Regioselectivity in isoquinoline alkaloid synthesis

Article abstract of DOI:10.1016/j.tetlet.2010.01.115

Regioselectivity in isoquinoline alkaloid synthesis is analyzed here. Our experiments have shown that substituents on the aromatic ring of the starting amine are determinant in isoquinoline synthesis. The use of dicyclohexylcarbodiimide in activating carb

Full text of DOI:10.1016/j.tetlet.2010.01.115

Tetrahedron Letters 51 (2010) 1774–1778
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Regioselectivity in isoquinoline alkaloid synthesis
*
Rodolfo Quevedo , Edwin Baquero, Mario Rodriguez
Departamento de Química, Facultad de ciencias, Universidad Nacional de Colombia, Cra 30 N° 45-03 Bogotá, Colombia
a r t i c l e i n f o
a b s t r a c t
Article history:
Regioselectivity in isoquinoline alkaloid synthesis is analyzed here. Our experiments have shown that
substituents on the aromatic ring of the starting amine are determinant in isoquinoline synthesis. The
use of dicyclohexylcarbodiimide in activating carboxylic acids for electrophilic aromatic substitution
reactions is presented for the first time.
Received 24 September 2009
Revised 25 January 2010
Accepted 28 January 2010
Available online 4 February 2010
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
as a catalyst in the experimental conditions used here. These re-
sults showed that the Pictet–Spengler reaction occurred in polar
The Pictet–Spengler reaction involves adding a phenylethyl-
amine to an aldehyde to form an iminium ion; the resulting elec-
trophile is suitable for aromatic substitution and ring closure.
This methodology has been used in alkaloid synthesis as the key
step in tetrahydroisoquinoline formation.1 Such reaction is usually
carried out in an aprotic solvent in the presence of an acid catalyst
and proceeds most smoothly when the phenylethylamine aromatic
ring is activated by an electron-donating substituent. Poor regiose-
lectivity may be observed in some cases.2
It has been demonstrated recently that this reaction does not
occur when tyrosine or tyramine is involved,3,4 that regioselectiv-
ity in Pictet–Spengler cyclization for isoquinoline synthesis de-
pends on activating the aromatic ring of phenylethylamine and
that the least sterically hindered ortho position is the predominant
cyclization site.5 It has also been proposed that using some catalyst
and dehydrating agents has increased the yield and regioselectivity
and that Brønsted acids, such as trifluoromethanesulfonic acid,
acetic acid, and trifluoroacetic acid, have not been effective.6 No
relationship has been established between this reaction’s pH and
yield or regioselectivity.7,8
The dopamine reaction was carried out in this work with some
aromatic aldehydes to establish the importance of substituents on
the aromatic ring of the starting phenylethylamine during the
course of the Pictet–Spengler reaction when working with protic
solvents and Brønsted acids as catalyst.
The dopamine hydrobromide reaction was carried out with se-
lected aldehydes, methanol as solvent, and acetic acid as catalyst;
the respective isoquinoline hydrobromide was obtained in all cases
in high yields and with complete regioselectivity (Table 1); the for-
mation of any other regioisomer was not detected. Starting phen-
ylethylamine was recovered in all cases; this did not react.
Product formation was not observed when an acid was not used
solvents when the starting phenylethylamine’s aromatic ring had
a substituent in position 3 activating the ring. It should be stressed
that the existence of electron-donating and electron-withdrawing
groups on aldehydes did not have an effect on regioselectivity.
