Hydrotrioxides rather than cyclic tetraoxides (tetraoxolanes) as the primary reaction intermediates in the low-temperature ozonation of aldehydes. The case of benzaldehyde

Article abstract of DOI:10.1021/jo801594n

(Chemical Equation Presented) We demonstrate in this work by theory and experiment that benzaldehyde hydrotrioxide (PhC(O)OOOH), the intermediate most likely formed in the low-temperature ozonation of benzaldehyde, is too unstable to be detected by NMR (¹H, ¹³C, and ¹⁷O) spectroscopy in various organic solvents at temperatures ≥ -80°C and that its previous detection must have been erroneous. Several plausible mechanisms for the formation of this polyoxide were explored by using density functional theory. We found that the formation of the hydrotrioxide involves the facile 1,3-dipolar insertion of ozone into the C-H bond (ΔH^? =11.1 kcal/mol) in a strongly exothermic process (ΔH_R = -57.0 kcal/mol). The hydrotrioxide then quickly decomposes in a second concerted, exothermic reaction involving an intramolecular H transfer to form benzoic acid and singlet oxygen (O₂(¹Δ_g)) (ΔH? = 5.6 kcal/mol), ΔH_R = -14.0 kcal/mol). The equilibrium is thus expected to be shifted toward the products; therefore, this intermediate cannot be observed experimentally. Peroxybenzoic acid, still another major reaction product formed in the ozonation reaction, is formed as a result of the surprising instability of the RC(O)O-OOH bond (ΔH _R = 23.5 kcal/mol), generating HOO^? and benzoyloxyl radicals. Both of these radicals can then initiate the chain autoxidation reaction sequence - the abstraction of a H atom from benzaldehyde to form either a benzoyl radical and HOOH or a benzoyl radical and benzoic acid. Because only very small amounts of HOOH were detected in the decomposition mixtures, the recombination of the benzoyl radical with the HOO^? radical (ΔH_R = -80.7 kcal/mol) appears to be the major source of peroxybenzoic acid. A theoretical investigation of the mechanistic possibility of the involvement of still another intermediate, a cyclic tetraoxide (tetraoxolane) formed as a primary product in the 1,3-dipolar cycloaddition of ozone to the carbonyl group of the aldehyde, revealed that the tetraoxide is a "real" molecular entity with the five-membered ring adopting an envelope conformation. The tetraoxide is destabilized by 7.0 kcal/mol relative to the reactant complex, and the transition state for its formation is 17.4 kcal/mol above the reactant complex, which, although accessible under the reaction conditions, is not expected to be competitive with the reaction generating the hydrotrioxide.

Full text of DOI:10.1021/jo801594n

Hydrotrioxides Rather than Cyclic Tetraoxides (Tetraoxolanes) as
the Primary Reaction Intermediates in the Low-Temperature
Ozonation of Aldehydes. The Case of Benzaldehyde
†
,†
†
,‡
Janez Cerkovnik, Bo zˇ o Plesni cˇ ar,* Jo zˇ e Koller, and Tell Tuttle*
Department of Chemistry, Faculty of Chemistry and Chemical Technology, UniVersity of Ljubljana,
P.O. Box 537, 1000 Ljubljana, SloVenia, and WestCHEM, Department of Pure and Applied Chemistry,
UniVersity of Strathclyde, 295 Cathedral Street, Glasgow G1 1XL, U.K.
bozo.plesnicar@fkkt.uni-lj.si; tell.tuttle@strath.ac.uk
ReceiVed July 21, 2008
We demonstrate in this work by theory and experiment that benzaldehyde hydrotrioxide (PhC(O)OOOH), the
intermediate most likely formed in the low-temperature ozonation of benzaldehyde, is too unstable to be detected
1
13
17
by NMR ( H, C, and O) spectroscopy in various organic solvents at temperatures g -80 °C and that its
previous detection must have been erroneous. Several plausible mechanisms for the formation of this polyoxide
were explored by using density functional theory. We found that the formation of the hydrotrioxide involves
q
the facile 1,3-dipolar insertion of ozone into the C-H bond (∆H ) 11.1 kcal/mol) in a strongly exothermic
process (∆H ) -57.0 kcal/mol). The hydrotrioxide then quickly decomposes in a second concerted, exothermic
R
1
reaction involving an intramolecular H transfer to form benzoic acid and singlet oxygen (O
.6 kcal/mol), ∆H ) -14.0 kcal/mol). The equilibrium is thus expected to be shifted toward the products;
therefore, this intermediate cannot be observed experimentally. Peroxybenzoic acid, still another major reaction
2
( ∆ )) (∆Hq )
g
5
R
product formed in the ozonation reaction, is formed as a result of the surprising instability of the RC(O)O-OOH
•
bond (∆H
R
) 23.5 kcal/mol), generating HOO and benzoyloxyl radicals. Both of these radicals can then
initiate the chain autoxidation reaction sequence—the abstraction of a H atom from benzaldehyde to form
either a benzoyl radical and HOOH or a benzoyl radical and benzoic acid. Because only very small amounts
of HOOH were detected in the decomposition mixtures, the recombination of the benzoyl radical with the
•
HOO radical (∆H
R
) -80.7 kcal/mol) appears to be the major source of peroxybenzoic acid. A theoretical
investigation of the mechanistic possibility of the involvement of still another intermediate, a cyclic tetraoxide
tetraoxolane) formed as a primary product in the 1,3-dipolar cycloaddition of ozone to the carbonyl group of
(
the aldehyde, revealed that the tetraoxide is a “real” molecular entity with the five-membered ring adopting
an envelope conformation. The tetraoxide is destabilized by 7.0 kcal/mol relative to the reactant complex, and
the transition state for its formation is 17.4 kcal/mol above the reactant complex, which, although accessible
under the reaction conditions, is not expected to be competitive with the reaction generating the hydrotrioxide.
