Carbon protonation of 2,4,6-triaminopyrimidines: Synthesis, NMR studies, and theoretical calculations

Article abstract of DOI:10.1021/jo0605703

Several C(5)-substituted 2,4,6-triaminopyrimidine derivatives and their HBF4 salts were synthesized to study the carbon protonation of the pyrimidine ring. NMR investigations in DMSO-d₆ prove experimentally that, in addition to the usual protonation at N(1), the compounds can be protonated at C(5) as well. We present several new stable cationic σ-complexes in the pyrimidine series, where C(5) protonation predominates over N(1) protonation. Quantum chemical calculations using the B3LYP/cc-pVDZ method were utilized in the gas phase and also in DMSO solvent with the polarized continuum model (PCM) method to rationalize the observed protonation behavior. Results of the calculations accord with the experimental observations and prove that combined steric and electronic effects are responsible for the observed C(5) protonation and for σ-complex stability. We demonstrate that C(5) protonation is a general feature of the 2,4,6-triaminopyrimidine system.

Full text of DOI:10.1021/jo0605703

Carbon Protonation of 2,4,6-Triaminopyrimidines: Synthesis, NMR
Studies, and Theoretical Calculations
Bala´zs Ne´meth,‡ Csaba We´ber,† Tama´s Veszpre´mi,‡ Tama´s Ga´ti,† and AÄ da´m Demeter*,†
Department of Inorganic Chemistry, Budapest UniVersity of Technology and Economics, Gelle´rt te´r 4,
H-1521 Budapest, Hungary, and Gedeon Richter Ltd., H-1475 Budapest, P.O. Box 27, Hungary
a.demeter@richter.hu
ReceiVed March 14, 2006
Several C(5)-substituted 2,4,6-triaminopyrimidine derivatives and their HBF4 salts were synthesized to
study the carbon protonation of the pyrimidine ring. NMR investigations in DMSO-d6 prove experimentally
that, in addition to the usual protonation at N(1), the compounds can be protonated at C(5) as well. We
present several new stable cationic σ-complexes in the pyrimidine series, where C(5) protonation
predominates over N(1) protonation. Quantum chemical calculations using the B3LYP/cc-pVDZ method
were utilized in the gas phase and also in DMSO solvent with the polarized continuum model (PCM)
method to rationalize the observed protonation behavior. Results of the calculations accord with the
experimental observations and prove that combined steric and electronic effects are responsible for the
observed C(5) protonation and for σ-complex stability. We demonstrate that C(5) protonation is a general
feature of the 2,4,6-triaminopyrimidine system.
Introduction
are also known as strong AGAT inhibitors.3 Furthermore,
hexaalkyl-2,4,6-triaminopyrimidines constitute important sub-
units of 21-aminosteroids (e.g., tirilazad) and 2-(aminomethyl)-
chromans that inhibit iron-dependent lipid peroxidation and
protect against central nervous system (CNS) trauma.4 Proto-
nation studies might provide basic insight into the mechanism
The biological importance of the pyrimidine system has
induced extensive research toward the understanding of structure-
activity relationships.1 Substituted 2,4-diaminopyrimidines (e.g.,
pyrimethamine and thrimethoprim), as well as the respective
heterocyclic analogues (e.g., piritrexim and methotrexate),
constitute a class of pharmacologically important dihydrofolate
reductase (DHFR) inhibitors.2 Substituted diaminopyrimidines
of action as clearly demonstrated for DHFR inhibitors.5
A
number of experimental and theoretical studies explored the site
of protonation for simple aminopyrimidine derivatives. They
indicated that protonation occurs at the pyrimidine ring nitrogen
[N(1) and/or N(3)] and/or at the exocyclic amino group both in
the gas phase and in solution.6
Recently, we have discovered that hexamethyl-2,4,6-triami-
nopyrimidine (4) (see Chart 1) could be protonated at the C(5)
position in water, leading to an equilibrium between the C(5)
* To whom correspondence should be addressed. Phone: +36-1-431 5922.
Fax: +36-1-431 6006.
‡ Budapest University of Technology and Economics.
† Gedeon Richter Ltd.
(1) For a comprehensive review of pyrimidines, see: (a) Brown, D. J.;
Evans, R. F.; Cowden, W. B.; Fenn, M. D. The Pyrimidines; Wiley-
Interscience: New York, 1994. (b) ComprehensiVe Heterocyclic Chemistry;
Katritzky, A. R., Rees, C. W., Scriven, E. F., Eds.; Pergamon: Oxford,
1996; Vol. 6, pp 93-233. (c) ComprehensiVe Heterocyclic Chemistry;
Katritzky, A. R., Rees, C. W., Eds.; Pergamon: Oxford, 1984; Vol. 3, pp
57-157.
(2) (a) Hitchings, G. H. Angew. Chem., Int. Ed. Engl. 1989, 28, 879-
885. (b) Hitchings, G. H. ReV. Port. Farm. 1991, 41, 21-29. (c) Zink, M.;
Lanig, H.; Troschu¨tz, R. Eur. J. Med. Chem. 2004, 39 (12), 1079-1088.
(d) Gangjee, A.; Zeng, Y.; McGuire, J. J.; Kisliuk, R. J. J. Med. Chem.
2005, 48 (16) 5329-5336 and references therein.
(3) Terashima, I.; Kohda, K. J. Med. Chem. 1998, 41, 503-508.
(4) (a) Jacobsen, E. J.; VanDoornik, F. J.; Ayer, D. E.; Belonga, K. L.;
Braughler, J. M.; Hall, E. D.; Houser, D. J. J. Med. Chem. 1992, 35, 4464-
4472. (b) Jacobsen, E. J.; McCall, J. M.; Ayer, D. E.; VanDoornik, F. J.;
Palmer, J. R.; Belonga, K. L.; Braughler, J. M.; Hall, E. D.; Houser, D. J.;
Krook, M. A.; Runger, T. A. J. Med. Chem. 1990, 33, 1145-1151.
(5) Feeney, K. Angew. Chem., Int. Ed. 2000, 39, 290-312.
10.1021/jo0605703 CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/02/2006
4910
J. Org. Chem. 2006, 71, 4910-4918

Carbon Protonation of 2,4,6-Triaminopyrimidines
CHART 1
CHART 2
SCHEME 1a
a Reagents and conditions. 5: MeI (5 equiv), CH2Cl2, 40 °C, 120 h. 6:
A method, t-BuI (3 equiv), DIPEA (3 equiv), THF, 8 h, 65 °C; A method,
isobutene, cat. p-TsOH, 65 °C, 8 h. 7: NaNO2 (2 equiv), CH2Cl2-AcOH-
H2O, 0-5 °C, 1 h. 8: NO2BF4‚Et2O (2.3 equiv), CH2Cl2, -70 °C, 4 h then
room temperature for 24 h. 9 (from 7): H2, Pd/C (20%), Ac2O (3 equiv),
THF, room temperature, 3 h. 10: NaOtBu (1.05 equiv), EtOH, reflux, 8 h.
