Activation of carbon dioxide by electrogenerated superoxide ion: A new carboxylating reagent

Full text of DOI:10.1021/jo9512057

380
J . Org. Chem. 1996, 61, 380-383
mate reaction product was gained; its tetramethylam-
Activa tion of Ca r bon Dioxid e by
monium salt was isolated from a nonelectrochemical
reaction between (Me4N)O2 and CO2, which allowed its
structural characterization. The overall stoichiometry of
the process was established (eq 1),
Electr ogen er a ted Su p er oxid e Ion : A New
Ca r boxyla tin g Rea gen t
Maria Antonietta Casadei, Stefania Cesa, and
Franco Micheletti Moracci*
2CO2 + 2O2•- f C2O62- + O2
(1)
Dipartimento di Studi di Chimica e Tecnologia delle
Sostanze Biologicamente Attive, Universita` di Roma “La
Sapienza”, P.le Aldo Moro 5, I-00185 Roma, Italy
and a reaction mechanism was suggested (eqs 2-4),
•-
involving nucleophilic addition of O2 to CO2 as the
primary step.
Achille Inesi*
•-
O2•- + CO2 f CO4
(2)
(3)
(4)
Dipartimento di Chimica, Ingegneria Chimica e Materiali,
Universita` dell’Aquila, I-67040 Monteluco di Roio,
L’Aquila, Italy
•-
CO4•- + CO2 f C2O6
Marta Feroci
C2O6•- + O2•- f C2O62- + O2
Dipartimento di ICMMPM, Universita` di Roma “La
Sapienza”, Via del Castro Laurenziano 7,
I-00161 Roma, Italy
The formation of CO2•- as an intermediate by electron
transfer from O2 to CO2 was ruled out,7 owing to the
•-
Received J uly 5, 1995
large difference (about 1V) between the reduction poten-
tials of O2 and CO2 in aprotic solvents.8 Hirobe et al.9
investigated the reactivity of the nonelectrochemical
KO2-CO2 system and evidenced its oxidizing power
toward olefins and sulfides which were converted into
oxiranes and sulfoxides, respectively. On the basis of
isotope incorporation using K18O2 as the reagent, a
reaction mechanism involving the [O2•-CO2] complex and
CO2 has a long history of use in organic synthesis
because it is an inexpensive reagent that is available on
a large scale. The thermodynamic stability and relative
kinetic inertness of CO2 require its preliminary activa-
tion, and electrochemical techniques provide some solu-
tions to the problem.1 CO2 has been activated either by
direct (at the electrode) or indirect (through heteroge-
neous or homogeneous catalysis) reduction. Further-
more, examples of both redox2,3 and chemical2,4 homoge-
neous catalysis have been described. Most of the latter
employed transition metals complexes, which can also
promote the insertion of CO2 into suitable organic
substrates.5
•-
CO2 as intermediates in the formation of a peroxycar-
bonate radical anion was suggest.
O2•- + CO2 f [O2•-CO2] f O2 + CO2
(5)
(6)
•-
•-
CO2•- + O2 f CO4
Among the possible activators of CO2, very little
attention has been paid to the superoxide ion, although
superoxide-activated CO2 was suggested to be responsible
for the vitamin K-dependent carboxylation of glutamic
-
-
This would then evolve to HCO4 and/or HC2O6 ions,
which were believed to be the oxidizing species.
