Synthesis of amides and lactams in supercritical carbon dioxide

Article abstract of DOI:10.1021/jo9021875

(Chemical Equation Presented) Supercritical carbon dioxide can be employed as an environmentally friendly alternative to conventional organic solvents for the synthesis of a variety of carboxylic amides. The addition of amines to ketenes generated in situ via the retro-ene reaction of alkynyl ethers provides amides in good yield, in many cases with ethylene or isobutylene as the only byproducts of the reaction. Reactions with ethoxy alkynes are performed at 120-130°C, whereas tert-butoxy derivatives undergo the retro-ene reaction at 90°C. With the exception of primary, unbranched amines, potential side reactions involving addition of the amines to carbon dioxide are not competitive with the desired C-N bond-forming reaction. The amide synthesis is applicable to the preparation of β-hydroxy and β-amino amide derivatives, as well as amides bearing isolated carbon-carbon double bonds. Preliminary experiments aimed at developing an intramolecular variant of this process to afford macrolactams suggest that the application of CO₂/co-solvent mixtures may offer advantages for the synthesis of large-ring compounds.

Full text of DOI:10.1021/jo9021875

pubs.acs.org/joc
Synthesis of Amides and Lactams in Supercritical Carbon Dioxide
Xiao Yin Mak,† Rocco P. Ciccolini,‡ Julia M. Robinson,† Jefferson W. Tester,*,‡,§ and
Rick L. Danheiser*,†
†Department of Chemistry and ‡Department of Chemical Engineering, Massachusetts Institute of
Technology, Cambridge, Massachusetts 02139. §Current address: School of Chemical and
Biomolecular Engineering, Cornell University, Ithaca, NY.
danheisr@mit.edu; jwt54@cornell.edu
Received October 10, 2009
Supercritical carbon dioxide can be employed as an environmentally friendly alternative to conven-
tional organic solvents for the synthesis of a variety of carboxylic amides. The addition of amines to
ketenes generated in situ via the retro-ene reaction of alkynyl ethers provides amides in good yield, in
many cases with ethylene or isobutylene as the only byproducts of the reaction. Reactions with ethoxy
alkynes are performed at 120-130 °C, whereas tert-butoxy derivatives undergo the retro-ene reaction
at90 °C. Withthe exceptionof primary, unbranchedamines, potential side reactionsinvolving addition
of the amines to carbon dioxide are not competitive with the desired C-N bond-forming reaction. The
amide synthesis is applicable to the preparation of β-hydroxy and β-amino amide derivatives, as well
as amides bearing isolated carbon-carbon double bonds. Preliminary experiments aimed at develop-
ing an intramolecular variant of this process to afford macrolactams suggest that the application of
CO2/co-solvent mixtures may offer advantages for the synthesis of large-ring compounds.
Introduction
scCO2 as a reaction solvent for a variety of synthetic transfor-
mations is now well documented. To date, however, only a
few examples have been reported of carbon-nitrogen bond
formation in scCO2,3 principally due to the facility of the
reaction of amines with this electrophilic solvent (vide infra).
One goal of our program is the development of general
strategies for the utilization of amines (and amine derivatives)
in scCO2, and the application of these strategies in C-N bond-
forming reactions and the synthesis of nitrogen heterocycles.
The carboxamide functional group occurs in the structure
of 25% of known pharmaceuticals according to a 1999
analysis of the Comprehensive Medicinal Chemistry data-
base.4 A recent study of the syntheses of drug candidates at
A major goal of research in “green chemistry” involves
the replacement of conventional organic solvents with more
“environmentally friendly” alternatives.1 Supercritical carbon
dioxide (scCO2) has attracted considerable attention in recent
years as an alternative to conventional solvents for organic
synthesis.2 This interest derives from the fact that carbon
dioxide is relatively nontoxic and nonflammable, inexpensive,
and widely available and poses minimal problems with regard
to waste disposal. The tunable solvent properties of scCO2 have
also attracted interest, as relatively small changes in tempera-
ture and pressure often allow for significant changes in visco-
sity, density, and self-diffusivity.2 The successful application of
(1) (a) Sheldon, R. A. Green Chem. 2005, 7, 267. (b) Green Reaction
Media in Organic Synthesis; Mikami, K., Ed.; Blackwell Publishing: Oxford,
2005. (c) Clark, J. H.; Tavener, S. J. Org. Process Res. Dev. 2007, 11, 149.
(2) (a) Tester, J. W.; Danheiser, R. L.; Weinstein, R. D.; Renslo, A.; Taylor, J.
D.; Steinfeld, J. I. Supercritical Fluids as Solvent Replacements in Chemical
Synthesis. In Green Chemical Syntheses and Processes; Anastas, P. T., Heine, L.
G., Williamson, T. C., Eds.; ACS Symposium Series 767; American Chemical
Society: Washington, DC, 2000; pp 270-291. (b) Chemical Synthesis Using
Supercritical Fluids; Jessop, P. G., Leitner, W., Eds.; Wiley-VCH; Weinheim,
1999. (c) Leitner, W. Top. Curr. Chem. 1999, 206, 108. (d) Oakes, R. S.; Clifford,
A. A.; Rayner, C. M. J. Chem. Soc., Perkin Trans. 1 2001, 917. (e) Beckman, E.
J. J. Supercrit. Fluids 2004, 28, 121. (f) Oakes, R. S.; Clifford, A. A.; Rayner, C.
M. Org. Process Res. Dev. 2007, 11, 121.
(3) For examples, see: (a) Jessop, P. G.; Hsiao, Y.; Ikariya, T.; Noyori, R.
J. Am. Chem. Soc. 1996, 118, 344. (b) Wittmann, K.; Wisniewski, W.;
Mynott, R.; Leitner, W.; Kranemann, C. L.; Rische, T.; Eilbracht, P.;
Kluwer, S.; Ernsting, J. M.; Elsevier, C. J. Chem.;Eur. J. 2001, 7, 4584.
(c) Shi, M.; Cui, S.-C.; Li, Q.-J. Tetrahedron 2004, 60, 6163. (d) Smith, C. J.;
Early, T. R.; Holmes, A. B.; Shute, R. E. Chem. Commun. 2004, 1976. (e)
Smith, C. J.; Tsang, M. W. S.; Holmes, A. B.; Danheiser, R. L.; Tester, J. W.
