of
the molecule under investigation must be low, a condi-
tion not often met in nonrigid small molecules. We
estimate the contribution of the phenyl ring current to
the chemical shift changes of the π-allyl protons, i.e. ∆δar
) δ - δref by using complex 17 as a reference (Table
2).27 Complex 16 is a more suitable reference for
complexes of ligands 2a and 5. The ligands 2a ,b and 8
give ∆δar,syn ≈ -1.6 to -1.8 ppm, ∆δar,anti ≈ -0.7 to -0.9
ppm, and ∆δar,H-2 ≈ -0.6 ppm. This is in fact in
excellent agreement with ∆δar calculated using the
J ohnson-Bovey model,28 based on the geometry deter-
mined for 15 by X-ray crystallography (Chart 2). It is
important to note that (i) corrections to the ring
current23c are not necessary and (ii) the coordinates for
the static crystallographic structure are applicable to
the solution structure although dynamic phenomena
exist (vide infra).
A larger distance between the phenyl rings of the
chelating ligand and the π-allyl ligand protons is sug-
gested by the small value ∆δarom,syn ) -0.20 measured
for 19, in agreement with its different geometry.29
Another explanation would be low complex stability.
Weak binding of 7, which was also indicated by easy
decomposition of the complex, could be due to the
“wrong” bite angle of this ligand,8d which is 39° as
opposed to 89.3° for the bispidine ligands.
F u r th er Effects. Ligand 8 as well as ligands 2a ,b
exhibited the unexpected tendency to form stable com-
plexes with the silver ions used to precipitate the
chloride of the starting (π-allyl)palladium chloride
dimers.30 The synthetically most readily accessible
bispidinone derivative 2a showed an unexpected reac-
tivity upon complexation, i.e. conversion of the carbonyl
group into the hydrate. Hydrate formation is revealed
by the disappearance of the keto carbonyl signal from
the 13C NMR spectrum and the appearance of a new
signal at δ 92.5, assigned to R2C(OH)2. In addition, a
NOE is detected between the OH protons and the axial
CH2 protons of the bispidine ligand. Finally, the
Ta ble 3. Cr ysta l Da ta a n d Str u ctu r e
Refin em en t for 15
empirical formula
fw
C24H29N2OPd+CF3SO3-‚CDCl3
701.93
temp
m
296(1) K
8.7 cm-1
radiation (λ, Å)
cryst system
space group
unit cell dimens
Mo KR (0.710 73)
monoclinic
P21
a ) 10.1746(3) Å, R ) 90°
b ) 17.5309(3) Å, â ) 95.0950(10)°
c ) 17.8717(4) Å, γ ) 90°
3175.18(13) Å3
4
V
Z
D(calcd)
abs coeff
1.4684(1) Mg/m3
0.87 mm-1
F(000)
1420
cryst size
θ range for data collcn
index ranges
0.4 × 0.12 × 0.08 mm
1.63-23.24°
-5 e h e 11, -19 e k e 19,
-19 e l e 19
11 843
reflcns collcd
indepdt reflcns
4478 [R(int) ) 0.0673]
abs corr
none
refinement method
data/restraints/param
goodness-of-fit on F2
final R indices (I > 2σI)
R indices (all data)
extinction coeff
full-matrix least-squares on F2
4478/0/332
1.079
R1 ) 0.0459, wR2 ) 0.1243
R1 ) 0.0526, wR2 ) 0.1292
0.0017(4)
largest diff peak and hole
0.362 and -0.641 e‚Å-3
cyclohexenyl)palladium complex (2.14(1) and 2.16(1)
Å).22 On the other hand, the distances between the
palladium atom and the π-allyl carbons are with 2.125-
(6) Å (terminal) to 2.128(5) Å (central) in the usual
range, i.e. slightly larger than in the mentioned bpy
complex and the chloro dimer but somewhat shorter
than in the spartein complex (2.14(2) Å and 2.09(2) Å).22
Finally, we note that the terminal π-allyl carbons are
positioned below the plane defined by Pd and the
nitrogen atoms (C-1′, -0.225(7) Å; C-3′, -0.233(7) Å;
Figure 2), and that the central carbon is above this plane
(C-2′, 0.319(9) Å). Again, this shows a somewhat
distorted geometry when compared to e.g. the (2,2′-
bipyridyl)(2-methyl-1,3-η3-propenyl)palladium complex,
where the corresponding values were ca. -0.2 and -0.1
Å for the terminal carbons and +0.5 Å for the central
carbon.4b
In conclusion, the arrangement of the 1,3-η3-propenyl
ligand in 15 is distorted as compared to other (π-allyl)-
palladium complexes. A likely source of this distortion
is the large steric interaction between the phenyl rings
and even such a small π-allyl ligand.
(23) (a) J ohnson, C. E.; Bovey, F. A. J . Chem. Phys. 1958, 29, 1012.
(b) Haigh, C. W.; Mallion, R. B. In Progress in NMR Spectroscopy;
Emsley, J . W., Feeney, J ., Sutcliffe, L. H., Eds.; Pergamon Press:
Oxford, U.K., 1980; Vol. 13, p 303. See also: (c) Agarwal, A.; Barnes,
J . A.; Fletcher, J . L.; McGlinchey, M. J . Can. J . Chem. 1977, 55, 2575.
(24) Applications for proteins are more common. For a comparison
of different models, see: Williamson, M. P.; Asakura, T. J . Magn.
Reson., Ser. B 1993, 101, 63.
(25) (a) Fukazawa, Y.; Usui, S.; Tanimoto, K.; Hirai, Y. J . Am. Chem.
Soc. 1994, 116, 8169. (b) Schneider, H.-J .; Ru¨diger, V.; Cuber, U. J .
Org. Chem. 1995, 60, 996. (c) Fukazawa, Y.; Ogata, K.; Usui, S. J .
Am. Chem. Soc. 1988, 110, 8692. (d) Estimation of the ring current in
various cyclic π-systems via the chemical shift changes imposed on
protons situated above or below these rings: De J ong, H. A. P.;
Wynberg, H. Tetrahedron 1965, 21, 515.
(26) (a) Cross, K. J .; Wright, P. E. J . Magn. Reson. 1985, 64, 220.
(b) Abraham, R. J .; Medforth, C. J . Magn. Reson. Chem. 1987, 25, 432.
(27) Assuming that ligand 3b in 17 causes similar electronic effects
but no ring current effects because of the long distance between its
phenyl rings and the π-allyl ligand.
(28) Calculations of ∆δar were done with QCPE program no. 536,
using the J ohnson-Bovey method.
Ch em ica l Sh ift Effects. In complexes between 1,3-
η3-propenyl and the N-phenylated ligands 2a ,b and 8
the chemical shift of the syn protons changes to lower
frequencies, as compared to chloro dimer 10. A much
smaller effect is observed for the N-benzylated ligands
3a ,b, and for the N-phenylated ligand 7. This observa-
tion can be rationalized by the aromatic ring current
effect, i.e. the chemical shift anisotropy of the phenyl
rings, which are located close to the π-allyl ligand in
complexes with 2a ,b and 8. Expectedly, the more
distant anti protons and the central proton H-2 are less
affected. It would be of practical value if this effect
could be quantified and then be utilized to map the
geometry of the π-allyl ligand. Several empirical models
(29) In 19, bidentate ligand 7 and metal form a five-membered ring
in contrast to the six-membered rings formed by the bispidine
derivatives and 8, which puts the phenyl rings further away from the
π-allyl ligand.
(30) No such complexes are formed with 7. Formation of silver
complexes can be avoided by first adding the silver salt to the chloro
dimer, waiting until all silver is precipitated as chloride, and then
adding the nitrogen ligand.
(22) Togni, A.; Rihs, G.; Ammann, C. Helv. Chim. Acta 1990, 73,
723.

(π-Allyl)palladium Complexes
Ta ble 4. Selected Geom etr y P a r a m eter s for Com p lex 15 a n d Rela ted Bisp id in e Der iva tives
15 N,N′-diphenylbispidine
MM+b
Organometallics, Vol. 16, No. 6, 1997 1171
X-ray crysta
PM3(TM)c
(L)CuCl2 X-ray crystd X-ray cryste
2.004, 1.965f
MM+b
Pd-N(7), Pd-N(3) (Å)
N(3)-N(7) (Å)
2.169(4), 2.184(4)
2.966
2.154
2.915
2.103
2.840
2.71
3.07
3.006
Pd-C(1′), Pd-C(3′) (Å)
Pd-C(2′) (Å)
2.126(6), 2.125(6)
2.128(5)
1.463, 1.471
1.487-1.506
85.93
92.1, 88.0
67.7
122.8(8)
2.240, 2.240 2.122
2.203
1.488
1.528
85.2
2.142
1.492
1.546-1.526
85.0
N-Cipso (Å)
N-CH2 (Å)
1.39
1.47
1.358
1.47
1.500
N-Pd-N (deg)
86.3 (N-Cu-N)
g
C(9)-N-Cipso-Cortho (deg)
60.9
68.7
81.2, 87.4
102.0
0.0
C(1′)-Pd-C(3′) (deg)
64.9
69.1
allyl vs N-Pd-Nh (deg)
129
120.8
104.0
121.8
+0.12
+0.92
dihedral between Ph planes (deg) 127.8
125.0i
104.0
C(1′), C(3′) vs N-Pd-N planej (Å) -0.225(7), -0.233(7) -0.77
C(2′) vs N-Pd-N planej (Å)
0.319(9)
-0.24
a
b
This investigation, X-ray crystallography. Hyperchem 4.0, MM+ parameter set37 with additional parameters for (π-allyl)Pd.38
c Spartan 4.1.1.39 L ) N,N′-dimethyl-1,5-diphenylbispidin-9-one.41 Ionic radii: Pd2+ ) 0.80 Å, Cu2+ ) 0.72 Å. e X-ray crystallography.10
d
f Cu-N. Cortho oriented toward the π-allyl base. Angle between plane of π-allyl system and plane containing both nitrogen atoms and
g
h
i
j
Pd. Angle between N-CH3 bonds. Negative value ) below plane.
