N-Thio- And N-selenophenacylaniidiries: Electrophilic activation as a route to some l-hetero-3-aza-4-dimethylaminobuta-l,3-dienes

Article abstract of DOI:10.1039/a903150b

The preparation of 2-phenyl-4-dimethylamino-l-aza-, 1-oxa-, 1-thia-, l-selena-3-azabuta-l,3-dienes as well as their 4-methyl derivatives is described following a new heteroatom interchange reaction process. Heteronucleophilic attack at one particular reactive site of bis-electrophilic amidinium salts is the key feature of the process. In addition, we also disclose that the substituted l-oxa-3-aza- and l,3-diazabuta-l,3-dienes can be obtained by a reactional transformation cascade initiated by either silver acetate addition or tosyl azide [3+2] cycloaddition onto the CS or CSe double bonds of the 1-thia- and l-selena-3-azabuta-l,3-diene analogues. The Royal Society of Chemistry 1999.

Full text of DOI:10.1039/a903150b

View Article Online / Journal Homepage / Table of Contents for this issue
N-Thio- and N-selenophenacylamidines: electrophilic activation as a
route to some 1-hetero-3-aza-4-dimethylaminobuta-1,3-dienes
Gabriel T. Manh,a Franck Purseigle,a Didier Dubreuil,a Jean Paul Pradère,*a André Guingant,a
Renée Danion-Bougot,b Daniel Danionb and Loic Toupetc
a Laboratoire de Synthèse Organique, UMR au CNRS 6513, 2 rue de la Houssinière, BP 92208,
44322 Nantes Cedex 3, France
b Laboratoire de Synthèse et Electrosynthèse Organiques, UMR au CNRS 6510,
Université de Rennes I-Beaulieu, 35042 Rennes Cedex, France
c Groupe Matière Condensée et Matériaux, UMR au CNRS 6626,
Université de Rennes I-Beaulieu, 35042 Rennes Cedex, France
Received (in Cambridge, UK) 20th April 1999, Accepted 30th July 1999
The preparation of 2-phenyl-4-dimethylamino-1-aza-, 1-oxa-, 1-thia-, 1-selena-3-azabuta-1,3-dienes as well as their
4-methyl derivatives is described following a new heteroatom interchange reaction process. Heteronucleophilic
attack at one particular reactive site of bis-electrophilic amidinium salts is the key feature of the process. In addition,
we also disclose that the substituted 1-oxa-3-aza- and 1,3-diazabuta-1,3-dienes can be obtained by a reactional
transformation cascade initiated by either silver acetate addition or tosyl azide [3ϩ2] cycloaddition onto the CS
or CSe double bonds of the 1-thia- and 1-selena-3-azabuta-1,3-diene analogues.
silver acetate nucleophilic attack or 2) tosyl azide [3 ϩ 2] dipolar
cycloaddition on 1,2, are also effective for the preparation of
1-oxa-3-aza- and 1,3-diazabuta-1,3-dienes 5 and 6, respectively.
Introduction
Heterobutadienes are useful building blocks in heterocyclic
chemistry and during the last few years they have increasingly
been used in [4 ϩ 2] cycloadditions to prepare a large variety of
natural products.1 Some reports from this laboratory disclosed
that efficient hetero Diels–Alder reactions could be effected by
treatment of substituted 1-thia-3-azabuta-1,3-dienes 1 (N-thio-
phenacylamidines) and 1-selena-3-azabuta-1,3-dienes 2 (N-sel-
enophenacylamidines) with several electron-poor dienophiles2
and such a methodology was also applied to the first asym-
metric synthesis of the 4H-dihydro-1,3-thiazine skeleton.3 The
reactivity of phenacylamidines such as 1 and 2 has mainly been
exploited for their ability to successfully participate in [4 ϩ 2]
cycloadditions. However, in the literature there are few
examples of the transformation of N-thioacylamidines into
five-membered heterocycles via their corresponding amidinium
and 1,2,4-dithiadiazolium salts.4 In this paper, we wish to report
a hitherto unexplored facet of amidinium salt chemistry in
showing that 3 and 4, derived from dienes 1 and 2, respectively,
are key intermediates for the facile transformation of dienes
1(2) into dienes 2(1) as well as dienes 1,2 into dienes 5,6 follow-
ing an overall heteroatom interchange reaction process5 as
portrayed in general terms in Scheme 1. Additionally, we also
report results which establish that strategies based on either 1)
Results and discussion
As previously reported for the N-thiophenacylformamidine 1a,6
its methyl-substituted analogue 1b and the related N-seleno-
phenacylformamidine 2b react with methyl iodide at room tem-
perature to give, almost quantitatively, the hetero substituted
amidinium salts 3b and 4b, respectively (Scheme 1, E = Me).
These salts are bis-electrophiles and, as such, are prone to
nucleophilic attack at both the C-2 and C-4 positions. Semi-
empirical calculations effected at the PM3 level on amidinium
salts 3a and 4a clearly indicated that the electron deficiency is
seen to be concentrated on C-4 and that the LUMO has the
greatest coefficient located on C-4 (Scheme 2).
Scheme 2
The same conclusions are also valid for both 3b and 4b
although calculations indicated less marked differences between
the C-2 and C-4 reactive sites.7 Consequently, it appears that,
under either charge or orbital control, nucleophilic species
should be preferentially directed to carbon C-4.
Although this situation seems, a priori, unfavourable to meet
our purpose, this may in fact not be the overall controlling
factor in determining the product structure since the hetero-
atom interchange at C-2 does not necessarily imply that the
incoming nucleophile would be kinetically directed to this
site. Indeed, in the case where the incoming nucleophile is
Scheme 1
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828
This journal is © The Royal Society of Chemistry 1999
2821

View Article Online
an ambident species of general formula HY–C(R)᎐X (X,Y᎐
᎐
᎐
heteroatoms), reversibility or possible rearrangement of the
primary addition product followed by an irreversible trans-
formation toward the targeted diene induced by a nucleophilic
entity (e.g. HY–C(R)᎐X itself) may well overcome the above
᎐
problem (Scheme 3; the overall process corresponds to an
addition of the nucleophile by its X terminus at C-2 of 3).
