Structure and reactivity of a chiral cyclononadienone

Article abstract of DOI:10.1016/j.tetlet.2009.02.017

The structure and reactivity of a highly enantioenriched dissymmetric cyclononadienone are described.

Full text of DOI:10.1016/j.tetlet.2009.02.017

Tetrahedron Letters 50 (2009) 1882–1885
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Structure and reactivity of a chiral cyclononadienone
*
Yue Zhang, Stephen D. Lotesta, Thomas J. Emge, Lawrence J. Williams
Contribution from the Department of Chemistry and Chemical Biology, Rutgers, The State University of New Jersey, Piscataway, NJ 08854, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The structure and reactivity of a highly enantioenriched dissymmetric cyclononadienone are described.
Received 1 January 2009
Revised 1 February 2009
Accepted 2 February 2009
Available online 8 February 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Synthesis
Nine-membered rings
Caryophylloids
Xenicanes
The problem of the synthesis of nine-membered cycloalkanes
primarily stems from destabilizing trans-annular interactions.1
Most synthetic strategies to prepare cyclononanes and their deriv-
atives deal with this structural congestion by initial formation of a
smaller ring to which a second ring is then fused (A?B, Scheme
1).2 In a subsequent maneuver, the small ring connectivity is re-
moved and the nine-membered ring connectivity is left intact
(B?C). The related strategy of temporary macrocycle formation
and then ring contraction to give a nine-membered ring is also
known.2b The 1964 and 2008 chemical syntheses of b-caryophyl-
lene by Corey and coworkers constitute the first and most recent
preparation of this nine-membered ring-containing natural prod-
uct and demonstrate the logic represented in A?B?C.3 The 2008
synthesis utilized ketone 1. Here, we describe our studies related
to the structure and reactivity of (À)-cyclononadienone 1, a nine-
membered ring-containing molecule that lacks the archetypal ste-
reogenic center or axis of chirality but is, nevertheless, chiral.
Our interest in the synthesis of nine-membered ring-containing
natural products was piqued by the recognition that strategies to
access such targets are lacking. We were further attracted to this
problem by reports of isolates, for example, 4, from the soft coral
Xenia elongata identified through the use of a novel assay for apop-
totic activity in cancer cells.4 Most of the molecules in this class,
including those that lack an isolated nine-membered ring, as in
umbellacins C (5) and D5 (6), have not been studied in detail due
to the paucity of materials obtained from the organisms that pro-
duce them coupled with the difficulty associated with their chem-
ical synthesis.
Cope and others recognized that in order for trans-cyclononene
and related ring systems to racemize, they must undergo a
potentially slow conformational interconversion (3?ent-3).6 This
X
X
B
( )m
( )m
( )n
A
C
O
Me
Me
O
O
Me
con-1
1
ent-1
Me
O
3
ent-3
2
Me
Me
Me
OH
Me
Me
OH
Me
Me
Me
HO
O
Me
OH
H
H
O
AcO
AcO
H
H
O
MeO
OMe
O
5
O
4
6
new isolate
umbellacin C
umbellacin D
* Corresponding author.
E-mail address: ljw@rutchem.rutgers.edu (L.J. Williams).
Scheme 1.
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.02.017

Y. Zhang et al. / Tetrahedron Letters 50 (2009) 1882–1885
1883
substance was shown to have a racemization half-life of approxi-
mately 4 min at 0 °C, and 6 s at 25 °C. We wondered if more func-
tionalized constructs, for example, 1 and analogous structures (e.g.,
2), would racemize more slowly than the parent compound. In the
case of 1, the racemization process could well be step-wise, for
example, conversion of 1 to con-1 and then to ent-1. In favorable
cases, species such as 1 could be prepared directly in enantioen-
riched form or resolved by dynamic kinetic methods. Importantly,
stereoselective transformations of this substance should be possi-
ble when racemization is slow. Hence, this strategy aims to convert
traditional chirality into conformational chirality7 and then to new
center chirality.
We prepared 1 and initiated the study of its structure and reac-
tivity. Originally, our synthesis focused on direct cyclization of an
acyclic precursor (8, Scheme 2). The presence of the trans-olefin
in 10 complicates this strategy, since the olefin geometry necessar-
ily projects the termini in opposite directions. Still, this route ben-
efits from avoidance of transannular strain. The single flask
conversion of known diyne 9 by way of a carbometalation/trans-
metalation/conjugate addition sequence8 gave ynal 10 (30%, unop-
timized). Unfortunately, cyclization of the desilylated ynal gave
complex mixtures of products that did not include the desired
cyclononenes (8?2, 7).