F igu r e 3. Exchange pattern of phenyl protons as a result
of phenyl rotation and ligand rotation.
Ch a r t 2
F igu r e 1. Molecular structure (ORTEP view) of the cation
of 15, showing 50% probability thermal ellipsoids.
parent ligand rotation (Figure 3).5b,33 As a result,
signals of bispidinone protons situated above or below
the metal coordination plane are averaged. Free ener-
gies of activation ∆Gq for this and additional processes
(vide infra) were obtained from signal coalescence (Table
5) and correlated nicely with the pKa values, i.e., the
base strength of the ligands. Of the bispidinone deriva-
tives 2 and 3, the one with higher pKa also has the
F igu r e 2. Overlay of structures of 15 from X-ray crystal-
lography (bold lines) and geometry optimized with PM3-
(TM) (narrow lines). One set of phenyl rings has been
removed from (a) for better visibility.
required change of its elemental composition could be
confirmed when this conversion had gone to completion
after a few days.31 Unfortunately, the carbonyl group
of the free ligand could not be protected,32 although
reduction with sodium borohydride yielded the corre-
sponding alcohols.15 Thus, 3,7-diphenyl-1,5-dimethyl-
bispidinone (2b) emerged as the candidate of choice for
an analytically applicable ligand.
(32) The reason was the low reactivity of 1,5-dicarboalkoxybispidi-
none derivatives, which has already been reported by Mannich upon
their first preparation: Mannich, C.; Veit, F. Ber. Dtsch. Chem. Ges.
1935, 68, 506. Commonly applied derivatization reactions such as
ketalization, reduction, and reaction with hydrazine and its derivatives
all failed. It has been suggested that the keto group in bispidinones
should behave more like an amide carbonyl group due to interaction
with the nitrogen lone pair.19 For transformations of related com-
pounds, see: (a) Ferris, J . P.; Miller, N. C. J . Am. Chem. Soc. 1963,
85, 1325. (b) House, H. O.; Mu¨ller, H. C. J . Org. Chem. 1962, 27, 4436.
(33) (a) Albinati, A.; Kunz, R. W.; Ammann, C. J .; Pregosin, P. S.
Organometallics 1991, 10, 1800. (b) Gogoll, A.; O¨ rnebro, J .; Grennberg,
H.; Ba¨ckvall, J . E. J . Am. Chem. Soc. 1994, 116, 3631. (c) Elguero, J .;
Fruchier, A.; de la Hoz, A.; J alo´n, F. A.; Manzano, B. R.; Otero, A.;
Go´mez-de la Torre, F. Chem. Ber. 1996, 129, 589. (d) For a summary,
see ref 2b.
Dyn a m ic Beh a vior . All complexes showed the
expected dynamic behavior, namely the so-called ap-
(31) The free ligand behaves similarly in the presence of acid, e.g.
addition of p-toluenesulfonic acid hydrate to a dimethyl sulfoxide
solution of 2a . For related observations, see: (a) Gonikberg, E. M.; le
Noble, W. J . J . Org. Chem. 1995, 60, 7751. (b) Halm, J . M.; le Noble,
W. J . J . Am. Chem. Soc. 1992, 114, 1916.

1172 Organometallics, Vol. 16, No. 6, 1997
Gogoll et al.
Ta ble 5. F r ee En er gies of Activa tion ∆Gq for
and meta proton signals being split into two separate
multiplets. Typical free energies of activation for the
rotation of an unsubstituted phenyl ring in aromatic
amines are around 46 kJ /mol, higher upon protona-
tion.34 In the present complexes, the energies are
virtually the same as those for the apparent ligand
rotation of the respective complex.35 Molecular models
of the complexes show that a phenyl rotation is sterically
impossible as long as both nitrogen atoms of the
chelating ligand are bound to the metal. This leaves
two explanations: Either the same process, i.e. dissocia-
tion of one nitrogen atom from the metal, facilitates both
processes or, as an alternative, the phenyl rings might
not rotate at all, and the positions of two ortho protons
relative to the π-allyl ligand are interchanged by the
apparent ligand rotation. Because of the π-allyl ligand
symmetry, these two possibilities cannot be distin-
guished (Figure 3). Such a distinction would be possible
with a π-allyl ligand without a plane of symmetry
through its center (vide infra).
The complexes with 7 and 8 showed a further dy-
namic process with higher free energy of activation
(7, 53 kJ /mol; 8, 60 kJ /mol). This process has to be
complete dissociation of the chelating ligand, since it
exchanges protons on different faces of the piperazine
or diazacyclooctane rings. As an example, the 1H NMR
spectrum of 20 at 25 °C shows two signals for the
N-CH2 protons (δ 4.02 and 3.48). At 50 °C, only one
signal is observed, as is in the spectrum of the free
ligand. In complex 19, the rates associated with the
dynamic processes are ca. 710 s-1 (apparent ligand
rotation, Tc ) -50 °C) and 650 s-1 (ligand dissociation,
Tc ) +5 °C); i.e., apparent ligand rotation does not
require complete dissociation. Complex 20 shows broad
signals for ligand 8 at all accessible temperatures. This
is due to the high conformational flexibility of 8, which
is lacking the bridge of the bispidine skeleton, resulting
in additional equilibria between conformers with the C3-
chains of the 1,5-diazacyclooctane ring folded either
toward or away from the metal (eq 2).36
Dyn a m ic P r ocesses in Com p lexes
a
[(L)(π-a llyl)P d ]CF 3SO3
apparent
ligand rotation
phenyl rotationb
∆Gq
∆Gq
complex ligand L (kJ /mol) Tc (K)c (kJ /mol)
Tc (K)c
13d
14
18
15
16
17
19
20
21
22
1
2a
5
2b
3a
3b
7e
8f
63.8
40
42
42.9
52.7
57.5
41.9
54
323
188
188
223
298
296
223
273
263
42.5
216
53g
278
54, 60g
56
293, 313g
263
2b
3b
57
72h
a
All measurements were performed for acetone-d6 solutions
except for the 1,5-dicarboxymethyl derivatives 2a and 3a (1:1
mixture of CBr2F2/CH2Cl2). ∆Gq values obtained from coalescence
of the equatorial methylene proton signals. For 7, the ring
methylene signals were used. b Coalescence of ortho proton signals.
d
c Coalescence temperature. Complex with (2S*,3R*-2-methoxy-
3,4,5-η3-hexenyl)Pd, coalescence of bipyridyl protons H-6 and H-6′.6
e Solvent CD2Cl2. f Solvent CDCl3. g Complete ligand dissociation.
h
From line shape analysis at 313 K.
higher ∆Gq value for its complex. This fits well with
our recent observation that the apparent ligand rotation
involves breaking of a metal-nitrogen bond.33b The
complexes of carbomethoxy derivatives 2a and 3a have
lower ∆Gq values than those of their methyl analogs 2b
and 3b. A likely reason is a reduced nitrogen electron
density for the â-keto dicarboxylic acid esters 2a and
3a , with a resulting weaker metal-nitrogen bond. The
difference is less pronounced for the N-phenyl deriva-
tives 2a ,b, since here the electron density is low already.
The complex with ligand 5, i.e. the hydrate of 2a , has a
higher ∆Gq than the one with 2a itself, slightly lower
than the value measured for 2b. This seems reasonable
since the number of carbonyl groups decreases in the
sequence 2a > 5 > 2b. For the complex with 8, a ∆Gq
value in the region of the N-benzyl derivatives 3 is not
surprising, since this ligand is lacking the electron-
withdrawing carbonyl substituents but has the N-
phenyl substituents instead. The ligands 7 and bpy do
not fit into this picture because their complexes have a
different geometry (7, vide infra) or different nitrogen
hybridization (bpy).
For the complexes with 2b and 3b, addition of excess
chelating ligand (ca. 0.2 equiv), chloride ions, or a trace
of hydrochloric acid did not significantly change the
observed ∆Gq values. However, exchange between free
and complexed chelating ligand was detected for 2b by
NOESY and saturation transfer experiments but not for
3b. This is in accordance with the higher pKa of 3b as
compared to 2b (Table 1); i.e., the nitrogen-metal bond
is stronger for 3b. Addition of chloride ions (0.5-2 equiv
of LiCl) induces dissociation of the chelating nitrogen
ligand. Addition of an excess of hydrochloric acid
triggers decomposition of the complexes. For the bpy
complex 12, addition of free ligand does decrease the
∆Gq value, in accordance with previous observations.33b
The observed dynamic behavior of the complexes thus
supports a mechanism of apparent ligand rotation via
dissociation of one side of the chelating ligand.
Calcu lated Geom etr ies. As a complementary source
of information, the complex geometries were investi-
gated by molecular modeling37,38 and by the semiem-
pirical method PM3(TM).39 Such a comparison of
(34) Sternhell, S. in Dynamic NMR Spectroscopy; J ackman, L. M.,
Cotton, F. A., Eds.; Academic Press: San Diego,CA, 1975; p 163.
(35) For the complex of 7, ∆Gq for phenyl rotation could not be
determined due to signal overlap.
(36) Musker, W. K. Coord. Chem. Rev. 1992, 117, 133.
(37) Using the MM+ force field in Hyperchem 4.0 (Hypercube Inc.,
Waterloo, Canada) with some additional parameters for (π-allyl)Pd
from ref 38.
(38) Norrby, P.-O.; Åkermark, B.; Hæffner, F.; Hansson, S.; Blomberg,
M. J . Am. Chem. Soc. 1994, 115, 4859.
(39) Using PM3(TM) as present in Spartan 4.1.1 (Wavefunction Inc.,
Irvine, CA).
