Pyridine-N-oxide as a mild reoxidant which transforms osmium-catalyzed oxidative cyclization

Article abstract of DOI:10.1002/anie.200705425

(Chemical Equation Presented) The PNOman can: The use of pyridine-N-oxide (PNO) transforms the catalytic oxidative cyclization to include the formation of pyrrolidines from N-Z-protected amino alcohols and amino acids (see scheme; Z = PhCH₂OCO). This new method expands the scope of the oxidative cyclization as illustrated with the synthesis of a range of di- and trisubstituted pyrrolidines with complete control of stereochemistry.

Full text of DOI:10.1002/anie.200705425

Communications
DOI: 10.1002/anie.200705425
Oxidative cyclization
Pyridine-N-Oxide as a Mild Reoxidant Which Transforms Osmium-
Catalyzed Oxidative Cyclization**
Timothy J. Donohoe,* Katherine M. P. Wheelhouse (nØe Gosby), Peter J. Lindsay-Scott,
Paul A. Glossop, Ian A. Nash, and Jeremy S. Parker
Recent reports from our group showed that the catalytic
oxidative cyclization of 1,5-dienes[1] directly to cis tetrahy-
drofurans (THFs) can be interrupted and that vicinal diols
derived from 1,5-dienes also cyclize to give THF rings in
excellent yields under the action of osmium tetroxide and an
acid.[2] In addition, we also disclosed that the corresponding
N-tosyl amino alcohols (tosyl = p-toluenesulfonyl) bearing a
pendant alkene cyclize to give the relevant pyrrolidine ring
systems 2, the first time that such oxidative methodology had
been used to form nitrogen heterocycles (Scheme 1).[3] Both
The advantage of using diols or amino alcohols as
precursors to the cyclization is that it allows the direct
formation of enantiopure heterocycles, including pyrrolidines,
by a pathway that is simply not viable from the cyclization of
1,5-dienes.[4]
However, this catalytic reaction does have some limita-
tions. Firstly, the cyclization of protected amines to form
pyrrolidines is restricted to amino alcohols with a sulfonamide
group on the nitrogen center; carbamates do not cyclize in
acceptable yield. Secondly, our studies revealed that OsVI is
the more active catalyst for the cyclization process, and the
presence of OsVIII is to be avoided because it causes unwanted
dihydroxylation of alkenes in the substrate.[3] A consequence
of this observation is that it is necessary to add an excess of a
sacrificial alkene (usually cinnamic acid) to the system so that
it is dihydroxylated in preference to the substrate, and thereby
increases the concentration of OsVI in the reaction. Such an
internal conflict within the reaction mixture is clearly less
than ideal.
Table 1: Cyclization reaction of N-tosyl-protected amino alcohols.
Scheme 1. Oxidative cyclization of hydroxysulfonamide substrates.
Ts =p-toluenesulfonyl; Z=PhCH2OCO.
Entry Conditions[a]
Yield of
4
cyclization reactions are notable for their stereospecificity
(syn addition across the tethered alkene) and stereoselectivity
(cis-2,5-heterocycles are formed exclusively). As expected,
the absolute stereochemistry of the starting materials is
transferred to the products, thus providing a short route to
valuable heterocycles with complete control of stereochem-
istry.
1
2
5% K2OsO2(OH)4, CSA, TMO, CH2Cl2 (48 h)
5% K2OsO2(OH)4, CSA, cinnamic acid, TMO, CH2Cl2 79%
69%
(16 h)
3
4
5% K2OsO2(OH)4, CSA, PNO, CH2Cl2 (4 h)
5% K2OsO2(OH)4, CSA, PNO, citric acid,[5] CH2Cl2
(2 h)
85%
98%
[a] Reactions run at room temperature. Tosyl=p-toluenesulfonyl; CSA=
(Æ)-camphorsulfonic acid; TMO=trimethylamine-N-oxide.
[*] T. J. Donohoe, K. M. P. Wheelhouse (nØe Gosby), P. J. Lindsay-Scott
Department of Chemistry, University of Oxford
Chemistry Research Laboratory
Mansfield Road, Oxford, OX1 3TA (UK)
Fax: (+44)1865 275708
E-mail: timothy.donohoe@chem.ox.ac.uk
P. A. Glossop
Pfizer Global R&D, Pfizer Limited
Ramsgate Road, Sandwich, Kent, CT13 9NJ (UK)
I. A. Nash, J. S. Parker
AstraZeneca
15 Stanhope Gate, W1K 1LN (UK)
[**] We thank the EPSRC, Pfizer, and AstraZeneca for supporting this
project. We thank Merck for unrestricted funding.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Scheme 2. Stereoselective oxidative cyclization.
2872
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2872 –2875

