Synthesis of substituted chromanones: An organocatalytic aldol/oxa-michael reaction

Article abstract of DOI:10.1021/ol101221c

(Equation Presented). A diastereoselective organocatalytic aldol/oxa-Michael reaction has been developed to efficiently deliver medicinally relevant 2,3-ring-substituted chromanones. Development of this synthetic strategy revealed an unexpected kinetic anti-Saytzeff elimination; an origin for the observed selectivity is suggested on the basis of the results of quantum chemical calculations. This unusual kinetic selectivity necessitated an isomerization protocol that in turn led to the discovery of an intriguing Pd-mediated isomerization/intramolecular Friedel-Crafts-type alkylation.

Full text of DOI:10.1021/ol101221c

ORGANIC
LETTERS
2010
Vol. 12, No. 15
3410-3413
Synthesis of Substituted Chromanones:
An Organocatalytic Aldol/oxa-Michael
Reaction
Jeffrey D. Butler, Wayne E. Conrad, Michael W. Lodewyk, James C. Fettinger,
Dean J. Tantillo, and Mark J. Kurth*
Department of Chemistry, UniVersity of California, One Shields AVenue,
DaVis, California 95616
mjkurth@ucdaVis.edu
Received May 26, 2010
ABSTRACT
A diastereoselective organocatalytic aldol/oxa-Michael reaction has been developed to efficiently deliver medicinally relevant 2,3-ring-substituted
chromanones. Development of this synthetic strategy revealed an unexpected kinetic anti-Saytzeff elimination; an origin for the observed
selectivity is suggested on the basis of the results of quantum chemical calculations. This unusual kinetic selectivity necessitated an isomerization
protocol that in turn led to the discovery of an intriguing Pd-mediated isomerization/intramolecular Friedel-Crafts-type alkylation.
Chromanones are medicinally pertinent heterocycles,1 and
the chroman parent system has been identified in natural
products such as sappanone B2 and robustadial3 in addition
to being a bioisostere for the hydantoin moiety.4 With this
biological relevance, chromanone synthesis has received
considerable attention by our group5 and others.6 Asymmetric
preparations of chromanones have also been reported.7 A
less explored aspect of the chemistry of this heterocycle is
the development of syntheses that impart R- and/or ꢀ-sub-
stitution or, more specifically, fused-ring chromanones (cou-
marin derivatives and tetrahydro-xanthones). These structures
are found in the cores of pseudobruceol-I8 and diversonol9
(Figure 1). Although the laboratories of Toste,10 Tietze,11
and Akiba12 have made advances in this area, our goal is
the development of an organocatalyzed aldol/oxa-Michael
reaction that delivers fused-ring chromanones (I) from
precursors of type II (Figure 1). Herein, we report a short
synthetic strategy for the diastereoselective preparation of
cyclopentane-fused chromanones using organocatalysis. As
this synthetic strategy developed, we encountered an unex-
pected anti-Saytzeff13 alkene formation that fortuitously led
to the discovery of a unique catalytic Pd-mediated isomer-
ization/intramolecular Friedel-Crafts-type alkylation.
(1) Kabbe, H. J.; Widdig, A. Angew. Chem., Int. Ed. Engl. 1982, 21,
247–256.
(2) Takikawa, H.; Suzuki, K. Org. Lett. 2007, 9, 2713–2716.
(3) Majewski, M.; Irvine, N. M.; Bantle, G. W. J. Org. Chem. 1994,
59, 6697–6702.
(4) Lipinski, C. A.; Aldinger, C. E.; Beyer, T. A.; Bordner, J.; Bussolotti,
D. F.; Inskeep, P. B.; Siegel, T. W. J. Med. Chem. 1992, 35, 2169–2177.
(5) Carpenter, R. D.; DeBerdt, P. B.; Holden, J. B.; Milinkevich, K. A.;
Min, T.; Willenbring, D.; Fettinger, J. C.; Tantillo, D. J.; Kurth, M. J.
J. Comb. Chem. 2008, 10, 225–229.
(8) Rashid, M. R.; Armstrong, J. A.; Gray, A. I.; Waterman, P. G.
Phytochemistry 1992, 31, 3583–3588.
(9) Turner, W. B. J. Chem. Soc., Perkin Trans. 1 1983, 1365–1368.
(10) Corkey, B. K.; Toste, F. D. J. Am. Chem. Soc. 2005, 127, 17168–
17169.
(6) (a) Chandrasekhar, S.; Vijeender, K.; Reddy, K. V. Tetrahedron Lett.
2005, 46, 6991–6993. (b) Graves, C. R.; Zeng, B.-S.; Nguyen, S. T. J. Am.
Chem. Soc. 2006, 128, 12596–12597.
(11) Tietze, L. F.; Spiegl, D. A.; Stecker, F.; Major, J.; Raith, C.; Brobe,
C. Chem.sEur. J. 2008, 14, 8956–8963.
(7) (a) Enders, D.; Breuer, K.; Runsink, J.; Teles, J. H. HelV. Chim.
Acta 1996, 79, 1899–1902. (b) McKervey, M. A.; Ye, T. J. Chem. Soc.,
Chem. Commun. 1992, 823–824. (c) Biddle, M. M.; Lin, M.; Scheidt, K. A.
J. Am. Chem. Soc. 2007, 129, 3830–3831. (d) Carpenter, R. D.; Fettinger,
J. C.; Lam, K. S.; Kurth, M. J. Angew. Chem., Int. Ed. 2008, 47, 6407–
6410.
(12) Lee, Y.-G.; Ishimaru, K.; Iwasaki, H.; Ohkata, K.; Akiba, K. J.
Org. Chem. 1991, 56, 2058–2066.
(13) (a) Saytzeff, A. M. Justus Liebigs Ann. Chem. 1875, 179, 296. (b)
Braida, B.; Prana, V.; Hiberty, P. C. Angew. Chem., Int. Ed. 2009, 48, 5724–
5728.
10.1021/ol101221c  2010 American Chemical Society
Published on Web 07/09/2010

