The prins reaction using ketones: Rationalization and application toward the synthesis of the portentol skeleton

Article abstract of DOI:10.1021/ol202829u

We report a TMSI-promoted Prins cyclization reaction with ketones as carbonyl partners to prepare polysubstituted chiral spirotetrahydropyrans. In the presence of racemic 2-methylcyclohexanone a dynamic kinetic resolution occurred affording one stereoisom

Full text of DOI:10.1021/ol202829u

ORGANIC
LETTERS
2012
Vol. 14, No. 1
58–61
The Prins Reaction Using Ketones:
Rationalization and Application toward the
Synthesis of the Portentol Skeleton
Maiwenn Jacolot,† Mickael Jean,† Nicolas Levoin,‡ and Pierre van de Weghe*,†
ꢀ
Universite de Rennes 1, UMR 6226, Sciences Chimiques de Rennes, Equipe PNSCM,
UFR des Sciences Biologiques et Pharmaceutiques, 2 avenue du Prof Leon Bernard,
ꢀ
F-35043 Rennes Cedex, France, and Bioprojet-Biotech, 4 rue du Chesnay Beauregard,
ꢀ
BP96205, F-35762 Saint-Gregoire, France
pierre.van-de-weghe@univ-rennes1.fr
Received October 20, 2011
ABSTRACT
We report a TMSI-promoted Prins cyclization reaction with ketones as carbonyl partners to prepare polysubstituted chiral spirotetrahydropyrans.
In the presence of racemic 2-methylcyclohexanone a dynamic kinetic resolution occurred affording one stereoisomer. The observed
enantiospecificity has been rationalized by DFT calculation.
Lichens are among the most fascinating organisms,
characterized by a symbiotic association between a fungus
and a green algae or a cyanobacterium producing many
biologically active metabolites. Recent studies have re-
ported that many lichens contain a stable consortium of
other microorganisms on the surface offering a wide mol-
ecular diversity.1 Among the numerous lichenic deriva-
tives, portentol 1 (Figure 1), a polypropionate isolated
from Roccella portentosa, possesses a fascinating structure
for which no synthesis (or synthetic approach) has been re-
ported to date.2 Despite the numerous major achievements
in total synthesis that have been reported over the past
half-century,3 the portentol structure presents a unique
challenge.
Retrosynthetic simplification of 1 can, in principle, be
rather straightforward and, in our hands, yielded the
spiroether 2 as a model substrate. Prins cyclization4 invol-
ving analkene appeared heretobethe method of choicefor
preparing a polyfunctionalized tetrahydropyran of this
sort, but despite significant advances in the Prins reaction
field,4 only a few examples that combine homoallylic alco-
hols and ketones are reported in the literature.5 Based on
this statement, we were intrigued by the possibility of con-
structingthe portentol spirotetrahydropyran skeleton via a
Prins cyclization. Thus polysubstituted chiral homoallylic
alcohol 3 and cyclohexanone 4, in the presence of a catalyst
or promoter, could potentially undergo a Prins cyclization
reaction to give 2 (Figure 1).
†
ꢀ
Universite de Rennes 1.
‡ Bioprojet-Biotech.
(1) (a) Boustie, J.; Grube, M. Plant Genetic Resources 2003, 3, 273–
287. (b) Podterob, A. P. Pharm. Chem. J. 2008, 42, 582–588. (c) Grube,
M.; Berg, G. Fung. Biol. Rev. 2009, 23, 72–85.
(2) (a) Aberhart, D. J.; Overton, K. H. Chem. Commun. 1969, 162–
163. (b) Aberhart, D. J.; Overton, K. H. Chem. Commun. 1970, 664–665.
(c) Aberhart, D. J.; Overton, K. H.; Huneck, C. J. Chem. Soc. (C) 1970,
1612–1623. (d) Ferguson, G.; Mackray, I. R. Chem. Commun. 1970, 665–
666. (e) Quilhot, W.; Garbarino, J. A.; Gambaro, V. J. Nat. Prod. 1983,
46, 594–595. (f) Pettit, G. R.; Zhang, Q.; Pinilla, V.; Herald, D. L.;
Doubek, D. L.; Duke, J. A. J. Nat. Prod. 2004, 67, 983–985.
(3) (a) Nicolaou, K. C.; Sorensen, E. Classics in Total Synthesis;
Wiley-VCH: 1996. (b) Nicolaou, K. C.; Snyder, S. A. Classics in Total
Synthesis II; Wiley-VCH: 2003. (c) Nicolaou, K. C.; Chen, J. S. Classics in
Total Synthesis III; Wiley-VCH: 2011.
(4) For a recent review, see: Olier, C.; Kaafarani, M.; Gastaldi, S.;
Bertrand, M. P. Tetrahedron 2010, 66, 413–445.
(5) For significant examples, see: (a) Nishizawa, M.; Shigarahi, T.;
Takao, H.; Imagawa, H.; Sugihara, T. Tetrahedron Lett. 1999, 40, 1153–
1156. (b) Sabitha, G.; Reddy, K. B.; Bhikshapathi, M.; Yadav, J. S.
Tetrahedron Lett. 2006, 47, 2807–2810. (c) Yadav, J. S.; Reddy, B. V. S.;
Krishna, V. H.; Swamy, T.; Kumar, G. G. K. S. N. Can. J. Chem. 2007,
85, 412–417. (d) Castaldi, M. P.; Troast, D. M.; Porco, J. A., Jr. Org.
Lett. 2009, 11, 3362–3365.
r
10.1021/ol202829u
Published on Web 11/30/2011
2011 American Chemical Society

