Aromatization and ring cyclization: A better understanding on the ring cyclization mechanism of 3-amino-6-hydrazino-1,2,4-triazin-5(2H)-one reacted with acetic acid in N,N-dimethylformamide

Article abstract of DOI:10.1002/jhet.5570420516

In this paper we report that the title compound (3) reacts with excess N,N-dimethylformamide (DMF) containing two equivalents of acetic acid to afford 6-amino-1,2,4-triazolo[3,4-f][1,2,4]triazin-8(7H)-one (1). When 3-amino-2-benzyl-6-hydrazino-1,2,4-triazin-5(2H)-one (6), the N-2 benzylated derivative of 3, is treated under the same conditions, ring Cyclization does not occur; instead, 3-amino-2-benzyl-6-(2-formyl-hydrazino)-1,2,4-triazin-5(2H)-one (7) is formed. Single-crystal X-ray analysis of a 3-ethyl derivative of compound 1 reveals the predominant tautomeric structure to be the 7H-tautomer (7H-1). From these results, we propose a reasonable cyclization mechanism that incorporates two important points: (1) the tautomerism of the N-2 hydrogen with the C-5 oxo group aromatizes the 1,2,4-triazine ring, and (2) the DMF is protonated by acetic acid on the nitrogen atom, then deamination occurs where DMF is attacked by the 6-hydrazino group of 3 or 6.

Full text of DOI:10.1002/jhet.5570420516

Jul-Aug 2005
Aromatization and Ring Cyclization: A better Understanding on the
Ring Cyclization Mechanism of 3-Amino-6-hydrazino-1,2,4-triazin-
5(2H)-one Reacted with Acetic Acid in N,N-Dimethylformamide
851
a
a
b
Long-Chih Hwang *, Rang-Rong Wu and Chun-Hsien Tu
a
Faculty of Medicinal and Applied Chemistry, Kaohsiung Medical University, Kaohsiung City 80708, Taiwan, ROC
e-mail: lnchhw@kmu.edu.tw
Department of Psychiatry, Veterans Hospital of Longchuan, Pingtung 912, Taiwan, ROC
b
Received November 5, 2004
In this paper we report that the title compound (3) reacts with excess N,N-dimethylformamide (DMF) con-
taining two equivalents of acetic acid to afford 6-amino-1,2,4-triazolo[3,4-f][1,2,4]triazin-8(7H)-one (1).
When 3-amino-2-benzyl-6-hydrazino-1,2,4-triazin-5(2H)-one (6), the N-2 benzylated derivative of 3, is
treated under the same conditions, ring cyclization does not occur; instead, 3-amino-2-benzyl-6-(2-formyl-
hydrazino)-1,2,4-triazin-5(2H)-one (7) is formed. Single-crystal X-ray analysis of a 3-ethyl derivative of
compound 1 reveals the predominant tautomeric structure to be the 7H-tautomer (7H-1). From these results,
we propose a reasonable cyclization mechanism that incorporates two important points: (1) the tautomerism
of the N-2 hydrogen with the C-5 oxo group aromatizes the 1,2,4-triazine ring, and (2) the DMF is proto-
nated by acetic acid on the nitrogen atom, then deamination occurs where DMF is attacked by the
6-hydrazino group of 3 or 6.
J. Heterocyclic Chem., 42, 851 (2005).
Introduction.
lents of acetic acid in excess N,N-dimethylformamide
(DMF) under reflux, and the ring closure at the N-1 nitro-
gen of the 1,2,4-triazine ring occurred to afford 1.
Lovelette suggested the heterobicyclic structure of 1 as the
5H-tautomer (5H-1) and described the probable mecha-
nism shown in Figure 1 [1]. As a part of our ongoing pro-
gram, we have recently described the prototropic tau-
tomerism of 6-amino-3-ethyl-1,2,4-triazolo[3,4-f][1,2,4]-
triazin-8(7H)-one (2), a 3-ethyl derivative of compound 1
[2]. In that study, the crystal structure of 2 was clarified by
single-crystal X-ray analysis, revealing the predominant
tautomeric structure as the 7H-tautomer, not the 5H-tau-
tomer as speculated by Lovelette. This result inspired us to
understand more about the actual mechanism of this ring
cyclization.