O
O
Me
Me
O
Me
OH
a
b
c
O
TBSO
HO
OH
OH
11
12
13
OH
OTs
Me
Me
OH
Me
OH
O
X e
d
H
TBSO
TBSO
OTBS
Me
14
15
O
O
OH
Me
Me
g
f
HO
OH
12
OH
17
OH
16
OTs
Me
OH
Me
h
O
i
18
1
An alternative route to enantioenriched 1 is shown in Scheme 3.
Use of ketone 11 followed by eventual fragmentation would enable
access to enones of type 1. At the outset of this work, it was unclear
that the dissymmetric structure would be sufficiently stable to be
converted stereoselectively to enantioenriched products; however,
this route coincided with that of Corey and Lavinov in their elegant
(and stereoselective!) syntheses of caryophyllene and coriaxenio-
lide.3c Briefly, ketone 11 was prepared in enantioenriched form,9
reduced, protected, reduced again (?14), and then activated for
fragmentation (?15).10 Tosylate 15 did not undergo fragmentation
under standard conditions11 or at elevated temperatures. Although
the geometry of 15 is not ideal for fragmentation,12 it was not obvi-
ous that the deviation accounted for this failure. Nevertheless, we
examined olefinic analog 17,13 which appeared to have a slightly
improved geometry for fragmentation.14 Pleasingly, treatment of
18 under standard conditions gave the desired cyclononadienone
1.
Scheme 3. Reactions and conditions: (a) NaBH4, methanol, 0 °C, 77%; (b) TBSCl,
imidazole, DMAP, CH2Cl2, 0 °C to rt, 82%; (c) Li(s), NH3(l), reflux, 47%; (d) TsCl,
pyridine, DMAP, CHCl3, 0 °C to rt, 70%; (e) KH, THF, rt to reflux; NaH, THF, rt to
reflux; NaH, DMF, rt; NaH, DMSO, rt; (f) (i) MsCl, TEA, 0 °C; (ii) DBU neat, 80 °C, 53%
(from 12); (g) Li(s), NH3(l), reflux, 40%; (h) TsCl, pyridine, DMAP, CHCl3, 0 °C to rt,
80%; (i) NaH, DMSO, rt, 60%.
The optical rotation [
a
D = À17.0] is in agreement with the earlier
report. The highest directly observed enantiopurity of 1 was 87%
ee. This was achieved by chiral HPLC analysis immediately upon
extraction from the fragmentation reaction followed by filtration
through a plug of silica gel. Alternatively, oxidative trapping of 1
with DMDO followed by flash column chromatography gave epox-
ide 1913 in 90% ee. Although a small degree of racemization takes
place during HPLC analysis, this method was used to evaluate the
stability of 1. Much slower than trans-cyclononene, compound 1
racemizes with t1/2 = 32 h, 23 °C (krac = 3.0 Â 10À6 sÀ1, Table 1 and
Fig. 1).
Further analysis provided several insights into the structure and
reactivity of 1. This substance is a low melting crystalline solid that
is difficult to manipulate in crystalline form, and we have been
unsuccessful in obtaining crystallographic data for this material.
NMR analysis strongly suggests that the structure represented
as 1 is the dominant conformer in solution.15 Presumably, 1 race-
mizes by way of con-1, though we have not observed this or other
related species. For example, variable temperature 1H NMR analy-
sis of 1 proved invariant from 25 °C to 55 °C. Consistent with these
observations, computational analysis suggests that 1 is more stable
than con-1 by 4.2 kcal/mol.16
As indicated above, DMDO epoxidation gave 19 in high yield
(96%). Indeed, dissymmetric ketone 1 participates in a number of
stereoselective transformations. Our initial attempts focused on
Diels–Alder reaction in the presence of cyclopentadiene, which
did not react at room temperature. Lewis acids Me2AlCl, Et2AlCl,
O
Me
8
Me
1 or 2
TMS
O
7
Me
a
c
O
TMS
EtZn
9
10
Me
Table 1
Racemization of 1
b
8
7
10
O
Entry
Time (h)
eea,b (%)
1
2
3
4
5
0
84.4
74.2
68.0
51.0
40.6
5.5
9.25
23
2
33.5
Scheme 2. Reactions and conditions: (a) (i) Cp2ZrCl2, AlMe3, (CH2Cl)2, 25 °C; (ii)
CuLi(hexynyl)2, acrolein, HMPA, TMSCl, À78 °C to À30 °C, 30%; (b) TBAF, AcOH, THF,
0 °C, 93%; (c) dicyclohexylboran, hexanes, diethylzinc, 0 °C.
a
Based on a reversible reaction model, y = Àln(ee%).
