An investigation into the allylic imidate rearrangement of trichloroacetimidates catalysed by cobalt oxazoline palladacycles

Article abstract of DOI:10.1002/chem.200700873

Dimeric palladacycles, di-μX-bis[({η⁵-(S)-( _pR)-2-[2′-(4′-methylethyl)oxazolinyl]cyclopentadienyl, 1-C.3′-N}(η⁴-tetraphenylcyclobutadiene)cobalt]dipalladium (COP-X), containing bridging groups X = OAc, Cl, Br, I, O₂

Full text of DOI:10.1002/chem.200700873

DOI: 10.1002/chem.200700873
An Investigation into the Allylic Imidate Rearrangement of
Trichloroacetimidates Catalysed by Cobalt Oxazoline Palladacycles
Hiroshi Nomura and Christopher J. Richards*[a]
Abstract: Dimeric palladacycles, di-m-
X-bis
[{h5-(S)-(pR)-2-[2’-(4’-methylethy-
variant of X (93–96%) and the yield
increased in the sequence I<p-
O2CC6H4F<OAc <O2CCF3 ꢀBrꢀCl.
With X=Cl (COP-Cl), the catalyst
loading was reduced to 0.25 mol%
(CH3CN/708C/48 h) and these condi-
tions applied to various trichloroaceti-
CH2CH=CH2, CH2OTBDMS) to give
the corresponding (S)-trichloroaceta-
mides (68–88% yield, 84–94% ee; ee=
enantiomeric excess). Addition of
COP-Cl to triphenylphosphinobenzoyl
NovaGel AM resin gave a recyclable
catalyst in which the ee was maintained
over three cycles (89–94%). Catalysis
with COP-OAc displayed a small posi-
tive non-linear effect. The factors re-
sponsible for the activity of COP-X are
discussed.
ACHTREUNG
l)oxazolinyl]cyclopentadienyl,1-C,3’-
N}(h4-tetraphenylcyclobutadiene)co-
balt]dipalladium (COP-X), containing
bridging groups X=OAc, Cl, Br, I,
O2CCF3, p-O2CC6H4F, were synthesised
and compared as catalysts for the
asymmetric allylic imidate rearrange-
midates
(R=nPr,
Me,
CH2Ph,
ment of (E)-Cl3CCACHTRE(UGN =NH)OCH2CH=
Keywords: allylic compounds
·
CHR with R=nPr. The enantiomeric
excess of the product (S)-Cl3CC(=
O)NHCHRCH=CH2 was essentially in-
asymmetric catalysis · ligand effects ·
metallocenes · palladium
Introduction
together with the measured activation parameters, were con-
sistent with a concerted pericyclic process. Alternatively, ad-
dition of mercury(II) salts resulted in rate accelerations of
up to 1012 for a reaction formulated as proceeding by means
of a two-step iminomercuration–deoxymercuration mecha-
nism. The discovery that palladium(II) complexes are supe-
rior catalysts for this reaction[2] led to the development of
chiral non-racemic palladium(II) species as catalysts for the
asymmetric allylic imidate rearrangement.[3] A major break-
through in the practicality of this process was the discovery
that cobalt oxazoline palladacycle 2b (COP-Cl),[4] readily
derived from 1 by diastereoselective palladation and ligand
exchange (Scheme 2),[5] gave high enantioselectivities (typi-
cally >90% ee; ee=enantiomeric excess) for the rearrange-
The allylic imidate rearrangement was first reported by
Overman in 1974 as a method for the conversion of allylic
alcohols into higher value allylic amides and amines
(Scheme 1).[1] This [3.3]-sigmatropic rearrangement was ini-
Scheme 1. The allylic imidate rearrangement.
tially carried out by thermolysis of trichloroacetimidic esters
of primary and secondary alcohols at a temperature of
1408C. The resulting regio- and stereoselectivities observed,
[a] H. Nomura, Dr. C. J. Richards
School of Biological and Chemical Sciences
Queen Mary, University of London
Mile End Road, London, E1 4NS (UK)
Fax : (+44)207-882-7427
E-mail: c.j.richards@qmul.ac.uk
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Scheme 2. Diastereoselective synthesis of cobalt oxazoline palladacycles.
10216
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10216 – 10224

FULL PAPER
ment of both (E)- and (Z)-trifluoroacetimidates (R=F, R1 =
p-MeOC6H4).[6] Crucially, and in contrast to previous appli-
cations of chiral palladacycle catalysts to the allylic imidate
rearrangement, the halide-bridged palladacycle 2b does not
require activation by addition of one equivalent of silver tri-
fluoroacetate. A subsequent report by Anderson and Over-
man that 2b also catalyses the rearrangement of trichloroa-
cetimidates (R=Cl, R1 =H) with high enantioselectivies[7,8]
is especially significant due to the relative ease with which
the product trichloroacetamide may be hydrolysed to re-
lease an allylic amine. Palladcycles 2a and 2b have also
been successfully applied as enantioselective catalysts of re-
lated reactions,[9] and 2a as a reagent in enantioselective
transcyclopalladation.[10]
The pure complex 2e was then obtained by recrystallisation.
Finally the trifluoroacetate derivative 2 f was obtained by
addition of one equivalent of silver trifluoroacetate to 2b
followed by filtration to remove the resulting precipitate of
silver chloride. Although this[6] and related trifluoroacetate-
bridged oxazoline-based palladacycles[3f,k] have previously
been generated and used in situ, we found the isolated com-
plex to be an air-stable non-hygroscopic complex, in
common with all the related COP-X derivatives.
In addition to these palladium dimers, we also synthesised
a monomeric complex (Scheme 4). Addition of one equiva-
In light of the observed differences in reactivity of these
palladacycles as a function of the bridging ligand X, we
were curious to investigate the effectiveness of other deriva-
tives with the aim of minimising the catalyst loading of
asymmetric trichloroacetimidate rearrangement. In addition,
we anticipated that this and related studies may shed light
on the factors responsible for catalyst activity. Our investiga-
tions to this effect are reported in this paper.[11]
Scheme 4. Synthesis of palladacycle–triphenylphosphine adducts 3.
lent of triphenylphosphine to 2b resulted in the rapid and
essentially quantitative formation of adduct 3b. The formu-
lation of this as containing the ligated triphenylphosphine
cis to the carbon–palladium bond is made by analogy to re-
lated non-metallocene aryl analogues.[13] In addition, the
chemical shift of one of the cyclopentadienyl hydrogens in
the 1H NMR spectrum is smaller by approximately d=1–
1.5 ppm compared to the corresponding signal for 2b. This
difference is due to the proximity of the cis-coordinated
phosphine to the hydrogen alpha to palladium, the upfield
shift resulting from the positive anisotropic effect of the ad-
jacent phenyl groups.
Results and Discussion
Synthesis of COP-X derivatives: A variety of COP-X deriv-
atives were readily synthesised from 2a and 2b (Scheme 3).
Bromide and iodide-containing compounds, 2c and 2d, re-
Application of COP-X derivatives to the allylic imidate re-
arrangement: To compare the effectiveness of these pallada-
cycles as allylic imidate rearrangement catalysts, complexes
2a–f and 3b, in addition to the related palladacycles 4,[8b]
Scheme 3. Synthesis of additional derivatives of COP-X.
spectively, were obtained in an analogous fashion to 2b by
simply adding the acetate-bridged complex 2a to an aque-
ous acetone solution of the appropriate sodium halide.
After stirring at room-temperature, the products were isolat-
ed by filtration. In common with chloride 2b, the bromide
and iodide-bridged complexes were obtained as mixtures of
cis/trans isomers with respect to the alignment of the two
carbon–nitrogen palladacycle chelates about the central (m-
Cl)2Pd2 moiety.[12] The p-fluorobenzoate derivative 2e was
obtained by dissolving 2a and one equivalent of p-fluoro-
benzoic acid in dichloromethane, and driving the ligand
metathesis reaction towards completion by repeatedly wash-
ing with water to remove the more water soluble acetic acid.
