Heterophase N-aminomethylation of 5-arylidenepseudothiohydantoins by arylamines and aqueous formaldehyde in aromatic solvents: Effect of substituents in the heterocyclic substrate and the aryl amine on the efficiency of the process

Article abstract of DOI:10.1007/s10593-006-0185-0

We have obtained a series of 3-aryl-7-arylidene-3,4-dihydro-2H-[1,3] thiazolo[3,2-a][1,3,5]triazin-6(7H)-ones by means of heterophase aminomethylation of 2-amino-5-arylidene-1,3-thiazol-4(5H)-ones with aqueous formaldehyde and aromatic amines in benzene or toluene. We explain the effect of substituents in the heterocyclic substrate and the aryl amine on the efficiency of the process within a detailed scheme for one of the possible aminomethylation reaction routes.

Full text of DOI:10.1007/s10593-006-0185-0

Chemistry of Heterocyclic Compounds, Vol. 42, No. 7, 2006
HETEROPHASE N-AMINOMETHYLATION
OF 5-ARYLIDENEPSEUDOTHIOHYDANTOINS
BY ARYLAMINES AND AQUEOUS FORMALDEHYDE
IN AROMATIC SOLVENTS: EFFECT OF SUBSTITUENTS
IN THE HETEROCYCLIC SUBSTRATE AND THE ARYL
AMINE ON THE EFFICIENCY OF THE PROCESS
S. M. Ramsh, N. L. Medvedskiy, and S. O. Uryupov
We have obtained a series of 3-aryl-7-arylidene-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-
6(7H)-ones by means of heterophase aminomethylation of 2-amino-5-arylidene-1,3-thiazol-4(5H)-ones
with aqueous formaldehyde and aromatic amines in benzene or toluene. We explain the effect of
substituents in the heterocyclic substrate and the aryl amine on the efficiency of the process within a
detailed scheme for one of the possible aminomethylation reaction routes.
Keywords: 2-amino-5-arylidene-1,3-thiazol-4(5H)-ones (5-arylidenepseudothiohydantoins), 3-aryl-
7-arylidene-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-ones, aminomethylation, synthesis.
We previously developed a method for obtaining derivatives of 3,4-dihydro-2H-[1,3]thiazolo[3,2-a]-
[1,3,5]triazin-6(7H)-one by N-aminomethylation of either 2-amino-1,3-thiazol-4(5H)-one (pseudothiohydantoin,
1) or its 5-arylidene derivative with aqueous formaldehyde and aliphatic or aromatic amines in ethanol [1, 2].
Subsequently high biological activity was observed in derivatives of 3,4-dihydro-2H-[1,3]thiazolo[3,2-
a][1,3,5]triazin-6(7H)-one [3]. Among other compounds, in [1] 7-benzylidene-3-phenyl-3,4-dihydro-2H-
[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-one (3a) was synthesized starting from 2-amino-5-benzylidene-1,3-
thiazol-4(5H)-one (5-benzylidenepseudothiohydantoin, 2a) and aniline. The aim of this work was to synthesize
analogs of compound 3a with substituents on both phenyl rings: 7-R1-benzylidene-3-R-phenyl-3,4-dihydro-2H-
[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-ones 3. In order to achieve this goal, the previously proposed [1, 2]
method for obtaining compounds 3 was modified. In this work, we also analyzed and explained the effect of the
R1 substituents in 2-amino-5-R1-benzylidene-1,3-thiazol-4(5H)-one (5-R1-benzylidenepseudothiohydantoin, 2)
and R in the arylamine on the efficiency of the aminomethylation process.
O
O
N
N
p-R1C6H4CHO
R1
H2N
S
2a-e
AcONa, AcOH
H2N
S
1
2 a R1 = H, b R1 = NEt2, c R1 = NMe2, d R1 = OMe, e R1 = NO2
__________________________________________________________________________________________
St. Petersburg State Technical University, St. Petersburg 198013, Russia; e-mail:
gsramsh@mail.wplus.net. Translated from Khimiya Geterotsiklicheskikh Soedinenii, No. 7, pp. 1095-1104,
July, 2006. Original article submitted June 20, 2005.
948
0009-3122/06/4207-0948©2006 Springer Science+Business Media, Inc.