A second dynamic process which can be observed in
complexes of N-phenylated ligands 2a ,b, 7, and 8 is
restricted phenyl rotation (Figure 3), resulting in ortho

(π-Allyl)palladium Complexes
Organometallics, Vol. 16, No. 6, 1997 1173
Ch a r t 3
observed and calculated geometries is also of a more
general interest. More recently calculated geometries
have been exploited to obtain detailed structural infor-
mation on palladium complexes when crystallographic
data were not available.40 Such information has then
been used to explain the chemical behavior of these
complexes. Key structural parameters thus obtained
for 15 are compared with the corresponding crystal-
lographic data in Table 4 and in Figure 2. Both
calculated structures show significant deviations from
the crystal structure, with the semiempirical method
performing better. In particular, the π-allyl ligand is
situated too far away from the metal, and its position
with respect to the N-Pd-N plane is too high (Figure
2a). Furthermore, the tilt of the phenyl rings against
a plane through the nitrogen atoms and C-9 of the
bispidine skeleton is smaller (68.7° compared to ca. 90°).
Another striking difference is the tilt of the π-allyl
ligand with respect to the N-Pd-N plane, which is 104°
(PM3) as compared to 122.8(8)° in the crystal. Further-
more, the angles of the phenyl rings toward each other
are smaller than in the crystal (121.8° for PM3, 127.8°
from the crystallographic structure). In conclusion, the
steric interactions between the two organic ligands on
the palladium are stronger in reality than the theoreti-
cal calculations suggest, illustrating the potential dan-
ger in conclusions based on such data for new types of
ligands. The observed deviations reemphasize also the
difference between this new type of ligand and previ-
ously used ligands, including e.g. sparteine,22 which has
the same backbone as the bispidine skeleton but lacks
the N-phenyl substituents. In fact, calculated geom-
etries for complexes of (1,3-η3-propenyl)palladium with
the N-benzyl ligands (i.e. 3) have the propenyl ligand
in a position closer to the one found in the crystal
structure of 15. Because of the structural deviations,
calculated ring current effects on the π-allyl protons
diverge from the observed values more than those based
on the crystallographic structure (Chart 2).42 As is
apparent from Figure 2, the larger distance of the π-allyl
ligand from the phenyl rings is in part compensated by
their tilt toward this ligand. Therefore, the calculated
values are sufficiently good to probe the stereochemistry
of the π-allyl ligand.43
Ta ble 6. Ch em ica l Sh ifts of th e π-Allyl Liga n d
P r oton s in (6-Dim eth yl-2-m eth ylen ebicyclo[3.1.1]-
1,3-η3-h ep ten yl)p a lla d iu m Com p lexesa
∆δar,
obsd
∆δar,
∆δar,
proton
11
21
22
calcdb,c
calcdb,d
1
3
2.32
4.08
1.79
1.79
2.08
2.38
2.64
3.68
2.96
1.35
0.97
1.95
3.67
-1.15
0.63
1.66
1.32
2.58
1.86
3.54
1.17
0.66
2.45
4.48
1.68
1.77
2.21
1.78
2.85
3.50
4.07
1.39
1.02
-0.50
-0.81
-2.83
-1.14
-0.55
-0.46
-0.27
-1.64
-0.53
-0.22
-0.36
-0.52
-0.82
-1.15
-1.20
-0.26
0.08
-0.24
-0.46
-3.47
-0.87
-0.41
-0.09
-0.09
-1.78
-0.89
-0.23
-0.24
4R
4â
5
7R
7â
10a
10b
Me-8
Me-9
0.05
-1.95
-0.44
-0.02
-0.09
a
b
CDCl3 solution, 25 °C. J ohnson-Bovey model.28 c Structure
derived using modified MM+ parameter set.37 Structure derived
d
from PM3(TM) parameter set in Spartan 4.1.1.39
We therefore prepared the complexes 21 and 22, derived
from â-pinene (Chart 3). Both bispidine derivatives 2b
and 3b formed stable complexes, and as expected, large
chemical shift changes were observed for those protons
in complex 21 which were close to the phenyl rings
(Table 6). In particular, the proton signals are dispersed
over a larger chemical shift range than in both complex
22 and the chloro dimer 11 (4.8 ppm as compared to
3.5 and 3.1 ppm, respectively). The signals of H-4R and
H-4â, which overlap in 11, are well separated in 21, and
coupling constants can be measured for several protons
because of the simpler multiplet patterns (Figure 4). The
methylene protons on C-7, C-4, and C-10 can easily be
distinguished due to the very different extent of their
chemical shift changes. To obtain a more quantitative
picture, we again relate the chemical shift changes to
the complex geometry, using complex 22 as a reference.
From the geometries obtained with PM3(TM),44 chemi-
cal shift changes due to the aromatic ring currents were
calculated as for 15 (Table 6). Observed and calculated
changes agree reasonably well, with the exception of
those protons located very close to the aromatic rings.
For the free ligands, there is also a difference between
the lowest energy conformer according to molecular
modeling37 and the conformer found in a previously
reported crystal structure.10 Molecular modeling has
the phenyl rings aligned in a plane with the two
nitrogen atoms and C-9 (dihedral angle C(9)-N-Cipso
-
Cortho ≈ 90°). In the crystal, they are aligned almost
perpendicular to this plane (Table 4).
F u r th er Com p lexes. With the characteristic fea-
tures of ligand 2b and its congeners established, it was
important to prove its complexation to palladium com-
plexes with sterically more demanding π-allyl ligands.
(40) (a) Pen˜a-Cabrera, E.; Norrby, P.-O.; Sjo¨gren, M.; Vitagliano,
A.; De Felice, V.; Oslob, J .; Ishii, S.; O’Neill, D.; A° kermark, B.; Helquist,
P. J . Am. Chem. Soc. 1996, 118, 4299. (b) Andersson, P. G.; Harden,
A.; Tanner, D.; Norrby, P.-O. Chem. Eur. J . 1995, 1, 12.
(41) Levina, O. I.; Potekhin, K. A.; Kurkutova, E. N.; Struchkov, Y.
T.; Zefirova, O. N.; Palyulin, V. A.; Zefirov, N. S. Dokl. Akad. Nauk
SSSR 1986, 289, 876.
(42) Some deviations are also to be expected because the protons
were placed at standard positions in the crystallographic structure,
i.e., assuming idealized geometry.
The proton with the largest ring current effect ∆δar
)
-2.83 (at 25°C), H-4R, is embedded in a cavity on top
of one phenyl ring (Figure 5). Ring current effects for
most protons are slightly larger at -55 °C (see Experi-
(44) To obtain a better geometry, the π-allyl carbons were fixed
during geometry optimization at positions corresponding to those found
in the crystals of 15.
(43) Deviations between observed and measured ring current effect
of ca. 0.1-0.2 ppm are common.23,25

1174 Organometallics, Vol. 16, No. 6, 1997
Gogoll et al.
F igu r e 4. 1H NMR spectra of (a) (2-methylene-6,6-dimethylbicyclo[3.1.1]hept-2,3,10-η3-enyl)palladium chloride dimer (11)
and (b) the corresponding complex 21 with ligand 2b (400 MHz, CDCl3 solution, 25 °C).
F igu r e 6. Phenyl ring positions relative to 1,3-η3-propenyl
ligands in complexes (a) 19, (b) 15, and (c) complex with
bis(arylimino)acenafthen (BIAN). Geometries shown are
from X-ray crystallography for 15 and from PM3(TM) for
15 and the bis(arylimino)acenafthen complex.
F igu r e 5. Space-filling presentation (right) of complex 21,
showing the close interaction between the phenyl ring and
proton H-4R (geometry optimized with PM3(TM)). Left:
Stick structure.
rotation as well as apparent ligand rotation do occur
simultaneously. The NOESY and ROESY experiments
run to detect this exchange were done with sufficiently
short mixing times as to avoid complete scrambling of
the information due to stepwise, indirect exchange such
as o-Ph-3 f o-Ph-7′ f o-Ph-7; i.e., a cross-peak between
o-Ph-3 and o-Ph-7 was not observed. It is reasonable
to assume that the 1,3-η3-propenyl ligand has the same
dynamic behavior. In those cases where ∆Gq could be
determined for both dynamic processes, the values are
the same within experimental error (Table 5). One may
therefore assume that they depend upon a similar rate-
determining step. This could be the breaking of one
Pd-N bond, in accordance with previous observations.33b
Con clu sion s. N,N′-Diphenylbispidinone derivatives
form stable complexes with even bulky (π-allyl)palla-
dium. Close steric interactions between the phenyl
rings and the π-allyl ligands on the palladium result in
large anisotropic chemical shift effects. This results
from two structural features: (i) the double chair
conformation of the N,N′-diphenyldiazacyclooctane mo-
tif, which forms a six-membered ring upon complex-
ation, i.e. counting the metal, the two nitrogen atoms,
and the C3 chain linking them; (ii) the one-carbon bridge
between C-1 and C-5 adding further rigidity to the
complex, which now resembles an adamantane-like
substructure. In contrast, N,N′-diphenylpiperazine (7)
is a less efficient ligand, because it forms a five-
membered ring upon complexation. This causes the
phenyl rings to bend away from the π-allyl ligand
(Figure 6).45 At first glance, these ligands might remind
one of the bis(arylimino)acenafthen (BIAN) ligands
mental Section), indicating the impact of ligand and
phenyl ring rotation at higer temperatures.
In comparison of 21 with 22, ∆Gq of apparent ligand
rotation again was higher for the complex with the
N-benzylated ligand (3b, complex 22, ∆Gq ) 72 kJ /mol)
than for the one with the N-phenylated ligand (2b,
complex 21, ∆Gq ) 57 kJ /mol). Interestingly, the
difference of ∆Gq between the two complexes, 15 kJ /mol,
is approximately the same as the one found between
17 and 15, 14.6 kJ /mol. This supports the previous
assumption that ∆Gq is related to ligand basicity, i.e.
pKa, but clearly steric factors are involved as well.
