General and rapid pyrimidine condensation by addressing the rate limiting aromatization

Article abstract of DOI:10.1021/ol500886a

The rate limiting aromatization within the condensation approach toward pyrimidines utilizing amidines and activated olefins was addressed to provide for a general and rapid process. A strong solvent effect was elucidated to affect the rate for the initia

Full text of DOI:10.1021/ol500886a

Letter
pubs.acs.org/OrgLett
General and Rapid Pyrimidine Condensation by Addressing the Rate
Limiting Aromatization
Daniel R. Fandrick,* Delia Reinhardt, Jean-Nicolas Desrosiers, Sanjit Sanyal, Keith R. Fandrick,
Shengli Ma, Nelu Grinberg, Heewon Lee, Jinhua J. Song, and Chris H. Senanayake
Chemical Development, Boehringer-Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road/P.O. Box 368, Ridgefield Connecticut
06877-0368, United States
S
* Supporting Information
ABSTRACT: The rate limiting aromatization within the condensa-
tion approach toward pyrimidines utilizing amidines and activated
olefins was addressed to provide for a general and rapid process. A
strong solvent effect was elucidated to affect the rate for the initial
alkoxide elimination from the intermediate Michael adduct wherein
polar aprotic solvents demonstrate an addition controlled aromatiza-
tion. Spectroscopic studies support a solvent dependent equilibrium
between the amidine and alkoxide base wherein the rate for
aromatization is optimal when the equilibrium toward the amidine
anion was strongly favored.
he biological importance of the pyrimidine core can be
exemplified by the essential vitamin B1 and its presence in
NMR analysis, Katritzky and Yousaf demonstrated both the
intermediate formation and subsequent alkoxide elimination
were slow operations for the Pinner pyrimidine synthesis
between an amidine and β-ketoketones.8 Application of
microwave technology has been applied to afford reasonable
reaction times,9 but no general and time efficient process
amenable for a conventional batch process has been reported.
General rapid approaches toward pyrimidines which provide
high throughput in regard to time and volume10 would provide
value toward library synthesis, discovery optimization, and
scaleup. Herein, we report an effective model system to identify
conditions to address the rate limiting aromatization and
development of a rapid pyrimidine synthesis from amidines and
activated olefins.
T
natural products.1 Furthermore, pharmaceutical applications of
the pyrimidine have generated both the blockbuster drugs
imatinib and rosuvastatin as well as led to the identification of
many potent and promising drug candidates.2 This interest in
pyrimidines provided the impetus for numerous and creative
methodologies3 to construct the pyrimidine including the use
of vinamidinium salts.4 However, the predominately practiced
strategy toward the heterocycle relied on the classical
condensation approach (Scheme 1) which requires the
Scheme 1. Representative Condensation Conditions toward
Pyrimidines
The model system to examine the Michael addition approach
toward pyrimidines employed the reaction between benzami-
dine and our recently reported fluorinated butenone 8.11
Reversed phase HPLC analysis of the intermediate generated
from the reaction between the components under neutral
conditions showed multiple broad peaks with a consistent mass
indicative of the cyclized adduct 9. The main limitation with
HPLC analysis was the inability to qualitatively determine the
rate for both the intermediate formation as well as the
subsequent aromatization. Accordingly, a react-IR analysis on
the reaction was conducted. Component analysis revealed a
rapid and complete Michael addition once the amidine salt was
neutralized by sodium ethoxide (Figure 1). Furthermore, this
intermediate was relatively stable, and at 35 °C, the
intermediate showed minimal decomposition over several
hours.12 The stability and rapid formation was also demon-
elimination of alkoxide substituents for aromatization.5
Although numerous variations are established to affect the
condensation,6,7 basic conditions employing a protic solvent
with a reaction time of several hours to days are the most
common. Even among the thousands of examples, few
mechanistic or optimization studies to address the rate limiting
operations are published. Through detailed and exemplary 13C
Received: March 25, 2014
Published: May 12, 2014
© 2014 American Chemical Society
2834
dx.doi.org/10.1021/ol500886a | Org. Lett. 2014, 16, 2834−2837

