One-Pot (1-Ethoxycarbonylcyclopropyl)-triphenylphosphonium tetrafluoroborate ring-opening and wittig reaction

Article abstract of DOI:10.1021/ol202207f

An efficient method was developed for the synthesis of 2-methylene-4- substituted ethyl butyrates via cyclopropyl opening followed by a Wittig reaction. The desired products were formed in a two-step, one-pot reaction sequence. Alternatively, the key inte

Full text of DOI:10.1021/ol202207f

ORGANIC
LETTERS
2011
Vol. 13, No. 19
5338–5341
One-Pot (1-Ethoxycarbonylcyclopropyl)-
triphenylphosphonium Tetrafluoroborate
Ring-Opening and Wittig Reaction
Seung Won Chung,* Mark S. Plummer, Laura A. McAllister, Robert M. Oliver III,
Joseph A. Abramite, Yue Shen, Jianmin Sun, Daniel P. Uccello, Joel T. Arcari,
Loren M. Price, and Justin I. Montgomery
Antibacterial Chemistry, Worldwide Research & Development, Pfizer, Inc.,
Eastern Point Road, Groton, Connecticut 06340, United States
seungwon.chung@pfizer.com
Received August 15, 2011
ABSTRACT
An efficient method was developed for the synthesis of 2-methylene-4-substituted ethyl butyrates via cyclopropyl opening followed by a Wittig
reaction. The desired products were formed in a two-step, one-pot reaction sequence. Alternatively, the key intermediate ylide 2 was isolable and
could be stored under oxygen-free conditions and subsequently utilized. A variety of nucleophiles were found to open the commercially available
cyclopropane 1. The resulting ylide reacted with aldehydes to provide E-olefinic products.
In practice, the (1-ethoxycarbonylcyclopropyl)tripheny-
lphosphonium tetrafluoroborate 1 has been widely used to
form carbocyclic and heterocyclic compounds.1 In the
reported examples, the cyclopropane opens with nucleo-
philes to provide an intermediate ylide that subsequently
undergoes an intramolecular Wittig reaction to form cyclic
systems. In this work, we report a novel application of this
reagent, a one-pot, tandem ring-opening Wittig sequence.
We adapted this methodology for our purpose by reacting
the intermediate ylide in an intermolecular fashion to
provide 2-methylene-4-substituted butyrates.
Reactions were monitored by LCÀMS, as both intermedi-
ate ylide 2 and final product 3 were easily detected in the
two-step, one-pot reaction (Figure 1).
Here, we react commercially available 1 with a number
of nucleophiles followed by the addition of unhindered
aldehydes to provide highly substituted butyrates with
good efficiency and operational ease. Various nucleophiles
have been employed to open the (1-ethoxycarbonylcyclo-
propyl)triphenylphosphonium salt to form the ylide.1aÀf
Figure 1. General one-pot reaction: (a) deprotonation, (b) open-
ing of (1-ethoxycarbonylcyclopropyl)triphenylphosphonium
tetrafluoroborate, (c) Wittig reaction.
To establish a general method, 3,4,5-trimethylpyrazole
was selected as a trial nucleophile for optimization in com-
bination with various bases (NaH, LiHMDS, K2CO3/
NaOH/TBAHS)2 and solvents (THF, toluene, DMF)
(Table 1).
(1) (a) Fuchs, P. L. J. Am. Chem. Soc. 1974, 96, 1607. (b) Dauben,
W. G.; Hart, D. J. Tetrahedron Lett. 1975, 49, 4353. (c) Flitsch, W.;
Pandl, K.; Russkamp, P. Liebigs Ann. Chem. 1983, 4, 529. (d) Flitsch,
W.; Russkamp, P. Liebigs Ann. Chem. 1983, 4, 521. (e) Barreau, M.;
Ponsinet, G. Tetrahedron Lett. 1985, 26, 5451. (f) Chatterjee, P.;
Murphy, P. J.; Pepe, R.; Shaw, M. J. Chem. Soc., Perkin Trans. 1
1994, 17, 2403. (g) Hersperger, R.; Janser, P.; Pfenninger, E.; Wuethrich,
H. J.; Miltz, W. Int. Pat. Appl. 077932, 2005.
The best yield of compound 4 was obtained when
sodium hydride in either THF or toluene was used.
r
10.1021/ol202207f
Published on Web 09/12/2011
2011 American Chemical Society

