Dual Palladium(II)/Tertiary Amine Catalysis for Asymmetric Regioselective Rearrangements of Allylic Carbamates

Article abstract of DOI:10.1002/chem.201600138

The streamlined catalytic access to enantiopure allylic amines as valuable precursors towards chiral β- and γ-aminoalcohols as well as α- and β-aminoacids is desirable for industrial purposes. In this article an enantioselective method is described that transforms achiral allylic alcohols and N-tosylisocyanate in a single step into highly enantioenriched N-tosyl protected allylic amines via an allylic carbamate intermediate. The latter is likely to undergo a cyclisation-induced [3,3]-rearrangement catalysed by a planar chiral pentaphenylferrocene palladacycle in cooperation with a tertiary amine base. The otherwise often indispensable activation of palladacycle catalysts by a silver salt is not required in the present case and there is also no need for an inert gas atmosphere. To further improve the synthetic value, the rearrangement was used to form dimethylaminosulfonyl-protected allylic amines, which can be deprotected under non-reductive conditions.

Full text of DOI:10.1002/chem.201600138

DOI: 10.1002/chem.201600138
Full Paper
&
Allylic Compounds
Dual Palladium(II)/Tertiary Amine Catalysis for Asymmetric
Regioselective Rearrangements of Allylic Carbamates
Johannes Moritz Bauer, Wolfgang Frey, and RenØ Peters*[a]
Dedicated to Professor Dieter Enders on the occasion of his 70th birthday
Abstract: The streamlined catalytic access to enantiopure al-
lylic amines as valuable precursors towards chiral b- and g-
aminoalcohols as well as a- and b-aminoacids is desirable
for industrial purposes. In this article an enantioselective
method is described that transforms achiral allylic alcohols
and N-tosylisocyanate in a single step into highly enantioen-
riched N-tosyl protected allylic amines via an allylic carba-
mate intermediate. The latter is likely to undergo a cyclisa-
tion-induced [3,3]-rearrangement catalysed by a planar chiral
pentaphenylferrocene palladacycle in cooperation with a ter-
tiary amine base. The otherwise often indispensable activa-
tion of palladacycle catalysts by a silver salt is not required
in the present case and there is also no need for an inert
gas atmosphere. To further improve the synthetic value, the
rearrangement was used to form dimethylaminosulfonyl-
protected allylic amines, which can be deprotected under
non-reductive conditions.
Introduction
Enantiopure a-branched allylic amines are employed as high-
value building blocks for technical scale applications[1] because
the synthetically versatile amino and olefin functionalities
allow for a number of subsequent synthetic manipulations. For
that reason chiral allylic amines have also been used in a large
number of natural product syntheses,[2] for a straightforward
access towards N-containing chiral heterocycles[3] as well as a-
or b-amino acids.[4] In addition, allylic amines have been de-
scribed as peptide isosters with potential biological activity.[5]
As a result of the synthetic utility of chiral enantiopure allylic
amines, a number of catalytic methods have been developed
for the enantioselective preparation including allylic substitu-
tions,[6] 1,2-additions of appropriate nucleophiles to imines,[7]
hydrogenations of dienamines,[8] hydroaminations of allenes[9]
and several types of rearrangements. The latter include
[1,3]-,[10] [2,3]-[11] and [3,3]-rearrangements.[12] Predominantly al-
lylic imidate substrates 5 (see Scheme 1, dashed box) in combi-
nation with carbophilic Lewis acids have been employed for
enantioselective [3,3]-rearrangements.[12] A key feature of these
reactions is a high regioselectivity, which has been explained
by a cyclisation-induced rearrangement mechanism.[13]
Scheme 1. Comparison of allylic imidates 5 and allylic carbamates 1 and
mechanistic idea of the decarboxylative PdII-catalysed rearrangement.
arylbenzamides, which are difficult to hydrolyse to the corre-
sponding allylic amines. Further milestones were then the suc-
cessful use of allylic trichloro-[15] (R3 =CCl3 in 5) as well as N-
aryl-[16] and N-alkyltrifluoroacetimidates[17] (R3 =CF3) in enantio-
selective rearrangements, because the trichloro- and trifluoroa-
cetamide products can be readily transformed into primary
and secondary amines, respectively. Very recently, the method
could also be extended to non-halogenated acetimidates (R3 =
CH2R) providing allylic acetamides, which can be deprotected
under mild enzymatic conditions.[18]
The first enantioselective allylic imidate rearrangements
were reported in 1997 by Overman et al.[14] In these initial reac-
tions N-arylbenzimidates (R3 =aryl in 5) were used providing N-
However, a common disadvantage of allylic imidate sub-
strates 5 is their relatively large sensitivity towards hydrolysis,
hampering their isolation and long-time storage. In addition,
the preparation of trifluoro- and non-halogenated allylic imi-
dates is tedious.
[a] Dr. J. M. Bauer, Dr. W. Frey, Prof. Dr. R. Peters
Institut für Organische Chemie, Universität Stuttgart
Pfaffenwaldring 55, 70569 Stuttgart (Germany)
E-mail: rene.peters@oc.uni-stuttgart.de
Our goal was to develop a [3,3]-rearrangement, which still
provides the general advantages noted for PdII-catalysed asym-
metric allylic imidate rearrangements such as high regio- and
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201600138.
Chem. Eur. J. 2016, 22, 5767 – 5777
5767
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
enantioselectivity and a broad compatibility with functional
groups, but uses more robust and more readily accessible sub-
strates. We envisioned to develop the first catalytic asymmetric
allylic carbamate rearrangements, because 1) the quite stable
carbamates 1 are very easily prepared by addition of alcohols
to inexpensive commercial isocyanates; 2) carbamates 1 resem-
ble allylic imidates; and 3) the initial carbaminic acid rearrange-
ment product 3 would spontaneously decarboxylate saving an
extra deprotection step (Scheme 1).
Table 1. Investigation of the decarboxylative allylic carbamate rearrange-
ment of model substrate 6a under neutral reaction conditions.
A thermal decarboxylative [3,3]-rearrangement of an N-phe-
nylallylcarbamate was already reported in 1968 and accelerat-
ed by NaH, still requiring a reaction temperature of 200–
2408C.[19] In 1991, Wang and Calabrese reported that BF3 can
promote decarboxylative rearrangements for substrates with
special substitution patterns to stabilise an allylic carbocation
intermediate.[20] In 2000, Lei and Liu reported a non-enantiose-
lective palladium(II)-catalysed rearrangement of N-tosylcarba-
mates.[21] High regioselectivity was accomplished by Pd(OAc)2
(5 mol%) as catalyst assisted by an excess of LiBr (4 equiv) in
DMF at 1008C.[22,23] Ten years later, Xing and Yang described
a gold(I) (5 mol%) catalysed reaction that made use of stoi-
chiometric amounts of iPr2NEt as base additive to form racemic
decarboxylative rearrangement products.[24]
Allylic carbamates have also been employed for Pd0-[25] and
IrI-catalysed[26] decarboxylative allylic substitutions, in which
linear products were preferentially formed in the first case and
chiral branched products in the latter through electrophilic p-
allyl complexes. Here we report the development of the first
catalytic asymmetric decarboxylative allylic carbamate rear-
rangements and the translation into a domino process[27] em-
ploying achiral allylic alcohols as substrates.[28] This process is
enabled by the cooperative action of PdII and a tertiary
amine.[29]
Entry
Precatalyst
[xmol%]
AgX
[ymol%]
Yield 7
Yield 8a
ee 8a
[%][a]
[%][a]
[%][b]
1
2
3
4
5
6
7
8
[FIP-Cl]2 (5)
[FIP-Cl]2 (5)
AgNO3 (10)
AgNO3 (20)
AgNO3 (10)
AgNO3 (20)
AgNO3 (10)
AgNO3 (20)
AgOTs (6)
2
6
0
2
17
15
0
0
0
0
0
10
12
0
–
–
–
–
9
10
–
[PPFIP-Cl]2 (5)
[PPFIP-Cl]2 (5)
[PPFOP-Cl]2 (5)
[PPFOP-Cl]2 (5)
[FBIP-Cl]2 (1.5)
[FBIPP-Cl]2 (1.5)
AgOTs (6)
0
0
–
[a] Yield determined by 1H NMR spectroscopy using an internal standard.
