Iodine-mediated electrophilic cyclization of 2-alkynyl-1-methylene azide aromatics leading to highly substituted isoquinolines and its application to the synthesis of norchelerythrine

Article abstract of DOI:10.1021/ja805326f

The reaction of 2-alkynyl-1-methylene azide aromatics 1 with iodine and/or other iodium donors, such as the Barluenga reagent (Py₂IBF ₄/HBF₄) and NIS, gave highly substituted cyclization products, namely, the 1,3-disubstituted 4-iodoisoquinolines 2, in good to high yields. Not only simple 2-alkynyl benzyl azides 1a-j and their substituted analogues 1k-u and 6 but also heteroaromatic analogues, including pyridine 8, pyrroles 10a-c, furane 10d, and thiophenes 10e-g, gave the corresponding isoquinoline derivatives in excellent to allowable yields. Electron-donating and electron-accepting substituents on the aromatic ring were equally tolerated, and either acidic or basic (or even neutral) reaction conditions, depending on the reactivity of the substrate, could be applied to smoothly convert the azide starting materials into the desired isoquinoline products in moderate to good yields. Limits were found only in connection with the substituent at the alkyne terminus, where electron-neutral or electron-donating substituents are clearly favored. The iodine-mediated electrophilic cyclization of 1 most probably proceeds through the iodonium ion intermediate 4 followed by nucleophilic cyclization of the azide and subsequent elimination of N₂. This new methodology was successfully applied to the short synthesis of norchelerythrine.

Full text of DOI:10.1021/ja805326f

Published on Web 10/23/2008
Iodine-Mediated Electrophilic Cyclization of
2-Alkynyl-1-methylene Azide Aromatics Leading to Highly
Substituted Isoquinolines and Its Application to the Synthesis
of Norchelerythrine
Dirk Fischer, Hisamitsu Tomeba, Nirmal K. Pahadi, Nitin T. Patil, Zhibao Huo, and
Yoshinori Yamamoto*
Department of Chemistry, Graduate School of Science, Tohoku UniVersity, Sendai 980-8578, Japan
Received July 10, 2008; E-mail: yoshi@mail.tains.tohoku.ac.jp
Abstract: The reaction of 2-alkynyl-1-methylene azide aromatics 1 with iodine and/or other iodium donors,
such as the Barluenga reagent (Py2IBF4/HBF4) and NIS, gave highly substituted cyclization products, namely,
the 1,3-disubstituted 4-iodoisoquinolines 2, in good to high yields. Not only simple 2-alkynyl benzyl azides
1a-j and their substituted analogues 1k-u and 6 but also heteroaromatic analogues, including pyridine 8,
pyrroles 10a-c, furane 10d, and thiophenes 10e-g, gave the corresponding isoquinoline derivatives in
excellent to allowable yields. Electron-donating and electron-accepting substituents on the aromatic ring
were equally tolerated, and either acidic or basic (or even neutral) reaction conditions, depending on the
reactivity of the substrate, could be applied to smoothly convert the azide starting materials into the desired
isoquinoline products in moderate to good yields. Limits were found only in connection with the substituent
at the alkyne terminus, where electron-neutral or electron-donating substituents are clearly favored. The
iodine-mediated electrophilic cyclization of 1 most probably proceeds through the iodonium ion intermediate
4 followed by nucleophilic cyclization of the azide and subsequent elimination of N2. This new methodology
was successfully applied to the short synthesis of norchelerythrine.
Introduction
synthesis, as well as for fine-tuning of the biological and/or
physical properties of those compounds for final application, a
Isoquinoline and heterocyclic pyridine derivatives are an
important class of alkaloids, commonly found in natural
products.1 Because of their biological activities, they are often
used as building blocks in pharmaceutical compounds.2 Fur-
thermore, they are utilized as chiral ligands for transition-metal
catalysts,3 and their iridium complexes are used in organic light-
emitting diodes.4 For delicate and sophisticated natural product
general and flexible approach to this class of heterocycles that
tolerates a wide variety of functional groups is highly desirable.
Although many methods for synthesizing isoquinoline ana-
logues are known, classic methods such as the Pomeranz-Fritsch
reaction have considerable drawbacks, including the use of
strong acids and elevated temperatures,5 which are not suitable
for sensitive substrates. For that reason, not all functional and
protecting groups are tolerated. In recent years, especially in
the laboratories of Larock and co-workers, new transition-metal
catalyzed reactions using phenyl acetylenes as starting materials
have been developed for the synthesis of substituted isoquino-
lines (Scheme 1).6 These reactions have proven to be extremely
efficient in the synthesis of a wide variety of 3,4-substituted
(1) Bentley, K. W. The Isoquinoline Alkaloids; Harwood Academic:
Amsterdam, 1998; Vol 1.