€
Org. Biomol. Chem. 2005, 3, 3767. (f) Dunetz, J. R.; Ciccolini, R. P.; Froling,
M.; Paap, S. M.; Allen, A. J.; Holmes, A. B.; Tester, J. W.; Danheiser, R. L.
Chem. Commun. 2005, 4465. (g) Fuchter, M. J.; Smith, C. J.; Tsang, M. W. S.;
Boyer, A.; Saubern, S.; Ryan, J. H.; Holmes, A. B. Chem. Commun. 2008, 65.
(4) Ghose, A. K.; Viswanadhan, V. N.; Wendoloski, J. J. J. Comb. Chem.
1999, 1, 55.
DOI: 10.1021/jo9021875
r
Published on Web 11/18/2009
J. Org. Chem. 2009, 74, 9381–9387 9381
2009 American Chemical Society

Mak et al.
JOCArticle
Pfizer, GSK, and AstraZeneca revealed that acylations of
amines (to produce amides) comprise fully 8% of all of the
reactions involved in the 128 syntheses surveyed!5 Amide
bond formation generally requires initial conversion of the
carboxylic acid component to an “activated” acyl derivative,
which is then reacted in situ or in a second step with an
amine.6 In the recent survey of drug candidate syntheses
cited above,5 acid chlorides were found to be the most
common acylating agents employed in amide bond forma-
tion. Mixed anhydrides were the next most frequently used
derivatives, followed by various activated carboxyl species
generated by the reaction of acids with “coupling agents”
such as carbonyldiimidazole and carbodiimides. Although
the use of acid chlorides and mixed anhydrides may be
economically attractive, none of these standard methods
are atom-economical, some involve toxic reagents (many
diimides are sensitizers), and all are carried out in volatile
organic solvents and result in the production of considerable
quantities of chemical waste.7
The aim of this investigation was the development of a
method for amide bond formation compatible with the use of
scCO2 as the reaction medium. In this initial study we
focused our attention on the reaction of amines with ketenes
generated in situ by the retro-ene reaction of alkynyl ethers
(eq 1). In this process, scCO2 would function as an environ-
mentally friendly “replacement solvent” and might also
provide important advantages in intramolecular applica-
tions leading to medium- and large-ring lactams (vide infra).
A particularly noteworthy feature of this atom-economical
approach would be the formation of ethylene as the only
byproduct of the reaction. Provided that side reactions were
minimal and starting materials were completely consumed,
simple depressurization could potentially provide the amide
products free from contamination by solvent residues in
what would constitute an exceptionally green process.
TABLE 1. Optimization of Conditions for Amide Synthesis
entry alkynyl ether solvent temp (°C) pressure (bar) yield (%)a
1
2
3
4
5
3a
3a
3a
3a
3b
toluene
CO2
CO2
CO2
CO2
120
120
120
130
90
1.3b
215
394
228
218
88
85
86
88
82
aIsolated yield of products purified by column chromatography on silica
gel. bCalculated value based on the vapor pressure of toluene at 120 °C.
nucleophilic amines react reversibly with carbon dioxide to
form carbamic acids of type 1 and ammonium carbamate salts
of type 2 (eq 2).8 This can pose a significant challenge to the
transfer of C-N bond-forming reactions from conventional
solvents to scCO2, as the formation of carbamic acids (and
salts) can inhibit the reaction of the amine in the desired
fashion and canleadtothe formation of undesired byproducts
and polymers. A key to the success of the proposed amide
synthesis was therefore to identify conditions under which the
reaction of the amine with carbon dioxide would not be
competitive with the desired C-N bond-forming reactions.
Results and Discussion
Optimization of Conditions for Amide Synthesis in scCO2.
Initial feasibility experiments were carried out using
1-ethoxy-1-octyne (3a) and the corresponding tert-butoxy
alkyne, 3b. Ethoxyoctyne 3a was conveniently prepared in
high yield via the alkylation of the lithium derivative of
commercially available ethoxyacetylene with 1-iodohexane
as previously reported by Kocienski.9 The tert-butoxy deri-
vative was prepared in a similar fashion by alkylation
of lithium tert-butoxyacetylide. Greene has developed an
efficient one-pot method for the generation of lithium
alkoxyacetylides that involves the addition of a potassium
alkoxide to dichloroacetylene (generated in situ from
trichloroethylene) followed by treatment with n-BuLi.10 Our
attempts to employ this one-pot procedure in the preparation
of 3b gave the desired alkyne in poor yield; however, improved
results were obtained by isolating the product of the first stage
of the reaction (tert-butyl 1,2-dichlorovinyl ether) and then
subjecting it to reaction with 2.4 equiv of n-BuLi followed by
iodohexane.11 Although not especially attractive from the
standpoint of “green chemistry”, these approaches to the
synthesis of 3a and 3b provided us with rapid access to
abundant quantities of the alkynes as required for our initial
feasibility and optimization studies.
At the outset of this study, we recognized that a potential
obstacle toward the successful implementation of the pro-
posed method would be the reactivity of carbon dioxide
toward basic amino groups. It is well documented that
(5) Carey, J. S.; Laffan, D.; Thomson, C.; Williams, M. T. Org. Biomol.
Chem. 2006, 4, 2337.
(6) For a recent review, see Montalbetti, C. A. G. N.; Falque, V.
Tetrahedron 2005, 61, 10827.
(7) For recent examples of environmentally friendly methods for amide
synthesis, see: (a) Gunanathan, C.; Ben-David, Y.; Milstein, D. Science 2007,
317, 790. (b) Nordstrøm, L. U.; Vogt, H.; Madsen, R. J. Am. Chem. Soc.
2008, 130, 17672. (c) Al-Zoubi, R. M.; Marion, O.; Hall, D. G. Angew.