Angewandte
Chemie
the relative ease of preparing them in enantiopure form
(Scheme 4). The cyclization reaction works well and
allows the preparation of a series of useful, enantiopure
cyclic amino acids; note also the successful cyclization
of amino amide 19. The change in the solvent and the
acid for this reaction enabled us to maximize solubility,
heating, and purification procedures.
The cyclization of aspartic acid derived amino acid
27[8] reveals that substituents on the backbone are
tolerated under the cyclization conditions. However,
we were intrigued by the notion of cyclizing a
compound containing two allyl groups in an attempt
to induce a diastereoselective cyclization, thus gaining
a free stereogenic center in the pyrrolidine product
(Scheme 5). Substrates 29 and 31 were prepared easily
and then cyclized by using the PNO conditions. In each
case a major product was formed with excellent
diastereoselectivity and X-ray crystallographic analysis
confirmed the trans relationship between the remain-
ing allyl group and the C2 methylene group.[9] From
transition-state models this outcome can be rational-
Scheme 3. Oxidative cyclization of carbamate-protected
amino alcohols. Ns=nitrobenzenesulfonyl;
Z=PhCH2OCO.
To overcome these problems we sought a
reoxidant that would oxidize Os to the VI oxidation
state (where it is capable of promoting an oxidative
cyclization), but not to the VIII oxidation state
(where it can also dihydroxylate the substrate).
After screening a range of reoxidants for osmium,
we can now report that pyridine-N-oxide (PNO) is
an effective reoxidant for the oxidative cyclization
reaction. The consequences of using this reoxidant
are that it engenders a much more powerful
oxidizing system, which does not require the
addition of a sacrificial alkene and does not cause
dihydroxylation of the substrate. A good example
of this effect can be seen in the formation of
pyrrolidine 4 from 3 (Table 1).
Hydroxysulfonamides 5 and 7, which each took
over 24 hours to react under the old conditions
(yields 78–80%), cyclized most efficiently and with
low catalyst loadings (Scheme 2).
In fact, the improvement that this new system
lends to the oxidative cyclization is so marked that
N-carbamate-protected amino alcohols can now be
cyclized in excellent yields and with low loadings of
the osmium catalyst (Scheme 3).[6] None of these
reactions, including cyclization of N-nosyl-pro-
tected (nosyl = nitrobenzenesulfonyl) substrate 11,
gave viable (or often any) yields of pyrrolidine rings
under the previous best set of conditions.
Armed with the new cyclization conditions we
were able to re-examine cyclization substrates that
were not successfully transformed in our early
work. We concentrated on amino acid substrates
because of the useful products that they form[7] and Scheme 4. Oxidative cyclization of amino acids.
Angew. Chem. Int. Ed. 2008, 47, 2872 –2875
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2873

