Intramolecular Diels-Alder reactions of cycloalkenones: Translation of high endo selectivity to trans junctions

Article abstract of DOI:10.1021/ja307708q

Intramolecular Diels-Alder reactions of cyclobutenone and larger cycloalkenones are described. High levels of endo addition attained from Lewis acid catalysis translate to trans hydrindene junctions upon fragmentation of the tricyclic adducts.

Full text of DOI:10.1021/ja307708q

Article
pubs.acs.org/JACS
Intramolecular Diels−Alder Reactions of Cycloalkenones: Translation
of High Endo Selectivity to Trans Junctions
Audrey G. Ross,† Xiaohua Li,†,‡ and Samuel J. Danishefsky*,†,§
†Department of Chemistry, Columbia University, Havemeyer Hall, 3000 Broadway, New York, New York 10027, United States
§Laboratory for Bioorganic Chemistry, Sloan-Kettering Institute for Cancer Research, 1275 York Avenue, New York, New York
10065, United States
S
* Supporting Information
ABSTRACT: Intramolecular Diels−Alder reactions of cyclo-
butenone and larger cycloalkenones are described. High levels
of endo addition attained from Lewis acid catalysis translate to
trans hydrindene junctions upon fragmentation of the tricyclic
adducts.
nectivity in the IMDA reaction is dictated by the same
arguments discussed above. In cases where the tethering
element is sufficiently flexible to allow for either endo or exo
presentation, endo/exo selectivity can be problematic (Figure
2).
INTRODUCTION
■
The role of the Diels−Alder reaction in organic chemistry is
well appreciated.1,2 Its particular impact on complex targeted
natural product synthesis arises, in the first instance, from its
capacity to deliver extensively functionalized six-membered
rings with defined sites of unsaturation and functional group
placements. Moreover, it provides the basis for coordinating
stereogenic centers in the emerging cyclohexene. These
configurational relationships are governed by the apparently
universal suprafacial nature of the Diels−Alder reaction, and by
the less universal mode of mutual presentation of its diene and
the dienophilic components (i.e., endo vs exo addition).
Thus, in the formation of cyclohexene 3 by Diels−Alder
logic, the relationships at C1 and C4 reflect the incident
stereochemistry of diene 1 (Figure 1, shown with Z-hydrogens
Figure 2. Cyclobutenone IMDA concept.
Figure 1. Stereochemistry of Diels−Alder reactions.
In an earlier study, we had demonstrated the powerful
dienophilicity of cyclobutenone in the context of intermolecular
Diels−Alder reactions.5 In those explorations, it was found that
the parent cyclobutenone not only is quite reactive, but also
exhibits high levels of endo control, even in uncatalyzed settings.
Following those findings, we wondered about the utility of a
cyclobutenone dienophilic moiety in the context of IMDA
reactions. To deal with this question, it would be necessary to
synthesize probe substrates with a 1,3-diene suitably tethered to
a carbon center of a dienophilic cyclobutenone. We would
attempt to demonstrate the anticipated cycloaddition. Assum-
at C1 and C4), which is transmitted to 3 via suprafacial
cycloaddition. The same governing principle dictates the
emerging relative configurations at C5 and C6 arising from
dienophile 2 (shown for illustration with two cis Y groups). The
mutual relationship of C1/C6 and C4/C5 reflects the outcome
of endo vs exo preference.
Generally, in thermally mediated intermolecular Diels−Alder
reactions with typical dienophilic activation, endo presentation
tends to be favored. Not infrequently, Lewis acid catalysis
enhances endo selectivity in Diels−Alder reactions.3
Intramolecular variations of the Diels−Alder reaction
(IMDA) are similarly recognized for their capacity to deliver
high-complexity cycloaddition products.4 Stereochemical con-
Received: August 7, 2012
Published: September 5, 2012
© 2012 American Chemical Society
16080
dx.doi.org/10.1021/ja307708q | J. Am. Chem. Soc. 2012, 134, 16080−16084

Journal of the American Chemical Society
Article
ing IMDA reaction did occur, we would hope to determine
endo/exo preferences in both catalyzed and uncatalyzed modes.
