Photochemistry of the Stilbenes in Methanol. Trapping the Common Phantom Singlet State

Article abstract of DOI:10.1021/acs.jpca.8b04011

A comparative study of the photochemistry of cis- and trans-stilbene in methanol shows that both isomers undergo methanol photoaddition giving similar yields of α-methoxybibenzyl in competition with cis-trans photoisomerization. Methanol addition occurs primarily following torsional relaxation of the lowest excited singlet states of each isomer, ¹c? and ¹t?, to a common twisted singlet excited state intermediate, ¹p?, initially called the phantom singlet state. The addition is consistent with the zwitterionic character of ¹p?. Ether forms by direct 1,2-addition of CH₃OH to the central carbon atoms and by 1,1-addition following rearrangement to 1-benzyl-1-phenylcarbene. Use of CD₃OD and GC/MS (gas chromatographic/mass spectroscopic) analysis of the ether products revealed that the ratio of carbene/direct addition pathways is higher starting from cis-stilbene. We conclude that ¹p? formed from ¹c? is hotter than ¹p? formed from ¹t?. Surprisingly, except for favoring the carbene pathway, the use of higher energy photons (254 vs 313 nm) does not affect the overall ether quantum yield starting from cis-stilbene but significantly enhances both pathways starting from trans-stilbene. It appears that carbene formation and direct methanol addition to higher trans-stilbene excited state(s) compete with relaxation to S₁. Substitution of D for the vinyl Hs of stilbene enhances the direct addition pathway more than 2-fold and strongly suppresses the carbene insertion pathway, revealing a large, kpc^d0/kpe^d2 = 6.3, primary deuterium isotope effect in the carbene rearrangement. The 2-fold increase in the ether quantum yield is due primarily to a 2.75-fold increase in the lifetime of ¹p? on deuterium substitution of the vinyl hydrogens.

Full text of DOI:10.1021/acs.jpca.8b04011

Subscriber access provided by University of Florida | Smathers Libraries
A: Spectroscopy, Molecular Structure, and Quantum Chemistry
Photochemistry of the Stilbenes in Methanol.
Trapping the Common Phantom Singlet State
Jack Saltiel, and Shipra Gupta
J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.8b04011 • Publication Date (Web): 21 Jun 2018
Downloaded from http://pubs.acs.org on June 22, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Photochemistry of the Stilbenes in Methanol.
Trapping the Common Phantom Singlet State
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Jack Saltiel,* and Shipra Gupta
Department of Chemistry and Biochemistry, Florida State University, Tallahassee, FL 32306-
4390
Corresponding Author
*
saltiel@chem.fsu.edu
ABSTRACT: A comparative study of the photochemistry of cis- and trans-stilbene in methanol
shows that both isomers undergo methanol photoaddition giving similar yields of α-
methoxybibenzyl in competition with cis-trans photoisomerization. Methanol addition occurs
1
primarily following torsional relaxation of the lowest excited singlet states of each isomer, c* and
1
1
t*, to a common twisted singlet excited state intermediate, p*, initially called the phantom singlet
1
state. The addition is consistent with the zwitterionic character of p*. Ether forms by direct 1,2-
addition of CH
benzyl-1-phenylcarbene. Use of CD
analysis of the ether products revealed that the ratio of carbene/direct addition pathways is higher
3
OH to the central carbon atoms and by 1,1-addition following rearrangement to 1-
3
OD and GC/MS (gas chromatographic/mass spectroscopic)
1
1
1
starting from cis-stilbene. We conclude that p* formed from c* is hotter than p* formed from
1t*. Surprisingly, except for favoring the carbene pathway, the use of higher energy photons (254
ACS Paragon Plus Environment
1

The Journal of Physical Chemistry
Page 2 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
vs 313 nm) does not affect the overall ether quantum yield starting from cis-stilbene, but
significantly enhances both pathways starting from trans-stilbene. It appears that carbene
formation and direct methanol addition to higher trans-stilbene excited state(s) compete with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
relaxation to S . Substitution of D for the vinyl Hs of stilbene enhances the direct addition pathway
1
more than two-fold and strongly suppresses the carbene insertion pathway, revealing a large,
d2
k
pcd0/kpe = 6.3, primary deuterium isotope effect in the carbene rearrangement. The two-fold
1
increase in the ether quantum yield is due primarily to a 2.75-fold increase in the lifetime of p*
on deuterium substitution of the vinyl hydrogens.
INTRODUCTION
Extensively studied, the stilbenes serve as the prototypical example of molecules undergoing cis-
trans photoisomerization.1-3 They have played a central role in studies of the effect of solvent
friction on S
torsional relaxation, the initial step in isomer interconversion.2,3 In hydrocarbon
1
solvents, torsional relaxation from the trans side, the major process competing with fluorescence
1
at all temperatures (Ts), is subject to an energy barrier that is in part intrinsic, 2.9 kcal/mol, and
 
in part diffusional, 0.4 E , where E
is the activation energy of viscous flow.4-6 Torsional
.7 Both

relaxation from the cis side is entirely diffusional and subject to most, if not all, of E
1
isomers give a twisted intermediate p*, initially named the phantom excited singlet state in
8,9
1
1
anticipation of its fleeting existence. Decay from p* slightly favors cis-stilbene, c, over trans-
1
1
1
stilbene, t, and there is a reverse torsional component from p* to t* that is responsible for a
1
1
7,10,11
minor adiabatic c*  t* reaction channel.
Conrotatory cyclization to 4a,4b-
1
dihydrophenanthrene, DHP, is a competing c* reaction, less prominent than cis-trans
photoisomerization.12-15 Ether formation was reported as a minor reaction channel on irradiation
ACS Paragon Plus Environment
2

Page 3 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
1
16
17,18
of t in methanol (MeOH) or 2,2,2-trifluoroethanol
(TFE). In the case of TFE, of the
1
mechanisms considered for the direct alcohol reaction with excited stilbene, protonation of t*
was favored.17,18
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We recently described our work on the mechanism of ether formation resulting from the
1
9
irradiation of the three 1,4-diphenyl-1.3-butadiene (DPB) isomers in ethanol. The key excited
state that reacted with the alcohol was proposed to be a twisted zwitterionic trans-phenallyl
cation/benzyl anion formed efficiently from trans,trans- and cis,trans-1,4-diphenyl-1.3-
19
butadiene and inefficiently from cis,cis-1,4-diphenyl-1.3-butadiene. A reevaluation of the
stilbene results led us to propose that alcohol addition to a zwitterionic twisted phantom singlet
1
9
state is a more likely path to ether. The results in this paper substantiate that proposal.
EXPERIMENTAL SECTION
Materials. The perhydrostilbenes, t- and c-Std0 and their α,α´-dideuterio derivatives, t- and c-Std2
were previously described.5,11 They were purified by chromatography with hexane eluent on
silica immediately before use. Methanol (Sigma Aldrich, Spectrometric Grade, Anhydrous),
methanol-d
4
(Cambridge Isotopes, 10x 0.5 mL pack), hexane (HPLC grade, EMD), cyclohexane
(
VWR, Spectrometric Grade, anhydrous) and silica, 230-400 mesh, (Sigma-Aldrich, technical
grade) were used as received.
Irradiation Procedures. Sample preparation and degassing procedures were described
20
previously. Sets of 13 mm o.d. Pyrex tubes fitted with standard taper joints and grease traps
were loaded with 2.0 mL aliquots of St solutions. They were degassed using 4 freeze-pump-thaw
-4
cycles to < 10 Torr and flame sealed at a constriction. All operations, including analyses, were
performed under nearly complete darkness (red light). Irradiations were carried out in a Moses
ACS Paragon Plus Environment
3

