Stereoselective O2-induced photoisomerization of all-trans-1,6-diphenyl-1,3,5-hexatriene

Article abstract of DOI:10.1139/v03-082

Irradiation of all-trans-1,6-diphenyl-1,3,5-hexatriene (ttt-DPH) in degassed acetonitrile (AN) gives ctt- and tct-DPH, relatively inefficiently, mainly via isomerization in the singlet excited state. The triplet contribution to the photoisomerization is small due to a very low intersystem crossing yield (φ_is = 0.01). Central bond isomerization is quenched in the presence of air by a factor of 1.4, consistent with the expected quenching of the lowest singlet and triplet excited states by oxygen. However, the presence of air enhances terminal bond photoisomerization by nearly twofold. Triplet-sensitized ttt-DPH photoisomerization favors tct-DPH formation and is quenched by oxygen. It follows that the interaction of singlet-excited ttt-DPH with O₂ suppresses isomerization to tct-DPH but opens a new isomerization pathway to ctt-DPH. The presence of dimethylfuran, a singlet O₂ trap, has no effect on the photoisomerization, eliminating the possible involvement of singlet O₂ in this new reaction, ttt-DPH radical cations are ruled out as intermediates because the presence of fumaronitrile, which leads to their formation, suppresses both central and terminal bond photoisomerizations. In contrast to acetonitrile, ctt-DPH formation is quenched by oxygen in methylcyclohexane, suggesting the requirement of a polar environment. Strikingly different deuterium isotope effects distinguish the direct and O₂-induced photoisomerization pathways. A comparative study of ttt-DPH-d₀ with ttt-DPH-d ₂ and ttt-DPH-d₄, involving deuteration of one and both terminal double bonds, reveals an inverse kinetic isotope effect (k. _H^ox/k_D^ox = 0.92) for the O ₂-induced reaction. An attractive mechanism for the new oxygen-induced photoisomerization involves charge transfer from the S ₁ state of ttt-DPH to oxygen followed by collapse of the exciplex to either a zwitterionic or a biradicaloid species through bonding at one of the benzylic positions. Rotation about the new single bond in this intermediate followed by reversion to DPH and O₂ gives the observed result.

Full text of DOI:10.1139/v03-082

673
Stereoselective O2-induced photoisomerization of
all-trans-1,6-diphenyl-1,3,5-hexatriene
Jack Saltiel, Govindarajan Krishnamoorthy, Zhennian Huang, Dong-Hoon Ko,
and Shujun Wang
Abstract: Irradiation of all-trans-1,6-diphenyl-1,3,5-hexatriene (ttt-DPH) in degassed acetonitrile (AN) gives ctt- and
tct-DPH, relatively inefficiently, mainly via isomerization in the singlet excited state. The triplet contribution to the
photoisomerization is small due to a very low intersystem crossing yield (φ = 0.01). Central bond isomerization is
is
quenched in the presence of air by a factor of 1.4, consistent with the expected quenching of the lowest singlet and
triplet excited states by oxygen. However, the presence of air enhances terminal bond photoisomerization by nearly
twofold. Triplet-sensitized ttt-DPH photoisomerization favors tct-DPH formation and is quenched by oxygen. It follows
that the interaction of singlet-excited ttt-DPH with O2 suppresses isomerization to tct-DPH but opens a new isomeriza-
tion pathway to ctt-DPH. The presence of dimethylfuran, a singlet O2 trap, has no effect on the photoisomerization,
eliminating the possible involvement of singlet O2 in this new reaction. ttt-DPH radical cations are ruled out as inter-
mediates because the presence of fumaronitrile, which leads to their formation, suppresses both central and terminal
bond photoisomerizations. In contrast to acetonitrile, ctt-DPH formation is quenched by oxygen in methylcyclohexane,
suggesting the requirement of a polar environment. Strikingly different deuterium isotope effects distinguish the direct
and O2-induced photoisomerization pathways. A comparative study of ttt-DPH-d0 with ttt-DPH-d2 and ttt-DPH-d4, in-
ox
volving deuteration of one and both terminal double bonds, reveals an inverse kinetic isotope effect (kHox/kD = 0.92)
for the O2-induced reaction. An attractive mechanism for the new oxygen-induced photoisomerization involves charge
transfer from the S1 state of ttt-DPH to oxygen followed by collapse of the exciplex to either a zwitterionic or a
biradicaloid species through bonding at one of the benzylic positions. Rotation about the new single bond in this inter-
mediate followed by reversion to DPH and O2 gives the observed result.
Key words: diphenylhexatrienes, trans-cis photoisomerization, oxygen sensit6iz7a9tion.
Résumé : L’irradiation du 1,6-diphénylhexa-1,3,5-triène complètement trans (ttt-DPH), dans de l’acétonitrile dégazé
(AN) conduit d’une façon relativement inefficace à la formation des ctt- et tct-DPH, principalement par le biais d’une
isomérisation dans l’état excité singulet. La contribution triplet à la photoisomérisation est faible en raison du très
faible rendement du passage intersystème (φ = 0,01). L’isomérisation de la liaison centrale est piégée en présence
is
d’air par un facteur de 1,4 qui est en accord le piégeage par l’oxygène des états excités singulet et triplet les plus bas.
Toutefois, la présence d’air augmente la photoisomérisation de la liaison terminale par un facteur égal pratiquement à
deux. La photoisomérisation du ttt-DPH sensibilisée au niveau triplet favorise la formation de tct-DPH et elle est
piégée par l’oxygène. Il en découle que l’interaction de l’oxygène avec le ttt-DPH excité à l’état singulet supprime
l’isomérisation en tct-DPH mais qu’elle conduit à une nouvelle voie d’isomérisation conduisant au ctt-DPH. La pré-
sence de diméthylfurane, un piège pour O2 à l’état singulet, n’a pas d’effet sur la photoisomérisation ce qui élimine la
possibilité d’implication du O2 à l’état singulet dans cette nouvelle réaction. On a éliminé la possibilité d’une implica-
tion des cations radicaux ttt-DPH comme intermédiaires sur la base du fait que la présence de fumaronitrile qui conduit
à leur formation supprime les photoisomérisations des liaisons tant terminale que centrale. Par opposition à ce qui est
observé avec l’acétonitrile, la formation de ctt-DPH est piégée par l’oxygène dans le méthylcyclohexane ce qui suggère
la nécessité d’un environnement polaire. Des effets isotopiques du deutérium très différents distinguent les voies de
photoisomérisations directe et induite par O2. Une étude comparative des ttt-DPH-d0, ttt-DPH-d2 et ttt-DPH-d4 impli-
quant la deutération d’une et des deux doubles liaisons terminales permet de mettre en évidence un effet isotopique
ox
cinétique inverse, kHox/kD = 0,92, pour la réaction induite par O2. Un mécanisme intéressant pour la nouvelle photoi-
somérisation induite par l’oxygène implique un transfert de charge de l’état S1 du ttt-DPH vers l’oxygène, suivi d’une
décomposition de l’exciplex en espèces soit zwitterionique ou biradicaloïde par le biais de la formation d’une liaison à
l’une des positions benzyliques. Une rotation autour de la nouvelle liaison simple de cette intermédiaire, suivie d’une
Received 24 February 2003. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 17 June 2003.
Dedicated to Professor Don Arnold for his contributions to chemistry.
J. Saltiel,1 G. Krishnamoorthy, Z. Huang, D.-H. Ko, and S. Wang. Department of Chemistry and Biochemistry, Florida State
University, Tallahassee, FL 32306-4390, U.S.A.
1Corresponding author (e-mail: saltiel@chem.fsu.edu).
Can. J. Chem. 81: 673–679 (2003)
doi: 10.1139/V03-082
© 2003 NRC Canada

674
Can. J. Chem. Vol. 81, 2003
réversion en DPH et en O2, conduit au résultat observé.
Mots clés : diphénylhexatriènes, photoisomérisation trans-cis, sensibilisation par l’oxygène.
[Traduit par la Rédaction] Saltiel et al.
using a known procedure (15). Lithium perchlorate (Spec-
trum, anhydrous ACS reagent) was used as received. The
syntheses of 1,2-dideuterio- (ttt-DPH-d2, 99.9%, HPLC) and
1,2,5,6-tetradeuterio-all-trans-1,6-diphenyl-1,3,5-hexatriene
(ttt-DPH-d4, 98.0%, HPLC, 1.7% tcc-DPH-d4, major con-
taminant) have been described (10). Prior to use in this
study, ttt-DPH-d4 was recrystallized three times to a final
purity of 99.4% with 0.5% tcc-DPH-d4, the major contami-
nant.
Introduction
all-trans-1,6-Diphenyl-1,3,5-hexatriene (ttt-DPH) has
been used as a model for the long polyenes related to the
retinols and vitamin A. It is the first member of the α,ω-
diphenylpolyene vinylogous family whose all-s-trans confor-
mer exhibits fluorescence both from the initially excited 1 1Bu
state and from a lower energy singlet excited state that, fol-
lowing Hudson and Kohler, is identified as the forbidden
1
doubly excited 2 Ag state (1–3). The lowest singlet excited
states of all longer diphenylpolyenes (1–3) and of all conju-
gated polyenes starting with 1,3-butadiene (4–7), have also
1
been identified as 2 Ag states. In degassed solutions the
trans → cis photoisomerization of ttt-DPH following direct
excitation occurs inefficiently by a combination of singlet
and triplet pathways (8). In the triplet state, the ground state
isomers of DPH exist as a mixture of freely equilibrating
conformers in which ttt, tct, and ctt geometries dominate (9).
In the absence of oxygen, triplet excitation transfer events
between relatively long-lived DPH triplets and DPH ground
states of different geometry are links in quantum chain
processes that enhance the contribution of triplets to the
photoisomerization, compensating somewhat for the low
intersystem crossing yields. The quenching of the triplets by
oxygen significantly decreases their lifetime and prevents
their participation in quantum chain events (9).
