A Practical synthesis of 3-Acyl cyclobutanones by [2 〈 2] annulation. Mechanism and utility of the zn(II)-Catalyzed condensation of r-Chloroenamines with electron-deficient alkenes

Article abstract of DOI:10.1021/jo102257k

New conditions for the conversion of simple tertiary amides to R-chloroenamines and their use in Zn(II)-catalyzed cycloaddition reactions with commercial α,β-unsaturated carbonyl compounds allows rapid, regiocontrolled access to 3-acyl cyclobutanones. Rea

Full text of DOI:10.1021/jo102257k

ARTICLE
pubs.acs.org/joc
A Practical Synthesis of 3-Acyl Cyclobutanones by [2 þ 2] Annulation.
Mechanism and Utility of the Zn(II)-Catalyzed Condensation of
r-Chloroenamines with Electron-Deficient Alkenes
Jeannette M. O’Brien and Jason S. Kingsbury*
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467, United States
S Supporting Information
b
ABSTRACT:
New conditions for the conversion of simple tertiary amides to R-chloroenamines and their use in Zn(II)-catalyzed cycloaddition
reactions with commercial R,β-unsaturated carbonyl compounds allows rapid, regiocontrolled access to 3-acyl cyclobutanones.
Reactions take place at ambient temperature without solvent, giving strained [2 þ 2] adducts with all-carbon-substituted quaternary
carbon atoms. Ab initio calculations of the putative keteniminium intermediate and studies with styrenyl olefins suggest a dual role
for Zn(OTf)2 during catalysis.
’ INTRODUCTION
carbofunctionalization of an alkene. Although assertions have
been made in support of concerted [π2s þ π2a] pericyclic
mechanisms7a,9 involving the CdC double bond and low-lying
unoccupied π*CdN (or π*CdO) orbitals, the literature records
ample evidence in favor of asynchronous10 or even stepwise
bond-forming events.7a,f For instance, keteniminium ions show
marked nonstereospecificity in intermolecular cycloaddition
with acyclic cis-substituted olefins.7e As shown in Scheme 2, a
stepwise mechanistic model also adequately explains the high
levels of asymmetric induction attainable in reaction of small
cycloalkenes (cyclopentene) or acyclic trans-alkenes with
the chiral, proline-derived keteniminium 3, namely: (1) least-
hindered approach of the reactants to facilitate initial bonding at
C1; (2) sterically preferred bond rotations (as indicated) in
carbonium ion 4 to establish the nucleophilic enamine system
prior to formation of the second C-C bond (f5); and finally
(3) hydrolytic removal of the pyrrolidine auxiliary to reveal
enantiomerically pure bicycloheptanone 6, whose absolute con-
figuration at the ring juncture was shown to be (1R,5S).7d
The electrophiles have been accessed by stoichiometric Zn-
promoted dechlorination of R-chloroenamines (see 2, prepared
by chlorination-deprotonation of 1)7b,e or by direct dehydra-
tion of the tertiary amides 1 with Tf2O and 2,4,6-collidine.7c,f
In 1981, it was reported that keteniminium cations cycloadd to
electron-poor double bonds, as well.11 As an example, R-chloro-
enamine 7 reacts with isopropenyl methyl ketone upon treat-
ment with 1.2 equiv of ZnCl2, presumably through the inter-
mediacy of keteniminium salt 8 (Scheme 3).11 Product is isolated
Complexity and diversity among bioactive natural products
with a four-carbon ring continue to stimulate the development of
improved methods1 for [2 þ 2] annulation.2 Providencin3 and
pauferrol A,4 for instance, are both unmet synthetic challenges5
and important lead structures for the treatment of human
cancers. In this article, we describe a catalytic, solvent-free
synthesis of 3-acyl cyclobutanones6 (A) from simple R,β-un-
saturated ketones, carboxylic esters, and amides (B, Scheme 1) as
substrates. Our approach is based on an underutilized variant of
the keteniminium-olefin cycloaddition developed by Ghosez
and co-workers.7 Not only do the products A represent versatile
intermediates for chemical synthesis, but the presence of a 1,4-
dicarbonyl points to an unusual pattern of reactivity. As opening
examples of catalysis for this transformation, it is hoped that this
work can inspire modern enantioselective entries to substituted
cyclobutanones since prior strategies have relied exclusively on
pyrrolidine-based chiral auxiliaries.7d,e,g
The synthesis of cyclobutanones from aldo- and keto-keteni-
minium salts is a valuable complement to the ketene-olefin
[2 þ 2] cycloaddition reaction.8 Each process generates four-
carbon ring strain and formally accomplishes the vicinal
Received: November 22, 2010
Published: February 23, 2011
r
2011 American Chemical Society
1662
dx.doi.org/10.1021/jo102257k J. Org. Chem. 2011, 76, 1662–1672
|

The Journal of Organic Chemistry
ARTICLE
Scheme 1
Table 1. Optimization of the Synthesis of 3-Acetyl-2,2-di-
methylcyclobutanone from MVK and 1-Chloro-1-N,N,2-
trimethylpropenylamine
entry catalyst loading solvent temp (°C) time conv (%)a yield (%)b
1
2
3
4
5
6
7
8
ZnCl2
AgOTf 1 equiv CH2Cl2
ZnCl2 1 equiv neat
1 equiv CH2Cl2
50
50
50
50
50
23
35
35
24 h
24 h
24 h
24 h
24 h
6 days
6 h
97
53
95
89
51
85
80
89
81
nd
80
nd
nd
nd
72
86
Zn(OTf)2 20 mol % neat
Zn(OTf)2 10 mol % neat
Zn(OTf)2 20 mol % neat
Zn(OTf)2 20 mol % neat
Zn(OTf)2 25 mol % neat
Scheme 2. Ghosez’s Enantioselective Intermolecular 4-C
Annulation through Cation-π Reaction in the Intermediate
Chiral Keteniminium 3
6 h
a Determined by 1H NMR analysis. b After chromatography on silica gel.
of MVK, 7, and ZnCl2 (ratio 1.2:1:1) at 50 °C under nitrogen
gives ketone 11 as a single regioisomer after a brief hydrolytic
workup with H2O.11 Several Ag(I) salts,7a,14 including AgOTf, are
effective promoters too, but conversion is lower under identical
conditions (entry 2). After seeing the benefit of higher concen-
trations, a run was performed without solvent. As shown in entry
3, high efficiency is maintained. We then surveyed a number15 of
other Lewisacids aiming to improve solubility (over ZnCl2) in the
oily reactant mixture. Zn(OTf)2 proved superior, and an early
attempt at catalysis was favorable; smooth conversion to 11 is
observed as shown in entry 4. Attempts to lower either the catalyst
loading or temperature led, respectively, to lower conversion
(51%, entry 5) or an exceedingly long reaction (6 days, entry 6).
However, we were able to shorten reaction time and enhance
efficiency by sonicating16 the neat mixture at 35 °C (entry 7). As a
final refinement, a slight increase in the catalyst loading (25 mol
%, entry 8) provides 11 in 86% yield at 89% conversion. On both
milligram and gram scale, reaction is clean and regioselective. The
only detectable impurity under the conditions of entry 8 is 8-
10% (isolated yield) of N,N-dimethylisobutyramide derived from
the hydrolysis of unreacted 7 during aqueous workup.
Scheme 3. Stoichiometric [2 þ 2] Annulation with an R,β-
Enone
Having arrived at practical conditions for efficient [2 þ 2]
cycloaddition, the scope with regard to the alkene reaction
component was evaluated, with results summarized in Table 2.
Thermal bicyclization is reported11 to give a 4:1 ratio of 9 and its
regioisomer 10 in low yield, yet Zn-catalyzed reaction of 7 and
2-propenyl methyl ketone provides only 9 (73% distilled yield at
87% conv) with sonication in the absence of solvent (entry 2).
Both methyl acrylate and methyl methacrylate are acceptable17
donor alkenes, affording cyclobutanones 12 and 13 in >60%
isolated yields (entries 3 and 4). A minor 2-acyl isomer was finally
observed for N,N-dimethylmethacrylamide as substrate (80%
yield, 3:1 ratio favoring 15, entry 6) in an event that was of
comparable efficiency to that involving the unsubstituted acrylic
amide (83% of 14, entry 5). The Weinreb amides of acrylic and
methacrylic acid are also suitable. Cyclobutanone 16 was isolated
in 64% distilled yield at 85% conversion as one regioisomer
(entry 7), and its quaternary variant 17 was recovered in 57%
yield (87% conv), again with perfect regiocontrol (entry 8). At
this time, the method is limited to reactants lacking substitution
at the β position; trans-methyl crotonate and cyclohexenone fail
to react. R,β-Unsaturated nitriles were also tested but did not
result in a productive transformation. It deserves note that in the
one case in which a minor product is observed (entry 6), the
regioisomers are separable by silica gel chromatography, yet
distillation remains an attractive purification method.
in 34% yield as an 80:20 mixture of regioisomers 9 and 10. To the
best of our knowledge, no other reports in this area have followed
in support of an alternative reaction pathway. We found the issue
of mechanism to be quite intriguing since an analogous stepwise
explanation invokes a carbocation that is either primary (for
minor regioisomer 10) or R to a carbonyl group (for 9). Because
of the simplicity of this approach to cyclobutane rings with a
versatile 1,4-dicarbonyl function and the chance that it might be
of value in the stereocontrolled synthesis of more complex
derivatives, the following studies were undertaken. Our results
show that the reaction can be catalytic in zinc(II) triflate and that
its strained cycloadducts participate in a variety of enabling
synthetic transformations.
’ RESULTS AND DISCUSSION
Catalysis of the Reaction with Electron-Poor Olefins. We
began our investigation by optimizing reaction conditions for
the preparation of 3-acetyl-2,2-dimethylcyclobutanone (see 11,
Table 2) from R-chloroenamine 7 and methyl vinyl ketone
(MVK). These experiments were facilitated by the commercial
availability of 7, a reagent useful in other contexts, including acid
chloride generation under neutral conditions12 and the halogena-
tion of delicate medium ring ethers and lactones.13 As shown in
entry 1 of Table 1, prolonged stirring of a sealed CH2Cl2 solution
1663
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
Table 2. Zn-Catalyzed, Solvent-Free Access to 3-Acyl
Cyclobutanones
the bond between C3 and C4 is significantly longer than the
remaining three: C1-C2 1.513 Å, C1-C4 1.526 Å, and C2-C3
1.537 Å. Although steric congestion imparted by the gem-
dimethyl unit could partly account for this, a stabilizing FMO
interaction involving the C3-C4 linkage and the carboxylic
π*CdO orbital is likely to be present given its favorable alignment
and the shortened distance between C3 and C5 (1.494 Å).