Chem., Int. Ed. 2008, 47, 2876. (d) Comerford, J. W.; Clark, J. H.; Macquarrie,
D. J.; Breeden, S. W. Chem. Commun. 2009, 2562.
(8) (a) Dell’Amico, D. B.; Calderazzo, F.; Labella, L.; Marchetti, F.;
Pampaloni, G. Chem. Rev. 2003, 103, 3857. (b) Aresta, M.; Ballivet-Tkatchenko,
D.; Dell’Amico, D. B.; Bonnet, M. C.; Boschi, D.; Calderazzo, F.; Faure, R.;
Labella, L.; Marchetti, F. Chem. Commun. 2000, 1099. (c) Park, J.-Y.; Yoon, S.
J.; Lee, H. Environ. Sci. Technol. 2003, 37, 1670. (d) Masuda, K.; Ito, Y.;
Horiguchi, M; Fujita, H. Tetrahedron 2005, 61, 213. (e) For a study of the effect
of increased steric demand on the formation of carbamic acids and carbamates
from primary amines, see: Fischer, H.; Gyllenhaal, O.; Vessman, J.; Albert, K.
Anal. Chem. 2003, 75, 622. (f) For a discussion on the effect of temperature and
solvation effects on carbamic acid and carbamate salt equilibrium, see: Dijkstra, Z.
J.; Doornbos, A. R.; Weyten, H.; Ernsting, J. M.; Elsevier, C. J.; Keurentjes, J. T. F.
J. Supercrit. Fluid. 2007, 41, 109 and references therein.
Table 1 summarizes the results of initial experiments to
establish the feasibility of carrying out the desired amide
(9) Pons, J.-M.; Kocienski, P. Tetrahedron Lett. 1989, 30, 1833.
(10) Denis, J.-N.; Moyano, A.; Greene, A. E. J. Org. Chem. 1987, 52,
3461.
(11) For details, see Supporting Information.
9382 J. Org. Chem. Vol. 74, No. 24, 2009

Mak et al.
JOCArticle
synthesis in supercritical carbon dioxide. Thermolysis of
ethoxyoctyne in the presence of N-benzylbutylamine was
examined first. A temperature of 120 °C was initially selected
for this reaction on the basis of a review of conditions
previously reported for related transformations involving
alkynyl ethers. The retro-ene reaction of alkynyl ethers to
form ketenes12 was first observed in the laboratories of
Ficini13 and Arens,14 and Ficini appears to have been the
first to employ this process for the synthesis of amides.
Subsequent researchers have applied this process to generate
ketenes for cyclizations leading to lactones,15 lactams,16a and
cyclic imides.16b Solvents typically employed in these reac-
tions include xylene, toluene, chloroform, and acetonitrile.
The retro-ene reaction of ethoxy alkynes typically takes place
at 100-120 °C, but more highly substituted ethers undergo
the reaction at lower temperatures, as low as 50 °C in the case
of tert-butyl derivatives.15a,b,17,18
We were pleased to find that addition of N-benzylbutyl-
amine to the ketene generated in situ from ethoxyoctyne
proceeds as cleanly and efficiently in scCO219 as the reaction
carried out via the conventional procedure using toluene in a
sealed tube (Table 1, entries 1-3). Particularly significant is
the fact that no interference from the reversible reaction of
the amine with CO2 (see eq 2) was observed in this case. The
optimal temperature for the reaction in scCO2 was 130 °C
(entry 4); at 110 °C unreacted alkynyl ether (ca. 25%) was
recovered even after 39 h. The use of the tert-butoxy octyne
3b permits the reaction to be carried out at 90 °C and also
affords the desired amide 5 in excellent yield (entry 5).
At 120 °C, the minimum pressure required to completely
solubilize the reactants was found to be ca. 215 bar (entry 2).
At this temperature and pressure, the reaction mixture is
initially monophasic and then becomes biphasic after ca. 5
h, at which point a new liquid phase is observed to form. As
shown inFigure 1a, eventually the reactionmixtureconsists of
a (top) lower-density, CO2-rich, supercritical-like phase and a
(bottom) higher-density, product amide-rich,20 liquid phase.
Initially, the appearance of a liquid phase led to some concern
with regard to the potential detrimental effects of phase
partitioning of the reactants. However, no difference in yield
was observed when the reaction was performed at 394 bar
(entry 3), a pressure at which a single phase was observed
throughout the entire duration of the reaction (Figure 1b).
FIGURE 1. Phase behavior observed for reactions in Table 1 entry
2 (a) and entry 3 (b). The solid object is a stir bar.
Notably, no ketene dimer or products of [2 þ 2] cycloaddi-
tion of the ketene and alkynyl ether starting material were
detected in the crude products of these reactions. Depressuri-
zation of the reaction mixture after 24 h provided the amide 5
as an oil, which was determined to be 95-98% pure by
1H NMR analysis. In these experiments the product was
generally found to be contaminated with some solid debris
originating from abrasion of the o-rings and reactor wall.
Consequently, the amide 5 was transferred out of the reactor
and subjected to column chromatography to remove this
material, resulting in the loss of ca. 5% of product as
estimated on the basis of control experiments.
Scope of the Reaction with Respect to Amine. As shown in
Table 2, a variety of amines participate in the desired amide
bond-forming reaction in supercritical carbon dioxide. In these
reactions the amine was heated with 1 equiv of ethoxyoctyne (3a)
at 120 or 130 °C for 24 h; the higher temperature was required in
some cases as ca. 5% of 3a remained after 24 h at 120 °C. The
reaction pressure (230-284 bar) was selected on the basis of
the minimum pressure sufficient to solubilize both the amine and
alkynyl ether and at least initially afford a homogeneous solution
at the elevated reaction temperature. In each case, however,
the product amides were observed to eventually separate from
the reaction mixture as a second, liquid phase.
As demonstrated by the examples in Table 1 and Table 2
(entries 1 and 2), secondary amines such as 4, 6, and 8 readily
undergo reaction in scCO2 to afford the desired amides in
good yield. The R-branched primary amines 10 and 14 also
gave the expected amides in good yield (entries 3 and 5);
however, in these cases insoluble solids that we suspect to be
carbamate salts were observed to form upon initial addition
of CO2 to the reactor at room temperature (Figure 2). As the
reactor temperature warmed to ca. 100 °C, these solids
disappeared, forming a liquid phase that dissolved in CO2
upon further heating to the final reaction temperature.8f
In the case of the less nucleophilic primary amine aniline,
amide formation also proceeded in fairly good yield
(entry 4). The limit with regard to the scope of this reaction
of primary amines in scCO2 was revealed by the reaction of
benzylamine (entry 6). As in the case of entries 3 and 5,
immediate precipitation of a carbamate salt was observed
upon introduction of CO2 to a mixture of this amine and
alkynyl ether at room temperature. However, in this case the
desired amide 17 was obtained in only 37-43% yield.