Introduction
oxides, ROOOH, in the reaction of C-H bonds in various
saturated compounds with ozone has already been documented,
the mechanistic details and the potential involvement of
hydrotrioxides in the ozonation of aldehydes were much less
explored.
3
Ozonation reactions play an important role in the polluted
atmosphere, wastewater purification, ozone toxicity studies, and
chemical synthesis.
1
-3
Although the involvement of hydrotri-
†
Proton NMR spectroscopic evidence for the existence of an
unusually stable hydrotrioxide (up to +10 °C) believed to be
University of Ljubljana.
University of Strathclyde.
‡
9
6
J. Org. Chem. 2009, 74, 96–101
10.1021/jo801594n CCC: $40.75  2009 American Chemical Society
Published on Web 11/14/2008

Low-Temperature Ozonation of Aldehydes
SCHEME 1. Ozonation of Benzaldehyde
of the obtained stationary points. Intrinsic reaction coordinate (IRC)
29,30
calculations
were used to confirm the connectivity between
minima and associated transition states. Calculated binding energies
have been corrected for basis set superposition error using the
31
counterpoise correction method. The effect of the solvent medium
on the systems investigated was included using the conductor-like
polarizable continuum model (C-PCM) applying the dielectric
3
2,33
constant of acetone (ꢀ ) 20.7).
The decomposition of the hydrotrioxide has been investigated
both as a concerted mechanism as well as a radical decomposition.
1
The concerted mechanisms result in the formation of the O
2
( ∆
g
),
which has singlet diradical character. Due to the deficiencies of
DFT when describing multireference systems such as O ( ∆ ), we
2 g
employed spin projection to provide an accurate description of
the complexes involving O ( ∆ ). At the B3LYP level, this
technique has been shown to provide reliable energetic data for
and is able to reasonably
reproduce (∆E ) 20.9 kcal/mol) the experimental triplet-singlet
4
,5
benzaldehyde hydrotrioxide was reported. We demonstrate
in this work by theory and experiment that benzaldehyde
hydrotrioxide (2, Scheme 1) is too unstable to be detected by
NMR spectroscopy in various organic solvents at temperatures
above -80 °C and that its previous detection must have been
erroneous.
1
3
4
1
2
g
1
35-37
complexes involving O
2
( ∆
g
)
3
8
splitting of 22.5 kcal/mol.
The mechanistic possibility, already proposed before but never
tested, of the involvement of still another intermediate, a cyclic
A recent contribution by Radom et al. assessed the ability of a
variety of density functionals to reproduce the bond dissociation
energies (BDEs) against experimental and high-level ab initio (W1)
6
-8
tetraoxide 3 (tetraoxolane)
formed as a primary product in
the cycloaddition of ozone to the carbonyl group of the aldehyde,
has also been explored.
39
data. We have, therefore, investigated the decomposition reaction
mechanisms that involve the dissociation into radical pairs against
the best-performing functionals identified in their work. These
41
It is known that benzaldehydes react with ozone at -70 to
40
+
10 °C in inert solvents to form the corresponding benzoic acids
include the restricted open-shell (RO)-BMK, the RO-MPWB1K,
42,43
9
,10
as the major products.
Various amounts of peroxybenzoic
and the unrestricted (U)-M05-2X functionals;
the mean absolute
acids were also reported to be among the reaction products.
Ozone is believed to attack the aldehydes as an electrophilic
reagent (Hammett F values of -1.1 to -0.6), and the reactions
are first-order each in ozone and aldehyde. Studies of the
deviation from the W1 (experimental) data across the set of 22
reactions is 0.69 (1.00) kcal/mol, 0.43 (0.74) kcal/mol, and 1.55
(1.33) kcal/mol for the three functionals, respectively. In previous
work, the RO-B3LYP functional was also investigated for the same
39
9
deuterium isotope effect showed relatively low kH/kD values of
(
12) Giamalva, D. H.; Church, D. F.; Pryor, W. A. J. Am. Chem. Soc. 1986,
1
.2-2.0 for these reactions when run with ozone-nitrogen
1
08, 7678.
