Aryl-Allene Cyclization via a Hg(OTf)2-Catalytic Pathway

Article abstract of DOI:10.1021/acs.orglett.6b01144

Hg(OTf)₂-catalyzed aryl-allene cyclization accompanied by formation of a quaternary carbon center has been realized. Deuterium-labeling experiments and computational modeling were used to propose a novel catalytic pathway involving direct H-transfer from the aromatic ring to the vinyl mercury moiety followed by mercury 1,2-migration.

Full text of DOI:10.1021/acs.orglett.6b01144

Letter
pubs.acs.org/OrgLett
Aryl−Allene Cyclization via a Hg(OTf)2‑Catalytic Pathway
Hirofumi Yamamoto,*,† Maho Ueda,† Naoto Yamasaki,† Akiyoshi Fujii,† Ikuo Sasaki,† Kazunobu Igawa,‡
Yusuke Kasai,† Hiroshi Imagawa,† and Mugio Nishizawa†
†Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770-8514, Japan
‡Institute for Materials Chemistry and Engineering, Kyushu University, 6-1 Kasuga-koen, Kasuga, Fukuoka 816-8580, Japan
S
* Supporting Information
ABSTRACT: Hg(OTf)2-catalyzed aryl−allene cyclization accompanied by formation of a quaternary carbon center has been
realized. Deuterium-labeling experiments and computational modeling were used to propose a novel catalytic pathway involving
direct H-transfer from the aromatic ring to the vinyl mercury moiety followed by mercury 1,2-migration.
he use of allene functionality to construct C−C bonds is a
aryl−diene cyclization,6d have been developed on the basis of
Twell-known and conventional method in the field of strong affinity of the mercury salt for C−C multiple bonds. In
synthetic chemistry.1 Recently, metal-catalyzed allene-cycliza-
connection with this previous work we examined the
corresponding Friedel−Crafts type 6-exo-trig cyclization of
aryl 1,1-disubstituted allene derivative 1d with the Hg(OTf)2
catalyst, which leads to isochromane derivative 2d in good yield
under mild conditions. A 1,1-disubstituted allene derivative 5
was initially designed as a substrate, which was reacted with 5
mol % of Hg(OTf)2 in CH3CN at room temperature for 2 h
(entry 1, Table 1). While a considerable amount of 5 was
recovered, a sole product obtained in 36% yield was analyzed to
be the desired isochromane derivative 6. Although the yield was
unsatisfactory, the result indicated that Hg(OTf)2 acted as a
catalyst to promote the desired cyclization.
In order to make the reaction synthetically valuable, reaction
conditions were optimized by investigating the effect of utilizing
different solvents. Using CH2Cl2, the yield of 6 was improved
to 75% (entry 2), whereas toluene or Et2O did not give any
products (entries 3 and 4). By contrast, CH3NO2 showed
remarkable reactivity, giving rise to 6 in 95% yield within 1.5 h
(entry 5). The comparative experiment with HOTf failed to
give the desired product (entry 6). Using HOTf as an additive
did not significantly change the yield of 6 (entry 7 vs 8). In light
of our previous success in the cycloisomerization of 2-(4-
pentynyl)furan with a combination of Hg(OAc)2 and Sc-
(OTf)3,7 the catalytic performance of other mercuric salts and
additives including Hg(OAc)2/Sc(OTf)3, HgCl2/AgOTf,
HgCl2/AgBF4, and HgCl2/AgNTf2 was examined (entries 9−
12). However, all showed inferior catalytic activities compared
to that of the Hg(OTf)2 catalyst. The Hg(OAc)2/Sc(OTf)3
combination also afforded the desired product in moderate
yield (entry 9). Platinum and gold catalysts (entries 13 and 15)
tions, such as exo-trig, exo-dig, endo-dig, and endo-trig type
cyclizations, have been widely studied with the aim of
synthesizing useful carbon frameworks.2 Indeed, this strategy
plays an important role in the synthesis of many natural
products and bioactive compounds.1d,3 Enantioselective cycliza-
tions have also been developed, which facilitate easy access to
some chiral inductions.4 Most of these reactions, however, used
naked 1a, 1,3-disubstituted 1b, and 1,3,3-trisubstituted allenes
1c as substrates, while the cyclization of 1,1-disubstituted
allenes 1d, especially exo-trig cyclization of 1d by the formation
of a quaternary carbon center,3f has not been widely reported
(eq 1, Scheme 1). We have contributed to this area through the
development of a Hg(OTf)2-catalyzed protocol and mecha-
nistic studies of these reactions.