Scheme 4
Scheme 5
Scheme 3
both cases, the accompanying by-products (benzonitrile and
methanethiol) are the same.
S→Se Heteroatom interconversion reaction process
The action of thiobenzamide on the selenoamidinium salt 4b
with the same reaction conditions proved to be less efficient
leading to N-thiophenacylamidine 1b in only 17% yield. Not
surprisingly, switching selenobenzamide for the less nucleo-
philic benzamide left 3b and 4b unchanged.
With the above considerations in mind, we first elected to study
the possibility of transforming dienes 1a,1b to dienes 2a,2b
(S→Se heteroatom interconversion). Indeed, the action of
hydroselenide on the thiomethylamidinium 3a led to the
expected formation of 2a isolated in 50% yield. A related
approach described by Liebscher, led to the formation of 2a
in 37% yield from the chloroamidinium salt precursor.8
However, the action of selenobenzamide on the thiomethyl-
amidinium salt 3b proved far more efficient, leading to
selenophenacylamidine 2b in up to 80% isolated yield. This
interconversion process may be interpreted in different ways
depending on the nature of the incoming nucleophile. In the
case where the nucleophile is NaSeH, the formation of diene 2a
is easily explained by the establishment of a rapid C-4→C-2
equilibration process followed by the loss of MeSH as shown in
Scheme 4.
The situation appears more complex when selenobenzamide
is the approaching nucleophile. Although an overall process
as depicted in Scheme 4 represents a plausible mechanistic
interpretation, provided the selenium atom is the nucleophilic
center, it was ruled out, however, on the basis of previous
investigations.4b As a result, the mechanism as shown in Scheme
3 (X = Se, Y = NH) could well explain the formation of diene
2b. β-Elimination of the thioimidate moiety in intermediate 7
rather than its [3,3] rearrangement could also help to account
for the product formation (Scheme 5). Experimental dis-
crimination between the two processes seems difficult since, in
S,Se→O and Se→S Heteroatom interconversion reaction
processes
In order to perform the S→O interconversion,9 amidinium salt
3a was first treated with a boiling aqueous ethanolic solution
for 30 min. A rather complex mixture formed from which imide
8, resulting from water addition at both electrophilic sites of 3a,
could be isolated in 42% yield. By contrast, when amidinium
salt 3a was subjected to the action of 2 equivalents of silver
acetate in dry acetonitrile, a regioselective reaction leading to
the formation of phenacylamidine 5a, as the sole isolated
product (40% yield), was observed. The formation of diene 5a
could be explained by the mechanism depicted in Scheme 3
(X = Y = O) or by a mechanism featuring a direct silver acetate
attack at C-2 to give the key intermediate 9 (Scheme 6). Though
plausible, the latter mechanism seems less probable in view of
results to be reported later in this paper which suggest that
sulfur complexation by a metal should preferentially direct the
nucleophile towards C-4.
Surprisingly, under the same reaction conditions, amidinium
salts 3b,4b were both transformed into compound 12 (mixture
of tautomer forms) in 60 and 85% yield, respectively. Distinc-
tion between each of the two forms of 12 was facilitated by
2822
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828

View Article Online
Fig. 1 ORTEP14 diagram of 12.
Scheme 6
direct comparison of their 13C NMR spectra with that of
compound 5b prepared by reaction of benzamide with di-
methylformamide dimethyl acetal (see Experimental section).
Additionally, an X-ray analysis of compound 12 revealed that,
in the solid state, it exists in the enamino ketone form (Fig. 1).
The formation of 12, resulting from both a heteroatom inter-
change (S,Se→O) and an acylation reaction at the methyl
substituent of the starting amidinium salts, can be accounted
for by a reaction mechanism featuring a) formation of the
O-acylamidinium salt intermediate 10; b) acetate-induced
deprotonation of the amidinium moiety leading to intermedi-
ate 11 and c) intramolecular capture of the acetyl residue by
the enamine moiety previously generated (Scheme 7).
Scheme 8
Scheme 9
the electrophilic properties of amidinium salts as reported
above to design a new and efficient preparation of these syn-
thetically useful dienes and, toward this end, we first explored
the behaviour of toluene-p-sulfonamide as the incoming
nucleophile. Its reaction with amidinium salt 3a in acetonitrile
containing triethylamine (2 equivalents) led to a mixture of
three compounds identified as being the N-tosylamidine 13
(38%), the dithioacetal derivative 14 (40%), and the 1,3-
diazabuta-1,3-diene 6a (18%). Although the marginal form-
ation of diene 6a was rather disappointing, this reaction is not
without interest from a mechanistic point of view. Thus, if it
were not possible to choose between the two reaction paths that
can operate in the reaction between an ambident nucleophile
and an amidinium salt such as 3, the formation of 14 clearly
shows that a non-ambident nucleophile, i.e. toluene-p-sulfon-
amide, was selectively directed toward the electrophilic C-4 site
of 3a (Scheme 10). Conversely, the small amount of diene 6a
probably results from attack at the electrophilic C-2 site of 3a.
Quite interestingly, enhancement in the site selectivity of the
attack at C-4 could be achieved when a metal with strong sulfur
affinity was added to the reaction medium. Thus, by the simple
action of adding mercury() chloride to the reaction mixture,
Scheme 7
The nature of both the metal counter ion and the heteroatom
in the incoming nucleophile greatly influenced the course and
efficiency of this transformation. Thus, switching silver acetate
for potassium acetate resulted in the formation of 12 in low
yield (17%) whereas the action of potassium thioacetate led to
the phenacylthioformamidine 1b in 57% isolated yield (Se→S
heteroatom interchange, Scheme 8).
