Triquinanes: A One-Pot IMDA-tandem metathesis cascade strategy: Ring-closing metathesis (RCM) dominates norbornene ROM!

Article abstract of DOI:10.1021/ol100150f

A general route to linear triquinanes including a formal synthesis of (±)-δ⁹⁽¹²⁾-capnellene is described. This cascade strategy combines an intramolecular Diels-Alder reaction tandem metathesis protocol to generate the linear cis-anti-cis-tricyclo[6.3.0.0 ^2,6]undecane skeleton directly. Significantly, the ring-closing-ring-opening-cross-metathesis (RCM-ROM-CM) sequence with our norbornene adducts is the only observed mechanism.

Full text of DOI:10.1021/ol100150f

ORGANIC
LETTERS
2010
Vol. 12, No. 8
1684-1687
Triquinanes: A “One-Pot” IMDA-Tandem
Metathesis Cascade Strategy: Ring-Closing
Metathesis (RCM) Dominates Norbornene
ROM!
Natalie N. M. Nguyen,† Mathieu Lecle`re, Nicole Stogaitis, and Alex G. Fallis*
Department of Chemistry, UniVersity of Ottawa, 10 Marie Curie, Ottawa, Ontario K1N 6N5, Canada
alexander.fallis@uottawa.ca
Received January 20, 2010
ABSTRACT
A general route to linear triquinanes including a formal synthesis of (()-∆9(12)-capnellene is described. This cascade strategy combines an intramolecular
Diels-Alder reaction tandem metathesis protocol to generate the linear cis-anti-cis-tricyclo[6.3.0.02,6]undecane skeleton directly. Significantly, the
ring-closing-ring-opening-cross-metathesis (RCM-ROM-CM) sequence with our norbornene adducts is the only observed mechanism.
The fascinating array of diverse terpenoid structures have captured
the imagination of organic chemists beyond their commercial
potential. The polyquinanes, like the steroids and triterpenes of an
earlier era, provide an impressive arena for the development of
synthetic strategies in a facile and efficient manner. The triquinane
family contains three dominant structural motifs. Among these
variations, the linear cis-anti-cis-tricyclo[6.3.0.02,6]undecane skel-
Figure 1. Triquinane structural motifs.
eton in the sequiterpene ∆(9,12)-capnellene (Figure 1) has received
the most attention as a synthetic target. An impressive number of
creative approaches are known.1
closing-ring-opening (RCM-ROM-CM) pathway dominates.
This mechanism may be influenced partly due to the cyclopentane
ring attached to the bridgehead, but this is not the only factor, as
we have established with metathesis competition and labeling
experiments. Consequently, these insights, particularly the behavior
of the norbornene, may be more important than our synthesis! A
synthetic tradition is that discoveries en route to a molecule reveal
unexpected results. Of course, the Woodward-Hoffmann rules are
the most famous example.
Previously, we have developed controlled intramolecular
Diels-Alder reactions of subsutituted cyclopentadienes by tether
length restriction to generate the desired product from the 1,5-
sigmatropic rearrangement(s). Alternatively, hydrogen migration
could be inhibited by a spiro-cyclopropane moiety.4,5
We anticipated that our “one-pot” linear quinane synthesis would
follow an intramolecular Diels-Alder-Ring-Opening Metathesis-
Ring-Closing Metathesis (IMDA-ROM-RCM) cascade strategy
based on literature precedent and the inherent strain2 in norbornenes
(Scheme 1).3
Instead, we have established, contrary to current dogma with
most ruthenium catalysts, the alternative cross-metathesis ring-
† Present address: Merck Frost Canada, 16711 Trans Canada Hwy,
Kirkland, QC H9H 3L1, Canada.
(1) Reviews: (a) Singh, V.; Thomas, B. Tetrahedron 1998, 54, 3647.
(b) Mehta, G. A.; Srikrishna, A. Chem. ReV. 1997, 97, 671(see the
Supporting Informationfor total syntheses of (()-D9(12)-capnellene). (c)
For the most recent synthesis, see: Lemie`re, G.; Gandon, V.; Cariou, K.;
Hours, A.; Fukuyama, T.; Dhimane, A.-L.; Fensterbank, L.; Malacria, M.