Changing the solvent from methylcyclohexane (MCH) to
acetonitrile (AN) increases the rates of photoisomerization
in the singlet excited state manifold, the increase being espe-
cially pronounced (20-fold) at the terminal double bonds (8,
10), thereby decreasing the singlet excited state lifetime and
the contribution of the competing decay channels of fluores-
cence (φ = 0.65 and 0.27 at 20°C in MCH (11) and AN
f
(10), respectively) and intersystem crossing (φ = 0.029 and
0.010 at 20°C in MCH (12) and AN (13, 14),isrespectively).
Photoisomerization in the ttt → ctt direction in AN is almost
entirely a singlet excited state process (98%) that is attenu-
ated by a factor of 1.36 on deuteration of the olefinic hydro-
gens of the terminal double bond (10).
We report here the effect of oxygen on the photo-
isomerization of ttt-DPH and its 1,2-dideuterio and 1,2,5,6-
tetradeuterio derivatives upon direct excitation in AN. In
contrast to the central bond whose photoisomerization is at-
tenuated, consistent with oxygen quenching of singlet and
triplet excited states, oxygen enhances photoisomerization at
the terminal double bonds. The mechanism of this new
photoisomerization pathway is addressed in this paper.
Irradiation procedure
Irradiations were carried out in a Moses merry-go-round
(16) apparatus immersed in a thermostated water bath. A
heating coil connected to a thermoregulator (Polyscience
Corporation) was used to control the temperature to 20 ±
0.1°C. The benzophenone-sensitized photoisomerization of
trans-stilbene was used for actinometry (φt→c = 0.55) (17,
18). A Hanovia medium pressure Hg lamp (200 W, Ace
Glass, Inc.) and Corning CS 7-37 and 0-52 filters were used
for excitation at 366 nm. Aliquots of actinometer solutions
(3.0 mL) were pipetted into Pyrex ampoules (13 mm o.d.),
degassed, and flame-sealed at a constriction. Air-saturated
DPH solutions (1.0 × 10–3 M in AN, 3 mL aliquots) were
Experimental section
Materials
Sources and purification procedures for all-trans-1,6-
diphenyl-1,3,5-hexatriene (100.0%, HPLC), benzophenone,
trans-stilbene, anthracene, and acetonitrile have been de-
scribed (8, 10). Acetonitrile was used as received or purified
© 2003 NRC Canada

Saltiel et al.
675
Table 1. PSS fractions and photoisomerization quantum yields
for ttt-DPH-dn in air-saturated AN (366 nm, 20.0°C).a
irradiated in 13 mm o.d. Pyrex ampoules lightly stoppered
with Teflon-tape-covered rubber stoppers. In some instances,
the ampoules were provided with fritted disks at the bottom
for bubbling air or oxygen throughout the irradiation. Sam-
ple preparation, degassing, and analysis were performed in
almost complete darkness (red light).
DPH
ctt + ttc
tct
ttt
PSS fractions
d0
0.430(0.239)
0.419(0.206)
0.395(0.183)
0.095(0.106)
0.100(0.106)
0.109(0.115)
φtct
0.400(0.635)
0.401(0.667)
0.402(0.661)
d2
Analytical procedures
d4
Actinometer solutions were analyzed by GLC and DPH
solutions by HPLC (λ mon = 350 nm) as previously described
(8, 9), except that on replacing AN with hexanes prior to
HPLC analyses, a stream of argon was used to evaporate the
solvent and care was exercised to avoid taking samples to
dryness (10). This procedure minimizes ctt-DPH losses.
Quantum yields
φ
ctt+ ttc
d0
d2
d4
0.121 (0.062)
0.119 (0.054)
0.114 (0.046)
0.025 (0.035)
0.026 (0.036)
0.028 (0.038)
aValues in parentheses are corresponding values under degassed
conditions from ref. (10); estimated uncertainties in absolute quantum
yields are ±5%.
Fluorescence measurements
Fluorescence spectra were measured with a Hitachi/
PerkinElmer MPF-2A fluorescence spectrophotometer as
previously described (19), or with a Hitachi F4500 fluori-
meter. Fluorescence spectra were recorded for both air-
saturated and Ar-bubbled solutions of DPH in AN in stan-
dard 1 cm2 quartz cells. Temperature was maintained at
20.0 ± 0.1°C using a Neslab-RTE 4DD constant temperature
circulation bath, and was monitored continuously during
each scan with an Omega Engineering Model 199 RTD digi-
tal thermometer in a reference cell placed in the same con-
stant temperature cell holder. Fluorescence lifetimes were
measured with a phase modulation Fluorolog-τ2 lifetime
spectrofluorometer (SPEX) equipped with a 450 W Xe arc
source, a Hamamatsu R928P photomultiplier tube, and a
Lasermetrics BNC1072FW Pockel cell (KD2PO4). The
Fluorolog-τ2 modulates the frequency of the excitation light
from 0.5 to 300 MHz. A glycogen solution (τ = 0 ns, scat-
tered light) was used as reference. Temperature was main-
tained at 20.0 ± 0.1°C and was monitored continuously
during each scan as described above. Air-saturated ttt-DPH
solutions (2.0 × 10–6 M) were employed. Software (DMF
3000F Spectroscopy) provided by the manufacturer was
used to analyze the data. Determinations of the quality of the
lifetimes were based on examination of the statistics of the
fit (a plot of the residual deviations vs. frequency) and re-
duced χ 2 values that were all close to unity.