Many aza and deaza analogues of purine and their nucle-
osides have attracted considerable interest due to their bio-
logical activities. Compound 1, 6-amino-1,2,4-tria-
zolo[3,4-f][1,2,4]triazin-8(5H)-one (4,8-diaza-9-deaza-
guanine), the isosteric isomer of guanine, has been previ-
ously synthesized by Lovelette [1]. He treated 3-amino-6-
hydrazino-1,2,4-triazin-5(2H)-one (3) with two equiva-
Results and Discussion.
Initially, we worked backward through the mechanism
from the 7H-tautomer 2 to the starting material compound
3, which exists predominantly as the 2H-tautomer accord-
ing to the X-ray crystallographic structure previously

852
L.-C. Hwang, R.-R. Wu and C.-H. Tu
Vol. 42
reported by us [3]. Pith et al. had presented similar results
based on ultraviolet spectra and ionization constants [4].
From our mechanistic analysis we noted that the ring
cyclization occurred by way of an aromatic 1,2,4-triazine
ring, that is, the 2H-5-one aromatized to the 5-ol.
Therefore, the cyclization was speculated to be dependent
on the 1,2,4-triazine ring aromatization. To prove this
hypothesis, our starting point was to replace the N-2-H of
3 with a benzyl group to prevent prototropic tautomerism
to the aromatic intermediate. Therefore, we synthesized
the N-2-benzyl derivative 6, and investigated the reactions
of 6 with acetic acid in DMF.
Scheme 1 shows our synthesis of 6 from 3-amino-6-
bromo-1,2,4-triazin-5(2H)-one (4). Treatment of 4 with
hexamethyldisilazane gave the silylated intermediate,
which was then reacted with benzyl bromide to give 3-
amino-2-benzyl-6-bromo-1,2,4-triazin-5(2H)-one (5).
1
The H nmr spectrum of 5 showed two singlets at δ 5.17
H
and 7.58 ppm corresponding to the methylene protons of
the benzyl group and the 3-amino group, respectively. The
remaining multiplets at δ 7.24 to 7.45 ppm were attrib-
H
uted to the aromatic protons of the benzyl group. The ben-
zylation site of 5 was established as N-2 (not N-4) based
1
on uv spectra [5-9] and 2D nmr [5,6]. The long-range H-
13
C heteronuclear correlation (HETCOR) nmr revealed
the connectivity of methylene protons (δ 5.17 ppm) on
H
the benzyl group to C-3 (δ 155.09 ppm) in the 1,2,4-tri-
C
azine ring. Hydrazination of 5 by nucleophilic substitution
with hydrazine monohydrate gave compound 6.
1
Characterization of 6 was done by the H nmr spectrum
which showed signals at δ 3.90 (s, 2H, NHNH ) and
nostic for the carbonyl absorption of fused triazines.
H
2
7.34 (s, 1H, NHNH ) for hydrazino group, revealing that
However, our ir spectrum of 7 showed an intense absorp-
2
-1
the bromine atom was displaced by hydrazine.
tion at 1666 cm , which suggested that ring closure didn't
1
occur. The H nmr spectrum of 7 showed two singlets at
δ 4.98 and 6.97 ppm corresponding to the methylene pro-
H
tons of the benzyl group and the 3-amino group, respec-
tively. The multiplets at δ 7.22 to 7.37 ppm corresponded
H
to aromatic protons of the benzyl group, and one singlet at
δ 7.98 ppm to the N-1 hydrogen of the 6-formylhydrazino
H
group. The doublets at δ 8.26 and 9.71 ppm (both with J
H
= 1.2 Hz) corresponded to N-2 hydrogen and the proton of
the formyl group, respectively. Additional supporting evi-
dence for 7 was found in the high-resolution mass spec-
trum showing the molecular ion at m/z 260.1021, confirm-
ing the precise molecular formula as C H N O (calcd.