J. Am. Chem. Soc. 2009, 131, 2993, and references cited therein.
(2) Recent G3 ring strain energy calculations for norbornene (18.8 kcal/
mol) norbornadiene (27.6 kcal/mol): Howell, J.; Goddard, J. D.; Tam, W.
Tetrahedron 2009, 65, 4562, and references therein.
Based in part on this experience, we have developed a direct
assembly of linear quinanes from the alkylation of the cyclopen-
tadienyl anion with an appropriate tosylate 4 to afford the tricyclic
10.1021/ol100150f  2010 American Chemical Society
Published on Web 03/15/2010

Scheme 1. Retro-Synthetic IMDA Tandem Metathesis to Linear
cis-anti-cis Triquinanes
Scheme 2. Synthesis of Cyclopentadienes 5a and 5b
structures 7. It is preferable to use the isomeric cyclopentadienes
5 (Scheme 2), from the alkylation, directly in the intramolecular
cycloaddition to the adducts 6 (microwave). The vessel is opened,
and a Grubbs catalyst is added to effect the anticipated tandem
ring-opening metathesis of the norbornenes followed by ring-
closing metathesis with the allyl substituent to generate the fused
tricyclic cyclopentenes 7 (Scheme 1).
the desired side chain tosylates 4, and the cyclopentadienyl anion
generated with NaH was exposed to 4a and 4b individually to
afford 5a/5b (99%). The isomer equilibrium ratio (room temper-
ature, 22 °C) in each series was ∼0.8:1.0 (1H NMR). This is
irrelevant as the cycloaddition preference is dictated by the tether
length. The adduct from the C2 isomer violates Bredt’s rule, while
an unactived olefin will not react with the C5 isomer. It was not
detected by 1H NMR spectroscopy.
We have demonstrated previously the advantage that microwave
irradiation is a superior heat source for thermally challenging IMDA
reactions.5,9 Modern microwave instruments have replaced
kitchen models.10,11 Consequently, this is the best method
for IMDA reaction of unactivated cyclopentadienes. Expo-
sure of 5a in chlorobenzene to microwave radiation equipped
with a spherical glass-coated Carboflon (for energy transfer)12
at 210 °C under 310 psi pressure for 1.5 h afforded the
desired adduct 6a in 80% yield (Table 1, entry 1).
We also conducted these reactions stepwise (cycloaddition
workup, subsequent tandem metathesis) to establish the best
conditions and unravel the metathesis pathways.
The 1,4-diene-ester 9a6 was prepared by orthoester Claisen
rearrangement of the corresponding 1,5-hexadien-3-ol (8a, Aldrich)
with catalytic propionic acid or phenol in the case of 8b7 to afford
9b8 in refluxing triethyl orthoacetate. The esters 9 were reduced
individually to the alcohols 10 with LiAlH4, tosylation afforded
(3) (a)Tandem norbornene metatheses: (a) Stille, J. R.; Santarsiero, B. D.;
Grubbs, R. H. J. Org. Chem. 1990, 55, 843. (b) Stragies, R.; Blechert, S.
Synlett 1998, 169. (c) Wrobleski, A.; Sahasrabudhe, K.; Aube´, J. J. Am.
Chem. Soc. 2002, 124, 9974. (d) Arjona, O.; Csakye, A. G.; Medel, R.;
Plumet, J. J. Org. Chem. 2002, 67, 1380. (e) Minger, T. L.; Phillips, A. J.
Tetrahedron Lett. 2002, 43, 5357. (f) Hart, A. C.; Phillips, A. J. J. Am.
Chem. Soc. 2006, 128, 1094. (g) Maechling, S.; Norman, S. E.; Mckendrick,
J. E.; Basra, S.; Koppner, K.; Blechert, S. Terahedron. Lett. 2006, 47, 189.
(h) Malik, C. K; Ghosh, S. Org. Lett. 2007, 9, 2537. (i) Maity, S.; Ghosh,
S. Tetrahedron Lett. 2008, 49, 1133. (j) Malik, C. K.; Yadav, R. N.; Drew,
M. G. B.; Ghosh, S. Tetrahedron Lett. 2009, 50, 5277. (k) , C. K.; Hossain,
Md. F.; Ghosh, S. J. Org. Chem. 2009, 74, 1957, and references cited therein.
