Pd(ii)-catalyzed carbonylation of N-unprotected arylethylamines

Article abstract of DOI:10.1039/c0cc03478a

An unprecedented NH₂-directed Pd(ii)-catalytic carbonylation of quaternary aromatic α-amino esters to yield 6-membered benzolactams has been developed. The reaction shows a strong bias to 6-membered lactams over 5-membered ones. The steric hind

Full text of DOI:10.1039/c0cc03478a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Preparation of benzolactams by Pd(II)-catalyzed carbonylation
of N-unprotected arylethylaminesw
a
a
a
b
a
a
´
Blanca Lopez, Aleix Rodriguez, David Santos, Joan Albert, Xavier Ariza, Jordi Garcia*
and Jaume Granell*b
Received 25th August 2010, Accepted 21st October 2010
DOI: 10.1039/c0cc03478a
An unprecedented NH2-directed Pd(II)-catalytic carbonylation
of quaternary aromatic a-amino esters to yield 6-membered
benzolactams has been developed. The reaction shows a strong
bias to 6-membered lactams over 5-membered ones. The steric
hindrance around the amino group seems to be pivotal for the
success of the process.
C–H activation/carbonylation of primary amines under a CO
environment has not been established.
Here we describe the first preparation of benzolactams
by palladium acetate-catalyzed aromatic carbonylation of
quaternary a-amino a-alkyl esters, by an unprecedented
process that uses NH2 as a directing group.
As part of an ongoing research project on bioorganometallic
chemistry,6 we attempted the cyclometallation of imines
The development of selective methods for the direct conver-
sion of carbon–hydrogen bonds into carbon–heteroatom and
carbon–carbon bonds remains a critical challenge in organic
chemistry. An interesting approach to address this issue
involves the use of aromatic substrates that contain coordinating
atoms (or directing groups).1 These ligands bind to the metal
center in the first step and a further rearrangement of some
atoms allows the C–H bond activation. The factors that
influence this well-known cyclometallation reaction are now
reasonably well understood.2
RCHQNC(Me)(CH2Ph)(COOMe) (R
=
4-ClC6H4 or
2,6-Cl2C6H3), derived from quaternary a-amino ester 1a, with
Pd(OAc)2 in toluene or AcOH. Unexpectedly, no imine
palladacycle was obtained, but 3a arising from the metallation
of the corresponding free amino ester was isolated after
subsequent reaction with LiBr (see Scheme 1). When we
applied the same conditions to free amino ester 1a, pallada-
cycle 3a was isolated again in 79% yield. We could extend the
process to amino esters 1b, 2a, and 2b. These results were
unexpected because it is not easy to metallate primary amines,7
and are in contrast with those obtained from the closely
related imines arising from the a-amino acids glycine, alanine,
valine and tyrosine methyl esters, which have been metallated
in good yields.6c
The transition-metal-catalyzed carbonylation of arenes
involving CO gas is also a significant chemical transformation
since it not only extends the carbon chain length but it also
introduces a synthetically versatile carbonyl group. Arenes
were first carbonylated by Fujiwara et al. in 1980 with
Pd(OAc)2 under 15 atm of CO using the arene substrates as
the solvent to obtain carboxylic acids.3 However, no control
over regioselectivity was observed in the case of substituted
arenes. This problem has been overcome by different research
groups through the directing group approach.1,4 Thus, Yu
et al. described very recently the palladium acetate-catalyzed
carbonylation of anilides to obtain N-acyl anthranilic acids.4d
In this regard, Orito et al., reported in 2004 the direct
carbonylation with CO of aromatic C–H bonds in N-alkyl-
o-arylalkylamines using a Pd(OAc)2/Cu(OAc)2/air system in
toluene solution at 120 1C, to obtain benzolactams.4b,5
However, the authors stated that ‘‘carbonylation of primary
amines, including benzylic amines or phenylethylamines,
under the same conditions, produced no benzolactams but
produced ureas in good yields’’.z Thus, a method for catalyzed
These results show that the presence of substituent R
in 1 and 2 plays a pivotal role in the metallation reaction,
probably due to the beneficial effects of steric bulk in avoiding
secondary reactions, but also by the Thorpe–Ingold effect2b,8
(or gem-dimethyl effect) which states that alkyl substituents
positioned on the acyclic carbon backbone improve the
outcome of organic cyclization reactions. A combination of
the different hydrolysis processesy that imines undergo and the
gem-dimethyl effect can explain our experimental findings.