Racemization rate found: k = 3.0 Â 10À6 sÀ1 (298 K).
b

1884
Y. Zhang et al. / Tetrahedron Letters 50 (2009) 1882–1885
are derived from Cope rearrangement of 1 followed by isomeriza-
-ln(ee%)
1
tion. Stereoselective hydride reduction of 1 gave the corresponding
alcohol 2213 as a crystalline solid with no deterioration in enantio-
meric purity, according to Mosher ester analysis (23).20 Allylation
produced 24 (75%). Compound 24 failed to undergo anion-acceler-
ated oxy-Cope rearrangement at low temperature, and at higher
temperature, only the transannular Cope product, trisubstituted
cyclopentene 26, was obtained. To further demonstrate the versa-
tility of 1, and to evaluate a potential alternative oxy-Cope se-
quence, 1 was epoxidized and then allylated to give 27. Upon
subjection to anion-accelerated oxy-Cope conditions, the transan-
nular epoxide-opened product 28 was cleanly obtained (90%).
Transannular cyclization of 19 to 28 occurs slowly at low temper-
ature. Indeed, when the allylation of 1 was run at 0 °C, 28 was ob-
tained in good yield (76%)21 (see Scheme 4).
These studies elaborate earlier findings on cyclononadienone 1,
including several stereoselective transformations and insight into
the structure of this interesting class of synthetically useful inter-
mediates. Importantly, 19–29 are derived from 1, and are obtained
in enantioenriched form. Many dissymmetric compounds related
to 1 should be readily accessible.
•
y=(0.0217)t+0.1767
0.8
0.6
0.4
R
2 =0.9996
•
•
•
0.2
•
0
0
10
20
time (h)
30
40
Figure 1. Plot of Àln(ee%) versus time.17
and AlCl3 also failed to promote this reaction.18 No cyclopentadi-
ene cycloaddition product was observed at elevated temperatures.
Acknowledgment
Instead, 20 and 21 were obtained.19 These
a,b-unsaturated ketones
Generous financial support from Merck & Co. is gratefully
acknowledged.
Me
O
O
a
References and notes
1. Prelog, V.; Schenker, K. Helv. Chim. Acta 1952, 35, 2044–2053; Still, W. C.;
Galynker, I. Tetrahedron 1981, 37, 3981–3996.
19
2. (a) Tatsuta, K.; Hosokawa, S. Chem. Rev. 2005, 105, 4707–4729; (b) Ohtsuka, Y.;
Niitsuma, S.; Tadokoro, H.; Hayashi, T.; Oishi, T. J. Org. Chem. 1984, 49, 2326–
2332.
3. (a) Corey, E. J.; Mitra, R. B.; Uda, H. J. Am. Chem. Soc. 1963, 85, 362–363; (b)
Corey, E. J.; Mitra, R. B.; Uda, H. J. Org. Chem. 1964, 86, 485–492; (c) Larionov, O.
V.; Corey, E. J. J. Am. Chem. Soc. 2008, 130, 2954–2955.
4. Andrianasolo, E. H.; Haramaty, L.; Degenhardt, K.; Mathew, R.; White, E.; Lutz,
R.; Falkowski, P. J. Nat. Prod. 2007, 70, 1551–1557.
5. El-Gramal, A. A. H.; Wang, S.-K.; Duh, C.-Y. J. Nat. Prod. 2006, 69, 338–341.
6. Cope, A. C.; Banholzer, K.; Keller, H.; Pawson, B. A.; Whang, J. J.; Winkler, H. J. S.
J. Am. Chem. Soc. 1965, 87, 3644–3649.
7. It has long been recognized that conformers will be chiral in the absence of free
rotation. van’t Hoff, J. H. The Arrangement of Atoms in Space; Longmans, Green,
and CO: 39 Paternoster Row, London, New York and Bombay, 1898. p 54.
8. Lipshutz, B. H.; Ellsworth, E. L. J. Am. Chem. Soc. 1990, 112, 7440–7441.
9. Hajos, Z. G.; Parish, D. R. Org. Synth. Col. 1990, 7, 363–368.
10. The route to 1 was not optimized in light of Ref. 3c.
11. Grob fragmentation conditions, see also Ref. 3a (a) Paquette, L. A.; Yang, Jiong.;
Long, Y. O. J. Am. Chem. Soc. 2002, 124, 6542–6543; (b) Winkler, J. D.; Quinn, K.
