Photochemistry of diketones: Observation of a triplet state-oxygen adduct

Article abstract of DOI:10.1021/ja048174u

The presence of triplet state-oxygen adducts (a species previously proposed but never observed in a direct manner) is readily observed following laser flash photolysis studies of 2,2′-thenil. We report on the kinetic and spectroscopic parameters characteristic of this transient adduct. Copyright

Full text of DOI:10.1021/ja048174u

Published on Web 06/25/2004
Photochemistry of Diketones: Observation of a Triplet State-Oxygen Adduct
Gonzalo Cosa and Juan C. Scaiano*
Department of Chemistry, UniVersity of Ottawa, 10 Marie Curie, Ottawa, Ontario K1N 6N5, Canada
Received March 30, 2004; E-mail: tito@photo.chem.uottawa.ca
Scheme 1 . Photooxidation of Diketones in the Presence of O2 and
An early report by Saltiel et al. reports on the photooxidation of
benzil by oxygen. This process was proposed to be mediated by a
biradical intermediate in the form of an excited R-diketone-oxygen
adduct.1 The sensitized photoepoxidation of olefins in the presence
of benzil or diacetyl and oxygen was also established to occur
through this biradical intermediate,2 within the general framework
of the photosensitized oxygen-transfer mechanism originally sug-
gested by Schenck.3,4
Olefins1,6,9
Later reports (see Scheme 1), based on the 1:2 stoichiometry of
R-diketone (1) consumed to epoxide (7) formed, have established
that the sensitizer-oxygen adduct (3) is only a precursor of the
epoxidating agent and does not play an active role in the olefin (6)
epoxidation. The active species was proposed to be an acylperoxyl
radical (5),5-8 formed from the cleavage of the biradical (3). The
acyl radical (4) that results from this cleavage is trapped by oxygen
producing a second epoxidizing agent.6
Scheme 2 . R-Cleavage vs Triplet-Oxygen Adduct Formation
Several results support the mechanism of Scheme 1: (1) the
reduced singlet oxygen lifetime sensitized by diketones following
laser excitation, explained by quenching of singlet oxygen by
acylperoxyl radicals,10 (2) the enhanced photodecomposition of
R-diketones under O2 vs N2,1,6 and (3) the reduced reactivity of
benzil in the presence of trans-stilbene, an efficient benzyl triplet
quencher.1,6 The observed epoxidation stoichiometry also supports
the proposed mechanism.8,9 However the triplet oxygen-adduct
intermediate (species 3 in Scheme 1) has never been observed.
With the aim of understanding better the photochemistry of
diketones in the presence of O2, we performed laser flash photolysis
(LFP) studies with 2,2′-thenil (TL) (see Scheme 2).
Table 1. Reactivity of 2,2′-Thenil with Various Substrates in
Acetonitrilea
quencher
kq/M-1 s-1d
O2
1.8 × 109
b
tert-stilbenec
1.0 × 1010
1,3-cyclohexadienec
troloxb
1.0 × 1010
8.2 × 109
Following irradiation of TL under nitrogen a transient absorption
spectrum is observed with maxima at 730 and 430 nm and a
shoulder at 530 nm, as well as bleaching with a minimum at 310
nm (data not shown). The reactivity of the transient was examined
in the presence of various quenchers; TL has a triplet energy (ET)
of 228 and 233 kJ/mol (from phosphorescence) in methylcyclo-
hexane and 4:1 methanol/ethanol, respectively. Thus, O2, 1,3-
cyclohexadiene (ET 219 kJ/mol11), and trans-stilbene (ET 208 kJ/
1,4-cyclohexadieneb
naphthaleneb
4.6 × 108
no quenching (,106)
a All the values obtained in acetonitrile; TL concentration in the range
of 9-30 × 10-5 M, laser 355 nm. b Monitored at 720 nm. c Monitored at
at 650 nm. d Estimated error e 10%.
the growth at 360 nm (rate constant of 8.2 × 106 s-1), matches the
triplet decay at 540 nm (rate constant of 7.5 × 106 s-1). The new
transient decays with rate constants of 3.3 × 106 s-1 and 1.0 × 106
s-1 in acetonitrile and toluene, respectively.