Scheme 1. Plausible Mechanism, ChairLike Transition State and Stereochemical Outcome
intermediate, the hydroxy product 10 was obtained along
with the fluoride compound 11.9 It is also noteworthy that
the optically active tetrahydropyrans 10 and 11 were for-
med with excellent diastereoselectivities(>98:2) aftercare-
ful examination of the crude reaction mixture by NMR.
Encouraged by this result we next investigated the re-
action with cyclohexanone 4 (Scheme 2, eq 2). Despite
numerous attempts to optimize the reaction, the conver-
sion remained low and the expected product 5 was isolated
as the sole diastereoisomer in moderate yield and only
traces of the fluoride derivative 12 were observed. Sabitha
et al.5b described an efficient synthesis of spiroethers from
various ketones and homoallylic alcohols in acetonitrile
using TMSI (generated in situ) as the promoter and nucle-
ophile source. With these reaction conditions in mind, we
decided to apply them to our substrates (Scheme 2, eq 3).
We were delighted to observe a complete conversion after a
few minutes of stirring. An exclusive equatorial attack of
the iodide on the carbocation species occurred; the iodos-
pirotetrahydropyran 13 was obtained as the major diaster-
eoisomer (>98:2) as established by NMR analysis of the
crude reaction mixture. It was isolated in good yield (78%).
Figure 1. Structure of portentol 1 and retrosynthetic approach.
The stereochemical outcome of the Prins cyclization
reaction can be anticipated relying on the accepted gener-
al mechanism and by invoking the chairlike transition
state (Scheme 1). The oxocarbenium ion intermediate TS1
would undergo a 6-endo cyclization to give selectively the
carbocationTS2orTS3 (respectivelyformedfrom (E)-3or
(Z)-3) that could then be trapped by various nucleophiles.
According to the Alder’s model,6 a preferential equator-
ial trapping of the nucleophile will usually occur. Therefore,
it would be necessary to carry out the Prins cyclization re-
action with the homoallylic alcohol (E)-3 leading probably
to 5 as the major product which could lead to the desired
spiroether 2 after an inversion of the absolute configuration
of the alcohol function. Detailed studies on the Prins cycli-
zation reaction involving polysubstituted homoallylic alco-
hols and ketones are reported herein.
Scheme 2. Initial Investigations
In our initial investigation, the search for optimal reac-
tion conditions was conducted using the enantiopure ho-
moallylic silylether (E)-87 and benzaldehyde 9 by varying
the promoter, solvent, and temperature.8 Boron trifluoride
etherate was proven to be an effective promoter for the
reaction process (Scheme 2, eq 1). Because water was
produced during the formation of the oxocarbenium ion
(6) Alder, R. W.; Harvey, J. N.; Oakley, M. T. J. Am. Chem. Soc.
2002, 124, 4960–4961.
(7) The isolation of the homoallylic alcohol 3 gave nonreproducible
yields. Moreover, whatever the reaction conditions used for the Prins
cyclization reaction, the homoallylic alcohol 3 decomposes quickly in
the reaction mixture leading to the expected product in very low yield.
Fortunately, the reaction works well using its silylether form.
(8) The most significant results are summarized in Supporting
Information.
(9) (a) Kataoka, K.; Ode, Y.; Matsumoto, M.; Nokami, J. Tetrahe-
dron 2006, 62, 2471–2483. (b) Launey, G. G.; Slawin, A. M. Z.;
O’Hagan, D. Beilstein J. Org. Chem. 2010, 6, 41.
With appropiate Prins cyclization reaction conditions in
hand, we focused on the scope of this transformation
(Table 1). As noted with the cyclohexanone 4, a single
Org. Lett., Vol. 14, No. 1, 2012
59

Table 1. Scope of the Prins Cyclization Reaction with TMSI
diastereoisomer was obtained for the reactions involving
(E)-8 and benzaldehyde 9 or cyclopentanone 15. The struc-
ture of these was confirmed by NMR analysis (Table 1,
entries 1 and 2). The formation of product occurs by a
chairlike transition state with equatorial trapping of the
halogen. A mixture of two diastereoisomers 18a and 18b in
a ratio of 1:1 was isolated after reaction with the isobutyl
methyl ketone (Table 1, entry 3). Use of the bicyclic norcam-
phor 19 led to a mixture of two diastereoisomers 20a and 20b
in a ratio of 6:4 (Table 1, entry 4). A more hindered ketone
like camphor 21 did not lead to any cyclized product, with no
reaction occurring under the TMSI-promoted Prins reaction
conditions. We also examined the effect of the olefin nature
on the Prins cyclization reaction using cyclohexanone 4 as
the ketone partner. Thus use of (Z)-8 afforded a mixture of
two diastereoisomers 22 and 13 (Table 1, entry 6). The
formation of the latter is probably due to a rapid cis/trans
isomerization of the double bond under the reaction condi-
tions. A monosubstituted alkene works well as demonstrated
by the reaction between 23 and 4 which provided 24 in good
yield (Table 1, entry 7). By way of contrast, no Prins cycli-
zation product was formed with a bulky substituent like a
phenyl group on the terminal olefin. We only observed loss
of the silyl protecting group from the starting material
(Table 1, entry 8). While we had never seen the formation
of a product resulting from the trapping of acetonitrile be-
fore (PrinsꢀRitter reaction),10 the addition of rac-26 to 4 in
acetonitrile produced a mixture of the iodo derivative rac-27
and the acetamide rac-28 in a 1:1 ratio (Table 1, entry 9).
Changing acetonitrile for methylene chloride as the solvent
allowed us to isolate rac-27 as a single diastereoisomer in
high yield (90%).
We then turned our attention to the Prins cyclization
reaction involving 2-substituted ketones (Scheme 3). Ad-
dition of 1 equiv of the racemic 2-methylcyclohexanone
rac-29 to 1 equiv of the enantiopure alkene (E)-8 in the pre-
sence of TMSI (generated in situ) in acetonitrile or methy-
lene chloride furnished 30 as essentially a single diastereoi-
somer (checked as usual by careful examination of the
crude reaction mixture by NMR) in good yield (eq 1).11 To
our knowledge, it is the first example of a Prins cyclization
reaction with dynamic kinetic resolution (DKR) of one
partner of the reaction.12
This remarkable phenomenon has been confirmed.
Starting from both the (S)-29 or (R)-29, the same product
30 was formed in good yield. It is likely that the slow-
reacting (S) enantiomer undergoes a very fast isomeriza-
tion through enolization into the fast-reacting enantiomer
(R) to lead finally to 30 solely. Moreover, the enantiomer
ent-30 was obtained in similar yield by the reaction be-
tween ent-(E)-8 and rac-29.13 The same process occurred
(11) The structure of 30 was established with the help of 2D NMR
and confirmed by X-ray analysis after its conversion into a 4-amide. See
Supporting Information.
(12) Noyori et al. reported a similar observation in the course of the
reduction of 2-isopropylcyclohexanone in the presence of an enantio-
pure ruthenium catalyst; see: Ohkuma, T.; Ooka, H.; Yamakawa, M.;
Ikariya, T.; Noyori, R. J. Org. Chem. 1996, 61, 4872–4873.
(13) Absolute configuration of (E)-8: 2R,3S; ent-(E)-8: 2S,3R. Op-
tical rotation of 30: [R] = þ18.0 (c 1.7, CHCl3); ent-30: [R] = ꢀ16.0
(c 1.3, CHCl3).
(10) For representative examples of PrinsꢀRitter reactions, see: (a)
Perron, F.; Albizati, K. F. J. Org. Chem. 1987, 52, 4128–4130. (b) Yadav,
J. S.; Reddy, S. B. V.; Aravind, S.; Kumar, G. G. K. S. N.; Reddy, G. M.
Tetrahedron 2008, 64, 3025–3031.
60
Org. Lett., Vol. 14, No. 1, 2012