We have found that the superoxide ion, generated by
electroreduction of O2 in dipolar aprotic solvents, acti-
vates CO2 to give a carboxylating reagent, which is able
to convert NH-protic acetamides or propanamides bear-
ing a leaving group at the ω position into oxazole or 1,3-
oxazine derivatives, respectively. Carboxamides were
selected as the substrate because we previously showed10
that acetamides of type 1-3 (Scheme 1) on treatment
with an electrogenerated base in the presence of CO2
yield oxazolidine-2,4-diones, a class of biologically active
substances.11 The process was supposed to involve an
intermediate carbamate ion cyclizing to the product via
intramolecular SN2. According to the first of the estab-
lished procedures (method A), the carboxylation process
can be accomplished by reducing a DMF (or MeCN)
solution of O2, CO2, and the substrate (1 or 2) at a
potential value (E ) -1.0 V vs SCE) corresponding to
the first reduction peak of O2. At this potential, neither
acid residues of prothrombin.6 To our knowledge, only
•-
two subsequent reports dealing with the system O2
+
CO2 are present in the literature. Sawyer et al.7 inves-
tigated the reactivity of electrogenerated superoxide ion
toward CO2 in aprotic solvents. Evidence supporting the
2-
formation of peroxydicarbonate ion C2O6 as the ulti-
(1) (a) Aresta, M.; Forti, G. Carbon Dioxide as a Source of Carbon;
Nato Asi Ser. C; Reidel: Dordrecht, 1987; Vol. 206. (b) Sullivan, B. P.;
Krist, K.; Guard, H. E. Electrochemical and Electrocatalytic Reactions
of Carbon Dioxide; Elsevier Science Publishers: Amsterdam, 1993.
(2) The meaning of this term has been stated in the following:
Andrieux, C. P.; Dumas-Bouchiat, J .-M.; Saveant, J .-M. J . Electroanal.
Chem. Interfacial Electrochem. 1978, 87, 39.
(3) (a) See: ref 5a in ref 4a. (b) Isse, A. A.; Gennaro, A.; Severin, M.
G.; Vianello, E. Proceeding of the International Conference on Carbon
Dioxide Utilisation; Bari: Italy, 1993; p 287 and references cited
therein.
(4) (a) Bhugun, I.; Lexa, D.; Saveant, J .-M. J . Am. Chem. Soc. 1994,
116, 5015. (b) Hammouche, M.; Lexa, D.; Momenteau, M.; Saveant,
J .-M. J . Am. Chem. Soc. 1991, 113, 8455 and references cited therein.
(c) Nagao, H.; Mizukawa, T.; Tanaka, K. Inorg. Chem. 1994, 33, 3415.
(d) Collomb-Dunand-Sauthier, M.-N.; Deroutier, A.; Ziessel, R. Inorg.
Chem. 1994, 33, 2961. (e) Smith, C. I.; Crayston, J . A.; Hay, R. W. J .
Chem. Soc., Dalton Trans. 1993, 3267.
(8) Amatore, C.; Saveant, J .-M. J . Am. Chem. Soc. 1981, 103, 5021.
(9) Nagano, T.; Yamamoto, H.; Hirobe, M. J . Am. Chem. Soc. 1990,
112, 3529.
(5) (a) Derien, S.; Clinet, J .-C.; Dun˜ach, E.; Perichon, J . J . Org.
Chem. 1993, 58, 2578 and references cited therein. (b) Tsuda, T. Gazz.
Chim. Ital. 1995, 125, 101 and references cited therein. (c) Taschedda,
P.; Dun˜ach, E.; J . Chem. Soc., Chem. Commun. 1995, 43. (d) Aresta,
M.; Quaranta, E.; Ciccarese, A. J . Mol. Catal. 1987, 41, 355.
(6) Esnouf, M. P.; Green, M. R.; Hill, H. A. O.; Irvine, G. B.; Walter,
S. J . Biochem. J . 1978, 174, 345.
(10) Casadei, M. A.; Cesa, S.; Inesi, A. Tetrahedron 1995, 51, 5891.
(11) (a) 3,5,5-Trimethyl (trimethadione), 5-ethyl-3,5-dimethyl
(paramethadione), and 3-allyl-5-methyl (malidone, 4f) derivatives have
been used as anticonvulsants, in particular in the treatment of absence
seizures.11b More recently, the herbicidal activity of several products
belonging to this class of compounds has been also disclosed. (b) Clark-
Lewis, J . W. Chem. Rev. 1958, 58, 63. Mercier, J . In Anticonvulsants
Drugs. International Encyclopedia of Pharmacology and Therapeutics;
Mercier, J ., Ed.; Pergamon Press: Oxford, 1973; Vol. 1, p 213.
(7) Roberts, J . L., J r.; Calderwood, T. S.; Sawyer, D. T. J . Am. Chem.
Soc. 1984, 106, 4667.