Communications
Scheme 6. Mechanistic probes for osmium oxidation. Bn=benzyl.
conditions, but again no oxidation of OsVI was observed). All
of this evidence points to a lack of OsVIII after the addition of
PNO.[14]
Scheme 5. Cyclization onto one of two diastereotopic allyl groups.
P=protecting group.
Given the likely dominance of OsVI in the above PNO-
driven reactions, we also present data to confirm that this
metal species is capable of accomplishing oxidative cycliza-
tion (Scheme 7). Hydroxytosylate 3 was subjected to reaction
with one equivalent of OsVI in the presence of tmeda as a
stabilizing ligand.[15] Even at near neutral pH, condensation of
the amino alcohol with Os was rapid, producing azaglycolate
ester 39 in 83% yield. Subsequent cyclization of OsVI ester 39
could be accomplished by reaction with an acid, thus
producing the pyrrolidine 4, as before, in a two step procedure
that validates the ability of OsVI to undergo both coordination
with the substrate and oxidative cyclization under acidic
conditions. Repetition of this sequence with the nosyl
analogue 11 (not shown) enabled us to obtain a crystal
structure (40) of the corresponding intermediate azaglycolate
osmate ester, before it was cyclized with acid.
ized by placing the bystander allyl group in the less crowded
exo position (A), rather than the endo position (B), where it
experiences a greater steric clash with the incipient pyrroli-
dine ring. The smaller hydrogen atom in 31 or hydroxy group
in 29 is then forced to adopt this more crowded position with a
smaller energetic penalty. The advantages of positioning an
allyl group onto the pyrrolidine backbone are manifold when
one considers opportunities for cross-metathesis, oxidative
cleavage, or alkene isomerization.
Our next task was to investigate more fully the role of
PNO in the above cyclization reactions. Initial observations
centered around the hypothesis that this oxidant does not
form OsVIII in the reaction mixture. To support this idea, we
report that PNO does not promote turnover in an UpJohn
style dihydroxylation reaction of alkenes.[10]
Next, we examined a stoichiometric model for the key
oxidation of OsVI to OsVIII which was modeled after a system
originally described by Sharpless and co-workers to probe the
effect of the second catalytic cycle in the asymmetric
dihydroxylation reaction.[11] The reaction of orange complex
33 with cyclohexene enabled the formation of (green) OsVI
compound 34 (Scheme 6).[12] When this complex was treated
with TMO a color change was observed to give putative
complex 35;[13] the formation of OsVIII was then verified as 35
was able to oxidize a second alkene that was added after the
color change was complete. Reductive cleavage of the
osmium gave 37 and 38 in a 1:1 ratio. Substitution of TMO
with PNO in this reaction did not cause the color change/
reoxidation of 34, and the osmium complex that resulted was
not able to oxidize a second alkene as observed in the first
experiment; indeed the alkene could be recovered (this
experiment was also repeated in the presence of six equiv-
alents of CSA to replicate the acidic oxidative cyclization
Scheme 7. Mechanistic probes for osmium oxidation.
tmeda=N,N,N’,N’-tetramethylethylenediamine.
2874 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2872 –2875

