Tandem Synthesis of 10-Dimethylaminobenzo[h]quinazolines from 2-Ketimino-1,8-bis(dimethylamino)naphthalenes via Nucleophilic Replacement of the Unactivated Aromatic NMe2 Group

Article abstract of DOI:10.1021/acs.orglett.6b01178

It has been found that 2-bromo-1,8-bis(dimethylamino)naphthalene on sequential treatment with n-BuLi and 2 equiv of the same or different aryl(hetaryl) cyanide as a result of [2 + 2 + 2] nucleophilic cascade annulation produces 10-dimethylaminobenzo[h]quinazolines, as yet unknown NMe₂/-N= analogues of the proton sponge. It is even more convenient to use preliminarily prepared 2-ketimino-1,8-bis(dimethylamino)naphthalenes as starting material. The substitution of both peri-NMe₂ groups furnishing quinazolino[7,8-h]quinazoline derivatives is also possible. The process is remarkable by surprisingly mild nucleophilic displacement of an unactivated aromatic NMe₂ group.

Full text of DOI:10.1021/acs.orglett.6b01178

Letter
pubs.acs.org/OrgLett
Tandem Synthesis of 10-Dimethylaminobenzo[h]quinazolines from
2‑Ketimino-1,8-bis(dimethylamino)naphthalenes via Nucleophilic
Replacement of the Unactivated Aromatic NMe2 Group
Vladimir Y. Mikshiev, Alexander S. Antonov, and Alexander F. Pozharskii*
Department of Organic Chemistry, Southern Federal University, Zorge str. 7, 344090 Rostov-on-Don, Russian Federation
S
* Supporting Information
ABSTRACT: It has been found that 2-bromo-1,8-bis-
(dimethylamino)naphthalene on sequential treatment with n-
BuLi and 2 equiv of the same or different aryl(hetaryl) cyanide
as a result of [2 + 2 + 2] nucleophilic cascade annulation
produces 10-dimethylaminobenzo[h]quinazolines, as yet un-
known NMe2/−N analogues of the proton sponge. It is
even more convenient to use preliminarily prepared 2-
ketimino-1,8-bis(dimethylamino)naphthalenes as starting material. The substitution of both peri-NMe2 groups furnishing
quinazolino[7,8-h]quinazoline derivatives is also possible. The process is remarkable by surprisingly mild nucleophilic
displacement of an unactivated aromatic NMe2 group.
1:RCN ratio to furnish, after quenching the crude reaction
ecause of the extremely high basicity of the dimethylamide
Banion (pK ≈ 35), the dimethylamino group is considered mixture with water, the corresponding imine 2 in good yield. In
a
some experiments, we also fixed by thin-layer chromatography
the appearance of a yellow spot going in front of a reddish-
colored zone of imine 2 (Supporting Information).
one of the worst leaving groups in nucleophilic displacement
reactions. Nevertheless, there exist quite a few examples of its
substitution. All of them are based on preliminary activation of
the NMe2 group via placing in a substrate molecule the strong
electron-accepting functionalities (NO2, COCF3, etc.),1 its
At first, we did not attach importance to this, believing that it
was a small admixture of ketone, which we usually received
separately and in good yield by acid hydrolysis of the
corresponding imine.4 Nonetheless, once (in experiment with
PhCN) we had studied a structure of the yellow byproduct, to
our surprise, we found that it was an unknown 10-
dimethylamino derivative of benzo[h]qunazoline 3 (R = Ph),
an intriguing representative of almost unexplored5 peri-NMe2/
−N analogues of the proton sponge. In additon, this reaction
is a new synthetic way to quinazolines as an important class of
nitrogen heterocycles having a considerable pharmaceutical
significance.6
Following the most probable mechanism of the formation of
3 shown in Scheme 2, the process should involve 2 equiv of
nitrile. With this point in mind, it seemed to us reasonable to
begin directly from authentic imine 2 already having in its
structure half of the necessary nitrile component. Accordingly, a
series of 2 was treated at −20 °C with butyllithium (for
bromides 2g,h with LDA) followed by 1.1 molar equiv of the
corresponding nitrile. The subsequent cyclization was con-
ducted at room temperature.
conversion into an NMe3 group,2 or acid catalysis.3 The two
+
last methods allow the NMe2 group to be eliminated in a form
of relatively stable trimethyl- or dimethylammonium ions. In
particular, the acid catalysis can be exemplified by our previous
observation that azomethines, hydrazones, and oximes derived
from 2(7)-carbonyl derivatives of 1,8-bis(dimethylamino)-
naphthalene (proton sponge) undergo intramolecular hetero-
cyclizations accompanying by substitution of the 1-NMe2 group
to produce a number of difficulty accessible heterocyclic
compounds and 1-naphthol derivatives.3 Herein, we report that
similar cyclizations can be surprisingly and easily realized under
strongly basic conditions when the NMe2 group is formally
unactivated and displaced as the Me2N− anion.
