Divergent Photocyclization/1,4-Sigmatropic Rearrangements for the Synthesis of Sesquiterpenoid Derivatives

Article abstract of DOI:10.1021/acs.orglett.6b03635

Combined experimental and computational efforts have demonstrated the utility of divergent photocyclization/1,4-sigmatropic rearrangement reactions for developing a general strategy toward the synthesis of cubebane-, spiroaxane-, and guaiane-type sesquiterpenes and related analogues. The configuration of the bridgehead substituent, the choice of solvent, and the wavelength of irradiation all impact diastereoselectivity in this tandem reaction process.

Full text of DOI:10.1021/acs.orglett.6b03635

Letter
pubs.acs.org/OrgLett
Divergent Photocyclization/1,4-Sigmatropic Rearrangements for the
Synthesis of Sesquiterpenoid Derivatives
Evgueni Gorobets,† Norman E. Wong,† Robert S. Paton,‡ and Darren J. Derksen*,†
†Department of Chemistry, University of Calgary, 2500 University Drive NW, Calgary, Alberta Canada T2N 1N4
‡Chemistry Research Laboratory, Oxford University, Mansfield Road, Oxford OX1 3TA, U.K.
S
* Supporting Information
ABSTRACT: Combined experimental and computational efforts have
demonstrated the utility of divergent photocyclization/1,4-sigmatropic
rearrangement reactions for developing a general strategy toward the
synthesis of cubebane-, spiroaxane-, and guaiane-type sesquiterpenes and
related analogues. The configuration of the bridgehead substituent, the
choice of solvent, and the wavelength of irradiation all impact
diastereoselectivity in this tandem reaction process.
ased on our interest in identifying natural products that
interact with transient receptor potential channels,1 we
required alternative strategies to access cubebane, guaiane, and
spiroaxane sesquiterpene frameworks for the synthesis of
analogues (Scheme 1, 4−6). Inspired by the well-known
been relatively underutilized in organic synthesis compared to
other rearrangement reactions.3 The rich chemistry of divinyl
ketones is well-known from elegant work on Nazarov chemistry
to generate complex cyclopentanones,4 and the related acid-
catalyzed dienone−phenol rearrangement has been investigated
extensively.5 Oxyallyl cation intermediates generated from
irradiation of monocyclic dienones have been used to concisely
generate complex polycylic frameworks and have been
examined computationally.6 Central to our synthetic strategy
is an understanding of the impact that existing stereocenters
have on the regio- and stereochemical outcome of the tandem
reaction. The bridgehead substituent (Scheme 1, 1-X) plays a
key role in the rearrangement of fused ring systems by dictating
the stereochemical outcome of the initial photocyclization
reaction. In santonin-derived syntheses, a methyl substituent is
utilized at the bridgehead position imposing severe limitations
on subsequent functionalization.2 Balancing substituent stability
with the ability to be removed under reductive reaction
conditions, we selected the nitrile moiety as a versatile
bridgehead functional group. From tetrahydrocarvone 7, the
nitrile moiety was introduced using tosyl cyanide and LDA (8),
followed by annulation using methyl vinyl ketone and catalytic
sodium methoxide (9/10, Scheme 2). Dehydration of 9 to form
the α,β-unsaturated ketone was followed by formation of
dienone 11 via halogenation and elimination (85% over three
steps). In order to obtain gram quantities of 10, a revised
synthetic strategy was also devised from (R)-carvone7,8 (see the
Supporting Information for details).Conversion of 10 to the
divinylketone 12 occurred in a similar fashion; however,
Saegusa oxidation was required as the analogous halogen
elimination strategy used for 9 gave low yields of the desired
product (5%). With the key precursors in hand, the central
B
Scheme 1. (Top) Photomediated Rearrangement of α-
Santonin. (Bottom) Retrosynthesis for the Divergent
Syntheses of Sesquiterpene Frameworks Using Tandem
Reactions
rearrangement of α-santonin to lumisantonin and related
applications in synthesis,2 we aimed to establish a general
strategy to access the functionalized tricyclic cores (2, 3) using
the tandem photocyclization/1,4-sigmatropic rearrangement
sequence. The [1,4]-sigmatropic rearrangement (1,4-SR)
reaction has intrigued synthetic chemists for decades but has
Received: December 6, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b03635
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Scheme 2
Scheme 3. Optimization of Irradiation Conditions (See the
SI for Details)
a
Key: (a) p-Tol-SO2CN, LDA, −78 °C, 85%; (b) NaOMe, methyl
vinyl ketone, 64% (5.3:1); (c) (i) TFAA, DIPEA, 93%, (ii) TMSOTf,
NEt3, (iii) Pd(OAc)2, 77% or NBS; LiCO3, 85%.