The Journal of Physical Chemistry
Page 4 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
21
merry-go-round apparatus immersed in a water bath thermostated to 22.0 °C. A heating coil
connected to a thermoregulator (Polyscience Corp.) was used to control the temperature to  0.5
°
C. The photoisomerization of t-St in CH
3
OH was used for actinometry, tc = 0.54.22-24 Hanovia
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
50-W medium pressure Hg lamps were employed. The 313 nm Hg line was isolated using a
20
potassium perchromate/potassium carbonate filter solution1 Irradiations in CD
carried out in N outgassed ampules using 3.0 mL aliquots at 23 0.5 °C. Small samples, 0.3 L,
were removed periodically for GC analysis.
3
OD were
2
To conserve the deuterated solvent, 254 nm irradiations of 0.50 mL aliquots of N
2
outgassed
CD OD solutions were carried out in capped quartz NMR tubes using a previously described
3
2
5
cylindrical merry-go-round attached to a stirring motor. The samples were placed at a distance
of 3 in from the Hg lamps. A 12-W low pressure Hanovia Hg lamp inserted in a cylindrical
Vycor glass sleeve was employed. Temperature was maintained in the 24.8-25.6 °C range.
26
1
Analytical Procedures. Analysis of product composition was by GC, by H NMR and, for
deuterated samples, by GC/MS. A Bruker/Varian CP-3800 gas chromatograph equipped with an
electronic integrator was used. Samples were concentrated to approximately 1 mL under a stream
of N and 0.1-0.2 µL aliquots were injected onto a Varian Factor Four column. Analyses were
2
o
o
o
carried out at an initial temperature of 100 C that was increased to 220 C using a 20 C/min
o
ramp. The injector inlet and the FID detector were both held at 250 C, and an initial injector
split ratio of 30:1 was increased to 60:1 after 0.75 min and decreased to 5:1 after 1.25 more min
1
to conserve carrier gas. A Bruker Avance III 500 MHz spectrometer was used for H NMR
spectra. GC response factors of the St isomers were assumed to be identical. The GC molar
response factor of the ether photoproduct, E, was determined to be 0.503 relative to t-St by
ACS Paragon Plus Environment
4

Page 5 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
1
relating the quantitative H NMR analysis of an irradiation mixture to GC peak areas for the
same mixture. This ratio of E/St molar response factors is in excellent agreement with 0.507, the
19
analogous E/DPB ratio. Accordingly, ether GC areas were multiplied by 1.98 before
7
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
calculating % molar conversions. Reduction of FID molar response factors with introduction of
heteroatoms in hydrocarbons is a known phenomenon.27,28
A Varian Cary 300B UV-Vis spectrophotometer was used to record absorption spectra at 20.5
oC using a thermostated cell holder.
RESULTS
Stilbene-d
0
– Methanol-d , 313 nm. Six pairs of sealed ampules containing degassed methanol
0
solutions were exposed to 313 nm excitation in parallel. Ampule pairs were periodically removed
from the merry-go-round for GC analysis. Initial concentrations were [c-St] = 0.0154 M and [t-St]
=
0.0140 M. Isomer contaminations at 0-time, determined by GC analysis, were 0.25% c-St in t-
St and 0.13% t-St in c-St. The quasi-photostationary state attained on prolonged irradiation of the
same stilbene solutions, (89.6  0.1)% c-St, is in reasonable agreement with the reported
2
0
t
photostationary state of 91.7% c-St in n-pentane. Photoisomerization conversions at time t, f
c
t
29
and f starting from t- and c-St, respectively, were corrected for back reaction and for initial
t
isomer contamination,20 f
0, eq 1, using the quasi-photostationary fraction of stilbene product, f
eq.
x
x
eq
0
fx  fx
corr
eq
fx  fx ln (
)
(1)
eq
t
fx  fx
Conversions to the ether, corrected for the difference in relative ether and stilbene GC molar
responses, and back reaction corrected conversions to the two stilbene isomers are linearly
dependent on time, Figure 1. Photoisomerization quantum yields are given in Table S1.
ACS Paragon Plus Environment
5

The Journal of Physical Chemistry
Page 6 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
7
6
5
4
3
2
1
0
b
1
0
8
6
4
2
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
20
40
60
0
20
40
60
t/min
t/min
Figure 1. Conversions in CH
3
OH starting from t- and c-St in panels a and b, respectively – lines
are blue for St and red for ether product (P). Ether % conversions are multiplied by 102.
Stilbene-d
initial concentrations of [c-St] = 0.0216 M and [t-St] = 0.0210 M were transferred into a pair of
Pyrex ampules. The ampules were outgassed with a gentle stream of N gas using latex tubing
0
– Methanol-d
4
, 313 nm. Aliquots, 3 mL, of stilbene solutions in CD
3
OD having
2
and sealed with a rubber septum. They were irradiated in parallel in the merry-go-round at 313
nm at 22.5 ± 0.5 °C and the progress of the reactions was monitored by GC using no more than
0
.3 L injections. The N
2
outgassing was repeated after each injection. Prolonged irradiation
under these conditions led to a quasi-PSS of 89.9% c-St and 10.1 % t-St. Isomer contaminations
1
0
8
6
4
2
0
8
6
4
2
0
b
a
0
5
10
0
5
10
t/h
t/h
Figure 2. Conversions at 313 nm in CD OD starting from t- and c-St in panels a and b,
3
respectively – lines are blue for St and red for ether. Ether % conversions are multiplied by 102.
ACS Paragon Plus Environment
6