(2) Selected papers: (a) Dzierszinski, F.; Coppin, A.; Mortuaire, M.;
Dewally, E.; Slomianny, C.; Ameisen, J.-C.; DeBels, F.; Tomavo, S.
Antimicrob. Agents Chemother. 2002, 46, 3197–3207. (b) Kletsas, D.;
Li, W.; Han, Z.; Papadopoulos, V. Biochem. Pharmacol. 2004, 67,
1927–1932. (c) Mach, U. R.; Hackling, A. E.; Perachon, S.; Ferry,
S.; Wermuth, C. G.; Schwartz, J.-C.; Sokoloff, P.; Stark, H. Chem-
BioChem 2004, 5, 508–518. (d) Muscarella, D. E.; O’Brian, K. A.;
Lemley, A. T.; Bloom, S. E. Toxicol. Sci. 2003, 73, 66–73.
(3) Selected papers: (a) Sweetman, B. A.; Mu¨ller-Bunz, H.; Guiry, P. J.
Tetrahedron Lett. 2005, 46, 4643–4646. (b) Durola, F.; Sauvage, J.-
P.; Wenger, O. S. Chem. Commun. 2006, 171–173. (c) Lim, C. W.;
Tissot, O.; Mattison, A.; Hooper, M. W.; Brown, J. M.; Cowley, A. R.;
Hulmes, D. I.; Blacker, A. J. Org. Process Res. DeV. 2003, 7, 379–
384. (d) Alcock, N. W.; Brown, J. M.; Hulmes, G. I. Tetrahedron:
Asymmetry 1993, 4, 743–756.
(5) (a) Whaley, W. M.; Govindachari. T. R. In Organic Reactions; Adams,
R., Ed.; Wiley: New York, 1951; Vol. 6, pp 151-190. (b) Whaley,
W. M.; Govindachari, T. R. In Organic Reactions; Adams, R., Ed.;
Wiley: New York, 1951; Vol. 6, pp 74-150. (c) Gensler, W. J. In
Organic Reactions; Adams, R., Ed.; Wiley: New York, 1951; Vol. 6,
pp 191-206. (d) Bentley, K. W. Nat. Prod. Rep. 2005, 22, 249–268.
(6) Selected papers: (a) Maassarani, F.; Pfeffer, M.; Le Borgne, G.
J. Chem. Soc., Chem. Commun. 1987, 565–567. (b) Wu, G.; Geib, S.;
Rheingold, A. L.; Heck, R. F. J. Org. Chem. 1988, 53, 3238–3241.
(c) Girling, I. R.; Widdowson, D. A. Tetrahedron Lett. 1982, 23, 4281–
4284. (d) Roesch, K. R.; Zhang, H.; Larock, R. C. J. Org. Chem.
1998, 63, 5306–5307. (e) Roesch, K. R.; Zhang, H.; Larock, R. C. J.
Org. Chem. 2001, 66, 8042–8051. (f) Dai, G.; Larock, R. C. J. Org.
Chem. 2002, 67, 7042–7047. (g) Huang, Q.; Larock, R. C. J. Org.
Chem. 2003, 68, 980–988. (h) Hashmi, A. S. K.; Schaefer, S.; Woelfe,
M.; Diez Gil, C.; Fischer, P.; Laguna, A.; Blanco, M. C.; Gimeno,
M. C. Angew. Chem., In. Ed. 2007, 46, 6184–6187.
(4) Selected papers: (a) Fang, K.-H.; Wu, L.-L.; Huang, Y.-T.; Yang, C.-
H.; Sun, I.-W. Inorg. Chim. Acta 2006, 359, 441–450. (b) Liu, S.-J.;
Zhao, Q.; Chen, R.-F.; Deng, Y.; Fan, Q.-L.; Li, F.-Y.; Wang, L.-H.;
Huang, C.-H.; Huang, W. Chem. Eur. J. 2006, 12, 4351–4361. (c)
Zhao, Q.; Liu, S.; Shi, M.; Wang, C.; Yu, M.; Li, L.; Li, F.; Yi, T.;
Huang, C. Inorg. Chem. 2006, 45, 6152–6160. (d) Tsuboyama, A.;
Iwawaki, H.; Furugori, M.; Mukaide, T.; Kamatani, J.; Igawa, S.;