[b] Enantiomeric excess of isolated product (by column chromatography)
determined by HPLC.
lished protocols, resulting in a chloride ligand exchange. Using
an excess of the silver salts per ferrocene unit, the monopalla-
dacycles also undergo an oxidation process. In the case of pen-
taphenylferrocene palladacycles we have previously shown
that paramagnetic PdIII complexes are generated under these
conditions as catalytically active species, while the ferrocene
core remained intact.[16i,32]
Results and Discussion
Development and optimisation of the decarboxylative car-
bamate rearrangement
Table 1 summarises the initial results obtained with the dif-
ferent catalysts in CH2Cl2 solutions at 608C using small pres-
sure tubes as reaction vials. For the monopalladacycles, the
use of both non-oxidised and oxidised activated catalysts was
investigated (Table 1, entries 1, 3, 5 and 2, 4, 6, respectively). In
most cases, none of the desired product 8a was obtained uti-
lising a monopalladacycle catalyst.[33] The same result was
found with both bismetallacycle catalysts (Table 1, entries 7
and 8). The only catalyst that delivered small amounts of rear-
rangement product 8a (ca. 10% yield), was the pentaphenyl-
ferrocene-based oxazoline monopalladacycle PPFOP[15e]
(Table 1, entries 5 and 6). However, the precatalyst activated by
AgNO3 provided only disappointing enantioselectivities (ca.
10% ee). In addition, relatively large amounts of sulfonamide 7
were detected as side product.
To develop an asymmetric decarboxylative allylic carbamate re-
arrangement, N-tosyl protected allyl carbamates were chosen
as substrates based on their reactivity in the previous non-
enantioselective protocols.[21,24] They were readily prepared as
geometrically pure isomers by addition of the corresponding
E-configured allylic alcohols to p-tosylisocyanate (for details
see the Supporting Information). The nPr-substituted olefin 6a
was chosen as a model substrate, because the alcohol precur-
sor is commercially available in isomerically pure form. Differ-
ent ferrocene-based planar chiral pallada- and platinacycles
previously developed by our research group[15e,16g,i,j] were then
screened for their efficiency as asymmetric catalysts (Table 1).
These metallacycles are chloride-bridged dimers, which
showed almost no catalytic activity in our previous investiga-
tions on allylic imidate rearrangements without a removal of
the chloride bridges. For that reason they were activated by
AgNO3 in CH2Cl2 (monopalladacycles)[15e,16i,30] and AgOTs in
MeCN (bismetallacycles)[16g,j,31] according to our previously pub-
Because PPFOP delivered the only active system in the initial
reactions, in the subsequent development we focused on this
catalyst type. To increase the reactivity, the use of Brønsted
base additives was studied (Table 2) to deprotonate the quite
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5768
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
different solvents were initially investigated under the condi-
tions of Table 2, entry 12. Whereas Lewis basic solvents signifi-
cantly lowered the activity (yield with DMF: 35%; MeCN: 22%),
non- or weakly-coordinating solvents were tolerated, but in
each case the efficiency was slightly lower than with CH2Cl2
(yield/ee with EtOAc: 86%/85%; THF: 80%/82%; toluene:
82%/84%; CHCl3: 79%/84%). Nevertheless, these data demon-
strate that a solvent more appropriate to a technical applica-
tion than CH2Cl2 like for instance EtOAc can be utilised.
Table 2. Investigation of different base additives.
Entry
AgNO3
[ymol%]
Base
[mol%]
Yield 7
[%][a]
Yield 8a
ee 8a
[%][a]
[%][b]
1
2
3
4
5
6
7
8
12
12
12
12
12
12
6
K2CO3 (100)
Cs2CO3 (100)
KOtBu (100)
2,4,6-collidine (100)
iPr2NEt (100)
PS[c] (100)
3
0
3
3
6
3
6
3
6
0
12
0
88
90
89
96
0
86
–
83
–
89
88
89
88
–
Different silver salts (AgX) were explored for the further opti-
misation (Table 3). In these trials the amount of the dimeric
precatalyst was reduced to 1 mol% to facilitate the identifica-
tion of differences in activity. 2 mol% of the corresponding
silver salts was used for the chloride exchange.
iPr2NEt (100)
PS[c] (100)
6
9[d]
10[d]
11
12
13
14
15
16
–
–
6
6
6
6
6
6
iPr2NEt (100)
PS[c] (100)
iPr2NEt (25)
PS[c] (25)
iPr2NEt (10)
PS[c] (10)
25
30
3
9
6
5
9
13
0
–
Table 3. Investigation of different silver salts for catalyst activation.
89
86
33
91
11
27
88
87
92
86
n.d.[e]
n.d.[e]
iPr2NEt (6)
PS[c] (6)
[a] Yield determined by 1H NMR spectroscopy using an internal standard.
[b] Enantiomeric excess of isolated product (by column chromatography)
determined by HPLC. [c] PS=proton sponge (1,8-bis(dimethylamino)-
naphthalene). [d] The reaction was performed in the absence of a pallada-
cycle catalyst. [e] Enantiomeric excess was not determined due to the
low product yield.
Entry
AgX
Yield 7
Yield 8a
ee 8a
[%][a]
[%][a]
[%][b]
1
2
3
4
5
6
7
8
AgNO3
8
4
86
94
65
77
62
44
65
48
86
66
59
94
58
77
85
78
84
n.d.[c]
n.d.[c]
82
n.d.[c]
88
n.d.[c]
n.d.[c]
90
AgO2CCF3
AgOAc
AgOTs
AgOMs
AgOTf
25
12
13
15
16
16
8
10
14
2
NH-acidic carbamate (pKa ꢀ3.7).[34] Whereas the reactivity was
still low when using the O-centred bases K2CO3, Cs2CO3 and
KOtBu (Table 2, entries 1–3), the enantioselectivity was massive-
ly improved in the presence of the salts. The enantioselectivity
was further improved by using neutral bulky organic N-bases,
and the rearrangement product was obtained in high yields
using either iPr2NEt (Table 2, entry 5) or proton sponge (PS,
1,8-bis(dimethylamino)naphthalene, entry 6). In contrast, there
was no reactivity with 2,4,6-collidine (Table 2, entry 4), maybe
due to catalyst inhibition. In the case of iPr2NEt or proton
sponge the initial catalyst oxidation state did not play a signifi-
cant role. The experiments using either the PdIII catalyst
(Table 2, entries 5 and 6) or the PdII catalyst (Table 2, entries 7
and 8) gave very similar results in terms of product yield, enan-
tioselectivity and regioselectivity. The linear product of
a formal [1,3]-rearrangement was only detected in trace quan-
tities. In the absence of the catalyst no rearrangement product
was detected for both N-bases, but decomposition towards
sulfonamide 7 was largely increased (Table 2, entries 9 and 10).
Stoichiometric quantities of the N bases are not necessary
because very similar reaction outcomes were also found with
25 mol% (Table 2, entries 11 and 12). The activity of PS also al-
lowed for the use of 10 mol% (Table 2, entry 14), whereas with
10 mol% iPr2NEt the reactivity was low (entry 13). Lower base
loadings than 10 mol% were not tolerated in each case
(Table 2, entries 15 and 16).
AgBF4
AgPF6
9
AgO2CC3F7
AgO2CC7F15
AgO2CCF3
AgO2CCF3
–
10
11[d]
12[e]
13[f]
13
86
[a] Yield determined by 1H NMR spectroscopy using an internal standard.
[b] Enantiomeric excess of isolated product (by column chromatography)
determined by HPLC. [c] Enantiomeric excess was not determined.
[d] 0.5 mol% of precatalyst and 1.0 mol% of the silver salt were used.