We suspect that with this less sterically hindered amine the
equilibrium described in eq 2 is shifted more in favor of the
carbamic acid, thereby inhibiting addition of the amine
to the transient ketene intermediate. In support of this
(12) Reviewed in: (a) Brandsma, L.; Bos, H. T.; Arens, J. F. In Chemistry of
Acetylenes; Viehe, H. G., Ed.; Marcel Dekker: New York, 1969; pp 808-811. (b)
George, D. M.; Danheiser, R. L. In Science of Synthesis; Danheiser R. L., Ed.;
Thieme: Stuttgart, 2006; Vol. 23, pp 55-56. (c) Tidwell, T. T. In Science of
Synthesis; Danheiser R. L., Ed.; Thieme: Stuttgart, 2006; Vol. 23, pp 588-590.
(13) Ficini, J. Bull. Soc. Chim. Fr. 1954, 1367.
(14) Nieuwenhuis, J.; Arens, J. F. Recl. Trav. Chim. Pays-Bas 1958, 77,
761.
(15) (a) Funk, R. L.; Abelman, M. M.; Jellison, K. M. Synlett 1989, 36. (b)
Magriotis, P. A.; Vourloumis, D.; Scott, M. E.; Tarli, A. Tetrahedron Lett.
1993, 34, 2071. (c) Liang, L.; Ramaseshan, M.; MaGee, D. I. Tetrahedron
1993, 49, 2159.
(16) (a) MaGee, D. I.; Ramaseshan, M. Synlett 1994, 743. (b) MaGee, D.
I.; Ramaseshan, M.; Leach, J. D. Can. J. Chem. 1995, 73, 2111.
ꢀ
(17) (a) Valenti, E.; Pericas, M. A.; Serratosa, F.; Mana, D. J. Chem. Res.,
Synop. 1990, 118. (b) Valentı, E.; Pericas, M. A.; Serratosa, F. J. Org. Chem.
~ꢀ
ꢀ
1990, 55, 395.
(18) van Daalen, J. J.; Kraak, A.; Arens, J. F. Recl. Trav. Chim. Pays-Bas
1961, 80, 810.
(19) All reactions in scCO2 were performed in a 25-mL Thar stainless steel
view cell reactor fitted with two coaxial sapphire windows to allow visual
monitoring of the reaction. A detailed description of the reactor setup is
included in Supporting Information.
(20) Amide 5 was found to be insoluble at 211 bar in scCO2 at 120 °C.
J. Org. Chem. Vol. 74, No. 24, 2009 9383

Mak et al.
JOCArticle
TABLE 2. Synthesis of Amides by Reaction of Amines and 1-Ethoxy-
1-octyne in Supercritical Carbon Dioxide
CO2-rich phase.8 Amide formation would also be less favor-
able in this scenario since partitioning of the amine to a
greater extent into the amide-rich liquid phase would reduce
the efficiency of its reaction with the ketene intermediate in
the upper CO2-rich phase. That effects such as these may not
be operating is suggested by the results of carrying out the
reaction of benzhydrylamine 10 at a pressure (208 bar) at
which the amine was observed to be insoluble in the CO2-rich
phase that alkynyl ether 3a (and the derived ketene) were
shown to be present in. In this experiment, the yield of amide
11 was unchanged from that observed under the monophasic
conditions of entry 3.
Scope of the Reaction with Respect to Alkynyl Ether.
Further studies focused on the application of this chemistry
to the preparation of amides starting from a variety of
alkynyl ether derivatives.21 Table 3 summarizes our results.
Reactions involving ethoxy alkynes were carried out at
130 °C, while the use of tert-butoxy alkynyl ethers
allows the reaction to be achieved at lower temperature
(entries 1 and 3). For example, primary amine 10 reacts with
tert-butoxy alkyne 3b at 90 °C (entry 1) to afford amide 11 in
essentially the same yield as that obtained via reaction at
130 °C using ethoxyoctyne 3a (Table 2, entry 3). Interes-
tingly, in the reaction at 90 °C the initially formed precipitate
(that we suspect to be carbamate salt) was observed to be
present for nearly the entire duration of the reaction, in
contrast to reaction at the higher temperature where the solid
eventually disappeared. The fact that the yield of amide
product was nearly identical in these two cases suggests that
in the lower temperature reaction sufficient free amine is
present via the equilibrium described in eq 2 to allow amide
formation to compete with undesired ketene dimerization
and cycloaddition pathways.
Entries 2 and 3 describe our attempts to apply this
chemistry to the preparation of β-hydroxy amide derivatives.
Reaction of the ethyl alkynyl ether 18 with N-benzylbutyl-
amine (4) afforded the desired amide 19 in 56% yield,
accompanied by 31% of the β-elimination product 20
(entry 2). Interestingly, when this reaction was carried out
in toluene (also at 130 °C), the desired amide was obtained in
only 10% yield together with 80% of the elimination pro-
duct. Improved results were obtained by carrying out the
reaction at lower temperature in scCO2 (entry 3). Thus,
reaction at 90 °C employing the tert-butoxy alkynyl ether
21 led to the desired amide 19 in 80% yield with only trace
amounts of the R,β-unsaturated byproduct observed under
these conditions.
aReaction conditions: 1:1 ratio of amine and 3a, scCO2 (230-284 bar,
see text), 130 °C (120 °C in the case of entry 2), 24 h. bIsolated yield of
products purified by column chromatography on silica gel unless
otherwise indicated. Isolated yield of product purified by trituration
with dichloromethane/hexanes.
c
FIGURE 2. Phase behavior initially observed for reactions in
Table 2, entries 3, 5, and 6.
