Silicon-assisted ethoxycarbonylmethylation of N-methylquinolinium and isoquinolinium iodides

Article abstract of DOI:10.1002/1099-0690(200008)2000:16<2915::AID-EJOC2915>3.0.CO;2-B

A new regioselective route to 2-ethoxycarbonylmethyl-1,2-dihydro-N-methylquinolines and 1-ethoxycarbonylmethyl-1,2-dihydro-N-methylisoquinolines is described starting from methylquinolinium or -isoquinolinium iodides and commercially available ethyltrimethylsilyl acetate (ETSA). The methylene carbanion was generated by fluorodesilylation using caesium fluoride. On exposure to air, the ethoxycarbonyl-methyl adducts were oxidised, leading to the corresponding alkylidene derivatives 4 and 5, whereas 2a in solution led slowly to its regioisomer 6a.

Full text of DOI:10.1002/1099-0690(200008)2000:16<2915::AID-EJOC2915>3.0.CO;2-B

FULL PAPER
Silicon-Assisted Ethoxycarbonylmethylation of N-Methylquinolinium and
Isoquinolinium Iodides
[a]
[a]
[a]
[a]
¨
Faıza Diaba, Cyril Le Houerou, Micheline Grignon-Dubois,* Bernadette Rezzonico,
and Pierre Gerval[a]
Keywords: Alkylations / Quinolinium salts / ETSA / Nucleophilic additions
A new regioselective route to 2-ethoxycarbonylmethyl-1,2-
dihydro-N-methylquinolines and 1-ethoxycarbonylmethyl-
methylene carbanion was generated by fluorodesilylation us-
ing caesium fluoride. On exposure to air, the ethoxycarbonyl-
1,2-dihydro-N-methylisoquinolines is described starting from methyl adducts were oxidised, leading to the corresponding
methylquinolinium or -isoquinolinium iodides and commer- alkylidene derivatives 4 and 5, whereas 2a in solution led
cially available ethyltrimethylsilyl acetate (ETSA). The
slowly to its regioisomer 6a.
Introduction
ylmethyl-1,2-dihydro-N-methyl-quinolines and -isoquinol-
ines starting from methylquinolinium or -isoquinolinium
iodides (1, 7) and commercially available ETSA in the pres-
ence of fluoride.[6]
1,2-Disubstituted-1,2-dihydroquinolines have aroused in-
creasing interest with regard to their pharmacological activ-
ity.[1] As part of a program directed toward the synthesis of
new drugs from functionalized heterocyclic derivatives, we
needed to introduce the ethoxycarbonylmethyl group at the
C-2 -position of the quinoline ring. Ethoxycarbonylmethyl-
ation of quinolinium derivatives has previously been
achieved by treating ethyl(tributylstannyl)acetate with quin-
oline and methyl chloroformate.[2] However, organostannyl
reagents are not suitable for drug synthesis. Indeed, they
are toxic, polluting, and often difficult to get rid of. Fur-
thermore, this process seems to be limited to quinolines ac-
tivated by alkyl chloroformate, whereas we needed an alkyl
group on the nitrogen atom. In contrast to organostannyl
reagents, organosilicon derivatives are well-known for their
ability to functionalise without side-effects of pollution or
toxicity. In the case of 1-methylquinoliniums, Fukuzumi et
al.[3] obtained the C-2- and/or C-4-alkoxycarbonylmethy-
lene adducts when using a large excess of ketene silyl acetal,
but this reagent is moisture-sensitive and has to be gener-
ated. Looking for a more manageable source of the alkyl
acetate anion equivalent, we turned to ethyl trimethylsilyl
acetate (ETSA), which is a stable, commercially available,
useful reagent.[4] In particular, it has been shown to add to
the carbonyl double bond in the presence of fluoride, lead-
ing to silyl-Reformatsky products.[5] In contrast, to the best
of our knowledge, it has never been used to introduce the
ethoxycarbonylmethyl moiety in the heteroaromatic series.
We report here a new regioselective route to ethoxycarbon-
Results and Discussion
As expected, ethyltrimethylsilyl acetate (ETSA) alone did
not react with compound 1a.[7] The same reaction con-
ducted with 1 equivalent of sodium hydroxide led to 2a, but
in only 8% yield, along with the 1-methyl-2-quinolone
(34%). In contrast, when 1a was reacted with ETSA in the
presence of dried alkali metal fluoride in acetonitrile solu-
tion at reflux, the nucleophilic addition of the methylene
anion was systematically observed, leading to 2a
(Scheme 1).[8] The reaction also works at room temperature,
but needs 5 h to go to completion instead of 2 h. As ex-
pected on the basis of their respective nucleophilic power,
caesium fluoride led to a better yield than potassium fluor-
ide (87 and 40%, respectively). Attempts to replace aceton-
itrile by dichloromethane or THF and alkali metal fluoride
by tetrabutylammonium fluoride were unsuccessful. These
observations show the importance of using a dissociating
solvent[9] and are consistent with a charge-controlled pro-
cess. The reactions were also conducted with the methyl-
quinoliniums 1b؊1e. We previously observed a dramatic ef-
fect of ultrasound upon the regiochemistry of nucleophilic
addition to 2-Me- and 3-Me-quinoliniums,[10] leading us to
systematically study the influence of the activation mode
(reflux or sonication at 10 °C). Examination of Table 1
shows that the experimental conditions here have very weak
effects on the regiochemistry of the addition, except with
1e. The C-2 adducts 2c and 2d were isolated as the only
product, while 2b and 2e were obtained in mixtures with
the C-4 regioisomer 3b and 3e, respectively. Formation of
3b is probably due to competition during the addition step
rather than to a C-2 Ǟ C-4 isomerization. Indeed, the iso-
mer ratio was practically the same when working at reflux
[a]
´
Laboratoire de Chimie Organique et Organometallique, CNRS
´
UMR 5802, Universite Bordeaux I
´
351, Cours de la Liberation, F-33405 Talence-Cedex, France
[b]
´ ´
Laboratoire de Chimie des Substances Vegetales Ϫ Institut du
´
Pin, Universite Bordeaux I
´
351, Cours de la Liberation, F-33405 Talence-Cedex, France
Fax: (internat.) ϩ 33-5/56846422
E-mail: m.grignon@ipin.u-bordeaux.fr
Eur. J. Org. Chem. 2000, 2915Ϫ2921
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1434Ϫ193X/00/0816Ϫ2915 $ 17.50ϩ.50/0
2915

F. Diaba, C. Le Houerou, M. Grignon-Dubois, B. Rezzonico, P. Gerval
FULL PAPER
or under sonochemical activation (84:16 and 82:18, respect- under ultrasound (2e/3e 71:29) and 56% yield under reflux
ively). Analogously, the 2,3-benzoquinolinium 1f and iso- conditions (2e/3e 79:21), whereas only traces of quinolone
quinolinium derivatives 7a/7b led respectively to 3f and 8a/ were detected. This particular behaviour of the 3-bromoqui-
8b. With these substrates, yields vary from 65 to 77% and nolinium could be related to the fact that 1e exhibits the
are systematically higher at reflux (Table 1). The regiochem- larger positive charge on C-2 in the whole series
ical assignments were based on NMR chemical shifts and (Table 2).[12a] Reacting 1e with caesium fluoride in aceton-
coupling constant values and on 1H, 1H and HMBC experi- itrile at reflux led to small amount of bromoquinolone (7%
ments.