R
CH2O aq.,
p-RC6H4NH2
O
N
N
2a-e
R1
N
S
3a-n
The starting arylidenepseudothiohydantoins 2a-e were obtained by the conventional method: boiling
pseudothiohydantoin 1 with aromatic aldehydes in glacial acetic acid in the presence of sodium acetate [4]. This
method is based on the Grenacher reaction, which in turn is a special case of the Knoevenagel reaction [5], a
typical case of nucleophilic addition to a carbonyl group occurring via a base catalysis mechanism [6].
We attempted to synthesize thiazolotriazines 3 according to the well-known method in [1]:
aminomethylation of compounds 2 with aqueous formaldehyde and aromatic amines in ethanol. The
"unsubstituted" thiazolotriazine 3a was obtained rather easily by such a method, although in order to achieve a
30% yield the reaction time was increased 20-fold compared with that indicated in [1]. However, as we
established, this method cannot be used as a universal method for obtaining analogs of compound 3a with
substituents on both phenyl rings. Attempts to synthesize compounds 3l (R = H, R1 = NO2) and 3n
(R = R1 = NO2) with the most unfavorable substituent combinations (see below) were unsuccessful due to the
high recovery (~85%) of the starting compound 2e and the meager yield, on the order of 1-2%. Obviously under
the given conditions, these compounds are formed too slowly, and prolonged heating in aqueous ethanol
probably does not promote an increase in yield due to its significant solvolysis.
We also could not obtain thiazolotriazine 3l by aminomethylation of substrate 2e with aniline and
paraform in benzene; the starting compound 2e did not dissolve, and remained unchanged even in boiling
solvent. The failure of this experiment may be explained as follows. We know [7] that the aminomethylation
reaction is accelerated in hydroxyl-containing media; furthermore, hydroxymethylation, which is also possible
both as one of the steps in the aminomethylation reaction and as one of the steps in formation of the target
material, also is favored by hydroxyl-containing solvents. Finally, the anhydroformaldehyde aniline formed from
paraform and aniline [1] in itself probably does not have aminomethylating properties, and the equilibrium
content of the aminomethylating species formed as a result of its hydration (methylolphenylamine
(anilinomethanol)) is too low in such a reaction medium.
Based on the above, we decided to carry out the aminomethylation reaction in an aromatic solvent
immiscible with water: benzene or toluene, using formaldehyde in the form of an aqueous solution as the
methylene component. Boiling the reaction mixture, besides accelerating the process, should ensure at least a
minimum concentration of substrate 2 in the liquid phase, while vigorous stirring should ensure its dispersion
and fine emulsification of the water in the organic solvent, with an increase in the contact surface area between
the solid and the two liquid phases. Such a synthesis procedure proved to be successful, and allowed us to obtain
a series of novel arylidene thiazolotriazines 3b-n.
We assume that owing to the presence of water in the reaction mixture, a concentration of the
aminomethylating species (methylol phenylamine) sufficient for the reaction to occur is maintained, and
aminomethylation occurs either in the aqueous phase or in the wet benzene, and the reaction product 3 generally
is deposited in the benzene phase, which to some degree prevents its solvolysis. Visual evidence for the extent of
aminomethylation was the gradual complete or partial disappearance of the precipitate of heterocyclic substrate
2.
The ratio of the reagents in each case was selected by trial and error, aiming for an acceptable "reaction
time–yield" combination.
Compound 3l was obtained in the highest yield with six times as much aminomethylating agents as
substrate 2e, and after boiling for 3.5 h, the recovery of starting compound 2e was 55% while the yield of
949