Because of the difference in ∆Gq, at room temperature
all methylene protons of 22 are nonisochronous, whereas
only two averaged signals of 4 protons each for the
equatorial and the axial bispidinone protons are ob-
served for 21. These signals are separated into indi-
vidual signals at -55 °C.
The lack of symmetry of the π-allyl ligand in 21 makes
it possible to determine whether phenyl rotation is
involved in the exchange of the phenyl protons or if they
exchange only via ligand rotation. If the π-allyl ligand
has the symmetry of 1,3-η3-propenyl (Cs, Figure 3, R )
R′), the exchange pattern of e.g. the ortho protons
resulting from phenyl rotation would be indistinguish-
able of the one observed if only ligand rotation took place
(vide supra), because protons in locations B and C have
the same chemical shift. This is not so for complex 21,
where all four ortho protons have different chemical
shifts. Hence, the observed direct exchange of e.g. o-Ph-3
with o-Ph-3′ as well as o-Ph-7′ proves that phenyl
(45) The angle between the two phenyl ring planes is ca. 127.8° for
15 and 155° for 19.

(π-Allyl)palladium Complexes
Organometallics, Vol. 16, No. 6, 1997 1175
were performed with the programs SHELXS-8656a and
SHELXL-9356b on F2. During the solution, the presence of
CDCl3 molecules became evident, the position of which could
not be refined due to its disordered orientation. Therefore,
the procedure SQUEEZE57 was applied. The positions of the
hydrogen atoms were calculated by assuming idealized geom-
etry (C-H ) 0.95 Å), methyl groups were located on circular
Fourier maps using local 3-fold symmetry, and then all
hydrogen atom positions were riding on the respective pivot
atoms. The final R indices are as follows: on observed data I
> 2σ(I), R1 ) 0.0459, wR2 ) 0.1243; on all data, R1 ) 0.0526,
wR2 ) 0.1292. Maximum and minimum peaks in the final
difference map were 0.362 and -0.641 e‚Å-3, respectively.
3,7-Diben zyl-1,5-d ica r bom eth oxybisp id in on e (3a ). To
methanol (400 mL) cooled on ice, were added dimethyl ac-
etonedicarboxylate (5.85 mL, 40 mmol), formaldehyde (14.8
mL 30% solution in water, 164 mmol), and benzylamine (9.0
mL, 82 mmol). The solution was stirred at 0 °C during 15 h.
Thereafter, the temperature was allowed to rise to room
temperature and the stirring continued for an additional 10
h. The solvent was evaporated under reduced pressure to
approximately half its volume, the precipitate was filtered off,
and the residue was washed with two portions (25 mL) of cold
methanol. The product was dried in vacuo overnight: White
powder (7.5 g, 17.2 mmol, 43%), mp 117 °C; 1H NMR (300
MHz, CDCl3) δ 7.31 (m, 10H, aromatic), 3.96 (s, 4H, benzylic),
3.86 (s, 6H, CH3), 3.28 (d, J ) 11.1 Hz, 4H, Hax), 3.15 (d, J )
11.1 Hz, 4H, Heq); 13C NMR (75.4 MHz, CDCl3) δ 203.4 (C-9),
170.2 (COO), 137.1 (Cipso), 128.7 (Cortho), 128.1 (Cmeta), 127.1
(Cpara), 60.7 PhCH2), 59.7 (CH2), 58.9 (C), 52.1 (CH3); IR (KBr)
1743, 1716 cm-1; MS m/e 436 (M+, 1.3%), 313, 218, 167, 91.
Anal. Calcd for C25H28N2O5: C, 68.79; H, 6.47; N, 6.42.
Found: C, 68.64; H, 6.41; N, 6.46.
introduced by Elsevier et al.,2j,46 but these have the
phenyl rings farther apart and at a much larger angle.47
Ligand 2b is therefore a promising new analytical tool
for the NMR spectroscopic characterization of (π-allyl)-
palladium complexes and as a chemical shift reagent
for ligands with complicated proton NMR spectra.
Exp er im en ta l Section
Gen er a l Exp er im en ta l Deta ils. NMR spectra were re-
corded for 1H at 300 and 400 MHz and for 13C at 75.4 and
100.6 MHz, at 25 °C unless stated otherwise. Chemical shifts
are indirectly referenced to TMS via the solvent signal (CDCl3,
7.26 and 77.00; acetone-d6, 2.04 and 206.0; CD2Cl2, 5.32 and
53.80). Mass spectra were recorded on a Finnigan MAT
INCOS 50 instrument at 70 eV (EI mode). IR spectra were
recorded on a Perkin-Elmer 1600 FTIR. Elemental analyses
were performed by Analytische Laboratorien, Engelskirchen,
Germany. The X-ray crystallographic analysis was done by
Dr. V. Langer, Dept. of Inorganic Chemistry, Chalmers
University of Technology and University of Go¨teborg, Go¨te-
borg, Sweden. Progress of the reactions was followed by TLC
on Merck precoated silica gel 60-F254 plates. For column
chromatography Merck Kieselgel 60 (230-400 mesh) was used.
All melting points are uncorrected. Commercially available
chemicals were used as supplied.
The pKa values were determined by titration of a solution
of the compound in dimethyl sulfoxide (dried with molecular
sieves and distilled, c ) 0.02-0.01 M, 25 °C) with a solution
of p-toluenesulfonic acid in the same solvent (c ) 0.1 M).18a
A
glass electrode (Schott pH-meter CG825) was used. The
reported values are estimated to be accurate within (0.2 units.
NMR signals were assigned from PECOSY,48 HSQC,49 HSBC,50
NOESY,51 ROESY,52 and NOE difference spectra.53 For
NOESY and ROESY experiments, mixing times between 0.5
and 1.2 s were used. Free energies of activation ∆Gq were
estimated from the coalescence behavior of the proton NMR
signals54 or by line shape analysis.55 Their accuracy is
estimated as (1 kJ /mol ((2 kJ /mol for 22).
Cr ysta llogr a p h ic An a lysis. Crystals of 15 were obtained
from a CDCl3 solution and are of needle shape. The largest
available crystal was selected for data collection and mounted
on a glass fiber. Measurements were made on a CCD-detector-
based SMART diffractometer (Siemens) using Mo KR radiation
(λ ) 0.710 73 Å) (sealed tube at 50 kV and 45 mA). Cell
parameters were obtained by least-squares refinement on 8188
reflections with I/σ(I) > 10. Crystal data, data collection
parameters, and results are listed in Tables 3 and 4. The
structure solution and full-matrix least-squares refinement
Meth ylen ea n ilin e (Tr im er ic) (9). To ethanol (200 mL)
were added aniline (9.0 mL, 0.1 mol) and formaldehyde (18.0
mL 30% solution in water, 0.2 mol). The solution was stirred
during 20 h at room temperature. Thereafter water (100 mL)
was added and the precipitate was immediately filtered off
and discarded. The filtrate was left until a new precipitate
formed. This precipitate was collected, and the solution was
left until a new precipitate was formed. The procedure was
repeated until no new precipitate was formed. The combined
solids were dried in vacuo over night: Colorless solid (3.0 g,
1
9.52 mmol, 29%), mp ) 136 °C (lit.58 mp 139.5 °C); H NMR
(300 MHz, CDCl3) δ 7.23 (m, 6H, m-Ph), 7.02 (m, 6H, o-Ph),
6.87 (m, 3H, p-Ph), 4.89 (s, 6H, CH2); 13C NMR (75.4 MHz,
CDCl3) δ 148.2, 129.1, 120.9, 117.7, 68.6; IR (KBr) 1596, 1498
cm-1; MS m/e 315 (M+, 0.8%), 105, 77.
3,7-Dip h en yl-1,5-d ica r bom eth oxybisp id in on e (2a ). To
methanol (100 mL) were added dimethyl acetonedicarboxylate
(0.88 mL, 6 mmol), formaldehyde (6.0 mL 30% solution in
water, 13 mmol), and methyleneaniline (trimeric) (1.4 g, 4.4
mmol). The solution was stirred for 20 h at room temperature.
The solvent was removed under reduced pressure. The
remaining oil was dissolved in chloroform (3 mL) and purified
by column chromatography (1:1 diethyl ether/n-pentane). The
first eluted fractions, containing mainly a mixture of the
desired product and methyleneaniline, were combined, and the
solvent was removed. The remaining solid was dissolved in
methanol (20 mL). When the sample was cooled to -20 °C,
precipitation of methyleneaniline and the desired product was
initiated. The solid was collected and dissolved in methanol/
chloroform (9:1, 10 mL). When the sample was cooled to -20
°C, the desired product crystallized as colorless needles which
(46) van Asselt, R.; Elsevier, C. J .; Smeets, W. J . J .; Spek, A. L.
Inorg. Chem. 1994, 33, 1521.
(47) This angle is determined from the geometry obtained with PM3-
(TM) as 170.8° in the (1,3-η3-propenyl)Pd complex; X-ray crystal-
lography of a corresponding Pd(0)-olefin complex46 found ca. 190°. In
the PM3(TM)-optimized structures of (L)[(1,3-η3-propenyl)palladium
complexes, the distance between the aromatic ring centers is ca. 8.26
Å for L ) BIAN, as compared to 8.06 Å for L ) 2b and 8.32 Å for L )
7.
(48) Mueller, L. J . Magn. Reson. 1987, 72, 191.
(49) (a) Bodenhausen, G.; Ruben, D. J . Chem. Phys. Lett. 1980, 69,
185. (b) Torres, A. M.; Nakashima, T. T.; McClung, R. E. D. J . Magn.
Reson. 1993, A102, 219.
(50) Ko¨ve´r, K. E.; Prakash, O.; Hruby, V. J . Magn. Reson. Chem.
1993, 31, 231.
(51) (a) States, D. J .; Haberkorn, R. A.; Ruben, D. J . J . Magn. Reson.
