Lewis acid-catalyzed or base-promoted regioselective cycloisomerization of N-imidoyl-o-alkynylanilines for synthesis of N-imidoyl-(1-H)-indoles and 4-alkylidene-3,4-dihydroquinazolines

Article abstract of DOI:10.1002/adsc.201401186

Product selectivity control for the synthesis of imidoylindoles and 4-alkylidenedihydroquinazolines from N-imidoyl-o-alkynylanilines via silver triflate-catalyzed cycloisomerization or tetrabutylammonium fluoride-promoted cyclization is described. The pro

Full text of DOI:10.1002/adsc.201401186

FULL PAPERS
DOI: 10.1002/adsc.201401186
Lewis Acid-Catalyzed or Base-Promoted Regioselective Cycloiso-
merization of N-Imidoyl-o-alkynylanilines for Synthesis of N-Imi-
doyl-(1H)-indoles and 4-Alkylidene-3,4-dihydroquinazolines
Takashi Otani,a,b,* Xue Jiang,a Kinryo Cho,a Rino Araki,a Noriki Kutsumura,a,c
and Takao Saitoa,*
a
Department of Chemistry, Faculty of Science, Tokyo University of Science, Kagurazaka, Shinjuku-ku, Tokyo 162-8601,
Japan
Fax : (+81)-3-5261-4631; phone: (+81)-3-3262-4271; e-mail: tsaito@rs.kagu.tus.ac.jp
Research Center for Chirality, Research Institute for Science & Technology, Tokyo University of Science, Kagurazaka,
b
Shinjuku, Tokyo 162-8601, Japan
Fax : (+81)-3-5261-4631; phone: (+81)-3-3262-4271; e-mail: totani@rs.tus.ac.jp
International Institute for Integrative Sleep Medicine (WPI-IIIS), University of Tsukuba, 1-1-1 Tennodai, Tsukuba,
c
Ibaraki 305-8577, Japan
Received: December 25, 2014; Revised: February 16, 2015; Published online: && &&, 0000
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201401186.
Abstract: Product selectivity control for the synthesis lyst/promoter used, and on the substituents on the
of imidoylindoles and 4-alkylidenedihydroquinazo- alkyne and amidine functions of the substrates.
lines from N-imidoyl-o-alkynylanilines via silver tri-
flate-catalyzed cycloisomerization or tetrabutylam-
monium fluoride-promoted cyclization is described. Keywords: cyclization; heterocycles; indoles; quina-
The product selectivity depends mainly on the cata- zolines; selectivity
Introduction
both multiple bond-forming reaction sites and ambi-
dent nucleophilic centers in the same molecule, and
The development of methodologies for highly selec- hence control of regioselectivity and stereoselectivity
tive synthesis of different products from the same if present in the processes is very important for the ef-
compounds by simple modification of catalysts, pro- fective synthesis of desired products. In this context,
moters, additives, and/or reaction conditions is we recently reported palladium-catalyzed highly
a highly attractive and challenging task.[1] Among regio- and stereoselective synthesis of (Z)-4-alkyli-
recent atom-economical methods for the synthesis of dene-4H-3,1-benzoxazines from N-acyl-2-alkynylani-
carbo- and heterocycles, transition metal- or Lewis lines (Scheme 1, path b).[3] This benzoxazine forma-
acid- or base-mediated transformations of unsaturated tion via the 6-exo-dig mode of cyclization by O-attack
species such as alkynes and alkenes reacting with an is still very rare and scarcely known,[4,5] but this cycli-
available nucleophilic function are powerful and at- zation bias was effectively controlled by use of
tractive tools.[2] In these cases, substrates often bear Pd(OAc)2 catalyst[3,6] together with acetic acid in pref-
Scheme 1. Catalyst-dependent, highly regio- and stereoselective cyclization modes.
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
ÞÞ
These are not the final page numbers!

FULL PAPERS
Takashi Otani et al.
Scheme 2. Four possible cycloisomerization modes of N-imidoyl-2-alkynylanilines I.
erence to the 5-endo-dig mode of cyclization by N-
attack leading to the usual formation of indoles
(Scheme 1, path a), for which a vast number of exam-
ples using a variety of catalysts has been reported.[6–8]
In continuation of our interest in exploring product
selectivity associated with such cyclization modes, we
aimed to examine experimentally Lewis acid- or base-
promoted cycloisomerization of N-imidoyl-2-alkynyla-
nilines (I) bearing various common substituents, such
as alkyl and aryl groups at R1, R2, and R3.[9] Scheme 2
illustrates four possible cyclization modes of I leading Scheme 3. (i) Ph3P/C2Cl6, CH2Cl2, r.t.; (ii) R2-NCO, CH2Cl2,
r.t.; (iii) R3-MgBr, THF, 08C ! r.t.
to indoles (II) via the 5-endo-dig mode (path a), 3,4-
dihydroquinazolines (III) via the 6-exo-dig mode
(path b), benzazetidines (IV) via the 4-exo-dig
mode[10] (path c), and 3H-benzo[d][1,3]diazepines (V) are shown in Table 1. From the data, the following
via the 7-endo-dig mode (path d). What controls can be deduced in general: (1) In all of the cases, cy-
these cycloisomerization modes? In this paper we cloisomerization products were the indole derivatives
report our results on the product selectivity for cycloi- (5a, 5b, and 5j) without any other cyclized prod-
somerizations of the simple substrate system I.
uct(s), such as quinazolines 6; (2) The relative reactiv-
ity of 1 for the reaction in the presence of the same
Lewis acid depended mainly on the properties and
positions of the substituents (R1, R2, and R3); (3)
silver triflate (AgOTf) was among the most effective
catalysts for the cycloisomerization leading to N-imi-
doylindoles 5 (entries 6, 13, and 19).
