Solution behavior and X-ray structure of cationic allylpalladium(II) complexes with iminophosphine ligands. Kinetics and mechanism of allyl amination by secondary amines

Article abstract of DOI:10.1021/om980858l

The solution behavior of the cationic complexes [Pd(η³-allyl)(P-N)]⁺ (P-N = o-(PPh₂)C₆H₄-CH=NR (R = C₆H₄OMe-4, Me, CMe₃, (R)-bornyl); allyl = propenyl (1a-4a) and 3-methyl-2-butenyl (1b-4b)) consists essentially of three dynamic processes: (i) a very fast conformational change of the P-N chelate ring, which moves above and below the P-Pd-N coordination plane, (ii) a relatively fast η³-η¹-η³ interconversion which brings about a syn-anti exchange only for the allylic protons cis to phosphorus; (iii) a slower apparent rotation of the η³-allyl ligand around its bond axis. For 1b-3b, two geometrical isomers are observed, the predominant one having the allyl CMe₂ group trans to phosphorus. The complexes 4a and 4b, containing the chiral (R)-bornyl group, are present in solution with two and four diastereomeric species, respectively.

Full text of DOI:10.1021/om980858l

Organometallics 1999, 18, 1137-1147
1137
Solu tion Beh a vior a n d X-r a y Str u ctu r e of Ca tion ic
Allylp a lla d iu m (II) Com p lexes w ith Im in op h osp h in e
Liga n d s. Kin etics a n d Mech a n ism of Allyl Am in a tion by
Secon d a r y Am in es
Bruno Crociani*,† and Simonetta Antonaroli
Dipartimento di Scienze e Tecnologie Chimiche, Universita` di Roma “Tor Vergata”,
00133 Rome, Italy
Giuliano Bandoli
Dipartimento di Scienze Farmaceutiche, Universita` di Padova, Padua, Italy
Luciano Canovese,‡ Fabiano Visentin, and Paolo Uguagliati
Dipartimento di Chimica, Universita` di Venezia, Venice, Italy
Received October 14, 1998
The solution behavior of the cationic complexes [Pd(η3-allyl)(P-N)]+ (P-N ) o-(PPh2)C6H4-
CHdNR (R ) C6H4OMe-4, Me, CMe3, (R)-bornyl); allyl ) propenyl (1a -4a ) and 3-methyl-
2-butenyl (1b-4b)) consists essentially of three dynamic processes: (i) a very fast
conformational change of the P-N chelate ring, which moves above and below the P-Pd-N
coordination plane, (ii) a relatively fast η3-η1-η3 interconversion which brings about a syn-
anti exchange only for the allylic protons cis to phosphorus; (iii) a slower apparent rotation
of the η3-allyl ligand around its bond axis. For 1b-3b, two geometrical isomers are observed,
the predominant one having the allyl CMe2 group trans to phosphorus. The complexes 4a
and 4b, containing the chiral (R)-bornyl group, are present in solution with two and four
diastereomeric species, respectively. The X-ray structural analysis of 4b(ClO4) shows the
presence of two diastereomeric molecules in the unit cell, both having distorted-square-
planar coordination geometries, characterized by rather elongated Pd-CMe2 bonds trans to
phosphorus and by a marked distortion of the allyl ligand, which is rotated away from the
PPh2 group. The complexes [Pd(η3-allyl)(P-N)]+ react with secondary amines HY in the
presence of fumaronitrile, yielding [Pd(η2-fn)(P-N)] and allylamines. Under pseudo-first-
order conditions the amination rates obey the laws kobs ) k2[HY] + k3[HY]2 for 1a -4a and
kobs ) k2[HY] for 1b, 3b, and 4b. The k2 term is related to direct bimolecular attack on a
terminal allyl carbon of the substrate, whereas the k3 term is ascribed to parallel attack by
a further amine molecule on the intermediate [Pd(allyl)(P-N)(HY)]+. The k2 values increase
with increasing basicity and decreasing steric hindrance of the amine, and with increasing
electron-withdrawing ability and increasing bulkiness of the P-N nitrogen substituent. The
higher amination rates for [Pd(η3-allyl)(P-N)]+, compared to [Pd(η3-allyl)(R-diimine)]+, are
essentially due to lack of displacement equilibria of the P-N ligand by amines.
In tr od u ction
(0) derivatives to generate in situ the catalyst (or
catalyst precursor), which was generally assumed to be
the cationic species [Pd(η3-allyl)(L-L′)]+. The cationic
complexes with chiral P-N ligands [Pd(η3-allyl)(P-N)]X
(X- ) SbF6-, PF6-, BF4-, CF3SO3-) have been isolated,
and their solution behavior has been studied by multi-
nuclear and multidimensional NMR spectrometry in
order to rationalize the observed high enantioselectivity
(ee up to 97-99%) on the basis of (i) the relative
concentrations of the configurational isomers resulting
from different geometries of the allyl group, (ii) the rates
of isomer interconversion, and (iii) the site and the
In recent years the palladium-catalyzed amination of
allyl substrates has been extensively studied for its
relevance to organic synthesis.1 Enantioselective allylic
amination has been the subject of considerable interest,
and several chiral ligands have been developed for this
particular application.2 Among these, diphosphines3 and
combined P,N donor4 bidentate ligands have been used
either in the reaction with [Pd(µ-Cl)(η3-allyl)]2 dimers
or in the oxidative addition of allyl esters to palladium-
† E-mail: crociani@stc.uniroma2.it.
‡ E-mail: cano@unive.it.
(1) (a) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1989, 28, 1173. (b)
Godleski, S. A. In Comprehensive Organic Synthesis; Trost, B. M.,
Fleming, I., Semmelbach, M. F., Eds.; Pergamon: Oxford, U.K., 1991;
Vol. 4, p 585. (c) Tsuji, J . Palladium Reagents and Catalysts. Innovation
in Organic Synthesis; Wiley: Chichester, U.K., 1995; p 290.
(2) (a) Frost, C. G.; Howarth, J .; Williams, J . M. J . Tetrahedron:
Asymmetry 1992, 3, 1089. (b) Williams, J . M. J . Synlett 1995, 705. (c)
Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395. (d) Tye,
H.; Smyth, D.; Eldred, C.; Wills, M. J . Chem. Soc., Chem. Commun.
1997, 1053.
10.1021/om980858l CCC: $18.00 © 1999 American Chemical Society
Publication on Web 02/27/1999

1138 Organometallics, Vol. 18, No. 7, 1999
Crociani et al.
relative rates of nucleophilic attack at the nonequivalent
terminal allyl carbons trans to donor atoms with dif-
ferent electronic and steric properties.4c,d,5 To the best
of our knowledge, however, no quantitative kinetic data
for the allyl amination of [Pd(η3-allyl)(P-N)]+ have been
reported so far.
We therefore decided to extend these studies to the
cationic complexes [Pd(η3-allyl)(P-N)]+ containing imi-
nophosphine ligands of the (2-(diphenylphosphino)-
benzylidene)amine type in order (i) to stabilize the com-
plex toward displacement of the P-N chelate ligand by
the entering amine and (ii) to investigate the influence
of the coordinated PPh2 unit on the reaction rates. As
indicated by substitution reactions on palladium(0)
adducts,10 iminophosphines are better chelating ligands
than 2-(iminomethyl)pyridines. On the other hand, it
is well-known that the PPh2 group in P-N ligands
enhances the reactivity of the corresponding (η3-allyl)-
palladium(II) cationic complexes toward nucleophilic
attack at the allyl moiety (and particularly at the ter-
minal carbon trans to phosphorus), due to the better
π-accepting properties and higher trans influence of the
phosphino group relative to the N-donor group.11
In the present work, we have also used an asym-
metrically substituted allyl ligand (namely, the 3-meth-
yl-2-butenyl ligand) in order to gain a better under-
standing of the dynamic processes involving isomer in-
terconversion for the cationic substrates [Pd(η3-allyl)-
(P-N)]+ and of factors affecting the regiochemistry of
allylic amination.
