Palladium-catalyzed aminocarbonylation of aryl chlorides at atmospheric pressure: The dual role of sodium phenoxide

Article abstract of DOI:10.1002/anie.200702943

(Chemical Equation Presented) No pressure, no worries: A general, functional-group-tolerant, mild system for the Pd-catalyzed carbonylation of aryl chlorides to the corresponding amides has been developed. The catalyst operates at 1 atm CO using an inexpensive, air-stable, and commercially available ligand (see scheme, Cy = cyclohexyl). Sodium phenoxide is a critical additive in this transformation; its role has been studied using in situ IR spectroscopy.

Full text of DOI:10.1002/anie.200702943

Communications
DOI: 10.1002/anie.200702943
Carbonylation
Palladium-Catalyzed Aminocarbonylation of Aryl Chlorides at
Atmospheric Pressure: The Dual Role of Sodium Phenoxide**
Joseph R. Martinelli, Thomas P. Clark, Donald A. Watson, Rachel H. Munday, and
Stephen L. Buchwald*
Since Heckꢀs seminal reports in 1974, palladium-catalyzed
carbonylation has emerged as an indispensable method for
the regioselective installation of carbonyl functional groups
and remains an area of active research.[1,2] Among the aryl
halides, aryl chlorides are particularly attractive starting
materials for these transformations, as they are both inex-
pensive and commercially abundant.[3] The relative chemical
inertness of the aryl chlorides also makes them more
challenging substrates,[3] and aryl chloride carbonylation
methodologies remain underdeveloped compared to those
of the more expensive aryl bromide,[4] iodide,[5] and triflate[6]
variants.
In particular, only limited methods exist for the prepara-
tion of amides from unactivated aryl chlorides (aminocarbo-
nylation).Nearly 20 years ago, Milstein and co-workers
demonstrated that the bulky, electron-rich bidentate phos-
phine ligand bis(diisopropylphophino)propane (dippp) is an
effective ligand in palladium-catalyzed aryl chloride carbon-
ylation.[7a] However, the reaction requires high temperature
(1508C) and high CO pressure (70 psi).While highly signifi-
cant, this study was limited in scope (it only included three
examples of aminocarbonylation) and did not address func-
tional-group compatibility, the effects of ortho substituents, or
the use of reagents other than dialkyl amines.Moreover, the
reaction conditions dramatically limit the utility of this
method in preparative chemistry because of the high temper-
ature, the need for specialized pressure reactors, and the
dangers associated with handling pressurized CO.Recently,
Beller and co-workers have shown that lower CO pressures
can be used in aryl chloride carbonlyation reactions (1 bar)
when electron-rich bidentate bisphosphine ligands based
upon ferrocene are employed.[7b,c] However, like Milsteinꢀs
report, this study focused primarily on the formation of esters
(alkoxycarbonylation), with only a single aminocarbonylation
(chlorobenzene and di-n-propylamine) being reported.Fur-
thermore, reactions in this later study also required high
reaction temperatures (1458C).Finally, Lagerlund and
Larhed have reported microwave conditions for aryl chloride
aminocarbonylation.[7d] Although greater substrate scope was
demonstrated in this study, hexacarbonylmolybdenum was
required as a CO surrogate, and the reaction cannot be
performed under standard thermal conditions.Like the
previous reports, high temperatures (1708C) are required.
Thus, further development in this area is clearly needed.
Herein, we report a general and practical method for the
aminocarbonylation of aryl chlorides.The reported reactions
proceed at atmospheric CO pressure and moderate temper-
atures, and these conditions have been employed for a wide
range of aryl chlorides and amines.Unlike many new methods
employing palladium catalysis, successful development of this
chemistry did not hinge on a particular precatalyst/ligand
combination.Rather, the critical advance was the use of
sodium phenoxide as the basic additive, which acts both as a
base and as an acyl transfer agent, thereby facilitating an
otherwise challenging transformation.
For the initial experiments, 4-n-butylchlorobenzene, 3-
chloroanisole, and morpholine were chosen as test substrates,
and reactions were conducted under balloons filled with CO
(Tables 1 and 2).As was found in previous studies, ligand
screening indicated that electron-rich, bulky bisphosphines
were the most effective.Herein, 1,3-bis(dicyclohexylphosphi-
no)propane[8] (dcpp) was the most efficient ligand of those
screened, and it could be introduced as the air-stable
tetrafluoroboric acid adduct.It is of note that, as in previous
Table 1: Optimization of atmospheric-pressure aminocarbonylation of 4-
n-butylchlorobenzene.
[*] Dr. J. R. Martinelli, Dr. T. P. Clark, Dr. D. A. Watson, Dr. R. H. Munday,
Prof. Dr. S. L. Buchwald
Department of Chemistry
Massachusetts Institute of Technology
77 Massachusetts Ave. Cambridge, MA 02139 (USA)
Fax: (+1)617-253-3297
Ligand
Conversion [%]
Yield (GC) [%]
1
6
0
2
1
0
3
6
0
4
8
4
5[a]
7
0
6
61
57
7
92
91
8
14
7
E-mail: sbuchwal@mit.edu
[a] [(Cy3P)2PdCl2] employed; Cy=cyclohexyl.
[**] We thank the NationalInstitutes of Heatlh (GM46059) and the
MIT–Singapore Alliance for funding. We are indebted to Merck,
Amgen, and Boehringer-Ingelheim for unrestricted support. We are
also extremely grateful to Nippon Chemical Co. for providing
generous donations of 6–8. We also thank BASF for a donation of
palladium acetate.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
8460
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 8460 –8463

Angewandte
Chemie
Table 3: Substrate scope of atmospheric-pressure aminocarbonylation
studies, a three-carbon linker between the phosphorous atoms
proved optimal.[7a–c] We believe that the ligand having this
linker, in addition to being electron-rich and bulky, may be a
critical requirement of the ligands in these transformations.[9]
Further optimization revealed that anhydrous DMSO was the
optimal solvent.In general, exclusion of extraneous water,
accomplished through the use of activated 4-Š molecular
sieves, was necessary to obtain high yields.
with 7.[a,b]
Entry ArClAmine
1
Product
d [%]
Yiel
As mentioned above, the identity of the base proved
critical to the success of the carbonylation reaction (Table 2).
Initial screening of standard basic additives failed to provide
88[c,d]
79[e]
2
Table 2: Effect of base on the aminocarbonylation of 3-chloroanisole.
3
4
88
85[e]
5
6
93[f]
97[c,g]
Entry
Base
Conversion
ArCl[%]
Yield [%]
10
9
11
7
8
93[e,h]
65
1
2
3
4
5
6
7
nPr3N
Na2CO3
Cs2CO3
K3PO4
11
94
>99
>99
78
11
52
8
62
25
70
–
–
–
–
–
–
–
–
38
n.d.[a]
n.d.
n.d.
n.d.
–
DBU
98[d]
92
Na2CO3/25% PhOH
PhONa
>99
>99
9
>99
8
9
>99
76
22
–
10
80
60
<2
n.d.
11
12
13
14
92
99
94
75
[a] Not determined.
high yields of the desired amide product.The use of sodium
carbonate resulted in the formation of a significant amount of
carboxylic acid.On the other hand, the use of anhydrous
organic bases, such as tri-n-propylamine or 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU), resulted in very low conversion of
the aryl chloride.We hypothesized that the problematic step
in the catalytic process might be transfer of the acyl group
from the palladium center to the amine and decided to
explore additives that might act as acyl transfer agents to
facilitate this process.Consistent with this reasoning, the
addition of phenol to the reaction mixture provided a modest
increase in the observed yield of amide 9.The use of
anhydrous sodium phenoxide (NaOPh) as base resulted in
even more dramatic improvements, providing complete
conversion of the aryl chloride and an excellent yield of the
desired amide.Sterically hindered phenoxides were also
examined but did not perform as well.
To determine the scope of this process, many different aryl
chlorides and amines were employed as substrates (Table 3).
Primary, secondary, and aromatic amines are all readily
converted into amides.Additionally, electron-rich, -neutral,
and -poor aryl chlorides are all compatible with these
aminocarbonylation conditions.Aryl chlorides with ortho
substituents were also well-tolerated (Table 3, entry 15).For
15
86[c,d]
[a] Reaction conditions: 2 mol% Pd(OAc)2, 4–5 mol% 7, 1 mmolaryl
chloride, 3 mmol amine, 2 mmol NaOPh, 150 mg 4-Š molecular sieves
in DMSO (1 mL) at 1008C for 15 h. [b] Yields (of isolated product) are an
average of two runs. [c] Reaction time 4 h. [d] Reaction temperature
1108C. [e] Reaction temperature 1208C. [f] Reaction time 3 h. [g] Less
catalyst used: 0.5 mol% Pd(OAc)2, 2 mol% 7. [h] 4 equiv of amine used.
these substrates, the reactions proceeded at 100–1108C.The
combination of a primary amine and an electron-poor aryl
chloride can be successfully transformed to the corresponding
benzamide in 4 h using only 0.5 mol% catalyst (Table 3,
entry 6).Not surprisingly, acyclic secondary amines require
higher temperatures to afford complete transformation to the
desired amide (Table 3, entries 2, 4, and 7).However, in no
case were temperatures in excess of 1208C required, which is
a considerably lower temperature than in previously reported
conditions.While methyl esters underwent amidation under
Angew. Chem. Int. Ed. 2007, 46, 8460 –8463ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8461

Communications
these reaction conditions, other functional groups, such as the
conversion (t1/2 ꢀ 12 m) of ester 13 to amide 14 was
observed.[10]
The above observations suggest that sodium phenoxide is
playing two distinct roles in the formation of amide 14
(Scheme 1).First, owing to its greater nucleophilicity com-
t-butyl ester and nitrile groups, remained intact (Table 3,
entries 8 and 14).Heteroaryl chlorides, such as 3-chloropyr-
idine and 2-chlorothiophene, were converted into the corre-
sponding benzamides without complications (Table 3,
entries 10–13).
To gain evidence for the proposed role of NaOPh in these
transformations, the reaction of 3-chloroanisole and di-n-
butylamine was monitored using in situ IR spectroscopy.In
the initial kinetic experiments, Pd(OAc)2 (2 mol%) and 7
(4 mol%) were used as the catalytic additives.These reac-
tions displayed a variable initiation period, presumably owing
to the reduction of Pd(OAc)2 to Pd0.To improve reproduci-
bility in the kinetic experiments, we prepared [(dcpp)PdPhCl]
(12) from dcpp and [(Ph3P)2PdPhCl] as the monotoluene
solvate.Using 12 (2 mol%) and 7 (2 mol%) as the precatalyst
eliminated the initiation period.
Scheme 1. Proposed reaction pathway for reactions involving acyclic
Figure 1 shows the reaction profile as determined by
in situ IR spectroscopy resulting from the combination of 3-
chloroanisole, di-n-butylamine, and NaOPh under catalytic
conditions at 1208C.A signal at 1736 cm À1, corresponding to
secondary amines.
pared to di-n-butylamine, NaOPh intercepts the palladium
acyl species resulting from oxidative addition of the aryl
chloride and migratory insertion of CO[11] and leads to the
formation of 13.[12,13] We suspect that this lower energy
pathway, involving intermediate ester 13, is the critical feature
that allows lower operational temperatures in this method
compared to previously reported systems that employ similar
ligand/metal systems.Second, NaOPh acts as a Brønsted base
in catalyzing the conversion of the intermediate phenyl ester
to the amide product.The phenoxide anion may also play a
role as a ligand to the metal center by displacement of
chloride.However, at this point we have no evidence to
support this suggestion.Ongoing studies are directed at
further elucidation of these subtle mechanistic details.
In conclusion, we have developed a general, practical
protocol for the aminocarbonylation of aryl chlorides at
atmospheric pressure of CO.Electron-deficient, -neutral and
-rich aryl chlorides were all successfully transformed into the
corresponding amides.Primary, a-branched primary, cyclic
secondary, acyclic secondary, and aryl amines were all
productive in the reaction.Furthermore, the process tolerates
functional groups and utilizes an inexpensive, air-stable, and
commercially available ligand salt.Like previous studies in
this area, the optimal ligand proved to be an electron-rich
bulky bisphosphine.However, the critical innovation in this
process was the use of sodium phenoxide as the basic additive.
In addition to acting as a base, this reagent facilitates acyl
transfer through the formation of phenyl esters as intermedi-
ates and catalyzes the conversion of phenyl ester into the final
amide product.This route has resulted in lower reaction
temperatures and CO pressures in the aminocarbonylation
process, which greatly improves both the practicality and
safety profile of this transformation, and for the first time has
resulted in a general method for the aminocarbonylation of
aryl chlorides.The dual role of NaOPh revealed during this
study underscores the concept that ligands alone, while
important for success in developing a method, are not the
only reaction parameter that needs to be considered.In this
case, consideration of the potential reactive intermediates
Figure 1. Kinetic profile for the reaction of di-n-butylamine and 3-
chloroanisole.
phenyl 3-methoxybenzoate 13, was observed at the beginning
of the reaction.This signal reached a maximum intensity after
about 1 h, and then slowly decayed.Additionally, a signal at
1632 cmÀ1, corresponding to amide 14, was observed a few
minutes after the beginning of the reaction and continued to
increase in intensity for approximately 5 h.At the end of the
reaction, the aryl chloride and 13 had been completely
consumed, and 14 was formed in 88% yield (determined by
GC).Thus, as predicted, phenyl ester 13 appears to be an
intermediate in the formation of amide 14.
To further confirm the intermediacy of the ester, we
examined the kinetics of the conversion of ester 13 to amide
14.Surprisingly, combining ester 13 and di-n-butylamine in
DMSO at 1208C resulted only in the very slow formation of
amide 14 (t1/2 ꢀ 10 h).However, when the same reaction was
conducted in the presence of NaOPh (1 equiv), rapid
8462 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 8460 –8463

Angewandte
Chemie
resulted in the introduction of an unusual basic additive and
allowed alternate reaction pathways to proceed, resulting in
the successful development of this methodology.
[1] a) A.Schoenberg, I.Bartoletti, R.F.Heck, J. Org. Chem. 1974,
39, 3318; b) A.Schoenberg, R.F.Heck, J. Org. Chem. 1974, 39,
3327.
[2] a) HM. .Colquhoun, DJ..Thompson, MV. .Twigg,
Carbon-
ylation, Direct Synthesis of Carbonyl Compounds, Plenum, New
York, 1991; b) M.Beller, W.Mägerlein, AF. .Indolese, C.
Fischer, Synthesis 2001, 1098.
Experimental Section
General procedure (NOTE: Carbon monoxide should only be
handled in a well-ventilated fume hood.): An oven-dried culture
tube (18 ” 150 mm, VWR) or screw-cap test tube equipped with a
teflon-coated magnetic stir bar was charged with 4-Š molecular sieves
and sealed with a 14/20 rubber septum (inverted) or screw cap and
teflon-lined septum.The tube was then evacuated, heated for about
1 min with a Bunsen burner or for about 2–3 min with a heat gun to
activate the molecular sieves; the tube was then backfilled with argon
and allowed to cool.Subsequently, the tube was taken into a glovebox
(a needle was inserted in the septum or the septum was removed upon
entering the antechamber) and charged with anhydrous NaOPh
(2 mmol, 2 equiv, 232 mg).The tube was resealed inside the glovebox
and removed.The solid reagents were added in air by briefly
removing the rubber septum: Pd(OAc)2 (2 mol%, 0.02 mmol,
0.02 equiv, 4.5 mg) and 1,3-bis(dicyclohexylphosphino)propa-
ne·2HBF4 salt (7, 4–5 mol%, 0.04–0.05 mmol, 0.04–0.05 equiv, 24.5–
30.6 mg). After the addition of all solid reagents, the rubber septum
was secured by wrapping with electrical tape.Next, all liquid reagents
were added by syringe: DMSO (1 mL), aryl chloride (1 mmol,
1 equiv; aryl chlorides which were solids at room temperature were
added during the initial charge), and amine (3 mmol, 3 equiv).Once
all reagents were added, a balloon was connected to the reaction
vessel using a short length of rubber tubing (ca.1 in.), a needle
adapter, and a 20 G needle.The inert atmosphere was then exchanged
for carbon monoxide by briefly exposing the reaction vessel to
vacuum (1–2 s) and backfilling with carbon monoxide; the balloon
was inflated with CO(g) directly following this atmosphere exchange.
