A relationship between the catalytic carbonylation of (N-arenesulfonyl)imides and iodonium ylides

Article abstract of DOI:10.1139/v01-074

Reactivities of various sulfonamide derivatives of general formula ArSO₂N=XPh_n (X = I; n = 1; X = S, Se, n = 2; X = P, As, n = 3) in catalytic carbonylations have been explained by structural properties of the substrates. Based on cr

Full text of DOI:10.1139/v01-074

649
A relationship between the catalytic carbonylation
of (N-arenesulfonyl)imides and iodonium ylides
Gábor Besenyei, András Neszmélyi, Sándor Holly, László I. Simándi, and
László Párkányi
Abstract: Reactivities of various sulfonamide derivatives of general formula ArSO2N=XPhn (X = I; n = 1; X = S, Se,
n = 2; X = P, As, n = 3) in catalytic carbonylations have been explained by structural properties of the substrates.
Based on crystallographic data reported in the literature, it has been concluded that the two-step catalytic oxidative
carbonylation of sulfonylimides proceeds only when the N=X double bond is strongly polarized. The structural require-
ment observed within the sulfonamide derivatives has been found to be applicable for iodonium ylides. Catalytic
carbonylation of the phenyliodonium methylide (3a) prepared from dimedone (5,5-dimethyl-1,3-cyclohexanedione) and
(diacetoxyiodo)benzene, resulted in a new spirobenzofuranone derivative (5a) which was characterized by X-ray single
1
crystal diffraction, JC,C coupling constants, 13C-T1 relaxation times, as well as by IR and Raman spectroscopic meth-
ods. Crystals of 5a are monoclinic, space group P21/c, a = 7.187(1), b = 21.129(1), and c = 10.636(1) Å, β =
92.81(1)°, V = 1613.2(3) Å3, and Z = 4. A ketene intermediate has been postulated in the catalytic carbonylation of the
iodonium ylide. A survey of databases has revealed that the triacyl dihydrofuranone moiety in 5a constitutes a new
structural unit.
Key words: oxidative carbonylation, palladium catalyst, iodonium ylides, spirobenzofuranone.
Résumé : Les réactivités de divers dérivés du sulfonamide, de formule générale ArSO2N=XPhn (X = I, n = 1; X = S,
Se, n = 2; X = P, As, n = 3), dans des réactions de carbonylations catalytiques sont expliquées par des propriétés
structurales des substrats. En se basant sur des données cristallographiques rapportées dans la littérature, on a conclu
que la réaction de carbonylation oxydante, catalysée et en deux étapes, de sulfonylimides ne se produit que lorsque la
double liaison N=X est fortement polarisée. On a trouvé que les conditions structurales observées dans les dérivés sul-
fonamides s’appliquent aussi aux ylures d’iodonium. La carbonylation catalytique du méthylure de phényliodonium (3a)
préparé à partir de la dimédone (5,5-diméthylcyclohexane-1,3-dione) et de (diacétoxyiodo)benzène, conduit à la forma-
tion d’un nouveau dérivé spirobenzofuranone (5a) qui a été caractérisé par diffraction des rayons X sur un cristal
1
unique, par les constantes de couplage JC,C, les temps de relaxation 13C-T1 ainsi que par spectroscopies IR et Raman.
Les cristaux de 5a sont monoclinique, groupe d’espace P21/c avec a = 7,187(1), b = 21,129(1) et c = 10,636(1) Å, β =
92,81°, V = 1613,2(3) Å3 et Z = 4. On suggère l’existence d’un intermédiaire cétène lors de la carbonylation cataly-
tique de l’ylure d’iodonium. Une revue des bases de données indique que la partie triacyldihydrofuranone du composé
5a correspond à une nouvelle unité structurale.
Mots clés : carbonylation oxydante, catalyseur de palladium, ylures d’iodonium, spirobenzofuranone.
[Traduit par la Rédaction] Besenyei et al. 654
Introduction
Reductive carbonylation of aromatic nitro and nitroso com-
pounds and oxidative carbonylation of aromatic amines are
the most representative cases of these catalytic transforma-
tions.
During our search for phosgene-free methods for the prepa-
ration of arenesulfonyl isocyanates, we have discovered three
new reactions leading to the formation of these target mole-
cules (3–5). The catalytic transformation of N-chloro-
arenesulfonamidates (1) and (N-arenesulfonyl)imides (2) are
shown by eqs. [1] and [2].