Moriyama, T.; Miura, S.; Takiguchi, T.; Okada, S.; Hoshino, M.; Ueno,
K. J. Am. Chem. Soc. 2003, 125, 12971–12979.
9
15720 J. AM. CHEM. SOC. 2008, 130, 15720–15725
10.1021/ja805326f CCC: $40.75
2008 American Chemical Society

Iodine-Mediated Electrophilic Cyclization
A R T I C L E S
Scheme 1. Transition-Metal Catalyzed 3,4-Disubstituted
Isoquinoline Synthesis
Table 1. Iodine-Mediated Cyclization of 2-Alkynylbenzyl Azides 1
yield (%)a
Scheme 2. PtBr2-Catalyzed Isoquinoline Synthesis
entry compd 1
R
I+
base
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
2
3
b
b
1
2
3
4
5
1a
1b
1c
1d
1e
1f
Ph
I2
I2
I2
-
-
-
-
-
95
94
95
82
p-OMe-Ph
p-CF3-Ph
1-cyclohexenyl I2
tert-butyl
2-pyridinyl
H
b,c
b
b
I2
I2
I2
71 (21d)
b
6
-
-
86
56
-
-
-
b,d
7
8
9
1g
1h
1i
K3PO4
isoquinolines and other carbo- and heterocycles.7 It is, however,
not possible to synthesize 1,3,4-trisubstituted isoquinolines by
these methods.
Py2IBF4 (NIS)f
-
-
-
(NaHCO3)f 55 (42)f
e
e
e
Me
Py2IBF4 (NIS)f
(NaHCO3)f 67 (66)f
e
e
e
e
n-butyl
CH2TMS
Py2IBF4 (NIS)f
(NaHCO3)f 69 (62)f
e
e
10
1j
a Isolated yield. b I2 (5 equiv), K3PO4 (5 equiv), CH2Cl2 (0.1 M), rt,
24 h, Ar. c Reaction time 72 h. d The crystal structure is given in the
Supporting Information for ref 7f. e Py2IBF4 (2 equiv), HBF4 in Et2O
(2 equiv), CH2Cl2, -78 °C, 1 h, Ar. f NIS (5 equiv), NaHCO3 (1 equiv),
(CH2Cl)2 (0.1 M), 50 °C, 72 h, Ar.
During our research on the synthesis of functionalized indene
derivates with a platinum catalyst,8 we observed in one case
the formation of a 1,3-substituted isoquinoline (Scheme 2).9 To
date only aldimine-type substrates were known to undergo the
transformation shown in Scheme 1, but the result shown in
Scheme 2 suggests that proper choice of substrates and/or
reagents may overcome that limitation to afford 1-substituted
isoquinolines.
metals11 exhibit strong Lewis acidity toward unsaturated C-C
bonds (π-electrophilic nature) as well as toward lone pairs of
electrons associated with heteroatoms (σ-electrophilic nature).
Especially, gold-catalyzed reactions of alkynes are of keen
interest in modern organic synthesis.10 However, in the case of
the gold-catalyzed reactions, gold is not incorporated into the
final product, and a vinyl-gold intermediate that may be formed
during the reaction is not easily trapped by an external
nucleophile; therefore, further functionalization at C4 is not
possible. It was expected that an iodo substituent would be
introduced at C4 in the iodine-mediated reaction.
In our case, the two methyl substituents of the -NMe2
functional group act as the leaving group, formally eliminating
ethane. Upon consideration of the pathway of the transformation,
it occurred to us that azide (-N3), with nitrogen as the leaving
group, instead of dimethylamine would be a far better nitrogen
source of the isoquinoline framework. Furthermore, to gain
access to 1,3,4-trisubstituted isoquinolines, we considered using
an electrophile as the reaction mediator, since it would be
incorporated into the final product. Iodonium is well-suited for
this purpose, and hence, it offers clear and powerful potential
for further assembly of molecular diversity at C4. It is well-
known that coinage metals (Cu, Ag, Au)10 and other related
Results and Discussion
The initial cyclization study of 1a using 5 equiv of iodine
and 5 equiv of K3PO4 as the base in CH2Cl2 provided the desired
isoquinoline 2a in an excellent yield of 95% (Table 1, entry 1).
Further optimization of the reaction conditions (e.g., smaller
amounts of iodine, different bases such as Al2O3 or NEt3, other
solvents such as THF, or changes of temperature) failed to
improve and in some cases even decreased the yield. Therefore,
we continued to use the initially tested reaction conditions for
the synthesis of 3,4-disubstitued isoquinolines.