1983, 48, 286. (b) Wider, G.; Macura, S.; Kumar, A.; Ernst, R. R.;
Wu¨thrich, K. J . Magn. Reson. 1984, 56, 207.
(52) Bothner-By, A. A.; Stephens, R. L.; Lee, J . M.; Warren, C. D.;
J eanloz, R. W. J . Am. Chem. Soc. 1984, 106, 811.
(56) (a) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467. (b)
Sheldrick, G. M. J . Appl. Crystallogr., Manuscript in preparation.
(57) SQEEZE57b is part of the the software PLATON57a which was
used for geometry calculation: (a) Spek, A. L. Acta Crystallogr. 1990,
A46, C34. (b) Spek, A. L. Abstracts of Papers; American Crystal-
lographic Association, Atlanta Meeting; American Crystallographic
Association: New York, 1994; Paper M05.
(53) Sanders, J . K. M.; Mersh, J . D. In Progress in Nuclear Magnetic
Resonance Spectroscopy; Emsley, J . W., Feeney, J ., Sutcliffe, L. H.,
Eds.; Pergamon Press: Oxford, U.K., 1982; Vol. 15, p 353.
(54) Anderson, J . E.; Lehn, J . M. J . Am. Chem. Soc. 1967, 89, 81.
(55) Gu¨nther, H. NMR Spectroscopy, 2nd ed.; Wiley: Chichester,
U.K., 1995; p 335. Spectra were simulated with: gNMR, v. 3.6.5;
Cherwell Scientific Publ.: Oxford, U.K., 1996.
(58) Miller, J . G.; Wagner, E. C. J . Am. Chem. Soc. 1932, 54, 3698.

1176 Organometallics, Vol. 16, No. 6, 1997
Gogoll et al.
were collected and recrystallized under the same conditions.
from diethyl ether, yielding colorless crystals which were
collected and dried in vacuo (1.0 g, 6.0 mmol, 69%): mp 61-
62 °C; 1H NMR (400 MHz, CDCl3) δ 3.36 (d, J ) 12.2 Hz, 4H,
The crystals were dried in vacuo (285 mg, 0.7 mmol, 12%):
1
mp 96 °C; H NMR (300 MHz, CDCl3) δ 7.27 (m, 4H, m-Ph),
7.01 (m, 4H, o-Ph), 6.94 (m, 2H, p-Ph), 4.13 (d, J ) 12.0 Hz,
H
eq), 2.95 (d, J ) 12.2 Hz, 4H, Hax), 2.67 (s, 2H, NH), 0.88 (s,
4H, Hax), 3.91 (d, J ) 12.0 Hz, 4H, Heq), 3.73 (s, 6H, CH3); 13
C
6H, CH3); 13C NMR (100.6 MHz, CDCl3) δ 215.4 (C-9), 61.8
NMR (75.4 MHz, CDCl3) δ 201.7 (C-9), 169.6 (COO), 149.6
(Cipso), 129.0 (Cmeta), 120.9 (Cpara), 117.4 (Cortho), 59.0 (C), 57.8
(CH2), 52.6 (CH3); IR (KBr) 1747, 1733 cm-1; MS m/e 408 (M+,
21.8%), 288, 120, 105, 77. Anal. Calcd for C23H24N2O5: C,
67.62; H, 5.93; N, 6.86. Found: C, 67.68; H, 5.92; N, 6.85.
(CH2), 49.4 (C), 17.2 (CH3); IR (KBr) 3353, 1700, 1209 cm-1
;
MS m/e 168 (M+, 9.6%), 138, 124, 69. Anal. Calcd for
C9H16N2O: C, 64.25; H, 9.59; N, 16.65. Found: C, 64.01; H,
9.42; N, 16.49.
3,7-Dip h en yl-1,5-d im eth ylbisp id in on e (2b). Following
a
general procedure for phenylation of amines,59 freshly
1,5-Dicar bom eth oxy-3,7-diazaadam an tan -9-on e (6). 3,7-
Dibenzyl-1,5-dicarbomethoxybispidinone (3a ) (1.96 g, 4 mmol)
was dissolved in acetic acid (20 mL) and ethanol (4 mL).
Palladium on activated carbon (10%, 0.2 g) was added. The
mixture was shaken under hydrogen pressure (3 atm) during
20 h. Thereafter, the mixture was filtered and the solid was
washed with acetic acid (10 mL) and ethanol (10 mL). The
solvent was evaporated, and the remaining oil was dissolved
in chloroform (15 mL) and washed with water (2 × 50 mL).
The chloroform phase was dried over MgSO4, and the solvent
was evaporated. The remaining product was dissolved in
chloroform/diethyl ether (1:1) and recrystallized at -20 °C. The
solid was collected and dried in vacuo (0.5 g, 1.9 mmol, 47%):
prepared sodium amide60 (3.9 g, 100 mmol) was suspended in
dry THF (10 mL) in a round-bottomed flask equipped with a
reflux condenser and a gas lock. The suspension was cooled
on ice and magnetically stirred. tert-Butanol (1.2 mL, 12
mmol) was added dropwise. 1,5-Dimetylbispidinone (0.9 g, 5
mmol) was added followed by phenyl bromide (1.26 mL, 12
mmol) that was added dropwise under 5 min. The tempera-
ture was raised and kept at 45 °C during 12 h under a nitrogen
atmosphere. The reaction mixture was cooled on ice, and
water (10 mL) was added in small portions. The reaction
mixture was poured into a separation funnel containing water
(100 mL), and the funnel was left for 30 min. The mixture
was then extracted with methylene chloride (2 × 50 mL). The
methylene chloride phases were combined, and the solvent was
removed under reduced pressure. The remaining oil was
purified by column chromatography (n-pentane/diethyl ether,
95:5). The appropriate fractions were combined, and the
solvent was removed under reduced pressure. The remaining
solid was recrystallized from ethanol, and repeated crystal-
lization gave the pure product as white needle-shaped crystals
(0.23 g, 0.72 mmol, 14%): mp 126 °C; 1H NMR (400 MHz,
CDCl3) δ 7.18 (m, 4H, m-Ph), 6.81 (m, 4H, o-Ph), 6.77 (m, 2H,
p-Ph), 3.93 (d, J ) 12.2 Hz, 4H, Heq), 3.18 (d, J ) 12.2 Hz, 4H,
1
mp 167 °C; H NMR (300 MHz, CDCl3) δ 4.05 (s, 2H, CH2),
3.76 (s, 6H, CH3), 3.70 (d, J ) 12.8 Hz, 2H, Heq), 3.60 (d, J )
12.8 Hz, 2H, Hax); 13C NMR (75.4 MHz, CDCl3) δ 200.3 (C-9),
168.8 (COO), 72.4 (N-CH2-N), 60.7 (CH2), 55.9 (C), 52.3
(CH3O); IR (KBr) 2958, 1732, 1699, 1443 cm-1; MS m/e 268
(M+, 2.7%), 240, 113, 59. Anal. Calcd for C12H16N2O5: C,
53.71; H, 6.01; N, 10.45. Found: C, 53.81; H, 6.15; N, 10.31.
3,7-Diben zyl-1,5-d im eth ylbisp id in on e (3b). A slight
modification of a literature procedure was employed.8c Ben-
zylamine (28.4 mL, 260 mmol) was added to 50 mL of ethanol
in a round-bottomed flask equipped with a reflux condenser.
The mixture was cooled on ice, and acetic acid (15 mL) was
added. Paraformaldehyde (15.0 g, 0.5 mol) and diethyl ketone
(13.2 mL, 125 mmol) were added, and the mixture was refluxed
for 15 h. The reaction mixture was then cooled on ice, and
300 mL of diethyl ether was added. Perchloric acid was added
until the pH was 3 or lower. After 12 h at -20 °C the solid
was collected and suspended in methylene chloride (200 mL).
The suspension was treated with 20% aqueous NaOH (500
mL). In the case of the remaining solid, more sodium
hydroxide (solid) was added until all precipitate was dissolved.
The aqueous phase was separated and extracted with meth-
ylene chloride (50 mL). The combined organic phases were
washed with water (2 × 300 mL). The methylene chloride was
removed under reduced pressure, and the resulting oil was
purified by column chromatography (diethyl ether/n-pentane,
1:1). The appropriate fractions were combined, and the solvent
was removed. Repeated crystallization from 95% ethanol at
-20 °C gave white crystals which were dried in vacuo (7.86 g,
45.2 mmol, 18%) (lit. 5.0%): mp 83-84 °C (lit.8c mp 84-86
°C); 1H NMR (400 MHz, CDCl3) δ 7.32 (m, 10H, aromatic),
3.53 (s, 4H, benzylic), 3.04 (d, J ) 11.0 Hz, 4H, Heq), 2.38 (d,
J ) 11.0 Hz, 4H, Hax), 0.96 (s, 6H, CH3); 13C NMR (100.6 MHz,
CDCl3) δ 215.0 (C-9), 138.4 (Cipso), 128.7 (Cortho), 128.2 (Cmeta),
127.0 (Cpara), 65.5 (CH2), 61.3 (PhCH2), 46.7 (C), 20.0 (CH3);
IR (KBr) 3024, 1720, 1653, 1601 cm-1; MS m/e 348 (M+, 0.8%),
214, 134, 91.
