An atom efficient route to N-aryl and N-alkyl pyrrolines by transition metal catalysis

Article abstract of DOI:10.1039/c0ob00383b

The synthesis of N-aryl, N-tosyl, and N-alkyl pyrrolines from allyl alcohols and amines has been developed. The reaction sequence includes a palladium-catalyzed allylation step in which non-manipulated allyl alcohol is used to generate the diallylated amine in good to excellent yield. An excess of allyl alcohol was necessary for efficient diallylation of the amine, where the excess alcohol could be recycled three times. For aryl and tosyl amines, Pd[P(OPh)₃]₄ was used and for benzyl and alkyl amines a catalytic system comprising Pd(OAc)₂, PⁿBu₃, and BEt₃ was used. Both the electronic properties and the steric influence of the amine affected the efficiency of the allylation. The isolated diallylated amines were transformed into their corresponding pyrrolines by ring-closing metathesis catalyzed by (H₂IMes)(PCy₃)Cl ₂RuCHPh in good to excellent yield. A one-pot reaction was developed in which aniline was transformed into the corresponding pyrroline without isolating the diallylated intermediate. This one-pot reaction was successfully scaled-up to 1 mL of aniline in which the N-phenyl pyrroline was isolated in 95% yield. The versatility of the reaction in which 3-methyl-1-phenyl pyrroline was prepared in two-steps was demonstrated.

Full text of DOI:10.1039/c0ob00383b

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 2548
www.rsc.org/obc
PAPER
An atom efficient route to N-aryl and N-alkyl pyrrolines by transition metal
catalysis†
Supaporn Sawadjoon and Joseph S. M. Samec*
Received 5th July 2010, Accepted 13th January 2011
DOI: 10.1039/c0ob00383b
The synthesis of N-aryl, N-tosyl, and N-alkyl pyrrolines from allyl alcohols and amines has been
developed. The reaction sequence includes a palladium-catalyzed allylation step in which
non-manipulated allyl alcohol is used to generate the diallylated amine in good to excellent yield. An
excess of allyl alcohol was necessary for efficient diallylation of the amine, where the excess alcohol
could be recycled three times. For aryl and tosyl amines, Pd[P(OPh)3]4 was used and for benzyl and
alkyl amines a catalytic system comprising Pd(OAc)2, PnBu3, and BEt3 was used. Both the electronic
properties and the steric influence of the amine affected the efficiency of the allylation. The isolated
diallylated amines were transformed into their corresponding pyrrolines by ring-closing metathesis
catalyzed by (H2IMes)(PCy3)Cl2RuCHPh in good to excellent yield. A one-pot reaction was developed
in which aniline was transformed into the corresponding pyrroline without isolating the diallylated
intermediate. This one-pot reaction was successfully scaled-up to 1 mL of aniline in which the N-phenyl
pyrroline was isolated in 95% yield. The versatility of the reaction in which 3-methyl-1-phenyl pyrroline
was prepared in two-steps was demonstrated.
the aromatic counterpart. Another synthetic route is to react an
arylamine with 1,4-dihalobutane. However, this procedure also
generates stoichiometric amounts of waste. Recently, a palladium-
catalyzed diallylation of tosylamides, followed by ring-closing
Introduction
N-aryl pyrrolidines are core structures in a wide range of
biologically active compounds (Fig. 1). These compounds have
metathesis (RCM) was reported.7 Allyl carbonates, generated from
allyl alcohol, were used as the allylating agent. Even though
the overall yield is 81%, this procedure only gives 40.5% atom
economy. An advantage with generating the pyrroline is that
the double bond can be functionalized in a subsequent step. We
wanted to expand the scope to include arylamines, benzylamines,
and alkylamines. Furthermore, it is desirable to substitute the allyl
carbonate that generates stoichiometric amounts of waste with
the less expensive and renewable allyl alcohol.8 This would give a
procedure with only ethene and water as side-products.
shown promising biological activity toward a plethora of diseases
through: (i) inhibition of the lethal protease activity in anthrax
toxin,1 (ii) inhibition of the Syk kinase related to thrombosis
and non Hodgkin’s lymphoma,2 (iii) alteration of the antagonistic
activity of neuropeptide Y receptor related to food intake, learning
activity, Alzheimers and Parkinson’s diseases,3 (iv) cytotoxic agent
activity against human microvascular endothelial cells related to
diabetes and tumor growth.4
Whereas the palladium-catalyzed Tsuji–Trost monoallylation of
amines using allyl halides, acetates, carbonates, etc is a well studied
reaction,9 the diallylation of amines using allyl alcohol is not.10,11
In most cases, an acid is added to the reaction mixture in order
to promote the transformation.11a,d,f As the reaction is reversible, a
drying-agent has been used to drive the reaction to completion.12
We describe an unprecedented synthetic route to pyrrolines in
which allyl alcohols are used. A broad substrate scope is presented,
where different anilines, benzylamine, tosyl amide, and alkyl amine
have been used. The reaction proceeds by a palladium-catalyzed
diallylation of the amine, followed by a ruthenium-catalyzed RCM
to generate the corresponding pyrrolines with only ethene and
water as side-products.