reaction product 3l after recrystallization was only 9%. When using stoichiometric amounts of reagents,
compound 3l was obtained in even lower yield.
In synthesis of other thiazolotriazines 3 (Tables 1 and 2), we noted that although excess
aminomethylating agents also promote an increase in the extent of conversion of the heterocyclic substrate 2,
increasing the rate and shifting the equilibrium for aminomethylation toward the reaction product 3, the product
of reaction between the arylamine and formaldehyde formed with an excess (anhydroformaldehyde aniline in the
case of aniline) interferes with separation of compound 3. So the rest of the thiazolotriazines 3 were obtained
using a 1.5-fold or slightly greater excess of formaldehyde (3-3.5 moles per mole substrate) and a stoichiometric
or slightly greater amount of arylamine (1-1.2 moles per mole substrate). We should point out that the yield only
very approximately reflects how readily reaction occurs and the extent of conversion of the starting compounds
2 to form target compound 3, while the reaction time is a more objective characteristic.
In order to obtain compounds 3a,b,d,e,i,j, brief boiling of the reaction mixture was required to
completely dissolve the starting compound (10-15 min). The boiling time was much longer for obtaining
compounds 3h,g,m (several hours), and the starting compound was completely dissolved only in the case of
compound 3h. The low yield of compound 3i (R = H, R1 = OMe) (only 14%) is obviously connected with the
complications involved in isolating and purifying it.
Compounds 3c,f,k,n, the products of aminomethylation of 5-arylidenepseudothiohydantoins 2a,b,d,e by
p-nitroaniline, are much more difficult to obtain than compound 3, which does not contain a nitro group. The
reaction time for obtaining these compounds is approximately an order of magnitude longer than when obtaining
compounds 3 (with no nitro group), and is of the same order of magnitude as the reaction time for obtaining
compounds 3l,m, which also contain a nitro group. The solubility of compounds 3 containing a nitro group is
extremely low in benzene, toluene, acetonitrile, acetone, chloroform, dioxane, DMF, DMSO, acetic acid,
1
ethanol, and ethyl acetate, which makes it very difficult to recrystallize them and to record their H NMR
spectra.
In the IR spectra of compounds 3a-n (Table 2), there are no absorption bands for stretching vibrations of
the NH groups, which are present in the spectra of the starting compounds 2a-e, while the frequencies of the
stretching vibrations for the C=O and C=N groups lie within the ranges 1690-1710 and 1625-1665 cm-1
respectively, which is typical of their analogs [1, 2].
1
The H NMR spectra of thiazolotriazines 3a-n in heated DMSO-d6 are similar to the spectra of
previously synthesized substituted thiazolotriazines [1, 2]. The signals from the protons of the methylene groups
2H-2 and 2H-4 appear at 5.22-5.54 ppm and 4.91-5.22 ppm respectively, while the signals from aromatic
protons of the 3-aryl and 7-arylidene groups and also the CH protons of the arylidene groups appear in the range
6.73-8.32 ppm. When a substituent is present, the aromatic protons appear as pairs of doublets. Overlap of
individual signals generally gives a complex multiplet pattern in the aromatic absorption region.
Thus from the synthesis results we trace out the following qualitative patterns, although not entirely
clearly (not surprisingly for a synthesis experiment rather than a specially devised physicochemical experiment).
The compounds formed most readily and/or in highest yield are 3e,h,j, i.e., the compounds where electron-donor
substituents are present both in the arylamine moiety and in the arylidene group. The "unsubstituted" compound
3a, and also compounds 3b,d,g,i with one electron-donor group in the molecule, are intermediate with respect to
how readily the reaction occurs and the achievable yield. Compounds 3c,f,k, which have a nitro group in the
arylamine moiety, are more difficult to obtain; and compounds 3l-n with a nitro group in the arylidene group are
the most difficult to obtain and have the lowest yields. Among the latter, compound 3n (with nitro groups on
both phenyl rings) has the poorest parameters.
Our earlier experiments on aminomethylation of pseudothiohydantoin [1] with aqueous formaldehyde
and aniline provided indirect evidence that of all the possible routes [8] for the aminomethylation reaction, the
route with intermediate formation of methylol arylamine is realized. The reaction of aniline with formaldehyde
occurs rather rapidly, and first we see precipitation from the reaction mixture of anhydroformaldehyde aniline, a
950

951

952

precursor of which is methylol arylamine*, which then is consumed in the reaction with the heterocyclic
substrate. This means, first of all, that the first step of aminomethylation (formation of methylol arylamine) is not
the rate-determining step, and the equilibrium for it is established rather rapidly; secondly, the equilibrium for
the second step (formation of the N-aminomethyl derivative), while it is generally achieved, is established
appreciably more slowly, and the rate of the process as a whole or the final equilibrium should be influenced by
electronic effects of substituents both on the arylamine and the heterocyclic component of the reaction.
The recently obtained experimental data also fit entirely into the likely scheme for the formation of the
end product of reaction 3 and the detailed scheme for the N-aminomethylation reaction with preliminary
equilibrium in the step of formation of the aminomethylating species methylol arylamine, rapidly established
before the rate-determining step of the reaction between this species and the heterocyclic substrate 2, which also
may be an equilibrium step. Our theoretical analysis in accordance with these schemes for the cross effect of
substituents R1 in heterocyclic substrate 2 and R in the arylamine on the rates and equilibria of the reactions of
aminomethylation and subsequent cyclization*2 predicts that donor substituents R facilitate the reaction and
conversely acceptor substituents R slow down the reaction in the aryl amine, in addition to the more noticeable
facilitating effect of donor substituents R1 and the retarding effect of acceptor substituents R1 on the heterocyclic
substrate 2, which on the whole is consistent with the presented results of the synthesis experiments.
We should point out that the precipitate of reaction product 3 formed from the reaction mixture during
the process may promote successful occurrence of the process and mask substituent effects.
EXPERIMENTAL
The 1H NMR spectra were recorded on a Bruker AM-200 (200 MHz) and a Bruker AM-300 (300 MHz)
in DMSO-d6, internal standard HMDS (δ 0.05 ppm). The IR spectra were obtained on an IKS-29 spectrometer in
nujol. TLC was run on Silufol UV-254 plates, eluent 1:2 acetone–hexane. The starting compounds 2a-e are not
chromatographically mobile in this solvent system. For low-solubility compounds 3c,f,k-n, we also used the
system 1:9 ethanol–chloroform.
Starting 5-Arylidenepseudothiohydantoins 2a-e were obtained according to a modified method from
[4]. Pseudothiohydantoin 1 (0.2 mol) was boiled with equimolar amounts of the aromatic aldehyde and
anhydrous sodium acetate in 100 mL glacial acetic acid with vigorous stirring. Then the reaction mixture was
cooled down. The precipitate formed was filtered out, washed with water, and recrystallized from acetic acid.
The precipitates of compounds 2a,b were washed before filtration from a 300 ml flask of water. The boiling time
was 80 min, 20 min, 20 min, 20 min, and 150 min respectively.
7-Benzylidene-3-phenyl-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-one (3a). Compound
2a (0.80 g, 3.9 mmol), aniline (0.47 g, 0.46 ml, 5.0 mmol), and formalin (1.10 ml, 13.6 mmol) in toluene (9 ml)
was boiled with stirring for 2 h. The precipitate formed was filtered out and recrystallized from toluene.
7-Benzylidene-3-(4-nitrophenyl)-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-one (3c).
Compound 2a (4.08 g, 20 mmol), p-nitroaniline (2.76 g, 20 mmol), and formalin (4.8 ml, 60 mmol) in benzene
(50 ml) were boiled with stirring for 1 h. The gummy precipitate falling out of the reaction mixture after cooling
was ground with hexane to a crystalline consistency, and the precipitate obtained was recrystallized from
acetonitrile.
7-[(4-Diethylamino)benzylidene]-3-phenyl-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-
one (3d). Compound 2b (5.51 g, 20 mmol), aniline (2.05 g, 2.01 ml, 22 mmol), and formalin (4.8 ml, 60 mmol)
in benzene (100 ml) were boiled with stirring for 10 min until the starting compound was completely dissolved.
_______
* Anhydroformaldehyde aniline plays the role of the depo form of its own precursor: methylol arylamine.
*2 Additional data can be obtained from the authors.
953