In 2003 we reported that Hg(OTf)2 showed high catalytic
activity for the carbocyclization of ω-arylalkyne 3 (eq 2).5 Since
that time, various Hg(OTf)2-catalyzed reactions,6 including
ene−yne cyclization,6b aryl−allyl alcohol cyclization,6c and
Scheme 1. Allene 6-exo-trig Cyclizations and Hg(OTf)2-
Catalyzed Aryl−Yne Cyclizations
Received: April 20, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b01144
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Table 1. Screening of Catalysts and Solvents
Table 2. Assessment of Hg(OTf)2-Catalyzed Cyclization in
MeNO2
a
b
yield (%)
entry
catalyst
additive
solvent
5
6
1
2
3
4
Hg(OTf)2
Hg(OTf)2
Hg(OTf)2
Hg(OTf)2
Hg(OTf)2
−
Hg(OTf)2
Hg(OTf)2
Hg(OAc)2
HgCl2
−
−
−
−
−
HOTf
MeCN
CH2Cl2
PhMe
54
36
75
0
trace
76
81
0
Et2O
0
c
5
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeOH
MeOH
CH2Cl2
CH2Cl2
CH2Cl2
PhMe
95
0
6
83
39
30
0
d
7
d
−
55
51
83
0
8
HOTf
Sc(OTf)3
AgOTf
AgBF4
AgN(Tf)2
−
9
10
11
12
13
14
15
16
17
18
84
98
92
64
67
84
50
56
83
76
42
69
HgCl2
0
HgCl2
0
PtCl2
0
PtCl2
AgOTf
0
AuCl3
−
0
AuCl3
AgSbF6
AgOTf
AgOTf
−
0
AuCl3
0
AuCl3
0
e
19
PPh3AuCl
PPh3AuCl
PPh3AuCl
PhMe
0
e
20
AgOTf
PhMe
22
10
e
21
AgBF4
PhMe
a
b
Reactions conducted with 0.02 mmol of substrate 5. Isolated yield.
c
Reaction conducted with 0.1 mmol of substrate 5, which was
d
complete within 1.5 h. 2 mol % of catalyst was used for the reaction.
e
Reaction carried out for 20 h.
in addition to other well-known combinations such as PtCl2/
AgOTf, AuCl3/AgSbF6, and AuCl3/AgOTf (entries 14 and 16−
18) were inefficient catalysts for the reaction. Under gold(I)
catalytic conditions with PPh3AuCl/AgOTf and PPh3AuCl/
AgBF4, the desired cyclization proceeded, but the reaction rates
were low i.e., 22% yield of 6 after 20 h (entries 20 and 21).
Thus, from among the several solvents and catalysts screened
in our study, Hg(OTf)2 in CH3NO2 was shown to be the most
suitable combination. Hence, the scope of the reaction was
examined using the optimized conditions. As shown in Table 2,
the corresponding cyclic products were obtained in excellent to
good yield in many cases. The reaction of mono-m-methoxy
substituted 7 with 5 mol % of the mercury catalyst proceeded
smoothly at room temperature affording 8 at the p-position for
the methoxy group in 84% yield with complete regioselectivity
(entry 1). Using n-butyl substituted allene 9 instead of
methylallene as the substrate, the desired 10 with a quaternary
carbon center was afforded in 78% yield (entry 2). Indole
derivative 11 also underwent the desired reaction (entry 3).
The cyclizations of dimethyl malonate derivative 13 as well as
sulfonamide substrates 15a−e gave the corresponding products
14 and 16a−e in good yield, respectively (entries 4 and 5).
However, it was found that the polar functional group at the γ-
position affected the course of the reaction. Indeed, the
cyclization of 17a, including p-toluenesulfonyl (tosyl) amide at
the γ-position, furnished no product even after 2 days (entry 6).
The reactivity of ether 17b also decreased in comparison with
that of 5, giving 18b in 23% yield after 24 h.
a
b
See Supporting Information for full experimental details. These
c
reactions were conducted with 0.05−0.1 mmol of substrate. Isolated
yield. Ts = p-toluenesulfonyl. Starting substrate was mainly
recovered.
d
e
Because of the characteristics and size of the sulfonyl group
at the β-position of 15b−e, there was no difference in reactivity.
Thus, 17a and 17b probably formed 21 in which Hg(OTf)2
interacted with the γ-nitrogen atom (or γ-oxygen atom) as well
as the allene moiety to define the location of the reactive site
for the phenyl group as being set apart from the allene moiety.
