A new mild method for the C-acylation of ketone enolates. a convenient synthesis of β-keto-esters,-thionoesters, and-thioesters

Article abstract of DOI:10.1021/ol303324a

A new method for ketone enolate C-acylation is described which utilizes alkyl pentafluorophenylcarbonates, thiocarbonates, and thionocarbonates as the reactive acylating agents, and MgBr₂·Et₂O, DMAP, and i-Pr₂NEt as the reagents for enolization. A wide range of ketones have been observed to undergo clean C-acylation via this protocol.

Full text of DOI:10.1021/ol303324a

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
A New Mild Method for the C‑Acylation of
Ketone Enolates. A Convenient Synthesis
of β‑Keto-Esters, -Thionoesters,
and -Thioesters
Karl J. Hale,* Milosz Grabski, and Jakub T. Flasz
The School of Chemistry and Chemical Engineering, and the CCRCB, Queen’s
University Belfast, Stranmillis Road, Belfast BT9 5AG, Northern Ireland, U.K.
k.j.hale@qub.ac.uk
Received December 4, 2012
ABSTRACT
A new method for ketone enolate C-acylation is described which utilizes alkyl pentafluorophenylcarbonates, thiocarbonates, and
thionocarbonates as the reactive acylating agents, and MgBr2 Et2O, DMAP, and i-Pr2NEt as the reagents for enolization. A wide range of
ketones have been observed to undergo clean C-acylation via this protocol.
3
A frequently called-upon reaction in modern-day or-
ganic synthesis is the C-acylation of a ketone enolate to
obtain the corresponding β-keto ester which, despite the
variety of methods that exist for effecting this trans-
formation,1ꢀ4 can sometimes present difficulties in certain
situations. One case in point can be found in our recent
synthetic route to the antitumor natural product (ꢀ)-
echinosporin,5 where there was a requirement to convert
ketone 1 into the β-keto ester enol 2, in high yield, to
satisfactorily progress the synthesis. A careful application
of all pre-existing ketone enolate C-acylation methods to 1
produced unsatisfactory outcomes, with respect to either
reaction yield or the formation of undesired byproducts.
Our first success in this direction came when LiHMDS
was used to enolize 1 in THF at ꢀ78 °C, and methyl
cyano-formate 6 (Mander’s reagent)4 was employed
for methoxycarbonylation. These conditions afforded the
desired β-keto ester enol 2 in 46% yield (Table 1, entry 2),
alongside a multitude of other undesired byproducts. Presum-
ably the latter arose from the starting ketone and the initially
formed products, both reacting with the cyanide ion that was
being liberated as the reaction progressed, an outcome that has
previously been documented6 for certain ketones when enolate
C-acylation is attempted with the Mander reagent and base.
Given the poor outcome of this process with 1, several
alternate C-alkoxycarbonylating agents were evaluated,
(4) For examples of the use of Mander’s reagent [MeOC(O)CN] for
enolate C-acylation, see: (a) Mander, L.; Sethi, S. P. Tetrahedron Lett.
1983, 24, 5425. (b) Hutt, O. E.; Mander, L. N. J. Org. Chem.. 2007, 72,
10130. (c) For a recent example of the use of LDA and MeOC(O)CN in
THF/DMPU for a ketone kinetic enolization and C-acylation, see:
Henderson, J. A.; Phillips, A. J. Angew. Chem., Int. Ed. 2008, 47, 8499.
(d) For the use of LDA/THF/HMPA and Mander’s reagent for ketone
C-acylation, see: Beshore, D. C.; Smith, A. B., III. J. Am. Chem. Soc.
2007, 129, 4148. (e) For another successful use of LDA/Mander’s
reagent/THF at ꢀ78 °C, see: Clive, D. L. J.; Sannigrahi, M.; Hisaindee,
S. J. Org. Chem. 2001, 66, 954. (f) For the concurrent occurrence of
competing O-acylation during an attempted enolate C-acylation with
MeOC(O)CN, see: Mander, L. N.; Thomson, R. J. Org. Chem. 2005, 70,
1651.
(1) For a previous report on the use of LiTMP and an alkyl
chloroformate for the C-acylation of ketones, see: Olofson, R. A.;
Cuomo, J.; Bauman, B. A. J. Org. Chem. 1978, 43, 2073. This paper
also records that less hindered amide bases such as LDA liberate amines
that can sometimes compete more favorably than the enolate itself for
the chloroformate.
(2) For examples of the use of acid chlorides to C-acylate Li-enolates,
see: Wiles, C.; Watts, P.; Haswell, S.; Pombo-Villar, E. Tetrahedron Lett.
2002, 43, 2945.
(5) Flasz, J. T.; Hale, K. J. Org. Lett. 2012, 14, 3024.
(3) For examples of the use of acyl imidazolides to C-acylate (or
O-acylate) ketone enolates, see: Trost, B. M.; Xu, J. J. Org. Chem. 2007,
(6) For the recent documentation of cyanohydrin byproducts being
formed as major reaction products during a KHMDS-mediated ketone
enolization and C-acylation with MeOC(O)CN, see: Kazimierski, A.;
72, 9372. This paper also reports that the addition of BF3 Et2O to such
reactions reverses the selectivity in favor of the exclusive formation of
enol carbonates in high yield.
3
_
Kazuza, Z.; Chmielewski, M. ARKIVOC 2004, 3, 213.
r
10.1021/ol303324a
XXXX American Chemical Society