The presence of two strongly activated positions on the dopa-
mine ring favored the formation of two isoquinoline regioisomers;
however, only the product of electrophilic aromatic substitution on
carbon 6 of phenylethylamine was obtained in all cases in the con-
ditions used here (Table 1). Such complete regioselectivity sug-
gested that the activating group in position 3 of the aromatic
ring was necessary for the Pictet–Spengler reaction to occur and,
in turn, direct the course of the reaction toward aromatic electro-
philic substitution in position 6. The distribution of Mulliken atom-
ic charges9 from intermediary imine arising from the Pictet–
Spengler reaction showed greater electronic density on carbon 6
(ꢀ0.0559) than on carbon 2 (ꢀ0.0262) of the ring when calculated
**
using PC-GAMESS software, B3LYP/6-31G basis set and suggested
that the reaction was favored at this point (Fig. 1). Calculations for
Schiff bases from different aldehydes with electron-donating or
electron-withdrawing groups presented a similar pattern; differ-
ences regarding the distribution of the charge between C-2 and
C-6 did not change appreciably. These results correlated with those
of the isolated products.
The same methodology was used for the 6,7-dimethoxyphenyl-
ethylamine reaction with 3-nitrobenzaldehyde; this reaction did
not lead to the expected isoquinoline but to the respective inter-
mediary Schiff base 8 (Scheme 1).
Calculations for compound 8 revealed a Mulliken charge distri-
bution similar to that observed for the Schiff bases from the dopa-
mine reaction with aldehydes. Such results led to proposing that
the 6,7-dimethoxyphenylethylamine reaction did not occur be-
cause the activation energy for cyclization was greater for the
Schiff base with methoxy groups on the aromatic ring.
Schiff base 8 was refluxed in 37% HCl to overcome the energy
barrier and obtain 6,7-dimethoxyphenylethylamine-derived tetra-
* Corresponding author. Fax: +571 316 5220.
E-mail address: arquevedop@unal.edu.co (R. Quevedo).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.01.115

R. Quevedo et al. / Tetrahedron Letters 51 (2010) 1774–1778
1775
Table 1
quired for electrophilic aromatic substitution. The same as in ear-
lier cases, the reaction occurred with complete regioselectivity.
The results presented up to this point have shown that Pictet–
Spengler cyclization yield depended only on the starting alde-
hydes’ electrophilic capacity, that activating the phenylethylamine
aromatic ring was important in the course of the reaction and that
the high electron density on carbon 6 of dopamine favored substi-
tution at this point. We propose that using protic solvents (such as
methanol) increases cyclization regioselectivity due to greater
intermediary Schiff base solvation on phenol hydroxyls; such sol-
vation generates greater steric impediment and directs the reac-
tion toward carbon 6, this being the least sterically hindered due
to the presence of other substituents.
Dopamine reaction with aldehydes
HO
NH+3 -
Br
HO
O
CH3OH
CH3COOH
70ºC
H
R
6,7-Dimethoxy-1,4-dihydro-(2H)-isoquinoline-3-one 10 was
synthesized from 3,4-dimethoxybenzylamine (Scheme 3) to dem-
onstrate that isoquinoline synthesis regioselectivity depends only
on factors which can increase steric impediment regarding the
electrophile, that it is not exclusive to the Pictet–Spengler reaction
and that it does not depend on the catalyst used. The same as in
Pictet–Spengler cyclization, only one of the expected regioisomers
was formed. The steric impediment generated by the alkyl halide–
aluminum chloride complex could provide a possible explanation
for total regioselectivity in this alkylation, confirming that previ-
ously proposed for obtaining tetrahydroisoquinolines.
6,7-Dimethoxy-1,2-dihydroisoquinolin-3,4-dione 11 was syn-
thesized from 3,4-dimethoxybenzylamine to observe whether total
regioselectivity also occurred in acylation reactions (Scheme 4).
The same as in the previous reactions, the acylation reaction
regioselectively produced the respective isoquinoline and (unex-
pectedly) in a single step. The strategy used showed that N,N-dicy-
HO
NH+2
-
NH+2
-
Br
Br
HO
HO
R
HO
R
1- 7
R
Yield (%)
50
1
2
H–
Cl
78
3
90
O2N
clohexylcarbodiimide commonly used as
a carboxylic acid
activator for amide formation is also useful for activating carbox-
ylic acids in electrophilic aromatic substitution reactions on
strongly activated rings and (as in previous cases) cyclization
was completely regioselective.
H3CO
4
5
63
52
HO
HO
This work has thus presented a simple and economic methodol-
ogy which does not require the synthesis and purification methods
which are completely isoquinoline regioselective. It has been
shown that electrophilic aromatic substitution reactions for iso-
quinoline formation only occurred on carbon 6 of the starting phen-
ylalkylamine when solvent, catalyst, or activator conditions
increased steric impediment at reaction sites; such regioselectivity
was not restricted to the Pictet–Spengler reaction. Using dicyclo-
hexylcarbodiimide as a carboxylic acid activator for electrophilic
aromatic substitution reactions has also been presented for the first
time. Applying this strategy to isoquinoline compound synthesis is
currently being investigated and will be reported in due course.
6
7
38
35
H3CO
H
N
OH
HO
2. Experimental
2.1. General procedure for dopamine reaction with aldehydes
Acetic acid (0.1ꢁ mol) was added to a solution of dopamine
hydrobromide (ꢁ mol) and the respective aldehyde (ꢁ mol) in
methanol; the mixture was refluxed for 48 h. The precipitated
product was filtered and washed with methanol.
2.2. 6,7-Dihydroxy-1,2,3,4-tetrahydroisoquinoline
hydrobromide 1
Figure 1. Mulliken atomic charge distribution for imine.
(50%, white solid, mp 256–257 °C). 1H NMR (D2O, 400 MHz):
2.88 (2H, t, J = 6.4 Hz), 3.38 (2H, t, J = 6.4 Hz), 4.13 (2H, s), 6.61
(1H, s), 6.66 (1H, s). 13C NMR (D2O, 100 MHz): 23.8 (C-4), 41.7
(C-3), 44.0 (C-1), 113.7 (C-8), 115.8 (C-5), 119.6 (C-9), 123.7 (C-
10), 143.0 (C-7), 143.8 (C-6).
hydroisoquinolines 9 and little time was taken to precipitate the
respective isoquinoline (Scheme 2).10 This experiment demon-
strated that isoquinoline not being obtained with methoxy groups
in the initial conditions was due to the high activation energy re-

1776
R. Quevedo et al. / Tetrahedron Letters 51 (2010) 1774–1778
H3CO
H3CO
O
H
H3CO
H3CO
NH2
N
CH3OH
70ºC
+
CH3COOH
NO2
90 %
NO2
8
Scheme 1. 3,4-Dimethoxyphenylethylamine and 3-nitrobenzaldehyde reaction.
O
H3CO
H3CO
H3CO
H3CO
Cl
+
NH2
HO
N
DCC/DCM
NO2
8
Cl
O
H3CO
H3CO
HCl 37 %
reflux
NH
AlCl3/eter
H3CO
-
NH+2
-
NH+2
Cl
Cl
H3CO
H3CO
OCH3
OCH3
O
O
H3CO
H3CO
H3CO
92 %
NH
NH
NO2
NO2
9
10 (69 %)
Scheme 2. 1-(3-Nitrophenyl)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline
9
synthesis.
Scheme 3. 6,7-Dimethoxy-1,4-dihydro-(2H)-isoquinolin-3-one 10 synthesis.
2.3. 6,7-Dihydroxy-1-(4-chlorophenyl)-1,2,3,4-
tetrahydroisoquinoline hydrobromide 2
6.77 (1H, s). 6.98 (2H, d, J = 8.8 Hz), 7.22 (2H, d, J = 8.8 Hz). 13C
NMR (D2O, 100 MHz): 23.9 (C-4), 39.0 (C-3), 55,4 (CH3–O), 58.3
(C-1), 114.9 (C-8), 115.4 (C-5), 123.0 (C-10), 128.6 (C-9), 143.1
(C-7), 144.5 (C-6), 114.5 (C-13, C-15), 124.6 (C-11),132.3 (C-12,
C-16), 159.9 (C-14).
(78%, cream-colored solid, mp 232–234 °C) 1H NMR (D2O,
400 MHz): 3.01 (2H, m), 3.37 (2H, m), 5.60 (1H, s), 6.29 (1H, s),
6.76 (1H, s). 7.23 (2H, d, J = 8.0 Hz), 7.40 (2H, d, J = 8.0 Hz). 13C
NMR (D2O, 100 MHz): 23.9 (C-4), 39.2 (C-3), 58.1 (C-1), 114.9 (C-
8), 115.5 (C-5), 122.4 (C-10), 124.6 (C-9), 143.2 (C-7), 144.5 (C-6),
129.2 (C-12, C-16), 131.3 (C-13, C-15),134.7 (C-14), 135.3 (C-11).
2.6. 6,7-Dihydroxy-1-(4-hydroxyphenyl)-1,2,3,4-
tetrahydroisoquinoline hydrobromide 5
(52%, coffee-colored solid, mp 257–259 °C) 1H NMR (D2O,
400 MHz): 2.93 (2H, m), 3.30 (2H, m), 5.44 (1H, s), 6.23 (1H, s),
6.68 (1H, s). 6.78 (2H, d, J = 8.4 Hz), 7.06 (2H, d, J = 8.4 Hz). 13C
NMR (D2O, 100 MHz): 24.0 (C-4), 39.1 (C-3), 58.4 (C-1), 115.0
(C-8), 115.5 (C-5), 124.6 (C-10), 133.1 (C-9), 143.1 (C-7), 144.3
(C-6), 115.9 (C-13, C-15), 128.0 (C-11), 131.5 (C-12, C-16),156.8
(C-14).
2.4. 6,7-Dihydroxy-1-(3-nitrophenyl)-1,2,3,4-
tetrahydroisoquinoline hydrobromide 3
(90%, cream-colored solid, mp 259–260 °C) 1H NMR (D2O,
400 MHz): 3.00 (2H, m), 3.38 (2H, m), 5.72 (1H, s), 6.23 (1H, s),
6.73 (1H, s). 7.59 (1H, t, J = 8.0 Hz), 7.67 (1H, d, J = 8.0 Hz), 8.00
(1H, d, J = 8.0 Hz), 8.22 (1H, d, J = 8.0 Hz).
2.7. 6,7-Dihydroxy-1-(4-hydroxy-3-methoxyphenyl)-1,2,3,4-
tetrahydroisoquinoline hydrobromide 6
2.5. 6,7-Dihydroxy-1-(4-methoxyphenyl)-1,2,3,4-
tetrahydroisoquinoline hydrobromide 4
(38%, cream-colored solid, mp 230 °C) 1H NMR (D2O, 400 MHz):
3.02 (2H, m), 3.40 (2H, m), 5.53 (1H, s), 6.35 (1H, s), 6.78 (1H, s).
6.80 (1H, d, J = 8.0 Hz), 6.86 (1H, s), 6.90 (1H, d, J = 8.0 Hz).
(63%, white solid, mp 238–239 °C) 1H NMR (D2O, 400 MHz):
3.02 (2H, m), 3.37 (2H, m), 3.78 (3H, S), 5.57 (1H, s), 6.32 (1H, s),

R. Quevedo et al. / Tetrahedron Letters 51 (2010) 1774–1778
1777
O
H3CO
H3CO
OH
+
HO
NH2
O
2 DCC/DCM
O
OCH3
O
O
H3CO
O
H3CO
H3CO
11 (73 %)
NH
NH
OCH3
+
O
OCH3
H3CO
NH
12 (11 %)
NH
H3CO
O
Scheme 4. 6,7-Dimethoxy-1,2-dihydroisoquinoline-3,4-dione 11 synthesis
2.8. 6,7-Dihydroxy-1-[4-(6,7-dihydroxy-1,2,3,4-
tetrahydroisoquinoline-1-yl)phenyl]-1,2,3,4-
tetrahydroisoquinoline dihydrobromide 7
bodiimide (DCC) (1.24 g, 6 mmol) in 5 mL dichloromethane (DCM).