Scheme 3. Case of 2-Methyl Cyclic Ketones
Scheme 4. Case of Menthone
with the olefin (E)-31 to give 32 as a single isomer. In the
absence of a substituent on the terminal position of the
olefin, the DKR does not seem to take place and a mixture
of two diastereoisomers was obtained in a moderate yield
(eq 3). Finally, the same process occurred with a five-mem-
bered ring with a lower yield (as observed previously).
While the Prins cyclization between the L-menthone 36
and (E)-8 led to the single diastereoisomer 37 after 20 min
at rt in good yield, the reaction was less facile and much
slower with the D-menthone 40 (Scheme 4). After the
reaction stirred overnight at 0 °C, only the spiroether 41
was formed and isolated in low yield.14 This product is the
result of a slow epimerization of the chiral center at the
R-position of the ketone function. These results support our
previous observations; a matched pair of reagents affords
the product in good yield whereas a mismatched pair in-
duces the epimerization of the R-chiral center of the ketone
before reaction (Scheme 3, eq 5).
Figure 2. Chairlike transition state calculated by DFT.
led to the differentiation observed for both enantiomers of
rac-29 during the reaction (Figure 2).
In conclusion we have described a TMSI-promoted
Prins cyclization reaction between substituted silyl ether
of homoallylic alcohol and ketones to afford polysubsti-
tuted spirotetrahydropyrans. In the case of 2-substituted
cyclic ketones, a DKR process has been observed. Addi-
tionally, DFT calculations of the Prins cyclization reaction
have been undertaken. Extension of this reaction to poly-
functionalized ketones and application to the synthesis of
portentol are inprogress and willbereportedindue course.
Acknowledgment. We are grateful for the support pro-
ꢀ
vided by Rennes Metropole, the Region Bretagne, and the
Universite de Rennes 1. We gratefully acknowledge the
ꢀ
ꢀ
DFT calculations have been used to study the Prins
cyclization reaction between (E)-8 and rac-29.15 The cal-
culated reaction path shows that the chairlike transition
state TS1-(R) and TS2-(R) are more stable than TS1-(S)
and TS2-(S) by 3.5 kcal/mol, confirming the preferential
formation of 30. This important energy difference surely
CRMPO for mass measurements and Thierry Roisnel
(UMR 6226, Rennes) for X-ray analytical data. We also
ꢁ
thank the Ministere de la Recherche for the fellowship to
Maiwenn Jacolot.
Supporting Information Available. Full details of the
preparation of starting materials and experimental proce-
dures including 1H, 13C NMR spectra. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(14) Only a part of 40 has been recovered; no trace of the olefin (E)-8
has been found in the crude reaction mixture.
(15) DFT calculations are detailed in the Supporting Information.
Org. Lett., Vol. 14, No. 1, 2012
61