1.2 × 10–3 M diethylamine (DEA) to 1.00 × 10–3 M ttt-DPH-
d0 in degassed or air-saturated AN solutions did not alter the
photoisomerization quantum yields.
Fluorescence measurements
The fluorescence of 1.0 × 10–3 M ttt-DPH-d0 in degassed
and air-saturated AN was found to be independent of [DEA]
up to 1.5 × 10–3 M. Fluorescence lifetimes at different tem-
peratures for air-saturated solutions are compared in Table 2
with those for degassed solutions (10). Excellent fits to the
single exponential decay model were obtained (χ 2 ≤ 1.2) and
the lifetimes were independent of excitation and monitoring
wavelengths, within experimental uncertainty. Reproducibi-
lity in independent measurements was better than ±0.1 ns.
Discussion
Background
The ground state of molecular oxygen is generally a very
effective quencher of electronically excited molecules (20).
Rate constants for oxygen quenching of singlet excited states
of aromatic hydrocarbons in hydrocarbon solvents are often
close to the diffusion-controlled limit. Corresponding values
for triplet state quenching are usually an order of magnitude
smaller owing to restrictions by spin statistical factors (20–
22). These quenching events are commonly associated with
formation of the lowest singlet state of the oxygen molecule
O2*(1∆g), sometimes through the intermediacy of the higher
Results
+
energy state O2*(1Σg ). In the case of singlet excited states of
Photochemical observations
Photostationary states were approached from the ttt-DPH-
dn (n = 2 and 4) side. Degassed DPH samples (1.0 × 10–3 M)
in air-saturated AN were irradiated for different time inter-
vals at 20.0°C until HPLC analysis revealed no further
change in the isomer composition. Photostationary state
fractions are shown in Table 1. Overall conversions for
quantum yield measurements were kept below 8% and were
corrected for back reaction as has been described (10).
Quantum yields are also shown in Table 1. The effect of ad-
ditives was studied for ttt-DPH-d0. The presence of up to
0.40 M LiClO4 had no effect on ctt/tct conversion ratios for
both air-saturated and degassed AN solutions. The presence
of 0.188 M 2,5-dimethylfuran (DMF) had no significant ef-
aromatic hydrocarbons, induced intersystem crossing has
been shown to accompany oxygen quenching in hydrocarbon
media (23), a pathway to triplets that is diminished by polar
solvents such as acetonitrile (23–25). The importance of a
charge transfer (CT) quenching pathway by oxygen was first
established by Kikuchi et al. (26) who showed that, for
anthracenecarbonitriles and acridinium salts, fluorescence
quenching rate constants in acetonitrile decrease as the free
energy for full electron transfer (∆GSCT) increases from neg-
ative to positive values. A comprehensive systematic study
of the contribution of CT in the quenching of aromatic hy-
drocarbon singlet and triplet states by oxygen in AN was
described by Abdel-Shafi and Wilkinson (27). With the ex-
ception of tetracene that undergoes significant chemical deg-
radation, the net rate constants for the triplet quenching
fect on photoisomerization quantum yields of 1.38 × 10–3
ttt-DPH-d0 in air-saturated AN. Similarly, addition of up to
M
© 2003 NRC Canada

676
Can. J. Chem. Vol. 81, 2003
Table 2. O2 effect on ttt-DPH-d0 fluorescence
ttt-DPH-dn (n = 0, 2, and 4) by molecular oxygen, a value
close to the diffusion-controlled limit.
lifetimes (366 nm).a
The range of attenuations of φttt → tct values by oxygen is
slightly larger (1.36 to 1.40) than the attenuations of fluores-
cence quantum yields and lifetimes, because in addition to
quenching the singlet excited state, oxygen dramatically
shortens the lifetime of DPH triplets, effectively eliminating
quantum chain contributions to the photoisomerization (9b).
Oxygen interferes with the photoisomerization of the central
bond of ttt-DPH-dn (n = 0, 2, and 4) only in so far as it deac-
tivates ttt-DPH-dn excited singlet (primarily) and triplet
states. Photoisomerization quantum yields in the presence of
FN in degassed and air-saturated AN solutions are consistent
with an O2 induced sixfold (compared with factors of 7 and
10 reported for cyclohexane and methanol, respectively, (12,
T (°C)
τ (Air, ns)
τ (Deg, ns)b
10.0
20.0
30.0
40.0
50.0
3.9
3.5
2.9
2.5
2.2
5.3
4.7
4.1
3.6
3.2
aReproducibility in independent measurements was
better than ±0.1 ns.
bFrom ref. (10).
pathway involving CT were shown to correlate well with
∆GTCT, indicating a transition state with 13.5% CT character
(27). Net rate constants for the analogous process involving
excited singlet states were as high as 1.6 × 1010 M–1 s–1 but
29)) enhancement of φ (34). In this case, however, this is a
is
minor perturbation because it only raises φ from 0.01 to
is
0.06.