Results and Discussion
Preparation of N-imidoyl-2-alkynylanilines 1
N-imidoyl-2-alkynylanilines 1 were prepared accord-
With the best Lewis acid, AgOTf, in hand for the
ing to the reactions from 2-alkynylanilines 2 as illus- representative model substrates (1a, 1b, and 1j), we
trated in Scheme 3 (see Supporting Information).
performed the reaction in the presence of 10 mol%
AgOTf using substrates 1 with a variety of substitu-
ents in combinations of aryl and alkyl groups at R1,
R2, and R3. The results are shown in Table 2. The data
suggest the following: (1) on the whole, indoles 5
were produced highly regioselectively, and only in the
Lewis acid-catalyzed cyclization of N-imidoyl-2-
alkynylanilines (1)
First, screening of Lewis acids was performed using cases of 1c, 1e and 1h were minor cyclization prod-
substrates 1a (R1 =4-Tol, R2 =R3 =Ph), 1b (R1 =4- ucts 6c, 6e and 6h observed (6c: 19%, 6e: 44% and
Tol, R2 =nPr, R3 =Ph), and 1j (R1 =tBu, R2 =nPr, 6h: 46% yield) (entries 3, 5 and 8). Comparison of
R3 =Ph) as model compounds. The reactions were entries 6–8 suggests that electronic effects of the sub-
carried out in the presence of various 10 mol% cata- stituent on R1 also affect the selectivity for the cycli-
lysts under the reaction conditions (solvent, tempera- zation; (2) the indoles 5 would be formed via the con-
ture, and reaction time) listed. Representative results ceivable Ag–alkyne complex species (Figure 1, A and
2
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Lewis Acid-Catalyzed or Base-Promoted Regioselective Cycloisomerization
B; LUMO-umpolung[11]) in the stereoelectronically
favorable 5-endo-dig cyclization mode (path a). The
bent structure of the complex B would be advanta-
geous for the 5-endo mode of cyclization with proxim-
ity for the N1-nucleophilic attack (path a).
Table 1. Screening of Lewis acids for cycloisomerization.
Interestingly, the 6-exo-dig mode of cyclization was
observed in the AgOTf-catalyzed reaction when sub-
strates 1y–b (R1 =H) were employed (Table 3). Ac-
tually, 4-methylenequinazolines 6y–b were exclusively
formed. This observation is consistent with the view
that substrate 1 with R1 =H prefers the 6-exo-dig
mode of cyclization via p-complex A and s-complex
C with an exo-oriented form easily leading to 6 (path
b), because of the less steric hindrance with H and
the greater stabilization energy gain resulting from
the benzylic carbocation present in the structure C[12]
rather than the other terminal less stabilized carbocat-
ion with R1 =H.
Entry Lewis acid
Solvent
Temp./time [h] Yield
For 1a: R1 =4-Tol, R2 =R3 =Ph
1
2
3
4
5
6
ZnCl2
In(OTf)3
Cu(OTf)2
toluene
(CH2Cl)2 reflux/19
CH2Cl2
r.t./90
87
57[a]
27[a]
64
r.t./27
reflux/2
r.t./7
[Rh(cod)Cl]2 CH2Cl2
Pd(OAc)2
AgOTf
CH2Cl2
CH2Cl2
87
r.t./0.7
99
For 1b: R1 =4-Tol, R2 =nPr, R3 =Ph
7
8
9
Me3SiOTf
In(OTf)3
Yb(OTf)3
Cu(OTf)2
[Rh(cod)Cl]2 CH2Cl2
Pd(OAc)2
AgOTf
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
r.t./72
r.t./26
r.t./70
r.t./96
r.t./0.5
r.t./1
57
44
14[a]
43
It is worth comparing our results with those of
other related work.[9] Medio-Simꢁn et al.[9a] reported
a mechanistic study on the cyclization that cationic
[IPrAu]SbF6 catalyst promoted 6-exo-dig mode cycli-
10
11
12
13
82
53
CH2Cl2
CH2Cl2
r.t./0.5
93
For 1j: R1 =tBu, R2 =nPr, R3 =Ph
zation
to
form
quinazolin-2-one,
whereas
14
15
16
17
18
19
ZnCl2
In(OTf)3
Cu(OTf)2
[Rh(cod)Cl]2 CH2Cl2
AgOTf
AgOTf
CH2Cl2
toluene
CH2Cl2
r.t./72
r.t./24
r.t./32
reflux/12
r.t./24
NR[b]
NR[b]
NR[b]
NR[b]
65
[tBu3PAu]SbF6 catalyst favored indole formation. Un-
fortunately, the results arose from the reaction of only
one terminal alkyne substrate (R1 =H) bearing a urea
nucleophile without any variation of substituents
(Scheme 4, eq. (1)).[9a] In contrast, gold(AuCl)-cata-
lyzed reactions of pyrazinone-bearing substrates af-
forded indoles, whereas the silver(AgOTf)-catalyzed
reaction preferentially produced quinazoline deriva-
tives (eq. (2)).[9b] Similar to the above AuCl-catalyzed
reaction in eq. (2), the gold(NaAuCl4)-catalyzed reac-
CH2Cl2
(CH2Cl)2 reflux/3
94
[a]
[b]
Considerable amounts of 1 were recovered.
Almost no reaction.
Table 2. Silver triflate-catalyzed cycloisomerization of 1 to give N-imidoylindoles 5.
