Intramolecular cycloadditions of cyclobutadiene with olefins

Article abstract of DOI:10.1021/ja0203162

Intramolecular cycloadditions between cyclobutadiene and olefins can provide highly functionalized cyclobutene-containing products. The outcome of the reaction depends on the nature of the tether connecting the two reactive partners in the cycloaddition.

Full text of DOI:10.1021/ja0203162

Published on Web 11/15/2002
Intramolecular Cycloadditions of Cyclobutadiene with Olefins
John Limanto,†,§ John A. Tallarico,†,| James R. Porter,†, Kelli S. Khuong,‡
K. N. Houk,*,‡ and Marc L. Snapper*,†
Contribution from the Department of Chemistry, Merkert Chemistry Center, Boston College,
2609 Beacon Street, Chestnut Hill, Massachusetts 02467, and Department of Chemistry and
Biochemistry, The UniVersity of California, Los Angeles, California 90095-1569
Received March 1, 2002. Revised Manuscript Received September 3, 2002
Abstract: Intramolecular cycloadditions between cyclobutadiene and olefins can provide highly functionalized
cyclobutene-containing products. The outcome of the reaction depends on the nature of the tether connecting
the two reactive partners in the cycloaddition. Electronically unactivated olefins attached to cyclobutadiene
through a three-atom, heteroatom-containing tether yield successfully the desired cycloadducts, whereas
the corresponding substrates without a heteroatom linkage or with a longer tether are less prone to undergo
the intramolecular cycloaddition. Calculations were used to help uncover some of the factors that influence
the course of the cycloaddition. Successful intramolecular reactions usually require either electronic activation
of the dienophile, conformational restriction of the tether, or a slower oxidation protocol. In general, a facile
intermolecular dimerization of cyclobutadiene is the major process that competes with the intramolecular
cycloaddition.
Introduction
complex [C4H4Fe(CO)3] (2, R ) H), for example, has sufficient
stability to tolerate a wide range of transformations without
Cyclobutadiene (1, R ) H) is a highly reactive, antiaromatic
species1 that undergoes rapid and facile dimerization (1 f 3,
eq 1).2 Nevertheless, cyclobutadiene has been observed at low
temperatures (i.e., 8 K, noble gas matrix),3 inferred through
Rebek’s three-phase test,4 and isolated inside Cram’s hemicar-
cerand molecular container.5 The reactivity of cyclobutadiene
can be modulated, however, through coordination to a metal
center.6 The metalloaromatic7 tricarbonylcyclobutadiene iron
disrupting the cyclobutadiene functionality, including electro-
philic aromatic substitution reactions,8 deprotonation of the
cyclobutadiene ring hydrogen(s) followed by trapping with
electrophiles,9 and Pd(0)-catalyzed C-C and C-N bond form-
ing reactions.10 In general, the iron tricarbonyl complexes
survive acidic, basic, and reducing environments, as well as
some mild oxidizing conditions. Treatment of these complexes
with cerium ammonium nitrate (CAN), however, can oxidize
the iron and liberate free cyclobutadiene.11 To a lesser extent,
FeCl3 and Pb(OAc)4 have also been used for this purpose.11
When cyclobutadiene is generated in the presence of olefins
and dienes, an intermolecular cycloaddition can lead to a variety
of cyclobutene-containing adducts (eqs 2 and 3).11
* To whom correspondence should be addressed. E-mail: marc.
snapper@bc.edu and houk@chem.ucla.edu.
† Boston College.
‡ The University of California, Los Angeles.
§ Current address: Merck Research Laboratories, P.O. Box 2000, R80Y-
240, Rahway, NJ 07065-0900.
| Current address: Harvard Institute for Chemistry and Cell Biology,
Harvard Medical School, 25 Shattuck St., Boston, MA 02115.
Current address: Infinity Pharmaceuticals, 650 Albany St., Boston,
MA 02118.
(1) (a) Maier, G. Angew. Chem. 1974, 86, 491. (b) Deniz, A. A.; Peters, K. S.;
Snyder, G. J. Science 1999, 286, 1119. (c) Bally, T.; Masamune, S.
Tetrahedron 1980, 36, 343.
(2) Li, Y.; Houk, K. N. J. Am. Chem. Soc. 1996, 118, 880 and references
therein.
(3) (a) Chapman, O. L.; McIntosh, C. L.; Pacansky, J. J. Am. Chem. Soc. 1973,
95, 614. (b) Chapman, O. L.; De La Cruz, D.; Roth, R.; Pacansky, J. J.
Am. Chem. Soc. 1973, 95, 1337. (c) Krantz, A.; Lin, C. Y.; Newton, M. D.
J. Am. Chem. Soc. 1973, 95, 2744.
(4) (a) Rebek, J., Jr.; Gavina, F. J. Am. Chem. Soc. 1974, 96, 7112. (b) Rebek,
J., Jr.; Gavina, F. J. Am. Chem. Soc. 1975, 97, 3453.
(5) Cyclobutadiene was generated by irradiation of R-pyrone trapped inside
the hemicarcerand host. (a) Cram, D. J. Nature 1992, 356, 29. (b) Cram,
D. J.; Tanner, M. E.; Thomas, R. Angew. Chem., Int. Ed. Engl. 1991, 30,
1024.
We envisioned that an intramolecular reaction between
cyclobutadiene and olefins could offer substantial control over
(6) (a) Emerson, G. F.; Watts, L.; Pettit, R. J. Am. Chem. Soc. 1965, 87, 131.
(b) Rosenblum, M.; Gatsonis, C. J. Am. Chem. Soc. 1967, 89, 5074. (c)
Paquette, L. A.; Wise, L. D. J. Am. Chem. Soc. 1967, 89, 6659. (d) Grubbs,
R. H. J. Am. Chem. Soc. 1970, 92, 6993. (e) Grubbs, R. H.; Grey, R. A. J.
Chem. Soc., Chem. Commun. 1973, 76. (f) Pettit, R. J. Organomet. Chem.
1975, 100, 205. (g) Fitzpatrick, J. D.; Watts, L.; Emerson, G. F.; Pettit, R.
J. Am. Chem. Soc. 1965, 87, 3254.
(8) For reviews on cyclobutadienylmetal complexes, see: Efraty, A. Chem.
ReV. 1977, 77, 691 and references therein.
(9) Bunz, U. Organometallics 1993, 12, 3594 and references therein.
(10) For Stille type couplings, see: (a) Bunz, U. H. F.; Enkelmann, V. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1653. (b) Wiegelmann, J. E. C.; Bunz, U.
H. F.; Schiel, P. Organometallics 1994, 13, 4649. For C-N bond forming
reactions, see: Wiegelmann-Kreiter, J. E. C.; Enkelmann, V.; Bunz, U. H.
F. Chem. Ber. 1995, 128, 1055.
(7) Bursten, B. E.; Fenske, R. F. Inorg. Chem. 1979, 18, 1760.
