Concise preparation of 8-trifluoromethyltetrahydro-6H-pyrido [1,2-a] pyrazine-6-one

Article abstract of DOI:10.1016/j.tetlet.2009.04.094

Methodology to rapidly and efficiently assemble 8-trifluoromethyltetrahydro-6H-pyrido [1,2-a] pyrazine-6-one (11) has been developed and its general applicability has been demonstrated.

Full text of DOI:10.1016/j.tetlet.2009.04.094

Tetrahedron Letters 50 (2009) 4050–4053
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Concise preparation of 8-trifluoromethyltetrahydro-6H-pyrido [1,2-a]
pyrazine-6-one q
*
Shankaran Kothandaraman , Deodialsingh Guiadeen, Gabor Butora, George Doss, Sander G. Mills,
Malcolm MacCoss, Lihu Yang
Department of Medicinal Chemistry, Merck and Co. Inc., RY 121-252, Rahway, NJ 07065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Methodology to rapidly and efficiently assemble 8-trifluoromethyltetrahydro-6H-pyrido [1,2-a] pyra-
zine-6-one (11) has been developed and its general applicability has been demonstrated.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 9 April 2009
Revised 22 April 2009
Accepted 23 April 2009
Available online 3 May 2009
We recently disclosed the preparation of MK-0812 (Fig. 1), a po-
tent chemokine (CCR2) receptor antagonist containing four chiral
centers and featuring a tetrahydro-3-trifluoromethyl-1, 6-naph-
thyridine nucleus.2 Ongoing issues related to pharmacokinetic
and ion-channel activities required further elaboration of this
structural class. We surmised that perhaps the inclusion of more
polar heterocycles such as that in analog 1 would be the most
promising area for further investigation.
Scheme 1 illustrates the original route used to access the target
heterocycle 113 that was required for the synthesis of 1. From the
commercially available 2, 6-dichloro-4-trifluoromethylpyridine
(2), a successive introduction of a tert-butoxy group (3) then a cy-
ano group, led to 4. Reduction of the cyano function by catalytic
reduction to give 5 was followed by sulfonylation of the primary
amine to afford 6. A two-step sequence of reactions involving the
alkylation of the sulfonamide 6 with 1,2-dibromoethane gave 7
which was followed by cleavage of the tert-butoxy group to afford
8. Base-induced facile ring closure of 8 yielded 9. The 2-nitrophen-
ylsulfonyl group in 9 was removed under standard conditions and
the final purification was aided by the temporary introduction of a
Boc-carbamate protecting group to initially afford 10 which was
then transformed to 11.
mono-N-Boc-ethylenediamine (14) with 2-nitrobenzenesulfonyl
chloride6 gave the sulfonamide 15. Its N-alkylation with propargyl
bromide afforded the intermediate 16, which was coupled to the
iodoester 177 under Sonogashira conditions8 to afford ynoate 18.
Cleavage of the Boc-protecting group produced the amine 12 as a
hydrochloride salt.
From the many possible metal catalysts that might trigger the
desired cyclization, we chose mercury, since this catalytic system
has been well studied in similar instances.9,10 Indeed, exposure of
12 to catalytic amounts of mercury(II) chloride at elevated temper-
ature induced a smooth cyclization, and the key intermediate 9
was obtained in high yield whose spectral data matched those of
similar analog prepared as described in Scheme 1. It was then sub-
sequently converted to 11 as described in Scheme 1.
When the key cyclization of 12 to 9 was performed in the ab-
sence of a catalyst, the reaction was found to be sluggish requiring
several hours to complete.11 To gain a deeper understanding of this
cascade reaction, a mechanistic study using high-resolution NMR
spectroscopy was undertaken. The key findings are summarized
in Scheme 3.
The hydrochloride salt of amine 12 was dissolved in dioxane-d8
and a stoichiometric amount of triethylamine-d15 was added, and
the reaction progress was monitored at ambient temperature by
1H NMR. The initial spectrum of 12 changed within minutes to
an intermediate that was assigned structure 19. Subsequently,
the intermediate 19 was found to convert to the final product 9
over a period of 2 days at room temperature.
Unfortunately, the starting material 2 was difficult to obtain in
large quantities4 and an alternative route to 11 was investigated.
We envisioned that a suitably functionalized ynoate such as 12,
could initially undergo an intramolecular conjugate addition of
the amino group to the triple bond5 to yield a piperazine derivative
13, and the heterocycle 9 would then form in a spontaneous ring
closure as shown in Figure 2.
The structure of the key intermediate 19 was assigned as fol-
lows: Isomeric structure 21 was eliminated because the NMR spec-
trum clearly indicated a presence of only one vinyl proton along
with four aliphatic methylene groups. Structures 19 and 20, both
consistent with this NMR pattern, were distinguished based on
3-bond 1H–13C correlation between the proton H-e and the ester
carbonyl C-j (HMBC). A strong NOE between H-a and H-d provided
further support for structure 19. In addition, the stereochemistry of
The requisite hydrochloride salt of ynoate 12 was prepared as
illustrated in Scheme 2. Reaction of commercially available
q
See Ref. 1.
* Corresponding author. Tel.: +732 594 1934.
E-mail address: kothandaraman_shankaran@merck.com (S. Kothandaraman).
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.04.094