and 3 was cleanly produced in 68% overall yield (Table 1,
entry 5). Despite the greatly improved outcome, this was a
level of performance that we wished to further enhance,
and it led us to examine whether a prolonged reaction time
might be beneficial. Extending the reaction time to 96 h now
produced the desired enol 2 exclusively in 92ꢀ95% yield.5
The great success of these refined kinetic enolization/
C-methoxycarbonylation conditions was striking, for they
left the adjacent R-keto stereocenter chemically undis-
turbed, and they produced an excellent yield of the desired
product 2 from a system that had consistently failed to give
good results when it was submitted to all pre-existing
enolate C-acylation methods. Recognizing the great po-
tential of this new transformation, we tested its synthetic
applicability on other substrates and report our findings.
Our early effort focused upon devising a good way of
preparing the various pentafluorophenyl alkyl carbonates,
thionocarbonates, and phenylthiocarbonates that would
be needed for our study (Scheme 1). For methyl penta-
fluorophenyl carbonate 8, allyl pentafluorophenyl carbo-
nate 9, and bromoethyl pentafluorophenyl carbonate 10,
these were all readily prepared by treating the correspond-
ing commercially available chloroformates with penta-
fluorophenol (1 equiv) and N-methylmorpholine (NMM)
in CH2Cl2 at 0 °C, and in the case of 10 and benzyl
pentafluorophenylcarbonate 11, a catalytic amount of
DMAP was also added. Via this approach, high yields of
the desired electrophiles were obtained (89ꢀ98%). For the
preparation of phenyl pentafluorophenyl thionocarbonate
13, O-phenyl chlorothionoformate 12 served as the active
electrophile for the pentafluorophenol, and DMAP was
once again added. For thiophenyl pentafluorophenyl car-
bonate 14, triphosgene was the initial electrophile, it being
sequentially reacted with pentafluorophenol and thiophe-
nol in CH2Cl2, in the presenceof NMM. Inthe caseof 13, it
was obtained in 95% yield as a crystalline solid, while 14
was isolated in 64% yield as an oil. All of these compounds
Table 1. Attempted Kinetic Enolization and C-Acylation of 1
including methoxycarbonyl imidazolide 7 and methyl
pentafluorophenyl carbonate 8. Both were deemed suit-
able because of their ability to liberate a non-nucleophilic
counteranion following enolate C-acylation. KHMDS
was the base initially selected for study, and with 7 as the
electrophile, no apparent reaction took place (Table 1,
entry 3). However, when 8 was deployed, the desired product
3 was formed cleanly in low yield (15%) (Table 1, entry 4).
The success of this reaction prompted us to seek other
alternative methods for kinetically enolizing 1 and sub-
sequently trapping with 8.
Inthisconnection, webecameinterested inseveral recent
reports from the Coltart laboratory,7 where MgBr2 OEt2
Scheme 1. Methods Used to Prepare 8ꢀ11, 13, and 14
3
and Hunig’s base were used to generate bromomagnesium
enolates from methyl ketones and thioesters. These were
then employed for Claisen condensations with alkyl penta-
fluorophenyl esters and N-acylbenzotriazoles to obtain
1,3-diketones and β-keto-thioesters in good yield. Notable
for their absence from these studies, however, were elec-
trophiles such as alkyl pentafluorophenyl carbonates,
thiophenyl pentafluorophenyl carbonates, and aryl penta-
fluorophenyl thionocarbonates, which did not auger well
for the problem we had at hand.
This observation notwithstanding, when we subjected 1
to Coltart’s “soft” enolization conditions in CH2Cl2, at rt,
in the presence of 8, we observed that no reaction was
occurring after 10 min; we therefore added DMAP, and
after stirring for a further 70 min, a separable mixture of 2
(7) (a) Lim, D.; Fang, F.; Zhou, G.; Coltart, D. Org. Lett. 2007, 9,
4139. (b) Zhou, G.; Lim, D.; Coltart, D. M. Synthesis 2009, 56.
B
Org. Lett., Vol. XX, No. XX, XXXX