9
mixtures.
(13) Giamalva, D. H.; Church, D. F.; Pryor, W. A. J. Org. Chem. 1988, 53,
429.
(14) Hellman, T. M.; Hamilton, G. A. J. Am. Chem. Soc. 1974, 96, 1530.
3
Aldehyde hydrotrioxides (ROOOH) were proposed as the
4,5,9-11
initial intermediates in these reactions.
These intermedi-
(
15) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
ates might be formed by either (a) a concerted 1,3-dipolar
insertion of ozone into the C-H bond of the aldehyde or (b)
the abstraction of either the H atom or the hydride ion from
R-H by ozone to first form the radical (R• •OOOH) or ionic
(16) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(
(
(
17) Hertwig, R. H.; Koch, W. Chem. Phys. Lett. 1997, 268, 345.
18) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
19) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys.
Chem. 1994, 98, 11623.
+
-
(20) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
pair (R OOOH), with subsequent collapse to the hydrotri-
(
21) Kohn, W.; Sham, L. J. Phys. ReV. 1965, 140, 1133.
1
2-14
oxide.
(22) Parr, R. G.; Yang, W. T. Density Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989.
(
(
(
23) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54, 724.
24) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
25) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257.
Computational Methods
The mechanism for the ozonation of benzaldehyde was inves-
21,22
(26) Cerkovnik, J.; Tuttle, T.; Kraka, E.; Lendero, N.; Plesni cˇ ar, B.; Cremer,
15-20
tigated using the B3LYP
level of density functional theory
D. J. Am. Chem. Soc. 2006, 128, 4090.
(27) Plesni cˇ ar, B.; Cerkovnik, J.; Tuttle, T.; Kraka, E.; Cremer, D. J. Am.
(
6
DFT) in combination with the double-ꢀ-polarized basis set
2
3-25
Chem. Soc. 2002, 124, 11260.
28) Plesni cˇ ar, B.; Tuttle, T.; Cerkovnik, J.; Koller, J.; Cremer, D. J. Am.
-31G(d,p).
Previous investigations on related ozonation
(
reactions have shown this level of theory to be sufficiently
Chem. Soc. 2003, 125, 11553.
(29) Fukui, K. J. Phys. Chem. 1970, 74, 4161.
(
(
(
(33) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003,
24, 669.
3,26-28
accurate.
All structures were optimized at this level of theory,
30) Fukui, K. Acc. Chem. Res. 1981, 14, 363.
31) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553.
32) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995.
and second derivatives were calculated to characterize the nature
(
1) Oppenl a¨ nder, T. Photochemical Purification of Water and Air: AdVanced
Oxidation Processes (AOPs): Principles, Reaction Mechanisms, Reactor Con-
cepts; Wiley-VCH: Weinheim, Germany, 2003.
(34) Wittbrodt, J. M.; Schlegel, H. B. J. Chem. Phys. 1996, 105, 6574.
(35) Kova cˇ i cˇ , S.; Koller, J.; Cerkovnik, J.; Tuttle, T.; Plesni cˇ ar, B. J. Phys.
Chem. A 2008, 112, 8129.
(36) Reddy, A. R.; Bendikov, M. Chem. Commun. 2006, 1179.
(37) Xu, X.; Muller, R. P.; Goddard, W. A., III. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 3376.
(38) CRC Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC Press
LLC: Boca Raton, FL, 2000.
(39) Menon, A. S.; Wood, G. P. F.; Moran, D.; Radom, L. J. Phys. Chem.
A 2007, 111, 13638.
(40) Boese, A. D.; Martin, J. M. L. J. Chem. Phys. 2004, 121, 3405.
(41) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2004, 108, 6908.
(42) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Phys. 2005, 123,
161103.
(
2) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics: From
Air Pollution to Climate Change; Wiley-Interscience: New York, 2006.
3) Tuttle, T.; Cerkovnik, J.; Plesni cˇ ar, B.; Cremer, D. J. Am. Chem. Soc.
004, 126, 16093, and references cited therein.
4) Stary, F. E.; Emge, D. E.; Murray, R. W. J. Am. Chem. Soc. 1974, 96,
671.
(
2
5
1
(
(
5) Stary, F. E.; Emge, D. E.; Murray, R. W. J. Am. Chem. Soc. 1976, 98,
880.
(
(
(
(
6) Klopman, G.; Joiner, C. M. J. Am. Chem. Soc. 1975, 97, 5287.
7) Nangia, P. S.; Benson, S. W. J. Am. Chem. Soc. 1980, 102, 3105.
8) Cremer, D. Isr. J. Chem. 1983, 23, 72.
9) Erickson, R. E.; Bakalik, D.; Richards, C.; Scanlon, M.; Huddleston, G.
