Copper(I) catalysts for the stereoselective addition of N-chloroamines to double bonds: A diastereoselective radical cyclisation

Article abstract of DOI:10.1002/1099-0690(200206)2002:11<1848::AID-EJOC1848>3.0.CO;2-V

Copper(I) catalysts for the diastereoselective radical cyclisation of N-chloro-N-pentenylamines have been developed. The stereoselectivity of the cyclisation depends upon the ligands employed, proving that the radical is bound to the catalyst during the formation of the new stereocentre and making a catalyst-influenced stereoselective radical reaction possible. The influence of the catalyst on the Beckwith-Houk transition states is discussed. Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002).

Full text of DOI:10.1002/1099-0690(200206)2002:11<1848::AID-EJOC1848>3.0.CO;2-V

FULL PAPER
Copper(I) Catalysts for the Stereoselective Addition of N-Chloroamines to
Double Bonds: A Diastereoselective Radical Cyclisation
Gerold Heuger,[a] Stefanie Kalsow,[a] and Richard Göttlich*[a]
Keywords: Homogeneous catalysis / Cyclization / Nitrogen heterocycles / Radicals / Diastereoselectivity
Copper(I) catalysts for the diastereoselective radical cyclis-
ation of N-chloro-N-pentenylamines have been developed.
The stereoselectivity of the cyclisation depends upon the li-
gands employed, proving that the radical is bound to the
catalyst during the formation of the new stereocentre and
making a catalyst-influenced stereoselective radical reaction
possible. The influence of the catalyst on the Beckwith−Houk
transition states is discussed.
( Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany,
2002)
Introduction
Pyrrolidines and piperidines are important substructures
of a variety of biologically active compounds.[1] We are in-
vestigating new methods for their stereoselective synthesis
by the catalysed intramolecular addition of nitrogenϪ
heteroatom bonds to double bonds.[2] This transformation
can be achieved via nitrogen-centred radicals under acidic
conditions, which are necessary to protonate the radical,
thus making it more electrophilic for an efficient addition
to a double bond.[3] In a recent publication[4] we reported
on milder reaction conditions using copper(I) salts as cata-
lysts for the radical cyclisation of N-chloro-N-pentenyl-
amines under aprotic conditions (Figure 1). In this reaction
the catalyst acts as an initiator of the radical reaction (A),
as activator for the aminyl radical cyclisation[5] (B) and as
scavenger for the carbon radical formed (C).[6]
The initial product, a 2-(chloromethyl)pyrrolidine, re-
arranges under these almost neutral conditions via an aziri-
dinium ion in two subsequent SN2 reactions[7] to the 3-chlo-
ropiperidine,[8] thereby allowing the synthesis of piperidines
via a kinetically favourable 5-exo cyclisation. As the aminyl
radical is coordinated to the catalyst in the addition step,[9]
an influence of the catalyst structure and ligand sphere on
the stereochemical outcome of the reaction can be expected.
This could allow an efficient radical cyclisation with a cata-
lyst-induced stereoselectivity.[10]
Figure 1. Catalytic cycle
tions can be greatly increased by using copper(I) catalysts
together with chelating ligands.
In a first set of experiments we analysed the diastereose-
lectivity of aminyl radical cyclisations using our previously
reported conditions[4] and various N-chloro-N-pentenylam-
ines containing a chiral centre. To do so, the chloroamines
1aϪ1f were synthesised. The precursors 1b, 1c and 1f were
easily available using a Johnson orthoester Claisen re-
arrangement[11] as the key step (Figure 2).[12]
Results and Discussion
Here we wish to report on this influence of the catalyst
and that the diastereoselectivity in aminyl radical cyclisa-
For the synthesis of 1b and 1c the initially formed esters
were saponified to the acids 2b and 2c. These were trans-
formed via the acyl chlorides into the corresponding N-bu-
tylamides 3b and 3c. Reduction of the amides gave the
amines 4b and 4c in good yield,[13] from which the N-chlo-
[a]
Organisch-Chemisches Institut der Westfälischen Wilhelms-
Universität Münster,
Corrensstr. 40, 48149 Münster, Germany
Fax: (internat.) ϩ 49-(0)251/833-6529
E-mail: gottlir@uni-muenster.de
1848
 WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 1434Ϫ193X/02/0611Ϫ1848 $ 20.00ϩ.50/0 Eur. J. Org. Chem. 2002, 1848Ϫ1854

Copper(I) Catalysts for the Stereoselective Addition of N-Chloroamines to Double Bonds
FULL PAPER
Figure 3. Synthesis of chloroamine 1d
Table 1. Cyclisation of the N-chloroamines using 10 mol % CuCl
in THF at 50 °C
Figure 2. Synthesis of the chloroamines by a Johnson orthoester
Claisen rearrangement
roamines were formed by reaction with NCS. For the syn-
thesis of 1f this procedure failed, due to isomerisation of
the double bond in conjugation with the aromatic ring. We
therefore reduced the initially formed ester 5 to the alcohol
6, which was either tosylated to 7 or transformed into the
iodide 8 prior to the reaction with butylamine to give the
amine 4f. Both reaction sequences gave satisfying yields, al-
though the workup of the product resulting from the tosyl-
ate proved to be difficult on a larger scale. The amine 4f
was transformed into the N-chloroamine 1f in good yield
by reaction with NCS.
For the synthesis of 1d a different reaction sequence was
applied. Alcohol 9 was obtained by the addition of a but-
enyl Grignard reagent to benzaldehyde. Substitution of the
hydroxy group with butylamine via the corresponding iod-
ide led to amine 4d (Figure 3).
reactions a remarkable excess of one of the two possible
diastereomers was observed (Table 1).
This diastereomeric ratio is notably higher than in the
cyclisation reactions of the corresponding protonated ami-
nyl radicals,[14] thus proving that the metal catalyst indeed
has a beneficial influence on the stereoselectivity of the
cyclisation. This effect results from the two
BeckwithϪHouk transition states[15] shown in Figure 4. We
believe that the copper(II) centre not only activates the ami-
nyl radical but also coordinates the alkene, thereby leading
to a more rigid transition state.