In previous papers we described the solution behavior
of the complexes [Pd(η3-allyl)(L-L′)]+ (L-L′ ) 2-(imi-
nomethyl)pyridine6 or 2-(thiomethyl)pyridine7), the ki-
netics of nucleophilic attack at the allyl group in these
complexes by secondary and tertiary amines yielding
zerovalent compounds [Pd(η2-ol)(L-L′)] in the presence
of activated olefins (ol),7,8 and the kinetics of oxidative
allyl transfer from allylammonium cations to [Pd(η2-ol)-
(L-L′)] which regenerates the starting derivatives [Pd-
(η3-allyl)(L-L′)]+.9 While the observed dynamic pro-
cesses consist essentially of an apparent rotation of the
η3-allyl ligand around its bond axis to the metal and/or
inversion at sulfur involving exchange of the two coor-
dinating sulfur lone pairs for the 2-(thiomethyl)pyridine
ligands, the kinetic data suggest the mechanism
Resu lts a n d Discu ssion
NMR Sp ectr a a n d Solu tion Beh a vior . The 1H and
31P{1H} NMR spectroscopic data for the η3-propenyl
complexes 1a (BF4)-4a (BF4) and for the corresponding
η3-3-methyl-2-butenyl derivatives 1b(BF4)-4b(BF4) are
summarized in Table 1, while the 13C{1H} NMR results
of some selected compounds are reported in Table 2. The
configurations of the complexes are sketched in Scheme
1, along with the numbering scheme for the allylic
protons and carbons.
For the η3-3-methyl-2-butenyl derivatives, two cis and
trans isomeric structures, II and III, are possible.
However, due to the nonplanarity of the chelate six-
membered ring of these complexes, as revealed by the
X-ray structure of 4b(ClO4) (see further), the imino
carbon and the ortho-disubstituted phenyl group of the
P-N ligand lie on the same side out of the N-Pd-P
coordination plane, whereas the nitrogen substituent R
and one of the phenyl groups of the PPh2 unit are in
pseudoaxial positions on the opposite side of the N-Pd-P
coordination plane. Such an asymmetric conformation
of the chelate P-N ligand, compounded with the dif-
ferent orientations of the η3-allyl group, increases to four
the number of possible isomers for each configuration
I-III as shown in Figure 1. The isomeric species A and
The nucleophilic attack occurs at a terminal allyl carbon
of the cationic substrate at different rates, depending
on the steric and electronic properties of L-L′ and Nu,
and it is also accompanied by reversible displacement
of the L-L′ ligand by the more coordinating amines.
(3) (a) Hayashi, T.; Yamamoto, A.; Ito, Y. Tetrahedron Lett. 1988,
29, 99. (b) Hayashi, T.; Yamamoto, A.; Ito, Y.; Nishioka, E.; Miura, H.;
Yanagi, K. J . Am. Chem. Soc. 1989, 111, 6301. (c) Hayashi, T.; Kishi,
K.; Yamamoto, A.; Ito, Y. Tetrahedron Lett. 1990, 31, 1743. (d) Trost,
B. M.; Van Vranken, D. L. Angew. Chem., Int. Ed. Engl. 1992, 31, 228.
(e) Trost, B. M.; Van Vranken, D. L.; Bingel, C. J . Am. Chem. Soc.
1992, 114, 9327 (f) Trost, B. M.; Van Vranken, D. L. J . Am. Chem.
Soc.1993, 115, 444. (g) Trost, B. M.; Bunt, R. C. J . Am. Chem. Soc.
1994, 116, 4089. (h) Trost, B. M.; Krueger, A. C.; Bunt, R. C.;
Zambrano, J . J . Am. Chem. Soc. 1996, 118, 6520. (i) Mori, M.; Kuroda,
S.; Zhang, C. S.; Sato, Y. J . Org. Chem. 1997, 62, 3263. (j) Uozumi, Y.;
Tanahashi, A.; Hayashi, T. J . Org. Chem. 1993, 58, 6826. (k) Yamazaki,
A.; Achiwa, K. Tetrahedron: Asymmetry 1995, 6, 1021. (l) Thorey, C.;
Wilken, J .; Henin, F.; Martens, J .; Mehler, T.; Muzart, J . Tetrahedron
Lett. 1995, 36, 5527. (m) Yamazaki, A.; Achiwa, I.; Achiwa, K.
Tetrahedron: Asymmetry 1996, 7, 403.
(4) (a) von Matt, P.; Loiseleur, O.; Kock, G.; Pfaltz, A.; Lefeber, C.;
Feucht, T.; Helmchen, G. Tetrahedron: Asymmetry 1994, 5, 573. (b)
J umnah, R.; Williams, A. C.; Williams, J . M. J . Synlett 1995, 821. (c)
Togni, A.; Burckhardt, U.; Gramlich, V.; Pregosin, P. S.; Salzmann, R.
J . Am. Chem. Soc. 1996, 118, 1031. (d) Burckhardt, U.; Gramlich, V.;
Hoffmann, P.; Nesper, R.; Pregosin, P. S.; Salzmann, R.; Togni, A.
Organometallics 1996, 15, 3496. (e) Burckhardt, U.; Baumann, M.;
Togni, A. Tetrahedron: Asymmetry 1997, 8, 155.
(7) (a) Canovese, L.; Visentin, F.; Uguagliati, P.; Chessa, G.; Luc-
chini, V.; Bandoli, G. Inorg. Chim. Acta 1998, 275, 385. (b) Canovese,
L.; Visentin, F.; Uguagliati, P.; Chessa, G.; Pesce, A. J . Organomet.
Chem. 1998, 566, 61.
(8) (a) Crociani, B.; Antonaroli, S.; Di Bianca, F.; Canovese, L.;
Visentin, F.; Uguagliati, P. J . Chem. Soc., Dalton Trans. 1994, 1145.
(b) Canovese, L.; Visentin, F.; Uguagliati, P.; Di Bianca, F.; Antonaroli,
S.; Crociani, B. J . Chem. Soc., Dalton Trans. 1994, 3113. (c) Canovese,
L.; Visentin, F.; Uguagliati, P.; Crociani, B.; Di Bianca, F. Inorg. Chim.
Acta 1995, 235, 45.
(9) Canovese, L.; Visentin, F.; Uguagliati, P.; Di Bianca, F.; Fontana,
A.; Crociani, B. J . Organomet. Chem. 1996, 508, 101.
(10) Antonaroli, S.; Crociani, B. J . Organomet. Chem. 1998, 560,
137.
(11) (a) Szabo`, K. J . Organometallics 1996, 15, 1128 and references
therein. (b) Oslob, J . D.; Åkermark, B.; Helquist, P.; Norrby, P. O.
Organometallics 1997, 16, 3015 and references therein.
