Radical migration of substituents of aryl groups on quinazolinones derived from N-Acyl cyanamides

Article abstract of DOI:10.1021/ja910653k

A newly designed radical cascade involving N-acyl cyanamides is reported. It builds on aromatic homolytic substitutions as intermediate events and leads to complex heteroaromatic structures via an unprecedented radical migration of a substituent on aryl groups of quinazolinones (hydrogen or alkyl). Mechanistic considerations are detailed, which allowed us to devise fine control over the domino processes. The latter could be predictably stopped at several stages, depending on the reaction conditions. Finally, a surgical introduction of a trifluoromethyl substituent on a quinazolinone was achieved via the reported migration.

Full text of DOI:10.1021/ja910653k

Published on Web 03/05/2010
Radical Migration of Substituents of Aryl Groups on
Quinazolinones Derived from N-Acyl Cyanamides
Marie-He´le`ne Larraufie, Christine Courillon,† Cyril Ollivier, Emmanuel Lacoˆte,*
Max Malacria,* and Louis Fensterbank*
UPMC UniV Paris 06, Institut Parisien de Chimie Mole´culaire (UMR CNRS 7201), C. 229,
4 Place Jussieu, 75005 Paris, France
Received December 17, 2009; E-mail: emmanuel.lacote@upmc.fr; max.malacria@upmc.fr;
louis.fensterbank@upmc.fr
Abstract: A newly designed radical cascade involving N-acyl cyanamides is reported. It builds on aromatic
homolytic substitutions as intermediate events and leads to complex heteroaromatic structures via an
unprecedented radical migration of a substituent on aryl groups of quinazolinones (hydrogen or alkyl).
Mechanistic considerations are detailed, which allowed us to devise fine control over the domino processes.
The latter could be predictably stopped at several stages, depending on the reaction conditions. Finally, a
surgical introduction of a trifluoromethyl substituent on a quinazolinone was achieved via the reported
migration.
Scheme 1. Formation of Quinazolinones via Radical Cascades
1. Introduction
Since the seminal total syntheses of (-)-heliotridine
(Hart),1 hirsutene (Curran),2 and prostaglandin F2R (Stork),3
radical chemistry has been fully accepted in the forum of
natural product synthesis.4 In particular, radicals are well
suited for rapid increase of molecular complexity from simple
starting materials via domino processes.5
En route toward the total synthesis of luotonin A, we
introduced N-acyl cyanamides as new radical acceptors.6 This
synthesis and the methodological examination that it spawned
featured a final aromatic homolytic substitution (Scheme 1).7
Interestingly, when o,o’-dimethyl derivative 1b was em-
ployed, a des-methyl quinozaline 2b was isolated in high
yield, presumably again via aromatic homolytic substitution.8
In both cases, the homolytic substitution was the terminal
process of the cascade. Equation 1 shows its general principle.
The radical extruded from the molecules after this kind of
aromatization cannot be exploited for further complexity
increase of the carbon skeleton.9 Alternatively, a radical
aromatization can be used to generate specially designed radicals
without tin-containing reagents as illustrated by the recent work
of Walton and Studer,10 but in those cases the aromatic moiety
was a byproduct that did not participate in further reactions (eq
2).
† Present address: Universite´ Paris-Diderot, ITODYS (UMR CNRS
7086), Baˆtiment Lavoisier, C. 7090, 15 rue Jean Antoine de Ba¨ıf, 75013
Paris, France.
(1) Hart, D. J. Science 1984, 223, 883–887.
(2) Curran, D. P.; Rakiewicz, D. M. Tetrahedron 1985, 41, 3943–3958.
(3) Stork, G.; Sher, P. M.; Chen, H.-L. J. Am. Chem. Soc. 1986, 108,
6384–6385.
(4) (a) Jasperse, C. P.; Curran, D. P.; Fevig, T. L. Chem. ReV. 1991, 91,
1237–1286. (b) Koert, U. Angew. Chem., Int. Ed. Engl. 1996, 35, 405–
407. (c) Aldabbagh, F.; Bowman, W. R. Contemp. Org. Synth. 1997,
261–280. (d) Radicals in Organic Synthesis; Renaud, P., Sibi, M. P.,
Eds.; Wiley-VCH: Weinheim, Germany, 2001; Vol. 2. (e) Renaud, P.
Chimia 2001, 55, 1045–1048.
(5) For reviews: (a) Albert, M.; Fensterbank, L.; Lacoˆte, E.; Malacria,
M. Top. Curr. Chem. 2006, 264, 1–62. (b) Nicolaou, K. C.; Edmonds,
D. J.; Bulger, P. G. Angew. Chem., Int. Ed. 2006, 118, 7134–7186,
See also: (c) Curran, D. P.; Ko, S.-B.; Josien, H. Angew. Chem., Int.
Ed. 1995, 34, 2683–2684. For contributions of our own group, see:
(d) Rychlet Elliott, M.; Dhimane, A. L.; Malacria, M. J. Am. Chem.
Soc. 1997, 119, 3427–3428. (e) Bogen, S.; Fensterbank, L.; Malacria,
M. J. Am. Chem. Soc. 1997, 119, 5037–5038. (f) Devin, P.;
Fensterbank, L.; Malacria, M. J. Org. Chem. 1998, 63, 6764–6765.
(g) Marion, F.; Coulomb, J.; Servais, A.; Courillon, C.; Fensterbank,
L.; Malacria, M. Tetrahedron 2006, 62, 3856–3871.
In this article, we report a newly designed radical cascade
(6) (a) Servais, A.; Azzouz, M.; Lopes, D.; Courillon, C.; Malacria, M.
Angew. Chem., Int. Ed. 2007, 46, 576-579. For other (ionic) reactions
involving cyanamides, see: (b) Giles, R. L.; Sullivan, J. D.; Steiner,
A. M.; Looper, R. E. Angew. Chem., Int. Ed. 2009, 48, 3116–3120.
(c) Giles, R. L.; Nkansah, R. A.; Looper, R. L. J. Org. Chem. 2010,
75, 261–264. Amidinyl radicals can also be formed directly: (d)
Gennet, D.; Zard, S. Z.; Zhang, H. Chem. Commun. 2003, 1870–1871.
involving N-acyl cyanamides. It builds on aromatic homolytic
(7) (a) Studer, A.; Bossart, M. Radicals in Organic Synthesis; Renaud,
P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, Germay, 2001; Vol. 2,
pp 62-80. (b) Bowman, W. R.; Storey, J. M. D. Chem. Soc. ReV.
2007, 1803–1822.
9
10.1021/ja910653k  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 4381–4387 4381

A R T I C L E S
Larraufie et al.
Scheme 2. General Principle of the Radical Migration
substitutions as intermediate events and leads to complex
heteroaromatic structures via an unprecedented radical migration
of a substituent on aryl groups of quinazolinones (Scheme 2).
We decided to introduce a suitably placed unsaturation of the
quinazoline derived from our previous cascade (precursors of
type A, Scheme 2). After iodine abstraction from the mediator
and a regioselective 5-exo-dig cyclization, the N-iminyl radical
C would undergo cyclization onto the aromatic ring. After
rearomatization, the radical R• extruded from intermediate D
would then re-add on E and lead to the formation of a new
bond in B. Reasoning that aromatic moieties help groups
additions onto conjugated double bonds, we thought that the
best point of attachment for the unsaturation destined to
participate to the final bond forming event was on the carbon
between both nitrogen atoms.
Table 1. Formal H• Migration
entry
solvent
n
yield of 4 (%)
byproducts
1
2
3
4
5
6
PhH
1.5
1.5
1.5
1
0.5
0.25
66
6 (20%)
toluene
t-BuOH
t-BuOH
t-BuOH
t-BuOH
59a
5 (18%)a, 6 (15%)a
6 (12%)a
77 (79)a
72a
6 (13%)a
64a
5 (13%)a, 6 (11%)a
3 (6%)a, 5 (8%)a, 6 (10%)a
2. Reductions
Simple reductions (formal H• migrations) were examined first
(Table 1).
57a
a NMR yields. Sulfolene was used as internal standard.
In a typical experiment, vinyl iodide 3 was allowed to react
with 2 equiv of tributyltin hydride (TBTH) in the presence of
1.5 equiv of AIBN in refluxing benzene. TBTH was added via
syringe pump (0.2 mmol/h, entry 1). The desired quinazoline 4
was isolated in 66% yield along with 20% of a byproduct 6
arising from endocyclic migration of the double bond before
final reduction. A rapid solvent screening indicated that a higher
boiling solvent was detrimental (Entry 2). tert-Butanol gave the
best isolated yields of 4 (77%, entry 3), with 12% of 6.
Interestingly, the quinazoline obtained are biologically active
molecules or analogs thereof.11
for rearomatization after radical addition to aryl moieties because
AIBN presumably traps a H• from the intermediate radical
bimolecularly to lead to the aromatization.12 This mechanism
was initially suggested by Curran,12a and Allin et al. were the
first to evidence the key role azo initiators played as traps for
the hydrogen atom transferred from radicals like F (Scheme
3). Aromatized product G was isolated together with hydrazine
H.13
In our case, the H• formally adds onto the substrate (vide
infra), so catalytic amounts of AIBN should have been sufficient
for the reactions to proceed. Gradual lowering of the amount
Next, we focused on the amount of AIBN required. Indeed,
stoichiometric amounts of AIBN are necessary in most cases
(11) Hermecz, I.; Vasvari-Debreczy, L.; Horvath, A.; Balogh, M.; Kokosi,
J.; De Vos, C.; Rodriguez, L. J. Med. Chem. 1987, 30, 1543–1549.
(12) For arguments in favor of a rearomatization step involving a
H-abstraction mechanism by AIBN, see: (a) Curran, D. P.; Yu, H.;
Liu, H. Tetrahedron 1994, 50, 7343–7366. (b) Curran, D. P.; Liu, H.
J. Chem. Soc., Perkin Trans. 1 1994, 1377–1393. (c) Beckwith,
A. L. J.; Bowry, V. W.; Bowman, W. R.; Mann, E.; Parr, J.; Storey,
J. M. D. Angew. Chem. Int. Ed 2004, 43, 95–98. When aryl iodides
are involved in intramolecular radical additions to quinolines, a SET
process may operate and only a catalytic amount of AIBN is required.
See: (d) Harrowven, D. C.; Sutton, B. J.; Coulton, S. Tetrahedron
2002, 58, 3387–3400. (e) Harrowven, D. C.; Guy, I. L.; Nanson, L.
Angew. Chem., Int. Ed. 2006, 45, 2242–2245. See also: (f) Bowma,
W. R.; Heaney, H.; Jordan, B. M. Tetrahedron 1991, 47, 10119–10128.
Oxygen can also be responsible for the aromatization. See: (g) Curran,
D. P.; Keller, A. I. J. Am. Chem. Soc. 2006, 128, 13706–13707.
(13) (a) Allin, S. M.; Barton, W. R. S.; Russell Bowman, W.; McInally,
T. Tetrahedron Lett. 2001, 42, 7887–7890. (b) Allin, S. M.; Barton,
W. R. S.; Russell Bowman, W.; Bridge (ne´e Mann), E.; Elsegood,
M. R. J.; McInally, T.; McKee, V. Tetrahedron 2008, 64, 7745–7758.
(8) (a) Beaume, A.; Courillon, C.; Derat, E.; Malacria, M. Chem. Eur. J.
2008, 14, 1238–1252. For previous examples of aromatization
prompted by alkyl and alkoxyl radical cleavage, see: (b) Rosa, A. M.;
Lobo, A. M.; Branco, P. S.; Prabhakar, S. Tetrahedron 1997, 53, 285–
298. (c) Harrowven, W. D. C.; Nunn, M. I. T.; Newman, N. A.;
Fenwick, D. R. Tetrahedron Lett. 2001, 42, 961–964. (d) Du, W.;
Curran, D. P. Synlett 2003, 1299–1302. (e) Bowman, W. R.; Fletcher,
A. J.; Lovell, P. J.; Pedersen, J. M. Synlett 2004, 1905–1908. (f)
Binmore, G.; Cardellini, L.; Walton, J. C. J. Chem. Soc., Perkin Trans.
2 1997, 757–762.
(9) A rapid radical H-atom donor such as PhSeH can block the aroma-
tization, see: Crich, D.; Grant, D.; Krishnamurthy, V.; Patel, M. Acc.
Chem. Res. 2007, 40, 453–463.
(10) For a review, see: (a) Walton, J. C.; Studer, A. Acc. Chem. Res. 2005,
38, 794–802. For representative reports: (b) Guin, J.; Fro¨hlich, R.;
Studer, A. Angew. Chem., Int. Ed. 2008, 47, 779–782. (c) Studer, A.;
Amrein, S.; Schleth, F.; Schulte, T.; Walton, J. C. J. Am. Chem. Soc.
2003, 125, 5726–5733. (d) Baguley, P. A.; Binmore, G.; Milne, A.;
Walton, J. C. Chem. Comm. 1996, 2199–2200. (e) ref 8f.
9
4382 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010

Radical Migrations on Quinazolinones
A R T I C L E S
Scheme 3. Evidence for H• Trapping in the Literaturea
the mediator (Scheme 5). This key hydrogen transfer is
reminiscent of the Mayo mechanism of initiation in the thermal
polymerization of styrene.14
Following the formation of intermediate cyclohexadienyl
radical 11,8 there are three distinct possibilities for the migration
to take place (Scheme 6). The hydrogen atom can migrate in
an intramolecular fashion (path A). It can ꢀ-eliminate from 11,
and subsequently re-add at the exomethylenic carbon of 8 (path
B).15 The transfer can occur in a bimolecular fashion. In this
latter scenario, some 8 would initially be formed from a transfer
of H• from 11 to AIBN. Then, 8 would abstract an H• from 11
thus leading to final radical 12 and another molecule of 8 (path
C).
In order to sort between path A and B/C, we carried out
double labeling experiments. Cyanamides 9 (pentadeuterated)
and 13 (p-methoxy) were chosen as substrates for this purpose.
Treated under the optimized cascade conditions devised previ-
ously, cyanamide 13 led to the expected product 14 in 72%
yield in refluxing t-butanol (Scheme 7).
a See ref 13.
of AIBN employed slightly eroded the efficiency of the reaction.
The NMR yields of 4 decreased from 79% to 72%, 64%, and
57% when the loading was decreased from 1.5 equiv to 1, 0.5,
and 0.25 equiv (entries 3-6). Below 1 equiv, some byproducts
arising from reduction without cyclization (5, 8%) as well as
some starting material (6%) were isolated.
We next carried out the reaction with half an equivalent each
of 9 and 13. If the migration had been fully intramolecular (path
A), then 9 should have led to 10 only via deuterium migration,
whereas 13 would have yielded 14, as before. Thus an equimolar
mixture of 10 and 14 would have been expected. On the other
hand, if the reaction followed paths B or C, then one would
have expected exchange, and the formation of 15 and 16.
In the event, an equimolar mixture of the four possible
compounds was isolated (Scheme 8). Consequently, path A
could be excluded, but we could not discriminate between the
two remaining paths at that stage.
3. Mechanistic Considerations
When cyanamide 3 was treated with tributyltin deuteride,
quinazolinone 7 was isolated in 55% yield with 83% deuterium
incorporation, as measured by 1H NMR (Scheme 4), along with
a minor amount of exo-methylene compound 8 (i.e., the product
expected from the aborted cascade). The deuterium is on the
five-membered ring. On the other hand, pentadeuterated cyan-
amide 9 led to quinazolinone 10 (48% yield in benzene, 59%
in t-butanol), where the deuterium is on the methyl group (>95%
deuterium incorporation measured by MS).
4. Interrupted Cascades
As mentioned earlier, the final reduction of the exomethylene
compound would arise from trapping of H• by a molecule of 8
formed in the early stages of the reaction. We imagined that
the bimolecular approach required in path C would be more
Those labeling experiments thus confirmed that the reduction
of the double bond proceeds via migration of a hydrogen atom
from the benzoyl group to the exo methylene moiety, followed
by radical termination by transfer of a second hydrogen from
Scheme 4. Labeling Experiments for the Migration of H•
Scheme 5. Origin of the Two Hydrogens in Quinazolinone 4
9
J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010 4383

A R T I C L E S
Larraufie et al.
Scheme 6. Possible Mechanisms for the H• Migration
Scheme 7. Modification of the Aromatic Ring
From a synthetic point of view, these results suggested that
the cascade could be stopped en route to the final product. With
a view of developing conditions to control it toward either one
or the other outcome, we endeavored to devise modifications
in our reaction that would deliver cleanly the unsaturated product
coming from the interrupted process.
Introduction of substituents in the meta position of the benzoyl
groups, that is ortho to the carbon undergoing addition seemed
a potential way of again blocking the bimolecular approach,
this time from the aromatic region of the substrates.
sensitive to steric hindrance than the addition of a putative H•
to the double bond (path B).
When an additional methyl substituent was installed onto the
double bond, the efficiency dropped sharply since only 27% of
the reduced 18 was isolated (Scheme 9). The major product
was the aborted adduct 19 (58%, 5:1 ds). Similarly, substrate
20 led to 6 in 82% yield.
Since steric hindrance in those two cases should not be a
preponderant factor for the radical addition of hydrogen atoms,
we believe this suggests that path C is the operating mechanism
for the final migration (see below).
Additional methyl substituents had no effect, since 63% of
reduced 22 were obtained from 21, along with the usual
migration product 23. Introduction of t-butyl groups proved
highly detrimental (Scheme 10). Only 33% of reduced 25 was
isolated from 24. Because the overall yield is low, it is not
possible to conclude whether steric hindrance on the aromatic
ring blocked the final reduction or not.
Scheme 8. Double-Labeling Experiment
Scheme 9. Influence of the Steric Hindrance
9
4384 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010

Radical Migrations on Quinazolinones
A R T I C L E S
Scheme 10. Steric Hindrance on the Aromatic Ring
genated cyanamide 29 (45%, Scheme 11). The two expected
cyclized products were also isolated (24% overall), albeit the
aborted cascade was predominant in this case (2:1). The low
efficiciency of the cascade is consistent with a slower 6-exo-
dig cyclization, although we were surprised by the intensity of
the rate reduction. Also, the increased amount of “interrupted”
product 28 may come from a more rapid migration of the double
bond to the endo position relative to the 5-6-6 series.
Scheme 11. Influence of the Ring Size
By switching the position of the double bond in starting
cyanamide 30, we could again take advantage of the blocking
of the final migration to obtain a good yield of 28 (77%).
Competition experiments were next devised with a view of
rerouting the hydrogen transfer toward a sacrificial acceptor.
In this way, the final transfer would be aborted, delivering the
unsaturated products. Benzylidenemalonodinitrile (31) was
chosen as the sacrificial acceptor.
Reaction of 3 in the presence of 5 equiv of the dinitrile 31
resulted in a complete suppression of the reduced product and
provided instead adduct 32 (Scheme 12). However, the resulting
exomethylene product 8 was not very stable under the reaction
The initial cyclization featured a 5-exo-dig closure eventually
leading to 5-6-6 tricyclic systems. We decided to extend our
investigations to the homologous 6-6-6 system, with an initial
6-exo-dig cyclization. Reaction of 26 delivered mostly dehalo-
Scheme 12. Use of a Sacrificial H• Acceptora
a Conditions: Bu3SnH (2 equiv) and AIBN (1.5 equiv).
9
J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010 4385

A R T I C L E S
Larraufie et al.
Table 2. Radical Migration of Alkyl Substituents
.
a Very slow addition (0.06 mmol/h) of Bu3SnH is required to avoid decomposition.
conditions; thus, the isolated yield was disappointing. The
outcome with the pentadeuterated derivative 9 was very similar,
furnishing D-32. Yet, no significant rate change was observed.
This likely indicates that the reduction is not the kinetically
determinant step of the cascade, which is probably the radical
addition to the aromatic moiety.
To our pleasure, the cyclization of 17 delivered 19 in 92%
yield (approximately 1:1 mixture of olefin isomers). The
selectivity was much improved under the “aborted” condi-
tions, since there was no trace of 18 (27% under the normal
conditions).
in refluxing benzene upon slow addition of TBTH in the
presence of AIBN (45%, Table 2, entry 1).16 t-BuOH only
led to a slight increase in yield (entry 2). Other alkyl radicals
(e.g., isopropyl) also were conducive to the migration (entry
3).
The most promising outcome was provided by the migra-
tion of the trifluoromethyl radical, which was obtained in
55% yield overall (entry 5), albeit in that case the reaction
in benzene was much less efficient (24%, entry 4).17 Given
the number of fluorine-containing molecules, especially with
CF3 substitution, and nitrogen heterocycles in active phar-
maceutical ingredients, this entry into variously substituted
fluorine-containing aromatic heterocycles should be of par-
ticular importance for medicinal chemists.18
5. Carbon-Substituent Migrations
With the migration of hydrogen atoms in hand, we ap-
proached the more synthetically useful migration of alkyl
substituents. We first investigated ortho,ortho′-disubstituted
benzoyl cyanamides (Table 2).
Lastly, we ran a competition experiment between hydrogen
atom and alkyl group migration (entries 6 and 7). As was
To our delight, the title cascades were successful. In a
typical experiment, a methyl group underwent the migration
(16) As pointed out by one referee, the addition of a methyl radical on a
simple alkene is slow and propagation of the radical chain would not
be operative (see for these kinetics: Zytowski, T.; Fischer, H. J. Am.
Chem. Soc. 1996, 118, 437–439). However, we can anticipate that 8
due to its highly conjugated character is a good acceptor of alkyl
radicals.
(17) A possible side reaction can be the CF3 radical addition on benzene:
(a) Stefani, A. P.; Szwarc, M. J. Am. Chem. Soc. 1962, 84, 3661–
3666. (b) Whittemore, I. M.; Stefani, A. P.; Szwarc, M. J. Am. Chem.
Soc. 1962, 84, 3799–3803.
(14) (a) Mayo, F. R. J. Am. Chem. Soc. 1968, 90, 1289–1295. (b) Chong,
Y. K.; Rizzardo, E.; Solomon, D. H. J. Am. Chem. Soc. 1983, 105,
7761–7762.
(15) As proposed by one referee, AIBN could also serve as relay for
hydrogen delivery, through the intervention of H-AIBN adduct (see
ref 13).
9
4386 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010

Radical Migrations on Quinazolinones
A R T I C L E S
anticipated, the reaction of cyanamide 41 led to the unique
formation of reduced 42 (58% in benzene, 69% in t-BuOH).
This can be explained by the more rapid radical addition at the
less substituted carbon atom of the aromatic ring.
this is unprecedented in a context of radical cascades and
should be of use to the broad community of synthetic as well
as medicinal chemists.
Acknowledgment. Dedicated to the memory of Keith Fagnou.
We thank UPMC, CNRS, IUF (M.M. and L.F.), and ANR (Grant
BLAN0309, Radicaux Verts) for funding. M.-H.L. has been
awarded a graduate fellowship by the Re´gion Ile-de-France, which
is gratefully acknowledged. Technical assistance (MS and elemental
analyses) was generously offered by FR 2769 and ICSN (Gif sur
Yvette, France). We specially thank Prof. Dennis P. Curran
(Pittsburgh) for helpful suggestions.
6. Conclusion
We have designed a novel radical cascade involving N-acyl
cyanamides. Heteroaromatic structures can be accessed via
a domino process that builds up to two C-C and one C-N
bonds. The central feature of the reaction is the radical
migration of hydrogen atoms or carbon substituents triggered
by rearomatization of a cyclohexadienyl radical generated
by radical addition to the aromatic ring. To our knowledge
Supporting Information Available: Procedures and charac-
terization of all new compounds (including copies of the NMR
spectra). This material is available free of charge via the Internet
at http://pubs.acs.org.
(18) (a) Dolbier, W. R., Jr. J. Fluor. Chem 2005, 126, 157–163. (b) Be´gue´,
J. P.; Bonnet-Delpon, D. J. Fluor. Chem. 2006, 127, 992–1012. (c)
Kirk, K. L. J. Fluor. Chem. 2006, 127, 1013–1029. (d) Billard, T.;
Langlois, B. R. Eur. J. Org. Chem. 2007, 891–897.
JA910653K
9
J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010 4387