The precipitate falling out from the hot organic layer as it cooled was recrystallized twice from benzene over
Al2O3.
Compounds 3g (50/15), 3b (20/50), 3j (20/50), 3h (11/50), and 3e (15/30) were obtained similarly. In
parentheses are given the amount of substrate (mmol)/amount of solvent (ml) used for the synthesis.
7-(4-Methoxybenzylidene)-3-phenyl-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]-triazin-6(7H)-one (3i).
Compound 2d (11.7 g, 50 mmol), aniline (5.40 g, 5.28 ml, 58 mmol), and formalin (12.0 ml, 150 mmol) in
benzene (100 ml) were boiled with stirring for 10 min until the starting compound was completely dissolved.
The organic layer was separated, the benzene was driven off at 25°C under vacuum almost to dryness, and the
residue was treated with benzene (75 ml). Hexane (100 ml) was added to the benzene extract. The mixture
obtained was brought to boiling over Al2O3 and filtered. The precipitate falling out as the filtrate cooled was
recrystallized from a 2:1 benzene–hexane mixture over Al2O3.
7-(4-Methoxybenzylidene-3-(4-nitrophenyl)-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7)-
one (3k). Compound 2d (1.87 g, 8 mmol), p-nitroaniline (1.10 g, 8 mmol), and formalin (1.9 ml, 24 mmol) in
benzene (50 ml) were boiled for 2 h with stirring, and then cooled down to room temperature. The precipitate
was filtered out and recrystallized from benzene.
Compound 3f was obtained similarly from compound 2b (3.58 g, 13 mmol), p-nitroaniline (1.80 g,
13 mmol), and formalin (3.6 ml, 45 mmol).
7-(4-Nitrobenzylidene)-3-phenyl-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-one (3l).
Compound 2e (3.75 g, 15 mmol), aniline (8.38 g, 8.20 ml, 90 mmol), and formalin (7.2 ml, 90 mmol) in benzene
(200 ml) were boiled for 3.5 h with stirring. The hot reaction mixture was filtered to remove unreacted starting
compound (2.10 g). The precipitate falling out from the filtrate after cooling was recrystallized from a 3:1
benzene–hexane mixture.
Compound 3n was obtained similarly from compound 2e (5.00 g, 20 mmol), p-nitroaniline (2.76 g,
20 mmol), and formalin (4.8 ml, 60 mmol).
3-(4-Methoxyphenyl)-7-(4-nitrobenzylidene)-3,4-dihydro-2H-[1,3]thiazolo[3,2-a][1,3,5]triazin-6(7H)-
one (3m). Compound 2e (5.00 g, 20 mmol), p-anisidine (2.50 g, 20 mmol), and formalin (4.8 ml, 60 mmol) in
benzene (50 ml) were boiled for 4 h with stirring. The reaction mixture was filtered to remove undissolved
starting compound. The filtrate was dried for 24 hours by calcium chloride, the benzene was driven off under
vacuum, and the residue was recrystallized from acetonitrile.
REFERENCES
1.
2.
3.
S. Yu. Solov'eva-Yavits, S. M. Ramsh, and A. I. Ginak, Khim. Geterotsikl. Soedin., 477 (1981) [Chem.
Heterocycl. Comp., 17, 340 (1981)].
S. Yu. Solov'eva-Yavits, S. M. Ramsh, and A. I. Ginak, Khim. Geterotsikl. Soedin., 1204 (1983). [Chem.
Heterocycl. Comp., 19, 961 (1983)].
S. M. Ramsh, T. S. Basieva, O. Yu. Uryupov, S. Yu. Solov'eva, V. M. Vinogradov, A. I. Ginak,
B. I. Krivoruchko, I. A. Shelkovnikov, and Yu. N. Shanin, USSR Inventor's Certificate 1095612; B. I.,
No. 20, 197 (1984).
4.
5.
D. Liberman, J. Himbert, and L. Hengl, Bull. Soc. Chim. Fr., 1120 (1948); Chem. Abstr., 43, 3819
(1949).
K. V. Vatsuro and G. L. Mishchenko, Name Reactions in Organic Chemistry [in Russian], Khimiya,
Moscow (1976), p. 152, p. 215.
6.
7.
8.
J. March, Organic Chemistry [Russian translation], Mir, Moscow (1987), Vol. 3, p. 379, p. 387.
V. Tychlopoulos and J. H. P. Tyman, Synth. Commun., 16, 1401 (1986).
M. Tramontini, Synthesis, 703 (1973).
954