Entry
1, 5
R1
R2
R3
Solvent
Temp./Time [h]
Yield [%]
1
2
3
4
5
6
7
8
a
b
c
d
e
f
g
h
i
j
k
l
m
n
4-Tol
4-Tol
4-Tol
4-Tol
Ph
Ph
nPr
Ph
nPr
Ph
Ph
Ph
Ph
Ph
nPr
Ph
nPr
Ph
Ph
Ph
Ph
Et
Et
Et
Ph
Ph
Ph
Ph
Ph
Et
Et
Ph
Ph
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
(CH2Cl)2
(CH2Cl)2
(CH2Cl)2
(CH2Cl)2
CH2Cl2
CH2Cl2
r.t./0.7
r.t./0.5
r.t./2
r.t./0.5
r.t./3
r.t./1
r.t./0.5
r.t./2
reflux/2
reflux/3
reflux/7.5
reflux/9
r.t./0.5
r.t./2
99
93
68[a]
86
49[a]
98
Ph
4-MeOC6H4
4-CF3C6H4
tBu
tBu
tBu
tBu
nPr
iPr
83
53[a]
78
9
10
11
12
13
14
94
67
65
76
97
[a]
Quinazolines 6c (E), 6e (E) and 6h (E) were formed in 19%, 44% and 46% yields, respectively.
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3
ÞÞ
These are not the final page numbers!

FULL PAPERS
Takashi Otani et al.
Figure 1. A possible mechanistic pathway leading to indoles 5 and quinazolines 6.
Table 3. Silver triflate-catalyzed cycloisomerization of 1 to tive results are shown in Table 4. The bases tBuONa,
give 4-methylenequinazolines 6.
NaH, and TBAF promoted the reaction; however,
bases K2CO3, Et3N, and DBU did not. The reactions
of substrate 1g in the presence of tBuONa, NaH, or
TBAF gave quinazolines 6g with good regioselectivity
(entries 1–4), whereas indoles 5i and 5m were exclu-
sively formed for the reactions of 1i and 1m in the
presence of the same bases (entries 5–8 and 9–12).
Among the bases used, TBAF was the best choice
and acetonitrile was the best solvent (entries 12–17).
Next, our efforts turned to reducing the amount of
base, TBAF, under suitable reaction conditions for
each substrate. The data (entries 3 vs. 4, 7 vs. 8, and
11 vs. 12) suggest that appropriate amounts of base
(TBAF) are necessary for each substrate 1 to main-
tain a high yield and selectivity of the products.
Entry
1, 6
R2
R3
Time [h]
Yield [%][a]
1
2
3
4
y
z
a
Ph
nPr
Ph
Ph
Ph
Et
Et
1
15
1
75
71
92
61
b
nPr
0.5
[a]
Not detected for 5.
On the basis of the results from the preliminary ex-
tion of the electron-withdrawing substituent RF-bear- periments (Table 4), we performed the TBAF-pro-
ing substrates preferred the formation of indoles (17 moted reaction in acetonitrile using substrates 1 with
examples), while InCl3 (CuSO4, etc.) or K2CO3 cata- a variety of substituents in combinations of aryl and
lysts preferentially produced quinazolines (three ex- alkyl groups at R1, R2, and R3. The best and represen-
amples; R1 =H or Ph) (eq. (3)).[9c] The AgNO3-cata- tative results for all of the various substrates 1 are
lyzed reaction of the guanidine- or isourea-bearing shown in Table 5. The data suggest the following. (1)
substrates (X=NR2 or OR; R1 =Ph, R2 =aryl group In cases where substrate 1 bears an alkyl group at
only) exclusively afforded indoles (23 examples), no substituent R1 and an aryl (Ph) group at R3 (en-
formation of quinazolines being described (eq. tries 1–7), the reaction produced indoles 5 exclusively
(4)).[9d,13] The reaction is highly regioselective, but un- in high yield even when R2 is either an aryl group (en-
fortunately the scope of the substituents seemed to be tries 1–3) or an alkyl group (entries 4–7). (2) Even
rather limited (R1 =Ph; R2 =aryl group only).
when R1 is an aryl group (entries 8–11), the reaction
of 1 (R3 =Ph) also produced indoles 5 regioselective-
ly. In these cases, a bulky (alkyl) group at R2 (tBu,
iPr) is necessary regioselectively to give indoles 5
(compare with the cases of less bulky R2 =nPr, vide
infra: entries 19–22). (3) In cases of substrates 1k and
Base-promoted cyclization of N-imidoyl-2-
alkynylanilines (1)
Next, we examined the base-promoted cycloisomeri- 1l when R1 (tBu) and R3 (Et) are both alkyl groups
zation of 1 to examine the efficiency and selectivity of (entries 12 and 13), the TBAF-promoted reaction did
the reaction (Table 4). Preliminary experiments for not proceed at all even on prolonged heating. This
the reaction showed that the reaction rates (effective- was compensated by the fact that AgOTf-catalyzed
ness, namely yield) and regioselectivity were signifi- reaction of the corresponding 1k, l indeed occurred
cantly dependent upon the substituents (R1, R2, and to give indoles 5k, l in 67 and 65% yields, respective-
R3) of 1, bases, and the amounts used. First, we select- ly (vide supra: Table 2, entries 11 and 12). (4) In con-
ed bases K2CO3, Et3N, DBU (1,8-diazabicy- trast to the above results, substrates 1 bearing at least
clo[5.4.0]undec-7-ene), tBuONa, NaH, and TBAF two aryl groups at R1 and R2, or R1 and R3 preferen-
(tetrabutylammonium fluoride) (3 equiv, 10 mol%) in tially produced quinazolines 6 in high yields (en-
a refluxing solvent such as acetonitrile, DMF, THF, tries 14–22), but for the latter cases (R1 and R3 =aryl
toluene, (CH2Cl)2, and CH2Cl2 in the reactions of group, entries 19–22), R2 should be a less bulky
model substrates 1g, 1i, and 1m. Several representa- (alkyl) group (nPr) (cf. entries 8–11). (5) When both
4
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Lewis Acid-Catalyzed or Base-Promoted Regioselective Cycloisomerization
Scheme 4. Reported examples of alternative cyclization modes in the catalyzed cycloisomerization of various o-alkynylani-
line derivatives.
