2-(Aminoaryl)alkanone O-phenyl oximes: Versatile reagents for syntheses of quinazolines

Article abstract of DOI:10.1039/b803630f

Microwave irradiations of 2-(aminoaryl)alkanone O-phenyl oximes and carbonyl compounds generate iminyl radicals in company with imines; iminyl on imine ring closure yields dihydroquinazolines or quinazolines when ZnCl ₂ is included in the mixtu

Full text of DOI:10.1039/b803630f

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
2-(Aminoaryl)alkanone O-phenyl oximes: versatile reagents for syntheses
of quinazolines
Fernando Portela-Cubillo,a Jackie S. Scottb and John C. Walton*a
Received (in Cambridge, UK) 3rd March 2008, Accepted 12th March 2008
First published as an Advance Article on the web 17th April 2008
DOI: 10.1039/b803630f
Microwave irradiations of 2-(aminoaryl)alkanone O-phenyl
oximes and carbonyl compounds generate iminyl radicals in
company with imines; iminyl on imine ring closure yields
dihydroquinazolines or quinazolines when ZnCl2 is included in
the mixture.
The quinazoline ring is an important structural motif in many
biologically active compounds. Interest in new ways of making
derivatives is high because anticancer, antibacterial, antic-
onvulsant and diuretic activities have been reported for var-
ious quinazoline alkaloids.1,2a–c For example, 2-substituted-4-
aminoquinazolines are known to be potent against several
bacterial strains.3 1,2-Dihydro-4-aminoquinazoline derivatives
are effective and selective inhibitors of inducible nitric oxide
synthase and show anti-inflammatory activity in vivo.4 With
the exception of the CuCl2 catalysed reaction of aldehydes
with anthranilamide,5 methods of forming the quinazoline
ring either require multi-step preparations of special re-
agents/reactants or give moderate yields.6
Scheme 1 Projected annulations of 2-aminoacetophenone O-phenyl
oxime with carbonyl compounds.
Thermolyses of model O-phenyl oxime ethers were shown to
release simple dialkyl- and diaryl-iminyl radicals together with
phenoxyl radicals. In fact, the N–O bond dissociation enthal-
pies [BDE(R2CN–OPh) B 140 kJ molÀ1 for R = Me or Ph]
were found to be lower than the O–O BDEs of dialkyl
peroxides [159–167 kJ molÀ1].7 We recently discovered that
for the special case of alkenone O-phenyl oximes, microwave
irradiation was a suitable method for release of unsaturated
iminyl radicals which ring closed to afford dihydropyrroles in
good yields.8,9
favours ring closure onto the NQC bond in the 6-endo-trig
mode because the resultant aminyl radical 4 will be strongly
resonance stabilised. Hydrogen abstraction from the toluene
solvent should then furnish dihydroquinazoline 5.
We conjectured, however, that the imine forming condensa-
tion would also be promoted by microwaves. There was a
possibility therefore of assimilating imine formation with
radical generation thus enabling the whole sequence to be
combined in one pot (route b, Scheme 1).
We reasoned that other heterocycles could probably be
made by using O-phenyl oximes containing different types of
radical acceptors in their side chains. In particular, by use of
imine functionality in the side chain, di-aza-heterocycles might
be accessible under mild, neutral conditions with all the
convenience and rapidity of a microwave-assisted process.
The O-phenyl oxime of 2-aminoacetophenone 1 can be made
in good yield by treatment of this ketone with commercially
available O-phenylhydroxylamine hydrochloride.7 Imines 2
could be made by condensation of 1 with aldehydes or ketones
in the conventional way (route a, Scheme 1). On irradiation
with microwaves the weak N–O bond of 2 should break
releasing iminyl radical 3 together with the resonance-stabi-
lised phenoxyl radical. The architecture of iminyl radical 3
To test this novel annulation sequence pent-4-enal (1 equiv.)
and oxime ether 1 (1 equiv.) in toluene solution with 1-ethyl-3-
methyl-1H-imidazol-3-ium hexafluorophosphate (emimPF6)
as ionic liquid, were irradiated with microwaves (nominally
300 MHz) in a Biotage Initiator closed reactor. Dihydroqui-
nazoline 5b (R1 = H, R2 = but-3-enyl) was indeed formed
with toluene as the H-atom donor solvent. After 30 min
irradiation the 1H NMR spectrum of the total product showed
a very clean mixture with essentially quantitative production
of 5b together with an equal amount of phenol.10 The latter is
formed by H-abstraction from solvent by the co-produced
phenoxyl radicals and is easily separated.