Figure 1. Fused chromanone containing natural products and
description of a linear synthetic approach.
Our synthetic efforts began with the lithiation of 2-bro-
mophenol and subsequent addition of 2-methylcyclohexanone
(Figure 2). After an acid-promoted (aq HCl) dehydration of
the resulting benzyl alcohol, we were surprised to find a
predominance of the anti-Saytzeff product in a 1:1.3 ratio
(entry 1, 1a:1b, Figure 2) as determined by relative 1H NMR
integration. This alkene mixture was not found in the reported
cyclopentene-substituted case.14 Lowering the temperature
of the dehydration step from 0 °C to -40 or -78 °C gave
a ∼1:3.1 (entry 2, 1a:1b) or 1:5.0 ratio (entry 3, 1a:1b),
respectively. These results showed no dependence on the acid
employed as H2SO4 and H3PO4 gave comparable ratios.
Methyl substitution in the 3-, 4-, or 5-position had little effect
on the mixture (entries 4-6). However, a sterically larger
i-Pr substituent at the 2-position generated a 1:19 ratio of
alkenes (entry 7). In addition, similar alkene mixtures were
found when 2-bromoanisole was used instead of 2-bro-
mophenol (entry 8).
Figure 2. General synthetic method: (a) regioselectivity and (b)
isomerization of cyclohexenylphenol derivatives.
While these mixtures may likely be avoided by employing
modern palladium coupling methods, we viewed this as an
opportunity to confirm our computational results and set out
to develop an isomerization method. Attempts to isomerize
1b f 1a under acidic conditions (TsOH, EtOH, reflux)
resulted in dihydrobenzofuran formation where the nu-
cleophic phenol intercepts the cationic intermediate upon
anti-Markovnikov protonation.15
In contrast, anisole 6b slowly isomerizes to the desired
tetrasubstituted alkene 6a (TsOH, EtOH, reflux, 14 d) without
heterocyclization. However, compatible methods to dem-
ethylate the o-methoxy group proved unreliable and difficult
to scale-up in our hands. Encouraged by the recent report
by Wang et al. of a mild ortho-enhanced isomerization,16
we employed their conditions (cat. PdCl2, FeCl3, DCE 60
°C) and found that 6a/b is efficiently isomerized to only 6a
in 4 h (Figure 2). However, isomerization attempts on the
phenol mixture 1a/1b resulted in formation of the afore-
mentioned oxy-cyclized product. To circumvent this, the
mixture of phenolic alkenes 1a/1b-4a/4b were mesyl-
protected and to our delight, nicely participated in the PdCl2
isomerization reactions to give 7a-10a. The mesyl group
was subsequently removed by hydrolysis yielding the desired
phenolic cyclic styrenes 1a-4a (Figure 2).
To better understand these unusual results, we utilized DFT
calculations (B3LYP/6-31+G(d,p), see Supporting Informa-
tion for details) to examine the energies and structures of
the possible alkene products (1a/1b) as well as the presumed
common benzylic cation precursor. The results of these
calculations indicate that alkene 1a is indeed the thermody-
namically favored product, lying approximately 2.3 kcal/mol
lower in energy than alkene 1b (see Supporting Information).
Shown in Figure 3 are five conformers (two chair
conformations and three twist-boat conformations) of the
benzylic cation that we were able to locate, along with their
computed relative energies. The results show that the three
lowest energy conformers (cat1.1, cat1.3, and cat1.4) have
the R1 methyl substituent in alignment with the empty
p-orbital of the carbocation, rather than the tertiary hydrogen
atom that would lead to the thermodynamic product upon
deprotonation. In fact, the two conformers (cat1.2 and cat1.5)
that have this hydrogen atom aligned with the empty p-orbital
of the carbocation lie more than 3 kcal/mol higher in energy
than the lowest energy conformer. We postulate that these
energy differences are likely manifested to some extent in
the barriers for deprotonation, consistent with alkene 1b
being the kinetically favored product for this reaction.
As delineated in Figure 4, 1a-4a were subjected to
ozonolysis conditions to yield linear diketo phenols 11-14.
(15) Le´vy, J.; Sigaut, F. Tetrahedron Lett. 1983, 24, 4987–4988, See
product characterization in Supporting Information.
(16) Fan, J.; Wan, C.; Wang, Q.; Gao, L.; Zheng, X.; Wang, Z. Org.
Biomol. Chem. 2009, 7, 3168–3172.
(14) Wang, Q.; Chen, C. Org. Lett. 2008, 10, 1223–1226.
Org. Lett., Vol. 12, No. 15, 2010
3411