From this amine the N-chloroamine 1d was obtained in
good yield under standard conditions. With the chloroam-
ines being readily available, we first tested our initially re-
ported conditions[4] for the cyclisation of 1aϪ1f. In all cases
moderate to good yields of the corresponding 3-chloropip-
Figure 4. BeckwithϪHouk transition states for the cyclisation of
eridines 10aϪ10f were obtained. Furthermore, in all these the aminyl radical; B leads to the preferred product
Eur. J. Org. Chem. 2002, 1848Ϫ1854
1849

G. Heuger, S. Kalsow, R. Göttlich
FULL PAPER
Figure 5. First experiments with addition of ligands
In the preferred chair-like transition state B all groups,
including the copper ion, are in an equatorial position. In
contrast to this, in the less favourable boat-like transition
state A, the coordination of the double bond by the metal
ion forces the sterically demanding copper salt into an unfa-
vourable pseudo-axial position. The transition state A
should therefore become even more unfavourable as the size
of the metal complex increases. This not only explains the
observed high diastereomeric excess, but, if this hypothesis
holds true, the addition of sterically demanding ligands to
the catalyst should lead to an increase in diastereoselectiv-
ity.
To verify this assumption chelating diamines were added
as ligands for the copper ion. Treatment of the chloroamine
1g with these catalysts leads to a low yield of piperidine
10g, along with unchanged chloroamine 1g and the corres-
ponding secondary amine 4g as the main products of the
reaction (Figure 5). The low yield of the cyclisation product
could be due to the reduced Lewis acidity of the copper
complex, which now cannot activate the aminyl radical for
the addition step.
Figure 6. Varying the metal salt and ligand
We therefore turned our attention towards more Lewis to any cyclisation product at all. Traces of water that might
acidic copper(I) salts with weakly coordinating counterions, have been present are also not responsible for the observed
such as copper(I) hexafluorophosphate.[16] With this com- yield, as addition of water to the reaction mixture reduced
plex as a catalyst, and without any addition of further li- the yield.[19] Therefore, the surprising difference between
gands, low yields were also obtained. However, this might TMEDA and TMCDA as ligands can only be attributed to
be due to a now too high Lewis acidity of the reaction mix- the fixed bite-angle of TMCDA, which seems to be ideal
ture; Newcomb has reported that Lewis acid catalysed ami- for this reaction.[20] We therefore checked the diastereoselec-
nyl radical cyclisations produce lower yields under too tivity of the cyclisation using 10 mol % catalyst together
Lewis acidic conditions.[17] Therefore, this copper salt with different amounts of ligand (Figure 7). To our surprise
seemed ideal for further studies, as the addition of ligands an excess of ligand was necessary to obtain an optimal dia-
should reduce the Lewis acidity of the catalyst. This proved stereoselectivity. This resulted in the formation of only one
to be correct and higher yields were obtained upon addition stereoisomer when 5 equiv. or more of ligand were used,
of co-ligands. We next checked the cyclisation reaction of whilst addition of further excess of ligand did not have a
chloroamine 1a with different ligands. In all cases, upon pronounced effect on the yield of the cyclisation.
using copper(I) hexafluorophosphate and chelating ligands,
The diastereoselectivity of the cyclisation step is consider-
the piperidine 10a formed, although the varying yields ob- ably higher than with copper(I) chloride alone or than
tained are surprising. Even though there seems to be no observed in the cyclisation of the corresponding carbon
great difference between TMCDA (trans-N,N,NЈ,NЈ-tetra- radicals.[21] This exceptionally high diastereoselectivity in a
methylcyclohexanediamine) and TMEDA, the former pro- radical cyclisation is probably due to the two
duced a much higher yield of cyclisation product when used BeckwithϪHouk transition states shown in Figure 4, with
as ligand (Figure 6).[18]
the copper ion generally avoiding a pseudo-axial position.
An explanation for this unusual effect might be that In the case of chloroamine 1a there is an additional effect;
TMCDA is a catalyst in its own right, reacting with the in transition state A the copper is in a gauche conformation
chloroamine by abstracting a chloronium ion, which, in to two alkyl groups, whereas in B there is only one of these
turn, could add to the double bond. However, performing unfavourable interactions, which favours B over A even
the reaction with TMCDA and no copper salt did not lead more strongly.
1850
Eur. J. Org. Chem. 2002, 1848Ϫ1854

Copper(I) Catalysts for the Stereoselective Addition of N-Chloroamines to Double Bonds
FULL PAPER
thesis of pharmacologically active compounds and natural
products.
Experimental Section
General: All solvents were purified by distillation and dried, if
necessary, prior to use. The reactions were carried out under va-
cuum in heat-dried glassware under argon. Products were purified
by flash chromatography on silica gel (40Ϫ63 µm). NMR spectra
were recorded with a Bruker WM 300, a Bruker AMX 400 or a
Varian Unity plus 600 spectrometer in CDCl3 with TMS as internal
standard. Mass spectra were recorded with a MAT 8200, a MAT
8230 or a Micromars Quattro LC spectrometer. Elemental analyses
were performed with a CHN-O Rapid from Foss-Heraeus or a Va-
rio EL III from Elementar Analysensysteme.
Figure 7. High diastereoselectivity with 5 equiv. of ligand
General Procedure for the Synthesis of Pentenoic Acids 2: The cor-
responding acid (acetic or propionic acid; 1 mL) was added to a
mixture of 0.5 mol of triethyl orthoester and 0.5 mol of the allylic
alcohol and the solution heated slowly to 150 °C. The reaction
mixture was kept at this temperature until 2 equiv. (1.0 mol) of
ethanol had distilled off (approx. 2Ϫ5 h). After cooling to room
We therefore decided to check the diastereoselectivity
without this additional ‘‘gauche effect’’, and thus per-
formed studies on the cyclisation of chloroamine 1e, in
which the directing group is one carbon atom further away
from the aminyl radical (Figure 8). As expected, the cyclis- temp., this crude pentenoic ester was added to a solution of potas-
sium hydroxide (40 g) in 300 mL of ethanol and the mixture heated
under reflux for 3 h. The solvent was then evaporated under va-
cuum and the residue dissolved in 300 mL of water. Concd. hydro-
chloric acid was added to this mixture at 0 °C until pH ϭ 1Ϫ2.