1,5-Dim eth ylbisp id in on e (4). A general procedure8b was
employed. 3,7-Dibenzyl-1,5-dimethylbispidinone (3b) (3.0 g,
8.6 mmol) was dissolved in a mixture of acetic acid (25 mL),
ethanol (5 mL), and perchloric acid (2 drops), and palladium
on activated carbon (10%, 0.3 g) was added. The mixture was
shaken under hydrogen pressure (3 atm) at room temperature
for 60 h. The reaction mixture was filtered, and the solid was
washed with ethanol (20 mL) and acetic acid (10 mL). The
filtrate was concentrated under reduced pressure and dissolved
in chloroform (10 mL). The solution was washed with 10%
aqueous NaOH (10 mL). The chloroform was removed under
reduced pressure, and the remaining oil was recrystallized
H
ax), 1.16 (s, 6H, CH3); 13C NMR (100.6 MHz, CDCl3) δ 214.0
(C-9), 149.5 (Cipso), 129.0 (Cmeta), 118.9 (Cpara), 115.4 (Cortho), 62.3
(CH2), 47.0 (C), 18.3 (CH3); IR (KBr) 1717, 1684, 1647, 1364
cm-1; MS m/e 320 (M+, 0.5%), 200, 130, 120, 77. Anal. Calcd
for C21H24N2O: C, 78.70; H, 7.55; N, 8.75. Found: C, 78.59;
H, 7.46; N, 8.84.
N,N′-Dip h en ylp ip er a zin e (7) was prepared according to
the literature:59 Yellowish crystals, mp 169-170 °C (lit.59 mp
166 °C); yield 7.55 g, 31.7 mmol, 84%; 1H NMR (400 MHz,
CDCl3) δ 7.30 (m, 4H, m-Ph), 7.00 (m, 4H, o-Ph), 6.90 (m, 2H,
p-Ph), 3.35 (s, 8H, CH2); 13C NMR (100.6 MHz, CDCl3) δ 151.2
(Cipso), 129.2 (Cmeta), 120.1 (Cpara), 116.3 (Cortho), 49.4 (CH2); IR
(KBr) 2831, 1598, 1496, 1447, 940 cm-1; MS m/e 238 (M+, 56%),
132, 105, 77.
N,N′-Dip h en yl-1,5-d ia za cycloocta n e (8). The dihydro-
bromide of 1,5-diazacyclooctane was obtained by condensation
of the disodium salt of N,N′-ditosyl-1,3-propanediamine with
1,3-ditosyloxypropane according to reported procedures61
(yield 60%), followed by detosylation with HBr/glacial acetic
acid61a (80 °C, 67 h, 89%).61c To a solution of tert-butanol (2.0
g, 27 mmol) in dry THF (22 mL) was added sodium amide (10.0
g, 256 mmol) with vigorous stirring (gas lock). 1,5-Diazacy-
clooctane dihydrobromide (3.84 g, 13.9 mmol) was added, and
after the initial gas evolution had ceased, the mixture was kept
at 45 °C. Bromobenzene (4.6 g, 29.3 mmol) was added in
portions during 30 min. The initially gray and then brown
suspension was stirred for another 5 h. Water (30 mL) was
added to the dark brown solution. The mixture was poured
into water (150 mL) and left for 2 h at room temperature, after
which time a dark brown solid had formed on the surface of a
yellow solution. The solid was filtered off on a Bu¨chner funnel
and dissolved in diethyl ether/ethanol (1:1). The remaining
solution was extracted with diethyl ether (50 mL) and the
(59) Cambe´re, P.; Derozier, N. Bull. Soc. Chim. Fr. 1969, 1737.
(60) Vogel, A. I. Textbook of Practical Organic Chemistry, 5th ed.;
Longman: Harlow, U.K., 1989; p 462.
(61) (a) Bo¨rjesson, L.; Welch, C. J . Acta Chem. Scand. 1991, 45, 621.
(b) Rickman, J . E.; Atkins, T. J . J . Am. Chem. Soc. 1974, 96, 2269.

(π-Allyl)palladium Complexes
Organometallics, Vol. 16, No. 6, 1997 1177
ether phase combined with the ether/ethanol solution. Filtra-
tion followed by evaporation yielded a yellowish-brown solid,
contaminated with a dark brown oil. The crude product (3.7
g) was dissolved in cyclohexane, the solution dried over sodium
sulfate, and recrystallized from diethyl ether: yield 2.77 g (10.8
Hz), 68.1 (C-1, C-3π-allyl), 61.5 (CH2), 57.0 (C), 53.7 (CH3); IR
(KBr) 2957, 1743, 1721, 1598, 1498 cm-1
.
(3,7-Dip h en yl-1,5-d im et h ylbisp id in on e)(1,3-η3-p r op e-
n yl)p a lla d iu m tr iflu or om eth a n esu lfon a te (15) was pre-
pared from 10 (18.3 mg, 50 µmol), 3,7-diphenyl-1,5-dimethyl-
bispidinone (2b ) (40.8 mg, 100 µmol), and silver trifluoro-
methanesulfonate (28.3 mg, 110 µmol) (50 mg, 70%): mp 109
°C dec; 1H NMR (400 MHz, CDCl3) δ 7.55 (m, 4H, o-Ph), 7.41
(m, 4H, m-Ph), 7.14 (m, 2H, p-Ph), 5.60 (tt, J ) 12.4, 6.8 Hz,
1H, H-2π-allyl), 4.88 (d, J ) 12.6 Hz, 4H, Heq), 3.07 (d, J ) 12.4
Hz, 2H, Hanti), 3.05 (d, J ) 12.6 Hz, 4H, Hax), 2.22 (d, J ) 6.8
Hz, 2H, Hsyn), 1.27 (s, 6H, CH3); 1H NMR (400 MHz, acetone-
d6) δ 7.68 (m, 4H, o-Ph), 7.42 (m, 4H, m-Ph), 7.16 (m, 2H, p-Ph),
5.72 (tt, 12.4, 6.6 Hz, 1H, H-2π-allyl), 4.98 (d, J ) 12.4 Hz, 4H,
1
mmol, 78%); mp 116 -117 °C; H NMR (400 MHz, CDCl3) δ
7.25 (m, 4H, m-Ph), 6.72 (m, 5H, o-Ph, p-Ph), 3.48 (m, 8H,
NCH2), 2.02 (m, 4H, CH2); 13C NMR (100.6 MHz, CDCl3) δ
147.5 (Cipso), 129.3 (Cmeta), 115.5 (Cpara), 111.3 (Cortho), 48.7
(NCH2), 25.2 (CH2); IR (KBr) 2872, 1471, 1360, 692 cm-1; MS
m/e 266 (M+, 5%), 237, 146, 104, 77. Anal. Calcd for
C
18H22N2: C, 81.16; H, 8.32; N, 10.52. Found: C, 81.00; H,
8.18; N, 10.56.
Bis[(1,3-η3-p r op en yl)p a lla d iu m ch lor id e] (10) was pre-
pared according to a literature procedure:62 mp 159 °C (lit.62
H
H
eq), 3.26 (d, J ) 12.4 Hz, 4H, Hax), 2.95 (d, J ) 12.4 Hz, 2H,
1
anti), 2.17 (d, J ) 6.6 Hz, 2H, Hsyn), 1.27 (s, 6H, CH3); 13C
mp 160 °C); H NMR (300 MHz, CDCl3) δ 5.46 (tt, J ) 12.1
Hz, 6.7, 1H, H-2π-allyl), 4.11 (d, J ) 6.7 Hz, 2H, Hsyn), 3.04 (d,
NMR (100.6 MHz, CDCl3) δ 210.1 (C-9), 154.6 (Cipso), 129.3
(Cmeta), 125.3 (Cpara), 120.7 C-2π-allyl), 117.4 (Cortho), 68.3 (C-1,
C-3π-allyl), 66.5 (CH2), 46.4 (C), 17.7 (CH3); IR (CDCl3) 2348,
1740, 1594, 1460 cm-1. Anal. Calcd for C25H29N2O4SF3Pd‚
CDCl3: C, 42.35; H, 4.24; N, 3.80. Found: C, 43.33; H, 4.09;
N, 3.69. Deviations are due to solvent inclusion; see the X-ray
crystallographic data.
J ) 12.1 Hz, 2H, Hanti); 13C NMR (100.6 MHz, CDCl3) δ 111.2
(CH), 63.0 (CH2); IR (KBr) 1457, 1382, 1022 cm-1
.
Bis[(2-m et h ylen e-6,6-d im et h ylb icyclo[3.1.1]h ep t -2,3,-
10-η3-en yl)p a lla d iu m ch lor id e] (11) was prepared from
â-pinene.63 1H NMR: Table 6.
Gen er a l Meth od for P r ep a r a tion of Com p lexes [(L)-
(1,3-η3-p r op en yl)P d ] CF 3SO3. To chloroform (3 mL) was
added chlorodimer 10 and a 10% excess of silver trifluo-
romethanesulfonate. The mixture was stirred during 1 min
at room temperature under a nitrogen atmosphere, and then
an equimolar amount of nitrogen ligand was added. After a
further 10 min the mixture was filtered. To the filtrate was
added diethyl ether to initiate crystallization, followed by
storage at -20 °C overnight. The obtained solids were
collected and dried in vacuo. Because of the straightforward
method of preparation, elemental analyses were performed
only for selected complexes.
(3,7-Dib en zyl-1,5-d ica r b om et h oxyb isp id in on e(1,3-η3-
p r op en yl)p a lla d iu m tr iflu or om eth a n esu lfon a te (16) was
prepared from 10 (18.3 mg, 50 µmol), 3,7-dibenzyl-1,5-dicar-
bomethoxybispidinone (3a ) (43.6 mg, 100 µmol), and silver
trifluoromethanesulfonate (28.3 mg, 110 µmol) (30 mg, 50%):
1
mp 156 °C; H NMR (400 MHz, CDCl3) δ 7.34-7.40 (m, 10H,
aromatic), 6.21 (tt, J ) 12.5 Hz, 7.1 Hz, 1H, H-2π-allyl), 4.69 (s,
4H, benzyl), 4.17 (d, J ) 13.0 Hz, 4H, Heq), 3.88 (d, J ) 7.1
Hz, 2H, Hsyn), 3.75 (d, J ) 12.5 Hz, 2H, Hanti), 3.67 (s, 6H, CH3),
3.56 (d, J ) 13.0 Hz, 4H, Hax); 13C NMR (100.6 MHz, CDCl3,
-55 °C) δ 199.9 (C-9), 165.8 (COO), 132.1 (Cortho), 129.4 (Cipso),
(Bip yr id yl)(1,3-η3-p r op en yl)p a lla d iu m tr iflu or om eth -
a n esu lfon a te (12) was prepared from 10 (36.6 mg, 100 µmol),
bipyridine (1) (31.2 mg, 200 µmol), and silver trifluoromethane-
sulfonate (56.6 mg, 220 µmol) according to the general
procedure but with methanol as solvent (80 mg, 76%): mp 288
128.7 (Cpara), 128.3 (Cmeta), 121.7 (C-2π-allyl), 120.3 (CF3, J CF
)
320 Hz), 70.9 (PhCH2), 65.0 (C-1π-allyl, C-3π-allyl), 58.1 (CH2),
57.8 (CH2), 56.9 (C), 56.8 (C), 53.6 (CH3); IR (KBr) 2360, 1746,
1724, 1497, 1255 cm-1
. Anal. Calcd for C29H33N2O8SF3-
Pd‚H2O: C, 46.38; H, 4.70; N, 3.73. Found: C, 46.51; H, 4.68;
N, 3.66.