(l) Hetero-norbornenes: Lui, Z.; Rainier, J. D. Org. Lett. 2005, 6, 131. (m)
Lui, Z.; Rainier, J. D. Org. Lett. 2006, 8, 459. (n) Carreras, J.; Avenoza, A.;
Busto, J. H.; Peregrina, J. M. Org. Lett. 2007, 9, 1235. (o) Calvet, C.;
Blanchard, N.; Kouklovsky, C. Org. Lett. 2007, 9, 1485. (p) Reviews: Arjona,
O.; Csaky, A. G.; Plumet, J. Eur. J. Org. Chem. 2003, 611. (q) Astruc, D.
New J. Chem. 2005, 29, 42. (r) Nicolaou, K. C.; Bulger, P. G.; Sarlah, D.
Angew. Chem., Int. Ed. 2005, 44, 4490. (s) Blechert, S.; Holub, N. Chem.
Asian J. 2007, 2, 1064.
Addition of a Grubbs catalyst (22 °C) (entries 4 and 5) under
ethylene assisted the ring-opening-ring-closing metathesis condi-
tions to afford triquinanes 7a and 7b directly (98-85%). The “one-
pot” reactions (entries 7-9) are less efficient, a consequence of
the absence of ethylene (entry 3) and trace cycloaddition impurities
that reduced the catalytic efficiency. Grubbs I catalysts were more
efficient than the second generation. The reduced yield of 6b (entry
2) reflects the steric interactions induced by geminal methyl groups.
COSY, HMQC, and NOESY data for 6 confirmed the regiochem-
istry and cis-anti-cis ring stereochemistry.
(4) (a) Breitholle, E. G.; Fallis, A. G. Can. J. Chem. 1976, 54, 1991.
(b) Breitholle, E. G.; Fallis, A. G. J. Org. Chem. 1978, 43, 1964. (c) Be´rube´,
G.; Fallis, A. G. Tetrahedron Lett. 1989, 30, 4045. (d) Be´rube´, G.; Fallis,
A. G. Can. J. Chem. 1991, 69, 77.
The reactions required to convert 7b to (()-∆9(12)-capnellene
are established. A parallel allylic oxidation enone reduction method
(5) (a) Attah-Poku, S. K.; Alward, S. J.; Fallis, A. G. Tetrahedron Lett.
1983, 24, 681. (b) Attah-Poku, A. K.; Gallacher, G.; Ng, S. S.; Taylor,
L. E. B.; Alward, S. J.; Fallis, A. G. Tetrahedron Lett. 1983, 24, 677. (c)
Alward, S. J.; Fallis, A. G. Can. J. Chem. 1984, 62, 121. (d) Attah-Poku,
S. K.; Antczak, K.; Alward, S. J.; Fallis, A. G. Can. J. Chem. 1984, 62,
1717. (e) Gallacher, G.; Ng, A. S.; Attah-Poku, S. K.; Antczak, K.; Alward,
S. J.; Kingston, J. F.; Fallis, A. G. Can. J. Chem. 1984, 62, 1709. (f) Antczak,
K.; Kingston, J. F.; Fallis, A. G. Can. J. Chem. 1985, 63, 993. (g) Antczak,
K.; Kingston, J. F.; Fallis, A. G.; Hanson, A. W. Can. J. Chem. 1987, 65,
114.
(9) (a) Lei, B.; Fallis, A. G. J. Am. Chem. Soc. 1990, 112, 4609. (b)
Lei, B.; Fallis, A. G. J. Org. Chem. 1993, 58, 2186
.
(10) (a) Lu, Y.-F.; Fallis, A. G. Tetrahedron Lett. 1993, 34, 3367. (b)
Lu, Y.-F.; Fallis, A. G. Can. J. Chem. 1995, 73, 2239. (c) Clay, M. A.;
Fallis, A. G. Can. J. Chem. 2005, 83, 559
.