9
14748
J. AM. CHEM. SOC. 2002, 124, 14748-14758
10.1021/ja0203162 CCC: $22.00 © 2002 American Chemical Society

Cycloadditions of Cyclobutadiene with Olefins
A R T I C L E S
chemo-, regio-, and stereoselectivity issues in these cycloaddi-
tions. Furthermore, the intramolecular variants could yield a
unique and rapid access into highly functionalized, cyclobutene-
containing cycloadducts. The strain associated with these
cycloadducts should provide novel opportunities for improved
access to several types of challenging synthetic targets.12
In support of the intramolecular reaction, Grubbs and co-
workers have shown that cyclobutadiene reacts with tethered
alkynes to yield aromatic systems (eqs 4 and 5).13 Presumably,
these reactions proceed through the desired cyclobutene-
containing Dewar benzene-intermediates (4 and 6), but rearrange
upon mild heating during the workup to the observed aromatic
products.
or the secondary hydroxyl or ether functionality adjacent to the
cyclobutadiene moiety in compound 10 inhibits the reaction,
substrate 11 was prepared to address this reactivity question.
Oxidation of complex 11 generates cycloadduct 13 in 74% yield
as a 3:1 mixture of diastereoisomers (eq 9). Evidently, this
example suggests that the intramolecular cycloaddition with
CAN tolerates a substituent adjacent to cyclobutadiene and,
moreover, requires the ether linkage to proceed.
Our preliminary studies have demonstrated the feasibility of
intramolecular reactions of cyclobutadienes with olefins.14
Moreover, we have shown that the resulting highly strained
cycloadducts provide unique and effective entries into seven-
and eight-membered ring systems.15 The concise nine-step
synthesis of asteriscanolide, featuring an intramolecular cyclo-
butadiene cycloaddition, is illustrative (eq 6).12
Examining and understanding the subtle, yet critical effect
of tether composition in the intramolecular cycloaddition is
necessary to exploit fully the utility of the methodology. Along
these lines, our studies of the factors that influence the
intramolecular cycloadditions of cyclobutadienes with olefins
are reported herein. In particular, the influence of tether length
and substituents, as well as the stereochemistry and electronic
properties of the cycloaddition partners in the intramolecular
cycloadditions, are described. In conjunction with experimental
studies, theoretical density functional calculations using B3LYP/
6-31G(d) were performed to quantify specific effects of the
tether and olefin substitution on the relative ease of intra- and
intermolecular cycloadditions of cyclobutadiene.
While offering effective access to functionalized cyclobutenes,
our results indicate that the success of these cycloadditions is
dependent on the nature of the tether connecting the two reactive
cycloaddition partners. For example, when 8 is treated with
CAN, cycloadduct 9 is generated in 85% yield (eq 7); however,
the identical oxidation of complex 10 yields only cyclobutadiene
dimers without any of the desired intramolecular cycloaddition
(eq 8). Because it was unclear whether the ether tether of
substrate 8 serves to facilitate the intramolecular cycloaddition
Results and Discussion
Oxidative Decomplexation of (CO)3Fe-Cyclobutadiene
Complexes. The major competing side reaction in the intramo-
lecular cycloadditions is the dimerization of cyclobutadiene (eq
1). In general, minimizing the concentration of free cyclobuta-
diene will serve to favor the intramolecular process over the
facile intermolecular side reaction. For substrates predisposed
toward the intramolecular pathway, a cerium ammonium nitrate
(CAN) oxidation of the iron complex provides the desired
cycloadducts rapidly in acceptable yields (5 equiv of CAN,
acetone, [1-2 mM] iron complex, room temperature, 15 min).
For substrates less prone to undergo the intramolecular reaction,
a slower trimethylamine-N-oxide (TMAO) oxidation affords
better yields of the desired cycloadducts (8-20 equiv of TMAO,
acetone, [2-20 mM] iron complex, reflux, 6-24 h). In either
case, excess oxidant is usually required for complete consump-
tion of the starting complex. Even under high dilution and slow
oxidation conditions, particularly poor cycloaddition substrates
(11) (a) Watts, L.; Fitzpatrick, J. D.; Pettit, R. J. Am. Chem. Soc. 1965, 87,
3253. (b) Grubbs, R. H.; Grey, R. A. J. Am. Chem. Soc. 1973, 95, 5765.
(c) Schmidt, E. K. G. Angew. Chem., Int. Ed. Engl. 1973, 12, 777. (d)
Schmidt, E. K. G. Chem. Ber. 1975, 108, 1599. (e) For a discussion of
cycloaddition reactions between cyclobutadiene and unsaturated compounds
coordinated to the iron metal, see: Ward, J. S.; Pettit, R. J. Am. Chem.
Soc. 1971, 93, 262.
(12) Limanto, J.; Snapper, M. L. J. Am. Chem. Soc. 2000, 122, 8071.
(13) (a) Grubbs, R. H.; Pancoast, T. A.; Grey, R. A. Tetrahedron Lett. 1974,
28, 2425. (b) Grubbs, R. H.; Pancoast, T. A. J. Am. Chem. Soc. 1977, 99,
2382.
(14) Tallarico, J. A.; Randall, M. L.; Snapper, M. L. J. Am. Chem. Soc. 1996,
118, 9196.
(15) (a) Snapper, M. L.; Tallarico, J. A.; Randall, M. L. J. Am. Chem. Soc.
1997, 119, 1478. (b) Randall, M. L.; Lo, P. C.-K.; Bonitatebus, P. J., Jr.;
Snapper, M. L. J. Am. Chem. Soc. 1999, 121, 4534. (c) Lo, P. C.-K.;
Snapper, M. L. Org. Lett. 2001, 3, 2819. (d) Deak, H. L.; Stokes, S. S.;
Snapper, M. L. J. Am. Chem. Soc. 2001, 123, 5152.
9
J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002 14749

A R T I C L E S
Limanto et al.
Table 2. Intramolecular Cycloadditions with Three-Atom
Table 1. Cycloadditions of Chiral Cyclobutadiene Complexes
Heteroatom-Containing Tethers
entry
SM (% ee)
conditionsa
yield
product ratio (ee)
1
2
3
4
14 (35%)
14 (35%)
18 (35%)
18 (35%)
CAN
TMAO
CAN
66%
55%
48%
54%
16:17 ) 16 (0%):1 (0%)
16:17 ) 14 (0%):1 (0%)
20:21 ) 1 (0%):1 (0%)
20:21 ) 1 (0%):1 (0%)
TMAO
a CAN (room temperature, 2 mM, 15 min); TMAO (56 °C, 20 mM,
6 h).
(see below) are unable to compete with the rapid intermolecular
dimerization. Nevertheless, the TMAO oxidative conditions
generally yield more favorable intramolecular cycloadditions
for a broader range of substrates.