Reaction of alkynyl ether 22 to form amide 23 proceeded
in good yield (entry 4). As expected, [2 þ 2] cycloaddition of
the ketene intermediate to the isolated carbon-carbon
double bonds was found not to be competitive with the
desired nucleophilic addition of the amine. Our focus
turned next to the synthesis of β-amino amides. For this
purpose, alkynyl ethers 24 and 26 were prepared via the
addition of a lithium ethoxyacetylide to the R-amido
sulfone 28, which serves in this reaction as an N-acyl imine
equivalent (Scheme 1).22 Reaction of these alkynyl ethers
with piperidine furnished the expected amides in good
hypothesis, we found that in toluene the reaction of benzyl-
amine with alkynyl ether 3a takes place efficiently at 130 °C
to afford amide 17 in 98% yield.8e
Alternative explanations to account for the lower yields in
the case of reactions of primary amines were ruled out by
control experiments. For example, initially we considered the
possibility that the lower yields observed for certain amines
could also be a consequence of phase partitioning effects.
More polar amines such as the unbranched primary amine 16
might be expected to partition more than less polar amines
into the amide-rich liquid phase that appears during
the course of the reaction. This could result in a shift of the
equilibrium of eq 2 more in favor of carbamic acid due to
the greater polarity of this phase as compared to the upper
(21) The alkynyl ethers were prepared via alkylation or addition reactions
of alkoxyacetylides. For details, see Supporting Information.
(22) Mecozzi, T.; Petrini, M. J. Org. Chem. 1999, 64, 8970.
9384 J. Org. Chem. Vol. 74, No. 24, 2009

Mak et al.
JOCArticle
TABLE 3. Synthesis of Amides in Supercritical Carbon Dioxide: Scope with Regard to Alkynyl Ether
aReaction conditions for entries 2-6: 1:1 ratio of amine and alkynyl ether, scCO2 (241-277 bar, see text), 130 °C, 24 h unless othewise indicated.
bIsolated yield of products purified by column chromatography on silica gel unless otherwise indicated. cIsolated yield of product purified by trituration
with dichloromethane/hexanes. dReaction conditions: 1:1 ratio of amine and alkynyl ether, scCO2 (218 bar), 90 °C, 24 h.
SCHEME 1. Synthesis of Alkynyl Ethers 24 and 26
yield (entries 5 and 6) with no evidence for formation of
β-elimination side products in either case.
Lactam Synthesis in Supercritical Carbon Dioxide. Having
completed our study of the scope of the intermolecular reaction
of amines with alkynyl ethers in scCO2, we next turned our
attention to extending this strategy to the synthesis of macro-
cyclic lactams.23 The efficient synthesis of medium- and large-
ring compounds typically requires the application of high
dilution conditions in order to minimize competitive inter-
molecular oligomerization processes.24 The use of scCO2
as a replacement solvent is obviously quite attractive for
reactions such as these in which large volumes of solvent are
a necessity.
Unfortunately, preliminary screening experiments indi-
cated that lactam 30 is formed in lower yield when scCO2 is
employed as the reaction medium as compared to the same
reaction performed in toluene in a sealed tube (Scheme 2).
For the reaction in scCO2, the reaction mixture was mono-
phasic for the first 5 h, after which time a film-like liquid
phase developed, believed to contain the dilactam 3125
and higher molecular weight oligomeric products. Some
improvement in lactam yield was observed when toluene
(10% by volume) was added as a co-solvent; under these
conditions a liquid phase did not appear until near the very
end of the reaction. This observation suggests that the lower
yield obtained in the absence of co-solvent may be due to the
preferential partitioning of the amine starting material 29
into the more polar and smaller volume liquid phase that
develops as the reaction proceeds. Migration of the starting
(23) For elegant examples of the synthesis of macrocyclic lactams (and
lactones) via intramolecular trapping of ketenes generated in situ by thermo-
lysis of dioxolenones, see: Boeckman, R. K.; Weidner, C. H.; Perni, R. B.;
Napier, J. J. J. Am. Chem. Soc. 1989, 111, 8036 and references cited therein.
€
(24) Reviewed in: (a) Knops, P.; Vogtle Top. Curr. Chem. 1991, 161, 1. (b)
Meng, O.; Hesse, M. Top. Curr. Chem. 1991, 161, 107. (c) Nubbemeyer, U.
Top. Curr. Chem. 2001, 216, 125.
(25) Dilactam 31 was found to be insoluble at 322 bar and 120 °C in
scCO2.
J. Org. Chem. Vol. 74, No. 24, 2009 9385

Mak et al.
JOCArticle
SCHEME 2
15 mL of CH2Cl2. The residual oil in the reactor was purified by
column chromatography on 12 g of silica gel (elution with 15%
EtOAc/hexanes) to provide 0.871 g (88%) of amide 5 as a yellow
oil: IR (neat) 2926, 1651, and 1453 cm-1. 1H NMR (400 MHz,
CDCl3): major rotamer δ 7.16-7.38 (m, 5H), 4.61 (s, 2H), 3.18
(t, J = 7.7 Hz, 2H), 2.38 (t, J = 7.6 Hz, 2H), 1.48-1.75 (m, 4H),
1.26-1.35 (m, 10H), and 0.85-0.94 (m, 6H); minor rotamer
δ 4.54 (s), 3.36 (t, J = 7.6 Hz), and 2.31 (t, J = 7.6 Hz). 13C
NMR (100 MHz, CDCl3): major rotamer δ 173.3, 138.3, 128.6,
128.1, 127.3, 48.2, 47.0, 33.3, 31.9, 29.7, 29.3, 25.8, 22.8, 20.2,
14.3, and 14.0; minor rotamer 173.6, 137.5, 129.0, 127.6, 126.3,
51.2, 46.1, 33.5, 30.8, 29.9, 29.5, 29.3, 25.6, 22.8, 20.4, 14.3, and
14.0. HRMS-ESI (m/z): [M þ Na]þ calcd for C19H31NO
312.2298; found 312.2292.
material 29 into the liquid phase would promote oligomer-
ization pathways as these intermolecular reactions would be
favored with the increase in molar concentration relative to
the initial concentration (0.002 M) of 29 in the bulk CO2-rich
phase.
Reaction of N-benzylbutylamine (0.506 g, 3.10 mmol) with
1-tert-butoxy-1-octyne 3b (0.565 g, 3.10 mmol) at 90 °C
(218 bar) for 24 h according to the General Procedure provided
ca. 1.0 g of a yellow-orange oil. Purification by column chro-
matography on 15 g of silica gel (elution with 10% EtOAc/
hexanes) afforded 0.738 g (82%) of amide 5 as a pale yellow oil.
1-Piperidinyloctanamide (7). Reaction of piperidine (0.34 mL,
3.40 mmol) with 1-ethoxy-1-octyne 3a (0.525 g, 3.40 mmol) at
130 °C (230 bar) for 24 h according to the General Procedure
provided 0.703 g of brown oil. Purification by column chromato-
graphy on 15 g of silica gel (gradient elution with 10-20% EtOAc/
hexanes) afforded 0.623 g (87%) of amide 7 as a pale yellow oil
with spectral data consistent with that previously reported.27
N-Butyl-N-(diphenylmethyl)octanamide (9). Reaction of
N-butylbenzhydrylamine 8 (0.551 g, 2.30 mmol) with 1-ethoxy-
1-octyne 3a (0.350 g, 2.27 mmol) at 120 °C (280 bar) for 24 h
according to the General Procedure provided 0.860 g of brown oil.
Purification by column chromatography on 25 of silica gel (elution
with 15% EtOAc/hexanes) afforded 0.720 g (87%) of 9 as a pale
Conclusions
From the standpoint of green chemistry, synthetic reac-
tions should ideally be carried out in the absence of
any solvent. Unfortunately, most synthetic transformations
require the use of a solvent in order to ensure adequate
mixing and contact between reactants (particularly solid
compounds), to facilitate transfer of materials, and to con-
trol reaction temperature. Consequently, much effort has
been devoted in recent years to the development of “alter-
native solvents” that are superior to conventional volatile
organic solvents in terms of environmental impact and
toxicity. Supercritical carbon dioxide ranks as one of the
most attractive alternative solvents with regard to these
considerations.26
In this paper, we have demonstrated that carbon dioxide in
its supercritical state is an environmentally attractive alter-
native to conventional liquid organic solvents for the syn-
thesis of a variety of carboxylic amides. The addition of
amines to ketenes generated in situ via the retro-ene reaction
of alkynyl ethers provides amides in good yield, in many
cases with ethylene or isobutylene as the only byproducts of
the reaction. With the exception of primary, unbranched
amines, the competitive side reaction of the amines with
carbon dioxide does not interfere with the desired C-N
bond-forming reaction. Preliminary experiments aimed
at developing an intramolecular variant of this process
to afford macrolactams suggest that the application of
CO2/co-solvent mixtures may offer advantages for the syn-
thesis of large-ring compounds.
yellow oil: IR (neat) 2920, 1644, and 1454 cm-1 1H NMR
.
(500 MHz, CDCl3) δ 7.28-7.37 (m, 6H), 7.15-7.19 (m, 4H),
6.29 (s, 1H), 3.24-3.31 (m, 2H), 2.44 (t, J = 7.5 Hz, 2H), 2.39
(t, minor rotamer), 1.73-1.77 (m, 2H), 1.65-1.68 (m, minor
rotamer), 1.25-1.36 (m, 9H), 0.81-0.99 (m, 6H), and 0.59-0.66
(m, 3H). 13C NMR (100 MHz, CDCl3): major rotamer δ 173.6,
140.1, 129.1, 128.4, 127.4, 60.8, 45.9, 33.7, 31.9, 31.8, 29.7, 29.5,
25.9, 22.8, 20.3, 14.2, and 13.5; minor rotamer δ139.7, 128.9, 128.6,
127.9, 64.9, 45.0, 33.9, 32.0, 31.7, 30.2, 29.6, 25.5, 20.4, and 13.6).
HRMS-ESI (m/z): [M þ H]þ calcd for C25H35NO 366.2791; found
366.2796.
N-(Diphenylmethyl)octanamide (11). Reaction of benzhydryl-
amine (0.585 g, 3.20 mmol) with 1-ethoxy-1-octyne 3a (0.493 g,
3.20 mmol) at 130 °C (284 bar) for 24 h according to the General
Procedure provided ca. 1.0 g of an orange solid. Purification by
two trituration cycles with CH2Cl2/hexanes afforded 0.786 g
(79%) of amide 11 as an off-white solid: mp 104-105 °C.
IR (thin film): 2922, 1637, and 1544 cm-1 1H NMR (400
.
MHz, CDCl3): δ 7.22-7.36 (m, 10H), 6.27 (app d, J = 7.9
Hz, 1H), 5.98 (app d, J = 7.2 Hz, 1H), 2.27 (t, J = 7.6 Hz, 2H),
1.65-1.70 (m, 2H), 1.27-1.30 (m, 8H), and 0.87-0.89 (m, 3H).
13C NMR (100 MHz, CDCl3): δ 172.4, 141.8, 128.8, 127.6, 56.9,
37.0, 31.9, 29.4, 29.2, 25.9, 22.8, and 14.3. HRMS-ESI (m/z): [M
þ Na]þ calcd for C21H27NO 332.1985; found 332.1976.
Reaction of benzhydrylamine (0.623 g, 3.40 mmol, 1.0 equiv)
with 1-tert-butoxy-1-octyne (3b) (0.620 g, 3.40 mmol, 1.0 equiv)
at 90 °C (227 bar) for 24 h according to the General Procedure
provided 1.08 g of a pale yellow solid. Purification by two
trituration cycles with CH2Cl2/hexanes afforded 0.830 g
(79%) of amide 11 as a white solid.
Experimental Section
General Procedure for the Synthesis of Amides in scCO2.
N-Benzyl-N-butyloctanamide (5). A 25-mL, stainless steel view
cell reactor (see Supporting Information for details on the
reactor setup) was charged with N-benzylbutylamine (4)
(0.556 g, 3.40 mmol) and 1-ethoxy-1-octyne 3a (0.525 g, 3.40
mmol). The reactor was pressurized to 50 bar with CO2, heated
to 130 °C, and then pressurized with additional CO2 to 228 bar.
The reaction mixture was stirred at 130 °C (228 bar) for 24 h and
then allowed to cool to rt. The reactor outlet valve was opened,
and the CO2 phase was vented through a bubbler containing
N-Phenyloctanamide (13). Reaction of aniline (0.289 g,
3.10 mmol) with 1-ethoxy-1-octyne 3a (0.478 g, 3.10 mmol) at
130 °C (235 bar) for 24 h according to the General Procedure
(26) Clark, J. H.; Tavener, S. J. Org. Process Res. Dev. 2007, 11, 149.
9386 J. Org. Chem. Vol. 74, No. 24, 2009
(27) Jensen, A. E.; Knochel, P. J. Org. Chem. 2002, 67, 79.

Mak et al.
JOCArticle
provided 0.833 g of a gray-yellow solid. Purification by column
chromatography on 15 g of silica (elution with 10% EtOAc/
hexanes) afforded 0.416 g (61%) of amide 13 as a pale yellow oil
with spectral data consistent with that previously reported.28
N-Cyclohexyloctanamide (15). Reaction of cyclohexylamine
(0.39 mL, 0.34 g, 3.4 mmol) with ethoxy-1-octyne 3a (0.235 g,
3.40 mmol) at 130 °C (230 bar) for 24 h according to the General
Procedure provided 0.682 g of a tan solid, which was dissolved in
CH2Cl2 and concentrated onto 1 g of silica gel. The free-flowing
powder was placed at the top of a column of 10 g of silica gel and
eluted with 10-20% EtOAc/hexanes to provide 0.564 g (74%)
of amide 15 as a white solid with spectral data consistent with
that previously reported.29
N-Benzyloctanamide (17). Reaction of benzylamine (0.332 g,
3.10 mmol) with 1-ethoxy-1-octyne 3a (0.478 g, 3.10 mmol) at
130 °C (284 bar) for 24 h according to the General Procedure
provided 0.432 g of a yellow solid. Purification by two tritura-
tion cycles with CH2Cl2/hexanes afforded 0.270 g (37%) of
amide 17 as a white solid with spectral data consistent with that
previously reported.30
N-Benzyl-N-butyl-3-(tert-butyldimethylsiloxy)pentanamide (19).
Reaction of N-benzylbutylamine (0.556 g, 3.40 mmol) with alkynyl
ether 18 (0.824 g, 3.40 mmol) at 130 °C (241 bar) for 24 h according
to the General Procedure provided 1.32 g of a dark orange oil.
Purification by column chromatography on 20 g of silica gel
(gradient elution with 5-10% EtOAc/hexanes) afforded 0.717 g
(56%) of the β-siloxy amide 19 as a pale yellow oil and 0.261 g
(31%) of the R,β-unsaturated amide 20 as a yellow oil. Amide 19:
IR (thin film) 2959, 1646, and 1463 cm-1. 1H NMR (400 MHz,
CDCl3): δ 7.24-7.38 (m, 4H), 7.18 (d, J = 7.0 Hz, 1H), 4.37-4.81
(m, 2H), 4.20-4.32 (m, 1H), 3.33-3.40 (m, 1H), 3.16-3.31 (m,
1H), 2.32-2.63 (m, 2H), 1.41-1.66 (m, 4H), 1.24-1.34 (m, 2H),
0.83-0.98 (m, 6H), 0.88 (s, 9H), 0.09 (s, 3H), 0.07 (s, minor
rotamer), and 0.04 (s, 3H). 13C NMR (100 MHz, CDCl3): major
rotamer δ 172.0, 138.2, 128.3, 127.6, 126.4, 71.4, 48.8, 46.4, 40.7,
31.1, 30.7, 26.1, 20.3, 18.3, 14.0, 9.6, -4.4, and -4.5; minor
rotamer δ 171.5, 137.5, 129.0, 128.6, 127.4, 51.5, 47.4, 40.4, 30.0,
20.5, 14.1, and 9.7. HRMS-ESI (m/z): [M þ Na]þ calcd for
C22H39NO2Si 400.2642; found 400.2626. Amide 20: IR (neat)
2962, 1660, 1615, and 1425 cm-1. 1H NMR (400 MHz, CDCl3)
δ 7.25-7.42 (m, 4H), 7.19 (d, J = 7.4 Hz, 1H), 7.03 (app ddd, J =
6.4, 13.3, 21.6 Hz, 1H), 6.27 (d, J = 15.0 Hz, 1H), 6.18 (d, minor
rotamer), 4.66 (s, 2H), 4.60 (s, minor rotamer), 3.40 (t, minor
rotamer), 3.24 (t, J = 7.3 Hz, 2H), 2.27 (app quint, J = 7.0 Hz,
2H), 2.19 (app quint, minor rotamer), 1.52-1.59 (m, 2H),
1.27-1.34 (m, 2H), 1.10 (t, J = 7.4 Hz, 3H), 1.01 (t, minor
rotamer), and 0.85-0.94 (m, 3H). 13C NMR (100 MHz, CDCl3):
major rotamer δ 167.0, 148.6, 138.2, 128.7, 128.2, 126.6, 119.4,
49.1, 47.1, 31.3, 25.8, 20.2, 14.0, and 12.8; minor rotamer δ 167.4,
137.6, 129.0, 127.6, 127.4, 119.7, 51.2, 46.5, 29.9, 26.1, 20.5, and
14.1. HRMS-ESI (m/z):[M þ Na]þ calcd for C16H23NO 268.1672;
found 268.1669.
(E)-4,9-Dimethyl-1-(piperidin-1-yl)deca-4,8-dien-1-one (23).