Parenthetically, it is worth noting that the S,Se→O hetero-
atom interchange may also be effected by treating the N-thio-
and N-selenophenacylamidines 1b,2b with silver acetate in dry
acetonitrile. In that case, the ability of the silver ion to coordin-
ate with the thioamide10 or selenoamide function is sufficient to
promote and drive the reaction selectively toward the formation
of 12 in 50 and 80% yield, respectively (Scheme 9).
S,Se→N Heteroatom interconversion reaction process
In the last few years, effort has been devoted toward the syn-
thesis of 1,3-diazabuta-1,3-dienes which are suitable precursors
for nitrogen-containing heterocycles.1,11 We have thought to use
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828
2823

View Article Online
Fig. 2 ORTEP14 diagram of 6a.
Scheme 10
the formation of diene 6a could be entirely suppressed, the
reaction leading to the thioacetal derivative 14 (32%).
The same reaction with amidinium salt 3b and toluene-p-
sulfonamide followed a different course leading, by preferen-
tial attack at C-2, to 1,3-diazabuta-1,3-diene 6b isolated in low
yield (20%) along with thioimidate 15 (15%). The presence of
a cumbersome methyl substituent at C-4 may discourage the
toluene-p-sulfonamide from approaching this centre or, if not,
it favours the equilibration process (Scheme 11).
Fig. 3 ORTEP14 diagram of 6b.
Scheme 12
recognised that, similarly to the results of a computational
study conducted at the PM3 level, both dienes possess 1Z,3E
configurations and adopt an s-trans conformation with signifi-
cant deviations from planarity.13
Scheme 11
Given the previous results that favoured the formation of the
thioacetal derivative 13 at the expense of 6a, a different strategy
had to be devised to obtain the latter compound. From hetero-
dienes 1a and 1b we anticipated that a cascade reaction pathway
featuring a tosyl azide [3 ϩ 2] cycloaddition, preferentially at
the CS (CSe) double bond,12 followed by ring contraction and
sulfur (selenium) extrusion processes could finally lead to 1,3-
diazadienes 6a,6b as shown in Scheme 12. Indeed, mixing
heterodiene 1a with tosyl azide (5 equivalents) in boiling toluene
for 4 hours, led to the expected formation of 6a in a satisfactory
yield of 75%. Heterodienes 1b,2b behaved the same way leading
to 1,3-diazadiene 6b in 87 and 94% yields, respectively.
Dienes 1 and 2 have already been reported to engage in
[4 ϩ 2] cycloaddition reactions,2,3 but the synthetic potential-
ities of dienes 5 and 6 have not yet been delineated. Preliminary
studies have revealed that, under experimental conditions that
permitted the easy condensation of the parent heterodienes 1a
and 1b with methyl vinyl ketone, dienes 5,6 were reluctant to
undergo cycloadditions. This is not really surprising in view of
the fact that 1) the s-trans conformation15 would inhibit the
[4 ϩ 2] cycloaddition process, 2) the known examples of cyclo-
condensation reactions involving 4-amino-1,3-diazabutadienes
worked satisfactorily with ketene-derivatives16 and a stepwise
mechanism was put forward to explain the results17 and 3)
dienes 5 and 6 are predicted to be far less amenable to cyclo-
The structures of dienes 6a and 6b were obtained by single
crystal X-ray analysis as depicted in Figs. 2 and 3. It can be
2824
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828

View Article Online
Atomic scattering factors were issued from International
Tables for X-Ray Crystallography.23 ORTEP views were realized
with PLATON98.24 All the calculations were performed on a
Silicon Graphics Indy computer.
Compound 6b. C18H21N3O2S, Mr = 343.44, orthorhombic,
P212121, a = 8.032(3), b = 13.044(2), c = 17.604(4) Å, V =
1844.4(9) ÅϪ3, Z = 4, Dx = 1.237 Mg mϪ3, λ(Mo-Kα) =
0.71073 Å, µ = 1.90 cmϪ1, F(000) = 728, T = 293 K. The sample
(0.30 × 0.30 × 0.45 mm) is studied on an automatic diffrac-
tometer CAD4 NONIUS with graphite monochromatized
Mo-Kα radiation.19 The cell parameters are obtained by fitting
a set of 25 high-theta reflections. The data collection (2θmax
=
50Њ, scan ω/2θ = 1, tmax = 60 s, range hkl: h 0.8 k 0.14 l 0.18,
intensity controls without appreciable decay (0.3%) gives 1840
reflections from which 1324 were independent with I > 2.0σ(I ).
After Lorentz and polarization corrections,20 the structure was
solved with SIR-9721 which reveals the non-hydrogen atoms of
the structure. After anisotropic refinement, all the hydrogen
atoms are found with a Fourier Difference. The whole structure
was refined with SHELXL9722 by the full-matrix least-square
techniques (use of F magnitude; x y, z, βij for S, C, O and N
atoms, H atoms in riding mode; 218 variables and 1324 obser-
Fig. 4 Energy levels of FMO for 1b, 2b, 5b, 6b.
addition than their S and Se congeners 1,2 by consideration of
frontier orbital levels (Fig. 4).
Conclusion
2
vations; calc w = 1/[σ2(Fo ) ϩ (0.0668P)2 ϩ 0.5051P] where
We have shown that heteroatom interchange between 1-thia-
and 1-selena-3-azabutadienes on the one hand (1→2) and
between 1-thia (selena) and 1-oxa-3-aza- or 1,3-diaza-buta-
dienes on the other hand (1,2→5 and 1,2→6) could be easily
executed via the transitory formation of amidinium salts 3 and
4. Although the latter examples embody two electrophilic sites,
conditions that allow the regioselective direction of nucleo-
philic species were found. Additionally, we have also shown
that, in cases where the precedented strategy could not be satis-
factorily applied, the potential properties of the CS or CSe
double bonds as dipolarophiles18 in dienes 1,2 could be advan-
tageously employed to reach the targeted dienes. In particular,
we have found that tosyl azide [3 ϩ 2] cycloaddition at the CS
(CSe) double bond of dienes 1,2 initiates a sequential reaction
process that finally provides 1,3-diazadienes 6a,6b in practically
useful yields. Further studies aimed at identifying the chemical
reactivity of dienes 5 and 6 are currently under way.