1
°C; H NMR (400 MHz, CDCl3) δ 8.82 (ddd, J ) 0.8, 1.7, 5.4
Hz, 2H, H-6bpy), 8.59 (ddd, J ) 0.8, 1.3, 8.1 Hz, 2H, H-3bpy),
8.25 (ddd, J ) 1.7, 7.7, 8.1 Hz, 2H, H-4bpy), 7.77 (ddd, J ) 1.3,
5.4, 7.7 Hz, 2H, H-5bpy), 6.01 (tt, J ) 7.1 Hz, 12.7 Hz 1H,
H-2π-allyl) 4.29 (d, J ) 7.1 Hz, 2H, Hsyn), 3.56 (d, J ) 12.7, 2H,
(3,7-Diben zyl-1,5-dim eth ylbispidin on e)(1,3-η3-pr open yl)-
p a lla d iu m tr iflu or om eth a n esu lfon a te (17) was prepared
from 10 (36.6 mg, 100 µmol), 3,7-dibenzyl-1,5-dimethylbispi-
dinone (3b) (69.6 mg, 200 µmol), and silver trifluoromethane-
sulfonate (56.6 mg, 220 µmol) (43 mg, 41%): mp 132 °C.
1H NMR (400 MHz, CDCl3) δ 7.27-7.37 (m, 10H, aromatic)
6.21 (tt, J ) 12.4, 6.8 Hz, 1H, H-2π-allyl), 4.61 (s, 4H, benzylic),
3.90 (br, 4H, Heq), 3.87 (d, J ) 6.8 Hz, 2H, Hsyn), 3.78 (d, 12.4
Hz, 2H, Hanti), 2.74 (d, J ) 12.4, 4H, Hax), 0.90 (s, 6H, CH3);
1H NMR (100.6 MHz, acetone-d6): δ 7.37-7.47 (m, 10H,
aromatic) 6.20 (tt, J ) 12.5, 7.1, 1H, H-2π-allyl), 4.81 (s, 4H,
benzylic), 4.09 (d, J ) 7.1, 2H, Hsyn), 4.00 (br s, 4H, Heq), 3.69
(d, J ) 12.5 Hz, 2H, Hanti), 2.86 (d, J ) 12.5, 4H, Hax), 0.86 (s,
6H, CH3); 13C NMR (100.6 MHz, CDCl3) δ 210.1 (C-9), 132.2,
129.4, 129.2, 128.7, 121.8 (C-2π-allyl), 71.1 (benzylic), 65.1 (C-
1,C-3π-allyl), 63.4 (CH2), 46.5 (C), 17.5 (CH3); IR (CDCl3) 2869,
1735, 1542, 1261 cm-1. Anal. Calcd for C27H33N2O4SF3Pd: C,
50.30; H, 5.16; N, 4.35. Found: C, 48.93; H, 4.92; N, 4.24.
H
anti); 1H NMR (400 MHz, acetone-d6) δ 9.05 (ddd, J ) 0.8,
1.6, 5.3 Hz, 2 H, H-6bpy), 8.68 (ddd, J ) 0.8, 1.2, 8.1 Hz, 2H,
H-3bpy), 8.40 (ddd, J ) 1.6, 7.7, 8.1 Hz, 2H, H-4bpy), 7.83 (ddd,
J ) 1.2, 5.3, 7.7 Hz, 2H, H-5bpy), 6.16 (tt, J ) 7.0 Hz, 12.7 Hz
1H, H-2π-allyl), 4.48 (d, J ) 7.0, 2H, Hsyn), 3.56 (d, J ) 12.7,
2H, Hanti); IR (KBr) 3082, 1599, 1493, 1471, 1445 cm-1
.
(Bip yr id yl)[(2S*,3R*)-(2-m eth oxy-3,4,5-η3-h exen yl)]p a l-
la d iu m ]tr iflu or om eth a n e su lfon a te (13) was prepared as
described previously.6
(3,7-Dip h en yl-1,5-d ica r bom eth oxybisp id in on e)(1,3-η3-
p r op en yl)p a lla d iu m tr iflu or om eth a n esu lfon a te (14) was
prepared from 10 (18.3 mg, 50 µmol), 3,7-diphenyl-1,5-dicar-
bomethoxybispidinone (2a ) (40.8 mg, 100 µmol), and silver
trifluoromethanesulfonate (28.3 mg, 110 µmol) (50 mg, 70%):
1
mp 137 °C; H NMR (400 MHz, CDCl3) δ 7.61 (m, 4H, o-Ph),
(3,7-Dip h en yl-1,5-d ica r bom eth oxybisp id in -9-d iol)(1,3-
η3-p r op en yl)p a lla d iu m tr iflu or om eth a n esu lfon a te (18)
was obtained in solution by storing 14 in chloroform at room
temperature during 48 h: 1H NMR (400 MHz, CDCl3) δ 7.57
(m, o-Ph), 7.41 (m, 4H, m-Ph), 7.13 (m, p-Ph), 5.52 (tt, J )
12.4, 6.6 Hz, 1H, H-2π-allyl), 5.26 (s, 2H, OH), 4.80 (d, J ) 12.2
Hz, 4H, Heq), 3.94 (s, 6H, CH3), 3.65 (d, J ) 12.2 Hz, 4H, Hax),
2.96 (d, J ) 12.4 Hz, 2H, Hanti), 2.11 (d, J ) 6.6 Hz, 2H, Hsyn);
13C NMR (100.6 MHz, CDCl3) 171.1 (COO), 155.3 (Cipso), 129.3
(Cmeta), 125.3 (Cpara), 117.4 (Cortho), 113.2 (C-2π-allyl), 92.6
(C(OH)2), 68.0 (C-1,C-3π-allyl), 57.5 (CH2), 53.9 (CH3), 50.6 (C).
Anal. Calcd for C27H29N2O8SF3Pd‚H2O; C, 44.85; H, 4.32; N,
3.87. Found: C, 45.12; H, 4.10; N, 3.94.
7.39 (m, 4H, m-Ph), 7.13 (m, 2H, p-Ph), 5.67 (tt, J ) 12.5, 6.9
Hz, 1H, H-2π-allyl), 5.03 (br, 4H, Heq), 3.88 (br s, 4H, Hax), 3.85
(s, 6H, CH3), 2.98 (d, J ) 12.5 Hz, 2H, Hanti), 2.21 (d, J ) 6.9
Hz, 2H, Hsyn); 1H NMR (400 MHz, 1:1 CBr2F2/CD2Cl2): δ 7.63
(m, 4H, o-Ph), 7.47 (m, 4H, m-Ph), 7.19 (m, 2H, p-Ph), 5.67
(tt, J ) 12.5, 6.9 Hz, 1H, H-2π-allyl), 5.18 (br, 4H, Heq), 3.97 (s,
6H, CH3), 3.70 (d, J ) 12.2, 4H, Hax), 3.03 (d, J ) 12.5, 2H,
H
anti), 2.27 (br, 2H, Hsyn); 13C NMR (100.6 MHz, CDCl3, -25
°C) δ 200.2 (C-9), 165.6 (COO), 154.1 (Cipso), 129.3 (Cmeta), 125.6
(Cpara), 120.7 (C-2π-allyl), 117.5 (Cortho), 120.5 (CF3, J CF ) 321
(62) Hu¨ttel, R.; Kratzer, J .; Bechter, M. Chem. Ber. 1961, 94, 766.
(63) Trost, B. M.; Strege, P. E. Tetrahedron 1974, 30, 2603.

1178 Organometallics, Vol. 16, No. 6, 1997
Gogoll et al.