Whether this constitutes an incipient potential for substitution
events at the quaternary carbon is an interesting prospect for
future study.
Synthesis and Participation of Other r-Chloroenamines.
Further exploration on the scope of this transformation, espe-
cially with respect to the R-chloroenamine component, has been
challenging for several reasons: (1) 7 is the only commercial
available derivative bearing the reactive function; (2) existing
protocols18 for R-chloroenamine synthesis are complicated by
moisture sensitivity and tedious purifications (see below); and
finally, (3) β-monosubstituted chloroenamines—precursors for
putative access to aldo-keteniminium cations—are reportedly18b
unstable; they undergo self-condensation upon concentration
and are best prepared and used in solution at low temperature.19
Still, we wish to relay some progress on this challenging front,
beginning with useful changes to the reported synthesis of 7 that
translate well to the preparation of other β-disubstituted R-
chloroenamines.
As shown below in eq 1, the required events are amide
chlorination and proton loss from the resulting chloroimidate.
Because of the hazards associated with phosgene (see Scheme 2),
POCl3 has been recommended as an alternative reagent for
chlorination. However, in certain cases,18b it is necessary to
remove unreacted POCl3 before addition of the base (typically
Et3N), and several standard chlorinating agents are known to
react with tertiary amides. We decided upon oxalyl chloride since
the byproducts of chlorination are gaseous and preliminary
experiments with 19 gave high efficiency even in the absence
of DMF as a catalyst. A hydrocarbon solvent was chosen to
promote precipitation of the crystalline iminium salt 20 as it
forms, allowing for quantitative recovery by cannula filtration. In
the same flask, 20 is suspended in pentane at -78 °C and treated
with a solution of anhydrous trimethylamine. Upon warming, the
insoluble chloroimidate gradually reacts to form 7 with simulta-
neous precipitation of trimethylammonium chloride. Filtration
and concentration under inert conditions gives pure, colorless
product with no contamination from the base or its hydrochlor-
ide. This alternative is an improvement over the original use of
Et3N, which in our hands afforded discolored samples of 7 and
proved difficult to cleanly separate from the product by distilla-
tion (bp 129-130 °C).20
a Determined by 1H NMR analysis. b After chromatography on silica gel.
c After vacuum distillation. d With a 3:1 regioisomeric ratio favoring
that shown.
Several ketones in Table 2 are waxy solids below 23 °C, but
attempts to obtain single crystals suitable for diffraction met with
difficulty. An X-ray structure was eventually obtained for car-
boxylic acid 18, confirming regiochemical assignments across the
series (Scheme 4). Since 18 is soluble in water, conditions for
ester hydrolysis in 12 were needed that avoided an aqueous
workup altogether. A simple solution emerged involving: (1)
treatment with LiOH in wet THF (23 °C, 18 h); (2) removal of
solvent to provide lithium 2,2-dimethyl-3-oxocyclobutanecar-
boxylate as an off-white solid; and (3) dissolution in cold THF,
protonation by brief treatment with anhydrous HCl, filtration of
insoluble LiCl, and concentration to a white solid (91% yield). As
shown, 18 packs as a dimer in the solid state with intermolecular
CO2H hydrogen bonding. Noteworthy is the considerable bond
length distortion evident in the four-membered ring. At 1.583 Å,
Evidence for the utility of the new haloenamine synthesis
is provided in Scheme 5. Cyclic derivatives 21 and 22, prepared
in high yield by the operations just detailed, were without
purification subjected to the optimized conditions for Zn-cata-
lyzed [2 þ 2] annulation. In the presence of methyl acrylate, 21
provides spirocycle 23 in 61% isolated yield. Exposure of 22 to
3-methylene-2-norbornanone gives tetracyclic diketone 24
(60%, 3:2 rr, 40% minor), whose structure was secured by
1664
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
Scheme 4
Scheme 6. Initial Studies on the Versatility of 3-Acyl
Cyclobutanones
Scheme 5. Zn-Catalyzed Synthesis of Spirocyclic
Cyclobutanones
Scheme 7. Seemingly Unlikely Paths for Regioselective [2 þ
2] Annulation
X-ray crystallography following crystallization from heptane. The
ready availability of the starting materials and mild conditions
associated with this reaction are advantageous. Another salient
feature pertains to rapid buildup of sterically crowded all-carbon-
substituted quaternary carbon atoms, some contiguous. Both
carbonyl groups in any of the products seem poised for additional
stereoselective bond-forming events, and ring strain can also be
quite serviceable for further elaboration.
Synthetic Utility of 3-Acyl Cyclobutanones. In support of
these assertions, we investigated the reactions shown in Scheme 6
starting with ketoester 12 as a representative synthon. After some
experimentation, Schmidt nitrogen insertion was achieved in wet
methanol with sulfuric acid as a promoter, giving 25 as a single
regioisomer. Baeyer-Villiger oxidation under standard condi-
tions provides 26 in 76% yield. Room temperature R-bromina-
tion with pyridinium perbromide proceeds quantitatively to give
27, a structure assigned as the unexpected cis-diastereomer on the
basis of NOE experiments. Subsequent ring contraction21 with
sodium methoxide affords dimethylcyclopropane dimethyl ester
28, a C2-symmetric structure possibly useful for making rigidified
diol ligands with altered bite angles. Although the basis for
diastereocontrol in the bromination event is not clear, we note
that a similar contrasteric outcome in R-hydroxylation would be
of relevance to the complex enecyclobutane core of providencin.3
Mechanistic Considerations for Cyclobutannulation. The
exclusive or predominant products of the Zn-catalyzed process
contain a 1,4-dicarbonyl and seem to derive from a polarity
inversion. As mentioned previously and depicted in Scheme 7, a
conventional ionic mechanism that invokes keteniminium 8 as a
potent electrophile appears unlikely; participation of the elec-
tron-poor olefin B as a nucleophile results in a buildup of positive
charge R to the carbonyl group (C), or in the case of minor
regioisomer E, at the β position which is necessarily a primary
carbon for substrates currently within the scope of this transfor-
mation (see D). It is also worth noting that were the Zn catalyst
to act as a Lewis acid for the carbonyl group, rather than
removing chloride from the starting R-chloroenamine 7, a
straightforward path to a cyclobutanone can be drawn with the
enamine unit in 7 through intermediates F and G (bottom of
Scheme 7). However, the 2-acyl isomer (E) would form if this
mechanism were operative, and the co-product is only observed
in two seemingly unrelated cases of cyclocondensation.
We therefore began to consider alternative mechanisms for
product formation that start with Zn-catalyzed dehalogenation to
give the putative keteniminium cation 8. Since the method is
general for mono- and disubstituted R-olefins with a ketone,
ester, or amide substituent (B, G = CH3, OCH3, or N(CH3)2),
we judged that initial association of a Lewis basic oxygen lone pair
at C1 of 8 would be facile (fH, Scheme 8). Further processing
of ion pair H is then possible in either of two ways. The 5-exo-trig
cyclization gives a tertiary cation (I) that could be quickly
trapped internally to give [2.1.1]bicyclic oxonium ion J, a species
that derives additional resonance stabilization if substituent G is
electron-donating. Subsequent iminium ion formation and hy-
drolysis deliver the observed regioisomer A. Alternatively,
6-endo-trig cyclization gives R-alkoxy iminium ion K in a step
1665
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
Scheme 8. Plausible Mechanism Based on Initial Cl- Ab-
straction by Zn(II)
Figure 1. DFT calculations for keteniminium HOMO and LUMO.
Scheme 9. Viable Mechanism Based on Lewis Acid-Carbo-
nyl Activation
that could also be viewed as a 6-π electrocyclic ring closure. A
stereoelectronically disfavored 4-endo-trig bicyclization would
then be needed to reach intermediate L, a potential precursor
to the unusual 2-acyl regioisomer E. This, together with the
forbidden nature of the ring closure when viewed as a disrotatory
4-π electrocyclization, can adequately explain the exclusive
formation of 3-acyl cyclobutanone A.
At this point, a cationic path through H, I, and J was viewed as
a possible scenario for Zn-catalyzed merger of the reactants, and
we could account for the production of small amounts of
regioisomer E in the cases of entry 6 (Table 2) and 22f24
(Scheme 5) by invoking the Michael addition/Mannich closure
pathway described above (bottom of Scheme 7). The fact that
cations H-J are either tertiary or resonance-stabilized, and thus
more stable than those in C or D (Scheme 7), lends additional
support to the mechanism in Scheme 8. However, it involves a
reactivity inversion for the enamine function, and it is not clear
why such a reversal would occur. The strained and congested
nature of bridged bicyclic intermediate J was also a concern. A
computational approach toward probing the electronic structure
in the keteniminium species 8 thus seemed appropriate for
gaining more insight on its role in this process.
DFT Calculations and Their Mechanistic Implications.
DFT calculations at the B3LYP/6-31G* level show that the
HOMO-LUMO gap in tetramethylketeniminium cation is 7.17
kcal/mol (Figure 1). This compares favorably to the result for
ketene itself of 7.26 kcal/mol. More noticeable differences can be
found upon comparison of the Mulliken populations in each
electrophile. Our calculations show that, for ketene, Mulliken
charges on the R and β carbon atoms are, respectively, 0.153 and
0.021, and the charge on the carbonyl oxygen is -0.241. In
contrast, the keteniminium bears charges of 0.074 and -0.092 at
the R and β positions and 0.166 for the nitrogen atom. The
localization of partial anionic charge at the β carbon is unusual
considering that the nitrogen lone pair is disposed completely
orthogonal to the C-C π bond, but we were more intrigued by
the apparent attenuation of electrophilicity at the R carbon of
keteniminium (relative to ketene). Each of the mechanisms
considered thus far begins by interaction of the R,β-unsaturated
carbonyl with the keteniminium LUMO. However, there is no
reason to suspect that the reaction could not involve the
keteniminium HOMO and the vacant π* orbital on the elec-
tron-deficient alkene.