[e] The reaction was performed at 508C. [f] The reaction was performed
in the presence of the non-activated palladacycle catalyst.
Because the anionic ligand XÀ has an influence on the elec-
tronic and steric properties of the catalytic centres, significant
differences in catalytic activity and enantioselectivity were ex-
pected. Using AgNO3 for the activation, the ee value decreased
to 77% with the lower catalyst loading, albeit the product
yield was still identical (Table 3, entry 1). Better data was ob-
tained with silver trifluoroacetate (Table 3, entry 2), which al-
lowed for 85% ee and a yield of 94%. The other silver salts
tested allowed for very similar enantioselectivities, the only ex-
ception being AgOAc, which had an ee value smaller than 80%
(Table 3, entry 3).
The amount of side product 7 formed was influenced by
AgX. In particular, with AgOAc a rather large amount of 7 was
found (Table 3, entry 3). The only silver salt found that allowed
for an activity comparable to silver trifluoroacetate was silver
Due to the better reactivity with PS as co-catalyst, the reac-
tion was further optimised using this base. For that purpose,
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5769
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
heptafluorobutyrate, which also had a slightly positive effect
on the enantioselectivity (Table 3, entry 9). Nevertheless, the
use of silver trifluoroacetate was more attractive to us for eco-
nomic reasons.
previously found to cause lower enantioselectivities, which was
explained by an intrinsically more difficult differentiation of the
enantiotopic olefin faces.[15,16]
Regarding the regioselectivity the branched chiral products
were usually strongly favoured with ratios of ꢁ20:1. Lower re-
gioselectivities of 8:1 and 12:1 were found for substrates carry-
ing the branched iPr and iBu residues, respectively (Table 4, en-
tries 9 and 10), in which the [3,3]-rearrangement is expected to
be more difficult for steric reasons.
To decrease costs, lower catalyst loadings were also studied.
However, the product yield was moderate starting from
0.5 mol% of the precatalyst, partly because of the formation of
more side product 7 (Table 3, entry 11). On the other hand, the
ee value could be increased to 90% by a decrease of the reac-
tion temperature to 508C (Table 3, entry 12). Surprisingly, mod-
erate product yields could even be obtained with the non-acti-
vated catalyst (Table 3, entry 13).
Substrates with aryl residues R directly connected to the
olefin moiety were, in general, not well tolerated and either
provided no desired product (electron-rich aryl residues) or the
products were formed with poor yields (as a result of side reac-
tions) and low enantioselectivity (ee<40%, data not shown).
The substrate preference is thus complementary to Ir-catalysed
allylic aminations, for which aromatic residues R usually provid-
ed higher regioselectivities than aliphatic groups.[6]
Scope and limitations
The optimised conditions presented in Table 3, entry 12, were
then applied to various allylic N-tosylcarbamates carrying dif-
ferent alkyl residues R at the olefin function (Table 4).
Moreover, we found that Z-configured allylic moieties are
not favourable. Even using larger quantities of catalyst and
base and a temperature of 608C the reaction was sluggish and
provided the product with only moderate enantioselectivity
(Scheme 2). It is noteworthy, however, that the title reaction is
Table 4. Investigation of different carbamate substrates.
Entry
6/8
R
Yield
[%][a]
rs[b]
ee
[%][c]
1
2
3
4
5
6
7
8
a
b
c
d
e
f
g
h
i
nPr
Et
Me
nPent
(CH2)2Ph
(CH2)2CO2Me
CH2OTBS
CH2OBn
iPr
94
88
88
91
96
94
77
87
45
90
25:1
30:1
70:1
24:1
34:1
26:1
20:1
20:1
8:1
90
88
63
88
92
89
98
98
86
88
Scheme 2. Rearrangement of a Z-configured substrate.
stereospecific, because the product was formed with opposite
absolute configuration depending on the olefin geometry of
the substrate. This points to a scenario in which 1) enantio-
face-selective olefin coordination might predetermine the con-
figurational outcome and 2) similar to the allylic imidate rear-
rangement, the carbamate rearrangement might proceed
through an early (half)chair-like transition state, in which E sub-
stituents could adopt an equatorial position, whereas Z sub-
stituents need to be accommodated in an axial position (see
below).[15,16] The latter would explain the higher activation bar-
riers with Z substrates.
9[d]
10
j
iBu
12:1
[a] Yield of isolated product. [b] Regioselectivity branched/linear product
determined by 1H NMR of the crude product. [c] Enantiomeric excess de-
termined by HPLC. [d] t=48 h at 608C.
In nearly all cases the products 8 were formed in good to
high yields. The only exception was noted for the iPr-substitut-
ed substrate 6i. The reaction was found to be more sluggish
with the sterically demanding a-branched substituent and in-
creased formation of side products was observed (Table 4,
entry 9). In contrast, in the presence of the b-branched iBu
substituent, the reactivity was still high. Enantiomeric excesses
were usually in a range of 86 to 98% (Table 4, entry 10). The
best enantioselectivities were found for two substrates with
protected hydroxymethyl residues R (Table 4, entries 7 and 8).
Also, an enolisable ester residue was well tolerated both in
terms of reactivity and enantioselectivity (Table 4, entry 6). A
moderate ee value was only found for substrate 6c carrying
a methyl group as the smallest residue R of the substrates in-
vestigated (Table 4, entry 3). Similarly, for allylic imidate rear-
rangements a methyl substituent at the olefin function was
The absolute configuration of 8a was determined to be R
by X-ray analysis (Figure 1).[35] In addition, for products 8a–d
and 8h–j the R configurations could be assigned by compari-
son to literature data (see the Supporting Information for de-
tails).
Development and application of a domino process
Since the N-tosylcarbamate formation between the allylic alco-
hols and N-tosylisocyanate proceeded smoothly and provided
very pure crude products, a one-pot protocol was envisioned
without the need for the isolation of the carbamate sub-
strate.[36] Stirring a solution of the allylic alcohol 9a and
pTsNCO for 30 min at room temperature in CH2Cl2 was fol-
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5770
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 5. Application of the cascade title reaction.
Entry
9/8
R
Yield
[%][a]
rs[b]
ee
[%][c]
Figure 1. Determination of the absolute configuration of 8a by X-ray crystal
structure analysis.
1
2
3
4
5
6
7
8
9
10
a
b
c
d
e
f
g
h
i
nPr
Et
Me
nPent
(CH2)2Ph
(CH2)2CO2Me
CH2OTBS
CH2OBn
iPr
80
88
90
85
86
80
83
90
52
79
20:1
16:1
40:1
22:1
25:1
26:1
20:1
20:1
7:1
90
91
72
93
92
90
98
98
92
89
lowed by addition of the catalyst (prepared from 3 mol% pre-
catalyst and 6 mol% AgO2CCF3) and base (20 mol% PS). After
18 h at 608C the decarboxylative rearrangement product was
formed in high yield with an ee of 89% (Scheme 3, Method A).
j
iBu
10:1
[a] Yield of isolated product. [b] Regioselectivity determined by 1H NMR
analysis of the crude product. [c] Enantiomeric excess determined by
HPLC.
formed in good to high yields and with high enantio- and re-
gioselectivity. For the difficult substrate 9c carrying a Me sub-
stituent R, the ee could be increased from 63 to 72% applying
this protocol (Table 5, entry 3). In all other examples the ee
values ranged from 89 to 98%. Similar to the use of isolated
carbamates, branched alkyl groups R had a negative impact on
the regioselectivity (Table 5, entries 9 and 10). On the other
hand, functional groups such as ester, ether and silyl ether
moieties were well accommodated (Table 5, entries 6–8).
The practicality of the rearrangement protocol was exam-
ined on gram scale for substrate 9d using a sealed pressure
tube under air (Scheme 4). With 5.24 mmol of this substrate,
1.366 g of 8d (92% yield) were obtained with 91% ee. A reduc-
tive sulfonamide deprotection was performed under standard
conditions[39] to release the free enantioenriched allylic amine
10d in high yield (Scheme 5). However, as these reductive con-
ditions should not be compatible with reduction-sensitive
functional groups, we were also interested in alternative pro-
tecting groups that might be removed under different condi-
tions.
