Gas-Phase Reactions of 1,2-Dimethylcyclopentene and of 2,6-Heptanedione with Ozone: Unprecedented Formation of an Ozonide by Ozone Treatment of a Diketone

Article abstract of DOI:10.1002/(SICI)1099-0690(199804)1998:4<627::AID-EJOC627>3.0.CO;2-N

Gas-phase ozonizations of 1,2-dimethylcyclopentene (1) and of 2,6-heptanedione (5) afforded in each case dimethylcyclopentene ozonide (2) in low yields. In the ozonization of 1, diketone 5 was formed as the single major product, along with nine "abnormal" ozonolysis products which were formed by oxidative cleavage of carbon-carbon single bonds.

Full text of DOI:10.1002/(SICI)1099-0690(199804)1998:4<627::AID-EJOC627>3.0.CO;2-N

FULL PAPER
Gas-Phase Reactions of 1,2-Dimethylcyclopentene and of 2,6-Heptanedione
with Ozone: Unprecedented Formation of an Ozonide by Ozone Treatment of
a Diketone
Karl Griesbaum*, Vasile Miclaus[°], In Chan Jung[°°], and Ralf-Olaf Quinkert
Engler-Bunte-Institut, Bereich Petrochemie, Universität Karlsruhe (TH),
D-76128 Karlsruhe, Germany
Received September 25, 1997
Keywords: Gas-phase chemistry / Ozonolysis / Ozonide / Cycloolefin / Diketone
Gas-phase ozonizations of 1,2-dimethylcyclopentene (1) and
diketone 5 was formed as the single major product, along
of 2,6-heptanedione (5) afforded in each case dimethylcyclo- with nine “abnormal” ozonolysis products which were for-
pentene ozonide (2) in low yields. In the ozonization of 1,
med by oxidative cleavage of carbonϪcarbon single bonds.
Introduction
dence: a) The GC peak for ozonide 2 disappeared when
the reaction mixture was treated with triphenylphosphane
(TPP). b) From the combined crude products of several
ozonolysis reactions of 1, the acids 9Ϫ12 were extracted
with aqueous NaHCO3 and identified on the basis of diag-
Gas-phase ozonolyses of olefins are of interest in connec-
tion with the role of such reactions in atmospheric chemis-
try. A large variety of olefins, in particular terpenes[1]
,
which are emitted to the atmosphere from coniferous trees,
have been ozonized, and several mechanisms have been pro-
posed for the course of such reactions[2][3][4]. It is a common
feature of all results reported that, in contrast to ozonolyses
of olefins in solution, peroxidic compounds were either
found as minor products or not at all. In particular, ozon-
ides, which are often the major products in liquid-phase
ozonolyses, have for a long time only been detected in gas-
phase ozonolyses of small acyclic olefins[5], and more re-
1
nostic signals in the H- and/or 13C-NMR spectra of the
mixture of acids. c) The neutral products 3Ϫ8 were isolated
or enriched to the extent that they could be unequivocally
1
assigned on the basis of their H- and/or 13C-NMR data.
cently of cyclic terpenes having exocyclic double bonds[6]
.
By contrast, bicyclic ozonides derived from ozonolyses of
cyclic olefins have not been reported, although a number of
terpenes having such structural units have been ozonized.
Since it is known, that 1,2-dimethylcyclopentene (1) pro-
vides the corresponding ozonide 2 in good yields in liquid-
phase ozonolyses[7], we have now examined the ozonolysis
of 1 in the gas phase.
Results and Discussion
Treatment of 1 with 0.8 molar equivalents of ozone at
room temperature in the gas phase gave a product mixture
in which, aside from residual 1, compounds 2 (0.5%), 3
(5%), 4 (1%), 5 (51%), 6 (2%), 7 (3%), 8 (9%), 9 (2%), 10
(5%) and 11 (5%) have been detected by GC analysis in the
proportions reported. In addition, the gas chromatogram
showed the presence of trace amounts, each, of ca. 20 un-
It is generally agreed that gas-phase ozonolyses of olefins
occur by the Criegee mechanism, though with the differ-
ence, that a diradical- rather than a zwitterion-type of car-
bonyl oxide is produced. Hence, ozonolysis of 1 should pro-
vide intermediate 13, and ozonide 2 could arise from intra-
molecular cycloaddition of the latter. It was, furthermore,
reported that “hot” radical-type carbonyl oxides may be
isomerized to dioxiranes, which can undergo several sub-
sequent reactions, including epoxidations and rearrange-
ments to give acids or esters, depending on the nature of the
1
known compounds, and H- as well as 13C-NMR analysis
showed the presence of 12. The identifications of the above
products are based on coinjections with authentic samples,
on GC/MS analyses and on the following additional evi-
[°] On leave from Babes-Bolyai University, Cluj-Napoca, Rouma-
nia.