Table 1. Solvent and Base Selection for the One-Pot Synthesisa
Table 2. Electronic Effect vs Steric Hindrancea
entry solvent
base
E/Z ratio isolated yield (%)
1
2
3
3
4
5
THF
NaH
1:0
1:0
1:0
1:0
1:0
60
57
47
32
34
0
toluene NaH
DMF
THF
NaH
LiHMDS
toluene LiHMDS
DMF K2CO3/NaOH/TBAHS
a Reaction conditions: (a) base/solvent/20 °C/30 min; (b) (1-car-
bethoxycyclopropyl)triphenylphosphonium tetrafluoroborate/solvent/
80 °C/2 h; (c) benzaldehyde/solvent/80 °C/overnight.
Notably, the (1-ethoxycarbonylcyclopropyl)triphenylpho-
sphonium salt does not form the desired ylide 2 with the
sodium salt of trimethylpyrazole at room temperature, so
moderate heating isrequired. Inaddition, the reaction with
LiHMDS resulted in diminished yields of isolated product.
The K2CO3/NaOH/TBAHS2 base system did not result in
the desired ylide 2, even after overnight reaction (via
LCÀMS), and did not provide the desired product 4. This
optimization showed that either THF or toluene withNaH
provides useful yields of butyrates. Interestingly, in each
case, only a single olefin isomer was detected in crude
product NMR spectra. After isolation of the product, the
olefin geometry was shown by an NOE experiment to have
an E-configuration (4a), which is consistent with recently
published work.3
a Reaction conditions: (a) preformed ylide 2a/aldehyde/toluene
(2 mL)/80 °C/overnight, (0.1 M).
The electronic and steric influence of aldehyde substitu-
ents on the reaction outcome was examined using 4-phe-
nylpyrazole as the nucleophile (Table 2).
In order to minimize variability between experiments,
we opted to utilize the isolated ylide 2a, which was syn-
thesized on a multigram scale (Supporting Information).
Electron-donating methoxy groups on the benzaldehyde
gave poorer yields when compared to the other benzalde-
hyde analogues tested, irrespective of the position of the
methoxy group on the aryl ring. The most significant
influence was observed when the methoxy group was in
the orthoposition, suggesting sterics may alsoplaya role in
the reaction outcome (entry 5, Table 2).
Electron-withdrawing groups on the benzaldehyde,
such as chloro and nitro, provided products in excellent
yield with no detrimental influence of ortho-substitution
(entries 3 and 4). The o-methyl benzaldehyde, however, did
result in a reduced yield compared to its meta and para
counterparts (entry 2, Table 2). In all cases, the E-olefin
geometry was the only isomer observed. Interestingly, the
p-nitro analogue (entry 4, Table 2) provided the E-olefin
isomer initially, but upon standing slow olefin isomeriza-
tion was observed.
Heterocyclic and aliphatic aldehydes were next investi-
gated to further test the reaction scope using the stepwise,
one-pot method described above (Table 3). Not surpris-
ingly, formaldehyde reacted effectively with the ylide 2a to
provide desired product (Table 3, entry 1). Formaldehyde
could be introduced either as p-formaldehyde or as an
aqueous solution of formaldehyde with equal effective-
ness. The pyridyl and pyrazolyl aldehydes (entries 2 and 6
Table 3) provided products in moderate yield. Homolo-
gues of the pyridyl aldehyde (entries 3 and 4, Table 3)
provided products in moderate yield, suggesting that alkyl
aldehydes might also provide products. The steric encom-
berance of the aldehyde in entry 5, possibly along with
electronics, effectively combined to prevent the reaction.
Along with aliphatic aldehydes, steric hindrance seems to
have a major influence, since good yields were observed
(2) (a) Padwa, A.; Brodney, M. A.; Satake, K.; Straub, C. S. J. Org.
Chem. 1999, 64, 4617. (b) Stafford, J. A.; McMurry, J. E. Tetrahedron
Lett. 1988, 29, 2531.
(3) Appel, R.; Loos, R.; Mayr, H. J. Am. Chem. Soc. 2009, 131, 704.
Org. Lett., Vol. 13, No. 19, 2011
5339