photorearrangement/1,4-SR sequence could be explored.
Irradiation conditions of 11 and 12 were evaluated by varying
the solvents and wavelengths of light (monitoring by 1H NMR
spectroscopy). Exposure of 11 in benzene to visible light (400−
700 nm) provided low conversion and selectivity to 13:14,
which improved slightly when UV-A (315−400 nm) or UV-B
light (280−315 nm, Scheme 3, entries 1−3) was used. To
minimize the amount of degradation byproducts observed on
an increased scale, it was necessary to recycle starting material
through the reaction conditions (entries 4 and 5). Changing the
solvent to acetonitrile had a marked effect of reaction rate and
selectivity. A consistent improvement in the conversion of
starting material was observed when the wavelength was
increased from visible to UV−C light (UV-C: 235−280 nm;
Scheme 3, entries 6−10). Contrary to the reactivity of 11, the
tandem reaction of 12 in benzene was highly selective for 16
regardless of the wavelength of light used (entries 11−13). The
reaction selectivity of 12−16 was essentially unaffected by C6
substituents as incorporation of a halogen produced the
respective bromo derivative9 of 16 (entries 14 and 15). The
6-methyl and 6-cyano substituents formed the respective
derivatives of 16, containing adjacent chiral quaternary centers,
in good yield using visible light (entries 16 and 17). Upon
switching the solvent to acetonitrile, 15 could be observed in
the course of the irradiation with visible light; however, the
ratio of 15/16 decreased until only 16 remained after full
conversion of starting material (entries 18−20). Similar results
were obtained when UV-A and UV-B light were used, requiring
significantly shorter reaction times (entries 21−23), while UV-
C was optimal for obtaining 15 (entries 24−26). Extending
reaction times using UV-C light led to complex mixtures of
degradation byproducts. Although substrate configuration plays
a major role in the outcome of the tandem reaction process, the
choice of solvent and wavelength of irradiation have a
significant impact on the regioselectivity observed in the
resulting products. A similar dependence of wavelength on
product selectivity in photochemical reactions has been
previously reported.10 After the tandem rearrangement
sequence, the choice of retention or fissure of the cyclopropane
moiety determines the natural product framework to be
obtained. Reduction of 13 was accomplished by hydrogenation
with Pd/C to produce 17 (Scheme 4). Generation of the
spirocycle 18 occurred by reductive ring-opening of 17 using
sodium napthalide at low temperature.11 Following literature
precedent for decyanation reactions, ketone 18 was protected
as the cyclic acetal before reaction with potassium/HMPA.12
Removal of the acetal under mildly acidic conditions generated
the spiroaxane-like compound 20. Compound 17 can also be
directly protected to form a cyclic acetal, and exposure to
dissolving metal conditions followed by acid treatment
completes the decyanation reaction to maintain the cube-
a
b
c
Type of light bulb used. Hours irradiated. Ratio based on crude 1H
NMR spectra. No ratio indicates maximum conversion to isolated
d
e
product. Yield of major product unless indicated. Benzene solvent.
f
g
h
5−70 mg scale. 480 mg scale, yields based on two recycles. Yield of
i
j
k
minor product. Acetonitrile solvent. 150−180 mg scale. Low
conversion with extensive product degradation observed.
bane-like ring system of known compound 19.13 Utilizing an
identical strategy, rearrangement product 15 was hydrogenated
to produce 21 in high yield. Acetal formation, decyanation, and
acid treatment to ketone 22 then allowed clean conversion to
the natural product cubebol (23) by treatment with MeLi and
CeCl3.13,14
Hydrogenation of 16 produced tricyclic compound 24
(Scheme 5). Reductive ring opening using sodium naphthalide
generated 25 as a separable but inconsequential mixture of
B
DOI: 10.1021/acs.orglett.6b03635
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
tertiary nitrile 24 occurs using conditions analogous to those
used for 17 and 21 (see the SI for details).