For R ) Ar (Table 1, entries 1-3), the cyclization of 1a-c
proceeded very smoothly to give the desired isoquinolines 2a-c
in yields of 94-95%, but in the case of 1c, which bears an
electron-withdrawing substituent, the reaction time had to be
extended to 3 days to reach full conversion. Besides the aromatic
substituents, the nonaromatic derivatives 1d and 1e (entries 4
and 5), bearing olefinic and sterically bulky tert-butyl substit-
uents, respectively, were tolerated without problems as well,
giving the desired products 2d and 2e in 82 and 71% yield,
respectively. It must nevertheless be mentioned that for 1e, the
reaction time was extended to 21 days to reach almost full
conversion. To our surprise, when R ) 2-pyridinyl (1f) and H
(7) Selected papers: (a) Zhang, D.; Liu, Z.; Yum, E. K.; Larock, R. C. J.
Org. Chem. 2007, 72, 251–262. (b) Yue, D.; Nicola, D. C.; Larrock,
R. C. J. Org. Chem. 2006, 71, 3381–3388. (c) Hessian, K. O.; Flynn,
B. L. Org. Lett. 2003, 5, 4377–4380. (d) Yao, X.; Li, C.-J. Org. Lett.
2006, 8, 1953–1955. (e) Asao, N.; Chan, S.; Takahashi, K.; Yamamoto,
Y. Tetrahedron 2005, 61, 11322–11326. (f) For a preliminary
communication of this full paper, see: Fischer, D.; Tomeba, H.; Pahadi,
N. K.; Patil, N. T.; Yamamoto, Y. Angew. Chem., Int. Ed. 2007, 46,
4764–4766. (g) Lyaskovskyy, V.; Froehlich, R.; Wuerthwein, E.-U.
Synlett 2007, 2733–2737. (h) Alonso, R.; Campos, P. J.; Garcia, B.;
Rodriguez, M. A. Org. Lett. 2006, 8, 3521–3523. (i) Asao, N.; Yudha
S, S.; Nogami, T.; Yamamoto, Y. Angew. Chem., Int. Ed. 2005, 44,
5526–5528. (j) Ohtaka, M.; Nakamura, H.; Yamamoto, Y. Tetrahedron
Lett. 2004, 45, 7339–7341. (k) Huo, Z.; Tomeba, H.; Yamamoto, Y.
Tetrahedron Lett. 2008, 49, 5531–5533.
(8) Bajracharya, G. B.; Pahadi, N. K.; Gridnev, I. D.; Yamamoto, Y. J.
Org. Chem. 2006, 71, 6204–6210.
(9) Pahadi, N. K. Ph.D. Thesis, Tohoku University (Sendai), 2005.
(10) Selected reviews and papers: (a) Lipschutz, B. A.; Yamamoto, Y.
Chem. ReV. 2008, 108, 2793–3442. (b) Hashmi, A. S. K. Chem. ReV.
2007, 107, 3180–3211. (c) Widenhoefer, R. A.; Han, X. Eur. J. Org.
Chem. 2006, 4555–4563. (d) Furstner, A.; Davies, P. W. Angew.
Chem., Int. Ed. 2007, 46, 3410–3449. (e) Jimenez-Nunez, E.;
Echavarren, A. M. Chem. Commun. 2007, 333–346. (f) Marion, N.;
Nolan, S. P. Angew. Chem., Int. Ed. 2007, 46, 2750–2752. (g) Marco-
Contelles, J.; Soriano, E. Chem.sEur. J. 2007, 13, 1350–1357. (h)
Patil, N. T.; Yamamoto, Y. ArkiVoc 2007, 10, 121–141. (i) Patil, N. T.;
Yamamoto, Y. ArkiVoc 2007, 5, 6–19. (j) Gorin, D. J.; Toste, F. D.
Nature 2007, 446, 395–403.
(11) (a) For computations on σ- versus π-electrophilicity of various metal
halides toward lone-pair σ electrons of carbonyl and imine groups
and toward π electrons of unsaturated C-C bonds, see: Yamamoto,
Y. J. Org. Chem. 2007, 72, 7817–7831. (b) Yamamoto, Y. J. Org.
Chem. 2008, 73, 5210.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15721

A R T I C L E S
Fischer et al.
Scheme 3. Plausible Reaction Mechanism
Table 2. Iodine-Mediated Synthesis of Highly Substituted
Isoquinolines
(1g) were tested, instead of the expected isoquinolines, the
corresponding bis(iodine) adducts 3f and 3g were obtained
selectively in 86 and 56% yield, respectively (entries 6 and 7).