(5) (a) Sprinz, J .; Kiefer, M.; Helmchen, G.; Reggelin, M.; Huttner,
G.; Walter, O.; Zsolnai, L. Tetrahedron Lett. 1994, 35, 1523. (b) Baltzer,
N.; Macko, L.; Shaffner, S.; Zehnder, M. Helv. Chim. Acta 1996, 79,
803. (c) Ferna`ndez-Galan, R.; J alo`n, F. A.; Manzano, B. R.; Rodr`ıguez-
de la Fuente, J .; Vrahami, M.; J edlicka, B.; Weissensteiner, W.; J ogl,
G. Organometallics 1997, 16, 3758.
(6) Crociani, B.; Antonaroli, S.; Paci, M.; Di Bianca, F.; Canovese,
L. Organometallics 1997, 16, 384.

Pd(II) Allyl Complexes with Iminophosphine Ligands
Ta ble 1. Selected 1H a n d 31P {1H} NMR Da ta a
iminophosphine protons allyl protons
NdCH other signals H2 H3s
Organometallics, Vol. 18, No. 7, 1999 1139
complex
isomerb
H1s
H1a
H3a
CH3
31P resonance
22.07 s
1a (BF4)
I
I
I
8.33 d 3.81s [OCH3]
4.05 dd
(7.0)c
3.65 dd
(9.5)c
5.84 m
5.88 m
5.74 m
3.42 d(br) 2.95 d(br)
(2.8)c
(6.0)d
(13.0)d
(7.0)d
(14.1)d
2a (BF4)
3a (BF4)
8.48e
4.02e [NCH3]
4.95 dd
(6.7)c
3.98 dd
(9.0)c
3.11 d(br)
(8.4)d
(6.7)d
(13.0)d
8.50 s
8.54 s
1.23 s [C(CH3)3]
4.95 dd
(6.9)c
3.78 dd
(10.6)c
(14.7)d
3.43 (br)
2.95 (br)
26.09 s
23.20 s
(6.9)d
4a (BF4) If
If,g
4.56 m(br) [NCH] 4.87 (br) 4.05 (br) 5.83 m
4.48 m(br) [NCH] 4.70 (br) 3.77 (br)
3.40 (br)
3.32 (br)
2.98 (br)
2.82 (br)
8.55 s
8.50 s
4.58 m [NCH]
4.45 m [NCH]
4.82 dd
(6.0)c
4.01 dd
(10.0)c
(13.5)d
3.68 dd
(10.4)c
(13.6)d
5.86 m
3.48 d(br) 3.03 d(br)
(5.0)d (12.2)d
3.35 d(br) 2.80 d(br)
(5.0)d (12.0)d
(6.0)d
4.66 dd
(6.0)c
(6.0)d
1b(BF4) II (93%)
8.39 d 3.80 s [OCH3]
5.28 t
3.12 (br)h 2.65 (br)i
0.96 dj
(10.3)c
1.28 dk
(5.9)c
25.02 s
20.37 s
23.50 s
19.07 s
31.01 s
24.87 s
25.62 s
(3.3)c
(10.0)d
III (7%)
8.29 d 3.82 s [OCH3]
mkl
3.70 dd
(9.8)c
5.60 dd
(8.2)d
0.92 dj
(9.8)c
(3.3)c
(14.5)d
(14.5)d
1.06 dk
(6.2)c
2b(BF4) II (77%)
8.56e
8.42e
3.81e [NCH3]
4.01e [NCH3]
5.40 t
2.77 (br)
2.07 dj
(11.0)c
1.59 dk
(6.0)c
(10.0)d
III (23%)
4.72 dd
(7.7)c
3.96 dd
(10.0)c
(13.9)d
5.62 dd
0.92 dj
(8.2)c
(7.7)d
1.11dk
(6.0)c
3b(BF4) II (82%)
8.55 d 1.20 s [C(CH3)3]
5.23 t
(9.7)d
2.75 (br)
2.50 (br)
2.04 dj
(11.5)c
1.50dk
(6.7)c
(3.6)c
III (18%)
8.49 d 1.26 s [C(CH3)3]
4.65 dd
(7.9)c
3.68 dd
(10.9)c
(14.1)d
5.39 dd
0.98 dj
(9.6)c
(4.9)c
(7.9)d
1.23 dk
(6.6)c
4b(BF4) II (83%)m
8.55 s
4.36 m(br) [NCH]
5.26 m(br) 2.97 (br)
2.40 (br)
1.93 dj
(10.05)c
1.54 dk
(5.6)c
III (17%)f 8.47 s
4.60 m [NCH]
4.58 m [NCH]
4.45 dd
(7.5)c
(7.5)d
mkn
4.06 dd
(10.0)c
(15.0)d
3.57 dd
(10.5)c
(14.0)d
5.60 dd
(7.6)d
mk
mk
21.29 s
20.32 s
(15.0)d
5.47 dd
(8.2)d
(14.0)d
a
In CDCl3 at 30 °C unless otherwise stated. Satisfactory integration values were obtained; coupling constants are given in Hz.
b
Abbreviations: s, singlet; d, doublet; t, triplet; dd, doublet of doublets; m, multiplet, mk, masked. See Scheme 1 for isomer structure
and numbering scheme; the percentages of isomer II and III are given in parentheses. c J (PH). J (HH). e Unresolved multiplet. f Two
d
g
h
i
diastereoisomers in 1:1 molar ratio. At -35 °C. At -35 °C the signal is a doublet with J (HH) ) 7.8 Hz. At -35 °C the signal is a
j
k
l
m
doublet with J (HH) ) 14.6 Hz. syn methyl group. anti methyl group. Masked by the signal at 3.80 ppm. Two diastereoisomers in
n
fast interconversion; at -40 °C the diastereoisomer ratio is ca. 4:1. Masked by the signal at 4.36 ppm.
B′ (or B and A′) differ from each other in the opposite
orientation of the allyl ligand relative to the rest of the
molecule. If they are both present in solution, they
should be observable in the NMR spectra, unless a fast
interconversion occurs. For R ) C6H4OMe-4, Me, and
CMe3, the isomers B and B′ are the enantiomeric forms
of A and A′, respectively, and they cannot be distin-
guished in the NMR spectra under the experimental
conditions used in this study.
For the cationic complexes 1a -3a , all having config-
1
uration I, the H NMR spectra are characterized by a
single set of proton resonances even at the lowest
temperature explored (-75 °C, in CD2Cl2). A single set
of 31P and 13C resonances is also observed at 30 °C.
These data appear to indicate either the presence of only
one pair of enantiomers A/B (or A′/B′) or the occurrence
of fast interconversion processes involving the four
isomers.

1140 Organometallics, Vol. 18, No. 7, 1999
Ta ble 2. Selected 13C{1H} NMR Da ta a
iminophosphine carbons
Crociani et al.
allyl carbons
C3
c
d
complex
isomerb
NdCH
other signals
C1
C2
CH3
1a (BF4)
I
174.98
55.65 [OCH3]
85.54 d
(27.4)
123.11
56.10
56.19
45.98
3a (BF4)
I
166.06 d
(5.9)
29.76 [C(CH3)3]
80.79 d
(30.4)
120.85 d
(6.5)
1b(BF4)
2b(BF4)
II
II
167.35 d
(4.7)
168.72 d
(6.2)
55.60 [OCH3]
53.02 [NCH3]
122.81 d
(25.4)
118.93 d
(26.7)
111.47 d
(5.2)
111.69 d
(5.7)
25.35, 20.72 d
(4.9)
26.92, 20.61 d
(5.3)
46.10 d
(4.5)
III
II
168.53 d
(3.7)
59.81 [NCH3]
73.54 d
(28.8)
mke
118.37 d
(6.0)
86.99 d
(6.6)
25.59, 20.92
3b(BF4)
166.30 d
(5.1)
30.08 [C(CH3)3]
29.91 [C(CH3)3]
110.06 d
(6.8)
40.39
28.60, 21.58 d
(5.0)
III
167.26 d
(5.1)
71.60 d
(30.6)
115.51 d
(6.5)
88.80
26.22, 21.69
a
b
In CDCl3 at 30 °C; J (PC) coupling constants (Hz) in parentheses. See Scheme 1 for isomer structure and numbering scheme. c trans
to phosphorus. cis to phosphorus. e Masked by the phenyl carbon resonances in the range 125-130 ppm.
d
Sch em e 1a
ligands, the 31P nucleus gives larger coupling constants
with 1H and 13C nuclei in trans positions.5b,5c,12 For
complex 2b, the above assignments are confirmed by
the interligand and intraligand NOEs observed for
1
isomers II and III in a 2D H NMR ROESY experiment
carried out on their equilibrium mixture in CDCl3 (Table
3). For each isomer II and III, the allylic methyl protons
exhibit long-range coupling constants of different mag-
nitudes with phosphorus. According to intraligand
NOEs, in both isomers II and III of complex 2b the syn
methyl protons are characterized by larger J (PH) values
than the corresponding anti methyl protons. Conse-
quently the related signals in the complexes containing
the η3-3-methyl-2-butenyl ligand were assigned on the
basis of the observed difference in the J (PH) values.
The variable-temperature spectra and the phase-
sensitive 2D ROESY spectrum of 2b suggest that the
cationic complexes are fluxional in solution. Two dy-
namic processes are observed: a relatively fast syn-anti
exchange of the H3s and H3a protons cis to phosphorus
and a slower P-N ligand site exchange (which may also
be viewed as a 180° rotation of the allyl ligand around
its bond axis to palladium). As can be seen in Table 1,
at 30 °C the H3s and H3a protons of complexes 1a -4a
and of isomers II of 1b-4b are generally detected as
two broad signals, which coalesce into a single broad
resonance for 2a and for isomer II of 2b. The fine
structure of these signals can be observed at lower
temperature (cf. the spectra of 4a and 1b at -35 °C).
The H3s T H3a exchange may be explained in terms of
a selective η3-η1-η3 interconversion of the allyl ligand,
which involves initial breaking of the Pd-C1 bond (trans
to P), rotation around the C2-C3 bond in the η1-allyl
transient, and re-formation of the η3 complex with
interchange of H3s and H3a protons. Such η3-η1-η3
interconversion has also been observed for other (η3-
allyl)palladium(II) complexes with bidentate P-N,5a,c
P-P,13 P-S,14 and P-O15 ligands. However, for the
isomers of type III (where the allyl terminus cis to
phosphorus is the CMe2 moiety) no syn-anti exchange
of either the H1s and H1a protons or the Me(s) and Me(a)
protons takes place at an appreciable rate, as shown
a
(R)-Bornyl ) endo-(1R)-1,7,7-trimethylbicyclo[2.2.1]hept-
2-yl.
For the 1b-3b derivatives, both cis and trans isomers
of configurations II and III are detected in solution (the
isomer II being in any case predominant). Each geo-
metrical isomer is characterized by a single set of proton
resonances in the temperature range from -35 to 30
°C in CDCl3, even though it may be present in solution
with four isomers, namely the two enantiomeric pairs
A, B and A′, B′ of Figure 1. A single set of phosphorus
and carbon resonances is also observed for each geo-
metrical isomer at 30 °C.
The assignment of the allyl resonances and of the
configurations II and III to the cis and trans isomers of
complexes 1b-4b is based essentially on the differences
in J (HH) values between the allylic protons and on the
differences in J (PH) and J (PC) values between phos-
phorus and allylic protons and carbons. It is well-
documented that in allyl complexes with phosphine
(12) (a) Powell, J .; Shaw, B. L. J . Chem. Soc. A 1967, 1839. (b)
Åkermark, B.; Krakenberger, B.; Hansson, S.; Vitagliano, A. Organo-
metallics 1987, 6, 620.

Pd(II) Allyl Complexes with Iminophosphine Ligands
Organometallics, Vol. 18, No. 7, 1999 1141
F igu r e 1. Side view from the allyl ligand toward the Pd atom for the four possible isomers of each configuration I-III.
The curved line represents the dCHC6H4- unit lying above or below the N-Pd-P plane. Legend: (i) interconversion
through a conformational change of the chelate P-N ligand (see text); (ii) interconversion through a selective η3-η1-η3
rearrangement of the allyl ligand (see text).
Ta ble 3. Selected 2D 1H Da ta a for 2b(BF 4)
On the other hand, the observation of exchange cross-
peaks between the signals of isomer II and the corre-
sponding signals of isomer III in the latter spectrum
(see Figure 2 for Me(s) and Me(a) signals) clearly indicates
that the isomers interconvert into each other through
a dynamic process which involves a P-N ligand site
exchange. Although no quantitative measurement was
carried out, the sharpness of the interconverting signals
points to a reduced rate for this process if compared to
the marked broadening of the signals caused by the
faster syn-anti exchange occurring simultaneously for
isomer II.
In a consideration of the four possible isomers of
Figure 1, the selective η3-η1-η3 process brings about
the A T B′ and A′ T B interconversions for the cationic
complexes 1a -3a and 1b-3b of configurations I and
II, respectively, while the P-N ligand site exchange
brings about the same type of interconversion for
complexes 1a -3a and the II T III interconversion for
isomer
intraligand NOEs
interligand NOEs
II
NdCH [8.56]‚‚‚NMe [3.81]
NMe [3.81]‚‚‚Me(s) [2.07]
H2 [5.40]‚‚‚Me(s) [2.07] (11.0)b NMe [3.81]‚‚‚Me(a) [1.59]
H3a + H3s [2.77]‚‚‚
Me(a) [1.59] (6.0)b
Me(a) [1.59]‚‚‚Me(s) [2.07]
III
NdCH [8.42]‚‚‚NMe [4.01]
H2 [5.62]‚‚‚H1s [4.72]
NMe [4.01]‚‚‚H1s [4.72]
NMe [4.01]‚‚‚H1a [3.96]
H2 [5.62]‚‚‚Me(s) [0.92] (8.2)b
H1a [3.96]‚‚‚Me(a) [1.11] (6.0)b
Me(a) [1.11]‚‚‚Me(s) [0.92]
a
Values in brackets are the proton chemical shifts in ppm.
b
J (PH).
by the absence of exchange cross-peaks between the H1s
and H1a signals and between the Me(s) and Me(a) signals
of isomer III in the phase-sensitive 2D ROESY spec-
trum of 2b (see Figure 2 for the methyl signal region).
(13) (a) Breutel, C.; Pregosin, P. S.; Salzmann, R.; Togni, A. J . Am.
Chem. Soc. 1994, 116, 4067. (b) Pregosin P. S.; Salzmann, R.; Togni,
A. Organometallics 1995, 14, 842. (c) Barbaro, P.; Pregosin P. S.;
Salzmann, R.; Albinati, A.; Kunz, R. W. Organometallics 1995, 14,
5160. (d) Pregosin, P. S.; Salzmann, R. Coord. Chem. Rev. 1996, 155,
35.
(14) Hermann, J .; Pregosin, P. S.; Salzmann, R.; Albinati, A.
Organometallics 1995, 14, 3311.
(15) Hosokawa, T.; Wakabayashi, Y.; Hosokawa, K.; Tsuji, T.;
Murabashi, S. I. J . Chem. Soc., Chem. Commun. 1996, 859.

1142 Organometallics, Vol. 18, No. 7, 1999
Crociani et al.
1
F igu r e 2. Phase-sensitive 2D H ROESY spectrum of 2b(BF4) in the region of the allylic methyl signals, in CDCl3 at 30
°C. Only positive NOEs are shown.
1b-3b. Preliminary crystallographic data on 1a (BF4)
have shown that in the solid state the complex is formed
by an equimolar mixture of the isomeric species B and
A′.16 The failure to detect such isomers (or those of types
interconverting with A′) and B (rapidly interconverting
with B′). As the temperature is raised, a progressive
line broadening is observed and eventually at 30 °C
some resonances coalesce into a single signal (cf. the
imino proton at 8.54 ppm) because of the increasing rate
of A T B′ and B T A′ interconversion mainly through
the η3-η1-η3 mechanism.