Scheme 4. (Top) Reduction and Ring Opening of 13
Followed by Reductive Decyanation To Generate
Spiroaxane- or Cubebane-Type Frameworks. (Bottom)
Synthesis of Cubebol
Given the utility of tricyclic frameworks 13−16 as
intermediates in complex molecule synthesis, we investigated
the competing mechanisms computationally (Figure 1).15 The
M06-2X exchange-correlation functional was used for all
geometry optimizations and the Karlsruhe def2-SZVP basis
set for all elements. Single-point energetics were further refined
with a larger def2-TZVP basis set and implicit SMD description
of acetonitrile.16 The uM06-2X functional has been used
previously by Houk and co-workers to describe the triplet PES
of the di-π-methane rearrangement of dibenzobarrelenes and
gives results comparable to those of multiconfigurational
(MCSCF) calculations.17 In accordance with Zimmerman’s
seminal experimental findings,18 C3−C5 bond formation in
triplet states of cyclohexadienones 11 and 12 is facile (barriers
of 3.6 and 6.1 kcal/mol, respectively). Recent MCSCF
computations on the santonin−lumisantonin rearrangement
show that intersystem crossing (ISC) to the singlet state can
occur either at this stage or following subsequent C−C
cleavage.19 The formation of Woodward−Hoffmann disallowed
products 14 and 16 is inconsistent with singlet ground-state
reactivity (i.e., a Zimmerman type A zwitterion) and suggests
ISC occurs after C−C cleavage. Therefore, assuming that
cyclopropyl ring-opening takes place in the triplet state, we
obtained competing transition structures (TSs) in which the
cleavage of either C−C bond takes place to give the two
product skeletons. The activation barriers for these TSs are all
feasible and, most importantly, are consistent with the observed
switch in product selectivity obtained from 11 and 12: the
computed selectivity is 3.3:1 (ΔΔG⧧ 0.7 kcal/mol) in favor of
cubebane 13 and 1:12.6 (ΔΔG⧧ 1.7 kcal/mol) in favor of
guaiane 16.20 This computed switch in selectivity was
independent of the level of theory used: uwB97XD, uB3LYP-
D3(BJ), and uB2-PLYP single-point energies for the TSs with
the def2-TZVP basis set all favor the formation of cubebane 13
(by 0.2−0.7 kcal/mol) and guaiane 16 (by 1.6−1.7 kcal/mol).
The triplet pathway for each diastereomer is thus predicted to
occur with opposite selectivity. Nevertheless, excited and
ground singlet-state reactivity is also plausible for both
diastereomers. The cyclopropane-opening TSs (A-B, A-C, D-
E, D-F) lead to triplet diradical intermediates which, following
Scheme 5. Reduction and Ring Opening of 16 Followed by
Decyanation To Form Guaiane-Type Scaffold
diastereomers (85:15), and a three-step sequence produced the
decyanated, guaiane-type compound 26. The decyanation of
Figure 1. Rearrangements of 11 and 12. SMD-uM062X/def2-TZVP//uM062X/def2-SVP relative Gibbs energies (kcal/mol).
C
DOI: 10.1021/acs.orglett.6b03635
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
B.; Munoz, M. C.; Pedro, J. R. J. Org. Chem. 2000, 65, 2138−2144.
̃
ISC, were found to evolve barrierlessly to singlet products.