showed high shelf stability, with 8ꢀ11 lasting perfectly well
for more than a year at rt without special precautions to
guard against decomposition. The S-containing carbo-
nates 13 and 14 were also stable at rt, but they did
decompose after prolonged storage in light.
Table 2. Outcomes of the MgBr2 Et2O/i-Pr2NEt/DMAP
Mediated C-Acylations of Various Ketones with 8 in CH2Cl2
3
We have screened a wide range of ketone substrates in
thisC-acylation with 8 (1.0ꢀ1.4 equiv) (Table2). Although
a substantial number of these reactions were successfully
performed at rt in CH2Cl2 with MgBr2 Et2O (2.5 equiv),
3
i-Pr2NEt (3 equiv), and DMAP (0.2 equiv) over 27ꢀ65 h
(entries 1, 2, 5, and 12), in some instances, it was more
advantageous to heat the reactions at reflux to shorten the
overall reaction time and boost the product yield. In other
cases, it was beneficial to increase the overall quantities of
reagent employed at rt. Generally yields were good which-
ever protocol was selected.
The rt process could be conducted on methyl ketone 15
without any difficulty (entry 1). Here, a 2.15:1 ketoꢀenol
mixture was obtained in 66% yield, while, for 2-acetylthio-
phene 20, a 23:1 ketoꢀenol mixture of C-acylated products
21 and 22 resulted in a 89% yield. In both systems, the
reactions proceeded cleanly. Propiophenone 18 also reacted
reasonably well, affording β-keto ester 19 in 61% yield.
1-Indanones were particularly good substrates for this
enolate C-acylation, with ketones 26, 29, and 32 all giving rise
to high yields of product, and 2-alkyl-substituted 1-indanones
working especially well. The latter behavior contrasted sharp-
ly with 2-methyl-1-tetralone 39 (entry 10), where the R-
substituent proved detrimental, due the development of
allylic strain in the enolate. Nonetheless, 39 was still a viable
substrate, delivering a workable 43ꢀ53% yield of 40.
We also investigated the feasibility of using R-unsubsti-
tuted enones as substrates and carvone 47 worked very well in
this regard (Table 2, entry 13, 68ꢀ94% yield of 48). Yet, we
must emphasize that not all complex enone substrates have
worked as well as carvone. For example, (þ)-4-cholesten-3-
one 49 produced a highly complex mixture of products when
enolate C-acylation was attempted with 8 at rt (Table 2, entry
14). The monosaccharide-derived ketone 508 also underwent
serious decomposition with 8 at rt (Table 2, entry 15).
Another key reaction parameter that we have carefully
investigated is how steric crowding around the R- and
β-carbons of the ketone affects the performance of these
bromomagnesium enolate acylation reactions with 8
(Scheme 2). In the case of (ꢀ)-menthone, the stereochemi-
cal integrity of the R-isopropyl stereocenter was also
compromised to some degree. Nonetheless, very useful
and workable results were still obtained with cyclopenta-
none systems, as evidenced by the formation of 57 in 67%
yield from (ꢀ)-R-thujone 56. Here the inclusion of catalytic
DMAP was found to be essential for reaction progression,
with its presence bolstering the yield of 57 from 17% to 67%
in a greatly reduced reaction time. Notably, this CO2Me was
installed next to a quaternary carbon.
bromides at other positions (Table 3, entry 4), all sur-
vived the application of our new enolate C-acylation
method. The capacity toefficiently install allyloxycarbonyl
and benzyloxycarbonyl functionality R- to a ketone is a
particularly useful and notable aspect of this invention
(Table 3, entries 1ꢀ3), as is the ability to construct
Importantly, 4-oxo-thiochromanones (Table 2, entries
11 and 12), and β-keto ester products with labile alkyl
(8) Manaviazar, S.; Frigerio, M.; Bhatia, G. S.; Hummersone, M. G.;
Aliev, A. E.; Hale, K. J. Org. Lett. 2006, 8, 4477.
Org. Lett., Vol. XX, No. XX, XXXX
C