S. Kothandaraman et al. / Tetrahedron Letters 50 (2009) 4050–4053
4051
O
O
OMe
OMe
H
N
H
N
CF3
CF3
N
N
N
O
O
N
O
1
MK-0812
Figure 1.
X
CF3
R1
CF3
CF3
CF3
N
R1N
R1N
iii
vi
vii
N
R2
N
HN
Br
N
Y
OtBu
O
O
2, X, =Cl
5, R1,R2=H
6, R1=2-NO2PhSO2, R2=H
7, R1=2-NO2PhSO2, R2=CH2CH2Br
8
9, R1=2-NO2PhSO2
10, R1=BOC
11, R1=H
i
ii
iv
v
viii
ix
3, X=Cl, Y=OtBu
4, X=CN, Y=OtBu
Scheme 1. Reagents and conditions: (i) KOtBu, THF, 0 °C; (ii) Zn(CN)2, (PPh3)4Pd, DMF, 90 °C; (iii) H2, RaNi, EtOH; (iv) 2-nitrophenylsulfonyl chloride, DIEA, DCM; (v) 1,2-
dibromoethane, K2CO3, DMF; (vi) TFA, rt, 10 min; (vii) K2CO3, THF, 60 °C; (viii) (a) PhSH, K2CO3, DMF; (b) BOC2O, K2CO3, DMF; (ix) HCl.
RN
CF3
CF3
CF3
RN
RN
NH
N
NH2 EtOOC
EtO
O
O
12, R=2-NO2PhSO2
13, R=2-NO2PhSO2
9, R=2-NO2PhSO2
Figure 2.
NHR
CF3
RN
R1N
RN
I
CF3
CF3
ii
iii
v
+
N
EtO
NHBOC
EtOOC
NHBOC
NHR2
O
O
9, R=NO2PhSO2
14, R =H
16, R=NO2PhSO2
17
18, R1=2-NO2PhSO2, R2=BOC
12, R1=2-NO2PhSO2, R2=H
i
iv
15, R=NO2PhSO2
Scheme 2. Reagents and conditions: (i) 2-nitrobenzenesulfonylchloride, CH2Cl2, TEA, 100%; (ii) propargyl bromide, K2CO3, DMF, 0 °C to rt, 96%; (iii) PdCl2 (PPh3)2, CuI, TEA,
THF, 70 °C, 71%; (iv) HCl, EtOAc, 0 °C to rt, 94%; and (v) HgCl2 (10 mol %), TEA, dioxane, 60–70 °C, 83%.
NO2
O2N
O2N
f
d
d
f
a
c
O
N
i
ii
a
g
g
O2S
CF3
S
CF3
O2S
N
N
O
CF3
j
b
N
e
N
b
e
COOEt
j
c
EtO
NH2
O
O
12
19
9
F3C
F3C
COOEt
COOEt
NO2
NO2
O
N
O
S
S
O
O
N
N
NH
20
21
Scheme 3. Reagents and conditions: (i) 12ÁHCl, dioxane-d8, TEA-d15, rt, (ii) dioxane-d8, rt, 48 h.

