Chiral anthracene and anthrone templates as stereocontrolling elements in Diels-Alder/retro Diels-Alder sequences

Article abstract of DOI:10.1016/j.bmc.2005.06.025

The development of chiral anthracene templates for use in Diels-Alder/retro Diels-Alder sequences is described. A summary of past results and new progress is reported.

Full text of DOI:10.1016/j.bmc.2005.06.025

Bioorganic & Medicinal Chemistry 13 (2005) 5299–5309
Chiral anthracene and anthrone templates
as stereocontrolling elements in Diels–Alder/retro
Diels–Alder sequences
Kerrie L. Burgess, Matthew S. Corbett, Paul Eugenio, Neil J. Lajkiewicz, Xiang Liu,
Amitav Sanyal, Wanlin Yan, Qian Yuan and John K. Snyder*
Department of Chemistry and the Center for Methodology and Library Development, Boston University,
590 Commonwealth Ave., Boston, MA 02215, USA
Received 5 April 2005; accepted 18 May 2005
Available online 19 July 2005
Abstract—The development of chiral anthracene templates for use in Diels–Alder/retro Diels–Alder sequences is described. A sum-
mary of past results and new progress is reported.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
library development,2 an application that was originally
inspired by ThebtaranonthÕs anthracene-templated
Diels–Alder/retro Diels–Alder sequence to racemic and
achiral cyclopentenones.3 Success in this endeavor re-
quires not only high diastereoselectivity in the cycload-
dition but also high regioselectivity in subsequent
transformations of the carbonyls of the mounted dieno-
phile. Subsequent cycloreversion,4 which must proceed
without racemization, then produces the target lactams
or lactones and regenerates the chiral anthracene tem-
plate. In this article, we summarize our work and also
describe new work using chiral anthrone templates.
Over the past few years, we have been developing chiral
anthracene templates as recyclable stereocontrolling ele-
ments in the preparation of butenolides, a,b-unsaturated
c-lactams, and polycyclic alkaloidal core structures
(Scheme 1).1 Our intent was to apply these chiral auxil-
iaries in Diels–Alder/retro Diels–Alder sequences for
2. Enantioselectivity in the cycloadditions of
1,5-disubstituted anthracenes
Our initial efforts focused on the discovery of homochi-
ral catalysts to promote the enantioselective cycloaddi-
tion of achiral 1,5-disubstituted anthracenes with
maleic anhydride (MA) and N-alkyl maleimides.5 De-
spite screening of numerous catalysts known to engineer
highly enantioselective Diels–Alder reactions with other
reactants,6 these efforts were basically unsuccessful. The
best results (50% ee, enantiomer not assigned) were
obtained in the cycloaddition of N-methylmaleimide
(NMM) with 1,5-diphenylanthracene (1) using (S)-BI-
NOL/Me2AlCl as the catalytic system7 (Scheme 2, Eq.
1). Further experimentation established that carbonyl
differentiation in the MAcycloadducts of 1 was also
Scheme 1.
*
Corresponding author. Tel.: +1 617 353 2621; fax: +1 617 353
6466; e-mail: jsnyder@bu.edu
0968-0896/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.06.025

5300
K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
Scheme 3.
of diastereoselection. Similar observations were also
reported by Jones with 3b who also studied the solvent
dependency of the diastereoselection.8,9 In addition,
moving the stereogenic center one atom away from the
anthracene ring (9) or removing the oxygen from the
stereogenic center (10) led to near complete losses in
diastereoselection (dr = 3:2 in both cases in cycloaddi-
tions with NMM, major diastereomers not assigned;
Scheme 4).
Scheme 2.
not synthetically valuable. For example, the highest level
of regioselectivity in the reduction of MAcycloadduct 2
was 3:1 with L-selectride but in <10% yield. Higher yield
was obtained with NaBH4 (55%) but the regioselectivity
dropped to 2:1 (regioselectivity not assigned; Scheme 2,
Eq. 2).
These observations led us to propose a transition state
that accounts for the observed diastereoselection. This
explanation for the diastereoselection requires kinetic
control in the cycloaddition, that was confirmed using
anthracene 5.1c In this model, the methoxymethyl group
of 5 (or the methyl group in 3a, not shown) is oriented
antiperiplanar to the approaching dienophile with facial
selectivity established by minimization of electrostatic
repulsion between the anthracene C-9 methoxyl oxygen
and the dienophile carbonyl oxygen (A vs B, Fig. 2).
This rationale, which derives from HoukÕs inside alkox-
ide effect10 can also be applied to the trifluoromethyl
analogues 4a and 6 wherein the electrostatic repulsion
between the trifluoromethyl group (D) is greater than
that suffered with the methoxyl oxygen (C),1b also giving
rise to the observed diastereoselection. Finally, with the
free alcohol in 3b, hydrogen bonding (in F) helps to sta-
bilize the alternative approach that now competes effec-
tively with the transition state E, giving rise to the
3. Chiral anthracene templates
With the disappointing enantioselectivity in the cycload-
ditions of 1,5-disubstituted anthracenes, we then turned
to homochiral anthracenes with stereogenic centers
located in C-9 substituents. Two types of anthracenes
were examined (Fig. 1): Type Awith alkyl substituents
(3–6) and Type B with heteroatom substituents (7 and
8). All of the Type A anthracenes (3a–6) underwent cy-
cloadditions with maleic anhydride and N-alkylmale-
imides with complete diastereoselectivity, producing
single products in high yield (75–99%), thereby accom-
plishing the first step in the sequence depicted in Scheme
1 with the illustrated diastereoselectivity.