J. Org. Chem. 1966, 31, 461.
(
(
10) White, H. M.; Bailey, P. S. J. Org. Chem. 1965, 30, 3037.
11) Syrov, A. A.; Tsyskovskii, V. K. Zh. Org. Khim. 1970, 6, 1392.
(43) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory Comput. 2006,
2, 364.
J. Org. Chem. Vol. 74, No. 1, 2009 97

Cerkovnik et al.
set of 22 reactions; the mean absolute deviation from the experi-
4
4
mental data for this functional is 1.31 kcal/mol. All structures
were optimized at the B3LYP level of theory with the 6-31G(d,p)
basis set. Single-point calculations were then carried out using the
desired functional in combination with the 6-311+G(3df,2p) basis
4
5
set, which is consistent with the approach employed by Radom
3
9
et al. Harmonic vibrational frequencies computed at the B3LYP/
6
-31G(d,p) level were used to provide zero-point energy and thermal
46
corrections. All calculations were performed in Gaussian 03.
Results and Discussion
Ozonation of Benzaldehyde (Experimental Results). Ozo-
nation of benzaldehyde with an ozone-nitrogen mixture, either
neat at -30 °C or in various solvents (0.1-1.0 M; acetone-d6,
methyl acetate, tert-butyl methyl ether, or methylene chloride-
d2) at -80 °C, produced the corresponding acid and peroxyacid
in a molar ratio of 1:4 ( 2 as the reaction products. The
1
peroxybenzoic acid was characterized by OOH H NMR
1
3
absorption at δ 13.2 ( 0.5 ppm, by CdO C NMR absorption
1
7
47
at 169 ppm, and by O NMR at 317 and 275 ppm. The OOH
1
H NMR signal, previously attributed to PhC(O)OOOH, exactly
matched the OOH H NMR signal of peroxybenzoic acid (as
well as the C and O NMR chemical shifts) under the
conditions we investigated. The corresponding acid was identi-
fied by OH H NMR absorption at δ 12.0 ( 0.5 ppm, by CdO
C NMR absorption at 167 ppm, and by O NMR at 252
1
1
3
17
1
FIGURE 1. Transition state structures and products of the optimized
ozonation mechanisms: (a) TS(1a-2) f 2, (b) TS(1b-2b) f 2b, (c)
TS(1c-3) f 3.
1
3
17
4
8
1
ppm. It is interesting to mention that the H NMR chemical
shifts for the OOH absorption of the peroxyacid and those for
the OH absorption of the acid remained unchanged during the
warming-up (from -80 to +30 °C) and cooling-down proce-
dures. Very small amounts of HOOH were also detected in the
decomposition mixtures. However, no other absorptions that
could be attributed to either the benzaldehyde hydrotrioxide 2
or the cyclic tetraoxide 3 could be detected by NMR (g 2 mol
of the aldehyde were investigated as the primary mechanistic
modes (Scheme 1). Additionally, the possibility of an initial
–
12-14
H• or H abstraction by ozone
with the the subsequent
collapse of the radical (R• •OOOH) or the ionic pair
+
–
(R OOOH) to form the hydrotrioxide 2 (Scheme 1) was
investigated. However, all attempts to locate the TS for an
independent abstraction process were unsuccessful.
1
13
17
%
; H, C, and O NMR) in solutions after the ozonation of
4
9
benzaldehyde 1 at temperatures g -80 °C.
The initial complex formed between ozone and the benzal-
dehyde is not strongly bound. At the B3LYP level, the binding
energy is -0.3 kcal/mol; however, this is most likely an
underestimation of the true binding energy of the two species.
DFT is unable to accurately describe dispersion-bound com-
plexes, and this type of interaction is expected to exist in the
benzaldehyde-ozone complex. Nonetheless, the complex, though
relatively weakly bound, provides a reliable representation of
the starting structure for the ozonation reaction. In the water-
catalyzed version of the reaction to form 2 (vida infra), the
increase in the binding energy is consistent with the formation
of H-bonds between the complex and H2O since DFT is able
to accurately represent such interactions.
1
The OOH H NMR absorption for peroxybenzoic acid
disappeared in the temperature range of -20 to +20 °C,
producing benzoic acid. The observed pseudo-first-order kinetics
and the activation parameters for the disappearance of peroxy-
benzoic acid in neat benzaldehyde (Baeyer-Villiger reaction)
were comparable to those for the decomposition of the purported
4
,5
benzaldehyde hydrotrioxide at approximately the same con-
centration and temperature range interval (see Table S1 in
Supporting Information).
Ozonation of Benzaldehyde (Theoretical Results). The 1,3-
dipolar insertions of ozone into either the C-H or CdO bond
(
44) Henry, D. J.; Parkinson, C. J.; Mayer, P. M.; Radom, L. J. Phys. Chem.
The insertion of ozone into the C-H bond (path a, Scheme
1) results in the formation of the hydrotrioxide of benzaldehyde
A 2001, 105, 6750.