Another mechanistic scenario therefore emerges based on the
alternative role for keteniminium as a nucleophile (Scheme 9).
The π component in 8 may engage the electrophilic β carbon of
substrate in an addition subject to LUMO lowering by the Lewis
acid. Subsequent internal trapping of the tertiary cation by the Zn
enolate (in M) would generate dimethyliminium N, the direct
product prior to the addition of water upon workup. Scrutiny
reveals vicinal dication character in intermediate M, so it is
probably best to view this as a concerted asynchronous process.
Importantly, the regiochemistry and timing of C-C bond forma-
tion with respect to keteniminium ion 8 is the same here as that
given in Scheme 8. However, the two mechanisms differ greatly
in their underlying basis for zinc(II) catalysis; the latter clearly
implies a dual role for Zn(OTf)2 in assisting both generation of
the keteniminium cation (8) and acceleration of its addition to
the various electrophilic Michael acceptors.
Evidence To Support a Dual Role for the Zinc(II) Catalyst.
In an effort to validate one of the two mechanistic pathways,
we searched for a means to separately evaluate the catalyst’s
role in both chloride abstraction from R-chloroenamine 7 and
carbonyl activation. It was known from work by Ghosez7a and
Weingarten14 that keteniminium salts form quantitatively (and
can indeed be observed and characterized spectroscopically)
upon exposure of 7 to an appropriate silver(I) salt in CH2Cl2 or
acetonitrile (AgBF4 and AgPF6). We chose to experiment with
AgOTf since triflate may best approximate the true counterion
under our reaction conditions. As indicated below, treatment of 7
with an equimolar amount of AgOTf in CH2Cl2 at 0 °C
immediately gives a voluminous AgCl precipitate that can be
filtered off under inert conditions; variable processing of the
supernatant 8 then gave the results entered in Table 3. In the
absence of Zn(OTf)2, a mixture of tetramethylketeniminium 8
1666
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
Table 3. Transformations of Keteniminium Formed by a Ag(I) Promoter
entry
conditions
“R” in product
conv to product (%)a
1
2
3
4
1 equiv of MVK, no catalyst
Ac
Ac
Ac
Ph
<5
80
75
83
1 equiv of MVK, 25 mol % of Zn(OTf)2
1 equiv of MVK, 1.25 equiv of AgOTf (@ start)
1 equiv of styrene, no catalytst
a Determined by 1H NMR analysis of the unpurified reaction mixture.
MVK, 83% conversion to 3-phenyl-2,2-dimethylcyclobutanone
is observed in the absence of Zn(OTf)2 (entry 4). This last result
implies that more electron-rich donor alkenes freely engage the
keteniminium through the typical electrophilic mechanism (see
Scheme 7); in the case of styrene, the penultimate cationic
intermediate is a stable benzylic cation. This finding also points
to a way to probe Zn(OTf)2’s role in keteniminium ion forma-
tion, namely, success in catalyzing the synthesis of cyclobuta-
nones starting from non-carbonyl-substituted olefins.
Table 4. Extending the Catalytic [2 þ 2] Annulation to
Nucleophilic Olefins
Data compiled in Table 4 indeed confirm that the Ghosez
cyclobutanone annulation with simple mono- and disubstituted
olefins is subject to catalysis by Zn(OTf)2. As entry 1 shows, a
stirred solution of cyclohexene, 7, and 25 mol % catalyst gives 88%
of bicyclooctanone 29 after 36 h at ambient temperature.
Cycloaddition of trans-1,3-decadiene proceeds to >98% conver-
sion in18h at23°C, affording the 1-octenyl-substituted ketone30
in 85% yield as a single regioisomer (entry 2). Additionally,
styrene, p-methoxystyrene, and p-nitrostyrene condense smoothly
with chloroenamine 7 to give the respective [2 þ 2] adducts in
good yield. All three of the latter reactions give quantitative
conversion under the indicated conditions (entries 3-5), but
monitoring each prior to completion allows for manifestation of
the expected rate differences. For instance, at a concentration of
0.25M inCDCl3, p-nitrostyrene gives 14%conversion to 33 in the
same time (2 h) that p-methoxystyrene gives 75% conversion to
32. Again, these data seem to be consistent with the hypothesis
that non-carbonyl-substituted alkene substrates react with the
keteniminium ion through a conventional ionic pathway in which
charge buildup at the benzylic position of the nucleophile is
influenced by the electron-releasing or electron-withdrawing
effects of the para-substituent.
a By 1H NMR analysis. b After silica gel chromatography. c Run at 40 °C.
’ CONCLUSIONS
and MVK gives <5% of 3-acetyl-2,2-dimethylcyclobutanone
(entry 1). In stark contrast, addition of substrate and 25 mol %
of catalyst to the solution of 8 allows for smooth conversion to
product, measured at 80% after 12 h at 40 °C (entry 2). These
two outcomes clearly implicate a Lewis acid-base association
between substrate and Zn(II) as a requirement for annulation,
and the closing entries of Table 3 further corroborate this idea. If
the experiment is initiated by a 25 mol % excess of AgOTf, a
comparable 75% conversion to product is observed (entry 3).
This suggests that Zn(OTf)2 is not alone in its ability to activate
the R,β-unsaturated carbonyl substrate as a Lewis acid. Finally, if
a less electron-poor olefin such as styrene is added to 8 instead of
We demonstrate zinc(II) catalysis in an unusual merger of
keteniminium ions22 and electron-poor alkenes. The former are
accessed by catalytic halogen abstraction from R-chloroena-
mines, themselves prepared by an improved one-pot protocol
from the corresponding tertiary amide. In all but two cases
studied to date, a 3-acyl cyclobutanone is the exclusive regio-
isomer formed, and we have shown such adducts to be useful in
the context of strain release. More importantly, a 1,4-dicarbonyl
inherent to the product structure points to an umpolung in
reactivity.23 On the basis of DFT calcluations and empirical
studies, a mechanism in which Zn(OTf)2 serves two functions is
1667
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
currently favored. The first involves abstraction of chloride anion
from the enamine starting material, giving tetrasubstituted
keteniminium ions as reactive intermediates. The second role
for the Lewis acid is to activate the R,β-unsaturated carbonyl
toward Michael addition by the modestly nucleophilic R carbon
of the keteniminium. This mechanistic proposal is supported by
the following facts: (1) β-substituted enones, acrylic esters, and
amides fail to react under the standard reaction conditions; (2)
the optimal catalyst loading (25 mol %) is relatively high,
reflecting its need to accomplish separate tasks during formal
[2 þ 2] cycloaddition; and (3) donor olefins that are poised to
react in accordance with an established7 stepwise pathway in
which keteniminium is the electrophile also react efficiently
under mild conditions with substoichiometric amounts of Zn-
(OTf)2. Some questions on scope and generality of the Zn-
catalyzed [2 þ 2] cycloaddition remain unanswered at this time,
including the performance of β-monosubstituted chloroena-
mines in the transformation and its preferred means of
enantiocontrol.
Infrared spectra were recorded on a FT-IR spectrophotometer, νmax
in cm-1. Bands are characterized as broad (br), strong (s), medium (m),
1
and weak (w). H NMR spectra were recorded on 400 or 500 MHz
spectrometers and are reported relative to deuterated solvent signals
(CDCl3: δ 7.26 ppm; C6D6: δ 7.16 ppm). Data are reported as follows:
chemical shift, multiplicity, coupling constant (hertz), and integration.
13C NMR spectra were recorded on 100 or 125 MHz spectrometers with
complete proton decoupling. Chemical shifts are reported in parts per
million from tetramethylsilane with solvent as the internal reference
(CDCl3: δ 77.16 ppm; C6D6: δ 128.39 ppm). High-resolution mass
spectra were obtained using the electrospray ionization method in
positive ion mode.
Experimental Procedures. 1-Chloro-N,N,2-trimethyl-1-prope-
nylamine (7). To a solution of N,N-dimethylisobutyramide (19,
0.500 mL, 3.87 mmol) in 3.87 mL of pentane under nitrogen at
-78 °C was added dropwise from a syringe neat oxalyl chloride
(0.327 mL, 3.87 mmol, 1.0 equiv). Upon slow (2 h) warming to
23 °C, a heterogeneous mixture was obtained that showed no
starting material remaining according to TLC analysis. The super-
natant was removed by cannula filtration under postitive nitrogen
pressure, and the solid was washed with three 5 mL volumes of
pentane. After removing the last pentane wash, the white solid was
suspended in 3.87 mL of pentane and cooled to -78 °C. With
vigorous stirring, cold trimethylamine (1.94 mL of 2.0 M solution,
3.87 mmol, 1.0 equiv) was then added dropwise and the hetero-
geneous mixture was transferred to a cold room (4 °C) for 12 h. The
reaction never achieves homogeneity since trimethylamine hydro-
chloride precipitates as the intermediate chloroimidate (20) enters
solution by deprotonation. Upon taking the sample into a glovebox,
additional pentane was used to filter the mixture through a fine
frit, affording a colorless filtrate that was concentrated in vacuo
at -78 °C (to discourage violent bumping). This operation provided
0.465 g of a colorless, ready-to-use sample of R-chloroeneamine 7
(90% yield), which can be stored for iterative use at -30 °C in a
glovebox freezer. 1H and 13C NMR data for this compound are
identical to that of vacuum-distilled commercial samples.