The practical importance of organic isocyanates and their
derivatives like urethanes and ureas has attracted continuous
interest in catalytic carbonylation at nitrogen atoms (1, 2).
Received October 13, 2000. Published on the NRC Research
Press Web site http://canjchem.nrc.ca on June 30, 2001.
Dedicated to Brian James on the occasion of his 65th
birthday.
G. Besenyei,1 A. Neszmélyi, S. Holly, L.I. Simándi,
and L. Párkányi. Chemical Research Center, Institute of
Chemistry, Hungarian Academy of Sciences, 1525 Budapest
P.O. Box 17, Hungary.
Pd−complex
[1] [ArSO2NCl]M + CO
→ ARSO2NCO + MCl
25° C
5−30 bar
1
1Corresponding author (telephone: (36-1) 325-7900 exts. 197
or 339; fax: (36-1) 325-7554; e-mail: besenyei@cric.chemres.hu).
+
+
M = Na , K
Can. J. Chem. 79: 649–654 (2001)
© 2001 NRC Canada
DOI: 10.1139/cjc-79-5/6-649

650
Can. J. Chem. Vol. 79, 2001
Table 1. A compilation of dNX bond lengths in (N-arenesulfonyl)imides (ArSO2N=XPhn).
dNX (obsd. or calcd.) (Å)
Compound
dNX (Å)
Single bond
Double bond
Ref.
Ar1SO2N=IAr2
CH3C6H4SO2N=SePh2
C6H5SO2N=SePh2
1.997(8)–2.039(3)
1.787(6)
1.82(1)
1.99
—
1.64
—
10–12
13
14
1.824(6)–1.846(4)
—
CH3C6H4SO2N=SPh2
CH3C6H4SO2N=PPh3
CH3C6H4SO2N=AsPh3
1.628(7)
1.579(4)
1.755(3)
1.76(2)
1.77(2)
1.87
1.53
1.60
—
6a, 8, 9
6
7
Pd−complex
S(IV)—N bond order in 2c can be estimated to be ca. 1.5
(d = 1.628(7) Å) (8, 9).
[2]
ArSO2 N=XPhn + CO
→
h
ArSO2NCO + P nX
25–50° C
2
On the contrary, available structural data for the [(N-
arenesulfonyl)imino]phenyliodinanes (ArSO2N=IPh) and the
theoretical calculations as well, have suggested little evi-
dence for the iodine-nitrogen multiple bonding (10–12). Sin-
gle crystal diffraction studies have revealed the strongly
polarized nature of the I=N bonds, which brings about sec-
ondary interactions between the amide nitrogen, sulfonyl ox-
ygen, and iodine atoms. In 4-CH3C6H4SO2N=IPh (2a, Ar =
4-CH3C6H4), the I—N bond length is 2.039(2) Å (12),
which together with other iodine—nitrogen covalent bonds
found in this series of imidocompounds (1.997(8)–2.027(3)
Å) (10, 11), compares well with the value of 1.99 Å calcu-
lated for a single bond (12). Although the nitrogen—sele-
nium bond length in 2b (Ar = 4-CH3C6H4) falls somewhat
out of the range observed for a selenium—nitrogen single
bond (1.824(6)–1.846(4) Å), the experimentally determined
value of 1.787(6) Å is still much greater than that predicted
for a selenium nitrogen double bond (1.64 Å) (13). X-ray
diffraction studies may indicate a further loss of Se—N mul-
tiple bonding in C6H5SO2N=SePh2 (1.82(1) Å) (14). These
observations suggest that resonance structure A makes a ma-
jor contribution to the distribution of valence electrons in 2a
and 2b, while formula B is thought to be a good approxima-
tion for the structural description of 2c–e. The N—Cl bond
length in the sodium N-chloro-p-toluenesulfonamidate 1
(1.750(2) Å) was found to be equal to the mean N—Cl dis-
tance in NCl3 and was described as a single bond (15).
30 bar
2a X = I, n =1; 2b X = Se, n = 2
Although the products are isocyanates, resembling those
of the catalytic carbonylation of nitro arenes, rxns. [1] and
[2] cannot be described as reductive carbonylations. As sub-
strates 1 and 2 are prepared via oxidative reactions, these
transformations of sulfonamide derivatives have been classi-
fied as two-step oxidative carbonylations and constitute a
new class of catalytic reactions taking place at nitrogen at-
oms.