[°°]
On leave from Hanseo University, Seosan, Korea.
Eur. J. Org. Chem. 1998, 627Ϫ629
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0404Ϫ0627 $ 17.50ϩ.50/0
627

K. Griesbaum, V. Miclaus, I. C. Jung, R.-O. Quinkert
FULL PAPER
substituents[3]. Hence, epoxide 3 could have been formed by Experimental Section
reaction of 1 with ozone and/or with the dioxirane 14, while
esters 6 and 7 could have been formed by rearragement of
14, possibly via intermediate 15. The formation of diol 4
General: The substrates 1[10] and 5[11] and authentic samples of
the not commercially available compounds 2[12], 3[13], 4[14a], 6[14b]
[14c], and 8[15], have been prepared by published procedures and
,
7
can be explained by reaction of 3 with water, which is were of > 98% purity. NMR spectra were obtained in CDCl3 with
TMS as internal reference with a Bruker AC 250 instrument. GC
analyses were performed with a Hewlett-Packard 5890A instru-
ment, using the following conditions: Column: 50 m, fused silica,
phenyl-5%-methylsilicone SE 54; 50°C for 10 min, then 50Ϫ180°C
at 7°C/min; injector temperature 200°C. GC/MS analyses (EI):
Hewlett-Packard system 5985B; GC conditions as above. HPLC
separations: Merck Hitachi chromatograph 655 A 11; column:
LiChrosorb Si 60, 25 cm, pentane/ether, 70:30. All ozonization re-
actions were carried out in a 6-l two-neck round-bottom flask,
equipped with a septum on one neck and a connection to a vacuum
line on the other neck.
known to be formed in gas-phase reactions of ozone with
hydrocarbons. The formation of diketone 5 as the single
major product is in line with the experience, that gas-phase
ozonolyses of olefins lead predominantly to the corre-
sponding carbonyl fragments.
For the formation of the abnormal ozonolysis products
8؊12, having less carbon atoms than substrate 1, we have
assumed the enols of 5 as precursors: Ozonolysis of enol 16
can produce acid 9 and the “hot” carbonyl oxide 17, which
Ϫ as has been reported previously[2] Ϫ can be converted
into formic acid (12). Similarly, ozonolysis of enol 18 can
afford 8, 10 and 11.
Ozonization of 1,2-Dimethylcyclopentene (1): The 6-l flask was
evacuated to 10Ϫ2 Torr at room temp., the vacuum line was closed
and 20 mg[16] (0.21 mmol) of the liquid substrate 1 was injected
through the septum. After 30 min, 170 ml of an O3/O2 mixture,
containing 1 mmol of O3/l, was injected from a gas syringe within
ca. 1 min. After a reaction time of 10 min, the flask was cooled in
an ice/water bath, and subsequently flushed with nitrogen for 20
min to remove volatiles, which were condensed by passing the nitro-
gen stream through a trap kept at Ϫ50°C. The 6-l flask was rinsed
with ether and the combined ether solutions from several experi-
ments were concentrated by distilling off the ether at room temp.
and reduced pressure to leave a liquid residue. GC and GC/MS
analysis showed that the cold trap contained a mixture of 1, 3 and
4, and the liquid residue contained 2 and 5 Ϫ 11 in the proportions
reported above. 1H- and 13C-NMR analysis showed that the liquid
residue contained 12, too. The liquid residue was dissolved in ether
and extracted with an aqueous solution of sodium bicarbonate.