For 4b at -40 °C, two diastereomeric species (ca. 4:1
molar ratio) of the major isomer II and two diastereo-
meric species (ca. 1:1 molar ratio) of the minor isomer
III are observed. At 30 °C, the diastereomeric species
of isomer II interconvert so rapidly as to give rise to
broad time-averaged 1H NMR signals, whereas the
diastereomeric species of isomer III are characterized
by fairly sharp resonances, which suggest a rather low
rate of interchange. Consistently, the 31P NMR spec-
trum at 30 °C shows an intense singlet for the rapidly
interconverting diastereomeric species of isomer II and
two 1:1 singlets of lower intensity for the diastereomeric
1
B′ and A) in the H NMR spectra of 1a and of the other
cationic complexes at low temperature (when the above
dynamic processes are frozen) is very likely due to a low-
energy conformational change of the chelate P-N ligand
whereby a fast A T A′ (and B T B′) interconversion
takes place. Such a process, involving partial rotations
of the PPh2 unit (around the bonds of phosphorus with
palladium and with the carbon atom of the ortho-
disubstituted phenyl group) and of the N-R moiety
(around the Pd-N bond), has been proposed to account
for the fluxional behavior of the complexes [RhCl(CO)-
(P-N)]17 and [PdMe(OR)(P-N)],18 containing a puck-
ered six-membered chelate ring with o-(diphenylphos-
phino)-N,N-dimethylbenzylamine as the P-N ligand.
The solution behavior of complexes 4a and 4b, con-
taining a chiral (R)-bornyl group, can be well-understood
by taking into account the observed dynamic processes.
For 4a at -35 °C, the proton resonances are split into
two sets of signals of ca. 1:1 relative intensity, indicating
the presence of two diastereomeric species A (rapidly
(16) Bandoli, G.; Crociani, B., preliminary results.
(17) Rauchfuss, T. B.; Patino, F. T.; Roundhill, D. M. Inorg. Chem.
1975, 14, 652.
(18) Kapteijn, G. M.; Spee, M. P. R.; Grove, D. M.; Kooijman, H.;
Spek, A.; van Koten, G. Organometallics 1996, 15, 1405.

Pd(II) Allyl Complexes with Iminophosphine Ligands
Organometallics, Vol. 18, No. 7, 1999 1143
Ta ble 4. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for 4b(ClO4)
the coordination geometry around the metal center is
distorted square planar, with the allyl ligand markedly
rotated away from the PPh2 group, as can be seen in
the side views of the molecules from the allyl ligand
toward the Pd atom:
molecule A
molecule B
Pd-P(1)
2.257(4)
2.16(1)
2.08(1)
2.14(2)
2.43(2)
1.26(2)
1.42(2)
1.36(2)
2.248(4)
2.13(1)
2.11(2)
2.20(1)
2.39(2)
1.27(2)
1.49(2)
1.38(2)
Pd-N(1)
Pd-C(8)
Pd-C(9)
Pd-C(10)
N(1)-C(7)
C(8)-C(9)
C(9)-C(10)
P(1)-Pd-N(1)
Pd-C(8)-C(9)
Pd-C(9)-C(10)
Pd-N(1)-C(7)
Pd-P(1)-C(1)
N(1)-C(7)-C(6)
P(1)-Pd-C(10)
N(1)-Pd-C(8)
C(8)-Pd-C(10)
87.1(3)
72.8(9)
84.8(9)
125.4(9)
104.0(4)
127(1)
161.9(4)
168.2(5)
64.3(6)
88.9(3)
73.1(9)
80.1(9)
126.9(9)
106.5(4)
128(1)
160.7(4)
170.8(6)
68.5(6)
1
species of isomer III. According to the 2D H NMR data
for 2b, which rule out any η3-η1-η3 process for isomer
III, the diastereoisomers III of 4b may interconvert
through a sequence of a slow P-N ligand site exchange
(III(A or A′) T II(B′ or B)), followed by a fast η3-η1-η3
rearrangement (II(B′ or B) T II(A or A′)), and finally
by another slow P-N ligand site exchange (II(A or A′)
T III(B′ or B)). The above rearrangements (starting
from isomer II) would also explain the presence of
isomers III in the equilibrium mixture in solution, even
though isomers III are not detected in the solid state,
as revealed by the X-ray structural analysis of 4b.
X-r a y Str u ctu r e of 4b(ClO4). The solid-state struc-
ture of the (R)-bornyl complex 4b was determined by
X-ray diffraction. For this study, the perchlorate salt
The C(8) and C(9) atoms are both below the N-Pd-P
plane (0.12 and 0.44 Å in A; 0.17 and 0.38 Å in B),
whereas the C(10) atom lies above such a plane (0.53 Å
in A; 0.54 Å in B). Similar rotation of the allyl moiety
with respect to its idealized position has also been
observed in related cationic complexes with P-N chelat-
ing ligands, containing the sterically demanding η3-1,3-
diphenylallyl fragment.4c,d,5b
-
The ortho-disubstituted phenyl ring plane (C(1)-C(6))
makes a dihedral angle of 54.4° (A) and 51.1° (B) with
the N-Pd-P coordination plane, while the allyl plane
(C(8)-C(9)-C(10)) makes a dihedral angle of 103.0° (A)
and 114.0° (B) with the same plane. The allyl ligand is
asymmetrically bound to the metal, as indicated by the
Pd-C(10) bond (trans to P) being markedly longer than
the Pd-C(8) bond (trans to N), in agreement with the
higher trans influence of the phosphorus donor atom.
The bond lengths and angles in the coordination sphere
are in line with literature data for analogous [Pd(η3-
allyl)(P-N)]+ complexes, with the outstanding exception
of the Pd-C(10) bonds (2.43 Å (A) and 2.39 Å (B)) which
are significantly longer than the reported Pd-C allyl
bond lengths trans to phosphorus (2.216-2.268 Å).4c,d,5b,c
We ascribe such lengthening and the marked rotation
of the allyl group in 4b to the steric repulsion between
the allyl CMe2 terminus and the bulky (R)-bornyl group.
On the other hand, the marked asymmetry of the allyl
group in the molecules A and B of complex 4b may also
be interpreted on the basis of a prevailing ene-yl
coordination of this ligand, with the double bond trans
to the phosphorus atom. This view is consistent with
the long Pd-C(10) bond distances (2.43 and 2.39 Å in
A and B, respectively) and also with the observed
differences in the carbon-carbon bond lengths of the
allylic unit (see Table 4).29
was chosen because the ClO4 anion is generally less
affected by thermal disorder than BF4-. Some selected
bond distances and angles are listed in Table 4. An
ORTEP plot of the asymmetric unit cell is reported in
Figure 3. There are two independent molecules in the
cell, which are not superimposable, as can be seen in
Figure 4, where superimposition of the (R)-bornyl
groups brings about a nearly specular arrangement of
the central metals and the coordinated ligands. Both
molecules, however, have the same coordination geom-
etry with the CMe2 allyl terminus trans to phosphorus
(structure II of Scheme 1) and the same orientation of
the allyl ligand relative to the N-Pd-P coordination
plane, with the central allyl carbon C(9) pointing out of
this plane on the same side as the (R)-bornyl and the
pseudoaxial P-Ph groups. The planar ring of the ortho-
disubstituted phenyl group lies out of the N-Pd-P
coordination plane, on the opposite side relative to the
central allyl carbon C(9). The two molecules are there-
fore the diastereoisomers A and B of Figure 1. Thus, in
the solid state the complex 4b is present as an equimo-
lar mixture of diastereoisomers A and B, with config-
uration II, whereas in CDCl3 solution it is present as a
mixture of both geometrical isomers II and III in a II/
III molar ratio of ca. 4.9:1. At variance with the solid
state, the two diastereomeric species of isomer II,
detected in solution at -40 °C, are in a 4:1 rather than
in a 1:1 molar ratio.
Kin etic Stu d ies
The η3-propenyl complexes 1a -4a react smoothly
with an excess of the secondary amines HY in CHCl3
If one considers the inner coordination sphere as
formed by the P- and N-donor atoms of the iminophos-
phine and by the terminal allyl carbons C(8) and C(10),

1144 Organometallics, Vol. 18, No. 7, 1999
Crociani et al.
F igu r e 3. ORTEP view of the unit cell of 4b(BF4), with the two independent molecules A and B. The phenyl rings of the
PPh2 group have been indicated as PH for clarity.
at 25 °C in the presence of fumaronitrile (fn) to give the
10
zerovalent complexes [Pd(η2-fn)(P-N)]
and the cor-
responding allylamine (eq 1; HY ) piperidine, diethyl-
amine, morpholine).
+fn(exc)
+ + HY(exc)
8
[Pd(η3-C3H5)(P-N)]
-H2Y+
1a -4a
[Pd(η2-fn)(P-N)] + CH2dCHCH2Y (1)
Under the same experimental conditions, the reac-
tions of the η3-3-methyl-2-butenyl derivatives 1b, 3b,
and 4b involve either nucleophilic attack at the allyl
termini (eq 2a; HY ) diisopropylamine, diethylamine,
piperidine) and/or deprotonation of the allyl ligand to
isoprene (eq 2b), depending on the steric requirements
of the amine.
F igu r e 4. Superposition of the (R)-bornyl groups of
molecule A (bold line) and molecule B (dotted line) for the
cation 4b.
essentially through eq 2a. In the latter case, the organic
products consist of the allylamine Me2CdCHCH2NEt2,
resulting from attack at the less substituted allyl car-
bon, and a small amount of isoprene (2-5%). The reac-
tions with the less sterically demanding piperidine are
very fast and proceed exclusively through eq 2a. In this
case, however, a mixture of the regioisomeric allyl-
amines Me2CdCHCH2Y and CH2dCHCMe2Y is ob-
tained.
1
According to H NMR spectra recorded at different
times for cationic complex/fn/HY reaction mixtures in
a 1:1.2:5 molar ratio in CDCl3 at 25 °C (with an initial
complex concentration of 2.5 × 10-2 mol dm-3), the reac-
tions with the bulkier diisopropylamine are slow and
proceed predominantly through eq 2b, whereas those
with diethylamine are much faster and go to completion
The molar ratio Me2CdCHCH2Y/CH2dCHCMe2Y ap-
pears to be kinetically controlled, as it depends not only
on the different rates of nucleophilic attack at the

Pd(II) Allyl Complexes with Iminophosphine Ligands
Organometallics, Vol. 18, No. 7, 1999 1145
different allylic termini of the η3-3-methyl-2-butenyl
group but also on the subsequent isomerization process
CH2dCHCMe2Y f Me2CdCHCH2Y, the rate of which
decreases with increasing concentration of fumaronitrile
and with decreasing temperature. This is clearly shown
by the results of reaction 2a for complex 3b under
different experimental conditions. When the reactants
are mixed in 3b/fn/piperidine ratios of 1:1.2:5 and 1:10:5
at 25 °C, the Me2CdCHCH2Y/CH2dCHCMe2Y ratios
(measured after 30 min, when the reaction is complete)
are 2:1 and 1:1, respectively. When the same reactants
are mixed in a 1:5:5 ratio at -30 °C, the observed
Me2CdCHCH2Y/CH2dCHCMe2Y ratio is 1:5 and re-
mains constant throughout the reaction.
Ta ble 5. Ra te Con sta n ts k2 a n d k3 for Rea ction s 1
a n d 2a in CHCl3 a t 25 °C
b
complex
aminea
k2 (mol-1dm3 s-1
)
k3 (mol-2 dm6 s-1
)
1a
pip
dea
morph
1.9 ( 0.1
0.21 ( 0.02
0.23 ( 0.02
72 ( 22
3.4 ( 1
12 ( 3
2a
3a
4a
1b
3b
pip
dea
0.45 ( 0.02
9 ( 2
0.032 ( 0.004
0.5 ( 0.1
pip
dea
3.4 ( 0.1
0.29 ( 0.02
39 ( 15
3 ( 1
pip
dea
3.9 ( 0.2
0.40 ( 0.04
93 ( 23
4.2 ( 1
pip
dea
0.031 ( 0.001
0.0018 ( 0.0005
Similar reactions involving deprotonation of η3-bound
allyl ligands or isomerization of allylamines have been
previously observed in the amination of various η3-
allylpalladium complexes.19
pip
dea
0.116 ( 0.003
0.011 ( 0.001
4b
pip
dea
0.076 ( 0.002
0.0076 ( 0.0002
1
a
From the H NMR studies of reactions 1 and 2a the
Abbreviations: Pip, piperidine; dea, diethylamine; morph,
b
morpholine. k3 ) k2′K (see text).
following features are also apparent: (i) the rate of the
η3-η1-η3 process for complexes 1a -4a and for isomer
II of complexes 1b, 3b, and 4b is greatly increased in
the presence of the amine HY; (ii) the II/III molar ratio
is not affected by the amine and remains constant
throughout the reaction, whereas the II T III inter-
conversion rate is slightly increased in the presence of
the amine; (iii) the chelate P-N ligand is not displaced
by the entering amine.
Sch em e 2
For quantitative kinetic measurements the course of
reactions 1 and 2a was followed by monitoring UV/vis
spectral changes (λ, 500-200 nm) of CHCl3 solutions of
the complexes ([Pd]0 ) 1 × 10-4 mol dm-3) in the
presence of fn ((2-8) × 10-4 mol dm-3) upon addition
of variable aliquots of excess HY (1 × 10-3-0.1 mol
dm-3). Under such pseudo-first-order conditions the
reactions went smoothly to completion, as indicated by
comparison of solution spectra after 7-8 half-lives with
those of the final products independently prepared.10
The conversion to the zerovalent [Pd(η2-fn)(P-N)] com-
plexes appears to obey the customary monoexponential
absorbance (A) vs time (t) relationship At ) A∞ + (A0 -
A∞) exp(-kobst). The pseudo-first-order rate constants
kobs were determined by nonlinear regression of absor-
bance At data to time. No dependence of the rates on
the fumaronitrile concentration could be detected in the
range investigated. For reaction 1 the kobs values fit the
two-term second- and third-order rate law
solvent, temperature, and concentration conditions.8 On
the other hand, the third-order rate law, observed for
the amination of [PdCl(η3-3-methyl-2-butenyl)(PPh3)]
with Me2NH, was ascribed to a slow deprotonation of
the intermediate formed in the rapid initial attack of
the amine on the allyl ligand.20 Since such a third-order
term is absent in rate law (4) and in that observed in
the reaction of [Pd(η3-allyl)(L-L′)]+ (L-L′ ) 2-(thio-
methyl)pyridine) with the same amines HY under the
same conditions,7 the above interpretations are to be
ruled out and an alternative mechanism is to be devised.
kobs ) k2 [HY] + k3[HY]2
(3)
1
In light of H NMR data we propose the mechanism
of Scheme 2, where the cationic substrate undergoes
nucleophilic attack by the entering amine at the allyl
ligand (step k2) or at the central metal (fast equilibrium
K), forming the intermediate [Pd(allyl)(P-N)(HY)]+ in
which the allyl moiety is attacked by a further amine
molecule (step k2′). Although such an intermediate could
not be detected in any 1H NMR spectrum of the reaction
mixtures, its formation (in low concentration) is sug-
gested by the rate increase for both the η3-η1-η3 and
II T III processes in the presence of the amine HY.
Accordingly, the intermediate may be conceived of as a
labile five-coordinate transient in equilibrium with an
η3-allyl species containing a P-monodentate iminophos-
whereas for reaction 2a only the second-order term is
observed:
kobs ) k2[HY]
(4)
The values of constants k2 and k3 are listed in Table 5.