In contrast, the β-functional group produces a phenyl group
close to the allene moiety through the formation of 22a.
Therefore, 19, possessing a nonligating methyl substituent at
the problem γ-position, reacted smoothly via 22b giving 20 as
the sole product in good yield with complete diastereoselec-
tivity.8
Considering the nature of the mercury salt reagent, these
catalytic reactions probably begin with Hg(OTf)2 activating the
B
DOI: 10.1021/acs.orglett.6b01144
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
internal double bond of the allene group 23-A, rather than the
external double bond, due to formation of a stable tertiary
carbenium intermediate species 23-B (Scheme 2). 23-B
insight into the course of the catalytic cycle. The reaction of 5-
H with Hg(OTf)2 in the presence of a stoichiometric amount
of DOTf resulted in the formation of normal 6-H, where
deuterium was not incorporated exclusively at the vinyl moiety
(eq 1). This result suggested that if 5 mol % of HOTf was
generated in situ through aromatization (step A), it did not
promote cleavage (step B) of the vinyl-Hg bond.9 Hence,
deuterium-labeled 5-D, which could be prepared in high purity
from 3,5-dihydroxybenzoic acid in five steps,10 was used as a
substrate. The reaction of 5-D with 5 mol % of Hg(OTf)2 in
CH3NO2 resulted in the production of 32% D-enriched 6-D
(eq 2).11 This observation established that the deuterium atom
originated from the aromatic ring and is directly used in the
cleavage of the vinyl−Hg bond of 23-C. It is known that a β-
mercury atom on a carbocation exerts a stabilizing effect (i.e.,
the so-called β-mercury effect13), which would assist in the
observed D-transfer reaction. DFT calculations using a
simplified methoxymercury model were carried out to obtain
further data regarding the reaction mechanism (Figure 1). 25-C
represents a likely cationic intermediate formed immediately
upon cyclization and is equivalent to 23-C in Scheme 2. The
lowest energy pathway of the H-transfer reaction of 25-C via a
sole first-order saddle point was estimated using the B3LYP/
CEP-121G method. The reaction proceeds through a transition
structure 25-D with an activation energy Ea of 19.0 kcal/mol,
which involves concomitant aryl carbon−hydrogen bond
cleavage (Caryl−H; 1.42 Å) and transferal of hydrogen to
vinyl carbon (H−Cvinyl; 1.51 Å). Eventually, 25-D shifts to the
secondary carbenium 25-G via mercury 1,2-migration, which is
computed by the IRC calculation as shown by the reaction path
geometries of 25-F12 and 25-E. The calculations revealed that
the H-transfer reaction of 25-C could proceed with a low
activation barrier and suggested that Hg(OTf)2 forms from 25-
G without generating HOTf. This suggestion is consistent with
the experimental finding that the extrinsic addition of HOTf
had no influence on the catalytic activity (entry 8, Table 1).
The mercury 1,2-migration (25-E → 25-G) is reasonable in
terms of cationic stability. The natural bond orbitals (NBO)
analysis to calculate the distribution of electron density clearly
recognized the existence of the secondary carbenium species in
the 25-G state.
Scheme 2. Mechanistic Studies of Hg(OTf)2-Catalyzed
Cyclization
a
b
20 times the amount of DOTf to generating HOTf. Starting
material 5-H was recovered in 60% yield.
In summary, we have demonstrated a novel cyclization of aryl
1,1-disubstituted allenes with formation of a quaternary carbon
center catalyzed by Hg(OTf)2 in CH3NO2. The experimental
and computational results indicate an interesting mechanistic
pathway, which involves novel direct H-transfer from the
promotes cyclization followed by the generation of organo-
Hg2+ intermediate 23-C. However, the likely mechanism for the
conversion of 23-C to 24 was unclear. Thus, several deuterium-
labeling experiments were performed to gain mechanistic
Figure 1. Transition states and energy changes for the H-transfer pathway from optimized structure 25-C. The white, gray, blue, and red in the
structures indicate hydrogen, carbon, mercury, and oxygen atoms, respectively. Energy is in kcal/mol with zero-point vibrational energy correction.
The NBO charges for carbon and mercury atoms in 25-E, -F, and -G are displayed.
C
DOI: 10.1021/acs.orglett.6b01144
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(3) Selected reports on synthetic studies using allene reactions:
(a) Gerstner, N. C.; Adams, C. S.; Grigg, R. D.; Tretbar, M.; Rigoli, J.
