A Method for Bischler-Napieralski-Type Synthesis of 3,4-Dihydroisoquinolines

Article abstract of DOI:10.1021/acs.orglett.9b00534

A new method for the Bischler-Napieralski-type synthesis of 3,4-dihydroisoquinolines was developed by a Tf₂O-promoted tandem annulation from phenylethanols and nitriles. Its success was mainly due to the fact that a phenonium ion was formed in the process and practically functioned as a stable and reactive primary phenylethyl carbocation.

Full text of DOI:10.1021/acs.orglett.9b00534

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
A Method for Bischler−Napieralski-Type Synthesis of 3,4-
Dihydroisoquinolines
Lin Min, Weiguang Yang, Yunxiang Weng, Weiping Zheng, Xinyan Wang,* and Yuefei Hu*
Key Laboratory of Bioorganic Phosphorus Chemistry and Chemical Biology (Ministry of Education), Department of Chemistry,
Tsinghua University, Beijing 100084, P.R. China
S
* Supporting Information
ABSTRACT: A new method for the Bischler−Napieralski-type
synthesis of 3,4-dihydroisoquinolines was developed by a Tf2O-
promoted tandem annulation from phenylethanols and nitriles. Its
success was mainly due to the fact that a phenonium ion was formed
in the process and practically functioned as a stable and reactive
primary phenylethyl carbocation.
ue to the unique structure of 3,4-dihydroisoquinolines 1,
they have been broadly used as versatile synthons in
organic synthesis. As shown in Scheme 1, their major
electrophilic aromatic substitution of N-arylethyl amides 2 to
produce the products 1 in the presence of dehydrating agents.
The mechanism studies for the B−N reaction proved that the N-
arylethyl nitrilium salts 3 generated from the dehydration of the
amides 2 were the key intermediates,7 and their reactivity was
influenced significantly by the counterions (X−).8 For example,
the classic B−N reactions usually were limited to the amides 2
bearing the electron-rich arenes as substrates (Scheme 2a, R1 =
alkoxyl, dialkoxyl or trialkoxyl) by using POCl3, P2O5, or POCl3,
etc. as dehydrating reagents.6a However, the high efficiency for
the amides 2 bearing normal arenes (Scheme 2b, R1 = H, Cl, or
Br) was achieved in Movassaghi’s work9 when Tf2O/2-ClPy was
used as a dehydrating reagent. This advantage may result from
the fact that the counterion TfO− in their nitrilium salts 3 has
almost no nucleophilicity.
D
Scheme 1. 3,4-Dihydroisoquinolines 1 as versatile synthons
Although different dehydrating agents were employed in the
classic B−N methods and Movassaghi’s method, they actually
used common substrates and pathways. From a synthetic point
of view, all these methods belong to linear synthesis. As a result,
their common substrates N-arylethyl amides 2 bearing the same
carbon numbers and functional groups as the target products 1
must be premade by a multistep synthesis.10 Herein, we report a
new method for the synthesis of 3,4-dihydroisoquinolines 1 via a
Tf2O-promoted tandem annulation from phenylethanols 4 and
nitriles 5. As shown in Scheme 2c, this method can be
considered as a modified B−N reaction because its key steps
involve the formation and the electrophilic aromatic sub-
stitutions of the nitrilium salts 3. However, this method was
carried out through a convergent synthesis and the use of easy
substrates.