To check if other sulfonylimides have a tendency to un-
dergo the reaction described by eq. [2], we prepared several
known imidocompounds of general formula 4-
CH3C6H4SO2N=XPhn (2c X = S, n = 2; 2d X = P, n = 3; 2e
X = As, n = 3) and subjected them to catalytic carbonyla-
tion. No CO consumption could be detected, however, under
reaction conditions successfully applied earlier for the cata-
lytic transformation of N-chloroarenesulfonamidates, [(N-
arenesulfonyl)imino]phenyliodinanes, and imido selenium
compounds.
It seemed to us of interest to find an explanation for why
compounds possessing seemingly close structural features
behave differently under the same reaction conditions. To
elucidate this problem, we examined various structural and
chemical properties of the substrates. As the reaction takes
place via the cleavage of nitrogen-heteroatom bonds, special
attention was paid to data characterizing the bonding proper-
ties within the SO2NX fragments.
Results and discussion
As a result of the long-standing interest toward the sulfo-
namide derivatives used here as model compounds, crystal-
lographic data are available for the typical representatives of
these substrates, providing us with the necessary structural
information to discuss a structure–reactivity relationship. A
compilation of selected crystallographic data is shown in Ta-
ble 1.
The phosphorus—nitrogen bond length of 2d (1.579(4) Å)
is much shorter than a P—N single bond (1.77(2) Å), there-
fore a bond order of 2 is most probable in the (N-p-
toluenesulfonyl)iminotriphenylphosphorane (6). A compari-
son of the experimentally determined arsine—nitrogen dis-
tance in 2e (1.755(3) Å) with the As—N single bond length
of ca. 1.87 Å makes feasible a similar conclusion (7). Based
on a systematic study of sulfur—nitrogen bonds, the
Apart from the crystallographic data, the other structural
parameter considered was the position of the νasSO2 band in
the IR spectra of the substrates (recorded in KBr pellets) be-
cause the wave number of this absorption can vary consider-
ably depending on the nature of the bonds surrounding the
sulfonyl group. The extent of these shifts, however, does not
have a predictive power by itself with respect to the reactiv-
ity of a given imide. The νasSO2 for the inert phosphorane 2d
and for the highly reactive [(N-p-toluenesulfonyl)imino]-
phenyliodinane were equally found to appear at 1266 cm–1.
This overview of the available structural data clearly
shows that the same structural formulas may hide completely
© 2001 NRC Canada

Besenyei et al.
651
different structural properties within the SO2NX system and
a bond order of close to unity is a precondition of the
carbonylation reaction. It seems plausible to assume that it is
the strength of the SO2N—X bond that is reflected by the re-
activity of the substrates, and strongly polarized, conse-
quently weaker, double bonds make the imide more reactive.
The polarization of the N—X bond is not important just for
weakening the bond but certainly facilitates the initial inter-
action with the catalyst.2 Different reactivities of N-
sulfonylimides in copper-catalyzed aziridination of olefins
have also been observed by Evans et al. (16) by demonstrat-
ing that 2a (Ar = p-tolyl) resulted in the highest yield, while
the imido sulfur compound PhMeS=NSO2Ph gave no reac-
tion in copper-catalyzed aziridination of olefins.
triacyl dihydrofuranone ring as well (shown as formula 6) is
a structural element that has not been observed in any cyclic
or open-chain molecule (“A” represents a hydrogen atom,
any organic moiety, or forms a ring with other units defined
as “A”). Therefore, catalytic carbonylation of iodonium
ylides can be viewed as a promising method to prepare for-
merly unknown heterocyclic compounds.
1
Based on two-dimensional routine H and 13C NMR meth-
ods (including HMQC) it was not feasible to conclusively
establish the connectivity of the carbon atoms. Determina-
1
tion of the JC,C coupling constants and T1 relaxation times
for all carbon atoms and the assignment of the ν(C=C) band
both in IR and Raman spectra together with Raman depolar-
ization measurements have unequivocally proved, however,
that the product is a hexahydrobenzofuranone derivative.