W.; Schomaker, J. M. Org. Lett. 2016, 18, 284−287. (b) Painter, T. O.;
Wang, L.; Majumder, S.; Xie, X.-Q.; Brummond, K. M. ACS Comb. Sci.
2011, 13, 166−174. (c) Zeldin, R. M.; Toste, F. D. Chem. Sci. 2011, 2,
1706−1709. (d) Bates, R. W.; Lu, Y. Org. Lett. 2010, 12, 3938−3941.
(e) Bates, R. W.; Dewey, M. R. Org. Lett. 2009, 11, 3706−3708.
(f) Liu, Z.; Wasmuth, A. S.; Nelson, S. G. J. Am. Chem. Soc. 2006, 128,
10352−10353.
aromatic ring to the vinyl mercury moiety. We also suggest that
Hg(OTf)2 forms as a catalytic complement from a cationic
vinyl mercury intermediate 25-G without generating HOTf.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.6b01144.
(4) Selected recent reports of the asymmetric cyclization of allenes:
(a) Aillard, P.; Retailleau, P.; Voituriez, A.; Marinetti, A. Chem. - Eur. J.
2015, 21, 11989−11993. (b) Jiang, T.; Bartholomeyzik, T.; Mazuela,
Detailed experimental procedures and characterization of
all new compounds (PDF)
J.; Willersinn, J.; Backvall, J.-E. Angew. Chem., Int. Ed. 2015, 54, 6024−
̈
6027. (c) Gonzal
(d) Liu, C.; Widenhoefer, R. A. Org. Lett. 2007, 9, 1935−1938.
(e) Tarselli, M. A.; Chianese, A. R.; Lee, S. J.; Gagne, M. R. Angew.
́
ez, A. Z.; Toste, F. D. Org. Lett. 2010, 12, 200−203.
AUTHOR INFORMATION
Corresponding Author
■
́
Chem., Int. Ed. 2007, 46, 6670−6673.
(5) Nishizawa, M.; Takao, H.; Yadav, V. K.; Imagawa, H.; Sugihara,
T. Org. Lett. 2003, 5, 4563−4565.
*E-mail: hirofumi@ph.bunri-u.ac.jp.
Notes
(6) (a) Nishizawa, M.; Imagawa, H.; Yamamoto, H. Org. Biomol.
Chem. 2010, 8, 511−521. (b) Nishizawa, M.; Yadav, V. K.;
Skwarczynski, M.; Takao, H.; Imagawa, H.; Sugihara, T. Org. Lett.
2003, 5, 1609−1611. (c) Namba, K.; Yamamoto, H.; Sasaki, I.; Mori,
K.; Imagawa, H.; Nishizawa, M. Org. Lett. 2008, 10, 1767−1770.
(d) Yamamoto, H.; Shiomi, S.; Odate, D.; Sasaki, I.; Namba, K.;
Imagawa, H.; Nishizawa, M. Chem. Lett. 2010, 39, 830−831.
(7) Yamamoto, H.; Sasaki, I.; Imagawa, H.; Nishizawa, M. Org. Lett.
2007, 9, 1399−1402.
(8) The relative configuration of 20 was identified through NOESY
experiments. The result strongly supports our hypothesis.
(9) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic
Molecules; Wiley: New York, 1980.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to Prof. Dr. Katsuhiko Tomooka of Kyushu
University for kind advice on computational chemistry. This
study was financially supported by a Grant-in-Aid (No.
23790034) from the MEXT (Ministry of Education, Culture,
Sports, Science, and Technology) of the Japanese Government.
REFERENCES
■
(1) For recent reviews on allene chemistry: (a) Soriano, E.;
(10) See Supporting Information with regard to the preparation of
deuterium-labeled 5-D.
Fernan
́
dez, I. Chem. Soc. Rev. 2014, 43, 3041−3105. (b) Lechel, T.;
Pfrengle, F.; Reissig, H.-U.; Zimmer, R. ChemCatChem 2013, 5, 2100−
2130. (c) Yu, S.; Ma, S. Angew. Chem., Int. Ed. 2012, 51, 3074−3112.
(d) Rivera-Fuentes, P.; Diederich, F. Angew. Chem., Int. Ed. 2012, 51,
2818−2828. (e) Krause, N.; Winter, C. Chem. Rev. 2011, 111, 1994−
2009. (f) Alcaide, B.; Almendros, P. Adv. Synth. Catal. 2011, 353,
2561−2576. (g) Alcaide, B.; Almendros, P.; Aragoncillo, C. Chem. Soc.
