Synthesis of oxacyclic scaffolds via dual ruthenium hydride/Bronsted acid-catalyzed isomerization/cyclization of allylic ethers

Article abstract of DOI:10.1002/chem.201304270

A ruthenium hydride/Bronsted acid-catalyzed tandem sequence is reported for the synthesis of 1,3,4,9-tetrahydropyrano[3,4-b]indoles (THPIs) and related oxacyclic scaffolds. The process was designed on the premise that readily available allylic ethers would undergo sequential isomerization, first to enol ethers (Ru catalysis), then to oxocarbenium ions (Bronsted acid catalysis) amenable to endo cyclization with tethered nucleophiles. This methodology provides not only an attractive alternative to the traditional oxa-Pictet-Spengler reaction for the synthesis of THPIs, but also convenient access to THPI congeners and other important oxacycles such as acetals.

Full text of DOI:10.1002/chem.201304270

DOI: 10.1002/chem.201304270
Communication
&
Tandem Catalysis
Synthesis of Oxacyclic Scaffolds via Dual Ruthenium Hydride/
Brønsted Acid-Catalyzed Isomerization/Cyclization of Allylic Ethers
Erhad Ascic,[a] Ragnhild G. Ohm,[a] Rico Petersen,[a] Mette R. Hansen,[a] Casper L. Hansen,[a]
Daniel Madsen,[a] David Tanner,[a] and Thomas E. Nielsen*[a, b]
that undergo nucleophilic attack by the electron-rich indole
Abstract: A ruthenium hydride/Brønsted acid-catalyzed
tandem sequence is reported for the synthesis of 1,3,4,9-
moiety. We now report a purely intramolecular alternative to
the traditional synthesis of THPIs starting from readily available
tetrahydropyrano[3,4-b]indoles (THPIs) and related oxacy-
indole-tethered allylic ethers, via tandem transition metal/
clic scaffolds. The process was designed on the premise
Brønsted acid-catalyzed isomerizations to form the key oxocar-
that readily available allylic ethers would undergo sequen-
benium intermediates (Scheme 1). The final step of the se-
tial isomerization, first to enol ethers (Ru catalysis), then to
quence shown in Scheme 1 is obviously related to the classical
oxocarbenium ions (Brønsted acid catalysis) amenable to
endo cyclization with tethered nucleophiles. This method-
ology provides not only an attractive alternative to the
traditional oxa-Pictet–Spengler reaction for the synthesis
of THPIs, but also convenient access to THPI congeners
and other important oxacycles such as acetals.
The 1,3,4,9-tetrahydropyrano[3,4-b]indole (THPI) skeleton
(Figure 1) is of considerable interest in the field of medicinal
chemistry, as exemplified by pharmaceuticals such as etodolac
(a nonstereoidal anti-inflammatory drug),[1] pemedolac (an ef-
fective analgesic agent),[2] and HCV-371 (which displays potent
antiviral activity).[3]
Scheme 1. Formation of oxacycles through metal-catalyzed isomerization se-
quences.
One time-honored synthetic route to this class of heterocy-
cles is the oxa-Pictet–Spengler reaction,[4] in which tryptophols
are condensed under acidic conditions with carbonyl com-
pounds (or masked forms thereof) to give oxocarbenium ions
Pictet–Spengler cyclization of tryptamine derivatives, which
normally involves acid-catalyzed intermolecular condensation
with carbonyl compounds to form the corresponding iminium
species. Recently, alternative methods to access such electro-
philes by means of metal-catalyzed isomerization of allylic
amines (via enamines) have been explored by a number of lab-
oratories, including our own.[5] While numerous publications
have dealt with the metal-catalyzed isomerization of allylic
ethers to enol ethers,[6] we are unaware of “one-pot” (tandem)
procedures for the sequential generation and reaction of oxo-
carbenium ions,[7,8] as depicted in Scheme 1.
Figure 1. THPI and some pharmaceutically important derivatives thereof.