4052
S. Kothandaraman et al. / Tetrahedron Letters 50 (2009) 4050–4053
Table 1
dence we postulate that the cyclization precursor 12 was initially
Select 13C and 1H NMR chemical shifts of 19 and 9a
transformed to 19 which over time produced the final bicycle 9.
The key NMR characteristics of compounds 19 and 9 are summa-
rized in Table 1.
The cascade cyclization approach described for the conversion
of 12 to 9 has been extended to other systems. These results are
summarized in Table 2.12
The efficient transformation of 22 to 23 illustrates that the phe-
nyl group can be a replacement for a trifluoromethyl group during
this cyclization. A non-regioselective cyclization of 24 led to the
fused heterocycle 25 and its regioisomer 26 thus demonstrating
the potential to expand this methodology to linear tricyclic hetero-
cycles, low yields not withstanding. We also show that this cycliza-
tion reaction can be performed in an intermolecular manner that
involves an external nucleophile (in the present case with benzyl
amine) as exemplified by the transformation of 27 to afford 28 in
a respectable yield. Pyridones such as 28 with aminomethyl
Position
19
9
13C
47.3
40.7
49.5
129.4
32.2
159.4
168.2
124.6
1H
13C
45.4
43.6
40.5
1H
a
b
c
d
e
g
3.91
3.27
3.83
6.58
3.61
—
4.53
3.68
4.11
6.32
6.75
—
97.9
115.8
144.9
160.5
124.5
j
—
—
—
—
CF3
a
In dioxane-d8, 600 MHz.
the double bond was determined to be E based on a strong 19F–1H
NOE between the trifluoromethyl group and H-d. Based on this evi-
Table 2
Extension of cascade cyclization to other systemsa,b,c
Entry
Starting materialsd,e
Coupling product
Cyclization product
RN
Ph
Ph
RN
RN
I
Ph
N
1
+
NHBoc
EtOOC
O
EtOOC
NHBoc
23, 58%
22, 47%
RN
N
O
COOMe
RN
RN
COOMe
25, 7%
I
2
+
RN
NHBoc
+
NHBoc
N
24, 85%
O
26, 20%
CF3
BocHN
CF3
I
CF3
BocHN
+
3f
BocHN
N
O
EtOOC
EtOOC
Bn
27, 51%
28, 73%
CF3
BocHN
I
CF3
CF3
N
+
4
NHBoc
O
EtOOC
EtOOC
29, 82%
30, 82%
BocNH
H
N
R
RN
RN
H
H
I
N
5
+
EtOOC
31, 38%
O
EtOOC
NHBoc
Product Not formed
a
Isolated yields of pure products that have been characterized by high-resolution NMR/MS.
Experimental conditions were similar to the one discussed for the transformation of 12 to 9.
Yields not optimized.
Reaction performed in accordance with Ref. 7.
R = 2-nitrobenzenesulfonyl.
b
c
d
e
f
The cascade cyclization involved benzylamine as an external nucleophilic source.