after 3 hours), which could result in the decomposition of
a fluoro C-2 adduct.[12b] Another possible fluoride source
could be the CsF/ETSA complex itself (Scheme 2). Indeed,
it has recently been shown that trimethylsilylacetonitrile re-
acts with tetramethylammonium fluoride in acetonitrile to
form a pentacoordinate silicon complex, which can act as a
source of either fluoride or cyanomethylcarbanion, de-
pending on the substrate.[13] Moreover, the authors also re-
ported that CsF, but not KF, led to pentacoordinate com-
plex when reacting with trimethylsilyltrifluoromethane in
acetonitrile.[13] Our results are in good agreement with
these findings.
During the purification step, an additional more polar
product was detected by TLC. It was isolated in low yields
ranging from 2Ϫ7% by alumina column chromatography
and identified as the N-methyl-2-ethoxycarbonylmethylid-
ene-dihydroquinolines 4, 5 (Figure 1). Their structure as-
signment was based on NMR, SM and elemental analysis.
Inspection of the chemical shift data allows a clear under-
standing of the regiochemistry of these products. It is worth
noting that in each case, the H-9 vinylic proton gives rise
to a shielded singlet, sensitive to the regiochemistry (δ ϭ
4.73Ϫ5.02 in 4; 5.60 in 5b). The same is true for the C-9
methine carbon (δ ϭ 81.9Ϫ83.8 in 4; 87.6 in 5b). Com-
pounds 4 and 5, are composed of a single isomer, showing
that the elimination process is stereospecific. The E/Z ste-
reochemistry of the exocyclic double bond was assigned
with the aid of 2D-NOESY experiments, showing the spa-
tial connectivity depicted in Figure 1. The E configuration
was confirmed by the presence of a strong H-9/NMe cross-
peak in compounds 4a, 4c, 4d and a H-5/H-9 cross-peak in
5b. In this last case, this assignment was confirmed by the
comparison of the 1H and 13C chemical shifts, in particular
for H-5 and H-9 (δ ϭ 7.94 and 5.60) with those described
for the (E)-4-ethoxycarbonylmethylidene-1,4-dihydro-1-
methylquinoline (δ ϭ 7.97 and 5.69).[14] From these results,
it seems that the elimination process is governed by steric
factors, the ethoxycarbonyl group lying away from the bulk-
ier substituent, namely the N-methyl group in 4 and the
aromatic ring in 5b.
Scheme 1
Table 1. Ethoxycarbonylmethylation of quinolinium iodides with
ETSA/MF/CH3CN
% of the crude Yield [%]
Substrate MF Method[a] Time [h] 2:3 [%]
2
3
1a
KF
B
3
100:0
100:0
100:0
84:16
82:18
100:0
100:0
100:0
100:0
82:18
71:29
79:21
0:100
0:100
2a, 40 Ϫ
2a, 87 Ϫ
2a, 84 Ϫ
2b, 51 3b, 10
2b, 58 3b, 14
2c, 81 Ϫ
2c, 73 Ϫ
2d, 71 Ϫ
2d, 67 Ϫ
2e, 9 3e, 2
2e, 19 3e, 8
2e, 44 3e,12
3f, 65
CsF A
CsF B
CsF A
CsF B
CsF A
CsF B
CsF A
CsF B
CsF A
2
1
1b
1c
1d
1e
2
2
2
3
1.5
3
3
KF
KF
CsF A
CsF B
A
B
3
3
1f
3
3
3f, 72
7a
7b
CsF A
CsF B
CsF A
CsF B
3
1
3
3
8a, 67
8a, 70
8b, 60
8b, 77
Spontaneous oxidation in air has been previously re-
ported by Leonard et al. in the case of dihydroquinoline
products arising from the reaction of quinolinium with
methylene-activated reagents such as nitromethane, ma-
[a]
A: sonication at 10 °C; B: stirring at reflux.
The dramatic effect of both the fluoride and the activa- lononitrile or ethylcyanoacetate.[15] This led us to examine
tion mode observed with the 3-bromoquinolinium 1e (Table the stability of compounds 2 and 3 in air. On exposure to
1) is worthy of note. Indeed, under sonochemical activation, air, all adducts led to oxidation products, except the acrid-
CsF led to the 3-bromo-N-methylquinol-2-one as the major ine (3f) and phenanthridine (8b) derivatives. However, the
product (29%)[11] along with the two ethoxycarbonylmethyl process was very slow, except in the case of 2a for which the
regioisomers (11% yield; 2e/3e 82:18). Replacing CsF by conversion rate reached 100% after ten days. In contrast, 2a
KF allowed us to obtain the expected adducts in 27% yield is stable towards caesium fluoride in acetonitrile solution,
2916
Eur. J. Org. Chem. 2000, 2915Ϫ2921

Ethoxycarbonylmethylation of N-Methylquinolinium- and Isoquinolinium Iodides
FULL PAPER
Table 2. C-2/C-4 Atomic charges in quinolinium ions and heat of formation of the C-2/C-4 adducts
R
Atomic charges in quinolinium
Heat of formation of the adducts [kcal·molϪ1
]
CϪ2
CϪ4
CϪ2
CϪ4
H
0.189
0.174
0.195
0.186
0.215
0.141
0.135
0.145
0.117
0.169
Ϫ63.5
Ϫ69.7
Ϫ72.5
Ϫ72.5
Ϫ55.8
Ϫ73
2-Me
3-Me
4-Me
3-Br
Ϫ79.7
Ϫ81.6
Ϫ79.2
Ϫ63.4
dition to the strong cross peak H-9/NMe, the correlation
map also reveals an H-3/Oet cross-peak. This slow but
quantitative 2a Ǟ 6a rearrangement was quite unexpected.
It could be the result of two successive hydride shifts by
[1,3] sigmatropic migrations of hydrogen or via the forma-
tion of a quinolinium-anion ion pair (Scheme 3). We first
used a frontier orbital approach to choose between these
two hypotheses. A [1,3] sigmatropic rearrangement can be
seen as occurring via an imaginary transition state, which
consists of a hydrogen atom and an allyl radical. AM1 cal-
culations of the quinolinyl radical led to predict for the first
step an antarafacial migration, which should be a thermal
process. However, 2a was recovered unchanged after 7 days
in dichloromethane or acetonitrile at reflux. Toluene was
then used as the solvent, and the reaction was monitored
by GPC. After 7 days at reflux, only 11% of 6a was present
along with 14% of 4a. After 21 days the composition of the
reaction mixture was: 2a (27%), 4a (38%) and 6a (33%).
These results show that increasing the temperature does not
dramatically accelerate the conversion, but instead leads to
the loss of selectivity. Moreover, CCM analysis reveals the
presence of tars.
Scheme 2
Figure 1. Stereochemistry for compounds 4a, c, d and 5b
and is recovered unchanged even after two days. An attempt
to dehydrogenate 2a with DDQ (1 equiv. in acetonitrile so-
lution) was unsuccessful. After 2 h at reflux, 2a remained
unchanged, whereas 2 equiv. of DDQ led to its complete
Scheme 3
From a mechanistic point of view, these results are con-
degradation. In contrast, KMnO4 in acetone allowed oxida- sistent with the formation of the ethoxycarbonylmethylene
tion of 2a at room temperature, but led to a 73:27 mixture anion by nucleophilic attack on the silicon, and its addition
of 4a and N-methylquinol-2-one (76% global yield). In ad- to the C-2 position, which is the most electrophilic site
dition, the evolution of 2a was monitored by GPC and (Table 2).[12] The regioselectivity of the reaction is remark-
NMR in CDCl3 solution. After 10 days, we only observed able. Indeed, the nucleophilic addition to quinolinium salts
traces of 4a, but surprisingly, its concentration did not in- is reported to be dependent on substituent effects as well as
crease and a new product identified as 6a was formed. Sev- on the nature of the nucleophilic reagent, leading to com-
enty days were necessary for the total conversion 2a Ǟ 6a. petition between C-2 and C-4 additions.[16,17] It is worth
No incorporation of deuterium was detected in 6a, clearly noting that we never observed isomerisation of the C-2 ad-
obviating the possibility of the solvent participating as a ducts during the reaction. Furthermore, comparison of the
reducing agent. A 2D-NOESY experiment showed an E results obtained at reflux and under sonochemical activa-
configuration of the exocyclic double bond as in 4a. In ad- tion shows that the activation process has no influence
Eur. J. Org. Chem. 2000, 2915Ϫ2921
2917

F. Diaba, C. Le Houerou, M. Grignon-Dubois, B. Rezzonico, P. Gerval
FULL PAPER
rich) was dried over molecular sieves. Commercial caesium- and
potassium fluoride (Aldrich, reagent ACS) were dried prior to use,
in a domestic microwave oven. Methiodides 1 and 6 were prepared
according to literature procedures.[19] by alkylation of the quinoline
with methyl iodide in acetone solution. The mixture was main-
tained at room temperature until the substrate had reacted com-
pletely, as monitored by TLC (SiO2; Et2O/CH2Cl2, 30:70 v/v). The
precipitated salts were filtered off and crystallized. Data for these
compounds are identical to those in the literature.[19]
upon the reaction outcome, as shown in Table 1. We previ-
ously described the trichloromethylation and acetonylation
of quinolinium moieties.[10] The C-2 adduct was obtained
as the only product under sonochemical activation, while
the C-4 adduct was generally isolated at reflux. This behavi-
our was interpreted as a thermodynamic process occurring
through the heterolytic dissociation of the C2ϪC9 bond in
the C-2 adducts, leading to an anionϪquinolinium cation
pair (Qϩ, RϪ). Calculation of the heats of formation of the
ethoxycarbonylmethylene derivatives showed that the C-4
regioisomer is the most stable (Table 2).[18] The magnitude
of this difference in energy is in the same range as for the
acetonyl derivatives (9.0 kcal.molϪ1), and greater than for
the trichloromethyl derivatives (3.2 kcal.molϪ1), which both
led to the C-2 Ǟ C-4 isomerisation. MNDO as well as AM1
calculations of the atomic charges for the C-2 adducts led
General Procedure: To a stirred solution of quinolinium iodide
(7.4 mmol) in acetonitrile (20 mL) was added fluoride (KF or CsF:
8.1 mmol) and ETSA (1.3 g, 8.1 mmol). The mixture was then
stirred at reflux or sonicated at 10 °C until the starting material
was completely consumed as monitored by TLC (SiO2, MeOH-
Me2CO, 10:90). The reaction mixture was filtered through Celite
and the filtrate was evaporated to dryness. The residue was taken
up in cyclohexane (100 mL) and insoluble materials, if present,
to a very similar electron distribution, regardless of the na- were removed by filtration and analyzed separately. The solution
was evaporated to dryness, leading to an oil which was purified
by chromatography, eluting typically with CH2Cl2 (Al2O3, activity
IIϪIII, 70Ϫ230 mesh). Yields are reported in Table 1.
ture of the C-2 substituent, showing that the thermal
stability of compounds 2aϪ2d is not governed by electronic
effects. These results show the poor migrating ability of the
ethoxycarbonylmethyl group relative to the trichloromethyl
or the acetonyl group, and could explain the formation of
4a or 6a through a common quinoliniumϪanion ion pair
2-Ethoxycarbonylmethyl-1,2-dihydro-1-methylquinoline (2a): Oil. Ϫ
Rf (SiO2, CH2Cl2) ϭ 0.82. Ϫ 1H NMR (250 MHz, CDCl3): δ ϭ
1.22 (3 H, t, J ϭ 7.1, CH2CH3), 2.44 and 2.56 (2 H, AB part of
resulting from the heterolytic dissociation of the C2ϪH ABX, J ϭ 14.4, 7.6 and 5.3, COCH2), 2.91 (3 H, s, NCH3), 4.05
and 4.12 (2 H, AB, J ϭ 10.7 and 7.1, CH3CH2), 4.47 (1 H, ddd,
J ϭ 7.6, 5.5 and 5.3, H-2), 5.78 (1 H, dd, J ϭ 9.5 and 5.5, H-3),
6.44 (1 H, d, J ϭ 9.5, H-4), 6.48 (1 H, d, J ϭ 8.1, H-8), 6.66 (1 H,
dd, J ϭ 7.3 and 7.2, H-6), 6.93 (1 H, dd, J ϭ 7.2, and 1.5, H-5),
7.12 (1 H, ddd, J ϭ 8.1, 7.3 and 1.5, H-7). Ϫ 13C NMR
(62.89 MHz, CDCl3): δ ϭ 14.2 (CH2CH3), 36.5 (NCH3), 38.0 (CO
CH2), 57.6 (C-2), 60.6 (CH3CH2), 111.0, 117.1, 123.8, 126.3, 126.9,
129.2 (CHsp2), 121.8, 144.2 (Cq), 171.4 (CO). Ϫ MS; m/z: 231 (Mϩ,
6.4), 144 (M Ϫ CH2COOEt, 100). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1630
(CϭC), 1725 (CϭO). Ϫ C14H17NO2.(231.3): calcd. C 72.70, H 7.41,
bond (Scheme 3).
Conclusion
ETSA has been shown to allow ethoxycarbonylmethyl-
ation of quinolines and isoquinolines. All the products de-
scribed here are new, except for 2a and 2d, which have been
previously isolated when treating 1a and 1d with the ketene N 6.06, O 13.83; found C 72.55, H 7.62, N 6.09, O 13.36.
triethylsilyl acetal of ethyl acetate, which is moisture-sensit-
2-Ethoxycarbonylmethyl-1,2-dihydro-1,2-dimethylquinoline (2b): Oil.
Ϫ Rf (SiO2, CH2Cl2) ϭ 0.69. Ϫ H NMR (250 MHz, CDCl3): δ ϭ
1.09 (3 H, t, J ϭ 7.1, CH2CH3), 1.56 (s, 3 H, CH3Ϫ2), 2.32 and
2.78 (2 H, AB syst., d, J ϭ 13.2, COCH2), 2.86 (3 H, s, NCH3),
3.92 and 3.96 (2 H, AB part of ABX3, J ϭ 10.7 and 7.1, CH3CH2),
5.52 (1 H, d, J ϭ 9.8, H-3), 6.36 (1 H, d, J ϭ 9.8, H-4), 6.50 (1 H,
ive and has to be generated.[3] Only the NMR yield was
1
reported, and the experimental protocol was not described.
Furthermore, 10 equivalents of the silyl reagent were neces-
sary for the reaction to occur, and it did not work with 2-
methylquinolinium. Our process, which is easy to handle
and only needs 1 equivalent of a stable and commercially d, J ϭ 8.2, H-8), 6.60 (1 H, dd, J ϭ 7.3 and 7.2, H-6), 6.86 (1 H,
dd, J ϭ 7.2, and 1.4, H-5), 7.08 (1 H, ddd, J ϭ 8.2, 7.3 and 1.4,
H-7). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.0 (CH2CH3), 30.9
(NCH3), 28.3 (CH3Ϫ2), 44.5 (CO CH2), 59.0 (C-2), 60.4
(CH3CH2), 110.3, 116.7, 125.0, 126.7, 128.8, 129.2 (CHsp2), 121.1,
144.8 (Cq), 170.8 (CO). Ϫ MS; m/z: 245 (Mϩ, 4.4), 158 (M Ϫ
CH2COOEt, 100). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1650 (CϭC), 1723
(CϭO). Ϫ C15H19NO2 (245.3): calcd. C 73.44, H 7.81, N 5.71, O
13.04; found C 73.36, H 8.01, N 5.43, O 13.11.
available silyl reagent, constitutes a major improvement for
synthesising these derivatives.
Experimental Section
General Remarks: Flash column chromatography techniques
(30 cm ϫ 2 cm column) were employed to purify crude products
using 70Ϫ230 mesh alumina (activity II-III, CH2Cl2) under posit-
ive air pressure. Ultrasound-promoted reactions were carried out
in a common ultrasonic laboratory cleaner filled with water and
2-Ethoxycarbonylmethyl-1,2-dihydro-1,3-dimethylquinoline (2c): Oil.
1
Ϫ Rf (SiO2, CH2Cl2) ϭ 0.65. Ϫ H NMR (250 MHz, CDCl3): δ ϭ
1.18 (3 H, t, J ϭ 7.1, CH2CH3), 1.88 (d, 3 H, J ϭ 1.2, CH3Ϫ2),
2.34 and 2.50 (2 H, AB part of ABX, J ϭ 14.3, 4.9 and 6.3,
maintained at a temperature of 0Ϫ5 C°. The reaction flask was COCH2), 2.92 (3 H, s, NCH3), 4.00 and 4.03 (2 H, AB part of
partially submerged in the sonicator water bath in a place that pro-
duced maximum agitation. 1H and 13C NMR spectra were recorded
at 250 and 63 MHz, respectively, with TMS as internal standard.
Elemental analyses were performed by Service Central d’Analyses
du CNRS (F-69390 Vernaison). Synthesis grade acetonitrile (Ald-
ABX3, J ϭ 10.7 and 7.1, CH3CH2), 4.29 (1 H, dd, J ϭ 6.3, 4.9, H-
2), 6.20 (1 H, q, J ϭ 1.2, H-4), 6.49 (1 H, dd, J ϭ 8.1, 0.6, H-8),
6.66 (1 H, ddd, J ϭ 7.4, 7.3 and 0.6, H-6), 6.89 (1 H, dd, J ϭ 7.3
and 1.3, H-5), 7.08 (1 H, ddd, J ϭ 8.1, 7.4 and 1.3, H-7). Ϫ 13C
NMR (62.89 MHz, CDCl3): δ ϭ 14.0 (CH2CH3), 20.8 (CH3Ϫ3),
2918
Eur. J. Org. Chem. 2000, 2915Ϫ2921

Ethoxycarbonylmethylation of N-Methylquinolinium- and Isoquinolinium Iodides
36.1 (CO CH2), 36.7 (NCH3), 60.7 (CH3CH2), 62.3 (C-2), 111.0, H, s, NCH3), 4.15 (2 H, q, J ϭ 7.1, CH3CH2), 4.61 (1 H, t, J ϭ
FULL PAPER
117.3, 122.2, 125.9, 128.2 (CHsp2), 122.8, 132.5, 142.5 (Cq), 172.1
7.6, H-9), 6.98Ϫ7.08 (4 H, m), 7.29Ϫ7.38(4 H, m). Ϫ 13C NMR
(CO). Ϫ MS; m/z: 245 (Mϩ, 1.9), 158 (M Ϫ CH2COOEt, 100). Ϫ (62.89 MHz, CDCl3): δ ϭ 14.3 (CH2CH3), 33.06 (NCH3), 40.86 (C-
IR (neat) ν˜max/cmϪ1 ϭ 1630 (CϭC), 1725 (CϭO). Ϫ C15H19NO2 9), 42.1 (CO CH2), 60.4 (CH3CH2), 112.4, 120.9, 127.5, 128.1
(245.3): calcd. C 73.44, H 7.81, N 5.71, O 13.04; found C 73.30, H
7.76, N 5.59, O 13.23.
(CHsp2), 126.3, 142.6 (Cq), 171.7 (CO). Ϫ IR (neat) ν˜max/cmϪ1
ϭ
1594 (CϭC), 1729 (CϭO). Ϫ C18H19NO2 (281.4): calcd. C 76.84, H
6.81, N 4.98, O 11.37; found C 76.45, H 7.01, N 4.83, O 11.58.
2-Ethoxycarbonylmethyl-1,2-dihydro-1,4-dimethylquinoline (2d): Oil.
1
Ϫ Rf (SiO2, CH2Cl2) ϭ 0.78. Ϫ H NMR (250 MHz, CDCl3): δ ϭ (E)-2-Ethoxycarbonylmethylidene-1,2-dihydro-1-methylquinoline
1.22 (3 H, t, J ϭ 7.1, CH2CH3), 2.04 (d, 3 H, J ϭ 1.4, CH3Ϫ4), (4a). ؊ By Aerobic Oxidation of 2a: A solution of 2a (500 mg,
2.39 and 2.53 (2 H, AB part of ABX, J ϭ 14.4, 7.5 and 5.3,
COCH2), 2.91 (3 H, s, NCH3), 4.04 and 4.12 (2 H, AB part of
2.2 mmol) in dichloromethane (10 mL) was deposited in a Petri
dish (6 cm diameter), which was kept at room temperature open
ABX3, J ϭ 10.9 and 7.1, CH3CH2), 4.40 (1 H, dddd, J ϭ 7.5, 5.8, to air. Slow evaporation of the solvent led to a thin film of 2a.
5.3 and 1.9, H-2), 5.65 (1 H, d, J ϭ 5.8, H-3), 6.50 (1 H, dd, J ϭ
7.8, 1.0, H-8), 6.71 (1 H, ddd, J ϭ 7.6, 7.5 and 1.0, H-6), 7.12 (1
H, dd, J ϭ 7.5 and 1.5, H-5), 7.14 (1 H, ddd, J ϭ 7.8, 7.6 and 1.5,
Crystallization occurred slowly owing to the formation of 4a. The
oxidation was completed within 10 days, affording pale yellow crys-
tals. NMR analysis of the crude product showed the formation of
H-7). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.2 (CH2CH3), 18.8 4a as the only product. It was purified by chromatography on silica
(CH3Ϫ4), 37.7 (CO CH2), 36.6 (NCH3), 60.5 (CH3CH2), 57.4 (C-
2), 111.0, 116.8, 121.4, 123.7, 128.9 (CHsp2), 123.3, 131.1, 144.1
(Cq), 171.6 (CO). Ϫ MS; m/z: 245 (Mϩ, 2.7), 158 (M Ϫ
CH2COOEt, 100). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1650 (CϭC), 1727
(CϭO). Ϫ C15H19NO2 (245.3): calcd. C 73.44, H 7.81, N 5.71, O
13.04; found C 73.55, H 7.69, N 5.56, O 13.08.
gel (CH2Cl2/EtOH 95:5) leading to 4a as a pale yellow solid
(435 mg, 87%).
By Oxidation with KMnO4: KMnO4 (300 mg, 1.8 mmol) was added
to a stirred solution of 2a (500 mg, 2.2 mmol) in acetone (10 mL).
Stirring was continued at room temperature for 18 h. Filtration of
the reaction mixture through Celite followed by workup identical
to the procedure described above gave 4a (280 mg, 56%) and N-
3-Bromo-2-ethoxycarbonylmethyl-1,2-dihydro-1-methylquinoline
(2e): Oil. Ϫ Rf (SiO2, CH2Cl2) ϭ 0.65. Ϫ 1H NMR (250 MHz, methylquinol-2-one (72 mg, 20%).[11]
CDCl3): δ ϭ 1.20 (3 H, t, J ϭ 7.1, CH2CH3), 2.50 (m, 2 H,
1
4a: Rf (SiO2, CH2Cl2) ϭ 0.25; m.p. 146 °C (C6H12/CH2Cl2). Ϫ H
NMR (250 MHz, CDCl3): δ ϭ 1.30 (3 H, t, J ϭ 7.1, CH2CH3),
3.42 (3 H, s, NCH3), 4.15 (2 H, q, J ϭ 7.1, CH3CH2), 4.81 (1 H,
d, J ϭ 0.6, H-9), 7.19 (1 H, dd, J ϭ 9.8, 0.6, H-4), 8.52 (1 H, d,
J ϭ 9.8, H-3), 7.19 (1 H, dd, J ϭ 7.9, 0.9, H-8), 7.08 (1 H, ddd,
J ϭ 7.4, 7.4 and 0.9, H-6), 7.35 (1 H, dd, J ϭ 7.4 and 1.6, H-5),
7.41 (1 H, ddd, J ϭ 7.9, 7.4 and 1.6, H-7). Ϫ 13C NMR
(62.89 MHz, CDCl3): δ ϭ 14.8 (CH2CH3), 34.5 (NCH3), 58.7
(CH3CH2), 83.3 (CH-9), 113.6, 121.4, 121.7, 128.2, 130.2, 131.6
(CHsp2), 122.7, 140.5, 153.0 (Cq), 168.5 (CO). Ϫ MS; m/z: 229
(Mϩ, 69.2), 184 (M Ϫ OEt, 67.1). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1621
(CϭC), 1728 (CϭO). Ϫ C14H15NO2 (229.3): calcd. C 73.34, H 6.59,
N 6.11, O 13.96; found C 72.95, H 6.71, N 6.06, O 14.06.
COCH2), 2.93 (3 H, s, NCH3), 4.01 and 4.08 (2 H, AB part of
ABX3, J ϭ 10.8 and 7.1, CH3CH2), 4.67 (1 H, dd, J ϭ 6.1, 5.2, H-
2), 6.76 (1 H, s, H-4), 6.52 (1 H, d, J ϭ 8.1, H-8), 6.67 (1 H, dd,
J ϭ 7.4, and 7.3, H-6), 6.88 (1 H, dd, J ϭ 7.4 and 1.4, H-5), 7.15
(1 H, ddd, J ϭ 8.1, 7.3 and 1.4, H-7). Ϫ 13C NMR (62.89 MHz,
CDCl3): δ ϭ 14.0 (CH2CH3), 36.2 (CO CH2), 37.2 (NCH3), 60.9
(CH3CH2), 64.8 (C-2), 111.9, 117.8, 126.4, 128.5, 129.5 (CHsp2),
115.5, 121.7, 142.0 (Cq), 171.0 (CO). Ϫ MS; m/z: 311 ({81Br}Mϩ,
6.2), 309 ({79Br}Mϩ, 6.2), 222 (M Ϫ CH2COOEt, 100). Ϫ IR (neat)
ν˜max/cmϪ1 ϭ 1620 (CϭC), 1725 (CϭO). Ϫ C14H16NBrO2 (310.2):
calcd. C 54.20, H 5.20, N 4.52, O 10.32; found C 53.95, H 5.31, N
4.43, O 10.46.
4-Ethoxycarbonylmethyl-1,4-dihydro-1,2-dimethylquinoline (3b): Oil.
Ϫ Rf (SiO2, CH2Cl2) ϭ 0.30. Ϫ H NMR (250 MHz, CDCl3): δ ϭ
(E)-2-Ethoxycarbonylmethylidene-1,2-dihydro-1,3-dimethylquinoline
(4c): Yellow solid, m.p. 64 °C (C6H12/CH2Cl2). Ϫ Rf (Al2O3,
CH2Cl2) ϭ 0.75. Ϫ H NMR (250 MHz, CDCl3): δ ϭ 1.30 (3 H,
1
1.22 (3 H, t, J ϭ 7.1, CH2CH3), 1.97 (br. s, 3 H, CH3Ϫ2), 2.44 (2
H, m, COCH2), 3.16 (3 H, s, NCH3), 4.12 (2 H, q, J ϭ 7.1,
CH3CH2), 4.19 (1 H, m, H-4), 4.61 (1 H, d, J ϭ 4.9, H-3), 7.1Ϫ7.6
(4 H, m). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.4 (CH2CH3),
20.1 (CH3Ϫ2), 32.7, 34.9, 44.5 (CO CH2), 60.3 (CH3CH2), 98.2,
111.6, 120.6, 126.9, 128.2 (CHsp2), 118.3, 138.0, 142.4 (Cq), 172.1
(CO). Ϫ MS; m/z: 245 (Mϩ), 158 (M Ϫ CH2COOEt, 100). Ϫ IR
(neat) ν˜max/cmϪ1 ϭ 1650 (CϭC), 1723 (CϭO). Ϫ C15H19NO2
(245.3): calcd. C 73.44, H 7.81, N 5.71, O 13.04; found C 73.56, H
7.65, N 5.56, O 12.89.
1
t, J ϭ 7.1, CH2CH3), 2.15 (s, 3 H, CH3Ϫ3), 3.59 (3 H, s, NCH3),
4.15 (2 H, q, J ϭ 7.1, CH3CH2), 5.02 (1 H, br. s, H-9), 7.06 (1 H,
s, H-4), 6.24 (1 H, d, J ϭ 8.3, H-8), 7.10 (1 H, dd, J ϭ 7.9 and 7.1,
H-6), 7.30 (1 H, dd, J ϭ 7.9 and 1.5, H-5), 7.39 (1 H, ddd, J ϭ
8.3, 7.1 and 1.5, H-7). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.6
(CH2CH3), 20.1 (CH3Ϫ3), 44.1 (NCH3), 58.9 (CH3CH2), 83.8 (CH-
9), 115.5, 122.2, 126.6, 129.1, 129.9 (CHsp2), 123.1, 131.5, 141.2,
153.8 (Cq), 166.9 (CO). Ϫ MS; m/z: 243 (Mϩ, 64.3), 198 (M Ϫ
OEt, 66). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1664 (CϭC), 1734 (CϭO). Ϫ
C15H17NO2 (243.3): calcd. C 74.05, H 7.04, N 5.76, O 13.15; found
3-Bromo-4-ethoxycarbonylmethyl-1,4-dihydro-1-methylquinoline
(3e): Oil. Ϫ 1H NMR (250 MHz, CDCl3): δ ϭ 1.30 (3 H, t, J ϭ C 73.80, H 7.19, N 5.51, O 12.92.
7.1, CH2CH3), 2.50 (m, 2 H, COCH2), 3.09 (3 H, s, NCH3), 4.15
(E)-2-Ethoxycarbonylmethylidene-1,2-dihydro-1,4-dimethylquinoline
(2 H, m, CH3CH2), 4.30 (1 H, dd, J ϭ 4.8, 7.7, H-4), 6.36 (1 H, s,
H-2), 6.60Ϫ7.12 (4 H, m). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ
14.2 (CH2CH3), 38.0 (NCH3), 42.5 (CO CH2), 43.2 (C-4), 60.5
(CH3CH2), 111.5, 121.6, 127.5, 128.8, 133.8 (CHsp2), 93.6, 122.2,
133.7, (Cq), 171.4 (CO). Ϫ C14H16NBrO2 (310.2): calcd. C 54.20,
H 5.20, N 4.52, O 10.32; found C 54.45, H 5.27, N 4.33, O 10.53.
(4d): Yellow solid, m.p. 102 °C (C6H12/CH2Cl2). Ϫ Rf (SiO2,
1
CH2Cl2) ϭ 0.32. Ϫ H NMR (250 MHz, CDCl3): δ ϭ 1.29 (3 H,
t, J ϭ 7.1, CH2CH3), 2.35 (s, 3 H, CH3Ϫ4), 3.39 (3 H, s, NCH3),
4.15 (2 H, q, J ϭ 7.1, CH3CH2), 4.73 (1 H, br. s, H-9), 8.43 (1 H,
s, H-3), 7.10 (1 H, dd, J ϭ 7.9, 7.3, H-6), 7.16 (1 H, d, J ϭ 8.2, H-
8), 7.49 (1 H, dd, J ϭ 7.9 and 1.5, H-5), 7.39 (1 H, ddd, J ϭ 8.2,
9-Ethoxycarbonylmethyl-9,10-dihydro-10-methylacridine (3f): Oil. Ϫ 7.3 and 1.5, H-7). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.8
Rf (SiO2, CH2Cl2) ϭ 0.77. Ϫ 1H NMR (250 MHz, CDCl3): δ ϭ 1.25 (CH2CH3), 19.2 (CH3Ϫ4), 34.5 (NCH3), 58.6 (CH3CH2), 81.9 (CH-
(3 H, t, J ϭ 7.1, CH2CH3), 2.64 (2 H, d, J ϭ 7.6, COCH2), 3.42 (3
9), 113.9, 120.4, 121.5, 124.6, 129.9 (CHsp2), 123.7, 138.6, 152.9
Eur. J. Org. Chem. 2000, 2915Ϫ2921
2919

F. Diaba, C. Le Houerou, M. Grignon-Dubois, B. Rezzonico, P. Gerval
FULL PAPER
(Cq), 168.6 (CO). Ϫ MS; m/z: 243 (Mϩ, 50.3), 198 (M Ϫ OEt, 66),
6.50Ϫ7.80 (8 H, m). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.1
171 (M Ϫ COOCH2CH2, 100). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1667 (Cϭ (CH2CH3), 36.7 (CO CH2), 37.4 (NCH3), 61.1 (C-6), 60.6
C), 1729 (CϭO). Ϫ C15H17NO2 (243.3): calcd. C 74.05, H 7.04, N (CH3CH2), 113.3, 118.4, 122.9, 123.4, 126.0, 127.1, 128.0, 129.4
5.76, O 13.15; found C 73.85, H 7.16, N 5.61, O 12.97.
(CHsp2), 122.5, 130.7, 134.9, 144.1 (Cq), 171.7 (CO). Ϫ MS; m/z:
281 (Mϩ, 3.4), 194 (M Ϫ CH2COOEt, 100). Ϫ IR (neat) ν˜max
/
(E)-4-Ethoxycarbonylmethylidene-1,4-dihydro-1,2-dimethylquinoline
(5b): Pale yellow solid, m.p. 110 °C (C6H12/CH2Cl2). Ϫ Rf (SiO2,
cmϪ1 ϭ 1725 (CϭO). Ϫ C18H19NO2 (281.4): calcd. C 76.84, H
6.81, N 4.98, O 11.37; found C 76.68, H 6.83, N 4.79, O 11.48.
1
CH2Cl2) ϭ 0.30. Ϫ H NMR (250 MHz, CDCl3): δ ϭ 1.30 (3 H,
t, J ϭ 7.1, CH2CH3), 2.34 (s, 3 H, CH3Ϫ2), 3.55 (3 H, s, NCH3),
4.17 (2 H, q, J ϭ 7.1, CH3CH2), 5.60 (1 H, br. s, H-9), 7.75 (1 H,
s, H-3), 7.18 (1 H, dd, J ϭ 7.5, 7.4, H-6), 7.27 (1 H, d, J ϭ 8.4, H-
8), 7.47 (1 H, dd, J ϭ 8.4 and 7.4, H-7), 7.94 (1 H, d, J ϭ 7.5, H-
5). Ϫ 13C NMR (62.89 MHz, CDCl3): δ ϭ 14.4 (CH2CH3), 21.6
(CH3Ϫ2), 33.6 (NCH3), 58.0 (CH3CH2), 87.6 (CH-9), 106.1, 114.5,
122.5, 123.8, 130.0 (CHsp2), 122.2, 139.6, 143.3, 145.4 (Cq), 169.6
(CO). Ϫ MS; m/z: 243 (Mϩ, 42.2), 198 (M Ϫ OEt, 61). Ϫ IR (neat)
ν˜max/cmϪ1 ϭ 1667 (CϭC), 1730 (CϭO). Ϫ C15H17NO2 (243.3): C
74.05, H 7.04, N 5.76, O 13.15; found C 73.98, H 7.11, N 5.69,
O 12.95.
Acknowledgments
Financial support by Conseil Regional d’Aquitaine is gratefully
acknowledged.
[1]
See as example: J. Babiak, H. M. Elokdah, C. P. Miller, T. S.
Sulkowski, U.S. US 5,939,435, 1999, American Home Product
Corporation, USA.
T. G. Murali Dhar, C. Gluchowski, Tetrahedron Lett. 1994,
35, 989Ϫ992.
[2]
[3] [3a]
J. Otera, Y. Wakahara, H. Kamei, T. Sato, H. Nazaki, S.
[3b]
Fukuzumi, Tetrahedron Lett. 1991, 32, 2405Ϫ2408. Ϫ
S.
Fukuzumi, M. Fujita, S. Noura, J. Otera, Chem. Lett. 1993,
(E)-2-Ethoxycarbonylmethylidene-1,2,3,4-tetrahydro-1-methyl-
quinoline (6a): A solution of 2a (100 mg, 0.43 mmol) in CDCl3
(1 mL) in an NMR tube was kept at room temperature. The 2aǞ6a
isomerization was monitored by 1H NMR spectroscopy, until 2a
had disappeared (70 d). The following 1H signals were used to con-
firm that conversion took place: for 2a, the NCH3 singlet at 2.91
and the H-3 dd at 5.78; for 6a, the NCH3 singlet at 3.30 and the
H-9 singlet at 4.95. The reaction was also followed by GC (capillary
column JWDB-5MS 30 m ϫ 0.25 mm ϫ 0.25 µm). The GC oven
temperature programming was 5 °C.minϪ1 starting from 220 °C
until 280 °C. Retention time (min) were respectively 2a 3.3, 6a 4.9,
4a 6.8. Within 10 days, the signals of 2a began to decrease, whereas
6a appeared. After 70 days on standing, the NMR spectra showed
the presence of only 6a. Evaporation of the solvent led to 6a
(98 mg, 98% yield). ). Ϫ Rf (SiO2, CH2Cl2) ϭ 0.62. Ϫ 1H NMR
(250 MHz, CDCl3): δ ϭ 1.30 (3 H, t, J ϭ 7.1, CH2CH3), 2.65 (2
H, m), 3.30 (3 H, s, NCH3), 3.43 (2 H, m), 4.14 (2 H, q, J ϭ 7.1,
CH3CH2), 4.95 (1 H, br. s, H-9), 5.35 (1 H, d, J ϭ 7.3, H-3), 6.95
(2 H, m), 7.11 (1 H, m),7.20 (1 H, m). Ϫ 13C NMR (62.89 MHz,
CDCl3): δ ϭ 14.6 (CH2CH3), 24.8 (CH2), 25.5 (CH2), 35.0 (NCH3),
58.8 (CH3CH2), 87.1, 115.1, 121.9, 127.2, 127.3 (CHsp2), 128.8,
141.6, 159.3, 168.6 (Cq). Ϫ MS; m/z: 231 (Mϩ, 100), 202 (M Ϫ Et,
11.6), 186 (M Ϫ OEt, 69.2). Ϫ C14H17NO2 (231.3): calcd. C 72.70,
H 7.41, N 6.06, O 13.83; found C 72.37, H 7.20, N 6.13, O 14.01.
1025Ϫ1028.
[4]
[5]
See as examples: D. J. Ager. In Encyclopaedia of Reagents for
Organic Synthesis, Ed. Paquette, John Wiley & Sons, Inc.: New
York, 1995, Vol. 4, pp 2523Ϫ2525 and references therein.
[5a] R. Latouche, F. Texier-Boullet, J. Hamelin, Bull. Soc. Chim.
[5b]
Fr. 1993, 130, 535Ϫ546. Ϫ
M. Bellassoued, N. Ozanne, J.
Org. Chem. 1995, 60, 6582Ϫ6584. Ϫ [5c] S. Jolivet, S. Abdallah-
El-Ayoubi, D. Mathe, F. Texier-Boullet, J. Hamelin, J. Chem.
[5d]
Research 1996, 300Ϫ301. Ϫ
E. Nakamura, M. Shimizu, I.
Kuwajima, Tetrahedron Lett. 1976, 20, 1699Ϫ1702.
For preliminary results, see M. Grignon-Dubois, F. Diaba, J.
Chem. Research 1998, 660Ϫ661.
[6]
[7]
Silicon reagents are known to be weaker nucleophiles than tin
reagents: see for example A. Hosomi, H. Iguchi, M. Endo, H.
Sakurai, Chem. Lett. 1979, 977Ϫ980.
[8]
The formation of a SiϪF bond (142 kcal.molϪ1) is usually a
highly exothermic process, which provides the driving force for
a number of useful synthetic reactions. For reviews, see as ex-
[8a]
amples:
W. P. Weber in Silicon Reagents for Organic Syn-
thesis, Springer Verlag, Berlin, 1983. Ϫ [8b] E. W. Colvin, Silicon
[8c]
Organic Synthesis, Butterworths, London, 1981. Ϫ
G. G.
[8d]
Yakobson, N. E. Akhmetova, Synthesis 1983, 169Ϫ184. Ϫ
J. H. Clark, Chem. Rev. 1980, 80, 429Ϫ452.
[9]
The larger dielectric constant of acetonitrile (ε ϭ 35.94) allows
for the formation of a solvent-separated ion pair, while THF
(ε ϭ 7.58) or dichloromethane (ε ϭ 8.93) should hinder the
dissociation. A. Loupy, B. Tchoubar, D. Astruc, Chem. Rev.