Page 7 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
at 0-time, determined by GC analysis, were 0.09% c-St in t-St and 0.39% t-St in c-St.
Product conversions over an irradiation period of 9 h are plotted against irradiation time in
Figure 2. Quantum yields based on the slopes of the lines in Figure 2 are given in Table S1.
Because there is a systematic deviation of the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
points from the photoisomerization lines in
panels a and b of Figure 2, isomer conversions in
the two directions were also plotted against each
other. The resulting line has a correlation
2
coefficient r = 0.999 and its slope, 0.676, gives ϕct = 0.37
5
, in very good agreement with the
value in Table S1. For the experiment in Figure 2, ratios of ether formed via carbene
rearrangement leading to 1,1-addition of CD OD and direct 1,2-addition of CD OD, 0.456 and
.678 starting from t- and c-St, respectively, were determined by GC/MS immediately after the
3
3
0
last irradiation period. They were
based on the ratios of the 125 to 124
mass base peaks after correction of the
125 peak for the expected 8.99% M + 1
13
2
17
contributions due to C, H and O
natural abundances, Scheme 1. In
Scheme 1 the base peak fragments are
circled. An independent experiment in
which GC/MS analysis was performed after St conversions close to the PSS value and 2.1%
ether yields had been achieved, the corrected 125/124 peak ratios were 0.555 and 0.529, starting
from t- and c-St, respectively. The ether forming steps are shown in Scheme 2.
ACS Paragon Plus Environment
7

The Journal of Physical Chemistry
Page 8 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Stilbene-d
0
– Methanol-d
4
, 254 nm. Aliquots, 0.50 mL, of c- and t-St solutions 0.0199 and
0
.0250 M, respectively, in anhydrous methanol were delivered into two quartz NMR tubes. The
methanol was completely evaporated and replaced with 0.50 mL of CD OD. Each tube was then
outgassed with a slow stream of N for 20 min to minimize solvent evaporation and capped. A
2-W low pressure Hg lamp placed in a Vycor sleeve (220 nm cutoff) was used to irradiate the
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
samples in parallel in a merry-go-round designed for NMR tubes. The progress of the reaction
was followed by GC analysis. Prolonged irradiation under these conditions led to a quasi-PSS of
4
0
7.4% c-St and 52.6 % t-St. Isomer contaminations at 0-time, determined by GC analysis, were
.09% c-St in t-St and 0.39% t-St in c-St. Product conversions over an irradiation period of 60
1
0
8
6
4
2
0
8
6
4
2
0
b
a
0
20
40
60
0
20
40
60
t/min
t/min
Figure 3. Conversions at 254 nm in CD OD starting from t- and c-St in panels a and b,
3
respectively – lines are blue for St and red for ether. Ether % conversions are multiplied by 102.
min are plotted against irradiation time in Figure 3. Quantum yields based on the slopes of the
lines in Figure 3 are given in Table S1. For the experiment in Figure 3, ratios of direct 1,2-
addition of CD
3
OD to 1,1-addition of CD
3
OD via carbene rearrangement of 0.82
4
and 1.0
3
starting from t- and c-St, respectively, were again based on the corrected ratios of the 124 to 125
mass peaks determined by GC/MS after the last irradiation period. Quantum yields from an
ACS Paragon Plus Environment
8

Page 9 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
independent experiment in which t-St CH
3
OH solutions were irradiated without the Vycor filter
sleeve are in excellent agreement with those in which the filter sleeve was employed, Table S1.
trans-Stilbene-d
to that described above for the 313 nm irradiation of the undeuterated stilbenes. Isomer
contamination at 0-time, determined by GC analysis, was 0.12% c-St-d . The irradiated methanol
solution was 0.0220 M t-St-d . Product conversions over an irradiation period of 122 min,
2
– Methanol-d
0
, 313 nm. The procedure used in this experiment is identical
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
corrected for back reaction are plotted against irradiation time in Figure 4. The quantum yield for
ether formation,
12
1
0
8
6
4
2
0
0
50
100
t/min
Figure 4. Conversions for the 313 nm irradiation of t-St-d
2
in CH
3
OH – lines are blue for c-St-d
2
and red for ether. Ether % conversions are multiplied by 102.
based on the slopes of the lines in Figure 4 and the assumption that ϕtc = 0.54 is 0.0057
9
, Table
S1. The ratio of carbene insertion to direct addition pathways to ether, 0.120, was based on the
GC/MS relative intensities of the 121 (via carbene insertion) and 122 (via direct addition) peaks.
DISCUSSION
ACS Paragon Plus Environment
9

The Journal of Physical Chemistry
Page 10 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
16
Laarhoven and coworkers used CH
1) direct 1,2-addition of CH O/D to the double bond to give threo and erythro ethers and (2)
1,1-addition of CH O/D via rearrangement to a carbene, followed by carbene insertion into
CH OD. Higher energy photons were found to favor the carbene over the direct ether pathway
3
OD to find two significant routes to α-methoxybibenzyl:
(
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
and the threo over the erythro adduct. Two reactive states were considered, in addition to the
lowest excited states, one of which is either a higher energy electronic state – Rydberg states
3
0
have been proposed in carbene formation from alkenes – or a vibrationally excited S state
1
involved in carbene and threo-ether formations. The results were deemed inconclusive as to the
mechanism of ether formation by direct methanol addition. Involvement of an excited singlet
1
zwitterionic twisted intermediate, p*, was ruled out and addition of methanol to planar t- and c-
1
1
St S
1
states, t* and c*, was considered a viable mechanism because 1,2-diphenylcyclopentene,
3
1
erroneously thought to have an obligatory planar excited state, also undergoes the reaction. Syn
1
1
additions of CH
3
OD to t* and c* were proposed as routes to threo and erythro ethers,
24
1
respectively. The possibility of addition to a biradical p* was also considered. What was
overlooked is that irradiation of t-St in methanol at 300 nm leads to continuous unabated ether
formation as a cis rich (~ 85% c-St) quasi-photostationary state is quickly approached and
maintained (see Figure 1 in ref 24). Formation of the ethers at an almost constant rate,
independent of stilbene isomer composition, indicates that the ethers form from the two isomers
with very similar efficiencies.
The photoaddition of 2,2,2-trifluoroethanol (TFE) to stilbene was reported, subsequently, by
Roberts and Pincock.17,18 In the more acidic TFE, St was found to give ether photoadduct faster
1
7
than in methanol. Use of TFE-OD established that, due to faster protonation, the carbene
1
8
pathway to ether is less important than in CH OH. The fluorescence of methoxy-substituted
3
ACS Paragon Plus Environment
1
0