In Figure 2-A and 2-B we present the transient absorption time
evolution, monitored at 340 nm, following excitation of TL under
oxygen and air-equilibrated acetonitrile solutions, respectively. In
the air-equilibrated sample, the growth rate constant (3.2 × 106
s-1) is ca. 5-fold smaller than that under oxygen (vide supra); the
decay rate constant (3.3 × 106 s-1) is, however, the same under
both conditions.
The quenching of TL triplet by oxygen in acetonitrile (resulting
in a singlet oxygen quantum yield of 0.47) is not purely photo-
physical; the appearance of the new band centered at 340 nm (vide
supra) indicates the occurrence of a photoreaction. The rate constant
of decay for this newly formed transient (3.3 × 106 s-1) is
independent of the oxygen concentration. This supports assignment
mol11) show quenching rate constants in excess of 109 M-1 s-1
,
while naphthalene (ET 253 kJ/mol,11 i.e., above the TL triplet) shows
no quenching. The quenching rate constants for trolox and for 1,4-
cyclohexadiene, both good hydrogen donors, are fast, supporting
an n,π* triplet. These results (see Table 1) establish the main
transient as the TL triplet state.
The short lifetime of the TL triplet under oxygen also supports
this assignment. In this case, not only the triplet-triplet absorption
is observed following excitation, but also a new band, with
maximum at 340 nm appears concurrently with TL triplet decay in
acetonitrile, as shown in Figure 1. Similar transient spectra are
obtained in toluene (see Supporting Information). The fitting of
these traces shows that the growth rate constant at 340 nm (1.5 ×
107 s-1) has the same value as the triplet decay rate constant
monitored at 420 nm (ca. 1.3 × 107 s-1) (see Supporting
Information). Similar observations were made in toluene, where
9
8636
J. AM. CHEM. SOC. 2004, 126, 8636-8637
10.1021/ja048174u CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
The results show that although the absorption spectra due to both
radicals are similar to that assigned to the triplet-oxygen adduct,
their time evolution differs significantly from that of the latter
intermediate (see Supporting Information). These radicals show half-
lives of at least 4 µs.
Product studies reveal the presence of 2-thiophenecarboxaldehyde
formed following irradiation under nitrogen with Luzchem UVA
lamps. Under oxygen the species formed is 2-thiophenecarboxylic
acid. From these studies we can propose Scheme 2 as the operative
one under the conditions employed, i.e. TL excitation under oxygen
atmosphere. The formation of radicals would mostly be due to the
cleavage of the triplet-oxygen adduct under these conditions.
We also attempted to observe the triplet-oxygen adduct for
benzil and diacetyl, where its involvement has been inferred
from product studies.1,2,6,9,10 Transient absorption spectra were
acquired for both species in acetonitrile under oxygen. There was
no spectroscopic evidence for a transient with characteristics
resembling those for the triplet-oxygen adduct observed in the TL
system (see Supporting Information). Failure to detect 3 in these
cases may be due either to weak absorbances or to short lifetimes.
TL may be particular in that its triplet energy is almost the same
as that for biacetyl (little triplet stabilization), but it bears an
electron-donating moiety.
Figure 1. Transient absorption spectra obtained following 355 nm excitation
of TL 1.21 × 10-4 M in acetonitrile, under O2, recorded (O) 28 ns (triplet
state contribution), (b) 140 ns, (0) 275 ns, (2) 590 ns after the laser pulse.
Figure 2. Transient absorption time evolution monitored at 340 nm,
following 355 nm laser irradiation of TL 9.9 × 10-5 M in acetonitrile, (A)
under O2 atmosphere, and (B) in an air-equilibrated solution.
In summary, our LFP studies have produced compelling evidence
consistent with the formation of a triplet R-diketone-oxygen adduct.