Table 1 shows that reactions with aliphatic aldehydes
delivered the corresponding dihydroquinazolines in high yields
(entries 1–3). Reaction of 1 with benzaldehyde gave a 72%
yield of the corresponding dihydroquinazoline (entry 4). For
other aromatic aldehydes, and for ketones, the reactions
tended to be incomplete and lower yields were obtained.
a University of St. Andrews, School of Chemistry, EaStChem, St.
Andrews, Fife, UK KY16 9ST. E-mail: jcw@st-and.ac.uk;
Fax: +44 (0)1334 463808; Tel: +44 (0)1334 463864
b GlaxoSmithKline, New Frontiers Science Park, Third Avenue,
Harlow, Essex, UK CM19 5AW
However,
cyclohexanone
afforded
an
interesting
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 2935–2937 | 2935

View Article Online
Table 1 Yields of dihydroquinazolines from reactions of 1
Entry Carbonyl compound Dihydroquinazoline Yield (mol%)a
1
2
3
4
91
94b
92b
72
PhCHO
5
78
a
Reaction conditions: emimPF6 (1 equiv.), solvent: PhMe, MW for
b
Fig. 1 Quinazolines prepared in ZnCl2 promoted reactions of 1 with
30 min at 160 1C. All yields are isolated yields. Solvent: t-BuOH.
aldehydes.
spirodihydroquinazoline (entry 5) in a very acceptable yield of
78%. The 1,2-dihydroquinazolines with only one substituent
at the 2-position (entries 1–4) slowly oxidised to the corres-
ponding quinazolines over a period of weeks when stored in
air at rt.
the fast imine formation equilibria ensure a constant supply
until reactant depletion sets in.
It is known that imine formation is promoted by inclusion
of zinc chloride.15 We surmised, therefore, that addition of
ZnCl2 as a promoter would improve the product yields when
less reactive carbonyl compounds were employed. When pent-
4-enal was reacted with 1 in the presence of 0.3 equivalents of
ZnCl2, a high yield was obtained, but the interesting outcome
was that the quinazoline, rather than the dihydroquinazoline,
was formed under these circumstances.
If H-atom abstraction from solvent by the iminyl radical
generated from 1 before imine formation had been important,
then 6a or its hydrolysis product 6b should have been formed.
Products and yields of quinazolines obtained from annula-
tions with a variety of aldehydes are shown in Fig. 1.16 High
yields of quinazolines 8a and 8b were obtained from 1 with the
aliphatic aldehydes pent-4-enal and cyclohexanecarbaldehyde,
respectively. The aromatic 4-nitrobenzaldehyde gave 8d also in
high yield (90%). Somewhat lower, but still useful, yields of
quinazolines were obtained on starting with aromatic alde-
hydes containing electron donating substituents, as 8e and 8f
demonstrate. Similarly, the annulation worked well with
pyridine-2-carbaldehyde (-8c).
Similarly, if H-atom abstraction from solvent by radical 3 had
been faster than ring closure then 7a, or its hydrolysis product
7b, should have been detected. In practise, neither 6a,b nor
7a,b were observed. Imine production from 1 and an aldehyde
takes place via a series of equilibria involving hemiaminal
formation, water loss and several protonation/deprotonation
steps. The dehydration step is acid catalysed and the rates of
individual steps depend on pH and on solvent.11 The absence
of 6a,b shows that the equilibria must be fast in comparison
with radical generation. The step 3 - 4 is a novel 6-endo-trig
cyclisation of an iminyl radical onto a CQN bond. Kinetic
data for this particular type of ring closure are not available.
However, the known rate constants for 6-endo-trig cyclisations
of C-centred radicals onto CQN bonds,12 and for iminyl
radicals onto CQC bonds,13,14 suggest 3 - 4 would be a fast
process at 160 1C with a rate constant at least equal to that of a
C-centred radical onto a CQC bond, and probably greater
because of the resonance stabilisation in 4. This is consistent
with the absence of compounds 7a,b in the product mixture.
Evidently imine is rapidly trapped by the cyclisation step but
To probe the sensitivity of the reaction to substituents in the
aryl ring, and at the 4-position of the quinazoline, O-phenyl
oxime ethers of 2-aminobenzaldehyde, 5-bromo-2-aminoace-
tophenone and 4,5-dimethoxy-2-aminoacetophenone were
prepared. The formation of quinazolines 8f, 8g and 8h all in
good yields (Fig. 1) showed that the reaction tolerates a good
range of functionality in the aminoketones as well as in the
partner aldehydes.
The change to aromatic quinazolines when ZnCl2 is in-
cluded might signal a change in mechanism. It is known from
previous work7,8 that the precursor O-phenyl oximes are
completely dissociated to radicals in o30 min under the
microwave conditions. To compete, any alternative mechan-
ism would have to be faster than this.17 Possibly the zinc bonds
ꢀc
This journal is The Royal Society of Chemistry 2008
2936 | Chem. Commun., 2008, 2935–2937

View Article Online
6. For reviews of quinazoline synthetic methods see:
(a) D. J. Connolly, D. Cusack, T. P. O’Sullivan and P. J. Guiry,
Tetrahedron, 2005, 61, 10153; (b) K. Undheim and T. Benneche, in
Comprehensive Heterocyclic Chemistry II, ed. A. J. Boulton,
Pergamon, Oxford, 1998, vol. 6, pp. 93.
7. J. A. Blake, D. A. Pratt, S. Lin, J. C. Walton, P. Mulder and
K. U. Ingold, J. Org. Chem., 2004, 69, 3112.
8. F. Portela-Cubillo, J. S. Scott and J. C. Walton, Chem. Commun.,
2007, 4041.
9. For alternative methods of generating iminyl radicals see:
(a) S. Z. Zard, Synlett, 1996, 1149; (b) L. E. Kaim and
C. Meyer, J. Org. Chem., 1996, 61, 1556; (c) K. Takai,
N. Katsura and Y. Kunisada, Chem. Commun., 2001, 1724;
(d) J. Boivin, E. Fouquet and S. Z. Zard, Tetrahedron Lett.,
1991, 32, 4299; (e) J. Boivin, A.-M. Schiano and S. Z. Zard,
Tetrahedron Lett., 1992, 33, 7849; (f) M. Kitamura, Y. Mori and
K. Narasaka, Tetrahedron Lett., 2005, 46, 2373; (g) M. Yoshida,
M. Kitamura and K. Narasaka, Bull. Chem. Soc. Jpn., 2003, 76,
2003; (h) R. Alonso, P. J. Campos, B. Garcia and
M. A. Rodriguez, Org. Lett., 2006, 8, 3521; (i) T. Mikami and
K. Narasaka, Chem. Lett., 2000, 338; (j) D. Nanni, P. Pareschi,
C. Rizzoli, P. Sgarabotto and A. Tundo, Tetrahedron, 1995, 51,
9045; (k) R. Leardini, H. McNab, M. Minozzi and D. Nanni,
J. Chem. Soc., Perkin Trans. 1, 2001, 1072.
10. The option of the aminyl radical 4b undergoing a second 5-exo-
trig cyclisation onto the butenyl chain to afford a tricyclo-radical,
with final production of a tetrahydropyrrolo[1,2-a]quinazoline, is
a clear possibility for this species. However, none of this product
was detected and evidently H-atom abstraction by 4b from the
toluene solvent is too fast for the second cyclisation to compete.
11. (a) J. Hine, M. S. Cholod and W. K. Chess Jr, J. Am. Chem. Soc.,
1973, 95, 4270; (b) J. Hine and Y. Chou, J. Org. Chem., 1981, 46,
649.
Scheme 2 Possible role of zinc chloride in quinazoline formation.
to the imine nitrogen prior to cyclisation i.e. 9 and the
subsequent ring closure gives amminium radical cation 10 in
a process akin to an iminium salt cyclisation. The adjacent
cation would considerably lower the pKa of the H-atom at
position 2 of the heterocycle 10. Proton loss would then yield
11 in
a process reminiscent of the Minisci reaction
(Scheme 2).18 Stabilised intermediate 11 might transfer an
electron to the starting oxime ether to give 12 or be converted
to 13 on exposure to oxygen during work-up. When the ZnCl2
is not present, the C–H at position 2 is not acidic, and hence
aromatisation to a quinazoline does not occur.19
Overall, this process is a two-stage synthetic route from
2-aminoarylalkenones via their O-phenyl oximes and thence
by a one pot procedure with carbonyl compounds to dihy-
droquinazolines or quinazolines. The process is of wide scope
and works well with alkyl, aryl and heterocyclic types of
aldehyde. The reaction has several advantages over existing
methods for quinazoline synthesis. It is rapid (30 min) and
requires no acids, bases, or toxic metals. It is comparatively
mild and high yielding. The O-phenyl oximes are easily made
in one step and can be stored indefinitely. Unlike many other
radical-mediated synthetic methods, no initiator is needed and
hence no by-products from initiator fragments contaminate
the system. There is evidently a promising future for micro-
wave-assistance in reactions where the initial step is homolysis
of a weak bond in a reactant molecule, O-phenyl oxime ethers
being prime examples.
12. S. Kim, K. S. Yoon and Y. S. Kim, Tetrahedron, 1997, 38, 73.
13. M.-H. Le Tadic-Biadatti, A.-C. Callier-Dublanchet, J. H. Horner,
B. Quiclet-Sire, S. Z. Zard and M. Newcomb, J. Org. Chem., 1997,
62, 559.
14. A. G. Fallis and I. M. Brinza, Tetrahedron, 1997, 53, 17543.
15. A. Pouilhes, Y. Langlois and A. Chiaroni, Synlett, 2003, 1488.
16. Experimental details for the preparation of 8d are typical of the
methodology for all the quinazolines. 4-Nitrobenzaldehyde (66
mg, 0.44 mmol) was added to a solution of 1-(2-aminophenyl)-
ethanone O-phenyl oxime 1 (100 mg, 0.44 mmol) in toluene (0.15
M), containing anhydrous ZnCl2 (17 mg, 0.13 mmol) and
emimPF6 (100 mg, 0.46 mmol) in a microwave vessel (2–5 cm3).
The vessel was sealed and subjected to microwave irradiation for
30 min at 160 1C in a Biotage Initiator system (nominally 300
MHz). After cooling, the ionic liquid was filtered off and the
toluene was removed under reduced pressure. The residue was
purified by flash column chromatography (5% EtOAc–hexane)
affording 4-methyl-2-(4-nitrophenyl)quinazoline 8d as a yellow
solid (105 mg, 90%). Mp 168–179 1C; nmax/cmÀ1 = 1597, 1569,
1548, 1340; 1H NMR (400 MHz, CDCl3), dH 2.96 (s, 3H, CH3),
7.58 (ddd, J = 8.2, 7.6, 1.2 Hz, 1H, CH), 7.84 (ddd, J = 8.5, 7.0,
1.4 Hz, 1H, CH), 8.02 (d, J = 7.8 Hz, 1H, CH), 8.04 (d, J = 7.8
Hz, 1H, CH), 8.26 (d, J = 9.0 Hz, 2H, CH), 8.72 (dt, J = 9.0, 2.0
Hz, 2H, CH); 13C NMR dC 21.5 (CH3), 122.3 (C), 122.6, 124.3,
126.9, 128.4, 128.3, 133.0 (CH), 143.0, 148.1, 149.2, 156.8, 167.8
(C); HRMS calcd for C15H12N3O2 (MH+) 266.0930, found
266.0936. Dihydroquinazolines were prepared in essentially the
same way, except that the zinc chloride was omitted.
We thank GSK and EaStChem for financial support and
Dr M. C. Clarke for loaning the microwave reactor.
Notes and references
1. J. P. Michael, Nat. Prod. Rep., 2003, 20, 476.
2. (a) J. H. Chan, J. S. Hong, L. F. Kuyper, M. L. Jones,
D. P. Baccanari, R. L. Tansik, C. M. Boytos, S. K. Rudolf and
A. D. Brown, J. Heterocycl. Chem., 1997, 34, 145;
(b) S. L. Gackenheimer, J. M. Schaus and D. R. Gehlert,
J. Pharmacol. Exp. Ther., 1995, 274, 1558; (c) R. O. Dempcy
and E. B. Skibo, Biochemistry, 1991, 30, 8480.
3. P.-P. Kung, M. D. Casper, K. L. Cook, L. Wilson-Lingardo,
L. M. Risen, T. A. Vickers, R. Ranken, L. B. Blyn, J. R. Wyatt,
P. D. Cook and D. J. Ecker, J. Med. Chem., 1999, 42, 4705.
4. A. C. Tinker, H. G. Beaton, N. Broughton-Smith, T. R. Cook,
S. L. Cooper, L. Fraser-Rae, K. Hallam, P. Framley, T. McInally,
D. J. Nicholls, A. D. Pimm and A. V. Wallace, J. Med. Chem.,
2003, 46, 913.
17. Electrocyclic ring closures of imines 2 to 3-phenoxy-2,3-dihydro-
quinazolines followed by elimination of PhOR1 or PhOR2 could
give the quinazoline products. However, precedents suggest it is
unlikely both these steps could be fast enough to compete with the
radical process (see ref. 20).
18. (a) F. Fontana, F. Minisci, M. C. N. Barbosa and E. Vismara,
J. Org. Chem., 1991, 56, 2866; (b) F. Minisci, F. Recupero,
C. Punta, C. Gambarotti, F. Antonietti, F. Fontana and
G. F. Pedulli, Chem. Commun., 2002, 2496.
5. R. J. Abdel-Jalil, W. Volter and M. Saeed, Tetrahedron Lett.,
2004, 45, 3475.
19. We thank an anonymous referee for suggesting this mechanism.
20. B. M. Trost and A. C. Gutierrez, Org. Lett., 2007, 9, 1473.
¨
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 2935–2937 | 2937