S. Kothandaraman et al. / Tetrahedron Letters 50 (2009) 4050–4053
4053
5. (a) Goerdeler, J.; Laqua, A.; Lindner, C. Chem. Ber. 1974, 107, 3518; (b) Merah, B.;
Texier, F. Bull. Soc. Chim. France 1980, 552; (c) Carr, R. M.; Norman, R. O. C.;
Vernon, J. M. J. Chem. Soc., Perkin Trans 1980, 156.
6. Fukuyama, T.; Jow, C.-K.; Cheung, M. Tetrahedron Lett. 1995, 36, 6373.
7. (a) Qing, F. L.; Zhang, T. M. Tetrahedron Lett. 1997, 38, 6729; (b) Wang, P. A.;
Deng, M. Z.; Pan, R. Q.; Zhang, S. Y. J. Fluorine Chem. 2003, 124, 93.
appendages are very hard to prepare by other methodologies. The
cyclization of 29 to 30 offers the possibility to expand on the scope
of this reaction further to give fused pyridones. Finally, the sub-
strate 31 surprisingly, which lacks substitution at the double bond,
was rapidly consumed but none of the desired product was
formed, and the reasons for this failure are not clear.
In conclusion, a highly efficient cascade cyclization pathway
that greatly improves the access to an important class of fused
and bridged heterocycles has been developed. We believe that
the present methodology has a significant scope and enormous po-
tential for the rapid generation of pharmacologically important
heterocycles.
8. Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975, 16, 4467.
9. (a) Luo, F. T.; Wang, R. T. Tetrahedron Lett. 1992, 33, 6835; (b) Jacobi, P. A.;
Brielmann, H. L.; Hauck, S. I. J. Org. Chem. 1996, 61, 5013; (c) Fairfax, D.; Stein, M.;
Livinghouse, T.; Jensen, M. Organometallics 1997, 16, 1523; (d) Zulys, A.; Dochnahl,
M.; Hollmann, D.; Lohnwitz, K.; Hermann, J. S.; Roesky, P. W.; Blechert, S. Angew.
Chem., Int. Ed. 2005, 44, 7794. and several references cited therein.
10. General procedures: Preparation of (18): To a suspension of alkyne 16 (17.3 g,
45.14 mmol) in anhydrous THF (125 ml) was sequentially added vinyl
iodoester 177 (14.6 g, 49.65 mmol), copper (I) iodide (0.89 g, 4.51 mmol),
Pd[(Ph3P)3]4 (2.62 g, 2.26 mmol), and potassium carbonate (24.95 g,
180.56 mmol) under nitrogen. The resulting reaction mixture was stirred at
70 °C for 5 h, filtered, concentrated, and flash chromatographed (hexane + 15–
25% ethyl acetate) to give 18 (17.69 g, 71%) as an oil. 1H NMR (CDCl3, 400 MHz):
d 8.00–8.02 (m, 1H), 7.64–7.59 (m, 3H), 6.55 (s, 1H), 4.69 (b,1H), 4.52 (s, 2H),
4.18 (q, J = 6.1 Hz, 2H), 3.55–3.58 (t, 2H), 3.36–3.38 (t, 2H), 1.39 (s, 9H), 1.24–
1.30 (t, J = 6.1 Hz, 3H). LC–MS for C22H26 F3N3O8S [M+H]+ calcd 550.14, found
450.05, (m-100, loss of t-butoxy group).
Supplementary data
Supplementary data (procedure, NMR data for the mechanism
and for the intermediates 9, 12, 14–19, and 22–31) have been
provided.
Preparation of (12): To a stirred solution of the N-Boc-protected amine 18
(17.69 g, 32.19 mmol) in EtOAc (100 ml) at 0 °C was added a saturated solution
of HCl (g)/ethyl acetate (200 ml). The reaction mixture was stirred at 0 °C for an
additional 2 h after which it was concentrated to give the hydrochloride salt of
12 (14.7 g, 94%) as a tan crystalline solid. LC–MS for C17H18 F3N3O6S [M+H]+
calcd 450.09, found 450.05.
Preparation of (9): A stirred solution of alkyne 12 (14.6 g, 30.04 mmol) in
anhydrous 1, 4 dioxane (100 ml) was treated with mercury(II) chloride (0.81 g,
3.00 mmol) and triethylamine (8. 22 ml, 60.08 mmol) under nitrogen. The
resulting suspension was then gently heated at 65 °C for 30 min, concentrated,
and flash chromatographed (methyl-t-butylether + 15–35% ethyl acetate) to
References and notes
1. Shankaran, K. Abstracts of papers, ORG-061, 236th National ACS Meeting at
Philadelphia, PA, August 17–21, 2008.
2. Yang, L. H.; Jiao, R. X.; Moyes, C.; Morriello, G.; Butora, G.; Shankaran, K.;
Pasternack, A.; Gobel, S.; Zhou, C. Y.; MacCoss, M.; Cumiskey, A. M.; Petersen, L.;
Forrest, M.; Ayala, J. M.; Jin, H.; DeMartino, J.; Mills, S.G. Abstracts of papers,
MEDI-018, 233rd National ACS Meeting, Chicago, IL, 2007.
3. Butora, G.; Guiadeen, D.; Shankaran, K.; MacCoss, M.; Mills, S.; Yang, L. H. PCT
International Application 2005072361.
4. Butora, G, unpublished results. When this investigation began 2 was available
from Lancaster Synthesis (currently Alfa-Aesar), but was discontinued later
during the scale-up studies toward the preparation of 1. Currently, it is
available again from Alfa-Aesar, Matrix, and others.
afford
9 (10.09 g, 83%) after trituration with diethyl ether. LC–MS for
C15H12F3N3O5S [M+H]+ calcd 404.04, found 404.05.
11. Longer reaction time inevitably also led to lower yields in addition to
intractable mixtures of products.
12. Cyclization procedure to afford the products (23, 25, 26, 28, and 30) was very
similar to the one discussed for 19 to 9 in Ref. 10.