The ether phase was concentrated by evaporation of ether and the
concentrate was separated by HPLC to give the following com-
pounds in the purities given in parentheses: 5 (88%), 6 (76%), 7
(24%) and 8 (12%). The aqueous extract was acidified, extracted
with ether and the ether was distilled off to leave a residue which
according to 1H-NMR analysis contained the following com-
pounds in the proportions given in parentheses: 9 (14%), 10 (34%),
11 (33%) and 12 (19%). Products 2Ϫ12 were assigned based on the
identity of the properties reported below with those of authentic
samples or with data reported in the literature.
To test this hypothesis, we have treated 5 with ozone at
the elevated temperature of 45°C in order to assure gas-
phase conditions. GC analysis showed indeed the formation
of 8 (0.3%), 9 (0.8%), 10 (1%) and 11 (1.3%), and 1H-NMR
analysis showed the presence of 12. To our surprise, how-
ever, the presence of ozonide 2 (0.8%) was also indicated
by GC analysis. This was confirmed by the isolation of 2
1
and its characterization by H-NMR spectroscopy and by
reduction with TPP to give 5.
The unprecedented formation of an ozonide by ozone
treatment of a diketone introduces a new aspect to organic
ozone chemistry and may have implications for the area of
atmospheric ozone chemistry. Since gas-phase reactions of
ozone with hydrocarbons are accompanied by the forma-
tion of OH radicals[8], which can subsequently form H2O2,
it appeared feasible, that ozonide 2 may have been formed
by addition of H2O2 to 5 via 19 as an intermediate. In the
liquid phase, such a reaction of 5 has indeed been realized,
but it was reported that the conversion of 19 into 2 required
the presence of the strong dehydrating reagent P2O5[9]. In
preliminary experiments, we have shown that 2 was indeed
formed from H2O2 and 5 in the gas phase at 45°C. Experi-
ments are underway to test the scope of this reaction with
other di- and also monocarbonyl compounds.
1,5-Dimethyl-6,7,8-trioxabicyclo[3.2.1]octane (2): GC: tR ϭ 16.2
min. Ϫ GC/MS; m/z (%): 144 (2) [M]ϩ, 128 (1) [M Ϫ O]ϩ, 112 (9)
[M Ϫ O2]ϩ, 97 (5), 85 (2), 71 (4), 43 (100).
1,5-Dimethyl-6-oxabicyclo[3.1.0]hexane (3): GC: tR ϭ 9.5 min.
Ϫ GC/MS; m/z (%): 112 (14) [M]ϩ, 97 (13) [M Ϫ CH3]ϩ, 71 (14),
69 (16), 55 (16), 43 (100), 41 (25), 39 (19).
trans-1,2-Dihydroxy-1,2-dimethylcyclopentane (4): GC: tR ϭ 18.6
min. Ϫ GC/MS; m/z (%): 130 (1) [M]ϩ, 112 (22), 97 (25), 71 (22),
69 (14), 58 (17), 57 (13), 55 (11), 43 (100), 41 (13).
2,6-Heptanedione (5): GC: tR ϭ 20.5 min. Ϫ GC/MS; m/z (%):
1
128 (4) [M]ϩ, 95 (5), 85 (4), 71 (9), 58 (17), 43 (100). Ϫ H NMR
(88% purity): δ ϭ 1.83 (quint, J ϭ 7.1 Hz, 2 H), 2.14[17] (s, 6
H), 2.48 (t, J ϭ 7.1 Hz, 4 H). Ϫ 13C NMR: δ ϭ 17.78, 29.86,
42.49, 208.19.
4-Oxopentyl Acetate (6): GC: tR ϭ 21.5 min. Ϫ GC/MS; m/z
Support by the Deutsche Forschungsgemeinschaft, the Korea Sci-
ence and Engineering Foundation and the Internationales Seminar
at the University of Karlsruhe is gratefully acknowledged.
(%): 144 (1) [M]ϩ, 101 (4), 87 (5), 84 (4), 61 (6), 58 (7), 43 (100).
Ϫ
1H NMR (76% purity): δ ϭ 1.93 (quint, J ϭ 6.9 Hz, 2 H),
2.05[17] (s, 3 H), 2.17 (s, 3 H), 2.54 (t, J ϭ 7.2 Hz, 2 H), 4.12 (t,
628
Eur. J. Org. Chem. 1998, 627Ϫ629

Unprecedented Formation of an Ozonide by Ozone Treatment of a Diketone
FULL PAPER
J ϭ 6.4 Hz, 2 H). Ϫ 13C NMR: δ ϭ 20.61, 22.66, 29.64, 39.65,
Reaction of 1,6-Heptanedione (5) with H2O2 in the Gas Phase: A
63.36, 170.66, 207.19. Ϫ Ref.[18] 1H NMR (CDCl3, TMS): δ ϭ few drops of 35% aqueous H2O2 and a few grains of 5 were placed
:
1.90 (t), 2.03 (s), 2.15 (s), 2.53 (t), 4.05 (t).
in a 500-ml round-bottom flask. The flask was cooled with liquid
nitrogen, evacuated to 10Ϫ2 Torr and placed into a water bath kept
at 45°C for 12 h. The flask was removed from the bath, filled with
nitrogen, rinsed with ether and the ether was distilled off at room
temp. and reduced pressure. 1H-NMR analysis showed the presence
of 2 (δ ϭ 1.50, s, CH3) and of 5 (δ ϭ 2.14, s, CH3) in a ratio of
ca. 1:5.
Methyl 5-Oxohexanoate (7): GC: tR ϭ 21.7 min. Ϫ GC/MS;
m/z (%): 144 (2) [M]ϩ, 113 (17), 101 (6), 87 (5), 85 (11), 74 (22), 59
1
(25), 55 (18), 43 (100). Ϫ H NMR (24% purity): δ ϭ 1.89 (quint,
J ϭ 7.2 Hz, 2 H), 2.15 (s, 3 H), 2.35 (t, J ϭ 7.2 Hz, 2 H), 2.52 (t,
J ϭ 7.2 Hz, 2 H), 3.67[17] (s, 3 H). Ϫ 13C NMR: δ ϭ 18.40, 29.18,
32.41, 41.70, 50.79, 172.86, 207.04.
4-Oxopentanal (8): GC: tR ϭ 12.3 min. Ϫ GC/MS; m/z (%): 100
(0.1) [M]ϩ, 72 (14), 57 (7), 43 (100), 29 (26). Ϫ 1H NMR (12%
purity): δ ϭ 2.21[17] (s, 3 H), 2.77 (s, 4 H), 9.81 (s, 1 H). Ϫ 13C
NMR: δ ϭ 29.76, 35.53, 37.45, 200.33, 206.29.
[1]
For a summary of literature see: K. Griesbaum, M. Hilß, J.
Bosch, Tetrahedron 1996, 47, 14813Ϫ14826.
[2]
H. E. OЈNeal, C. Blumstein, Int. J. Chem. Kin. 1973, Vol. V,
397Ϫ413.
[3]
M. C. Dodge, R. R. Arnts, Int. J. Chem. Kin. 1979, Vol. XI,
5-Oxohexanoic Acid (9): GC: tR ϭ 27.5 min. Ϫ GC/MS; m/z (%):
130 (0.3) [M]ϩ, 112 (8), 70 (5), 60 (5), 45 (15), 43 (100), 42 (15), 41
399Ϫ410.
[4]
R. Atkinson, W. P. Carter, Chem. Rev. 1984, 84, 437Ϫ470.
1
(9). Ϫ H NMR (14% purity): δ ϭ 1.90 (quint, J ϭ 7.1 Hz, 2 H),
[5]
P. Neeb, O. Horie, G. K. Moortgat, Tetrahedron Lett. 1996, 37,
2.16[17] (s, 3 H), 2.40 (t, J ϭ 7.2 Hz, 2 H), 2.55 (t, J ϭ 7.2 Hz, 2
H). Ϫ 13C NMR: δ ϭ 18.51, 29.78, 32.88, 42.23, 179.00, 208.24.
9297Ϫ9300, and literature cited therein.
[6]
K. Griesbaum, I. C. Jung, V. Miclaus, Environ. Sci. Technol.,
in print.
[7]
4-Oxopentanoic Acid (10): GC: tR ϭ 23.4 min. Ϫ GC/MS; m/z
R. Criegee, G. Lohaus, Justus Liebigs Ann. Chem. 1953, 583,
(%): 116 (2) [M]ϩ, 101 (2), 73 (3), 56 (18), 55 (8), 45 (12), 43 (100).
12Ϫ36.