Fig. 1 Core structure of biologically active compounds.1–4
The great potential in these compounds requires new divergent
atom efficient synthetic routes.5 Traditionally, these compounds
are synthesized by aromatic nucleophilic substitution of aryl
halides by a cyclic amine.6 Besides giving stoichiometric amounts
of waste this synthetic route limits the presence of halides on
Department of Biochemistry and Organic Chemistry, Box 576, 751 23
Uppsala University, Sweden. E-mail: joseph.samec@biorg.uu.se; Fax: +46
18-4713800; Tel: +46 18-471 3817
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c0ob00383b
2548 | Org. Biomol. Chem., 2011, 9, 2548–2554
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Results and discussion
Palladium-catalyzed allylation of amines
A. Ligand effect on Pd for the allylation of aniline (2a).
Initially, two different palladium-catalysts were studied in the
allylation of 2a. Allyl alcohol (1) (2 equiv.) was added to a
11e
toluene solution of 2a (1 equiv.) and Pd(PPh3)4 or Pd[P(OPh)3]4
(2 mol%), and the reaction was performed at 80 ◦C. The
Pd(PPh3)4 catalyst gave poor results (30% conv., 4 h) whereas the
Pd[P(OPh)3]4 catalyst gave above 95% conversion of 2a to afford
a 3 : 2 ratio of monoallylated (3a) : diallylated (4a) amine after
1 h (Scheme 1). It is hypothesized that the difficulty with using
allylic alcohols result from the oxidative addition step, i.e. the
carbon–oxygen bond cleavage, to generate the p-allyl palladium
intermediate. However, a more electron donating PPh3 ligand
would promote this step better than the more electron deficient
P(OPh)3 ligand. Clearly, the oxidative addition step is not the
limiting step in this reaction. The ease of oxidative addition is
also supported by Yamamoto’s report where 1 was oxidat◦ively
added to Pd(PCy3)2 to generate Pd(diallylether)(PCy3) at 30 C.13
Another interpretation of the difficulty with using allyl alcohols as
substrates is that the generated ether or diallylated amine chelate
to the metal that poisons the catalyst. Yamamoto reported a
Fig. 2 Dependence of 1 for the formation of 4a.
performed using 8 equivalents of 1 and the reaction was run
for 1 h after which the conversion was analyzed by H NMR
1
where >95% conversion to the desired 4a and no 3a was observed.
The reaction mixture was cooled down and the solvent mixture
(toluene, 1, and formed water) was vacuum transferred to another
Schlenk-tube containing Pd[P(OPh)3]4. Model substrate 2a and
two additional equivalents of 1 were added to the condensed
reaction mixture. This solution was degassed by 3-pump-thaw
cycles and the reaction was run for 1 h. We recycled the solvent
mixture three times without observing any deviation in reactivity
between the runs, demonstrating the ability to perform an atom
economic process (Fig. 3). Note, that every run yields 2 equivalents
of water. In the fifth run, 10 equivalents of water in the reaction
mixture did not influence the reactivity negatively demonstrating
the robustness of the reaction system.12
◦
temperature of 150 C for thermolysis of Pd(diallylether)(PCy3).
However, to generate such an intermediate requires that one
ligand dissociates. When using electron-deficient phosphites that
are more p-acidic, due to strong back-donation, the dissociation of
ligands may be prevented. An alternative mechanism proceeding
through a metal hydride has also been proposed.11b,c
Scheme 1 Allylation of 2a.
B. Effect of 1. With 2 equivalents of 1, full conversion of
2a to a mixture of 3a and 4a in a 3 : 2 ratio was observed after
60 min (Fig. 2). After 4 h, this ratio was increased to 3 : 1 in
favor of 4a. This is in conflict to the report by Ikariya, where
84% yield of diallylated product and no 3a was reported. When
4 equivalents of 1 was used, full conversion of 2a to a mixture
of 3a and 4a in a 1 : 4 ratio was observed after 60 min. When 8
equivalents of 1 was used, >95% conversion to 4a was observed
within 1 h, and no 3a was detected in the 1H NMR spectrum. One
explanation for the requirement of an excess amount of 1 is that
the reaction is reversible and the excess alcohol employed is used
to shift the equilibrium towards product.12 To test this hypothesis,
the reaction using 2 equivalents of 1 was performed in the presence
of molecular sieves. However, this led to no formation of 4a and
Fig. 3 Recycling of the solvent system.
D. Electronic and steric influence of the amine. Next, the
correlation between the electronic and steric properties of the
anilines and the rate for the formation of diallylated amine was
investigated. From the report of Ikariya and coworkers only one
example with 2a was presented.11e As the oxidative addition was
probably not the rate-limiting step, it was hypothesized that the
nucleophilic attack or the proton transfer by the amine was the
rate-determining step. Further insight into the nucleophilic attack
was gained by varying the electronic and steric properties of the
aniline by varying the substitution in the para- and ortho-positions.