Allylic ether 1a was chosen as substrate for preliminary
studies of the proposed tandem process (Table 1). Based on
previous experience,[5] optimization was carried out using
5 mol% of Ru-, Pd- and Rh-based catalysts in refluxing toluene,
and initial results indicated that RuHCl(CO)(PPh3)3 was extreme-
ly efficient for the isomerization of 1a to 1a’, the latter being
formed as a 2:3 (E:Z) mixture of diastereomers in near-quanti-
tative yield within 1 hour (Table 1, entry 1). Further transforma-
tion to THPI 2a did not occur even after prolonged reaction
time (22 h), indicating that the metal catalyst alone could not
promote the second step of the desired tandem sequence.
However, to our delight, the addition of 5 mol% of diphenyl
[a] Dr. E. Ascic, R. G. Ohm, Dr. R. Petersen, Dr. M. R. Hansen, C. L. Hansen,
D. Madsen, Prof. Dr. D. Tanner, Prof. Dr. T. E. Nielsen
Department of Chemistry
Technical University of Denmark
2800 Kgs. Lyngby (Denmark)
Fax: (+45)45-93-39-68
E-mail: ten@kemi.dtu.dk
[b] Prof. Dr. T. E. Nielsen
Singapore Centre on Environmental Life Sciences Engineering
Nanyang Technological University
Singapore 637551 (Singapore)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304270.
Chem. Eur. J. 2014, 20, 3297 – 3300
3297
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 1. Screening of catalysts for the formation of THPI 2a.
Table 2. Substrate scope of the tandem sequence to THBIs.
Entry
Catalyst
Ratio 1a/1a’/2a[a]
1 h
Entry Substrate R1
R2
R3
R4
Product, yield [%][a]
6 h
22 h
1[b]
2[b]
3[b]
4[b]
5[d]
6[d]
7[d]
RuHCl(CO)(PPh3)3
Pd(PPh3)4
Rh(Ph3P)3Cl
Grubbs I
Grubbs II
0:100:0 (0:0:100)[c]
65:29:6
66:30:4
100:0:0
16:84:0
41:45:14
62:38:0
0:95:5
13:19:68
24:65:11
–
–
–
–
0:95:5
0:0:100
0:100:00
–
–
–
–
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1 f
1g
1h
1i
H
Me
Et
CH2OTBDPS
CH2CN
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
2a, 80
2b, 83
2c, 79
2d, 88
2e, 60[b]
2 f, 41
2g, 62
2h, 63
2i, 70
7-Et
4-OBn
5-OMe
6-iPr
7-Et
H
H
H
7-Et
Hoveyda–Grubbs I
Hoveyda–Grubbs II
[a] Determined by RP-HPLC (UV detection at 215 nm). [b] The product
mixture was very clean (>85% of 1a/1a’/2a in the reaction mixture).
[c] The reaction was carried out with 5 mol% of diphenyl phosphate and
5 mol% of RuHCl(CO)(PPh3)3. [d] The purity of the product mixture was
low (<50% of 1a/1a’/2a in the reaction mixture) due to formation of
the cross-metathesis dimer.
9
H
10
11
12
13
1j
1k
1l
Me
Ph
H
2j, 75
2k, 65
Me 2l, 53[c]
1m
Me Me 2m, 35[d]
[a] Yield of the isolated products after flash column chromatography.
[b] 2 equivalent of TFA and 10 mol% of RuHCl(CO)(PPh3)3 was added.
[c] 10 mol% of the catalysts were used. [d] 30 mol% of the catalyst was
used in refluxing m-xylene.
phosphate at the beginning of the reaction allowed the
tandem process to proceed smoothly with a full and clean
conversion of 1a to the THPI product 2a in 1 hour. It was
noted that Pd(PPh3)4 mediated a clean conversion of 1a to 2a,
albeit 22 h was required for full conversion (Table 1, entry 2).