(11) Microwave-assisted organic synthesis: (a) Kappe, C. O. Angew.
Chem., Int. Ed. 2004, 43, 6250. (b) Loupy, A.; Perreux, L. Tetrahedron
2001, 57, 9199. (c) Lidstro¨m, P.; Tierny, J.; Wathey, B.; Westman, J.
Tetrahedron 2001, 57, 9225. (d) Loupy, A.; Petit, A.; Hamelin, J; Texier-
Boullet, F.; Jacquault, P.; Mathe´, D. Synthesis 1998, 1213. (e) Majetich,
(6) Chattopadhyay, A.; Mamdapur, A. V. R. Indian J. Chem. 1988, 27,
169.
(7) (a) Antonsson, T.; Moberg, C.; Tottie, L. J. Org. Chem. 1989, 54,
4914. (b) Gau, A.-H.; Lin, G.-L.; Uang, B.-J.; Liao, F.-L.; Wang, S.-L. J.
Org. Chem. 1999, 64, 2194.
G. R.; Hichs, J. J. MicrowaVe Power 1995, 30, 27
.
(8) Kondo, K.; Matsui, K.; Takahatake, Y. Tetrahedron Lett. 1976, 48,
4359.
(12) The glass encased Carboflon absorbs microwave energy, and the
temperature and pressure are monitored by a fiber optic probe.
Org. Lett., Vol. 12, No. 8, 2010
1685

Table 1. Stepwise and “One-Pot” IMDA-Metathesis Tandem Reactions
entry
reactants
solvent
catalyst
temp, °C (psi)
time (h)
yielda (%)
1
2
3
4
5
6
7
8
9
DA-(5a-6a)
DA-(5b-6b)
Met-(6a-7a)
Met-(6a-7a)
Met-(6b-7b)
Met-(6b-7b)
DAM(5a-7a
DAM(5a-7a
DAM(5b-7b
PhCl
PhCl
PhH
PhH
PhH
PhH
PhCl
PhMe
PhCl
210 (310)
210 (310)
22 (atm)
22 (atm)
22 (atm)
22 (atm)
210 (310)
200 (200)
210 (310)
1.5
1.5
2
2
1
1.5
2 + 2
2 + 1
2 + 2
80
54
85
Grubbs I
Grubbs I
Grubbs I
Grubbs II
Grubbs I
Grubbs II
Grubbs I
98b
98b
85b
65
15
42
a Isolated yield. b Reaction run with 1 atm of ethylene.
was used for ceratopicanol.13 Grubbs and Stille reduced the vinyl
group to the methyl substituent in capnellene,14 and installation of
the exo-cyclic employed a Wittig reaction15 or Tebbe reagent14 to
reach the target.
Scheme 3. Box A: Anticipated Steps in ROM-RCM Tandem
Pathway. Box B: Key Steps Observed in RCM-ROM-CM
Pathway
Norbornene Metathesis: A Conundrum Is Resolved.
ROM is driven enthalpically by ring strain release while RCM is
entropically driven by ethylene production.16 Alkene substitution
decreases the olefin metathesis rate. Thus, initial metathesis occurs
preferentially at monosubstituted acyclic alkenes rather than di-
substituted, cyclic olefins, depending on ring size and strain. In
principle, the metatheses reactions from adducts 6 may follow
different mechanistic pathways, whether the Ru-metallo-cyclobu-
tane is generated at the strained, norbornene double bond or the
allyl group preferentially. Previous research demonstrated the
norbornene ring strain often activates the cyclic double bond in
favor of another alkene or allyl group.16
tive C7-substituted examples are described in their (+)-africanol
synthesis.18b Although the catalysts and steric environments differ
from the Ru experiments in Table 1, this is consistent with our
conclusions below for the endo-attachment of the allyl group.
Koreeda and Holtsclaw19 observed that endo-face attack was
absent with 7,7-disubstituted norbornenes (Grubbs I or II catalysts).