Clearly, the more that is understood about the mechanism of
the intramolecular cycloaddition, the better the reaction outcome
can be predicted. One longstanding concern has been the role
of iron in the cycloaddition. Previous studies on intermolecular
cycloadditions with olefins and intramolecular cycloadditions
with alkynes have indicated that the iron does not play a
stereochemical role in the CAN-promoted cycloadditions.13
As shown in Table 1, our findings support the conclusion
that both the CAN- and the TMAO-promoted intramolecular
cycloadditions of cyclobutadienes with olefins likely proceed
through liberation of a free cyclobutadiene, without involvement
of a chiral iron complex. Specifically, when enantiomerically
enriched disubstituted complexes 14 and 18 (both 35% ee) were
treated with either CAN or TMAO under the usual reaction
conditions, a regioisomeric mixture of racemic cycloadducts was
obtained (16:17, 20:21). These results suggest the intramolecular
cycloaddition involves an achiral cyclobutadiene species such
as cyclobutadiene 15 (i.e., not 19).
Three-Atom Heteroatom-Containing Tethers. Our initial
investigation focused on substrates containing a three-atom
tether between the cyclobutadiene and the olefinic partner. Table
2 summarizes the results of these cycloadditions with an ether-
or amine-containing linkage. Electronic activation of the ole-
fins is not necessary for these intramolecular cycloadditions
to proceed efficiently. The transformation yields the cyclo-
butene-contained cycloadducts in moderate to excellent yields
(20-93%). Entries 2 and 3 indicate that the reaction proceeds
stereospecifically, results that are consistent with a concerted
cycloaddition of cyclobutadiene reacting from the rectangular
singlet ground state.16
a Reaction conditions: method A, CAN (5 equiv), acetone (1-2 mM),
room temperature, 15 min; method B, CAN (10 equiv), DMF (1 mM), 80
°C, 5 min; method C, TMAO (10 equiv), acetone (45 mM), reflux, 12 h.
b Isolated yields.
substrate 32 (entry 6) was prepared to study the possibility of
a competing intramolecular “[5 + 2]” cycloaddition.17 Under
the typical reaction conditions, however, the “[4 + 2]” (cy-
clobutadiene functioning as a diene) cycloadduct 33 was
obtained exclusively in 90% yield. The possibility of employing
an allene as a cycloaddition partner was also investigated;
subjection of substrate 34 to CAN afforded a mixture of
diastereomeric cyclobutene cycloadducts 35 (4.5:1 dr), albeit
in only 35% yield (entry 7).
The disubstituted complex 36 (entry 8), which was used in
the total synthesis of (+)-asteriscanolide,12 was less prone to
undergo the desired intramolecular cycloaddition under the
typical CAN-promoted reaction conditions, presumably due to
additional ring strain. The yield for this cycloaddition was
improved when the reaction was carried out in DMF at 80 °C
(20 f 49%). To improve further the reaction, several other
oxidants were screened. Trimethylamine N-oxide (TMAO) was
found to promote an efficient intramolecular cycloaddition of
complex 36 at higher concentrations (i.e., 45 mM), without
The cycloaddition clearly tolerates sterically encumbered
olefins, as illustrated in entry 5. In this example, two adjacent
quaternary carbon centers (C1, C2) are generated simultaneously
in the newly formed cyclobutane 31. The kojic acid-derived
(16) (a) Dewar, M. J. S.; Gleicher, G. J. J. Am. Chem. Soc. 1965, 87, 3255. (b)
Reeves, P.; Henery, J.; Pettit, R. J. Am. Chem. Soc. 1969, 91, 5888. (c)
Reeves, P.; Devon, T.; Pettit, R. J. Am. Chem. Soc. 1969, 91, 5890. (d)
Balkova, A.; Bartlett, R. J. J. Chem. Phys. 1994, 101, 8972.
(17) (a) Rodriguez, J. R.; Rumbo, A.; Castedo, L.; Mascarenas, J. L. J. Org.
Chem. 1999, 64, 4560. (b) Wender, P. A.; Mascarenas, J. L. Tetrahedron
Lett. 1992, 33, 2115. (c) McBride, B. J.; Garst, M. E. Tetrahedron 1993,
49, 2839.
9
14750 J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002

Cycloadditions of Cyclobutadiene with Olefins
A R T I C L E S
Table 4. Intramolecular Cycloadditions with Three-Atom
Table 3. Intramolecular Cycloadditions of Carboxylate Derivatives
All-Carbon Tethers
a Reaction conditions: method A, CAN (5 equiv), acetone (1-2 mM),
room temperature, 15 min; method B, CAN (5 equiv), CH3CN (1 mM),
room temperature, 15 min; method C, CAN (10 equiv), DMSO:H2O (1:1),
room temperature, 15 min; method D, TMAO (8-10 equiv), acetone (20
mM), reflux, 12 h. b Isolated yields. c 20-30% starting material recovered,
as well as 65% trans-cinnamyl alcohol.
formation of cyclobutadiene-dimer byproducts. The reaction was
carried out in refluxing acetone for 6-12 h with excess oxidant
to give the desired cycloadduct 37 in 64% yield. The fact that
the transformation can be carried out at a higher concentration
and requires longer reaction times to consume starting materials
suggests that the reactive cyclobutadiene is generated more
slowly as compared to under the CAN conditions, and, hence,
self-dimerization of cyclobutadiene is kept to a minimum.
To expand the scope of the cycloaddition methodology,
nitrogen-tethered intramolecular cycloadditions were examined.
Concerned that free amines may undergo oxidation, the sul-
fonamide complex 38 was prepared for the initial studies.
Subjection of 38 to the CAN-promoted reaction conditions
afforded the desired cyclobutene cycloadduct 39 in 93% yield
(entry 9, Table 2). Under similar conditions, the N,N-diallyl-
amine complex 40 also afforded the corresponding cyclobutene
41 as the major product in 70% yield (entry 10). Again,
electronic activation of the dienophile does not appear to be
essential for the cycloaddition. These results suggest that
oxidation of the amine is either slower than oxidative liberation
of the cyclobutadiene ligand or the nitrogen is protected under
the moderately acidic CAN reaction conditions.
We next investigated intramolecular cycloadditions of iron
complexes linked to olefins through a carboxylate-containing
tether (Table 3). Subjection of ester 42 to the typical CAN
reaction conditions afforded the desired lactone 43 in 55% yield
(entry 1). The yield of the reaction appears to be dependent on
solvent, with the polar mixture of DMSO:H2O (1:1) giving the
best results (88% yield).18 When the ester functionality is
transposed as in complex 44, however, none of the desired
cyclobutene lactone 45 was obtained with either the CAN- or
the TMAO-based oxidations. Moreover, incomplete consump-
tion of starting material was encountered (70-80% conversion),
and trans-cinnamyl alcohol (∼65%) was recovered upon
aqueous workup, suggesting that hydrolysis of the ester occurred
a Reaction conditions: method A, CAN (5 equiv), acetone (1-10 mM),
room temperature, 15 min; method B, TMAO (10 equiv), refluxing acetone
(20 mM), 6-8 h; method C, CAN (10 equiv), refluxing acetone:CH2Cl2
(7:1, 1 mM), 3 min. b Isolated yields.
under the reaction conditions. The corresponding N,N-diallyl-
amide complex 46, on the other hand, afforded cyclobutene 47
in modest yield under either CAN or TMAO reaction conditions
(entry 3, Table 3).