Page 11 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
trans-stilbenes is quenched by TFE in acetonitrile, and formation of cation intermediates was
1
8
observed by pulsed laser excitation. Protonation of S
1
followed by nucleophilic attack on the
resulting cation was assumed to apply also to the parent t-St. However, here too, the fact that the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
1
rate of ether formation was found to be independent of cis/trans St composition, points to p*
as a common reactive stilbene intermediate. To the extent that the initial S states of the stilbene
isomers are involved in ether formation, one would expect t* to play a more prominent role than
1
1
1
1
OH, 43 ps at 297 K,32,33 is almost 10 times longer than the
2
c* because the lifetime of t* in CH
3
1
34-39
lifetime of c*, 0.48±0.02 ps under similar conditions.
It was proposed that the shorter
stilbene S
1
lifetimes in more polar solvents generally, and in methanol specifically, were due to
the lowering of the energy of the potential energy surface along torsional coordinates leading to a
1
23.24
relatively polar p*.
This is not easily established because experimental Arrhenius activation
energies for torsional relaxation in solution involve coupled solvent/solute dynamics and do not
readily reveal solute-specific properties. We had shown that experimental activation enthalpies,
, the activation energy for viscous flow.4,5

ΔHobsd, in the n-alkane family depend linearly on E
They can be calculated by adding 0.4 E
relaxation.4,5 In changing the solvent from hexane to methanol

to 2.9 kcal/mol, the intrinsic energy barrier for torsional
4
32,33
the activation energy for the
1
1
t*  p* process increases from 4.1 to 4.6 kcal/mol (the value of the Arrhenius activation
4
energy in hexane is the activation enthalpy plus 0.6 kcal/mol for RT). The activation energy for
1
1
the t*  p* process in methanol is anomalously high compared to higher members of the n-
2
1

alcohol series where the linear ΔHobsd vs E relationship is not obeyed. The short t* lifetimes in
methanol are due to compensation of the larger activation energies by even larger pre-
exponential factors. The corresponding effective activation energy change on changing the
1
1
solvent for the c*  p* process from hexane to methanol is not known. However, it has been
ACS Paragon Plus Environment
1
1

The Journal of Physical Chemistry
Page 12 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
7
shown that in hydrocarbon solvents that process experiences no intrinsic barrier. It occurs as a
7
diffusion controlled process, subject to almost the entire activation energy of viscous flow. No
such simple relationship exists in alcohols where stronger specific solute/solvent and
solvent/solvent hydrogen bonding interactions are involved whose disruption as the molecule
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
proceeds to a zwitterionic p* probably fail to keep up with the molecular motion of the solute.
Solvent reorganization contributes significantly to the activation enthalpies and entropies of
1
1
torsional relaxation, especially in the much faster c*  p* direction. Torsional barriers starting
1
1
38,40
from t* and c* have also been estimated from isoviscosity Arrhenius plots.
They are of
dubious physical significance, not least of all, because of their reliance on sheer viscosity, a
macroscopic property, as a measure of solvent friction.2,4-6 At least in the family of n-alkanes,
torsional rate constants are much better behaved when microviscosities based on empirical solute
diffusion coefficients are employed instead of sheer viscosity.4-6 However, there is no reason to
expect microfriction approaches to work with alcohols. Fleming and coworkers, Waldeck and
3
coworkers and others have discussed these issues in greater detail. There is ample evidence that
in the gas phase the OH proton of methanol hydrogen bonds to the π electron cloud of
benzene.41,42 But as the number of methanol molecules in the complex increases the nature of the
interaction with benzene changes as methanol molecules hydrogen bond more strongly to each
other than to benzene. Similar hydrogen bonding should also involve the phenyl groups of
stilbene. To account for the methanol solvent effect in solution one would have to know how
hydrogen bonding interactions change as each excited stilbene isomer undergoes torsional
1
motion over a polar/polarizable transition state to a zwitterionic p*.
1
1
The p* intermediate. Early c-St transient absorption studies assigned a broad spectrum to c*
with max ~ 700 nm whose decay,  = 1.0 ps in hexane, gives rise to a transient with max ~ 335
c
ACS Paragon Plus Environment
1
2

Page 13 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
34
nm that was attributed to hot t-St product. An upper limit of 0.15 ps was estimated for the
1
34
7
lifetime, 
p
, of p*. Our estimates of  of 0.21 to 0.26 ps in hexane were based on steady state
p
1
1
1
7,10,11
measurements of the efficiency of the adiabatic c*  p*  t* channel.
A c-St transient
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
3
absorption study employed more sophisticated instrumentation to refine and extend the initial
transient observations.34-37 In hexane, the c* transient, max = 635 nm, was observed to decay to
1
1
1
1
the 335 nm transient, now assigned to p*, which subsequently decays to hot ground state c/ t
product whose cooling leads to sharpening of the absorption band. Two kinetics models were
employed to account for the time evolution of the transients. The authors expressed preference
1
1
for the model in which the c* to p* conversion is reversible over the usual model of sequential
1
1
1
43
irreversible c*  p*  p steps. Use of the reversible kinetics model gives  = 0.33 ps in
p
4
3
hexane at room temperature. But, this reversibility model can be discarded because, assuming a
1
common p* intermediate, reversibility requires that excitation of t-St give measurable amounts
of dihydrophenanthrene, DHP, in contradiction to steady state observations that established long
4
4
45-47
ago that DHP forms exclusively from c-St.
The induction period in the time evolution of
16
the DHP photoproduct in methanol starting from t-St demonstrates the obligatory excitation of
1
c-St in DHP formation. Assuming irreversible formation of p* gives 
p
= 0.23 ps in excellent
agreement with our estimate of 
.7 The biexponential decay model was also used by the authors
p
48
in a more recent publication. Observation of DHP absorption following t-St excitation and the
1
1
absence of t* transient absorption following c-St excitation again led to the conclusion that t*
1
1
 p*  c* adiabatic isomerization occurs in hexane, whereas the reverse adiabatic pathway
1
1
1
48
c*  p*  t* does not. The authors did not consider that, based on our steady state
7
1
1
fluorescence observations, only 0.3% of the molecules that reach p* give t* and such a small
yield would be undetectable in transient absorption. The observation of DHP following 1t
ACS Paragon Plus Environment
1
3

