Titanium(IV) isopropoxide mediated synthesis of pyrimidin-4-ones

Article abstract of DOI:10.1021/ol100624p

A novel, one-step method for the synthesis of tri- and tetrasubstituted pyrimidin-4-ones is reported. This method involves a titanium(IV)-mediated cyclization involving two sequential condensations of primary and β-ketoamides. The reaction is operationally facile, readily scalable, and offers rapid entry into differentially substituted pyrimidin-4-one scaffolds. The high functional group compatibility allows for substantial diversification in the products generated from this transformation.

Full text of DOI:10.1021/ol100624p

ORGANIC
LETTERS
2010
Vol. 12, No. 10
2270-2273
Titanium(IV) Isopropoxide Mediated
Synthesis of Pyrimidin-4-ones
Joshi M. Ramanjulu,* Michael P. DeMartino, Yunfeng Lan, and Robert Marquis
Department of Medicinal Chemistry, Immuno-Inflamation CEDD, GlaxoSmithKline,
1250 South CollegeVille Road, CollegeVille, PennsylVania 19426
joshi.m.ramanjulu@gsk.com
Received March 16, 2010
ABSTRACT
A novel, one-step method for the synthesis of tri- and tetrasubstituted pyrimidin-4-ones is reported. This method involves a titanium(IV)-
mediated cyclization involving two sequential condensations of primary and ꢀ-ketoamides. The reaction is operationally facile, readily scalable,
and offers rapid entry into differentially substituted pyrimidin-4-one scaffolds. The high functional group compatibility allows for substantial
diversification in the products generated from this transformation.
The pyrimidin-4-one scaffold is well represented in natural
and pharmaceutically relevant chemical matter. Physio-
chemically, this moiety represents an excellent drug-like
template with which structurally diverse analogues can be
generated and assessed. As such, this scaffold has shown
immense value in the context of drug discovery, manifested
through its presence in marketed drugs and clinical candidates
(Figure 1). For instance, Isentress (3) and Risperidone (4)
are marketed for the treatment of Schizophrenia and HIV,
respectively.1 The additional examples depicted span a
variety of clinical indications and therapeutic targets.2-7
Additional synthetic procedures to access these structures
could allow for even greater exploitation of this chemical
template.
While a variety of methods have been developed for the
synthesis of bicyclic quinazolinones and tetrahydroquinazoli-
(1) Madaan, V. Drugs Today 2009, 45 (1), 55–62.
(2) Nugent, R. A.; Schlachter, S. T. PCT Int. Appl. 1995,
WO9511235A1
(3) Bagli, J.; Bogri, T.; Palameta, B.; Rakhit, S.; Peseckis, S.; McQuillan,
J.; Lee, D. K. H. J. Med. Chem. 1988, 31, 814
(4) Belyk, K. M.; Morrison, H. G.; Jones, P.; Summa, V. PCT Int. Appl.
2006, WO2006060712A2
(5) Madhavan, G. R.; Chakrabarti, R.; Vikramadithyan, R. K.; Mamidi,
R. N. V. S.; Balaraju, V.; Rajesh, B. M.; Misra, P.; Kumar, S. K. B.; Lohray,
.
.
Figure 1. Pharmaceutically active pyrimidin-4-ones.
.
B. B.; Lohray, V. B.; Rajagopalan, R. Bioorg. Med. Chem. 2002, 2671
(6) Hoornaert, C.; Wick, A. Eur. Pat. Appl. 1994, EP 607077 A1
(7) Reboud-Ravaux, M. J. Soc. Biol. 2001, 195, 143
.
nones, the known methods to prepare analogous differentially
and/or tetrasubstituted pyrimidin-4-ones are limited. The most
.
.
10.1021/ol100624p  2010 American Chemical Society
Published on Web 04/23/2010