11 12
6 2
260.1022). Taken together, 7 was assigned as 3-amino-2-
benzyl-6-(2-formylhydrazino)-1,2,4-triazin-5(2H)-one;
the ring closed compound 8, the acetylhydrazino com-
pound 9, and the 2-(N,N-dimethylaminomethylene)-
hydrazino compound 10 were not observed (Scheme 2).
A reasonable mechanism explaining the deamination of
DMF by 6 to form 7 is shown in Figure 2. We proposed
that the DMF was protonated by acetic acid on the nitrogen
atom, not the oxygen atom, in the first step. The carbonyl
When 6 was reacted with excess DMF containing two
equivalents of acetic acid, 7 was isolated (Scheme 2).
According to the reports of Lovelette [1] and Becker et al.
-1
[10], the intense infrared absorption at 1725 cm is diag-

Jul-Aug 2005
A better Understanding on the Ring Cyclization Mechanism
853
carbon of the protonated DMF was then attacked by the 6-
hydrazino group of 6. Finally the dimethylamine was lost
and 7 formed.
lar ion at m/z 152.0446 corresponding to the molecular for-
mula C H N O (calcd.152.0447). From these spectral and
4
4 6
analytical data, the product structure was assigned as 6-
amino-1,2,4-triazolo[3,4-f][1,2,4]triazin-8-one (1). The
major prototropic tautomer of 1 was conjectured through
the X-ray structure of 2 as the 7H-tautomer [2].
Accordingly, the nomenclature of compound 1 is 6-amino-
1,2,4-triazolo[3,4-f][1,2,4]triazin-8(7H)-one.
We then repeated the reaction with compound 3 (Scheme
1
3) to afford 1, which was characterized by the H nmr spec-
trum [δ 6.36 (s, 2H, NH ), 9.00 (s, 1H, 3-H), 11.55 (br s,
H
2
13
1H, NH)]. The proton-decoupled C nmr and DEPT of 1
indicated three resonances (with δ of both C-3 and C-8a at
C
139.70 ppm), including one methylidene (C-3) and three
Figure 3 shows the proposed cyclization mechanism for
3 refluxed with acetic acid in DMF to form the 7H-tau-
tomer 1. Lovelette [1] had previously described the proba-
ble ring cyclization mechanism for 5H-tautomer 1 (Figure
1). He suggested that DMF was protonated by acetic acid
on the oxygen atom in the first step, partly based upon the
report by Fraenkel et al. [11], who suggested that DMF was
monoprotonated on the oxygen atom in strong acids (e.g.
100% sulfuric acid). However, the chemical reaction from 3
to 1 (Scheme 3) was done in a weaker acid (acetic acid in
DMF). Moreover, when 6 was reacted with acetic acid in
DMF, no prototropic tautomerism from the 2H-5-one to the
5-ol was possible and, therefore, no ring cyclization
occurred (Scheme 2). This observation suggested that DMF
was protonated on the nitrogen atom. These findings corre-
sponded well with our previous hypothesis, that is, the ring
cyclization was dependent on the resonance of the 2H-5-
quarternary carbons (δ of C-8a, C-6 and C-8 at 139.70,
C
151.07 and 152.10 ppm, respectively). Additionally, the
occurrence of ring closure was confirmed by the intense
-1
infrared absorption at 1719 cm corresponding to the car-
bonyl of similar fused triazines, as reported by Lovelette
[1]. The high-resolution mass spectrum showed a molecu-

854
L.-C. Hwang, R.-R. Wu and C.-H. Tu
Vol. 42
ethyldisilazane (HMDS; 25 mL, 118 mmoles), and a catalytic
amount of ammonium sulfate (60 mg, 0.45 mmoles) was added.
The mixture was heated under reflux for 4 hours with the exclu-
sion of moisture until a clear solution was obtained. The excess
HMDS was removed under reduced pressure to give the sily-
lated intermediate as an oil, which was then dissolved in dry
toluene (30 mL). To this stirred solution was added a solution of
benzyl bromide (3.42 g, 20 mmoles) in dry toluene (15 mL). The
reaction mixture was refluxed for 16 hours (monitored by tlc).