Table 3. Phenylpyrazole Nucleophile with Variety of Aldehydesa
Table 4. Various Nucleophiles with Benzaldehydea,b
a Reaction conditions: (a) NaH/THF or toluene/0 °C/30 min; (b)
(1-carbethoxycyclopropyl)triphenylphosphonium tetrafluoroborate/
THF or toluene/80 °C/2 h; (c) benzaldehyde/THF or toluene/80 °C/
overnight. b See Scheme 2 for details.
2, 4, and 6, Table 4) were successfully employed. They
ranged from morpholine and azide to a benzenesulfona-
mide anion and clearly demonstrated the diversity of
reactants capable of providing substituted butyrate pro-
ducts. A single oxygen nucleophile was employed in the
form of phenol, and it provided a moderate yield of
product (entry 3, Table 4). Finally, two different sulfur
nucleophiles were tested, a sulfinic acid and a thiol (entries
5 and 7, Table 4), and both provided the desired product
in useful quantities. The benzenesulfonamide in entry 6
(Table 4) was successful but was also found to react further
to give a cyclized lactam product as a 3:1 ratio (linear to
lactam) by LCÀMS (Scheme 1).6 However, the de-
scribed linear product could be isolated by careful
chromatography.
a Reaction conditions: (a) NaH/THF/20 °C/30 min; (b) (1-car-
bethoxycyclopropyl)triphenylphosphonium tetrafluoroborate/THF/
80 °C/2 h; (c) aldehyde/80 °C/overnight.
when less encumbered aldehydes were employed (entries 7,
11, and 12, Table 3). Little or no product forms with more
bulky aliphatic side chains (entries 8À10, Table 3), and
heating to 160 °C in DMF was ineffective at increasing
product formation. Likewise, we had no success with
ketones as Wittig partners.4 From a practical standpoint,
product separation from starting aliphatic aldehyde and
triphenylphosphine oxide was challenging when 4-phenyl-
pyrazole was utilized as the nucleophile, resulting in re-
duced isolated yields for entries 2, 6, 10, and 11, even when
scavenging resins were used to remove triphenylphosphine
oxide.5
Scheme 1. Linear vs Cyclized Product of Sulfonamide
Finally, we examined a variety of nucleophiles for open-
ing the cyclopropane ring in 1 (Table 4). For each of the
nucleophiles, the two-step, one-pot reaction gave approxi-
mately 50À70% isolated yield of desired product. Four
different classes of nitrogen-based nucleophiles (entries 1,
Unlike the other examples in Table 4, the sulfonamide
has two potentially acidic protons and provided two
products from the reaction in THF. When the reaction
was carried out with less than 1.0 equiv of sodium hydride
(4) (a) Fitjer, L.; Quabeck, U. Synth. Commun. 1985, 15, 855. (b)
Olah, G. A.; Krishnamurthy, V. V. J. Am. Chem. Soc. 1982, 104, 3987.
(5) Lipshutz, B. H.; Blomgren, P. A. Org. Lett. 2001, 3, 1869.
5340
Org. Lett., Vol. 13, No. 19, 2011

in THF, only the linear product was formed. However,
when greater than 1.0 equiv of sodium hydride (i.e., 1.2 or
2.0 equiv) was used, both linear and cyclized product
formed (Scheme 1). Interestingly, when the solvent was
changed to DMF, only the linear product was formed
regardless of the amount of base utilized. When the
isolated linear product was treated with NaH (0.1 equiv)
in THF, the cyclized product formed while the same
reaction in DMF did not form cyclized product.6a
nucleophile. Benzylamine was employed asthe nucleophile
in our typical one-pot reaction (Scheme 2). Benzylamine
was heated directly with cyclopropane 1 in DMF at 80 °C
to give 8b in 75% isolated yield.6c,d We were unable to
detect any of the 5-endo-ring product (8a) by NMR or
LCÀMS (details in the Supporting Information).6a
Overall, we were able to synthesize 2-methylene-4-sub-
stituted ethyl butyrates via a one-pot ring-opening of
(1-ethoxycarbonylcyclopropyl)triphenylphosphonium tetra-
fluoroborate followed by a Wittig reaction. Products were
isolated in moderate yield and good purity. We have
described general conditions where a variety of nucleo-
philes generated with THF/NaH provided a product.
Furthermore, when appropriate nucleophiles were utilized
a clean cyclization to form a pyrrolidinone ring was
possible.
Scheme 2. Application of Reaction
Acknowledgment. We thank our colleagues in the Pfizer
Antibacterial Chemistry Department for their support of
this work.
Prompted by the in situ cyclization observed with the
benzenesulfonamide to give the lactam, we wanted to
explore the generality of this transformation with another
Supporting Information Available. Experimental pro-
cedures, characterization data for all new compounds,
descriptions of stereochemical assignments, and copies of
1H and 13C NMR spectra for all new compounds reported
in the text. This material is available free of charge via the
Internet at http://pubs.acs.org.
(6) (a) Ichikawa, J.; Lapointe, G.; Iwai, Y. Chem. Commun.
(Cambridge, U.K.) 2007, 26, 2698. (b) Vicario, J. L.; Badia, D.; Carrillo,
L. J. Org. Chem. 2001, 66, 5801. This molecule is similar to 8b; see:(c)
Iwamoto, K.; Kojima, M.; Chatani, N.; Murai, S. J. Org. Chem. 2001,
66, 169. (d) Li, S.; Yuan, W.; Ma, S. Angew. Chem., Int. Ed. 2011, 50,
2578.
Org. Lett., Vol. 13, No. 19, 2011
5341