5
[14] and 6,[15] were applied as catalysts for the rearrangement
of trichloroacetimidate 7a (R=nPr, Scheme 5). Acetonitrile
was chosen as the solvent for this investigation for two rea-
Chem. Eur. J. 2007, 13, 10216 – 10224
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10217

C. J. Richards and H. Nomura
in acetonitrile. In this respect, the hexafluoroacetylacetonate
derivative 4, which had previously been synthesised to in-
crease catalyst solubility,[8b] offered no particular advantage
over its synthetic precursors, either 2a or 2b (entry 8). The
triphenylphosphine adduct 3b gave a significantly reduced
yield despite a longer reaction time (entry 7). A notable fea-
ture of all these reactions is the consistently high enantio-
meric excesses obtained (92–96%), revealing that enantiose-
lectivity is essentially independent of the other ligand(s) co-
ordinated to the palladacycle C–N chelate.
Scheme 5. Palladacycle-catalysed rearrangement of trichloroacetimidates
7.
Finally, the two related palladacycles 5 and 6 were tested
for comparison by using the same conditions. The former,
which contains the opposite configuration of planar chirality
(S,pS-5 compared to S,pR-2b) resulted in a low yield of the
opposite enantiomer of amide 8a (entry 9). That the planar
chirality is important to the enantioselectivity displayed by
2b is supported by the result obtained with 6. This was only
marginally less selective in catalysing the formation of the
same enantiomer of product as 2b, despite the remoteness
of the element of central chirality from palladium. However
the low yield obtained with 6 precluded further investiga-
tions with this palladacycle.
sons. Firstly, the boiling point (828C) is higher than that of
dichloromethane (408C), the solvent generally used for this
transformation.[7,16] It was anticipated that a higher boiling
point could help to shorten the reaction time when low cata-
lyst loadings were employed. Secondly, it was reasoned that
this coordinating solvent[17] would aid the stability of the pal-
ladacycle at higher reaction temperatures.
Initially the new and known palladacycles were compared
by employing a catalyst loading of 0.5 mol% (i.e. 1 mol%/
Pd) in acetonitrile at 508C for 24 h (Table 1). This revealed
Having identified that catalyst 2b can be successfully used
in acetonitrile, we next examined lowering the catalyst load-
ing to 0.25 mol% and compensating for this by increasing
the reaction temperature (Table 2). Examination of the
Table 1. Catalysis of the rearrangement of trichloroacetimidate 7a to
amide 8a (R=nPr).[a]
Entry
Cat.
t [h]
Yield [%][b]
ee [%][c]
AHCTREUNG
(config.)[c]
1
2
3
4
5
6
7
8
9
2a (X=OAc)
2b (X=Cl)
2c (X=Br)
2d (X=I)
24
24
24
24
24
24
48
24
48
24
61
72
69
11
28
70
33
79
25
9
94 (S)
93 (S)
96 (S)
93 (S)
94 (S)
93 (S)
92 (S)
94 (S)
28 (R)
86 (S)
Table 2. Optimisation of time and temperature for the rearrangement of
trichloroacetimidate 7a to amide 8a (R=nPr) catalysed by 0.25 mol% of
2b (COP-Cl).[a]
Entry
t [h]
T [8C]
Yield [%][b]
ee [%][c]
ACHTREUNG
(config.)[c]
2e (X=p-O2CC6H4F)
2 f (X=O2CCF3)
3b[d]
4[d]
5
1
2
3
4
5
6
24
48
24
48
24
48
60
60
70
70
80
80
65
71
67
95 (S)
90 (S)
93 (S)
92 (S)
89 (S)
92 (S)
10
6
79
63[d]
85[d]
[a] Compound 7a (2.6 mmol), catalyst (0.5 mol%), CH3CN (1 mL), 508C.
[b] Isolated yield after chromatography. [c] Determined by HPLC analy-
sis. [d] 1 mol% catalyst.
[a] Compound 7a (2.6 mmol), 2b (0.25 mol%), CH3CN (1 mL). [b] Iso-
lated yield after chromatography. [c] Determined by HPLC analysis.
[d] Calculated yield based on mass recovery, but HPLC analysis revealed
the presence of an unidentified impurity.
a small difference in yield between the acetate 2a and chlo-
ride 2b-bridged palladacycles (entries 1 and 2). The bro-
mide-bridged species 2c (entry 3) was similar to 2b, and
both were superior to the corresponding iodide 2d (entry 4)
such that the relationship between the halogen and the per-
centage yield obtained is ClꢀBr>I. The para-fluoroben-
zoate-bridged species 2e was relatively ineffective (entry 5),
and the trifluoroacetate derivative 2 f gave a similar yield to
both 2b and 2c (entry 6). This latter result was surprising in
light of the previous requirement to convert chloride bridg-
ed oxazoline-based palladacycles into their trifluoroacetate
congeners in order to achieve catalysis.[3f,k] The lower than
expected yield obtained with 2 f may be a consequence of
the relative insolubility of this catalyst under the conditions
employed, as apart from 2 f, all of the other catalysts com-
pletely dissolved once the substrate allylic imidate (2.6m)
had been added to a solution/suspension of the palladacycle
crude reaction mixture by 1H NMR spectroscopy revealed
the reaction to be incomplete after both 24 and 48 h at 608C
(entries 1 and 2) and 24 h at 708C (entry 3), but that a reac-
tion time of 48 h at 708C resulted in essentially complete
conversion (entry 4) and a yield of 79% after isolation by
column chromatography. Increasing the temperature to
808C further increased the rate of conversion as determined
1
by H NMR spectroscopy, but resulted in a reduction in the
isolated yield due to the formation of an unidentified by-
product (entries 5 and 6). A notable feature of these results
is the consistency of the enantiomer ratio of 8a over the
range of temperatures employed, revealing that the differen-
ces in free energy between the two transition states
10218
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10216 – 10224

Trichloroacetimidates
FULL PAPER
(DDGÀ=2.17 KcalmolÀ1 at 708C) must be dominated by
the entropy term TDDS°.
The lower catalyst loading of 0.25 mol% was then applied
to a variety of other trichloroacetimidates to determine the
general usefulness of these new conditions (Table 3). Methyl
A solid-supported catalyst: On completion of this larger-
scale reaction with COP-Cl 2b, column chromatography was
used to recover a mixture of yellow metallocene based com-
pounds which contained
a significant amount of 2b
(ꢀ30%). This prompted us to generate a solid-supported
derivative with the potential to be more easily recycled and
reused. Addition of triphenylphosphinobenzoyl NovaGel
AM resin (0.4–0.7 mmolgÀ1)[18] (triphenylphosphine Nova-
Gel) to 2b in acetonitrile led to the isolation of 10, the poly-
mer supported analogue of the triphenylphosphine adduct
3b (Scheme 7). The loading was determined by microanaly-
Table 3. Catalysis of the rearrangement in Scheme 5 by 0.25 mol% of 2b
(COP-Cl).[a]
Entry Substrate (R)
t (h) Product/yield [%][b] ee [%]
(config.)[c]
AHCTREUNG
1
2
3
4
5
6
7b (R=Me)
48
48
8b/82
8c/68
8d/68
8e/72
8 f/31
8a/21
89 (S)[c]
90 (S)[d]
84 (S)[d]
94 (S)[c]
0
7c (R=CH2Ph)
7d (R=CH2CHCH2)[e] 72
7e (R=CH2OTBDMS) 48
7 f (R=Ph)[f]
9
96
72
67 (R)[c]
[a] Compound 7 or 9 (2.6 mmol), 2b (0.25 mol%), CH3CN (1 mL), 708C
[b] Isolated yield after chromatography. [c] Determined by HPLC analy-
sis. [d] Assigned by analogy. [e] Additional 2b (0.5 mol%) added after
48 h and the reaction maintained at 708C for a further 24 h. [f] 2b
(0.5 mol% plus additional 0.5 mol% added after 48 h and the reaction
maintained at 708C for a further 48 h).
Scheme 7. Synthesis of a solid-supported palladacycle.
(entry 1), benzyl (entry 2), allyl (entry 3) and protected hy-
droxymethyl (entry 4)-substituted trichloroacetimidates 7b–
e all gave the corresponding amides with enantioselectivities
similar to, or just slightly less than, that obtained with 7a.