Organic Letters
Letter
tion subjected the Michael adduct 9 generated from the
reaction between the butenone 8 and free base amidine 1b in
DMSO to catalytic sodium ethoxide (Figure 3). Both the
Figure 1. React-IR analysis for the Michael addition conducted with
benzamidine 1 (1.02 equiv), butenone 8 (1.0 equiv), and NaOEt (21
wt % in EtOH, 1.0 equiv) in EtOH. Relative component analysis and
identities confirmed by HPLC and LC-MS.
Figure 3. React-IR analysis on the pyrimidine condensation. Reaction
conducted with benzamidine 1b (free base, 1.02 equiv), butanone 8
(1.0 equiv), and NaOEt (21 wt % in EtOH, 0.2 equiv) in DMSO.
Relative component analysis and identities confirmed by HPLC and
LC-MS.
strated in DMSO wherein NMR and mass spectroscopic
analysis supported the cyclized intermediate 912 consistent with
the Katritzky and Yousaf report.8
After establishment of a rapid intermediate formation, the
rate limiting aromatization was addressed. The most influential
parameter examined12 was the solvent choice under basic
conditions. Alcoholic solvents are the most commonly
employed solvents for the pyrimidine synthesis. However, the
model system under the conventional conditions5−7 demon-
strated a sluggish aromatization and provided only a moderate
yield for the pyrimidine (Figure 2). Alternatively, polar aprotic
Michael addition and subsequent aromatization exhibited
addition control behavior.12 Accordingly, the aromatization
promoted by the necessary alkoxide eliminations demonstrated
a strong rate dependence on the solvent choice under basic
conditions.
The developed pyrimidine synthesis from amidines and
activated olefins proved general (Table 1). The challenge with a
general process was establishing conditions to tolerate the
breath and quality of the available activated olefins. A process
wherein the Michael addition was initiated prior to the base
promoted aromatization proved more robust than conducting
both operations concurrently. Accordingly, the olefin was
subjected to the free base amidine, either by utilizing the free
base or by neutralizing the hydrochloride salt with sodium
ethoxide, prior to subjecting the reaction to the basic
conditions. To compensate for the activated olefin quality,
which was observed to contain impurities that would quench
the base, the sodium ethoxide charge to promote the
aromatization was adjusted based on the individual reaction
from 0.1 to 1.3 equiv. For example, butenones 17 (50 wt %), 19
(80 wt %), and 27 (35 wt %) were able to be directly utilized in
the process without purification after appropriate adjustment to
the base stoichiometry. Flavone 35 was also competent for the
rapid pyrimidine synthesis. Due to the observed slower Michael
addition and resulting phenol functional group related to the
system, the process was modified by use of an excess of sodium
tert-butoxide to afford a moderate yield within 15 min (eq 1).
Figure 2. Solvent effect on the pyrimidine aromatization.12 Reactions
conducted with benzamidine 1 (1.02 equiv), butenone 8 (1.0 equiv),
and NaOEt (21 wt % in EtOH, 1.0 equiv) in the indicated solvent.
Pyrimidine 10 yield determined by HPLC (220 nm) assay utilizing m-
xylene as an internal standard.
solvents such as DMSO, DMF, or DMAc demonstrated a rapid
and essentially addition controlled reaction. Other commonly
employed solvents such as THF and toluene also demonstrate a
rapid aromatization but provided only poor to moderate yields
due to the reaction stalling at partial conversions. The
aromatization rate for the pyrimidine formation was also
qualitatively assessed through react-IR analysis. The examina-
2835
dx.doi.org/10.1021/ol500886a | Org. Lett. 2014, 16, 2834−2837

Organic Letters
Letter
a
A reasonable rationalization for the rapid aromatization
observed in the condensation approach toward pyrimidines
with an alkoxide base in DMSO relates to the solvent effect
driving the acid−base equilibrium to favor the amidine anion
and associated increase in the rate for the initial alkoxide
elimination (Scheme 2). Based on Katritzky and Yousaf’s
Table 1. Pyrimidine Scope
Scheme 2. Proposed Mechanism for the Pyrimidine
Condensation
analysis8 and utilization of a system wherein the Michael
addition was rapid, the reasonable rate limiting step for the
condensation process was the aromatization. Under basic
conditions, the elimination would occur through the amidine
anion generated from the base wherein the rate is dependent
upon the concentration of the anion. Solvent choice is well-
known to influence acid−base equilibria as well as dramatically
affect reaction rates.13 The equilibrium between the amidine
anion and base may also show a similar solvent dependence.
Direct examination of the anionic intermediate 9 was not
feasible due to the rapid resulting reaction. However, the free
base benzamidine is a reasonable model for the intermediate 9
in acid−base equilibria. The established pKa values of methanol,
tert-butanol, and benzamidine in DMSO are 29.0, 32.2,14 and
26.7,15 respectively, to support a strong equilibrium toward the
amidine anion with alkoxide bases in polar aprotic solvents. The
corresponding equilibrium in methanol was qualitatively
assessed through spectroscopy. The amidine anion generated
in DMSO with sodium tert-butoxide was apparent by the
respective aromatic proton shielding upon deprotonation
1
through H NMR as well as the associated IR absorbance
change of the amidine function group (Figures 4 and 5). When
the amidine was subjected to sodium tert-butoxide in methanol,
which generates sodium methoxide, the spectra supported
a
Reactions were conducted by charging NaOEt (21 wt % in EtOH, 0.1
to 1.3 equiv based on individual reaction12) to a 15 min aged mixture
of NaOEt (21 wt % in EtOH, n equiv), electrophile, and amidine n-
HCl in DMSO at 50 °C. Completion of each reaction was determined
b
c
by HPLC. Isolated yields. 2.94 equiv of butenone 8 utilized.
d
Reaction conducted by aging a mixture of the amidine, electrophile,
and NaOEt (21 wt % in EtOH, 1.1 equiv) for 15 min at 50 °C.
The condensation with a water miscible solvent was also
amenable for a direct crystallization process. Simple addition of
water to the parent reaction between benzamidine 1b and
butenone 8 catalyzed by sodium ethoxide enabled the
precipitation of pure pyrimidine 10 with good recovery (eq 2).
Figure 4. 1H NMR at 500 MHz analysis for the treatment of
benzamidine 1b with (red) and without (blue) sodium tert-butoxide
(1.1 equiv) in the indicated solvent at ambient temperature.
2836
dx.doi.org/10.1021/ol500886a | Org. Lett. 2014, 16, 2834−2837