transformations included (a) the dehydroaromatization of the
B-ring,1 (b) the hydrogenation of a double bond,2 (c) the
addition of a double bond,3 and (d) the cycloaddition of a
double bond.4 Many 3,4-dihydroisoquinolines 1 have been used
in the total synthesis of the bioactive natural alkaloids.5
Numerous methods for the synthesis of 3,4-dihydroisoquino-
lines 1 have been developed in the literature. The most
prominent among them is the Bischler−Napieralski reaction
(B−N reaction) for its practicability and reliability.5,6 As shown
in Scheme 2a, the B−N reaction is an intramolecular
Scheme 2. Classic and Modified B−N Methods
Investigation indicated that, like the dehydration of amides,
the N-alkylation of nitriles with carbocations is also an efficient
method for the formation of nitrilium salts.8 But this method has
not been efficiently used for the synthesis of 3,4-dihydroisoqui-
nolines 1 thus far in the literature. For example, Lora-Tamayo11
Received: February 11, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b00534
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
et al. in 1960 reported a synthesis of 3,4-dihydroisoquinolines 1
via a SnCl4-promoted cyclization of 2-phenylethyl chlorides 6
and nitriles 5. As shown in Scheme 3a, they proposed that a
shown in Scheme 5a, the desired product 1-phenyl-3,4-
dihydroisoquinoline (1ba) was obtained only in 9% yield. To
Scheme 5. Primary Conditional Tests
Scheme 3. Lora-Tamayo’s Method and Ho’s Method
our surprise, 1ba was obtained in 26% yield by using the
phenylethyl triflate (12a) generated in situ from phenylethanol
(4b) and Tf2O (Scheme 5b). This result suggested that the
conversion of triflate 12a into its phenonium ion 13a may be
catalyzed by TfOH released from the triflation of 4b.18 This
phenomenon was confirmed by the fact that the yield of 1ba was
tripled (29%) when 1 equiv of TfOH was employed in Scheme
5a. As was expected, 7-methoxy-3,4-dihydroisoquinoline (1aa)
was obtained in 32% yield when electron-rich 4-methoxyphe-
nylethanol (4a) was used as a substrate (Scheme 5c).
primary phenylethyl carbocation 7 was formed by a normal
ionization of 6 and then quickly underwent an N-alkylation with
nitrile 5 to give an N-phenylethyl nitrilium salt 3. However, this
method has rarely been used in past decades12 due to its hash
conditions. However, Ho13 et al. in 2003 reported that 2-methyl-
2-phenylethyl chloride (8) reacted with benzonitrile (5a) under
Lora-Tamayo’s conditions to yield 1-phenyl-3-methyl-3,4-
dihydroisoquinoline (11) (Scheme 3b). They proved that the
migration of the methyl group in this reaction resulted from the
formation of a stable phenonium ion 9, which was the real
reactive intermediate rather than the unstable primary
carbocation. In literature, the phenonium ions have been well
studied,14 and they functioned as stable and reactive primary
phenylethyl carbocations driven by the relief of the cyclopropyl
ring strain and recovery of the aromatic benene.15
Thus, the conditions for the synthesis of 1aa were optimized
as shown in Table 1. The ratios of 4a to 5a (entries 1−4) showed
a
Table 1. Effects of the Substrates and Additives
In another example, the N-alkylation of nitriles with alkyl
triflates has often been used for the formation of nitrilium salts.8
However, this method is usually limited to the formation of N-
methylnitrilium salts (RCN+−Me) by using methyl triflate
(MeOTf) as an alkylating reagent.16 This result may be caused
by the fact that the carbocations formed from the alkyl triflates
bearing a carbon chain longer than two carbons may easily carry
out the undesired carbocation rearrangements during the N-
alkylations. As a result, 2-phenylethyl triflates 12 have been
never used to react with nitriles 5 for the synthesis of 3,4-
dihydroisoquinolines 1, even though some of them were known
products17 (Scheme 4a). However, all the above investigations
b
entry
4a/5a (mole)
additive
1aa (%)
1
2
3
4
5
6
7
8
1:1
32
50
56
33
70
75
80
84
87
85
85
75
1:1.2
1:1.5
1.5:1
1:1.5
1:1.5
1:1.5
1:1.5
1:1.5
1:1.5
1:1.5
1:1.5
DABCO
Et3N
DBU
DAMP
2-FPy
2-ClPy
2-IPy
9
10
11
12
Scheme 4. Hypothesis for the Synthesis of 1 from 12
2,6-Cl2Py
a
The mixture of 4a and 5a in DCE (3 mL) was treated with Tf2O (2
mmol) at 0 °C for 30 min with or without an additive (1 mmol) and
b
then was heated at 80 °C for 12 h. Isolated yields.