The presence of the third-order term k3 in the rate
law, observed for the reaction of [Pd(η3-allyl)(L-L′)]+
(L-L′ ) 2-(iminomethyl)pyridine) with secondary amines
HY, was initially interpreted as resulting from a bimo-
lecular attack by a hydrogen-bonded amine dimer in
equilibrium with the monomer under the prevailing
(19) (a) Åkermark, B.; Vitagliano, A. Organometallics 1985, 4, 1275.
(b) Åkermark, B.; Zetterberg, K.; Hansson, S.; Krakenberger, B.;
Vitagliano, A. J . Organomet. Chem. 1987, 335, 133.
(20) Vitagliano, A.; Åkermark, B. J . Organomet. Chem. 1988, 349,
C22.

1146 Organometallics, Vol. 18, No. 7, 1999
Crociani et al.
phine ligand and/or with an η1-allyl species containing
a bidentate iminophosphine ligand. The k2 and k2′ steps
yield palladium(0) products with an η2-bound allylam-
monium group, which are rapidly converted to the final
derivative [Pd(η2-fn)(P-N)] by the more π-accepting
fumaronitrile ligand.
4a may occur at both terminal allylic carbons (i.e., trans
to P and trans to N) unless one of the two carbons is
preferentially attacked on steric grounds. Therefore, the
k2 values in Table 5 should be considered as the sum of
the two contributions for 1a -3a , while for 4a each
diastereoisomer will contribute to the experimental
overall value. Moreover, due allowance being made for
the different structural features of the P-N ligands, the
selectivity in the nucleophilic attack at the allylic carbon
trans to P, observed in the palladium-catalyzed asym-
metric allylic amination with chiral ferrocenyl pyra-
zole-phosphine ligands,4c is likely to be caused by steric
rather than electronic factors.
On the other hand, the higher reaction rates observed
for 1a -4a compared to those for complexes [Pd(η3-
C3H5)(N-N′)]+ are mainly due to the lack of displace-
ment of the P-N ligands by the secondary amines that
we have used in this work. In fact, in the reactions of
complexes [Pd(η3-C3H5)(N-N′)]+ the allylic amination
is accompanied by reversible displacement of the N-N′
ligand by amines:8
According to this mechanism, the rate law should be
k2[HY] + k2′K[HY]2
kobs
)
(5)
1 + K[HY]
which reduces to eq 3 with k2′K ) k3 if the term K[HY]
is much lower than 1, as is expected from the inability
to detect appreciable concentrations of the intermediate
[Pd(allyl)(P-N)(HY)]+. In this context the observed rate
law 4 is a particular case of eq 3 when k2′K[HY] , k2.
Table 5 shows that the k2 term increases with
increasing basicity and decreasing steric demands of the
amine, in agreement with our previous kinetic studies.7,8
A marked decrease is observed on going from η3-
propenyl complexes 1a -4a to the corresponding η3-3-
methyl-2-butenyl derivatives 1b, 3b, and 4b, as a result
of increased steric hindrance and decreased electrophilic
character of the allyl fragment brought about by the
methyl substituents.
KE
[Pd(η3-C3H5)(N-N′)]+ + 2HY y z
[Pd(η3-C3H5)(HY)2]+ + N-N′ (6)
A notable influence on k2 is also exerted by the imino
nitrogen substituent R of the P-N ligand. As can be
seen in Table 5, the k2 term increases with increasing
electron-withdrawing ability of R: cf. 1a (R ) C6H4OMe-
4) vs 2a (R ) Me). Surprisingly, with the bulkier and
more electron donating CMe3 and (R)-bornyl groups (3a
and 4a , respectively) higher k2 values are observed. A
similar effect is detected on going from 1b to 3b and
4b. These findings can be rationalized by invoking some
sort of distortion of the η3-bound allyl ligand in the
substrates 3a , 3b and 4a , 4b, which leads to an early,
product-like transition state with marked allyl carbon
reactivity. This ground-state allylic distortion is pur-
ported to play a major role in palladium-catalyzed
asymmetric allylic aminations.4c,21 Such interpretation
is further supported by the marked distortion of the
allylic ligand observed in the solid-state structure of
complex 4b(ClO4) (vide supra). Accordingly, the amina-
tion rates of complexes [Pd(η3-allyl)(L-L′)]+ (L-L′ )
2-(thiomethyl)pyridine) were found to increase with
increasing bulkiness of the pyridine group.7
Accordingly, the following experimental rate law was
observed:
k2[HY] + k3[HY]2
kobs
)
(7)
[HY]2
1 + K
E[N-N′]
where the ratio 1/(1 + KE[HY]2/[N-N′]) represents the
fraction of starting substrate that is reactive toward the
amination, the bis(amine) displacement product [Pd(η3-
C3H5)(HY)2]+ being unreactive.8
For complexes 1b and 3b, for which the isomer ratio
II/III remains constant throughout the reaction, the
nucleophilic attack by diethylamine may take place at
the unsubstituted allyl carbon of both isomers, at rates
lower than that of interconversion. Thus, the observed
k2 values are again to be considered as the sum of the
two contributions, unless one isomer is by far preferred
for steric reasons. For 4b the contribution to the
experimental k2 value is even more complicated by the
presence of two diastereoisomers for each structure of
types II and III. On the other hand, the k2 values for
the reaction of 1b, 3b, and 4b with piperidine should
be taken as merely indicative, since the amination
occurs at both allyl carbon termini for each isomeric
species of the substrate and is further complicated by
isomerization of the allylamine products.
On the basis of the proposed mechanism, the qua-
dratic k3 term is the product of the equilibrium constant
K and the second-order rate constant k2′. Provided the
K values are low, detection of the k3 term for the η3-
propenyl complexes 1a -4a implies in any case high
values for k2′. The enhanced reactivity of the intermedi-
ate [Pd(allyl)(P-N)(HY)]+ may arise from a markedly
distorted allyl group in a five-coordinate species. In this
context, the absence of the quadratic k3 term in the
reactions of 1b, 3b, and 4b could be explained by a
When the k2 data gathered so far for allyl amination
of complexes [Pd(η3-C3H5)(L-L′)]+ by diethylamine,
piperidine, and morpholine are compared, the following
ranges are observed: L-L′ ) 2-(iminomethyl)pyridine
(N-N′), 1.2 × 10-2-0.4;8 iminophosphine (P-N), 3.2 ×
10-2-3.9 (this work); 2-(thiomethyl)pyridine (N-S),
7
0.2-10.0 mol-1 dm3 s-1
.
These values correspond to
the following gross reactivity order of L-L′: N-S >
P-N > N-N′. It is noteworthy that replacement of a
pyridine group in N-N′ by a tertiary phosphine group
in P-N does not bring about any outstanding increase
in the k2 term, at variance with what expected from the
higher trans influence of the phosphine group, which
should render the allylic carbon trans to P more
susceptible to nucleophilic attack.11 The modest increase
of the k2 values on going from the N-N′ to the P-N
complexes suggests that the nucleophilic attack on 1a -
(21) Blo¨chl, P. E.; Togni, A. Organometallics 1996, 15, 4125.