11: NaOtBu (0.97 equiv), EtOH, reflux, 8 h.
Results and Discussion
and N(1) protonated forms.7 We have also shown that a sterically
congested 1,1-bis(pyrimidin-5-yl)-2-chloroethene-type derivative
is, peculiarly, fully monoprotonated at the C(5) position and
forms a stable cationic σ-complex. A detailed NMR analysis
with an emphasis on the story behind our structural conclusions
regarding C(5) protonation has also been given.8 We proposed
that both steric and electronic effects might be responsible for
the observed C(5) protonation, and they could account for the
stability of the σ-complex in the bispyrimidine derivative.
To explore these effects in detail, we initiated a systematic
study of the protonation behavior of the 2,4,6-triaminopyrimi-
dine system by NMR and quantum chemical calculations. To
this end, we synthesized a selected set of simple C(5)-substituted
2,4,6-triaminopyrimidine derivatives (Chart 1) and their respec-
tive HBF4 salts. Compound 6 was failed to be prepared and
was only theoretically studied. Our ultimate goal was to find
new stable cationic σ-complexes among simple triaminopyri-
midines and to provide experimental and theoretical evidence
for their stability.
The investigated molecules and the numbering of atoms are
shown in Chart 1. Throughout this paper, the protonated forms
are denoted with the site of protonation after the number
assigned to each molecule. For example, 4N(1) denotes molecule
4 (base form) that is protonated at N(1), and 4C(5) stands for
the C(5) protonated form (Chart 2). Accordingly, e.g., the signal
of N(1) of compound 4 corresponds to the 15N NMR chemical
shift of the nitrogen on position 1. Because most of the base
forms of the investigated molecules are practically insoluble in
water, the protonation behavior was investigated in dimethyl
sulfoxide (DMSO) with the respective HBF4 salts. HBF4 was
chosen for this study to obtain sharp NMR signals.
Synthesis. 1, 2 and 4 as well as the HBF4 salt of 1 are known
compounds,10 and 3, 5, and 7-11 as well as the respective HBF4
salts of 2-5 and 7-11 are newly synthesized in the pyrimidine
series.
Two different approaches were used for the synthesis of
5-substituted 2,4,6-triaminopyrimidines (Scheme 1). First, 5-sub-
stituted 2,4,6-tris(dimethylamino)pyrimidines were synthesized
by reaction of 4 with the appropriate electrophiles.11 Second,
in the case of 5-substituted 4,6-di- and 2,4,6-triaminopyrim-
idines, condensation of the respective malonitrile and amidine
derivatives proved to be successful. A detailed description of
the applied reaction conditions and the method of preparation
is given in the Experimental Section.
More than a hundred years after its first synthesis (1899),9
we show a so far unexplored chemical character of the
pyrimidine ring.
(6) (a) Threadgill, M. D.; Griffin, R. J.; Stevens, M. F. G.; Wong, S. K.
J. Chem. Soc., Perkin Trans. I 1987, 2229-2234. (b) Sta¨deli, W.;
Philipsborn, W.; Wick, A.; Kompis, I. HelV. Chim. Acta 1980, 63, 504-
522. (c) Riand, J.; Chenon, M.-T. Th.; Lumbroso-Bader, N. J. Am. Chem.
Soc. 1977, 99, 6838-6845. (d) Wagner, R.; von Philipsborn, W. HelV. Chim.
Acta 1970, 53, 299-320. (e) Nguyen, V. Q.; Turee`ek, F. J. Am. Chem.
Soc. 1997, 119, 2280-2290. (f) Riand, J.; Coupry, C.; Chenon, M.-T. J.
Chem. Soc., Perkin Trans. 2 1981, 783-788. (g) Griffiths, D. V.; Swetnam,
S. P. J. Chem. Soc., Chem. Commun. 1981, 1224-1225.
(7) Demeter, AÄ .; We´ber, Cs.; Brlik, J. J. Am. Chem. Soc. 2003, 125,
2535-2540.
It is interesting to note that methylation of 4 with iodomethane
afforded a 2:1 mixture of N(7)- and C(5)-methylated products
(10) (a) Gabriel, S. Chem. Ber. 1901, 34, 3362-3366. (b) Gerngross,
O. Chem. Ber. 1905, 38, 3394-3408. (c) Wheland, R. S.; Martin, E. L. J.
Org. Chem. 1975, 40, 3101-3109. (d) Mohrle, H.; Von der Lieck-
Waldheim, Sci. Pharm. 1997, 65 (3), 93-98. (e) Joshi, A. A.; Narkhede,
S. S.; Viswanathan, C. L. Bioorg. Med. Chem. Lett. 2005, 15, 73-76.
(11) Hemmerich, P.; Prijs, B.; Erlenmeyer, H. HelV. Chim. Acta. 1959,
42, 1604-1611.
(8) Demeter, AÄ .; We´ber, Cs. Concepts Magn. Reson., Part A 2004, 22
(1) 12-24.
(9) Gabriel, S.; Colman, J. Chem. Ber. 1899, 32, 1525-1538.
J. Org. Chem, Vol. 71, No. 13, 2006 4911

Ne´meth et. al
TABLE 1. 1H NMR Chemical Shifts in ppma
N(7)H2
N(8,9)H2
or
or
H(5)
N(7)(CH3)2
N(8,9)(CH3)2
C(10)H3
C(10)(CH3)3
others
1
4.86
4.96
5.49
6.31
5.24
6.96
5.06
6.86
8.31
3.01
3.13
3.27
3.01
3.27
5.58
6.38
5.51
6.90
5.15
6.58
x 8.64, y 8.85
2.94
3.08 br
x 3.19, y 3.22
2.83
1N(1)
2
2N(1)
3
3N(1)
3C(5)
4
4N(1)
4C(5)
5
1.69
1.76
N(1,3)H, 10.44 br
N(1,3)H, 9.94 br
1.36
1.37
0.98
3.08
4.95
5.17
3.79
1.91
1.25 d
5C(5)
7
4.06 q
x 3.20, y 3.28
x 3.11, y 3.16 x 2.78, y 3.04
x 3.30, 3.44 br
7O
3.26
x 3.12, y 3.22
x 3.27, y 3.32
3.09
x 3.31, y 3.43
2.90
2.94
x 3.18, y 3.19
5.15
NOH, 14.37
8
3.12
3.30
3.01
3.03
3.30
8C(5)
9major
9minor
9C(5)
10
6.97
NHCOCH3, 8.54; NHCOCH3, 1.90
NHCOCH3, 8.16; NHCOCH3, 1.37
NHCOCH3, 9.00 d; NHCOCH3, 1.88
Pyr-CH2, 1.75-1.84 m; Pyr-NCH2, 3.28-3.35 m
Pyr-CH2, 1.80-2.00 m; Pyr-NCHx,y, 3.42-3.56 m,
3.64-3.76 m
SCH3, 2.32
SCH3, 2.53; N(1,3)H, 12.52 br
SCH3, 2.49
5.69 d
3.09
1.37
0.98
10C(5)
x 8.63, y 8.94
11
11N(1)
11C(5)
5.70
7.00
x 9.72, y 9.97
1.37
1.36
0.97
3.27
a The signals are singlets unless otherwise noted. d ) doublet, q ) quartet, m ) multiplet, br ) broad. x,y denotes peaks due to chemically nonequivalent
Me groups or H atoms within a given NMe2 or NH2 group.