Scheme 3. Development of a one-pot protocol using activated (Method A)
or non-activated catalyst (Method B).
The operational simplicity would, of course, be further im-
proved by avoiding the catalyst activation with a silver salt. As
promising results were already obtained in the initial studies
(Table 3, entry 13), the carbamate rearrangement of 6a to 8a
was reinvestigated using the non-activated catalyst in different
solvents at higher temperatures to further increase the reactivi-
ty. The best results were obtained in CHCl3 at 858C (in a sealed
tube). Under conditions otherwise identical to Table 3,
entry 13, compound 8a was then obtained in 88% yield and
with 90% ee. The one-pot version was thus repeated in CHCl3
with the non-activated catalyst. After carbamate generation at
room temperature, [PPFOP-Cl]2 (3 mol%) and PS (20 mol%)
were added and after 18 h at 858C the product was isolated in
89% yield and with 88% ee (Scheme 3, Method B).
Due to the high efficiency of the one-pot version, the possi-
bility of a domino reaction[37] was studied without a separate
allylic carbamate preformation. All reaction components were
thus mixed at once. Gratifyingly, this operationally very simple
protocol also provided the rearrangement product 8a with
high yield and enantioselectivity using 1 mol% of [PPFOP-Cl]2,
even when conducted under air (Table 5, entry 1).[38]
Scheme 4. Gram-scale experiment of the domino carbamate formation/de-
carboxylative rearrangement.
This protocol was then used for different E-configured allylic
alcohol substrates 9 (Table 5). Gratifyingly, the results in terms
of yield, regio- and enantioselectivity for all ten substrates
were very similar to those presented in Table 4, in which the
isolated carbamates 6 were employed in combination with the
activated catalyst. Hence, again the products were usually
Scheme 5. Reductive sulfonamide cleavage to release a free allylic amine.
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5771
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Variation of the N-protecting group
ate choice, in particular because it is also well established as
a synthetically attractive protecting group as it can be re-
moved under non-reductive conditions.[40] In our preliminary
studies, this substrate type was prepared over two steps from
the corresponding allylic alcohol 9, CDI and Me2NSO2NH2 in
low overall yields (ca. 20%).[28] As anticipated, a positive influ-
ence of the strongly electron-donating amino group at the sul-
fonyl moiety was indeed observed in terms of enantioselectivi-
ty as shown below. For that reason a more efficient route to
this substrate class was developed. In a one-pot protocol, the
allylic alcohols were first treated with chlorosulfonylisocyanate
(CSI) to form a chlorosulfamoyl intermediate, which was then
trapped by Me2NH (Table 7).
To increase the flexibility of possible protecting group strat-
egies, a number of different allylic carbamate moieties were in-
vestigated. Unfortunately, N-benzyl-, N-carboxyl- (i.e., N-carbon-
yloxybenzyl, N-carbonyloxyphenyl, N-carbonyloxymethyl and
N-carbonyloxytrichloromethyl) and N-aryl- (4-MeOC6H4, 4-
O2NC6H4, 3,5-(F3C)2C6H3) protected allylic carbamates failed to
form the rearrangement products in more than trace quantities
using the PPFOP catalyst system in combination with PS. Con-
sequently, the applicability of different N-sulfonyl residues was
investigated (Table 6).
Table 6. Investigation of the electronic effect of different N-sulfonyl resi-
dues.
Table 7. Synthesis of the dimethylaminosulfonyl protected allylic sub-
strates.
Entry
R
Starting material/product Yield
[%][a]
rs[b]
ee
[%][c]
Entry
9/21
R
Yield
[%][a]
1
2
3
4
5
6
OMe 11/16
83
94
84
93
89
24:1
16:1
17:1
14:1
19:1
87
85
87
82
82
1
2
3
4
5
6
7
8
9
10
a
b
c
d
e
f
g
h
i
nPr
Et
Me
nPent
(CH2)2Ph
(CH2)2CO2Me
CH2OTBS
CH2OBn
iPr
92
99
84
89
90
85
94
99
96
97
Me
H
F
6a/8a
12/17
13/18
14/19
Cl
NO2 15/20
62 (80)[d] 19:1 (9:1) 73 (84)[d]
[a] Yield of isolated product. [b] Regioselectivity determined by 1H NMR
analysis of the crude product. [c] Enantiomeric excess determined by
HPLC. [d] Data in brackets with non-activated catalyst (1 mol%) in CHCl3
at 808C for 24 h.
j
iBu
[a] Yield of isolated product.
The N-sulfonyl-substituted substrates 11–15 were either pre-
pared from the allylic alcohol 9a and the corresponding iso-
cyanates (like with TsNCO) or using the corresponding primary
sulfonamide plus CDI (1,1’-carbonyldiimidazole) as dehydrating
agent (for details see the Supporting Information).
Different substrates 21 including those functionalised by
ester, silylether or ether groups were accessible by this proce-
dure through a smooth transformation. In general, the N,N-di-
methylaminosulfonyl-protected allylic carbamates 21 showed
lower reactivity in the catalytic asymmetric rearrangement
than the N-tosyl-protected substrates 6, therefore they re-
quired a reaction temperature of 608C with the PPFOP catalyst
activated by silver trifluoroacetate (for N-tosyl usually 508C
was sufficient under these conditions, see Table 4). Despite
this, the products were typically formed in high yields and
with improved enantioselectivities (Table 8). For instance, the
substrate most difficult in terms of enantioselection equipped
with a methyl residue R (21c) still allowed for an ee of 82%. For
the other products the ee values were in the range of 92–99%.
Despite the increased reaction temperature, functional
groups were well tolerated (Table 8, entries 6–8) and also the
branched/linear product ratios were usually high (18–76:1)
with the exception of substrate 21i carrying the sterically
more demanding iPr residue as R (4:1, Table 8, entry 9). In the
latter case, a reaction temperature of 808C was required for
a useful yield after 72 h.
All investigated alternative sulfonyl-protected substrates 11–
15 formed the corresponding products 16–20 with good re-
gioselectivities in CH2Cl2 at 608C in a sealed tube using activat-
ed catalyst (14:1–24:1). As a trend it was observed that N-aryl-
sulfonyl substrates with electron-donating residues (Table 6,
entries 1 and 2) allowed for somewhat better enantioselectivi-
ties than those with electron-withdrawing residues (Table 6,
entries 4–6). The NH acidity and the resultant nucleophilicity/
Lewis basicity of the deprotonated substrates thus seem to be
important for high enantioselection in the CÀN bond-forming
step. The enantioselectivity was the lowest with the para-nosyl
N-protecting group (entry 6, 73%). However, in this case the
enantioselectivity could be improved to 84% ee by use of the
non-activated catalyst in CHCl3 at 808C (Table 6, entry 6, data
in brackets).
Due to the observed electronic effect, the more electron-rich
N,N-dimethylaminosulfonyl moiety seemed to be an appropri-
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5772
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 8. Investigation of different dimethylaminosulfonyl protected allylic
substrates 21 in the decarboxylative rearrangement followed by deprotec-
tion.
Table 9. Investigation of different dimethylaminosulfonyl-protected allylic
substrates in the decarboxylative rearrangement using [PPFOP-Cl]2.
Entry
2
R
[PPFOP-Cl]2
[xmol%]
Yield
[%][a]
rs[b]
ee
[%][c]
Entry 21/22/10
R
T
Yield
rs[b]
ee
Yield
[8C] 22 [%][a]
[%][c] 10 [%][a]
1
2
3
4
5
6
7
8
9
10
a
b
c
d
e
f
g
h
i
nPr
Et
Me
nPent
(CH2)2Ph
(CH2)2CO2Me
CH2OTBS
CH2OBn
iPr
2
1
1
2
1
1
2
2
2
2
82
84
95
95
93
83
73
94
15:1
17:1
37:1
14:1
24:1
16:1
15:1
22:1
5:1
92
94
84
94
96[d]
n.d.