The structure of 3g was determined unambiguously by X-ray
crystal structure analysis (see the Supporting Information for
ref 7f). Attempts to convert 3f and 3g to the desired isoquino-
lines 2f and 2g by prolonged heating or abstraction of iodide
by silver resulted only in decomposition of 3f and 3g. When
small- or medium-sized aliphatic substituents such as Me,
n-butyl, and CH2TMS (1h-j) were used, the desired cyclization
products (2h-j) were obtained by the use of the Barluenga
reagent or NIS (entries 8-10).
yield of
2 (%)a
entry compd 1
R1
R2
R3
R4
R5
I+
b
b
b
b
1
2
1k
1l
Me
n-hexyl
-(CH2)3-
H H
H H
H
H H
H H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
69
73
68
I2
I2
I2
I2
3
4
5
6
7
8
9
1m
1n
1o
1p
1q
1r
1s
c
Ph
68
d
CH2OAc
Me
cyclopropyl H
61
58
62
59
66
58
72
0
Py2IBF4
b
-O-CH2-O-
-O-CH2-O-
-O-CH2-O-
-O-CH2-O-
NO2
I2
I2
I2
I2
I2
I2
I2
I2
b
b
Bn
H
H
H H
b
isobutyl
Me
cyclopropyl H H
Me
b
10
11
12
13
1t
1u
1v
b
NO2
b
H OMe OMe
H H -CHdCH-CHdCH-
OMe
b
1w Me
0
In view of these results, we assume the reaction mechanism
to be working as shown in Scheme 3.12 At first the electrophilic
I+ activates the alkyne via a cyclic iodonium ion, opening the
way for the nucleophilic cyclization of the azide at C2′ (4).
When the intermediate 5 loses N2 and H+, the isoquinoline 2 is
formed. The proximity of the pyridinyl substituent to the reaction
center enables the basic nitrogen atom to coordinate to the
iodonium ion, thereby holding it close to C2′ and preventing
the azide from forming the new C2′-N bond. In the case of
the smallest substituent, R ) H, the cyclic iodonium ion 4 is
twisted, strengthening the C2′-I bond and preventing closure
of the nucleophilic ring. Therefore, the iodide anion in the
solution has enough time to add at the cationic center at C1′.
To prevent the formation of the bis(iodine) adduct 3,
iodonium sources with less nucleophilic counterions were tested.
While ICl led to a mixture of iodine- and chloride-substituted
isoquinolines and NBS gave no conversion, the Barluenga
reagent (Py2IBF4/HBF4) in CH2Cl2 at -78 °C for 1 h and NIS
in (CH2Cl)2 at 50 °C for 72 h gave the desired isoquinolines
2h-j without formation of the bis(iodine) adducts (entries
8-10). The Barluenga reagent afforded the products 2h-j in
55, 67, and 69% yield, while NIS gave the products in 42, 66,
and 62% yield, respectively.
Of course, we were more interested in the synthesis of even
more highly substituted isoquinolines; therefore, we used 1k
as a model substrate to optimize the synthesis of 1,3,4-
substituted isoquinolines. To our surprise, we had to change
the base from K3PO4 (used for the reactions in Table 1) to
NaHCO3 to achieve the best results, isolating the desired product
2k in 69% yield (Table 2, entry 1). As we could prove, the
substituents R1-R4 had only a very small influence on the yield
of the products. Irrespective of whether electron-rich, electron-
deficient, or electron-neutral substituents on the benzene ring
were used, the yields of the highly substituted products 2k-u
were in general between 60-70%. The reactions of 1l and 1m
having aliphatic substituents at R1 proceeded without problems
and gave yields of 73 and 68%, respectively (entries 2 and 3).
It should be noted that it was even possible to form the tricyclic
product 2m in 68% yield (entry 3).
a Isolated yield. b I2 (5 equiv), NaHCO3 (1 equiv), CH2Cl2 (0.1 M),
rt, 24 h, Ar. c Two side products were formed that could not be
separated out; the yield of 2n was determined by NMR. d Py2IBF4 (2
equiv), HBF4 in Et2O (2 equiv), CH2Cl2, -78 °C, 1 h, Ar.
However, problems were encountered with substrates 1n and
1o. In the case of the phenyl-substituted substrate 1n (entry 4),
two unidentified side products were formed that could not be
separated from the desired product. Thus, the yield of 68% for
the desired product 2n itself was not an isolated yield but was
determined by NMR analysis of the mixture. In the other
problematic case, substrate 1o was not stable toward basic
reaction conditions, and therefore, it was treated with the acidic
Barluenga reagent, forming the product 2o in 61% yield (entry
5). The Barluenga reagent was also tested for entries 3 and 7,
and we obtained the desired products in yields similar to those
achieved using I2. In general, a rather large excess of I2 (5 equiv)
was used in the cyclization; if we used 2 equiv of I2 in the
reaction of 1k, it took 3 days to complete the reaction. Use of
5 equiv of I2 was required in order to enhance the reaction speed.