Model substrate 2a was rapidly transformed into 4a in above
95% conversion within 30 min (Table 1, entry 1). The ratio of 4a : 3a
1
instead diallyl ether was observed as the only product in the H
NMR spectrum.11c
C. Recycling the excess of 1. The excess amount of 1 used
in the reaction is negative for the atom economy and therefore
recycling of the alcohol was investigated. Allylation of 2a was
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2548–2554 | 2549
©

View Article Online
Table 1 Effect of substituent in the allylation
Table 2 Palladium-catalyzed allylation of amines
entry substrate Substituent conv. 30 min (%) ratio (4 : 3) (di : mono)
entry substrate R
[Pd] (mol%) time (h) product yielda (%)
1
2
3
4
5
2a
2b
2c
2d
2e
H
p-F
p-MeO
p-Me
o-Me
95
85
49
95
93
1.9 : 1
0.6 : 1
0.16 : 1
0.9 : 1
0.3 : 1
1
2
3
4
5
6
7
8
2a
2b
2c
2d
2e
2f
Ph
p-F-Ph
p-MeO-Ph 5
p-Me-Ph
o-Me-Ph
Mes
(Ph)2CH
Ts
Bn
Cy
2
2
1
8
4
4
4a
4b
4c
4d
4e
4f
95
98
>95b
95
2
2
10
2
10
5
8
57
52
10
22
6
15
19
The reactions were run using 7 mmol of 1, 0.875 mmol of 2a, and 2 mol%
of Pd[P(OPh)3]4 in 1.25 mL of toluene at 80 ◦C. Reactions were monitored
by 1H NMR spectroscopy using mesitylene as internal standard.
2g
2h
2i
4g
4h
4i
88
40b
97c
92c
9
10
2j
5
4j
a The reactions were run using 7 mmol of 1, 0.875 mmol of 2a, and 2 mol%
of Pd[P(OPh)3]4 in 1.25 mL of toluene at 80 ◦C. Yields refer to isolated
yields. b Conversion by 1H NMR. c The reactions were run using 7 mmol
of 1, 0.875 mmol of 2a, and 5 mol% of Pd(OAc)2, 20 mol% PnBu3, and
22 mol% of BEt3 in 2 mL of THF at 66 ◦C.
was 1.9 : 1. The p-fluoro substituted aniline (2b) was transformed
more slowly (85% conversion within 30 min). The ratio of di : mono
(4b : 3b) allylation was 0.6 : 1. The p-Me substituted aniline (2d)
was transformed in above 95% conversion within 30 min in analogy
to aniline. However, the ratio of di : mono (4d : 3d) allylation was
0.9 : 1. The o-Me substituted aniline (2e) was also transformed
in above 93% conversion within 30 min. However, the ratio of
di : mono (4e : 3e) allylation was 0.3 : 1. The p-methoxy substituted
aniline (2c) was transformed in 49% conversion within 30 min
in a 0.16 : 1 ratio di : mono (4c:3c). Clearly, the nucleophile has
a profound effect in efficiency of the diallylation of anilines by
Pd[P(OPh)3]4, where both electron donating and withdrawing
substituents lower the conversion to the diallylated product. It
is noteworthy that the first allylation step may not be influenced
by the electronic properties of the aniline to the same extent as the
second allylation step. The electronic influence of the amine may
affect both the nucleophilicity of the amine and also the ability for
the amine to work as a proton mediator in the removal of water.
We are currently investigating the mechanism. The steric influence
of the substitution pattern of the aniline also has a profound effect
of the second allylation step, in which 2d proceeds to above 50%
conversion whereas 2e proceeds to 23% conversion within 30 min.
In this case, the nucleophilicity is most likely influenced by the
steric properties of the amine.
of catalyst (Table 2, entry 5). Clearly, the steric influence of the
ortho substituent has a dramatic effect of the efficiency of the
allylation. The diallylation of mesitylamine (2f) proceeds in 52%
yield after 10 h using 10 mol% of catalyst (Table 2, entry 6). The low
efficiency in terms of yield, reaction time, and catalyst loading may
be explained by both steric and electronic effects. The major side
product was the monoallylated mesitylamine (3f). The diallylation
of aminodiphenylmethane (2g) was performed using 2 mol% of
catalyst. The reaction was run for 22 h and the diallylated product
(4g) was isolated in 88% yield (Table 2, entry 7). The diallylation
of electron deficient p-toluenesulfonamide (2h) proceed in 40%
yield after 6 h using 10 mol% of Pd[P(OPh)3]4 (Table 2, entry
8). The major side product was the monoallylated amide. The
lower efficiency may be explained by the lower nucleophilicity
of the sulfonamide. The diallylation of benzylamine (2i) did not
proceed using Pd[P(OPh)3]4 as catalyst. One explanation is that
2i or product (4i) coordinates palladium and inhibits catalysis.
Instead the palladium phosphite catalyst was exchanged for a
catalytic system comprising Pd(OAc)2, PnBu3, and BEt3.14 With
this palladium source, both benzylamines and alkylamines were
efficiently transformed. Benzylamine was diallylated in 97% yield
using 5 mol% of palladium-catalyst in 15 h (Table 2, entry 9). The
allylation was performed on cyclohexylamine (2j) to generate the
diallylated product (4j) in 92% yield after 19 h (Table 2, entry 10).