Three-Atom Tethers, All-Carbon Linkage. We also inves-
tigated the feasibility of intramolecular cycloadditions of
cyclobutadiene complexes bearing all-carbon tethers. While
heteroatom-containing three-atom tether substrates undergo
efficient CAN-promoted cycloadditions, the all-carbon tether
substrates, such as 48 and 51 (entries 1 and 2, Table 4, as well
as 10, eq 7), fail to undergo an efficient intramolecular
cycloaddition under similar reaction conditions. These findings
suggest that under fast oxidative decomplexation of the iron
complex by CAN, intermolecular dimerization of the cyclo-
butadiene competes with the intramolecular process even under
high dilution reaction conditions. When the reaction was carried
out using a slower oxidant TMAO, however, the desired
cycloadducts were obtained as a mixture of diastereomers in
49% yield (49:50 ) 2.9:1).
There are numerous differences between the heteroatom- and
the all-carbon-containing tethers that could account for this
reactivity difference. The ability of the oxygen in the tether to
(18) Jung, M. E.; Gervay, J. J. Am. Chem. Soc. 1989, 111, 5469.
9
J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002 14751

A R T I C L E S
Limanto et al.
Table 5. Cycloadditions with Enone Substrates
coordinate to Lewis acidic species may influence the course of
the reaction. However, as will be described in the theoretical
section, calculations on the corresponding transition states
indicate that the shorter C-O bond lengths and compressed
C-O-C bond angles may play a role favoring the intramo-
lecular cycloaddition process.19
One way to enhance the intramolecular Diels-Alder reaction
is to electronically activate the dienophile (i.e., lower LUMO)
through conjugation with a phenyl group. As shown in entry 3
of Table 4, treatment of the trans-phenyl substrate 54 with CAN
or TMAO affords the desired cycloadduct 55 in 53-55% yield,
with no indication of intermolecular side products. The cis-
phenyl isomer 56, however, failed to undergo the desired
intramolecular cycloaddition under CAN reaction conditions.
With the slower oxidant TMAO, the corresponding cyclobutene
cycloadduct 57 was obtained in 53% yield (entry 4). These
findings suggest that secondary orbital interactions between the
cyclobutadiene and the dienophile, enjoyed in the trans- but
not the cis-isomers, may be helpful, but are not required for a
favorable intramolecular process. In addition, while electronic
activation of trans-olefin substrates (48 or 54) is not essential
for a successful intramolecular cycloaddition, some activation
is necessary for the corresponding cis-olefin isomers (complex
51 vs 56).
Changing the substituent on the olefin from a phenyl to an
ester group provides additional electronic activation of the olefin
(i.e., further lowers LUMO) and increases the efficiency of the
intramolecular cycloaddition. For example, treatment of trans-
methyl ester complex 58 with CAN (2 mM, room temperature)
affords the diastereomeric cycloadducts (59:60 ) 2.3:1) in 92%
yield (entry 5, Table 4). A similar yield was obtained when
this reaction was carried out at higher concentrations. In
comparison, oxidative decomplexation of the corresponding cis-
methyl ester 61 under high dilution (CAN, 2 mM, room
temperature) affords the diastereomeric cycloadducts in a
somewhat lower yield (66%, 62:63/2.5:1, entry 6). These results
suggest that, while secondary orbital interactions between two
reacting partners in trans-isomers could favor the intramolecular
cyclizations, electronic activation of either the cis- or the trans-
dienophile is an overriding factor in favoring the intramolecular
reaction.
Another way to favor the intramolecular cycloaddition relative
to competitive intermolecular processes is to limit the degrees
of freedom of the tether. The conformationally restricted and
electronically activated complex 64 (entry 7, Table 4), however,
affords cyclobutene 65 in only 34% yield in refluxing acetone:
CH2Cl2 [7:1] (CAN, 1 mM, 3 min). These results (as well as
those reported below) indicate a limitation on using functionality
within the connecting tether to actiVate the olefin toward
cycloaddition.
a Reaction conditions: method A, CAN, acetone (1 mM), room tem-
perature; method B, TMAO, acetone (5-20 mM), reflux; method C, CAN,
MeOH (1 mM), room temperature. b Isolated yields. c Overall yield from
the corresponding enone C4H4Fe(CO)3 substrate; the number in parentheses
represents percent conversion during ketalization of enones 66, 68, and 70.
d Isolated as the corresponding ketones 67, 69, and 71.
or 70 under typical (acetone, 1 mM, room temperature) or more
vigorous (refluxing acetone, acetonitrile, dioxane) CAN reaction
conditions did not afford any of the desired ketone cycloadducts
(entries 1-3, method A, Table 5); in all cases, only cyclobuta-
diene dimers were detected.
The failure of these CAN-promoted intramolecular cycload-
ditions can be attributed perhaps to disfavorable torsional and
angle strains associated with the tethered carbonyl in the
transition state of the requisite rotamer for the cycloaddition.
A way to resolve this limitation is to convert the ketone into an
sp3-hybridized functional group. In this regard, we prepared the
dimethyl ketal complexes 74-76. Not only would the dimethyl
ketal help to release any strain associated with the tethered
carbonyl, it would also increase the population of the reactive
rotamer through Thorpe-Ingold and reactive rotamer effects.18
Subjection of the dimethyl ketal complexes 74-76 to the CAN-
promoted cycloaddition conditions (MeOH, 1 mM, room
temperature) indeed afforded the corresponding dimethyl ketal
cycloadducts, which, with the exception of the phenyl cycload-
duct derived from 74, were difficult to isolate without some
hydrolysis of the ketal functionality. In practice, these adducts
were subjected to a mild deprotection protocol (SiO2/10%
aqueous oxalic acid, CH2Cl2, room temperature), yielding the
corresponding ketone cycloadducts 67, 69, and 71 in 20-70%
yield for the three steps (entries 5-7, Table 5). It is interesting
to note that under these deprotection conditions, no rearrange-
ments associated with carbonium ion character adjacent to the
cyclobutane ring were observed.
As mentioned, substrates with all-carbon, three-atom tethers
that possess electronic activation of the dienophile external to
the connecting tether undergo a facile intramolecular cycload-
dition (entries 5 and 6, Table 4), especially under the faster
oxidative decomplexation conditions of CAN. Given these
observations, we investigated the feasibility of cycloadditions
involving electronically activated enone substrates, where the
carbonyl functionality is part of the tether (i.e., internal with
respect to the olefin, Table 5). Subjection of complexes 66, 68,
(19) Jung, M. E.; Kiankarimi, M. J. Org. Chem. 1998, 63, 2968.