The starting point for this work was our recent synthesis of
unexpectedly stable 1,8-bis(dimethylamino)naphthalene-2-keti-
mines 2 on treatment of 2-lithium derivative 1 with various
nitriles (Scheme 1).4 We usually employed the equimolar
Scheme 1. Reaction of 2-Lithio-1,8-
bis(dimethylamino)naphthalene with Nitriles
As seen from Table 1, such a procedure provides a moderate
to high yield of 3 and can be applied to a variety of aromatic
and heteroaromatic nitriles (SI). Unfortunately, the method
does not work for the most alkyl cyanides owing to their
conversion under strongly basic conditions into a low
Received: April 22, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b01178
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Possible Mechanism for the Formation of 3
Figure 1. Molecular structure of 3a.
a
Table 1. Interaction of Ketimines 2 with Base and Nitriles
Figure 2. Molecular structure of 3k.
cycle. Compounds 3k and 3l are also valuable because the
presence of the bromine atom in position 9 permits their
additional functionalization. Indeed, using this opportunity, we
have transformed 3k in a one-flask manner into compound 7
(SI, S6), which is the first derivative of the unknown
quinazolino[7,8-h]quinazoline system (Scheme 3).
imine
base
nitrile (R1CN)
PhCN
product (mp, °C) yield (%)
2a
2a
2a
2b
2b
2c
2c
2d
2e
2f
n-BuLi
n-BuLi
n-BuLi
n-BuLi
n-BuLi
n-BuLi
n-BuLi
n-BuLi
n-BuLi
n-BuLi
LDA
3a (101−102)
3b (130−131)
3c (113−114)
3d (131−133)
3e (144−146)
3f (127−129)
3g (39−42)
83
36
40
60
77
61
64
42
87
36
40
47
4-MeOC6H4CN
3-cyanopyridine
4-MeOC6H4CN
PhCN
Scheme 3. Synthesis of 2,4,9,11-
Tetraphenylquinazolino[7,8-h]quinazoline 7
2-cyanothiophene
3-cyanopyridine
1-cyanonaphthalene
PhCN
3h (212−215)
3i (108−109)
3j (77−78)
2-cyanofuran
PhCN
2g
2h
3k (164−165)
3l (155−156)
LDA
4-MeOC6H4CN
a
For chromatographically pure isolated sample.
One of the central points of our study consists of questioning
why the NMe2 group in ketimines 2 is replaced under such
mild conditions, in contrast with those which are commonly
demanded in all of the above cited cases.1−3 We believe that a
combination of factors is responsible for this. The first, and
perhaps less important, is metallocatalysis. Recently, we
established that 2-lithium sponge 1 exists in the solid as a
dimer with the out-inverted 1-NMe2 group forming the
coordination bond N → Li.8 One can assume that similar
coordination might activate the NMe2 group for replacement as
shown in structures 5a and 5b (Scheme 2). The second and
probably much more essential factor can be connected with the
extremely low stability of σ-complex 6 due to its steric crowding
and oversaturation by lone electron pairs of the nitrogen atoms
together with the negatively charged naphthalene ring. The
third and likely decisive factor is the high basicity
(nucleophilicity) of the side nitrogen atoms in lithium imides
4 and 5. Indeed, as we have shown recently,4 the peri-NMe2
groups in 2 are effectively conjugated with the exocyclic CN
electrophilic carbanion. The only exception was the synthesis of
compound 3i with a tert-butyl group in position 4. Our attempt
to introduce the t-Bu group in position 2 failed, apparently due
to steric reasons. We also tried to utilize a one-pot modification
of the synthesis starting directly from 1, but it gave comparable
yields of 3 only upon using the same nitrile on both stages. In
addition, isolation and purification of the product in these cases
were rather complicated.
Structures of all compounds 3 were confirmed by elemental
analysis and spectral measurements (SI, Figures S1−S32). For
quinazolines 3a and 3k, the single-crystal X-ray studies were
also performed (Figures 1 and 2; SI, Table S1). As expected (cf.
ref 7), a strong “buttressing effect” appeared in bromide 3k. In
particular, it results in considerable flattening the 10-NMe2
group (ΣN3 value for 3k is of 359.5° against 343.2° for 3a) and
elongation of the N1···N3 distance (2.828 Å vs 2.746 Å for 3a)
due to pressure of the N-methyl groups on the pyrimidine
B
DOI: 10.1021/acs.orglett.6b01178
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
bond. This leads to an appreciable overflow of some electron
density to the imino group and even to preferable protonation
of the latter over peri-NMe2 groups. There are no doubts that
such electron transfer in conjunction with the negative charge
on terminal nitrogen atoms in imides 4 and 5 should greatly
enhance nucleophilicity of the side chains.9 Interestingly, that
one of the reviewers has suggested treating transiton 5 → 6 as
the 6π-electrocyclization event, which may also help explain its
facility.
We believe the rate-limiting step in conversion 2 → 3 is the
formation of 5 from 4, while the stages of nucleophilic addition
5 → 6 and aromatization 6 → 3 proceed relatively fast. In
particular, this is confirmed by a strong slowing reaction at the
interaction of imine 2a with less electrophilic 4-methoxybenzo-
nitrile to produce 3b (Figure 3, lower curve). A significant
demonstration of unusual ease of nucleophilic displacement of
the formally unactivated aromatic NMe2 group on the last stage
of the process, (2) the first example of nucleophilic substitution
of the two NMe2 groups in the same substrate, and (3) a new
approach to the synthesis of difficultly accessible benzo[h]-
quinazoline derivatives. Several arguments were put forward
that the success of the disclosed transformations originates
from the proton sponge nature of the used substrates.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.6b01178.
Technical details and experimental procedures; spectral
data for all new compounds (PDF)
Crystallographic data for 3a and 3k (CIF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: apozharskii@sfedu.ru.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Russian Foundation for Basic
Research (project No 14-03-00010). We thank Drs. Anna V.
Gulevskaya and Valery A. Ozeryanskii (Department of Organic
Chemistry, Southern Federal University) for helpful discus-
sions, the Scientific and Educational Laboratory of Resonance
Spectroscopy (Department of Natural and High Molecular
Compounds Chemistry, Southern Federal University) for
NMR measurements, and Dr. Kyrill Y. Suponitsky (A. N.
Nesmeyanov Institute of Organoelement Compounds, Mos-
cow) for X-ray measurements.
Figure 3. NMR monitoring of conversion 2 → 3 for some selected 2-
ketimines.
slowdown in the process also occurs after about 50−60%
conversion of 2. We attribute this to the gradual accumulation
in the reaction mixture of the dimethylamide anion, which
reacts with nitrile to convert it into an amidine byproduct
(experimental evidence: SI, Figure S33).
Of three aforementioned factors favoring easy nucleophilic
substitution of the 1-NMe2 group in imides 4 at their
interaction with nitriles, at least two, the second and the
third, stem from the proton sponge nature of the substrates. In
accord with this point, we found that this reaction cannot be
applied to benzene series. Thus, on treatment of 2-
(dimethylamino)benzophenonimine10 with n-BuLi and then
PhCN, only the starting material was recovered from the
reaction mixture. Finally, it should be noted that Li et al. have
recently elaborated copper-catalyzed synthesis of quinazolines
via [2 + 2 + 2] annulations of diaryliodonium salts and two
nitriles.11 Their approach is based exclusively on electrophilic
reactions and does not address the displacement of the NMe2
group.
In summary, it was found that interaction of 2-lithio-1,8-
bis(dimethylamino)naphthalene with an excess of aryl or
hetaryl cyanide proceeds as a [2 + 2 + 2] nucleophilic cascade
annulation of two nitrile molecules to a naphthalene system and
results in the formation of 10-(dimethylamino)benzo[h]-
quinazoline derivatives. However, the best way for carrying
out the reaction is to start from preliminarily prepared 2-
ketimino-1,8-bis(dimethylamino)naphthalenes. It allows not
only the use of two different nitriles but also enhances the
yield of the reaction product up to 36−87%. The most
innovative findings of the work are considered to be (1) a
REFERENCES
■
(1) Hojo, M.; Masuda, R.; Okada, E.; Miya, H. Synthesis 1989, 1989,
870−873. (b) Sekiguchi, S.; Suzuki, T.; Hirosawa, Y.; Ishikura, H. J.
Org. Chem. 1990, 55, 1829−1834. (c) Okada, E.; Otsuki, Y.;
Shinohara, M.; Medebielle, M.; Shimizu, Y.; Takeuchi, H. Tetrahedron
Lett. 2003, 44, 741−744. (d) Sekiguchi, S.; Suzuki, T.; Hosokawa, M. J.
Chem. Soc., Perkin Trans. 2 1989, 1783−1788. (e) Sekiguchi, S.;
Hosokawa, M.; Suzuki, T.; Sato, M. J. Chem. Soc., Perkin Trans. 2 1993,
1111−1118.
(2) Zhang, H.; Hagihara, S.; Itami, K. Chem. - Eur. J. 2015, 21,
16796−16800.
(3) Povalyakhina, M. A.; Antonov, A. S.; Dyablo, O. V.; Ozeryanskii,
V. A.; Pozharskii, A. F. J. Org. Chem. 2011, 76, 7157−7166.
(4) Antonov, A. S.; Mikshiev, V. Y.; Pozharskii, A. F.; Ozeryanskii, V.
A. Synthesis 2014, 46, 3273−3282.
(5) Pozharskii, A. F.; Ozeryanskii, V. A.; Antonov, A. S.; Mikshiev, V.
Y.; Chernyshev, A. V.; Metelitsa, A. V.; Fedik, N. S.; Borodkin, G. S.;
Dyablo, O. V. J. Org. Chem. 2016, submitted for publication.
(6) For recent reviews, see: (a) Khan, I.; Ibrar, A.; Ahmed, W.; Saeed,
A. Eur. J. Med. Chem. 2015, 90, 124−169. (b) Shakhidoyatov, Kh. M.;
Elmuradov, B. Zh. Chem. Nat. Compd. 2014, 50, 781−800. (c) Bilbro,
J.; Mart, M.; Kyprianou, N. Anticancer Res. 2013, 4695−4700.
(d) Wang, D.; Gao, F. Chem. Cent. J. 2013, 7, 95. (e) Zhang, Y.;
Wang, P.; Hu, D.; Jin, L.; Yang, S. Youji Huaxue 2012, 32, 444−461.
(7) (a) Pozharskii, A. F.; Ryabtsova, O. V.; Ozeryanskii, V. A.;
Degtyarev, A. V.; Kazheva, O. N.; Alexandrov, G. G.; Dyachenko, O. A.
J. Org. Chem. 2003, 68, 10109−10122. (b) Pozharskii, A. F.;
C
DOI: 10.1021/acs.orglett.6b01178
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Povalyakhina, M. A.; Degtyarev, A. V.; Ryabtsova, O. V.; Ozeryanskii,
V. A.; Dyablo, O. V.; Tkachuk, A. V.; Kazheva, O. N.; Chekhlov, A. N.;
Dyachenko, O. A. Org. Biomol. Chem. 2011, 9, 1887−1900.
(8) Antonov, A. S.; Pozharskii, A. F.; Ozeryanskii, V. A.; Filarowski,
A.; Suponitsky, K. Yu.; Tolstoy, P. M.; Vovk, M. A. Dalton Trans.
2015, 44, 17756−17766.
(9) Using the database Reaxys, we have found in the literature only
one mention concerning 1,3-diazabutadiene derivatives: Sharma, P.;
Kumar, A.; Rane, N.; Gurram, V. Tetrahedron 2005, 61, 4237−4248
These derivatives were employed as rather active reagents in the [4 +
2]-cycloaddition reactions. Actually, the 1,3-diazabutadiene structural
fragment in 5 can be considered an N-arylidenamidine known as a
highly basic compound already as neutral. Evidently, in anionic form
their basicity and nucleophilicity are greatly increased..
(10) Ozeryanskii, V. A.; Pozharskii, A. F.; Antonov, A. S.; Filarowski,
A. Org. Biomol. Chem. 2014, 12, 2360−2369.
(11) Su, X.; Chen, C.; Wang, Y.; Chen, J.; Lou, Z.; Li, M. Chem.
Commun. 2013, 49, 6752−6754.
D
DOI: 10.1021/acs.orglett.6b01178
Org. Lett. XXXX, XXX, XXX−XXX