0022-3263/96/1961-0380$12.00/0 © 1996 American Chemical Society

Notes
J . Org. Chem., Vol. 61, No. 1, 1996 381
Ta ble 2. Ca r boxyla tion of Am id es 1-3 by
Electr or ed u ction of O2 in th e P r esen ce of CO2 F ollow ed
by Ad d ition of th e Su bstr a tea
Sch em e 1
products
entry
substrate
nb
(yield, %)c
1
2
3
4
5
6
7
8
9
1a
1a
2a
2c
3a
3c
2i
1.0
1.5
1.0
1.2
1.0
1.2
1.0
1.0
1.0
4a (65), 1a (13)
4a (82)
4a (94)
4c (57)
4a (97)
4c (36)
7 (89)
6 (27), 3g (54)
4h (43), 5 (40)
3g
3h
a
b
DMF-0.1 M TEAP, Hg cathode, E ) -1.0 V. Faraday/mol
of substrate. c HPLC analysis.
increasing the electron-withdrawing ability of the N-
substituent, the yield of 4 sharply decreases, also in the
case where the better leaving group is present (entries
14 and 15). These results are consistent with a reaction
pathway involving a preliminary deprotonation of the
amide NH, followed by carboxylation and cyclization of
the resulting carbamate ion. The question arises whether
the superoxide itself or basic species arising from its
reaction with CO2 are responsible for the deprotonation
of the NH group. Information about the reactivity of the
Ta ble 1. Ca r boxyla tion of Am id es 1 a n d 2 by
Electr or ed u ction of O2 in th e P r esen ce of CO2 a n d th e
Su bstr a te
•-
system O2 + CO2 + substrate can be gained by cyclic
voltammetry experiments (Hg cathode, ν ) 0.2 V s-1).
The addition of 1a to a saturated DMF solution of O2 (c
) 4.8 × 10-3 M) causes an increase of the current value
pertinent to the first reduction peak of O2 which is 40%
of the initial value when [1a ] ) 5 × 10-3 M and 50% when
[1a ] ) 1 × 10-2 M. Under these condition, the second
reduction and the oxidation peaks of O2 as well as the
reduction peak of 1a disappear. Subsequent addition of
1a does not further modify the voltammogram. These
SSEa/
products
entry substrate cathode
E or I
nb
(yield, %)c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1e
2a
2c
2d
2e
2f
A/Hg
B/Hg
A/C
A/C
A/C
-1.0 V
-1.0 V
-1.0 V
1.0 4a (85)
1.0 4a (86)
1.0 4a (87)
10 mA cm-2 0.5 4a (22), 1a (68)
10 mA cm-2 1.0 4a (60), 1a (32)
10 mA cm-2 1.5 4a (70), 1a (17)
10 mA cm-2 2.0 4a (80)
A/C
A/C
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
A/Hg
10 mA cm-2 0.5 4a (48), 1a (46)
10 mA cm-2 1.0 4a (82), 1a (12)
10 mA cm-2 1.5 4a (80)
•-
results indicate that O2 is able to deprotonate 1a
according to the known7 overall eq 7.
10 mA cm-2 2.0 4a (71)
2O•- + 2 HA f H2O2 + O2 + 2A-
(7)
-1.0 V
-1.0 V
-1.0 V
-1.0 V
-1.0 V
-1.0 V
-1.0 V
1.0 4b (88)
1.2 4e (64), 1e (29)
1.0 4a (94)
1.0 4c (71)
1.2 4d (96)
1.2 4e (98)
1.0 4f (93)
As reported,7 the addition of CO2 to a DMF solution of
O2 modifies the voltammogram of the latter; under our
conditions, when [CO2]/[O2] ) 4 the current value of the
first reduction peak of O2 is 5-fold the initial value.