The reaction tube was then submerged in a preheated oil bath (100–
1208C).The reaction mixture was heated while being stirred
vigorously for 15 h or until the aryl halide had been completely
consumed as judged by GC analysis.The reaction mixture was then
allowed to cool to room temperature, diluted with methylene chloride
or ethyl acetate (ca.10 mL), filtered through a plug of celite (eluting
with methylene chloride or ethyl acetate), and concentrated under
reduced pressure.The crude material was purified by flash chroma-
tography on silica gel.
[3] A.F.Littke, G.C.Fu,
Angew. Chem. 2002, 114, 4350; Angew.
Chem. Int. Ed. 2002, 41, 4176.
[4] a) T.Ishiyama, H.Kizaki, T.Hayashi, A.Suzuki, N.Miyaura,
J.
Org. Chem. 1998, 63, 4726; b) N.-F. K. Kaiser, A. Hallberg, M.
Larhed, J. Comb. Chem. 2002, 4, 109; c) K.Kumar, A.Zapf, D.
Michalik, A.Tillack, T.Heinrich, H.Böttcher, M.Arlt, M.
Beller, Org. Lett. 2004, 6, 7; d) J.Albaneze-Walker, C.Bazaral, T.
Leavey, P.G.Dormer, J.A.Murry, Org. Lett. 2004, 6, 2097; e) S.
Klaus, H.Neumann, A.Zapf, D.Strübing, S.Hübner, J.Almena,
T.Riermeier, P.Groß, M.Sarich, W.-R.Krahnert, K.Rossen, M.
Beller, Angew. Chem. 2006, 118, 161; Angew. Chem. Int. Ed.
2006, 45, 154; f) J.R. Martinelli, D.M.M. Freckmann, S.L.
Buchwald, Org. Lett. 2006, 8, 4843.
[5] a) M.S.M. Ahmed, A. Mori, Org. Lett. 2003, 5, 3057; b) G.
Abbiati, A.Arcadi, V.Canevari, L.Capezzuto, E.Rossi, J. Org.
Chem. 2005, 70, 6460.
[6] a) S.Cacchi, P.G.Ciattini, E.Morera, G.Ortar,
Tetrahedron
Lett. 1986, 27, 3931; b) O.Rahman, T.Kihlberg, B.Lꢁngström, J.
Org. Chem. 2003, 68, 3558.
[7] a) Y.Ben-David, M.Portnoy, D.Milstein,
J. Am. Chem. Soc.
1989, 111, 8742; b) W.Mägerlein, AF. .Indolese, M.Beller,
Angew. Chem. 2001, 113, 2940; Angew. Chem. Int. Ed. 2001, 40,
2856; c) W.Mägerlein, A.F.Indolese, M.Beller,
J. Organomet.
Chem. 2002, 641, 30; d) O.Lagerlund, M.Larhed, J. Comb.
Chem. 2006, 8, 4.
[8] a) Commercially available from Nippon Chemical Co.(103099-
52-1); b) for
Information.
a convenient synthesis, see the Supporting
[9] The complex reported by Milstein[7a] and the ferrocene-based
ligand reported by Beller[7b,c] can be employed and afford similar
results.
[10] Initial efforts to measure rate constants have proven very
sensitive, and we have thus far been unable to determine reagent
order.
[11] a) P.E.Garrou, R.F.Heck, J. Am. Chem. Soc. 1976, 98, 4115;
b) D.Milstein, Acc. Chem. Res. 1988, 21, 428.
[12] For previous examples of the production of aryl esters from aryl
bromides and iodides using carbonylation, see: a) Y.Kubota, T.
Hanaoka, K.Takeuchi, Y.Sugi, Synlett 1994, 515; b) Y.Kubota,
Received: July 3, 2007
Published online: September 27, 2007
T.Hanaoka, K.Takeuchi, Y.Sugi,
L187.
J. Mol. Catal. A 1996, 111,
[13] For an example of reductive elimination of aryl carboxylates,
Keywords: amides · aryl chlorides · carbonylation · palladium ·
phosphines
see: S.Komiya, Y.Akai, K.Tanaka,
1130.
Organometallics 1985, 4,
.
Angew. Chem. Int. Ed. 2007, 46, 8460 –8463ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8463