9
14752 J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002

Cycloadditions of Cyclobutadiene with Olefins
A R T I C L E S
Table 6. Intramolecular Cycloadditions with Four-Atom Ethereal
Tethers
Along the same lines, a gem-dialkyl substituent on the tether
should increase the population of the reactive conformer for
the cycloaddition. In this regard, while phenyl enone substrate
66 failed to give the desired cycloadduct under CAN-promoted
reaction conditions, the dimethyl phenyl enone substrate 77
afforded the corresponding ketone adduct 78 in 83% yield (entry
8, Table 5).
When an additional electron-withdrawing group (-CO2Me)
is present on the dienophile, such as in complex 72, the
corresponding cyclobutene cycloadduct 73 is obtained in high
yield (91%, entry 4). When the slower oxidant TMAO is used
in refluxing acetone, enone complexes 66, 68, and 70 underwent
the desired [4 + 2] cycloadditions, affording the corresponding
cyclobutene cycloadducts 67, 69, and 71 in 27-70% yields
(entries 1-3, method B, Table 5). Substrates bearing a bulkier
substituent on the olefin, such as the tert-butyl group in complex
70, required higher dilutions (5 mM) and longer reaction times,
and only a 27% yield of cycloadduct 71, together with
significant amounts of the corresponding cyclobutadiene dimers,
was obtained (entry 3, method B, Table 5).
In general, the success of intramolecular cycloadditions of
three-atom-tethered iron cyclobutadiene complexes depends on
several factors. Electronic activation of the dienophile, favorable
secondary orbital interactions (trans-olefin), restrictions in the
degrees of freedom of the connecting tether, and/or incorporation
of a geminal substitution along the tether all serve to facilitate
the intramolecular cycloaddition when the fast oxidant CAN is
employed. In cases lacking some of these favorable elements,
the slower oxidant TMAO may be used to accomplish the
desired transformation.
a Reaction conditions: method A, CAN, acetone (1 mM), room tem-
perature, 15 min; method B, TMAO (20 equiv), refluxing acetone (1 mM),
24 h; method C, TMAO (8 equiv), refluxing acetone (20 mM), 6 h. b Isolated
yields.
Four-Atom Heteroatom-Containing Tethers. Inserting an
additional atom in the tether to generate a four-atom connection
between the cycloaddition partners increases the degrees of
freedom and number of possible nonproductive rotamers, which
can result generally in a less efficient intramolecular process.
Under typical CAN oxidation conditions, complex 79 failed to
undergo the desired cycloaddition (entry 1, method A, Table
6); only dimers were obtained as judged by the 1H NMR
spectrum of the reaction mixture. On the other hand, switching
to the slower TMAO oxidation leads to the formation of some
of the corresponding cyclobutene cycloadduct 80 (entry 1,
method B). Optimization of the reaction conditions provides a
modest 22% yield of cycloadduct 80. When the dienophile is
activated, such as in trans-phenyl complex 81, the cycloaddition
proceeds more smoothly, affording the corresponding cycload-
duct 82 under either CAN (63%) or TMAO (55%) reaction
conditions (entry 2). More activated dienophiles, such as in
trans-methyl ester complex 83, also provide the desired cy-
clobutene 84 in a 65% yield (entry 3) under CAN-promoted
reaction conditions. These results suggest that electronic activa-
tion of the reacting dienophile in the four-atom tether substrates
is essential for a successful and synthetically useful intramo-
lecular cycloaddition.
mM, refluxing acetone, 24 h), complex 85 afforded a 17% yield
of cycloadduct 86. Interestingly, under these reaction conditions,
the cis-methyl ester substrate 87 gave two cycloadducts (88:
89/3:1; entry 5) in a modest 21% yield. The major cycloadduct
(88) was the product of a Type II intramolecular cycloaddition.20
A role for the beneficial secondary orbital interaction is also
consistent with the trans-olefinic substrates undergoing the
intramolecular cycloaddition in higher yields as compared with
the corresponding cis-olefin-containing compounds. Neverthe-
less, there may be additional unfavorable steric interactions in
the transition state of the cis-isomers that may also slow the
intramolecular reaction relative to competitive intermolecular
processes. Complex 90, which has both a cis- and a trans-
substituent on the pendent olefin, was prepared to help dif-
ferentiate between these effects. If it is the favorable electronic
interactions of the trans-substituent that lead to a successful
intramolecular cycloaddition, then 90 should cyclize smoothly.
Alternatively, if the cis-substituent suffers unfavorable steric
interactions in the transition state, then the intramolecular
reaction will suffer relative to competitive intermolecular
processes. Subjecting complex 90 to the CAN cycloaddition
conditions did not result in the formation of any of the desired
cyclobutene cycloadduct; again, only the corresponding cyclo-
butadiene dimers were obtained. With TMAO (1 mM, refluxing
acetone, 24 h), cyclobutene cycloadduct 91 was obtained, albeit
in only 30% yield (entry 6, Table 6). These findings suggest
that while activation (i.e., primary and secondary orbital
While activated trans-olefinic substrates afford the cy-
clobutene adducts in reasonable yields, the cis-isomers undergo
a much less efficient intramolecular cycloaddition. For example,
subjection of cis-phenyl substrate 85 (entry 4) or cis-methyl
ester 87 (entry 5) to CAN did not result in the formation of the
intramolecular cycloadducts; in each case, only dimers were
obtained. Under the TMAO-promoted reaction conditions (1
(20) Bear, B. R.; Sparks, S. M.; Shea, K. J. Angew. Chem., Int. Ed. 2001, 40,
821 and references therein.
9
J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002 14753

A R T I C L E S
Limanto et al.
Table 8. Transformations Used To Examine Transition States
Table 7. Cycloadditions with Four-Atom All-Carbon Tethers
a Reaction conditions: method A, CAN (5 equiv), acetone (1 mM), room
temperature. b Isolated yields. c The yield represents a two-step sequence:
Knovenagel condensation of the corresponding iron aldehyde with Mel-
drum’s acid followed by CAN-promoted cycloaddition.
interactions) of four-atom tether substrates is essential for
efficient cycloadditions, steric factors (i.e., nonbonded interac-
tions) also play a significant role in dictating the outcome of
the reaction.
Four-Atom Tethers, All-Carbon Linkage. Intramolecular
cycloadditions between cyclobutadiene and olefins connected
through an all-carbon four-atom tether were also examined
(Table 7). All studies involving these substrates were performed
using CAN to promote the oxidative decomplexations and
intramolecular cycloadditions.
These studies suggest that the intramolecular cycloadditions
of all-carbon four-atom-tethered substrates require strongly
activated dienophiles to compete favorably with the intermo-
lecular cyclobutadiene dimerization process. In addition, unlike
the three-atom-tethered examples, the incorporation of a rigid
group in the four-atom tether, such as a phenyl ring, which
decreases conformational freedom, can improve the efficacy of
the intramolecular process.