The solvent was evaporated to afford the crude product as an oil,
which was then applied to a silica gel (230-400 mesh) column.
one to the 5-ol in the 1,2,4-triazine ring, which aromatized
the system. Aromatization placed more electron density in
the 1,2,4-triazine ring of 3-amino-6-(2-formylhydrazino)-
1,2,4-triazin-5(2H)-one (11). In such a case, the sequential
deprotonation of the N-1 hydrogen of 6-formylhydrazino
group gave the electron pair into 1,2,4-triazine ring via π-
electron resonance. The 1,2,4-triazine ring then served as a
nucleophile, allowing for the intramolecular nucleophilic
cyclization at N-1 of 3-amino-6-(2-formylhydrazino)-
1,2,4-triazin-5-ol (12), followed by dehydration and tau-
tomerization to the predominant form 7H-tautomer 1.
The column was eluted with CHCl and MeOH (60:1) and the
3
proper fractions were combined and evaporated. The residue
thus obtained was recrystallized from CH Cl and EtOH (5:1) to
Conclusions.
2
2
give 5 (0.90 g, 32%) as bright white crystals, mp 227-229° (from
As we know, aromaticity has ubiquitous influence in mol-
ecular stability and reactivity, the reaction products to be
expected, the symmetry and geometry of molecules, and
various physicochemical properties [12,13]. In this research
two important ideas were presented: (1) that the tau-
tomerism of the N-2 hydrogen with the C-5 oxo group
induced aromatization in the 1,2,4-triazine ring, and (2) that
DMF was protonated by acetic acid on the nitrogen atom,
promoting deamination upon attack by the 6-hydrazino
group of 3 or 6. Our investigations revealed that the aroma-
tization and resonance influenced the reactivity of 1,2,4-tri-
azine π-deficient heterocyclic molecule. Evidence was pre-
sented to suggest that the cyclization is dependent on the
aromatization of the 1,2,4-triazine ring. Additionally, study-
ing and expounding the correct dominant tautomeric struc-
ture makes it possible to properly interpret the detailed
mechanisms of reactions of tautomeric heterocycles [14].
–1
CH Cl -EtOH); ir (KBr): (ν cm ) 3338, 3122, 1635, 1521,
2
2
,
1441, 1149, 1046; uv: (0.1 N HCl): λ
nm (logε) 208 (3.53),
max
270 (2.83); (H O): λ
nm (logε) 206 (3.51), 271 (2.81); (0.1 N
2
max
1
NaOH): λ
nm (logε) 231 (3.93), 260 (3.56); H nmr (200
max
MHz): δ 5.17 (s, 2H, CH ), 7.24-7.45 (m, 5H, C H ), 7.58 (s,
2
6 5
13
2H, NH ); C nmr (200 MHz): δ 55.59 (CH ), 135.52 (C-6),
2
2
+
155.09 (C-3), 157.85 (C-5); ms: m/z, 282 (M +1, 5.3), 280 (5.6),
175 (25), 106 (41), 91 (100), 69 (35), 65 (48).
Anal. Calcd. for C H N OBr: C, 42.73; H, 3.23; N, 19.93.
10
9 4
Found: C, 42.59; H, 3.20; N, 19.96.
3-Amino-2-benzyl-6-hydrazino-1,2,4-triazin-5(2H)-one (6).