J.; MacKinnon, C. H.; Hiscock, S. D.; McLaughlin, E. C. Org Lett. 2003, 5, 1805–
1808; (c) Molander, G. A.; Huérou, Y. L.; Brown, G. A. J. Org. Chem. 2001, 66,
4511–4516.
Xb
no Diels-Alder product observed
Me
Me
Me
Me
O
c
+
Me
O
O
1
20
21
O
22, R = H
23, R =
d
e
F3C
MeO
RO
Me
H
Ph
Me
HO
Me
X g
f
O
12. Molecular modeling suggests that 15 is highly constrained and that the
(HO)CCCO(Ts) torsion angle is 139° (see Ref. 16).
H
24
25
13. This compound was characterized by single crystal X-ray diffraction.
Crystallographic data for compound 17, 19, and 22 have been deposited with
the Cambridge Crystallographic Data Center, Nos. CCDC 712607 (17), CCDC
712606 (19), and CCDC 712605 (22). Copies of the data can be obtained, free of
charge, on application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK, (fax:
+44-(0)1223-336033 or e-mail: deposit@ccdc.cam.ac.Uk).
Me
h
24
19
OH
26
Me
14. Molecular modeling suggests that 18 is more highly constrained than 15, and
the (HO)CCCO(Ts) torsion angle for 18 is better suited for fragmentation (165°)
(see Ref. 16).
O
Me
OH
15. NOSY analysis shows the olefin protons as proximal, cf. 3c.
16. Calculations used (DFT-B3LYP 6-31G(d, p)). (a) All structures were fully
optimized by analytical gradient methods using the GAUSSIAN 03 suites, (a)
Frisch, M. J.; Trucks, G. W.; Schlege, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.,
et al GAUSSIAN 03, Revision E.01; Gaussian: Wallingford, CT, 2004; Density
functional (DFT) calculations used the exchange potentials of: (b) Becke, A. D. J.
Chem. Phys. 1993, 98, 5648; the correlation functional of: (c) Lee, C.; Yang, W.;
Parr, R. G. Phys. Rev. B 1988, 37, 785.
17. Bada, J. L.; Protsch, R. Proc. Nat. Acad. Sci. 1973, 70, 1331–1334. We define half-
life as ee = 50%.
18. (a) Murray, L. M.; O’Brien, P.; Taylor, R. J. K. Org. Lett. 2003, 5, 1943–1946;
(b) Jeroncic, L. O.; Cabal, M-P.; Danishefsky, S. J. J. Org. Chem. 1991, 56,
O
HO
i
j
27
28
Scheme 4. Reactions and conditions: (a) DMDO, CHCl3, 0 °C, 96%; (b) cyclopenta-
diene, rt/Lewis acid (see text), toluene, À78 °C to rt; (c) cyclopentadiene, toluene,
130 °C, 60%, 2:1;3 (d) NaBH4, methanol, 0 °C, 93%; (e) (R)-Mosher’s acid, DCU, DMAP,
CH2Cl2, rt, 80%; (f) allylmagnesium chloride, THF, 0 °C to rt, 75%; (g) KH, 18-crown-
6, THF, reflux; (h) toluene, 130 °C; (i) allylmagnesium chloride, THF, À78 °C, 75%; (j)
KH, 18-crown-6, THF, À78 °C to rt, 90%.

Y. Zhang et al. / Tetrahedron Letters 50 (2009) 1882–1885
1885
387–395; c Although these Lewis acids failed, trityl perchlorate was effective at
promoting conjugate addition to this enone (Ref. 3c).
19. Assignments of 20 and 21 were based on the enone proton signals
(5.88 ppm for cis, 20, and 6.72 ppm for trans, 21) by analogy to: Taber, D. F.;
Jiang, Q.; Chen, B.; Zhang, W.; Campbell, C. L. J. Org. Chem. 2002, 67, 4821–
4827.
(22 mg, 0.146 mmol) in THF (0.5 mL) at 0 °C was added allylmagnesium
chloride (2.0 M, in THF, 0.11 ml, 0.22 mmol). After 1 h, the reaction was
quenched with saturated NH4Cl, extracted with ether (3 Â 10 mL), the organic
fractions were combined, dried over Na2SO4, filtered, concentrated, and then
the residue was purified by flash chromatography (3% EtOAc/hexanes) to yield
alcohol 24 (21 mg, 75% yield) as a colorless oil. IR m
max(neat)/cmÀ1 3562, 2924,
20. For example, reduction of 1 (in one instance enone of 60% ee was used)
provided 22 from which 23 was prepared as a mixture of diastereomers
(dr = 4:1, corresponding to 60% ee of 22). Crystallographic structure
determination of 22 revealed this material to be a disordered solid composed
of both enantiomers (see Ref. 13).