significant influences to the results, and the best result was
obtained at 1:1.5 (entry 3). In order to modulate the strength of
TfOH, several amine-based additives were tested as acid-
binding agents. As shown in entries 5−7, the yield of 1aa was
significantly improved by adding DABCO, Et3N, or DBU. The
pyridine derivatives seemed to be better additives to give higher
yield of 1aa (entries 8−12). Finally, the conditions in entry 9
were chosen for further conditional tests.
enabled us to realize that the carbocations generated from 2-
phenylethyl triflates 12 may be stabilized by forming the
corresponding phenonium ions 13, which then react with
nitriles 5 to yield 3,4-dihydroisoquinolines 1 (Scheme 4b).
Thus, the mixture of phenylethyl triflate (12a) and
benzonitrile (5a) in DCE was heated at 80 °C for 12 h. As
As shown in Table 2, the yield of 1aa was decreased by
increasing or decreasing the reaction temperatures (entries 2
and 3). The yield of 1aa was decreased by reducing the reaction
B
DOI: 10.1021/acs.orglett.9b00534
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
substituted 4-trifluoromethylbenzonitrile (4j) and 4-nitro-
benzonitrile (4k), respectively.
Table 2. Effects of the Temperature, Time, and Solvent
As shown in part b of Scheme 6, like Movassaghi’s method, the
non-electron-rich phenylethanol (4b) reacted with nitriles 5a−h
to give the corresponding products 1ba−bh in satisfactory yields
under standard conditions. Due to the electronic effects, the
product 1da was synthesized from m-methylphenylethanol (4d)
in a lower yield (58%) compared to the products 1ca (81%) and
1ea (74%). By prolonging the reaction time, the halogen-
substituted phenylethanols 4g and 4h were successfully
converted into the corresponding products 1ga and 1ha. To
our surprise, the product 1ia was synthesized smoothly, and its
terminal alkyne stayed under these conditions. Under standard
conditions, the products 1aa and 1ba were prepared on 2 g
scales in 82% and 68% yields, respectively.
However, the phenylethanols bearing strong electron-with-
drawing-groups, such as acyl, cyano, or nitro groups, were not
suitable substrates for this method. Complicated mixtures were
obtained when benzyl alcohol (PhCH2OH) or 3-phenyl-1-
propanol [Ph(CH2)3OH] were used as a substrate to react with
5a, possibly because they could not form the phenonium ions.
Thus, a possible pathway was proposed for our new method. As
shown in Scheme 7, phenylethanol 4 initially was triflated by
b
entry
temp (°C)
time (h)
solvent
1aa (%)
1
2
3
4
5
6
7
80
70
90
80
80
80
80
12
12
12
10
14
12
12
DCE
DCE
DCE
DCE
DCE
toluene
CHCl3
87
58
86
74
87
65
63
a
The mixture of 4a (1 mmol), 5a (1.5 mmol), and 2-FPy (1 mmol) in
DCE (3 mL) was treated by Tf2O (2 mmol) at 0 °C for 30 min and
then was heated at the given temperatures and times. Isolated yields.
b
time, but no improvement was observed by extending the
reaction time (entries 4 and 5). Compared to DCE, PhMe and
CHCl3 were quite poor solvents for this method (entries 6 and
7). Finally, the conditions in entry 1 were assigned as the
standard conditions.