The solution was extracted three times with 150 mL of TBDME
(tert-butyl methyl ether), the combined organic phases were dried
with magnesium sulfate, the solvent was evaporated and the
product purified by distillation.
ation of this chloroamine led to reduced, but still high, dia-
stereoselectivities, although the increase in stereoselectivity
upon going from simple copper(I) chloride to CuPF6 to-
gether with chelating ligands as catalyst is still remarkable
and can again be attributed to the bulky ligand sphere of
the catalyst, which avoids the pseudo-axial orientation of
transition state A. In this system TMEDA proved to be the
best ligand and a diastereomeric ratio of 10:1 was achieved,
which is still extraordinarily high compared to the cyclis-
ation of carbon radicals.[21]
2-Methylpent-4-enoic Acid (2b): Yield: 37.2 g (77% over 2 steps) of
a colourless liquid, b.p. 102Ϫ104° C (50 mbar). 1H NMR (300
MHz, CDCl3): δ ϭ 1.19 (d, J ϭ 6.9 Hz, 3H), 2.21 (m, 1 H), 2.44
(m, 1 H), 2.54 (m, 1 H), 5.07 (m, 2 H), 5.77 (ddt, J ϭ J ϭ 17.7, 10.2,
6.9 Hz, 1 H), 11.53 (s, 1 H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ
16.2, 37.4, 39.2, 117.1, 135.1, 182.7 ppm. These data are in accord-
ance with published data.[22]
3-Methylpent-4-enoic Acid (2c): Yield: 18.4 g (34% over 2 steps) of
a colourless liquid, b.p. 72 °C (20 mbar). 1H NMR (300 MHz,
CDCl3): δ ϭ 1.10 (d, J ϭ 6.6 Hz, 3 H), 2.32 (dd, J ϭ 15.1, 7.2 Hz,
1 H), 2.43 (dd, J ϭ 15.1, 7.2 Hz, 1 H), 2.70 (m, 1 H), 4.98 (dt, J ϭ
10.2, 1.3 Hz, 1 H), 5.05 (dt, J ϭ 17.1, 1.3 Hz, 1 H), 5.79 (ddd, J ϭ
17.4, 10.3, 7.5 Hz, 1 H), 11.44 (s, 1 H) ppm. 13C NMR (75 MHz,
CDCl3): δ ϭ 19.6, 34.1, 41.0, 113.5, 142.1, 178.2 ppm. These data
are in accordance with published data.[22]
General Procedure for the Synthesis of N-Butylpent-4-enamides 3:
Thionyl chloride (0.22 mol) was added to 0.2 mol of the carboxylic
acid, the mixture stirred for 1 h, and then heated to 50 °C for 4 h
to complete the reaction. The resulting acyl chloride was slowly
added to butylamine (2 mol), cooled in an ice bath, and the re-
sulting mixture stirred overnight at room temp. Excess butylamine
was removed in vacuo and the residue taken up in 200 mL of
TBDME. This solution was extracted twice with 100 mL of 2 
hydrochloric acid, dried with sodium sulfate and the solvent re-
moved in vacuo. The residue, a yellow oil, was purified by bulb-to-
bulb distillation.
Figure 8. Diastereoselectivity in the cyclisation of chloroamine 1e
Conclusion
In summary we have demonstrated the influence of li-
gands on the stereochemistry of the copper(I)-catalysed
radical cyclisation of unsaturated N-chloroamines, enabling
us to perform a stereoselective catalytic radical reaction. We
believe that this new and highly stereoselective synthesis of
N-Butyl-2-methylpent-4-enamide (3b): Yield: 30.1 g (88% over 2
1
steps) of a colourless oil. H NMR (300 MHz, CDCl3): δ ϭ 0.92
functionalised piperidines will prove valuable for the syn- (t, J ϭ 7.2 Hz), 1.14 (d, J ϭ 6.6 Hz), 1.34 (m, 2 H), 1.46 (m, 2 H),
Eur. J. Org. Chem. 2002, 1848Ϫ1854 1851

G. Heuger, S. Kalsow, R. Göttlich
FULL PAPER
2.14 (m, 1 H), 2.23 (sept, J ϭ 6.6 Hz, 1 H), 2.38 (m, 1 H), 3.24 (m, 142.5, 171.7 ppm. These data are in accordance with published
2 H), 5.05 (m, 2 H), 5.57 (s, 1 H), 5.75 (ddt, J ϭ 17.4, 10.2, 6.9 Hz,
1 H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 13.7, 17.3, 20.0, 31.7,
38.4, 39.1, 41.3, 116.7, 135.9, 175.0 ppm.
data.[23]
3-Phenylpent-4-en-1-ol (6): A solution of ethyl 3-phenylpent-4-eno-
ate (5) (14.3 g, 70 mmol) in 50 mL of TBDME was added to a
suspension of lithium aluminium hydride (2.65 g, 70 mmol) in 100
mL of TBDME at 0 °C through a dropping funnel. The reaction
mixture was stirred overnight at room temperature before being
poured onto 200 mL of ice/water. Sulfuric acid (5 ) was added
until the alumina salts dissolved. The phases were then separated
and the aqueous phase extracted twice with 100 mL of TBDME.
The combined organic phases were dried with sodium sulfate, the
solvent removed in vacuo and the residue purified by flash chroma-
tography to give 10.58 g (65 mmol, 93%) of a colourless oil. 1H
NMR (300 MHz, CDCl3): δ ϭ 1.66 (s, 1 H), 1.94 (m, 2 H), 3.45
(q, J ϭ 7.2 Hz, 1 H), 3.61 (m, 2 H), 5.06 (m, 2 H), 5.96 (ddd, J ϭ
17.6, 10.0, 7.6 Hz, 1 H), 7.21 (m, 5 H) ppm. 13C NMR (75 MHz,
CDCl3): δ ϭ 38.0, 46.2, 60.8, 114.3, 126.3, 127.6, 128.5, 141.8,
143.7 ppm. C11H14O (162.2): calcd. C 81.44, H 8.70; found C 81.46,
H 8.75. These data are in accordance with published data.[23,24]
N-Butyl-3-methylpent-4-enamide (3c): Yield: 30.9 g (91% over 2
1
steps) as a colourless oil. H NMR (300 MHz, CDCl3): δ ϭ 0.92
(t, J ϭ 7.2 Hz), 1.05 (d, J ϭ 6.6 Hz), 1.34 (m, 2 H), 1.46 (m, 2 H),
2.09 (dd, J ϭ 13.8, 7.5 Hz, 1 H), 2.20 (dd, J ϭ 13.9, 7.5 Hz, 1 H),
2.69 (m, 1 H), 3.24 (dt, J ϭ 6.9, 5.7 Hz), 5.00 (m, 2 H), 5.61 (s, 1
H), 5.78 (m, 1 H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 13.6,
19.6, 20.0, 31.7, 34.7, 39.1, 43.9, 113.3, 142.9, 171.6 ppm.