The additional double bond in substrate 7d inhibited the re-
action requiring the use of a higher catalyst loading. The
phenyl-substituted trichloroacetimidate 7 f resulted in a low
yield and no enantioselectivity (entry 5). It has been noted
previously that a tertiary or phenyl substituent dramatically
slows the reaction, as does the use of (Z)-trichloroacetimi-
dates.[7] Although an example of the latter (9) gave (R)-8a
in 67% ee (entry 6), it is clear that the inherent lack of reac-
tivity of these substrates could not be overcome.
sis to be approximately 0.6 mmolgÀ1. As the catalytic activi-
ty of 3b is somewhat less than that of 2b, the catalysis po-
tential of 10 for the rearrangement of trichloroacetimidate
7a was tested with a loading of approximately 1.7 mol% of
phosphine coordinated palladacycle monomer (Table 4).
After heating at 508C for 48 h, the product 8a was isolated
Table 4. Sequential application of 10 to the catalysed rearrangement of
trichloroacetimidate 7a to amide 8a (R=nPr).[a]
Run
t [h]
Yield [%][b]
ee [%]
ACHTREUNG
(config.)[c]
The applicability of the new conditions to a larger scale
reaction was also examined (Scheme 6). Heating 4.84 g of
1
2
3
48
48
48
91
27
17
94 (S)
93 (S)
89 (S)
[a] Compound 7a (2.6 mmol), 10 (0.120 g), CH3CN (1 mL). [b] Isolated
yield after chromatography. [c] Determined by HPLC analysis.
in 91% yield and 94% ee (entry 1). After filtering and wash-
ing the polymer it was again employed for the rearrange-
ment of 7a (entry 2), and although the yields of 8a obtained
in this and a subsequent run (entry 3) were rather low, the
enantioselectivities obtained were consistently high. These
results suggest that alternative modes of immobilisation, not
involving deactivating palladium coordination, may lead to
practical recyclable catalysts for the allylic imidate rear-
rangement and related reactions.
Scheme 6. Larger-scale rearrangement of 7a with 0.25 mol% 2b (COP-
Cl).
7a with just 72.4 mg (i.e. 0.25 mol%) of COP-Cl 2b led to
the isolation of the product amide in 88% yield and
92% ee. Use of 0.25 mol% of COP-OAc 2a with 4.73 g of
7a resulted in a 62% isolated yield and an 89% ee. These
results highlight the superior activity of the chloride-bridged
palladacycle compared to the acetate analogue.
Rationalisation of reactivity trends: A mechanism for the
palladium(II) and mercury(II)-catalysed rearrangement of
allylic imidates has been proposed by Overman (Scheme 8).
Chem. Eur. J. 2007, 13, 10216 – 10224
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10219

C. J. Richards and H. Nomura
Unlike ferrocene-based palladacycles, these cobalt sys-
tems do not require activation by addition of one equivalent
of silver trifluoroacetate. The need for this in the ferrocene
series may be due to the higher lability (with respect to
ligand substitution) of the introduced trifluoroacetate group,
but it could also be due to the oxidation of the ferrocene
into a ferrocenium ion.[24] It has been noted that the cyclo-
pentadienyl ligand of (h4-tetraphenylcyclobutadiene)(h5-cy-
clopentadienyl)cobalt is electron poor compared to ferro-
Scheme 8. Cyclisation-induced rearrangement mechanism.
1
cene, as revealed by a comparison of H NMR spectroscopic
data, and the values of pKa and pKb for hydroxy and amino-
substituted metallocenes respectively.[25] The electronic
properties of the nitrogen ligand also appears to be impor-
tant, as the higher activity observed with 2b compared to 6
is likely due to differences between the coordinating nitro-
gen of the oxazoline and imidazole groups respectively, as
reflected in the lower basicity of the former.[26]
The evidence for cyclisation-induced rearrangement (CIR)
catalysis in this and other [3,3] sigmatropic rearrangements
has been reviewed[19] and followed its initial proposal by
Henry as the mechanistic explanation of palladium acetate
catalysed intramolecular allylic ester rearrangements.[20] Use
of a chiral trichloroacetimidate substrate (S)-12 (88% ee)
with the achiral catalyst [PdCl
2A
We then considered whether the bridging groups, X, sig-
nificantly influence the Lewis acidity of the metal in the pal-
ladacycles studied for catalyst activity. To investigate this,
the same method used for the synthesis of triphenylphos-
phine adduct 3b was used to generate small-scale samples of
oxazoline complexes 3a, 3c, 3d and 3 f (Scheme 4), together
with a corresponding complex 11 derived from the imida-
zole analogue 6. Examination of their 31P NMR spectra re-
vealed little difference in the chemical shifts of the coordi-
nated phosphorus atoms (Table 5), and no correlation be-
in the formation of (S,E)-13, also of 88% ee (Scheme 9).[2d]
Table 5. 31P NMR spectroscopic chemical shifts for triphenylphosphine
coordinated palladacycles.
Entry
Compound
Chemical shift [ppm]
1
2
3
4
5
6
3a (X=OAc)
3b (X=Cl)
3c (X=Br)
3d (X=I)
3 f (X=O2CCF3)
11
32.88
33.87
34.02
33.40
32.49
34.02
tween chemical shift and catalyst activity. This suggests that
there is little difference between the Lewis acidity of a chlo-
ride versus, for example, an iodide coordinated palladacycle.
Instead, to account for the different yields observed with
COP-X it was noted that the leaving group ability of the
halogens in associative substitution reactions of square-
planar palladium(II) complexes increases in the order Cl>
Br>I.[27] Although the differences are relatively small, the
nature of the entering group having a larger effect, the rate
Scheme 9. The possible pathways for the conversion of (S)-13 into (S,E)-
13.
There are two possible CIR mechanistic explanations of this
outcome. In pathway (i), initial palladium coordination to
give diastereoisomer A must be followed by anti-iminopalla-
dation and anti-deoxypalladation to give (S,E)-13. In path-
way (ii), formation of diastereoisomer B requires subsequent
syn-iminopalladation and syn-deoxypalladation. Studies on
the addition of amines[21] and other nucleophiles[22] to palla-
dium coordinated alkenes have revealed these Wacker-type
reactions to proceed with anti selectivity, outcomes that
point to (i) as the pathway for allylic imidate rearrange-
ment.[23] In support of this is the all-equatorial disposition of
the substituents in the chair-like intermediate (e.g. A’) aris-
ing from (E)-allylic imidates.
of substitution of [Pt
ACHTRE(UGN dien)Cl]Cl is about three and a half
times faster than [Pt
AHCTREUNG
electron deficiency of cobalt metallocenes appears to be im-
portant for the reactivity they display, the comparison of a
series of COP-X derivatives points to the leaving group abil-
ity of the bridging ligand as another factor influencing activ-
ity.
To study the reaction further, COP-Cl was dissolved in
1
CD3CN and examined by H NMR spectroscopy. The pres-
10220
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10216 – 10224

Trichloroacetimidates
FULL PAPER
ence of cis and trans isomers (ratio=1:0.5, it is not known
which is which) revealed the maintenance of the chloride-
bridged dimeric structures in this solvent. Following addition
of COP-Cl 2b to 7a (0.016m) in CD3CN (7a/2b 3:1) and
standing of the resulting solution at room temperature over-
1
night, H NMR spectroscopy revealed a 75% conversion of
7a into 8a, with unchanged COP-Cl 2b (isomer ratio=
1:0.7) and no other palladacycle species present. This indi-
cates that under these conditions, the resting state of the cat-
alyst in the presence of the substrate and the product is as
the chloride-bridged dimer.
The formation of dimeric and higher-order structures by
bridging ligands such as halogens or alkoxides is associated
with striking non-linear effects (NLEs) in asymmetric catal-
ysis.[29] The possibility of this occurring in the palladacycle-
catalysed allylic imidate rearrangement was examined with
the acetate-bridged complex 2a rather than the correspond-
ing chloride 2b in order to avoid possible complications aris-
ing from the presence of cis and trans isomers in the latter.