R2 and R3 are alkyl groups (nPr and Et) such as 1d changed to Z/E=8:92. This observation suggests that
(entry 23), the reaction also afforded quinazoline 6d, the Z-isomer is a kinetic product and the E-isomer is
albeit in 37% yield. (6) In accordance with the silver a thermodynamic product. Furthermore, it is notewor-
triflate-catalyzed reaction, substrate 1y readily under- thy that these Z-isomers were actually rather stable
went 6-exo-dig cyclization in the presence of without isomerization in the presence of TBAF. In
10 mol% of TBAF in refluxing acetonitrile for 10 min fact, this observation was proved by tracing the ratio
1
to give product 6y exclusively (entry 24).
of E- and Z-isomers of 6a by H NMR spectroscopy.
The thus formed 4-(R1-methylene)quinazolines 6 The ratio of Z- and E-isomers (Z/E=87:13) of 6a dis-
were obtained as a mixture of E- and Z-isomers, solved in CDCl3 in the presence of TBAF (20 mol%)
where the Z-isomers were in the majority. However, did not changed at all after 20 h (Figure 2(b)). Fur-
the ratios seemed to be dependent mainly on the sub- thermore, when 1a was treated in the presence of
stituents R1 and R2 and circumstances. Interestingly, TBAF (5 mol%) in CD3CN on heating at 808C in an
the ratio gradually changed in solution. For example, NMR tube, 1H NMR measurements after cooling
the isolated quinazoline 6a (entry 14) in a ratio of Z/ showed that 1a was consumed and the formed Z- and
E=92:8 before silica gel chromatography treatment[14] E-isomers of 6a were in a ratio of 93:7 after 20 min,
was dissolved in CDCl3 and allowed to stand at room and in the same ratio even after 50 min and 6 h at
temperature for 50 h. During that time the ratios 808C, and after an additional 16 h at room tempera-
were traced by measuring integration of the olefinic ture (Figure 2(c)). As shown in Figure 3, because of
1
protons for both isomers by H NMR spectroscopy. the steric repulsion between the aryl groups of R1 (4-
The results are shown in Figure 2(a). The amount of Tol, Ph, 4-MeOC6H4, 4-CF3C6H4) and R2 (Ph) present
the major Z-isomer (Z/E=92:8) in the beginning in the Z-isomer (Z-6a), the Z-isomer is less stable
slowly decreased with isomerization to the E-isomer, than the E-isomer with relatively less steric repulsion
which increased with time, and after 50 h the ratio
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
5
ÞÞ
These are not the final page numbers!

FULL PAPERS
Takashi Otani et al.
Table 4. Screening of bases and solvents in base-promoted cycloisomerization using model substrates 1g, 1i, and 1m.
Entry
R1
Base[a]
Solvent
Time [h]
Yield [%]
(3 equiv–10 mol%)
5
6
1
2
3
4
5
6
7
8
4-MeOC6H4 (1g)
tBuONa (3 equiv)
NaH (3 equiv)
TBAF (3 equiv)
TBAF (1.5 equiv)
tBuONa (3 equiv)
NaH (3 equiv)
TBAF (3 equiv)
TBAF (10 mol%)
tBuONa (3 equiv)
NaH (3 equiv)
TBAF (3 equiv)
TBAF (10 mol%)
TBAF (10 mol%)
TBAF (10 mol%)
TBAF (10 mol%)
TBAF (10 mol%)
TBAF (10 mol%)
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
DMF
33
28
1
12
6
9
72
60
89
87
0
0
0
1
8
tBu (1i)
33
30
11
22
29
15
0.4
0.5
26
13
21
23
36
93
97
99
9
11
50
99
99
82
4
ND[b,c]
ND[b,c]
ND[b,c]
0
9
nPr (1m)
10
11
12
13
14
15
16
17
0
ND[b,c]
ND[b,c]
ND[b,c]
ND[b,c]
ND[b,c]
THF
toluene
(CH2Cl)2
CH2Cl2
28
ND[b]
4
[a]
TBAF (1 mol/L THF solution).
Not detected.
Considerable amounts of starting material 1 remained.
[b]
[c]
between the R1 (4-Tol, Ph, 4-MeOC6H4, 4-CF3C6H4) N acceptor,[15] thereby Z-6 preferably formed by the
group and the peri-H5 of the benzo group.
protonation. Thus, in such basic conditions the kineti-
The olefinic protons present in the deshielded zone cally controlled, predominant formation of the Z-
of the Z-isomers appeared at d 6.20–6.63 as a singlet isomer was observed. In fact, the Z-isomers of 6,
and those of the E-isomers at d 5.00–5.26 as a singlet having an enamine structure, were stable under the
(for 6a, c, f–h, entries 14–18), which also supports the basic conditions described above. When the substitu-
stereochemical assignments of the Z- and E-isomers ents R1 and R2 are both aryl groups, the Z-isomers of
(see Supporting Information). In contrast to the 6 in a solution (CDCl3) or by silica gel (chromatogra-
above observations, for instance, both the Z- and E- phy)[14] slowly isomerized to the corresponding ther-
isomers of 6b (R1 =4-Tol, R2 =nPr, R3 =Ph) were modynamically stable E-isomers.
stable in CDCl3 solution at room temperature for
a long time without any isomerization, as shown in
Figure 2(d). This is probably due to the much lower Conclusions
steric hindrance between the R1 (4-Tol) and R2 (nPr)
groups in the Z-isomer (for 6b, d, w, x, entries 19–23) In conclusion, we have developed a AgOTf-catalyzed
relative to the high barrier of the isomeric rotation and TBAF-mediated cycloisomerization reaction of
(Figure 3).