commonly employed routes are depicted in Figure 2. Route
A involves the condensation of a Pinner-derived amidine with
a ꢀ-ketoester to generate the pyrimidin-4-one scaffold but
generally suffers from low yields and often poor “O” vs “N”
selectivity in the subsequent alkylation step.8,9 While Route
B avoids this difficult alkylation, it is not without problems
including low yields, scope limitation (restricted to malonyl
dichlorides, R2 ) H), and long reaction times.10 Routes C-E
share the underlying difficulty of enamine preparation which,
in turn, is used to generate linear precursors such as an
enamide ester (Route D) or an enediamide (Route E) prior
to ring formation. Although many methods have been
published for these transformations, they are generally low
yielding and highly substrate dependent.11,12 During the
course of this work, these routes were evaluated to assess
feasibility for rapid analogue generation and were found to
be inadequate in some or all of the following aspects:
functional group compatibility, incorporation of structural
diversity, scalability, number of steps, and yields. In this
paper, a highly efficient method for the preparation of
differentially and tetrasubstituted pyrimidin-4-ones from
ꢀ-ketoamide and primary amides mediated by titanium(IV)
isopropoxide is disclosed.
initially prepared using a two-step process coupling amines
with ꢀ-ketoacids (not pictured), but this tactic resulted in
low yields.13 A more concise approach was then exploited:
microwave irradiation of ꢀ-ketoesters with amines provided
ꢀ-ketoamides directly in reasonable yields (Scheme 1).14 This
method is compatible with a variety of functionalities and
allows for rapid access to the required ꢀ-ketoamides.15
Alternatively, preparation of ꢀ-ketoamides devoid of C-2
substitution was achieved in high yields upon treatment of
diketene with various amines.15
Scheme 1. Preparation of ꢀ-Ketoamides
Pyrimidin-4-one formation was realized through the tita-
nium(IV) isopropoxide mediated union and cyclization of
ꢀ-ketoamides and primary amides. A variety of alternate
Lewis acids were screened to identify a suitable activator
for this transformation, although none facilitated pyrimidin-
4-one formation.16 Ti(Oi-Pr)4 is a versatile reagent uniquely
suited for this one-pot transformation, acting as both a Lewis
acid, which activates otherwise poorly nucleophilic amides,
and a mild dehydrating reagent. The relatively mild nature
of this reagent suggested a high degree of functional group
tolerance could be achieved. As this trait is paramount for
any methodology to be employed in a pharmaceutical setting,
an initial screening was conducted to assesses functional
group tolerance (Figure 3). Unoptimized conditions (includ-
ing use of a large excess of Ti(Oi-Pr)4, 15 equiv, and primary
amide, 4 equiv) were used to clearly demonstrate this
tolerability as unprotected N-H heterocycles (10a-c),
phenols (10g),17 and basic moieties (10c, 10d, 10f) did not
hinder the reaction. Even extremely electron-rich aromatic
amides such as thiophene and oxazole proceeded smoothly
(10e and 10f). Importantly, the crystallinity of thiophene 10e
allowed for structural verification by X-ray crystallography
and, thus, unambiguous regiochemical confirmation (Figure
4). Confident that the method would tolerate structural
diversity, focus then shifted to optimization.
Figure 2. Common pyrimidin-4-one synthetic methods.
Initial attempts were made to reduce the requisite stoi-
chiometry of Ti(Oi-Pr)4 by partially replacing with high
boiling solvents such as 1,2-dichloroethane, toluene, and
xylenes; however, no obvious advantage was observed for
Access to a diverse collection of starting materials is
critical to render any method as practical. While many
primary amides are commercially available, the same cannot
be said for ꢀ-ketoamides. As such, ꢀ-ketoamides were
(13) Grayson, D. H.; Tuite, M. R. J. Chem. Soc., Perkin Trans. 1 1986,
2137.
(8) Pinner, A. Chem. Ber. 1889, 22, 1612
(9) Salimbeni, A.; Canevotti, R.; Paleari, F.; Poma, D.; Caliari, S.; Fici,
F.; Cirillo, R.; Renzetti, A.; Subissi, A.; Belvisi, L.; Bravi, G.; Scolastico,
.
(14) Addition of ethanol aids in absorption of microwave irradiation.
(15) See Supporting Information for general procedure.
(16) (a) Other Lewis acids tried [H+, Sc(Tf)3, ZnCl2, Zn(Tf)2, MgSO4,
Al(Oi-Pr)3, Zr(OEt)4] generated no detectable/trace amounts of product by
LCMS analysis. (b) Han, C.; Lee, J. P.; Lobkovsky, E.; Porco, J. A., Jr.
J. Am. Chem. Soc. 2005, 127, 10039.
C.; Giachetti, A. J. Med. Chem. 1995, 38, 4806
.
(10) Jezewski, A.; Jurczak, J.; Lidert, Z.; Tice, C. M. J. Heterocycl.
Chem. 2001, 38, 645.
(11) Jeong, J. U.; Chen, X.; Rahman, A.; Yamashita, D.; Luengo, J. I.
(17) It is known that the transesterification of titanium(IV) isopropoxide
is known to be kinetically slow: Paquette, L. Handbook of reagents for
organic synthesis - ActiVating agents and protecting groups; John Wiley
& Sons: New York, 1999; pp 389-394.
Org. Lett. 2004, 6, 1013
(12) Takahashi, T.; Hirokami, S.; Nagata, M. J. Chem. Soc., Perkin
Trans. 1 1998, 2653
.
.
Org. Lett., Vol. 12, No. 10, 2010
2271