(
(
45) Krishnan, R.; Frisch, M. J.; Pople, J. A. J. Chem. Phys. 1980, 72, 4244.
46) Frisch, M. J. et al. Gaussian 03, version C.01; Gaussian, Inc.:
(
1
2, Figure 1). This reaction occurs with a low barrier (∆Hq )
Wallingford, CT, 2004.
47) Antolini, L.; Benassi, R.; Ghelli, S.; Folli, U.; Sbardellati, S.; Taddei,
F. J. Chem. Soc., Perkin Trans. 2 1992, 1907.
48) Balakrishnan, P.; Baumstark, A. L.; Boykin, D. W. Org. Magn. Reson.
984, 22, 753.
49) A direct spectroscopic detection of O
1.1 kcal/mol, Table 1) and is strongly exothermic (∆HR )
(
-57.0 kcal/mol, Table 1). These results are consistent with a
relatively low kinetic isotope effect observed in the ozonation
of benzaldehyde. The inclusion of solvent effects, via the
(
1
9
(
2
(1∆
g
) in the thermodecomposition
and UV photolysis of benzaldehyde hydrotrioxide at-10 °C in diethyl ether
continuum model, does not have a substantive effect on the
calculated reaction energetics—the activation energy is slightly
increased, and the reaction becomes slightly more exothermic
was reported: Chou, P. T.; Martinez, M. L.; Studer, S. L. Chem. Phys. Lett.
1
1
2 g
990, 174, 46. However, our attempts to detect O ( ∆ ) in the decomposition
of products obtained by the low-temperature ozonation of benzaldehyde (tert-
butyl methyl ether, methyl acetate, acetone;-80 °C; an attempted detection of
,10-dimethylanthracene-9,10-endoperoxide ( Kotani, H.; Ohkubo, K.; Fukuzumi,
(Table 1). This indicates that the polarizing nature of the solvent
9
S. J. Am. Chem. Soc. 2004, 126, 15999.) unambiguously failed. Chou et al. most
has little effect on the energetics of the reaction; however, water
may play an additional explicit role in catalyzing the reaction.
We, therefore, considered the possible role of a water molecule
in the ozonation of benzaldehyde 1.
1
2 g
likely detected O ( ∆ ) during the thermal decomposition of a more stable ether
hydrotrioxide (at-10 °C) formed in the ozonation of benzaldehyde in diethyl
5
0
ether as solvent. Namely, we also detected 9,10-dimethylanthracene-9,10-
endoperoxide in this solvent but, indicatively, not in other solvents investigated.
9
8
J. Org. Chem. Vol. 74, No. 1, 2009

Low-Temperature Ozonation of Aldehydes
a
TABLE 1. Energetics of Ozonation Reactions
8.2 kcal/mol lower in energy than 2 (Figure 2). The intramo-
lecular H transfer has a low activation enthalpy (∆H ) 5.6
q
∆
E
∆H
∆G
∆Gsolv
BE
kcal/mol), and the resulting formation of benzoic acid is
exothermic (∆HR ) -14.0 kcal/mol). As outlined in the
Computational Methods, spin projection was used in the
calculation of the activation and reaction enthalpies for the final
1
a
0
0
11.1
0
15.1
0
13.2
-0.3
TS(1a-2)
2
1
TS(1b-2b)
2
1
14.6
-58.0
0
23.6
-71.0
0
-57.0
-52.6
-61.0
b
0
0
0
-5.1
18.8
-69.7
0
17.4
7.0
23.0
-66.0
0
22.1
11.5
21.0
-73.3
0
17.7
6.0
1
step in this reaction, where O2( ∆g) is formed. There is a small
b
c
-13.8
-1.7
fraction of spin contamination (fsc) in the TS (fsc ) 0.17),
whereas the product complex has a much larger contamination
(fsc ) 0.97, see Supporting Information for details on the
calculation of fsc and spin-projected energies).
TS(1c-3)
3
18.2
6.4
a
Relative energies, enthalpies, and free enthalpies in kcal/mol. ∆Gsolv
values are the relative free energies from the solvent phase single-point
calculations of the gas phase optimized structures. Binding energies
All subsequent barriers for the decomposition of 2 are less
than those required for the initial ozonation reaction, and,
therefore, the peak contribution of the hydrotrioxide is expected
to be small due to the rapid decomposition process. This is also
true for the water-catalyzed decomposition reaction when
compared with the activation enthalpy required to form 2b (see
Supporting Information). Given that the benzoic acid complex
is favored from both a kinetic and thermodynamic perspective
and that the reaction from 1 can occur spontaneously, it is not
surprising that we are not able to detect intermediate 2. The
(
BE) are BSSE-corrected using the counterpoise correction.
In Table 1, the compounds listed as 1b, TS(1b-2b), and 2b
represent the same reaction as indicated by path a in Scheme 1,
but with an additional water molecule (see Figure 1b). The
inclusion of the water molecule did not have a favorable effect
on the activation enthalpy, which was increased by 7.7 kcal/
mol (∆Hq ) 18.8 kcal/mol, Table 1). However, the reaction is
slightly more exothermic due in part to the more favorable
H-bond complex formed in the product, as evidenced by the
increased binding energy of 2b relative to 1b.
1
final complex of benzoic acid and O ( ∆ ) dissociates in a
2
g
barrierless process (∆H ) -1.7 kcal/mol) to further stabilize
the singlet diradical character of O2.
The possibility of inserting the ozone molecule into the CdO
bond (path b, Scheme 1), where the final product is the cyclic
tetraoxide (tetraoxolane 3, Scheme 1), was also considered. The
calculated product for this pathway indicates that 3 is a stable
system, with the five-membered ring adopting an envelope
conformation (Figure 1c). This is the most stable conformation;
however, the reaction is destabilized by 7.0 kcal/mol (∆H, Table
Formation of Peroxybenzoic Acid. Another major product
resulting from the ozonation reaction is peroxybenzoic acid,
which is formed as a result of the surprising instability of the
RC(O)O-OOH bond in 2. As a result of this instability, 2
decomposes into hydroperoxyl and benzoyloxyl radicals (Scheme
2a) rather than into benzoylperoxyl and hydroxyl radicals (as
10,53
suggested previously).
Both of these radicals can then initiate
1
) relative to the reactant complex. The transition state in
the chain autoxidation reaction—the abstraction of a H atom
from benzaldehyde to form either a benzoyl radical and HOOH
(Scheme 2b) or a benzoyl radical and benzoic acid (Scheme
2c). However, because only very small amounts of hydrogen
peroxide were detected in the decomposition mixtures, it appears
that an alternative reaction involving the hydroperoxyl radical
(Scheme 2d) must be an important one. This reaction leads to
one of the major observed products of the ozonation reaction,
peroxybenzoic acid. The peroxyacid may also be formed by an
additional radical reaction (Scheme 2e); benzoyl radicals could
also react with triplet oxygen (formed from singlet oxygen and,
much less likely, present as an impurity in the ozone-nitrogen
stream) to form benzoylperoxyl radicals. The abstraction of a
hydrogen atom from benzaldehyde by the benzoylperoxyl radical
is then able to produce the observed peroxybenzoic acid
(Scheme 2f).
51
forming 3 is 17.4 kcal/mol above the reactant complex, which,
although accessible under the reaction conditions, is not expected
5
2
to be competitive with the reaction generating 2 (for the
calculated IR spectra for 2 and 3, see Supporting Information).
Furthermore, the inclusion of solvent effects via the C-PCM
calculation is unable to lower the barrier to cyclization. The
activation energy in forming 3 is slightly increased to 17.7 kcal/
mol (∆Gsolv, Table 1) when the effect of the polar solvent
(
acetone) is included.
Formation of Benzoic Acid. The concerted decomposition
of the hydrotrioxide results in the formation of benzoic acid
and singlet oxygen (O2( ∆g)). The H transfer can occur
intramolecularly (Figure 2) or in a water-assisted, concerted
mechanism (see Supporting Information). However, in analogy
with the formation of 2, the calculated activation enthalpy (∆Hq
1
)
8.9 kcal/mol) for the decomposition reaction indicates that
We have also considered the addition of ozone to the benzoyl
radical formed in the reactions outlined in Scheme 2b,c,f.
However, we were unable to optimize the RC(O)OOO• inter-
mediate as a stable molecular entity. Such a radical most likely
the presence of a water molecule is unfavorable (Figure S1 in
Supporting Information). Nonetheless, in both cases (with and
without H2O), the reaction is facile.
The ozonation reaction (path a, Scheme 1) results in an open-
chain structure (i.e., the trans form of ROOOH, see Figure 1a);
however, the intramolecular H transfer requires the cis form of
the OOOH chain. The rotation of the trans to cis isomer of 2
proceeds over two moderate barriers, with the final cis product
immediately dissociates into RC(O)O• + O . It is, therefore,
2
unlikely that the reaction is a two-step process, but rather, a
direct O abstraction from ozone may take place to yield the
benzoyloxyl radical that is shown as one of the products in
Scheme 2a.
To assess the equilibrium of the competing reaction mech-
anisms in Scheme 2, we have calculated the relative stabilities
(
50) For ether hydrotrioxides, see: Plesni cˇ ar, B.; Cerkovnik, J.; Tekavec, T.;
Koller, J. Chem.;Eur. J. 2000, 6, 809.
51) Benson et al. (see ref 7) estimated a minimum activation energy of 10.6
kcal/mol for the formation of the cyclic tetraoxide of acetaldehyde.
52) Klopman et al. (see ref 6) reported relatively fast exchange between
O-enriched isobutyraldehyde and ozone in a reversible equilibrium (via cyclic
tetraoxide). Our preliminary experiments showed no exchange between benzal-
(
(53) The bond dissociation enthalpy for the RC(O)OO-OH bond was
calculated in the same manner and has a value of 37.4 kcal/mol. For comparison,
the bond dissociation enthalpy for the HO-OOH bond was calculated to be
31.1 kcal/mol (31.7 ( 1.4 kcal/mol; Shum, L. G. S.; Benson, S. W. J. Phys.
Chem. 1983, 87, 3479).
(
8
1
1
8
dehyde and O-labeled ozone at-70 ( 10 °C.
J. Org. Chem. Vol. 74, No. 1, 2009 99

Cerkovnik et al.
FIGURE 2. Energy profile for the concerted reaction of ozone and benzaldehyde to form benzoic acid and singlet oxygen (O
2
(1∆
g
)).
SCHEME 2. Radical Decomposition Reactions of 2
TABLE 2. Relative Enthalpies of Radical Reactions Described in
a
Scheme 2
RO-B3LYP RO-BMK RO-MPWB1K U-M05-2X average
∆
HR
18.8
7.4
24.7
11.8
23.9
9.6
26.6
10.2
24
10
(
Scheme 2a)
∆
∆
∆
∆
∆
∆
∆
Hq
Scheme 2b)
HR
Scheme 2b)
HR
Scheme 2c)
(
5.3
5.4
6.0
5.7
6
(
-16.3
-76.2
-32.7
3.7
-21.6
-81.8
-37.9
7.6
-19.4
-82.4
-36.3
5.4
-24.1
-82.6
-35.2
6.0
-20
-81
-36
6
(
HR
(
Scheme 2d)
HR
(
Scheme 2e)
Hq
(
Scheme 2f)
HR
-4.6
-5.1
-5.3
-4.9
-5
(
Scheme 2f)
a
Results are reported for each functional as relative enthalpies
(
∆H(298)) in kcal/mol.
The initial dissociation reaction of 2 to form the benzoyloxyl
radical and the hydroperoxyl radical (Scheme 2a) requires an
and activation barriers (where applicable) for these mechanisms.
The results from the single-point calculations using the density
functionals outlined in the Computational Methods are reported
in Table 2. The RO-BMK and RO-MPWB1K methods are
generally found to slightly overestimate the BDE relative to
the W1 results, whereas RO-B3LYP and U-M05-2X generally
enthalpy of ca. 24 kcal/mol, which is the lowest BDE (RO-OOH)
5
3
value reported of any hydrotrioxide studied until now. The
dissociation enthalpies predicted by the four methods span a
range of ca. 8 kcal/mol, with the two less-reliable functionals
(RO-B3LYP and U-M05-2X) predicting the highest and lowest
values. The average value obtained is slightly less than the RO-
BMK and RO-MPWB1K results, which were found by Radom
et al. to be the most accurate functionals and generally to slightly
39,44
underestimate the BDE relative to the W1 results.
Therefore,
we have taken a simple average of the four results for each
reaction, and it is these average values that are discussed below.
1
00 J. Org. Chem. Vol. 74, No. 1, 2009

Low-Temperature Ozonation of Aldehydes
3
9
overestimated the BDEs. Although the dissociation enthalpy
is relatively high, the formation of 2 is a strongly exothermic
process, and as such, there is sufficient energy in the system to
allow the dissociation to occur. Moreover, the entropic gain from
the dissociation is also a driving force behind this reaction. Thus,
based on the low BDE and the favorable entropic effects, the
RC(O)O-OOH bond should dissociate spontaneously under the
of the hydrotrioxide to form a benzoyloxyl radical and a
hydroperoxyl radical is accessible as a result of the exothermicity
of the ozonation reaction and is driven by the entropic gain
resulting from dissociation. The resulting benzoyloxyl radical
is able to react with benzaldehyde in an exothermic reaction to
generate the benzoyl radical that, in turn, reacts in a strongly
exothermic reaction with the hydroperoxyl radical to form the
observed peroxybenzoic acid. The formation of peroxybenzoic
acid through the reaction of O2 with benzoyl radicals to generate
benzoylperoxyl radicals and the subsequent abstraction of a H
atom from benzaldehyde was also investigated and found to be
a viable additional mechanism by which peroxybenzoic acid is
produced. Both radical mechanisms will be operative and
essentially diffusion-controlled; that is, the reaction with O2 or
hydroperoxyl radicals requires diffusion out of the solvent cage
after the initial decomposition and hydrogen abstraction from
benzaldehyde. Thus, although the reaction between the benzoyl
and hydroperoxyl radicals is more exothermic, we expect both
radical mechanisms to be competitive and irreversible.
Although the cyclic tetraoxide 3 is a stable molecular entity,
the transition state for its formation is too high to compete with
the reaction generating the hydrotrioxide 2. However, both
intermediates, the hydrotrioxide 2 and the cyclic tetraoxide 3,
should, at least potentially, be detected by IR matrix spectro-
scopic techniques.
5
3
reaction conditions.
The formation of hydrogen peroxide via Scheme 2b is limited
due to the endothermicity of the reaction. The barrier to the
abstraction of the H atom from benzaldehyde by the hydrop-
eroxyl radical is moderate (∆Hq ≈ 10 kcal/mol), however, the
endothermicity of the reaction implies that the reverse reaction
has a relatively low barrier of ca. 4 kcal/mol. Thus, the
equilibrium for such an abstraction reaction would be shifted
toward the reactants. On the other hand, the benzoyloxyl radical
resulting from the initial dissociation is able to easily abstract
a H atom from benzaldehyde (Scheme 2c) in a strongly
5
4
exothermic reaction (∆HR ≈ -20 kcal/mol), generating
benzoic acid and the benzoyl radical.
The unreacted hydroperoxyl radical from the initial dissocia-
tion is able to react with the benzoyl radical formed in Scheme
2
c in a barrierless, strongly exothermic reaction (∆HR ≈ -80
kcal/mol), resulting in the observed peroxybenzoic acid. The
unfavorable equilibrium in the reaction outlined in Scheme 2b
and the strong exothermicity of the recombination reaction of
Scheme 2d are consistent with the observation that only very
small amounts of hydrogen peroxide were detected in the
decomposition mixtures.
Experimental Section
Ozonations were performed by using ozone-nitrogen mixtures
that were prepared by ozone absorption on silica gel at -78 °C
and blowing with nitrogen gas. Ozone was released from silica gel
We also considered a second competitive mechanism that may
also contribute to the formation of peroxybenzoic acid;the
benzoyl radicals reacting with triplet oxygen in a strongly
exothermic reaction (∆HR ≈ -36 kcal/mol) to form benzoylp-
eroxyl radicals (Scheme 2e). The abstraction of a hydrogen atom
from benzaldehyde by the benzoylperoxyl radical to generate
peroxybenzoic acid occurs with only a moderate barrier (∆Hq
3
,11
by raising the temperature of the bath
and delivered either into
the neat benzaldehyde or into a solution of the aldehyde in the
appropriate solvent (for details, see Supporting Information).
Reaction products (i.e., benzoic, peroxybenzoic acids, and
hydrogen peroxide) formed in the ozonation of benzaldehyde were
1
13
17
determined by H, C, and O NMR spectroscopy using reference
materials. Reaction products were also methylated using a
diazomethane-diethyl ether solution and analyzed by GC/MS. In
this way, benzoic acid was detected as the methyl ester.
≈
6 kcal/mol) in a slightly exothermic reaction (∆HR ≈ -5
kcal/mol, Scheme 2f).
The kinetics of the disappearance of peroxybenzoic acid was
1
measured by following the decay of the OOH signal by H NMR
Conclusions
using TMS as the internal standard. Kinetic and activation
parameters for the first-order reaction were obtained by standard
The formation of the hydrotrioxide of benzaldehyde through
ozonation was found to occur in a facile mechanism involving
the 1,3-dipolar insertion of ozone into the C-H bond. The
inclusion of solvent effects via a continuum solvent model had
a minor destabilizing effect on the reaction enthalpies, and the
presence of an explicit water molecule was also found to be
mildly destabilizing. The hydrotrioxide is formed in an exo-
thermic reaction; however, this intermediate can then decompose
in a second concerted, exothermic mechanism to form benzoic
3
procedures.
WARNING. On several occasions, we encountered violent
explosive) decompositions of mixtures obtained during the low-
temperature ozonation (-80 °C) of either neat benzaldehyde or
solutions of benzaldehyde with an ozone-nitrogen mixture.
(
Acknowledgment. The authors thank the Slovenian Research
Agency for the financial support (J1-9410) and Dr. J. Plavec
Slovenian NMR Centre, National Institute of Chemistry,
(
1
acid and singlet oxygen (O2( ∆g)). Given the low barrier and
17
Ljubljana) for running O NMR spectra.
the exothermicity of the latter phase of the reaction, the
equilibrium is expected to be shifted strongly toward the
products of the reaction, and the intermediate is not observed
experimentally under conditions investigated in this work.
Peroxybenzoic acid was formed from the ozonation reaction
as a result of the instability of the RC(O)O-OOH bond and
subsequent autoxidation reaction sequence. The decomposition
Supporting Information Available: General experimental
methods, experimental kinetic data, calculations of spin-
projected energies, the effect of explicit water molecule in the
ozonation of 1, the calculated IR spectra for 2 and 3, and the
calculated geometries and energies of all structures (in xyz
format). This material is available free of charge via the Internet
at http://pubs.acs.org.
(
54) Attempts to locate a TS for this reaction were unsuccessful because the
H abstraction occurs in a barrierless process.
JO801594N
J. Org. Chem. Vol. 74, No. 1, 2009 101