P = (Fo2 ϩ 2Fc2)/3 with the resulting R = 0.033, Rw = 0.1070 and
Sw = 1.18 (residual ∆ρ ≤ 0.18 e ÅϪ3).
Atomic scattering factors were issued from International
Tables for X-Ray Crystallography.23 ORTEP views were realized
with PLATON98.24 All the calculations were performed on a
Silicon Graphics Indy computer.
Compound 12. C13H16N2O2ؒH2O, Mr = 250.29, monoclinic,
P21/n, a = 12.638(7), b = 7.998(2), c = 12.968(1) Å, β = 93.15(1)Њ,
V = 1308.8(8) ÅϪ3, Z = 4, Dx = 1.270 Mg mϪ3, λ(Mo-Kα) =
0.71073 Å, µ = 0.91 cmϪ1, F(000) = 536, T = 293 K. The sample
(0.42 × 0.35 × 0.33 mm) is studied on automatic diffractometer
CAD4 NONIUS with graphite monochromatized Mo-Kα
radiation.19 The cell parameters are obtained by fitting a set
of 25 high-theta reflections. The data collection (2θmax = 54Њ,
scan ω/2θ = 1, tmax = 60 s, range hkl: h 0.16 k 0.8 l Ϫ16.16,
intensity controls without appreciable decay (1.1%) gives 2770
reflections from which 2651 were independent (2143 with
I > 1.5σ(I )). After Lorentz and polarization corrections,20 the
structure was solved with SIR-9721 which reveals the non-
hydrogen atoms of the structure. After anisotropic refinement,
all the hydrogen atoms are found with a Fourier Difference. The
whole structure was refined with SHELXL 9722 by the full-
matrix least-square techniques (use of F magnitude; x, y, z,
βij for C, O and N atoms, x, y, z for H atoms; 213 variables
and 2651 observations; ω = 1/[σ2(Fo)2 ϩ (0.0889P)2 ϩ 1.2514P]
where P = (Fo2 ϩ 2Fc2)/3 with the resulting R = 0.061, Rω =
0.166 and Sω = 1.043 (residual ∆ρ ≤ 0.76 e ÅϪ3).
Experimental
X-Ray analysis†
Compound 6a. C17H19N3O2S, Mr = 329.4, monoclinic, P21/n,
a = 13.330(2), b = 8.404(2), c = 15.100(2) Å, β = 100.20(1)Њ,
V = 1664.8(5) ÅϪ3, Z = 4, Dx = 1.314 Mg mϪ3, λ(Mo-Kα) =
0.71073Å, µ = 2.07 cmϪ1, F(000) = 696, T = 293 K. The sample
(0.50 × 0.35 × 0.20 mm) is studied on an automatic diffrac-
tometer CAD4 NONIUS with graphite monochromatized
Mo-Kα radiation.19 The cell parameters are obtained by fitting
a set of 25 high-theta reflections. The data collection (2θmax
=
Atomic scattering factors were issued from International
Tables for X-Ray Crystallography.23 ORTEP views were realized
with PLATON98.24 All the calculations were performed on a
Silicon Graphics Indy computer.
50Њ, scan ω/2θ = 1, tmax = 60 s, range hkl: h 0.15 k 0.9 l Ϫ17.17,
intensity controls without appreciable decay (0.2%) gives
3041 reflections of which 2909 were independent (2031 with
I > 2.0σ(I )). After Lorentz and polarization corrections,20
the structure was solved with SIR-9721 which reveals the non-
hydrogen atoms of the structure. After anisotropic refinement,
all the hydrogen atoms are found with a Fourier Difference.
The whole structure was refined with SHELXL9722 by the
full-matrix least-square techniques (use of F magnitude; x, y,
z, βij for S, C, O and N atoms, x, y, z in riding mode for methyl
H atoms, x, y, z for the other H atoms; 239 variables and
General procedures
NMR spectra (1H at 200 MHz and 13C at 50.3 MHz) were
recorded with a Bruker AC 200 spectrometer. Chemical shifts
(δ) are expressed in parts per million (ppm) referenced to
residual chloroform (7.27 ppm). Coupling constants, J, are
given in hertz (Hz). Multiplicities are recorded as s (singlet), d
(doublet), t (triplet), q (quartet), and m (multiplet). Mass
spectra m/z (% base peak) were recorded on an HP 5889A spec-
trometer EI (70 eV). Infra-red spectra were carried out on a
Bruker IFS 45WHR Fourier transform IR spectrophotometer.
Melting points were determined on a C. REICHERT micro-
scope apparatus and are uncorrected. Elemental analyses were
2
2909 observations; calc w = 1/[σ2(Fo ) ϩ (0.0599P)2 ϩ 0.6055P]
where P = (Fo2 ϩ 2Fc2)/3 with the resulting R = 0.037, Rw =
0.077 and Sw = 1.029 (residual ∆ρ ≤ 0.35 e ÅϪ3).
† CCDC reference number 207/354. See http://www.rsc.org/suppdata/
p1/1999/2821 for crystallographic files in .cif format.
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828
2825

View Article Online
carried out on a Perkin-Elmer 2400, C, H, N elemental analyser
at the Microanalyse Service of CNRS (Vernaison, France).
Tetrahydrofuran (THF) was prepared by pre-drying on KOH
pellets followed by distillation from Na–benzophenone. Diethyl
ether was distilled from Na–benzophenone. Acetonitrile, ethyl
acetate and methylene chloride were dried by distillation
over P2O5 and toluene was distilled from sodium. All solvents
were freshly distilled by standard methods prior to use. Flash
chromatography was performed on silica gel Merck 60 230–400
mesh. Thin layer chromatography was performed on precoated
plates of silica gel 60F254 (Merck, Art 7735).