E. Scope of the amines. After this initial study in which
the rate of the allylation of the anilines was studied the scope
of the amines was investigated with regard to the isolated
diallylated amine. Most diallylated amines were isolated by column
chromatography in good to excellent yields (Table 2). Model
substrate 2a was diallylated in 1 h, using 2 mol% of Pd[P(OPh)3]4
and 4a was isolated in 95% yield after column chromatography
(Table 2, entry 1). The p-F substituted 2b was diallylated in 8
h using 2 mol% of catalyst and 4b was isolated in 98% yield
(Table 2, entry 2). The longer reaction time compared to 2a is
in accordance to the lower reaction rate (Table 1, entries 1 and
2). In the case of the p-MeO aniline, 5 mol% of Pd[P(OPh)3]4 was
used and the diallylated product was generated in full conversion
after 4 h (Table 2, entry 3). Attempts to isolate the product by
column chromatography led to decomposition. Diallylation of p-
Me substitued 2d was performed using 2 mol% of catalyst for
4 h to generate 4d in 95% yield (Table 2, entry 4). The o-Me
substituted 2e was allylated in 57% yield after 8 h using 2 mol%
F. Ring-closing metathesis of diallylated aryl, tosyl and alkyl
amines. RCM was performed on the diallylated amines with
(H2IMes)(PCy3)Cl2RuCHPh (Table 3) used to prepare the
pyrrolines.15 The reactions were monitored by 1H NMR by
integrating known signals of the starting material and the product
using mesitylene as internal standard. Substrate 4a was trans-
formed into N-phenyl pyrroline (5a) in above 95% conversion
◦
within 1 h at 30 C using 2 mol% of catalyst (Table 3, entry 1).
The pyrroline 5a was isolated by column chromatography in 75%
yield. The diallylated p-F aniline 4b was transformed by RCM to
afford the corresponding pyrroline (5b) in above 95% conversion
over night using 2 mol% of (H2IMes)(PCy3)Cl2RuCHPh (Table
2550 | Org. Biomol. Chem., 2011, 9, 2548–2554
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
temperature was decreased to 25 ◦C and the observed rate at this
temperature was 4.2 ¥ 10-8 mol s-1 (eqn (3)).
Table 3 Ruthenium-catalyzed ring-closing metathesis to generate the
pyrrolines
(1)
(2)
entry substrate R
[Ru] (mol%) time (h) product conva (%)b
1
2
3
4
5
6
4a
4b
4c
4h
4i
Ph
2
2
5
2
10
10
1
5a
5b
5c
5h
5i
>95
>95
85
p-F-Ph
p-MeO-Ph
Ts
Bn
Cy
12
12
1
1.5
1
(3)
>95
70c
4j
5j
70c
a Reactions were run using 5 ¥ 10-5 mol of diallylated amine and 2–
10 mol% of (H2IMes)(PCy3)Cl2RuCHPh in DCM at 30 ◦C. b The yield was
determined by 1H NMR using mesitylene as internal standard. c Reactions
were run using microwave irradiation at 50 ◦C.
Finally, a one-pot procedure was designed. From the kinetic
data, it was clear that the palladium-catalyst and ruthenium-
catalyst were not compatible at the same reaction temperature.
Therefore, a catalytic system was designed where the allylation
of 2a was followed by RCM. The allylation was performed using
2 mol% of Pd[P(OPh)3]4 at 65 ◦C for 1 h. Then the temperature was
decreased to 30 ◦C and 5 mol% of (H2IMes)(PCy3)Cl2RuCHPh
was added to the reaction mixture via syringe. After 2 h the reaction
was finished with 95% overall yield of 2a to 5a was observed
using mesitylene as internal standard (Scheme 2). Encouraged by
the results we decided to scale-up this reaction to 1 mL of 2a.
The diallylation was performed using 2 mol% of Pd[P(OPh)3]4 at
80 ◦C for 4 h. The reaction mixture was cooled down to 30 ◦C
and 5 mol% of (H2IMes)(PCy3)Cl2RuCHPh was added to the
reaction mixture and it was stirred for 3 h. Pyrroline 5a was
isolated in 95% yield by column chromatography. The remaining
5% was the corresponding pyrrole and starting material. If the
generated water is excluded in the reaction, the atom economy of
this transformation is 80%.
3, entry 2). The scope of this approach, when compared to other
methodologies involving nucleophilic aromatic substitutions, is
nicely demonstrated by the addition of a fluorine atom on the
arylamine.6 In the case of the diallylated p-methoxy aniline 4c, a
crude mixture from the diallylation was used and RCM generated
a mixture of desired pyrroline 5c and the aromatized pyrrole (5c¢)
in 85% conversion after 12 h using 5 mol% of catalyst (Table
2, entry 3). The RCM of the diallylated p-toluenesulfonamide
(4h) generated the corresponding pyrroline (5h) in above 95%
conversion (Table 3, entry 4). The more basic benzylamine
derivative 4i required protonation by HCl prior to RCM.16 The
reaction was run for 90 min at 50 ◦C using microwave irradiation to
generate the pyrroline (5i) in 70% conversion.17 After protonation,
the diallylated cyclohexylamine was ring-closed to afford a 3 : 2
mixture of pyrrole and pyrroline (5j) in 70% conversion after 1 h
at room temperature. Prolonged reaction time favored formation
of pyrrole.