The Journal of Physical Chemistry
Page 14 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
excitation (Fig. 16 in ref 48) may be due to excitation of photoproduct c in the laser beam path
by a second photon.
1
A recent study found biexponential c* fluorescence decay in cyclohexane with 
c
= 1.2 ps and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
′

c
= 0.23 ps and assigned the 0.23 ps process to the initial partitioning between
1
photoisomerization and photocyclization pathways from c*′, an early excited state, and the 1.2
1
ps process to the subsequent c* that undergoes torsional relaxation along the
49
photoisomerization pathway. Although resolved fluorescence spectra associated with the two
1
components are very similar, it was proposed that they originate from significantly different c*′
1
and c* geometries on an energetically flat potential energy surface. However, geometries of
excited c-St that differ in the degree of planarity would be expected to have different
fluorescence spectra and if they lay on an energetically flat potential energy surface should be
present as an equilibrium mixture. Decay to and from an equilibrium mixture involving two
1c*conformers, one poised for torsional relaxation and the other poised for cyclization to DHP,
1
could account for biexponential c* fluorescence and transient absorption decays.
The Kinetics of Ether Formation. The results in Table S1 are generally consistent with the
1
6
observations of Laarhoven and coworkers, but their more quantitative nature reveals hitherto
hidden mechanistic details. The observation of similar ether quantum yields from t- and c-St is
1
most important because it implicates p* as the major ether precursor by direct methanol 1,2-
1
addition and by 1,1-adition following p* rearrangement to a carbene. As reported by Laarhoven
16
and coworkers, higher energy photons give higher ether yields. If the higher energy pathway to
1
1
ether involved formation of hot p*, it would be more likely to originate from c* because, absent
1
1
1
1
an intrinsic torsional barrier, the c*  p* process is faster than the activated t*  p* process.
ACS Paragon Plus Environment
1
4

Page 15 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
The larger contribution of the carbene pathway to ether on 313 nm excitation of c-St in CD OD
3
is consistent with this expectation. We were surprised, therefore, to find that the enhanced ether
yield at 254 nm is due entirely to excitation of t-St. It appears that there is a pathway to ethers
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
that involves direct methanol addition and carbene formation via a higher excited state of 1t.
1
1
Whether these reactions occur in the initial transoid higher excited state, t*´, or following t*´
1
1
p*´ torsional relaxation of this state to a higher energy twisted intermediate, p*´, remains to be
established. In addition to insertion to methanol, Laarhoven and coworkers established that the
carbene undergoes rearrangement back to ground state stilbenes.16
Zwitterionic twisted intermediates, first proposed by Dauben and coworkers to explain the
5
0
stereospecific photocyclization of trans-3-ethylidenecyclooctene and the photoaddition of
methanol to 1,3-dienes,51,52 found initial theoretical support in Salem’s sudden polarization close
5
3
1
to orthogonal olefin geometries. The zwitterionic nature of p*, the twisted stilbene conical
intersection, CI,54-57 is supported by recent theoretical calculations that predict decay through a
1
CI attained in the B
u
1
S state by central bond twisting along with pyramidalization of one of the
benzylic moieties.58-62 There is also ample experimental evidence for the involvement of
zwitterionic twisted intermediates in the
63
photoisomerizations of 1,3-dienes, 1,2-
6
4
19
diarylethenes, 1,4-diphenylbutadiene and
1,6-diphenylhexatriene.65,66 Laarhoven and
1
coworkers suggested the possibility that the rearrangement of p* to the carbene involves a
1
6
1
vibrationally hot diradical intermediate. Hydride shift in a zwitterionic p* provides a more
attractive pathway, Scheme 3.
ACS Paragon Plus Environment
1
5

The Journal of Physical Chemistry
Page 16 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The well-established stilbene photoisomerization mechanism is augmented in eqs 2 – 16 to
account for carbene and ether formation,
1
1
1
t + h´  t*´
(2)
(3)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
t + h  t*
1
1
t*´  t*
(4)
1
1
t*  t + h
(5)
(6)
1
1
t*  p*
1
1
1
p*   t + (1 - ) c
(7)
1
1
c + h  c*
(8)
1
1
c*  p*
(9)
1
c*  DHP
p* + MeOH  E12
(10)
1
(11)
(12)
(13)
(14)
(15)
(16)
1
1
p*  Carb
Carb + MeOH  E11
1
1
1
1
Carb  t + c  P
t*´ + MeOH  E12
1
1
1
t*´  Carb
where  is the fraction of twisted excited singlets that decays to t and ktt, k
1
f
, ktp, k
d
, kcp, kcDHP, kpe
,
k
pc, kec, kcr, kt´d and k t´c are the rate constants assigned to eqs 4 - 16, respectively. All the rate
constants are unimolecular except for kpe and kt´d the rate constants for methanol 1,2-addition to
1
1
p* and t*, respectively, which are pseudo-unimolecular. DHP stands for dihydrophenanthrene,
E12 and E11, designate ether product formed by direct 1,2-addition and by 1,1-carbene insertion,
1
respectively, and Carb stands for singlet 1-benzyl-1-phenylcarbene. P stands for phenanthrene
1
formed from Carb via the sequence of 1,2-hydride shift to the stilbenes, photochemical
1
conversion of c-St to DHP and oxidation of the latter (see below). DHP forms from c* only, eq
ACS Paragon Plus Environment
1
6

Page 17 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
0. Omitted steps from the above mechanism are c* fluorescence and the adiabatic p*  t*
conversion. The low quantum yields of those steps, fc ~ 10 and c*t* ~ 0.002, in hexane,7,10,11
that are likely to be even smaller in the more polar methanol solvent, renders them negligible
within the accuracy of our GC measurements. By eq 16 we do not wish to imply that the higher
t-St excited state goes directly to carbene. It may do so via a higher electronic or vibrational 1p
state.
-4
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Application of the steady state approximation on all excited species gives:
(
17)
tc  (1  )ktp f
  k 
(18)
ct
cp
c



t]
ct
c
t
(
) 

(19)
s
[
c]
tc
where tc and ct, are the photoisomerization quantum yields in the trans  cis and cis  trans
1
directions, respectively, α is the fraction of twisted intermediates, p*, that gives t-St, 
f
is (k +
f
-1
1
-1
1
k
tp) , the fluorescence lifetime of t*, and  is (kcDHP + kcp) , the lifetime of c*. Strictly
c
speaking, the right sides of eqs 17 and 18 should be multiplied by k
d
/(k + kpc + kpe), the fraction
d
1
of p* that decays to the stilbenes. That fraction is too close to unity for its use to be justified.
Examination of the quantum yields in Table S1 shows that use of tc = 0.54 for actinometry gives

ct = 0.38 for both 313 and 254 nm excitation, in agreement with previous photoisomerization
24
23
quantum yield measurements in methanol: tc = 0.54 ± 0.05 and ct = 0.38 ± 0.02. Those
values are about 8% higher than values in hydrocarbon solvents,20,22.45 due to more rapid
torsional relaxation in methanol,23,24 consistent with the observation of the shorter t* and c*
1
1
8
-1
5
lifetimes. The radiative rate constants, k , in hexane, 6.3 x 10 s , is known and its value in
f
ACS Paragon Plus Environment
1
7