thus providing a convenient synthetic pathway to 5. Under the
same conditions, methylation of 1 gave only the N(1)-methylated
product (see Supporting Information). The difference in the
observed reactivity can be rationalized as follows. In 1, the
nucleophilic character of C(5) is weak compared to N(1), thus
giving the N(1)-methylated derivative. In 4, on the other hand,
C(5) is a strong nucleophilic center leading to 5C(5). Because
the N(1) position is sterically hindered by the bulky (dimethyl)-
amino groups at C(2) and C(4,6), the less stable N(7)-methylated
derivative is formed. Tert-butylation of 4 was attempted with
tert-butyl-bromide and with the tert-butyl cation generated from
isobutene, but formation of 6 could not be detected. HBF4 salts
of the respective bases were prepared by adding the acid in
diethyl ether to the dichloromethane solution of the respective
aminopyrimidine. 9 was synthesized from 7 in two steps.
Catalytic hydrogenation of the nitroso compound 7 was followed
by one-pot acylation of the 5-amino derivative with acetic
anhydride.
NMR Spectroscopy. The increased electron density of the
2,4,6-triaminopyrimidine system, as compared to pyrimidine,
becomes clearly visible by inspection of the respective chemical
shifts (Tables 1-3). On the 15N NMR spectra, the electron-
donating resonance effect of the (dimethyl)amino groups results
in a significant upfield shift of N(1,3) (Table 3) relative to
pyrimidine.6b Moreover, a substantial upfield shift could be
observed on the 13C NMR spectra for C(5) (1 ∆δ -47.1 ppm,
4 ∆δ -51.2 ppm, Table 2), which indicates increased electron
density at this position as well. It is intriguing to investigate
the effect of the alkyl substituents at C(5) on the NMR chemical
shifts. We observed, in general, that substitution at C(5) results
in a significant downfield shift of C(5) and a small downfield
shift of N(1,3) (Tables 2 and 3), which points to decreased
electron density at these atoms relative to the unsubstituted
derivatives. We also observed an upfield shift at C(4,6) for 2
and 3; however, in 5, a significant downfield shift was detected.
This can be rationalized intuitively; in 5, the steric effect prevails
over the electronic effect and indicates less conjugation of the
(dimethyl)amino groups at C(4,6). The upfield shift of N(8,9)
supports this assumption. These observations are in line with
the calculated geometries and NPA charges (see below).
The base forms exhibit spectral symmetry with respect to
the C(5)-C(2) axis except for 7. In 7, the conjugation of the
electron-withdrawing nitroso group with the pyrimidine system
is strong enough to cause slow NO rotation, i.e., slow mutual
exchange on the NMR chemical shift time scale, which results
in nonequivalence of the C(4) and C(6) as well as of the N(1)
and N(3) atoms and also of the C(4)-NMe2 and C(6)-NMe2
groups. Moreover, the methyl groups within the (dimethyl)-
amino groups at the C(2), C(4), and C(6) atoms also show
chemical shift nonequivalence due to the double bond character
of the C-NMe2 bonds. Rotation of the (dimethyl)amino groups
for the other investigated bases is fast on the NMR chemical
shift time scale. In the 1H, 13C, and 15N NMR spectra of 9, two
signal sets were observed due to the cis and trans amide
rotamers. On the basis of the measured strong NH-COCH3
NOE, we assign the major species (80%) to the trans amide
rotamer.
NMR investigations of the respective HBF4 salts expose the
protonation behavior of the studied compounds. Because of the
relatively large activation energy associated with carbon pro-
tonation of aromatic systems (i.e., sigma complex formation),
C(5) protonation is a slow process. Therefore, the exchange
between C(5) and N(1) protonated forms is slow on the NMR
chemical shift time scale and separate signals are observed.
Integration of the equilibrium intensities gives the population
ratio indicative of the thermodynamic stability of the carbon
4912 J. Org. Chem., Vol. 71, No. 13, 2006

Carbon Protonation of 2,4,6-Triaminopyrimidines
TABLE 2. 13C NMR Chemical Shifts in ppma
C(2)
C(4)
C(6)
C(5)
N(7)(CH3)2
N(8,9)(CH3)2
C(10)H3
C(10)(CH3)3
C(10)(CH3)3
others
1
163.1
157.7
160.6
152.4
159.2
167.4
160.9
152.5
163.0
158.5
162.0
158.9
164.5
161.6
162.2
157.7
162.0
172.3
163.7
74.8
73.7
80.2
79.3
93.5
48.9
70.7
71.7
25.3
87.2
28.2
140.1
1N(1)
2
9.0
8.2
2N(1)
3
3C(5)
4
4N(1)
4C(5)
5
5C(5)
7
32.1
36.9
31.3
27.7
36.2
36.6
161.6
154.7
37.7
38.7 br, 37.6 br
37.7
166.5
167.7
170.9
37.0
36.2
40.3
16.2
17.0
37.0
x 37.9, y 37.8
x 41.5, y 36.7 br
x 37.8, y 42.4 br
x 41.8, y 37.3
x 38.1, y 41.0
39.8
152.5
156.5
160.7
164.8
x 36.5, y 36.6
7O
159.1
160.2
137.9
x 37.5, y 37.6
8
158.5
160.6
158.0
104.1
69.7
90.4
36.6
37.6
36.3
8C(5)
9major
158.2
163.8
x 39.1, y 39.3
40.0
COCH3, 168.1;
COCH3, 22.2
COCH3, 172.4;
COCH3, 19.5
COCH3, 168.7;
COCH3, 21.9
Pyr-CH2, 25.0;
Pyr-NCH2, 45.7
Pyr-CH2, 24.4;
Pyr-NCH2, 47.5
SCH3, 12.8
9minor
9C(5)
10
158.1
161.3
157.1
162.3
163.7
165.8
161.6
171.3
92.4
37.2
92.8
49.1
36.3
37.3
40.0
x 38.3, y 38.3
32.0
36.5
31.3
27.6
10C(5)
11
11N(1)
11C(5)
163.7
157.8
189.1
161.1
157.0 br
172.9
97.2
95.7
49.9
32.2
32.3
38.1
30.5
29.6
27.5
SCH3, 12.8
SCH3, 13.9
a br ) broad. x,y denotes peaks due to chemically nonequivalent Me groups within a given NMe2 group.
TABLE 3. 15N NMR Chemical Shifts in ppma
population ratio can be fine tuned with the appropriate sub-
stituent. The stronger and weaker electron-donating pyrrolidine
and SMe groups at C(2) in 10 and 11 can increase or decrease
the relative amount of the C(5) protonated form relative to 3,
indicating a strong electronic substituent effect. It is interesting
to note that 7 protonates fully at the nitroso oxygen. We
observed that the HBF4 salt of 8 was not stable in DMSO-d6.