The great tendency showed by these free amino esters
to undergo cyclopalladation prompted us to study their
palladium-catalyzed NH2-directed carbonylation at low
pressure. First, we confirmed that palladacycle 3a was easily
carbonylated to benzolactams 5a with CO (1 atm) in different
a
´
`
´
Departament de Quımica Organica and IBUB, Fac. Quımica,
´
`
Universitat de Barcelona, Martı i Franques 1, 08028-Barcelona,
Spain. E-mail: jordigarciagomez@ub.es; Fax: +34 93 3397878;
Tel: +34 93 4034819
b
´
`
´
Departament de Quımica Inorganica and IBUB, Fac. Quımica,
Universitat de Barcelona, Martı´ i Franque`s 1, 08028-Barcelona,
Spain. E-mail: jaume.granell@qi.ub.es; Fax: +34 93 4111492;
Tel: +34 93 4039139
w Electronic supplementary information (ESI) available: Experimental
details for the synthesis and characterization of lactams 5a–c, 5e, 5f,
10–14 and 6c and palladacycles 3a, 3b, 4a and 4b. See DOI: 10.1039/
c0cc03478a
Scheme 1 Stepwise transformation of quaternary a-amino esters 1
and 2 into benzolactams.
c
1054 Chem. Commun., 2011, 47, 1054–1056
This journal is The Royal Society of Chemistry 2011

View Article Online
Table 2 Carbonylation of phenylethylamines
Fig. 1 Urea 7a and acetamide 8a.
Lactam/
acetamide
ratio
solvents even at rt. Then we investigated the palladium
acetate-catalyzed carbonylation of racemic amino acid 1a,
using Cu(OAc)2/O2 as the oxidant in toluene. Unfortunately,
urea 7a (Fig. 1) was obtained in different experimental condi-
tions but not the expected benzolactam 5a. Since a strong
accelerating effect in some reactions catalyzed by Pd(OAc)2
in AcOH has been reported,9 we changed to this solvent.
Fortunately, we obtained the desired lactam 5a but it was
contaminated with the acetamide 8a, with the best result being
a ratio of 5a/8a = 64/36 in 91% yield.
Yield
(%)
Entry
R
R0
Lactam
1
2
3
4
5
6
7
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
Me
Propyl
Bn
p-MeO–Bn
H
Allyl
Me
Bn
5b
5c
5d/5d0
5e
—
98
93
80a
91
—
89
85
100 : 0
100 : 0
100 : 0
46 : 54
b
9
10
82 : 18
50 : 50c
CH2OH
a
b
Mixture of regioisomers (5d/5d0 = 6 : 4). Complex mixture of
c
compounds. Lactam 10 is not acetylated on the hydroxyl group.
Since the formation of acetamide may be favoured by the
Cu(II) salts which can enhance the amide bond formation,10z
we required an alternative oxidant. After some trials, we
obtained good results using benzoquinone as the oxidant
(Table 1).
presence of a neighboring coordinating hydroxymethyl or
allyl group erodes or inhibits completely the formation of
benzalactam (Table 2, entries 7 and 5, respectively). It should
be also noted that the presence of MeO, CN or F groups on
the aromatic ring (5d0 and 11–13 in Scheme 2) is compatible
with the formation of lactam, as shown in Scheme 2.
As shown, the best results were obtained with a 1.5 ꢀ 10ꢁ2
M
solution of 1a in refluxing AcOH using Pd(OAc)2
(5% molar) and an amino ester/benzoquinone molar ratio of
1 : 2. Under these conditions the yield was 91% and the
benzolactam/acetamide ratio was 90 : 10 (entry 3).
On the other hand, the reaction is highly sensitive to the size
of the benzolactam formed. Although a 57% yield of the
5-membered lactam 14 was obtained from triphenylmethyl-
amine with a total selectivity, no lactams were detected when
the reaction was performed with ligands 2a, 2b and 2e (Fig. 2).