The Journal of Physical Chemistry
Page 18 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
-1
= 3.75 x 108n1.65
,
methanol 6.0 x 10 s , can be estimated using the empirical relationship k
f
8
-1
1
where 3.75 x 10 s is the radiative decay of t* under isolated conditions and n is the refractive
5
32
index of the solvent. The decrease in 
f
from 76 to 43 ps at 24 °C, should lead to a similar
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
decrease in 
f
from 0.047 in hexane to 0.026 in methanol. Since, to a very good approximation,
k
tp

f
= 1 - 
f
, α = 0.45 can be calculated using eq 17. Use of eq 18 and ct = 0.38 then leads to the
1
conclusion that 84.4% of c* undergo torsional relaxation and the remaining 15.6% decay via the
1
DHP channel. A 70%/30% torsional/DHP channel split was proposed for c* decay in
hydrocarbon media.67,68 This was probably based on the lower ct = 0.3520,63 in hydrocarbon
68
solvents and the usual assumption that α = 0.5. Use of tc = 0.50 and 
f
= 0.047 gives α = 0.47
5
in hexane, the procedure we applied for the results in methanol gives a 74%/26% torsional/DHP
channel split in hexane. The observed decrease of the DHP quantum yield from 0.15 in hexane
5
23
to 0.07
7
in methanol is in excellent agreement with the attenuation of the DHP decay channel.
1
1
1
37,68
Since c*  DHP* is the process that competes with torsional relaxation to p*,
about half
1
1
the c* molecules that reach DHP* give ground state DHP instead of one third as previously
proposed.7,69
Similar small ether quantum yields of 0.0029 and 0.0020 from t- and c-St, respectively, are
1
consistent with the involvement of p* in ether formation. They represent the sum of ether
1
formed by direct addition of methanol to p* and carbene insertion in methanol. The
contributions of the two pathways will be considered below. The conclusion that the lowest
1
excited state of t-St, t*, does not react with methanol validates previous interpretations of
fluorescence lifetimes that consider torsional relaxation as the sole competing radiationless
channel in alcohols.3,4,24,33,40
ACS Paragon Plus Environment
1
8

Page 19 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
1
The Common p* Intermediate. Use of CD
3
OD as solvent allowed separation of ether
quantum yields from t- and c-St, te and ce, respectively, into contributions of ether formed by
direct methanol addition, te and ce , and of ether formed via carbene insertion, te and cec,
Table S1. For a truly common intermediate, the ratio of the two ether forming paths should be
identical starting from either isomer. The fact that the carbene insertion product is favored more
d
d
c
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
starting from c-St, does suggest that at 313 nm the p* that forms from c* is initially
vibrationally hot and more prone to rearrangement.
Applying the steady state approximation to all reactive species, the mechanism in eqs 2-16
gives
d
ce

 k  k 
(
20)
cp
c
pe
p
c
  k  k  k 
(21)
ce
cp
c
pc
p
ec carb
d
te

 k  k   k 
(22)
(23)
tp
t
pe
p
t'e t'
c
te

 k  k  k 
 k  k 
tp
t
pc
p
ec carb
t'c t' ec carb
1
1
where 
c
, 
t
and carb are the lifetimes of c*, t* and the carbene. The two ether pathways cannot
be distinguished when CH
3
OH is the solvent and observed ether quantum yields are the sum of
eqs 20 and 21 for c-St and the sum of eqs 22 and 23 for t-St. With λexc = 313 nm, the results of
the first experiment in Table S1 give indistinguishable ether/isomerization quantum yield ratios
1
of 0.0054 and 0.0055 starting from t- and c-St, respectively. This would be expected if p* were
the common ether precursor and if the partitioning fraction, α, were equal to 0.5. Our derived
value of α = 0.45 indicates that those ether/isomerization ratios should differ by about 18%,
(te/tc)/(ce/ct) = α/(1-α) = 0.82. In view of the experimental uncertainty in our ether yields and
ACS Paragon Plus Environment
1
9

The Journal of Physical Chemistry
Page 20 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the small deviation of this quantity from unity, the results of the first experiment in Table S1 are
1
consistent with the conclusion that with low energy photons p* and its carbene rearrangement
1
product are the sole ether precursors. We can rule out the possibility that the t*´ pathway, eqs 2,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
5, 16, contributes to ether formation at 313 nm because that would increase te and diminish,
instead of enhance, the (te/tc) ratio.
1
6
The Excitation Wavelength Effect. Laarhoven and coworkers irradiated t-St in CH OD and
3
c
/
d. on changing
found a significant increase in the carbene insertion/direct addition ratio, 
e
e
λ
exc, from 0.6 ± 0.1 at 360 nm to >3 at 185 nm. There was also an increase in the threo/erythro
ratio, k /k , of the 1,2-addition product from 0.7 at 360 nm to 1.8 at 254 nm. Following
t
e
consideration of alternative mechanisms, they proposed a mechanism involving syn addition of
1
CH
3
OD to “planar, unrelaxed (π,π*) states” of t- and c-St resulting in threo and erythro ether,
respectively, and competing 1,2-addition and carbene formation involving a vibrationally hot
twisted diradical excited state.16
Selected results for irradiations in CD OD from Table S1 are shown in Table 1. They allow
3
evaluation of the λexc effect on the reactivity of the two isomers separately. Although not in
quantitative agreement with the earlier work,16,17 they confirm the reported trend in the carbene
Table 1. The Excitation Energy Effect on Product Quantum Yields
Total
direct
carbene
Starting from
λ
exc/nm

e

e

e
c-St
t-St
c-St
t-St
313
313
254
254
0.0013
0.0016
0.0012
0.0030
0.00078
0.0011
0.00058
0.0013
0.00053
0.00049
0.00057
0.0017
ACS Paragon Plus Environment
2
0

Page 21 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
insertion/direct addition ratio on changing λexc. Laarhoven and coworkers reported a change in
c
d
the 
other hand, reported 
progress to a cis-rich quasi-photostationary state considered. In fact the latter value is in good
e
/
e
ratio from 0.9 ± 0.3 at 300 nm to 2.0 ± 0.5 at 254 nm. Roberts and Pincock, on the
c
d
18
e
/ = 0.64 at 300 nm. In neither of these studies was the extent of
e
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
c
d
agreement with 
e
/ = 0.67 at 313 nm that we observed for c-St, Table 1. Our isomer specific
e
GC/MS ratios indicate that at 313 nm the direct addition is favored more starting from t-St as
1
1
would be expected if formation of thermally equilibrated p* from t* underwent less carbene
rearrangement. Calculation of partial quantum yields at 313 nm was carried out using the
observed GC/MS ratios for the two isomers. For t-St we observe a doubling of the ether quantum
yield on increasing the excitation energy with most of the increase, a factor of 3.5 (0.00049 at
3
13 nm and 0.0017 at 254 nm), in the carbene path. Changes for c-St are much more modest.
The overall ether quantum yields at the two wavelengths are the same within experimental error
but there is a shift in the partial ether quantum yields in favor of the carbene path at 254 nm. We
conclude that following c-St excitation at either wavelength, ether formation from both paths
1
involves only the zwitterionic p* intermediate which is not thermally relaxed and whose initial
energy increases with increasing excitation energy. At 313 nm in CD
3
OD, (te/tc)/(ce/ct) =
1
0
.86, is very close to the expected α/(1-α) = 0.82 ratio and it follows that p* also accounts for
most if not all of the ether formed on excitation of t-St at 313 nm. Excess vibrational energy in
1
1
p*favors the carbene pathway and it is likely that the product distribution obtained from t*
1
formed on 313 nm excitation reflects the behavior of vibrationally relaxed p*. It is the
enhanced ether quantum yields for t-St excitation in CD
3
OD at 254 nm and Laarhoven’s
1
observation of more threo ether in CH
3
OD that require the involvement of t*´, a higher t-St
excited state.
ACS Paragon Plus Environment
2
1

The Journal of Physical Chemistry
Page 22 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Solvent Deuterium Isotope Effects. Protonation of diphenylcarbene by alcohols was initially
70
proposed by Kirmse based on indirect measurements and later confirmed spectroscopically by
the observation of diphenylcarbenium ions.71-73 Diphenylcarbene has a triplet ground state in
aprotic solvents, but in protic solvents strong hydrogen bonding selectively stabilizes the singlet
state such that singlet diphenylcarbene is the ground state in the presence of methanol.74,75
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The deuterium isotope effect on the carbene insertion into OH vs OD, kecH/kec = 1.4±0.2,
D
1
6
determined by Laarhoven and coworkers using a 1:1 CH
3
OH/ CH
3
OD, is in good agreement
with previous measurements for singlet carbene insertion into methanol.76-80 The kecH/kec = 1.7
D
value determined by Kohler and coworkers by direct transient observation of the 1,1-
8
0
diphenylmethylcation formed by protonation of singlet diphenylcarbene by methanol is
probably the most reliable. It will be used here because, based on the comparison of transient
kinetics in the pure CH
relevant to our work.
3
OH and CH OD solvents instead of in their 1:1 mixture, it is more
3
Laarhoven and coworkers generated the 1-benzyl-1-phenylcarbene by exciting 1,1-
diphenylethene in methanol. Formation of 1-methoxy-1,1-diphenylethane, the Markovnikov
product, competes with 1,2-phenyl shift, resulting in the same carbene that is obtained from the
stilbenes. The carbene gives the expected 1-methoxy-1,2-diphenylethane ether and phenanthrene
in a 2:1 ratio. Formation of phenanthrene reveals that carbene insertion into methanol occurs in
competition with carbene rearrangement to ground state stilbenes that do not accumulate in more
1
6
than trace amounts under the reaction conditions. It follows that the carbene reacts with
methanol twice as fast as it rearranges, i.e., kec/kcr = 2, where kec and kcr are the rate constant of
the carbene reactions in eqs 13 and 14, respectively. It should be pointed out that Tomioka and
coworkers generated the same carbene in CH
OD by photolysis of 1,2-diphenyldiazoethane.81
3
ACS Paragon Plus Environment
2
2

Page 23 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Ether and the stilbenes formed in 57 and 43% yield, respectively and there was 11% D
incorporation in the stilbenes. No D incorporation in the stilbene products was detected by
Laarhoven and coworkers.16
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Relevant results from Table S1 are shown in Table 2. They show that changing the solvent
from CH
3
OH to CD
3
OD diminishes the overall ether quantum yields obtained on 313 nm
excitation by factors of 1.4 and 1.8 starting from c- and t-St, respectively. Use of kecH/kec = 1.7
D
and kecH/kcr = 2 allows calculation of the D isotope effect on the lifetime of the carbene,
H
D

carb /carb = (kecD + kcr)/( kecH + kcr) = 0.72
5
and on the ratios of ether formation via the carbene
Table 2. The Solvent D-Isotope Effect on Ether Quantum Yields
Total
direct
carbene
Starting from
c-St
Solvent

e

e

e
CH
CH
CD
CD
3
3
3
3
OH
OH
OD
OD
0.0021
0.0029
0.0013
0.0016
0.0011
0.0018
4
7
0.0009
0.0010
0.0053
3
t-St
3
c-St
t-St
0.00078
0.0011
0.00049
and direct pathways, (te /ted)
c
/(te /ted) = (kecH carbH/ kecD carbD) = 1.23. The slower carbene
 
c
H
D
reaction with CD
3
OD is partially compensated for by the longer carbene lifetime in that solvent.
OD for t-St gives (te /ted)
c
Applying this ratio on our ratio of ether quantum yields in CD
3
H
=
0
.54
9
in CH
3
OH. The total ether quantum yield for t-St in CH
3
OH can now be decomposed into
c
d
the contributions of the two ether pathways: te = 0.010
3
and te = 0.0018
7
. These values are
entered in Tables 1S and 2. It follows that the D isotope effect on the quantum yields for 1,2-
d
d
addition of methanol starting from t-St is te-H /te-D = 1.7. At 313 nm the ratio of the quantum
ACS Paragon Plus Environment
2
3

The Journal of Physical Chemistry
Page 24 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
yields for direct CH
3
OH and CD
3
OD addition to p* is given by eq 24 because we need only
consider the first term in eq 22.
d
te-H
d
H
H
H
H
p
H


k  k 
k
k
k
tp
f
pe
pe

=
(24)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
D
D
f
D
pe
D
p
D
pe

k

te-D
tp
1
1
1
Neglecting solvent effects on the rate of formation of p* and on the lifetimes of t* and p*, the
quantum yield ratio gives the ratio of the rate constants for direct CH
3
OH and CD OD addition
3
to p*: kpeH/kpe = 1.7. This D isotope effect is identical to that measured by Kohler and
1
D
coworkers for protonation of diphenylcarbene in CH
3
OH vs CH
3
OD, 78 strongly suggesting that
1p* in the same two solvents also undergoes protonation. The protonation pathway had been
disfavored earlier because there was no change in the rate of product formation when the
-3
SO
.16
irradiation of t-St in CH
3
OH was carried out in the presence of 8 x 10 M H
2
4
1
The lifetime of p*, 
p
, in CH
3
OH is not known. The pseudo-unimolecular rate constant for
1
9 -1
ether formation by direct CH
3
OH addition to p*, kpe-H = 8.3 x 10 s , can be estimated roughly
1
7,43
by using 
p
= 0.23 ps. the lifetime of p* in hexane. This is an upper limit because the lifetime
1
of p* in CH
3
CN, a more polar solvent, is about twice as long as in pentane (see SI in ref 43).
Assuming that the quantum yield of ether formation by this pathway is determined by the rate
1
1
constant of protonation of p*, the rate constant of protonation of p* in CH OH is more than a
3
80
factor of 10 slower than the rate constant of diphenylcarbene protonation. Judging from the
1
suppression of the carbene pathway to ether and neglecting solvent effects on the lifetime of p*,
1
OH than in CH
OH.18
protonation of p* is 3.3 times faster in CF
3
CH
2
3
ACS Paragon Plus Environment
2
4

