N-Acyl imine and enamide intermediates in the palladium-catalyzed amidocarbonylation reaction

Article abstract of DOI:10.1016/S0040-4039(01)00466-X

N-Acyl imines and enamides have been synthesized and subjected to the reaction conditions for palladium-catalyzed amidocarbonylation of aldehydes. These compounds were competent substrates resulting in the formation of N-acyl amino acids; however, the presence of water was found to be necessary. Direct study of the same amidocarbonylation reaction revealed that enamides could be detected during the course of the reaction. A slight enhancement in the yield of the amidocarbonylation is observed in the presence of radical inhibitors ruling out a meaningful radical pathway. The results are most consistent with a mechanism involving complex equilibration of the starting materials to a number of intermediates which can converge to a haloamidal that subsequently undergoes a palladium insertion.

Full text of DOI:10.1016/S0040-4039(01)00466-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 3403–3406
N-Acyl imine and enamide intermediates in the
palladium-catalyzed amidocarbonylation reaction
Dana A. Freed and Marisa C. Kozlowski*
Department of Chemistry, Roy and Diana Vagelos Laboratories, University of Pennsylvania, Philadelphia, PA 19104, USA
Received 28 February 2001; revised 19 March 2001; accepted 20 March 2001
Abstract—N-Acyl imines and enamides have been synthesized and subjected to the reaction conditions for palladium-catalyzed
amidocarbonylation of aldehydes. These compounds were competent substrates resulting in the formation of N-acyl amino acids;
however, the presence of water was found to be necessary. Direct study of the same amidocarbonylation reaction revealed that
enamides could be detected during the course of the reaction. A slight enhancement in the yield of the amidocarbonylation is
observed in the presence of radical inhibitors ruling out a meaningful radical pathway. The results are most consistent with a
mechanism involving complex equilibration of the starting materials to a number of intermediates which can converge to a
haloamidal that subsequently undergoes a palladium insertion. © 2001 Elsevier Science Ltd. All rights reserved.
The synthesis of amino acids is important in many fields
of chemistry. Currently, there are few general routes to
a-amino acids. Strecker cyanation, homologation of
glycine derivatives, amination of enolates, and the hydro-
genation of dehydroamino acids are among the more
versatile of the known methods.1 While the Strecker
synthesis is certainly the simplest, the potential of intro-
ducing carbonyl equivalents other than cyanide, such as
CO and CO2, to form amino acids remains attractive.
haloamidal intermediate (2) to palladium is followed by
CO insertion and hydrolysis to yield the N-acyl amino
acid 5.
Other pathways can also be considered for this reaction
such as elimination of the halide from the haloamidal to
form iminium ion 6, to which a palladium species might
add (Scheme 2). Isomerization of this iminium ion would
lead to enamide (7), to which a HꢀPd might add in a
regioselective manner.5 The role of enamide intermedi-
ates in the palladium-catalyzed reaction also has implica-
tions for the use of a-substituted aldehydes. In order to
probe these possibilities, N-acyl imines and enamides
have been synthesized and subjected to the conditions
of the palladium-catalyzed amidocarbonylation reac-
tion.
The use of carbon monoxide to achieve such a goal
utilizing a cobalt catalyst was first described by Waka-
matsu in 1970.2,3 Recently, Beller et al. showed that the
amidocarbonylation reaction proceeds with palladium
catalysts against a wider variety of substrates (Eq. (1)).4
This paper outlines mechanistic studies of several aspects
of the palladium-catalyzed amidocarbonylation reaction.
O
O
O
PdBr2, PPh3, LiBr, CO
H2SO4, NMP, 120 °C
+
HN
R2
Beller et al. have proposed the following mechanism
(Scheme 1) for the palladium-catalyzed reaction mecha-
nism.4 In this mechanism, oxidative addition of the
(1)
R1
H
H2N
R2
R1
CO2H
HBr
-H2O
R2
O
R2
O
R2
O
R2
O
R2
O
BrPd
O
O
H2O
-Pd0
HO
+
Pd0 Br
CO
HO
R1
NH
H
Br
NH
H
NH
H
Pd NH
NH
R1
H
H2N
R2
O
O
R1
R1
R1
H
R1
H
1
2
3
4
5
Scheme 1.
* Corresponding author.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00466-X

3404
D. A. Freed, M. C. Kozlowski / Tetrahedron Letters 42 (2001) 3403–3406
O
O
R2
O
R2
O
O
R2
H
+
-HBr
R2
HN
R2
-Br
-
N
PdX
HPdX
Br
NH
H
XPd
R1
NH
H
XPd
R1
NH
H
if R1 = CH2R3
R1
H
H
R1
R3
3
6
2
7
3
Scheme 2.