Figure 3. Computed structures and energies (B3LYP/6-31+G(d,p))
for the cationic precursor to 1a/1b. The benzylic cation center and
σ-bonds aligned with the empty p-orbital are highlighted in orange.
Distances are shown in Å, and energies are in kcal/mol.
Gratifyingly, treatment of 11 with pyrrolidine at 50 °C in
MeOH delivered the fused chromanone 15 as a single
diastereomer in 87% yield and in an efficient 61% overall
yield. To explore the impact of cyclopentane ring substitution
on diastereoselectivity, racemic methyl-substituted diketo
phenols 12-14 were exposed to the chromanone-forming
conditions (pyrrolidine, MeOH, 50 °C) to give methyl-
substituted chromanones 16-18a/18b. As outlined in Figure
4, methyl substitution at the 3-position gave good diaste-
reoselectivity (a 9:1 ratio of 16a:16b). Methyl substitution
at either the 4- (17a/17b) or 5-position (18a/18b) shows
reduced selectivity; 2.3:1 and 5:1, respectively. To further
assess the diastereoselectivity and scope of this strategy, we
wanted to explore chiral diones 19 and 20. Treatment of the
protected alkene mixture 5a/5b from (+)-menthone with the
isomerization reagents and subsequent hydrolysis, yields only
21 (Figure 5), whose structure was confirmed by X-ray
crystallography. This result represents a unique isomerization/
Friedel-Crafts-type alkylation.
To access the desired diones 19 and 20, an alternate
synthesis was required. The Baeyer-Villiger oxidation
products of (+)- and (-)-menthone (22 and 23, Figure 4),
synthesized via literature protocol,17 were treated with the
Li-exchanged product of 2-bromophenol and BuLi. Subse-
quent oxidation afforded diones 19 and 20, which were
exposed to the chromanone forming conditions yielding only
24 and enantiomer 25, respectively.
Figure 4. General synthetic method for tetrahydrocyclopenta-
[b]chromanones.
mediated process. Upon cyclization/dehydration, the forward
pathway could diverge. Possibilities include conjugate ad-
dition to the iminium followed by hydrolysis or, alternatively,
hydrolysis of the iminium and Michael addition to the enone.
Although both give the same end product, catalyst involve-
ment in each is distinctly different. That said, understanding
the role(s) of the catalyst and reaction pathway(s) is vital.
Conveniently, the 9:1 diastereomeric mixture of 16a:16b
is separable by standard flash chromatography and allowed
us the opportunity to probe reaction reversibility and gain
insight as to the role(s) of the pyrrolidine catalyst in this
tandem reaction. Several informative experiments were
conducted; these conditions and results are summarized in
Figure 7.
In the absence of catalyst, (Expt. A, Figure 7) 16a showed
no reaction. However, upon addition of catalyst (Expts. B
and C), both 16a and 16b returned to the thermodynamic
9:1 ratio of 16a:16b. If catalytic Et3N is added to 16a without
catalyst (Expt. D), a 9:1 ratio is again formed. For Expts.
Next, we probed the role(s) of the organocatalyst in our
newly developed aldol/oxa-Michael reaction. As depicted in
Figure 6, we only observed products (chromanone and
uncyclized enone) formed by chemoselective enamine for-
mation, exo-trig cyclization, and dehydration.18 It is note-
worthy that no reaction is observed (11 f 15) if catalytic
Et3N is used in place of pyrrolidine, evidence of an enamine-
(17) For oxidation of menthol, see: Sanborn, L. T. Org. Synth. 1929, 9,
52. For Baeyer-Villiger of menthone, see: (a) Yadav, J. S.; Reddy, B. V. S.;
Basak, A. K.; Narsaiah, A. V. Chem. Lett. 2004, 33, 248–249. (b) Kotsuki,
H.; Arimura, K.; Araki, T.; Shinohara, T. Synlett 1999, 4, 462–464.
Figure 5. Synthesis of hexahydrofluorenol derivative.
3412
Org. Lett., Vol. 12, No. 15, 2010