The resulting mixture was shaken at room temperature for 3 h and
filtered. The solvent was evaporated and the product extracted
with isopropyl acetate. The isopropyl acetate was evaporated off;
the crude product was dissolved in 5 mL ethyl ether and treated
with 0.03 g AlCl3. The resulting mixture was shaken for 6 h; once
this time had elapsed, the product was isolated and purified by sil-
ica gel column chromatography, eluted with isopropyl acetate.
(35%, cream-colored solid, mp 235–236 °C) 1H NMR (D2O,
400 MHz): 2.97 (4H, m), 3.30 (4H, m), 5.59 (2H, s), 6.15 (2H, s),
6.74 (2H, s), 7.15 (4H, s). 13C NMR (D2O, 100 MHz): 23.9 (C-4,C-
40), 39.0 (C-3, C-30), 58.1 (C-1, C-10), 115.0 (C-8, C-80), 115.6 (C-5,
C-50), 122.2 (C-10, C-100), 124.6 (C-9, C-90), 143.2 (C-7, C-70),
144.5 (C-6, C-60), 130,5(C-12, C-13, C-15, C-16).
2.13. 6,7-Dimethoxy-1,4-dihydro-(2H)-isoquinoline-3-one 10
2.9. Procedure for 6,7-dimethoxyphenylethylamine reaction
with 3-nitrobenzaldehyde
(0.43 g, 69%, white solid, mp 118–120 °C). 1H NMR (400 MHz,
CDCl3): 4.44 (2H, s), 3.89 (6H, s), 4.11 (2H, s), 6.81 (1H, s), 6.84
(1H, s). 13C NMR (100 MHz, CDCl3): 42.6 (C-4), 43.7 (C-1), 55.9
(CH3–O), 56.0 (CH3–O), 11.2 (C-8), 111.3 (C-5), 129.8 (C-9, C-10),
148.7 (C-7), 149.3 (C-6), 165.7 (C-3).
A solution of 6,7-dimethoxyphenylethylamine (ꢁ mol) and 3-
nitrobenzaldehyde (ꢁ mol) in methanol was refluxed for 72 h.
The solvent was evaporated and the crude product was treated
with 37% HCl; the resulting mixture was refluxed for 1 h. The pre-
cipitated product was filtered and washed with methanol.
2.14. Synthesis of 6,7-dimethoxy-(1H)-isoquinoline-3,4-dione
11 (Scheme 4)
2.10. N-[2-(3,4-Dimethoxyphenyl)ethyl]-N-[(3-nitrophenyl)
methylene]amine 8
3,4-Dimethoxybenzylamine (0.50 g, 3 mmol) was added to a
solution of oxalic acid (0.30 g, 3,3 mmol) and dicyclohexylcarbodi-
imide (DCC) (1.38 g, 6,7 mmol) in 5 mL dichloromethane (DCM).
The resulting mixture was shaken at room temperature for 3 h
and filtered; the DCM was evaporated. Product 11 was extracted
with isopropyl acetate; sub product 12 corresponded to an insolu-
ble solid in isopropyl acetate.
(90%, yellow solid, mp 91–92 °C) 1H NMR (CDCl3, 400 MHz): 3.00
(2H, t, J = 6.8 Hz), 3.91 (2H, t, J = 6.8 Hz), 6.76 (1H, s), 6.79 (1H, d,
J = 7.6 Hz). 6.81 (1H, d, J = 7.6 Hz), 7.60 (1H, t, J = 8.0 Hz), 8.03 (1H,
d, J = 8.0 Hz), 8.21 (1H, s), 8.27 (1H, d, J = 8.0 Hz), 8.57 (1H, s).