N-Acyl imine 8 was selected for study since it is rela-
tively stable, unlike most N-acyl imines, and can be
conveniently generated (Scheme 3).6 When this imine
was subjected to the amidocarbonylation conditions in
the presence of water, the amidocarbonylation product
9 was produced in ꢀ10% conversion as determined by
gas chromatography. The corresponding amidocar-
bonylation reaction from p-methoxybenzaldehyde and
benzamide proceeded with a similar conversion (18%)
under the same conditions.7 Thus, imine 8 appears to
be a substrate for the amidocarbonylation.
were also found to produce the enamide 11 in 30%
conversion when subjected to the amidocarbonylation
conditions without the palladium catalyst.
A benefit of using enamides as substrates is that the
role of water in the amidocarbonylation mechanism can
be delineated.10,11 The N-acyl amino acid 5 observed in
these reactions could arise from the reductive elimina-
tion of palladium to the acyl bromide followed by
hydrolysis (path a, Scheme 4), from the direct hydroly-
sis of the acylpalladium species (path b), or from
hydrolysis of an oxazolone intermediate arising from
intramolecular displacement of the palladium (path c).
In all cases, hydrolysis would be accomplished by the
water released from the aldehyde upon formation of the
bromoamidal 2. With enamides, this water is not
present and it may be possible to trap the intermediate
leading to the final product with other nucleophiles. To
this end, the enamide reaction was performed in the
presence of MeOH (entry 3, Table 1) which resulted in
a very high conversion (78%). Surprisingly, only 21%
was the methyl ester and the remaining 57% was the
acid. In an identical reaction performed under rigor-
ously anhydrous conditions (entry 4), the conversion
dropped substantially to 9% of the acid and 1% of the
methyl ester. Apparently the trace amount of water
introduced during assembly of the high pressure
apparatus is necessary to allow formation of the N-acyl
amino acid. The successful simulation of the aldehyde
reaction conditions (entry 1) by the addition of one
equivalent of water to the enamide reaction (entry 6),
provided further evidence that water is not detrimental.
Next, enamide substrates were examined and the read-
ily synthesized trans-enamide 118 was subjected to a
series of carbonylation conditions (Table 1). Entry 1
shows the results from the corresponding amidocar-
bonylation reaction between phenylacetaldehyde and
acetamide and entry 2 shows the results for the enamide
under the identical reaction conditions.9 The amidocar-
bonylation product 9 is generated in similar yield from
both substrates.
The above result does not distinguish between forma-
tion of bromoamidal 2 (Scheme 1) from the enamide by
addition of bromide present in the reaction mixture or
addition of a PdꢀH species to the enamide; however, it
does indicate that enamides are competent substrates
for the amidocarbonylation reaction. As such, a-substi-
tuted aldehyde substrates would be epimerized in the
amidocarbonylation via enamide formation. Since this
issue is only relevant if enamides form under the reac-
tion conditions, phenylacetaldehyde and acetamide
were subjected to the amidocarbonylation conditions
and the reaction was halted prior to completion (Eq.
(2)). Analysis of the crude reaction mixture by 1H
NMR spectroscopy against an internal standard indi-
cated the presence of phenylacetaldehyde (5%), N-acyl
amino acid 10 (33%), and enamide 11 (39%). In a
further experiment, phenylacetaldehyde and acetamide
O
The lack of product formation under anhydrous
conditions10 in the presence of methanol suggests that
there is quite a difference in the hydrolysis versus
methanolysis rate of the true intermediate.12 This points
away from paths a and c since methanolysis of the
respective acyl bromide (12) and oxazolone (13) inter-
O
O
O
O
O
a
N
Ph
b
b
HO2C
HN
Ph
+
C
+
H
N
Ph
H
H2N
Ph
H
O
H
36%
MeO
18%
MeO
MeO
MeO
8
9
Scheme 3. (a) Ether, reflux; (b) 1 mol% (PPh3)2PdBr2, 35 mol% LiBr, 1 mol% H2SO4, CO, NMP, 120°C.
O
O
1 mol% (Ph3P)2PdBr2
O
O
35 mol% LiBr, 60 bar CO
+
HN
CH3
HN
CH3
+
Ph
1 mol%H2SO4
NMP, 120 °C
H
H2N
CH3
Ph
(2)
CO2H
11
39%
10
Ph
33%

D. A. Freed, M. C. Kozlowski / Tetrahedron Letters 42 (2001) 3403–3406
Table 1. Amidocarbonylation of enamide substrates (Eq. (3))a
3405
O
N
H
CH3
1 mol% (PPh3)2PdBr2, LiBr,
11
O
(3)
OR
HO
CO (60 bar), H+, NMP, 120 °C
N
H
CH3
O
O
+
O
H2N
CH3
10
H
Entry
Substrate
LiBr (mol%)
H+ (mol%)
Additive (mol%)
Conv. (%)b
1
2
3
4
5
6
7
8
9
Aldehyde/amide
Enamide (11)
Enamide (11)
Enamide (11)
Enamide (11)
Enamide (11)
Enamide (11)
Enamide (11)
Enamide (11)
Enamide (11)
35
35
35
35
35
35
0
100
35
35
1 H2SO4
1 H2SO4
1 H2SO4
1 H2SO4
1 H2SO4
1 H2SO4
1 H2SO4
1 H2SO4
100 H2SO4
10 BF3·OEt2
–
–
33
36
100 MeOH
78 (21 ester)
10 (B1 ester)
100 MeOH, anhyd.c
1000 MeOH
100 H2O
–
–
–
–
0
26
8
32
27
39
10
a Performed using anhydrous solvents and reagents in a dried glass liner.
b Conversions are the average from two trials and were measured by 1H NMR (see Ref. 9).
c Performed under rigorously anhydrous conditions, in which the autoclave was heated overnight at 80°C, cooled under vacuum, and assembled
inside an inert atmosphere box.
O
R2
Br
closely parallel those observed in the reaction of the
aldehyde and the amide.14 For example, removal of
LiBr caused a drop in conversion since there is less
halide for formation of the bromoamidal intermediate.
On the other hand, the addition of a full equivalent of
LiBr does not enhance the conversion, which also
points away from an acyl bromide product. The use of
a full equivalent of protic acid results in a mild decrease
in conversion while the use of BF3·OEt2 instead of
H2SO4 is not detrimental. These similar reaction pat-
terns suggest that the enamide and the aldehyde/amide
react via a similar mechanism.
Pd
NH
H
O
-Pd0
-
-[PdIIBr]
O
R1
4
R2
a
c
O
R2
-
O
-[PdIIBr]
Br
H2O
b
N
H
NH
H
O
R2
H2O
-
O
R1
R1
HO
H2O
12
-Br
13
NH
H
O
R1
5
Scheme 4.
In a further experiment, radical inhibitors (hydro-
quinone and BHT) were added to the palladium-cata-
lyzed amidocarbonylation reaction to discern if the
bromoamidal intermediate 2 underwent radical cleav-
age over the course of the reaction. With both
inhibitors, somewhat higher yields (64–65% versus 55%)
of 14 were observed indicating that the formation of
radicals leads to deleterious side reactions (Eq. (4)).
mediates should be facile. On the other hand, the direct
cleavage of the acylpalladium 4 by a palladium-associ-
ated water (path b) would account for this reactivity
pattern.13 Alternatively, water may be necessary in
order for the enamide to react; perhaps a water
molecule remains in close proximity to the palladium
throughout the cycle.11 Regardless, the enhancement in
yield upon addition of MeOH (entry 3) is difficult to
explain. The use of further MeOH (10-fold excess, entry
5) results in no conversion which may be due to a
solvent effect. Consistent with this, when methanol was
used instead of NMP as solvent in the aldehyde ami-
docarbonylation no reaction was observed.
In conclusion, N-acyl imines and enamides have been
shown to lead to the same amidocarbonylation prod-
ucts as produced from aldehydes and amides. Addition-
ally, an enamide byproduct was observed in the
amidocarbonylation of phenylacetaldehyde providing
evidence that enamide intermediates are relevant to
these reactions. Further experiments with enamides
The results from varying the amounts of LiBr and acid
(entries 7–10 of Table 1) with the enamide substrate
have shown that water is critical for the reaction consis-
O
1 mol% (Ph3P)2PdBr2
35 mol% LiBr, 60 bar CO
O
O
no inhibitor
10% hydroquinone 65%
10% BHT 64%
55%
+
HN
CH3
CO2H
(4)
Ph
H
H2N
CH3
1 mol%H2SO4
Ph
NMP, 120 °C
14

3406
D. A. Freed, M. C. Kozlowski / Tetrahedron Letters 42 (2001) 3403–3406
tent with direct hydrolysis of an acylpalladium species
to form the carboxylic acid product. Overall the experi-
ments described above indicate that the amidocarbonyl-
ation largely proceeds as outlined in Scheme 1.