Alternative diastereomers to those observed experimentally
were investigated for these stepwise rearrangements, although
in every case these were substantially disfavored by more than
25 kcal/mol. In addition, a singlet, concerted [1,4]-sigmatropic
rearrangement was also found for the formation of cubebane
product in each case: although this pathway could take place if
ISC occurs earlier, these TSs were found to be less stable than
the triplet TSs discussed above. Our combined experimental
and computational efforts indicate that cyclohexyl substituents
exert a controlling steric influence on product selectivity. In the
favored TS A−C, the cyclohexyl ring is half-chair-shaped with
(c) Greene, A. E.; Edgar, M. T. J. Org. Chem. 1989, 54, 1468−1470.
(d) Edgar, M. T.; Greene, A. E.; Crabbe, P. J. Org. Chem. 1979, 44,
159−160. (e) Marshall, J. A.; Johnson, P. C. J. Org. Chem. 1970, 35,
192−196. (f) White, E. H.; Marx, J. N. J. Am. Chem. Soc. 1967, 89,
5511−5513.
(3) (a) Tantillo, D. J. Acc. Chem. Res. 2016, 49, 741−749. (b) Houk,
K. N.; Li, Y.; Evanseck, J. D. Angew. Chem., Int. Ed. Engl. 1992, 31,
682−708. (c) Padwa, A.; Bur, S. K. Tetrahedron 2007, 63, 5341−5378.
(4) (a) Tius, M. A. Chem. Soc. Rev. 2014, 43, 2979−3002.
(b) Spencer, W. T., III; Vaidya, T.; Frontier, A. J. Eur. J. Org. Chem.
2013, 2013, 3621−3633. (c) Vaidya, T.; Eisenberg, R.; Frontier, A. J.
ChemCatChem 2011, 3, 1531−1548. (d) Marx, V. M.; Burnell, D. J. J.
Am. Chem. Soc. 2010, 132, 1685−1689. (e) Nakanishi, W.; West, F. G.
Curr. Opin. Drug. Discovery Devel. 2009, 12, 732−751. (f) Tius, M. A.
Eur. J. Org. Chem. 2005, 2005, 2193−2206.
(5) (a) Vitullo, V. P.; Grossman, N. J. Am. Chem. Soc. 1972, 94,
3844−3848. (b) Schultz, A. G.; Green, N. J. J. Am. Chem. Soc. 1992,
114, 1824−1829. (c) Schultz, A. G.; Hardinger, S. A. J. Org. Chem.
1991, 56, 1105−1111.
(6) (a) Bos, P. H.; Antalek, M. T.; Porco, J. A., Jr.; Stephenson, C. R.
J. J. Am. Chem. Soc. 2013, 135, 17978−17982. (b) Dauben, W. G.;
i
pseudoequatorial Me and Pr groups; in the disfavored TS D−
F, a twist-boat conformation results in which there is an
unfavorably small contact (2.17 Å) between the cyclo-
i
pentadienyl and Pr substituents (see the SI). We speculate
that the increased formation of cubebane-type products at
lower wavelengths may occur due to C−C formation in an
excited singlet state and that a singlet [1,4]-rearrangement
could also bypass the involvement of the triplet pathway. The
tandem photocyclization/1,4-sigmatropic rearrangement strat-
egy is a viable method to access densely functionalized
polycyclic frameworks found in cubebane-, spiroaxane-, and
guaiane-type sesquiterpenes. Computational evidence is con-
sistent with singlet and triplet mechanisms competing to form
reaction products depending on conformational effects of the
substituents. With new insights gained from the tandem
cyclization rearrangements of 11 and 12, we are focusing our
efforts on developing novel methods to further control
selectivity.
Hecht, S. J. Org. Chem. 1998, 63, 6102−6107. (c) Gom
́
ez, I.; Olivella,
A. J. Am. Chem. Soc. 2002, 124,
S.; Reguero, M.; Riera, A.; Sole,
15375−15384.
́
(7) Shipe, W. D.; Sorensen, E. J. J. Am. Chem. Soc. 2006, 128, 7025−
7035.
(8) Ayad, T.; Genisson, Y.; Baltas, M. Org. Biomol. Chem. 2005, 3,
2626−2631.