A solution of compound 5 (0.84 g, 3 mmoles) and hydrazine
monohydrate (0.3 g, 6 mmoles) in ethanol (20 mL) was refluxed for
16 hours (monitored by tlc). The solvent was evaporated to 5 mL,
and then cooled. The precipitate was collected and washed with
water, followed by recrystallization from ethanol to give 6 (0.48 g,
–1
69%), mp 245-247° (from EtOH); ir (KBr): (ν cm ) 3419, 3326,
,
1562, 1496, 1056; uv: (0.1 N HCl): λ
nm (logε) 208 (4.24), 280
max
(3.73); (H O): λ
nm (logε) 220 (3.23), 292 (2.76); (0.1 N
2
max
1
NaOH): λ
nm (logε) 226 (3.19), 290 (2.69); H nmr (400 MHz):
max
EXPERIMENTAL
δ 3.90 (s, 2H, NHNH ), 5.01 (s, 2H, CH ), 6.83 (s, 2H, NH ), 7.28-
2
2
2
13
7.38 (m, 5H, C H ), 7.34 (s, 1H, NHNH ); C nmr (200 MHz): δ
6
5
2
Melting points were measured on a YANACO micromelting
point apparatus and were uncorrected. The ir spectra were taken
with potassium bromide (KBr) discs on a Perkin-Elmer FTIR 1650
spectrophotometer and the uv absorption spectra were recorded on
54.96 (CH ), 147.60 (C-6), 154.05 (C-3), 158.69 (C-5); ms: m/z
2
+
232 (M , 45), 141 (85), 106 (16), 91 (100), 69 (21), 65 (33).
Anal. Calcd. for C
H N O: C, 51.72; H, 5.21; N, 36.19.
10 12 6
Found: C, 51.61; H, 5.19; N, 35.97.
1
13
a Shimadzu UV-Visible spectrophotometer. The H nmr and
C
3-Amino-2-benzyl-6-(2-formylhydrazino)-1,2,4-triazin-5(2H)-
one (7).
nmr spectra were obtained in dimethylsulfoxide-d (DMSO-d ) on
6
6
a Varian Gemini 200 (200 MHz) or a Varian Unity 400 (400 MHz)
spectrometer. Chemical shifts are expressed in ppm (δ) with
Compound 6 (0.70 g, 3 mmoles) in acetic acid (0.36 g, 6
mmoles) and N,N-dimethylformamide (15 mL) were refluxed for
12 hours. Cooling gave white crystals which were recrystallized
from ethanol to afford 7 (0.48 g, 62%), mp 245-246° (from
tetramethylsilane (Me Si) as an internal standard. Coupling con-
4
stants J are given in Hertz (Hz); multiplicities were recorded as
broad peaks (br), singlet (s), doublets (d), triplets (t), quartets (q),
and multiplets (m). Thin layer chromatography (tlc) analyses were
–1
EtOH); ir (KBr): (ν cm ) 3296, 3194, 1666, 1509, 1374; uv:
,
performed on silica gel plates (Merck 60 F , 0.2 mm thickness),
(0.1 N HCl): λ
nm (logε) 207 (4.09), 291 (3.59) ; (H O): λ
254
max 2 max
nm (logε) 222 (4.04), 284 (3.51); (0.1 N NaOH): λ
nm (logε)
and the components were detected by uv light (254 nm). MS and
hrms spectra were obtained on a Quattro VG-5022 spectrometer
and VG 70-250S GC/MS, respectively, with an ionization potential
of 70 eV. Elemental analyses were performed on a Heraeus CHN-
O-Rapid elemental analyzer. All the solvents used were dried and
distilled under argon prior to use.
max
1
230 (4.27), 266 (3.97), 329 (3.85); H nmr (400 MHz): δ 4.98 (s,
2H, CH ), 6.97 (s, 2H, NH ), 7.22-7.37 (m, 5H, C H ), 7.98 (s,
2
2
6 5
1H, NHNHCO), 8.26 (d, 1H, J = 1.2 Hz, NHNHCO), 9.71 (d, 1H,
13
J = 1.2 Hz, CHO); C nmr (200 MHz): δ 54.95 (CH ), 145.04
2
(C-6), 154.51 (C-3), 159.99 (C-5), 167.55 (CO); ms: m/z 260
(M , 3), 141 (10), 91 (100), 69 (15), 65 (27); hrms: m/z Calcd. for
+
3-Amino-2-benzyl-6-bromo-1,2,4-triazin-5(2H)-one (5).
C H N O : 260.1022. Found: 260.1021.