Pd(II) Allyl Complexes with Iminophosphine Ligands
Organometallics, Vol. 18, No. 7, 1999 1147
Ta ble 6. Elem en ta l An a lysis Da ta (%)a
Ta ble 7. Cr ysta l Da ta a n d Deta ils of Da ta
Collection for 4b(ClO4)
complex
C
H
N
formula
space group
cryst syst
fw
a (Å)
b (Å)
C34H41ClNO4PPd
P1
triclinic
T (°C)
λ (Å)
25(2)
0.710 73
2
1b(BF4)
2b(BF4)
3b(BF4)
4b(BF4)
56.8 (56.60)
53.2 (53.08)
55.1 (55.33)
59.5 (59.36)
4.7 (4.75)
4.9 (4.81)
5.6 (5.47)
6.1 (6.01)
2.1 (2.13)
2.4 (2.48)
2.2 (2.30)
2.1 (2.04)
Z
700.5
F(000)
V (Å3)
Dcalcd (g/cm3)
724
10.460(6)
10.938(5)
15.785(7)
97.05(4)
107.15(4)
104.58(4)
1632(1)
1.426
7.37
0.041
0.104
1.053
a
Calculated values in parentheses.
c (Å)
µ (cm-1
)
R (deg)
â (deg)
γ (deg)
Ra
considerable decrease in the k2′ values, which parallels
the corresponding decrease observed for k2 values.
a
Rw
GOFb
a
2
2
2
R ) ∑(|Fo| - |Fc|)/∑(|Fo|); Rw ) [∑[w(|Fo| - |Fc| )]2/∑(w|Fo| )2]1/2
.
b
2
2
GOF ) [∑w(|Fo| - |Fc| )2/(Nobs - Npar)]1/2
.
Exp er im en ta l Section
Ma ter ia ls. The iminophosphines,10 the dimers [Pd(µ-Cl)-
(η3-allyl)]2 (allyl ) C3H5 and 1,1-Me2C3H3),22 and the complexes
1a (BF4)-4a (BF4)10 were prepared by published methods. All
other chemicals were commercial grade and were purified or
dried by standard methods, when required.23 The complexes
1b(BF4)-4b(BF4) were prepared from the reaction of [Pd(µ-
Cl)(η3-1,1-Me2C3H3)]2 (0.25 mmol) with the appropriate imi-
nophosphine P-N (0.5 mmol) in CH2Cl2 (20 mL), followed by
addition of an excess of NaBF4 (2 mmol) dissolved in 10 mL of
methanol. The reaction mixture was worked up as previously
described for 1a (BF4)-4a (BF4),10 to give the required products
(yields in the range 68-87%, based on the theoretical amount),
which were purified by reprecipitation from a CH2Cl2/Et2O
solvent mixture. Because of the straightforward method of
preparation, the cationic complexes can be completely char-
acterized by their multinuclear NMR spectral data. However,
elemental analysis, IR spectra in the solid state, and molar
conductivity measurements were also carried out in order to
Utrecht, The Netherlands. Elaboration of the spectrum was
carried out on a Digital Graphic workstation. A sine bell
apodization function, shifted by π/3, was applied in both
dimensions. A zero filling was applied in t1 in order to obtain
a real 1K × 1K matrix.
X-r a y St r u ct u r a l An a lysis of 4b(ClO4). Intensity data
were collected on a Siemens Nicolet R3m/V diffractometer
using the ω-2θ scan mode. The unit cell was determined by
the automatic indexing of 50 centered reflections. Crystal
decay was negligible, and correction was deemed unnecessary,
while an empirical absorption correction based on Ψ-scans was
applied (Tmax ) 0.951, Tmin ) 0.356). A total of 5155 reflections
were collected in the range 4 < 2θ < 48°, of which 4722
reflections with I > 2σ(I) were used for the structure deter-
mination. The structure was solved by the Patterson method
using SHELXTL/PC25 and refined by the full-matrix least-
squares method on F2 using SHELXL-93.26 All non-hydrogen
atoms were refined anisotropically, and the positions of the
hydrogen atoms were located geometrically and constrained
to ride on the carbon atoms to which they are bonded with
their thermal parameters fixed at values of 1.2 or 1.5 (for
methyl groups) of their parent atoms. Their contributions were
added to the structure factor calculations, but their positions
were not refined. The chirality was defined from the Flack
coefficient, 0.03(4).27 The final difference map was featureless,
apart from some peaks (up to 0.888 e Å-3) in the vicinity of
-
ascertain the composition, the presence of the BF4 anion (ν-
(B-F) around 1060 cm-1), and the nature of uni-univalent
electrolytes (molar conductivities in the range 91-98 Ω-1 cm2
mol-1 for 1 × 10-3 mol L-1 MeOH solutions at 25 °C). The
elemental analysis data of the η3-2-methyl-2-propenyl com-
plexes are given in Table 6. The complex 4b(ClO4) was
prepared similarly, using NaClO4‚H2O instead of NaBF4 (yield
1
90%). The H and 31P NMR spectra of 4b(ClO4) were found to
be identical with those of 4b(BF4). Clear, yellow crystals of
4b(ClO4) suitable for X-ray analysis were obtained by slow
diffusion of diethyl ether into a dilute dichloromethane solu-
tion.
-
the ClO4 tetrahedrons. The crystal data and details of data
collection are listed in Table 7.
Kin etic Mea su r em en ts. The kinetics of allyl amination
were studied by addition of known aliquots of amine solution
to a solution of the complex under study and fumaronitrile in
the thermostated cell compartment of a Perkin-Elmer Lambda
40 spectrophotometer (25 °C). Mathematical and statistical
data analysis was carried out on a personal computer by means
of a locally adapted version of Marquardt’s algorithm written
in TURBOBASIC (Borland).28
1
NMR Mea su r em en ts. The H,13C{1H}, and 31P{1H} NMR
spectra were recorded on a Bruker AM 400 spectrometer,
operating at 400.13, 100.61, and 161.98 MHz, respectively.
Chemical shifts (ppm) are given relative to Me4Si (1H and 13C
NMR) and 85% H3PO4 (31P NMR). The amination reactions
were monitored by recording 1H NMR spectral changes of
CDCl3 solutions, prepared by dissolving 0.05 mmol of the allyl
complex and the appropriate amounts of amine and fumaroni-
trile in 2 mL of the solvent. The 1H 2D ROESY spectrum of
2b(BF4) in CDCl3 was obtained in the phase-sensitive mode
using the TPPI phase cycle with the ROESY pulse sequence,
modified to eliminate the offset dependence of cross-peak
intensity.24 A total of 128 transients on a size of 2K were
accumulated in the phase-sensitive mode for 512 experiments.
A spin-lock period corresponding to a transverse mixing time
of 0.3 s was applied. Data were processed with the 2D NMR
program TRITON, licensed by the University of Utrecht,
Ack n ow led gm en t. Financial support by the Italian
Ministero per l’Universita` e la Ricerca Scientifica e
Tecnologica is gratefully aknowledged.
Su p p or tin g In for m a tion Ava ila ble: Tables of complete
crystallographic experimental details, positional parameters
for all atoms, bond distances and angles, anisotropic thermal
parameters, and hydrogen atom coordinates. This material is
available free of charge via the Internet at http://pubs.acs.org.
OM980858L
(22) (a) Hartley, F. R.; J ones, S. R. J . Organomet. Chem. 1974, 66,
469. (b) Powell, J .; Shaw, B. L. J . Chem. Soc. A 1967, 1839.
(23) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon Press: NewYork, 1988.
(24) (a) Marion, D.; Wuthrich, K. Biochem. Biophys. Res. Commun.
1983, 113, 964. (b) Bothner-By, A. A.; Stephens, R. L.; Lee, J .; Warren,
C. D.; J eanloz, R. W. J . Am. Chem. Soc. 1984, 105, 811. (c) Griesinger,
C.; Ernst, R. R. J . Magn. Reson. 1987, 75, 261.
(25) Sheldrick, G. M. SHELXTL/PC Version 5.03; Siemens Analyti-
cal X-ray Instruments Inc., Madison, WI, 1994.
(26) Sheldrick, G. M. SHELX-93; University of Gottingen, Gottingen,
Germany, 1993.
(27) Flack, H. D. Acta Crystallogr. 1983, A39, 876.
(28) Marquardt, D. W. SIAM J . Appl. Math. 1963, 11, 431.
(29) We thank Prof. P. S. Pregosin for this suggestion.