Finally, this study would compare the IMDA characteristics of a
cyclobutenone moiety with those of more traditional cyclo-
pentenone- and cyclohexenone-type dienophiles. Though these
cycloalkenones might be powerful Diels−Alder synthons on
paper, the latter two are notoriously sluggish dienophiles.6−9
For reasons of design and synthetic convenience, we
proposed to place the diene-containing tether at C3 of the
cyclobutenone (see system 4, Figure 2). In addition to
exploring this uncharted IMDA reaction type, and addressing
several questions of interest to students of the Diels−Alder
reaction, there was another consideration. Figure 2 spells out
the stereochemical consequences of the endo vs exo modalities
of diene/dienophile presentations, leading to 5 and 6,
respectively. It was hoped that the ketone function of the
IMDA product would provide a site for cleavage of the
cyclobutanone A-ring. This possibility is generalized in the form
of the respective bicyclic degradation products 7 and 8, to be
retrieved from 5 and 6, respectively. Thus, while the
cycloaddition, per se, creates the AB fusion, it is the coalescence
of the tether, the γ-carbon of the diene, and the β-carbon of the
activated dienophile which gives rise to a new ring. In our
construction, this peripherally formed ring is referred to as the
C-ring. It is also fused to the B-ring.
Herein, we report the following: (i) reduction to practice of
the central cyclobutenone IMDA hypothesis, (ii) high levels of
endo stereoselection and extension even with the more classical
cyclopentenone and cyclohexenone dienophilic moieties in the
β-tethering arrangement employed here, and (iii) an illustrative
method for obtaining trans-fused hydrindenes from the
resultant tricyclic IMDA products.
It is easily recognized that the trans ring junction found in 7,
and the nonparallel relationship of the double bond at the BC
junction, are inaccessible from conventional Diels−Alder
reaction logic. In the case of cis-fused 8, the double bond is,
again, nonparallel to the junction. Thus, 8 is also not directly
accessible by a classical [4+2] step.
Figure 3. Preparation of cyclobutenone substrates. Key: (a) t-BuLi, 9,
then 10; (b) TFAA/NaHCO3 or trace BF3·OEt2, 56−66% yield; (c)
Zn, AcOH, TMEDA, MeOH, 64−69% yield.
arrangement where the diene was to be separated from the
β-carbon of the cyclobutenone moiety by a tether of three
methylene groups. Following the IMDA reaction, a five-
membered C-ring would be produced by the coalescence
described in Figure 2. The method of synthesis of such
precursor substrates centered on reaction of a lithium derivative
of the type 9 (R = H or Me) with the known 4-chloro-3-
ethoxycyclobutenone 10.14 Following 1,2-addition and unravel-
ing mediated by trifluoroacetic acid (TFAA) or trace BF3·OEt2,
there was obtained the pre-IMDA substrate 12, which could be
reductively dechlorinated by the action of zinc in acetic acid,
tertiary amine, and alcoholic solvent.15 Two probe substrates
13 and 14, prepared by this method, were to be studied in
detail as to their performance in IMDA reactions (vide infra).
As the study was expanded to include cyclopentenone and
cyclohexenone as the dienophiles in the IMDA reaction, a
second protocol was adopted (Figure 4). Reaction of 9 with the
A complementary motivation underlying these studies was
that of enhancing the reach of the pattern recognition analysis
(PRA) approach to retrosynthetic design.10 We view PRA as an
aid to the powerful dominant logic of prioritized strategic bond
disconnections formulated by Corey and associates.11−13 In
PRA, one seeks to identify subunits, i.e. motifs, within a more
complex target. Identification of such patterns may bring to
light a strategic framework for building outward from the
pattern in advancing to the target. Alternatively, one might
identify several recognizable patterns that can be melded by
appropriate chemistry.