Finally, if both 1a and CO2 are simultaneously added to
the solution of O2 and the molar ratio [1a ]/[CO2] is of the
same order of that used during the electrolyses (1:4), the
current value is 4- to 5-fold the initial value. As a whole,
the voltammetric data suggest that the superoxide
preferentially reacts with CO2 also in the presence of the
substrate. To support this conclusion, the electrolytic
procedure has been modified in that the substrate is
added after the current is switched off and the solution
is bubbled with N2 (method B). Before adding the
substrate, the absence of O2, O2•-, CO2, and, more
generally, any species reducible between 0.0 V and the
discharge of the supporting electrolyte was ascertained
by voltammetry. According to this procedure, the oxazo-
lidine-2,4-diones are formed in yields comparable with
those obtained according to procedure A (Table 2, entries
1-6), thus supporting the starting hypothesis. It should
be noted that by using procedure B, substrates bearing
an electrophore more easily reducible than O2 can also
be carboxylated, greatly increasing the usefulness of the
method.
a
Solvent-supporting electrolyte system: A ) DMF-0.1 M
b
TEAP; B ) MeCN-0.1 M TEAP. Faraday/mol of substrate.
c HPLC or GC analysis.
CO2 nor the substrate undergoes reduction, thus assuring
that the overall process is initiated exclusively by O2
•-
.
The influence of various experimental parameters has
been checked by using chloroacetamide 1a (Table 1,
entries 1-11). Under potentiostatic control, high yields
of oxazolidinedione 4a are obtained independent of the
nature of the cathode and the solvent (entries 1-3). The
nature of the electrode appears to play a role in the
electrolyses carried out under galvanostatic control. In
fact, when reticulated carbon is used as the cathode, the
attainment of high yields of 4a involves the halving of
the current yield (entries 4-7), whereas on a Hg cathode
both chemical and current yields are high (entries 8-11).
As expected, the nature of both the N-substituent and
the nucleofugal influences the yield of 4. The first being
the same, the substitution of chlorine for a tosyloxy group
causes an increase in the yield of 4, which is very sharp
when the cyclization involves a secondary carbon (cf.
entries 14 and 17 vs 1 and 13). On the other hand, by
According to Sawyer,7 the activation of CO2 by super-
2-
oxide would result in the formation of the C2O6 ion as

382 J . Org. Chem., Vol. 61, No. 1, 1996
Notes
CH3CN-H2O (35:65) mixture was used as the eluent in the
analysis of the solutions containing 4a and 4e. In all the other
analyses, a CH3CN-H2O mixture in a linear gradient from 1:9
to 9:1 in 25 min was employed. The flow of the eluent was
the final product which, therefore, would be regarded as
the reagent responsible for the carboxylation carried out
according to procedure B. However, it cannot be excluded
that, when proceeding according to method A, other
always 1 mL min-1
. GC analyses were carried out using a J
•-
intermediates of the reaction between O2 and CO2, as
and W fused silica megabore DB-WAX (30 m) column in the
temperature range 130-150 °C, depending on the nature of the
oxazolidinedione. Quantitative HPLC and GC analyses were
carried out with the internal standard method using authentic
samples of every compounds. N,N-Dimethylformamide (DMF,
Aldrich), acetonitrile (MeCN, Carlo Erba), and tetraethylam-
monium perchlorate (TEAP, Fluka) were purified as already
described.13 The solution of the supporting electrolyte (0.1 M)
in the chosen solvent was percolated on activated alumina just
before using.
Rea gen ts. N-Benzylchloroacetamide (1a ),14 N-allylchloro-
acetamide (1b),15 N-benzyl-2-chloropropanamide (1e),16 N-ben-
zylbromoacetamide (3a ),17 bromoacetanilide (3c),18 N-benzyl-2-
bromo-2-methylpropanamide (3g),19 and N-benzyl-3-bromo-
propanamide (3h )20 were obtained through the reaction between
the opportune chloro(bromo)acyl chloride(bromide) and amine.
N-Benzyl-O-tosylglycolamide (2a ),21 N-phenyl-O-tosylglycola-
mide (2c),21 N-methyl-O-tosyllactamide (2d ),10 N-benzyl-O-to-
syllactamide (2e),10 and N-allyl-O-tosyllactamide (2f)10 were
synthesized by reacting the corresponding bromo amide with
silver toluene-4-sulfonate as previously described.21
those described by Sawyer7 and Hirobe,9 may be involved
in the carboxylation process leading to 4. Indeed, ad-
ditional reaction pathways must be operative during
carboxylation by procedure A since only in this case was
H2O2 detected in the electrolyzed solutions (20-25% with
respect to the number of Faraday consumed, as ascer-
tained by routine titrimetric analyses).12 The simplest
explanation for the formation of H2O2 during procedure
A is that competitive deprotonation of the substrate by
the superoxide can also occur, although the intervention
of other basic species cannot be a priori discarded.