In a freshly prepared solution, the presence of 8C(5) (7%) could
be detected together with the base form 8. The solution slowly
transforms to the HBF4 salt of 4, which can be rationalized by
elimination of NO2+ from the intermediate 8C(5). Because NO2
is a good leaving group, the elimination is favored in this case,
and the σ-complex was of only transitory existence.
N(1)
N(3)
N(7)
N(8,9)
others
1
-190
-305 -307
-305 -302
-306 -306
-303 -297
-308 -300
-268 -257
-317 -319
1N(1)
2
-213
-188
-228
-182
-184
-193
2N(1)
3
3C(5)
4
4N(1)
4C(5)
5
5C(5)
7
7O
8
8C(5)
9major
9minor
9C(5)
10
10C(5)
11
-299 -198 -307 -313
nd
-276 -268
-317 -324
-277 -268
-178
-195
-192 -189 -294 -279, -289 NO, nd
-182 -183 -273 -262, -271 NOH, 23
-185
-193
-182
-183
-191
-183
-185
-155
nd
-296 -280
-271 -260
-317 -322
-315 -321
-275 -268
-295 -300
-251 -258
-296
NO2, -27
NO2, -16
NHCOCH3, -263
NHCOCH3, -262
NHCOCH3, -265
The C(5) protonated forms give characteristic NMR spectra.
The H(5) proton appears as a sharp singlet for 3C(5), 4C(5),
8C(5), 10C(5), and 11C(5), whereas it shows up as a quartet
for 5C(5) (3J ) 7.4 Hz) and as a doublet for 9C(5) (3J ) 9.2
Hz) due to the C(5)H-CH3 and C(5)H-NHCOCH3 vicinal
couplings, respectively. The C(5) carbon experiences a sub-
stantial upfield shift (e.g., 45 and 59 ppm for 4 and 5) relative
to the base form as a result of the sp2 to sp3 hybridization change
(Table 2). 13C and 15N chemical shift changes indicate effective
delocalization of the positive charge on the σ-complex frame:
C(2), C(4), and C(6) show characteristic downfield shifts in line
11N(1)
11C(5)
-288
-243
-168
a nd ) not detected.
and the nitrogen protonated species (Table 7, see below).
Reaching thermal equilibrium was generally fast (from a few
minutes to hours at 30 °C), and it was further checked after
several weeks. In the case of 3, we observed a somewhat slower
process. The first major species 3N(1) (kinetically controlled
form) was gradually transformed to the thermodynamically more
stable 3C(5) form, attaining the equilibrium after a few weeks
at room temperature.
It has been found that 5, 9, and 10 are fully C(5) protonated,
that 1 and 2 are fully N(1) protonated, and that 3, 4, and 11
exhibit equilibrium between both forms under the applied
experimental conditions (Table 7, see below). The C(5)H/N(1)H
with their positively charged character (Table 2). Similarly, 15
N
chemical shifts of N(7) and N(8,9) experience substantial
downfield shifts, and N(1,3) shows a small, but characteristic,
upfield shift (Table 3) with respect to the base forms. The
observed chemical shift changes are in line with the calculated
NPA charges and the alternating charge distribution (see below).
It is interesting to note that in the C(5) protonated form the
substituents at C(5) assume a quasiaxial orientation (Figure 1b),
whereas in the base form they are aligned with the pyrimidine
ring plane. This geometrical change gives rise to the observed
J. Org. Chem, Vol. 71, No. 13, 2006 4913

Ne´meth et. al
(CH3)2 pairs. The slower exchange in 4N(1) can be attributed
to the steric hindrance of the N(1) and N(3) positions due to
the bulky (dimethyl)amino groups.
It is also interesting to discuss the oxygen protonation of 7.
In the base, conjugation of the nitroso group with the pyrimidine
system is strong resulting in the partial double bond character
of the C(5)-NO bond. This shows up in the relatively large
(ca. 140 ppm) chemical shift of C(5) relative to 4. Upon
protonation at the nitroso oxygen, the resulting system can be
described with the C(5)dN-OH oxime-type bond and the
delocalization of the positive charge in the pyrimidine ring
similar to that observed for the C(5) protonated forms. Similar
to the base form (7), 7O is asymmetrical with respect to the
C(5)-C(2) axis, and rotation of the (dimethyl)amino groups is
a slow process as indicated by the observed chemical shift
nonequivalence of the methyl signals.
FIGURE 1. Optimized geometry of (a) 5N(1) and (b) 5C(5) using
B3LYP/cc-pVDZ.
CHART 3
Computational Results. All the calculated geometrical data
are summarized in the Supporting Information. Geometry
optimization of the base forms using the B3LYP/cc-pVDZ
method showed that fully planar rings with coplanar amino or
(dimethyl)amino groups do not present local minima on the
potential energy surface (PES), and the exocyclic nitrogens are
slightly pyramidal. The pyramidal nitrogens allow various
conformers with different orientation of the (dimethyl)amino
groups. However, it was found that the PES is shallow and the
calculated conformers are energetically very close to each other.
Therefore, in our study, we selected and compared the same
conformer of all the compounds. In this conformer, the lone
pairs of exocyclic 8 and 9 nitrogens are on the same side of the
ring plane. The nearly planar arrangement of the (dimethyl)-
amino groups indicates substantial conjugation with the pyri-
midine ring. In the case of 9, the four possible conformers were
also studied. The steric effect of the C(5) substituents hampers
full delocalization, hence it causes a slight shift of the C(5) atom
out of the ring plane and a small twist of the (dimethyl)amino
groups at C(4) and C(6) with respect to the ring plane (Figure
1). The steric effect increases in the order of 1 < 2 < 3 and 4
< 5 < 6 and is stronger in the (dimethyl)amino series (4-6).
While in the N(1) protonated derivatives, the geometries of
the (dimethyl)amino groups are similar to the base forms (Figure
1a), protonation at C(5) results in a planar structure of the frame,
with the exception of the C(5) atom (Figure 1b). In the C(5)
protonated forms, conjugation of the C(4,6) (dimethyl)amino
group is stronger compared to the base forms which could
explain the observed stability of the σ-complex.
The differences in geometry and also natural population
analysis (NPA)12 charges between gas-phase and solvated-phase
structures are negligible (see Supporting Information). As
expected, the ring and exocyclic nitrogen atoms are highly
negative. However, it was found that the charge on C(5) is also
strongly negative in all bases which can be explained by the
electron-donating effects of different substituents. Earlier cal-
culations suggested that the negative charge on C(5) in the
pyrimidine ring undergoes monotonic increases while connecting
more amino substituents.13 We found that the charge in
substituted 2,4,6-triaminopyrimidines is competitive with that
of N(1,3) which rationalizes the protonation at C(5). In our
study, protonation was theoretically examined at each of the
six negative centers of the framework (Table 4).
upfield shift of the 1H signals due to the Me and tBu groups in
3, 5, 10, and 11 (Table 1) relative to the respective base because
these protons experience the anisotropic shielding and deshield-
ing effects of the pyrimidine ring in the C(5) protonated and
base forms, respectively. In line with our previous observations,7
rotation of the N(8,9) (dimethyl)amino groups in the studied
σ-complexes is slow on the NMR chemical shift time scale and
1
separate H and 13C signals are detected. The assignment of
the signals due to the chemically nonequivalent H atoms or Me
groups within NH2 or NMe2 groups is based on the measured
C(5)H-N(8,9)Mex NOE. Accordingly, the upfield Hx atom or
Mex group is close in space to C(5)H (see Tables 1 and 2). The
double bond character indicates that conjugation of the (dim-
ethyl)amino groups stabilizes the positively charged σ-complex.
Carbon protonation of 9 gives rise to four possible conformers
as shown in Chart 3.
However, only one signal set was observed indicating either
slow exchange with a single predominant conformation or fast
equilibrium between these forms on the NMR chemical shift
time scale. The observed strong NH-COCH3 NOE and the lack
of dipolar interaction between C(5)H and COCH3 point to a
trans amide bond (form II and IV). On the basis of the measured
relatively large C(5)H-NHCOCH3 vicinal coupling (3J ) 9.2
Hz) and the observed strong C(5)H and NHCOCH3 NOE, we
conclude that under the given conditions form IV is the
predominant conformer with a coplanar arrangement of the C(5)-
H and NHCOCH3 protons according to the calculated geometry
of conformer IV (see Supporting Information).
The N(1) protonated form is characteristic of the observed
upfield shifts of C(2), C(4,6), and N(1,3) and the downfield
shift of N(7) and N(8,9) as well as of N(7)H2/(CH3)2 and N(8,9)-
H2/(CH3)2 relative to the base (Tables 1-3). In the case of 1N-
(1), 2N(1) and 11N(1), the mutual exchange between the N(1)
and N(3) protonated forms is fast on the NMR chemical shift
time scale and apparent spectral symmetry with respect to the
C(5)-C(2) axis is observed due to signal averaging. In the case
of 1N(1), the N(1)H signal is broadened to the extent that it
escapes detection; on the other hand, the N(1)H signal of 2N-
(1) and 11N(1) are observable as a broad singlet signal. For
4N(1), however, the exchange is slow on the 13C and 15N
chemical shift time scale and separate signals are observed for
the N(1) and N(3), C(4) and C(6), and N(7)(CH3)2 and N(8)-
(12) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,
83, 735-746.
(13) Barfield, M.; Fagerness, P. J. Am. Chem. Soc. 1997, 119, 8699-
8711.
4914 J. Org. Chem., Vol. 71, No. 13, 2006

Carbon Protonation of 2,4,6-Triaminopyrimidines
TABLE 4. NPA Charges of 1 and Its Protonated Derivativesa
TABLE 5. Calculated Proton Affinitiesa
1
1C(5)
1N(1)
1N(7)
1N(8)
C(5)
N(1)
N(7)
N(8)
O
N(1)
C(2)
N(3)
C(4)
C(5)
C(6)
N(7)
N(8)
N(9)
H(10)
H+
-0.657
0.691
-0.657
0.487
-0.459
0.487
-0.821
-0.818
-0.818
0.222
-0.630
0.730
-0.630
0.578
-0.586
0.578
-0.724
-0.752
-0.752
0.292
-0.636
0.736
-0.615
0.517
-0.427
0.509
-0.796
-0.758
-0.794
0.259
-0.618
0.649
-0.619
0.491
-0.404
0.490
-0.771
-0.780
-0.780
0.251
-0.606
0.701
-0.632
0.419
-0.424
0.515
-0.781
-0.764
-0.776
0.252
1
2
3
4
5
6
7
8
9
990.1
988.0
1005.2
1010.0
1012.1
1034.6
1030.4
1033.1
993.2
928.1
934.3
936.2
985.4
976.5
975.0
940.7
933.4
910.2
917.8
927.5
966.9
977.2
992.6
948.2
942.5
1033.6
1034.4
1058.0
1110.6
982.1
1006.4
1056.4
1056.7
996.1
1048.4
989.0
965.9
983.6
1025.0
1031.1
1007.2
10
11
980.8
944.5
910.6
0.292
0.431
0.470
0.462
a NPA charges were calculated at the B3LYP/cc-pVDZ level in the gas
phase.
a The B3LYP/cc-pVDZ method was used. The values are in kJ/mol and
belong to the most stable studied conformers (see Supporting Information).
The PA of 7N(10) is 1011.4 kJ/mol.
centers in the frame are in nitrogen atoms. In all molecules, the
MEP values indicate more negative nitrogens than C(5) which
suggests that N(1) protonation is favorable under kinetic control.
As a conclusion, when C(5) protonation is favored, it should
be under thermodynamical control. This phenomenon was
clearly observed by NMR spectroscopy and discussed in the
NMR section. Note that the MEP values in DMSO are normally
smaller than those in the gas phase, although the tendency of
the charge distribution is the same in both phases.
The gas-phase proton affinity (PA) has been calculated for
the following reaction:17
B + H+ a BH+
FIGURE 2. Molecular electrostatic potential of (a) 1 and (b)
pyrimidine. The MEP values are plotted on a 0.01 electron density
value, calculated by the polarized continuum model (PCM; DMSO
solution) with B3LYP/cc-pVDZ.
where B is the base form and BH+ is the protonated form. The
PAs were calculated using eq 1:
In the carbon protonated forms, the charge on C(5) decreases
which confirms the appearance of a new delocalization in these
structures because the positive charge is shared in the molecular
frame. This can be seen on the charges of the exocyclic nitrogen
atoms; hence their growth is around 0.07-0.1. On the other
hand, if a nitrogen atom is protonated, the conjugation of the
positive charge is less effective. This can also be obtained from
the calculated nucleus-independent chemical shifts (NICS)14 (see
Supporting Information). In comparison, for pyrimidine, all base
forms which have an electron-donating substituent at C(5) are
less aromatic (-7.02 (pyrimidine) changes between -4 and -6)
and the aromaticity decreases with the N(1) protonation.
Electron-withdrawing substituents in 7 and 8 highly decrease
NICS values to -2.22 and -3.36, respectively. All carbon
protonated forms and 7O have positive NICS indices which
indicate that the aromatic system indeed ceases.
The charge distribution was also investigated with molecular
electrostatic potential (MEP),15 mapped on an isoelectron density
surface for all the studied molecules. Figure 2 shows the results
for 1 and, as a comparison, for pyrimidine. All the other MEPs
are in the Supporting Information.
A previous study showed that protonation at the C(5) position
is not predictable by the MEP of nonsubstituted pyrimidine.16
Our calculations show that in the substituted derivatives C(5)
has a strongly negative MEP value and all the other nucleophilic
5
2
PA ) -∆H ) ∆Etot + ∆ZPE + RT
(1)
where ∆H is the enthalpy change of the protonation reaction,
Etot is the electron energy at 0 K, ZPE is the zero point
correction, R is the molar gas constant, and T is the absolute
temperature in kelvin.