C10H19NO (169.3): calcd. C 70.96, H 11.31, N 8.27; found C 70.67,
H 11.33, N 8.01.
General Procedure for the Synthesis of N-Butylpent-4-enylamines 4:
80 mmol of the amide, dissolved in 50 mL of THF, was added
dropwise to a suspension of lithium aluminium hydride (14.0 g,
0.37mol) in 300 mL of THF at 0 °C. The resulting mixture was
then heated under reflux overnight and the excess hydride destroyed
by careful addition of a solution of potassium hydroxide (7.0 g) in
33 mL of water. The resulting suspension was heated for another
hour under reflux and then filtered hot through a Büchner funnel.
The solid was taken up in 200 mL of THF, heated and filtered
through a Büchner funnel again. The combined liquid phases were
dried with sodium sulfate, the solvent was removed in vacuo and
the residue purified by distillation.
1-Tosyloxy-3-phenylpent-4-ene (7): Pyridine (8 mL, 0.1 mol) and 3-
phenyl-4-pentenol (9.7 g, 0.060 mol) were added at 0 °C to a solu-
tion of p-toluenesulfonyl chloride (19 g, 0.1 mol) in CH2Cl2
(150 mL). The reaction mixture was stirred overnight at room temp.
and then poured into water (150 mL). The organic layer was separ-
ated and the aqueous layer extracted twice with CH2Cl2 (50 mL
each). The combined organic layers were dried with Na2SO4 and
the solvent was removed in vacuo. The residue was purified by flash
chromatography to give 16.2 g (85%) of a colourless solid. 1H
NMR (300 MHz, CDCl3): δ ϭ 2.02 (m, 2 H), 2.11 (s, 3 H), 3.37
(q, J ϭ 7.8 Hz, 1 H), 3.98 (m, 2 H), 5.00 (m, 2 H), 5.85 (ddd, J ϭ
17.7, 10.5, 7.5 Hz, 1 H), 7.06 (m, 2 H), 7.24 (m, 5 H), 7.72 (m, 2
H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 21.5, 34.2, 45.4, 68.4,
115.1, 126.6, 127.5, 127.8, 128.6, 129.7, 133.2, 140.4, 142.4, 144.6
ppm. C18H20SO3 (316.4): calcd. C 68.33, H 6.37; found C 68.35, H
6.47. These data are in accordance with published data.[24]
N-Butyl-2-methylpent-4-enylamine (4b): Yield: 9.29 g (75%) of a
colourless liquid, b.p. 85 °C (50 mbar). 1H NMR (300 MHz,
CDCl3): δ ϭ 0.91 (d, J ϭ 6.6 Hz, 3 H), 0.92 (t, J ϭ 7.2 Hz, 3 H),
1.25 (br. s, 1 H), 1.28Ϫ1.55 (m, 4 H) 1.72 (pseudo-oct, J ϭ 6.6 Hz,
1 H), 1.92 (m, 1 H), 2.15 (m, 1 H), 2.41 (dd, J ϭ 11.7, 7.2 Hz, 1
H), 2.55 (dd, J ϭ 11.7, 6.0 Hz, 1 H), 2.60 (m, 2 H), 5.02 (m, 2 H),
5.79 (ddt, J ϭ 17.1, 10.2, 6.9 Hz, 1 H) ppm. 13C NMR (75 MHz,
CDCl3): δ ϭ 14.0, 18.0, 20.5, 32.3, 33.2, 39.5, 49.9, 56.1, 115.7,
137.2 ppm. C10H21N (155.3): calcd. C 77.35, H 13.63, N 9.02;
found C 76.99, H 13.71, N 8.77. Furthermore the hydrochloride of
the amine was prepared. C10H22ClN (191.7): calcd. C 62.64, H
11.57, N 7.30; found C 62.57, H 11.91, N 6.90.
1-Iodo-3-phenylpent-4-ene (8): Iodine (6.50 g, 25.5 mmol) was ad-
ded to a mixture of 3-phenylpent-4-en-1-ol (6) (3.24 g, 20 mmol),
imidazole (2.04 g, 30 mmol) and triphenylphosphane (6.72 g,
25.5 mmol) in 120 mL of toluene/acetonitrile (5:1) and the brown
mixture heated under reflux for 5 min. After cooling to room tem-
perature, 50 mL of TBDME was added and the solution extracted
three times with a 10% aqueous solution of sodium sulfite (50 mL
each). The organic layer was washed with water and brine (50 mL
each) and dried with sodium sulfate. After removing the solvent in
vacuo, 4.23 g (15.5 mmol, 78%) of the pure iodide 8 was obtained
N-Butyl-3-methylpent-4-enylamine (4c): Yield: 8.51 g (68%) of a col-
1
ourless liquid, b.p. 77 °C (50 mbar). H NMR (300 MHz, CDCl3):
δ ϭ 0.92 (t, J ϭ 7.2 Hz, 3 H), 1.01 (d, J ϭ 6.3 Hz, 3 H), 1.23 (s, 1
H), 1.35 (m, 2 H), 1.47 (m, 2 H), 1.51 (t, J ϭ 6.9 Hz, 2 H), 2.20
(sept, J ϭ 6.9 Hz, 1 H), 2.60 (m, 4 H) 4.95 (m, 2 H), 5.71 (ddd,
J ϭ 17.3, 10.4, 7.5 Hz, 1 H) ppm. 13C NMR (75 MHz, CDCl3):
δ ϭ 13.9, 20.3, 20.5, 32.3, 36.1, 36.9, 48.0, 49.8, 112.6, 144.4 ppm.