Rev. 2010, 39, 783−816. (h) Brasholz, M.; Reissig, H.-U.; Zimmer, R.
Acc. Chem. Res. 2009, 42, 45−56. (i) Widenhoefer, R. A. Chem. - Eur. J.
2008, 14, 5382−5391. (j) Ma, S. Chem. Rev. 2005, 105, 2829−2872.
(k) Modern Allene Chemistry; Krause, N., Hashmi, A. S. K., Eds.;
Wiley-VCH: Weinheim, 2004.
(11) 5-D is moisture sensitive and gradually changes to 5-H. Thus,
68% of the protonated 6-D was probably generated from the
cyclization of 5-D whose o-deuterium converted to protons derived
from the water that could not be completely removed from the
MeNO2 solvent. MeNO2 might also act as a competitive proton
source.
(12) The NBO analysis suggests that the cationic charges of the
obtained geometries 25-E, -F, and -G are delocalized by the
coordination of MeOHg+ to vinyl double bond, in which the
electronic occupancy of the π-orbitals are ca. 1.70 (typically 1.95−
2.00).
(2) Selected recent reports of the carbocyclization of allenes:
(13) (a) Lambert, J. B.; Emblidge, R. W. J. Phys. Org. Chem. 1993, 6,
555−560. (b) Schaleger, L. L.; Turner, M. A.; Chamberlin, T. C.;
Kreevoy, M. M. J. Org. Chem. 1962, 27, 3421−3425.
(a) Haraburda, E.; Lledo,
́
A.; Roglans, A.; Pla-Quintana, A. Org. Lett.
e, G.;
2015, 17, 2882−2885. (b) Ondet, P.; Joffrin, A.; Diaf, I.; Lemier
̀
Dunach, E. Org. Lett. 2015, 17, 1002−1005. (c) Alcaide, B.;
̃
Almendros, P.; Carrascosa, R.; Casarrubios, L.; Soriano, E. Chem. -
Eur. J. 2015, 21, 2200−2213. (d) Qiu, Y.; Fu, C.; Zhang, X.; Ma, S.
Chem. - Eur. J. 2014, 20, 10314−10322. (e) Xia, X.-F.; Wang, Y.-Q.;
Zhang, L.-L.; Song, X.-R.; Liu, X.-Y.; Liang, Y.-M. Chem. - Eur. J. 2014,
20, 5087−5091. (f) Pitaval, A.; Leboeuf, D.; Ceccon, J.; Echavarren, A.
M. Org. Lett. 2013, 15, 4580−4583. (g) Rigoli, J. W.; Weatherly, C. D.;
Vo, B. T.; Neale, S.; Meis, A. R.; Schomaker, J. M. Org. Lett. 2013, 15,
290−293. (h) Amako, Y.; Hori, H.; Arai, S.; Nishida, A. J. Org. Chem.
2013, 78, 10763−10775. (i) Liu, Y.; Zhu, J.; Qian, J.; Xu, Z. J. Org.
Chem. 2012, 77, 5411−5417. (j) Persson, A. K. Å.; Backvall, J.-E.
̈
Angew. Chem., Int. Ed. 2010, 49, 4624−4627. (k) Mo, J.; Lee, P. H.
Org. Lett. 2010, 12, 2570−2573. (l) Kong, W.; Fu, C.; Ma, S. Eur. J.
́
Org. Chem. 2010, 2010, 6545−6555. (m) Mauleon, P.; Zeldin, R. M.;
Gonzal
(n) Kong, W.; Fu, C.; Ma, S. Chem. Commun. 2009, 4572−4574.
(o) Weber, D.; Tarselli, M. A.; Gagne, M. R. Angew. Chem., Int. Ed.
2009, 48, 5733−5736. (p) Weber, D.; Gagne, M. R. Org. Lett. 2009,
11, 4962−4965. (q) Tarselli, M. A.; Gagne, M. R. J. Org. Chem. 2008,
́
ez, A. Z.; Toste, F. D. J. Am. Chem. Soc. 2009, 131, 6348−6349.
́
́
́
73, 2439−2441. (r) Watanabe, T.; Oishi, S.; Fujii, N.; Ohno, H. Org.
Lett. 2007, 9, 4821−4824. (t) Molander, G. A.; Cormier, E. P. J. Org.
Chem. 2005, 70, 2622−2626.
D
DOI: 10.1021/acs.orglett.6b01144
Org. Lett. XXXX, XXX, XXX−XXX