At this time, we realized that we had found a new method for
an efficient synthesis of 3,4-dihydroisoquinolines 1. To the best
of our knowledge, no such method has been reported in
literature so far. Thus, the scope of this method was tested under
standard conditions. As shown in part a of Scheme 6, 4-
Scheme 7. Proposed Pathway for the Formation of 1
Scheme 6. Scope of the Substrates and the Products
Tf2O to form phenylethyl triflate 12. Promoted by TfOH and
heating, the triflate 12 lost its triflate group to give the
phenonium ion 13. Then the nitrile 5 was N-alkylated by
phenonium ion 13 to produce a reactive intermediate nitrilium
salt 3. Finally, nitrilium salt 3 undergent an intramolecular
electrophilic aromatic substitution to give the target product 3,4-
dihydroisoquinoline 1.
To confirm that the phenonium ion 13 is the active
intermediate for our method, three controlled experiments
were made under the standard conditions. As was expected, 3,3-
and 4,4-dideuterio products (15 and 16) were obtained as 1:1
mixture in 61% yield when 1,1-dideuterio-2-phenylethanol (14)
was employed as a substrate (Scheme 8a). 1-Phenyl-3-methyl-
3,4-dihydroisoquinoline (11) was obtained as single product
when 2-methyl- or 1-methyl-2-phenylethanol (17 or 18) was
employed as a substrate (Scheme 8b,c). The migration of the
deuterium and the methyl group in Scheme 8a,b indicated that
the phenonium ion 13 is indeed the active intermediate for our
method. The low yields of product 11 in Schemes 8b,c may be
caused by the fact that the corresponding nitrilium ion (the
analogues of 3) was subjected to a retro-Ritter reaction. This
phenomenon was fully in agreement with the early studies of the
B−N reaction where the nitrilium ion with a more substituted
carbon bonded to the N-atom has more tendency to undergo a
retro-Ritter reaction.7a,19 In fact, this is also the reason why the
B−N reaction is mainly used for the synthesis of 3,4-
nonsubstituted 3,4-dihydroisoquinolines.5,6,9
methoxyphenylethanol (4a) reacted smoothly with different
nitriles 5a−n to yield the corresponding products 1aa−an. In
these transformations, the steric and electronic effects of nitriles
were observed clearly. For example, 1aa was obtained in 88%
yield from p-methylbenzonitrile (5a), while 1ad was obtained in
only 56% yield from o-methylbenzonitrile (5d). As was
expected, 1aj (70%) and 1ak (60%) were synthesized in
relatively lower yields from the electron-withdrawing-group
C
DOI: 10.1021/acs.orglett.9b00534
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
7003−7007. (b) Geffe, M.; Opatz, T. Org. Lett. 2014, 16, 5282−5285.
(c) Verzijl, G. K. M.; de Vries, A. H. M.; de Vries, J. G.; Kapitan, P.; Dax,
T.; Helms, M.; Nazir, Z.; Skranc, W.; Imboden, C.; Stichler, J.; Ward, R.
A.; Abele, S.; Lefort, L. Org. Process Res. Dev. 2013, 17, 1531−1539.
(d) Chang, M.; Li, W.; Zhang, X. Angew. Chem., Int. Ed. 2011, 50,
10679−10681.
Scheme 8. Three Controlled Experiments and Results
(3) For selected recent references, see: (a) Dasgupta, S.; Liu, J.;
Shoffler, C. A.; Yap, G. P. A.; Watson, M. P. Org. Lett. 2016, 18, 6006−
6009. (b) Luu, H.-T.; Wiesler, S.; Frey, G.; Streuff, J. Org. Lett. 2015, 17,
2478−2481.
(4) For selected recent references, see: (a) Jarvis, C. L.; Jemal, N. M.;
Knapp, S.; Seidel, D. Org. Biomol. Chem. 2018, 16, 4231−4235.
(b) Tan, W. W.; Yoshikai, N. Chem. Sci. 2015, 6, 6448−6455.
(c) Unsworth, W. P.; Coulthard, G.; Kitsiou, C.; Taylor, R. J. K. J. Org.
Chem. 2014, 79, 1368−1376. (d) Eschenbrenner-Lux, V.; Kuchler, P.;
̈
Ziegler, S.; Kumar, K.; Waldmann, H. Angew. Chem., Int. Ed. 2014, 53,
2134−2137.
(5) For selected reviews, see: (a) Chrzanowska, M.; Grajewska, A.;
Rozwadowska, M. D. Chem. Rev. 2016, 116, 12369−12465. (b) Heravi,
M. M.; Nazari, N. Curr. Org. Chem. 2015, 19, 2358−2408.
(6) (a) Whaley, W. M.; Govindachari, T. R. Org. React. 1951, 6, 74−
150 (a review) . (b) Bischler, A.; Napieralski, B. Ber. Dtsch. Chem. Ges.
1893, 26, 1903−1908.
In conclusion, a new method for the Bischler−Napieralski-
type synthesis of 3,4-dihydroisoquinolines was developed by a
Tf2O-promoted tandem annulation from phenylethanols and
nitriles. Its success is mainly due to the formation of the stable
phenonium ions, by which the N-alkylation of nitriles was
achieved efficiently to give the key intermediate N-phenylethyl
nitrilium triflate. The method has two distinctive advantages: (a)
it changes the Bischler−Napieralski reaction from a linear
synthesis into a convergent synthesis and (b) it changes the
substrates from N-phenylethyl amides into phenylethanols.
Since both advantages practically promote the substrate and
product diversities of the Bischler−Napieralski reaction, we may
expect that this method will have widespread applications in
organic synthesis.
(7) (a) Fodor, G.; Nagubandi, S. Tetrahedron 1980, 36, 1279−1300 (a
review) . (b) Fodor, G.; Gal, J.; Phillips, B. A. Angew. Chem., Int. Ed.
Engl. 1972, 11, 919−920.
(8) For selected reviews, see: (a) van Dijk, T.; Slootweg, J. C.;
Lammertsma, K. Org. Biomol. Chem. 2017, 15, 10134−10144.
(b) Hegarty, A. F. Acc. Chem. Res. 1980, 13, 448−454.
(9) Movassaghi, M.; Hill, M. D. Org. Lett. 2008, 10, 3485−3488.
(10) In the B−N reaction, most N-arylethyl amides 2 are prepared by
acylation of the corresponding arylethylamines. Since most arylethyl-
amines are not commercially available, they need to be synthesized also.
(11) Lora-Tamayo, M.; Madron
1960, 93, 289−297.
ero, R.; Munoz, G. G. Chem. Ber.
̃ ̃
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.9b00534.
■
S
(12) Van Binst, G.; Baert, R. B. J. Heterocycl. Chem. 1975, 12, 1165−
1174.
(13) Ho, T.-L.; Chein, R.-J. J. Org. Chem. 2004, 69, 591−592.
(14) (a) Prakash, G. K. S.; Reddy, V. P. In Carbocation Chemistry;
Olah, G. A., Prakash, G. K. S., Eds.; John Wiley & Sons, Inc.: New York,
2004; pp73−101. (b) Olah, G. A.; Porter, R. D. J. Am. Chem. Soc. 1970,
92, 7627−7629. (c) Cram, D. J. J. Am. Chem. Soc. 1949, 71, 3863−3870.
(15) For a review, see: Tsuji, Y.; Richard, J. P. J. Phys. Org. Chem. 2016,
29, 557−564.
Experiments, characterization, and H and 13C NMR
spectra for all products 1aa-1an, 1ba-1bh, 1ca-1ja, 11, 14,
and the mixture of 15/16 (1:1) (PDF)
1
AUTHOR INFORMATION
Corresponding Authors
■
(16) (a) Rong, M. K.; van Duin, K.; van Dijk, T.; de Pater, J. J. M.;
Deelman, B.-J.; Nieger, M.; Ehlers, A. W.; Slootweg, J. C.; Lammertsma,
K. Organometallics 2017, 36, 1079−1090. (b) Liu, Y.; Yi, X.; Luo, X.; Xi,
C. J. Org. Chem. 2017, 82, 11391−11398. (c) Yan, X.; Zou, S.; Zhao, P.;
Xi, C. Chem. Commun. 2014, 50, 2775−2777. (d) Hodges, L. M.;
Gonzalez, J.; Koontz, J. I.; Myers, W. H.; Harman, W. D. J. Org. Chem.
1995, 60, 2125−2146. (e) Booth, B. L.; Jibodu, K. O.; Proenca, M. F. J.
Chem. Soc., Chem. Commun. 1980, 1151−1153.
*E-mail: yfh@mail.tsinghua.edu.cn.
*E-mail: wangxinyan@mail.tsinghua.edu.cn.
ORCID
Xinyan Wang: 0000-0002-7030-0418
Notes
(17) (a) Liu, S.; Zeng, X.; Hammond, G. B.; Xu, B. Adv. Synth. Catal.
2018, 360, 3667−3671. (b) Jin, L.; Hao, W.; Xu, J.; Sun, N.; Hu, B.;
Shen, Z.; Mo, W.; Hu, X. Chem. Commun. 2017, 53, 4124−4127.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the National Natural Science
Foundation of China (Nos. 21472107 and 21372142).
̌
́
(c) Seki, H.; Pellett, S.; Silhar, P.; Stowe, G. N.; Blanco, B.; Lardy, M. A.;
Johnson, E. A.; Janda, K. D. Bioorg. Med. Chem. 2014, 22, 1208−1217.
(d) Dang, H.; Mailig, M.; Lalic, G. Angew. Chem., Int. Ed. 2014, 53,
6473−6476. (e) Wang, Y.; Frattarelli, D. L.; Facchetti, A.; Cariati, E.;
Tordin, E.; Ugo, R.; Zuccaccia, C.; Macchioni, A.; Wegener, S. L.; Stern,
C. L.; Ratner, M. A.; Marks, T. J. J. Phys. Chem. C 2008, 112, 8005−
8015.
■
REFERENCES
■
(1) For selected recent references, see: (a) Zheng, B.; Trieu, T. H.; Li,
F.-L.; Zhu, X.-L.; He, Y.-G.; Fan, Q.-Q.; Shi, X.-X. ACS Omega 2018, 3,
8243−8252. (b) Tanabe, G.; Sugano, Y.; Shirato, M.; Sonoda, N.;
Tsutsui, N.; Morikawa, T.; Ninomiya, K.; Masayuki, N.; Yoshikawa, M.;
Muraoka, O. J. Nat. Prod. 2015, 78, 1536−1542. (c) Chaitanya, M.;
Yadagiri, D.; Anbarasan, P. Org. Lett. 2013, 15, 4960−4963. (d) Awuah,
E.; Capretta, A. J. Org. Chem. 2010, 75, 5627−5634.
(18) Booth, B. L.; Haszeldine, R. N.; Laali, K. J. Chem. Soc., Perkin
Trans. 1 1980, 2887−2893.
(19) (a) Larsen, R. D.; Reamer, R. A.; Corley, E. G.; Davis, P.;
Grabowski, E. J. J.; Reider, P. J.; Shinkai, I. J. Org. Chem. 1991, 56,
6034−6038. (b) Nagubandi, S.; Fodor, G. J. Heterocycl. Chem. 1980, 17,
1457−1463.
(2) For selected recent references, see: (a) Zhu, J.; Tan, H.; Yang, L.;
Dai, Z.; Zhu, L.; Ma, H.; Deng, Z.; Tian, Z.; Qu, X. ACS Catal. 2017, 7,
D
DOI: 10.1021/acs.orglett.9b00534
Org. Lett. XXXX, XXX, XXX−XXX