Wilkinson’s catalyst and Grubbs-type catalysts are known to be
effective for the isomerization of allylic amines into iminium in-
termediates.[5] However, whereas Rh(Ph3P)3Cl led to incomplete
reaction due to slow isomerization (Table 1, entry 3), the use of
Ru alkylidenes led to substantial formation of the cross-meta-
thesis dimer (Table 1, entries 4–7).
synthesis of THPIs. For example, when tryptophol 3 is reacted
with propanal 4 in the presence of the Lewis or Brønsted acids
commonly used for this transformation, the bis-
(indolyl)methane byproduct 5 is invariably formed together
with 2a (Scheme 2), under standard reaction conditions. In our
hands, the reaction was generally characterized by an incom-
With suitable reaction conditions in hand, the substrate
scope was then examined. As is evident from the results pre-
sented in Table 2, a combination of 5 mol% of RuHCl(CO)-
(PPh3)3 and 5 mol% of diphenyl phosphate in refluxing toluene
proved to be effective for a wide range of indole substrates.
The variation of R1 in substrates 1a–d was well tolerated af-
fording THPIs 2a–d in 79–88% yield (Table 2, entries 1–4). Sub-
strate 1e was not converted to 2e under these reaction condi-
tions. However, a clean isomerization to the acrylonitrile inter-
Scheme 2. Oxa-Pictet–Spengler cyclization of tryptophol 3.
plete conversion of the starting material, accompanied by for-
mation of numerous, unidentified byproducts, and the maxi-
mum yield of 2a was only 35%.[9] Formation of compounds
such as 5 under classical oxa-Pictet–Spengler conditions is well
documented.[10]
1
mediate was evident from UPLC- MS and H NMR analyses, in-
dicating that a stronger Brønsted acid was required to bring
the double bond out of conjugation towards 2e. The problem
was solved by treating 1e with 10 mol% RuHCl(CO)(PPh3)3 at
reflux for 1 hour, followed by addition of TFA (2 equiv), which
afforded 2e in 60% yield (Table 2, entry 5). Substrates 1 f–
i bearing substitution on the indole core were also suitable for
the reaction, providing the corresponding products 2 f–i in
slightly lower yields (41–70%; Table 2, entries 6–9). Variation of
R3 was also tolerated which led to products 2j and 2k in good
yields (65–75%). On the other hand, when changing the
alkene substitution pattern (substrates 1l and 1m), approxi-
mately 15–30 mol% of catalyst was needed to obtain products
2l and 2m, in 53 and 35% yield, respectively (Table 2, en-
tries 12 and 13). The present method offers distinct advantages
over the conventional oxa-Pictet–Spengler cyclization for the
The success of the present catalytic tandem approach to
THPIs prompted the investigation of an enantioselective ver-
sion. Recently, Terada and Toda demonstrated that RuHCl(CO)-
(PPh3)3 was compatible with chiral phosphoric acids to pro-
mote the synthesis of tetrahydroisoquinolines from allylic
amines in up to 47% ee.[5e] In addition, List and co-workers
have developed a new type of chiral acid (imidodiphosphoric
acid) that effectively catalyzes oxocarbenium ion cyclizations
to form enantiopure acetals and spiroacetals.[11] After screening
a range of acid catalysts,[12] we found that the use of 5 mol%
imidodiphosphoric acid 14 together with 5 mol% of
Chem. Eur. J. 2014, 20, 3297 – 3300
www.chemeurj.org
3298
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 3. Substrate scope of the tandem sequence for the synthesis of
acetals.
Entry Substrate
R1 R2 R3 R4 n dr
Product, yield [%][a]
2:1 7a, 72
1
2
3
4
5
6
7
8
9
6a
6b
6c
6d
6e
6 f
6g
6h
6i
Ph
Ph
Bn
H
H
H
H
Ph H
H
H
0
0
0
0
1
1
>20:1 7b, 95
1:1 7c, 89
>20:1 7d, 83
8:1 7e, 83
2:1 7 f, 68
H
Me Me Ph H
Ph
H
H
H
H
H
H
H
Ph H
H
Scheme 3. Asymmetric version of the tandem sequence for the synthesis of
THIP 2b.