CT
could not be determined over a sufficiently large ∆GS
The results in Table 1 reveal a strikingly different O2 ef-
fect on terminal bond photoisomerization. Instead of the ex-
pected ~1.3-fold attenuation, φctt values are enhanced by
factors of 2 to 2.5, progressively increasing with terminal
bond deuteration. Correspondingly, the presence of air
strongly shifts photostationary state (PSS) compositions to
larger ctt-DPH contributions at the expense of ttt-DPH. Oxy-
gen enhanced trans to cis 1,2-diarylethene photoisomeriza-
tion quantum yields have previously been observed, but were
generally attributed to O2 induced intersystem crossing (35–
38). For instance, in the case of trans-4,4′-diphenylstilbene,
quenching of the highly fluorescent singlet excited state by
~1 × 10–2 M O2 in methylcyclohexane increases φt → c 18-
fold from 0.0025 to 0.045. However, because DPH triplets in
AN favor decay to tct-DPH over decay to ctt-DPH by at
least a factor of 5 (9a), O2-catalyzed intersystem crossing
would be expected to increase φtct more than φctt . Instead,
we observe enhancement of the latter and quenching of the
former. We conclude that the differential O2 effects on cen-
tral and terminal bond photoisomerization rule out the O2-in-
duced intersystem crossing pathway in ttt-DPH.
In the Background section, we reviewed evidence for the
formation of radical cations (2ttt-DPH·+) in AN solutions.
We recently evaluated and ruled out their possible involve-
ment in the photoisomerization of ttt-DPH following direct
excitation in degassed AN (8). Briefly, fluorescence and
photoisomerization quantum yields, measured as a function
of fumaronitrile (FN) concentration, show that 2ttt-DPH·+
radical cations formed from the 1(ttt-DPH–FN)* exciplex
(13) undergo no photoisomerization (8). It is not surprising,
therefore, that we find that addition of up to 0.015 M
diethylamine (DEA) has no effect on the photoisomerization
range to establish a meaningful logarithmic dependence on
this quantity (27). Except in the case of 2,6-
dimethoxynaphthalene (DMN) for which the quenching
2
event is accompanied by a small yield of DMN·+ (3 × 10–3)
(25), attempts to detect the radical cations of the donors and
·–
2O2 have failed, despite the exothermicity of full electron
transfer in AN for most of the excited singlet donors studied
(25, 27).
Energy transfer from the triplet state of ttt-DPH to O2 in
air-saturated solutions yields O2*(1∆g) with unit efficiency
(28). Quenching of the singlet state of ttt-DPH leads to
O2*(1∆g) less efficiently and the efficiency decreases with
increasing solvent polarity (0.5 and 0.3 in cyclohexane and
methanol, respectively) (29). Competing with O2*(1∆g) for-
mation is O2-induced intersystem crossing leading to as
much as a 10-fold increase in φ (0.015 to 0.15 for degassed
is
vs. air-saturated methanol (12)).
Two-photon laser excitation of ttt-DPH gives the radical
cation (2ttt-DPH·+) in a variety of solvents (30, 31). In AN
the lifetime of the radical cation (7.3 µs) is insensitive to the
presence of air (31, 32). The radical cation is also observed
1
on quenching of ttt-DPH* by electron acceptors in polar
solvents including AN (12, 13). It isomerizes to a new tran-
sient, assigned as a product of trans → cis photoisomeri-
zation, which reverts thermally to 2ttt-DPH·+, and reacts
efficiently with electron donors to give back the starting ttt-
DPH (31).
Oxygen effects on ttt-DPH photoisomerization
We reported earlier that the fluorescence quantum yields
and lifetimes of ttt-DPH-dn (n = 0, 2, and 4) in AN are
nearly independent of deuterium substitution at the terminal
double bonds (10). It was not surprising, therefore, that oxy-
gen in air-saturated AN solutions at 20°C attenuates fluores-
cence quantum yields and lifetimes of the three ttt-DPH-dn
(n = 0, 2, and 4) by the same 1.31 ± 0.01 factor, within ex-
perimental uncertainty (10). The 1.34 ratio of the lifetimes at
20°C in Table 2 is in reasonable agreement with the quan-
tum yield measurements and the effect is similar at different
temperatures. Use of 1.9 ± 0.2 × 10–3 for the oxygen concen-
tration in air-saturated AN (33) gives kqox = (3.5 ± 0.4) ×
1010 M–1 s–1 for the quenching of the excited singlet states of
quantum yields and, in particular, on the φ :φ ratio in ei-
tct
ther degassed or air-saturated AN. At thectthigh concentra-
2
tions employed, DEA, a known, effective quencher of ttt-
DPH·+ radical cations, kq = 1.05 × 109 M–1 s–1 (31a), should
have intercepted them all, in view of their long lifetime (7.3 µs)
1
(31). In the unlikely event that electron transfer from ttt-
2
DPH* to O2 contributes to the formation of ttt-DPH·+ radi-
cal cations, they are not involved in trans → cis photo-
isomerization.
The possibility that O2 opens a thermal free radical iso-
merization process can be ruled out because the observed
© 2003 NRC Canada

Saltiel et al.
677
Scheme 1.
Table 3. Dissection of quantum yields for terminal bond
photoisomerization into normal and oxygen pathways.
S
ox
ox
DPH
(1 – fox)
φ
φ
kHox/kD
fad
ctt+ ttc
ctt+ ttc
d0
d2
d4
0.762
0.759
0.759
0.047
0.041
0.035
0.074
0.078
0.079
0.62
0.65
0.66
0.90
0.94
(10). Analysis of φctt values in the presence of air is
accomplished by dissecting them into contributions from the
1
normal pathway, involving the fraction of ttt-DPH* not in-
tercepted by oxygen (1 – fox), and the new pathway, involv-
1
ing the fraction of ttt-DPH* that is intercepted by oxygen
(fox) and follows the pathway in Scheme1. Since nearly all
terminal bond isomerization in degassed solution is a singlet
excited state reaction, we expect that, in the presence of air,
the normal pathway will experience the same attenuation as
fluorescence. Substitution of kqox = 3.5 × 1010 M–1 s–1 for the
quenching of the excited singlet states of ttt-DPH-dn (n = 0,
2, and 4), [O2] = 1.9 × 10–3 M, and the singlet excited state
formation of more cis double bonds moves the system up
hill thermodynamically (9b). We are left with addition fol-
lowed by elimination starting from 1ttt-DPH*–O2 interac-
tion, as a viable Schenck-type mechanism (39) for the O2-
induced photoisomerization (Scheme 1). Although the
adduct (adducts for ttt-DPH-d2) in Scheme 1 is shown as a
zwitterionic species because this process is much more pro-
nounced in AN than in MCH,2 the possibility that the adduct
is biradicaloid cannot be ruled out. Bond formation at the
benzylic positions should be favored because it leads to the
more delocalized carbon-centered cation or radical interme-
diate. Rotation about the new single bond followed by loss
of O2 would explain enhanced ctt-DPH formation. In princi-
ple, the intermediate could form either from the interaction
of 1ttt-DPH* with ground state O2 (3Σg) or by the addition of
singlet 1O2* (1∆g) to ground state ttt-DPH. However, if 1O2*
were responsible for enhanced values under direct excitation
0
lifetimes (τf ) (10) in eq. [1] gives the (1 – fox) values listed
in
0
0
[1]
(1 – fox) = kqoxτf [O2]/(1 + kqoxτf [O2])
Table 3. Quantum yields for terminal bond isomerization by
the normal singlet excited state pathway (φctt + ttcS) obtained
by multiplying the values under degassed conditions (given
in parentheses in Table 1) by the (1 – fox) attenuation factors,
were subtracted from observed quantum yields in the pres-
ence of air to obtain φctt + ttcox, the quantum yields for the O2-
induced pathway (Table 3). Assuming that Scheme 1 ap-
plies, the ratio of the rate constant for bond formation at a
benzylic position bearing hydrogen (kHox), to the analogous
rate constant for a benzylic position bearing deuterium
(kDox) is given by
conditions, then O2 would enhance φ values even more un-
ox
ox
kHox/kD = (α D/α H)(φcttd0/φcttd4)(τd4ox/τd0
)
ctt
[2]
der triplet-sensitized conditions because the efficiency of
3
1O2* formation is higher when ttt-DPH* is quenched by
where the α s are decay fractions to ctt-DPH for the
deuterated and undeuterated DPH–O2 adducts, the quantum
yields are the values for the O2-induced pathway from Ta-
ble 3, and the fluorescence lifetimes are in the presence of
air. Taking α D = α H and substituting for the other quantities
in eq. [2] using the data for ttt-DPH-d0 and ttt-DPH-d4 gives
1
3O2 (28, 29). It follows that the involvement of O2* can be
ruled out because we have found no photoisomerization en-
hancement of any kind under triplet-sensitized conditions in
air-saturated AN (9). As a further precaution, we now report
that addition of 0.188 M 2,5-dimethylfuran (DMF), a very
kHox/kD = 0.94. This requires that formation of the DPH–
ox
1
efficient O2* scavenger (28, 41), has no effect on the φctt :
O2 exciplex be somewhat reversible, as no intermolecular
deuterium isotope effect would be expected if exciplex for-
mation were irreversible and product determining. A similar
treatment using the data for ttt-DPH-d0 and ttt-DPH-d2 gives
φtct ratio in air-saturated AN.
Deuterium isotope effect
The small increasing trend in φtct values under both de-
gassed and air-saturated conditions with increasing deute-
rium substitution is consistent with the effect of deuterium
ox
kHox/kD = 0.90. The inverse secondary deuterium isotope
effect is consistent with the expected sp2 to sp3 rehybri-
dization in the transition states leading to the adducts (42,
43).
1
on the lifetime of ttt-DPH* (10). Terminal bond deuterium
substitution does not affect rate constants for central bond
isomerization. In contrast, φctt values decrease strongly in
degassed solutions and show a less-pronounced decrease in
the presence of air. The torsional relaxation rate constants
1
Significant fractions of ttt-DPH*–O2 quenching interac-
tions must follow the adduct pathway to isomerization.
ox
These fractions (fad) can be estimated from fox/(α φctt + ttc
)
1
for dihydro- and dideuterio-terminal bonds in ttt-DPH* are
by assuming equal partitioning of the adducts to cis and
subject to a normal deuterium isotope effect (kH/kD = 1.36)
trans double bonds (α = 0.5). The resulting values are rela-
2 In MCH, the presence of air attenuates tct- and ctt-DPH formation by factors of 0.35 and 0.58, respectively (34, 40).
© 2003 NRC Canada

678
Can. J. Chem. Vol. 81, 2003
12. S.K. Chattopadhyay, P.K. Das, and G.L. Hug. J. Am. Chem.
Soc. 104, 4507 (1982).
tively insensitive to deuterium substitution falling in a nar-
row 0.62 to 0.66 range (Table 3).
13. F.W. Schael. Ph.D. Dissertation, Technische Universität, Carlo-
Wilhelmina zu Braunschweig, Germany, 1995.
14. (a) F.W. Schael and H.-G. Löhmannsröben. Chem. Phys. 206,
193 (1996); (b) F.W. Schael, J. Kühster, and H.-G.
Löhmannsröben. Chem. Phys. 218, 175 (1997). (c) F.W. Schael
and H.-G. Löhmannsröben. J. Photochem. Photobiol. A, 105,
317 (1997).
Trapping the exciplex
Initial indications that the O2-induced photoisomerization
–
pathway is quenched when Li+ClO4 is included in AN (34),
were not confirmed when great care was taken to avoid ctt-
DPH loss in sample preparation for HPLC analysis. Because
no changes in φctt :φtct ratios were observed even when a sat-
15. J.F. O’Donnel, J.T. Ayres, and C.K. Mann. Anal. Chem. 37,
1161 (1965).
–
urated Li+ClO4 solution (~0.4 M) was used in air-saturated
AN, we can conclude that Li+ fails to intercept an ion pair in
this case. This approach has been shown to succeed in sev-
eral instances (44). Apparently, in our case, either charge
transfer is not complete in the exciplex, or the exciplex is
too short-lived. An interesting anomaly is our preliminary
observation that on attempting to increase the contribution of
the O2-induced photoisomerization pathway by increasing
the O2 concentration (bubbling O2 during the irradiation) we
were surprised to find lower φctt :φtct ratios than in the pres-
ence of air. If the mechanism in Scheme 1 is correct, this
result suggests that, at the higher concentration, oxygen
quenches the ttt-DPH–O2 exciplex. While oxygen quenching
of singlet exciplexes is well-known (45), quenching of an
oxciplex is unprecedented and remains to be established.
16. F.G. Moses, R.S.H. Liu, and B.M. Monroe. Mol. Photochem.
1, 245 (1969).
17. H.A. Hammond, D.E. DeMeyer, and J.L.R. Williams. J. Am.
Chem. Soc. 91, 5180 (1969).
18. D. Valentine, Jr. and G.S. Hammond. J. Am. Chem. Soc. 94,
3449 (1972).
19. J. Saltiel, D.F. Sears, Jr., J.-O. Choi, Y.-P. Sun, and D.W.
Eaker. J. Phys. Chem. 98, 35 (1994); 98, 8260 (1994).
20. (a) I.B. Berlman. Handbook of fluorescence spectra of aro-
matic molecules. 2nd ed. Academic Press, New York. 1971;
(b) J. Saltiel and B.W. Atwater. Adv. Photochem. 14, 1 (1988);
(c) F. Wilkinson, W.P. Helman, and A.B. Ross. J. Phys. Chem.
Ref. Data, 22, 113 (1993).
21. (a) G. Porter and M.R. Wright. Discuss. Faraday Soc. 27, 18
(1959); (b) B. Stevens and B.E. Algar. Ann. N. Y. Acad. Sci.
171, 50 (1970).
Acknowledgments
22. O.L.J. Gijzeman, F. Kaufman, and G. Porter. J. Chem. Soc.,
Faraday Trans. 2, 69, 708 (1973).
The National Science Foundation, most recently by Grant
CHE 9985895, supported this research. JS dedicates this
paper to his friend Don Arnold as an expression of his admi-
ration for Don’s many seminal contributions to photo-
chemistry.
23. R. Potashnik, C.R. Goldschmidt, and M. Ottolenghi. Chem.
Phys. Lett. 9, 424 (1971).
24. (a) F. Wilkinson, D.J. McGarvey, and A.F. Olea. J. Am. Chem.
Soc. 115, 12 144 (1993); (b) A.F. Olea and F. Wilkinson. J.
Phys. Chem. 99, 4518 (1995).
25. C. Sato, K. Kikuchi, C. Sato, Y. Takahashi, and T. Miyashi. J.
Phys. Chem. 99, 16 925 (1995).
26. K. Kikuchi, K. Okamura, M. Watabe, H. Ikeda, Y. Takahashi,
and T. Miyashi. J. Am. Chem. Soc. 115, 5180 (1993).