They confirmed the significant steric interference of the cross ring
endo-hydrogens. When a Ru-alkylidene species was generated on
a C7 syn-allyl tether ROM attacked the top face of the norbornene
olefin.
Based on these precedents and lack of C7 substitution,
we anticipated that the initial ruthenium-alkylidene species
A initiated the ring-opening cascade from adduct 6a by
preferential association with the norbornene (B). Subsequent
regioselective, cycloaddition to C was influenced by the
cyclopentane ring. ROM to D and subsequent RCM of the
allyl moiety via intermediate E should afford 7a (Scheme
3, box A).
Moore and Rule17 examined the ROMP reactivity of endo- and
exo-dicyclopentadiene and norbornene. The rate constants were k
) 0.019, 0.37, and 0.43 s-1, respectively, and established the rate
retardation of an endo-substituent. Hoveyda, Schrock, and co-
workers examined 7-syn-(1-butenyl)-7-anti-(1-oxaallyl)norbornene
in competitive metathesis ARCM-AROM transformations.18a In
all cases, the RCM between the two C7 terminal alkenes dominated
except with their unique adamtanylimido Mo complex for which
ROM of their norbornene was preferred (4:1). Additional, competi-
(13) Baralotto, C.; Chanon, M.; Julliard, M. J. Org. Chem. 1996, 61,
3576.
(14) Stille, J. R.; Grubbs, R. H. J. Am. Chem. Soc. 1986, 108, 855. This
research also utilized the titanocene alkylidene complex to open their
norbornene synthon.
(15) (a) Sternbach, D. D.; Hughes, J. W.; Burdi, D. F. J. Org. Chem.
1984, 49, 201. (b) Sternbach, D. D.; Hughes, J. W.; Burdi, D. F.; Banks,
B. A. J. Am. Chem. Soc. 1985, 107, 2149.
(16) Zuercher, W. J.; Hashimoto, M.; R. Grubbs, R. H. J. Am. Chem.
Soc. 1996, 118, 6634.
Others have suggested that it is often not obvious with strained
cyclic olefins whether ring cleavage always occurs first.3a,r,q,18b
Consequently, we suspected some of the double-metathesis product
might be initiated at the allyl moiety, particularly if this pathway
was facilitated by potential endo-stabilization in complex G from
(17) Rule, J. D.; Moore, J. S. Macromolecules 2002, 35, 7878.
(18) (a) Tsang, W. C. P.; Jernelius, J. A.; Cortez, G. A.; Weatherhead,
G. S.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2003, 124, 2591.
(b) Weatherhead, G. S.; Cortez, G. A.; Scrock, R. R.; Hoveydayda, A. M.
Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5805.
(19) Holtsclaw, J.; Koreeda, M. Org. Lett. 2004, 6, 3719.
1686
Org. Lett., Vol. 12, No. 8, 2010

Scheme 4. Truncated Route to Deuterated Compounds
metallocycle F (Scheme 3, box B). This mechanism requires either
sufficient ethylene or a related alkene source (7a?) be available
for completion of the final metathesis step to the vinyl product.
These suspicions were confirmed by low-emperature NMR
experiments, deuterium-labeling studies (Scheme 4), and competi-
tive metatheses reaction comparisons.
Figure 2.
1H NMR spectra of alkylidene ruthenium complexes using 6a
(a), 6a-D3 (b), or a 1/1 mixture of 1-hexene and norbornadiene (c).
Labeling experiments confirmed the dominance of the
initial allyl RCM reaction. Deuterated adduct 6a-D3 and
triquinane 7a-D3 were synthesized from 11,24 and the -18
°C experiment was repeated. The absence of the ruthenium
At -78 to -35 °C, the initial metatheses with 6a were
completely suppressed (CD2Cl2). The reagents were mixed (NMR
tube, -78 °C, no CH2dCH2). The spectra were recorded at -18
°C, the temperature at which the reaction commenced.20 After 5
1
alkylidene hydrogen signal in the H NMR spectrum in
Figure 2 (spectrum b) proved the initial ring-metathesis
reaction had indeed proceeded via a ring-closing route to F,
followed by the complexed intermediate G to H to I to
quinane 7a-D3. This proves the mechanism in Scheme 3,
box B, is the only one involved in these transformations.