99
99
88
1
2
3
4
5
6
7
8
9
10
a
b
c
d
e
f
g
h
i
nPr
Et
Me
nPent
(CH2)2Ph
60
60
60
60
50
85
94
95
95
92
95
90
96
24:1 94
24:1 93
76:1 82
24:1 94
75
94[d]
73[d]
96
39:1 98[e] 98
(CH2)2CO2Me 60
18:1 n.d.
28:1 99
24:1 99
-
89
91
85[d]
90[d]
CH2OTBS
CH2OBn
iPr
60
60
80
60
36[e]
66[e]
45[f]
90[g]
4:1
92
j
iBu
15:1
93
j
iBu
18:1 94
[a] Yield of isolated product. [b] Regioselectivity branched/linear product
determined by H NMR analysis of the crude product 22. [c] Enantiomeric
excess determined by HPLC after N-deprotection and subsequent tosyl
protection. [d] Enantiomeric excess of 22e determined by HPLC. [e] The
rearrangement was performed for 72 h.
1
[a] Yield of isolated product. [b] Regioselectivity branched/linear product
determined by 1H NMR analysis of the crude product 22. [c] Enantiomeric
excess determined by HPLC after deprotection of 22 and subsequent tosy-
lation. [d] Isolated as hydrochloride salt. [e] Enantiomeric excess of 22e de-
termined directly by HPLC. [f] The rearrangement was performed in CHCl3
for 72 h. [g] The rearrangement was performed for 72 h.
strate 22c, the highest ee value for all experiments with R=
Me (84%, Table 9, entry 3) was attained under these conditions
(88–99% ee for the other residues R). The observed substitu-
tion-dependent reactivity and selectivity trends are otherwise
similar as before.
For the determination of the ee values, most substrates were
deprotected, and only for 22e were we able to determine the
ee directly from the rearrangement product by HPLC. Depro-
tection was in most cases achieved in good to very high yields
under standard conditions using 1,3-diaminopropane as re-
agent.[41] Allyl amines with low boiling points were isolated as
their hydrochloride salts to avoid a loss of material. The N-de-
protection of the ester-containing 22 f was problematic, result-
ing in decomposition. The enantiomeric excess of 22 f could
not be determined thus far.
Mechanistic considerations
To gain information how the title reaction proceeds, different
mechanistic studies were executed. The observation that the
reaction is stereospecific, providing different major product
enantiomers depending on the configuration of the olefin
moiety pointed to a mechanistic similarity to the PdII-catalysed
allylic imidate rearrangements. For the latter, the stereospeci-
ficity has been explained by a cyclic 6-membered (half)chair-
like transition state, in which differentiation of the enantiotopic
olefin faces by a chiral catalyst allows for enantiocontrol.[16d,13]
Different enantiomers result from an accommodation of olefin
E residues in the equatorial position, whereas Z residues adopt
an axial position. This would also explain the lower reactivity
of the Z substrates in the present study by 1,3-diaxial interac-
tions in the transition state of the CÀN bond formation.
In the other cases (if not indicated otherwise) ee values were
determined by HPLC after tosylation of the free amino group
(see the Supporting Information). This revealed that the rear-
rangements of dimethylaminosulfonyl and tosyl-protected al-
lylic substrates both formed the new stereocenter with identi-
cal absolute configuration (comparison of optical rotation and
HPLC data). The configurational outcome was further con-
firmed by comparison of samples of free amine 10d prepared
by either deprotection of 8d (Scheme 5) or 22d (Table 8,
entry 4).
Moreover, in the case of 22e we could show that the dime-
thylaminosulfonyl deprotection proceeded without significant
partial racemisation; identical ee values were determined for
22e itself and for the corresponding N-tosyl-protected 8e ob-
tained by deprotection of 22e and subsequent tosylation.
To further increase the practicality of this version, the use of
the non-activated catalyst was also investigated (Table 9). For
comparable yields an increased catalyst loading was some-
times necessary and all reactions were performed at 808C in
CHCl3. Despite the harsher conditions, high regio- and enantio-
selectivities were again achieved. In fact, for the difficult sub-
To control the assumption of a rearrangement reaction,
cross-over experiments were conducted, which should provide
information if the reaction proceeds through an inter- or intra-
molecular pathway. A 1:1 mixture of substrates 6c and 12 dif-
fering in the N-sulfonyl protecting group and the olefin sub-
stituent was applied to quasi-neat reaction conditions
(Scheme 6). Whereas for an intramolecular pathway only two
reaction products 8c and 17 would be expected, an intermo-
lecular reaction should also provide two additional cross-over
products. The latter were only detectable in trace amounts (<
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5773
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 6. Cross-over experiment supporting an intramolecular reaction
pathway.
1% in GC-MS), whereas both products expected for an intra-
molecular pathway were produced in similarly high quantities.
These data are thus in agreement with a rearrangement mech-
anism.
A pseudo-zero-order kinetic dependence on the carbamate
substrate is suggested by 1H NMR monitoring at a reaction
temperature of 358C using activated catalyst (Figure 2). There
is an almost linear relationship of the product amount formed
Figure 3. ESI-MS of the putative substrate catalyst adduct.
The substrate saturation and loss of the anion XÀ might be
explained by the formation of a chelate complex 24 as the
resting state, in which both the olefin and the anionic carba-
mate moiety are coordinated to the palladium centre thus re-
placing the anionic ligand (Scheme 7).
Such a chelate might explain the observation that the anion-
ic ligand XÀ is less important for the enantioselectivity than is
usually found for other applications of this catalyst type, be-
cause it is replaced by the deprotonated substrate in the enan-
tioface-selective coordination. In this coordination the neutral
olefin moiety should coordinate trans to the oxazoline N-donor
based on the coordination behaviour typically found for ferro-
cene palladacycles.[15e,42] Like in other PdII-olefin complexes, the
olefin moiety should be perpendicular to the Pd-square
plane,[43] that is, parallel to the ferrocene axis in the present
case. For steric reasons we expect that the allylic C-1 methylene
points towards the massive C5Ph5 ligand because of its relative-
ly low steric demand, whereas repulsive interactions between
the N-sulfonyl residue of the substrate and the ferrocene core
are minimised (Scheme 7). Also, for steric reasons the coordina-
tion of the enantiotopic olefin face is expected to be unfavour-
able.
Figure 2. Relationship of product yield and time.
and the reaction time for yields up to about 80%. This is in ac-
cordance with a substrate saturation of the catalyst. Catalyst
decomposition does not seem to be very critical as it would
cause a decrease of the reaction rate with increasing conver-
sion. The high affinity for the substrate to bind to the catalyst
is in contrast to allylic imidate rearrangements using the same
catalyst in which this behaviour was not found.[15e]
Substrate saturation for the allylic carbamate rearrangement
is also in agreement with ESI-HRMS measurements. After a reac-
tion time of 30 min at 358C using substrate 6a and 5 mol% of
[PPFOP-O2CCF3]2, a deprotonated substrate molecule appears
to be bound to the palladacycle in the dominating ferrocene
palladacycle detected (not only carrying the C,N-ligand,
Figure 3).
In the chelate complex 24 the deprotonated carbamate-N
should be cis to the oxazoline N. It might then attack the
olefin through an inner-sphere mechanism to generate the CÀ
N bond. However, this possibility seems to be less likely be-
cause olefin insertions into PdÀY bonds typically proceed by
concerted mechanisms through planar 4-membered cyclic
transitions states.[43] That means for the present case that in
the reactive conformation the C=C double bond would need
to be in the PdII-square plane necessitating a rotation of the
Only traces of the initially generated activated catalyst spe-
cies [PPFOP-O2CCF3]2 were found to be present under these
reaction conditions as judged from 19F NMR experiments. The
latter also suggest that the large majority of trifluoroacetate
anions are not coordinating any more to the catalyst mole-
cules, probably because they have been replaced by the sub-
strate (for details see the Supporting Information).