’ EXPERIMENTAL SECTION
Materials and Methods. Unless stated otherwise, reactions were
carried out in flame-dried glassware under an atmosphere of nitrogen
with standard Schlenk or vacuum line techniques. A preferred experi-
mental setup for sonication consists of a variable output Misonix
3000 sonicator equipped with an external water circulator to control
the temperature of the bath in which the cup horn (sonar probe) is
immersed. Dry and degassed solvents (CH2Cl2, pentane, and THF)
were dispensed from a solvent purification system based on activated
alumina columns.24 Materials obtained from commercial sources
were usually purified before use. 1-Chloro-1-N,N,2-trimethylprope-
nylamine was carefully degassed under vacuum at -78 °C and then
distilled with warming to 23 °C through a short path with the
receiving flask cooled to -78 °C. A dry trimethylamine solution (2.0
M in pentane) was prepared as follows: (1) commercial gas was
condensed onto potassium hydroxide pellets at -78 °C; (2) the
resulting colorless mixture was stirred cold for 10 min and then
distilled through a short path with warming into a preweighed flask
kept at -78 °C; and (3) the resulting colorless liquid (8.87 g, 150
mmol) was dissolved in 75.0 mL of pentane and transferred by
cannula to a dry Shlenk flask that could be stored at -20 °C for
repeated use. N,N-Dimethylisobutyramide was distilled under va-
cuum over calcium hydride. Zinc(II) triflate was rigorously dried
under vacuum in an Abderhalden pistol containing P2O5 at 80 °C.
N,N-Dimethylcyclohexanecarboxamide was prepared by catalytic
hydrogenation of benzoic acid25 and subsequent chlorination/
amidation according to a procedure given below for N,N-dimethyl-
cycloheptanecarboxamide. Methyl vinyl ketone, 3-methyl-3-buten-
2-one, methyl methacrylate, methyl acrylate, N,N-dimethylmetha-
crylamide, and N,N-dimethylacrylamide were freed of stabilizers by
vacuum distillation at 23 °C over calcium sulfate through a short path
with the receiving flask kept at -78 °C. Oxalyl chloride was distilled under
nitrogen. Anhydrous ethyl acetate (for the bromination of 12) was distilled
under nitrogen from CaH2. Pyridinium perbromide recrystallized as bright
orange needles from cyclohexane. Cyclohexene was distilled under nitrogen.
trans-1,3-Decadiene, styrene, p-methoxystyrene, and p-nitrostyrene were
vacuum distilled.
3-Acetyl-2,2-dimethylcyclobutanone (11)
Representative Thermal Procedure for Formal [2 þ 2] Cycloaddi-
tion. In a glovebox, a flame-dried 2 dram (7.4 mL) vial equipped with a
stir bar was charged successively with methyl vinyl ketone (0.254 mL,
3.13 mmol, 1.25 equiv), 1-chloro-N,N,2-trimethyl-1-propenylamine
(7, 0.331 mL, 2.50 mmol), and zinc(II) chloride (0.341 g, 2.50 mmol,
1.0 equiv). After sealing the vial with a rubber septum and removing it
from the glovebox, 2.5 mL of CH2Cl2 was added under positive nitrogen
pressure in a fume hood. The resulting heterogeneous mixture was
tightly capped and stirred at 50 °C for 24 h, a point at which the solids
had dissolved to form a yellow solution; 1.1 mL of 10:1 THF-H2O was
then added to the solution to hydrolyze the iminium salt of the product.
After 1 h of stirring, the mixture was diluted with 20 mL of CH2Cl2 and
dried over magnesium sulfate. Vacuum filtration and concentration
afforded a yellow oil which showed 97% conversion to product ketone
(based on relative integration of 1H NMR signals corresponding to N,N-
dimethylisobutyramide (19), the product of R-chloroenamine hydro-
lysis). Purification was achieved by passage through a short plug (2.5 cm
height, 7.5 cm diameter) of silica gel in 3:1 hexanes/ethyl acetate (TLC
Rf = 0.34 for ketone, 0.08 for amide). Concentration of the chromato-
graphy fractions yielded 0.283 g of product (81% yield) as a colorless oil:
IR (thin film) 2966 (s), 2924 (m), 2869 (w), 1782 (s), 1708 (s), 1641
(w), 1464 (m), 1442 (w), 1383 (m), 1364 (s), 1176 (s), 1160 (m), 1126
(m), 1066 (s); 1H NMR (400 MHz, CDCl3) δ 3.56 (dd, J = 17.9, 7.9 Hz,
1H), 3.09 (dd, J = 8.4, 7.9 Hz, 1H), 2.78 (dd, J = 17.9, 8.4 Hz, 1H), 2.16
(s, 3H), 1.32 (s, 3H), 0.98 (s, 3H); 13C NMR (100 MHz, CDCl3) δ
210.3, 205.4, 64.1, 47.7, 43.1, 31.2, 23.2, 18.2; HRMS (ESIþ) calcd for
C8H13O2þ [M þ H]þ 141.0916, found 141.0915.
Column chromatography was performed with normal phase silica gel
60 (230-400 mesh ASTM) and driven with compressed air. Analytical
thin-layer chromatography was carried out with silica gel 60 F254
precoated plates (250 μm thickness) and potassium permanganate or
ceric ammonium molybdate (CAM) stains for spot visualization.
1668
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
Methyl 2,2-dimethyl-3-oxocyclobutanecarboxylate (12)
N,N-dimethylacrylamide (3.13 mmol, 1.25 equiv), 0.331 mL of
1-chloro-N,N,2-trimethyl-1-propenylamine (7, 2.50 mmol), and 0.227
g of zinc(II) triflate (0.625 mmol, 0.25 equiv). Upon obtaining an orange
oil after the hydrolytic workup (96% conversion by 1H NMR analysis), a
Kugelrohr distillation apparatus was used for purification. First, N,N-
dimethylisobutyramide (19) and the unreacted N,N-dimethylacryla-
mide were removed at 1.8 Torr and 50 °C without isolation. Upon
raising the temperature to 125 °C, the product distilled at 1.8 Torr over 6
h. Careful transfer and weighing of this distilled fraction delivered 0.353
g of a colorless oil (83% yield): IR (thin film) 2964 (w), 2929 (w), 2868
(w), 1779 (s), 1636 (s), 1499 (w), 1463 (w), 1401 (w), 1263 (w), 1153
(m), 1072 (m); 1H NMR (400 MHz, CDCl3) δ 3.89 (dd, J = 18.1, 7.1
Hz, 1H), 3.18 (dd, J = 8.8, 7.1 Hz, 1H), 3.06 (s, 3H), 3.02 (s, 3H), 3.01
(dd, J = 18.1, 8.8 Hz, 1H), 1.37 (s, 3H), 1.06 (s, 3H); 13C NMR (100
MHz, CDCl3) δ 212.2, 170.7, 63.5, 4þ5.7, 39.2, 37.4, 35.9, 23.3, 18.4;
HRMS (ESIþ) calcd for C9H16NO2 [M þ H]þ 170.1181, found
170.1173.
N,N,1,2,2-Pentamethyl-3-oxocyclobutanecarboxamide (15). The
sonication procedure described above was applied with 0.376 mL of
N,N-dimethylmethacrylamide (3.00 mmol, 1.2 equiv), 0.331 mL of
1-chloro-N,N,2-trimethyl-1-propenylamine (7, 2.50 mmol), and 0.227 g
of zinc(II) triflate (0.625 mmol, 0.25 equiv). Upon obtaining an orange
oil after the hydrolytic workup (90% conversion by 1H NMR analysis), a
Kugelrohr distillation apparatus was used for purification. First, N,N-
dimethylisobutyramide (19) and the unreacted N,N-dimethylmethacry-
lamide were removed at 1.8 Torr and 50 °C without isolation. Upon
raising the temperature to 150 °C, the product distilled at 1.8 Torr over 6
h. Careful transfer and weighing of this distilled fraction delivered 0.330
g (80% yield) of a colorless oil that consisted of a pure 3:1 mixture of
regioisomeric cyclobutanones. Characterization follows for the major
isomer, separable with 2:2:3:1 CH2Cl2/Et2O/hexanes/NH4OH as
chromatographic eluant (TLC Rf = 0.33): IR (thin film) 2962 (w),
2928 (w), 2971 (w), 1779 (s), 1628 (s), 1496 (w), 1462 (w), 1396 (m),
1382 (m), 1270 (w), 1158 (w), 1144 (w), 1114 (w), 1093 (m), 1065
(w), 938 (w), 671 (w); 1H NMR (500 MHz, CDCl3) δ 4.17 (d, J = 17.6
Hz, 1H), 2.99 (s, 3H), 2.99 (s, 3H), 2.62 (d, J = 17.6 Hz, 1H), 1.41 (s,
3H), 1.22 (s, 3H), 1.19 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 212.6,
173.8, 63.0, 55.9, 43.7, 37.7, 36.6, 21.7, 21.2, 17.1; HRMS (ESIþ) calcd
for C10H18NO2þ [M þ H]þ 184.1338, found 184.1341.
Representative Sonication Procedure for Formal [2 þ 2] Cycload-
dition. In a glovebox, a flame-dried 2 dram (7.4 mL) vial equipped with a
stir bar was charged successively with methyl acrylate (0.281 mL, 3.12 mmol,
1.25 equiv), 1-chloro-N,N,2-trimethyl-1-propenylamine (7, 0.331 mL, 2.50
mmol), and zinc(II) triflate (0.227 g, 0.625 mmol, 0.25 equiv). The mixture
was briefly stirred to blend its contents, and the vial was sealed tightly with a
Teflon-lined screw cap. After removing the sample from the glovebox, the
reaction was sonicated at 35 °C for 6 h (180 W, Misonix Sonicator 3000),
during which time the solution turned yellow and, after 2 h, solidified; 1.1 mL
of 10:1 THF-H2O was then added to the mixture to hydrolyze the iminium
salt of the product. After 1 h of stirring, the mixture was diluted with 20 mL of
CH2Cl2 and dried over magnesium sulfate. Vacuum filtration and solvent
removal on a rotary evaporator afforded a yellow oil showing >98%
conversion to product ketone based on the lack of detectable signals
corresponding to N,N-dimethylisobutyramide (19), the product of R-
chloroenamine hydrolysis. Purification was achieved by filtration through a
short plug (2.5 cm height, 7.5 cm diameter) of silica gel in 9:1 hexanes/ethyl
acetate (TLC Rf = 0.31 for ketone, 0.0 for amide). Concentration of fractions
gave 0.257 g of a colorless oil (66% yield): IR (thin film) 2965 (s), 2930 (m),
2870 (w), 1788 (s), 1735 (s), 1463 (m), 1438 (s), 1383 (m), 1353 (m),
1273 (w), 1208 (s), 1180 (s), 1163 (m), 1131 (w), 1110 (w), 1065 (m),
1040 (m); 1H NMR (500 MHz, CDCl3) δ 3.73 (s, 3H), 3.52 (dd, J = 17.9,
7.3 Hz, 1H), 3.08 (dd, J = 17.9, 9.0 Hz, 1H), 2.94 (dd, J = 9.0, 7.3 Hz, 1H),
1.29 (s, 3H), 1.10 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 211.0, 172.5,
þ
64.3, 51.9, 44.7, 40.6, 23.0, 18.4; HRMS (ESIþ) calcd for C8H13O3
[M þ H]þ 157.0865, found 157.0858.