For the conversion of 7a into 8a, the use of COP-OAc with
enantiomeric excesses of 25 and 50% resulted in slightly
higher values of product ee than calculated by linear extrap-
olation of the value of 89% ee obtained with enantiopure
2a (Table 6). The origin of this outcome was investigated by
Scheme 10. Substitution chemistry in related palladacycles and platinacy-
cles.[31]
the former with 2,6-dimethylpyridine gave a monomeric pal-
ladium complex with the chloride trans to the metal–carbon
bond. This stereochemistry is the same as displayed by 3,
and has been observed with other monomeric adducts of
palladacycles containing nitrogen donor ligands.[13,32] In con-
trast, the platinacycle 16 gave the monomeric adduct with
the chloride cis to the metal–carbon bond. This outcome, in
which the bridging chloride ligand trans to the metal–carbon
bond has been substituted, is in agreement with the estab-
lished order of kinetic trans effects.[33] Prolonged reaction of
this kinetic product with 2,6-dimethylpyridine resulted in its
conversion to the isomeric thermodynamic product.[31] That
the kinetic product was not observed in the analogous palla-
dium chemistry is consistent with the much greater kinetic
lability of palladium compared to platinum complexes in
ligand substitution and isomerisation reactions.
Table 6. Catalysis of the rearrangement of trichloroacetimidate 7a to
amide 8a with non-enantiopure 2a (COP-OAc).[a]
Entry Cat. ee
8a found ee
[%]
8a expected ee
Conv.
[%][d]
[%][b]
[%][c]
1
2
3
4
100
75
50
89
66
55
39
–
–
>99
85
67
45
22
The crystal structure of 4[8b] reveals the palladium–oxygen
bond lengths of the hexafluoroacetylacetonate ligand trans
to carbon and nitrogen as 2.102(4) and 2.020 Š(4) respec-
tively. Thus, the metallocene ligand has a larger trans influ-
ence than the oxazoline and most likely a larger trans effect
given: 1) the frequent correlation between these thermody-
namic and kinetic effects and 2) the similarity of this C–N
ligand to that in palladacycle 14. Thus, the COP-X dimers
react by introduction of either a solvent or substrate mole-
cule to give a monomeric species initially containing this
substituent trans to the metallocene ligand.[34] The lability of
the bridging ligands X to substitution influences the yield
obtained in a COP-X-catalysed allylic imidate rearrange-
ment.
25
75
[a] Compound 7a (2.6 mmol), 2a (0.25 mol%), CH3CN, (1 mL), 708C,
48 h. [b] Obtained by combining (S,pR)-2a and (R,pS)-2a in appropriate
amounts. [c] From linear extrapolation of 89%. [d] Determined by
1H NMR spectroscopy.
examining the 1H NMR spectrum of racemic COP-OAc
which was found to be identical to the spectrum of enantio-
pure COP-OAc. This may be accounted for by either the ab-
sence of the S,pR/R,pS (meso) diastereoisomer, or by it pro-
ducing an identical spectrum to the S,pR/S,pR (or R,pS/R,pS)-
acetate-bridged dimer. Furthermore, no additional signals
1
were observed in the H NMR spectrum of 50% ee COP-
OAc compared to enantiopure COP-OAc when the spectra
were recorded in CD3CN in the presence of excess 7a. The
observed small positive NLE may be accounted for by the
operation of either an (ML)2 system or the reservoir effect
model, both of which require the formation of a meso
isomer.[30] The former is consistent with dimeric COP-Cl as
the actual catalyst, the later with an unobserved monomeric
solvated species as the starting point of the catalytic cycle.
The substitution chemistry of chloride-bridged nitrogen
coordinating palladacycles and platinacycles has been stud-
ied with complexes 14 and 15 (Scheme 10).[31] Reaction of
Conclusion
The screening of a range of COP-X-dimeric-bridged palla-
dacycles as catalysts for the allylic imidate rearrangement of
a trichloroacetimidate revealed an optimum group X to be
chloride, the yield of the product trichloroacetamide increas-
ing in the series I<p-O2CC6H4F<OAc <O2CCF3 ꢀBrꢀCl.
In contrast, the enantioselectivity was independent of the
identity of X. With acetonitrile as the solvent, the enantiose-
Chem. Eur. J. 2007, 13, 10216 – 10224
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10221

C. J. Richards and H. Nomura
0.3H), 3.27–3.22 (m, 1H), 3.30 (t, 1H, J=9.3 Hz), 3.39 (t, 0.3H, J=
8.1 Hz), 4.22–4.25 (m, 1H+0.3H), 4.38 (brt, 0.3H), 4.46 (brt, 1H), 4.64
(d, 0.3H, J=2.2, Hz), 4.68 (d, 1H, J=2.2 Hz), 4.95 (brs, 1H), 5.06 (brs,
0.3H), 7.16–7.26 (m, 12H+3.6H), 7.55–7.71 ppm (m, 8H+2.4H);
13C NMR (CDCl3): d=14.1, 19.0, 30.1, 66.8, 71.5, 76.3, 76.4, 80.6, 85.4,
86.7, 91.2, 103.1, 126.6, 128.4, 129.5, 129.6, 135.5, 135.7, 171.5 ppm; IR
(KBr): n˜ =3050, 2950, 1603 (C=N), 1500, 1365, 1180 cmÀ1; MS (MALDI):
m/z (%): 1648.0 [M+]; elemental analysis calcd (%) for
C78H66Co2I2N2O2Pd2·2H2O: C 55.63, H 4.19, N 1.66; found: C 55.80, H
3.97, N 1.58.
lectivity of the COP-Cl-catalysed reaction was essentially in-
variant over the temperature range 50–808C, and with a cat-
alyst loading of only 0.25 mol%, the optimum yield was ob-
tained at 708C with a reaction time of 48 h. These condi-
tions were successfully applied to a range of trichloroacet-
imidates to give the product amides (S)-Cl3CC(=
O)NHCHRCH=CH2 in good yield and enantioselectivity
(85–94% ee). The potential for catalyst recycling was dem-
onstrated by combination of COP-Cl with a polymer sup-
ported phosphine, triphenylphosphine NovaGel and isola-
tion of the resulting adduct by filtration on completion of
the reaction. The activity of the COP-Cl catalyst is ascribed
to the relatively good leaving group ability of the chloride
ligand, and to the relative electron deficiency of the cobalt
metallocene and oxazoline components of the chelating pal-
ladium ligand. A small positive non-linear effect was ob-
served with COP-OAc.
[{h5-(S)-(pR)-2-[2’-(4’-methylethyl)oxazolinyl]-
Di-m-4-fluorobenzoylatobisCAHTREUNG
cyclopentadienyl,1-C,3’-N}(h4-tetraphenylcyclobutadiene)cobalt]dipalla-
dium (2e): p-Fluorobenzoic acid (0.057 g, 0.4 mmol) was added in one
portion to a solution of 2a (0.30 g, 0.2 mmol) in CH2Cl2 (3 mL). The solu-
tion was stirred overnight at room temperature, washed with water (2”
3 mL), dried (Na2SO4), filtered and the solvent removed in vacuo to give
2e as an orange crystalline solid (0.31 g, 93%). M.p. 185–1878C; [a]D22
=
+865 (c=0.164 in CHCl3); 1H NMR (CDCl3): d=0.03 (d, 3H, J=
6.6 Hz), 0.33 (d, 3H, J=7.1 Hz), 3.16 (m, 1H), 1.78 (m, 1H), 3.52 (t, 1H,
J=9.4 Hz), 4.09–4.14 (m, 1H), 4.30 (t, 1H, J=2.5 Hz), 4.41 (d, 1H, J=
1.5 Hz), 4.66 (d, 1H, J=2.0 Hz), 6.97, (t, 2H, J=8.8 Hz), 7.10–7.26 (m,
12H), 7.58–7.73 ppm (m, 10H); 13C NMR (CDCl3): d=13.2. 18.7, 28.9,
65.3, 71.1, 75.7, 78.5, 83.3, 85.6, 86.8, 99.2, 114.1, 114.5, 126.1, 128.0, 129.2,
131.6, 132.2, 135.9, 171.7, 174.3 ppm; IR (KBr) n˜ =3060, 2950, 1600 (C=
N), 1575, 1495, 1395, 1365, 1220, 1180, 1150, 1065 cmÀ1; MS (FAB): m/z
(%): 1672.4 [M+]; elemental analysis calcd (%) for C92H74Co2F2N2O6Pd2:
C 66.08, H 4.46, N 1.68; found: C 66.03, H 4.66, N 1.56.