7.58–7.78 (5H, m, C6H5), 8.57 (1H, s, H-4); δC 17.1 (SMe,
1
1JCH = 144 Hz), 39.6 and 45.8 (NMe2, JCH = 143 Hz), 129.5,
129.9, 134.3, 134.7 (Ph), 160.3 (C-4, 1JCH = 192 Hz), 196.9 (C-2);
νmax(KBr)/cmϪ1 1649, 1578, 1556 (for IR spectra values of
amidines and imidates see ref. 25).
3b. Mp 163–164 ЊC (from diethyl ether) (Found: C, 41.35; H,
4.95; N, 8.10. C12H17N2SI requires C, 41.39; H, 4.92; N, 8.04; S,
9.21%); δH 2.35 and 2.66 (6H, s, 2CH3), 3.60 and 3.78 (6H, 2s,
NMe2), 7.65 (s, 5H, Ph); δC 16.1 and 22.3 (2Me), 42.2 and 43.5
(NMe2), 127.3, 130.0, 133.8 and 137.5 (Ph), 171.3 and 181.3
(C-2 and C-4); νmax(KBr)/cmϪ1 1624 and 1590.
4-Dimethylamino-4-methyl-2-phenyl-1-thia-3-azabuta-1,3-diene
1b
4b. Mp 111–113 ЊC (from diethyl ether) (Found: C, 35.95;
H, 4.54; N, 7.74. C12H17N2SeI requires C, 36.48; H, 4.34; N,
7.09%); δH 2.47 and 2.57 (6H, 2s, 2Me), 3.55 and 3.79 (6H, 2s,
NMe2), 7.65 (5H, s, Ph); δC 10.4 and 21.8 (2Me), 42.1 and 43.5
(NMe2), 127.0, 129.8, 133.7 and 135.3 (Ph), 170.6 and 181.4
(C-2 and C-4); νmax(KBr)/cmϪ1 1626, 1575.
Method A. The heterodiene 1b (90% yield) was obtained by
condensation of N,N-dimethylacetamide dimethyl acetal in
excess thiobenzamide using a procedure described in the
literature.2a
Method B. Methylselenoamidinium salt 4b (600 mg, 1.5
mmol) was dissolved in dry CH3CN (20 ml). Potassium thio-
acetate (420 mg, 2.95 mmol) was added under nitrogen and the
mixture was stirred at room temperature for 2 h. The mixture
was filtered on a pad of Celite to remove insoluble non-organic
material and the solvent was removed in vacuo. Column
chromatography afforded the title diene (4:1 petroleum ether–
ethyl acetate) in 57% yield. From 3b, an identical procedure led
to 1b in 58% yield.
4-Dimethylamino-2-phenyl-1-oxa-3-azabuta-1,3-diene 5a
Methylthioamidinium salt 3a (534 mg, 1.6 mmol) was dissolved
in dry CH3CN (20 ml). Silver acetate (534 mg, 3.2 mmol) was
added under nitrogen and the mixture was stirred at room tem-
perature for 30 min. The mixture was then filtered on a pad of
Celite to remove insoluble non-organic material and the solvent
was removed in vacuo. Column chromatography of the crude
material afforded the title diene (3:2 petroleum ether–ethyl
acetate) in 40% yield. Colourless needles, mp 76 ЊC (from
diethyl ether) (Found: C, 68.48; H, 6.82; N, 15.36. C10H12N2O
requires C, 68.16; H, 6.86; N, 15.90; O, 9.08%). The physical
and spectral properties of isolated 5a were in accordance with
literature values.26
Method C. Methylselenoamidinium salt 4b (800 mg, 2 mmol)
was dissolved in dry CH2Cl2 (20 ml). Thiobenzamide (275 mg,
3.2 mmol) and triethylamine (4 mmol) were added successively
under a nitrogen atmosphere. The mixture was stirred at room
temperature for 4 h then concentrated under reduced pressure.
The crude product was purified by flash chromatography (4:1
petroleum ether–ethyl acetate) to give 1b in 17% yield. Mp
112 ЊC (from ethanol) (Found: C, 64.29; H, 6.69; S, 15.68.
C11H14N2S requires C, 64.04; H, 6.84; N, 13.58; S, 15.54%);
δC 18.0 (Me), 39.2 (NMe2), 127.6, 128.8, 131.8 and 142.6 (Ph),
167.8 (C-4) and 202.7 (C-2); m/z 206 (Mϩ, 70%), 173 (100), 129
(26), 121 (61), 77 (45), 56 (30); νmax(KBr)/cmϪ1 1624 and 1590.
General procedures for N-(p-tolylsulfonyl)-1,3-diazabuta-1,3-
dienes 6 (and adducts 13, 14, 15)
Method A. To a stirred solution of amidinium salt 3a (668
mg, 2 mmol) dissolved in dry CH3CN (20 mL) were sub-
sequently added, under a dry nitrogen atmosphere, toluene-p-
sulfonamide (340 mg, 2 mmol), Et3N (0.56 mL, 4 mmol) and
mercury() chloride (944 mg, 2 mmol). The mixture was stirred
at room temperature for 1.5 h and then concentrated under
reduced pressure. The crude product was purified by flash
chromatography (3:2 petroleum ether–ethyl acetate) to give 14
in 32% yield. In the absence of mercury salt, the isolated reac-
tion products were identified as diazabutadiene 6a, N-tosyl
amidine 13 and formamidine 14 (18, 38 and 40% yields, respec-
tively). Starting from the amidinium salt 3b, the same experi-
mental procedure led to the formation of diazabutadiene 6b
isolated in 20% yield along with N-tosyl methylthiobenzimidate
15 (15% yield).