These results are in sharp contrast with those reported by
Orito et al.5 in the related carbonylation of secondary amines
using Cu(II) as co-oxidant in which the benzolactams of
5 members are clearly favored over the 6-membered analogues.
We assumed that the greater reactivity of 6-membered pallada-
cycles could explain this result. To assess this assumption, we
In a next step, we successfully expanded the process to other
racemic phenylethylamines. The results are summarized in
Table 2.
A factor that plays a crucial role in the process is the steric
hindrance due to the R and R0 groups. Thus, the carbonyla-
tion of methyl phenylalaninate gave a low benzolactam/
acetamide ratio (Table 2, R0 = H, entry 4). This ratio
improved to 9 : 1 in compound 1a (Table 1, entry 3, R0 = Me)
and no acetamides were found in the preparation of 5b–d
bearing larger R groups (Table 2, entries 1–3, R0 = propyl,
benzyl and p-methoxybenzyl respectively), showing that an
increase in the steric hindrance around the amino group
prevents competitive acetylation. Interestingly, the presence
of the ester group (R a CO2Me) is not essential for the success
of the catalytic carbonylation (Table 2, entry 6). However, the
Table 1 Optimization of carbonylation of 1a
Benzoquinone
(% molar)
Yield
(%)
5a/8a/7a
ratio
Scheme 2 Benzolactams substituted on the aromatic ring.
Entry
t/h
T/1C
1a
2
6
6
6
6
6
3
Reflux
65
Reflux
Reflux
Reflux
Reflux
100
100
200
100
135
200
98
95
91
92
98
94
80 : 20 : —
58 : 42 : —
90 : 10 : —
70 : 30 : —
86 : 14 : —
84 : 16 : —
3
4b
5
6
a
b
2% molar of Pd(OAc)2. Two-fold 1a and benzoquinone
concentration.
Fig. 2 Benzolactam 14 and methyl phenylglycinates 2.
Chem. Commun., 2011, 47, 1054–1056 1055
c
This journal is The Royal Society of Chemistry 2011

View Article Online
Financial support from the Ministerio de Educacion y
´
Ciencia (CTQ2009-09692 and CTQ2009-11501) is gratefully
acknowledged.
Notes and references
z Pd(II) catalysts are readily reduced by CO, in a reaction that
also affords Ac2O, which could cause secondary reactions with
primary amines. See: I. I. Moiseev, Pure Appl. Chem., 1989, 61,
1755–1762.
y Mechanistic studies demonstrate that palladation of imines is a
somewhat complex process in which acidolysis of Pd–C and CQN
bonds and substitution on the acetato bridging positions are possible.
See: M. Gomez, J. Granell and M. Martinez, Eur. J. Inorg. Chem.,
2000, 217–224.
z In a blank experiment, the treatment of 1a with Cu(AcO)2 (20%
molar) in refluxing AcOH for 6 h in the absence of Pd(II) salts or
CO gave acetamide 8a in 96% yield. However, the aforementioned
CO-mediated reduction of Pd(AcO)2 to Pd(0) and Ac2O, which in turn
can acetylate 1a, should also be considered.
8 Preparation of 5f afforded also the acetamide of 15 (18%) as a
by-product.
** The carbonylation of 16 in the same conditions but in the presence
of benzoquinone (200% molar) gave 5f as the major product indicating
that benzoquinone favors the palladacycles equilibrium.
Scheme 3 Catalytic and stepwise carbonylation of 15.
compared the reactivity of 5- and 6-membered cyclopalladated
derivatives 3b and 4b (see Scheme 1) in front of CO at a moderate
temperature. Thus, after 1 hour of reaction at 50 1C, the
6-membered benzolactam 5b was obtained in 80% from 3b
whereas 4b afforded only 10% of 6b, in agreement with the
catalytic results.