The results of spectroscopic studies carried out on bulk
samples are in agreement with the results of crystallographic
studies on a single crystal. NMR data are collected in
Table 2. Crystallographic data, data collection parameters,
and selected bond lengths and angles are summarized in Ta-
bles 3 and 4. A molecular diagram of 5a is shown in Fig. 1.
Apart from the large difference between the angles C(3)-
C(3a)-C(4) (130.8°) and C(7)-C(7a)-O(1) (118.3°), com-
pound 5a is quite regular from a structural point of view.
The atoms in the hexahydrobenzofuranone ring are coplanar
with the exception of the C(6) atom that bends above the
plane by 0.6 Å and this feature turns this portion of the mol-
ecule into an envelope conformation. The cyclohexanedione
moiety adopts a chair conformation with the two carbonyl
groups pointing toward the side of the O(1) atom.
All these observations may be viewed as indirect evidence
for the intermediacy of palladium–nitrene complexes that
may emerge in the catalytic N-carbonylation of sulfonamide
derivatives.
[(N-arenesulfonyl)imino]phenyliodinanes and phenyliodo-
nium methylides of general formulas ArSO2N=IPh and
E1E2C=IPh (E represents electron-withdrawing groups like
RC(O), RSO2, RFSO2), respectively, have been mentioned as
nitrogen and carbon analogues of iodosylbenzene (PhIO)
and have been postulated as sources of arylsulfonylnitrene
(16) and diacylcarbene species (17). Considerations regard-
ing the structural properties of iodonium ylides (18) and
available crystallographic data for them (19) have turned our
attention toward the structural similarities of 2a and
iodonium ylides and have made the latter promising candi-
dates as substrates for two-step oxidative carbonylations.
The relaxation times for methylene and methyl carbon at-
oms (see Fig. 2) result in average values of 0.58 s and 3.2 s,
respectively. The averaging of carbon chemical shifts in the
saturated rings reflect conformational transitions (envelope–
envelope transitions probably). In addition, the methyl sub-
stituents have a nearly complete rotational freedom (methyl
relaxation times are approximately nine times longer than
expected for a rigid methyl carbon). From these observations
and the individual T1 values of the protonated carbons along
the skeleton we may deduce that the microdynamics of the
molecule in solution corresponds to that of a virtual rigid
body.
[3]
Iodonium ylides were first described in the pioneering
works of Neiland and co–workers (20), and their chemical
and physical properties have been extensively studied (18,
21). However, to our knowledge, catalytic carbonylation of
phenyliodonium methylides has not been described.
As expected, catalytic carbonylation of the iodonium ylide
3a prepared from 5,5-dimethyl-1,3-cyclohexanedione and
(diacetoxyiodo)benzene was accompanied by CO consump-
tion, but the IR spectrum of the final reaction mixture did
not reveal any absorption that could have been explained by
the formation of ketene 4. Based on X-ray crystallographic
data as well as on NMR and IR-Raman studies, structure 5a
was ascribed to the product. The carbonyl group in position
3 clearly indicates that the isolated compound is a product of
catalytic carbonylation.
We feel noteworthy that the NMR and IR-Raman spectro-
scopic data supported by crystallographic investigations re-
ported here for 5a may be of help in recognizing similar
structural units.
We suppose that 5a is formed via the intermediate ketene
4, which is thought to be the primary product of the reac-
tion. The postulated formation of 4 is in agreement with pre-
vious reports on the intermediacy of diketocarbene species
generated from 3a. Compound 5a seems to be a product of
the interaction of 4 with the unreacted starting substrate 3a.
The feasibility of a thermal reaction of diphenylketene with
3a has been demonstrated (23) and was found to produce
heterocycles formally composed of the ketene and the
alicyclic fragment of the iodonium ylide. Compound 5a is,
however, structurally different from those described as major
products of the interaction of 3a with diphenylketene. It
seems reasonable to postulate that the presence of a
A survey of various databases has shown (22) that com-
pound 5a has not been synthesized earlier although
benzofuranones and their derivatives have been the subject
of extensive research for decades. Moreover, a recent search
of the CA on-line registry file has revealed that the central
2 This point was raised by one of the reviewers.
© 2001 NRC Canada

652
Can. J. Chem. Vol. 79, 2001
Table 2. NMR spectroscopic data of compound 5a.
1H NMR
13C NMRa
Carbon
No.