C10H21N (155.3): calcd. C 77.35, H 13.63, N 9.02; found C 76.94, H
13.96, N 8.65. The hydrochloride of the amine was also prepared.
C10H22ClN (191.7): calcd. C 62.64, H 11.57, N 7.30; found C 62.28,
H 11.92, N 6.91.
1
by flash chromatography. H NMR (300 MHz, CDCl3): δ ϭ 2.21
(m, 2 H), 3.07 (m, 2 H), 3.43 (q, J ϭ 7.2 Hz, 1 H), 5.10 (m, 2 H),
5.91 (ddd, J ϭ 17.4, 10.2, 7.8 Hz, 1 H), 7.25 (m, 5 H) ppm. 13C
NMR (75 MHz, CDCl3): δ ϭ 4.6, 38.8, 50.0, 115.1, 126.7, 127.6,
128.7, 140.4, 142.5 ppm. C11H13I (272.1): calcd. C 48.55, H 4.82;
found C 48.56, H 4.99. These data are in accordance with pub-
lished data.[23]
Ethyl 3-Phenylpent-4-enoate (5): A mixture of triethyl orthoacetate
(0.5 mol), cinnamic alcohol (0.5 mol) and 1 mL of acetic acid was
heated slowly to 150 °C. The reaction mixture was kept at this
temperature until 2 equiv. (1.0 mol) of ethanol had distilled off
(approx. 2 h). Some additional ethanol was then removed in vacuo
to leave the pure product as a yellow oil which was used without
N-Butyl-3-phenylpent-4-enylamine (4f): The iodide
14.7 mmol) was dissolved in 50 mL of dichloromethane and this
solution added to 50 mL of butylamine at 0 °C. After stirring the
8 (4.00 g,
further purification (98.9 g, 97%). 1H NMR (300 MHz, CDCl3): reaction mixture overnight, the solvent and excess butylamine were
δ ϭ 1.16 (t, J ϭ 7.2 Hz, 3 H), 2.68 (dd, J ϭ 15.0, 7.5 Hz, 1 H),
2.74 (dd, J ϭ 15.0, 7.8 Hz, 1 H), 3.86 (q, J ϭ 7.2 Hz, 1 H), 4.06
removed in vacuo and the residue was taken up in 100 mL of
TBDME. This solution was extracted three times with 100 mL of
(q, J ϭ 7.2 Hz, 2 H), 5.04 (m, 1 H), 5.09 (m, 1 H), 5.98 (ddd, J ϭ 5% aqueous sodium hydroxide solution, dried with sodium sulfate
17.4, 10.2, 7.2 Hz, 1 H),7.29 (m, 5 H) ppm. 13C NMR (75 MHz, and the solvent removed in vacuo. From the residue 3.12 g
CDCl3): δ ϭ 14.1, 40.3, 45.6, 60.3, 114.7, 126.7, 127.6, 128.5, 140.3, (14.4 mmol, 99%) of the amine 4f was obtained after flash chroma-
1852
Eur. J. Org. Chem. 2002, 1848Ϫ1854

Copper(I) Catalysts for the Stereoselective Addition of N-Chloroamines to Double Bonds
FULL PAPER
1
tography. H NMR (300 MHz, CDCl3): δ ϭ 0.89 (t, J ϭ 7.2 Hz, 3 N-Butyl-N-chloro-2-methylpent-4-enylamine (1b): Yield: 5.0 g (88%)
1
H), 1.01 (br. s, 1 H), 1.32 (m, 2 H), 1.44 (m, 2 H), 1.89 (m, 2 H),
of a colourless liquid. H NMR (300 MHz, CDCl3): δ ϭ 0.94 (d,
2.54 (m, 2 H), 2.61 (m, 2 H), 3.33 (q, J ϭ 7.5 Hz, 1 H), 5.03 (m, 2 J ϭ 6.3 Hz, 3 H), 0.95 (t, J ϭ 7.3 Hz, 3 H), 1.39 (pseudo-sext,
H), 5.96 (ddd, J ϭ 18.0, 10.5, 7.8 Hz, 1 H), 7.20 (m, 5 H) ppm.
7.2 Hz, 2 H), 1.66 (m, 2 H), 1.88Ϫ2.12 (m, 2 H), 2.26 (m, 1 H),
13C NMR (75 MHz, CDCl3): δ ϭ 13.9, 20.4, 32.3, 35.6, 47.8, 48.0, 2.66 (dd, J ϭ 12.9, 6.9 Hz, 1 H), 2.80 (dd, J ϭ 13.2, 6.3 Hz, 1 H),
49.6, 114.0, 126.2, 127.5, 128.4, 142.1, 144.1 ppm. C15H23N (217.4): 2.93 (m, 2 H), 5.04 (m, 2 H), 5.81 (m, 1H) ppm. 13C NMR (75
calcd. C 82.89, H 10.67, N 6.44; found C 82.59, H 10.80, N 6.24.
MHz, CDCl3): δ ϭ 13.9, 17.4, 20.0, 30.0, 31.6, 38.8, 64.5, 70.1,
116.1, 136.6 ppm.
1-Phenylpent-4-en-1-ol (9): 1-Bromo-3-butene (25.0 g, 135 mmol) in
50 mL of diethyl ether was added at 0 °C to a suspension of mag-
nesium turnings (3.28 g, 135 mmol) in 150 mL of diethyl ether over
1 h. The solution was stirred at room temperature for 2 h and then
cooled to 0 °C again before benzaldehyde (10.6 g, 100 mmol) was
added slowly. The reaction mixture was stirred overnight and then
poured onto 200 mL of ice/water. Sulfuric acid (5 ) was added
until the magnesium salts dissolved. After separation of the phases,
the aqueous phase was extracted twice with 50 mL of TBDME.