Experimental Section
General: Dichloromethane and acetonitrile were distilled from calcium
hydride under an atmosphere of nitrogen. Petroleum ether refers to that
fraction boiling in the range of 40–608C. Column chromatography was
performed on silica gel (40–63 mm). All reactions were performed under
an atmosphere of nitrogen. All NMR spectra were recorded on a Jeol
JNM-EX 270 MHz spectrometer. Optical rotations were measured on a
Jasco P-1010 instrument and IR spectra were recorded on a Shimadzu
FTIR-8300 spectrometer. Melting points are uncorrected. Mass spectra
were recorded by the EPSRC National Mass Spectrometry Service
Centre and elemental analyses were performed at University College,
London. Compounds 2a,[5a] 2b[6] 4,[8b] 5,[14] 6[6] 7a,[1b] 7b,[1b] 7e,[7] 7 f,[1b]
8a,[1b] 8b,[35] 8e,[7] 8 f[1b] and 9[9b] have been previously described. The en-
antiomeric excesses of 8a, 8b and 8e were determined as previously re-
ported.[7]
[{h5-(S)-(pR)-2-[2’-(4’-methylethyl)oxazolinyl]cy-
Di-m-trifluoroacetatobisACHTREUNG
clopentadienyl,1-C,3’-N}(h4-tetraphenylcyclobutadiene)cobalt]dipalladi-
um (2 f): Silver trifluoroacetate (0.09 g, 0.41 mmol) was added to a solu-
tion of 2b (0.30 g, 0.2 mmol) in CH2Cl2 (6 mL) in one portion and the re-
action mixture was stirred overnight at room temperature. The resulting
precipitate of silver chloride was removed by suction filtration and the
solvent removed in vacuo to give 2 f as a brown glassy solid (0.15 g, 44%
yield). M.p. 232–2358C (decomp); [a]2D2 =+712 (c=0.092 in CHCl3);
1H NMR (CDCl3): d=À0.03 (d, 3H, J=6.6 Hz), 0.44 (d, 3H, J=7.1 Hz),
1.58–1.63 (m, 1H), 2.90 (dt, 1H, J=8.6, 2.7 Hz, 3.37 (t, 1H, J=9.1 Hz),
4.09 (dd, 1H, J=8.3, 3.7 Hz), 4.26 (t, 1H, J=2.7 Hz), 4.72 (d, 1H, J=2.0,
Hz), 4.79 (d, 1H, J=1.7 Hz), 7.17–7.28 (m, 12H), 7.56–7.59 ppm (m,
8H); 13C NMR (CDCl3) d=12.9, 13.0, 18.2, 29.4, 64.6, 71.3, 76.6, 79.5,
84.1, 84.5, 85.7, 96.7, 104.2, 126.4, 128.1, 129.1, 135.3, 171.0 ppm; IR
(KBr): n˜ =2960, 1680, 1600 (C=N), 1510, 1450, 1365, 1200, 1145 cmÀ1; MS
(FAB): m/z (%): 809.1 [1/2M+]; elemental analysis calcd (%) for
C82H66Co2F6N2O6Pd2·CH2Cl2: C 58.47, H 4.02, N 1.64; found: C 58.13, H
3.94, N 1.53.
Di-m-bromobisACHTREUNG
[{h5-(S)-(pR)-2-[2’-(4’-methylethyl)oxazolinyl]cyclopenta-
dienyl,1-C,3’-N}(h4-tetraphenylcyclobutadiene)cobalt]dipalladium (2c):
Aqueous sodium bromide (2m, 0.5 mL, 1 mmol) was added to a flask
containing 2a (0.15 g, 0.1 mmol) and acetone (1 mL) and the resulting
heterogeneous mixture was stirred at room temperature for 4 h. The re-
sulting yellow solid was filtered, washed with water (2.1 mL) and acetone
(0.3 mL) and dried in vacuo to give 2c as a mustard-coloured solid
(0.16 g, >99%). M.p. 209–2108C; [a]2D2 =+1137 (c=0.218 in CHCl3);
exists as a 1.0:0.7 mixture of dimers in CDCl3; 1H NMR (CDCl3): d=
0.67–0.71 (m, 3H+2.1H), 0.74 (d, 3H, J=6.8 Hz), 0.80 (d, 2.1H, J=
7.0 Hz), 2.10–2.33 (m, 1H+0.7H), 3.00–3.15 (m, 1H+0.7H), 3.30 (t, 1H,
J=9.0 Hz), 3.40 (t, 0.7H, J=8.9 Hz), 4.16–4.22 (m, 1H+0.7H), 4.28 (t,
0.7H, J=2.5 Hz), 4.41 (t, 1H, J=2.5 Hz), 4.66 (d, 0.7H, J=2.0 Hz), 4.69
(d, 1H, J=2.0 Hz), 4.96 (d, 1H, J=1.5 Hz), 5.00 (d, 0.7H, J=1.5 Hz),
7.16–7.28 (m, 12H+8.4H), 7.56–7.66 ppm (m, 8H+5.6H); 13C NMR
(CDCl3): d=14.0, 14.2, 18.8, 29.1, 29.3, 65.7, 71.2, 71.3, 76.3, 76.6, 80.3,
84.2, 84.5, 85.1, 85.2, 87.1, 88.2, 99.9, 126.2, 128.0, 129.2, 129.3, 135.2,
135.3, 170.8 ppm; IR (KBr): n˜ =3060, 2960, 1602 (C=N), 1500, 1365,
1180 cmÀ1; MS (MALDI): m/z (%): 1554.1 [M+]; elemental analysis
calcd (%) for C78H66Br2Co2N2O2Pd2·H2O: C 59.60, H 4.36, N 1.78; found:
C 59.24, H 4.24, N 1.71.
[{h5-(S)-(pR)-2-[2’-(4’-methylethyl)oxazolinyl]cyclopentadienyl,1-
ChloroACHTREUNG
C,3’-N}(h4-tetraphenylcyclobutadiene)cobalt]triphenylphosphinepalladi-
um (3): Triphenylphosphine (0.051 g, 0.19 mmol) was added to a solution
of 2b (0.14 g, 0.10 mmol) in CH2Cl2 (7.5 mL) in one portion and the reac-
tion mixture was stirred for 3 h at room temperature. The solvent was re-
moved in vacuo to give 3 as a red–brown glassy solid (0.18 g, 94%). M.p.
1
>2508C; [a]2D2 =+761 (c=0.18 in CHCl3); H NMR (CDCl3): d=0.79 (d,
3H, J=7.0 Hz), 0.82 (d, 3H, J=7 Hz), 2.87 (m, 1H), 3.13 (d, 1H, J=
1.5 Hz), 3.40 (t, 1H, J=8.6 Hz), 3.67 (dt, 1H, J=9.1, 3.7 Hz), 4.25 (t, 1H,
J=4.2 Hz), 4.30 (t, 1H, J=2.4 Hz), 4.65 (d, 1H, J=1.7 Hz), 7.14–
7.49 ppm (m, 35H); 13C NMR (CDCl3) d=14.0, 19.0, 28.8, 66.3, 71.4,
75.3, 80.3, 84.7, 87.6, 126.3, 127.9, 128.1, 129.1, 130.3, 131.5, 132.2, 135.1,
135.3, 135.4, 171.5 ppm; 31P NMR (CDCl3): d=33.87 ppm; IR (KBr): n˜ =
3050, 2950, 1605 (C=N), 1495, 1435, 1365, 1165 cmÀ1; MS (FAB): m/z
(%): 993.1 [M+]; elemental analysis calcd (%) for C57H48ClCoNOPPd: C
68.82, H 4.86, N 1.41; found: C 68.52, H 4.71, N 1.29.