δ (ppm)
2JH,H (Hz)
δ (ppm)
1JC,C (Hz)
2 or 1′
3
3a
4
5
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
13.6
107.0
181.4
108.5
202.6
39.5
33.4
27.9
27.9
51.5
J2,3 = 34, J2,2′ ≈ 34, J2,6′ ≈ 34
J3,2 = 32, J3,3a = 61
J3a,3 = 61, J3a,4 ≈ 61, J3a,7a ≈ 61
J4,3a = 64, J4,5 = 44
J5,4 = 44, J5,6 = 33
J6,5 = 33, J6,61 = 36, J6,62 ≈ 36, J6,7 ≈ 33
J61,6 = 36
J62,6 = 36
2.91 (2H) (s)
—
1.18 (3H) (s)
1.18 (3H) (s)
2.38 (2H) (s)
—
6
61
62
7
7a
2′
3′
J7,6 = 33, J7,7a = 38
J7a,7 = 41, J7a,3a = 59
J2′,2 = 35, J2′,3′ = 42
189.7
191.1
50.0
—
2.56 (1H),
3.26 (1H)
—
0.96 (3H) (s)
1.30 (3H) (s)
2.56 (1H),
3.26 (1H)
—
J3′,2′ = 42, J3′,4′ = 32
4′
—
—
—
13.6
30.2
26.6
29.8
50.0
J4′,3′ = 32, J4′,5′ = 32, J4′,4′1 = 35
J4′,4′ = 36
J4′2,4′ not determined
J5′,4′ = 32, J5′,6′ = 42
4′1
4′2
5′
6 ′
—
191.1
J6′,5′ = 42, J6′,2 = 35
aSpectra were recorded at 500–125 MHz (DRX-500, Bruker) at room temperature, 400 mg of 5a in 2 mL of CDCl3 with
1
internal TMS as reference; JC,C coupling constants were measured using the one-dimensional INADEQUATE method.
Table 4. Selected bond lengths (Å) and angles
(°) for 5a.
Table 3. Crystallographic data and data collection parameters for
compound 5a.
Bond length (Å)
C(3)—O(3)
Empirical formula
Formula weight
C17H20O5
304.33
1.209(1)
1.215(2)
1.356(2)
1.449(2)
1.462(2)
1.203(2)
1.205(2)
1.574(2)
1.430(1)
C(4)—O(4)
C(3a)—C(7a)
C(3)—C(3a)
C(3a)—C(4)
C(2′)—O(2′)
C(6′)—O(6′)
C(2)—C(3)
C(2)—O(1)
Bond angle (°)
C(5)-C(4)-C(3a)
C(4)-C(3a)-C(3)
C(3a)-C(3)-C(2)
C(4)-C(3a)-C(7a)
C(7)-C(7a)-O(1)
O(1)-C(2)-C(2′)-O(2′)
O(1)-C(2)-C(6′)-O(6′)
Crystal system
Space group
Unit cell dimensions
a (Å)
b (Å)
c (Å)
β (°)
Volume (Å3)
Monoclinic
P21/c
7.187(1)
21.129(1)
10.636(1)
92.81(1)
1613.2(3)
4
1.253
0.759
648
Z
Density (calcd.) (Mg m–3)
Absorption coefficient, µ (mm–1)
F(000)
114.0(1)
130.8(1)
103.1(1)
120.9(1)
118.3(1)
–18.9(2)
9.7(2)
Theta range for data collection (°) 4.18 ≤ θ ≤ 74.86
Index ranges
0 ≤ h ≤ 8; 0 ≤ k ≤ 26;
–13 ≤ l ≤ 13
3564
0.966
4
Reflections collected
Completeness to 2θ
Decay (%)
Independent reflections
Reflections I > 2 σ(I)
Data, restraints, parameters
Goodness-of-fit on F2
Final R indices [I > 2σ(I)]
R indices (all data)
Max and mean shift–esd
Extinction coefficient
Largest difference peak and hole
(e Å–3)
3197 [R(int) = 0.0093]
2637
3197, 0, 205
palladium catalyst induces a reaction path different from that
followed in the thermal interaction.
1.067
Further support for the intermediacy of 4 was obtained
when carbonylation of 3a was accomplished in the presence
of 2-amino-4-methyl-6-methylthiotriazine. The solid mate-
rial isolated upon carbonylation could be identified as amide
7. It is believed that 7 is formed by trapping the primarily
formed ketene 4 by the aminotriazine reactant to give an N-
substituted-1,3-cyclohexanedione-2-carboxamide derivative,
which is rearranged into its enol form.