For the substrates 1p-u having either electron-rich or
electron-deficient substituents on the benzene ring, we also tested
further substituents for R1 and found that in each case we were
able to form the desired product with a yield between 60 and
70% (entries 6-11). A general rule, it was not possible to
determine which substituent for R1 is more favored. It seems
that all substituents at R1 are equally tolerated, making this new
methodology very useful for further application. Unfortunately,
we did find some limitations with respect to substitution at the
C5 position of the aromatic ring (1v and 1w): substituents other
than R5 ) H were not tolerated (entries 12-13). It is most
probably the case that as a result of steric hindrance between
the substituent R5 and the iodonium ion of intermediate 4, C2′
is turned away from the azide. This prevents the closure of the
nucleophilic ring by the azide at C2′ that is necessary to form
the isoquinoline ring. For that reason, the products 2v and 2w
were not formed.
While we could show that the substituents R1-R4 had only
a small influence on the outcome of the reaction, we found that
in the case of the 3,4-di-RO substituents on the aromatic ring
(1p, 6a, and 6b), the electronic properties of the phenyl
(12) A similar reaction pathway as described takes place; see: Gorin, D. J.;
Davis, N. R.; Toste, F. D. J. Am. Chem. Soc. 2005, 127, 11260–11261.
9
15722 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Iodine-Mediated Electrophilic Cyclization
A R T I C L E S
Table 3. Influence of the Phenyl Substituent of the R Group
Table 4. Iodine-Mediated Cyclization of Five-Membered
Heterocyclic Derivatives
entry
starting material
R
product, yield (%)a,b
entry
compd 10
X
R1
R2
time (h)
yield of 11 (%)a,b
1
2
3
1p
6a
6b
Ph
p-tolyl
p-COOEt-Ph
2p, 58
7a, 93
7b, 11
1
2
3
4
5
6
7
10a
10b
10c
10d
10e
10f
N-Ts Ph
N-Ts Ph
N-Ts n-butyl
H
2.5
1
3
18
24
1.5
5.5
72
77
25
61
71
26
46
n-hexyl
H
H
H
a Isolated yield. b I2 (5 equiv), NaHCO3 (1 equiv), CH2Cl2 (0.1 M),
rt, 24 h, Ar.
O
S
S
S
Ph
Ph
Ph
n-butyl
Me
H
Scheme 4. Synthesis of 7,8-Substituted 1,6-Naphthylpyridine
10g
a Isolated yield. b I2 (5 equiv), NaHCO3 (1 equiv), CH3NO2, 100 °C.
than 2.5 h. The thiophene substrate 10e worked equally well,
forming the product 11e in 71% yield within 24 h. The furan
substrate 10d was also able to form the isoquinoline ring, but
only in 61% yield after 18 h. Different results were obtained
when we used an alkyl substituent for R2. While the pyrrole
10b (entry 2) exhibited even better performance than the parent
compound 10a, forming the desired product 11b in 77% yield
in less than 1 h, the thiophene analogue 10f gave 11f only in
an unacceptable yield of 26% (entry 6). Relatively small alkyl
substituents at R1 were not tolerated: both 10c and 10g (entries
3 and 7) gave low yields (25 and 46%, respectively).
To prove that our new method can be successfully applied
to the synthesis of more complicated substrates, we chose as
the target molecule norchelerythrine 12, which exhibits potent
pharmacological activities such as antitumor and antiviral
activities (Scheme 5).13
The retrosynthetic analysis of 12 is shown in Scheme 5. At
first, we break ring C between C11 and C11′. This gives us the
3,4-disubstituted isoquinoline 13, which most probably can be
synthesized through our new method. The key azide substrate
therefore has to be the diaryl acetylene 14, which can be derived
from the commercially available benzaldehydes 15 and 16.
Compound 17, which had been prepared previously,14 was
synthesized via Sonogashira coupling between 6-bromopiperonal
15 and TMS-acetylene with a yield of 69% (Scheme 6)
followed by reduction of the aldehyde and deprotection of the
acetylene in one step using NaBH4 in MeOH. Protection of the
resulting benzyl alcohol with TBDMSCl gave 17 in a very high
yield. Sonogashira coupling between 17 and 16 gave the
corresponding diaryl acetylene in 81% yield. The reduction of
the aldehyde with NaBH4 in MeOH/THF gave the benzyl
alcohol in 94% yield, and this was followed by the formation
of key substrate 14 in 91% yield using DPPA. Accordingly, 14
was synthesized from 15 in six steps in a total yield of 44%. In
addition, no purification by chromatography was necessary to
this point because all of the products could be purified by
crystallization.
substituents at the alkyne terminus (R) exerted a strong influence
on the yields of the isoquinolines (Table 3).