To test the effectiveness and generality of the electro-
chemical procedure of carboxamides carboxylation, a few
other experiments were carried out. 1a was submitted
to carboxylation in a nonelectrochemical system contain-
ing KO2, CO2, and 18-crown-6 ether. If an equimolar
amount of KO2 was used, oxazolidinedione 4a was ob-
tained in 20% yield; values comparable with those
observed in the electrochemical procedures are only
attained when using a molar excess of KO2. Finally, the
electrochemical procedure was extended to other NH-
protic carboxamides. In principle, any of such substrates
bearing a leaving group at the appropriate position
should undergo the carboxylation-cyclization process.
However, the steric hindrance and the length of the chain
(and hence the size of the cycle to be formed) play an
essential role in the course of the reaction (Table 2,
entries 7-9). Apparently, when the cyclization rate of
the intermediate carbamate ion is lowered, competitive
reaction patterns become important, owing to the revers-
ibility of the first carboxylation step.
N-Benzyl-2,2-dimethyl-3-(tosyloxy)propanamide (2i) was ob-
tained with the same procedure as an oil: IR (neat) 3330, 1640,
1600, 1530 cm-1 1H NMR δ 1.22 (s, 6H), 2.44 (s, 3H), 4.01 (s,
;
2H), 4.40 (d, 2H), 6.05 (s, 1H), 7.2-7.4 (m, 7H), 7.73 (d, 2H).
Anal. Calcd for 19H23NO4S: C, 63.17; H, 6.42; N, 3.88.
C
Found: C, 63.37; H, 6.50; N, 3.70.
Electr oca r boxyla tion of 1,2 by r ed u ction of O2 in th e
P r esen ce of Both CO2 a n d th e Su bstr a te (Meth od A). The
controlled-potential electrolyses were carried out at -1.0 V on
solutions of the substrate (2 mmol) in DMF(MeCN)-0.1 M TEAP
(50 mL), where O2 and CO2 were simultaneously bubbling.
Under galvanostatic conditions, the electrolyses were carried out
at
a
current density of 10 mA cm-2
.
At the end of the
electrolysis, the solution was stirred overnight at room temper-
ature.22 A sample of the electrolyzed solution (2 mL) was taken
off for HPLC or GC analyses, and the solvent was removed under
reduced pressure from the remaining solution. The residue was
extracted with Et2O (5 × 30 mL), H2O (100 mL) was added to
the insoluble portion in ether, and the mixture was extracted
with CHCl3 (3 × 50 mL). The extracts were dried (Na2SO4), and
the solvent was evaporated under reduced pressure. The
residues were analyzed by IR, 1H NMR, and TLC and were
combined if they had the same composition. Column chroma-
tography of the mixtures allowed the isolation of oxazolidinedi-
ones 4a -f, which were identified by comparison with authentic
samples.10 The results of the HPLC and GC analyses carried
out on the electrolyzed solutions are reported in Table 1.
Further studies using different classes of substrates
to establish the scope and generality of this new car-
boxylating procedure are in progress.
Con clu sion s
A new mild, safe carboxylating reagent is available
from the electrochemical reduction of O2 in the presence
of CO2. When applied to NH-protic carboxamides bearing
a leaving group at the carbon atom adjacent to the
carbonyl, the procedure affords oxazolidine-2,4-diones in
high to excellent yields. The substrate can be added to
the solution at the beginning of the electrolysis or, taking
advantage of the stability of the reagent, after the current
is switched off, thus making the carboxylation process
independent of the presence of electrophores in the
substrate, which are even more easily reducible than O2.