G. One-pot two-step procedure to pyrrolines from 1 and 2a.
Encouraged by the results of this two-step procedure, we explored
a one-pot procedure to make pyrrolines. There is a need to
design an efficient system for the tandem catalytic diallylation
of amines followed by ring-closing metathesis. As such, the rates
for the individual steps, i.e. palladium-catalyzed monoallylation,
diallylation and also ruthenium-catalyzed RCM were studied. The
allylation is reversible and therefore the initial kinetics (<30%
conversion) were measured.12 The reactions were run at 65 ◦C
Scheme 2 One-pot catalytic system.
H. Preparation of 3-methyl-1-phenyl-pyrroline (5¢a). Next
the methodology was expanded to compounds that are dif-
ficult to synthesize by conventional methods. For example,
the synthesis of 5¢a has only been reported using lithium
trimethylsilyldiazomethane and non-commercially available N,N-
disubstituted a-amino ketones.18 Model substrate 2a was monoal-
lylated by 2-methyl-2-propen-1-ol (6) (0.75 equiv.) using 2 mol%
1
in C6D6 using 2 mol% of catalyst and monitored by H NMR
using mesitylene as internal standard. The monoallylated aniline
3a gives a characteristic signal in the 1H NMR spectrum at d 3.50
ppm and the diallylated aniline 4a gives a characteristic signal at d
3.73 ppm. The monoallylation of 2a proceeded with an observed
rate of 1.3 ¥ 10-8 mol s-1 (eqn (1)). The diallylation of 3a proceeded
with an observed rate of 4.5 ¥ 10-8 mol s-1 (eqn (2)). The rate of
the RCM at this temperature was too fast to measure. Within the
time before we were able to lock and shim (2 min) the reaction
had proceeded to full conversion giving 5a with a characteristic
◦
of Pd[P(OPh)3]4 at 80 C in 74% isolated yield. The correspond-
ing N-(2-methyl-2-propenyl)aniline (3¢a) was allylated by 1 to
generate N-(2-methyl-2-propenyl)(allyl)aniline (4¢a) in 80% yield.
The remaining 10% was N-(2-methyl-2-propenyl)2aniline (4¢¢a).
These two compounds were not easily separated. Instead the
reaction mixture was transferred to a new reaction flask with
1
signal in the H NMR spectrum at d 3.80 ppm. Therefore, the
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2548–2554 | 2551
©

View Article Online
2 mol% (H2IMes)(PCy3)Cl2RuCHPh where 4¢a was ring-closed
at a higher rate to generate the trisubstituted pyrroline in favour
of generating the tetrasubstituted pyrroline from N-(2-methyl-2-
propenyl)2aniline (4¢¢a). After purification by column chromatog-
raphy the desired 3-Me-1-Ph pyrroline (5¢a) was isolated in 83%
yield.
were purchased from commercial suppliers and used without
further purification. Characterization data for 4a,11e 4b,21 4c,11a
4d,11a4e,224f,234g24 4h,25 4i,14 4j,14 5a,26 5b,27 5c¢,28 5h,25 5i,29 5j,30
5¢a18 was previously reported.
Preparation of N,N-diallylaniline (4a)
A flame-dried Schlenk tube was charged with Pd(dba)2 (10 mg,
0.0175 mmol), dissolved in 0.4 mL of DCM, and P(OPh)3 (18
mL, 0.07 mmol) was added via syringe. The slurry was degassed by
three freeze-pump-thaw cycles and stirred at room temperature for
30 min. The solvent was removed in vacuo. Toluene (1.25 mL) and
aniline (82 mL, 0.875 mmol) were added to the tube via syringe. The
solution was degassed by three freeze-pump-thaw cycles. Degassed
allyl alcohol (476 mL, 7 mmol) was added via syringe and the
◦
resulting mixture was stirred at 80 C for 4 h and monitored by
1H NMR and TLC. The mixture was concentrated and the crude
reaction mixture was purified by column chromatography on silica
gel using chloroform–pentane as eluent to afford a yellowish clear
oil as the pure product (0.144 g, 95% yield).
Preparation of N-allyl-N-benzylprop-2-en-1-amine (4i)
A flame-dried Schlenk tube was charged with Pd(OAc)2 (9.8 mg,
0.0438 mmol), THF (2 mL), and benzylamine (94 mL, 0.875
mmol). The slurry was degassed by three freeze-pump-thaw cycles.
Degassed allyl alcohol (476 mL, 7 mmol), BEt3 (263 mL, 0.2 mmol),
and PnBu3 (44 mL, 0.175 mmol) were added via syringe and stirred
at 66 ◦C for 15 h. The reaction was monitored by 1H NMR
and TLC. The mixture was concentrated in vacuo and the crude
reaction mixture was purified by column chromatography on silica
gel using chloroform–pentane as eluent to afford a clear oil as the
pure product (0.164 g, 97% yield).
Scheme 3 Generation of 5¢a.
Conclusions
The catalytic system where the palladium-catalyzed diallylation
of amines is followed by a ruthenium-catalyzed ring-closing
metathesis shows high atom economy. Only water and ethene
are generated as side-products. Furthermore, the procedure offers
synthetic advantages over stoichiometric methodologies. The
catalytic procedure expands the functional group tolerance on
the aryl group of the amine. For example, a fluorine atom may be
present on the aromatic ring. Compared to the allylation protocol
using allyl carbonates this novel methodology widened the scope
of amines to anilines, benzylamines and also alkyl amines. A one-
pot procedure where aniline was converted to N-Ph pyrroline in
95% overall conversion was demonstrated. 3-Me-1-Ph pyrroline
was conveniently synthesized using the protocol in 61% overall
yield.
Preparation of 1-phenyl pyrroline (5a)
A
flame-dried
Schlenk
tube
was
charged
with
(H2IMes)(PCy3)Cl2RuCHPh (0.8 mg, 1 mmol), 1 mL of a
DCM solution of diallylated aniline (0.05 M, 5 ¥ 10-5 mol), was
added and the reaction was performed at 30 ◦C. The yield was
determined by 1H NMR using mesitylene as internal standard.
Preparation of 1-phenyl pyrroline (5a) from aniline (2a)
Ruthenium catalyst (37 mg, 0.0438 mmol) was weighed into
a round bottom flask. The unpurified crude mixture of N,N-
diallylaniline prepared as described above, was added via syringe
and stirred at room temperature for 3 h. The reaction was
monitored by 1H NMR.
Experimental section
General remarks
Preparation of 1-phenyl pyrroline (5a) from 1 mL aniline (2a)
All reactions were conducted in oven-dried glassware under an
argon atmosphere using standard Schlenk techniques. Solvents
were purified by passage through alumina19 or distilled according
A flame-dried Schlenk tube was charged with Pd(dba)2 (120.8 mg,
0.21 mmol), dissolved in 4 mL of DCM, and P(OPh)3 (220 mL,
0.84 mmol) was added via syringe. The slurry was degassed by
three freeze-pump-thaw cycles and stirred at room temperature for
30 min. The solvent was removed in vacuo. Toluene (10 mL) and
aniline (957 mL, 10.5 mmol) were added to the tube via syringe. The
solution was degassed by three freeze-pump-thaw cycles. Degassed
allyl alcohol (5.713 mL, 84 mmol) was added via syringe and
the resulting mixture was stirred at 80 ◦C for 4 h. Ruthenium
1
to literature procedures. H NMR spectra were recorded on a
300 or 400 MHz spectrometer in CDCl3 at room temperature
and were referenced by the residual solvent signal. Allyl al-
cohol was distilled from magnesium and stored under argon.
Aniline was distilled under reduced pressure and stored under
argon. (H2IMes)(PCy3)Cl2RuCHPh20 and Pd[P(OPh)3]4 were
11c
prepared according to literature procedures. All other reagents
2552 | Org. Biomol. Chem., 2011, 9, 2548–2554
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
catalyst (446 mg, 0.525 mmol) was weighed into a round bottom
flask. The unpurified crude mixture of N,N-diallylaniline prepared
as described above, was added via syringe and stirred at room
temperature for 3 h. The mixture was concentrated and the crude
reaction mixture was purified by column chromatography on silica
gel using chloroform–pentane as eluent to afford the pure product
(1.447 g, 95% yield).
Recycling experiment
Four flame-dried Schlenk tubes were charged with Pd(dba)2
(10 mg, 0.0175 mmol), dissolved in 0.4 mL of DCM, and P(OPh)3
(18 mL, 0.07 mmol) was added via syringe. The slurry was degassed
by three freeze-pump-thaw cycles and stirred at room temperature
for 30 min. The solvent was removed in vacuo. Toluene (1.25 mL)
and aniline (82 mL, 0.875 mmol) were added to the first tube via
syringe. The solution was degassed by three freeze-pump-thaw
cycles. Degassed allyl alcohol (476 mL, 7 mmol) was added via
syringe and the resulting mixture was stirred at 80 ◦C for 1 h and
monitored by 1H NMR. The second Schlenk tube was cooled to
Preparation of 1-benzyl pyrroline (5i)
A microwave vial (BiotageTM) was charged with N-allyl-N-
benzylprop-2-en-1-amine (10 mg, 0.053 mmol) in 1 mL of a
dried DCM and 100 mL of 1 M HCl. The solution was stirred
at room temperature for 1 h. The mixture was evaporated
and azeotropically dried with toluene twice. Under an argon
atmosphere, ruthenium catalyst (4.5 mg, 0.0053 mmol) dissolved in
0.5 mL DCM was added to the protonated amine. The microwave
irradiation was carried out for 90 min (50 ◦C, 0–1 bar). The
crude mixture was treated with NaHCO3 and extracted with Et2O.
The residue was evaporated in vacuo. The reaction mixture was
analyzed by 1H NMR.
◦
-196 C. The remaining allyl alcohol and toluene from the first
tube were vacuum transferred to the second tube. After adding
aniline (82 mL, 0.875 mmol) and allyl alcohol (119 mL, 1.75 mmol)
via syringe, the reaction was run as above.