Di-m-iodobisACHTREUNG
[{h5-(S)-(pR)-2-[2’-(4’-methylethyl)oxazolinyl]cyclopenta-
dienyl,1-C,3’-N}(h4-tetraphenylcyclobutadiene)cobalt]dipalladium (2d):
By using the same procedure for the synthesis of 2c, 2a (0.15 g,
0.1 mmol) and aqueous sodium iodide (2m, 0.5 mL, 1 mmol) gave 2d as a
mustard-coloured solid (0.16 g, 96%). M.p. 230–2328C; [a]2D2 =+1325
(c=0.196 in CHCl3); exists as a 1.0:0.3 mixture of dimers in CDCl3;
1H NMR (CDCl3): d=0.65 (d, 0.9H, J=7.1 Hz), 0.72 (d, 3H+0.9H, J=
6.6 Hz), 0.80 (d, 3H, J=6.9 Hz), 2.40–2.19 (m, 1H+0.3H), 3.11–3.07 (m,
General method for the palladacycle-catalysed conversion of trichloro-
acetimidate 7a into trichloroacetamide 8a: Trichloroacetimidate 7a
(0.634 g, 2.6 mmol), the palladacycle catalyst (0.5 or 0.25 mol%) and ace-
tonitrile (1 mL) were added to a round-bottomed flask containing a stir
bar. The flask was sealed with a polyethylene stopper and placed in an
oil bath preheated to the required temperature (50–808C). After either
24 or 48 h the flask was removed from the heating bath, cooled to room
10222
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10216 – 10224

Trichloroacetimidates
FULL PAPER
temperature and concentrated in vacuo to give a brown oil. Column
chromatography (3% EtOAc/hexanes) gave 8a as a pale-yellow oil.
at room temperature. Removal of the solvent in vacuo gave 10 (0.120 g)
as brown beads: elemental analysis (%) NovaGel found: C 73.48, H 7.62,
N 1.23; calcd for NovaGel+0.6 mmol COP-Cl: C 70.57, H 6.68, N 1.44;
found: C 70.77, H 6.46, N 1.68.
(E)-2,2,2-Trichloroacetimidic acid but-4-phenyl-2-enyl ester (7c): A solu-
tion of trans-4-phenyl-2-buten-1-ol[36] (1.90 g, 12.8 mmol) and DBU
(DBU=1,8-diazabicyclo
A
Conversion of trichloroacetimidate 7a into trichloroacetamide 8a with
supported catalyst 10: Trichloroacetimidate 7a (0.73 g, 3.0 mmol), the po-
lymer supported catalyst 10 (120 mg, 0.05 mmol of palladacycle mono-
mer) and acetonitrile (1 mL) were added to a round-bottomed flask con-
taining a stir bar. The flask was sealed with a polyethylene stopper and
placed in an oil bath preheated to 508C. After 48 h, the flask was re-
moved from the heating bath, cooled to room temperature and the super-
natant decanted from the polymer which was washed with additional ace-
tonitrile (4 mL). Following combination of the acetonitrile solutions, the
rearranged product 8a was isolated as described above. After drying, the
product was used in two further sequential reactions under the same con-
ditions. Run 1 (8b: 91% yield, 94% ee); run 2 (27% yield, 93% ee);
run 3 (17% yield, 89% ee).
CH2Cl2 (97 mL) was cooled to 08C, and to this was added dropwise tri-
chloroacetonitrile (1.92 mL, 19.2 mmol), maintaining the temperature of
the reaction mixture below 58C. After stirring for 1 h at 08C, the solvent
was removed in vacuo and the residue column chromatographed (neutral
alumina, 2% EtOAc/petroleum ether) to give 7c as a yellowish oil
1
(2.76 g, 74%). H NMR (CDCl3): d=3.45 (d, 2H, J=6 Hz,), 4.82 (d, 2H,
J=6 Hz,), 5.79 (m, 1H), 6.06 (m, 1H), 7.35–7.21 (m, 5H), 8.33 ppm (brs,
1H); 13C NMR (CDCl3): d=38.7, 69.6, 91.5, 124.6, 126.3, 128.60, 128.69,
135.0, 139.6, 162.6 ppm; IR (film) n˜ =3340, 3030, 2940, 1660, 1490, 1450,
1380, 1290, 1080, 980, 910 cmÀ1
(E)-2,2,2-Trichloroacetimidic acid hex-2,5-dienyl ester (7d): A solution of
(E)-hexa-2,5-dien-1-ol[37] (0.22 g, 2.2 mmol) and DBU (0.07 mL,
0.5 mmol) in CH2Cl2 (11 mL) was cooled to 08C, and to this was added
dropwise trichloroacetnitrile (0.33 mL, 3.3 mmol), maintaining the tem-
perature of the reaction mixture below 58C. The resulting orange solu-
tion was stirred for 1 h at 08C. The solvent was removed in vacuo and
the residue column chromatographed (neutral alumina, 2% EtOAc/pe-
troleum ether) to give 7d as a colourless oil (0.18 g, 33%). 1H NMR
(CDCl3): d=2.83 (t, 2H, J=5.4 Hz), 4.75 (d, 2H, J=8.9 Hz), 5.03 (dd,
1H, J=8.9, 1.7 Hz), 5.05 (dd, 1H, J=18.7, 1.7 Hz), 5.70 (m, 1H), 5.83
(m, 2H), 8.27 ppm (brs, 1H); 13C NMR (CDCl3): d=36.3, 69.7, 91.7,
115.9, 124.3, 134.0, 135.8, 162.6 ppm; IR (film): n˜ =3344, 2947, 1662, 1450,
General method for the synthesis of 3a, 3c, 3d, 3 f and 11: Triphenyl-
phosphine (5.3 mg. 0.02 mmol) was added to a solution of either 2a, 2c,
2d, 2 f or 6 (0.01 mmol) in CH2Cl2 (0.7 mL) and the resulting solution
was stirred at room temperature for 3 h prior to the removal of the sol-
vent in vacuo.
1
Compound 3a (X=OAc): H NMR (CDCl3): d=0.73 (d, 3H, J=7.1 Hz),
0.80 (d, 3H, J=6.6 Hz), 1.42 (s, 1H), 2.38–2.30 (m, 1H), 3.21 (t, 1H, J=
8.9 Hz), 3.55 (d, 1H, J=1.9 Hz), 3.82–3.75 (m, 1H), 4.12 (dd, 1H, J=8.6,
4.9 Hz), 4.38 (brt, 1H), 4.61 (d, 1H, J=2.4 Hz), 7.17–7.47 ppm (35H, m);
31P NMR (CDCl3): d=32.88 ppm.
Compound 3c (X=Br): 1H NMR (CDCl3): d=0.79 (d, 3H, J=4.0 Hz),
0.81 (d, 3H, J=4 Hz), 2.89 (m, 1H), 3.07 (d, 1H, J=1.4 Hz), 3.33 (t, 1H,
J=9 Hz), 3.73 (m, 1H), 4.26 (t, 1H, J=4.0 Hz), 4.30 (brt, 1H), 4.66 (d,
1H, J=1.9 Hz), 7.13–7.41 ppm (m, 35H); 31P NMR (CDCl3): d=
34.02 ppm.
Compound 3d (X=I): 1H NMR (CDCl3): d=0.79 (d, 3H, J=3.4 Hz),
0.81 (d, 3H, J=3.4 Hz), 2.92 (m, 1H), 3.08 (d, 1H, J=1.9 Hz), 3.23 (t,
1H, J=9 Hz), 3.86 (m, 1H), 4.29 (t, 1H, J=3.2 Hz), 4.31 ppm (brt, 1H),
4.66 (d, 1H, J=1.9 Hz), 7.19–7.48 ppm (m, 35H); 31P NMR (CDCl3) d=
33.39 ppm.
1290, 1070, 975, 920, 830, 795, 650 cmÀ1
General method for the COP-Cl-catalysed conversion of trichloroacet-
imidates 7b–c,e–f into trichloroacetamides 8b–c,e–f: Trichloroacetimi-
date 7 (2.6 mmol), (S)-COP-Cl (9.5 mg, 6.5 mmol) and acetonitrile (1 mL)
were added to a round-bottomed flask containing a stir bar. The flask
was sealed with a polyethylene stopper and placed in an oil bath preheat-
ed to 708C. After 48 h, the flask was removed from the heating bath,
cooled to room temperature and concentrated in vacuo to give a brown
oil. Column chromatography (3% EtOAc/hexanes) gave 8 as a pale-
yellow oil.