11 12
6 2
Compounds 3 and 4 were prepared as described by Lovelette
[1]. Compound 4 (1.90 g, 10 mmoles) was suspended in hexam-
Anal. Calcd. for C H N O •1/2 H O: C, 49.07; H, 4.83; N,
31.21. Found: C, 49.14; H, 4.75; N, 31.19.
11 12 6 2 2

Jul-Aug 2005
A better Understanding on the Ring Cyclization Mechanism
855
Wang, Analytical Sciences, 18, 853 (2002).
6-Amino-1,2,4-triazolo[3,4-f][1,2,4]triazin-8(7H)-one (1).
[3] L. C. Hwang, J. H. Wang, C. C. Tzeng, G. H. Lee and S.
M. Peng, Analytical Sciences, 18, 723 (2002).
[4] J. Pitha, P. Fiedler and J. Gut, Collect. Czech. Chem.
Commun., 31, 1864 (1966).
[5] L. C. Hwang, D. C. Wei, M. C. Cheng, Y. Wang and C. C.
Tzeng, Nucleosides & Nucleotides, 13, 2185 (1994).
[6] L. C. Hwang, C. J. Wang, G. H. Lee, Y. Wang and C. C.
Tzeng, Heterocycles, 41, 293 (1995).
[7] C. C. Tzeng, L. C. Hwang, C. C. Chen and D. C. Wei,
Nucleosides & Nucleotides, 14, 1425 (1995).
Compound 3 (0.71 g, 5 mmoles) in acetic acid (10 mmoles) and
N,N-dimethylformamide (15 mL) were gently refluxed for 12
hours. Cooling gave compound 1 (0.51 g, 67%) as white crystals,
–1
mp >300° (from H O); ir (KBr): (ν cm ) 3331, 3186, 1719, 1663,
2
,
1351, 1293, 1090; uv: (0.1 N HCl): λ
nm (logε) 213 (3.38), 302
max
(2.89); (H O): λ
nm (logε) 217 (3.57), 296 (3.01); (0.1 N
2
max
1
NaOH): λ
nm (logε) 225 (4.20), 282 (3.65); H nmr (200
max
MHz): δ 6.36 (s, 2H, NH ), 9.00 (s, 1H, 3-H), 11.55 (br s, 1H, NH);
2
13
C nmr (200 MHz): δ 139.70 (C-3 and C-8a), 151.07 (C-6),
+
152.10 (C-8); ms: m/z 152 (M , 3), 128 (100), 100 (34), 69 (12);
[8] L. C. Hwang, E. C. Wang, K. H. Lee and C. C. Tzeng,
Chin. Pharm. J., 47, 1 (1995).
hrms: m/z Calcd. for C H N O: 152.0447. Found: 152.0446.
4
4 6
Anal. Calcd. for C H N O: C, 31.58; H, 2.65; N, 55.25.
4
4 6
[9] R. H. Hall, J. Am. Chem. Soc., 80, 1145 (1958).
[10] H. G. O. Becker, D. Beyer, G. Israel, R. Muller, W.
Riediger and H. J. Timpe, J. Prakt. Chem., 312, 669 (1970).
[11] G. Fraenkel and C. Franconi, J. Am. Chem. Soc., 82, 4478
(1960).
Found: C, 31.43; H, 2.68; N, 55.07.
Acknowledgement.
We are grateful to the National Science Council, Taiwan,
ROC, for support under grant NSC 88-2113-M-037-013.
[12] A. R. Katritzky, P. Barczynski, G. Musumarra, D. Pisano
and M. Szafran, J. Am. Chem. Soc., 111, 7 (1989).
REFERENCES AND NOTES
[13]
A. R. Katritzky, M. Karelson and N. Malhotra,
Heterocycles, 32, 127 (1991).
[1] C. A. Lovelette, J. Heterocyclic Chem., 16, 555 (1979).
[2] L. C. Hwang, C. H. Tu, J. H. Wang, G. H. Lee and Y.
[14] J. Elguero, A. R. Katritzky and O. V. Denisko, Adv.
Heterocycl. Chem., 76, 1 (2000).