R1 = 0.0338, wR2 = 0.0958
R1 = 0.0453, wR2 = 0.1001
0.099, 0.007
0.0031(6)
0.176 and –0.121
© 2001 NRC Canada

Besenyei et al.
653
Fig. 2. NMR relaxation times (T1, s) of the carbon atoms of 5a
measured at 125 MHz and room temperature in 0.2 M CDCl3
solution. Underlined numbers represent the 13C NMR chemical
shifts in ppm.
Fig. 1. Molecular structure of the spirobenzofuranone derivative
5a. Atomic displacement parameters represent 50% probabilities.
Our preliminary experiments with phenyliodonium
methylide 3b, synthesized from dibenzoylmethane and
(diacetoxyiodo)benzene, have demonstrated that the catalytic
reaction is accompanied by a pressure drop indicating the in-
corporation of carbon monoxide from the gas phase. The
isolated product has proved to contain two major compo-
nents, one of them having a molecular mass of 472.1311
amu (±5 ppm). The expected product (5b), formally com-
posed of two dibenzoyl methanide moieties and one mole-
cule of carbon monoxide, corresponds to the formula
C31H20O5 (theoretical molecular mass: 472.1311 amu).
volume of 45 mL. NMR spectra were recorded on Bruker
DRX-500 and Varian Unity Inova (400 MHz) spectrometers.
IR spectroscopic investigations were carried out on a Magna
750 (Nicolet) fourier-transform spectrometer equipped with
a DTGS detector at a resolution of 4 cm–1. Raman shifts
were measured in the range of 4000–100 cm–1 on a Nicolet
System 950 FT Raman spectrometer using Nd:YAG laser.
Mass spectroscopic analyses were performed on a VG ZAB-
2SEQ instrument.
X-ray diffraction studies
Intensity data were collected on an Enraf–Nonius CAD4
diffractometer (graphite monochromator, Cu Kα radiation,
λ = 1.54180 Å) at 293(2) K using ω–2θ scans. An experi-
mental (ψ-scan) absorption correction (25) was applied to
the data (the min and max transmission factors were 0.896
and 0.998, respectively). The structure was solved by direct
methods (26) (and subsequent difference syntheses). The
structure was refined on F2 (27). Hydrogen atomic positions
were calculated from assumed geometries. Hydrogen atoms
were included in structure factor calculations but they were
not refined. The isotropic displacement parameters of the hy-
drogen atoms were approximated from the Ueq value of the
atom they were bonded to. Data collection and refinement
data are shown in Table 3.
This strongly suggests that rxn. [3] takes place with 3b as
well but a combination of crystallographic and nonroutine
spectroscopic methods should be applied to obtain unques-
tionable proof for the connectivity of the quaternary carbon
atoms. Further studies on catalytic carbonylation of
iodonium ylides are in progress in our laboratory.
Carbonylation of iodonium ylide 3a
1.84 g (5.4 mmol) of ylide 3a and 0.1 g of [PdCl2(PhCN)2]
were charged in a pressure vessel equipped with a glass lin-
ing and 10 mL of dried (stored over activated 3 Å molecular
sieve) CH2Cl2 was added. The reactor was flushed and pres-
surized to 32 bar with CO. The reaction mixture was stirred
at 20°C until the pressure drop ceased. Evaporation of the
solvent and precipitation with ethanol resulted in a grayish
powder (414 mg), which was redissolved in dichloromethane,
filtered through a Celite™ layer and precipitated again with
EtOH. Yield: 355 mg (43%). Anal. calcd. for C17H20O5:
C 67.09, H 6.62; found: C 67.11, H 6.52.