3-Acetyl-2,2,3-trimethylcyclobutanone (9). The sonication proce-
dure described above was applied with 0.252 g of 3-methyl-3-buten-2-
one (3.00 mmol, 1.2 equiv), 0.331 mL of 1-chloro-N,N,2-trimethyl-1-
propenylamine (7, 2.50 mmol), and 0.227 g of zinc(II) triflate (0.625
mmol, 0.25 equiv). Upon obtaining a yellow oil after the hydrolytic
workup (87% conversion by 1H NMR analysis), a Kugelrohr distillation
apparatus was used for purification. First, N,N-dimethylisobutyramide
(19) and the unreacted 3-methyl-3-buten-2-one was removed at 1.8
Torr and 50 °C without isolation. Upon raising the temperature to
75 °C, the product distilled at 1.8 Torr over 6 h. Careful transfer and
weighing of this distilled fraction delivered 0.283 g of a colorless oil (73%
yield): IR (thin film) 2968 (m), 2929 (m), 2874 (w), 1779 (s), 1702 (s),
1460 (m), 1424 (w), 1404 (w), 1386 (m), 1381 (m), 1357 (m), 1279
N-Methoxy-N,2,2-trimethyl-3-oxocyclobutanecarboxamide (16).
The sonication procedure described above was applied with 0.345 g of
N-methoxy-N-methylacrylamide (3.00 mmol, 1.2 equiv), 0.331 mL of
1-chloro-N,N,2-trimethyl-1-propenylamine (7, 2.50 mmol), and 0.227 g
of zinc(II) triflate (0.625 mmol, 0.25 equiv). Upon obtaining an oil after
the hydrolytic workup (85% conversion by 1H NMR analysis), a
Kugelrohr distillation apparatus was used for purification. First, N,N-
dimethylisobutyramide (19) and the unreacted N-methoxy-N-methyla-
crylamide were removed at 1.8 Torr and 50 °C without isolation. Upon
raising the temperature to 125 °C, the product distilled at 1.8 Torr over 6
h. Careful transfer and weighing of this distilled fraction delivered 0.294
g of a colorless oil (64% yield): IR (thin film) 2967 (w), 2931 (w), 2870
(w), 1779 (s), 1655 (s), 1462 (m), 1444 (m), 1420 (m), 1383 (m), 1298
(w), 1258 (w), 1178 (w), 1163 (w), 1119 (w), 1067 (m), 1023 (w), 977
(w), 955 (w); 1H NMR (500 MHz, CDCl3) δ 3.72 (dd, J = 17.9, 7.6 Hz,
1H), 3.71 (s, 3H), 3.23-3.18 (m, 4H), 2.97 (dd, J = 18.1, 8.8 Hz, 1H),
1.31 (s, 3H), 1.04 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 212.8, 172.8,
64.5, 61.1, 44.0, 39.0, 39.1, 32.5, 23.0, 18.3; HRMS (ESIþ) calcd for
C9H16NO3þ [M þ H]þ 186.1130, found 186.1127.
1
(w), 1184 (w), 1154 (m), 1122 (w), 1099 (w); H NMR (400 MHz,
CDCl3) δ 3.91 (d, J = 17.4 Hz, 1H), 2.38 (d, J = 17.4 Hz, 1H), 2.17 (s,
3H), 1.40 (s, 3H), 1.18 (s, 3H), 1.08 (s, 3H); 13C NMR (100 MHz,
CDCl3) δ 210.7, 208.3, 64.0, 50.8, 48.1, 27.9, 20.9, 19.9, 17.3; HRMS
(ESIþ) calcd for C9H15O2þ [M þ H]þ 155.1072, found 155.1077.
Methyl 1,2,2-trimethyl-3-oxocyclobutanecarboxylate (13). The so-
nication procedure described above was applied with 0.321 mL of
methyl methacrylate (3.00 mmol, 1.2 equiv), 0.331 mL of 1-chloro-N,
N,2-trimethyl-1-propenylamine (7, 2.50 mmol), and 0.227 g of zinc(II)
triflate (0.625 mmol, 0.25 equiv). Upon obtaining an orange oil (91%
conversion by 1H NMR analysis) after the hydrolytic workup, purifica-
tion was achieved by filtration through a short plug (2.5 cm height, 7.5
cm diameter) of silica gel in 10:1 hexanes/ethyl acetate (TLC Rf = 0.32
for ketone, 0.0 for amide). Concentration of fractions yielded 0.269 g of a
light yellow oil (63% yield): IR (thin film) 2970 (m), 2938 (m), 2874
(w), 1785 (s), 1732 (s), 1460 (m), 1436 (m), 1403 (w), 1384 (m), 1365
(w), 1296 (s), 1200 (s), 1157 (m), 1134 (s), 1104 (m), 1062 (m), 993
(w); 1H NMR (500 MHz, CDCl3) δ 3.78 (d, J = 17.4 Hz, 1H), 3.72 (br s,
3H), 2.62 (dd, J = 17.6, 11.5 Hz, 1H), 1.38 (s, 3H), 1.11 (d, J = 1.5 Hz,
3H), 1.10 (d, J = 1.5 Hz, 3H); 13C NMR (125 MHz, CDCl3) δ 211.5,
174.9, 64.7, 52.6, 52.1, 42.5, 20.9, 20.0, 17.5; HRMS (ESIþ) calcd for
C9H15O3þ [M þ H]þ 171.1021, found 171.1022.
N-Methoxy-N-methylmethacrylamide. A synthesis was adapted
from that reported for N-methoxy-N-methylacrylamide.26 The following
modifications were made: (1) the reaction was carried out on 12.3 mmol
scale with respect to the acid chloride; (2) 2.0 equiv of N,O-dimethylhy-
droxylamine hydrochloride was used; (3) 4.0 equiv of triethylamine was
substituted for pyridine; and (4) THF was the solvent (0.5 M): IR (thin
N,N,2,2-Tetramethyl-3-oxocyclobutanecarboxamide
sonication procedure described above was applied with 0.310 g of
(14). The
1669
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
1-Chloro-1-cycloheptylidene-N,N-dimethylmethanamine
film) 2973 (w), 2937 (w), 1655 (s), 1629 (m), 1458 (m), 1421 (w),
1377 (m), 1262 (w), 1177 (w), 1149 (w), 1118 (w), 1092 (w), 998 (w),
(21). This R-chloroenamine is prepared from 0.846 g (5.0 mmol) of N,
N-dimethylcycloheptanecarboxamide by direct analogy to the synthesis
of 7, yielding 0.798 g (85% yield) of 21 as a colorless oil: 1H NMR (400
MHz, C6D6) δ 2.42 (dt, J = 17.6, 6.2 Hz, 2H), 2.28 (s, 6H), 1.57-1.15
(m, 8H), 0.87 (dd, J = 7.1, 6.8 Hz, 2H); 13C NMR (100 MHz, CDCl3) δ
133.6, 43.3, 33.0, 32.1, 30.2, 29.7, 28.9, 27.2. IR and mass spectral data
are not recorded; the compound undergoes rapid and quantitative
hydrolysis upon exposure to air and moisture.
1
945 (w), 917 (w), 888 (w); H NMR (400 MHz, CDCl3) δ 5.29
(quartet, J = 1.1 Hz, 1H), 5.22 (quintet, J = 1.7 Hz, 1H), 3.64 (s, 3H),
3.23 (s, 3H), 1.97 (dd, J = 1.7, 1.1 Hz, 3H); 13C NMR (100 MHz,
CDCl3) δ 171.6, 140.3, 117.5, 61.4, 33.5, 20.1; HRMS (ESIþ) calcd for
C6H12NO2þ [M þ H]þ 130.0868, found 130.0861.
N-Methoxy-N,1,2,2-tetramethyl-3-oxocyclobutanecarboxamide (17).
The sonication procedure described above was applied with 0.387 g of
N-methoxy-N-methylmethacrylamide (3.00 mmol, 1.2 equiv), 0.331
mL of 1-chloro-N,N,2-trimethyl-1-propenylamine (7, 2.50 mmol), and
0.227 g of zinc(II) triflate (0.625 mmol, 0.25 equiv). Upon obtaining a
yellow oil after the hydrolytic workup (87% conversion by 1H NMR
analysis), a Kugelrohr distillation apparatus was used for purification.