Of course, for PRA to aid in retrosynthetic analysis, the
patterns must themselves be accessible by synthesis. Thus, it
was anticipated that, by identifying useful motifs for PRA and
by charting new pathways for their synthesis (utilizing new
insights and new protocols), the potential applicability of PRA
as a resource in retrosynthetic analysis would be significantly
enhanced. In terms of PRA, 7 corresponds to a trans-iso-Diels−
Alder motif, while 8 can be seen as a cis-iso-Diels−Alder
structure.
Figure 4. Preparation of IMDA substrates. Key: (a) t-BuLi, 9, then
cycloalkenone; (b) PCC, 58−69% yield over two steps.
appropriate cycloalkenone gave rise to the expected allylic
tertiary alcohols 15. After oxidative transposition of these
intermediates,16 probe systems 16−19 were in hand.
Subsequently, this method was extended to cover the cases of
9 with four methylene groups in the cyclobutenone, as well as
cyclopentenone and cyclohexenone settings (vide infra).
We first examined the cyclobutenone-containing dienophile
13 (Figure 5). It was found that, under BF3·OEt2 mediation,
IMDA reaction occurred in good yield, giving rise to 20 with
high levels of stereoselection. Thus, endo selection in the
RESULTS AND DISCUSSION
■
Two routes for producing the required tethered probe
structures to facilitate the study were employed. Starting with
the case of cyclobutenone (Figure 3), we selected an
16081
dx.doi.org/10.1021/ja307708q | J. Am. Chem. Soc. 2012, 134, 16080−16084

Journal of the American Chemical Society
Article
Figure 5. Lewis acid (LA)-catalyzed IMDA reactions.
Figure 6. Thermal IMDA reactions.
cycloaddition step, giving rise to the cis AB junction, had also
established the trans BC relationship. Similarly, the terminal
pentadienyl compound 14 also underwent IMDA reaction
mediated by BF3·OEt2. Once again, the process was highly endo
selective, giving rise to 21 in high yield. In both cases there
seemed to be some decomposition of these vulnerable
substrates. We found no indication for formation of a closely
related stereoisomeric (exo) product as a factor in yield erosion.
It was of interest to determine whether the finding of strict
endo cycloaddition, in the cases of 13 and 14, extends to more
traditional dienophilic cycloalkenones, which are significantly
less reactive than is cyclobutenone.17 Moreover, in the few
literature cases where the performance of five- and six-
membered cycloalkenone dienophiles in IMDA reactions
were evaluated (though in a different tethering mode), levels
of endo selectivity were not easily interpretable, let alone
predictable.18−20
Happily, BF3·OEt2 catalysis of substrates 16−19 also gave
rise to apparently single cycloaddition products (22−25, Figure
5). Though rather more forcing conditions were required, and
somewhat diminished yields were observed in the larger
cycloalkenone IMDA reactions, the logic described above
does serve well to provide excellent means of generating
peripheral (BC) trans junctions. Perhaps this advance could be
used toward the development of enantioselective catalysis
conditions for cycloalkenones in the future.
We next asked whether the high yields and margins of endo
stereoselection manifested in these BF3-mediated IMDA
reactions would extend to uncatalyzed conditions (Figure 6).
As above, the two cyclobutenone tethered cases 13 and 14
produced 20 and 21, respectively. Again, these were the only
products found. Not surprisingly, absent catalysis, significantly
higher temperatures were required to effect IMDA reaction.
The yields of IMDA products 20 and 21 were reduced to 35%
and 54%, respectively. Substrate decomposition, rather than
product instability, was shown to be the cause of the diminished
yields, since 20 and 21, on resubjection to the reaction
conditions, were actually quite stable.
dimethylethyl)-4-methylphenol (butylated hydroxytoluene, or
BHT). Diastereomeric mixtures were encountered in the ratios
indicated (see exo isomers 26−29, in addition to the major endo
products 22−25). In summary, then, as regards the IMDA
reactions in the cases studied thus far, only Lewis acid-catalyzed
protocols provide yields and stereospecificities suitable for
further development.
The tethering described in systems 13−19 has involved three
methylene groups between the diene and dienophilic elements
of the IMDA reaction. Thus, the collateral C-ring formed in the
Lewis acid-directed mode is five-membered and trans-fused to
the concurrently emerging B-ring; i.e., a “trans hydrindenoid”
submotif is elaborated. We wondered about extending the size
of the tether by one methylene group, thereby providing access
to a six-membered collateral C-ring. Following degradation of
the A-ring (see Figure 2), a trans-fused octalin would emerge.
Toward that end, we prepared compounds 30−32,21 using
much the same chemistry described earlier (see 13, 14, and
16−19).
We were most surprised to find that, upon reaction of
compound 30, which we expected to be the best prospect of
these substrates under Lewis acid-catalyzed conditions (first
with zinc chloride as the catalyst), the previously described
compound 21 was obtained (in 53% yield, Figure 7). When the
reaction was conducted, as before, with BF3·OEt2 as the
catalyst, the same product was produced, but now only in 22%
yield. Further, 21 was also the only product obtained, albeit in
poor yield, from thermal IMDA reactions of 30. Under both
conditions, no isomerized 14 was recovered; the mass balance
A more complicated situation arose following IMDA
reactions of dienylcycloenones 16−19 under thermal con-
ditions. Here, again, we obtained IMDA products in each case.
However, high temperatures were required in uncatalyzed
reactions (ca. 200 °C) along with antioxidant 2,6-bis(1,1-
Figure 7. Tandem diene migration−IMDA. Thermal and Lewis acid
(LA)-catalyzed IMDA conditions for the appropriate dienophile ring
size are as shown in Figures 5 and 6.
16082
dx.doi.org/10.1021/ja307708q | J. Am. Chem. Soc. 2012, 134, 16080−16084

Journal of the American Chemical Society
Article
loss was from substrate decomposition. Apparently, Lewis acid
or thermolysis provoked isomerization of the butadienyl four-
carbon tether, providing the pentadienyl three-carbon tether
(see compound 14). This tandem diene migration−IMDA
process was at the time unexpected but is, in retrospect, not
unprecedented.22−26
While four E,Z stereopermutations of 14 are theoretically
possible, apparently only the E,E isomer undergoes cyclo-
addition under these conditions, thereby affording the
previously observed 21 in low yield. It should be recognized
that, in the case of our earlier three-carbon tether substrate
compounds, similar isomerization was also in principle possible.
Such a migration, for instance in the case of 13, would have
generated a two-carbon tether of the cyclobutenone, which was
independently confirmed to not undergo isomerization or
IMDA reaction under the conditions employed in this study.21
Since the conversion of 13→20 had occurred in fairly high
yield, diene migration must be much slower than cycloaddition.
Assuming that the migration is roughly independent of the
differing tether size (i.e., 13 and 30 might be seen as equally
prone to double bond migration), the direct IMDA reaction in
the case of 30 (as opposed to 13) is far slower than in the
analogous double bond migration. Put differently, in changing
the tether from three to four methylene groups, the rate of
IMDA cycloaddition is dramatically attenuated.
Figure 8. Illustrative fragmentation of A-ring to trans BC hydrindenes.
Given this interesting, albeit nonhelpful sequence in
cyclobutenone substrate 30, it was not surprising that the
same behavior was observed from the four-methylene-tethered
cyclopentenone and cyclohexenone systems 31 and 32. In
these cases, cycloaddition under thermal conditions or Lewis
acid mediation gave rise, in reduced yields, to the previously
encountered endo IMDA products 24 and 25 arising from
precursors 18 and 19, respectively.
could also retain the tricycle or otherwise transform the
product, for example by carbon- or nitrogen-based ring
expansions.
CONCLUSION
■
In summary then, the proposed logic at the core of the proposal
been reduced to practice. The first examples of cyclobutenone
as an IMDA dienophile are described. As in bimolecular Diels−
Alder reactions, cyclobutenone is more reactive and endo-
selective than are analogous larger cycloalkenone dienophiles. A
remarkably high level of endo stereogovernance in the Lewis
acid-catalyzed IMDA reaction was observed. These advances
provide an exploitable route to otherwise difficultly accessible
trans hydrindenoid systems, of interest in our laboratory.29−32
The resident double bond in such structures provides
additional opportunities for diversification. Given the stereo-
selectivity now attainable, the idea of using IMDA reactions to
generate peripheral trans fusions is apt to be a resource in the
retrosynthetic analysis of contemporary targets.
Finally, we address the issue of exposing the target angularly
substituted and functionalized hydrindenes. As discussed
earlier, the plan was to exploit the ketone found in the A-ring
corresponding to the original dienophile (see 4→5, Figure 2).
There are a variety of ways by which ring disconnection or
derivatization could be accomplished. Here, for initial
demonstration purposes, we chose a sequence consisting of
Baeyer−Villiger oxidation to produce the erstwhile A-ring as a
lactone, followed by reductive cleavage to generate a diol with
differentiable primary and secondary alcohols (Figure 8). In
practice, in the compounds bearing no methyl groups in the B-
ring (i.e., 20, 22, and 24), this chemistry occurred quite
smoothly. Regiospecific-type Baeyer−Villiger oxidations could
be conducted without complications. In the case of the
cyclobutenone-containing product tricyclic 21 containing an A-
ring methyl group, facile Baeyer−Villiger oxidation of the
cyclobutenone could be accomplished with hydrogen peroxide
as in substrate 20. However, in the cases of 23 and 25,
attempted Baeyer−Villiger oxidations with mCPBA were
frustrated by epoxidation of the resident carbon−carbon
double bond. In the case of 23, the desired Baeyer−Villiger
oxidation could be achieved through the use of bis-
(trimethylsilyl) peroxide (BTSP) with 5 mol % Bi(OTf)3.27,28
However, the seemingly obvious extension to 25 took a
surprising turn: There was observed, transiently, what appeared
to be a nonproductive Baeyer−Villiger peroxide intermediate,
eventually giving rise to the isoable α-hydroxyketone 43.
Clearly, in principle, there are a wealth of methods for
degrading cyclic ketones or α-hydroxyketones to unveil
diversity in the trans hydrenoid iso-Diels−Alder pattern. One
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures; spectral and other characterization
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
s-danishefsky@ski.mskcc.org
■
Present Address
‡Department of Chemistry, The University of Toledo, Toledo,
OH 43606
Notes
The authors declare no competing financial interest.
16083
dx.doi.org/10.1021/ja307708q | J. Am. Chem. Soc. 2012, 134, 16080−16084

Journal of the American Chemical Society
Article
(32) Peng, F.; Grote, R. E.; Danishefsky, S. J. Tetrahedron Lett. 2011,
52, 3957−3959.
ACKNOWLEDGMENTS
■
Support was provided by the NIH (HL25848 to S.J.D.) and the
NSF (predoctoral fellowship to A.G.R.). We thank Drs. J.
Decatur and Y. Itagaki for NMR and mass spectrometric
assistance, respectively; W. Sattler (Parkin group, Columbia
University) for X-ray experiments (NSF, CHE-0619638); and
William F. Berkowitz for helpful discussions.
REFERENCES
■
(1) Corey, E. J. Angew. Chem., Int. Ed. 2002, 41, 1650−1667.
(2) Nicolaou, K. C.; Snyder, S. A.; Montagnon, T.;
Vassilikogiannakis, G. Angew. Chem., Int. Ed. 2002, 41, 1668−1698.
(3) Roush, W. R. Comprehensive organic synthesis; Pergamon Press:
Oxford, 1991; Vol. 5, pp 513−550.
(4) Takao, K.; Munakata, R.; Tadano, K. Chem. Rev. 2005, 105,
4779−4807.
(5) Li, X.; Danishefsky, S. J. J. Am. Chem. Soc. 2010, 132, 11004−
11005.
(6) Fringuelli, F.; Pizzo, F.; Taticchi, A.; Halls, T. D. J.; Wenkert, E. J.
Org. Chem. 1982, 47, 5056−5065.
(7) Bartlett, P. D.; Woods, G. F. J. Am. Chem. Soc. 1940, 62, 2933−
2938.
(8) Fringuelli, F.; Pizzo, F.; Taticchi, A.; Wenkert, E. Synth. Commun.
1979, 9, 391−394.
(9) Ihara, M.; Kawaguchi, A.; Ueda, H.; Chihiro, M.; Fukumoto, K.;
Kametani, T. J. Chem. Soc., Perkin Trans. 1 1987, 1331−1337.
(10) Wilson, R. M.; Danishefsky, S. J. J. Org. Chem. 2007, 72, 4293−
4305.
(11) Corey, E. J. Chem. Soc. Rev. 1988, 17, 111−133.
(12) Corey, E. J. Angew. Chem., Int. Ed. 1991, 30, 455−465.
(13) Corey, E. J.; Chen, X.-M. The logic of chemical synthesis; Wiley:
New York, 1995.
(14) Dudley, G. B.; Takaki, K. S.; Cha, D. D.; Danheiser, R. L. Org.
Lett. 2000, 2, 3407−3410.
(15) Danheiser, R. L.; Savariar, S.; Cha, D. D. Org. Synth. 1990, 68,
32−40.
(16) Dauben, W. G.; Michno, D. M. J. Org. Chem. 1977, 42, 682−
685.
(17) Paton, R. S.; Kim, S.; Ross, A. G.; Danishefsky, S. J.; Houk, K. N.
Angew. Chem., Int. Ed. 2011, 50, 10366−10368.
(18) Shiina, J.; Nishiyama, S. Tetrahedron 2003, 59, 6039−6044.
(19) Jauch, J. Angew. Chem., Int. Ed. 2000, 39, 2764−2765.
(20) Grieco, P. A.; Kaufman, M. D.; Daeuble, J. F.; Saito, N. J. Am.
Chem. Soc. 1996, 118, 2095−2096.
(21) See Supporting Information for details: (a) synthesis, (b)
spectral documentation and other characterization.
(22) Borch, R. F.; Evans, A. J.; Wade, J. J. J. Am. Chem. Soc. 1977, 99,
1612−1619.
(23) Marinier, A.; Baettig, K.; Dallaire, C.; Pitteloud, R.;
Deslongchamps, P. Can. J. Chem.-Rev. Can. Chim. 1989, 67, 1609−
1617.
(24) Grieco, P. A.; Beck, J. P.; Handy, S. T.; Saito, N.; Daeuble, J. F.;
Campaigne, E. E. Tetrahedron Lett. 1994, 35, 6783−6786.
(25) Matikainen, J.; Kaltia, S.; Hamalainen, M.; Hase, T. Tetrahedron
1997, 53, 4531−4538.
(26) Humphrey, J. M.; Liao, Y. S.; Ali, A.; Rein, T.; Wong, Y. L.;
Chen, H. J.; Courtney, A. K.; Martin, S. F. J. Am. Chem. Soc. 2002, 124,
8584−8592.
(27) Suzuki, M.; Takada, H.; Noyori, R. J. Org. Chem. 1982, 47, 902−
904.
(28) Dembech, P.; Ricci, A.; Seconi, G.; Taddei, M. Org. Synth. 1997,
74, 84.
(29) Kim, W. H.; Lee, J. H.; Danishefsky, S. J. J. Am. Chem. Soc. 2009,
131, 12576−12578.
(30) Lee, J. H.; Zhang, Y.; Danishefsky, S. J. J. Am. Chem. Soc. 2010,
132, 14330−14333.
(31) Paton, R. S.; Mackey, J. L.; Kim, W. H.; Lee, J. H.; Danishefsky,
S. J.; Houk, K. N. J. Am. Chem. Soc. 2010, 132, 9335−9340.
16084
dx.doi.org/10.1021/ja307708q | J. Am. Chem. Soc. 2012, 134, 16080−16084