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5774
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tion of the linear isomer by a formal 1,3-rearrangement could
be explained by a competing allylic substitution through a tra-
ditional p-allyl-Pd complex, which might be generated by oxi-
dative addition from Pd0 formed by partial decomposition of
the palladacycle catalyst.[44] However, as shown above, catalyst
decomposition is not likely to play a significant role as judged
from the kinetic studies.
Alternatively, an oxo-rearrangement might be involved in
the generation of the linear isomer (Scheme 8).
Scheme 7. Proposed mechanism of the palladacycle/base-catalysed decar-
boxylative allylic carbamate rearrangement.
Scheme 8. Rationale for the formation of the formal 1,3-rearrangement
product.
coordinated olefin fragment by around 908.[43] Based on the
steric demand of the shielding C5Ph5 ligand, this is expected to
cause repulsive interactions between the catalyst core and the
substrate. For R=Me the smallest repulsive interaction of this
type would be expected, which might at least partly explain
the lower enantioselectivities in that case, because this path
would be less disfavoured than with bulkier residues R. We
thus favour an alternative scenario of an outer-sphere mecha-
nism that requires a preceding dissociation of the carbamate
N-atom, which might be triggered by the reversible recoordi-
nation of the anionic ligand XÀ. This might also explain the in-
fluence of XÀ on the reactivity. The anionic N-centre would
thus attack from the face remote to the metal. For that reason,
inner- and outer-sphere mechanisms should generate different
enantiomers as major products assuming that the same olefin
face is coordinating. Only the proposed outer-sphere pathway
is in agreement with the determined R-configuration of the
major product enantiomers based on the expected enantio-
face-selective olefin coordination. The lower reactivity of the
more electron-rich and thus more nucleophilic dimethylamino-
sulfonyl-protected substrates 21 suggests that the N-decom-
plexation (and not the CÀN bond formation) might be the
rate-limiting step using activated catalysts [PPFOP-X]2.
Similar oxo-rearrangements have been reported by Overman
for PdII-catalysed rearrangements of allylic N,N-dialkylcarba-
mates.[45] Intramolecular attack of a carbamate oxygen at C=C
would initially cause an isomerisation of the linear to the
branched carbamate derivative like 28i. In the presence of
a bulky branched olefin substituent such as iPr, this competing
reaction path might gain in importance, as the nucleophilic O-
centre in 25i is sterically significantly less demanding than the
tosyl-protected N-centre. The thus-generated branched carba-
mate 28i could then undergo a decarboxylative [3,3]-rear-
rangement, in which the carbamate N-atom is attacking the
less-hindered terminal olefin moiety to finally form the linear
allylic amine derivative regio-8i.
Conclusion
We have reported the development and application of a step-
economic enantio- and regioselective catalytic method to
transform achiral allylic alcohols in a single step into tosyl-pro-
tected chiral allylic amines. These reactions are likely to pro-
ceed through a PdII/tertiary amine catalysed cyclisation-in-
duced decarboxylative [3,3]-rearrangement of in situ-generated
allylic carbamate intermediates, explaining the preference for
the branched allylic product regioisomers and the observed
stereospecificity. Enantioselectivity is explained in analogy to
allylic imidate rearrangements using the same catalyst type by
enantioface-selective olefin coordination and a subsequent
outer-sphere attack of the nucleophilic deprotonated N-centre
to the coordinated olefin. Our mechanistic studies have shown
a substrate saturation of the catalyst up to high conversions,
which can be explained by a two-point coordination of the
The subsequent CÀN bond formation would provide the
cyclic aminopalladated intermediate 26 featuring a s-alkyl-Pd
bond. The deprotonated carbaminic acid derivative 27 is gen-
erated by a ring-opening b-elimination. Decomplexation, de-
carboxylation and protonation furnish the N-sulfonylated allylic
amine 8 and regenerate the palladium(II) and base catalysts for
the next turnovers.
As shown above, the ratio of branched to linear amine is de-
pendent on the steric bulk of the olefin substituent R. Forma-
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5775
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
substrate in the resting state. The operational simplicity is no-
table, in particular, as there is no need for catalyst activation
and an inert gas atmosphere. Whereas the allylic carbamate re-
arrangement could not be successfully applied to N-carbonyl-
protected substrates so far, its synthetic utility was further en-
hanced by the effective use of N,N-dimethylaminosulfonyl-pro-
tected allylic carbamates, because this protective group can be
removed under non-reductive conditions. The dimethylamino-
sulfonyl group thus behaves complementary to the tosyl
moiety. In addition, the dimethylamino group allowed for a fur-
ther improvement of the enantioselectivity. We believe that
the title reaction belongs to the most practical catalytic asym-
metric ways to form highly enantioenriched allylic amines.
Keywords: allylic compounds
enantioselectivity · palladium · regioselectivity
·
domino reactions
·
[1] S. V. Chandramouli, T. A. Ayers, X.-D. Wu, L. T. Tran, J. H. Peers, R. Disan-
to, F. Roberts, N. Kumar, Y. Jiang, N. Choy, C. Pemberton, M. R. Powers,
A. J. Gardetto, G. A. D’Netto, X. Chen, J. Gamboa, D. Ngo, W. Copeland,
D. E. Rudisill, A. W. Bridge, B. J. Vanasse, D. J. Lythgoe, Org. Process Res.
Dev. 2012, 16, 484.
[2] Selected examples: a) M. Nakanishi, M. Mori, Angew. Chem. Int. Ed.
2002, 41, 1934; Angew. Chem. 2002, 114, 2014; b) M. Mori, M. Nakanishi,
D. Kajishima, Y. Sato, Org. Lett. 2001, 3, 1913; c) C. Welter, R. M. Moreno,
S. Streiff, G. Helmchen, Org. Biomol. Chem. 2005, 3, 3266.
[3] Review: S. Nag, S. Batra, Tetrahedron 2011, 67, 8959.
[4] Selected early studies: a) T. Hayashi, A. Yamamoto, Y. Ito, E. Nishioka, H.
Miura, K. Yanagi, J. Am. Chem. Soc. 1989, 111, 6301; b) R. Jumnah, J. M. J.
Williams, A. C. Williams, Tetrahedron Lett. 1993, 34, 6619; c) J. F. Bower,
R. Jumnah, A. C. Williams, J. M. J. Williams, J. Chem. Soc. Perkin Trans.
1 1997, 1411; d) K. Burgess, L. T. Liu, B. Pal, J. Org. Chem. 1993, 58, 4758.
[5] Review: E. M. Skoda, G. C. Davis, P. Wipf, Org. Process Res. Dev. 2012, 16,
26.
[6] Selected examples with different metals: Pd: a) I. Dubovyk, I. D. G.
Watson, A. K. Yudin, J. Org. Chem. 2013, 78, 1559; Ir: b) H. Miyabe, Y.
Takemoto, Synlett 2005, 1641; c) G. Helmchen, A. Dahnz, P. Dübon, M.
Schelwies, R. Weihofen, Chem. Commun. 2007, 675; d) R. Takeuchi, S.
Kezuka, Synthesis 2006, 3349; e) O. V. Singh, H. Han, J. Am. Chem. Soc.
2007, 129, 774; f) R. Weihofen, O. Tverskoy, G. Helmchen, Angew. Chem.
Int. Ed. 2006, 45, 5546; Angew. Chem. 2006, 118, 5673; g) R. Weihofen,
A. Dahnz, O. Tverskoy, G. Helmchen, Chem. Commun. 2005, 3541; h) Y.
Yamashita, A. Gopalarathnam, J. F. Hartwig, J. Am. Chem. Soc. 2007, 129,
7508; i) M. Roggen, E. M. Carreira, J. Am. Chem. Soc. 2010, 132, 11917;
Ru: j) M. Kawatsura, K. Uchida, S. Terasaki, H. Tsuji, M. Minakawa, T. Itoh,
Org. Lett. 2014, 16, 1470.
[7] Selected examples: a) J. R. Porter, W. G. Wirschun, K. W. Kuntz, M. L.