8. (a) Dakin, H. D.; West, R. J. Biol. Chem. 1928, 745; (b)
Dakin, H. D.; West, R. J. Biol. Chem. 1928, 757; (c)
Redeker, U.; Engel, N.; Steglich, W. Tetrahedron Lett.
1981, 22, 4263.
9. This reaction was not run to completion in order to
determine if enamide was formed. With further conver-
sion, isolated yields of 85% have been reported for 10
(Ref. 4a). The same reaction conditions were employed
for all entries in Table 1: the substrates (2.5 mmol each of
aldehyde and acetamide or 2.5 mmol enamide) were
placed in an oven-dried glass liner under N2 and were
combined with distilled NMP (5 mL), PdBr2 (1 mol%),
Ph3P (2 mol%), anhydrous LiBr (35 mol%), and H2SO4 (1
mol%). The reaction vessel was placed in a high pressure
reactor (Parr No. N4767 or 4712) which was purged with
CO, pressurized to 60 bar with CO, and magnetically
stirred for 14 h at 120°C (bath temperature). After cool-
ing to rt, the reactor was depressurized and iPrOH (5.0
mmol) was added as an internal standard. The amounts
of aldehyde, enamide, N-acyl amino acid, and N-acyl
amino ester were determined by integration of the crude
Acknowledgements
Financial support was provided by the University of
Pennsylvania, the National Science Foundation (CHE-
9730576), Merck Research Laboratories, and DuPont.
References
1. For an overview, see: (a) Williams, R. M. Synthesis of
Optically Active h-Amino Acids; Pergamon: Oxford,
1989. For a homologation review, see: (b) Abellan, T.;
Chinchilla, R.; Galindo, N.; Guillena, G.; Najera, C.;
Sansano, J. M. Eur. J. Org. Chem. 2000, 2689. For a
cyanation review, see: (c) Dyker, G. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1700.
1
reaction mixture H NMR signals (CDCl3).
10. Water has been shown to be necessary in the cobalt-cata-
lyzed amidocarbonylation. When bisamidal substrates are
used instead of the aldehyde, no product formation is
observed. See: Parnaud, J.-J.; Campari, G.; Pino, P. J.
Mol. Catal. 1979, 6, 341.
11. Water has been hypothesized to be associated with the
cobalt throughout the course of the reaction and to
prevent hydrogenation of the metal carbonyl intermedi-
ate. See: Ojima, I.; Zhang, Z. Organometallics 1990, 9,
3122.
12. Methanol also gives lower yields when N-bromomethyl
phthalimide is used as a substrate. See Ref. 4b.
13. Lin, Y.-S.; Yamamoto, A. Organometallics 1998, 17,
3466.
14. In a separate series of experiments (data not shown), the
same reactivity trends were observed when the corre-
sponding aldehyde and amide were employed as starting
materials using the reaction conditions described in the
text.
2. Wakamatsu, H.; Uda, J.; Yamakami, N. J. Chem. Soc.,
Chem. Commun. 1971, 1540.
3. (a) Ojima, I. Chem Rev. 1988, 88, 1011; (b) Knifton, J. F.
In Applied Homogeneous Catalysis with Organometallic
Compounds; Cornils, B.; Herrmann, W. A., Eds.; VCH:
Weinheim, 1996; (c) Ku¨hlein, K.; Geissler, H. In Transi-
tion Metals for Organic Synthesis; Beller, M.; Bolm, C.,
Eds.; Wiley-VCH: Weinheim, 1998; p. 79.
4. (a) Beller, M.; Eckert, M.; Vollmu¨ller, F.; Bogdanovic, S.;
Geissler, H. Angew Chem., Int. Ed. Engl. 1997, 36, 1494;
(b) Beller, M.; Eckert, M. Angew Chem., Int. Ed. 2000,
39, 1011.
5. The intermediacy of an enamide may explain the lack of
reactivity of aryl aldehydes in the cobalt-catalyzed reac-
tion. See: Magnus, P.; Slater, M. Tetrahedron Lett. 1987,
28, 2829.
6. Arbuzov, B. A.; Zobova, N. N.; Sofronaova, O. V. Ser.
Khim. 1979, 2, 443.
7. Unoptimized yield. See Ref. 9 for procedures.
.
.