bulky groups such as i-propyl (entry 2) and c-propyl (entry
3) are detrimental to the reaction resulting in lower yields.
The scope of the primary amide was also investigated with
a wide variety of aliphatic and aromatic substituents (R3).
Generally, the yields are moderate to high irrespective of
the steric and electronic nature of the substituent (entries
5-14). An exception was observed with pivalamide (entry
9) where the bulky tert-butyl group exceeds the upper limit
of steric tolerance for this transformation. Notably, the
presence of heterocycles did not diminish the yields in spite
of their potential to chelate with the Ti center (entries
11-14).
Figure 3. Synthesis of heterocycle-containing pyrimidin-4-ones.
All products isolated and characterized; yields not determined
inclusion of such solvents. Importantly, the observation that
the starting materials were readily dissolved in Ti(Oi-Pr)4
alone at elevated temperatures led to the minimization of
the stiochiometry to just 4 equiv. The amount of primary
amide was also lowered (1.2 equiv) without diminishing
yields, and the reaction time was reduced to 24 h. Because
the unoptimized conditions also included a workup procedure
that stifled crude isolation and purification (distillation
followed by messy aqueous workup and chromatography),
rigorous optimization was required. Release of the organic
material from the resultant titanium salts seemed to be a
likely cause of the isolation difficulties. It was discovered
that an acidic aqueous quench (2 N HCl) of the reaction
followed by addition of toluene (with prolonged phase
separation) ameliorated this problem. Further aqueous extrac-
tion with dichloromethane was required to complete the mass
balance recovery. This two-step process allowed for facile
chromatographic purification of the product pyrimidin-4-
ones. These modifications resulted in an operationally facile
and robust experimental procedure affording high isolated
yields.15
Table 1. Scope Exploration: Variation of R3 and R4
entry
R4
R3
13 (yield %)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
(CH2)2Ph
i-Pr
Cy
Ph
Ph
Ph
Ph
Me
Et
i-Pr
Cy
t-Bu
Bn
13a (64)
13b (7)
13c (42)
13d (38)a
13e (22)
13f (41)
13g (63)
13h (55)
13i (trace)b
13j (63)
13k (60)
13l (54)
13m (74)
13n (72)
Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph
4-pyridyl
2-pyrizyl
2-furyl
2-thienyl
a Reaction conversion was higher; purification problematic. b Product
not isolated; trace product seen in LCMS.
The scope of the methodology was further expanded
through the synthesis of tetrasubstituted pyrimidin-4-ones
with R1 substitution arising from the ꢀ-ketoamide (Table
2). Yields did not suffer as a result of this functionalization;
in fact, substitution at this position generally led to a
considerable increase in isolated yield (compare Table 1,
entry 1, and Table 2, entry 1), the rationale for which is
discussed below. Interestingly, aliphatic and aromatic primary
amides performed similarly (entries 1 and 2 vs 4 and 5), as
the electronic nature of the benzamides had little bearing on
the reaction outcome (entries 6-8).
Figure 4. ORTEP representation for compound 10e.
Finally, the scalability of this methodology was demon-
strated with the synthesis of highly functionalized pyrimi-
dinone 17 (Scheme 2). The synthesis commenced with
commercially available diketene, which, upon treatment with
phenethylamine, produced ꢀ-ketoamide 11a in quantitative
yield. 11a (127 g) was condensed with 3-fluoro-2-hydroxy-
benzamide18 using 15 equiv of titanium tetraisopropoxide
(unoptimized conditions) generating the pyrimidin-4-one
product. The phenol was directly benzylated to produce 16
Initial scope exploration focused on unsubstituted ꢀ-ke-
toamides while varying the R4 group. Table 1 details the
exemplars of ꢀ-ketoamides with differing substitution at the
amide nitrogen (entries 1-4). Application of the optimized
reaction conditions led to isolation of the desired pyrimidin-
4-ones in reasonable yields. Condensation with benzamide
(entry 4) demonstrates that this method tolerates inherently
poor nucleophiles such as aryl amides. However, sterically
2272
Org. Lett., Vol. 12, No. 10, 2010