Snapper, A. H. Hoveyda, J. Am. Chem. Soc. 2000, 122, 2657; b) S. Hirner,
A. Kolb, J. Westmeier, S. Gebhardt, S. Middel, K. Harms, P. von Zezsch-
witz, Org. Lett. 2014, 16, 3162; c) M.-Y. Ngai, A. Barchuk, M. J. Krische, J.
Am. Chem. Soc. 2007, 129, 12644; d) B. Gopula, C.-W. Chiang, W.-Z. Lee,
T.-S. Kuo, P.-Y. Wu, J. P. Henschke, H.-L. Wu, Org. Lett. 2014, 16, 632; e) Y.
Wei, M. Shi, Chem. Rev. 2013, 113, 6659; f) K. Hyodo, S. Nakamura, N.
Shibata, Angew. Chem. Int. Ed. 2012, 51, 10337; Angew. Chem. 2012, 124,
10483.
Experimental Section
General procedure for the catalytic asymmetric decarboxyla-
tive rearrangement of N-sulfonyl-protected allylic carba-
mates by non-activated [PPFOP-Cl]2
The corresponding N-sulfonyl protected carbamate (1 equiv),
proton
sponge
(1,8-bis(N,N-dimethylamino)naphthalene)
(0.2 equiv), and [PPFOP-Cl]2 (usually 1–2 mol%) were charged to
a screw-cap vial. Vacuum was then applied and the vial was subse-
quently flushed with nitrogen (3 times repeated). Then CHCl3 (typi-
cally 150 mL per 100 mmol) was added, the vial was closed and the
mixture was stirred for the indicated time at the indicated temper-
ature. Afterwards the mixture was diluted with CH2Cl2 and washed
with aqueous 1m HCl. The aqueous layer was extracted twice with
CH2Cl2, the combined organic layers were dried over Na2SO4 and
the solvent was removed under reduced pressure. After determina-
tion of the regioselectivity by 1H NMR spectroscopy, the product
was isolated by silica gel chromatography. The purified samples
were used to determine the ee value by HPLC.
General domino procedure for the catalytic asymmetric syn-
thesis of N-tosyl-protected allylic amines starting from allylic
alcohols
[8] a) T.-L. Liu, C.-J. Wang, X. Zhang, Angew. Chem. Int. Ed. 2013, 52, 8416;
Angew. Chem. 2013, 125, 8574; b) M. Zhou, T.-L. Liu, M. Cao, Z. Xue, H.
Lv, X. Zhang, Org. Lett. 2014, 16, 3484.
[9] a) M. L. Cooke, K. Xu, B. Breit, Angew. Chem. Int. Ed. 2012, 51, 10876;
Angew. Chem. 2012, 124, 11034; b) K. Xu, Y.-H. Wang, V. Khakyzadeh, B.
Breit, Chem. Sci. 2016, DOI: 10.1039/C5SC04984A.
[10] S. Kobbelgaard, S. Brandes, K. A. Jørgensen, Chem. Eur. J. 2008, 14,
1464.
[11] H. Bao, U. K. Tambar, J. Am. Chem. Soc. 2012, 134, 18495.
[12] Selected reviews: a) H. Nomura, C. J. Richards, Chem. Asian J. 2010, 5,
1726; b) T. K. Hollis, L. E. Overman, J. Organomet. Chem. 1999, 576, 290;
c) L. E. Overman, N. E. Carpenter, Org. React. 2005, 66, 1; d) R. Peters,
D. F. Fischer, S. Jautze, Top. Organomet. Chem. 2011, 33, 139.
[13] M. P. Watson, L. E. Overman, R. G. Bergman, J. Am. Chem. Soc. 2007, 129,
5031.
A dry screw-cap vial was charged under air with [PPFOP-Cl]2
(1 mol%), proton sponge (1,8-bis(N,N-dimethylamino)naphthalene)
(0.2 equiv), the corresponding allylic alcohol (9, 1 equiv) and p-to-
sylisocyanate (1.0 equiv). Subsequently, CHCl3 (100 mL per 68 mmol
allylic alcohol) was added. The vial was closed and the mixture
stirred for 24 h at 808C. The solution was then diluted with CH2Cl2
and washed with aqueous 1m HCl. The layers were separated and
the aqueous phase was extracted twice with CH2Cl2. The combined
organic layers were dried over Na2SO4 and the solvent removed
under reduced pressure. After determination of the regioselectivity
by 1H NMR spectroscopy, the product was isolated by silica gel
chromatography. The purified samples were used to determine the
ee values by HPLC.
[14] M. Calter, T. K. Hollis, L. E. Overman, J. Ziller, G. G. Zipp, J. Org. Chem.
1997, 62, 1449.
[15] Trichloroacetimidates: a) Y. Donde, L. E. Overman, J. Am. Chem. Soc.
1999, 121, 2933; b) C. E. Anderson, L. E. Overman, J. Am. Chem. Soc.
2003, 125, 12412; c) S. F. Kirsch, L. E. Overman, M. P. Watson, J. Org.
Chem. 2004, 69, 8101; d) H. Nomura, C. J. Richards, Chem. Eur. J. 2007,
13, 10216; e) D. F. Fischer, A. Barakat, Z.-q. Xin, M. E. Weiss, R. Peters,
Chem. Eur. J. 2009, 15, 8722.
Acknowledgements
[16] Trifluoroacetimidates: see ref. [15e] and: a) L. E. Overman, C. E. Owen,
M. M. Pavan, C. J. Richards, Org. Lett. 2003, 5, 1809; b) R. S. Prasad, C. E.
Anderson, C. J. Richards, L. E. Overman, Organometallics 2005, 24, 77;
c) C. E. Anderson, Y. Donde, C. J. Douglas, L. E. Overman, J. Org. Chem.
2005, 70, 648; d) M. E. Weiss, D. F. Fischer, Z.-q. Xin, S. Jautze, W. B.
This work was financially supported by the Deutsche For-
schungsgemeinschaft (DFG, PE 818/4-1 and PE 818/4-2). The
Fonds der Chemischen Industrie (FCI) and the Carl-Zeiss-Stif-
tung are acknowledged for Ph.D. fellowships to J.M.B.
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5776
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Schweizer, R. Peters, Angew. Chem. Int. Ed. 2006, 45, 5694; Angew.
Chem. 2006, 118, 5823; e) S. Jautze, P. Seiler, R. Peters, Angew. Chem. Int.
Ed. 2007, 46, 1260; Angew. Chem. 2007, 119, 1282; f) D. F. Fischer, Z.-q.
Xin, R. Peters, Angew. Chem. Int. Ed. 2007, 46, 7704; Angew. Chem. 2007,
119, 7848; g) S. Jautze, P. Seiler, R. Peters, Chem. Eur. J. 2008, 14, 1430;
h) R. Peters, Z.-q. Xin, F. Maier, Chem. Asian J. 2010, 5, 1770; i) S. H. Eitel,
M. Bauer, D. Schweinfurth, N. Deibel, B. Sarkar, H. Kelm, H.-J. Krüger, W.
Frey, R. Peters, J. Am. Chem. Soc. 2012, 134, 4683; j) M. Weiss, W. Frey, R.
Peters, Organometallics 2012, 31, 6365.
2013, 19, 4128; h) M. Kondo, T. Nishi, T. Hatanaka, Y. Funahashi, S. Naka-
mura, Angew. Chem. Int. Ed. 2015, 54, 8198; Angew. Chem. 2015, 127,
8316; i) M. Kondo, N. Kobayashi, T. Hatanaka, Y. Funahashi, S. Nakamura,
Chem. Eur. J. 2015, 21, 9066.
[34] L. D. Taylor, M. Pluhar, L. E. Rubin, J. Polym. Sci. B Polym. Lett. 1967, 5,
77.
[35] CCDC 990155 (8a), contains the supplementary crystallographic data
for this paper. These data are provided free of charge by The Cam-
bridge Crystallographic Data Centre.
[17] See ref. [15e] and Z.-q. Xin, D. F. Fischer, R. Peters, Synlett 2008, 1495.
[18] J. M. Bauer, R. Peters, Catal. Sci. Technol. 2015, 5, 2340.