In summary, a cascade intramolecular Diels-Alder-tandem
metathesis sequence provided a rapid, stereospecific entry to
triquinanes (cis-anti-cis-tricyclo[6.3.0.02,6]undecanes).25 Our experi-
ments established the ring-closing metathesis-ring-opening me-
tathesis (RCM-ROM) allyl group pathway and excluded the
norbornene ROM alternative. Consequently, ring strain2 in nor-
bornenes may be a less important driving force than previously
assumed for metathesis initiation with ruthenium catalysts.
1
min, H NMR analysis revealed the presence of the alkylidene
ruthenium complex G from reaction with the vinyl side chain as
a triplet signal at 19.2 ppm (J ) 5.0 Hz), in addition to catalyst
signal(s). This corresponds to the pattern expected from G, not B
(Figure 2, spectrum a).21,22
Competition experiments established a terminal alkene
(allyl) was more reactive than norbornadiene at -18 °C.
Norbornadiene was selected due to its symmetry and higher
strain energy,2 plus the absence of endo-H nonbonded
interactions compared to norbornene. Parallel and blank
experiments with equimolar norbornadiene and 1-hexene
revealed the alkylidene ruthenium complex with the allyl
group dominated the norbornadiene ring opening, despite two
reactive sites, by a ratio of ∼6:1 (Figure 2, spectrum c).23
This implies that bicyclo[2.2.1]heptenes react more slowly
with a metathesis catalyst than generally assumed. The
reactivity effects induced by the steric environment of a
double bond are difficult to predict. A competition between
3,3-dimethylbutene and norbornadiene at -18 °C established
both were inert. No alkylidene ruthenium complexes were
Acknowledgment. We are grateful to the Natural Sciences
and Engineering Research Council of Canada (NSERC) for
financial support of this research, S. K. Collins (University of
Montreal) for fruitful discussions, and the reviewers for their
constructive suggestions.
Supporting Information Available: Experimental details
and spectral characterization for obtained compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
1
detected by H NMR spectroscopy.
OL100150F
(20) A solution of 6a or 6a-D3 (5 mg, 29 µmol) in 0.3 mL of degassed
CD2Cl2 was cooled to -78 °C. A solution of benzylidenebis(tricyclohexyl-
phosphine)dichlororuthenium (24 mg, 29 µmol) in 0.4 mL of degassed
CD2Cl2 was slowly added with vigorous stirring. The suspension was then
transferred in a NMR tube and quickly introduced to the precooled
(-18 °C) NMR probe.
(23) These results indicate the rate of the initial complexation-metallo-
cycloaddition step determines the preferred pathway. This was confirmed
by the detection of both competitive species. Consequently, the potential
back reactions are not a complication.
(24) Synthesis of 11: (a) Zibuck, R.; Streiber, J. Org. Synth. 1992, 71,
236. (b) Sannigrahi, M.; Mayhew, D. L.; Clive, D. J. J. Org. Chem. 1999,
64, 2776. Deuteride reduction: Kalvim, D. M.; Woodard, R. W. Tetrahedron
Lett. 1984, 40, 3385.
(21) The alkylidene ruthenium complex has been isolated (X-ray) from
the reaction of a specially selected tricyclobutene with a full equivalent of
catalyst: Tallarico, J. L.; Bonitatebus, P. J., Jr.; M. Snapper, M. L. J. Am.
Chem. Soc. 1997, 119, 5157.
(25) Crinipellins possess a rare angular tetraquinane skeleton for which only
one total synthesis of (()-crinipellin is known: Piers, E.; Renaud, J. J. Org.
Chem. 1993, 58, 11. A second allyl group in the cyclopentadienyl tether should
facilitate the extension of our strategy to tetraquinane frameworks.
(22) Our result assumes the intermediates in the postulated pathways
have sufficient lifetimes to be detected on the NMR time scale, although
the back reactions may be fast.
Org. Lett., Vol. 12, No. 8, 2010
1687