Table 3. MgBr2 Et2O/i-Pr2NEt/DMAP Mediated Kinetic
C-Acylation with 9, 10, 11, 13, and 14 in CH2Cl2
Scheme 2. Some Examples of the MgBr2 Et2O/i-Pr2NEt/
DMAP/8 Mediated Kinetic C-Acylation of Hindered Ketones
3
3
β-keto-thioesters (Table 3, entry 5) and rarely encountered
β-keto-thionoesters9 (entry 6) which, because of their
thermal instability, must frequently be prepared and pur-
ified at rt. The fact that thioesters undergo Pd(0)-catalyzed
reduction to aldehydes with Et3SiH makes this a potentially
useful new way of introducing a masked aldehyde group-
ing into an organic molecule.10 Thioesters also undergo
Pd(0)-mediated LiebeskindꢀSrogl cross-couplings with
boronic acids.11 Nicolaou et al. have additionally found
multiple new synthetic uses for thionoesters in their
brevetoxin work.12 β-Keto-allyl esters such as 61 are also
exceedingly useful substrates for the StoltzꢀTrost
Pd(0)-catalyzed asymmetric decarboxylative C-allylation
reaction.13
Although, it is possible that such enolate C-acylation
reactions proceed by rearrangement of an initially for-
med enol carbonate, they might equally well be occurring
directly, through chelated transition states involving the
bromomagnesium enolate and the protonated alkyl penta-
fluorophenyl carbonate or an alkoxycarbonyl-pyridinium
ion14 in what, essentially, would be a temporarily tethered
intramolecular cyclization.15 Possibly, all these processes
are occurring concurrently to different degrees.
In conclusion, while we cannot claim universality of
scope for our new ketone enolate C-acylation method,
we do believe that it will prove useful in some situations
where long-established enolate C-acylation technology
has already faltered, as it did for us with 1.5 We thus
recommend this convenient new rt C-acylation proce-
dure, as both a method of primary choice and one of
final resort.
(9) Sato, M.; Ban, H.; Uehara, F.; Keneko, C. Chem. Commun. 1996,775.
(10) Et3SiH/Pd(0)-thioester reduction: Fukuyama, T.; Lin, S.-C.; Li,
L. J. Am. Chem. Soc. 1990, 112, 7050.
(11) Liebeskind, L. S.; Srogl, J. J. Am. Chem. Soc. 2000, 122, 11260.
(12) (a) Nicolaou, K. C.; McGarry, D. G.; Somers, P. K.; Veale,
C. A.; Furst, G. T. J. Am. Chem. Soc. 1987, 109, 2504. (b) Nicolaou,
K. C.; Satao, M.; Theodorakis, E. A.; Miller, N. D. J. Chem. Soc., Chem.
Commun. 1995, 1583.
(13) (a) Mohr, J. T.; Behenna, D. C.; Harned, A. M.; Stoltz, B. M.
Angew. Chem., Int. Ed. 2005, 44, 6924. (b) Trost, B. M.; Xu, J. J. Am.
Chem. Soc. 2005, 127, 17180. (c) Franckevicius, V.; Cuthbertson, J. D.;
Pickworth, M.; Pugh, D. S.; Taylor, R. J. K. Org. Lett. 2011, 13, 4264.
(14) (a) For DMAP mechanisms for O-acylation, see: Xu, S.; Held, I.;
Kempf, B.; Mayr, H.; Steglich, W.; Zipse, H. Chem.;Eur. J. 2005, 11,
4751. (b) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M.
Bull. Chem. Soc. Jpn. 1979, 52, 1989.
(15) Such a chelate could operate in the following way:
Acknowledgment. We thank QUB and the ACS for
financial support and the EPSRC National Mass Spectro-
metry Service for HRMS assistance.
Supporting Information Available. Full experimental
procedures and spectra for all compounds prepared. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX