Medium effects on the direct cis-trans photoisomerization of 1,4-diphenyl-1,3-butadiene in solution

Article abstract of DOI:10.1021/jp111482m

Quantum yields for the photoisomerization of trans,trans-1,4-diphenyl-1,3- butadiene (tt-DPB), determined in benzene, cyclohexane, methylcyclohexane, hexane, and perfluorohexane, confirm the low values reported earlier for benzene and cyclohexane and reveal even lower values in the last two solvents. In contrast to trans-stilbene (t-St), fluorescence and torsional relaxation leading to photoisomerization do not account exclusively for S₁tt-DPB decay. Competing radiationless singlet excited-state decay pathways exist in tt-DPB, which do not lead to photoisomerization and may not involve large-amplitude torsional motions. Our results invalidate analyses of tt-DPB fluorescence quantum yields and lifetimes that assign all radiationless decay to the isomerization channel. Gas-phase chromatography analysis of tt-DPB photoisomerization in hexane shows the reaction to be concentration-independent and reveals, for the first time, a significant, two-bond photoisomerization pathway, β_tt→tc = 0.092 and β_tt→cc = 0.020. The dominant one-bond-twist (OBT) process is accompanied by a bicycle pedal (BP) process that accounts for almost 20% of tt-DPB photoisomerization. The OBT tt-DPB photoisomerization quantum yield is largest in benzene (Bz) and smallest in perfluorohexane (PFH). Contrary to expectations, reduction in medium friction in PFH is accompanied by a decrease in β_tt→tc. The 1¹B_u/2¹A_g order and energy gap appear to control the contribution of torsional relaxation to radiationless decay. Lowering the 1¹B_u energy as in Bz favors photoisomerization. Reversal of the 1¹B_u/2 ¹A_g order in PFH is accompanied by short τ_f and small β_f and β_tt→tc values that suggest the presence of competing 2¹A_g → 1 ¹Ag relaxation paths that are unproductive with respect to photoisomerization. We conclude that the Birks extension to diphenylpolyenes of the Orlandi-Siebrand cis-trans photoisomerization mechanism is not valid. Photoisomerization appears to occur in the 1^IB_u state, and we argue that this applies to t-St as well.

Full text of DOI:10.1021/jp111482m

ARTICLE
pubs.acs.org/JPCA
Medium Effects on the Direct Cis-Trans Photoisomerization of
1,4-Diphenyl-1,3-butadiene in Solution
Jack Saltiel,*,† Talapragada R. S. Krishna,† Kritapas Laohhasurayotin,† Yanjun Ren,† Kathleen Phipps,†
Paul H. Davis,‡ and W. Atom Yee‡
†Departments of Chemistry and Biochemistry, Florida State University, Tallahassee, Florida 32306-4390, United States
‡Santa Clara University, Santa Clara, California 95053-0365, United States
ABSTRACT: Quantum yields for the photoisomerization of
trans,trans-1,4-diphenyl-1,3-butadiene (tt-DPB), determined in
benzene, cyclohexane, methylcyclohexane, hexane, and per-
fluorohexane, confirm the low values reported earlier for
benzene and cyclohexane and reveal even lower values in the
last two solvents. In contrast to trans-stilbene (t-St), fluores-
cence and torsional relaxation leading to photoisomerization do not account exclusively for S1 tt-DPB decay. Competing
radiationless singlet excited-state decay pathways exist in tt-DPB, which do not lead to photoisomerization and may not involve
large-amplitude torsional motions. Our results invalidate analyses of tt-DPB fluorescence quantum yields and lifetimes that assign all
radiationless decay to the isomerization channel. Gas-phase chromatography analysis of tt-DPB photoisomerization in hexane shows
the reaction to be concentration-independent and reveals, for the first time, a significant, two-bond photoisomerization pathway,
φttftc = 0.092 and φttfcc = 0.020. The dominant one-bond-twist (OBT) process is accompanied by a bicycle pedal (BP) process that
accounts for almost 20% of tt-DPB photoisomerization. The OBT tt-DPB photoisomerization quantum yield is largest in benzene
(Bz) and smallest in perfluorohexane (PFH). Contrary to expectations, reduction in medium friction in PFH is accompanied by a
decrease in φttftc. The 11Bu/21Ag order and energy gap appear to control the contribution of torsional relaxation to radiationless
decay. Lowering the 11Bu energy as in Bz favors photoisomerization. Reversal of the 11Bu/21Ag order in PFH is accompanied by
short τf and small φf and φttftc values that suggest the presence of competing 21Ag f 11Ag relaxation paths that are unproductive
with respect to photoisomerization. We conclude that the Birks extension to diphenylpolyenes of the Orlandi-Siebrand cis-trans
photoisomerization mechanism is not valid. Photoisomerization appears to occur in the 1IBu state, and we argue that this applies to t-
St as well.
’ INTRODUCTION
are 11Bu and 21Ag, respectively.10 The energies of the 11Bu and
21Ag states were proposed to increase and decrease, respectively,
with twisting about the central bond. The OS photoisomeriza-
tion model was extended to the higher members of the diphe-
nylpolyene family by Birks,11 who interpreted the temperature
dependencies of their fluorescence quantum yields and lifetimes
by assuming that fluorescence and photoisomerization are the
sole significant decay processes, as in t-St.9
Because a large-amplitude motion is associated with their
photoisomerization, t-St and trans,trans-1,4-diphenyl-1,3-buta-
diene (tt-DPB), the two lowest members of the series, have been
widely used as probes for testing theories on medium effects. In
the main, those studies have followed Birk’s lead in assigning radia-
tionless decay rate constants inferred from fluorescence quantum
yields and lifetimes to torsional relaxation, exclusively.12-16
We have shown that the assumption of photoisomerization/
fluorescence complementarity is not valid for the 1,6-diphenyl-
1,3,5-hexatriene isomers.17 The results presented in this paper
on the photoisomerization of the DPB isomers show that the
The photochemistry and photophysics of R,ω-diphenylpo-
lyenes have been studied extensively because these molecules are
considered to be models for the retinyl polyenes that are related
to vitamin A and the visual pigments.1,2 A small intrinsic torsional
energy barrier in the lowest excited singlet state (S1) of t-St was
proposed by one of us long ago.3 It accounts for the temperature
dependencies of photoisomerization quantum yields4 and fluor-
escence quantum yields5 and lifetimes.6,7 On the basis of
symmetry, excited states of planar all-trans-diphenylpolyenes
are designated as n1Ag or n1Bu. Hudson and Kohler first pointed
out that, in contrast to t-St, for which the lowest excited state that
is observed in both absorption and fluorescence is the 11Bu state,
S1 for longer R,ω-diphenylpolyenes is the 21Ag state.8,9 Thus, for
the all-trans isomers of 1,6-diphenyl-1,3,5-hexatriene (ttt-DPH)
and 1,8-diphenyl-1,3,5,7-octatetraene, excitation to the 11Bu
state gives rise to the vibronically resolved bands of the lowest-
energy-allowed electronic transition in the absorption spectrum,
but fluorescence comes predominantly from the energetically
lower-lying 21Ag state.8,9
Orlandi and Siebrand (OS) attributed the torsional energy
barrier along the isomerization coordinate of t-St to an assumed
crossing of S1 and S2 potential energy curves whose symmetries
Received: December 2, 2010
Revised:
January 21, 2011
Published: February 25, 2011
r
2011 American Chemical Society
2120
dx.doi.org/10.1021/jp111482m J. Phys. Chem. A 2011, 115, 2120–2129
|

The Journal of Physical Chemistry A
ARTICLE
photoisomerization/fluorescence complementarity assumption
fails in this case also.
’ EXPERIMENTAL SECTION
Materials. tt-DPB (Aldrich, reagent) was purified by chro-
matography on alumina and recrystallized from ethanol; for
experiments at Santa Clara University (SCU), it was recrystal-
lized twice from ethanol. cc-DPB was synthesized using a
modification of a known procedure,18 as previously described.19
ct-DPB was prepared photochemically from tt-DPB followed by
chromatographic separation of the photostationary mixture by a
procedure analogous to that previously described.20 Cyclohexane
(Fisher, HPLC grade) was distilled over P2O5 before use.
Methylcyclohexane (Aldrich, spectrograde) was used as received.
Perfluorohexane (Aldrich) was refluxed over KMnO4 and distilled.
The sources of other materials and the purification methods used
in this work were as previously described.20-22
Irradiation Procedures. Sample preparation and degassing
procedures employed at Florida State University (FSU) have
been described.21,22 Sets of 13 mm o.d. Pyrex tubes fitted with
standard taper joints and grease traps were loaded with 3 mL
aliquots of each concentration of DPB and trans-stilbene. They
were degassed using at least six freeze-pump-thaw cycles to
about 10-5 Torr and flame-sealed at a constriction. All opera-
tions, including analyses, were performed under nearly complete
darkness (red light). UV absorption measurements in the course
of the irradiation were carried out either by providing the Pyrex
irradiation ampules with side arms consisting of a graded seal and
a quartz UV cell or by periodically opening ampules and
recording the UV spectra in standard UV cells. Irradiations were
carried out in a Moses merry-go-round23 apparatus immersed in
a thermostatted water bath. A heating coil connected to a
thermoregulator (Polyscience Corp.) was used to control the
temperature to (0.5 °C. The photoisomerization of trans-
stilbene was used for actinometry, φtfc = 0.52.21 Hanovia 450
W medium-pressure Hg lamps (Ace Glass, Inc.) were employed.
The 313 nm Hg line was isolated using a potassium perchro-
mate/potassium carbonate filter solution in parallel with Corning
CS 7-54 filter plates, as previously described.21
Photoisomerization in methylcyclohexane (MCH) was car-
ried out at SCU as previously described,20 except that this time,
excitation was with a 337 nm pulsed nitrogen laser (Avco-Everett).
Potassium ferrioxalate was used for actinometry. Stirred Ar-
outgassed samples were irradiated in a 1 cm quartz stoppered
fluorescence cell inside of a brass cell holder. Temperature was
controlled with a Neslab Endocal refrigerated circulating bath
system using propylene glycol.
Analytical Procedures. Analysis of stilbene isomerization
was by GC, as described previously.24 Analysis of DPB isomer-
ization was either by GC, by HPLC, or by principal component
analysis (PCA) of UV spectral matrixes. A Varian 3300 gas
chromatograph equipped with an electronic integrator was used
with a DB-5 capillary column (17 m). Analyses were carried out
at 12 psi of He and an initial temperature of 170 °C held for 10
min followed by 1 °C/min ramp to 190 °C. Detector and injector
temperatures were set at 250 °C. Pentane was replaced by CCl4
prior to GC analysis, and cyclohexane and benzene were replaced
by hexane prior to HPLC analysis. Solvent removal was achieved
by evaporation under a stream of Ar, and care was taken to avoid
dryness. The HPLC analyses were done at ambient temperature
on a normal-phase SI column (Beckman, ultrasphere, analytical,
Figure 1. Beer-Lambert plot in PFH (garnet line through first four
points).
4.6 ꢀ 250 mm) using optima hexanes (Fisher, optima) as the
mobile phase (absorbance monitored at 315 nm). Electronic
absorption spectra were recorded at 20 °C on a Varian Cary
300B UV-vis spectrophotometer using a thermostatted cell
holder.
Absorption spectra in MCH were measured as a function of
irradiation time with a Shimadzu UV-3101PC spectrophotometer.
DPB isomer concentrations were determined by analysis of the
absorption spectra. Quantum yields for photoisomerization of
tt- f ct-DPB were determined by use of Zimmerman’s method.25
Fluorescence Measurements. Fluorescence spectra were
measured with a Hitachi F-4500 spectrophotometer equipped
with a 150 W Xe arc source and a Hamamatsu R3788 photo-
multiplier tube. Fluorescence spectra were recorded for de-
gassed, air-saturated, and Ar-bubbled solutions of tt-DPB in
benzene in standard 1 cm2 quartz cells. There was no oxygen
effect on fluorescence intensity. 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 digital
thermometer in a reference cell placed in the same constant-
temperature cell holder. Absorption spectra were recorded with
a Cary 300 UV-vis spectrophotometer at 20.0 °C. Absor-
bances at the excitation wavelengths were less than 0.1. Quinine
sulfate in 1 N H2SO4 (φf = 0.546, at 25 °C)26,27 was used as a
reference standard. The solvent index of refraction correction
was applied.
’ RESULTS
Solvent Effect on tt-DPB Absorptivity. tt-DPB solutions in
hexane (H), cyclohexane (CH), benzene (Bz), and perfluor-
ohexane (PFH) were made by dilution of CH stock solutions
([tt-DPB] = 7.21 ꢀ 10-3 M for CH, H, and Bzand 9.87 ꢀ 10-3
M
for PFH) using λ-pipettes and volumetric flasks. In CH, H, and Bz,
the five identical tt-DPB concentrations employed in the range of
0.721-3.60 ꢀ 10-5 M gave excellent Beer-Lambert plots (CH
contamination was in the0.10-0.5% range in Bzand H and 0.25-
4% in PFH). The plot in PFH has a downward curvature due to
saturationbutislinearforthethreelowestconcentrations,Figure1.
The Beer-Lambert plots in the other three solvents were strictly
linear throughout the concentration range. Molar absorptivity
coefficients, ε, at λmax are given in Table 1.
Fluorescence Quantum Yield. The fluorescence quantum
yield of tt-DPB in Bz was determined by exciting degassed,
Ar-outgassed, and air-saturated solutions at 315, 330, and
345 nm. There was no oxygen-quenching effect within the
2121
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
Table 1. Solvent Effect on Molar Absorptivities of tt-DPBa
solvent
λmax, nm
10-4εmax, M-1 cm-1
10-4ε313tt, M-1 cm-1
10-4ε313ct, M-1 cm-1
PFH
H
318.5
328.0
330.4
334.5
5.82
4.82
4.62
4.44
3.22
2.98
3.05
3.00
2.76
5.51 (5.60)b
5.50 (5.42)c
5.10
CH
Bz
a The last two columns show molar absorptivity values at 313 nm for tt- and ct-DPB, respectively. b From ref 18. c From ref 20.
Table 2. tt-DPB Photoisomerization in Degassed Hexane,
313 nm, ∼22 °Ca
t
t
corr
corr
t, min
fct
fcc
fct
fcc
[tt-DPB] = 5.02 ꢀ 10-4
0
M
0
0.0072(7)
0.143(2)
0.249(2)
0.320(4)
0.489(4)
0.697(1)
0.698(5)
0
0
20
0.00291(2)
0.00532(9)
0.00692(30)
0.0106(5)
0.153
0.301
0.421
0.834
0.00311
0.00606
0.00825
0.01434
40
60
120
900
1200
0.0221(4)
0.0232(2)
[tt-DPB] = 3.01 ꢀ 10-3 M
0
0
0.018
0
0
20
0.0391(14)
0.068(2)
0.0875(4)
0.148(18)
0.501(1)
0.56(3)
0.00048(2)
0.00132(26)
0.00157(15)
0.00313(7)
0.0128(3)
0.022
0.0533
0.0753
0.148
0.000485
0.00136
0.00163
0.00337
40
Figure 2. Photoisomerization of tt-DPB in H (upper and lower panels
are for [tt-DPB] = 3.01 ꢀ 10-3 and 5.02 ꢀ 10-4 M, respectively); b and
d panels are expanded plots of conversions to cc-DPB.
60
120
900
1200
0.0159(5)
drawn through conversions that are corrected for zero-time
contamination and back reaction. Direct formation of cc-DPB
is demonstrated in panels b and d of Figure 2 by the absence of an
induction period; the cc-DPB versus time plots are linear from
the onset.
a Values in parentheses are deviations from the mean for multiple GC
injections; parallel irradiation of the trans-stilbene actinometer for 40
and 1200 min gave 38.5(3) and 93.3(3)% cis-stilbene, respectively.
experimental uncertainty of the measurements, and none
was expected in view of the reported very short (29 ps)12 tt-
DPB fluorescence lifetime. The average of the nine values
gave φf = 0.251 ( 0.015.
Photochemical Observations. Hexane. Concentrations of
tt-DPB in hexane (3.01 ꢀ 10-3 and 5.02 ꢀ 10-4 M) were
selected to roughly correspond to the highest and lowest con-
centrations employed in earlier work.28 An intermediate con-
centration (1.69 ꢀ 10-3 M) of trans-stilbene in pentane was used
for actinometry (φtfc = 0.52).21 Ampules, immersed in water at
room temperature (∼22 °C), were irradiated in parallel in the
merry-go-round. Conversions were measured by GC as a func-
tion of time, Table 2. They were corrected for zero-time isomer
content and back reaction21,29 using eq 1
Cyclohexane, Methylcyclohexane, and Benzene
Temperature E€ect. Degassed aliquots (3 mL) of tt-DPB solu-
tions in CH and Bz were irradiated in triplicate at 10 °C intervals
in the thermostatted water bath in which the merry-go-round was
immersed. Quantum yields were based on conversions for the
first two tubes, which were irradiated for 50 and 120 min,
respectively. The third tubes of each set were irradiated for
29-58 h to yield the PSS isomer compositions used for back
reaction corrections. Ampules containing trans-stilbene in pen-
tane were irradiated in parallel for actinometry. Isomer composi-
tions were determined by HPLC for DPB and by GC for stilbene
(see above). The solvents, cyclohexane and benzene, were carefully
replaced with hexane, and the solutions were concentrated to a
maximum volume of 1 mL under a stream of Ar. HPLC measure-
ments were performed with the UV detector set at 315 nm. Hexane
was used as the eluent. HPLC areas were converted to relative molar
contributions with the use of the molar absorptivities in ref 18. All
operations were carried out under red light. The lower absorbance of
cc-DPB (a factor of 2 lower than that for tt-DPB18 at 315 nm) coupled
with its inefficient formation made it impractical to determine
conversions of that isomer for quantum yield purposes. The actin-
ometer φtfc = 0.52 at 30 °C21 was adjusted for the T change on the
basis of the T dependence of φf of trans-stilbene (the change was
0.509-0.524 in the 10-40 °C range).30 Identical changes were
predicted using the Arrhenius parameters in ref 21 or the Eyring
!
0
fcxeq - fcx
corr
fcx
¼ fcxeq ln
ð1Þ
fcxeq - fcx
t
with the poststationary state (PSS) isomer composition fractions
determined for the lower [DPB]: 0.279, 0.698, and 0.0226 for
eq
ftteq, fct and fcceq, respectively. The other symbols in eq 1,
fcx0, fcxt and fcxcorr, where x is t or c, are the zero-time fraction,
the observed fraction at time t, and the corrected fraction,
respectively.
Plots of the results in Table 2 are shown in Figure 2. Curves are
drawn through observed conversions, and straight lines are
2122
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
Table 3. Photoisomerization Quantum Yields of tt-DPB in
parameters in ref 30. Averages of duplicate quantum yields are shown
in Table 3.
CH and Bza
Quantum yields determined at SCU in MCH with the use of
Zimmerman’s method25 are shown in Table 4. Light intensity was
determined with the use of potassium ferrioxalate actinometry.31
Because a considerably longer excitation wavelength (337 nm)
was employed than in the earlier work (325 nm),20 the φttfct
value was also determined in CH at room temperature with
337 nm excitation and was found to be 0.11, that is, identical to
the previously published value.20
PSSb: ftteq, fcteq, fcc
eq
φttfct
T, °C
CH
Bz
CH
0.270, 0.711, 0.0192
Bz
10.0 0.0975(25)
20.0 0.115(1)
30.0 0.126(0)
40.0 0.136(0)
0.216(1) 0.319, 0.661, 0.0195 0.341, 0.646, 0.0127
0.260(2) 0.276, 0.699, 0.0249 0.306, 0.673, 0.0208
0.300(5) 0.239, 0.732, 0.0297 0.294, 0.684, 0.0221
Perfluorohexane. The low solubility of tt-DPB in PFH pre-
cluded use of GC or HPLC analyses for the determination of
isomer distributions in that solvent. Ampules containing solu-
tions of the DPB isomers were provided with 1 cm square quartz
cell side arms for measurement of UV spectra in the course of the
irradiation. Samples of tt-DPB in PFH were irradiated in the
merry-go-round at room temperature (∼23 °C) in parallel with
tt-DPB samples in CH and in Bz. In addition, trans-stilbene
actinometry was employed. Sets of UV spectra obtained in the
course of 313 nm irradiation of each DPB isomer are shown in
the panels b-d of Figure 3. They were recorded at 25.0 °C in a
thermostatted cell holder. Pure DPB isomer spectra in PFH are
shown in Figure 3a. Analogous pure component spectra and
spectral evolutions starting from tt- and cc-DPB were observed
in the other solvents, in agreement with previously reported
results.18,20,32
a The concentrations employed were in the (1.05-1.27) ꢀ 10-3 and
(1.8-2.8) ꢀ 10-4 M ranges for t-St and tt-DPB, respectively; values in
parentheses are deviations from the mean in the last significant figures
shown for duplicate determinations. b Irradiation times for PSS deter-
minations (given in ftteq, fcteq, fcceq sequence) were 58, 40, 40, and 29 h for
the 10, 20, 30, and 40 °C samples, respectively.
Table 4. Photoisomerization Quantum Yields of tt-DPB in
MCHa
b
b
T, °C
φttfct
108kd , s-1
T, °C
φttfct
108kd , s-1
-5.3
7.2
0.054 (7)
0.072 (11)
0.120 (15)
11.66
13.39
16.33
40.0
50.0
0.146 (19)
0.178 (21)
23.77
29.15
20.0
Spectral sets in each solvent were subjected to principal
component analysis (PCA), as previously described.33 Three
principal component matrices were obtained in each solvent. Com-
ponent fractional contributions, fxxn, for the normalized (to unit
a Values in parentheses are deviations from the mean in the last
significant figures shown for duplicate determinations. b Interpolated
fluorescence decay rate constants from the Supporting Information in
ref 16 (see text).
Figure 3. Photoisomerization of the DPBs in PFH; (a_ Pure isomer spectra tt-, ct-, and cc-DPB in red, green, and blue, respectively; (b-d) starting from
tt-, ct-, and cc-DPB, respectively.
2123
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
area) spectra used in PCA were converted to molar fractional
contributions, fxxt, by dividing each by the expected molar area
(prior to normalization) of the corresponding pure component
spectrum, Ai, at the initial [DPB] and adjusting to unity the sum
of the resulting values, eq 2.
revealed that excitation of photoproduct ct-DPB converts it
mainly to tt-DPB, the yields of the two products were combined
to obtain the true measure of ct-DPB formation, Figure 5b.
Table 5 gives the slopes s of the lines in Figure 5, which are the
relative photoisomerization rates. They were converted to
photoisomerization quantum yields using t-St actinometry,
Table 5. Comparison with the more accurate quantum yields
in Table 3 shows the φttfct values to be in reasonable agreement,
with small deviations in opposite directions in CH and in Bz. The
estimated uncertainty in the PFH values is (15%.
fcxn=Acx
t
fcx
¼
ð2Þ
∑fin=Ai
i
Viewed in eigenvector space, the plot of the combination
coefficients, Figure 4, for the normalized spectra in Figure 2 is
instructive. All spectra fall on the normalization plane defined by
the three principal eigenvectors. The three pure component
spectra define the corners of the normalization triangle contain-
ing eigenvector combination coefficients for spectra of all real
mixtures of the three isomers. Starting from tt- and cc-DPB, initial
mixture spectra fall on the tt-/ct-DPB and cc-/ct-DPB edges of
the triangle, consistent with one-photon/one-bond photoisome-
rization, as initially reported by Zechmeister and co-workers.18
The PCA treatment of the UV spectra is not sufficiently accurate
to reveal the small two-bond tt- f cc-DPB photoisomerization
channel that was observed by GC analysis in H.
Figure 5 shows relative photoisomerization rates in CH, PFH,
and Bz, starting from tt- and cc-DPB. PCA derived conversions
starting from tt-DPB were corrected for back reaction, eq 1. No
back reaction corrections were applied to cc-DPB photoisome-
rization conversions because the low contribution of that isomer
in the PSS ensures that its formation is virtually irreversible.
However, because the PCA treatments of the spectral sets
’ DISCUSSION
The cis-trans photoisomerization of the DPBs was first
studied by Zechmeister and co-workers in H.18 They showed
that upon irradiation in hexane, both tt- andcc-DPB are converted to
the same ct-DPB-rich photostationary state.18b Photoisomeriza-
tion quantum yields, φttfct = 0.25 and φctftt = 0.19, were first
reported in Bz by Whitten and co-workers.32 They reported that
excitation of tt-DPB at either 313 or 366 nm in Bz leads to quasi-
stationary states that, in addition to tt-DPB, include 60 and 64%
ct-DPB, respectively, and 1-2% cc-DPB. They adopted the
mechanism involving one bond twist (OBT) to phenallyl/benzyl
intermediates that had been proposed for the photoisomeriza-
tion of excited singlet34 and triplet states35 of alkyl-substituted
1,3-butadienes, Chart 1.32 A more complete study of DPB
photoisomerization in CH was reported by Yee and co-
workers.20 Surprisingly, φttfct = 0.11, obtained upon 325 nm
excitation, was even lower than the value in Bz. Quantum yields
starting from the other isomers were similarly low, φccfct = 0.20
and φctftt = 0.04. Accordingly, the conclusion in both studies was
that, in Bz and CH, significant radiationless decay channels of the
Table 5. PCA-Derived Isomerization Rates of tt- and cc-DPB
in CH, PFH, and Bz, 313 nm, 23 °Ca
105[DPB]0, M
s, %/min
tt-DPB cc-DPB
eq
solvent tt-DPB cc-DPB
fct
φttfct φccfct
CH
PFH
Bz
3.60
3.11
3.49
6.19
6.31
6.60
0.003549 0.004353 0.71 0.087 0.19
0.003552 0.003814 0.721 0.076 0.17
0.010585 0.009077 0.64 0.265 0.42
Figure 4. Combination coefficient plot from the PCA treatment of the
spectra in Figure 3.
a Parallel irradiation (30 min) of [t-St] = 2.38 ꢀ 10-3 M gave 9.12%
(9.58%, corrected) c-St.
Figure 5. Relative tt- and cc-DPB photoisomerization conversions in CH (garnet 9), PFH (blue (), and Bz (green 2), panels a and b, respectively.
2124
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
Chart 1
Table 6. Solvent Effect on tt-DPB Photoisomerization and
Fluorescence Quantum Yields, ∼20 °Ca
solvent τf/ns
φf
108kf, s-1
φttfct
φttfcc
φctftt
φccfct φic
Bz 0.30 0.25(2)
MCH 0.59 0.36
CH 0.57 0.41
8.3
6.1
7.2
7.5
1.9
0.25, 0.25b
0.12
0.16, 0.19b 0.42 0.25
0.40
0.115, 0.11c
0.066, 0.04c 0.19 0.36
H
0.48 0.36
0.092, 0.084d 0.020 0.060
0.076
0.48
PFH 0.20 0.038
0.15
0.17 0.81
a In H and CH, τf and φf values are averages of those compiled in Table 2
of ref 16; PFH values are from ref 16; the τf in Bz is the average from refs
singlet excited states of the DPB isomers compete with torsional
relaxation leading to isomerization. Related observations by
G€orner led to the same conclusion.36 In agreement with Whitten
and co-workers, Yee et al. proposed that tt- and cc-DPB photo-
isomerize via the intermediates in Chart 1 by the OBT processes
c
12 and 37. and the φf is from this work. b From ref 32. From ref 20.
d From ref 28.
1tt* f tp* and cc* f cp*. Both intermediates are accessible
from 1ct*. Because cc-DPB was not detected by HPLC analysis at
300 nm, the cis-phenallyl/benzyl intermediate was assumed to
decay exclusively to ct-DPB. More recently, Krishna and Singh
reported a sharp increase in tt-DPB photoisomerization quantum
yields (330 nm, hexane, 25 °C) as the concentration of tt-DPB
was increased (a factor of 4.67 in φttfct over the 0.62-3.08 ꢀ
10-3 M range).28
1
1
1
Chart 2
contributions of direct formation of cc-DPB in the other solvents.
Simultaneous two-bond photoisomerization is consistent with
Warshel’s bicycle pedal (BP) mechanism.38 This volume-con-
serving process was proposed for space-confining media such as
the protein environments of the retinyl moieties in rhodopsin
and bacteriorhodopsin that could inhibit the large torsional
displacements to orthogonal geometries that are associated with
the usual OBT mechanism. BP predicts the net chemical process
in the light-adaptation photocycles of bacteriorhodopsin (13-cis-
15-syn f 13-trans-15-anti)39 and Anabaena sensory rhodopsin
(13-trans-15-anti f 13-cis-15-syn).40,41 The BP mechanism
reduces free volume requirements by confining most of the
motion to the vicinity of the isomerizing double bonds while
minimizing the motion of bulky substituents. Indeed, we recently
observed two-bond DPB photoisomerization in the opposite
direction (cc-DPB f tt-DPB) in glassy media at low T42 and in
the solid state.43 There are other solid-state examples of photo-
isomerization consistent with the BP mechanism.44 Less com-
mon are reports of BP products in solution. The interconversion
of ctt- and tct-DPH could involve simultaneous two-bond
photoisomerization.17a,b Another possible example is the adia-
batic photoisomerization of cis,cis-1,4-di(3,5-di-tert-butylstyryl)-
benzene to trans,trans-1,4-di(3,5-di-tert-butylstyryl)benzene.45
Related studies in the distyrylbenzene family have revealed
similar results.46 Most intriguing is Whitten’s proposal of BP
photoisomerization for the tt-DPB moiety covalently embedded
in the alkyl chain of the carboxylic acid 1, Chart 2.47 The change
in mechanism from OBT to BP was proposed to account for the
much lower T sensitivity of the radiationless decay of 1 relative to
the unsubstituted parent tt-DPB.
Quantum Yields. The combination coefficient plot for PFH
in Figure 4 is a representative example. We obtained similar plots
in CH and Bz. Starting from the tt- and cc-DPB tetrahedron
corners, spectral points initially move on the tt-/ct- and cc-/ct-
DPB edges, respectively, consistent with one-photon/one-bond
photoisomerization, as had been previously concluded.18,20,32
However, as the GC results in H show, due to the very small cc-
DPB contribution to photostationary states, the adherence of
initial points to two-component edges can be deceiving. Accord-
ingly, UV analysis fails to reveal the presence of significant decay
channels to cc-DPB.
The photoisomerization quantum yields determined in our
work at ambient T are summarized in Table 6 together with
literature values of fluorescence lifetimes and quantum yields
(except that φf in Bz is from this work). Our φttfct values in Bz
and CH agree exactly with the literature values. In contrast to the
report by Krishna and Singh,28 we find that φttfct in H is not
concentration-dependent. Our 0.092 value is somewhat larger
than 0.084, the value reported for the highest concentration
employed in that work.28 If the proposed interpretation of
fluorescence data16 were correct, one would expect that φttfct
would be largest in PFH. On the contrary, that is the solvent in
which the φttfct value is smallest. Our photoisomerization
quantum yields starting from cc-DPB exhibit the same trend,
being largest in Bz and smallest in PFH. Despite the considerable
range in quantum yields, PSS compositions are relatively solvent-
independent (see the Hexane subsection and Tables 3 and 5).
The contribution of ct-DPB is largest in all cases, and the
contribution of cc-DPB never exceeds 3%. Consistent with past
practice,20,32 we neglected the small contribution of cc-DPB and
used the molar absorptivities at 313 nm and the equilibrium
isomer fractions determined in this work to calculate the φctftt
values in Table 6
Medium constraints are absent in solution, and sequential,
instead of concerted, isomerization of the two bonds in S1
cannot be ruled out. The two-bond interconversion of trans,
trans- and cis,cis-2,4-hexadiene triplets was shown to proceed
via rapid equilibration of trans-allylmethylene and cis-allyl-
methylene triplet intermediates,35c and an analogous process
could account for the formation of cc-DPB from tt-DPB.
Possible tt-DPB photoisomerization pathways are shown in
Chart 3.
!
eq
313
εtt
f
tt
313
eq
φct f tt ¼ φtt f ct
fct
ð3Þ
εct
Bicycle Pedal Channel. Noteworthy is our observation that
∼20% of the photoisomerization of tt-DPB in H goes directly to
cc-DPB. GC analysis would probably have revealed small
Solvent Effect on the 11Bu/21Ag State Order. Before con-
sidering the significance of the solvent effect on the DPB
2125
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
Chart 3
fluorescence arises from the small equilibrium population of
molecules in the 11Bu state. Accordingly, kfeff, the effective
radiative rate constant, is given by
kf ¼ kf e-ΔE=RT
ð4Þ
eff
B
where kfB is the radiative rate constant of the 11Bu state and
ΔE is the 21Ag/11Bu energy gap. The theoretical value of
kfB = 6.7 ꢀ 108 s-1 in CH, based on the Birks-Dyson
modification57 of the Strickler-Berg relationship,58 is in
reasonable agreement with the experimental values in Table 6
for the hydrocarbon media. Taking into account the smaller
PFH index of refraction and the larger tt-DPB molar absorp-
tivities in PFH (Table 1), the kf value for H of 7.5 ꢀ 108 s-1 in
Table 6 was adjusted to 6.8 ꢀ 108 s-1 for PFH. Substitution
of that kfB value and ΔE = 0.7 kcal/mol from Figure 6 into
eq 4 gives kfeff = 2.0 ꢀ 107 s-1, in excellent agreement with
the value for PFH in Table 6.
We note that our interpretation of the radiative rate constant
of tt-DPB in PFH differs sharply from that given by Maroncelli
and co-workers.16 They assumed that the tt-DPB radiative rate
constant is a continuous function of solvent polarizability and
fitted values in hydrocarbon and fluorocarbon solvents in a single
plot that includes the experimental value of 1.54 ꢀ 107 s-1 that
had been assigned to the 21Ag f 11Ag transition of the isolated
molecule in the gas phase.54 Their model assumes the lowest
excited singlet state to be a single, strongly mixed (21Ag þ 11Bu)
state rather than an equilibrium mixture of two distinct states,
one mainly Ag-like and the other mainly Bu-like, as assumed here.
Competing Radiationless Decay Channels. The internal
conversion quantum yields listed in the last column of Table 6,
were obtained from the experimental fluorescence and photo-
isomerization quantum yields using
Figure 6. Medium effect on the lowest two singlet excited states of tt-
DPB.
photoisomerization quantum yields, we must briefly review the
extensive literature on the influence of the solvent on the relative
energy of the two lowest singlet excited states of tt-DPB.48
Absorption and fluorescence of tt-DPB involve the symmetry-
allowed 11Ag-11Bu transition in all solvents studied thus
far.12-16,49,50 However, one-photon51,52 and two-photon52,53
fluorescence excitation spectra of tt-DPB in the gas phase, seeded
in He in a jet expansion, firmly establish that the 21Ag state lies 3.3
kcal/mol below the 11Bu state under isolated molecule condi-
tions. In the free jet expansion where tt-DPB molecules have low
vibrational and rotational temperatures and there are no colli-
sions, mode-selective fluorescence lifetimes, ranging up to 100
ns,51-53 are consistent with a forbidden electronic transition, and
the fluorescence quantum yield is unity at 0 excess energy.54 The
sharp decrease in the fluorescence quantum yield and lifetime
with increasing excitation energy has been attributed to the onset
of a nonradiative channel associated with isomerization at about
1100 cm-1 above the 2lAg zero-point energy level. Excitation of a
static tt-DPB gas at 95 °C gives the 21Ag f 11Ag emission
spectrum, but upon thermal equilibration in the presence of
added PFH vapor, that spectrum is replaced by the allowed 11Bu
f 11Ag emission spectrum.49,50
φic ¼ 1 - φf - 2φtt f ct - φtt f cc
ð5Þ
The 11Bu r 11Ag transition energy decreases with increas-
ing medium polarizability, being linearly dependent on
R = (n2 - 1)/(n2 þ 2), where n is the refractive index.49,55,56
On the other hand, the forbidden 21Ag r 11Ag transition
energy is almost polarizability-independent. Consequently, as
shown in Figure 6, the low polarizability of PFH leads to
reversal of the 21Ag/11Bu energy order so that, whereas S1 is
the 11Bu state in all hydrocarbon solvents, S1 is the 21Ag state
in PFH.49,50 We can now readily understand the sharp drop
in the kf value of tt-DPB in PFH, Table 6. Most excited singlet
states occupy the nonfluorescent 21Ag state in PFH, and
In eq 5, it is assumed that the trans-phenallyl/benzyl intermedi-
ate, 1tp-DPB*, in Chart 3 decays with equal probability to tt-DPB
and ct-DPB, as suggested by Yee and co-workers.1,20 The small
difference in the φic = 0.36 value for CH in Table 6 and the
reported 0.34 value is due to use of φf = 0.42 and inclusion of a
small intersystem crossing contribution, φis = 0.02, in the earlier
work.20 Because experimental values for φf of tt-DPB in CH are in
the 0.39-0.44 range,16 we used the average value of 0.41.
The φic’s in Table 6 range from a low value of 0.25 in Bz to a
high value of 0.81 in PFH; the opposite trend applies to the
2126
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
Chart 4
based on the assignment of all radiationless decay to torsional
relaxation. Whether the competing internal conversion pathway
involves decay of the 21Ag state directly to the 11Ag state or via
the 11Bu state remains to be established and should be a challenge
for theory to predict.
Temperature Effects on Nonradiative Decay. The T de-
pendence of torsional decay rate constants can be derived from
the photoisomerization quantum yields in CH (Table 3) and
MCH (Table 4) and the known T dependence of the fluores-
cence lifetimes12,16 using
photoisomerization quantum yields. Photoisomerization is most
efficient in the highly polarizable Bz in which the 11Bu state is the
lowest excited singlet state, Figure 6, and the ΔEab gap is largest,
and it is least efficient in PFH in which the 21Ag state is the lowest
excited singlet state. We conclude that photoisomerization does
not involve crossing from the 11Bu into the 21Ag state as
proposed in the Birks extension of the OS mechanism to R,ω-
diphenylpolyenes.10,11 The most reasonable alternative explana-
tion is that the 11Bu state is the reactive state, as had been
proposed by Troe and co-workers.14a
Rktors ¼ φtt f ct=τf
ð6Þ
Table 4 includes fluorescence decay rate constants, interpolated
from the experimental values in ref 16 using a trinomial fit of the
decay constant dependence on T. Arrhenius plots of the product
of torsional rate constants and the decay fraction R of the
1
assumed tp-DPB* intermediate, Rktors, and of the total radia-
1
The deactivation of tt-DPB* by an internal conversion step
tionless decay rate constants, knr, from refs 12 and 16, are shown
in Figure 7. In both CH and MCH, those plots are almost parallel
but have significantly different intercepts. In CH, the activation
energies, Ea, are 5.3 and 4.9 kcal/mol, and pre-exponential
factors, A, are 8.0 ꢀ 1012 and 1.7 ꢀ 1012 s-1 (assuming R =
0.5, independent of T) for knr and ktors, respectively. The cor-
responding values in MCH are Ea = 5.4 and 6.6 kcal/mol, and the
pre-exponential factors are A = 1.04 ꢀ 1013 and 3.10 ꢀ 1013 s-1 in
the same order. In each solvent, the Ea values for knr and ktors are
the same within experimental uncertainty, especially if one
considers that we have confined our treatment of the decay rate
constants in MCH in ref 16 to those provided in the Supporting
Information for that paper. The more extensive data set plotted
in Figure 7 of that paper gives Ea = 6.4 kcal/mol and A = 5.45 ꢀ
1013 s-1, but the three additional knr values at lower T than that in
Figure 7b, where the fluorescence lifetime is dominated by
radiative decay, are subject to larger experimental uncertainty.
The finding of experimentally indistinguishable activation en-
ergies for overall radiationless decay as for photoisomerization is
consistent with Troe’s proposal that decay from the trans-
phenallyl/benzyl intermediate gives mainly tt-DPB. If 1tt-DPB*
f 1tp-DPB* were the only radiationless decay process, we could
infer from the ratio of the pre-exponential values that R would
have to be as small as 0.10 and 0.28 in CH and MCH,
respectively. However, as pointed out above, this is not likely
in view of the small φctftt values in Table 6. We refrain from
attaching undue significance to the A values because narrow T
ranges are involved, and small errors in the slopes of the Arrhenius
plots in Figure 7 lead to large errors in the intercepts.
not leading to photoisomerization was first proposed to account
for smaller than expected photoisomerization quantum yields in
Bz by Whitten and co-workers32 and in CH by Yee and co-
workers.20 A kic = 2.5 ꢀ 108 s-1 value in H was calculated by
Velsko and Fleming on the basis of the nonlinear T dependence
of total nonradiative decay constants12 in H, and similar values
were estimated in the n-alkanes by Troe and co-workers.14a
Because those estimated kic values are too small to account for the
low photoisomerization quantum yields, Troe and co-workers
suggested that the assumption in eq 5 of equal partitioning from
the twisted singlet excited intermediate to the ground states of tt-
and ct-DPB isomers does not hold for DPB. They proposed a
decay ratio for 1tp-DPB* (Chart 3) of 3-4 in favor of tt-DPB and
assumed that torsional relaxation to the twisted intermediate
(presumably, 1tt-DPB* f 1tp-DPB*) is the dominant nonradia-
tive decay path of the planar excited state. We can offer three
objections to this interpretation. First, decay from twisted triplet
and singlet excited stilbene intermediates, although usually assumed
to give the two ground-state isomers with equal probability,
actually favors cis-stilbene, the higher-energy isomer.3b,21 If the
1
analogy holds, decay from tp-DPB* should favor the higher-
energy ct-DPB isomer. Second, the corollary to the proposed
1
preferential decay of tp-DPB* to tt-DPB is that, assuming the
same intermediate, ct-DPB f tt-DPB photoisomerization quan-
tum yields should be unusually large. The low φctftt values in
Table 6 show that, in all solvents studied thus far, this expectation
is not borne out. Third, fluorescence lifetimes of 1,5-diphenyl-
2,3,4,6,7,8-hexahydronaphthalene (HHN, Chart 4), indicate that
internal conversion rate constants exhibit similar Arrhenius
behavior as those of tt-DPB and are larger than those of tt-
DPB, despite the fact that in HHN, photoisomerization is
structurally inhibited.59 Alkyl substitution in this structural analogue
of tt-DPB is expected to increase the energy gap between the
21Ag and 11Bu states,22 and it is reasonable to conclude that facile
radiationless decay pathways are available to the 11Bu state of
HHN that do not require large torsional displacement in the
diene unit. It is difficult to understand, however, why such
pathways are not similarly effective in 1.
If reactive radiationless decay from 1tt-DPB* leads to ct-DPB
via 1tp-DPB*, it follows that unreactive radiationless decay does
not reach that intermediate. Assuming that the identity of the
activation energies is not a coincidence, one can imagine that, in
hydrocarbon solvents, unreactive and reactive radiationless de-
cays of 1tt-DPB* are on the same trajectory, but torsional
displacement in the shared transition state stops short of that
required for reaching a 90° twisted intermediate. Photoisome-
rization might then occur via inefficient leakage from that
1
1
transition state to tp-DPB*. Alternatively, tp-DPB* may not
be on the reaction pathway, and photoisomerization may occur
inefficiently, following internal conversion, by continuation in
the ground state of the small excited-transition-state torsional
displacement. Momentum considerations for continuation of
motion from the excited to the ground state may apply in the
As can be seen in Table 6, 81% of 1tt-DPB* molecules in PFH
undergo internal conversion without photoisomerization. The
1
dominant contribution of an unreactive tt-DPB* radiationless
decay pathway to the overall decay rate in PFH brings into
question the significance of theoretical treatments16 that are
2127
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
Figure 7. Arrhenius plots of Rktors (red 9) and knr (blue () in CH, panel a, and MCH, panel b.
latter case.60 Decay via a conical intersection that is attained by a
relatively small torsional distortion would be consistent with
observations on HHN, the structurally constrained tt-DPB
analogue, Chart 4. We stress here that proper interpretation of
photoisomerization is most efficient in Bz, where S1 is the 11Bu
state and the ΔEab energy gap is largest, and least efficient in
PFH, where S1 is the 21Ag state. Because efficient population of
the 21Ag state hampers photoisomerization instead of facilitating
it, the Birks extension11 of the OS model10 for photoisomeriza-
tion to R,ω-diphenylpolyenes is not valid and must be
abandoned.
1
the effect of medium friction on tt-DPB* radiationless decay
must be based on knowledge of the molecular motion with which
the medium interferes. We have shown that the usual assumption
that that motion is torsional displacement leading to 1tp-DPB* is
not correct.
The OS model for stilbene photoisomerization10 is, itself, of
questionable validity. The increased polarizability accompanying
the change from the gas phase to a hydrocarbon medium
stabilizes the 11Bu state of t-St by about 5 kcal/mol.56 Because,
by analogy with tt-DPB and ttt-DPH, the 21Ag state should be
insensitive to the polarizability change, the OS model predicts a
substantial increase in the torsional barrier of t-St upon moving
from the isolated molecule in the gas phase to solution. The
actual change, from 3.3 ( 0.2 kcal/mol2,62,63 in the gas phase to
an intrinsic barrier of 2.9 kcal/mol in n-alkanes,30,61b is, if anything,
in the opposite direction. The possibility that the torsional barrier
is created by the crossing of the 11Bu state with a higher Ag
excited state64 is no more attractive because that barrier should
also increase with increasing medium polarizability. It appears,
therefore, that the torsional barrier is a property of the 11Bu state,
a position that was championed by Troe and Weitzel.65
The large contributions of 1tt-DPB* radiationless decay path-
ways that do not lead to photoisomerization, in all solvents
studied, and especially in PFH, bring into question explanations
of medium effects on 1tt-DPB* fluorescence lifetimes that assign
all radiationless decay to the photoisomerization channel.
1
Our results on the T dependence of tt-DPB* radiationless
decay are in contrast to observations on tt-DPH, for which
photoisomerization was shown to be subject to much higher
activation energies than those obtained from the overall radia-
tionless decay rate constants.17c
The possible T dependence of the relative contribution of
OBT and BP photoisomerization channels remains to be estab-
lished. Simultaneous two-bond photoisomerization accounts for
20% of the photoisomerization events in H at ∼23 °C, and its
contribution may change with T and with medium viscosity.
Preliminary attempts to measure the T dependence of photo-
isomerization quantum yields in PFH have been stymied pri-
marily due to solubility problems. However, the shorter
fluorescence lifetime and smaller photoisomerization quantum
yield in that solvent at 20 °C lead us to expect large discrepancies
between activation parameters for the photoisomerization path
and for overall radiationless decay. The low photoisomerization
quantum yield in this solvent shows that ready access to the 21Ag
state suppresses the photoisomerization channel instead of
enhancing it, as would be expected if the Birks extension of the
OS mechanism applied. Consistent with this conclusion is the
observation that photoisomerization is most efficient in Bz, the
solvent in which the 21Ag state is least accessible. We conclude
that the 21Ag state is not involved as a photoisomerization
intermediate. Fast 21Ag f 11Ag internal conversion could
account for the short tt-DPB fluorescence lifetime in PFH.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: saltiel@chem.fsu.edu.
’ ACKNOWLEDGMENT
This research was supported by the National Science Foundation,
most recently by Grant No. CHE-0846636.
’ CONCLUSIONS
Our observations on the photoisomerization of the 1,4-di-
phenylbutadiene isomers in solution complement previous work
on the stilbenes2,3,7,21,61 and the 1,6-diphenylhexatrienes.17,22
The first three members of the R,ω-diphenylpolyene family
exhibit strikingly different photophysical and photochemical
behavior. Controlling factors are the proximity and energetic
order of the 11Bu and 21Ag states, their two lowest electronically
excited singlet states. Photoisomerization in the singlet-state
manifold is most efficient in t-St, for which S1 is the 11Bu state,
and least efficient in ttt-DPH, for which S1 is the 21Ag
state, although in ttt-DPH, reversing the 21Ag/11Bu state
order does not enhance photoisomerization.22 In tt-DPB,
’ REFERENCES
(1) Allen, M. T.; Whitten, D. G. Chem. Rev. 1989, 89, 1691–1702.
(2) Saltiel, J.; Sun, Y.-P. In Photochromism, Molecules and Systems; D€urr,
H. Bouas-Laurent, H., Eds.; Elsevier: Amsterdam, The Netherlands, 1990;
pp 64-164.
(3) (a) Saltiel, J. J. Am. Chem. Soc. 1967, 89, 1036–1037. (b) Saltiel, J.
J. Am. Chem. Soc. 1968, 90, 6394–6400.
(4) Malkin, S.; Fischer, E. J. Phys. Chem. 1964, 68, 1153.
(5) Sharafi, S.; Muszkat, K. A. J. Am. Chem. Soc. 1971, 93, 4119.
(6) Sumitani, A.; Nakashima, N.; Yoshihara, K.; Nagakura, S. Chem.
Phys. Lett. 1977, 51, 183.
2128
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129

The Journal of Physical Chemistry A
ARTICLE
(7) (a) Charlton, J. L.; Saltiel, J. J. Phys. Chem. 1977, 81, 1940–1944.
(b) Saltiel, J.; Charlton, J. L. Cis-Trans Isomerization of Olefins. In
Rearrangements in Ground and Excited States; de Mayo, P., Ed.; Academic
Press: New York, 1980; Vol III, pp 25-89.
J. Am. Chem. Soc. 1972, 94, 5903–5905. (d) Saltiel, J.; Townsend, D. E.;
Sykes, A. J. Am. Chem. Soc. 1973, 95, 5968–5973.
(36) G€orner, H. J. Photochem. 1982, 19, 343–356.
(37) Keery, K. M.; Fleming, G. R. Chem. Phys. Lett. 1982,
93, 322–326.
(38) Warshel, A. Nature (London) 1976, 260, 679–683.
(39) Harbison, G. S.; Smith, S. O.; Pardoen, J. A.; Winkel, C.;
Lugtenburg, J.; Herzfeld, J.; Mathies, R.; Griffin, R. G. Proc. Natl. Acad.
Sci. U.S.A. 1984, 81, 1706–1709.
(40) Sineshchekov, O. A.; Trivedi, V. D.; Sasaki, J.; Spudich, J. L.
J. Biol. Chem. 2005, 280, 14663–14668.
(41) Wada, Y.; Kawanabe, A.; Furutani, Y.; Kandori, H.; Ohtani, H.
Chem. Phys. Lett. 2008, 453, 105–108.
(42) (a) Saltiel, J.; Krishna, T. S. R.; Turek, A. M.; Clark, R. J. J. Chem.
Soc.: Chem. Commun. 2006, 1506–1508. (b) Saltiel, J.; Bremer, M. A.;
Laohhasurayotin, S.; Krishna, T. S. R. Angew. Chem., Int. Ed. 2008,
47, 1237–1240.
(43) (a) Saltiel, J.; Krishna, T. S. R.; Clark, R. J. J. Phys. Chem. A 2006,
110, 1694–1697. (b) Saltiel, J.; Krishna, T. S. R.; Laohhasurayotin, S.;
Fort, K.; Clark, R. J. J. Phys. Chem. A 2008, 112, 199–209.
(44) (a) Odani, T.; Matsumoto, A.; Sada, K.; Miyata, M. Chem.
Commun. 2001, 2004–2005 and references therein.. (b) Matsumoto, A.
Top. Curr. Chem. 2005, 254, 263–305. (c) Nishizawa, N.; Furukawa, D.;
Kobatake, S.; Matsumoto, A. Cryst. Growth Des. 2010, 10, 3203–3210.
(d) Sonoda, Y.; Kawanishi, Y.; Tsuzuki, S.; Goto, M. J. Org. Chem. 2005,
70, 9755–9763.
(8) Hudson, B. S.; Kohler, B. E. Annu. Rev. Phys. Chem. 1974,
25, 437–460.
(9) Hudson, B. S.; Kohler, B. E.; Schulten, K. In Excited States; Lim,
E. C., Ed.; Academic Press: New York, 1982; Vol. 6, pp 1-95.
(10) Orlandi, G.; Siebrand, W. Chem. Phys. Lett. 1975, 30, 352–354.
(11) (a) Birks, J. B.; Birch, D. J. S. Chem. Phys. Lett. 1975, 31, 608–610.
(b) Birks, J. B.; Tripathy, G. N. R.; Lumb, M. D. Chem. Phys. 1978,
33, 185–194. (c) Birks, J. B. Chem. Phys. Lett. 1978, 54, 430–434.
(12) Velsko, S. P.; Fleming, G. R. J. Chem. Phys. 1982, 76, 3553–3562.
(13) Birch, D. J. S.; Imhof, R. E. Chem. Phys. Lett. 1982, 88, 243–247.
(14) (a) Gehrke, C.; Mohrschladt, R.; Schroeder, J.; Troe, J.;
V€ohringer, P. Chem. Phys. 1991, 152, 45–56. (b) Mohrschladt, R.;
Schroeder, J.; Schwarzer, D.; Troe, J.; V€ohriger, P. J. Chem. Phys. 1994,
101, 7566–7579. (c) Schroeder, J.; Steinel, T.; Troe, J. J. Phys. Chem. A
2002, 106, 5510–5516 and references cited..
(15) (a) Anderton, R. M.; Kauffmann, J. E. J. Phys. Chem. 1994,
98, 12117–12124. (b) Anderton, R. M.; Kauffmann, J. E. J. Phys. Chem.
1994, 98, 12125–12132. (c) Anderton, R. M.; Kauffmann, J. E. J. Phys.
Chem. 1995, 99, 14628–14631.
(16) Dahl, K.; Biswas, R.; Maroncelli, M. J. Phys. Chem. B 2003,
107, 7838–7853.
(17) (a) Saltiel, J.; Ko, D.-H.; Fleming, S. A. J. Am. Chem. Soc. 1994,
116, 4099–4100. (b) Saltiel, J.; Wang, S.; Watkins, L. P.; Ko, D.-H.
J. Phys. Chem. A 2000, 104, 11443–11450. (c) Saltiel, J.; Krishna-
moorthy, G.; Huang, Z; Ko, D.-H.; Wang, S. J. Phys. Chem. A 2003,
107, 3178–3186.
(18) (a) Sandoval, A.; Zechmeister, L. J. Am. Chem. Soc. 1947,
69, 553–557. (b) Pickard, J. H.; Wille, B.; Zechmeister, L. J. Am. Chem.
Soc. 1948, 70, 1938–1944.
(19) Saltiel, J.; Dmitrenko, O.; Pillai, Z. S.; Klima, R.; Wang, S.;
Wharton, T.; Huang, Z.-N.; van de Burgt, L. J.; Arranz, J. Photochem.
Photobiol. Sci. 2008, 7, 566–577.
(20) Yee, W. A.; Hug, S. J.; Kliger, D. S. J. Am. Chem. Soc. 1988,
110, 2164–2169.
(21) Saltiel, J.; Marinari, A.; Chang, D. W.-L.; Mitchener, J. C.;
Megarity, E. D. J. Am. Chem. Soc. 1979, 101, 2982–2996.
(22) Saltiel, J.; Wang, S. Photochem. Photobiol. Sci. 2006, 5, 883–895.
(23) Moses, F. G.; Liu, R. S. H.; Monroe, B. M. Mol. Photochem.
1969, 1, 245–249.
(24) Saltiel, J.; Ganapathy, S.; Werking, C. J. Phys. Chem. 1987,
91, 2755–2758.
(25) Zimmerman, G.; Chow, L.-Y.; Paik, U.-J. J. Am. Chem. Soc.
1958, 80, 3528–3531.
(26) Melhuish, W. H. J. Phys. Chem. 1961, 65, 229–235.
(27) Meech, S. R.; Phillips, D. J. Photochem. 1983, 23, 193–217.
(28) Krishna, T. S. R.; Singh, A. K. Indian J. Chem. 1994,
33B, 413–414.
(29) Lamola, A. A.; G. S. Hammond, G. S. J. Chem. Phys. 1965,
43, 2129–2135.
(45) Sandros, K.; Sundhal, M.; Wennerstr€om, O.; Norinder, U.
J. Am. Chem. Soc. 1990, 112, 3082.
(46) (a) Ciorba, S.; Galiazzo, G.; Mazzucato, U.; Spalletti, A. J. Phys.
Chem. A 2010, 114, 10761–10768. (b) Bartocci, G.; Ciorba, S.;
Mazzucato, U.; Spalletti, A. J. Phys. Chem. A 2009, 113, 8557–8568
and previous papers in this series..
(47) Allen, M. T.; Miola, L.; Whitten, D. G. J. Phys. Chem. 1987,
91, 6099–6102.
(48) For a comprehensive critical review of spectroscopic literature
through 1989 see ref 2.
(49) Itoh, T.; Kohler, B. E. J. Phys. Chem. 1988, 92, 1807–1813.
(50) Itoh, T. Chem. Phys. Lett. 2001, 342, 550–554.
(51) Heimbrook, L. A.; Kohler, B. E.; Spiglianin, T. A. Proc. Natl.
Acad. Sci. U.S.A. 1983, 80, 4580–4584.
(52) Shepanski, J. F.; Keelan, B. W.; Zewail, A. H. Chem. Phys. Lett.
1983, 103, 9–14.
(53) Horwitz, J. S.; Kohler, B. E.; Spiglanin, T. A. J. Chem. Phys. 1985,
83, 2186–2190.
(54) Amirav, A.; Sonnenschein, M.; Jortner, J. Chem. Phys. 1986,
102, 305–312.
(55) Hudson, B. S.; Kohler, B. E. J. Chem. Phys. 1973, 59, 4984–5002.
(56) Sklar, L. A.; Hudson, B. S.; Petersen, M.; Diamond, J. Biochem-
istry 1977, 16, 813–818.
(57) Birks, J. B.; Dyson, D. J. Proc. R. Soc. London, Ser. A 1963,
275, 135–148.
(58) Strickler, S. J.; Berg, R. A. J. Chem. Phys. 1963, 37, 814–822.
(59) Yee, W. A.; Horwitz, J. S.; Goldbeck, R. A.; Einterz, C. M.;
Kliger, D. S. J. Phys. Chem. 1983, 87, 380–382.
(30) (a) Saltiel, J.; Sun, Y.-P. J. Phys. Chem. 1989, 93, 6246–6250. (b)
Sun, Y.-P.; Saltiel, J. J. Phys. Chem. 1989, 93, 8310–8316.
(31) Murov, S. L. Handbook of Photochemistry; Marcel-Dekker: New
York, 1973.
(32) Eastman, L. R., Jr.; Zarnegar, B. M.; Butler, J. M.; Whitten, D. G.
J. Am. Chem. Soc. 1974, 96, 2281–2282.
(33) (a) Saltiel, J.; Cires, E.; Turek, A. J. Am. Chem. Soc. 2003,
125, 2866–2867. (b) Sun, Y.-P.; Sears, D. F., Jr.; Saltiel, J. Anal. Chem.
1987, 59, 2515–2519.
(34) Saltiel, J.; Metts, L.; Wrighton, M. J. Am. Chem. Soc. 1970,
92, 3227–3229.
(60) Weiss, R. M.; Warshel, A. J. Am. Chem. Soc. 1979,
101, 6131–6133.
(61) (a) Saltiel, J.; Waller, A. S.; Sears, D. F., Jr. J. Am. Chem. Soc.
1993, 115, 2453–2465. (b) Saltiel, J.; Waller, A. S.; Sears, D. F., Jr.;
Garrett, C. Z. J. Phys. Chem. 1993, 97, 2516–2522.
(62) (a) Syage, J. A.; Lambert, W. R.; Felker, P. M.; Zewail, A. H.;
Hochstrasser, R. M. Chem. Phys. Lett. 1982, 88, 266–270. (b) Syage, J. A.;
Felker, P. M.; Zewail, A. H. J. Chem. Phys. 1984, 81, 4706–4723. (c)
Felker, P. M.; Zewail, A. H. J. Chem. Phys. 1985, 89, 5402–5411.
(63) Majors, T. J.; Even, U.; Jortner, J. J. Chem. Phys. 1984,
81, 2330–2338.
(64) Hohlneicher, G.; Dick, B. J. Photochem. 1984, 27, 215–231.
(65) Troe, J.; Weitzel, K. M. J. Chem. Phys. 1988, 68, 7030–7039.
(35) (a) Saltiel, J.; Metts, L.; Wrighton, M. J. Am. Chem. Soc. 1969,
91, 5684–5685. (b) Saltiel, J.; Metts, L.; Sykes, A.; Wrighton, M. J. Am.
Chem. Soc. 1971, 93, 5302–5303. (c) Saltiel, J.; Rousseau, A. D.; Sykes, A.
2129
dx.doi.org/10.1021/jp111482m |J. Phys. Chem. A 2011, 115, 2120–2129