[19] M. E. Synerholm, N. W. Gilman, J. W. Morgan, R. K. Hill, J. Org. Chem.
1968, 33, 1111.
[20] C. L. J. Wang, J. C. Calabrese, J. Org. Chem. 1991, 56, 4341.
[21] A. Lei, X. Lu, Org. Lett. 2000, 2, 2357.
[22] For a related protocol see: S. Christie, A. Warrington, C. Lunniss, Synthe-
sis 2009, 148.
[23] For a PdII-catalyzed asymmetric synthesis of oxazolidinones see: a) L. E.
Overman, T. P. Remarchuk, J. Am. Chem. Soc. 2002, 124, 12; b) S. F.
Kirsch, L. E. Overman, J. Org. Chem. 2005, 70, 2859.
[36] For in situ carbamate formation in: a) aminopalladations, see ref. [23a];
b) allylic substitutions: B. M. Trost, D. E. Patterson, J. Org. Chem. 1998,
63, 1339.
[37] Selected reviews: a) Domino Reactions: Concepts for Efficient Organic
Synthesis (Ed.: L. F. Tietze), Wiley-VCH, Weinheim, 2014; b) L.-Q. Lu, J.-R.
Chen, W.-J. Xiao, Acc. Chem. Res. 2012, 45, 1278; c) H. Clavier, H. Pellissi-
er, Adv. Synth. Catal. 2012, 354, 3347.
[38] Catalytic asymmetric intermolecular allylic substitutions using N-centred
nucleophiles and allylic alcohols: a) see ref. [6h]; b) C. Defieber, M. A.
Ariger, P. Moriel, E. M. Carreira, Angew. Chem. Int. Ed. 2007, 46, 3139;
Angew. Chem. 2007, 119, 3200.
[24] D. Xing, D. Yang, Org. Lett. 2010, 12, 1068.
[25] S. R. Mellegaard-Waetzig, D. K. Rayabarapu, J. A. Tunge, Synlett 2005,
2759.
[39] S. Tsushima, Y. Yamada, T. Onami, K. Oshima, M. O. Chaney, N. D. Jones,
J. K. Swartzendruber, Bull. Chem. Soc. Jpn. 1989, 62, 1167. A sample with
ee=93% was used.
[26] a) O. V. Singh, H. Han, Org. Lett. 2007, 9, 4801; b) see ref. [6e].
[27] For a review about cascades involving catalysis and sigmatropic rear-
rangements, see: A. C. Jones, J. A. May, R. Sarpong, B. M. Stoltz, Angew.
Chem. Int. Ed. 2014, 53, 2556; Angew. Chem. 2014, 126, 2590.
[28] Preliminary results have been reported as a communication: J. M.
Bauer, W. Frey, R. Peters, Angew. Chem. Int. Ed. 2014, 53, 7634; Angew.
Chem. 2014, 126, 7764.
[40] R. B. C. Jagt, P. Y. Toullec, D. Geerdink, J. G. de Vries, B. L. Feringa, A. J.
Minnaard, Angew. Chem. Int. Ed. 2006, 45, 2789; Angew. Chem. 2006,
118, 2855.
[41] S. Källstrçm, T. Saloranta, A. J. Minnaard, R. Leino, Tetrahedron Lett.
2007, 48, 6958.
[42] C. J. Richards, in Chiral Ferrocenes in Asymmetric Catalysis, (Eds.: L.-X.
Dai, X.-L. Hou), Wiley-VCH, Weinheim, 2010, pp. 337–368.
[43] a) Handbook of Organopalladium Chemistry for Organic Synthesis (Ed.: E.-
i. Negishi), Wiley, New York, 2002; b) K. N. Fanning, A. G. Jamieson, A.
Sutherland, Curr. Org. Chem. 2006, 10, 1007.
[44] One referee suggested to also consider an alternative dissociative
mechanism for the enantio- and regioselective product formation that
involves an in situ-generated chiral Pd0 catalyst species. In this case an
allyl palladium complex and a carbamate anion might form that could
combine prior to decarboxylation. The observation of an intramolecular
course might be explained by Coulomb interactions between the allyl
complex and the carbamate anion to avoid a dissociation prior to com-
bination in a regioselective manner. We think that such a scenario is
unlikely because our palladacycle catalysts are not stable on the oxida-
tion state of Pd0 and Pd black is then usually formed (see, for example,
ref. [31e]). Moreover, such an allyl complex would be uncharged due to
the anionic C,N-ligand in contrast to the large majority of Pd0 catalysts
used for allylic substitutions (see for example, ref. [43a], [6a] and refer-
ences therein). With a Pd0 catalyst the linear isomer might be expected
to dominate. In addition, competition of intra- and intermolecular path-
ways might be expected for a dissociative pathway depending on the
concentration used. However, the results in Scheme 6 were obtained
under almost neat reaction conditions.
[29] Cooperative Catalysis-Designing Efficient Catalysts for Synthesis, (Ed.: R.
Peters), Wiley-VCH, Weinheim, 2015.
[30] Selected catalytic applications of ferrocenyl monopalladacycles in asym-
metric catalysis: a) C. Schrapel, R. Peters, Angew. Chem. Int. Ed. 2015, 54,
10289; Angew. Chem. 2015, 127, 10428; b) T. Hellmuth, W. Frey, R.
Peters, Angew. Chem. Int. Ed. 2015, 54, 2788; Angew. Chem. 2015, 127,
2829; c) M. Weber, R. Peters, J. Org. Chem. 2012, 77, 10846.
[31] Selected catalytic applications of metallocendiyl bispalladacycles in
asymmetric catalysis: a) M. Weiss, R. Peters, ACS Catal. 2015, 5, 310; b) T.
Hellmuth, S. Rieckhoff, M. Weiss, K. Dorst, W. Frey, R. Peters, ACS Catal.
2014, 4, 1850; c) M. Weber, W. Frey, R. Peters, Angew. Chem. Int. Ed.
2013, 52, 13223; Angew. Chem. 2013, 125, 13465; d) M. Weber, W. Frey,
R. Peters, Chem. Eur. J. 2013, 19, 8342; e) S. H. Eitel, S. Jautze, W. Frey, R.
Peters, Chem. Sci. 2013, 4, 2218; f) M. Weber, S. Jautze, W. Frey, R.
Peters, Chem. Eur. J. 2012, 18, 14792.
[32] A similar observation was recently also made for ferrocenyl gold com-
plexes: M. Ayerbe Garcia, W. Frey, M. Ringenberg, M. Schwilk, R. Peters,
Chem. Commun. 2015, 51, 16806–16809.
[33] For applications of imidazolines as chiral ligands in asymmetric cataly-
sis, see, for example, a) review: H. Liu, D.-M. Du, Adv. Synth. Catal. 2009,
351, 489; b) H. Huang, R. Peters, Angew. Chem. Int. Ed. 2009, 48, 604;
Angew. Chem. 2009, 121, 612; c) H. Liu, D.-M. Du, Adv. Synth. Catal.
2010, 352, 1113; d) M. Ohara, S. Nakamura, N. Shibata, Adv. Synth. Catal.
2011, 353, 3285; e) K. Hyodo, S. Nakamura, K. Tsuji, T. Ogawa, Y. Funaha-
shi, N. Shibata, Adv. Synth. Catal. 2011, 353, 3385; f) S. Nakamura, K.
Hyodo, M. Nakamura, D. Nakane, H. Masuda, Chem. Eur. J. 2013, 19,
7304; g) K. Hyodo, M. Kondo, Y. Funahashi, S. Nakamura, Chem. Eur. J.
[45] a) L. E. Overman, C. B. Campbell, J. Org. Chem. 1976, 41, 3338; b) L. E.
Overman, C. B. Campbell, F. M. Knoll, J. Am. Chem. Soc. 1978, 100, 4822.
Received: January 12, 2016
Published online on March 15, 2016
Chem. Eur. J. 2016, 22, 5767 – 5777
www.chemeurj.org
5777
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim