Regioselective demethylation of quinoline derivatives. A DFT rationalization

Article abstract of DOI:10.1016/j.molstruc.2016.03.034

Demethylation of compound 2,7-dimethoxyquinoline-3-carbaldehyde 1, is carried out using BBr₃. However, all attempts led, either to the starting material or to the regioselective demethylation at position 2 affording the product 4a. The nature (donor or acceptor) and the position of the R (CHO or CN) group is likely to play a role in the preventing the demethylation at position 7. To address this phenomena, the demethylation of 2-chloro-7-methoxyquinoline-3-carbaldehyde 2 and 2,7-dimethoxyquinoline-3-carbaldehyde 3 has been carried out. To support the results obtained, theoretical computations at DFT level (vide infra) have been carried out upon compound 1. The exploration of how the gas-phase demethylation process on Quinoline can be affected at a position 7 center by stepwise substation effects using different electro-donor and attractor groups, show that demethylation process seems to be more favorable when substituent is an electro-donor. This is sustained by bond energy and thermodynamic analyses (vide infra).

Full text of DOI:10.1016/j.molstruc.2016.03.034

Accepted Manuscript
Regioselective demethylation of quinoline derivatives. A DFT rationalization
Belferdi Fatiha, Merabet Naima, Belkhiri Lotfi, Douara Bachir
PII:
S0022-2860(16)30229-0
10.1016/j.molstruc.2016.03.034
MOLSTR 22345
DOI:
Reference:
To appear in: Journal of Molecular Structure
Received Date: 5 January 2016
Revised Date: 29 February 2016
Accepted Date: 11 March 2016
Please cite this article as: B. Fatiha, M. Naima, B. Lotfi, D. Bachir, Regioselective demethylation of
quinoline derivatives. A DFT rationalization, Journal of Molecular Structure (2016), doi: 10.1016/
j.molstruc.2016.03.034.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
REGIOSELECTIVE DEMETHYLATION OF QUINOLINE DERIVATIVES. A DFT
RATIONALIZATION
Belferdi Fatihaa, Merabet Naimaa, Belkhiri Lotfib and Douara Bachira
aLaboratoire de Pharmacologie et de Phytochimie, Université de Jijel, Algeria
b URCHEMS, Département de Chimie, Université des Frères Mentouri, 25017 Constantine, Algeria.
*corresponding authors: Dr. Merabet-Bouraoui Naima, meraboura@gmail.com, tel : +213553271796
Pr. Belkhiri Lotfi, lotfi.belkhiri@umc.edu.dz; tel : +213781088541
Abstract
Demethylation of compound 2,7-dimethoxyquinoline-3-carbaldehyde 1 , is carried out
using BBr3. However, all attempts led, either to the starting material or to the regioselective
demethylation at position 2 affording the product 4a. The nature (donor or acceptor) and the
position of the R (CHO or CN) group is likely to play a role in the preventing the
demethylation at position 7. To address this phenomena, the demethylation of 2-chloro-7-
methoxyquinoline-3-carbaldehyde 2 and 2,7-dimethoxyquinoline-3-carbaldehyde 3 has been
carried out. To support the results obtained, theoretical computations at DFT level (vide infra)
have been carried out upon compound 1. The exploration of how the gas-phase demethylation
process on Quinoline can be affected at a position 7 center by stepwise substation effects
using different electro-donor and attractor groups, show that demethylation process seems to
be more favorable when substituent is an electro-donor. This is sustained by bond energy and
thermodynamic analyses (vide infra).
Keywords: Regioselective, Demethylation, Quinoline, Donor/Acceptor, Modeling, DFT.
Introduction
The interest of heterocyclic compounds is crucial and among them, the Quinoline derivatives
have attracted great attention and were the subject of several and various studies during
decades [1]. Demethylation of aromatic methyl ethers is often encountered in heterocyclic
chemistry and has been intensively investigated; this is illustrated by its rich literature.
Lewis acids like boron tribromide (BBr3) [2], BeCl2 [3], aluminum chloride (AlCl3) [4], (AlI3)
[5], (AlBr3) [6], (MgI2) [7], AlH2Cl [8] are among the most used reagents. Other methods use
strong Bronsted acids such as HBr [9], H2SO4 [10], HCl [11] or HI via C6H11I/DMF [12].
1

ACCEPTED MANUSCRIPT
Silicium, tain and sulphur systems are also used as demethylating agents e.g., SnO2 [13],
SiCl4/NaI [14], PhSSiMe3 [15], NaCN/DMSO [16], Me3SiI [17], iPrSNa [18], NaN(SiMe3)2
[19] and DABCO/MeSiI [20].
However, in this work all attempts to the demethylation of 2,7-dimethoxyquinoline-3-
carbaldehyde 1 did not occur on the 7-methoxy group (benzene ring) but occurred on 2-
methoxy group (pyridine ring) leading to the regioselective demethylation. This is probably
due to the implication of oxygen lone pair of the 7-methoxy group, in ring resonance with an
attractor group such as CHO or CN, at the 3 position of the Quinoline (Scheme 2).
Furthermore, we performed the demethylation of two other quinoline derivatives i.e., 2-
chloro-7-methoxyquinoline-3-carbaldehyde 2 and 2,7-dimethoxyquinoline-3-carbonitrile 3 as
:
illustrated in scheme 1 and scheme 2
5
4
CN
CHO
CHO
Cl
4'
3
6
2
7
8'
N
N
1
N
MeO
OMe
MeO
OMe
8
MeO
3
2
1
(Scheme 1)
starting material
R
R
X
N
MeO
OH
R
4a R=CHO
4b R= CN
N
MeO
1 X=OMe
R=CHO
2 X=Cl
R=CHO
3 X=OMe
R=CN
N
HO
OH
(Scheme 2)
In the cases of the electro-donor groups, to our knowledge, no such influence is reported in
the literature. However, as reported in previous works [2fg, 4, 16] a possible complete
demethylation of the methoxy groups is expected. For this purpose, no supplementary
2

ACCEPTED MANUSCRIPT
experiments for electro-donor group have been carried out and a predictive study of these
effects is reported only at the theoretical part.
Indeed, in order to get clear cut-vision on this regioselectivity, theoretical calculations have
been carried out at DFT/B3LYP level of theory to rationalize the electronic and/or steric
effects which could contribute to this Quinoline demethylation.
Experimental section
N-(3-methoxyphenyl) acetamide 5: In a 500 mL flask, distilled water (250 mL) and
concentrated hydrochloric acid (9.2 mL, 109 mmol, 1.01 eq) was introduced, 3-methoxy
aniline (12.17 mL, 108 mmol, 1 eq) was added with stirring, and the reaction mixture was
heated at 50 ° C for 10 minutes. After returning to room temperature, acetic anhydride (10.2
mL, 108 mmol, 1 eq) and sodium acetate (16.40 g, 200 mmol, 1.85 eq) were added, then the
mixture was stirring for 20 minutes. The solid was filtered, dried, and crystallized from
(water/ethanol: (80:20)) to give a brown powder. Yield: 75%; Rf : 0.40 (ethyl
acetate/cyclohexane (6:4)), m.p: 82°C, NMR1H δ (ppm) (300 MHz, CDCl3): 7.50 (s, 1H,
NH), 6.5-7.4 (m, 4H, Ar-H); 3.90 (s, 3H, OCH3); 2.20 (s, 3H, COCH3).
2-chloro-7-methoxyquinoline-3-carbaldehyde 2: In a 100 mL two-necked flask was
introduced dry N, N-Dimethylformamide (DMF) (10.45 mL, 135 mmol. 2.5 eq), and cooled at
0°C. under magnetic stirring, was added dropwise phosphorus trichloride oxide (POCl3)
(35.34 mL, 378 mmol. 7.0 eq). When the addition was completed, the reaction mixture was
allowed to attain room temperature and stirred at this temperature for 30 minutes. Then, 3-
methoxy acetanilide (8.92 g, 54 mmol, 1.0 eq) was added and the mixture was heated to 75 °
C for 16 hours. The reaction progress was monitored by thin layer chromatography (TLC).
The mixture was poured slowly into crushed ice for 30 min at 0 <T <10° C. The resulting
solid filtered, washed well with water and dried. The compound was crystallized from ethyl
acetate to give a yellow powder: Yield: 62 %, Rf : 0.81 (ethyl acetate/cyclohexane (6:4)), m.p
193°C. NMR 1H δ (ppm) (300 MHz, CDCl3): 10.42 (s, 1H, CHO), 8.60 (s, 1H, H4), 7.78 (d,
1H, H5), 7.35 (s, 1H, H8), 7.20 (d, 1H, H6), 3.90 (s, 3H, OCH3).
2,7-dimethoxyquinoline-3-carbaldehyde 1: To a solution of 2-chloro-7-methoxyquinoline-
3-carbaldehyde (0.23 g, 1.04 mmol, 1 eq) in methanol (6 mL) was added sodium hydroxide
(0.20 g, 5 mmol, 4.8 eq). The reaction mixture was refluxed for 2 hours. The resulting
solution is poured into ice. The precipitate was filtered crystallized from methanol to give a
yellow powder. Yield: 95%; m.p 160 °C; NMR 1H δ (ppm) (300 MHz, CDCl3): 10.35 (s, 1H,
3

ACCEPTED MANUSCRIPT
CHO), 8.45 (s, 1H, H4), 7.68 (d, 1H, H5), 7.18 (d, 1H, H8), 7.0 (d, 1H, H6), 4.10 (s, 3H, 2-
OCH3), 3.90 (s, 3H, 7-OCH3).
2,7-dimethoxyquinoline-3-carbonitrile 3: To a solution of 2,7 dimethoxyquinoline-3-
carbaldehyde (0.25 g, 1.15 mmol, 1 eq) in tetrahydrofuran (THF) (6 mL), was added a
solution of ammonium hydroxide (7 mL, 33 %) and iodine (0.32 g, 1.26 mmol, 1.1 eq), the
mixture was stirred at RT for 11 hours, the resulting solution was poured into ice. The
precipitated product was filtered, washed and dried to give an orange powder. Yield: 100%;
Rf : 0.75 (ethyl acetate/cyclohexane (6:4)); m.p: 139.7 °C, IR (KBr; ν cm-1): 2240 (νCN), 1247
(νC-O), 1135 (νC-N), NMR 1H δ (ppm) (500 MHz, CDCl3): 8.29 (s, 1H, H4); 7.60 (dd, 1H, H5);
7.20 (d, 1H, H8); 7.10 (dd, 1H, H6); 4.10 (s, 3H, 2-OCH3); 3.91 (s, 3H, 7-OCH3), NMR 13C δ
(ppm) (500MHz, CDCl3): 54.52 (2-OCH3); 55.88 (7-OCH3 ); 95.23 (C4); 106.73 (C3);
116.00 (C8); 118.10 (CN); 118.41 (C4’); 129.40 (C6); 144.61 (C5); 150.13 (C8’); 160.82
(C7); 163.82 (C2).
Demethylation with BBr3 [2f]: To a solution of 2,7-dimethoxyquinoline-3-carbaldehyde 1
(0.42 g, 1.97 mmol, 1 eq) in dichloromethane (20 mL) at -78 °C was added dropwise Boron
tribromide (BBr3, 1M) in dichloromethane (1.5 mL, 13.8 mmol, 7 eq). The reaction mixture
was allowed to attain room temperature and stirred at that temperature overnight. The mixture
was concentrated under reduced pressure to give a solid. The solid was treated with
hydrochloric acid (25 mL, 5%), dissolved in ethyl acetate (50 mL), washed with brine (3 x 50
mL) and water (3 x 50 mL), dried over anhydrous magnesium sulfate, and concentrated under
reduced pressure. After separation by chromatography we obtained 0.26 g of 2-hydroxy-7-
methoxyquinoline-3-carbaldehyde 4a. Yield: 50%; Rf : 0,25 (ethyl acetate: cyclohexane
(8/2)); NMR 1H δ (ppm) (300 MHz, CDCl3): 10.35 (s, 1H, CHO), 10.10 (s, H, NH lactam form
)
8.35 (s, 1H, H4), 7.58 (m, 1H, H5), 6.80 (s, 1H, H8), 6.60 (m, 1H, H6), 3.90 (s, 3H, 7-OCH3).
Demethylation with HBr [9]: A solution of 2-chloro-7-methoxyquinoline-3-carbaldehyde 2
(0.10 g, 0.45 mmol) and hydrobromic acid (1.03 mL) in acetic acid (33%) was stirred at
reflux. The reaction was controlled by thin layer chromatography (TLC). After 20 hours, the
reaction mixture was treated by water, and extracted with ethyl acetate to give the 2-hydroxy-
7-methoxyquinoline-3-carbaldehyde 4a, Yield: 31%.
Demethylation with HCl [11]: A solution of 2-chloro-7-methoxyquinoline-3-carbaldehyde 2
(0.10 g, 0.45 mmol) and hydrochloric acid (2 mL, 37%) was refluxed at 120°C. The reaction
was controlled by thin layer chromatography (TLC). The reaction mixture was treated with
4

ACCEPTED MANUSCRIPT
water, and extracted with ethyl acetate to give the 2-hydroxy-7-methoxyquinoline-3-
carbaldehyde 4a, Yield: 67%.
Demethylation with iodocyclohexane/DMF [12]: To
a solution of 2-chloro-7-
methoxyquinoline-3-carbaldehyde 2 (0.25 g, 1.13 mmol, 1 eq) in N,N-Dimethylformamide
(DMF), iodocyclohexane (1.52 mL, 11.30 mmol, 10 eq) was added. The mixture was refluxed
under argon for 14 hours. The reaction was controlled by thin layer chromatography (TLC).
The mixture was cooled, poured into water (25 mL) and extracted with ethyl acetate (3 x 20
mL). The organic layer was washed with saturated sodium bicarbonate and brine, dried over
sodium sulfate and filtered. The filtrate was concentrated and dried in vacuum to give a
mixture of compounds.
Demethylation with AlCl3 [4a]: 2-chloro-7-methoxyquinoline-3-carbaldehyde 2 (0.09 g, 0.40
mmol, 1 eq) was dissolved in dichloromethane (2 mL) and cooled to 0°C. Anhydrous
aluminum chloride (AlCl3) (0.21 g, 1.60 mmol, 4 eq) in dichloromethane (0.5 mL) was then
added dropwise to the stirred solution. The reaction was controlled by thin layer
chromatography (TLC). After, the reaction mixture reached room temperature, the mixture
was further stirred up to 24 hours. Water was carefully added to decompose excess of
anhydrous aluminum chloride. The reaction mixture was transferred into a separating funnel
and after adjusting the aqueous phase with 1N sodium hydroxide to pH of 11-12, the mixture
was then shacked. The organic phase was discarded and the aqueous phase was re-adjusted
with 1N hydrochloric acid to pH of 8-9 which affords precipitation. The precipitate was
dissolved in methylene chloride, dried over anhydrous sodium sulfate and dried in vacuum to
dryness to give the starting material 2.
Demethylation with AlCl3/DMS [4b]: Dimethyl sulphide (6 mL) was added dropwise to a
suspension of anhydrous aluminum chloride (AlCl3) (0.73 g, 5.65 mmol, 2.5 eq) in
dichloromethane at 0°C with stirring to yield a completely dissolved AlCl3/DMS solution. To
this solution, 2-chloro-7-methoxyquinoline-3-carbaldehyde 2 (0.50 g, 2.26 mmol, 1eq) in
dichloromethane (9 mL), was added over a period of 10 minutes at the same temperature.
Then the reaction mixture was allowed to rise to room temperature (30 °C) and stirring
continued 24 hours. The reaction was controlled by thin layer chromatography (TLC). The
reaction mixture was quenched with cold 1N hydrochloric acid (10 mL) and extracted with
dichloromethane (2 x 15mL) to give the starting material 2.
5

ACCEPTED MANUSCRIPT
Demethylation with AlCl3/Thiourea [4c]: To a mixture of anhydrous aluminum chloride
(AlCl3) (0.24 g, 18 mmol, 4 eq) and thiourea (0.07 g, 0.90 mmol, 2 eq), 2-chloro-7-
methoxyquinoline-3-carbaldehyde 2 (0.1 g, 0.45mmol, 1eq) was added. The reaction mixture
was heated to 90°C, and maintained at this temperature for 2 hours. The reaction was
controlled by thin layer chromatography (TLC). The mixture is then cooled to room
temperature and hydrochloric acid (20 mL, 5 %) was added. The precipitated product was
filtered, washed and dried to give the starting material 2.
Demethylation with [TMAH][Al2Cl7]
a. Preparation of ionic liquide [TMAH][Al2Cl7]
Anhydrous aluminum chloride (AlCl3) (1.73 g, 13.35 mmol, 2 eq) was dissolved in dry
dichloromethane and put into an ice-bath. When the mixture was cooled, trimethylamine
hydrochloride (0.62 g, 6.48 mmol, 1 eq) was added slowly. That reaction mixture was stirred
for 2 hours at room temperature. The dichloromethane was evaporated. This ionic liquid
[TMAH] [Al2Cl7] was used for the demethylation reaction.
b. Demethylation [4d]
2-chloro-7-methoxyquinoline-3-carbaldehyde 2 (0.30 g, 1.35 mmol) was mixed with dry
dichloromethane, and then above prepared ionic liquid, [TMAH] [Al2Cl7] (6.2 mL) was added
into it. The reaction mixture was refluxed in 50°C and monitored by thin layer
chromatography (TLC). After 20 hours, the reaction mixture was cooled to room temperature
and treated with dilute hydrochloric acid and extracted with dichloromethane and dried over
anhydrous sodium sulfate to give starting material 2.
Demethylation with NaCN/DMSO [16]: 2-Chloro-7-methoxyquinoline-3-carbaldehyde 2
(0.20 g, 0.90 mmol, 1 eq) and sodium cyanide (0.25 g, 5.10 mmol, 5.6 eq) was added to
dimethyl sulfoxide (1.35 mL). The mixture was refluxed under argon at 180°C for 5 h. The
reaction was controlled by thin layer chromatography (TLC). The reaction mixture was
acidified with 6N hydrochloric acid. The precipitate was washed with water, dried and filtered
to give the starting material 2.
Demethylation of 2,7-dimethoxyquinoline-3-carbonitrile 4b : A solution of 2,7-
dimethoxyquinoline-3-carbonitrile 3 (0.08 g, 0.36 mmol, 1 eq) and dichloromethane (3 mL)
was cooled at 0°C. Anhydrous aluminum chloride (AlCl3) (0.18 g, 1.4 mmol, 4 eq) in
6

ACCEPTED MANUSCRIPT
dichloromethane (3 mL) was added dropwise for 4 hours. The reaction mixture was allowed
to attain room temperature and stirred for 22 h. water was added to decompose the excess of
aluminum chloride. The mixture was basified to pH 11-12 with addition of 1N sodium
hydroxide , the aqueous layer was treated with 1N hydrochloric acid to pH de 8-9 to give 2-
hydroxy-7-methoxyquinoline-3-carbonitrile 4b as a white precipitate Yield: 85%, Rf : 0.35
(ethyl acetate:cyclohexane (6/4)); NMR 1H δ (ppm) (500 MHz, CDCl3): 8.4 (s, 1H, H4); 7.5
(d, 1H, H5); 6.8 (m, 2H, H8 and H6); 3.9 (s, 7-OCH3).
Computational Details. All calculations were performed at Density Functional Theory DFT
[21] (vide infra) using the Amsterdam Density Functional (ADF) [22] Energies and gradients
were calculated using the Local Density Approximation (LDA, Slater exchange and VWN
correlation) and gradient corrections with B3LYP [23] hybrid method due to Becke exchange
and Lee, Yang and Parr correlation was used self-consistently. The B3LYP computations are
all electron ones. For all elements, Triple–ζ Slater–type valence orbitals (STO-TZP)
augmented by one set of polarization functions were used (2p and 3d sets on H, C, O, and N
atoms) and taken from the ADF/B3LYP/TZP database.
As ADF program supplies an energetic decomposition of the metal-ligand bonding into
chemically useful terms, we have carried out spin-restricted fragment calculations considering
the two molecular moieties in interaction, according to the reaction scheme formally:
Me + Quinoleine-R → Me-Quinoleine-R
were R = CN, CHO, NO2 for attractor groups and CH2CN, CH2OH, NH2CH3 and for donors
ones. We remind that this energetic decomposition, which is based on the transition-state
method developed by Morokuma, then by Ziegler et al. [24], provides insights into the
balance of the different bonding electronic or electrostatic factors at work between the
fragments in a molecule.
Thus, within this scheme, the resulting total bonding energy TBEfrag between two fragments
can be decomposed into two terms as TBEfrag = Esteric+ Eorb, where the Esteric term is, in our
case, the steric interaction energy between the two fragments Me and Quinoleine-R, and Eorb
is the orbital (covalent) contribution to the Me-O methoxy bond.
7

ACCEPTED MANUSCRIPT
The steric energy term (Esteric = EPauli+ EES) is itself decomposed into a destabilizing term
EPauli, the electronic repulsion due to the Pauli principle, and EES, the stabilizing electrostatic
energy between the two fragments. The bonding energies have been computed at the
DFT/B3LYP/TZP level.
Thermodynamic terms, i.e., reaction enthalpies (ꢀH), entropies (ꢀS) and Gibbs free energies
(ꢀG) change, were calculated at 298.15 K and 1 atm (vide infra) from the computed fragment
bond energies (ꢀEfrag) and vibrational frequencies. This was achieved using standard
thermochemistry relations for an ideal gas, according to the following equation:
∆H298 = ∆Efrag + ∆U + ∆(pV) = ∆Efrag + ∆E(pV)
298
∆U = ∆Etrans298 + ∆Erot298 + ∆Etrans298 + ∆Evib
with ∆U is the internal energy at gas phase and the quantity pV/n = RT = 8.314472 * 298.15 /
4.184.103= 0.592 kcal/mol
Gibbs free energies G = H - TS or ∆G = ∆H - T∆S denotes the change for the reaction scheme
between reactants (CH3O-Q) and products (HO-Q), expressed by the equations as follow:
∆G = Gf(HO-Q) - Gi(CH3O-Q) were Gi and Gf denote initial and final Gibbs terms.
Q-OH
Q-OCH3
∆Efrag = TBEfrag
- TBEfrag
Results and discussion
Chemistry
Our general synthetic approach of the studied compounds is depicted in scheme 3.
According to literature [25], we obtained 1 from meta-anisidine via three step reaction;
compound 5 was obtained by treatment of meta-anisidine by anhydride acetic acid and
compound 2 by Vilsmeir-Haack reaction. The nucleophilic substitution of Cl by OMe [26]
within compound 2 leads to compound 1. Compound 3 was resulted from a transformation of
CHO into CN in compound 1 [27].
8

ACCEPTED MANUSCRIPT
CN
CHO
HCl, H2O
NH3 ( 30 %)
2, THF
POCl3 / DMF (7:3)
Ac2O, AcONa
O
N
NH2
MeO
MeO
H
I
N
N
MeO
Cl
MeO
Cl
5
2
6
MeOH / NaOH
reflux 2 h
MeOH / NaOH
reflux 2 h
CN
CHO
OMe
NH3 ( 30 %)
I2, THF
N
N
MeO
OMe
3
+
other product
MeO
3
1
(Scheme 3)
The compound 3 was prepared by reaction of the 2 with NH3/I2 to give first over the
corresponding carbonitrile 6 in quantitative yield, but nucleophilic substitution taking place
on Chloride atom gave two products including 3. In order to improve the yield of 3, we
choose to transform, first 2 into 1 afterwards, we conducted the nucleophilic substitution to
get the target compound 3.
Demethylation of 2,7-dimethoxyquinoline-3-carbaldehyde 1 (Figure 1) with Boron tribromide
BBr3 resulted on product 4a corresponding to the regioselective demethylation at position 2,
the effects of BBr3 concentration, temperature and time reaction were examined and the
results are summarized in table 1:
CHO
N
MeO
OMe
1
(Figure 1)
Table 1: Mole ratio of BBr3 reagent, condition and yield of starting material 1 demethylation.
Starting
material
Reagents
Temperature
Time
Products
BBr3, 7 eq
-78°C to RT
-25°C to RT
26 h
74 h
4a (50%)
4a
Entry 1
Entry 2
1
1
BBr3, 11 eq
BBr3, 11 eq
Reflux (40°C)
8 h
4a (53%)
Entry 3
1
9

ACCEPTED MANUSCRIPT
On the three mentioned inputs, reagents condition, temperature and reaction time were varied
in order to demethylate the 2,7-dimethoxyquinoline-3-carbaldehyde 1 at position 7, as aimed
initially. Indeed, the first attempt (Entry1: 7eq of BBr3, -78°C to RT, 26 h) did not give
demethylation at position 7, but only at position 2 (product 4a). Thus, we slightly increased
temperature, time and quantity of BBr3 (Entry 2: 11 eq (BBr3), -25 to RT, 74 h), however,
demethylation at position 7 failed once again; the reaction was then carried out at reflux, with
11eq of BBr3 for 8 h (Entry 3) and at each time, the product 4a is formed.
We conclude that although the BBr3 is the most used demethylation agent, it failed to
demethylate 2,7-dimethoxyquinoline-3-carbaldehyde 1 at position 7, even if varying reaction
conditions. These justify the further theoretical calculations to shed light on this phenomenon.
As the demethylation at position 7 is still the focus of our work, it was conducted on
compound 2 with a methoxy group at position 7 and a Chloro group at position 2 (Figure
2),using various demethylating agents (Table 2).
CHO
N
MeO
Cl
2
(Figure 2)
Table 2: Mole ratio of different reagents, condition and yield of starting material 2
demethylation.
Starting
Reagents
Temperature
Time
Products obtained
material
2
HBr/ AcOH (33%)
HBr/AcOH (33%)
HCl (2N)
Reflux
Reflux
8 h
20 h
1h 30
Starting material
4a (31%)
Entry 1
Entry 2
Entry 3
2
2
Reflux
Reflux
Starting material
HCl (4N)
2 h
4a (60%)
4a (67%)
Entry 4
Entry 5
2
2
HCl (37%)
Reflux 2 h, then
RT, 12 h
14 h
CyI/DMF
Reflux
RT
14 h
48 h
24 h
24 h
Mixture of products
Starting material
Starting material
Starting material
Entry 6
Entry 7
Entry 8
Entry 9
2
2
2
2
AlCl3/CH2Cl2 (4 eq)
AlCl3/CH2Cl2 (5 eq)
AlCl3/DMS
RT
0°C to TA
10

ACCEPTED MANUSCRIPT
AlCl3/thiourea
[TMAH][Al2Cl7]
NaCN/DMSO
Reflux
Starting material
Starting material
Starting material
Entry 10
Entry 11
Entry 12
5 h
2
2
Reflux (50°C)
Reflux (180°C)
20 h
5 h
2
Similarly to conditions reported in the Table 1, reaction conditions reported in the Table 2
were varied to achieve demethylation at position 7. In these cases we have used three sets of
demethylation agents summarized as follow:
•
Bronsted acids (HBr, HCl and HI) (Entries 1-6); time reaction, temperature and
concentration were varied according to literature. Except for entry 6 which gives a
mixture of products, all other attempts led to either the starting material or the product
4a (formed by the hydrolysis of C-Cl bond). Demethylation at position 7 is not
observed.
•
•
Lewis acid AlCl3 with different solvents (Entries 7-10), the reaction time and
temperature were varied according to literature, but none of these give demethylation
at position 7.
Finally, other systems (Entries 11-12) were used, but demethylation at position 7
failed.
In order to complete our study, we performed the demethylation of another quinoline
derivative 2,7-dimethoxyquinoline-3-carbonitrile 3 (Figure 3). Reagents and conditions were
varied (Table 3):
CN
N
MeO
OMe
3
(Figure 3)
Table 3: Mole ratio of different reagents, condition of starting material 3 demethylation.
Starting
material
Reagents
Temperature Time
Products
AlCl3/CH2Cl2
RT
22 h
2 h
Entry 1
Entry 2
3
4b
HCl
Reflux
Mixture of products
3
11

ACCEPTED MANUSCRIPT
CyI/DMF
Reflux
Reflux
14 h
8 h
Mixture of products
Mixture of products
Entry 3
Entry 4
3
3
NaCN/DMSO
we first worked with a Lewis acid AlCl3, which is a powerful demethylation agent and it is
generally used at low T or RT, but the conditions of temperature and reaction time in Entry 1
led, after 22 hours at RT, to 4b (product of demethylation at position 2). Thus, owing to the
literature, other reagents and stronger reaction conditions (reflux, and time) as in entries (2, 3,
and 4) were used unsuccessfully, to demethylate at position 7.
The results reported in Table 2 and Table 3 confirm the previous results (Table 1) showing
that no demethylation occurs at position 7.
Given the results above, we suggested that the methoxy group on benzene ring does not
demethylate because of the presence of acceptor groups such as CHO or CN at the 3 position
of quinoline. This behaviour is explained by the mesomeric effect which makes the electron
pair, on the benzene oxygen of 7-methoxy group, less available and prevents any
demethylation. Therefore, the reaction with BBr3 (scheme 5) (for example) cannot take place
with the methoxy involved in ring resonance with the CHO or CN, but it occurs with the
pyridine 2-methoxy group. As aforementioned, the facile demethylation of 2-methoxy group
can be explained by the tautomeric equilibrium lactame-lactime which is displaced toward the
more stable lactam form as depicted in scheme 4, and is likely to be the driving force for the
demethylation. This point will be supported by theoretical calculations.
R
R
O
N
N
MeO
OH
MeO
H
R=CHO, CN
(
Scheme 4)
12

ACCEPTED MANUSCRIPT
O
C
O
H
C
H
N
O
O
N
O
O
H3C
CH3
H3C
CH3
BBr3
O
O
C
C
H
H
- CH3Br
N
BBr3
MeO
N
O
MeO
O
Br2B
3 H2O
H3C
O
C
O
C
H
H
+ H3 BO3 + 2 HBr
N
N
MeO
O
H
MeO
OH
(Scheme 5)
Theoretical analysis
Molecular geometries: Optimized geometries were obtained at DFT/B3LYP/TZP level of
calculations and depicted on figure 4. The most significant optimized parameters; bond
distances (Å) and angles (°) and in addition, Total Bonding Energy (TBEfrag) are reported in
the table 4 (see computational details) and concern three attractor (A) type i.e. R = CN, CHO,
NO2 and three donors (D) on Quinoline-R (R = CH2CN, CH2OH, NHCH3) derivatives.
Table 4: Structural and energetic analysis of Quinoline-R series
∆Efrag
R
C7-O
O-Me
C7-O-Me
(eV)
kcal/mol
-116.29
CN (A)
1.350
1.338
1.337
1.342
1.411
1.406
1.407
1.405
118.6
116.5
116.6
116.2
-5.043
CHO (A)
NO2 (A)
CH2CN (D)
-5.276
-7.709
-5.200
-121.66
-122.47
-119.91
13

ACCEPTED MANUSCRIPT
CH2OH (D)
NHCH3 (D)
1.343
1.354
1.404
1.400
116.3
116.1
-5.149
-3.286
-118.73
-75.77
(A) Acceptor ; (D) Donor
N
CN
CHO
NO2
CH2CN
CH2OH
NHCH3
Figure 4: DFT/B3LYP optimized molecular structures of Quinoleine-R series
As shown in the table 4, for the acceptor substituent R, the C7-O bond distances seem to
decrease slightly according the order CN > CHO > NO2 i.e. from 1.350 to 1.337 Å. In
opposite, the TBEfrag increases in absolute value revealing, that the NO2 derivative lead to the
most stable (or most exothermic) C7-O methoxy coordination (-122.47 kcal/mol).
In the case of donor groups (R = CH2CN, CH2OH, NHCH3), the C7-O bond distances
increase slightly but significantly i.e., form 1.342 to 1.354 Å with the donation ability
variation CH2CN < CH2OH < NHCH3. These results shown that the NHCH3 amine donor
group leads to the longest bond C7-O methoxy coordination and so, to the lower TBEfrag in
absolute value (-75.77 kcal/mol) suggesting that is of mostly σ-single bond character. This
result is in agreement with the aforementioned available or unavailable lone pair on C7-
oxygen atom owing to donor/acceptor R group (see scheme 5). The electronic structure
analysis will shed light on these points.
Electronic structure analysis
To get a clearer picture of donor/acceptor effects on the C7-methoxy coordination, electronic
structure analyses using Mulliken population analysis [28] (MPA), have been considered at
DFT/B3LYP/TZP level of theory (see computational details). Despite its drawback, the MPA
14

ACCEPTED MANUSCRIPT
remains useful tool for comparative structural study. The MPA results are reported in the table
5 and concern atomic net charge and overlap atom-atom populations.
As shown by the results in table 5, the comparison of the net charge bearing by the O(C7)
oxygen-methoxy groups between the donor (D) and acceptor (A) group derivatives, show no
significant difference as the values are quasi-similar. However, the C7-O overlap populations
are slightly higher for acceptor than for donor cases, e.g. 0.207 vs. 0.197 for CN vs. CH2CN.
This agrees well with the slight greater multiple bond character of the former.
Table 5: Mulliken electronic structure analysis
Overlap
Net charges
population
R
O(C7)
-0.68
-0.66
-0.66
-0.66
-0.66
-0.67
C7-O
0.207
0.204
0.204
0.197
0.203
0.198
CN (A)
CHO (A)
NO2 (A)
CH2CN (D)
CH2OH (D)
NHCH3 (D)
Energetic analysis
In order to rationalize demethylation ability towards methoxy C7 site, we have carried out
energy fragmentation of the molecule according to the following scheme:
Me + quinoline-R → Me-quinoline-R
with R = CN, CHO, NO2 for acceptors and CH2CN, CH2OH and NH2CH3 for donors. The
results of frequencies calculations and fragment bonding energy (∆Efrag), are reported in the
table 6 and the latter is computed as follow:
Q-OH
Q-OCH3
∆Efrag = TBEfrag
- TBEfrag
TBEfragQ-OH and TBEfragQ-OCH3 correspond to the computed fragment bonding energies of the
Q-OH and its Q-OCH3 methoxy derivatives, respectively, in the Quinoline molecule as shown
in the figure 5.
15

ACCEPTED MANUSCRIPT
4
5
R
3
6
2
7
N
1
MeO
OMe
8
Figure 5: R = CN, CHO, NO2 for acceptors and CH2CN, CH2OH, NHCH3 for donors
The most silent result is the energy ∆Efrag(kcal/mol) of acceptor R groups is slightly more
important in absolute value (more exothermic) than in the donor cases. This is sustained by
the enthalpy reaction ∆H (vide infra) values which are also more negatives in the acceptor
case, suggesting that fragmentation (Me-O cleaving) is slightly more favorable when R is a
donor in agreement with experimental finding (for R acceptor).
As we can see in the table 6, the enthalpy reaction ∆H increases significantly in the order
NHCH3 > CH2CN > CH2OH for donor and decreases for the acceptor groups in the order
CHO > CN > NO2. These results reveal that the NHCH3 is the stronger donor group and the
CHO analogous is the more attractor one.
Table 6: Thermodynamic data (kcal/mol), Internal energy, Enthalpy, entropy and Gibbs free
energy for the product HO-Quinoline and reactant MeO-Quinoline molecules for different
donor and attractor R substituent.
Internal
R
Energy
∆Efrag
U
H
∆H
S
G
∆G
Acceptor
16.77
CHO
CN
137.12
130.68
132.96
-120.94 16.18
-116.20 14.48
-122.47 10.49
-45.98
-38.28
-31.95
118.14
117.51
116.67
-18.45
-19.96
-23.70
-43.97
-36.19
-31.89
15.07
11.08
NO2
Donor
NHCH3
CH2CN
CH2OH
159.36
201.31
151.99
-112.68 46.68
-119.98 81.33
-118.74 33.25
47.27
-30.20
-30.93
-32.74
123.01
130.88
119.32
10.59
42.89
-1.72
-28.31
-28.98
-30.72
81.92
33.84
16

ACCEPTED MANUSCRIPT
Furthermore, when R is an attractor, the enthalpy reaction (∆H) values are more negatives
(more exothermic) than in the donor cases. This sustain once again that demethylation is more
difficult to achieve with acceptor groups relatively to a donor ones, which is in agreement
with experimental trends.
Comparing the Gibbs free energy (∆G) change between acceptor (A) and donor (D) groups, it
reveals that the ∆G (kcal/mol) values are more negatives in the former, which indicates that
the Me-O methoxy bond formation reaction (Me + quinoline-R → Me-quinoline-R) is
stronger when R is an acceptor than in the donor case. This can be illustrated by the histogram
depicted on the figure 6 which shows the greater Gibbs free energy (-∆G) change in absolute
value for the donor than for the acceptor group.
Figure 6: variation of the Gibbs free energy (-∆G) change upon the R group
It is worth noting that interplay experience-DFT, is likely to explain the dependence of the
observed regio-selectivitive demethylation on electron donor/acceptor capability of the R
substituent. Interestingly, the bonding energy and thermodynamic analysis corroborate well
electronic and structural analysis. Both enthalpy reaction (∆H) and Gibbs free energy (∆G)
change between acceptor (A) and donor (D) groups, sustain the demethylation preference
related to the acceptor or donor nature of the Quinoline R substituent, supporting the
experimental finding.
Conclusion
The regioselective demethylation of 2,7-dimethoxyquinoline-3-carbaldehyde 1 and 2,7-
dimethoxyquinoline-3-carbonitrile 3 has been achieved with Lewis acids. BBr3 afforded
17

ACCEPTED MANUSCRIPT
selectively the demethylated isomer at position 2. However, the 2-chloro-7-
methoxyquinoline-3-carbaldehyde 2 survived throughout all attempts of demethylation at
position 7, leading to the starting material or to 4a. In order to explain (in vide infra) the
regioselective ability at position 7 corresponding to Quinoline-R derivatives, DFT
computations have been carried out using the B3LYP method on different acceptor (CHO,
CN and NO2) and donor (NHCH3, CH2CN and CH2OH) R groups. Computational results are
in good agreement with the experimental trends showing that selectivity observed in the
demethylation of the Quinoline-R derivatives is easier when R is donor than acceptor group.
Furthermore, DFT calculations explain that the observed selectivity depends on electron
delocalization capability of the R substituent. Indeed, the computed bonding energy and
thermodynamic analysis corroborate well electronic and structural analysis. This interplay
experience-DFT prompts us to explore more deeply this dependence of the regio-selectivitive
demethylation on electron donor/acceptor capability of the R substituent.
Acknowledgements
The authors gratefully acknowledge the funding support from the Algerian National
Administration of Scientific Research NASR and CNEPRU Grants No. E001720090011 and
No. E00920140093). Computing facilities were provided by Computing Centre of
URCHEMS at Frères Mentouri University of Constantine1. The authors are really grateful to
Dr. Gilles Hanquet from ULP Strasbourg (France) for the experimental support and the
offered facilities. We also address our acknowledgements to Dr. Boujemaa Manhour from
Reims University-(France) for his help.
References
[1] (a) A. Marella, O. P. Tanwar, R. Saha,. M. Rahmat Ali,. S. Srivastava, M. Akhter, M.
Shaquiquzzaman, M. M. Alam, Saudi Pharm. J. 21 (2013) 1-12. (b) S. Madapa, Z. Tusi,
S. Batra, Curr. Org. Chem. 12 (2008) 1116-1183.
[2] (a) J. F. W. McOmie, M. L. Watts, D. E. West, Tetrahedron. 24 (1968) 2289-2292. (b)
Y. Liu, Y. Feng, R.Wang, Y. Gao, L. Lai, Bioorg Med Chem Lett. 11 (2001) 1639-
1641. (c) A. H. Clark, J. D. McCorvy, W. K. Williams, Bioorg Med Chem. 20 (2012)
6366-6374. (d) S.N. Calderon, R.B. Rothman, J. Med. Chem. 37 (1994) 2125-2128. (e)
C. Che, J. Xiang, G. X. Wang, R. Fathi, J. M. Quan, Z. Yang, J. Comb. Chem. 9 (2007)
982–989. (f) K. M. Markovich, V. Tantishaiyakul, A. Hamada, D. D. Miller, J. Med.
18

ACCEPTED MANUSCRIPT
Chem. 35 (1992) 466-479. (g) C. Li, R. J. Angelici, Tetrahedron Lett. 39 (1998) 4235-
4238.
[3] H. Sharghi, F. Tamaddon, Tetrahedron. 52 (1996) 3623 -3640.
[4] (a) A. Asnawi, A. Nawawi, R. E. Kartasasmita, S. Ibrahim, J. Sci. 43 (2011) 43-50. (b).
G. Gopalakrishnan, V. Kasinath, N.D. Pradeep Singh, R. Thirumurugan, Molecules. 5
(2000) 880-885. (c) A. Vass, J. Dudas, L. Borbely, F. Haasz, P. Jekkel, U. S. Patent
0039643 (2008). (d) P. S. Shinde, S. Shinde, A. S. Renge, G. H. Patil, A. B. Rode, R. R.
Pawar, Lett. Org. Chem. (2009) 6. (e) B. Arifin, D. F. Tang, S. S. Achmadi, Indones. J.
Chem. 15 (2015) 77-85
[5] (a) H. Y. Sun, C.F. Xiao, Y.C. Cai, Y. Chen, W. Wei, X.K. Liu, Z.L. Lv, Y. Zou, Chem.
Pharm. Bull. 58 (2010) 1492-1496. (b) M. Vivekananda, J. Ramesh, Tetrahedron Lett.
25 (1984) 3497-3500.
[6] H. Wilhelm, L.A. Wessjohann, Tetrahedron. 62 (2006) 6961-6966.
[7] S. Yamaguchi, M. Nedachi, H. Yokoyama, Y. Hirai, Tetrahedron Lett. 40 (1999) 7363-
7365.
[8] J. F. Couture, E. Collazo, P. A. Ortiz-Tello, Nat. Struct. Mol. Biol. 14 (2007) 689-695
[9] (a) H. Y. Choi, B. S. Lee, D. Y. Chi, D. J. Kim, Heterocycles. 48 (1998) 2647-2652. (b)
G. R. Pettit, H. Hoffmann, D. L. Herald, J. Org. Chem. 69 (2004) 2251-2256.
[10] (a) B. Küenburg, L. Czollner, J. Fröhlich, U. Jordis, Org Process Res Dev. 3 (1999)
425-431. (b) D. A. Chaplin, N. Fraser, P. D. Tiffin, Tetrahedron Lett. 38 (1997) 7931-
7932. (c) M. Al-. Omar, O. A. Al-Deeb, H. A. Al-Khamees, A. A. El-Emam,
phosphorus, Sulfur, and Silicon. 179 (2004) 2509-2517.
[11] (a) P. P. Kulkarni, A. J. Kadam, R. B. Mane, Uday V. Desai, J. Chem. Research (S).
(1999) 394-395. (b) V. V. Russkikh, A. V. Konstantinova, Russ. J. Org. Chem. 41
(2005) 57-60. (c) N. Zhang, B. Wu, A. Wissner, D. W. Powell. Bioorg Med Chem Lett.
12 (2002) 423–425. (d) T. Sugaya, Y. Mimura, Y. Shida, J .Med. Chem. 37 (1994)
1028-1032.
[12] L. Zuo, S. Yao, W. Wang, W. Duan, Tetrahedron Lett. 49 (2008) 4054-4056.
[13] T. Tetsuya, O. Tsuyaho, O. IWAO, Tetrahedron Lett. 60 (2004) 10681-10693.
19

ACCEPTED MANUSCRIPT
[14] M. V. Bhatt, S. S. El-Morey, Communication synthesis. (1982) 1048-1050
[15] S. Hanessian, Y. Guindon, Tetrahedron Lett. 21 (1980) 2305-2308
[16] J. R. McCarthy, J. L. Moore, R. J. Cregge, Tetrahedron Lett. 52 (1978) 5183-5186.
[17] (a) S. C. Cho, M. Z. Sultan, S. S. Moon, Bull. Korean Chem. Soc. 29 (2008) 1587-1590.
(b) M. E. Jung, M. A. Lyster, J. Org. Chem. 42 (1977) 3761-3764.
[18] J. Keith, R. Xavier, R. Sharon, Org. Biomol. Chem. 1 (2003) 4380-4383.
[19] J. R. Hwu, F. F. Wong, J. J. Huang, S. C. Tsay, J. Org. Chem. 62 (1997) 4097-4104.
[20] J. I. Minamikawa, A. Brossi, Can. J. Chem. 57 (1979) 1720- 1722
[21] (a) C. F. Fonseca Guerra, J. G. Snijders, G. te Velde, E. J. Baerends, Towards an order-
N DFT method, Theor. Chem. Acc. 99 (1998) 391. (b) G. te Velde, F. M. Bickelhaupt,
S. J. A. van Gisbergen, C. F. Guerra, E. J. Baerends, J. G. Snijders, T. Ziegler,
Chemistry with ADF, Jour. Comput. Chem. 22 (2001) 931.
[22] ADF2014.07, SCM; Theoretical Chemistry, Vrije University: Amsterdam, The
Netherlands. http://www.scm.com
[23] (a) A. D. Becke, J. Chem. Phys. 98 (1993) 5648-5652. (b) C. Lee, W. Yang, R. G. Parr,
Phys. Rev. B. 37 (1988) 785-789.
[24] (a) K. Morokuma, J. Chem. Phys. 55 (1971) 1236. (b) K. Kitaura, K. Morokuma, Int. J.
Quantum Chem. 10 (1976) 325. (c) T. Ziegler, A. Rauk, Theor. Chim. Acta. 46 (1977)
1.
[25] O. Meth-Cohn, B. Narine, B. Tarnowsky, J. Chem. Soc. Perkin Trans 1. (1981) 1520-
1530
[26] R. Gatti, V. Cavrini, P. Roveri, D. Matteuzzi, P. Brigidi, Eur. J. Med. Chem. Chim.
Ther. (5), (1984) 468- 474.
[27] S. Upadhyay, A. Chandra, R. M. Singh, Indian. J. Chem. 48 (2009) 152-154
[28] R. S. Mulliken, J. Chem. Phys. 23 (1955) 1833.
20

ACCEPTED MANUSCRIPT
Highlights
•
•
•
•
Regioselective demethylation of quinoline derivatives was achieved.
No demethylation at position 7 whatever conditions reaction used when R acceptor.
DFT/B3LYP computations have been carried out on different acceptor and donor groups.
Theoretical results are in good agreement with experimental achieved for R acceptor.