Figure 6. Proposed mechanism of tetrahydro-cyclopenta[b]chro-
manone formation.
E-G, the 16a/16b mixture was treated with KOH in MeOH.
Through a presumed E1cB mechanism, the 16a/16b mixture
was quantitatively converted to enone phenol 26. No reaction
of 26 was observed in MeOH alone (Expt. E). Upon addition
of pyrrolidine, 26 was cleanly converted back to a 9:1
mixture of 16a and 16b (Expt. F). However, 26 treated with
catalytic Et3N afforded the same ratio of 16a:16b.
These experiments give insight into the possible dual role
played by the pyrrolidine catalyst in this tandem aldol/oxa-
Michael reaction and confirm that 1,4-addition to the iminium
(Figure 6) is not a requirement. In addition, Michael addition
reversibility was corroborated.
This reversibility allows for the preparation of a single
diastereomer by iterative separation and resubjection of the
minor isomer to the reaction conditions. It unfortunately also
means that enantioselective chromanone formation from an
achiral diketo phenol fails (attempts on 11 with L-proline
and L-prolinol gave racemic 15). We suspect this equilibra-
tion occurs via an E1cB mechanism where the pyrrolidine
catalyst (Expts. B, C, and F) or Et3N (Expts. D and G) acts
simply as a base.
Figure 7. Experiments performed to investigate Michael revers-
ibility and catalyst function.
the roles of the pyrrolidine catalyst. In addition, our synthetic
strategy revealed interesting anti-Saytzeff dehydration prod-
ucts, whose isomerization led to the discovery of an
intriguing Pd-mediated intramolecular Friedel-Crafts alky-
lation. Current efforts in our laboratory are focused on this
unique cyclization and these results will be the topic of a
future report.
Acknowledgment. The authors dedicate this letter to
Professor Thomas R. Hoye (University of Minnesota), a
mentor and friend, on the occasion of his 60th birthday.
Financial support provided by the National Science Founda-
tion (CHE-0910870, CHE-0449845) and the National Insti-
tutes of Health (GM089153).
Ultimately, our goal of developing an organocatalyzed
aldol/oxa-Michael reaction to deliver ring-fused chromanones
of synthetic and biological importance has been realized. This
reaction sequence was probed and revealed evidence as to
Supporting Information Available: Full experimental
details, characterization data (1H NMR, 13C NMR, IR, and
LC/MS) for all new compounds, and computational details.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Mukherjee, S.; Yang, J. W.; Hoffmann, S.; List, B. Chem. ReV.
2007, 107, 5471–5569.
OL101221C
Org. Lett., Vol. 12, No. 15, 2010
3413