H
Ph
Ph
H
H
H
Ph 1
–
7g, 44
H
H
2
3
6:5 7h, 85
1:1 7i, 49
RuHCl(CO)(PPh3)3 in refluxing toluene provided the THPI 2b
with an ee value of 47% and in 75% isolated yield (Scheme 3).
Current studies are being focused on improvement of the
enantioselectivity.
10
6j, (m=1)
6k, (m=2)
7j, (m=1), 13
7k, (m=2), 29
Returning to the racemic mode of catalysis, it was found
that substrates that are less nucleophilic than indole could also
be accommodated. Reactions with phenyl, benzothiophene, or
thiophene moieties generally benefited from the addition of
TsOH (5-100 mol%) together with RuHCl(CO)(PPh3)3 (5 mol%),
providing a range of novel heterocycles, usually in high isolat-
ed yields (Scheme 4).
11
[a] Yield of the isolated products after flash column chromatography.
oxacycles from allylic ethers.[13] The process, co-catalyzed by
a ruthenium hydride species and a Brønsted acid, provides
convenient access to a variety of tetrahydropyranoindoles and
related heterocyclic scaffolds in good to excellent yields. This
method is generally superior to the traditional oxa-Pictet–
Spengler reaction, since it circumvents the formation of bis-
(indolyl)methane byproducts. Finally, use of a chiral imidodi-
phosphoric acid in tandem with the ruthenium catalyst provid-
ed up to 47% ee in the synthesis of a THIP containing a quater-
nary stereogenic center. Efforts are now being made to expand
the scope and improve the efficiency of the enantioselective
process.
Experimental Section
General procedure for the Ru hydride/Brønst-
ed acid catalyzed tandem sequence
Scheme 4. Substrate scope of the tandem sequence for the synthesis of
fused oxacycles. [a] 1 equivalent of TsOH was added.
RuHCl(CO)(PPh3)3 (5 mol%) and diphenyl phosphate (5 mol%) were
added at room temperature to a solution of allyl ether substrate in
dry toluene (0.1m) and under argon atmosphere. The reaction mix-
ture was heated to 1158C and was followed by TLC. Upon full con-
sumption of the starting material, the reaction was cooled to room
temperature and the contents were purified by flash column chro-
matography on silica gel with n-hexane/EtOAc, to obtain the de-
sired product.
The new tandem process was also applicable to oxygen-
based nucleophiles to provide an entry to acetals (Table 3).
Upon exposure to 5 mol% of RuHCl(CO)(PPh3)3 and 5 mol% di-
phenyl phosphate in refluxing toluene, the alcohol subtrates
6a–k underwent cyclization to form 5–8 membered acetals
7a–k in 13–95% yield (Table 3, entries 1–11). In general, mix-
tures of diastereomers were formed as expected (Table 3, en-
tries 1, 3, 6, 8, and 9), although the cyclization of the more
highly substituted substrate 6d occurred with excellent diaste-
reoselectivity (Table 3, entry 4).
Acknowledgements
The Danish Council for Independent Research (Technology and
Production), the DSF Center for Antimicrobial Research, and
the Lundbeck Foundation are gratefully acknowledged for fi-
nancial support.
In summary, we have demonstrated an isomerization/oxocar-
benium ion cyclization tandem sequence for the synthesis of
Chem. Eur. J. 2014, 20, 3297 – 3300
www.chemeurj.org
3299
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[7] For a report on the isomerization of allylic ethers and formation of ace-
tals catalyzed by octacarbonyldicobalt, see: B. H. J. Chang, Organomet.
Chem. 1995, 492, 31.
Keywords: cyclization · isomerization · oxacycles · Pictet–
Spengler reaction · ruthenium
[8] For the synthesis of acetals via addition of alcohols to O-allylic com-
pounds, see: a) S. Krompiec, M. Penkala, E. Kowalska, R. Penczek, P.
Bujak, W. Danikiewicz, G. Spꢁlnik, A. Kita, I. Grudzka, Monatsh. Chem.
2011, 142, 1241; b) S. Krompiec, R. Penczek, P. Bujak, E. Kubik, J. Malarz,
M. Penkala, M. Krompiec, N, Kuznik, H, Maciejewski, Tetrahedron Lett.
2009, 50, 1193; c) S. Krompiec, R. Penczek, M. Penkala, M. Krompiec, J.
Rzepa, M. Matlengiewicz, J. Jaworska, S. Baj, J. Mol. Catal. A 2008, 290,
15.
[9] For a full account on reaction conditions tried for this transformation
see the Supporting Information.
[10] a) B. Bouguerne, C. Lherbet, M. Baltas, Lett. Org. Chem. 2010, 7, 420;
b) R. Ghorbani-Vaghei, H. Veisi, Mol. Diversity 2010, 14, 87; c) G. V. M.
Sharma, J. J. Reddy, P. S. Lakshmi, R. P. Krishna, Tetrahedron Lett. 2004,
45, 7729.
[1] C. A. Demerson, N. A. Abraham, G. Schilling, R. R. Martel, C. J. Asciak, J.
Med. Chem. 1983, 26, 1778.
[2] A. H. Katz, C. A. Demerson, C. C. Shaw, A. A. Asselin, L. G. Humber, K. M.
Conway, G. Gavin, C. Guinosso, N. P. Jensen, D. Mobilio, R. Noureldin, J.
Schmid, U. Shah, D. Van Engen, T. T. Chau, B. M. Weichman, J. Med.
Chem. 1988, 31, 1244.
[3] A. Gopalsamy, G. Ciszeski, K. Park, J. W. Ellingboe, J. Blom, S. Insaf, J. Up-
eslacis, T. S. Mansour, G. Krishnamurthy, M. Damarla, Y. Pyatski, D. Ho, J.
Med. Chem. 2004, 47, 6603.
[4] Fore recent reviews on oxa-Pictet–Spengler cyclizations, see: a) E. L.
Larghi, T. S. Kaufman, Eur. J. Org. Chem. 2011, 5195; b) E. L. Larghi, T. S.
Kaufman, Synthesis 2006, 187.
[5] a) K. Sorimachi, M. Terada, J. Am. Chem. Soc. 2008, 130, 14452; b) E.
Ascic, C. L. Hansen, S. T. Le Quement, T. E. Nielsen, Chem. Commun.
2012, 48, 3345; c) E. Ascic, S. T. Le Quement, M. Ishoey, M. Daugaard,
T. E. Nielsen, ACS Comb. Sci. 2012, 14, 253; d) E. Ascic, J. F. Jensen, T. E.
Nielsen, Angew. Chem. 2011, 123, 5294; Angew. Chem. Int. Ed. 2011, 50,
5188; e) Y. Toda, M. Terada, Synlett 2013, 24, 752; f) Q. Cai, X. W. Liang,
S. G. Wang, J. W. Zhang, X. Zhang, S. L. You, Org. Lett. 2012, 14, 5022;
g) Q. Cai, X. W. Liang, S. G. Wang, S. L. You, Org. Biomol. Chem. 2013, 11,
1602.
ˇ
´
[11] a) J. H. Kim, I. Coric, S. Vellalath, B. List, Angew. Chem. Int. Ed. 2013, 52,
ˇ
´
4474; b) I. Coric, B. List, Nature 2012, 483, 315.
[12] For a full account on the reaction optimization see the Supporting In-
formation.
[13] While this manuscript was under review, a related tandem isomeriza-
tion was reported in: V. M. Lombardo, C. D. Thomas, K. A. Scheidt,
Angew. Chem. Int. Ed. 2013, 52, 12910.
[6] For a review on transition metal-catalyzed isomerization of allyl ethers,
see: N. Kuznik, S. Krompiec, Coord. Chem. Rev. 2007, 251, 222.
Received: October 31, 2013
Published online on February 24, 2014
Chem. Eur. J. 2014, 20, 3297 – 3300
www.chemeurj.org
3300
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim