Unusual radical addition on a heteroaromatic nitrogen. A convenient access to new pyrimidine derivatives

Article abstract of DOI:10.1039/c0cc00533a

A radical generated on the side-chain of a 2- or 4-N-alkylamino-4,6- or 2,6-dichloropyrimidine can cyclise on the nitrogen or on the carbon of the pyrimidine ring depending on its position and on whether an acetyl or a Boc group is present on the extra-nuclear nitrogen.

Full text of DOI:10.1039/c0cc00533a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Unusual radical addition on a heteroaromatic nitrogen. A convenient
access to new pyrimidine derivativeswz
Yann Laot, Laurent Petit and Samir Z. Zard*
Received 22nd March 2010, Accepted 9th June 2010
DOI: 10.1039/c0cc00533a
A radical generated on the side-chain of a 2- or 4-N-alkylamino-
4,6- or 2,6-dichloropyrimidine can cyclise on the nitrogen or on
the carbon of the pyrimidine ring depending on its position and
on whether an acetyl or a Boc group is present on the extra-
nuclear nitrogen.
While developing a new radical based synthetic route to
azaindolines, we made an unexpected observation. Whereas
the radical ring-closure of N-acetyl xanthate derivative 1a
furnished the expected azaindoline 2a in good yield upon
treatment with stoichiometric amounts of lauroyl peroxide,
the N-Boc analogue 1b produced very little of the corresponding
azaindoline 2b (Scheme 1). Instead, we were surprised to
isolate pyridone 3 as the major product.1 For reasons still
unclear, the mere replacement of an acetyl group by a carbamate
encourages the intermediate carbon radical to undergo a
hitherto unprecedented attack on the pyridine nitrogen.2 We
also found that the presence of the chlorine (or a fluorine) is
necessary for this transformation.
Scheme 2
afforded the expected adducts 9a–g in generally good yield
(Table 1).4
We were pleased to find that treatment of these xanthates
with stoichiometric amounts of lauroyl peroxide in ethyl acetate
resulted indeed in cyclisation on the nitrogen atom, giving rise
to the various pyrimidinones 10a–g displayed in Table 1.
Two comments are worth making. The first relates to the
yields, which are generally significantly higher than in the
pyridine series, reflecting a cleaner reaction and fewer side-
products. The second, more important observation concerns
the formation of only one isomeric pyrimidinone. According
to the mechanistic manifold pictured in Scheme 3, cyclisation
of radical 4 leads to radical 11, and this is then oxidised into
the corresponding allylic cation 12 by electron transfer to the
peroxide.
In order to better understand this new reactivity mode, we
examined the behaviour of radical 4 attached to position 2 of
the pyrimidine ring. Because of the symmetry, this radical has
no choice but to cyclise on either of the two nitrogen atoms,
unless the presence of the N-acetyl (rather than a carbamate)
somehow intrinsically inhibits the ring-closure, possibly
through unfavourable geometrical constraints.
The synthesis of the required precursors was accomplished
starting from 2,4,6-trichloropyrimidine 5. Substitution of one
chlorine with allylamine furnished an easily separable 1 : 2
mixture of the desired 2-N-allylamino-4,6-dichloropyridine 6a
and its regioisomer 6b (Scheme 2).3 Acetylation of the former
gave 7a and radical addition with a number of xanthates 8a–g
Quenching with the laurate anion could thus give rise in
principle to two regioisomeric pyrimidinones 10 and 13,
depending on whether attack occurs at C-4 or C-6 of the
pyrimidine ring. The former is the least hindered position and
presumably the most favoured for attachment of the laurate
residue.
The fact that the monosubstitution reaction of 2,4,6-tri-
chloropyrimidine 5 with allylamine produces two regioisomers
6a and 6b allowed us to contrast the above results with those
from the regioisomeric N-acetyl-N-allyl-4-aminopyrimidine
derivative 7b. Thus, while the intermolecular addition of the
xanthates proceeded in the usual manner, albeit somewhat
less efficiently, the ring-closure of the resulting adducts 14a–e
occurred only on the carbon to give diazaindolines 15a–e
(Scheme 4). No cyclisation on the nitrogen was observed.
However, when the acetyl group in 7b was replaced by a
Boc group, the reaction of the corresponding xanthate adduct
16 gave a mixture of C-cyclised product 17 and two regioisomeric
N-cyclised products 18 and 19 in a modest combined yield of
66%. The last two compounds were difficult to separate and
only 19 could be obtained in sufficiently pure form for
complete characterisation. The quenching of the intermediate
cationic species by the laurate is obviously not very selective in
Scheme 1
`
Laboratoire de Synthese Organique, CNRS UMR 7652,
Ecole Polytechnique, 91128 Palaiseau, France.
E-mail: zard@poly.polytechnique.fr; Fax: +33 16933 5972;
Tel: +33 16933 5971
w This paper is dedicated with affection to the memory of our friend,
Prof. Pascal Le Floch.
z Electronic supplementary information (ESI) available: Experimental
details and copies of spectra. See DOI: 10.1039/c0cc00533a
ꢀc
This journal is The Royal Society of Chemistry 2010
5784 | Chem. Commun., 2010, 46, 5784–5786

View Article Online
Table 1 Synthesis of pyrimidinones 10a–g.
Xanthates (8a–g) Adducts
Pyrimidinones
Scheme 3
Scheme 4
this series, in contrast to the reaction leading to pyrimidinones
10. In any event, the very unusual and still little understood role
of the carbamate manifests itself strongly, even if the dichotomy
in reactivity is not as clear-cut as in the pyridine case.
Beyond the mechanistic insight provided by these preliminary
experiments, this metal-free radical based process represents a
new entry into pyrimidinones 10 and diazaindolines 15. Both
of these motifs are highly privileged structures by medicinal
chemists.5 The present route is convergent, uses very cheap
reagents, and allows the construction of numerous new and
highly functionalised pyrimidone scaffolds. Diversity can be
easily introduced by modifying the starting xanthate 8 or by
modifying the end product 10, either through further trans-
formation of the numerous functional groups that can be
introduced by the xanthate component or by substitution of
the chlorine atom with various nucleophiles. In the case of
diazaindolines 15, the two C–Cl residues would be expected to
offer the opportunity to exploit very different reactivity levels.
Notes and references
1 M. El Qacemi, L. Ricard and S. Z. Zard, Chem. Commun., 2006, 4422.
2 We wish to acknowledge in this respect an ongoing collaboration
with Dr Michele Coote, a theoretical chemist at the Australian
National University.
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 5784–5786 | 5785

View Article Online
3 M. Koga and S. Schneller, J. Heterocycl. Chem., 1992, 29, 1741.
4 For reviews of the xanthate transfer, see: S. Z. Zard, Angew. Chem.,
Int. Ed. Engl., 1997, 36, 672; S. Z. Zard, in Xanthates and Related
Derivatives as Radical Precursors in Radicals in Organic Synthesis,
ed. P. Renaud, and M. P. Sibi, Wiley-VCH, Weinheim, 2001, vol. 1,
p. 90; B. Quiclet-Sire and S. Z. Zard, Top. Curr. Chem., 2006, 264,
201; B. Quiclet-Sire and S. Z. Zard, Chem.–Eur. J., 2006, 12, 6002.
5 W. Wierenga, Pharmacol. Ther., 1985, 30, 67; J. S. Taylor and
M. P. Cohrs, J. Am. Chem. Soc., 1987, 109, 2834;
D. A. Stringfellow, S. D. Weed, H. E. Renis and E. E. Eidson,
J. Med. Chem., 1980, 23, 239; G. Roma, M. Di Braccio, A. Carrieri,
G. Grossi, G. Leoncini, M. G. Signorello and A. Carotti, Bioorg.
Med. Chem. Lett., 2003, 11, 123; D. K. Whelligan, D. W. Thomson,
D. Taylor and S. Hoelder, J. Org. Chem., 2010, 75, 11 and references
therein; S. Wang, A. Folkes, I. Chuckowree, X. Cockcroft, S. Sohal,
W. Miller, J. Milton, S. P. Wren, N. Vicker, P. Depledge, J. Scott,
L. Smith, H. Jones, P. Mistry, R. Faint, D. Thompson and S. Cocks,
J. Med. Chem., 2004, 47, 1329; P. M. Traxler, P. Furet, H. Mett,
E. Buchdunger, T. Meyer and N. Lydon, J. Med. Chem., 1996, 39,
2285.
ꢀc
This journal is The Royal Society of Chemistry 2010
5786 | Chem. Commun., 2010, 46, 5784–5786