Compound 3 f (X=OCOCF3): 1H NMR (CDCl3): d=0.72 (d, 3H, J=
7.1 Hz), 0.81 (d, 3H, J=6.9 Hz), 2.27–2.21 (m, 1H), 3.19 (t, 1H, J=
9.1 Hz), 3.66–3.62 (m, 2H), 4.12 (dd, 1H, J=8.6, 4.6 Hz), 4.41 (brt, 1H),
4.66 (d, 1H, J=2.7 Hz), 7.44–7.16 ppm (m, 35H); 31P NMR (CDCl3): d=
32.49 ppm.
Compound 11: 1H NMR (CDCl3): d=0.65 (d, 3H, J=6.9 Hz), 0.72 (s,
9H), 3.10 (d, 1H, J=2.2 Hz), 3.49 (q, 1H, J=6.7 Hz), 3.89 (d, 1H, J=
2.5 Hz), 4.24 (t, 1H, J=2.2 Hz), 7.14–7.49 ppm (m, 37H); 31P NMR
(CDCl3): d=34.02 ppm.
(S)-2,2,2-Trichloro-N-(1-benzylallyl)acetamide (8c): Isolated as a pale-
yellow oil (0.52 g, 68%). 94% ee; [a]2D2 =+17.4 (c=0.228 in CHCl3);
1H NMR (CDCl3): d=2.92 (dd, 1H, J=13.6, 6.6 Hz), 3.01 (dd, 1H, J=
13.8, 6.6 Hz), 4.76–4.67 (m, 1H), 5.18 (dd, 1H, J=17.7, 1.2 Hz), 5.19 (dd,
1H, J=10.3, 1.2 Hz), 5.85 (ddd, 1H, J=16.5, 10.8, 5.4 Hz), 6.56 (brs,
1H), 7.33–7.17 ppm (m, 5H); 13C NMR (CDCl3): d=40.5, 54.1, 92.7,
116.5, 127.1, 128.7, 129.5, 135.8, 136.0, 161.1 ppm; IR (film): n˜ =3300,
3060, 2920, 1685, 1650, 1520, 1440, 1410, 1350, 1260, 1085, 1015, 990, 880,
820 cmÀ1; MS (APCI): m/z (%): 309 (95) [M++NH3]; elemental analysis
calcd (%) for C12H12Cl3NO: C 49.26, H 4.13, N 4.79; found: C 49.40, H
4.21, N 4.64.
(S)-2,2,2-Trichloro-N-[1-(2-propenyl)allyl]acetamide (8d): Trichloroacet-
imidate 7d (0.200 g, 0.82 mmol), (S)-COP-Cl 2b (3.0 mg, 2 mmol) and
acetonitrile (0.31 mL) were added to a round-bottomed flask containing
a stir bar. The flask was sealed with a polyethylene stopper and placed in
an oil bath preheated to 708C. After 48 h, additional (S)-COP-Cl 2b
(3.0 mg, 2 mmol) was added and heating at 708C was maintained for a
further 24 h. After cooling to room temperature, the solvent was re-
moved in vacuo and the residual brown oil column chromatographed
(silica gel, 3% EtOAc/hexane) to give 8d as a pale-yellow oil (0.08 g,
Acknowledgement
We thank the EPSRC National Mass Spectrometry Service Centre, Uni-
versity of Wales, Swansea.
[1] a) L. E. Overman, J. Am. Chem. Soc. 1974, 96, 597–599; b) L. E.
Overman, J. Am. Chem. Soc. 1976, 98, 2901–2910; c) L. E. Over-
man, Acc. Chem. Res. 1980, 13, 218–224.
1
40%). 94% ee; [a]2D2 =+14.0 (c=0.428 in CHCl3); H NMR (CDCl3): d=
2.50–2.33 (m, 2H), 4.56–4.47 (m, 1H), 5.16 (dd, 1H, J=15.4, 1 Hz), 5.17
(dd, 1H, J=10.7, 1 Hz), 5.22 (dd, 1H, J=16.9, 1 Hz), 5.24 (dd, 1H, J=
6.6, 1 Hz), 5.89–5.68 (m, 2H), 6.63 ppm (brs, 1H); 13C NMR (CDCl3):
d=38.5, 52.2, 92.8, 116.2, 119.5, 132.6, 135.9, 161.1 ppm; IR (film): n˜ =
3330, 2925, 1700, 1650, 1500, 1440, 1250, 990, 920, 820, 750, 680 cmÀ1; MS
(APCI): m/z (%): 259 (100) [M++NH3]; HRMS (ES+): m/z: calcd for
C8H14Cl3N2O: 259.0166 [M++NH4]; found: 259.0165.
[2] a) T. Ikariya, Y. Ishikawa, K. Hirai, S. Yoshikawa, Chem. Lett. 1982,
1815–1818; b) T. G. Schenck, B. Bosnich, J. Am. Chem. Soc. 1985,
107, 2058–2066; c) P. Metz, C. Mues, A. Schoop, Tetrahedron 1992,
48, 1071–1080; d) M. Mehmandoust, Y. Petit, M. LarchevÞque, Tet-
rahedron Lett. 1992, 33, 4313–4316; e) J. Gonda, A.-C. Helland, B.
ˇ
Ernst, D. Bellus, Synthesis 1993, 729–734; f) L. E. Overman, G. G.
Zipp, J. Org. Chem. 1997, 62, 2288–2291.
Synthesis of 10 and application to the rearrangement of 7a: A mixture of
(S)-COP-Cl 2b (0.040 g, 27 mmol), triphenylphosphine NovaGel (0.080 g,
0.4–0.7 mmolgÀ1, ꢀ44 mmol) and acetonitrile (5 mL) were stirred for 3 h
[3] a) M. Calter, T. K. Hollis, L. E. Overman, J. Ziller, G. G. Zipp, J.
Org. Chem. 1997, 62, 1449–1456; b) T. K. Hollis, L. E. Overman,
Tetrahedron Lett. 1997, 38, 8837–8840; c) Y. Uozumi, K. Kato, T.
Chem. Eur. J. 2007, 13, 10216 – 10224
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10223

C. J. Richards and H. Nomura
Hayashi, Tetrahedron: Asymmetry, 1998, 9, 1065–1072; d) F. Cohen,
L. E. Overman, Tetrahedron: Asymmetry, 1998, 9, 3213–3222;
e) T. K. Hollis, L. E. Overman, J. Organomet. Chem. 1999, 576, 290–
299; f) Y. Donde, L. E. Overman, J. Am. Chem. Soc. 1999, 121,
2933–2934; g) Y. Jiang, J. M. Longmire, X. Zhang, Tetrahedron Lett.
1999, 40, 1449–1450; h) P.-H. Leung, K.-H. Ng, Y. Li, A. J. P. White,
D. J. Williams, Chem. Commun. 1999, 2435–2436; i) J. Kang, K. H.
Yew, T. H. Kim, D. H. Choi, Tetrahedron Lett. 2002, 43, 9509–9512;
j) A. Moyano, M. Rosol, R. M. Moreno, C. López, M. A. Maestro,
Angew. Chem. 2005, 117, 1899–1903; Angew. Chem. Int. Ed. 2005,
44, 1865–1869; k) C. E. Anderson, Y. Donde, C. J. Douglas, L. E.
Overman, J. Org. Chem. 2005, 70, 648–657; l) R. Peters, Z. Xin,
D. F. Fischer, W. B. Schweizer, Organometallics, 2006, 25, 2917–
2920; m) M. E. Weiss, D. F. Fischer, Z. Xin, S. Jautze, W. B. Schweiz-
er, R. Peters, Angew. Chem. 2006, 118, 5823–5827; Angew. Chem.
Int. Ed. 2006, 45, 5694–5698; n) S. Jautze, P. Seiler, R. Peters,
Angew. Chem. 2007, 119, 1282–1286; Angew. Chem. Int. Ed. 2007,
46, 1260–1264.
[20] a) P. M. Henry, J. Am. Chem. Soc. 1972, 94, 5200–5206; b) P. M.
Henry, Acc. Chem. Res. 1973, 6, 16–24.
[21] a) B. Škermark, J. E. Bäckvall, K. Siirala-HansØn, K. Sjçberg, K.
Zetterberg, Tetrahedron Lett. 1974, 15, 1363–1366; b) B. Škermark,
J. E. Bäckvall, L. S. Hegedus, K. Zetterberg, K. Siirala-HansØn, K.
Sjçberg, J. Organomet. Chem. 1974, 72, 127–138.