The combined organic phases were dried with sodium sulfate and
the solvent was removed in vacuo. The residue was subjected to
flash chromatography to yield 13.7 g (84 mmol, 84%) of pure 1-
phenylpent-4-en-1-ol (9). 1H NMR (300 MHz, CDCl3): δ ϭ 1.78
(m, 1 H), 1.85 (m, 1 H), 2.10 (m, 2 H), 2.16 (s, 1 H), 4.63 (dd, J ϭ
7.5, 5.7 Hz, 1 H), 4.98 (m, 2 H), 5.81 (ddt, J ϭ 17.1, 10.2, 6.9 Hz,
1 H), 7.30 (m, 5 H) ppm. 13C NMR (75 MHz, CDCl3): δ ϭ 29.9,
38.0, 73.9, 114.8, 125.9, 127.5, 128.4, 138.1, 144.6 ppm. These data
are in accordance with published data.[25]
N-Butyl-N-chloro-3-methylpent-4-enylamine (1c): Yield: 5.1 g (90%)
of a colourless liquid. H NMR (300 MHz, CDCl3): δ ϭ 0.93 (t,
1
J ϭ 7.2 Hz, 3 H), 1.03 (d, J ϭ 7.2 Hz, 3 H), 1.37 (sext, 7.2 Hz, 2
H), 1.58Ϫ1.78 (m, 4 H), 2.25 (m, 1 H), 2.91 (t, J ϭ 7.0 Hz, 4 H),
4.97 (m, 2 H), 5.96 (dd, J ϭ 10.4, 7.8 Hz, 1 H) ppm. 13C NMR (75
MHz, CDCl3): δ ϭ 13.9, 20.0, 20.4, 30.0, 34.5, 35.5, 62.3, 64.2,
113.1, 144.0 ppm.
N-Butyl-N-chloro-1-phenylpent-4-enylamine (1d): Yield: 6.86 g
(91%) of a colourless liquid. 1H NMR (400 MHz, CDCl3): δ ϭ
0.86 (t, J ϭ 6.8 Hz, 3 H), 1.29 (m, 2 H), 1.60 (m, 2 H), 1.92 (m, 1
H), 1.95 (m, 2 H), 2.29 (m, 1 H), 2.72 (m, 2 H), 3.84 (m, 1 H), 4.96
(m, 2 H), 5.78 (ddt, J ϭ 17.6, 10.4, 6.8 Hz, 1 H), 7.32 (m, 5 H)
ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 13.9, 19.9, 29.8, 30.4,
33.0, 60.2, 73.6, 114.9, 127.9, 128.1, 128.9, 138.1, 139.1 ppm.
N-Butyl-N-chloro-3-phenylpent-4-enylamine (1f): Yield: 6.41 g
(85%) of a colourless liquid. 1H NMR (300 MHz, CDCl3): δ ϭ
0.91 (t, J ϭ 7.5 Hz, 3 H), 1.34 (m, 2 H), 1.62 (m, 2 H), 2.06 (m, 2
H), 2.86 (m, 4 H) 3.41 (q, J ϭ 7.2 Hz, 1 H), 5.05 (m, 2 H), 5.95
(ddd, J ϭ 16.8, 10.5, 7.5 Hz, 1 H), 7.22 (m, 5 H) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 13.8, 19.9, 30.0, 33.3, 62.0, 63.3, 64.1,
114.4, 126.3, 127.6, 128.5, 141.7, 143.7 ppm.
N-Butyl-1-phenylpent-4-enylamine (4d): Iodine (6.5 g, 25.5 mmol)
was added to a mixture of 1-phenylpent-4-en-1-ol (9) (3.24 g,
20 mmol), imidazole (2.04 g, 30 mmol) and triphenylphosphane
(6.72 g, 25.5 mmol) in 120 mL of toluene/acetonitrile (5:1) and the
brown mixture heated under reflux for 5 min. After cooling to
room temperature, 50 mL of TBDME was added and the solution
extracted three times with a 10% aqueous solution of sodium sulfite
(50 mL each). The organic layer was washed with water and brine
(50 mL each) and dried with sodium sulfate. After removing the
solvent in vacuo, the crude residue was taken up in 50 mL of
dichloromethane. This solution was added to 50 mL of butylamine
at 0 °C. After stirring the reaction mixture overnight, the solvent
and excess butylamine were removed in vacuo and the residue was
taken up in 100 mL of TBDME. This solution was extracted three
times with 100 mL of 5% aqueous sodium hydroxide solution, dried
with sodium sulfate and the solvent removed in vacuo. From the
residue 1.14 g (5.24 mmol, 26% over two steps) of the amine 4d
were obtained after flash chromatography. 1H NMR (400 MHz,
CDCl3): δ ϭ 0.86 (t, J ϭ 7.2 Hz, 3 H), 1.28 (m, 2 H), 1.40 (br. s,
1 H), 1.43 (m, 2 H), 1.74 (m, 1 H), 1.84 (m, 1 H), 1.95 (m, 2 H),
2.42 (m, 2 H), 3.59 (dd, J ϭ 8.4, 5.6 Hz, 1 H), 4.95 (m, 2 H), 5.77
(ddt, J ϭ 17.2, 10.4, 6.8 Hz, 1 H), 7.29 (m, 5 H) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 13.9, 20.4, 30.5, 32.2, 37.1, 47.4, 63.0,
114.6, 127.0, 127.3, 128.3, 138.4, 144.4 ppm. C15H23N (217.4):
calcd. C 82.89, H 10.67, N 6.44; found C 82.57, H 10.84, N 6.53.
General Procedure for the Cyclisation Reaction Leading to 3-Chloro-
piperidines 10: The N-chloroamine (3 mmol) was added at 60 °C
(temperature of the oil bath) to a suspension of copper(I) chloride
(30 mg) in 5 mL of THF. The solution immediately turned blue-
green and it was kept at the same temperature for 12 h. After cool-
ing to room temp., silica (approx. 3 g) was added and the solvent
removed in vacuo. The product was isolated from the residue
bound to the silica by flash chromatography [eluent: pentane/
TBDME (10:1 then 3:1 and 1:1)].