Organic Letters
Letter
(4) Frutos, R. P.; Wei, X.; Patel, N. D.; Tampone, T. G.; Mulder, J.
A.; Busacca, C. A.; Senanayake, C. H. J. Org. Chem. 2013, 78, 5800−
5803.
(5) (a) Gogsig, T. M.; Nielsen, D. U.; Lindhardt, A. T.; Skrydstrup,
T. Org. Lett. 2012, 14, 2536−2539. (b) Kotljarov, A.; Irgashev, R. A.;
Iaroshenko, V. O.; Sevenard, D. V.; Sosnovskikh, V. Y. Synthesis 2009,
3233−3242. (c) Iaroshenko, V. O.; Specowius, V.; Vlach, K.; Vilches-
Herrera, M.; Ostrovskyi, D.; Mkrtchyan, S.; Villinger, A.; Langer, P.
Tetrahedron 2011, 67, 5663−5677.
(6) For selected examples of condensations towards pyrimidines with
β-alkoxybutenones, see: (a) Bellur, E.; Langer, P. Tetrahedron 2006,
62, 5426−5434. (b) Flores, A. F. C.; Pizzuti, L.; Brondani, S.; Rossato,
M.; Zanatta, N.; Martins, M. A. P. J. Braz. Chem. Soc. 2007, 18, 1316−
1321. (c) Muller, K.; Schubert, A.; Jozak, T.; Ahrens-Botzong, A.;
Schuenemann, V.; Thiel, W. R. ChemCatChem 2011, 3, 887−892.
(7) For examples of rapid condensations which proceed by alternate
elimination mechanisms than those described in Scheme 2, see:
(a) Cernuchova, P.; Vo-Thanh, G.; Milata, V.; Loupy, A.; Jantova, S.;
Theiszova, M. Tetrahedron 2005, 61, 5379−5387. (b) Danagulyan, G.
G.; Mkrtchyan, A. D.; Sahakyan, L. G. Chemistry of Heterocyclic
Compounds; Springer: New York, NY, 2005; Vol. 41, pp 262−264.
(c) Zanatta, N.; Fantinel, L.; Lourega, R. V.; Bonacorso, H. G.;
Martins, M. A. P. Synthesis 2008, 358−362.
Figure 5. IR analysis for the treatment of benzamidine 1b with (red)
and without (black) sodium tert-butoxide (1.1 equiv) in the indicated
solvent at ambient temperature.
predominately the free base amidine with little observable
amidine anion. Accordingly, the amidine−alkoxide acid−base
equilibrium was solvent dependent wherein the amidine anion
necessary for the initial alkoxide elimination was strongly
favored in DMSO. Other solvent effects13,16 may also
contribute to the reaction rate differences, but the correlation
between the anion equilibria and reaction behavior reasonably
rationalizes the solvent effect on the rate of the pyrimidine
condensation.
In conclusion, the rate limiting aromatization within the
condensation approach toward pyrimdines utilizing amidines
and activated olefins was addressed to provide for a general and
rapid process. Employing polar aprotic solvents with an
alkoxide base provided a reaction wherein optimal yields were
obtained in DMSO. The process demonstrated a broad
tolerance toward the substrates and allowed the generation of
a diverse pyrimidine series. Mechanistic analysis on the acid−
base equilibrium between an alkoxide base and benzamidine
demonstrated a pronounced solvent effect wherein DMSO
strongly favors the amidine anion necessary to promote the
initial alkoxide elimination. Extension of factors to bias the
acid−base equilibria necessary for base promoted condensa-
tions should also facilitate other related heterocycle syntheses.
(8) (a) Katritzky, A. R.; Yousaf, T. I. Can. J. Chem. 1986, 64, 2087−