4-Dimethylamino-4-methyl-2-phenyl-1-selena-3-azabuta-1,3-
diene 2b
Method A. Using a procedure similar to that described above
for the synthesis of 1b, compound 2b was obtained from
selenobenzamide in 92% yield.2h
Method B. Methylthioamidinium salt 3b (350 mg, 1 mmol)
was dissolved in dry CH2Cl2 (20 ml). Selenobenzamide (184 mg,
1 mmol) and triethylamine (2 mmol) were added successively
under nitrogen. The mixture was stirred at room temperature
for 2 h then concentrated under reduced pressure. The crude
product was purified by flash chromatography (7:3 petroleum
ether–ethyl acetate; Rf 0.16) to give 2b in 80% yield. (Physical
properties of 2b can be found in reference 2h.)
Method B. To a solution of N-thiophenacylformamidine 1a
in dry toluene (30 mL) was added toluene-p-sulfonyl azide27
(1 g, 5 mmol) under a nitrogen atmosphere. The mixture was
stirred at reflux for 4 h and then concentrated under reduced
pressure. The crude product was purified by flash chromato-
graphy (3:2 petroleum ether–ethyl acetate) to give 6a in 75%
yield. Starting from N-thio- or N-selenophenacylformamidine
1b, 2b, 1,3-diazabuta-1,3-diene 6b could be obtained in 87
and 94% yields respectively, following the same experimental
procedure as above.
General procedure for methylheteroamidinium salts 3a, 3b and 4b
The heterodiene precursors 1a, 1b or 2b (1 mmol) were dis-
solved in methyl iodide used as solvent. The colourless solu-
tions were stirred at rt for 30 min under nitrogen. In all cases the
addition of diethyl ether to the mixture induced the precipi-
tation of the amidinium salts. After filtration, the solids were
washed with diethyl ether.
4-Dimethylamino-2-phenyl-1-(p-tolylsulfonyl)-1,3-diazabuta-
1,3-diene 6a. Mp 170–171 ЊC (from light petroleum) (Found:
C, 61.69; H, 5.78; N, 12.56. C17H19N3O2S requires C, 62.01; H,
5.78; N, 12.77%); δH 2.37 (3H, s, Me), 2.97 and 3.38 (6H, 2s,
3a.6 Mp 173 ЊC (from diethyl ether) (Found: C, 39.03; H,
4.54; N, 7.92. C11H15N2SI requires C, 39.53; H, 4.52; N, 8.38; S,
9.59%); δH 2.75 (3H, s, SMe), 3.64 and 3.73 (6H, 2s, NMe2),
NMe ), 7.14–7.70 (9H, m, Ph, C H ), 8.37 (1H, s, CH᎐N);
᎐
2
6
4
δC 21.5 (Me), 38.8 and 41.1 (NMe2), 127.0, 127.8, 129.2, 129.5,
2826
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828

View Article Online
130.9, 131.0, 138.3, 142.6 (Ph, C6H4), 169.5 and 174.1 (C-2,
C-4); m/z 329 (Mϩ, 9%), 238 (33), 174 (95), 173 (25), 172 (22),
119 (43), 118 (51), 116 (41), 104 (100), 91 (92), 77 (31), 71 (76),
65 (27), 44 (23); νmax(KBr)/cmϪ1 1597, 1520.
br s, CH᎐N); δ 14.0 (SMe), 21.7 (Me), 77.4 (PhC), 126.6,
᎐
C
127.0, 127.7, 128.5, 129.2, 129.4, 129.8, 138.3, 139.1, 143.4 (Ph,
1
C H ), 160.1 (CH᎐N, JCH = 183 Hz); m/z 317 (Mϩ ϩ 1, 30%),
᎐
6
4
155 (64), 104 (24), 91 (100), 65 (26); νmax(KBr)/cmϪ1 1615
(C᎐N), 3285 (NH).
᎐
4-Dimethylamino-2-phenyl-1-(p-tolylsulfonyl)-1,3-diazapenta-
1,3-diene 6b. Mp 125–126 ЊC (from light petroleum) (Found:
C, 63.16; H, 6.24; N, 11.93. C18H21N3O2S requires C, 62.95; H,
6.16; N, 12.23%); δH 2.05 and 2.39 (6H, 2s, 2Me), 3.04 and 3.11
(6H, 2s, NMe2), 7.22–7.93 (9H, m, Ph, C6H4); δC 19.6 (Me), 21.0
(Me), 38.1 (NMe2), 126.5, 127.8, 128.5, 128.8, 131.8, 135.7,
139.9, 141.6 (Ph, C6H4), 158.6 and 166.9 (C-4, C-2); m/z 344
(Mϩ ϩ 1, 3%), 188 (96), 176 (30), 173 (34), 155 (21), 104 (57),
91 (100), 65 (25), 56 (22), 44 (43); νmax(KBr)/cmϪ1 1614, 1583.
N-(p-Tolylsulfonyl)methylthiobenzimidate 15
Mp 111 ЊC (from light petroleum) (Found: C, 58.72; H, 4.89;
S, 20.52. C15H15NO2S2 requires C, 58.99; H, 4.95; N, 4.59; O,
10.48; S, 20.99%); δH 2.39 (6H, s, MeC6H4, SMe), 7.21–7.78
(9H, m, Ph and C6H4); δC 15.9 and 21.5 (2Me), 126.9, 127.5,
128.1, 129.2, 131.4, 136.1, 138.7, 143.2 (Ph, C H ), 185.7 (C᎐N);
᎐
6
4
m/z 306 (CI, Mϩ, 100%), 258 (35), 155 (91), 105 (30), 91 (100),
77 (20), 65 (35); νmax(KBr)/cmϪ1 1596, 1546.