N-Butyl-3-chloro-5-methylpiperidine (10b): 1H NMR (600 MHz,
CDCl3): δ ϭ 0.89 (d, J ϭ 7.2 Hz, 3 H), 0.90 (t, J ϭ 7.2 Hz, 3 H),
1.15 (q, J ϭ 12.0 Hz, 1 H), 1.29 (sext, J ϭ 7.8 Hz, 2 H), 1.45 (m,
2 H), 1.54 (t, J ϭ 10.8 Hz, 1 H), 1.76 (m, 1 H), 1.90 (t, J ϭ 10.8 Hz,
1 H), 2.20 [d (with fine-splitting), J ϭ 12.8 Hz, 1 H], 2.35 (td, J ϭ
7.8, 1.2 Hz, 2 H), 2.80 [d (with fine-splitting), J ϭ 10.8 Hz, 1 H),
3.18 [d (with fine-splitting), J ϭ 10.8 Hz, 1 H), 3.94 (tt, J ϭ 11.2,
4.4 Hz, 1 H) ppm. 13C NMR (150 MHz, CDCl3): δ ϭ 14.0, 19.0,
20.7, 29.0, 31.4, 44.0, 55.4, 57.9, 60.6, 61.3 ppm. Minor isomer
1
10bЈ: H NMR (600 MHz, CDCl3): δ ϭ 0.89 (t, J ϭ 7.8 Hz, 3 H),
0.91 (d, J ϭ 7.2 Hz, 3 H), 1.30 (m, 2 H), 1.47 (m, 3 H), 1.82 (m, 1
H), 1.91 (dt, J ϭ 14.4, 4.2 Hz, 1 H), 2.18 (m, 1 H), 2.39 (m, 2 H),
2.70 [d (broad), 9.6 Hz, 1 H), 2.84 [d (broad), 7.8 Hz, 1 H), 4.29
(quint, J ϭ 3.9 Hz, 1 H) ppm. 13C NMR (150 MHz, CDCl3): δ ϭ
13.9, 18.7, 20.6, 26.7, 28.8, 40.7, 56.1, 58.1, 59.9, 60.9 ppm.
C10H20ClN (189.7): calcd. C 63.61, H 10.63, N 7.38; found C 63.16,
H 11.15, N 7.26. The hydrochloride salt was also prepared.
C10H21Cl2N (189.7): calcd. C 53.10, H 9.36, N 6.19; found C 53.04,
H 9.20, N 6.02.
General Procedure for the Synthesis of N-Butyl-N-chloropent-4-enyl-
amines 1: NCS (33 mmol) was added at 0 °C to a solution of amine
(30 mmol) in 100 mL of dichloromethane and the resulting suspen-
sion stirred at this temperature for 2 h. The solvent was removed
in vacuo and the residue taken up in 100 mL of pentane and fil-
tered. The residue was washed with another 50 mL of pentane and
the liquid phases were combined. After removal of the solvent in
vacuo, the residue was purified by flash chromatography [eluent:
pentane/TBDME (10:1)]. The chloroamines were obtained as col-
ourless oils that could be stored at Ϫ30 °C. Due to their instability,
satisfactory elemental analyses of the N-chloroamines could not
be obtained.
N-Butyl-3-chloro-4-methylpiperidine (10c): 1H NMR (600 MHz,
CDCl3): δ ϭ 0.90 (t, J ϭ 7.2 Hz, 3 H), 1.09 (d, J ϭ 6.0 Hz, 3 H),
1.29 (sext, J ϭ 7.8 Hz, 2 H), 1.44 (m, 1 H), 1.48 (m, 2 H), 1.50 (m,
Eur. J. Org. Chem. 2002, 1848Ϫ1854
1853

G. Heuger, S. Kalsow, R. Göttlich
FULL PAPER
Ϫ London Ϫ Sydney Ϫ Toronto, 1973; J. K. Kochi, Acc. Chem.
Res. 1974, 7, 351.
This means following an SN2 mechanism with two times inver-
sion, of course the formation of the aziridinium ion is a first-
order reaction.
R. C. Fuson, C. L. Zirkle, J. Am. Chem. Soc. 1948, 70,
2760Ϫ2762. C. F. Hammer, S. R. Heller, J. H. Craig, Tetrahed-
ron 1972, 28, 239Ϫ253.
M. Newcomb, T. M Deeb, D. J. Marquardt, Tetrahedron 1990,
46, 2317Ϫ2328. J. H. Horner, F. N. Martinez, O. M Musa,. M.
Newcomb, H. E. Shahin, J. Am. Chem. Soc. 1995, 117,
11124Ϫ11133.
1 H), 1.75 [d (with fine-splitting), J ϭ 13.8 Hz, 1 H), 1.99 (td, J ϭ
12.0, 2.4 Hz, 1 H), 2.09 (t, J ϭ 11.2 Hz, 1 H), 2.37 (m, 2 H), 2.88
[d (with fine-splitting), J ϭ 11.4 Hz, 1 H), 3.21 (ddd, J ϭ 10.8, 4.2,
1.8 Hz, 1 H), 3.61 (td, J ϭ 10.8, 4.2 Hz, 1 H) ppm. 13C NMR (150
MHz, CDCl3): δ ϭ 13.9, 19.1, 20.6, 28.9, 33.7, 39.8, 53.3, 57.8,
[7]
[8]
[9]
1
61.5, 63.1 ppm. Minor isomer 10cЈ: H NMR (600 MHz, CDCl3)
(selected signal): δ ϭ 4.20 (m, 1 H). C10H20ClN (189.7): calcd. C
63.61, H 10.63, N 7.38; found C 63.34, H 11.02, N 7.07.The hydro-
chloride salt was also prepared. C10H21Cl2N (226.2): calcd. C
53.10, H 9.36, N 6.19; found C 52.91, H 9.44, N 5.97.
[10]
[11]
N-Butyl-3-chloro-5-phenylpiperidine (10d): 1H NMR (400 MHz,
CDCl3): δ ϭ 0.76 (t, J ϭ 7.2 Hz, 3 H), 1.13 (m, 2 H), 1.34 (m, 2
H), 1.66 (m, 2 H), 1.81 (m, 1 H), 1.96 (ddd, J ϭ 12.8, 8.4, 4.8 Hz,
1 H), 2.23 (t, J ϭ 10.8 Hz, 1 H), 2.28 (m, 1 H), 2.40 (ddd, J ϭ
16.0, 13.2, 6.8 Hz, 1 H), 3.05 (dd, J ϭ 7.2, 3.2 Hz, 1 H), 3.43 [d
(with fine splitting), J ϭ 10.8 Hz, 1 H), 4.06 (tt, J ϭ 10.8, 4.0 Hz,
1 H), 7.29 (m, 5 H) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 13.9,
20.3, 28.1, 36.0, 36.2, 54.3, 56.0, 60.7, 67.5, 127.1, 127.5, 128.4,
143.9 ppm. HRMS (Quattro-LC, Electrospray): m/z calcd. for
C15H22NCl ϩ Hϩ 252.1519/254.1494; found 252.1498/254.1473.