Kinetic studies
Aniline allylation. 0.6 mL of aniline solution (0.145 M in C6D6)
was added to an the NMR tube with Pd[P(OPh)3]4 (0.00174 mmol)
under an argon atmosphere. Allyl alcohol (59 mL, 0.87 mmol)
was added via syringe, the NMR tube was shaken, and inserted
into the spectrometer preheated to 65 ◦C. The kinetics of the
allylation were determined by 1H NMR spectroscopy. The reaction
was followed with 1H NMR spectroscopy by integrating the ratio
of N-allylaniline at d 3.50 ppm. and N,N-diallylaniline at d 3.73
ppm and the mesitylene signal at d 2.21 ppm.
Preparation of 3-methyl-1-phenyl pyrroline (5¢a)
A flame-dried Schlenk tube was charged with Pd(dba)2 (30 mg,
0.0525 mmol), dissolved in 0.5 mL of DCM, and P(OPh)3 (55 mL,
0.21 mmol) was added via syringe. The slurry was degassed by
three freeze-pump-thaw cycles and stirred at room temperature
for 30 min. The solvent was removed in vacuo. Toluene (1.5 mL)
and aniline (360 mL, 3.95 mmol) were added to the tube via
syringe. The solution was degassed by three freeze-pump-thaw
cycles. 2-methyl-2-propen-1-ol (222 mL, 2.625 mmol) was added
via syringe and the resulting mixture was stirred at 80 ◦C for 4 h and
N-allylaniline allylation. 0.6 mL of N-allylaniline (0.145 M
in C6D6) was added to an the NMR tube with Pd[P(OPh)3]4
(0.00174 mmol) under an argon atmosphere. Allyl alcohol (59
mL, 0.87 mmol) was added, the tube was shaken, and inserted
into the spectrometer preheated to 65 ◦C. The kinetics of the
transformation from N-allylaniline to N,N-diallylaniline was
determined by H NMR spectroscopy by integrating ratio of N-
allylaniline at d 3.50 ppm. and N,N-diallylaniline at d 3.73 ppm
and the mesitylene signal at d 2.21 ppm.
1
1
monitored by H NMR. The mixture was concentrated and the
crude reaction mixture was purified by column chromatography
on silica gel using chloroform–pentane as eluent to afford a
colorless oil as the pure product (0.285 g, 74% yield). A flame-
dried Schlenk tube was charged with Pd(dba)2 (7.8 mg, 0.0136
mmol), dissolved in 0.4 mL of DCM, and P(OPh)3 (14 mL, 0.0544
mmol) was added via syringe. The slurry was degassed by three
freeze-pump-thaw cycles and stirred at room temperature for 30
min. The solvent was removed in vacuo. Toluene (1 mL) and N-
(2-methyl-2-propenyl)aniline (100 mg, 0.68 mmol) were added to
the tube via syringe. The solution was degassed by three freeze-
pump-thaw cycles. Degassed allyl alcohol (278 mL, 4.08 mmol)
was added via syringe and the resulting mixture was stirred at
80 ◦C for 2 h and monitored by 1H NMR. The mixture was con-
centrated and the crude reaction mixture was purified by column
chromatography on silica gel using chloroform–pentane as eluent
to afford a mixture of N-(2-methyl-2-propenyl)2aniline and N-
(2-methyl-2-propenyl)(allyl)aniline. Ruthenium catalyst (18.5 mg,
0.0218 mmol) dissolved in 2 mL DCM was added to the mixture
and the reaction was run for 4 h at room temperature. The mixture
was concentrated and the crude reaction mixture was purified by
column chromatography on silica gel using chloroform–pentane
as eluent to afford the pure product (75.7 mg, 83% yield).
Ring-closing metathesis. 0.6 mL of N,N-diallylaniline (0.145
in C6D6) was added to an the NMR tube with
M
(H2IMes)(PCy3)Cl2RuCHPh (1.45 mg, 0.00174 mmol) under an
argon atmosphere. The tube was shaken and inserted into the
spectrometer at room temperature. The kinetics of the formation
of 1-phenyl-2,5-dihydro-1H-pyrrole from N,N-diallylaniline was
1
determined by using H NMR spectroscopy by integrating the
signal of N,N-diallylaniline at d 3.73 ppm, 1-phenyl-2,5-dihydro-
1H-pyrrole at d 3.80 ppm and the mesitylene signal at d 2.21 ppm.
Acknowledgements
Financial support from the Department of Biochemistry and
Organic Chemistry, Uppsala University and Magnus Bergwalls
Stiftelse is gratefully acknowledged. We thank Prof. L. Kuo for
fruitful discussions.
Notes and references
1 M. Pellecchia, US 2010/0016292 A1.
2 Z. J. Jia, C. Venkataramani, W. Huang, M. Mehrotra, Y. Song, Q. Xu,
S. M. Bauer, and A. Pandey, WO 2009/136995 A2.
3 T. Okuno, N. Kouyama, and M. Sakagami, WO2007/125952 A1.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 2548–2554 | 2553
©

View Article Online
4 J. L. Kane, B. H. Hirth, C. Yee, M. M. Staveski, H.-P. Biemann, and
M. J. Pregel, US 2007/0254894 A1.
5 (a) R. A. Sheldon, Pure Appl. Chem., 2000, 72, 1233; (b) B. M. Trost,
Angew. Chem., Int. Ed. Engl., 1995, 34, 259.