[8]
1H NMR (34% purity): δ ϭ 2.21[17] (s, 3 H), 2.60Ϫ2.65 (m, 2
D. Grosjean, E. L. Williams, E. Grosjean, Environ. Sci. Technol.
Ϫ
1993, 27, 830Ϫ840.
H), 2.74Ϫ2.79 (m, 2 H). Ϫ 13C NMR: δ ϭ 27.79, 29.72, 37.71,
[9]
R. Criegee, G. Lohaus, Chem. Ber. 1953, 86, 1Ϫ4.
[10]
178.25, 206.58.
G. Büchi, P.-S. Chu, Tetrahedron 1981, 34, 4509Ϫ4514.
[11]
W. Ried, W. Kunstmann, Chem. Ber. 1967, 100, 605Ϫ610.
Acetic Acid (11): GC: tR ϭ 6.3 min. Ϫ GC/MS; m/z (%): 60 (34)
[12]
R. Criegee, G. Blust, G. Lohaus, Justus Liebigs Ann. Chem.
1
[M]ϩ, 45 (100), 43 (98), 29 (11), 15 (19). Ϫ H NMR (33% purity):
1953, 583, 2Ϫ6.
[13]
δ ϭ 2.10[17] (s). Ϫ 13C NMR: δ ϭ 20.63, 177.85.
L. E. Friedrich, R. A. Fiato, J. Am. Chem. Soc. 1974, 96,
5783Ϫ5787.
Formic Acid (12): 1H NMR (19% purity): δ ϭ 8.05[17] (s). Ϫ 13C
NMR: δ ϭ 165.69.
[14] [14a]
Prepared according to a general procedure in: Organikum,
17th ed., VEB Deutscher Verlag der Wissenschaften, Berlin,
[14b]
1988, p. 258Ϫ259. Ϫ
Prepared according to a general pro-
Ozonization of 2,6-Heptanedione (5): The solid diketone 5 (100
mg; 0.8 mmol) was placed in the 6-l flask, the flask was cooled
with liquid nitrogen, evacuated to 10Ϫ1 Torr and completely sub-
mersed in a water bath, which was kept at 45°C. After 5 was volatil-
ized, 500 ml of a O3/O2 mixture was added as described above, and
the mixture was kept at 45°C for 2 d. Then, the flask was removed
from the bath, filled with nitrogen and rinsed with ether. From the
combined ether solution of several runs, ether was distilled off, the
residue was examined by GC and GC/MS analysis and sub-
cedure in: Organikum, 17th ed., VEB Deutscher Verlag der Wis-
senschaften, Berlin, 1988, p. 405. Ϫ
[14c]
Prepared according to
a general procedure in: Organikum, 17th ed., VEB Deutscher
Verlag der Wissenschaften, Berlin, 1988, p. 423Ϫ424.
K. Griesbaum, G. Kiesel, Chem. Ber. 1989, 122, 145Ϫ150.
In reactions, which were carried out with 40 mg of 1, the mix-
ture became hazy immediately after ozone was injected and,
hence, exclusive gas-phase ozonolysis of 1 could not be assured
at such concentrations.
[15]
[16]
[17]
Signal used for the determination of the proportion of the com-
pound in the mixture of 5؊8 and in the mixture of 9؊12,
respectively.
K. Gollnick, K. Knutzen-Mies, J. Org. Chem. 1991, 56,
4017Ϫ4027.
1
sequently separated by the HPLC method to afford ozonide 2: H
[18]
[19]
NMR (70% purity): δ ϭ 1.50 (s, 6 H), 1.58Ϫ1.90 (m, 5 H),
2.10Ϫ2.30 (m, 1 H)[19]
.
H. Mayr, J. Baran, E. Will, H. Yamakoshi, K. Teshima, M.
Nojima, J. Org. Chem. 1994, 59, 5055Ϫ5058 reported different
data [δ ϭ 1.38 (s, 6 H), 1.5Ϫ1.9 (m, 6 H)], but we have con-
firmed our data with an authentic sample which was prepared
Reduction of 2: A sample of 2 in ca. 0.5 ml of CDCl3 was ad-
mixed with triphenyl phosphane in an NMR tube. 1H-NMR analy-
sis after ca. 30 min showed the presence of 5 as the sole product
of reduction.
by a published procedure[12]
.
[97287]
Eur. J. Org. Chem. 1998, 627Ϫ629
629