1992, 1141Ϫ1165 and references therein.
[10] [10a] F. Diaba, I. Lewis, M. Grignon-Dubois, S. Navarre, J. Org.
[10b]
Chem. 1996, 61, 4830Ϫ4832. Ϫ
M. Grignon-Dubois, F.
1-Ethoxycarbonylmethyl-1,2-dihydro-2-methylisoquinoline (8a): Oil.
Diaba, M.-C. Grelier-Marly, Synthesis 1994, 800Ϫ804.
Melting point as well as spectroscopic data are identical to lit-
erature: La Coste, W. Ber. 1882, 15, 186. M. Grignon-Dubois,
A. Meola, Synth. Commun. 1995, 25 (19), 2999Ϫ3006.
1
Ϫ Rf (SiO2, CH2Cl2) ϭ 0.68. Ϫ H NMR (250 MHz, CDCl3): δ ϭ
[11]
1.17 (3 H, t, J ϭ 7.1, CH2CH3), 2.50 and 2.71 (2 H, AB part of
ABX, J ϭ 14.1, 7.1 and 6.2, COCH2), 2.94 (3 H, s, NCH3), 4.04
(2 H, q, J ϭ 7.1, CH3CH2), 4.80 (1 H, ddd, J ϭ 7.1, 6.2 and 1.1,
H-1), 5.35 (1 H, d, J ϭ 7.3, H-3), 6.03 (1 H, dd, J ϭ 7.3 and 1.3,
H-4), 6.80Ϫ6.96 (2 H, m), 6.98 (1 H, ddd, J ϭ 7.5, 7.2 and 1.3),
7.10 (1 H, ddd, J ϭ 7.3, 7.2, and 1.8). Ϫ 13C NMR (62.89 MHz,
CDCl3): δ ϭ 14.1 (CH2CH3), 36.3 (CO CH2), 40.5 (NCH3), 59.3
(C-1), 60.5 (CH3CH2), 97.6, 122.7, 124.7, 125.6, 127.6, 136.1
(CHsp2), 128.1, 132.4 (Cq), 171.7 (CO). Ϫ MS; m/z: 231 (Mϩ, 7.6),
144 (M Ϫ CH2COOEt, 100). Ϫ IR (neat) ν˜max/cmϪ1 ϭ 1725 (Cϭ
O). Ϫ C14H17NO2 (231.3): calcd. C 72.70, H 7.41, N 6.06, O 13.83;
found C 72.39, H 7.25, N 6.16, O 14.06.
[12] [12a] Atomic charges in quinoliniums were calculated with Hyp-
erchem, version 5 (Hypercube, Inc. Gainsville, 32601 Fl). The
results come from an MNDO single-point calculation of the
optimized geometry (final gradient Յ 0.005 kcal. AϪ1molϪ1),
˚
using an RHF Hamiltonian. In the case of the C-2 adducts,
the AM1 method was also used. Ϫ [12b] These kinds of derivat-
ives appear to be unknown in the literature.
[13]
[14]
D. J. Adams, J. H. Clark, L. B. Hansen, V. C. Sanders, S. Tav-
ener, J. Fluor. Chem. 1998, 92, 123Ϫ125.
J. Levillain, M. Vazeux, Synthesis 1995, 56Ϫ62 .
[15] [15a]
N. J. Leonard, R. L. Foster, J. Am. Chem. Soc. 1951, 73,
[15b]
3325Ϫ3329. Ϫ
N. J. Leonard, R. L. Foster, J. Am. Chem.
[15c]
Soc.1952, 74, 2110Ϫ2111. Ϫ
O. N. Chupakhin, V. N.
Charushin, Tetrahedron 1988, 44, 1Ϫ34.
6-Ethoxycarbonylmethyl-5,6-dihydro-5-methylphenanthridine (8b):
[16]
[17]
[16a]
1
For reviews, see:
N. Y. Sidgewick, F. R. S. Sidgewick, The
Oil. Ϫ Rf (SiO2, CH2Cl2) ϭ 0.56. Ϫ H NMR (250 MHz, CDCl3):
Organic Chemistry of Nitrogen; Clarendon Press, Oxford, 1966,
δ ϭ 1.17 (3 H, t, J ϭ 7.1, CH2CH3), 2.40 and 2.63 (2 H, AB part
of ABX, J ϭ 14.2, 7.9 and 5.6, COCH2), 3.04 (3 H, s, NCH3), 4.04
(2 H, q, J ϭ 7.1, CH3CH2), 4.82 (1 H, dd, J ϭ 7.9, 5.6, H-6),
[16b]
p 718. Ϫ
See, for example:
J. Gurnos, Quinolines, Wiley, New York, 1982.
[17a]
´
J. Metzger, H. Larive, E.-J. Vincent, R.
Dennilauler, R. Baralle, C. Gaurat, Bull. Soc. Chim. Fr. 1967,
2920
Eur. J. Org. Chem. 2000, 2915Ϫ2921

Ethoxycarbonylmethylation of N-Methylquinolinium- and Isoquinolinium Iodides
FULL PAPER
[17b]
30Ϫ40. Ϫ
G. T. Pilyugin, B. M. Gutsulyak, Usp. Khim.
version 4.1 and GMMX, version 1.0 (from Serena Software,
P.O. Box 3076, Bloomington, IN 47402Ϫ3076). Both use the
features of Allinger’s MMX force-field, including pi-valence
electron self-consistent field calculations.
O. Doebner, W. Miller, Ber. 1883, 16, 2464Ϫ2472. Ϫ
C. F. Duffin, Adv. Heterocycl. Chem. 1964, 3, 1Ϫ56.
[17c]
1963, 32 (4), 389Ϫ432 (Chem. Abstr. 1963, 59, 3889b). Ϫ
J. W. Bunting, W. G. Meathrel, Tetrahedron Lett. 1971,
133Ϫ136. Ϫ [17d] S. Fukuzumi, S. Noura, J. Chem. Soc., Chem.
[17e]
[19] [19a]
[19b]
Commun. 1994, 287Ϫ288. Ϫ
Bull. 1990, 38, 2577Ϫ2580.
M. Maeda, Chem. Pharm.
[18]
Heats of formation for the pair of regioisomers were calculated
on a 486 PC-compatible running the programs PCMODEL,
January 23, 2000
[O99623]
Eur. J. Org. Chem. 2000, 2915Ϫ2921
2921