The structural requirements for high diastereoselectivity
in the cycloadditions of the chiral anthracenes were also
probed. With the unprotected alcohols 3b and 4b, the
cycloadditions with NMM showed poor diastereoselec-
tivity (1:1.7 and 1:1.9, respectively, Scheme 3). Further-
more, the minor product in both these cycloadditions
corresponded to the sole diastereomers produced with
the methyl ethers 3a and 4a with regard to the sense
Scheme 4.
Figure 1. Chiral anthracene templates; the number in parentheses is
relative rate of cycloadditions with NMM relative to unsubstituted
anthracene from competition experiments.
Figure 2. Proposed transition states for the cycloadditions of 5 (A/B),
4a/6 (C/D), and 3b (E/F).

K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
5301
observed mixture of diastereomeric cycloadducts depict-
ed in Scheme 3.8,10a,11
invariably led to c,c-dialkyl lactones as the major or sole
products (not shown).
From these preliminary studies, 5 emerged as our
leading chiral anthracene for use as a template in
Diels–Alder/retro Diels–Alder sequences. Dimethoxy-
ethylanthracene 5 was easy to prepare in a very high
ee12 and was quite stable to storage under ambient
conditions, though we still stored 5 under nitrogen.
Despite this stability, 5 was also more reactive than
anthracene itself in competition experiments in cyc-
loadditions with NMM. Indeed, competition experi-
ments established the relative rates (Fig. 1), with the
lower reactivity of 4a proving to be the main draw-
back in the use of this commercially available chiral
anthracene (PirkleÕs alcohol13).
With the high regio- and stereoselectivity of Grignard
reagent additions to the maleimides mounted on the
anthracene templates established, a divergent strategy
emerged to prepare alkaloidal core structures
enantioselectively by annulating rings of varying sizes
onto the mounted pyrrolidinone. In this approach, cyc-
loadducts 16 formed from a maleimide substituted on
the nitrogen with terminal alkenyl groups can be sub-
jected to one of the two procedures. In the first sequence
(Scheme 6), reduction or Grignard reagent addition fol-
lowed by allylation and ring closing metathesis (RCM)
yields the lactam 18, with cycloreversion producing the
target bicyclic alkaloidal system 20.14 The ring size of
the annulation can be systematically varied by changing
ÔnÕ in the maleimides. Transformation of the metathesis-
formed alkene lends further diversification to the chem-
istry. Cycloreversion then produces the bicyclic targets.
With the excellent diastereoselection in the cycloaddi-
tions of chiral anthracenes 3–6 established, regioselec-
tive transformations of the original dienophilic
subunits, now mounted on the anthracene templates,
were probed using cycloadducts 11 and 13 formed from
template 5. Reductions with superhydride and additions
of Grignard reagents showed excellent regioselectivity,
occurring exclusively at the carbonyl remote to the ori-
ginal C-9 anthracene substituent with the maleimide cy-
cloadducts 11 (Scheme 5, Eq. 1).1c Only nucleophilic
addition from the top face of the carbonyl, the anthra-
cene template blocking the approach to the bottom face,
were observed. We attributed the regioselectivity to
developing repulsive interactions between the chiral sub-
stituent and the carbonyl oxygen in the transition state
resulting from addition to the near carbonyl group as
this group changes from sp2 to sp3 hybridization. In
contrast, Li(t-BuO)3AlH reduction of the anhydride 13
gave a mixture of regioisomers 14 and 15 (63% and
12% isolated yields, respectively; Scheme 5, Eq. 2), with
the hydroxyl group b-oriented.5 Presumably, top face
reduction still occurred, with isomerization to the more
stable anomers resulting during workup. With the more
electrophilic anhydride, in comparison to imides 11, the
lower regioselectivity in the reduction may be the result
of an earlier transition state, with correspondingly less
sp3 character developed in the transforming carbonyl,
and therefore less repulsion with the stereogenic anthra-
cene substituent. Addition of Grignard reagents to 13
An alternative sequence can also be envisioned (Scheme
7) wherein addition of a Grignard reagent bearing a ter-
minal alkenyl group, followed by silane reduction or
addition of a second alkyl nucleophile, then RCM pro-
duces the same system 20 but with the stereogenic center
inverted in comparison to the first route leading to 18
Scheme 6.
Scheme 5.
Scheme 7.

5302
K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
illustrated in Scheme 6. Thus, both enantiomers of 20
can in principle be produced from the same chiral
anthracene template. Furthermore, in this latter strate-
gy, the location of the double bond following RCM
can also be systematically altered by changing ÔmÕ and
ÔnÕ, but keeping Ôm + nÕ as constant.
We successfully probed the suitability of the first alter-
native of this strategy in the synthesis of indolizidone
and pyrroloazepinone core structures.1c N-Vinyl-, N-al-
lyl-, and N-(3-butenyl)maleimide cycloadducts 16
(Scheme 6 above, n = 0, 1, and 2, respectively) were pre-
pared, and all underwent highly regio- and stereoselec-
tive addition of phenyl or methyl Grignard reagents,
or superhydride reduction in excellent yields. Allylations
also were successful (72–94%) with the transformed
N-allyl and N-butenyl cycloadducts, but not with the
N-vinyl analogue. Following ring closure with GrubbsÕ
I catalyst,15 hydrogenation, and cycloreversion using
flash vacuum pyrolysis16 (FVP, 400 °C, 4 min) produced
bicyclic lactams 20a–e (Fig. 3).
Scheme 9.
