An unexpectedly facile cyclization of polyhydric alcohols

Article abstract of DOI:10.1021/ol9013427

Image Presented Contrary to previous reports in the literature, the reaction of polyhydric alcohols such as sorbitol or mannitol gives good yields of the tetrahydroxyoxepane derivative in the presence of an acid catalyst, in refluxing toluene, with complete retention of stereochemistry.

Full text of DOI:10.1021/ol9013427

ORGANIC
LETTERS
2009
Vol. 11, No. 16
3722-3725
An Unexpectedly Facile Cyclization of
Polyhydric Alcohols
Christopher Pavlik, Amber Onorato, Steve Castro, Martha Morton, Mark Peczuh,
and Michael B. Smith*
Department of Chemistry, UniVersity of Connecticut, 55 N. EagleVille Road,
Storrs, Connecticut 06269-3060
michael.smith@uconn.edu
Received June 15, 2009
ABSTRACT
Contrary to previous reports in the literature, the reaction of polyhydric alcohols such as sorbitol or mannitol gives good yields of the
tetrahydroxyoxepane derivative in the presence of an acid catalyst, in refluxing toluene, with complete retention of stereochemistry.
Of the methods that are available for the preparation of cyclic
ethers, the intramolecular Williamson ether synthesis1 and
cyclodehydration of alcohols or diols are probably the most
prevalent.2 Formation of larger ring ethers is more difficult
by this method, but the formation of five- and six-membered
ring ethers from alcohols, using Pb(OAc)4, HgO, or AgOAc,
has been reported.3a Some examples have been reported for
the preparation of the seven-membered ring ethers known
as oxepanes.3b-d Seven- and eight-membered ring ethers
have been identified as important structural motifs for several
natural products, including cyclic ethers derived from Lau-
rencia4 and marine polycyclic ethers.5 Yields of the larger
ring ethers formed via cyclodehydration of diols tend to be
rather poor. In one report done in supercritical fluids, 1,7-
hexanediol was converted to the oxepane in only 8% yield
along with 3% of 2-ethyltetrahydrofuran,6 and mixtures of
products are common. Other methodology has been used,
such as the Lewis acid catalyzed cyclization of ω-alkenyl
acetals,7 which can generate good yields of substituted
oxepanes. Good yields of substituted oxepanes were obtained
by the N-iodocollidine salt mediated cyclization of ω-alkenyl
alcohols and dienyl alcohols.8 Other methods include an
intramolecular Horner-Wadsworth-Emmons cyclization9
and a ring-closing metathesis strategy from dienyl ethers.10
The synthesis of polycyclic ether natural products has been
accomplished using acid-catalyzed cyclization of intermedi-
ates that have multiple epoxide units. This approach generates
seven- and eight-membered ring ethers that are highly
substituted and with stereocontrol of the ring closure.11
(1) Smith, M. B.; March, J. March’s AdVanced Organic Chemistry, 6th
ed.; Wiley Interscience, New York, 2007, section 10-8, p 529. See: Kim,
K. M.; Jeon, D. J.; Ryu, E. K. Synthesis 1998, 835 for cyclization to an
alkene in the presence of a catalytic amount of iodine.
(4) (a) Moore, R. E. In Marine Natural Products; Scheuer, P. J., Ed.;
Academic: New York, 1978; Vol. 1, pp 43-121. (b) Erickson, K. In Marine
Natural Products; Scheuer, P. J., Ed.; Academic: New York, 1983; Vol. 5,
pp 131-257; (c) Faulkner, D. J. Nat. Prod. Rep. 1996, 13, 75–126.
(5) See Kadota, I.; Yamamoto, Y. Acc. Chem. Res. 2005, 38, 423–432.
(6) Gray, W. K.; Smail, F. R.; Hitzler, M. G.; Ross, S. K.; Poliakoff,
M. J. Am. Chem. Soc. 1999, 121, 10711–10718.
(2) (a) For a list of reagents used to convert alcohols and phenols to
ethers, see: Larock, R. C. ComprehensiVe Organic Transformations, 2nd
ed.; Wiley-VCH, NY, 1999; pp 890-893. (b) Kotkar, D.; Ghosh, P. K.
J. Chem. Soc., Chem. Commun. 1986, 650; (c) Tagliavini, G.; Marton, D.;
Furlani, D. Tetrahedron 1989, 45, 1187–1196. (d) Traynelis, V. J.;
Hergenrother, W. L.; Hanson, H. T.; Valicenti, J. A. J. Org. Chem. 1964,
29, 123–129. (e) Costa, A.; Riego, J. M. Synth. Commun. 1987, 17, 1373–
1376.
(7) (a) Castaneda, A.; Kucera, D. J.; Overman, L. E. J. Org. Chem. 1989,
54, 5695–5707. (b) Blumenkopf, T. A.; Look, G. C.; Overman, L. E. J. Am.
Chem. Soc. 1990, 112, 4399–4403. (c) Berger, D.; Overman, L. W.;
Renhowe, P. A. J. Am. Chem. Soc. 1997, 119, 2446–2452.
(8) Brunel, Y.; Rousseau, G. J. Org. Chem. 1996, 61, 5793–5800.
(9) See Taillier, C.; Bellostra, V.; Cossy, J. Org. Lett. 2004, 6, 2149–
2151.
(3) (a) Smith, M. B.; March, J. March’s AdVanced Organic Chemistry,
6th ed.; Wiley: New York, 2007; pp 965-966. (b) Vowinkel, E. Chem.
Ber. 1962, 95, 2997–3002. (c) Vowinkel, E. Chem. Ber. 1963, 96, 1702–
1711. (d) Vowinkel, E. Chem. Ber. 1966, 99, 42–47.
(10) See: Crimmins, M. T.; DeBaillie, A. C. Org. Lett. 2004, 5, 3009–
3011, and ref 13 therein.
10.1021/ol9013427 CCC: $40.75
Published on Web 07/29/2009
 2009 American Chemical Society

Scheme 1
.
Reported Acid-Catalyzed Cyclization of Sorbitol
Scheme 2. Literature Synthesis of Tetrahydroxyoxepanes
Recent work in our laboratory includes an investigation
into the role of alcohol additives with conducting polymers,
and their ability to significantly enhance conductivity. As
part of that investigation, we examined the reaction of
sorbitol and other polyhydric alcohols with acid. Sorbitol is
an important additive to conducting polymers, and the
literature is clear in terms of the acid-catalyzed cyclization
reactions. In a synthesis of C2-symmetric diphosphinite
ligands derived from carbohydrates by Diaz, Castillo´n, and
co-workers, heating sorbitol (1, also known as D-glucitol)
with tosic acid in xylene led to a mixture of products.
Subsequent treatment with acetone and acid, separation and
purification, and then passage through an ion exchange
column led to 2,5-anhydro-L-iditol (2) in a low 5% yield.12
An identical experiment was reported in a synthesis of linear,
branched, and cyclic oligoglycerol derivatives, and when
sorbitol was heated with methanesulfonic acid with isolation
procedures similar to those reported by Diaz and Castillo´n,
a 3% yield of 2 was obtained and it was suggested that 1,5-
anhydrohexitrol and 1,4-anhydrohexitol (sorbitan) were
produced as minor products.13 The presence of these latter
products was inferred from previous work without further
proof. In these and other studies, the only reported products
were five- and six-membered ring ethers, with no mention
of oxepane derivatives.
Although direct cyclization of sorbitol to an oxepane has
not been reported in the literature, there is a multistep
synthesis of tetrahydroxyoxepanes from either sorbitol or
mannitol. Starting with sorbitol (1), the fully protected
triisopropylidene 3 was prepared and then reacted with HCl
in MeOH to yield 4. This 3,4-isopropylidene product was
formed in only 27% yield as a mixture of the different
isomers, presumably separated by recrystallization. The
primary alcohol moieties were converted to the tosylates,
resulting in 5. The yield was about 50% and the product
decomposed quickly, requiring that it be used immediately.
Compound 5 reacted with sodium methoxide in CHCl3 to
yield the bis(epoxide) product 6 in a yield of 81%. Finally
reaction with NaOH at approximately 55 °C for 10 h was
followed by saturation with carbon dioxide to a pH of 8 and
purification by slow recrystallization for several days to give
the targeted tetrahydroxyoxepane, 7 (1,6-anhydrosorbitol),
in very poor yield.14 Stereochemistry is shown only for 1
and 7, but the sequence produced enantiopure 7.14
Given the literature precedent, we were surprised to find
that the reaction of sorbitol with 1% triflic acid in toluene at
reflux gave a single isolated product in high yield. Initially,
we anticipated that five- or six-membered ring anhydro
sugars were formed.15 Initial attempts to derivatize the
unknown product by formation of the corresponding ac-
etonide, by reaction with acetone and acid or with dimethox-
ypropane and acid, failed to give an isolable product. We
therefore prepared a derivative using an excess of 3-fluo-
robenzoyl chloride to properly identify the product.16 Proton
NMR showed that the derivative was too symmetrical relative
to products reported in the literature. The derivative did not
have a hydroxymethyl unit, clearly indicating that our initial
expectation was incorrect.15b Careful NMR analysis of the
anhydro sugar product showed it to be a tetra(3-fluoroben-
zoate) derivative of an oxepane, 1,6-anhydrosorbitol (7).17
Once the product was identified, determination of a yield
was possible, and the conversion of 1 to 7 proceeded in 86%
yield. This assignment resolved structural issues for the acid-
catalyzed reaction. We next examined the reaction of
D-mannitol (8) with triflic acid in refluxing toluene and
obtained a product in 89% yield that was clearly different
than the reaction with sorbitol and proved to be tetrahy-
droxyoxepane 9 (1,6-anhydromannitol).
In our acid-catalyzed reactions of 1 and 8, the stereo-
chemistry of the hydroxyl units in 7 and 9 were apparently
retained. To confirm this, we prepared a synthetic standard
(14) (a) Vargha, L.; Kasztreiner, E. Chem. Ber. 1960, 93, 1608–1616.
(b) Vargha, L.; Kasztreiner, E. Chem. Ber. 1959, 92, 2506–2515.
(15) (a) ORGN 101, Chemical Reactivity of Poly(ethylenedioxy)-
thiophene (PEDOT), Smith, M. B.; Sotzing, G. A.; Onorato, A.; Kumar,
A.; Delude, C. 232nd ACS National Meeting, San Francisco, CA, September
10, 2006. 234th ACS National Meeting, Boston, MA, August 20, 2007. (b)
ORGN 359, The Chemical Reactivity Of Sorbitol With PEDOT; Onorato,
A.; Navarathne, D.; Smith, M. B.; Sotzing, G. A.
(11) See for example, McDonald, F. E.; Wang, X.; Do, N.; Hardcastle,
K. I. Org. Lett. 2000, 2, 2917–2919.
(16) Yang, J.; Morton, M. D.; Hill, D. W.; Grant, D. F. Chem. Phys.
Lipids 2006, 140, 75–87.
(12) Aghmiz, M.; Aghmiz, A.; D´ıaz, Y.; Masdeu-Bulto´, A.; Claver, C.;
Castillo´n, S. J. Org. Chem. 2004, 69, 7502–7510. Also see: Giacometti, J.;
Milin, C.; Wolf, N.; Giacometti, F. J. Agric. Food Chem. 1996, 44, 3950–
3954.
(17) (a) Soha´r, P.; Vargha, L.; Kasztreiner, E. Tetrahedron 1964, 20,
647–653. (b) Vargha, L.; Kasztreiner, E. Chem. Ber. 1960, 93, 1608–1616.
(18) (a) Cohen, N.; Banner, B. L.; Laurenzano, A. J.; Carozza, L. Org.
Synth. Coll. Vol. 7 1990, 297–302. (b) Inoue, T.; Kitagawa, O.; Oda, Y.;
Taguchi, T. J. Org. Chem. 1996, 61, 8256–8263.
(13) Cassel, S.; Debaig, C.; Benvegnu, T.; Chaimbault, P.; Lafosse, M.;
Plusquellec, D.; Rollin, P. Eur. J. Org. Chem. 2001, 87, 5–896.
Org. Lett., Vol. 11, No. 16, 2009
3723

Scheme 3
.
Acid-Catalyzed Cyclodehydration of Sorbitol and
Mannitol
Scheme 4
.
Multistep Synthesis of the 1,6-Anhydromannitol
Synthetic Standard
of the symmetrical mannitol derivative 9 by a six-step
sequence starting with 2,3-O-isopropylidene-D-erythrono-
lactone, 10,18 based on an approach reported by Sturino and
co-workers.19 Reduction with diisobutyl-aluminum hydride
gave lactol 11 in 88% yield, and a Wittig reaction gave
alkene-alcohol 12 in 50% yield. Subsequent treatment with
base and then allyl bromide gave allylic ether 13 in 62%
yield. Ring-closing metathesis with the Grubbs’ II ruthenium
catalyst led to 14 in 66% yield. Dihydroxylation with osmium
tetroxide gave 1, which was deprotected without further
purification, to give tetrahydroxyoxepane 9 in 96% yield.
Synthetic 9 was identical in all respects with that obtained
by the reaction of 8 with p-TsOH or triflic acid in a single
step. With this synthesis, we have demonstrated that D-
mannitol (8) reacts with an acid catalyst to give tetrahy-
droxyoxepane, 9, with complete retention of the stereochem-
ical centers. Given the stereoselective nature of the mannitol
reaction and the NMR identification of the sorbitol product,
we are confident in the structural integrity of 7 as the sorbitol
product.
Confirmation that polyols are converted to the cyclic ether
under mild conditions and in good yield prompted us to examine
the reaction of other commercially available polyhydric alcohols
with p-TsOH as well as triflic acid. We observed that these
polyols undergo cyclodehydration with high stereoselectivity,
both with a catalytic amount of acid or with 1 molar equiv of
the acid. With 1% p-TsOH, 1% triflic acid or with 1 full equiv
of p-TsOH, sorbitol (1) is converted to 7 and D-mannitol (8) is
converted to 9. These results are shown in Table 1. In addition,
An important question remains. Why have previous experi-
ments failed to observe this transformation? The answer may
never be known, but previous experiments used more vigorous
conditions in some cases and derivatized the products prior to
isolation. Our control experiments have shown that oxepane 7
decomposes when heated with sulfuric acid, although it is stable
to heating with p-TsOH. Perhaps more relevant is our observa-
tion that 7 did not react with acetone and an acid catalyst or
with dimethoxypropane and an acid catalyst under standard
conditions. It is possible that previous work selectively deriva-
tized the reported products, and it is important to note that the
isolated yields reported in the literature were rather low.
Scheme 5
.
Acid-Catalyzed Conversion of Erythritol and Threitol
to Dihydroxytetrahydrofurans
Despite previous reports, polyols are cleanly converted to
the cyclic ethers under our conditions.20 The relevance of polyol
cyclodehydration to our ongoing work with conducting polymer
additives is under investigation. More pertinent to this discussion
Table 1. Cyclodehydration of Polyhydric Alcohols
cyclic ether
polyol
1% TsOH
1 equiv TsOH
1% TfOH
1
8
16
18
7 (79%)
9 (80%)
17 (70%)
19 (71%)
7 (84%)
9 (76%)
7 (86%)
9 (89%)
17 (75%)
19 (73%)
(19) Wong, J. C. Y.; Lacombe, P.; Sturino, C. F. Tetrahedron Lett. 1999,
40, 8751–8754.
(20) A catalytic amount of toluenesulfonic acid (1% mmol in 5 mL of
toluene) was added to a solution of D-sorbitol (0.18 g, 1 mmol) in 60 mL
of toluene via syringe, under a positive pressure of nitrogen. The solution
was heated to reflux overnight and cooled, and the solution was treated
with 10 mg of NaHCO3. The particulate matter was subsequently removed
by filtration and washed with diethyl ether (10 mL). The toluene filtrate
was concentrated in vacuo without heat to give 1,6-anhydrosorbitol 7 as a
pale red waxy solid (129 mg, 0.786 mmol, 79%). Mp, 29-30°C. 1H NMR
(CD3OD): δ 4.54 (t, J ) 4.7 Hz, 1H), 4.38 (d, J ) 3.7 Hz, 1H), 4.28 (dt,
J ) 5.4, 6.9 Hz, 1H), 4.2 (s, 1H), 3.9-3.8 (m, 3H), 3.45 (dt, J ) 7.5, 8.3
Hz, 1H). 13C NMR (CD3OD): δ 88.3, 81.9, 76.2, 75.2, 72.2, 71.4. MS (EI)
m/z 164, 147, 129, 111 102, 87, 69.
we have shown that meso-erythritol (16) reacts with either
p-TsOH or triflic acid to form 2,3-dihydroxytetrahydrofuran 17.
In addition, the acid-catalyzed reaction of threitol (18) leads to
19. In future work, these reactions may be extended to more
highly functionalized polyols and even sugars.
3724
Org. Lett., Vol. 11, No. 16, 2009

is the observation that polyols are converted to cyclic ethers in
good yield and with retention of stereochemistry. These
transformations may have value in the synthesis of natural
products that bear these structural motifs.
Supporting Information Available: Full experimental
details for all cyclodehydration experiments, details of the
synthesis of 9 from 10, and spectral data for the tetra(3-
fluorobenzoate derivative of 7. This material is available free
of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We thank Dr. Srikanth Rapole, who
obtained the exact mass determination of the sorbitol tetra(3-
fluorobenzoate) derivative.
OL9013427
Org. Lett., Vol. 11, No. 16, 2009
3725