simple N-imidoyl-2-alkynylanilines bearing various
Based on the above observations in the TBAF-pro- common substituents, and evaluated experimentally
moted cyclization, we propose a plausible mechanistic the reactivity and product selectivity associated with
pathway (Scheme 5). Initially, the fluoride anion de- the Baldwin cyclization modes. By these methods, in-
protonates the amidine-NH to form the intermediate doles and 4-methylenequinazolines were synthesized
anion A, which possesses an ambident nucleophilic in high yields with high regio- and stereoselectivities,
N-center. The adjacent nitrogen nucleophile attacks and with remarkable tolerance to various substituents
the proximal inner alkyne carbon to form the cyclized by a combination of substrate and catalyst control.
carbanion B, in which the Z-form is favored due to Although at the present stage we could not have suffi-
a stabilization by hyperconjugative nC- ! s*CꢀN inter- cient theoretical evidence nor any experimental evi-
action of anionic carbon orbital and antiperiplanar C– dence of potential key intermediates to account for
6
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Lewis Acid-Catalyzed or Base-Promoted Regioselective Cycloisomerization
Table 5. TBAF-catalyzed cycloisomerization of 1 to give indoles 5 and/or quinazolines 6.
Entry
1
R1
R2
R3
TBAF
Temp/Time [h]
Yield [%]
amounts
5
6
1
2
3
4
5
6
7
8
1i
tBu
nPr
cHex
tBu
tBu
cHex
nPr
4-Tol
4-Tol
Ph
4-MeOC6H4
tBu
tBu
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Et
Et
Ph
Ph
Ph
Ph
Et
Ph
Ph
Ph
Ph
Et
Ph
1.5 equiv
10 mol%
0.5 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
3.0 equiv
5 mol%
reflux/5
reflux/0.5
reflux/3
97
99
97
86
87
87
97
76
0
0
0
0
0
0
0
12
14
16
9
1m
1o
1j
1p
1q
1r
1s
1t
1u
1v
1k
1l
1a
1 f
1g
1h
1c
1b
1b
1w
1x
1d
1y
nPr
iPr
iPr
iPr
iPr
tBu
iPr
iPr
Ph
nPr
Ph
Ph
Ph
Ph
reflux/20
reflux/44
reflux/18
reflux/27
reflux/26
reflux/26
reflux/8
reflux/33
reflux/55
reflux/55
reflux/0.3
reflux/0.2
reflux/0.3
reflux/0.2
reflux/0.4
reflux/2
9
80
75
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
86
ND[a,b]
ND[a,b]
ND[a,b]
96
96
95
96
95
89
96
89
91
37[c]
91
ND[a,b]
4-Tol
Ph
3
2
3
0
5 mol%
4-MeOC6H4
4-CF3C6H4
4-Tol
4-Tol
4-Tol
Ph
4-MeOC6H4
4-Tol
H
10 mol%
1.0 mol%
10 mol%
0.5 equiv
1.5 equiv
1.0 equiv
0.5 equiv
1.5 equiv
10 mol%
Ph
trace
8
3
4
nPr
nPr
nPr
nPr
nPr
Ph
r.t./48
reflux/2
reflux/0.5
reflux/10
reflux/0.2
8
trace
0
[a]
Not detected for 5/6.
Considerable amounts of starting material 1 remained.
Intractable complex mixture remained.
[b]
[c]
and the aqueous layer was extracted with ethyl acetate (3 ꢂ
5 mL). Combined organic layers were washed with saturated
aqueous sodium chloride (5 mL) and water (5 mL), and
dried (MgSO4). Evaporation of the solvent in vacuo and
column chromatography of the residue on silica gel (ethyl
acetate/hexane=1:16) gave 5a (29.7 mg, 99%).
the regioselectivity of the reaction, we could predict
some preferential regioselectivity and E/Z-stereose-
lectivity, viz. formation of indoles vs. quinazolines,
when our simple substrates were used. With our re-
sults and the precedent results in mind, we conclude
that a more detailed study will be necessary for more
general and detailed explanation of the reaction.
Tetrabutylammonium fluoride-promoted cyclization
of N-imidoyl-2-alkynylanilines 1. A typical procedure
(Table 5, entry 14):
Experimental Section
To a solution of benzimidamide 1a (45.0 mg, 0.116 mmol) in
acetonitrile (1.2 mL), a THF solution (1.0m, 5.8 mL) of
TBAF (0.0058 mmol, 5 mol%) was added at 08C with stir-
ring under an argon atmosphere. The reaction mixture was
heated under reflux (808C) with stirring for 20 min, and
after cooling to 08C, quenched by addition of saturated
aqueous ammonium chloride (5 mL). The reaction mixture
was extracted with ethyl acetate (2 ꢂ 5 mL). The combined
organic layers were washed with saturated aqueous sodium
chloride (10 mL) and water (5 mL), and dried (MgSO4).
Evaporation of the solvent in vacuo and column chromatog-
raphy of the residue on silica gel (ethyl acetate/hexane=
Silver triflate-catalyzed cyclization of N-imidoyl-2-
alkynylanilines 1. A typical procedure (Table 2,
entry 1):
To a solution of (E)-N’-phenyl-N-[2-(p-tolylethynyl)phenyl]-
benzimidamide 1a (30.0 mg, 0.078 mmol) in dry dichlorome-
thane (1 mL), AgOTf (2.0 mg, 0.0078 mmol, 10 mol%) was
added, and the reaction mixture was stirred at room temper-
ature for 40 min under an argon atmosphere. The reaction
mixture was quenched by addition of saturated aqueous
sodium carbonate (5 mL), the organic layer was separated,
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
7
ÞÞ
These are not the final page numbers!

FULL PAPERS
Takashi Otani et al.
Figure 2. (a) Variation of mole fractions of E- and Z-isomers (6a) with time (h) in CDCl3 at room temperature. (b) Variation
of mole fractions of E- and Z-isomers (6a) with time (h) in the presence of TBAF in CDCl3 at room temperature. (c) For-
mation of 6a from 1a in the presence of TBAF (10 mol%) in CD3CN on heating at 808C, and the subsequent mole fractions
of E- and Z-isomers with time (min). (d) Variation of mole fractions of E- and Z-isomers (6b) with time (h) in CDCl3 at
room temperature.
Figure 3. Chemical shifts of the olefinic protons in the E-
and Z-isomers.
1:8) gave indole 5a (1.35 mg, 0.00349 mmol, 3%) as color-
less crystals and quinazoline 6a (43.2 mg, 0.111 mmol, 96%)
as yellow crystals (recrystallized from ethyl acetate/hexane,
or 1,2-dichloroethane/hexane for X-ray crystallography of
6g).
Scheme 5. A plausible mechanistic pathway leading to Z-
and E-isomers (6) for the TBAF-promoted cyclization.
8
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Lewis Acid-Catalyzed or Base-Promoted Regioselective Cycloisomerization
Experimental details and copies of 1H/13C NMR spectra
of all new compounds are available as supporting informa-
tion.
Sinai, A. Mꢄszꢅros, T. Gꢅti, V. Kudar, A. Pallꢁ, Z.
Novꢅk, Org. Lett. 2013, 15, 5654; c) J. K. Vandavasi, K.-
K. Kuo, W.-P. Hu, H.-C. Shen, W.-S. Lo, J.-J. Wang,
Org. Biomol. Chem. 2013, 11, 6520; d) W.-C. Lee, H.-C.
Shen, W.-P. Hu, W.-S. Lo, C. Murali, J. K. Vandavasi, J.-
J. Wang, Adv. Synth. Catal. 2012, 354, 2218.
References
[6] Other palladium (such as PdCl2), platinum, copper,
indium, TBAF, tBuOK and IPy2BF4 catalysts predomi-
nantly produced indoles; Pd: a) D. E. Rudisill, J. K.
Stille, J. Org. Chem. 1989, 54, 5856; b) K. Iritani, S.
Matsubara, K. Utimoto, Tetrahedron Lett. 1988, 29,
1799; c) E. C. Taylor, A. H. Katz, H. Salgado-Zamora,
A. McKillop, Tetrahedron Lett. 1985, 26, 5963; d) S. Ye,
Q. Ding, Z. Wang, H. Zhou, J. Wu, Org. Biomol.
Chem. 2008, 6, 4406; e) I. Ambrogio, S. Cacchi, G. Fab-
rizi, Org. Lett. 2006, 8, 2083; f) A. Arcadi, S. Cacchi, G.
Fabrizi, F. Marinelli, L. M. Parisi, J. Org. Chem. 2005,
70, 6213; g) S. Cacchi, G. Fabrizi, P. Pace, J. Org. Chem.
1998, 63, 1001; Pt: h) T. Shimada, I. Nakamura, Y. Ya-
mamoto, J. Am. Chem. Soc. 2004, 126, 10546; Cu: i) K.
Hiroya, S. Itoh, T. Sakamoto, J. Org. Chem. 2004, 69,
1126; j) K. Hiroya, S. Itoh, M. Ozawa, Y. Kanamori, T.
Sakamoto, Tetrahedron Lett. 2002, 43, 1277; InBr3:
k) N. Sakai, K. Annaka, A. Fujita, A. Sato, T. Konaka-
hara, J. Org. Chem. 2008, 73, 4160; l) N. Sakai, K.
Annaka, T. Konakahara, Tetrahedron Lett. 2006, 47,
631; TBAF: m) A. Yasuhara, N. Suzuki, T. Yoshino, Y.
Takeda, T. Sakamoto, Tetrahedron Lett. 2002, 43, 6579;
n) A. Yasuhara, Y. Kanamori, M. Kaneko, A. Numata,
Y. Kondo, T. Sakamoto, J. Chem. Soc. Perkin Trans.
1 1999, 529; IPy2BF4: o) J. Barluenga, M. Trincado, E.
Rubio, J. Gonzꢆlez, Angew. Chem. 2003, 115, 2508;
Angew. Chem. Int. Ed. 2003, 42, 2406; tBuOK: p) L.-P.
Sun, X.-H. Huang, W.-M. Dai, Tetrahedron 2004, 60,
10983.
[7] For reviews (indole synthesis): a) P. M. Barbour, L. J.
Marholz, L. Chang, W. Xu, X. Wang, Chem. Lett. 2014,
43, 572; b) M. Shiri, Chem. Rev. 2012, 112, 3508; c) K.
Chen, X. Liu, L. Lin, X. Feng, Chem. Sci. 2012, 3, 327;
d) F. Zhou, Y.-L. Liu, J. Zhou, Adv. Synth. Catal. 2010,
352, 1381; e) A. Millemaggi, R. J. K. Taylor, Eur. J.
Org. Chem. 2010, 4527; f) K. Krꢇger (nꢄe Alex), A.
Tillack, M. Beller, Adv. Synth. Catal. 2008, 350, 2153;
g) G. R. Humphrey, J. T. Kuethe, Chem. Rev. 2006, 106,
2875; h) G. Zeni, R. C. Larock, Chem. Rev. 2006, 106,
4644; i) S. Cacchi, G. Fabrizi, Chem. Rev. 2005, 105,
2873; (functionalization of indoles): j) D. Zhang, H.
Song, Y. Qin, Acc. Chem. Res. 2011, 44, 447; k) G. Bar-
toli, G. Bencivenni, R. Dalpozzo, Chem. Soc. Rev. 2010,
39, 4449; l) L. Joucla, L. Djakovitch, Adv. Synth. Catal.
2009, 351, 673.
[1] For selected recent examples, see: a) L.-L. Pan, Y.