Table 2. Preparation of Tetrasubstituted Pyrimidin-4-ones
Scheme 3. Proposed Reaction Mechanism
entry
R1
R4
R3
15 (yield %)
1
2
3
4
5
6
7
8
Me
Et
Ph
Me
Et
Me
Me
Me
(CH2)2Ph
(CH2)2Ph(o-F)
(CH2)2Ph
(CH2)2Ph
(CH2)2Ph(o-F)
(CH2)2Ph
Ph
Ph
Ph
Et
Et
4-OMe-Ph
4-F-Ph
2-furyl
15a (79)
15b (66)
15c (42)
15d (79)
15e (61)
15f (65)
15g (84)
15h (74)
(CH2)2Ph
(CH2)2Ph
While no rigorous studies were undertaken explicitly to
elucidate the underlying mechanism of this transformation,
the proposed reaction mechanism above is based on the
literature, substrate trends, and observed intermediates. For
example, the negligible conversion with pivalamide (Table
1, entry 9) demonstrates that the initial bimolecular conden-
sation is impeded by the sterics and suggests that this is the
rate-determining step, a notion also supported by the lack
of build-up of V during the reaction. Moreover, yields trend
higher for substituted ꢀ-ketoamides (R1 * H). This would
indicate intermediate V, stabilized by an intramolecular
hydrogen bond, would require a 180° bond rotation to
populate the reactive conformation in which the reacting
atoms are in close proximity. This rotation would be more
energetically favored upon the introduction of R1 substitution
due to 1,3-allylic strain. Although not conclusively proven,
these data are supportive of the proposed mechanism above.
(133 g) in 52% for two steps. Bromination of pyrimidin-4-
one 16 offered a handle to introduce a fourth substituent at
an alternate stage in the synthesis. Suzuki coupling with
5-methylthiophene boronic acid yielded 17, and subsequent
acidic debenzylation completed the synthesis of 18 (40 g)
in six steps and 32% overall.
Scheme 2. Scalability of Pyrimidin-4-one: Scale-up of 18
In conclusion, we have developed a novel preparation of
pyrimidin-4-ones. This method utilizes a double condensation
of two distinct amides facilitated by titanium tetraisopro-
poxide. The pyrimidin-4-one formation is accomplished
through the ability of Ti(Oi-Pr)4 to function dually as a Lewis
acid and dehydrating agent in the same transfomation. The
relatively mild nature of the method tolerates a wide array
of functionality and allows for rapid and efficient entry into
tri- and tetrasubsituted pyrimidin-4-ones in a single step from
simple reagents. The procedure was also demonstrated to
be amenable for multigram synthesis. The pharmaceutical
relevance of the pyrimidin-4-one scaffold coupled with the
structural diversity that is accessible via this method will
allow for widespread application of this technology.
A proposed reaction mechanism is detailed below (Scheme
3). Titanium tetraisopropoxide is likely to chelate the primary
amide II generating III, thus increasing its nucleophilicity.
ꢀ-Ketoamide I could form a bidentate chelated intermediate
IV augmenting its electrophilicity. The resultant enamine III
then condenses with the chelated ꢀ-ketoamide IV, with
subsequent dehydration (facilitated by the metal), generating
the enediamide V. The formation of intermediate V has been
confirmed by isolation and confirmation with LCMS and 1D/
2D NMR experiments. The presence of excess titanium
allows for another chelation event with enediamide V/VI
and sets the stage for a second, intramolecular, condensation.
The nucleophilic amide nitrogen attacks the pendant carbo-
nyl, and subsequent dehydration generates the product
pyrimidin-4-one, VII.
Acknowledgment. We thank Robert Trout, Arun Kaura,
Art Shu, and Thomas Ku for their assistance.
Supporting Information Available: Experimental pro-
cedures and characterization of all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(18) For synthesis of 3-fluoro-2-hydroxy benzamide, see: patent
WO2007062370A2.
OL100624P
Org. Lett., Vol. 12, No. 10, 2010
2273