N-Acylbenzamide 8
References
To a solution of methylthioamidinium salt 3a (534 mg, 1.6
mmol) in dry acetonitrile were added silver nitrate (533 mg, 3.2
mmol) and water (1 mL). The mixture was stirred at room
temperature for 30 min and then concentrated under reduced
pressure. The crude product was flash-chromatographed (3:2
petroleum ether–ethyl acetate) to give 8 (42% yield) as white
crystals. Mp 110–111 ЊC (from diethyl ether) (Found: C, 64.56;
1 (a) D. L. Boger and S. M. Weinreb, in Hetero Diels–Alder
Methodology in Organic Synthesis, ed. H. H. Wasserman, Academic
Press, San Diego, 1987; (b) W. Carruthers, in Cycloaddition
Reactions in Organic Synthesis, ed. J. E. Baldwin and P. D. Magnus,
Pergamon Press, Oxford, 1990, vol. 8; (c) D. L. Boger, in
Comprehensive Organic Synthesis, ed. L. A. Paquette, Pergamon
Press, Oxford, 1991, vol. 5, p. 451; (d) H. Waldmann, Synthesis,
1994, 535; (e) L. Tietze and G. Kettschau, Top. Curr. Chem., 1997,
189, 1.
2 (a) J. C. Meslin and H. Quiniou, Bull. Soc. Chim. Fr., 1979, 116, 347;
(b) J. C. Rozé, J. P. Pradère, G. Duguay, A. Guevel, C. G. Tea and
H. Quiniou, Sulfur Lett., 1983, 1, 115; (c) C. G. Tea, J. P. Pradère
and H. Quiniou, J. Org. Chem., 1985, 50, 1545; (d) J. P. Pradère,
J. C. Rozé, H. Quiniou, R. Danion-Bougot, D. Danion and
L. Toupet, Can. J. Chem., 1986, 64, 597; (e) C. Cellerin, C. G. Tea,
J. P. Pradère, A. Guingant and P. Guenot, Sulfur Lett., 1988, 8,
205; ( f ) J. P. Pradère, F. Tonnard, A. Abouelfida, C. Cellerin,
M. Andriamanamihaja, B. Jousseaume, L. Toupet and P. Guénot,
Bull. Soc. Chim. Fr., 1993, 130, 610; (g) D. Dubreuil, J. P. Pradère,
N. Giraudeau, M. Goli and F. Tonnard, Tetrahedron Lett., 1995,
36, 237; (h) F. Purseigle, D. Dubreuil, A. Marchand, J. P. Pradère,
M. Goli and L. Toupet, Tetrahedron, 1998, 54, 2545.
3 A. Marchand, D. Mauger, A. Guingant and J. P. Pradère,
Tetrahedron: Asymmetry, 1995, 6, 853.
4 (a) J. Liebscher, M. Pätzel and Y. F. Kelboro, Synthesis, 1989, 672;
(b) C. Cellerin, J. P. Pradère, D. Danion and F. Tonnard, C. R. Acad.
Sci. Paris, 1991, 313, Série II, 517; (c) A. Rolfs and J. Liebscher,
J. Org. Chem., 1997, 62, 3480.
5 For a preliminary study about the heteroatom interconversion
reaction concept as shown in Scheme 1, see reference 4a. See also
for related work: (a) H. G. Gold, Angew. Chem., 1960, 72, 956;
(b) J. Liebscher and H. Hartmann, Synthesis, 1976, 403.
6 J. C. Meslin, A. Reliquet, F. Reliquet and C. G. Tea, C. R. Acad. Sci.
Paris, 1978, 286, Série C, 397.
7 (a) Semi empirical PM3 calculations were carried out with
SPARTAN software (4.0 version) on a Silicon Graphics Indy. For
PM3 calculations see: J. J. P. Stewart, Comput. Chem., 1989, 42, 12;
(b) F. Purseigle, PhD Thesis (30/01/97), University of Nantes.
8 J. Liebscher, Synthesis, 1988, 655.
H, 4.86; N, 9.40. C8H7NO2 requires C, 64.42; H, 4.73; N,
1
9.39%); δC 128.1, 129.0, 131.0, 133.9 (Ph), 164.6 (HCO, JCH
=
208 Hz), 166.8 (CO); the other spectral properties of 8 (1H
NMR and IR) were in full accordance with literature values.28
Compounds 12
Using a procedure similar to that described above for the syn-
thesis of 5a, compound 12 (two tautomer forms in solution,
vide infra) was obtained either from N-heterophenacylamidines
1b and 2b (50 and 80% yields, respectively) or from amidinium
salts 3b and 4b (60 and 85% yields, respectively) (Found: C,
67.60; H, 7.21; N, 12.35; O, 13.78. C13H16N2O2 requires C,
67.22; H, 6.94; N, 12.06; O, 13.78%); m/z 232 (Mϩ, 25%), 189
(25), 105 (100), 77 (57), 44 (41); νmax(film)/cmϪ1 1721, 1695 and
1623.
N-(1-Dimethylamino-3-oxobutylidene)benzamide. δH 2.17
(3H, s, Me), 2.9 and 3.1 (6H, 2s, NMe2), 3.8 (2H, s, CH-2), 7.2–
1
8.1 (5H, 2m, Ph); δC 31.8 (Me, JCH = 126.2 Hz), 39.3 (NMe2
1JCH = 142.2 Hz), 45.8 (CH2, JCH = 128.2 Hz), 128.9, 129.3
1
130.3, 138.5 (Ph), 164.0, 176.3, 201.7 (C᎐N, 2 C᎐O).
᎐
᎐
N-(1-Dimethylamino-3-oxobut-1-enyl)benzamide
(enamino
ketone form). δH 1.9 (3H, s, Me), 2.8 (6H, s, NMe2), 4.7 (1H, s,
CH), 13.4 (1H, s, NH), 7.3 and 7.4 (5H, 2m, Ph); δC 30.6 (Me,
1
1JCH = 126.2 Hz), 40.9 (NMe2, JCH = 142.2 Hz), 85.8 (C᎐CH,
᎐
1JCH = 159.4 Hz), 129.0, 129.5, 133.4, 137.1 (Ph), 158.5, 166.8,
195.3 (2 C᎐O, C᎐CH).