The results show that in the studied molecules the PA of C(5)
and N(1) is commensurable, and the protonation of exocyclic
nitrogens is not favored (Table 5).
1N(1) is the most stable of the protonated forms of 1. The
electron-donating effect of the methyl group slightly raises the
PA at N(1), N(7), and N(8) but not at C(5); therefore, in the
case of 2, carbon protonation is not expected. In 3, because of
the bulky tert-butyl group, the steric effect of the C(5)
substituent can appear as a relevant factor and causes the highest
PA at C(5). In the cases of 4-6, the tendency is almost the
same as that in 1-3, but changing the amino groups to
(dimethyl)amino groups raises the PAs and the relevance of
the steric effects. By comparing 7 and 8 with other studied
molecules, it can be seen that the electron-withdrawing groups
decrease the PA at all positions. In both cases, oxygen
protonation was also studied and we found that in 7 this site
has the highest PA. It can be concluded that C(5) protonation
is favorable with large and electron-donating substituents.
Therefore, it is expected that the bulky acetamino substituent
causes high PA at the C(5) position in 9. The PA at the N(1)/
C(5) position can be fine tuned not only by C(5) but also by
(14) Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.; Jiao, H.; Hommes,
N. J. R. v. E. J. Am. Chem. Soc. 1996, 118, 6317-6318.
(15) Na´ray-Szabo´, G.; Ferenczy, Gy. G. Chem. ReV. 1995, 95, 829-
847.
(16) Shukla, M. K.; Mishra, P. C. J. Chem. Inf. Comput. Sci. 1998, 38,
678-684.
(17) Szulejko, J. E.; McMahon, T. B. J. Am. Chem. Soc. 1993, 115,
7839-7848.
J. Org. Chem, Vol. 71, No. 13, 2006 4915

Ne´meth et. al
TABLE 6. Calculated Relative Gibbs Free Energies (in DMSO)
and Relative PAsa
calculated Boltzmann distribution is very sensitive to slight
energy differences. Considering the inherent calculation errors,
the applied model reproduces the experimental protonation well.
If, however, there are two energetically close forms, a precise
population ratio cannot be predicted.
∆Gb
∆PAc
C(5) N(1) N(7) N(8)
O
C(5) N(1) N(7) N(8)
O
1
2
3
4
5
6
7
8
9
18.7
31.3
0.0
0.0 74.4 84.6
0.0 77.6 88.0
2.4 79.1 96.0
1.9 62.5 70.4
15.1
22.0
0.0 77.1 95.0
0.0 75.6 92.2
0.0 21.5 97.4 106.1
0.2 0.0 49.2 67.7
0.0 27.7 81.5 80.8
0.0 77.6 135.7 118.0
Conclusions
0.0
0.0 27.6 85.3 69.0
0.0 64.2 123.6 107.2
54.0 36.0 87.9 82.3 0.0 66.3 55.3 107.7 100.3 0.0
0.0 13.2 56.7 54.8 22.4 0.0 22.8 73.0 63.9 17.4
The protonation of a series of different C(5) substituted 2,4,6-
triaminopyrimidine derivatives was investigated. For this pur-
pose, we synthesized several new 2,4-diamino- and 2,4,6-
triaminopyrimidines and their respective HBF4 salts. The site
0.0 24.7
0.0 31.4
90.5
1
of the protonation was studied experimentally by H, 13C, and
10 0.0 19.5 83.3 90.4
11 17.0 0.0 77.6
0.0 25.6 75.9 112.3
15N NMR spectroscopy and theoretically by using PCM B3LYP/
cc-pVDZ quantum chemical methods. Generally, protonation
sites were found at the C(5) and N(1) positions. The C(5)H/
N(1)H population ratio strongly depends on the substituents
present. The results of the quantum chemical calculations are
in accord with the experimental observations and prove that
combined steric and electronic effects are responsible for the
observed C(5) protonation and for σ-complex stability. The steric
clash between the nearly coplanar substituents at C(5) and at
C(4,6) in the base is partly relieved in the σ-complex, where
the C(5) substituent assumes quasiaxial orientation perpendicular
to the ring plane. This gives rise to stronger conjugation of the
(dimethyl)amino substituents at C(4,6) relative to the base,
leading to electronic stabilization of the σ-complex through an
effective charge delocalization. Our results prove that C(5)
protonation is a general feature of the 2,4,6-triaminopyrimidine
system. The observed carbon protonation receives attention in
connection with structure-activity relationship studies of
pharmaceutically active compounds containing the 2,4,6-tri-
aminopyrimidine moiety.
11.0
0.0
96.6
a Values are in kJ/mol and belong to the most stable studied conformers
(see Supporting Information). b The PCM method was used at the B3LYP/
cc-pVDZ level. c The B3LYP/cc-pVDZ level in the gas phase.
TABLE 7. Experimental and Calculated Boltzmann Distributions
calculated distributiona
measured distributionb
C(5)
N(1)
O
C(5)
N(1)
O
1
2
3
4
5
6
7c
8
9
0
0
71
100
100
29
32
0
0
0
0
0
nd
nd
96
10
100
100
100
4
90
nd
68
100
100
0
100
100
100
0
100
0
nd
d
100
100
9
nd
nd
nd
nd
91
100
10
11
0
100
a Values are in %. The PCM method was used at the B3LYP/cc-pVDZ
level. The values belong to the most stable studied conformers (see
Supporting Information). b Values measured by NMR in DMSO-d6 with
integration of the respective forms. nd ) not detected. d ) detected. c The
calculated Boltzmann distribution of 7N(10) is 0.0%.
Experimental Section
The general spectroscopic and computational methods are given
in the Supporting Information.
C(2) substitution, as 3, 10, and 11 show. The electron donation
of C(2) substituents raises the PA in the order of 11 < 3 < 10.
This effect is larger at C(5) than at N(1); as a consequence, in
10, the C(5) protonation overrides N(1). The high PA at C(5)
in the crowded C(5) substituted molecules suggests that the
stability effect of new delocalization and the termination of the
steric strain are comparable with the stabilization effect of the
aromatic system.
Synthetic Methods. 5-tert-Butyl-2,4,6-triaminopyrimidine (3).
Sodium tert-butylate (0.42 g, 4.37 mmol) was dissolved in abs
ethanol (2 mL). Guanidine hydrochloride (0.36 g, 3.77 mmol) and
tert-butylmalonitrile10c (0.5 g, 4.1 mmol) were added, and the
reaction mixture was stirred for 2 h at 90 °C under reflux. The
solvent was evaporated in a vacuum. The residue was stirred with
water (3 mL), and the resultant solid was filtered off, washed with
water (2 × 1 mL), stirred with diethyl ether (5 mL), and filtered
off again to give 3 as a yellowish solid. Yield 67%. Mp 232-233
°C. IR (KBr, cm-1): ν ) 3518, 3467, 3393, 3349, 3177, 2956,
1662, 1635, 1565, 1406, 1037, 804. HRMS m/z calcd for C8H15N5,
181.1322; found, 181.1319.