2.11. 6,7-Dihydroxy-1-(3-nitrophenyl)-1,2,3,4-
tetrahydroisoquinoline hydrobromide 9
2.15. 6,7-Dimethoxy-1,2-dihydroisoquinoline-3,4-dione 11
(0.49 g, 73%, white solid, mp 213–214 °C). 1H NMR (400 MHz,
CDCl3/CD3OD (5:5)): 3.88 (3H, s), 3.89 (3H, s), 4.34 (2H, d,
J = 5.68 Hz), 4.73 (1H, s), 6.82 (1H, s), 6.84 (1H, s). 13C NMR
(100 MHz, CDCl3/CD3OD (5:5)): 43.6 (C-1), 55.2 (CH3–O), 55.4
(CH3–O), 110.9 (C-5), 111.5 (C-8), 119.4 (C-10), 132.7 (C-9), 148.0
(C-6), 149.0 (C-7), 158.5 (C-3, C-4).
(92%, cream-colored solid, mp 258–260 °C) 1H NMR (D2O,
400 MHz): 3.22 (2H, m), 3.51 (2H, m), 3.63 (3H,s), 3.91 (3H,s),
5.94 (1H, s), 6.44 (1H, s), 7.03 (1H, s). 7.70 (1H, t, J = 8.0 Hz), 7.82
(1H, d, J = 8.0 Hz), 8.19 (1H, s), 8.40 (1H, d, J = 8.0 Hz).
2.12. Synthesis of 6,7-dimethoxy-1,4-dihydro-(2H)-
isoquinoline-3-one 10 (Scheme 3)
2.16. N,N0-Bis(3,4-dimethoxybenzylethanodiamide) 12
3,4-Dimethoxybenzylamine (0.50 g, 3 mmol) was added to a
solution of chloroacetic acid (0.57 g, 6 mmol) and dicyclohexylcar-
(0.06 g, 11%, white solid, mp 160–162 °C). 1H NMR (400 MHz,
CD3OD): 3.83 (3H, s), 3.84 (3H, s), 4.28 (2H, s), 4.65 (2H, s), 6.84

1778
R. Quevedo et al. / Tetrahedron Letters 51 (2010) 1774–1778
(2H, d, J = 8.0 Hz), 6.86 (2H, s), 6.89 (2H, d, J = 8 Hz). 13C NMR
(101 MHz, CD3OD): 43.7 (C-Ph), 55.9 (CH3–O), 111.2 (C-2), 111.3
(C-5), 119.4 (C-6), 129.8 (C-1), 148.7 (C-4), 149.3 (C-3), 165.7
(CO).
References and notes
1. Cox, E. D.; Cook, J. M. Chem. Rev. 1995, 95, 1797–1842.
2. Sperlinga, E.; Kosson, P.; Urbanczyk-Lipkowska, Z.; Ronsisvalle, G.; Carrc, D.;
Lipkowskia, A. Bioorg. Med. Chem. Lett. 2005, 15, 2467–2469.
3. Quevedo, R.; Moreno, B. Tetrahedron Lett. 2009, 50, 936–938.
4. Quevedo, R.; Ortiz, I.; Reyes, A. Tetrahedron Lett. 2010, 51, 1216–1219.
5. Cho, S.; Song, S.; Hur, E.; Chen, M.; Joo, W.; Falck, J.; Yoon, Y.; Shin, D.
Tetrahedron Lett. 2001, 42, 6251–6253.
Acknowledgments
6. Manabe, K.; Nobutou, D.; Kobayashi, S. Bioorg. Med. Chem. Lett. 2005, 13, 5154–
5158.
7. Bates, H.; Bagheri, K.; Vertino, P. J. Org. Chem. 1986, 51, 3061–3063.
8. Vanden Eynden, M.; Stambuli, J. Org. Lett. 2008, 10, 5289–5291.
9. Saleh, B.; Essa, A.; Al-Shawi, S.; Jalbout, A. J. Mol. Struct.: THEOCHEM 2009, 909,
107–110.
We wish to acknowledge the Universidad Nacional de Colombia
for financing this work (DIB research project No. 8003424) and the
Chemistry Department for providing the laboratory facilities. We
would also like to thank Mr. Jason Garry for carefully proofreading
this Letter.
10. Zhong, M.; Villani, F.; Marzouq, R. Org. Proc. Res. Dev. 2007, 11, 463–465.