[22] a) J. K. Stille, R. Divakaruni, J. Am. Chem. Soc. 1978, 100, 1303–
1304; b) J. E. Bäckvall, B. Škermark, S. O. Ljunggren, J. Am. Chem.
Soc. 1979, 101, 2411–2416; c) B. Škermark, B. C. Sçderberg, S. S.
Hall, Organometallics, 1987, 6, 2608–2610.
[23] Examples of palladium-catalysed addition of nucleophiles to alkenes
that proceed by means of cis palladation have recently been report-
ed. Oxygen nucleophiles: a) T. Hayashi, K. Yamasaki, M. Mimura,
Y. Uozumi, J. Am. Chem. Soc. 2004, 126, 3036–3037; b) M. B. Hay,
A. R. Hardin, J. P. Wolfe, J. Org. Chem. 2005, 70, 3099–3107;
c) R. M. Trend, Y. K. Ramtohul, B. M. Stoltz, J. Am. Chem. Soc.
2005, 127, 17778–17788. Nitrogen nucleophiles: d) J. E. Ney, J. P.
Wolfe, Angew. Chem. 2004, 116, 3689–3692; Angew. Chem. Int. Ed.
2004, 43, 3605–3608; e) G. Liu, S. S. Stahl, J. Am. Chem. Soc. 2007,
129, 6328–6335.
[24] It has been noted that addition of an excess of silver trifluoroacetate
to an iodide-bridged ferrocene oxazoline palladacycle results in the
formation of a ferrocenium cation, see reference [3k].
[25] M. D. Rausch, R. A. Genetti, J. Org. Chem. 1970, 35, 3888–3897.
[26] The pKa values for the conjugate acids of 2-methyloxazoline and
imidazole are 5.5 and 7.75 respectively, see: a) M. A. Weinberger, R.
Greenhalgh, Can. J. Chem. 1963, 41, 1038–1041; b) T. C. Bruice,
G. L. Schmir, J. Am. Chem. Soc. 1958, 80, 148–156.
[4] COP=cobalt oxazoline palladacycle. For 2b the abbreviation
[COP-Cl]2 has also been used to highlight the dimeric nature of this
palladacycle.
[5] a) A. M. Stevens, C. J. Richards, Organometallics 1999, 18, 1346–
1348; b) C. E. Anderson, L. E. Overman, C. J. Richards, M. P.
Watson, N. White, Org. Synth. 2007, 84, 139–147; c) C. E. Anderson,
S. F. Kirsch, L. E. Overman, C. J. Richards, M. P. Watson, Org.
Synth. 2007, 84, 148–155.
[6] L. E. Overman, C. E. Owen, M. M. Pavan, C. J. Richards, Org. Lett.
2003, 5, 1809–1812.
[7] C. E. Anderson, L. E. Overman, J. Am. Chem. Soc. 2003, 125,
12412–12413.
[27] D. F. Shriver, P. W. Atkins, C. H. Langford, Inorganic Chemistry,
Oxford University Press, 1994.
[8] For other applications of cobalt oxazoline palladacycles as catalysts
for the allylic imidate rearrangement see: a) J. Kang, T. H. Kim,
K. H. Yew, W. K. Lee, Tetrahedron: Asymmetry 2003, 14, 415–418;
b) S. F. Kirsch, L. E. Overman, M. P. Watson, J. Org. Chem. 2004,
69, 8101–8104.
[9] a) S. F. Kirsch, L. E. Overman, J. Org. Chem. 2005, 70, 2859–2861;
b) S. F. Kirsch, L. E. Overman, J. Am. Chem. Soc. 2005, 127, 2866–
2867; c) E. E. Lee, R. A. Batey, J. Am. Chem. Soc. 2005, 127, 14887–
14893; d) S. F. Kirsch, L. E. Overman, N. S. White, Org. Lett. 2007,
9, 911–913.
[10] F. X. Roca, M. Motevalli, C. J. Richards, J. Am. Chem. Soc. 2005,
127, 2388–2389.
[11] During the preparation of this manuscript, a model for the enantio-
selectivity of COP-Cl-catalysed rearrangement of trichloroacetimi-
dates was reported. See: M. P. Watson, L. E. Overman, R. G. Berg-
man, J. Am. Chem. Soc. 2007, 129, 5031–5044.
[12] Ratios: 2a (1:0), 2b (1:0.7), 2c (1:0.7), 2d (1:0.3), 2e (1:0), 2 f (1:0).
Isomer mixtures (cis/trans, also referred to as syn/anti) are a feature
of many bridged palladacycles.
[13] a) T. Izumi. H. Watabe, A. Kasahara, Bull. Chem. Soc. Jpn. 1981, 54,
1711–1714; b) I. P. Smoliakova, K. J. Keuseman, D. C. Haagenson,
D. M. Wellmann, P. B. Colligan, N. A. Kataeva, A. V. Churakov,
L. G. Kuz’mina, V. V. Dunina, J. Organomet. Chem. 2000, 603, 86–
97.
[28] F. Basolo, H. B. Gray, R. G. Pearson, J. Am. Chem. Soc. 1960, 82,
4200–4203.
[29] a) M. Avalos, R. Babiano, P. Cintas, J. L. JimØnez, J. C. Palacios, Tet-
rahedron: Asymmetry 1997, 8, 2997–3017; b) C. Girard, H. B.
Kagan, Angew. Chem. 1998, 110, 3088–3127; Angew. Chem. Int. Ed.
1998, 37, 2922–2959.
[30] For a statistical distribution of S,pR/S,pR (R,R) R,pS/R,pS (S,S) and
S,pR/R,pS (R,S) dimers (K=4), eeaux =50% gives EEprod =55% (EE
= extrapolated enantiomeric yield) with g=0.5 (g=kRS/kRR). Alter-
natively, for g=0 an EEprod of 54% is calculated with eeaux =50% if
the component of the meso (R,S) diastereoisomer is 20%. See: D.
Guillaneux, S.-H. Zhao, O. Samuel, D. Rainford, H. B. Kagan, J.
Am. Chem. Soc. 1994, 116, 9430–9439.
[31] D. St. C. Black, G. B. Deacon, G. L. Edwards, Aust. J. Chem. 1994,
47, 217–227.
[32] a) B. N. Cockburn, D. V. Howe, T. Keating, B. F. G. Johnson, J.
Lewis, J. Chem. Soc. Dalton Trans. 1973, 404–410.
[33] Trans effect for PtII: NH3 ꢀamine<ClÀ ꢀBrÀ <SCNÀ ꢀIÀ ꢀNO2À ꢀ
PhÀ <MeÀ, see: F. Basolo, R. G. Pearson, Prog. Inorg. Chem. 1962,
4, 381–453.
[34] Models for the origin of enantioselection have been proposed in
which the alkene component of the allylic imidate substrate is coor-
dinated cis to the metallocene.[3m,11] This is compatible with a mecha-
nism in which initial coordination of the substrate is followed by iso-
merisation to the thermodynamic cis isomer, as presumably is also
the case with phosphine adducts 3.
[14] R. S. Prasad, C. E. Anderson, C. J. Richards, L. E. Overman, Orga-
nometallics 2005, 24, 77–81.
[15] G. Jones, C. J. Richards, Organometallics 2001, 20, 1251–1254.
[16] Acetonitrile has been used with catalyst 4 (1 mol% loading) but
only at 508C: see reference [8b].
[35] F. Cramer, N. Hennrich, Chem. Ber. 1961, 94, 976–989.
[36] P. Bulger, M. G. Moloney, P. C. Trippier, Org. Biomol. Chem. 2003,
1, 3726–3737.
[17] I. Persson, Pure Appl. Chem. 1986, 58, 1153–1161.
[18] Y. Uozumi, H. Danjo, T. Hayashi, Tetrahedron Lett. 1997, 38, 3557–
3560.
[37] B. C. Ranu, A. Majee, Chem. Commun. 1997, 1225–1226.
[19] L. E. Overman, Angew. Chem. 1984, 96, 565–572; Angew. Chem.
Int. Ed. Engl. 1984, 23, 579–586.
Received: June 8, 2007
Published online: September 24, 2007
10224
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 10216 – 10224