5-Methyl-2,4,6-tris(dimethylamino)pyrimidine (5). Iodomethane
(1.55 mL, 25 mmol) was added to a solution of 2,4,6-tris-
(dimethylamino)pyrimidine (1.04 g, 5 mmol) in dichloromethane
(5 mL). The solution was refluxed for 120 h at 40 °C, and then the
solvent was removed in a vacuum to give a crude product, which
contained two main components (TLC/ethyl acetate/MeOH/25%
NH3 solution in water ) 10:0.1:0.01). The two components of the
crude product were separated by silica gel gradient flash chroma-
tography (eluent A: ethyl acetate/MeOH/25% NH3 solution in water
) 10:0.1:0.01; eluent B: ethyl acetate/MeOH/25% NH3 solution
in water ) 10:3:0.1) to give 475 and 878 mg of product,
respectively. These products were dissolved in dichloromethane (10
mL), and the solutions were extracted with 20 m/m% sodium
carbonate solution in water (2 × 10 mL) and dried over anhydrous
sodium sulfate. The drying agent was filtered off, and the solvents
were evaporated in a vacuum to give 284 and 521 mg of methylated
Considering the Boltzmann distribution, the population ratio
of the different protonated forms can be calculated as follows:
+
N(BH1 )
∆G12
) exp -
(2)
[
]
+
RT
N(BH2 )
where N(BH1+) and N(BH2+) are the populations of the two
protonated forms and ∆G12 is the Gibbs free energy difference.
∆G12 was calculated using the PCM method which includes
solvation effects. The calculated distribution can be directly
compared with the protonation distribution measured by NMR.
In Table 6 the relative energy values in the gas phase and in
DMSO solution are compared. The correlation between ∆G in
DMSO and ∆PA is linear with 0.96 R2 (Chart SI 1, see
Supporting Information), which shows that the gas-phase
calculations simulate the protonation site well.
The experimental and calculated Boltzmann distributions are
given in Table 7. Because of the exponential function, the
4916 J. Org. Chem., Vol. 71, No. 13, 2006

Carbon Protonation of 2,4,6-Triaminopyrimidines
pyrimidine derivative, respectively. The major product (N(7)-
methylated derivative) was obtained as a semisolid. It proved to
be unstable, and it was not characterized in detail. Its structure was
confirmed by NMR (see Supporting Information). On the basis of
NMR, the minor product proved to be the 5-methyl-2,4,6-tris-
(dimethylamino)pyrimidine (5). Yield: 25.3%. Mp 97-98 °C. IR
(KBr, cm-1): ν ) 3429, 2945, 2861, 1565, 1383, 1140, 1070, 1032,
795. HRMS m/z calcd for C11H21N5, 223.1791; found, 223.1797.
2,4,6-Tris(dimethylamino)-5-nitrosopyrimidine (7). 2,4,6-Tris-
(dimethylamino)pyrimidine (1.5 g, 7.17 mmol) was dissolved in a
mixture of dichloromethane (12 mL) and acetic acid (2.5 mL). A
solution of NaNO2 (0.7 g, 14 mmol) in water (5 mL) was added
dropwise at 0-5 °C over a period of 15 min, and then the resultant
mixture was stirred for 1 h at the same temperature. The reaction
mixture was diluted with water (25 mL) and dichloromethane (25
mL) followed by basifying the solution to pH 10 with sodium
carbonate. The organic layer was separated and dried over
anhydrous sodium sulfate, and the solvent was evaporated to give
7. The crude product was purified by silica gel column chroma-
tography (100 g of silica gel, CHCl3/MeOH ) 99:1) to give 1.28
g of product as a blue solid. Yield: 75%. Mp 191-192 °C. IR
(KBr, cm-1): ν ) 2924, 1510, 1555, 1398, 1247, 1071, 778. HRMS
m/z calcd for C10H18N6O, 238.1537; found, 238.1544.
5-Nitro-2,4,6-tris(dimethylamino)pyrimidine (8). 2,4,6-Tris-
(dimethylamino)pyrimidine (500 mg, 2.39 mmol) was dissolved
in dichloromethane (5.0 mL). The solution was cooled to -70 °C
under an argon atmosphere, and solid nitronium tetrafluoroborate
diethyl etherate (740 mg, 5.52 mmol) was added in five portions
over a period of 2 h. The resultant mixture was stirred at -70 °C
for 4 h, allowed to warm, and then stirred for 24 h at room
temperature. Dichloromethane (15 mL) was added, and the resultant
solution was extracted with a solution of sodium carbonate (c )
20 m/m%, 3 × 5 mL). The organic layer was dried over anhydrous
sodium sulfate and filtered, and the solvent was removed in a
vacuum to give a crude product (520 mg), which was purified by
gradient flash chromatography (40 g of silica gel; eluent A:
n-hexane; eluent B: ethyl acetate). Yield: 35%. Description: white
solid. Mp 209-210 °C. IR (KBr, cm-1): ν ) 2931, 1558, 1477,
1396, 1312, 1277, 1057, 777, 722. HRMS m/z calcd for C10H18N6O2,
254.1486; found, 254.1479.
5-Acetylamino-2,4,6-tris(dimethylamino)pyrimidine (9). 5-Ni-
troso-2,4,6-tris(dimethylamino)pyrimidine (7, 0.92 g, 3.86 mmol)
was dissolved in a mixture of tetrahydrofuran (100 mL) and acetic
anhydride (1.2 mL, 11.1 mmol). Palladium on carbon (0.2 g,
loading: 10 m/m%) was added, and hydrogen was bubbled through
the solution for 3 h. The catalyst was filtered off and washed with
methanol, and the resultant solution was concentrated in a vacuum.
The residue was stirred with ethanol, and the solid was filtered off
and washed with ethanol and diethyl ether to give 9 as a white
solid. Yield: 32%. Mp 213-214 °C. IR (KBr, cm-1): ν ) 3295,
2940, 2874, 1660, 1559, 1506, 1281, 1053, 783. MS m/z calcd for
C12H22N6O, 266.1850; found, 266.1844.
5-tert-Butyl-4,6-diamino-2-(pyrrolidin-1-yl)pyrimidine (10).
Sodium tert-butylate (1.26 g, 13 mmol) was dissolved in abs ethanol
(12 mL). tert-Butylmalonitrile10c (2.0 g, 12.3 mmol) and pyrrolidine-
1-carboxamidine hydrogen sulfate18 (1.5 g, 13 mmol) were added,
and the mixture was stirred for 8 h at 90 °C under reflux. The
solvent was removed in a vacuum, and water (15 mL) was added.
The mixture was stirred for 2 h at room temperature. The solid
was filtered off, washed with water, and recrystallized from boiling
ethanol to give 10 as a white solid. Yield: 64%. Mp 222-223 °C.