There are recent reports on the diastereoselectivity of aminyl
˚
radical cyclisations by Somfai: A. Sjöholm, M. Hemmerling,
N. Pradeille, P. Somfai, J. Chem. Soc., Perkin Trans. 1 2001,
˚
891. M. Hemmerling, A, Sjöholm, P. Somfai, Tetrahedron:
Asymmetry 1999, 10, 4091.
W. S. Johnson, L. Wertheman, W. R. Bartlett, T. J. Brockson,
T. Li, D. J. Faulkner, M. R. Petersen, J. Am. Chem. Soc. 1970,
92, 741.
[12]
[13]
The synthesis of the chloroamines 1a and 1e has been described
by us previously (ref.[3]).
An alternative approach, reduction of the initially formed es-
ters followed by tosylation of the alcohols and substitution
with butylamine can be monitored for small-scale reactions; on
a larger scale, this procedure, though involving more steps,
proved to be more efficient.
M. Newcomb, D. J. Marquardt, T. M. Deeb, Tetrahedron 1990,
46, 2329.
N-Butyl-3-chloro-4-phenylpiperidine (10f): 1H NMR (300 MHz,
CDCl3): δ ϭ 0.93 (t, J ϭ 7.2 Hz, 3 H), 1.35 (m, 2 H), 1.51 (m, 2
H), 1.88 (m, 2 H), 2.10 (m, 1 H), 2.24 (t, J ϭ 10.8 Hz, 1 H), 2.43
(m, 2 H), 2.61 (m, 1 H), 2.98 [d, (with fine-splitting), 11.1 Hz, 1
H), 3.38 (ddd, J ϭ 11.1, 4.5, 1.5 Hz, 1 H), 4.14 (td, J ϭ 11.1,
4.5 Hz, 1 H), 7.30 (m, 5H ppm. 13C NMR (75 MHz, CDCl3): δ ϭ
13.9, 20.6, 29.0, 34.4, 52.1, 53.4, 57.8, 60.4, 61.9, 126.9, 127.5,
128.4, 142.4 ppm.
[14]
[15]
A. L. J. Beckwith, C. J. Easton, A. K. Serelis, J. Chem. Soc.,
Chem. Commun. 1980, 482. A. L. J. Beckwith, T. Lawrence, A.
K. Serelis, J. Chem. Soc., Chem. Commun. 1980, 484. D. C.
Spellmeyer, K. N. Houk, J. Org. Chem. 1987, 52, 959.
We used the commercially available tetrakis(acetonitrile) com-
plex.
M. Newcomb, C. Ha, Tetrahedron Lett. 1991, 32, 6493Ϫ6496.
C. Ha, O. M. Musa, F. N. Martinez, M. Newcomb, J. Org.
Chem. 1997, 62, 2704Ϫ2710.
Changing the reaction conditions (solvent, temperature or
amount of catalyst) did result in lower yield in the system using
TMCDA as a ligand. Therefore, this yield is optimized.
This result was expected, because under protic conditions the
aminyl radical is protonated and ammonium radicals cyclise
with poor diastereoselectivity.
To date we have no good explanation for the varying yields
using different ligands. Studies towards a better understanding
of the effect of ligand size and bite angle are currently being
performed in our laboratory.
[16]
[17]
Acknowledgments
[18]
[19]
[20]
We thank the Fonds der Chemischen Industrie, the Deutsche For-
schungsgemeinschaft, the SFB 424 as well as the Alfried Krupp
von Bohlen und Halbach Stiftung for financial support.
[1]
P. Hamann, Nachr. Chem. Tech. Lab. 1990, 38, 342.
R. Göttlich, M. Noack, Tetrahedron Lett. 2001, 42, 7771; M.
[2]
Noack, R. Göttlich, Chem. Commun. 2002, 563; J. Helaja, R.
Göttlich, Chem. Commun., in press.
For aminyl radical reactions see: F. Minisci, Synthesis 1973, 1;
[3]
[21]
[22]
L. Stella, Angew. Chem. 1982, 95, 368; Angew. Chem. Int. Ed.
Engl. 1983, 22, 337; J. L. Esker, M. Newcomb, Adv. Heterocycl.
Chem. 1993, 58, 1; L. Stella in Radicals in Organic Synthesis,
Wiley-VCH, 2001, vol. 2, p. 407 and references cited therein.
For an overview see: D. P. Curran, N. A. Porter, B. Giese, Ste-
reochemistry of Radical Reactions, Wiley-VCH, Weinheim,
1996.
A. L. J. Beckwith, A. A. Zavitsas, J. Am. Chem. Soc. 1986, 108,
8230Ϫ8234; H. J. Günther; E. Guntrum, V. Jäger, Liebigs Ann.
Chem. 1984, 15.
[4]
R. Göttlich, Synthesis 2000, 1561.
[5]
The equilibrium between an uncoordinated aminyl radical and
[23]
[24]
[25]
the cyclised product prefers the ring-opened product and thus
does not lead efficiently to product formation, see: M. New-
comb, M. T. Burchill, T. M. Deeb, J. Am. Chem. Soc. 1988,
110, 6528.
P. H. Dussault, U. R. Zope, J. Org. Chem. 1995, 60, 8218.
J. Hartung, M. Hiller, P. Schmidt, Chem. Eur. J. 1996, 2, 1014.
V. H. Rawal, S. P. Singh, C. Dufour, C. Michoud, J. Org. Chem.
1993, 58, 7718.
[6]
Copper(II) salts are efficient scavengers for carbon radicals,
Received December 28, 2001
see: J. K. Kochi, Free Radicals, John Wiley & Sons, New York
[O01600]
1854
Eur. J. Org. Chem. 2002, 1848Ϫ1854