[4]
Experimental
Physical data of compound 5a
Iodonium ylides 3a and b were prepared according to de-
scribed procedures (24). Catalytic carbonylations were car-
ried out in a pressure vessel (Parr Model 4712) with a total
EI-MS (70 eV) m/z (%) (MF = C17H20O5, molecular ion =
304.1448): 304 (M+, 38), 289 (55), 276 (11), 261 (15), 207
(15), 193 (64), 83 (100). IR (cm–1) (17 mg of 5a in 1 mL of
© 2001 NRC Canada

654
Can. J. Chem. Vol. 79, 2001
CH2Cl2, KBr cell, d = 0.016 cm): 1757 (vw) and 1732 (vs)
(νsCO and νasCO of 2′ and 6′ carbonyls), 1717 (s) and 1676
(m) (νCO of carbonyl groups at positions 3 and 4), 1592
(νC=C). Raman-IR (measured on the same solution, cm–1):
1755 (s) (ρ = 0.19) and 1733 (m) (ρ = 0.72) (νsCO and νasCO
of 2′ and 6′ carbonyls), 1719 (m) (ρ = 0.47) and 1677 (s) (ρ =
0.60) (νCO of carbonyl groups at 3 and 4), 1594 (s) (ρ =
0.22) (νC=C); (ρ denotes the depolarization ratio). NMR
data are shown in Table 2.
Hungarian Research Fund (OTKA Grants T 25 870 and
T 25 967).
References
1. F. Paul. Coord. Chem. Rev. 203, 269 (2000).
2. F. Ragaini and S. Cenini. J. Mol. Catal. A: Chem. 109, 1
(1996).
3. G. Besenyei, S. Németh, and L.I. Simándi. Angew. Chem. Int.
Ed. Engl. 29, 1147 (1990).
4. G. Besenyei and L.I. Simándi. Tetrahedron Lett. 34, 2839
(1993).
5. G. Besenyei, S. Németh, and L.I. Simándi. Tetrahedron Lett.
35, 9609 (1994).
Carbonylation of iodonium ylide 3a in the presence of
2-amino-4-methyl-6-methylthiotriazine
2.02 g (5.9 mmol) of iodonium ylide 3a and 0.95 g
(6.0 mmol) of 2-amino-4-methyl-6-methylthiotriazine were
carbonylated overnight in 10 mL of CH2Cl2 in the presence
of 96 mg of [PdCl2(PhCN)2]. The solvent was distilled off
under vacuum and the oily residue was treated with diethyl
ether. The white precipitate proved to be an analytically pure
compound (0.74 g, 39%). Based on MS, NMR, and IR data
the product was formulated as N-(4-methyl-6-methylthio-tri-
azin-2-yl)-5,5-dimethyl-1,3-cyclohexanedione-2-carboxamide
(7) observed in its enol form in solution, as shown by NMR
spectra. Anal. calcd. for C14H18N4O3S: C 52.16, H 5.63, N
17.38, S 9.96; found: C 52.45, H 5.87, N 17.46, S 9.85.
6. (a) A.F. Cameron, N.J. Hair, and D.G. Morris. Acta
Crystallogr. Sect. B: Struct. Crystallogr. Cryst. Chem. B30,
221 (1974); (b) F.H. Allen, O. Kennard, D.G. Watson, L.
Brammer, A.G. Orpen, and R. Taylor. J. Chem. Soc. Perkin
Trans. 2, S1 (1987).
7. H.-J. Mai, F. Weller, K. Dehnicke, C. Maichle-Mössmer, and J.
Strähle. Z. Anorg. Allg. Chem. 620, 851 (1994).
8. A. Kálmán, B. Duffin, and Á. Kucsman. Acta Crystallogr.
Sect. B: Struct. Crystallogr. Cryst. Chem. B27, 586 (1971).
9. A. Kálmán, G. Argay, E. Fischer, and G. Rembarz. Acta
Crystallogr. B35, 860 (1979).
10. M. Boucher, D. Macikenas, T. Ren, and J.D. Protasiewicz. J.
Am. Chem. Soc. 119, 9366 (1997).
11. R.L. Cicero, D. Zhao, and J.D. Protasiewicz. Inorg. Chem. 35,
275 (1997).
12. A.K. Mishra, M.M. Olmstead, J.J. Ellison, and P.P. Power.
Inorg. Chem. 34, 3210 (1995).
13. H.W. Roesky, K-L. Weber, U. Seseke, W. Pinkert, M.
Noltemeyer, W. Clegg, and G.M. Sheldrick. J. Chem. Soc.
Dalton Trans. 565 (1985) and refs. therein.
14. H.G. Furmanova, R.G. Gerr, V.P. Krasnov, and V.I. Naddaka.
Zh. Strukt. Khim. 25, 166 (1984).