Electr oca r boxyla tion of 1-3 by Red u ction of O2 in th e
P r esen ce of CO2, F ollow ed by Ad d ition of th e Su bstr a te
(Meth od B). The controlled-potential electrolyses were carried
out at -1.0 V on DMF-0.1 M TEAP solutions (50 mL), where
O2 and CO2 were simultaneously bubbling. At the end of the
electrolysis, N2 was flowing through the solution for 5 min, the
substrate (2 mmol) was added, and the solution was stirred
overnight at room temperature.22 The mixture was worked up
as described above. The residue from the reaction of 2i was
Exp er im en ta l Section
Gen er a l. The electrochemical apparatus, the cells, and the
reference electrode as well as the IR, HPLC and GC instruments
were described elsewhere.13 The values of the working potential
are given relative to SCE. Column chromatography (cc) was
performed on Merck silica gel 70-230 mesh; 1H NMR spectra
were acquired using an AC 200 Bruker spectrometer and Me4-
Si as the internal standard. HPLC analyses were carried out
using a Merck Hibar LiChrocart (250-4; 5 µm) RP-18 column; a
(14) J acobs, W. A.; Heidelberger, H. J . Biol. Chem. 1915, 20, 685.
(15) Harries, C.; Petersen, J . Ber. Dtsch. Chem. Ges. 1910, 43, 634.
(16) Kusher, S.; Cassel, R. I.; Morton, J ., II, Williams, J . H. J . Org.
Chem. 1951, 16, 1283.
(17) Huisgen, R.; Reimlinger, H. Liebigs Ann. Chem. 1956, 599, 161.
(18) Abenius, P. W. J . Prakt. Chem. 1889, 40, 428.
(19) Bischoff, C. A. Ber. Dtsch. Chem. Ges. 1898, 31, 3236.
(20) Chiavarelli, S.; Marini-Bettolo, G. B. Gazz. Chim. Ital. 1951,
81, 89.
(21) Casadei, M. A.; Cesa, S.; Inesi, A.; Micheletti Moracci, F. New
J . Chem. 1994, 18, 915.
(22) The solutions were analyzed after a 12 h delay to ensure that,
independent of the nature of the substrate, the reaction goes to
completion.
(12) Permanganate and iodimetric titrations have been also used
in ref 7 to establish the peroxide content of (Me4N)2C2O6.
(13) Casadei, M. A.; Inesi, A.; J ugelt, W.; Micheletti Moracci, F. J .
Chem. Soc., Perkin Trans. 1 1992, 2001.

Notes
J . Org. Chem., Vol. 61, No. 1, 1996 383
1-benzyl-3,3-dimethyl-2-azetidinone (7).23 The mixture from 3g
was resolved by cc (petroleum ether-acetone (7:3) as eluent) into
the starting amide and N-benzylmethacrylamide (6).24 Column
chromatography of the mixture from 3h (chloroform-acetone
(9:1) as eluent) gave 3-benzyl-3,4,5,6-tetrahydro-2H-1,3-oxazine-
2,4-dione (4h )25 and N-benzylacrylamide (5).26 The results of
the HPLC analyses carried out on the electrolyzed solutions are
reported in Table 2.
Ca r boxyla tion of 1a w ith KO2-CO2. To a mixture of KO2
(30 mg, 0.42 mmol) and 18-crown-6 (100 mg, 0.39 mmol) in DMF
(10 mL) in which CO2 was bubbling was added 1a (71 mg, 0.39
mmol). After the mixture was stirred overnight at room tem-
perature, HPLC showed the presence of 4a (20%) and unreacted
1a (68%). Upon an increase in the amount of KO2 (120 mg, 1.7
mmol) and 18-crown-6 (200 mg, 0.78 mmol), 4a was formed in
80% yield.
Ack n ow led gm en t. This work was supported by
research grants from MURST (1994, 60%) and CNR,
Rome, Italy. The authors wish to thank Miss Sonia
Renzetti for typing the manuscript.
(23) Okawara, T.; Matsuda, T.; Furukawa, M. Chem. Pharm. Bull.
1982, 30, 1225.
(24) Berzuglyi, V. D.; Alekseeva, T. A.; Kruglyak, C. P. Ukr. Khim.
Zh. 1964, 31, 500; C. A. 1965, 63, 8172 d.
(25) Yogo, M.; Hirota, K.; Senda, S. J . Heterocycl. Chem. 1981, 18,
1095.
(26) Bose, A. K.; Manhas, M. S. J . Org. Chem. 1962, 27, 1244.
J O9512057