6 L. Zhang, B. Qiu, X. Li, X. Wang, J. Li, Y. Zhang, J. Liu, J. Li and J.
Shen, Molecules, 2006, 11, 988.
7 S. Cerero, J. Cortes, M. Moreno-Manas, R. Pleixats and A. Roglans,
Tetrahedron, 1998, 54, 14869.
8 (a) E. Arceo, P. Marsden, R. G. Bergman and J. A. Ellman, Chem.
Commun., 2009, 3357; (b) O. Kamm and C. S. Marvel, Org. Synth.,
1941 Coll Vol. 1, 42, 1921, Vol. 1, 15.
9 (a) J. Tsuji, Acc. Chem. Res., 1969, 2, 144; (b) B. M. Trost and D. L.
Van Vranken, Chem. Rev., 1996, 96, 395; (c) M. Johansen and K. A.
Jorgensen, Chem. Rev., 1998, 98, 1689.
15 (a) R. H. Grubbs, (Ed), Handbook of Metathesis, Wiley-VCH, Wein-
heim, 2003; (b) G. C. Vougioukalakis and R. H. Grubbs, Chem. Rev.,
2010, 110, 1746; (c) T. M. Trnka and R. H. Grubbs, Acc. Chem.
Res., 2001, 34, 18; (d) R. Stu¨rmer, B. Scha¨fer, V. Wolfart, H. Stahr,
U. Kazmaier and G. Helmchen, Synthesis, 2001, 46.
16 G. C. Fu, S. T. Nguyen and R. H. Grubbs, J. Am. Chem. Soc., 1993,
115, 9856.
17 (a) D. Balan and H. Adolfsson, Tetrahedron Lett., 2004, 45, 3089; (b) C.
Yang, W. V. Murray and L. J. Wilson, Tetrahedron Lett., 2003, 44,
1783.
18 H. Ogawa, T. Aoyama and T. Shiori, Heterocycles, 1996, 42,
75.
19 A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen and F. J.
Timmers, Organometallics, 1996, 15, 1518.
10 (a) Y. Tamaru, Eur. J. Org. Chem., 2005, 2647; (b) J. Muzart,
Tetrahedron, 2005, 61, 4179; (c) J. Muzart, Eur. J. Org. Chem., 2007,
3077.
20 A. P. Blum, T. Ritter and R. H. Grubbs, Organometallics, 2007, 26,
2122.
21 K. H. Kang, K. I. Choi;, H. Y. Koh, Y. Kim, B. Y. Chung and Y. S.
Cho, Synth. Commun., 2001, 31, 2277.
22 J. Jacob and W. D. Jones, J. Org. Chem., 2003, 68, 3563.
23 P. Barmettler and H.-J Hansen, Helv. Chim. Acta, 1990, 73, 1515.
24 O. Rahman, T. Kihlberg and B. La˚ngstro¨m, Org. Biomol. Chem., 2004,
2, 1612.
25 C. Che, W. Li, S. Lin, J. Chen, J. Zheng, J. C. Wu, Q. Zheng, G. Zhang,
Z. Yang and B. Jiang, Chem. Commun., 2009, 5990.
26 L. Adak, K. Chattopadhyay and B. C. Ranu, J. Org. Chem., 2009, 74,
3982.
27 L. A. Gharat, N. K. Joshi, J. M. Gajera, and P. S. Yadav,
WO/2008/010061 A2..
28 H. C. Ma and X. Z. Jiang, J. Org. Chem., 2007, 72, 8943.
29 A. L. Miller, II and N. B. Bowden, Chem. Commun., 2007, 2051.
30 Z. Ding and J. J. Tufariello, Synth. Commun., 1990, 20, 227.
11 (a) S.-C. Yang and C.-W. Hung, Synthesis, 1999, 1747; (b) F. Ozawa, H.
Okamoto, S. Kawagishi, S. Yamamoto, T. Minami and M. Yoshifuji,
J. Am. Chem. Soc., 2002, 124, 10968; (c) F. Ozawa, T. Ishiyama, S.
Yamamoto, S. Kawagishi and H. Murakami, Organometallics, 2004,
23, 1698; (d) H. Kinoshita, H. Shinokubo and K. Oshima, Org. Lett.,
2004, 6, 4085; (e) Y. Kayaki, T. Koda and T. Ikariya, J. Org. Chem.,
2004, 69, 2595; (f) S.-C. Yang, Y.-C. Hsu and K.-H. Gan, Tetrahedron,
2006, 62, 3949.
12 S. Lemarie-Audoire, M. Savignac, J. P. Genet and J.-M. Bernard,
Tetrahedron Lett., 1995, 36, 1267.
13 T. Yamamoto, M. Akimoto, O. Saito and A. Yamamoto,
Organometallics, 1986, 5, 1559.
14 M. Kimura, M. Futamata, K. Shibata and Y. Tamaru, Chem. Commun.,
2003, 234.
2554 | Org. Biomol. Chem., 2011, 9, 2548–2554
This journal is
The Royal Society of Chemistry 2011
©