15. M.M. Olmstead and P.P. Power. Inorg. Chem. 25, 4057 (1986).
16. D.E. Evans, M.M. Faul, and M.T. Biledeau. J. Am. Chem. Soc.
116, 2742 (1994).
17. Y. Hayasi, T. Okada, and M. Kawanisi. Bull. Chem. Soc. Jpn.
43, 2506 (1970).
18. G.F. Koser. In The chemistry of functional groups. Supplement
D. Edited by S. Patai and Z. Rappoport. Wiley, New York.
1983. pp. 721–811.
19. (a) R.-Y. Yang, L.-X. Dai, and C.-G. Chen. J. Chem. Soc.
Chem. Commun. 1487 (1992); (b) S.-Z. Zhu and Q.-Y. Chen.
J. Chem. Soc. Chem. Commun. 1459 (1992); (c) U. Drück and
W. Littke. Acta Crystallogr. Sect. B: Struct. Crystallogr. Cryst.
Chem. B34, 3092 (1978).
Physical data of compound 7
EI-MS (70 eV) m/z (%): 322 (M+, 86), 307 (4), 294 (4),
266 (22), 251 (18), 183 (25), 156 (100), 110 (27). IR
(KBr, cm–1): 3500–2800 (OH, NH, CH), 1661, 1624, 1579,
1
1526, 1479, 1423, 1387, 1276. H NMR (CDCl3) δ: 1.12
(6H, s), 2.44 (2H, s), 2.57 (6H, s, methyl and S-methyl pro-
tons on the triazine ring), 2.59 (2H, s), 12.33 (1H, NH),
16.77 (1H, OH). 13C NMR (CDCl3) δ: 197.3, 194.5, 183.5,
176.4, 170.0, 161.7, 103.2, 51.2, 45.2, 30.9, 28.0 (2C), 25.7,
13.4 (SCH3).
Conclusions
A
survey of available structural data on (N-
arenesulfonyl)imides and the outcome of carbonylation ex-
periments with them and with iodonium ylides have shown
that: (i) the feasibility of two-step catalytic oxidative
carbonylation is not restricted to sulfonamide derivatives;
and (ii) the occurrence of transformation of this kind is asso-
ciated with structural properties of the substrates rather than
with the presence of special functional groups. We may con-
clude as well that other ylides like (ArSO2)2C=IPh or
(RFSO2)2C=IPh (where RF represents a perfluoroalkyl
group), may also prove suitable starting materials for cata-
lytic carbonylation. The catalytic transformation of
iodonium ylides described here is expected to open a new
route for the preparation of heterocyclic compounds.
20. E. Gudriniece, O. Neiland and G. Vanag. Zh. Obshch. Khim.
27, 2737 (1957).
21. A. Varvoglis. Tetrahedron, 53, 1179 (1997).
22. (a) CA on-line registry file; (b) CSD, Ver. 5.19. April, 2000.
23. G.F. Koser and S.-M. Yu. J. Org. Chem. 40, 1166 (1975).
24. (a) O. Neiland. Zh. Org. Khim. 1, 1858 (1965); (b) O.J.
Neiland and B.J. Karele. Zh. Org. Khim. 7, 1611 (1971).
25. A.C. North, D.C. Philips, and F. Mathews. Acta Crystallogr.
Sect. A: Fundam. Crystallogr. A24, 350 (1968).
26. G.M. Sheldrick. SHELXS-97. Program for crystal structure so-
lution. University of Göttingen, Germany. 1997.
27. G.M. Sheldrick. SHELXL-97. Program for crystal structure re-
finement. University of Göttingen, Germany. 1997.
Acknowledgements
The authors are indebted to Dr. Pál Kolonits, Department
of Organic Chemistry, Technical University, Budapest for
taking the nonroutine NMR spectra and for his invaluable
help in their interpretation. This work was supported by the
© 2001 NRC Canada

This article has been cited by:
1. Anna Antos, Yiannis Elemes, Adonis Michaelides, John, A. Nyxas, Stavroula Skoulika, Lazaros P. Hadjiarapoglou. 2012. The
Question of Electrophilic vs Nucleophilic Addition of Cyclic β-Dicarbonyl Phenyliodonium Ylides: Electrophilic Cycloaddition
of Diphenylketene. The Journal of Organic Chemistry 77:23, 10949-10954. [CrossRef]