1 T. W. Lyons and M. S. Sanford, Chem. Rev., 2010, 110, 1147–1169.
2 (a) J. Dupont, C. S. Consorti and J. Spencer, Chem. Rev., 2005,
105, 2527–2571; (b) X. Chen, K. M. Engle, D.-H. Wang and
J.-. Q. Yu, Angew. Chem., Int. Ed., 2009, 48, 5094–5115;
(c) M. Albrecht, Chem. Rev., 2010, 110, 576–623.
3 Y. Fujiwara, T. Kawauchi and H. Taniguchi, J. Chem. Soc., Chem.
Commun., 1980, 220–221.
4 (a) T. Asaumi, T. Matsuo, T. Fukuyama, Y. Ie, F. Kakiuchi and
N. Chatani, J. Org. Chem., 2004, 69, 4433–4440; (b) K. Orito,
A. Horibata, T. Nakamura, H. Ushito, H. Nagasaki, M. Yuguchi,
S. Yamashita and M. Tokuda, J. Am. Chem. Soc., 2004, 126,
A further experiment was specially significant. Catalytic
carbonylation of the amino ester 15, in which both 5-
and 6-membered metallacycles could be attained, gave only
the lactam 5f arising from the 6-membered metallacycle.8
However, the stoichiometric cyclometallation of 15 gave the
favored 5-membered palladacycle 16 as a major product.
Carbonylation of a sample of pure 16 in refluxing AcOH in
the absence of benzoquinone afforded the 5-membered lactam
6c in 86% yield as well as a minor amount (6%) of the isomer
5f (Scheme 3).** Thus both benzolactams sizes (6 or 5) are
attainable depending on the carbonylation method (catalytic
or stepwise).
14342–14343; (c) J. Vicente, I. Saura-Llamas, J. Garcıa-Lopez and
´ ´
D. Bautista, Organometallics, 2007, 26, 2768–2776; (d) R. Giri,
J. K. Lam and J.-Q. Yu, J. Am. Chem. Soc., 2010, 132, 686–693.
5 K. Orito, M. Miyazawa, T. Nakamura, A. Horibata, H. Ushito,
H. Nagasaki, M. Yuguchi, S. Yamashita, T. Yamazaki and
M. Tokuda, J. Org. Chem., 2006, 71, 5951–5958.
These experiments suggest that the 5-membered and the
6-membered palladacycles are probably in equilibrium under
catalytic conditions but the latter reacts more quickly with
CO, thus shifting the equilibrium to afford 5f as the only
stereoisomer.
6 (a) J. Albert, J. M. Cadena, A. Gonza
M. Font-Bardia, Chem. Commun., 2003, 528–529; (b) J. Albert,
J. M. Cadena, A. Gonzalez, J. Granell, X. Solans and
M. Font-Bardia, Chem.–Eur. J., 2006, 12, 887–894; (c) J. Albert,
M. Crespo, J. Granell, J. Rodrıguez, T. Calvet, S. Zafrilla,
´
lez, J. Granell, X. Solans and
´
´
In summary, we have demonstrated that an adequate
selection of the R groups positioned on the acyclic carbon
backbone of phenylethylamines and benzylamines allows an
unprecedented NH2-directed catalytic carbonylation with high
selectivity and yield. In this regard, an unexpected strong bias
to the 6-membered lactams over 5-membered ones has been
observed in our substrates. Studies designed to expand the
process to other organic derivatives of interest are currently
under way.
M. Font-Bardia and X. Solans, Organometallics, 2010, 29,
214–225.
7 J. Vicente and I. Saura-Llamas, Comments Inorg. Chem., 2007, 28,
39–72.
8 (a) R. M. Beesley, C. K. Ingold and J. F. Thorpe, J. Chem. Soc.,
1915, 107, 1080–1106. See, also: J. Kostal and W. L. Jorgensen,
J. Am. Chem. Soc., 2010, 132, 8766–8773.
9 F. Nahra, F. Liron, G. Prestat, C. Mealli, A. Messaoudi and
G. Poli, Chem.–Eur. J., 2009, 15, 11078–11082.
10 S. Yamada, S. Terashima and Wagatsuma, Tetrahedron Lett.,
1970, 11, 1501–1504.
c
1056 Chem. Commun., 2011, 47, 1054–1056
This journal is The Royal Society of Chemistry 2011