Furthermore, the chemoselectivity of subsequent car-
bonyl transformations of the mounted dienophile
through reduction or Grignard reagent addition were
now controlled by the dienophilic C-2 substituent while
the anthracene template provided the stereocontrol,
yielding solely additions to the carbonyl remote to the
C-2 substituent from the top face away from the aryl
rings. For example, radical allylation17 of 26g followed
by the addition of PhMgBr or n-BuLi gave the corre-
sponding adducts 28 and 29, respectively (Scheme 10)
in good yields as the sole isomers. Subsequent treatment
of 28 with SnCl4, followed by the addition of MeMgBr
to the generated acyliminium ion18 produced c,c-disub-
stituted lactam 30 (Scheme 11, Eq. 1, 81%). Cyclorever-
sion via FVP then yielded the desired a,b-unsaturated
lactam 31. Attempts to apply the same sequence to butyl
adduct 29, however, led only to elimination (Scheme 11,
Eq. 2). This proved to be a general problem with those
acyliminium ions capable of dehydrating into the side
chain as further described below.
Chiral anthracene 5 was fully recovered in all cases. At-
tempts to prepare pyrrolizidone scaffolds from the N-
allylmaleimide cycloadduct by vinyl Grignard reagent
addition, reduction, and RCM failed. Thus, attempts
to reduce vinyl adduct 21a led only to 24, and 21a
proved to be resistant to RCM with the Grubbs I cata-
lyst,14b producing only 25 in low yield with loss of the
stereogenic center at C-8 after chromatography on silica
gel (Scheme 8).
a-Substituted cycloadducts 26 were also readily ob-
tained beginning with 2-substituted maleic anhydrides
or maleimides. In the limited cases examined to date,
the cycloaddition invariably proceeded with excellent
stereo- and regioselectivity, providing only those regio-
isomers with the dienophile C-2 substituent remote to
the anthracene C-9 stereogenic substituent (Scheme 9).
Figure 3. Indolizidones and pyrroloazepinones prepared from the
general sequence shown in Scheme 6. Numbers in parentheses are
overall yields including the cycloaddition to form 16 (six steps).
Scheme 10.
Scheme 8.
Scheme 11.

K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
5303
Stereoselectivity in the enolate chemistry of racemic lac-
tone 33 prepared from the NaBH4 reduction of the
anthracene/maleic anhydride cycloadduct was also
examined as a model prior to using chiral anthracenes
in similar chemistry. Reactions of the enolate with alde-
hydes proceeded only in poor yields19 and with stereo-
selectivity that proved to be acceptable only with the
most sterically encumbered electrophile, pivaldehyde
(Scheme 12). With sterically less demanding aldehydes,
the stereoselectivity was unsatisfactory (63.5:1). Ben-
zoylation of the enolate of 33 was slightly more success-
ful, however, leading to b0-ketolactone 35 (Scheme 13).
With nonchelating reducing agents such as NaBH4
and KBHEt3, 35 underwent hydride reduction with
opposite facial selectivity in comparison with chelation
controlled reductions with L-selectride, or even better
Zn(BH4)2. In these reactions, the anthracene template
blocked the ÔbottomÕ face of the carbonyl, requiring a
ÔtopÕ face approach by the reducing agent. Without che-
lation, the favored anti-orientation of the carbonyl oxy-
gens of 35(A) exposes the opposite face of the carbonyl
in comparison with the syn-orientation (B) maintained
by chelation (here with Zn).
Scheme 14.
that produced separable cycloadducts 38a and b in a
4:1 ratio, each as a single diastereomer, followed by
MeMgBr addition to 38a. Acylation of the enolate
of 36 with octanoyl chloride failed to produce the
expected b0-ketolactone (Scheme 15). Presumably the
added steric encumberance of the anthracene C-9 sub-
stituent further reduces the reactivity at this site in
comparison with 33 (Scheme 13). Reaction of the
lithio enolate of 36 with octanal, however, did proceed
in a modest 44% yield to give separable hydroxylac-
tones 39a and b in an unattractive 2:1 ratio, unfortu-
nately with the unnatural diastereomer 39b as the
major product. Cycloreversion (FVP, 230–240 °C,
15 min) then produced acaterin 40a and its diastereo-
mer 40b. This work pointed out the somewhat disap-
pointing regioselectivity in the cycloadditions of 5
with nonsymmetric dienophiles such as 37 (Scheme
14), as well as the relatively poor yields in working
with the enolates of the lactones derived from the cyc-
loadducts of maleic anhydride.
This enolate approach to b0-alcohols was subsequently
applied to the synthesis of the natural acyl-CoA:cho-
lesterol transferase inhibitor acaterin.20 Thus, treat-
ment of 14, whose preparation was described earlier
by the reduction of maleic anhydride adduct 13
(Scheme 5), with excess MeMgBr (2.2 equiv) gave lac-
tone 36 (Scheme 14). This lactone was also prepared
by the cycloaddition of 5 with acetoxyfuranone 37
In addition to the poor enolate chemistry of the lactones
described above, other limitations in these anthracene
templates were found. First, attempts to capture the
acyliminium ion, generated upon treatment of the
c-hydroxylactams with a Lewis acid, using organometal-
lic reagents (such as Grignard reagents, organolithium
reagents, cuprates, or even Et3SiH), led to dehydration
Scheme 12.
Scheme 13.
Scheme 15.

5304
K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
Scheme 18.
11, Eq. 1) was also successful. In this example, complex-
ation of the Lewis acid between the methoxy groups as
well as the hydroxyl oxygen of 28 would position the
Sn(IV) optimally for acyliminium ion formation
(Scheme 18). Thus, with the cycloadducts of 2-substitut-
ed maleimides, which direct additions to the carbonyl
above the stereogenic C-9 substituent, the dimethoxy
element also assists in further transformations through
the corresponding acyliminium ion, as long as dehydra-
tion cannot compete.
Scheme 16.
Asecond limitation was the relatively high temperatures
necessary to achieve the retro Diels–Alder reactions. We
initially addressed the issue of more facile cyclorever-
sions by preparation of 9,10-disubstituted chiral
anthracenes 6, 44, and later, 49 (see below). Kirby and
co-workers21 and Keck and co-workers22 had also used
a 9,10-disubstituted anthracene (9,10-dimethylanthra-
cene) to trap and subsequently release highly reactive
acylnitroso reagents and thioaldehydes, and Kirby had
noted that cycloreversions with 9,10-dimethylanthra-
cene were slightly faster than with anthracene itself.21c
Another purpose in using 6 was enhanced reactivity rel-
ative to 4a by substituting C-10 with a methyl group.
This latter effect was indeed observed as the relative
reactivities of 4a:6 was 1:10 in competition experiments
with NMM.
or rearrangement in those cases with alkyl side chains
(Scheme 16, Eqs. 1 and 2, respectively, also see Scheme
11, Eq. 2). This stood in contrast to the success with the
Lewis acid catalyzed allylations (Schemes 6 and Scheme
16, Eq. 3), as well as the MeMgBr addition to the acyli-
minium ion generated from 28 (Scheme 11, Eq. 1) illus-
trated earlier. Similar results were also obtained with
analogous adducts capable of undergoing dehydration,
which were prepared from unsubstituted anthracenes,
suggesting that the limited ability to generate and syn-
thetically exploit the acyliminium may be an inherent
problem with the anthracene system.
With acyliminium ions that cannot rearrange or under-
go elimination (Scheme 17, Eq. 1), Grignard reagent
addition did succeed, though often only in poor yield
(e.g., 41, 27%). By comparison, the unsubstituted
anthracene analogue 42 underwent the same MeMgBr
addition in 76% yield (Scheme 17, Eq. 2). The lower
yield with the chiral template 12a in these examples
was thought to be due to complexation of the Lewis acid
between the lactam carbonyl oxygen and the auxiliary
methoxy groups, which inhibits formation of the desired
acyliminium ion. Supporting this suggestion was the ear-
lier observation that Sn(IV) promoted acyliminium ion
formation followed by MeMgBr addition to 28 (Scheme
Cycloadditions of 6, 44, and 49 with maleic anhydride,
N-alkylmaleimides, and 2-bromomaleic anhydride were
all completely diastereoselective. Regioselectivities in
subsequent carbonyl transformations were also excel-
lent, as illustrated for 44 (Scheme 19).
Unfortunately, there was insufficient moderation in the
conditions for the cycloreversion with 9,10-disubstituted
anthracene templates. For example, cycloadditions of
Scheme 17.
Scheme 19.

K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
5305
microwave promoted cycloreversion were observed with
graphite as a support with 19b and e, while with 19a, no
indolizidinone 20a was isolated, which was not stable to
microwave conditions. In all cases, however, the yields
of the lactams were higher with FVP (400 °C, 4–
5 min). Moreover, with FVP the chiral anthracene tem-
plate was fully recovered, but was completely lost with
microwave irradiation.
These relatively harsh cycloreversion conditions stimu-
lated further research into other chiral anthracene tem-
plates. We turned to type B chiral anthracenes 7a and
8 substituted at C-9 with lone pair donors (nitrogen
and oxygen) in an effort to develop a more reactive tem-
plate for both the original cycloaddition and the cyclo-
reversion. Czarnik has demonstrated that electron
donating substituents at anthraceneÕs C-9 (and C-10)
position accelerate the cycloreversion,25 though they
also increase the susceptibility of the anthracene to
oxidation.26
Scheme 20.
the unsymmetric dienophile 37 with 6 and 49 were inves-
tigated for comparison with the results described earlier
with anthracene 5 (see Scheme 14). The regioselectivity
observed with these 9,10-disubstituted anthracenes
(Scheme 20) was comparable to that obtained with 5.
Cycloadduct 50a, the major regioisomer in the cycload-
dition with 6, was subsequently transformed into the
butenolide 53, a precursor to a rove beetle pheromone,23
as previously reported.1b Thus, the addition of n-octyl
Grignard reagent to 50a gave 52, which was followed
by FVP, now at 190–200 °C, to yield 53. While the tem-
perature required for cycloreversion with this 9,10-
disubstituted anthracene template was reduced in
comparison to those with 3a and 5, high temperatures
were still necessary. Some cycloreversions could be
achieved in refluxing xylenes or by FVP at 190–200 °C,
but typically the best yields still required FVP at
400 °C (4 min).
Initially, we examined 9-acyloxyanthracenes 8a and b,
both easily prepared from anthrone (Scheme 21). Trost
et al.,27 Siegel and Thornton,28 and others29 had shown
that the corresponding mandelate-type 1-acyloxybutadi-
enes can participate in Diels–Alder cycloadditions with
excellent diastereoselectivities. Furthermore, oxyanion
promoted cycloreversions should be easily achieved, fol-
lowing desired transformations of the mounted dieno-
philes, by basic hydrolysis of the esters, analogous to
the work of Knapp et al.30 and Bunnage and Nico-
laou.31 Unfortunately, the cycloadditions of 8a and b
with maleic anhydride and NMM proceeded with poor
diastereoselectivity, presumably due to the temperatures
necessary to achieve acceptable yields of cycloadduct in
a reasonable amount of time. Lewis acid catalysis did
not improve the results.
From this work, anthracene 5 emerged as the preferred
template due to its ease of preparation, stability to stor-
age, and greater range of chemistry toleration with
excellent regioselectivity in carbonyl transformations
subsequent to the cycloaddition. Considerable effort
was therefore expended to accomplish the cyclorever-
sion under microwave irradiation24 using this template
with adducts 19 (Fig. 4). Various conditions were exam-
ined including irradiation neat at ambient atmosphere,
under vacuum, and adsorbed on silica gel, Montmoril-
lonite KSF, Montmorillonite K10, and graphite, all with
and without Lewis acids present. Best results for
Aminoanthracene 7a was then prepared via Pd-cata-
lyzed aryl amination32 beginning with 9-bromoanthra-
cene.1d This anthracene was considered
a good
candidate to serve as a chiral stereocontrolling template
based on the work of Rawal who has shown the effec-
tiveness of homochiral 1-aminobutadienes in controlling
the stereochemistry of cycloadditions.33 As anticipated,
7a proved to be sensitive to oxygen, producing anthra-
quinone upon standing under ambient conditions.
Figure 4. Comparison
cycloreversion.
of
FVP
with
microwave-induced
Scheme 21.

5306
K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
Storage under nitrogen at 0 °C was adequate, but it was
found to be easier to simply proceed to the cycloaddi-
tion with maleic anhydride or an N-alkylmaleimide
(Scheme 22); the cycloadducts were stable under ambi-
ent conditions for extended periods of time. The cyc-
loadditions proceeded at room temperature (rt) over
48 h, a significantly lower temperature than achieved
with other chiral anthracenes. Unlike the cycloadditions
of other chiral anthracenes 3–6, however, trace amounts
of the minor, stereoisomeric cycloadduct could be
detected in the NMR spectra of the reaction mixture.
Increasing the reaction temperature to accelerate the
cycloaddition unfortunately led to increasing amounts
of the minor diastereomer. For example, the diastereo-
selectivity of the cycloaddition of 7a with NMM
dropped from 10:1 at rt to 6:1 at 60 °C. The cycloaddi-
tion with acetoxyfuranone 37 was also not very selective,
producing a 2.5:1 ratio of 56a and b at rt (Scheme 22,
Eq. 2). The corresponding N-methylaminoanthracene
7b, also prepared by Pd-catalyzed aryl amination, gave
poor diastereoselectiviy in the cycloaddition with
NMM at rt (1.2:1, Scheme 21, Eq. 3).34
Scheme 23.
recovered by trapping as the NMM (or other maleimide)
cycloadduct in a minimum 70% yield, less than the
recovery of 5, that was essentially quantitative. Given
the greater stability of 5, its greater recovery from the
cycloreversion, and greater versatility in Grignard re-
agent additions to its cycloadducts, and the fact that
FVP was still required for optimal cycloreversion using
7a, 5 remains as the chiral anthracene template with
which we are pursuing other synthetic goals. In further
efforts to achieve cycloreversions under milder condi-
tions, we examined chiral anthrone templates as
described in the following section.
As with anthracenes 3–6, reductions or the additions of
Grignard reagents to adducts 55 occurred with exclusive
stereoselectivity at the carbonyl group remote to the
anthracene stereogenic substituent. However, Grignard
reagent additions were accompanied by some unwanted
cycloreversion, resulting in lower yields (57–78% for
methyl and phenyl Grignard reagents) in comparison
with the additions to the adducts of template 5 (83 –
99+%). Subsequent allylation, alkene transformations,
and FVP, now at 190 °C (10 min), yielded the target but-
enolides and a,b-unsaturated c-lactams as illustrated for
58a–c (Scheme 23).1d
4. Chiral anthrone templates
RickbornÕs extensive investigation into the cycloaddi-
tions of anthrone,35 coupled with KnappÕs et al.30 and
Bunnage and NicolaouÕs,31 demonstrated success with
the oxyanion promoted cycloreversions of 9-hydroxyan-
thracene cycloadducts led us to prepare and investigate
the cycloadditions of chiral anthrones. Effort to date
has focused primarily on anthrone 62, since the 1-meth-
oxyethyl stereogenic substituent had proven effective in
exerting excellent stereocontrol over the cycloadditions
of anthracene 3a. The preparation of 62 was accom-
plished in two ways. In the first route, anthrone was
While the amino substituent facilitated the cyclorever-
sion, FVP was still required for optimal yield, albeit at
lower temperatures. Aminoanthracene 7a could be
Scheme 22.
Scheme 24.

K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
5307
acylated via its in situ generated TMS ether, giving
anthronyl ketone 59 (Scheme 24, Path A).36 Asymmetric
reduction following CoreyÕs procedure37 reported for
similar substrates proceeded with excellent enantioselec-
tivity (>95% ee by HPLC). Selective methylation of the
alcohol proved troublesome due to competing methyla-
tion of the anthrone carbonyl oxygen. Indirect methyla-
tion to give 62 finally succeeded by formation of the
methylthiomethyl ether 6138 followed by Raney-Ni
reduction.
Alternatively, 62 was also prepared from 3a by singlet
oxygen cycloaddition,39 followed by reduction of the
resultant cyclic peroxide and dehydration of 63 cata-
lyzed by ZnCl2 adsorbed onto molecular sieves (Scheme
24, Path B). This latter route was more practical on a
larger scale.
Scheme 26.
ring-opened product 68 as a diastereomeric mixture
(Scheme 26, Eq. 1), parallelling observations previously
recorded by both Rickborn34,39 and Kagan and co-
workers.41 Allylation of model anthrone adduct 70
formed after MeMgBr addition to 69, however, was suc-
cessful with cycloreversion then producing racemic lac-
tam 71 (Scheme 26, Eq. 2). Therefore, it was
concluded that the difficulty in working with the acyli-
minium ions from 65 to 67 was a consequence of the
conjugative continuum extending from the alcohol oxy-
gen lone pair through the sigma framework to the acyli-
minium ion (inset). With the acyliminium ion generated
at the alternative carbonyl group as from 69, this contin-
uum does not exist, and ring opening is not a competi-
tive pathway. Therefore, we have turned our attention
to the asymmetric cycloaddition between anthrone itself
and N-alkylmaleimides.
Cycloadditions of 62 with maleimides such as NMM
proceeded in quantitative yields and complete diastereo-
selectivity under Rickborn conditions40 to give adducts
64 (Scheme 25). Reactions with maleic anhydride were
not successful due to the instability of this dienophile
to the reaction conditions. Though the highly hindered
tertiary alcohol of 64 could not be protected (Rickborn
previously noted the lack of reactivity of this alcohol,40)
64 smoothly underwent the addition of Grignard re-
agents to the top face of the cycloadduct at the carbonyl
remote to the chiral substituent with complete regiose-
lectivity. Surprisingly, no trace of cycloreversion prod-
uct could be detected. Since both Knapp and Nicolaou
had triggered their cycloreversions from anthrone cy-
cloadducts via oxygen anion generation with KH, we
reasoned that a cation effect may account for the lack
of cycloreversion from the magnesium alkoxides pro-
duced in the Grignard reagent additions. Treatment of
adducts 65 with KH did indeed result in rapid cyclore-
version (at rt) but resulted in racemic lactams 66. Pre-
sumably, the oxyanion-triggered cycloreversion is
sensitive to the degree of covalency of the alkoxide Ôsalt,Õ
with the less covalent K+ salt allowing for the facile ret-
ro Diels–Alder reaction. Unfortunately, the basic condi-
tions of the cycloreversion also led to rapid racemization
of the hydroxylactams 66.
Both Kagan and co-workers41,42 and Yamamoto and co-
workers,43 have reported sufficiently intriguing enanti-
oselectivities in the cycloadditions of anthrones with
N-alkylmaleimides to warrant further investigation
(Scheme 27). Kagan observed a maximum 61% ee using
quinidine (72) as the basic catalyst to promote the cyclo-
Attempts to utilize acyliminium ion chemistry, as was
successful with the chiral anthracene adducts derived
from 5, were not successful with adducts related to 65.
For example, attempted cyclization of 67 led only to
Scheme 25.
Scheme 27.

5308
K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
Tetrahedron Lett. 2003, 44, 931; (c) Burgess, K. L.;
Lajkiewicz, N. J.; Sanyal, A.; Yan, W.; Snyder, J. K.
Org. Lett. 2005, 7, 31; (d) Sanyal, A.; Yuan, Q.; Snyder, J.
K. Tetrahedron Lett. 2005, 46, 2475.
addition, while Yamamoto achieved 74% ee in the same
reaction using diol 73 as the catalyst with NMM as the
dienophile. In our own efforts, we have obtained a 60%
ee using CoreyÕs ligand 60 as the catalyst in the reaction
with NMM. Given the excellent diastereo- and regiose-
lectivity in the carbonyl transformations of the dienophile
mounted on the anthrone, along with the extremely facile
oxyanion-initiated cycloreversions as illustrated in
Schemes 25 and 26, we are continuing efforts to optimize
the enantioselectivity in this cycloaddition.
2. For reviews of chiral dienes as auxiliaries in Diels–Alder/
retro Diels–Alder sequences: (a) Winterfeldt, E. Chem.
Rev. 1993, 93, 827; (b) Winterfeldt, E.; Borm, C.; Nerenz,
F. Adv. Asymm. Syn. 1997, 2, 1; (c) Klunder, A. J. H.;
Zhu, J.; Zwanenburg, B. Chem. Rev. 1999, 99, 1163; (d) ,
For example, since these reviews:Goldenstein, K.; Fend-
ert, T.; Proksch, P.; Winterfeldt, E. Tetrahedron 2000, 56,
4173; (e) Tran-Huu-Dau, M.-E.; Wartchow, R.; Winter-
feldt, E.; Wong, Y.-S. Chem. Eur. J. 2001, 7, 2349; (f)
Wolter, M.; Borm, C.; Merten, E.; Wartchow, R.;
Winterfeldt, E. Eur. J. Org. Chem. 2001, 4051; (g)
Knappwost-Gieseke, C.; Nerenz, F.; Wartchow, R.; Win-
terfeldt, E. Chem. Eur. J. 2003, 9, 3849.
3. (a) Jenkitkasemwong, Y.; Thebtaranonth, Y.; Wajirum,
N. Tetrahedron Lett. 1979, 20, 1615; (b) Siwapinyoyos,
T.; Thebtaranonth, Y. J. Org. Chem. 1982, 47, 598; (c)
Kodpinid, M.; Siwapinyoyos, T.; Thebtaranonth, Y. J.
Am. Chem. Soc. 1984, 106, 4862; (d) Lertvorachon, J.;
Meepowpan, P.; Thebtaranonth, Y. Tetrahedron 1998,
54, 14341; (e) Lertvorachon, J.; Thebtaranonth, Y.;
Thongpanchang, T.; Thongyoo, P. J. Org. Chem. 2001,
66, 4692.
4. Reviews of retro Diels–Alder reactions: (a) Kwart, H.;
King, K. Chem. Rev. 1968, 68, 415; (b) Ripoll, J. L.;
Rouessac, A.; Rouessac, F. Tetrahedron 1978, 34, 19; (c)
Lasne, M. C.; Ripoll, J.-L. Synthesis 1985, 121; (d)
Rickborn, B. Org. React. 1998, 52, 1; (e) Rickborn, B.
Org. React. 1998, 53, 223.
5. Sanyal, A. PhD Dissertation, Boston University, 2002.
6. Some reviews of catalytic, enantioselective Diels–Alder
reactions: (a) Kagan, H. B.; Riant, O. Chem. Rev. 1992,
92, 1007; (b) Dias, L. C. J. Braz. Chem. Soc. 1997, 8, 289;
(c) Carmona, D.; Lamata, M. P.; Oro, L. A. Coord. Chem.
Rev. 2000, 200–202, 717; (d) Corey, E. J. Angew. Chem.,
Int. Ed. 2002, 41, 1650.
5. Conclusions
Chiral anthracenes have shown great promise as stereo-
controlling templates in the Diels–Alder/retro Diels–
Alder sequence. The applications with maleimide
dienophiles have shown success in the synthesis of alka-
loidal core structures, such as indolizidinones and
pyrroloazepinones, as well as in the preparation of a,
b-unsaturated c-lactams. The chiral anthracene that
has currently emerged for applications in synthesis is
dimethoxyethylanthracene 5. This anthracene has shown
the widest range of compatibility with carbonyl transfor-
mations subsequent to the cycloaddition and is also
benchtop stable. The main drawback with 5 is the
requirement for FVP for the cycloreversion. To over-
come this drawback, current efforts are also devoted to
develop the asymmetric cycloadditions of anthrone.
The anthrone template holds high potential due to the
facile cycloadditions, the high regioselectivity in subse-
quent carbonyl transformations including Grignard
reagent additions, and facile cycloreversion upon oxyan-
ion generation.
7. Examples using BINOL-aluminum complexes as catalysts
in asymmetric Diels–Alder reactions: (a) Graven, A.;
Johannsen, M.; Joergensen, K. A. J. Chem. Soc., Chem.
Commun. 1996, 2373; (b) Johannsen, M.; Yao, S.; Graven,
A.; Jorgensen, K. A. Pure Appl. Chem. 1998, 70, 1117; (c)
Morgan, P. E.; McCague, R.; Whiting, A. J. Chem. Soc.,
Perkin Trans. 1 2000, 515; (d) Bertozzi, F.; Olsson, R.;
Frejd, T. Org. Lett. 2000, 2, 1283.
Dedication
This manuscript is dedicated to Prof. Koji Nakanishi of
Columbia University, not only on the occasion of the
awarding of the well-deserved Tetrahedron prize, but
also for his exceptional mentorship and inspiration he
has provided since we first met more than 20 years
ago. J.K.S. is eternally grateful for the opportunities
Professor Nakanishi has afforded him to begin his re-
search career, which remains one of the most stimulating
experiences of his life.
8. Atherton, J. C. C.; Jones, S. Tetrahedron Lett. 2001, 42,
8239.
9. For related work of Jones on chiral anthracenes: (a) Jones,
S.; Atherton, J. C. C. Tetrahedron: Asymmetry 2001, 12,
1117; (b) Atherton, J. C. C.; Jones, S. J. Chem. Soc.,
Perkin Trans. 1 2002, 2166; (c) Atherton, J. C. C.; Jones, S.
Tetrahedron Lett. 2002, 43, 9097; (d) Bawa, R. A.; Jones,
S. Tetrahedron 2004, 60, 2765; (e) Jones, S.; Ojea-Jimenez,
I. Polycyclic Aromat. Compd. 2005, 25, 1.
Acknowledgments
10. (a) Houk, K. N.; Moses, S. R.; Wu, Y. D.; Rondan, N. G.;
Jaeger, V.; Schohe, R.; Fronczek, F. R. J. Am. Chem. Soc.
1984, 106, 3880; (b) Houk, K. N.; Duh, H. Y.; Wu, Y. D.;
Moses, S. R. J. Am. Chem. Soc. 1986, 108, 2754; (c)
Haller, J.; Niwayama, S.; Duh, H.-Y.; Houk, K. N. J. Org.
Chem. 1997, 62, 5728.
11. For other examples of hydroxyl-directed cycloadditions:
(a) Trost, B. M.; Lee, D. C. J. Org. Chem. 1989, 54, 2271;
(b) Curran, D. P.; Choi, S.-M.; Gothe, S. A.; Lin, F. T. J.
Org. Chem. 1990, 55, 3710; (c) Also, see: Barriault, L.;
Thomas, J. D. O.; Clement, R. J. Org. Chem. 2003, 68,
2317; (d) Barriault, L.; Ang, P. J. A.; Lavigne, R. M. A.
Org. Lett. 2004, 6, 1317.
We thank the donors of the Petroleum Research Fund,
administered by the American Chemical Society (ACS-
PRF 35222-AC), NIGMS CMLD initiative (P50
GM067041), Boston UniversityÕs Undergraduate Re-
search Opportunity Program (UROP), and the Pfizer
PREPARE program for financial support.
References and notes
1. Our work: (a) Sanyal, A.; Snyder, J. K. Org. Lett. 2000, 2,
2527; (b) Corbett, M. S.; Liu, X.; Sanyal, A.; Snyder, J. K.

K. L. Burgess et al. / Bioorg. Med. Chem. 13 (2005) 5299–5309
5309
12. Corey, E. J.; Noe, M. C. J. Am. Chem. Soc. 1996, 118,
11038.
25. Chung, Y.-S.; Duerr, B. F.; Nanjappan, P.; Czarnik, A.
W. J. Org. Chem. 1988, 53, 1334–1336.
13. (a) For a review: Pirkle, W. H.; Hoover, D. J. Top.
Stereochem. 1982, 13, 263; (b) , For use of 4b as a chiral
auxiliaryCarriere, A.; Virgili, A. Tetrahedron: Asymmetry
1996, 7, 227; (c) Carriere, A.; Virgili, A.; Figueredo, M.
Tetrahedron: Asymmetry 1996, 7, 2793.
14. For related approaches utilizing RCM: (a) Martin, S. F.;
Chen, H.-J.; Courtney, A. K.; Liao, Y.; Paetzel, M.;
Ramser, M. N.; Wagman, A. S. Tetrahedron 1996, 52,
7251; (b) Lim, S. H.; Ma, S.; Beak, P. J. Org. Chem. 2001,
66, 9056; (c) Hanessian, S.; Margarita, R.; Hall, A.;
Johnstone, S.; Tremblay, M.; Parlanti, L. J. Am. Chem.
Soc. 2002, 124, 13342; (d) Hanessian, S.; Sailes, H.;
Munro, A.; Therrien, E. J. Org. Chem. 2003, 68, 7219.
15. Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413.
16. Schiess, P.; Barve, P. V.; Dussy, F. E.; Pfiffner, A. Org.
Synth. 1995, 72, 116.
17. Keck, G. E.; Yates, J. B. J. Am. Chem. Soc. 1982, 104, 5829.
18. For a review of acyliminium ion chemistry: Speckamp, W.
N.; Moolenaar, M. J. Tetrahedron 2000, 56, 3817.
19. Examples of aldol reactions of c-lactones: (a) Widdowson,
D. A.; Wiebecke, G. H.; Williams, D. J. Tetrahedron Lett.
1982, 23, 4285; (b) Anceau, C.; Dauphin, G.; Coudert, G.;
Guillaumet, G. Bull. Soc. Chim. Fr. 1993, 18, 291; (c)
Anceau, C.; Dauphin, G.; Coudert, G.; Guillaumet, G.
Bull. Soc. Chim. Fr. 1994, 131, 291.
26. (a) Stevens, B.; Perez, S. R.; Ors, J. A. J. Am. Chem. Soc.
1974, 96, 6846; (b) Sweger, R. W.; Czarnik, A. W. J. Am.
Chem. Soc. 1991, 113, 1523.
27. (a) Trost, B. M.; Godleski, S. A.; Genet, J. P. J. Am.
Chem. Soc. 1978, 100, 3930; (b) Trost, B. M.; OÕKrongly,
D.; Belletire, J. L. J. Am. Chem. Soc. 1980, 102, 7595; (c)
Tucker, J. A.; Houk, K. N.; Trost, B. M. J. Am. Chem.
Soc. 1990, 112, 5465.
28. (a) Siegel, C.; Thornton, E. R. Tetrahedron Lett. 1988, 29,
5225; (b) Tripathy, R.; Carroll, P. J.; Thornton, E. R. J.
Am. Chem. Soc. 1991, 113, 7630.
29. Masamune, S.; Choy, W.; Petersen, J. S.; Sita, L. R.
Angew. Chem., Int. Ed. Engl. 1985, 24, 1.
30. Knapp, S.; Ornaf, R. M.; Rodriques, K. E. J. Am. Chem.
Soc. 1983, 105, 5494.
31. Bunnage, M. E.; Nicolaou, K. C. Angew. Chem., Int. Ed.
1996, 35, 1110.
32. For the procedure followed: Wagaw, S.; Rennels, R. A.;
Buchwald, S. L. J. Am. Chem. Soc. 1997, 119, 8451.
33. For lead references: (a) Kozmin, S. A.; Rawal, V. H. J.
Am. Chem. Soc. 1997, 119, 7165; (b) Kozmin, S. A.;
Rawal, V. H. J. Am. Chem. Soc. 1999, 121, 9562; (c) Janey,
J. M.; Iwama, T.; Kozmin, S. A.; Rawal, V. H. J. Org.
Chem. 2000, 65, 9059.
34. Yuan, Q.; Snyder, J. K. Unpublished results.
35. (a) Koerner, M.; Rickborn, B. J. Org. Chem. 1989, 54, 6;
(b) Koerner, M.; Rickborn, B. J. Org. Chem. 1990, 55,
2662.
20. (a) Naganuma, S.; Sakai, K.; Hasumi, K.; Endo, A. J.
Antibiot. 1992, 45, 1216; (b) Ishigami, K.; Kitahara, T.
Tetrahedron 1995, 51, 6431.
21. (a) Kirby, G. W. Chem. Soc. Rev. 1977, 6, 1; (b) Bladon,
C. M.; Ferguson, I. E. G.; Kirby, G. W.; Lochead, A. W.;
McDougall, D. C. J. Chem. Soc., Perkin Trans. 1 1985,
1541; (c) Choi, S. S. M.; Kirby, G. W. J. Chem. Soc.,
Perkin Trans. I 1991, 3225.
36. Adapted from: Camaioni, D. M.; Alnajjar, M. S. J. Org.
Chem. 1985, 50, 4456.
37. Corey, E. J.; Link, J. O. Tetrahedron Lett. 1990, 31, 601.
38. Medina, J. C.; Salomon, M.; Kyler, K. S. Tetrahedron
Lett. 1988, 29, 3773.
22. (a) Keck, G. E.; Webb, R. Tetrahedron Lett. 1979, 20,
1185; (b) Keck, G. E.; Webb, R. R., II J. Am. Chem. Soc.
1981, 103, 3173; (c) Keck, G. E.; Webb, R. R., II J. Org.
Chem. 1982, 47, 1302; (d) Keck, G. E.; Yates, J. B. J. Org.
Chem. 1982, 47, 3590.
39. Murray, R. W.; Kaplan, M. J. Am. Chem. Soc. 1968, 90,
537.
40. Harrison, R.; Rickborn, B. Org. Lett. 2002, 4, 1711.
41. Riant, O.; Kagan, H. B.; Ricard, L. Tetrahedron 1994, 50,
4543.
23. Bloch, R.; Gilbert, L. J. Org. Chem. 1987, 52, 4603.
24. Garrigues, P.; Garrigues, B. C. R. Acad. Sci., Paris, IIC
1998, 545.
42. Riant, O.; Kagan, H. B. Tetrahedron Lett. 1989, 30, 7403.
43. Uemae, K.; Masuda, S.; Yamamoto, Y. J. Chem. Soc.,
Perkin Trans. 1 2001, 1002.