(9) Halogenation procedure adapted from: Nicolaou, K. C.; Lim, Y.
H.; Becker, J. Angew. Chem., Int. Ed. 2009, 48, 3444−3448.
(10) Stichnoth, D.; Kolle, P.; Kimbrough, T. J.; Riedle, E.; de Vivie-
̈
Riedle, R.; Trauner, D. Nat. Commun. 2014, 5, 5597−5601.
(11) Magnus, P.; Booth, J.; Diorazio, L.; Donohoe, T.; Lynch, V.;
Magnus, N.; Mendoza, J.; Pye, P.; Tarrant, J. Tetrahedron 1996, 52,
14103−14146.
ASSOCIATED CONTENT
* Supporting Information
■
S
(12) (a) Kametani, T. J. Org. Chem. 1982, 47, 2331−2342.
(b) Kametani, T.; Suzuki, K.; Nemoto, H. J. Am. Chem. Soc. 1981,
103, 2890−2891. (c) Rojas, G.; Wagener, K. B. J. Org. Chem. 2008, 73,
4962−4970.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.6b03635.
(13) Hodgson, D. M.; Salik, S.; Fox, D. J. J. Org. Chem. 2010, 75,
2157−2168.
Experimental procedures, computational details, Carte-
sian coordinates, absolute energies, and UV−vis spectra
(PDF)
(14) (a) Tanaka, A.; Tanaka, R.; Uda, H.; Yoshikoshi, A. J. Chem. Soc.,
Perkin Trans. 1 1972, 1721−1727. (b) Piers, E.; Britton, R. W.; De
1H and 13C spectra for all new compounds (PDF)
Waal, W. Can. J. Chem. 1971, 49, 12−19. (c) Furstner, A.; Hannen, P.
̈
Chem. - Eur. J. 2006, 12, 3006−3019. (d) Fehr, C.; Galindo, J. Angew.
Chem., Int. Ed. 2006, 45, 2901−2904.
AUTHOR INFORMATION
■
(15) Calculations were performed with Gaussian09 rev. D01: see the
SI for full details.
Corresponding Author
*E-mail: dderksen@ucalgary.ca.
ORCID
(16) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−
241. (b) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2009, 113, 6378−6396.
Robert S. Paton: 0000-0002-0104-4166
Darren J. Derksen: 0000-0002-5945-6921
Notes
(17) (a) Matute, R. A.; Houk, K. N. Angew. Chem., Int. Ed. 2012, 51,
13097−13100. (b) Matute, R. A.; Garcia-Garibay, M. A.; Houk, K. N.
́ ́
Org. Lett. 2014, 16, 5232−5234. (c) Jimenez-Oses, G.; Liu, P.; Matute,
R. A.; Houk, K. N. Angew. Chem., Int. Ed. 2014, 53, 8664−8667.
(18) Zimmerman, H. E.; Pasteris, R. J. J. Org. Chem. 1980, 45, 4864−
4876.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(19) Chen, X.; Rinkevicius, Z.; Luo, Y.; Ågren, H.; Cao, Z.
ChemPhysChem 2012, 13, 353−362.
(20) Peng, Q.; Duarte, F.; Paton, R. S. Chem. Soc. Rev. 2016, 45,
6093−6107. Selectivities were obtained from application of TST,
focusing solely on the triplet-state TSs.
D.J.D. holds the Alberta Children’s Hospital Foundation Junior
Chair in Medicinal Chemistry and is supported by funding from
the Alberta Children’s Hospital Research Institute and the
University of Calgary.
REFERENCES
■
(1) LeGay, C. M.; Gorobets, E.; Iftinca, M.; Ramachandran, R.;
Altier, C.; Derksen, D. J. Org. Lett. 2016, 18, 2746−2749.
(2) (a) Bach, T.; Hehn, J. P. Angew. Chem., Int. Ed. 2011, 50, 1000−
1045. (b) Blay, G.; Bargues, V.; Cardona, L.; Collado, A. M.; García,
D
DOI: 10.1021/acs.orglett.6b03635
Org. Lett. XXXX, XXX, XXX−XXX