2093. (b) Kereselidze, J.; Kvaraia, M.; Pachulia, Z.; Zarkua, T. Bull.
Georg. Natl. Acad. Sci. 2011, 5, 69−72.
(9) For selected examples, see: (a) Luo, G.; Chen, L.; Poindexter, G.
S. Tetrahedron Lett. 2002, 43, 5739−5742. (b) El Latiff, F. M. A.;
Barsy, M. A.; Aref, A. M.; Sadek, K. U. Green Chem. 2002, 4, 196−198.
(c) Bagley, M. C.; Hughes, D. D.; Lubinu, M. C.; Merritt, E. A.;
Taylor, P. H.; Tomkinson, N. C. O. QSAR Comb. Sci. 2004, 23, 859−
867.
(10) Anderson, N. G. Practical Process Research & Development;
Academic Press: San Diego, 2000.
(11) Desrosiers, J.-N.; Kelly, C. B.; Fandrick, D. R.; Nummy, L.;
Campbell, S. J.; Wei, X.; Sarvestani, M.; Lee, H.; Sienkiewicz, A.;
Sanyal, S.; Zeng, X.; Grinberg, N.; Ma, S.; Song, J. J.; Senanayake, C.
H. Org. Lett. 2014, 16, 1724−1727.
(12) See Supporting Information for additional React-IR analyses,
intermediate characterization, optimization studies, and procedures.
(13) For selected references and examples, see: (a) Miller, J.; Parker,
A. J. J. Am. Chem. Soc. 1961, 83, 117−123. (b) Parker, A. J. Q. Rec.
Chem. Soc. 1962, 16, 163−187. (c) Wilson, S. R. Org. React. 1993, 43,
93−250. (d) Isanbor, C.; Babatunde, A. I. J. Phys. Org. Chem. 2009, 22,
1078−1085. (e) Ji, P.; Atherton, J. H.; Page, M. I. J. Org. Chem. 2011,
76, 3286−3295.
(14) (a) http://www.chem.wisc.edu/areas/reich/pkatable/index.
htm, accessed Mar. 24, 2014. (b) Olmstead, W. N.; Margolin, Z.;
Bordwell, F. G. J. Org. Chem. 1980, 45, 3295−3299.
(15) Bordwell, F. G.; Ji, G.-Z. J. Am. Chem. Soc. 1991, 113, 8398−
8401.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental details including additional react-IR experiments,
1
copies of HPLC, H and 13C NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: daniel.fandrick@boehringer-ingelheim.com.
Notes
(16) For selected references and citations within, see: (a) Humeres,
E.; Nunes, R. J.; Machado, V. G.; Gasques, M. D. G.; Machado, C. J.
Org. Chem. 2001, 66, 1163−1170. (b) McManus, S. P.; Somani, S.;
Harris, J. M.; McGill, R. A. J. Org. Chem. 2004, 69, 8865−8873.
The authors declare no competing financial interest.
REFERENCES
■
(1) (a) Lagoja, I. M. Chemistry & Biodiversity 2005, 2, 1−50.
(b) Zhao, L.; Ma, X.-D.; Chen, F.-E. Org. Process Res. Dev. 2012, 16,
57−60.
(2) Mah, R. Pyrimidine-Based Kinase Inhibitors in Cancer Chemo-
therapy. In Bioactive Heterocyclic Compound Classes; Dinges, J.,
Lamberth, C., Eds.; Wiley-VCH: Weinheim, 2012; pp 257−273.
(3) For selected recent examples, see: (a) Sasada, T.; Kobayashi, F.;
Sakai, N.; Konakahara, T. Org. Lett. 2009, 11, 2161−2164. (b) Gayon,
E.; Szymczyk, M.; Gerard, H.; Vrancken, E.; Campagne, J.-M. J. Org.
Chem. 2012, 77, 9205−9220. (c) Sedenkova, K. N.; Averina, E. B.;
Grishin, Y. K.; Kutateladze, A. G.; Rybakov, V. B.; Kuznetsova, T. S.;
Zefirov, N. S. J. Org. Chem. 2012, 77, 9893−9899. (d) Anderson, E. D.;
Boger, D. L. J. Am. Chem. Soc. 2011, 133, 12285−12292.
2837
dx.doi.org/10.1021/ol500886a | Org. Lett. 2014, 16, 2834−2837