Theoretical Calculations. Geometry optimizations were
performed for all reactants, transition structures, and prod-
ucts using RB3LYP/6-31G(d) as implemented in Gaussian
98.21 When an unrestricted wave function was found to be
more stable, a reoptimization was performed using UB3LYP/
6-31G(d). All stationary points were characterized with fre-
quency calculations.
As shown in Table 8, the calculations focused on four pairs
of model systems, each corresponding to a set of specific
experimental reactions. In each pair, the first entries (A, C, E,
G) model a reaction in which the intramolecular cycloaddition
was successful, whereas the second systems (B, D, F, H) model
an apparently similar reaction in which intermolecular dimer-
ization, instead of intramolecular cycloaddition, was the major
process. Entries A and B specifically address the reactivity of
an ether versus an all-carbon tether. Systems C and D examine
the effect of transposing the three-atom carboxylate tether.
Systems E and F probe the consequence of changing an sp3-
Unlike the electronically activated three-atom tether methyl
ester 58 (entry 4, Table 4), the corresponding four-atom tether
substrate 92 afforded only a trace of the expected cycloadduct
93 under identical reaction conditions (entry 1, Table 7).
Resorting to refluxing acetone or CH3CN did not improve the
intramolecular cycloaddition. When the tether was rigidified by
incorporation of a phenyl ring as in complex 94, however, the
desired cycloadduct 95 was obtained in 88% yield (entry 2).
Evidently, restricting the degrees of freedom in the four-atom
tether favors the intramolecular process, presumably by increas-
ing the relative population of the reactive conformer.
The failure of methyl ester 92 to undergo efficient intra-
molecular cycloaddition prompted us to investigate further
the effects of electronic activation of the tethered olefin. We
reason that structural modifications to the dienophile that
decrease the HOMO-LUMO gap of the cycloaddition part-
ners could outweigh the extra entropic freedom of the longer
tether. On the basis of this assumption, substrate 96 was pre-
pared via a Knovenagel condensation of the corresponding
four-carbon-tethered aldehyde. Because of the reactivity of 96
toward Michael additions, the compound was subjected to the
typical CAN-promoted cycloaddition conditions without rigor-
ous purification. The reaction proceeded to give the corre-
sponding cyclobutene cycloadduct 97 as the only isolable
monomeric product in 33% yield for the two steps (entry 3,
Table 7).
(21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.9;
Gaussian, Inc.: Pittsburgh, PA, 1998.
9
14754 J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002

Cycloadditions of Cyclobutadiene with Olefins
A R T I C L E S
Figure 2. Transition state geometries for intermolecular cycloadditions
between substituted cyclobutadienes and olefins.
Figure 1. Calculated transition states for reactions A and B.
hybridized carbon in the tether to an sp2 carbon. Systems G
and H study the influence of E versus Z stereochemistry of four-
atom-tethered alkenes.
Because cyclobutadiene dimerization is the major competing
side reaction, several other model systems were examined
computationally. Using systems I through K, we established
the influence of alkyl and carboxylate substituents.
Figure 3. Calculated transition states for reactions C and D.
force along the series. For example, the ∆Grxn changes from
-42.1 kcal/mol (cyclobutadiene + ethylene) to -39.4, -37.8,
and finally -33.8 kcal/mol (methylcyclobutadiene + propene).
The second additional set of calculations involved a simple
comparison of dimethyl ether and propane (see inset Figure 1)
to the ether tether and the three-carbon tether. The optimized
geometry of dimethyl ether is very similar to that of the ether
tether in A-TS. On the other hand, the optimized geometry of
propane deviates from that of the all-carbon tether in B-TS. It
appears that B-TS is more asynchronous and higher in energy
than A-TS because the natural length of the propyl tether is
longer than the natural length of the ether tether. This is partially
compensated for by a compression of the abc in B-TS, but
only at the expense of angle strain that destabilizes the TS. Note
the difference in tether lengths arises from C-O and C-C bond
length differences.
The transition structures for intramolecular [4 + 2] cycload-
dition for systems A and B are shown in Figure 1. Two sets of
energies (relative to the reactant) are provided for each transition
structure. The first energy value (∆E + ZPEq) is the calculated
electronic energy with zero point energy correction. The second
energy value (∆Gq) is the calculated free energy at 25 °C. Free
energies were computed for the conversion of global minimum
to transition state.
A comparison of A-TS and B-TS reveals that the intramo-
lecular cycloaddition is more favorable with the ether tether as
compared to the all-carbon tether (∆∆Gq ) 4.4 kcal/mol). A-TS
is more synchronous than B-TS, and the incipient five-
membered ring is more fully formed in A-TS. To evaluate the
structural features of these two transition states, two additional
sets of calculations were performed.
First, the ideal bond lengths of a cyclobutadiene-alkene [4
+ 2] transition structure were determined in the absence of a
tether. Examples of such transition structures are shown in Fig-
ure 2. In the case of cyclobutadiene plus ethylene, a concerted,
synchronous TS is found.22 The bond lengths clearly illustrate
that this is a prototypical [4 + 2] transition structure, and not a
[2 + 2] transition structure. Substitution of a methyl group on
either the alkene or the cyclobutadiene leads to an increase in
asynchronicity and activation energy. Methylcyclobutadiene plus
propene has the most asynchronous TS, and its electronic
activation barrier (∆E + ZPEq ) 14.7 kcal/mol) is higher than
either of the intramolecular cases A-TS or B-TS (6.2 and 10.9
kcal/mol, respectively). Methyl substitution has a significant
effect on the activation barrier of the intermolecular cycload-
dition and is partly due to a decrease in thermodynamic driving
The above calculations corroborate experimental results
obtained for reactions of 8 and 10. Assuming that the rate of
dimerization is approximately equal for both systems, we expect
a reactant with an ether tether would produce more of the desired
intramolecular adduct.
Figure 3 compares three-atom ester tethers connected in two
ways. The C-TS and D-TS have very similar energies (∆∆Gq
) 0.4 kcal/mol) with C-TS being slightly favored. These two
transition structures are significantly higher in energy than either
A-TS or B-TS; this is counterintuitive because we generally
expect electron-withdrawing groups (such as an ester) to activate
a [4 + 2] cycloaddition. However, as shown in the box on the
right of Figure 3, the ester tether is too short to connect the
(22) A stepwise, diradical mechanism was also studied. The rate-limiting step
for the diradical pathway was the addition of ethylene to cyclobutadiene
to form a diradical intermediate. This transition structure had the following
energy using UB3LYP/6-31G(d): ∆E + ZPEq ) 12.0 kcal/mol, ∆Gq
22.5 kcal/mol.
)
9
J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002 14755

A R T I C L E S
Limanto et al.
Figure 4. Calculated transition states for reactions E and F.
reacting groups without strain, and the ester is forced into the
unfavorable S-cis conformation in the TS.23
Figure 5. Calculated transition states for reactions G and H.