IR (KBr, cm-1): ν ) 3543, 3282, 3153, 2948, 2853, 1617, 1539,
1346, 1185, 796. HRMS m/z calcd for C12H21N5, 235.1791; found,
235.1791.
(2 mL). tert-Butylmalonitrile10c (0.5 g, 4.1 mmol) and thiocarbamide
(0.3 g, 3.95 mmol) were added, and the mixture was stirred for 5
h at 90 °C under reflux. The solvent was removed in a vacuum,
and water (2 mL) and a solution of hydrochloric acid (0.87 mL, c
) 16 m/m%, 4.1 mmol HCl) were added dropwise. The mixture
was stirred for 2 h at room temperature. The solid was filtered off
and washed with water to give 5-tert-butyl-4,6-diamino-2-thiopy-
rimidine (0.38 g, yield: 47%), which was used without further
purification for preparation of 11.
Sodium tert-butylate (0.20 g, 2.08 mmol) was dissolved in abs
ethanol (3.5 mL). 5-tert-Butyl-4,6-diamino-2-thiopyrimidine (0.35
g, 1.77 mmol) and iodomethane (0.115 mL, 1.85 mmol) were added,
and the mixture was stirred at room temperature for 2 h. The solvent
was evaporated in a vacuum. The residue was dissolved in
dichloromethane (5 mL), and the dichloromethane solution was
extracted with water (3×5 mL) and dried over anhydrous sodium
sulfate. The drying agent was filtered off, and the solvent was
removed in a vacuum to give a crude product, which was
recrystallized from ethanol. Yield: 64% (referring to 2-thiopyri-
midine derivative), 30% (referring to tert-butylmalonitrile). Descrip-
tion: white solid. Mp 180-181 °C. IR (KBr, cm-1): ν ) 3538,
3417, 3302, 3172, 2954, 2875, 1631, 1596, 1536, 1436, 1398, 1308,
956, 759. HRMS m/z calcd for C9H16N4S, 212.1090; found,
212.1079.
General Procedure for the Preparation of the HBF4 Salts of
Compounds 1-11. The aminopyrimidine derivative (2 mmol) was
dissolved in methanol (5.0 mL) at room temperature. The solution
was cooled to 0 °C under an argon atmosphere, and then a solution
of tetrafluoroboric acid diethyl etherate in diethyl ether (2.1 mL, c
) 1 M) was added dropwise over a period of 5 min. The solvent
was removed in a vacuum, and the residue was stirred with diethyl
ether (3 mL). The solid was filtered off, washed with diethyl ether,
and dried in a vacuum to give the HBF4 salt of the aminopyrimidine.
5-Methyl-2,4,6-triaminopyrimidinium Tetrafluoroborate (2‚
HBF4). Yield: 87%. Mp 245-246 °C. Description: white solid.
IR (KBr, cm-1): ν ) 3474, 3147, 3389, 3210, 2945, 1638, 1452,
1367, 1054 (BF4-), 708. HRMS m/z calcd for C5H9N5, 139.0852;
found, 139.0853.
5-tert-Butyl-2,4,6-triaminopyrimidinium Tetrafluoroborate
(3‚HBF4). Yield: 89%. Mp 209-210 °C. Description: white solid.
IR (KBr, cm-1): ν ) 3549, 3533, 3458, 3406, 3209, 2969, 1681,
1630, 1435, 1360, 1100 and 1056 (BF4-), 993, 776. HRMS m/z
calcd for C8H15N5, 181.1322; found, 181.1336.
2,4,6-Tris(dimethylamino)pyrimidinium Tetrafluoroborate
(4‚HBF4). Yield: 84%. Mp 243-245 °C. Description: white solid.
IR (KBr, cm-1): ν ) 3418, 2949, 1632, 1583, 1412, 1334, 1175,
1057 (BF4-), 766. HRMS m/z calcd for C10H19N5, 209.1635; found,
209.1621.
2,4,6-Tris(dimethylamino)-5-methylpyrimidinium Tetrafluo-
roborate (5‚HBF4). Yield: 78%. Mp 177-178 °C. Description:
white solid. IR (KBr, cm-1): ν ) 3400, 2946, 2770, 1594, 1393,
1054 (BF4-), 756. HRMS m/z calcd for C11H21N5, 223.1791; found,
223.1786.
2,4,6-Tris(dimethylamino)-5-nitrosopyrimidinium Tetrafluo-
roborate (7‚HBF4). Yield: 81%. Mp 196-197 °C. Description:
blue solid. IR (KBr, cm-1): ν ) 3324, 2941, 1649, 1578, 1372,
1075 (BF4-), 1005, 764. HRMS m/z calcd for C10H18N6O, 238.1537;
found, 238.1535.
5-Acetylamino-2,4,6-tris(dimethylamino)pyrimidinium Tet-
rafluoroborate (9‚HBF4). Yield: 92%. Mp 244-246 °C. Descrip-
tion: white solid. IR (KBr, cm-1): ν ) 3379, 2953, 1692, 1530,
1594, 1390, 1291, 1069 (BF4-), 788. HRMS m/z calcd for
C12H22N6O, 266.1850; found, 266.1851.
5-tert-Butyl-4,6-diamino-2-methylsulfanylpyrimidine (11). So-
dium tert-butylate (0.38 g, 3.96 mmol) was dissolved in abs ethanol
5-tert-Butyl-4,6-diamino-2-(pyrrolidin-1-yl)pyrimidinium Tera-
fluoroborate (10‚HBF4). Yield: 91%. Mp 241 °C. Description:
white solid. IR (KBr, cm-1): ν ) 3368, 3180, 2973, 1670, 1586,
1364, 1057 (BF4-), 999, 753. HRMS m/z calcd for C12H21N5,
235.1791; found, 235.1797.
(18) Bannard, R. A. B.; Casselman, A. A.; Cockburn, W. F.; Brown, G.
M. Can. J. Chem. 1958, 36, 1541-1549.
J. Org. Chem, Vol. 71, No. 13, 2006 4917

Ne´meth et. al
5-tert-Butyl-4,6-diamino-2-methylsulfanylpyrimidinium Tera-
fluoroborate (11‚HBF4). Yield: 82%. Mp 135-136 °C. Descrip-
tion: white solid. IR (KBr, cm-1): ν ) 3506, 3383, 2965, 2936,
1615, 1594, 1530, 1338, 1223, 1085 (BF4-), 776. HRMS m/z calcd
for C9H16N4S, 212.1090; found, 212.1072.
Supporting Information Available: Spectroscopic methods,
computational methods, additional synthetic methods, NMR and
IR spectra of the studied molecules, NPA charges, NICS indices,
MEPs, chart of the correlation between ∆G in DMSO and ∆H in
the gas phase, details of the calculation of temperature correction,
Cartesian coordinates, and total energies. This material is available
free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. The authors are grateful to Ga´bor Czira
for the HRMS measurements and to Be´la Hegedus for the IR
spectra. B.N. and T.V. acknowledge financial support from
OTKA Grant T048796.
JO0605703
4918 J. Org. Chem., Vol. 71, No. 13, 2006