Yang, K. M. Merz Jr, Biochemistry 2014, 53, 6126; b) G.
Bianchi, M. Chiarini, F. Marinelli, L. Rossi, A. Arcadi,
Adv. Synth. Catal. 2010, 352, 136; c) A. Gimeno, M.
Medio-Simꢁn, C. Ramꢃrez de Arellano, G. Asencio,
A. B. Cuenca, Org. Lett. 2010, 12, 1900; d) F. Barabꢄ, P.
Levesque, I. Korobkov, L. Barriault, Org. Lett. 2011,
13, 5580; e) A.-L. Girard, T. Enomoto, S. Yokouchi, C.
Tsukano, Y. Takemoto, Chem. Asian. J. 2011, 6, 1321;
f) M. Yoshida, Y. Fujino, T. Doi, Org. Lett. 2011, 13,
4526; g) M. Yoshida, Y. Fujino, K. Saito, T. Doi, Tetra-
hedron 2011, 67, 9993; h) M. Bian, W. Yao, H. Ding, C.
Ma, J. Org. Chem. 2010, 75, 269; i) G. Zhou, J. Zhang,
Chem. Commun. 2010, 46, 6593; j) Y. Xiao, J. Zhang,
Chem. Commun. 2009, 3594; k) K.-S. Masters, B. L.
Flynn, Adv. Synth. Catal. 2009, 351, 530; l) L. Liu, J.
Zhang, Angew. Chem. 2009, 121, 6209; Angew. Chem.
Int. Ed. 2009, 48, 6093; m) N. Iwasawa, K. Maeyama,
H. Kusama, Org. Lett. 2001, 3, 3871; n) K. Lee, P. H.
Lee, Adv. Synth. Catal. 2007, 349, 2092; o) M. Uchiya-
ma, H. Ozawa, K. Takuma, Y. Matsumoto, M. Yone-
hara, K. Hirota, T. Sakamoto, Org. Lett. 2006, 8, 5517
and references cited therein. For our previous exam-
ples: p) T. Saito, H. Ohmori, E. Furuno, S. Motoki, J.
Chem. Soc. Chem. Commun. 1992, 22 (thermal electro-
cyclization); q) T. Saito, H. Ohmori, T. Ohkubo, S.
Motoki, J. Chem. Soc. Chem. Commun. 1993, 1802
(Lewis acid-induced intramolecular Diels–Alder).
[2] a) B. Godoi, R. F. Schumacher, G. Zeni, Chem. Rev.
2011, 111, 2937; b) N. T. Patil, Y. Yamamoto, Chem.
Rev. 2008, 108, 3395; c) G. Zeni, R. C. Larock, Chem.
Rev. 2004, 104, 2285; d) I. Nakamura, Y. Yamamoto,
Chem. Rev. 2004, 104, 2127; e) T. L. Gilchrist, J. Chem.
Soc. Perkin Trans. 1 2001, 2491; f) G. W. Gribble, J.
Chem. Soc. Perkin Trans. 1 2000, 1045.
[3] T. Saito, S. Ogawa, N. Takei, N. Kutsumura, T. Otani,
Org. Lett. 2011, 13, 1089.
[4] In the following literature, formation of 4H-3,1-benzox-
azine via a similar cyclization mode is described,
though clearly the whole reaction is not cycloisomeriza-
tion: a) S. Cacchi, G. Fabrizi, L. M. Parisi, Org. Lett.
2003, 5, 3843 (by-product in 2-aryl and 2-heteroaryl
indole synthesis); b) S. Cacchi, G. Fabrizi, F. Marinelli,
L. Moro, P. Pace, Synlett 1997, 1363 (by-product in aryl
indole synthesis); c) S. Cacchi, G. Fabrizi, F. Marinelli,
Synlett 1999, 401 (by-product in 2-arylquinoline synthe-
sis); d) M. Costa, N. D. Ca, B. Gabriele, C. Massera, G.
Salerno, M. Soliani, J. Org. Chem. 2004, 69, 2469 (in
the synthesis of benzoxazine, quinolin-2-one, and qui-
nolin-4-one).
[8] Recent reports for indole synthesis: a) G. Liu, G. Xu, J.
Li, D. Ding, J. Sun, Org. Biomol. Chem. 2014, 12, 1387
(Cu); b) R.-Y. Tang, X.-K. Guo, J.-N. Xiang, J. H. Li, J.
Org. Chem. 2013, 78, 11163 (Ag); c) M. Chiarucci, R.
Mocci, L.-D. Syntrivanis, G. Cera, A. Mazzanti, M.
Bandini, Angew. Chem. 2013, 125, 11050; Angew.
Chem. Int. Ed. 2013, 52, 10850 (Au); d) M. Chiarucci,
E. Matteucci, G. Cera, G. Fabrizi, M. Bandini, Chem.
Asian J. 2013, 8, 1776 (Au); e) P. P. Sharp, M. G. Ban-
well, J. Renner, K. Lohmann, A. C. Willis, Org. Lett.
2013, 15, 2616 (Au); f) W. Yang, T. Wang, Y. Yu, S. Shi,
[5] Recently, iron-, copper-, or iodine-mediated regioselec-
tive 6-exo-dig mode cyclization of N-acyl-2-alkynylani-
lines for synthesis of 4-methylene-4H-3,1-benzoxazines
has been reported: a) A. L. Stein, F. N. Bilheri, D. F.
Back, G. Zeni, Adv. Synth. Catal. 2014, 356, 501; b) A.
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
9
ÞÞ
These are not the final page numbers!

FULL PAPERS
Takashi Otani et al.
T. Zhang, A. S. K. Hashmi, Adv. Synth. Catal. 2013,
355, 1523 (Pt); g) H. Kusama, H. Sogo, K. Saito, T.
Suga, N. Iwasawa, Synlett 2013, 24, 1364 (Pt); h) G.
Qiu, C. Chen, L. Yao, J. Wu, Adv. Synth. Catal. 2013,
355, 1579 (Pd); i) J. Liu, X. Xie, Y. Liu, Chem.
Commun. 2013, 49, 11794; j) M. Xu, K. Xu, S. Wang,
Z.-J. Yao, Tetrahedron Lett. 2013, 54, 4675 (ZnBr2);
k) S. Cacchi, G. Fabrizi, A. Goggiamani, A. Perboni, A.
Sferrazza, P. Stabile, Org. Lett. 2010, 12, 3279; l) P. Ko-
thandaraman, W. Rao, S. J. Foo, P. W. H. Chan, Angew.
Chem. 2010, 122, 4723; Angew. Chem. Int. Ed. 2010, 49,
4619; m) R. Bernini, G. Fabrizi, A. Sferrazza, S. Cacchi,
Angew. Chem. 2009, 121, 8222; Angew. Chem. Int. Ed.
2009, 48, 8078; n) W. Yu, Y. Du, K. Zhao, Org. Lett.
2009, 11, 2417; o) T. Miura, T. Toyoshima, Y. Ito, M.
Murakami, Chem. Lett. 2009, 38, 1174; p) G. Cantagrel,
B. de Carnꢄ-Carnavalet, C. Meyer, J. Cossy, Org. Lett.
2009, 11, 4262; q) Y.-X. Jia, E. P. Kꢇndig, Angew.
Chem. 2009, 121, 1664; Angew. Chem. Int. Ed. 2009, 48,
1636.
c) J. Zhu, H. Xie, Z. Chen, S. Li, Y. Wu, Org. Biomol.
Chem. 2012, 10, 516; and sec-amino or alkoxyl group in
R3 (AgNO3): d) N.-Y. Huang, M.-G. Liu, M.-W. Ding, J.
Org. Chem. 2009, 74, 6874.
[10] Based on Baldwinꢈs rule, it was recently proposed that
the 3- and 4-exo-dig cyclization modes are to be revised
as “allowed” and the 3- and 4-endo-dig cyclization
modes as “disallowed”; a) K. Gilmore, I. V. Alabugin,
Chem. Rev. 2011, 111, 6513; b) I. V. Alabugin, K. Gil-
more, J. Am. Chem. Soc. 2011, 133, 12608.
[11] For discussion, see I. V. Alabugin, K. Gilmore, Chem.
Commun. 2013, 49, 11246. We thank one of the refer-
ees for this suggestion.
[12] For a mechanistic discussion, see reference 9a. Metal
acetylide (deprotonated anionic complex) entails un-
favorable build-up of positive charge at a terminal ace-
tylenic carbon for the N-nucleophilic attack. If addi-
tional AgOTf activates the metal acetylide as a s,p-
double complex form, formation of indole will be fa-
vored.[9a]
[9] Recently, coinage metal (Au, Ag, Cu)- or K2CO3-cata-
lyzed cyclizations of N-imidoyl-substituted o-alkynyla-
nilines of type (I) leading to indole and/or quinazoline
derivatives have been reported. In these precedents the
substrates bear special substituents: a) one substrate of
R1 =H, R2 =Ph, R3 =O (urea-type) only: gold(I)-cata-
lyzed cyclizations 2-ethynylphenylurea to give indole
(5-endo-dig) and quinazolinone (6-exo-dig); A.
Gimeno, A. B. Cuenca, S. Suꢅrez-Pantiga, C. R. de Ar-
ellano, M. Medio-Simꢁn, G. Asensio, Chem. Eur. J.
2014, 20, 683; b) 6-chloropyrazin-3-one-2-yl group at R2
and R3 (AgOTf or AuCl/TFA + microwave radiation):
D. D. Vachhani, V. P. Mehta, S. G. Modha, K. Van
Hecke, L. VanMeervelt, E. V. Van der Eycken, Adv.
Synth. Catal. 2012, 354, 1593; trifluoroacetamidino or
[13] Previously, we reported the formation of 4-alkyldihy-
droquinazolines via 6-exo-trig mode cyclization, namely
via intramolecular aza-conjugate addition of o-amidino-
cinnamates (X=NR2 or OR) starting from the corre-
sponding carbodiimides as a tandem intermolecular N-
or O-nucleophile addition/intramolecular aza-Michael
addition in one pot: T. Saito, K. Tsuda, Y. Saito, Tetra-
hedron Lett. 1996, 37, 209.
[14] We observed that the Z-isomers (6a, c, f, g) gradually
isomerize to the E-isomers during silica gel chromatog-
raphy. Rf values of the E- and Z-isomers are essentially
overlapped or very close to each other.
[15] For the discussion of stereoelectronic features in exo-
dig cyclizations of anionic N-centered nucleophiles,
see: S. F. Vasilevsky, T. F. Mikhailovskaya, V. I. Mama-
tyuk, G. E. Salnikov, G. A. Bogdanchikov, M. Manohar-
an, I. V. Alabugin, J. Org. Chem. 2009, 74, 8106.
bromodifluoroacetamidino
group
in
R3
(NaAuCl4·2H2O, Cu(OAc)2, InCl3·3H2O or K2CO3, I2):
10
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Lewis Acid-Catalyzed or Base-Promoted Regioselective
Cycloisomerization of N-Imidoyl-o-alkynylanilines for
Synthesis of N-Imidoyl-(1H)-indoles and 4-Alkylidene-3,4-
dihydroquinazolines
11
Adv. Synth. Catal. 2015, 357, 1 – 11
Takashi Otani,* Xue Jiang, Kinryo Cho, Rino Araki,
Noriki Kutsumura, Takao Saito*
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
11
ÞÞ
These are not the final page numbers!