27. A.A. Abdel-Shafi and F. Wilkinson. J. Phys. Chem. A, 104,
5747 (2000).
28. K.C. Wu and A.M. Trozzolo. J. Phys. Chem. 83, 3180 (1979).
29. S.K. Chattopadhyay, C.V. Kumar, and P.K. Das. J. Phys. Chem.
89, 670 (1985).
30. M. Almgren and J.K. Thomas. Photochem. Photobiol. 31, 329
(1980).
31. (a) Z. Wang and W.G. McGimpsey. J. Phys. Chem. 97, 3324
(1993); (b) Z. Wang and W.G. McGimpsey. J. Phys. Chem. 97,
5054 (1993).
32. H. Görner. J. Photochem. 19, 343 (1982).
33. W.D.K. Clark and C. Steel. J. Am. Chem. Soc. 93, 6347 (1971).
34. S. Wang. Ph.D. Dissertation, Florida State University, Talla-
hassee, Florida, 1998.
35. P. Bortolus and G. Galiazzo. J. Photochem. 2, 361 (1974).
36. G. Fischer and E. Fischer. Mol. Photochem. 6, 463 (1974).
37. G.G. Aloisi, U. Mazzucato, J.B. Birks, and L. Minuti. J. Am.
Chem. Soc. 99, 6340 (1977).
38. For a review see: J. Saltiel, J.L. Charlton. In Rearrangements
in ground and excited states. Vol. 3. Edited by P. de Mayo. Ac-
ademic Press, New York. 1980. Essay 14.
References
1. B.S. Hudson and B.E. Kohler. Annu. Rev. Phys. Chem. 25,
437 (1974).
2. B.S. Hudson, B.E. Kohler, and K. Schulten. In Excited states.
Vol. 6. Edited by E.C. Lim. Academic Press, New York. 1982.
pp. 1–95.
3. J. Saltiel and Y.-P. Sun. In Photochromism, molecules and sys-
tems. Edited by H. Dürr and H. Bouas-Laurent. Elsevier, Am-
sterdam. 1990. pp. 64–164.
4. W. Fuß, S. Lochbrunner, W.E. Schmid, and K.L. Kompa. J.
Phys. Chem. A, 102, 9334 (1998).
5. W. Fuß, W.E. Schmid, and S.A. Trushin. J. Chem. Phys. 112,
8347 (2000).
6. For a recent review of polyene photophysics, see: W. Fuß, Y.
Haas, and S. Zilberg. Chem. Phys. 259, 273 (2000).
7. (a) S.H. Pullen, N.A. Anderson, L.A. Walker, II, and R.J.
Sension. J. Chem. Phys. 108, 556 (1998); (b) N.A. Anderson,
J.J. Shiang, and R.J. Sension. J. Phys. Chem. A, 103, 10 730
(1999).
8. J. Saltiel, S. Wang, L.P. Watkins, and D.-H. Ko. J. Phys. Chem.
A, 104, 11 443 (2000).
9. (a) J. Saltiel, S. Wang, D.-H. Ko, and D.A. Gormin. J. Phys.
Chem. A, 102, 5383 (1998); (b) J. Saltiel, J.M. Crowder, and
S. Wang. J. Am. Chem. Soc. 121, 895 (1999).
10. J. Saltiel, G. Krishnamoorthy, Z. Huang, D.-H. Ko, and S.
Wang. J. Phys. Chem. A, 107, 3178 (2003).
39. G.O. Schenck and R. Steinmetz. Bull. Soc. Chim. Belg. 71,
781 (1962).
40. D.-H. Ko. Ph.D. Dissertation, Florida State University, Talla-
hassee, Florida, 1997.
11. E.D. Cehelnik, R.B. Cundall, J.R. Lockwood, and T.J. Palmer.
J. Phys. Chem. 79, 1369 (1975).
© 2003 NRC Canada

Saltiel et al.
679
41. K. Gollnick, T. Franken, G. Schade, and G. Dorhofer. Ann.
N.Y. Acad. Sci. 171, 89 (1970).
42. A. Streitwieser, Jr., R.H. Jagow, R.C. Fahey, and S. Suzuki. J.
Am. Chem. Soc. 80, 2326 (1958).
43. For relevant early examples, see: M. Matsuoka and M. Szwarc.
J. Am. Chem. Soc. 83, 1260 (1961).
(c) M. Mac, P. Kwiatkowski, and A.M. Turek. Chem. Phys.
Lett. 250, 104 (1996); (d) M. Mac and J. Wirz. Photochem.
Photobiol. Sci. 1, 24 (2002).
45. For a typical example, see: J.L. Charlton, D.E. Townsend, B.D.
Watson, P. Shannon, J. Kowalewska, and J. Saltiel. J. Am.
Chem. Soc. 99, 5992 (1977).
44. (a) J. Gersdorf and J. Mattay. J. Photochem. 28, 405 (1985);
(b) M. Mac and J. Wirz. Chem. Phys. Lett. 211, 20 (1993);
© 2003 NRC Canada