While the substrate 1p gave the isoquinoline 2p in 58% yield
(see Table 2, entry 6), the yield of the isoquinoline 7a increased
dramatically to 93% when the electron-rich substrate 6a was
used (Table 3, entry 2). This is the best yield among all of the
reactions examined by the present methodologies. In contrast,
when 6b with the π-electron-withdrawing ester was tested, the
yield of the product 7b dropped to only 11%. This indicates
that while the substituents R1-R4 at the benzylic and aromatic
positions can be chosen without restrictions, only the substituent
R at the alkyne seems to be important for the outcome of the
reaction. It should be noted that in the case of the simple
substrate 1c that has R1-R4 ) H and only a σ-electron-
withdrawing p-Ph-CF3 group at the alkynyl terminus, the
electronic property of R did not exert any influence on the
product yield (Table 1, entries 1-3, 1a-c). Its only effect is to
significantly increase the reaction time required for complete
conversion. We therefore attribute the high yield of product 2c
to the better stability of substrate 1c under the reaction conditions
applied.
Besides substituted isoquinolines, we were extremely inter-
ested in the synthesis of substituted bicyclic heteropyridine
derivatives, which might show interesting biological properties
(Scheme 4). As expected, unlike the substrate 1f, where the
nitrogen of pyridine was incorporated into the alkynyl terminus
R, the substrate 8 was smoothly converted into 1,6-naphthylpy-
ridine 9 in 92% yield without formation of the bis(iodine)
adduct. Perhaps the coordination of the nitrogen of the pyridine
ring of 8 to the iodonium ion of 4 facilitates the attack of the
azide nitrogen at the C2′ cationic center.
Nevertheless, additional optimizations of the reaction condi-
tions were necessary when the five-membered heterocyclic
derivatives 10 were used instead of the phenyl substrates 1. The
use of the original standard reaction conditions optimized for 1
resulted in only minor formation of the desired products 11,
but we soon found that this problem could be overcome through
the use of CH3NO2 at 100 °C instead of CH2Cl2 at room
temperature (Table 4). In the above experiments (Tables 1-3),
we have known that the most favored combination of the
substituents is Ph as R1 and H as R2. Actually, the pyrrole
substrate 10a gave the desired product 11a in 72% yield in less
The key reaction from 14 to the 3,4-disubstituted isoquinoline
18 proceeded perfectly smoothly under the previously optimized
(13) (a) Chen, J.-J.; Fang, H.-Y.; Duh, C.-Y.; Chen, I.-S. Planta Med. 2005,
5, 470–475. (b) For a recent synthesis of norchelerythrine, see:
Harayama, T.; Akamatsu, H.; Okamura, K.; Miyagoe, T.; Akiyama,
T.; Abe, H.; Takeuchi, Y. J. Chem. Soc., Perkin Trans. I 2001, 523–
528.
(14) Odedra, A.; Datta, S.; Liu, R.-S. J. Org. Chem. 2007, 72, 3289–3292.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15723

A R T I C L E S
Fischer et al.
Scheme 5. Retrosynthetic Analysis of Norchelerythrine 12
Scheme 6. Synthesis of the Key Intermediate 14a
Scheme 8. Formation of the Styrene Derivative 13a
a (a) Bu4NF·3H2O (1.5 equiv), THF, 0.5 h, 92%; (b) DMSO (4 equiv),
(COCl)2 (2 equiv), CH2Cl2, -78 °C, followed by benzyl alcohol, 0.5 h,
NEt3 (3 equiv), 1 h, aqueous workup only (!); (c) the benzaldehyde was
immediately taken up in THF and added to a Wittig solution [PPh3CH3Br
(10 equiv), KOtBu (10 equiv), THF], 0 °C, 14 h, dark, 87% over two steps.
Scheme 9. Photocyclization of 13 To Form the Ring C of 12a
a (a) 15, TMS-acetylene (1.3 equiv), PdCl2 (2.5 mol %), PPh3 (5 mol
%), CuCl (3.6 mol %), NEt3 (3 equiv), CH3CN, ∆, 16 h, 69%; (b) NaBH4
(0.75 equiv), MeOH, 0 °C, 3 days, 92%; (c) TBDMSCl (1.3 equiv),
imidazole (1.8 equiv), THF, 0 °C, 99%; (d) 16 (0.85 equiv), PdCl2 (5 mol
%), PPh3 (10 mol %), NEt3 (3 equiv), CH3CN, ∆, 16 h, 81%; (e) NaBH4
(0.85 equiv), MeOH/THF (1:1.25), 0 °C, 3 h, 94%; (f) (PhO)2PON3 (1.25
equiv), DBU (1.35 equiv), toluene, 16 h, 91%.
a (a) hν, NEt3 (2 equiv), CHCl3, 24 h, 57%.