Angewandte
Chemie
B. Brunner, S. Streiff, P. Dübon, G. Helmchen, Org. Lett. 2005, 7,
1239 – 1242; d) F. A. Davis, H. Xu, Y. Wu, J. Zhang, Org. Lett.
2006, 8, 2273 – 2276.
To conclude, we have shown that PNO is a mild new
reoxidant for oxidative cyclization that dramatically improves
the range of substrates that are compatible with this reaction;
N-Z carbamate-protected amino alcohols and amino acids
were cyclized for the first time. The role of PNO was
investigated and it appears to act as a reoxidant for osmium
that does not produce OsVIII in situ. This reoxidation has
beneficial consequences for the efficiency of the cyclization
and has ramifications for the use of this chemistry on a large
scale.
[5] For the benefits that citric acid provides to osmium-catalyzed
dihydroxylations, see P. Dupau, R. Epple, A. A. Thomas, V. V.
Fokin, K. B Sharpless, Adv. Synth. Catal. 2002, 344, 421 – 433.
[6] There is good evidence to suggest that protected amino alcohols
are ligands for osmium: M. A. Andersson, R. Epple, V. V. Fokin,
K. B. Sharpless, Angew. Chem. 2002, 114, 490 – 493; Angew.
Chem. Int. Ed. 2002, 41, 472 – 475.
[7] Unnatural proline-like amino acids are useful for the synthesis of
peptides of therapeutic value and also in organic synthesis as
catalysts and reagents in their own right: N. Vignola, B. List, J.
Am. Chem. Soc. 2004, 126, 450 – 451.
Received: November 27, 2007
Revised: January 17, 2007
Published online: March 5, 2008
[8] I. B. Parr, S. K. Boehlein, A. B. Dribben, S. M. Schuster, N. G. J.
Richards, J. Med. Chem. 1996, 39, 2367 – 2378.
[9] The stereochemistry of all new pyrrolidine products was proven
by correlation to compounds on which an X-ray crystal structure
had been obtained, except compound 20 which was assigned by
analogy. Enantiomeric excess measurements were done by using
1HNMR spectroscopy and an enantiopure additive, and in
comparison with a racemic standard; greater than 95% ee means
that the other isomer was not detected by this technique.
[10] Oxidation of cyclohexene in acetone/water with catalytic OsO4
and PNO as reoxidant produces no detectable diol under a
variety of conditions (which include heating or the addition of
acid).
[11] J. S. M. Wai, I. Markꢀ, J. M. Svendsen, M. G. Finn, E. N.
Jacobsen, and K. B. Sharpless, J. Am. Chem. Soc. 1989, 111,
1123 – 1125.
[12] M. Schröder, A. J. Neielson, W. P. Griffith, J. Chem. Soc. Dalton
Trans. 1979, 1607 – 1611.
[13] See: a) reference [6]; b) P. Wu, R. Hilgraf, V. V. Fokin, Adv.
Synth. Catal. 2006, 348, 1079 – 1085.
[14] a) K. S. Coleman, L. J. L. Bedel, J. A. Osborne, C. R. Acad. Sci.
Ser. IIc 2000, 765 – 769; b) J. Muldoon, S. N. Brown, Org. Lett.
2002, 4, 1043 – 1045; c) K. Muꢁiz, I. Almodovar, J. Streuff, M.
Nieger, Adv. Synth. Catal. 2006, 348, 1831 – 1835.
[15] a) J. A. Ragazzo, E. J. Behrman, Bioinorg. Chem. 1976, 5, 343 –
352; b) T. J. Donohoe, M. J. Chughtai, D. J. Klauber, D. Griffin,
A. D. Campbell, J. Am. Chem. Soc. 2006, 128, 2514 – 2515.
Keywords: homogeneous catalysis · osmium · oxidation ·
pyrrolidines
.
[1] a) P. H. J. Carlsen, T. Katsuki, V. S. Martin, K. B. Sharpless, J.
Org. Chem. 1981, 46, 3936 – 3938; b) M. de ChampdorØ, M.
Lasalvia, V. Piccialli, Tetrahedron Lett. 1998, 39, 9781 – 9784;
c) V. Piccialli, N. Cavallo, Tetrahedron Lett. 2001, 42, 4695 – 4699;
d) L. Albarella, D. Musumeci, D. Sica, Eur. J. Org. Chem. 2001,
997 – 1003; e) T. J. Donohoe, S. Butterworth, Angew. Chem.
2003, 115, 978 – 981; Angew. Chem. Int. Ed. 2003, 42, 948 – 951;
f) S. Roth, S. Göhler, H. Cheng, C. B. W. Stark, Eur. J. Org.
Chem. 2005, 4109 – 4118.
[2] T. J. Donohoe, S. Butterworth, Angew. Chem. 2005, 117, 4844 –
4846; Angew. Chem. Int. Ed. 2005, 44, 4766 – 4768. See also D. M.
Walba, G. S. Stoudt, Tetrahedron Lett. 1982, 23, 727 – 730.
[3] T. J. Donohoe, G. H. Churchill, K. M. P. Wheelhouse (nØe
Gosby), P. A. Glossop, Angew. Chem. 2006, 118, 8193 – 8196;
Angew. Chem. Int. Ed. 2006, 45, 8025 – 8028.
[4] For related methods of making enantiopure THFs, see:
a) R. C. D. Brown, J. F. Keily, Angew. Chem. 2001, 113, 4628 –
4630; Angew. Chem. Int. Ed. 2001, 40, 4496 – 4498; b) G. D.
Head, W. G. Whittingham, R. C. D. Brown, Synlett 2004, 1437 –
1439. For enantiopure pyrrolidines, see: c) C. Welter, A. Dahnz,
Angew. Chem. Int. Ed. 2008, 47, 2872 –2875
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2875