First, N,N-dimethylisobutyramide (19) and the unreacted N-methoxy-
N-methylmethacrylamide were removed at 1.8 Torr and 50 °C without
isolation. Upon raising the temperature to 150 °C, the product distilled
at 1.8 Torr over 6 h. Careful transfer and weighing of this distilled
fraction delivered 0.283 g of a colorless oil (57% yield): IR (thin film)
2968 (w), 2935 (w), 2873 (w), 1777 (s), 1716 (w), 1648 (s), 1459 (m),
1406 (w), 1375 (m), 1362 (m), 1261 (w), 1154 (m), 1110 (w), 1063
Methyl 3-Oxospiro[3.6]decane-1-carboxylate (23). The sonication
procedure described above was applied with 0.281 mL of methyl acrylate
(3.00 mmol, 1.2 equiv), 0.469 g of 21 (2.50 mmol), and 0.227 g of
zinc(II) triflate (0.625 mmol, 0.25 equiv). Upon obtaining a light orange
oil after the hydrolytic workup (95% conversion by 1H NMR analysis),
21 and the unreacted methyl acrylate were removed at 1.8 Torr and
80 °C without isolation. Passage of the remaining undistilled fraction
through a plug of silica gel in 10:1 hexanes/ethyl acetate (TLC Rf = 0.30)
provided 0.320 g of a colorless oil (61% yield): IR (thin film) 2926 (m),
2857 (w), 1781 (s), 1732 (s), 1645 (w), 1461 (w), 1438 (w), 1353 (w),
1
1267 (w), 1200 (m), 1174 (m), 1089 (w), 1029 (w); H NMR (500
MHz, CDCl3) δ 3.72 (s, 3H), 3.42 (dd, J = 17.7, 7.2 Hz, 1H), 3.02 (dd,
J = 17.7, 8.9 Hz, 1H), 2.91 (dd, J = 8.9, 7.2 Hz, 1H), 1.95 (ddd, J = 11.5,
9.5, 2.2 Hz, 1H), 1.81-1.46 (m, 10H), 1.40 (m, 1H); 13C NMR (125
MHz, CDCl3) δ 211.6, 172.9, 72.5, 52.0, 44.4, 41.5, 35.9, 30.8, 29.3, 29.1,
23.4, 23.2; HRMS (ESIþ) calcd for C12H19O3þ [M þ H]þ 211.1334,
found 211.1341.
1-Chloro-1-cyclohexylidene-N,N-dimethylmethanamine (22). This
R-chloroeneamine is prepared from 0.776 g (5.0 mmol) of N,N-dimethyl-
cyclohexanecarboxamide by direct analogy to the synthesis of 7, yielding
0.485 g (56% yield) of 22 as a colorless oil: 1H NMR (400 MHz, C6D6) δ
2.33 (dd, J = 38.8, 1.2 Hz, 1H), 2.24-2.19 (m, 2H), 1.63 (s, 6H), 1.33-
1.04 (m, 6H), 0.73 (dd, J=7.2, 6.8 Hz, 1H); 13CNMR(100MHz, C6D6) δ
138.9, 131.7, 43.4, 31. 5, 31.2, 28.5, 27.7, 27.3. IR and mass spectral data are
not recorded; the compound undergoes rapid and quantitative hydrolysis
upon exposure to air and moisture.
1
(w), 1002 (m), 751 (w), 730 (w), 657 (w), 426 (w); H NMR (500
MHz, CDCl3) δ 4.00 (d, J = 17.6 Hz, 1H), 3.72 (s, 3H), 3.21 (s, 3H),
2.51 (d, J = 17.6 Hz, 1H), 1.40 (s, 3H), 1.13 (s, 3H), 1.12 (s, 3H); 13C
NMR (125 MHz, CDCl3) δ 213.4, 176.6, 64.3, 60.7, 54þ.1, 43.6, 33.1,
21.1, 20.4, 16.8; HRMS (ESIþ) calcd for C10H18NO3 [M þ H]þ
200.1287, found 200.1285.
2,2-Dimethyl-3-oxocyclobutanecarboxylic acid (18). To a solution
of 0.570 g (3.65 mmol) of 12 in 5.1 mL of 7:1 THF-H2O was added
lithium hydroxide (0.087 g, 3.65 mmol, 1.0 equiv). The solution was
stirred for 18 h sealed under nitrogen at 23 °C, and the solvent was
removed in vacuo. The off-white solid residue that remained (lithium
2,2-dimethyl-3-oxocyclobutanecarboxylate) was then dissolved in 7.3
mL of THF at 0 °C and protonated by brief treatment (5 min) with
anhydrous HCl (generated from NaCl and H2SO4 and gently bubbled
through the solution). Passage through a cotton plug (to remove
insoluble lithium chloride) and concentration afforded 0.470 g (91%
yield) of the acid as a white solid (mp = 66-69 °C): IR (thin film) 3176
(br), 2967 (w), 2928 (w), 2871 (w), 1786 (s), 1706 (s), 1494 (w), 1464
(w), 1444 (w), 1383 (w), 1327 (w), 1280 (w), 1188 (w), 1130 (w), 1065
(w), 796 (w); 1H NMR (400 MHz, CDCl3) δ 11.00 (br, 1H), 3.51 (dd, J
= 18.0, 7.2 Hz, 1H), 3.12 (dd, J = 18.0, 8.8 Hz, 1H), 3.01 (dd, J = 8.8, 7.2
Hz, 1H), 1.34 (s, 3H), 1.20 (s, 3H); 13C NMR (100 MHz, CDCl3) δ
210.6, 178.0, 64.9, 44.7, 40.8, 23.3, 18.5; HRMS (ESIþ) calcd for
C7H11O3þ [M þ H]þ 143.0708, found 143.0705.
Pentacyclic Diketone 24. The sonication procedure described above
was applied with 0.0350 mL of 3-methylene-2-norbornanone (0.286
mmol, 1.0 equiv), 0.0500 g of 1-chloro-1-cyclohexylidene-N,N-di-
methylmethanamine (22, 0.286 mmol, 1.0 equiv), and 0.026 g of
zinc(II) triflate (0.072 mmol, 0.25 equiv). Upon obtaining a yellow oil
1
after the hydrolytic workup (>98% conversion by H NMR analysis),
column chromatography in 10:1 hexanes/ethyl acetate (TLC Rf = 0.20
for minor regioisomer, 0.27 for major regioisomer) furnished 0.041 g of
major product (60% yield) and 0.027 g of minor product (40% yield),
both as colorless crystalline solids. Major regioisomer (mp = 74-
75 °C): IR (thin film) 2932 (m), 2877 (w), 2856 (w), 1773 (s), 1735
(s), 1451 (w), 1382 (w), 1304 (w), 1291 (w), 1276 (w), 1108 (w); 1H
NMR (400 MHz, CDCl3) δ 2.93 (d, J = 17.4 Hz, 1H), 2.87 (d, J = 17.4
Hz, 1H), 2.77 (br m, 1H), 2.63 (br m, 1H), 2.21 (br d, J = 16.4 Hz, 1H),
1.96 (br d, J = 12.0 Hz, 1H), 1.90-0.82 (m, 14H); 13C NMR (100 MHz,
CDCl3) δ 220.1, 210.3, 68.1, 53.9, 49.7, 47.3, 40.9, 36.2, 31.0, 30.3, 26.8,
25.8, 23.8, 23.7, 23.5; HRMS (ESIþ) calcd for C15H21O2þ [M þ H]þ
233.1542, found 233.1543. Minor regioisomer (mp = 72-74 °C): IR
(thin film) 2930 (m), 2877 (w), 2856 (w), 1769 (s), 1732 (s), 1447 (m),
1392 (w), 1274 (w), 1204 (w), 1167 (w), 1151 (w), 1097 (m), 1041
(m), 1021 (m), 948 (w), 910 (w); 1H NMR (500 MHz, CDCl3) δ 3.02
(d, J = 16.6 Hz, 1H), 2.71 (d, J = 16.4 Hz, 1H), 2.68 (ddd, J = 19.8, 4.6,
1.2 Hz 1H), 2.45 (br d, J = 13.2 Hz, 1H), 1.96 (br d, J = 13.0 Hz, 1H),
1.92-1.42 (m, 13H), 1.11 (qt, J = 13.5, 4.2 Hz, 1H), 0.85 (ddd, J = 13.0,
8.8, 4.2 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ 218.8, 208.6, 66.8,
52.3, 50.9, 50.8, 42.7, 35.8, 30.9, 29.3, 26.0, 24.4, 23.9, 23.6, 23.5; HRMS
(ESIþ) calcd for C15H21O2þ [M þ H]þ 233.1542, found 233.1548.
Methyl 2,2-Dimethyl-5-oxopyrrolidine-3-carboxylate (25). To a
stirring solution of 12 (17.0 mg, 0.106 mmol) in 0.42 mL of 3:1
N,N-Dimethylcycloheptanecarboxamide. To a stirring solution of
oxalyl chloride (3.33 mL, 39.3 mmol, 1.0 equiv) in pentane (78.6 mL, 0.5
M) at -78 °C was added dropwise 5.40 mL of cycloheptanecarboxylic
acid (Aldrich, 39.3 mmol, 1.0 equiv). After stirring for 2 h with slow
warming of the reaction to 23 °C, all starting material was judged to be
consumed by TLC analysis. At this point, the solution was recooled to -
78 °C and treated with 49.1 mL (98.3 mmol, 2.5 equiv) of a 2.0 M
solution of dimethylamine in pentane. A white precipitate formed
immediately, and the solution was stirred at -78 °C for an additional
6 h before cannula filtration and concentration of the filtrate on a
rotovap. Column chromatography in 1:1 hexanes/ethyl acetate (TLC Rf
= 0.33) furnished a light yellow oil that was further purified by short path
vacuum distillation to give 5.55 g of a colorless oil (84% yield): IR (thin
film) 2921 (m), 2856 (m), 1640 (s), 1495 (w), 1461 (w), 1411 (w),
1397 (w), 1362 (w), 1255 (w), 1178 (w), 1135 (w), 1107 (w), 1056 (w),
1024 (w); 1H NMR (400 MHz, CDCl3) δ 3.03 (s, 3H), 2.92 (s, 3H),
2.68-2.62 (m, 1H), 1.82-1.40 (m, 12 H); 13C NMR (100 MHz,
CDCl3) δ 218.6, 42.0, 37.4, 35.7, 31.4, 28.4, 27.0; HRMS (ESIþ) calcd
for C10H20NOþ [M þ H]þ 170.1545, found 170.1550.
1670
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
MeOH-H2O, was added 60.0 μL of sulfuric acid (0.211 mmol, 2.0
equiv) at 23 °C. Commercial sodium azide was then added directly to
the reaction mixture as a solid in one portion (8.0 mg, 0.13 mmol, 1.2
equiv), and the mixture was heated to 85 °C. After 3 h of stirring, an
additional 1.2 equiv of sodium azide was added, and the reaction was
stirred for 15 h. The sulfuric acid was then quenched by the dropwise
addition of 1 mL of saturated sodium bicarbonate at 0 °C, and the
mixture was transferred to a small separatory funnel with 3 mL of ethyl
acetate. After removing the organic layer, the aqueous layer was washed
three times with 1 mL portions of ethyl acetate. Pooled organic layers
were then dried over magnesium sulfate, filtered, and concentrated to
yield 11.0 mg of a light yellow oil (>95% pure by NMR analysis, 62% yield):
IR (thin film) 2982 (w), 1777 (s), 1735 (s), 1437 (w), 1391 (w), 1376 (w),
1271 (m), 1259 (m), 1215 (m), 1172 (w), 1119 (m), 1088 (w), 1018 (w),
1001(w), 965 (w), 929 (w); 1H NMR (400 MHz, CDCl3) δ 3.76 (s, 3H),
3.20 (dd, J = 9.7, 8.6 Hz, 1H), 3.09 (dd, J = 17.8, 9.7 Hz, 1H), 2.71 (dd, J =
17.8, 8.6 Hz, 1H), 1.61 (s, 3H), 1.31 (s, 3H); 13C NMR (100 MHz,
CDCl3) δ 173.9, 170.4, 52.6, 50.6, 32.0, 29.9, 28.7, 23.5; HRMS (ESIþ)
calcd for C8H14NO3þ [M þ H]þ 172.0974, found 172.0979.
Concentration of fractions yielded 177 mg of a colorless oil (85% yield):
IR (thin film) 2958 (m), 2924 (s), 2856 (m), 1778 (s), 1461 (m), 1379
(w), 1363(w), 1161(w), 1062(m), 968(m), 757(m), 733 (m);1HNMR
(500 MHz, CDCl3) δ 5.55-5.43 (m, 2H), 3.12 (dd, J = 17.4, 9.1 Hz, 1H),
2.94 (dd, J = 17.6, 7.8 Hz, 1H), 2.67-2.61 (m, 1H), 2.06-2.02 (m, 2H),
1.39-1.22 (m, 8H), 1.17(s,3H), 1.02(s,3H), 0.87(t,J=6.7Hz, 3H);13C
NMR (125 MHz, CDCl3) δ214.8, 133.3, 128.4, 63.1, 47.9, 39.2, 32.7, 31.8,
29.5, 28.9, 23.0, 22.8, 18.5, 14.2; HRMS (ESIþ) calcd for C14H25Oþ
[M þ H]þ 209.1905, found 209.1907.
Fused Bicyclobutanone 29:. IR (thin film) 2930 (m), 2859 (w),
1763 (s), 1450 (m), 1367 (w), 1313 (w), 1161 (w), 1012 (m), 971 (w),
916 (w), 904 (w), 834 (w); 1H NMR (500 MHz, CDCl3) δ 3.53-3.46
(m, 1H), 2.12-2.05 (m, 1H), 1.94-1.87 (m, 1H), 1.86-1.78 (m, 1H),
1.57-1.37 (m, 3H), 1.29 (s, 3H), 1.19-1.05 (m, 3H), 1.02 (s, 3H); 13C
NMR (125 MHz, CDCl3) δ 216.2, 61.0, 52.5, 34.4, 26.4, 24.5, 22.29,
22.28, 20.5, 16.9; HRMS (ESIþ) calcd for C10H17Oþ [M þ H]þ
153.1279, found 153.1240.
2,2-Dimethyl-3-phenylcyclobutanone (31):. IR (thin film) 2961
(w), 2924 (w), 2864 (w), 1775 (s), 1497 (w), 1460 (w), 1380 (w),
1364 (w), 1261 (w), 1159 (w), 1068 (m), 1031 (w), 985 (w), 943 (w),
909 (w), 848 (w), 801 (w), 764 (m), 733 (w), 700 (s), 650 (w), 568 (w);
1H NMR (500 MHz, CDCl3) δ 7.36 (dd, J = 7.6, 7.3 Hz, 2H), 7.29-7.24
(m, 1H), 7.20 (d, J = 7.3 Hz, 2H), 3.47 (dd, J = 16.4, 8.3 Hz, 1H), 3.41
(dd, J = 16.6, 8.3 Hz, 1H), 3.29 (dd, J = 16.1, 8.3 Hz, 1H), 1.36 (s, 3H),
0.79 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 213.7, 139.0, 128.5, 127.7,
126.8, 63.8, 46.0, 41.6, 23.6, 19.0; HRMS (ESIþ) calcd for C12H15Oþ
[M þ H]þ 175.1123, found 175.1137.
cis-Methyl 4-Bromo-2,2-dimethyl-3-oxocyclobutanecarboxylate
(27). Pyridinium perbromide (0.126 g, 0.394 mmol, 1.0 equiv) was
added to 12 (51.4 mg, 0.394 mmol) in 0.66 mL of anhydrous ethyl
acetate under inert atmosphere. After 48 h of stirring at 23 °C, the turbid
yellow reaction mixture was diluted with 4 mL of hexanes, dried over
magnesium sulfate, filtered, and then concentrated by rotary evaporation
to afford 74.0 mg of a light orange oil (>96% pure by 1H NMR analysis,
96% yield, exclusively the cis-diastereomer by NOE experiments): IR
(thin film) 2957 (w), 2931 (w), 2871 (w), 1797 (s), 1735 (s), 1461 (w),
1437 (w), 1386 (w), 1354 (w), 1260 (m), 1212 (m), 1177 (w), 1093
(w), 1053 (w), 978 (w), 849 (w); 1H NMR (400 MHz, CDCl3) δ 5.43
(d, J = 7.9 Hz, 1H), 3.82 (s, 3H), 3.11 (d, J = 7.9 Hz, 1H), 1.44 (s, 3H),
1.18 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 203.4, 169.8, 62.5, 52.7,
52.0, 46.3, 24.1, 18.9; LRMS (ESIþ) calcd for C8H12BrO3þ [M þ H]þ
234.9970, found 235.0201.
Dimethyltrans-3,3-Dimethylcyclopropane-1,2-dicarboxylate (28).
A flame-dried 2 dram (7.4 mL) vial equipped with a stir bar was charged
with a solution of cyclobutanone 27 (9.1 mg, 0.039 mmol, 1 equiv) in
0.30 mL of methanol and cooled to 0 °C. Sodium methoxide (10.7 μL of
a 30% solution in methanol, 0.0580 mmol, 1.5 equiv) was then added
dropwise by syringe. The reaction was stirred for 1 h at 0 °C and
quenched with 1.0 mL of a saturated solution of ammonium chloride.
The resulting mixture was diluted with water (5 mL) and in a separatory
funnel, and the aqueous layer was washed with three 6 mL volumes of
Et2O. Combined organic layers were dried over magnesium sulfate,
filtered, and concentrated to an oil (>98% yield). Spectroscopic data for
this material matched that reported in the literature.27
2,2-Dimethyl-3-(p-methoxyphenyl)cyclobutanone (32):. IR (thin
film) 2960 (w), 2926 (w), 2865 (w), 2837 (w), 1775 (s), 1612 (w),
1583 (w), 1513 (s), 1461 (w), 1443 (w), 1422 (w), 1381 (w), 1363 (w),
1302 (w), 1248 (s), 1180 (w), 1160 (m), 1035 (w), 1066 (m), 826 (m);
1H NMR (500 MHz, CDCl3) δ 7.11 (d, J = 8.3 Hz, 2H), 6.89 (d, J = 8.8
Hz, 2H), 3.81 (s, 3H), 3.40 (dd, J = 15.5, 7.7 Hz, 1H), 3.35 (dd, J = 15.9,
7.6 Hz, 1H), 3.27 (dd, J = 15.5, 8.2 Hz, 1H), 1.33 (s, 3H), 0.78 (s, 3H);
13C NMR (125 MHz, CDCl3) δ 214.1, 158.5, 131.1, 128.7, 114.0, 63.7,
55.4, 46.4, 40.9, 23.6, 19.0; LRMS (ESIþ) calcd for C13H17O2þ [M þ
H]þ 205.1229, found 205.1239.
2,2-Dimethyl-3-(p-nitrophenyl)cyclobutanone (33):. mp = 78-
82 °C; IR (thin film) 2963 (w), 2927 (w), 2866 (w), 1775 (s), 1599
(m), 1514 (s), 1496 (w), 1461 (w), 1382 (w), 1343 (s), 1293 (w), 1262
(w), 1184 (w), 1060 (w), 1139 (w), 1110 (w), 1066 (w), 1014 (w), 988
(w), 942 (w), 866 (w), 853 (m), 759 (m), 713 (w), 698 (w); 1H NMR
(500 MHz, CDCl3) δ 8.23 (d, J = 8.1 Hz, 2H), 7.36 (d, J = 8.6 Hz, 2H),
3.54-3.45 (m, 2H), 1.40 (s, 3H), 0.80 (s, 3H); 13C NMR (100 MHz,
CDCl3) δ 211.6, 147.0, 147.0, 128.6, 123.9, 64.8, 46.2, 41.8, 23.7, 19.1;
þ
HRMS (ESIþ) calcd for C12H14NO3 [M þ H]þ 220.0974, found
220.0972.
2,2 -Dimethyl-3-(1-octenyl)-cyclobutanone (30)
Representative Procedure for Catalytic [2 þ 2] Cycloaddition with
a Non-Carbonyl-Substituted Alkene. In a glovebox, a flame-dried 2
dram (7.4 mL) vial equipped with a small stir bar was charged in
succession with trans-1,3-decadiene (190 mg, 1.0 mmol, 1.0 equiv),
1-chloro-N,N,2-trimethyl-1-propenylamine (136 mg, 1.0 mmol), and
zinc(II) triflate (91.0 mg, 0.250 mmol, 0.25 equiv). After sealing the vial
with a rubber septum and removing it from the glovebox, 1.0 mL of
CH2Cl2 was added under positive nitrogen pressure in a fume hood,
dissolving the reactants and forming a light yellow homogeneous
solution. The vial was then sealed with a Teflon-lined screw cap and
stirred for 18 h at 23 °C. Then, 0.5 mL of 10:1 THF-H2O was added to
the solution to hydrolyze the iminium salt of the product. After 1 h of
stirring, the mixture was diluted with 3 mL of CH2Cl2 and dried over
magnesium sulfate. Vacuum filtration and concentration delivered a yellow
oil that showed >98% conversion to product (based on the lack of
detectable signals corresponding to 19). Purification was achieved by silica
gel chromatography in 25:1 hexanes/ethyl acetate (TLC Rf = 0.33).