Reaction of piperidine (0.264 g, 3.10 mmol) with alkynyl ether
22 (0.640 g, 3.10 mmol) at 130 °C (277 bar) for 24 h according to
the General Procedure provided 0.757 g of a brown oil. Pur-
ification by column chromatography on 18 g of silica gel (elution
with 20% EtOAc/hexanes) afforded 0.552 g (68%) of amide 23
1
as a yellow oil: IR (thin film) 2933, 1646, and 1436 cm-1. H
NMR (400 MHz, CDCl3) δ 5.14 (bs, 1H), 5.08 (t, J = 6.0 Hz,
1H), 3.55 (t, J = 5.5 Hz, 2H), 3.39 (t, J = 5.4 Hz, 2H), 2.33 (app
s, 4H), 2.03-2.08 (m, 2H), 1.95-1.99 (m, 2H), 1.67 (s, 3H),
1.62 (s, 3H), 1.59 (s, 3H), and 1.49 (m, 6H). 13C NMR (100 MHz,
CDCl3) δ 171.2, 136.3, 131.5, 124.4, 123.2, 46.8, 42.7, 39.9, 33.6,
26.8, 26.9, 25.9, 25.7, 24.7, 24.2, 17.8, and 16.2. HRMS-ESI (m/z):
[M þ Na]þ calcd for C17H29NO 286.2141; found 286.2152.
5-Oxo-5-(piperidin-1-yl)pentan-3-ylcarbamic Acid Ethyl Ester
(25). Reaction of piperidine (0.34 mL, 0.29 g, 3.4 mmol) with
alkynyl ether 24 (0.677 g, 3.40 mmol, 1.0 equiv) at 130 °C
(242 bar) for 24 h according to the General Procedure provided
0.974 g of a viscous dark brown oil. Column chromatography on
40 g of acetone-deactivated silica gel (gradient elution with
20-75% EtOAc/hexanes) provided 0.576 g (66%) of amide 25
as a yellow-orange oil: IR (neat) 3314, 1718, 1628, 1533, and
1240 cm-1. 1H NMR (400 MHz, CDCl3): δ 5.53 (d, J = 7.9 Hz,
1 H), 4.06 (q, J = 6.9 Hz, 2 H), 3.74-3.83 (m, 1 H), 3.46-3.57
(m, 2 H), 3.39 (t, J = 4.7 Hz, 2 H), 2.62 (dd, J = 15.3, 4.9 Hz,
1 H), 2.47 (dd, J = 15.3, 5.7 Hz, 1 H), 1.48-1.69 (m, 8 H), 1.21 (t,
J = 7.1 Hz, 3 H), 0.92 (t, J = 7.4 Hz, 3 H). 13C NMR (100 MHz,
CDCl3): δ 168.9, 156.5, 60.6, 50.1, 42.6, 37.0, 26.6, 25.7, 24.6,
14.7, 11.0. HRMS-ESI (m/z): [M þ H]þ calcd for C13H25N2O3
257.1860, found 257.1850.
5-Oxo-5-(piperidin-1-yl)pentan-3-yl-N-methylcarbamic Acid
Ethyl Ester (27). Reaction of piperidine (0.34 mL, 0.29 g,
3.4 mmol) with alkynyl ether 26 (0.725 g, 3.40 mmol) at
130 °C (222 bar) for 24 h according to the General Procedure
provided 1.007 g of a brown oil. Column chromatography on 40
g of acetone-deactivated silica gel (gradient elution with
25-60% EtOAc/hexanes) provided 0.778 g (85%) of amide 27
as a yellow oil: IR (neat) 1695, 1640, and 1254 cm-1. 1H NMR
(400 MHz, CDCl3): δ 4.21-4.32 (m, 1 H), 4.16-4.20 (m, minor
rotamer), 4.15 (q, J = 7.0 Hz, minor rotamer), 4.09 (q, J =
7.2 Hz, 2 H), 3.35-3.62 (m, 4 H), 2.79 (s, 3 H), 2.66 (dd, J = 14.1,
7.9 Hz, 1 H), 2.57 (dd, J = 14.5, 6.5 Hz, minor rotamer), 2.46 (dd,
J = 14.2, 7.0 Hz, 1 H), 2.41 (dd, J = 14.5, 7.7 Hz, minor rotamer),
1.42-1.71 (m, 8 H), 1.23 (t, J = 7.1 Hz, 3 H), 0.87 (t, J = 7.3 Hz,
3 H), 0.85 (t, J = 7.2 Hz, minor rotamer). 13C NMR (100 MHz,
CDCl3): δ 168.9, 156.6, 61.0, 56.0, 47.0, 42.6, 37.2, 26.6, 25.4, 24.5,
14.6, 10.9; minor rotamer δ 168.7, 156.7, 61.2, 55.1, 46.9, 42.7, 37.3,
26.5, 25.0, 14.7, 10.8. HRMS-ESI (m/z): [M þ Na]þ calcd for
C14H26N2NaO3 293.1841, found 293.1843.
Acknowledgment. We thank the Cambridge-MIT Institute
for generous financial support. R.P.C. thanks the United States
Environmental Protection Agency under the Science to Achieve
Results (STAR) Graduate Fellowship Program for funding.
X.Y.M. and R.P.C. are grateful to the Martin Family Society of
Fellows for Sustainability for fellowships. J.M.R. is supported
by a National Science Foundation Graduate Fellowship.
Reaction of N-benzylbutylamine (0.555 g, 3.40 mmol) and
alkynyl ether 21 (0.920 g, 3.40 mmol) at 90 °C (218 bar) for 24 h
according to the General Procedure provided 1.35 g of a yellow
oil. Purification by column chromatography on 25 g of silica gel
(gradient elution with 5-10% EtOAc/hexanes) afforded 1.03 g
(80%) of the β-siloxy-amide 19 as a pale yellow oil and 0.027 g
(3%) of the R,β-unsaturated amide 20 as a yellow oil.
(28) Ogawa, T.; Hikasa, T.; Ikegami, T.; Ono, N.; Suzuki, H. J. Chem.
Soc., Perkin Trans. 1 1994, 3473.
(29) Lucking, U.; Tucci, F. C.; Rudkevich, D. M.; Rebek, J. J. Am. Chem.
Soc. 2000, 122, 8880.
(30) Kunishima, M.; Watanabe, Y.; Terao, K.; Tani, S. Eur. J. Org.
Chem. 2004, 27, 4535.
Supporting Information Available: Detailed description
of the reactor setup, experimental procedures and characteriza-
tion data for the preparation of all alkynyl ethers and amide
products. This material is available free of charge via the
Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 24, 2009 9387