The similar activation energies for C and D appear to
contradict the experimental results in which a carboxylate-
tethered compound 42 gives a high yield of desired product
but a carboxylate-tethered compound 44 gives none of the
desired product. It is possible that the conjugation of the
cyclobutadiene to the carbonyl retards the oxidation process;
consequently, ester hydrolysis occurs faster than the intramo-
lecular cycloaddition. The cyclobutadiene-Fe(CO)3 might also
activate the ester toward nucleophilic attack. In addition, the
phenyl substituent on the olefin may also influence the relative
reactivity of this system. These calculations show, nevertheless,
that there is no inherent difference in cycloaddition reactivity
between C and D; the ester tethers increase the activation energy
substantially because they are too short to be accommodated
without strain in the transition state, and the syn-conformer is
unfavorable.
Transition structures E-TS and F-TS illustrate clearly that
an sp3-hybridized carbon of a ketal can be incorporated more
easily into the three-atom tether than an sp2-hybridized carbon
of a ketone (Figure 4). The activation energy for E-TS, ∆Gq )
10.4 kcal/mol, and the lengths of the forming σ bonds are similar
to those of B-TS, which is expected because both tethers are
composed of three sp3 carbons. Replacing the hydrate func-
tionality (model for ketal) with the corresponding ketone
introduces significant angle strain into the [4 + 2] transition
structure, resulting in F-TS being 4 kcal/mol higher in energy
than E-TS. The additional strain induced by the ketone is
reflected in the compressed angle of the ketone ( abc ) 114°).
These calculations support the experimental result that dimethyl
ketal complexes (74-76) undergo the intramolecular cycload-
dition more readily than their ketone analogues (66, 68, 70).
Because systems G and H have a longer tether (four atoms
instead of three), two intramolecular cycloadditions are now
feasible. These two reactions are known as Type I and Type II
intramolecular cycloadditions. Figure 5 shows the four relevant
transition structures for systems G and H. In each case,
conformational searches were performed to determine the lowest
energy conformation of the tether. G-TS-I and H-TS-I are the
lowest energy conformations corresponding to Type I, and
G-TS-II and H-TS-II are the lowest energy conformations
Figure 6. Calculated energy differences between endo- and exo-geometries,
as well as olefin stereochemistry in cycloaddition transition states.
corresponding to Type II. For each system, the Type I transition
structure is lower in energy than the Type II transition. The
results for G are consistent with the experimental results shown
for the reaction of 83 in which only the Type I cycloadduct
was isolated. On the other hand, the calculations for H appear
to disagree with the experimental observation that the product
ratio of Type I to Type II cycloadduct was 1.0:3.3. The second
important trend is that the E-alkene G leads to lower energy
transition structures as compared to the Z-alkene H. This result
is consistent with the experimental finding that 83 gives
relatively high yields of the desired adduct relative to dimer-
ization, whereas 87 gives relatively poor yields of cycloadduct
relative to dimerization.
As a quantitative measure of the factors that contribute to
the energy differences between the cis and trans systems,
calculations on simplified systems were done (Figure 6). The
first comparison is between cis- and trans-methyl crotonate, and
the cis-isomer is found to be 1.6 kcal/mol higher in energy than
the trans-isomer. The second comparison is between an endo
(23) Tantillo, D. J.; Houk, K. N.; Jung, M. E. J. Org. Chem. 2001, 66, 1938.
9
14756 J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002

Cycloadditions of Cyclobutadiene with Olefins
A R T I C L E S
Table 9. Predicted Relative Rates of Cycloadditions
Figure 7. Transition states for dimerization of substituted cyclobutadienes.
versus and exo cycloaddition involving methyl acrylate as the
dienophile. The endo orientation is preferred by only 0.6 kcal/
mol relative to the exo orientation in the [4 + 2] transition state.
Third, the transition states involving trans-crotonate versus cis-
crotonate differ by 2.8 kcal/mol. This indicates that ap-
proximately 0.6 kcal/mol of additional strain in the transition
state can be attributed to the cis- versus the trans-configuration.
Because cyclobutadiene dimerization is the major competing
side reaction, the energetics of the bimolecular process were
also probed. The dimerizations of three model systems (cy-
clobutadiene, methylcyclobutadiene, and methyl cyclobutadi-
enecarboxylate) are illustrated in Figure 7. Only the singlet
energy surface was explored because singlet cyclobutadiene has
been calculated to be significantly lower in energy than the
triplet. Bartlett’s MRCCSD(T) + ZPE calculations indicate that
the rectangular singlet ground state is 12.5 kcal/mol below the
triplet. Even at the square geometry, the singlet state is 8.8 kcal/
mol lower in energy than the triplet state.24
The discussion begins with the dimerization of parent
cyclobutadiene, which was studied previously with HF calcula-
tions.2 The most striking feature of the dimerization transition
structure is that is does not resemble a typical [4 + 2]
cycloaddition. It actually resembles the Cope transition state,
and the motion associated with the imaginary frequency is
highlighted in Figure 7. This is similar to the dimerization of
cyclopentadiene,25 in which a bispericyclic transition structure
leads to the formation of only one σ bond. A subsequent valley
ridge inflection breaks the C2 symmetry by mixing in the Cope
TS, and either the [4 + 2] or the [2 + 4] cyclopentadiene dimer
is formed. The main difference for the cyclobutadiene dimer-
ization is that there is no barrier to formation of the first σ bond.
Similar transition structures are found for methylcyclobuta-
diene and methylcyclobutadiene-carboxylate. One additional
complication is that the methyl or ester groups can adopt various
substitution patterns about each four-membered ring. A total
of seven dimers are possible for each system, and one is shown
in Figure 7 as a representative example. The nature of the
substituent has very little effect on the activation free energy
a Calculations predict no intramolecular cycloadditions, but experimen-
tally a good yield of cycloadduct 43 is obtained from 42.
of the dimerization. On the basis of these calculations, dimer-
ization of monosubstituted cyclobutadiene is expected to be
barrierless and, therefore, diffusion controlled.
Table 9 summarizes the predicted rate constants at room
temperature, using the computed values of ∆Gq found here and
transition state theory.26 The table orders the reactions according
to increasing activation free energy. From the calculated
activation barriers for systems A-H, the experimental yields
can be explained if the rate of cyclobutadiene dimerization is
slower than the intramolecular cycloadditions of A, G, and E,
but faster than the intramolecular cycloadditions of B, H, F, C,
and D. The intramolecular cycloadditions A, G, and E occur in
good yields, while dimerization appears to be the major reaction
for B, H, F, C, and D. This suggests that the first-order rate
constant for cyclobutadiene dimerization under diffusion control
in acetone is approximately 103 s-1, which allows for reaction
E to produce intramolecular cycloadducts but for reaction B to
produce mainly dimer. The value of 103 s-1 appears low because
the bimolecular diffusion rate constant is expected to be on the
order of 108 M-1 s-1 (the typical diffusion rate constant in water)
and the Fe(CO)3-cyclobutadiene complex is present in milli-
molar concentration. However, assuming that free cyclobuta-
diene is reacting as it is formed and that only 10-5 M is present
at any given time during the reaction, the diffusion rate could
(24) Balkova´, A.; Bartlett, R. J. J. Chem. Phys. 1994, 101, 8972.
(25) Caramella, P.; Quadrelli, P.; Toma, L. J. Am. Chem. Soc. 2002, 124, 1130.
(26) The first-order rate constants were evaluated using the Eyring equation: k
) (kbT/h) exp(-∆Gq/RT).
9
J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002 14757

A R T I C L E S
Limanto et al.
be 10-3 s-1, so that the fastest intramolecular cycloadditions,
such as A, G, and E, can compete with dimerization.
along with H2O and NaHCO3 (aq). The organic layer was separated,
and the aqueous layer was back-washed with CH2Cl2. The combined
organic layers were then dried over Na2SO4 and concentrated. The
resulting ether was purified by silica gel chromatography.
Conclusions
General Procedure for Cyclobutadienyl Alkylation. Tricarbonyl-
[η4-(methoxymethyl)cyclobutadiene] iron (1.0 equiv) was dissolved in
CH2Cl2 (0.5 M) under an Ar atmosphere at room temperature with
stirring, resulting in a yellow homogeneous solution. Addition of the
enol silyl ether (1.3 equiv) followed by boron trifluoride-diethyl etherate
(1.1 equiv) resulted in a dark brown solution. The reaction mixture
was stirred for 0.5 h and quenched with saturated NaHCO3 (aq). The
reaction mixture was diluted with CH2Cl2 and poured into a separatory
funnel containing water and CH2Cl2, the organic layer was separated,
and the aqueous layers were extracted with CH2Cl2. The combined
organic layers were then washed with brine and dried with anhydrous
sodium sulfate. The solution was filtered, concentrated under vacuum,
and purified by silica gel column chromatography. All compounds were
stored under a N2 atmosphere at -20 °C.
Oxidative decomposition of iron-cyclobutadiene complexes
with CAN or TMAO liberates free cyclobutadiene, which can
be trapped intramolecularly with various olefins to afford
potentially useful and highly functionalized cyclobutene-
containing cycloadducts. Generally, cyclobutadiene complexes
that fail to undergo the intramolecular cycloadditions under the
CAN-promoted reaction conditions (room temperature, acetone
or MeOH, 1 mM) react to afford the corresponding cycloadducts
when the slower oxidant, TMAO, is employed (acetone, 1-40
mM, 56 °C, 12-24 h). Electronically unactivated olefins,
connected through a three-atom etherate tether, have successfully
trapped the free cyclobutadiene to give [4 + 2] cycloadducts
in good yields. With four-atom ether tether substrates, on the
other hand, electronic activation and trans-configuration of the
tethered olefin are required for an efficient intramolecular
cycloaddition. Unactivated trans-olefin or activated cis-olefin
complexes afford only moderate yields of the corresponding
cyclobutene cycloadducts under TMAO-promoted cycloaddition
conditions.
In the carbocyclic series of substrates, electronically activated
three-atom-tethered cis- and trans-olefins efficiently trap the
free cyclobutadiene under fast CAN reaction conditions, provid-
ing good yields of the corresponding cycloadducts. On the other
hand, enone substrates, where the carbonyl group is part of the
tether, require one of the following: (1) conversion of the
carbonyl group into a dimethyl ketal functionality, (2) incor-
poration of an additional external electron withdrawing group
(i.e., CO2Me), (3) geminal dialkyl substituents in the tether, or
(4) employment of TMAO as a slower oxidant. Furthermore,
substrates with unactivated dienophiles connected through a
three-atom tether require the use of TMAO. Incorporating rigid
functional groups into the tether, as well as the presence of
strongly activated dienophiles, are essential for efficient CAN-
promoted intramolecular cycloadditions of four-atom, all-carbon
tether substrates.
Intramolecular Cyclobutadiene Cycloaddition Methods. The
intramolecular cycloadditions were promoted by oxidative removal of
the iron from its corresponding cyclobutadiene ligand according to one
of the following methods:
Method A (CAN-Promoted Cycloadditions). To a solution of the
iron complex (1.0 equiv) under a N2 atmosphere in either HPLC-grade
acetone or MeOH (1-3 mM) was added CAN (5.0 equiv) as a solid
over 3 min at room temperature. The resulting orange solution was
stirred for 10 min, and then quenched with saturated NaHCO3 and
transferred to a separatory funnel containing Et2O and H2O. The aqueous
layer was separated and extracted with Et2O (2×). The combined
organic layers were washed with H2O and brine, dried over MgSO4/
K2CO3, filtered and concentrated in vacuo, and purified by silica gel
flash column chromatography.
Method B (TMAO-Promoted Cycloadditions). To a solution of
the iron complex (1.0 equiv) under a N2 atmosphere in HPLC-grade
acetone (1-40 mM) was added a portion of TMAO (4-10 equiv) all
at once. The resulting reaction mixture was refluxed for 6 h, treated
with a second portion of TMAO (4-10 equiv), and refluxed for an
additional 6-18 h. Once all of the starting material has been consumed
as judged by TLC and GC, the resulting brown suspension was cooled
to room temperature and then transferred to a separatory funnel
containing Et2O and saturated NaHCO3. The aqueous layer was
separated and extracted with Et2O. The combined organic layers were
washed with H2O, brine, dried over MgSO4/K2CO3, filtered, concen-
trated, and purified by silica gel chromatography.
Theoretical considerations of the transition states in these
cycloadditions provide useful insight into the factors influencing
the course of the reaction. Moreover, the calculations offer a
means to predict the success of the intramolecular cycloadditions
versus competing intermolecular dimerization processes. The
information and understanding offered by these studies should
help advance further the methodology, as well as promote its
application toward new challenges in synthesis.
Acknowledgment. M.L.S. thanks the National Institutes of
Health (GM62824), the National Science Foundation (CHE-
0132221), and the ACS-PRF for financial support. K.N.H.
thanks the National Science Foundation (CHE-9986344) for
financial support. K.S.K. thanks the National Science Foundation
for a Graduate Fellowship. Michelle (Randall) Heffernan, Patrick
J. Morris, and Anthony Messina are acknowledged for their
experimental contributions. In addition, we thank Prof. Larry
T. Scott for insightful suggestions.
Experimental Section27
General Procedure for Cyclobutadienyl Etherification. Sulfuric
acid (0.20 equiv) was added to a solution of the tricarbonyl[η4-
(hydroxymethyl)cyclobutadiene] iron (1 equiv) in the allylic alcohol
(solvent) at 0 °C. The reaction was judged complete by TLC usually
after approximately 90 min at 0 °C. CH2Cl2 was added to the reaction
Supporting Information Available: Experimental procedures,
data on new compounds, and computational results (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
(27) See the Supporting Information for general experimental and computational
methods as well as details about specific compounds.
JA0203162
9
14758 J. AM. CHEM. SOC. VOL. 124, NO. 49, 2002