’ ASSOCIATED CONTENT
Supporting Information. Copies of H and 13C NMR
1
S
b
spectra for all compounds; 1- and 2-D NOESY spectra for 27;
CIF files for 18 and 24. This material is available free of charge via
the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: jason.kingsbury@bc.edu.
’ ACKNOWLEDGMENT
We gratefully acknowledge Boston College for startup funds,
and Dr. Bo Li, X-ray Crystallography Facility Director, Boston
1671
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672

The Journal of Organic Chemistry
ARTICLE
College, for X-ray analyses. We also thank Dr. Adil Zhugralin for
performing the ab initio calculations. J.M.O. is a John LaMattina
graduate research fellow. Instrumentation in the B.C. Mass Spec
Facility is supported by the NSF (DBI-0619576).
(12) Devos, A.; Remion, J.; Frisque-Hesbain, A.-M.; Colens, A.;
Ghosez, L. Chem. Commun. 1979, 1180–1181.
(13) Bendall, J. G.; Payne, A. N.; Screen, T. E. O.; Holmes, A. B.
Chem. Commun. 1997, 1067–1068.
(14) Weingarten, H. J. Org. Chem. 1970, 35, 3970–3971.
(15) Other Lewis acids screened include AlCl3, TiCl4, SnCl4, and
Sc(OTf)3.
’ REFERENCES
(16) For other representative examples of [2 þ 2] cycloaddition
reactions that benefit from ultrasonic activation, see: (a) Ref 7g. (b) Chen,
X.-T.; Bhattacharya, S. K.; Zhou, B.; Gutteridge, C. E.; Pettus, T. R. R.;
Danishefsky, S. J. J. Am. Chem. Soc. 1999, 121, 6563–6579.
(17) A competing oligomerization of the acrylic ester may account
for the diminished yield in these cases since a highly polar yet intractable
fraction is visible as a baseline spot in the TLC of the crude reaction
mixtures.
(18) (a) Haveaux, B.; Dekoker, A.; Rens, M.; Sidani, A. R.; Toye, J.;
Ghosez, L. Organic Syntheses; Wiley & Sons: New York, 1988; Collect.
Vol. No. 6, pp 282-286. (b) Ghosez, L.; George-Koch, I.; Patiny, L.;
Houtekie, M.; Bovy, P.; Nshimyumukiza, P.; Phan, T. Tetrahedron 1998,
54, 9207–9222.
(19) Our attempt to use a stereoisomeric mixture of the β-benzyl, N,
N-dimethyl R-chloroenamine without isolation was not successful,
presumably due to competing oligomerization of this reactant at the
warmer temperatures required for cyclocondensation. In addition, N,N-
diethyl-2-phenylpropanamide failed to chlorinate efficiently under stan-
dard conditions with oxalyl chloride, presumably due to steric hindrance.
(20) Checkers of the Organic Syntheses prep (ref 18a) also observed
that early fractions of 7 boiling at 100-125 °C were mixtures of product
and triethylamine.
(21) Chen, B.-C.; Ngu, K.; Guo, P.; Liu, W.; Sundeen, J. E.;
Weinstein, D. S.; Atwal, K. S.; Ahmad, S. Tetrahedron Lett. 2001,
42, 6227–6229.
(22) The intermediates play a role in other reactions that begin with
protonation or alkylation of nucleophilic ynamides. See: (a) Zhang, Y.;
Hsung, R. P.; Zhang, X.; Huang, J.; Slafer, B. W.; Davis, A. Org. Lett.
2005, 7, 1047–1050 for a specific example. See: (b) DeKorver, K. A.; Li,
H.; Lohse, A. G.; Hayashi, R.; Shi, Z.; Zhang, Y.; Hsung, R. P. Chem. Rev.
2010, 110, 5064–5106 for a recent review.
(1) (a) Ischay, M. A.; Lu, Z.; Yoon, T. P. J. Am. Chem. Soc. 2010,
132, 8572–8574. For details on a stereocontrolled approach to
tetrasubstituted cyclobutanes and its applications in synthesis, see: (b)
Baran, P. S.; Li, K.; O’Malley, D. P.; Mitsos, C. Angew. Chem., Int. Ed.
2006, 45, 249–252. (c) O’Malley, D. P.; Li, K.; Maue, M. A.; Zografos,
L.; Baran, P. S. J. Am. Chem. Soc. 2007, 129, 4762–4775. For Brønsted or
Lewis acid-catalyzed [2 þ 2] cyclizations based on tandem Michael-
Aldol sequences, see: (d) Inanaga, K.; Takasu, K.; Ihara, M. J. Am. Chem.
Soc. 2005, 127, 3668–3669. (e) Boxer, M. B.; Yamamoto, H. Org. Lett.
2005, 7, 3127–3129.
(2) Reviews on [2 þ 2] photocycloaddition: (a) Crimmins, M. T.;
Reinhold, T. L. Org. React. 1993, 44, 297–588. (b) Bach, T. Synthesis
1998, 683–703. (c) Hoffman, N. Chem. Rev. 2008, 108, 1052–1103.
Select recent studies: (d) Mangion, I. K.; MacMillan, D. W. C. J. Am.
Chem. Soc. 2005, 127, 3696–3697. (e) Ng, S. M.; Bader, S. J.; Snapper,
M. L. J. Am. Chem. Soc. 2006, 128, 7315–7319. (f) Shen, R.; Corey, E. J.
Org. Lett. 2007, 9, 1057–1059.
(3) (a) Marrero, J.; Rodríguez, A. D.; Baran, P.; Raptis, R. G. Org.
Lett. 2003, 5, 2551–2554. For synthetic studies toward providencin, see:
(b) Bray, C. D.; Pattenden, G. Tetrahedron Lett. 2006, 47, 3937–3939.
(c) Schweizer, E.; Gaich, T.; Brecker, L.; Mulzer, J. Synthesis 2007,
3807–3814. (d) Gaich, T.; Arion, V.; Mulzer, J. Heterocycles 2007,
74, 885–862. (e) White, J. D.; Jana, S. Org. Lett. 2009, 11, 1433–1436.
A recent review on the chemistry of furanocembranes and pseudopteranes:
(f) Roethle, P. A.; Trauner, D. Nat. Prod. Rep. 2008, 25, 298–317.
(4) Nozaki, H.; Hayashi, K.-I.; Kido, M.; Kakumoto, K.; Ikeda, S.;
Matsuura, N.; Tani, H.; Takaoka, D.; Iinuma, M.; Akao, Y. Tetrahedron
Lett. 2007, 48, 8290–8292. C8/C80 relative configuration is still
unknown.
(5) Recent total syntheses of other cyclobutane-containing targets:
(a) Tsai, A. S.; Bergman, R. G.; Ellman, J. A. J. Am. Chem. Soc. 2008,
130, 6316–6317. (c) Nicolaou, K. C.; Wu, T. R.; Sarlah, D.; Shaw, D. M.;
Rowcliffe, E.; Burton, D. R. J. Am. Chem. Soc. 2008, 130, 11114–11121.
(6) Other catalytic entries to [2 þ 2] cycloaddition rely on ynamide
functionality; see: (a) Li, H.; Hsung, R. P.; DeKorver, K. A.; Wei, Y. Org.
Lett. 2010, 12, 3780–3783. (b) Couty, S.; Meyer, C.; Cossy, J. Angew
Chem., Int. Ed. 2006, 45, 6726–6730 and references therein.
(7) (a) Marchand-Brynaert, J.; Ghosez, L. J. Am. Chem. Soc. 1972,
94, 2870–2872. (b) De Poortere, M.; Marchand-Brynaert, J.; Ghosez, L.
Angew. Chem., Int. Ed. Engl. 1974, 13, 267. (c) Falmagne, J.-B.; Escudero,
J.; Taleb-Sahraoui, S.; Ghosez, L. Angew. Chem., Int. Ed. Engl. 1981,
20, 879–880. (d) Houge, C.; Frisque-Hesbain, A. M.; Mockel, A.;
Ghosez, L. J. Am. Chem. Soc. 1982, 104, 2920–2921. (e) Saimoto, H.;
Houge, C.; Hesbain-Frisque, A. M.; Mockel, A.; Ghosez, L. Tetrahedron
Lett. 1983, 24, 2251–2254. (f) Markꢀo, I.; Ronsmans, B.; Hesbain-
Frisque, A. M.; Dumas, S.; Ghosez, L. J. Am. Chem. Soc. 1985,
107, 2192–2194. See also: (g) Chen, L.-Y.; Ghosez, L. Tetrahedron
Lett. 1990, 31, 4467–4470.
(23) For another unexpected regiochemical outcome in a ketene
cycloaddition reaction involving an electron-rich double bond, see:
Cagnon, J. R.; Bideau, F. L.; Marchand-Brynaert; Ghosez, L. Tetrahedron
Lett. 1997, 38, 2291–2294.
(24) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518–1520.
(25) Maegawa, T.; Akashi, A.; Sajiki, H. Synlett 2006, 1440–1442.
(26) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981,
22, 3815–3818.
(27) Sengmany, S.; Lꢀeonel, E.; Paugam, J. P.; Nꢀedꢀelec, J.-Y. Tetra-
hedron 2002, 58, 271–277.
(8) For a review, see: (a) Snider, B. B. Chem. Rev. 1988, 88, 793–811.
For an application in total synthesis relying on unconventional condi-
tions for ketene formation, see: (b) Mushti, C. S.; Kim, J.-H.; Corey, E. J.
J. Am. Chem. Soc. 2006, 128, 14050–14052.
(9) (a) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Academic Press: New York, 1969. (b) Fleming, I. Frontier
Orbitals and Organic Chemical Reactions; John Wiley & Sons: Chichester,
U.K., 1976; pp 143-147.
(10) Corey, E. J.; Desai, M. C. Tetrahedron Lett. 1985,
26, 3535–3538.
(11) Heine, H.-G.; Hartmann, W. Angew. Chem., Int. Ed. Engl. 1981,
20, 782–783.
1672
dx.doi.org/10.1021/jo102257k |J. Org. Chem. 2011, 76, 1662–1672