(N ,N ′-D ip h e n y lp ip e r a zin e )(1,3-η3-p r o p e n y l)p a lla -
d iu m tr iflu or om eth a n e su lfon a te (19) was prepared from
10 (36.6 mg, 100 µmol), N,N′-diphenylpiperazine (7) (48.0 mg,
200 µmol), and silver trifluoromethanesulfonate (56.6 mg, 220
µmol): yield 50 mg (70%); mp 118 °C; 1H NMR (400 MHz,
CDCl3) δ 7.42 (m, 8H, o-Ph, m-Ph), 7.19 (m, 2H, p-Ph), 5.58
(d, 12.1, 1H, H-4axL), 3.17 (d, 12.2, 1H, H-6axL), 3.00 (d, 12.7,
1H, H-2axL), 2.97 (d, 12.4, 1H, H-8axL), 2.54 (ddd, 4.9, 6.5, 9.0
Hz, 1H, H-7â), 1.94 (dd, 4.9, 4.9, 1H, H-1), 1.77 (s, 1H, H-10a),
1.62 (m, 1H, H-5), 1.26 (s, 3H, Me-1L/Me-5L), 1.20 (s, 3H, Me-
1L/Me-5L), 1.18 (d, 9.2 Hz, 1H, H-7R), 1.11 (s, 3H, Me-8), 0.60
(s, 3H, Me-9), 0.56 (dd, 5.1, 18.4 Hz, 1H, H-4â), -1.31 (dd, 3.8,
18.4 Hz, 1H, H-4R). 13C NMR (CDCl3, -55 °C) δ 210.9, 154.6,
152.0, 145.2, 130.5, 129.9, 128.5, 127.9, 125.4, 124.8, 122.8,
122.0, 119.4, 114.5, 73.6, 69.0, 68.4, 66.6, 66.0, 65.5, 46.4, 45.8,
45.0, 38.7, 37.3, 35.1, 25.9, 25.4, 21.4, 18.6, 17.2; IR (CDCl3)
(tt, J ) 12.1, 6.7 Hz, 1H, H-2π-allyl), 3.98 (d, J ) 6.7 Hz, 2H,
1
H
syn), 3.74 (br, 8H, CH2), 3.05 (d, J ) 12.1 Hz, 2H, Hanti); H
NMR (400 MHz, acetone-d6) δ 7.35 (m, 8H, o-Ph, m-Ph), 7.05
(m, 2H, p-Ph), 5.91 (tt, J ) 12.1, 6.8 Hz, 1H, H-2π-allyl), 4.12
(d, J ) 6.8 Hz, 2H, Hsyn), 3.28 (d, J ) 12.1 Hz, 2H, Hanti), 3.67
(br s, 8H, CH2); IR (CDCl3) 1732, 1598, 1495, 1214 cm-1. Low
stability prevented elemental analysis.
2977, 1792, 1739, 1496, 1461 cm-1
.
(3,7-Diben zyl-1,5-d im eth ylbisp id in on e)(2-m eth ylen e-
6,6-d im eth ylbicyclo[3.1.1]h ep t-2,3,10-η3-en yl)p a lla d iu m
tr iflu or om eth a n esu lfon a te (22) was prepared from 11 (27.8
mg, 100 µmol), 3b (34.8 mg, 100 µmol), and silver trifluo-
romethanesulfonate (28.3 mg, 110 µmol) according to the
(N,N′-Dip h en yld ia za cycloocta n e)(1,3-η3-p r op en yl)p a l-
la d iu m Tr iflu or om eth a n esu lfon a te (20). To 2 mL of
chloroform was added 10 (18.3 mg, 50 µmol) and silver
trifluoromethanesulfonate (56.6 mg, 100 µmol). The mixture
was magnetically stirred during 10 min at room temperature
under a nitrogen atmosphere and was thereafter filtered. To
the filtrate was added N,N′-diphenyl-diazacyclooctane (8) (26.6
mg, 100 µmol). This mixture was stirred for an additional 10
min under a nitrogen atmosphere. Diethyl ether was added
drop by drop to induce crystallization, and the solution was
stored at -20 °C overnight. The light yellow crystals were
collected and dried in vacuo (18 mg, 44%): mp 138 °C; 1H NMR
(400 MHz, CDCl3) δ 7.00-7.32 (m, 10 H, aromatic), 5.57 (tt, J
) 12.3, 6.7 Hz, 1H, H-2π-allyl), 4.02 (br, 4H, CH2), 3.48 (br, 4H,
CH2), 2.86 (d, J ) 12.3 Hz, 2 H, Hanti), 2.46 (br, 4 H, CH2),
2.28 (br, 2H, Hsyn); 1H NMR (400 MHz, acetone-d6) δ 7.20-
6.60 (m, broad, 10 H, aromatic), 5.90 (m, 1H, H-2π-allyl), 4.02
(br, 4H, CH2), 3.48 (br, 4H, CH2), 2.92 (d, J ) 11.7 Hz, 2H,
1
general procedure (36 mg, 49%): colorless oil; H NMR (400
MHz, CDCl3, 25 °C) δ 7.26-7.39 (m, 10H, PhL), 4.66 (d, 13.4,
1H, Bz-7L), 4.48 (m, 2H, Bz-3L), 4.48 (m, 1H, H-3), 4.46 (m,
1H, H-8eqL), 4.41 (dd, 2.3, 12.4, 1H, H-2eqL), 4.32 (d, 13.4, 1H,
Bz-7L), 4.07 (s, 1H, H-10b), 3.50 (s, 1H, H-10a), 3.43 (dd, 2.3,
12.5, 1H, H-4eqL), 3.19 (dd, 2.4, 12.4, 1H, H-6eqL), 2.87 (d, 12.4,
1H, H-6axL), 2.85 (m, 1H, H-7â), 2.77 (d, 12.6, 1H, H-8axL), 2.76
(d, 12.4, 1H, H-2axL), 2.68 (d, 12.5, 1H, H-4axL), 2.45 (dd, 5.4,
5.4, 1H, H-1), 2.21 (m, 1H, H-5), 1.78 (d, 9.7, 1H, H-7R), 1.77
(m, 1H, H-4â), 1.68 (dd, 4.0, 18.3, 1H, H-4R), 1.39 (s, 3H, Me-
8), 1.02 (s, 3H, Me-9), 0.96 (s, 3H, Me-5L), 0.84 (s, 3H, Me-1L);
(400 MHz, CDCl3 -50 °C) δ 7.44-7.23 (m, 10H, PhL), 4.62 (d,
13.4, 1H, Bz-3L), 4.54 (d, 13.6, 1H, Bz-7L), 4.44 (m, 1H, H-3),
4.44 (m, 1H, H-2eqL), 4.43 (m, 1H, Bz-7L), 4.41 (m, 1H, H-8eqL),
4.29 (d, 13.4, 1H, Bz-3L), 4.02 (s, 1H, H-10b), 3.46 (s, 1H,
H-10a), 3.40 (d, 12.7, 1H, H-6eqL), 3.15 (d, 12.2, 1H, H-4eqL),
2.88 (d, 12.2, 1H, H-4axL), 2.80 (m, 1H, H-7â), 2.76 (two d, 12.7,
2H, H-2axL, H-2axL), 2.73 (d, 12.7, 1H, H-6axL), 2.43 (dd, 5.3,
5.3, 1H, H-1), 2.18 (br, 1H, H-5), 1.76 (m, 1H, H-4â), 1.68 (m,
2H, H-4R, H-7R), 1.37 (s, 3H, Me-8), 1.00 (s, 3H, Me-9), 0.95
(s, 3H, Me-5L), 0.84 (s, 3H, Me-1L); 13C NMR (CDCl3, -55 °C)
δ 210.2, 145.8, 132.5 (2C), 131.7 (2C), 129.8, 129.4, 129.3,
129.1, 128.7 (2C). 128.6 (2C), 71.2, 70.5, 69.2, 64.7, 64.4, 64.1,
64.0, 62.2, 46.7, 46.3, 46.2, 39.8, 38.4, 34.4, 29.6, 25.9, 21.9,
H
anti) 2.46 (br, 4H, CH2), 1.98 (br, 4H, Hsyn); IR (KBr) 1595,
1504, 1267, 1160 cm-1. Anal. Calcd for C22H27N2F3PdSO3: C,
46.94; H, 4.83; N, 4.98. Found: C, 46.80; H, 4.98; N, 5.08.
(3,7-Dip h en yl-1,5-d im eth ylbisp id in on e)(2-m eth ylen e-
6,6-d im eth ylbicyclo[3.1.1]h ep t-2,3,10-η3-en yl)p a lla d iu m
tr iflu or om eth a n esu lfon a te (21) was prepared from 11 (13.9
mg, 50 µmol), 2b (16 mg, 50 µmol), and silver trifluo-
romethanesulfonate (14.1 mg, 55 µmol) according to the
general procedure (25 mg, 70%): mp 130-132 °C; 1H NMR
(400 MHz, CDCl3, 25 °C) δ 7.57 (m, 4H, o-PhL), 7.40 (m, 4H,
m-PhL), 7.12 (m, 2H, p-PhL), 4.90 (br, 4H, HeqL), 3.67 (d, 5.2,
1H, H-3), 3.54 (br s, 1H, H-10b), 3.08 (br, 4H, HaxL), 2.58 (ddd,
5.5, 6.2, 9.5, 1H, H-7â), 1.95 (dd, 5.2, 5.5, 1H, H-1), 1.86 (s,
1H, H-10a), 1.66 (br s, 1H, H-5), 1.32 (d, 9.5, 1H, H-7R), 1.25
(s, 6H, MeL), 1.17 (s, 3H, Me-8), 0.66 (s, 3H, Me-9), 0.63 (dd,
5.4, 18.8, 1H, H-4â), -1.15 (dd, 1.6, 18.8, 1H, H-4R); (400 MHz,
CDCl3 -55°): δ 8.21 (1H, o-Ph-3), 7.86 (1H, o-Ph-7), 7.49 (1H,
m-Ph-3), 7.45 (1H, m-Ph-7), 7.30 (1H, m-Ph-7′), 7.27 (1H,
m-Ph-3′), 7.12 (1H, p-Ph-3), 7.08 (1H, p-Ph-7), 7.05 (1H, o-Ph-
7′), 7.00 (1H, o-Ph-3′), 5.48 (d, J ) 12.7 Hz, 1H, H-2eqL), 5.39
(d, 12.4, 1H, H-8eqL), 4.30 (d, 12.2, 1H, H-6eqL), 3.86 (d, 12.1,
1H, H-4eqL), 3.58 (d, 5.1 Hz, 1H, H-3), 3.42 (s, 1H, H-10b), 3.29
17.9, 17.1; IR (CDCl3) 2935, 1736, 1584, 1260 cm-1
.
Ack n ow led gm en t. We thank the Swedish Natural
Science Research Council and Magn. Bergvalls Stiftelse
for financial support.
Su p p or tin g In for m a tion Ava ila ble: Tables of positonal
and thermal parameters and bond distances and angles (4
pages). Ordering information is given on any current mast-
head page.
OM9609527