The possibility of monitoring this new transient has enabled us to
establish some of its decay parameters, such as activation energy
and preexponential factor. Its lifetime (particularly at low temper-
ature) is long enough to make additional reactivity studies possible.
of structure 3 in Schemes 1 and 2 to the observed transient as
proposed by Saltiel and later adopted by Bartlett and others.1-4
To gain knowledge on the dynamics of the triplet-oxygen
adduct, we performed temperature-dependence studies on the decay
rate constant of this transient to determine its characteristic
Arrhenius parameters. From this we obtain activation energy values
of 28.2 and 34.3 kJ/mol in acetonitrile and toluene, respectively.
The respective preexponential values obtained are 3.2 × 1011 s-1
and 1.4 × 1012 s-1 for acetonitrile and toluene. This difference is
within the experimental error of the ordinate. These values are
reasonable for a unimolecular rearrangement, most probably the
cleavage of the triplet-oxygen adduct (3) into products (4) and
(5), according to Scheme 1.
Acknowledgment. J.C.S. thanks NSERC (Canada) for generous
support. G.C. thanks the Ontario Graduate Scholarship Program
for a Post-Graduate scholarship.
Supporting Information Available: Experimental details and
kinetic and spectral data. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) Saltiel, J.; Curtis, H. C. Mol. Photochem. 1969, 1, 239-243.
(2) Shimizu, N.; Bartlett, P. D. J. Am. Chem. Soc. 1976, 98, 4193-4200.
(3) Gollnick, K.; Schenck, G. O. Pure Appl. Chem. 1964, 9, 507-525.
(4) Gollnick, K. AdV. Photochem. 1968, 6, 1-122.
(5) Sawaki, Y.; Ogata, Y. J. Am. Chem. Soc. 1981, 103, 2049-2053.
(6) Sawaki, Y.; Foote, C. S. J. Org. Chem. 1983, 48, 4934-4940.
(7) Sawaki, Y.; Ogata, Y. J. Org. Chem. 1984, 49, 3344-3349.
(8) Bartlett, P. D.; Roof, A. A. M.; Shimizu, N. J. Am. Chem. Soc. 1982,
104, 3130-3131.
We also considered two alternative intermediates that could
account for the observed transient spectroscopy: (1) formation of
an acyl radical produced from R-cleavage of the triplet diketone
and (2) formation of an acylperoxyl radical produced following
the reaction of the acyl radical and oxygen. Given the high reactivity
of the acyl radical toward oxygen,12 we can rule out the first
possibility since the decay of the signal assigned to the triplet-
oxygen adduct does not depend on oxygen concentration. The
second possibility is not so easily eliminated. Acyl radicals are
(9) Sawaki, Y. Tetrahedron 1985, 41, 2199-2205.
(10) Darmanyan, A. P.; Foote, C. S.; Jardon, P. J. Phys. Chem. 1995, 99,
11854-11859.
(11) Carmichael, I.; Hug, G. L. In Handbook of Organic Photochemistry;
Scaiano, J. C., Ed.; CRC Press: Boca Raton, Florida, 1989; Vol. I, pp
369-403.
(12) Brown, C. E.; Neville, A. G.; Rayner, D. M.; Ingold, K. U.; Lusztyk, J.
Austr. J. Chem. 1995, 48, 363-379.
trapped by oxygen with a rate constant of 1.8 × 109 M-1 s-1 12
,
(13) Hancock-Chen, T.; Scaiano, J. C. Photochem. Photobiol. 1998, 67, 174-
about the same value we measured for oxygen quenching of TL
triplets. To rule out the acyl-peroxyl radical as responsible for
the observed growth band, experiments were recorded with di-tert-
butyl peroxide and 2-thiophenecarboxaldehyde, as well as with 2,2′-
thenoin. These substrates are convenient sources for acyl and the
acyl-peroxyl radicals under N2 and O2 atmosphere, respectively.13-16
178.
(14) Chatgilialoglu, C.; Lunazzi, L.; Macciantelli, D.; Placucci, G. J. Am. Chem.
Soc. 1984, 106, 5252-5256.
(15) Fouassier, J. P.; Lougnot, D. J.; Bassi, G. L.; Nicora, C. Polym. Commun.
1989, 30, 245-248.
(16) Phan, X. T. J. Radiat. Curing 1986, 13, 11-17.
JA048174U
9
J. AM. CHEM. SOC. VOL. 126, NO. 28, 2004 8637