9 For a recent review on conversion of the thiocarbonyl group into the
carbonyl group, see: A. Corsaro and V. Pistarà, Tetrahedron, 1998,
54, 15027.
10 M. Avalos, R. Babiano, C. J. Duran, J. L. Jiménez and J. C. Palacios,
Tetrahedron Lett., 1994, 35, 477.
11 J. Barluenga and M. Thomas, in Advances in Heterocyclic
Chemistry, ed. A. R. Katritzky, Academic Press, San Diego, 1993,
vol. 57, p. 61.
12 H. Schaumann, in The chemistry of double-bonded functional groups,
ed. S. Patai, Interscience Publication, John Wiley and Sons,
New York, 1989, vol. 2, ch. 17, p. 1338.
᎐
᎐
N-(p-Tolylsulfonyl)-NЈ,NЈ-dimethylformamidine 13
Mp 113–114 ЊC (from light petroleum) (Found: C, 53.39; H,
6.29; N, 12.51; S, 14.47. C10H14N2O2S requires C, 53.08; H, 6.24;
N, 12.38; O, 14.14; S, 14.17%); δH 2.39 (3H, s, Me), 3.00 and
3.11 (6H, 2s, NMe2), 7.23–7.78 (4H, 2d, C6H4), 8.12 (1H, s,
1
CH᎐N); δ 21.4 (Me, JCH = 127 Hz), 35.4 and 41.4 (NMe2,
᎐
C
1JCH = 138.5 Hz), 126.4, 129.3, 139.6, 142.4 (C6H4), 159.1
(CH᎐N, JCH = 184 Hz); m/z 226 (Mϩ, 27%), 91 (73), 71 (100),
13 Selected torsion angles: N1–C2–N3–C4 equal to Ϫ153.76Њ in 6a and
123.47Њ in 6b; C2–N3–C4–N2 equal to Ϫ170.83Њ in 6a and 166.54Њ
in 6b.
1
᎐
65 (42), 44 (83); νmax(KBr)/cmϪ1 1626.
14 C. K. Johnson, ORTEP Report ORNL-3794, Oak Ridge National
Laboratory, Tenessee, USA, 1965.
N-(p-Tolylsulfonyl)-NЈ-[bis(methylthio)phenylmethyl]form-
amidine 14
15 S. N. Mazumdar and M. P. Mahajan, Synthesis, 1990, 417.
16 (a) P. Luthardt and E. U. Würthwein, Tetrahedron Lett., 1988, 29,
921; (b) S. N. Mazumdar and M. P. Mahajan, Tetrahedron, 1991,
47, 1473; (c) S. N. Mazumdar, S. Mukherjee, A. Z. Sharma,
D. Sengupta and M. P. Mahajan, Tetrahedron, 1994, 50, 7579; (d)
P. D. Dey, A. K. Sharma, P. V. Bharatam and M. P. Mahajan,
Mp 60–61 ЊC (from diethyl ether) (Found: C, 53.67; H, 5.10;
N, 7.35; S, 24.98. C17H20N2O2S3 requires C, 53.66; H, 5.30; N,
7.36; O, 8.41; S, 25.27%); δH 1.98 (6H, s, SMe), 2.42 (3H, s, Me),
6.91 (1H, br s, NH), 7.27–7.79 (9H, m, Ph and C6H4), 8.78 (1H,
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828
2827

View Article Online
Tetrahedron, 1997, 53, 13829; (e) A. K. Sharma and M. P. Mahajan,
Tetrahedron, 1997, 53, 13841.
17 (a) K. Burger, E. Huber, N. Sewald and H. Partscht, Chem. Ztg.,
1986, 110, 83; (b) W. M. F. Fabian and G. Kollenz, J. Phys. Org.
Chem., 1994, 7, 1; (c) E. Rossi, G. Abbiati and E. Pini, Tetrahedron,
1997, 53, 14107.
18 For dipolarophilic properties of 1 see: R. Danion-Bougot, R.
Tuloup, D. Danion, J. P. Pradère and F. Tonnard, Sulfur Lett., 1989,
9, 245.
23 International tables for X-ray crystallography, Kynoch Press (present
distributor D. Reidel, Dordrecht), Birmingham, 1974, vol. IV.
24 A. L Spek, PLATON. A multipose crystallographic tool, 1998,
Utrecht University, Utrecht, The Netherlands.
25 G. Häfelinger, in The chemistry of amidines and imidates, ed. S.
Patai, Interscience Publication, John Wiley and Sons, New York,
1975, ch. 1, p. 53.
26 A. J. Blake, H. McNab and M. E. Ann Murray, J. Chem. Soc.,
Perkin Trans. 1, 1989, 589.
19 C. K. Fair, MolEN. An Interactive Intelligent System for Crystal
Structure Analysis, Enraf-Nonius, Delft, The Netherlands, 1990.
20 A. L. Spek, HELENA. Program for the handling of CAD4-
Diffractomer output SHELX(S/L), 1997, Utrecht University,
Utrecht, The Netherlands.
27 M. Regitz, J. Hocker and A. Liedhegener, Organic Syntheses,
ed. J. Wiley and Sons, New York, 1988, Coll. Vol. V, p. 179.
28 B. Seiller, D. Heins, C. Bruneau and P. H. Dixneuf, Tetrahedron,
1995, 51, 10901.
21 A. Altomare, M. C. Burla, M. Camalli, G. Cascarano, C.
Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori and
R. Spagna, J. Appl. Crystallogr., 1998, 31, 74.
22 G. M. Sheldrick, SHELX93. Program for the refinement of crystal
structures, 1993, University of Göttingen, Germany.
Paper 9/03150B
2828
J. Chem. Soc., Perkin Trans. 1, 1999, 2821–2828