Scheme 7. Iodine-Mediated Cyclization of 14a
considered using a light-induced radical cyclization/elimination
to form ring C (Scheme 9). We expected this to be possible
given the special instability of iodine compounds toward light.
To our surprise, we found no previous report of a similar
reaction using iodine as a light-sensitive group for the formation
of a phenyl ring. One report about the formation of a heterocycle
using a similar approach was later found.15
To test our photocyclization idea, we irradiated 13 in an NMR
tube in CDCl3. The NMR spectra taken after 8 h showed only
the desired natural product 12 without any of the styrene
derivative 13.16 Under optimized conditions with 2 equiv of
NEt3, we isolated a 10:1 mixture of the products 6-endo 12 and
5-exo 24 (see the Supporting Information), with an isolated yield
of 57% for the desired natural product 12; the two regioisomeric
products 12 and 24 could be separated by a column chroma-
tography. Therefore, we not only successfully applied the newly
developed method to form the highly substituted isoquinoline
12 but also found a new way to form a phenyl ring from the
iodoarylstyrene under mild conditions using the unique proper-
ties of the light-sensitive iodine moiety.
a (a) I2 (5 equiv), K3PO4 (5 equiv), CH2Cl2, 24 h, 94%.
reaction conditions without further modifications, giving the
desired product 18 in an almost quantitative yield of 94%
(Scheme 7).
To conclude the synthesis of the natural product, all that was
left was the formation of the last ring C. First, the TBDMS
group of 18 was deprotected with Bu4NF, forming the free
benzyl alcohol in 92% yield (Scheme 8). Afterward, the only
problematic step was encountered. Surprisingly, after oxidation
of the alcohol, the resulting benzaldehyde proved to be highly
unstable, giving black tar upon standing. After many trials, we
found that the problem could be avoided by use of the following
procedure: the unstable benzaldehyde obtained through Swern
oxidation was purified only by a quick aqueous workup and
immediately added to a Wittig ylide solution (10 equiv) to form
the stable styrene derivative 13 in 87% yield over two steps
(Scheme 8).
In conclusion, we have presented a general and flexible
approach to highly substituted isoquinoline building blocks,
which can then be further functionalized. Depending on the
To conclude the synthesis of 12, we examined Heck-type
cyclizations of 13 for many trials, but only the corresponding
5-exo-product 24 was obtained. After many attempts, we finally
(15) Layman, W. J., Jr.; Greenwood, T. D.; Downey, A. L.; Wolfe, J. F. J.
Org. Chem. 2005, 70, 9147–9155.
9
15724 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Iodine-Mediated Electrophilic Cyclization
A R T I C L E S
Barluenga reagent (Py2IBF4) (2 equiv) under argon was dissolved
in CH2Cl2 (0.16 M) and cooled to -78 °C, and HBF4 (54% in
Et2O) (2 equiv) was added. The solution was transferred to azide
1o (25 mg, 82 µmol) dissolved in CH2Cl2 (1.33 M) at -78 °C.
After 1 h, the reaction was quenched with Na2S2O3 solution. The
product was extracted with ethyl acetate, dried over MgSO4, and
purified by column chromatography (SiO2) to yield 20.1 mg of 2o
(61%).
requirements of the substrates employed, acidic, basic, or neutral
reaction conditions can be used, making this method compatible
with labile and highly functionalized molecules. Because the
reaction conditions are mild, a wide variety of substituents are
tolerated. The free choice of substituents along the benzyl area
and the possible application to the five-membered heteroaromatic
substrates especially make this method useful for further use in
the synthesis of small molecules, natural product synthesis, and
technical applications.
Supporting Information Available: Experimental procedures
and spectroscopic data for all of the new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Experimental Section
Azide 1k (7.22 mg, 0.29 µmol) was dissolved in CH2Cl2 (0.1
M), and NaHCO3 (1 equiv) was added. Iodine (5 equiv) was added,
and the solution was stirred in the dark for 24 h. After complete
conversion, as tested by TLC, the reaction was quenched with
Na2S2O3 solution. The product was extracted with ethyl acetate,
dried over MgSO4, and purified by column chromatography (SiO2)
to yield 6.9 mg of 2k (69%).
JA805326F
(16) Tousek, J.; Dostal, J.; Marek, R. J. Mol. Struct. 2004, 689, 115–120.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15725