1725, 1638, 1448, 996, 915, 743; dH (500 MHz, CDCl3) 5.83 (1H, dddd, J = 17.0,
10.2, 8.5, 6.1 Hz), 5.37 (1H, dd, J = 12.7, 1.2 Hz), 5.29–5.03 (4H, m), 2.43 (1H tdd,
J = 12.6, 11.9, 6.3), 2.32 (1H, ddt, J = 13.6, 6.1, 1.3 Hz), 2.20, (1H, qdd, J = 5.5, 2.5,
1.3 Hz), 2.17–2.05 (4H, m), 1.93 (1H, ddd, J = 13.8, 11.9, 6.3 Hz), 1.85 (1H, ddd,
J = 13.8, 6.3, 1.6 Hz), 1.74 (3H, t, J = 1.2 Hz), 1.56 (1H, br s), 1.58–1.51 (1H, dt,
m); dC (125 MHz, CDCl3) 139.9, 134.1, 131.5, 127.9, 125,4, 118.6, 76.4, 48.6,
44.0, 36.2, 27.1, 25.1 17.0; m/z (ESIMS) found: 175 (MÀOH)+; calcd: 175. (c) 28:
To a solution of epoxide 19 (20 mg, 0.133 mmol) in THF (0.5 mL) at 0 °C was
21. Preparation and characterization of 19, 24, and 28. (a) 19: To cyclononadienone
1
(50 mg, 0.333 mmol) was added DMDO in chloroform (2.5 ml, 0.2 M,
0.5 mmol) at 0 °C. After 5 min, the reaction mixture was concentrated and
purified by flash chromatography (20% EtOAc/hexanes) to give epoxide 19
added allylmagnesium chloride (2.0 M in THF, 90 ll, 0.18 mmol). The mixture
(53 mg, 96%) as a white crystalline solid. IR
m
max(neat)/cmÀ1 2936, 1690, 1625,
was stirred at 0 °C for 1 h. The reaction was then quenched with saturated
NH4Cl, extracted with ether (3 Â 10 mL), the organic fractions were combined,
dried over Na2SO4, filtered, concentrated, and then the residue was purified by
flash chromatography (25% EtOAc/hexanes) to yield bicyclic alcohol 28 (19 mg,
1453, 1177, 980, 747; dH (500 MHz, CDCl3) 6.28 (1H, dd, J = 12.3, 1.3 Hz), 5.81
(1H, ddd, J = 12.2, 9.1, 7.6 Hz), 3.10 (1H, dd, J = 11.8, 2.3 Hz), 2.77 (1H, td,
J = 11.5, 7.7 Hz), 2.77 (1H, td, J = 11.5, 7.7 Hz), 2.72–2.62 (2H, m), 2.40 (1H, ddd,
J = 11.4, 7.5, 1.5 Hz), 2.28–2.20 (1H, m), 2.16–2.05 (2H, m), 1.65 (1H, qd,
J = 11.7, 7.5 Hz), 1.20 (3H, s), 0.97 (1H, dd, J = 12.5, 11.8 Hz); dC (125 MHz,
CDCl3) 206.5, 136.8, 134.9, 61.1, 59.3, 37.4, 35.9, 24.3, 21.8, 17.0; m/z (ESIMS)
found: 167 (M+H)+; calcd: 167. HPLC: Daicel Chiralpak AS-H, n-hexane/i-
PrOH = 90/10, Flow Rate = 1 mL/min, UV = 230 nm, tR = 13.1 min and
tR = 27.9 min (major, 90% ee). (b) 24: To a solution of cyclononadienone 1
76% yield) as a white solid. IR m
max(neat)/cmÀ1 3424, 2932, 1639, 1440, 1078,
912; dH (500 MHz, CDCl3) 5.95–5.80 (2H, m), 5.13–4.99 (3H, m), 3.35 (1H, br s),
2.61–2.51 (1H, m), 2.31–2.14 (3H, m), 1.93–1.66 (5H, m), 1.64–1.55 (1H, m),
1.32–1.25 (1H, m), 1.20 (3H, s); dC (125 MHz, CDCl3) 134.4, 134.0, 130.3, 117.4,
78.2, 75.6, 70.0, 48.6, 38.8, 27.0, 26.5, 23.3, 21.1; m/z (ESIMS) found: 231
(M+Na)+; calcd: 231.