Page 25 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Intramolecular Deuterium Isotope Effect. Replacement of the vinyl Hs of t-St with D affects
the ether quantum yields by changing the lifetimes of the excited states and by diminishing the
1
stilbene/carbene rearrangement. Neglecting the involvement of t*´, and using d0 and d2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
subscripts to distinguish parameters of undeuterated from those of dideuterated stilbene
(

ted)d0/(ted)d2, the ratio of the ether quantum yields by direct addition of CH
OH at 313 nm in eq
3
25 is based on the first term in eq 22. Relevant ether formation quantum yields from Table S1 are
d
ted0
d
ktpd0 kped0 pd0


fd0

(25)
ktpd2 kped2 pd2
ted2
fd2
shown in Table 3. The fluorescence lifetimes of trans-α,α´-dideuteriostilbene, t-Std2, are known
Table 3. Intramolecular D Isotope Effects on Ether Formation
Total
direct
carbene
Starting from

e

e

e
t-Std0
t-Std2
0.0029
0.0058
0.00187
0.00518
0.00103
0.00062
3
2
to be consistently higher than those of the parent, t-Std0, under a variety of conditions. A
lifetime increase from 82.9 to 119.2 ps in hexane at 23 °C can be compared with a corresponding
32
change from 42.6 ps for t-Std0 to 61.1 ps for t-Std2 in methanol at 24 °C. The increase is a factor
of 1.43 in both solvents. Applying this ratio on fd0 = 0.026 gives fd2 = 0.037 for t-Std2. Although
the change in fluorescence quantum yield is significant, their absolute values are small and have
1
a negligible effect on the efficiency of reaching p*, (ktpd0

fd0)/( ktpd2

fd2) = (1-fd0)/(1-fd2) =
1
1
.011. The predicted change is a 1.1% decrease in the population of p* on dideuteration. This
small change would not be detected within the experimental error of our ether quantum yield
1
measurements. Protonation of p* involves no hybridization change at the reaction site and,
ACS Paragon Plus Environment
2
5

The Journal of Physical Chemistry
Page 26 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
consequently, the secondary D isotope effect should be negligible, (kped0/kped2) = 1. It follows
that, with the 1.1% modification, the ratio of the quantum yields for direct ether formation,
1

ted2

ted0 = 2.7
7
, gives the ratio of the lifetimes of p*, pd2/pd0 = 2.7
4
. This is in reasonable
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
agreement with the factor of 2 pd2/pd0 ratio estimated in hexane from the quantum yields of t*
1
1
1
7
fluorescence due to the adiabatic c*  p*  t* reaction channel.
The ratio of the quantum yields of ether formed by the carbene pathway at 313 nm in eq 26 can
be based on the first term in eq 23. With the exceptions of kpc, kec and carb, the effect of D
c
ktpd0 kpcd0 pd0 kecd0 carbd0


fd0
ted0
c
ted2

(26)
ktpd2 kpcd2 pd2 kecd2 carbd2
fd2
substitution for the vinyl Hs on the parameters in eq 26 is known. The kecdo/kecd2 ratio of the rate
constants for the ether forming reactions, for the undeuterated and dideuterated stilbenes is not
expected to deviate from unity because it is only subject to small secondary D isotope effects and
1
the D in the carbene is not at the reaction site. The rate constants for the rearrangements of p* to
the carbene, kpc, and of the carbene back to the stilbene, k
r
, on the other hand are subject to
primary D isotope effects and should decrease significantly on deuteration. Assuming krdo/krd2
=
6
.3, gives carbd0/carbd2 = 0.72. Use of our quantum yield ratio for the carbene pathway,
(

tec)d0/(tec)d2 = 1.66 from Table 3 then gives kpcdo/kpcd2 = 6.3. In view of the fact that the two
rearrangement reactions involve the same D shifting in opposite directions, it would not be
surprising if they experienced similar D isotope effects. However, because the lifetime of the
carbene depends primarily on its reaction with methanol, it is rather insensitive to the magnitude
of the D isotope effect on the rate constant of its rearrangement. Assuming the extreme case of
k
rd2/krd0 = 0 gives carbd0/carbd2 = 0.67 and results in a minor increase from 6.3 to 6.8 in kpcdo/kpcd2.
ACS Paragon Plus Environment
2
6

Page 27 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
CONCLUSIONS
By studying the addition of CH
3
OH separately to c- and t-St we showed that ether formation and
1
cis-trans photoisomerization proceed via torsional relaxation to the common p* twisted
intermediate, the phantom singlet state. This short-lived intermediate is common in that it decays
predominately to the cis and trans ground state isomers via a zwitterionic conical intersection as
predicted by theory57,58 in a 45/55 decay ratio favoring c-St. Inefficient ether formation occurs
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
via protonation of p* and via competing hydride shift to a singlet carbene that also gives ether
via protonation. The carbene pathway is favored by higher excitation energy, especially starting
from t-St, indicating that at least some of the hydride shift occurs in competition with vibronic
1
1
relaxation of hot p*. Protonation of p* in pure methanol is about ten times slower than the
80
extremely fast protonation of singlet diphenylcarbene, whose reactivity, we expect, is similar to
that of the singlet benzylphenylcarbene that is involved in our study. The conclusion of Kohler
and coworkers that, in pure methanol, solvent reorganization controls the rate of protonation of
80
74,75
singlet diphenylcarbene, has received strong recent experimental and theoretical support.
Not only does the formation of the carbene-methanol hydrogen bonded complex lead to a singlet
74
spin state, but its fate depends on the hydrogen-bonding disposition of nearby methanol
7
5
molecules. Earlier workers had also proposed differences in reactivity between singlet carbenes
and methanol monomers and oligomers.79,82 We expect similar solvent reorganization
considerations to apply to the very fast torsional relaxation and protonation reactions that are
described here.
There are strong indications that adduct formation via trapping of the zwitterionic twisted
intermediates in alkene photoisomerization is a general reaction. We have already mentioned the
1
9
addition of ethanol to the DPBs that inspired this work. Noteworthy is also the photohydration
ACS Paragon Plus Environment
2
7

The Journal of Physical Chemistry
Page 28 of 40
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
83
of 3-hydroxystilbene in 1:1 CH
3
CN/H
2
O. Although a mechanism involving hydration of the
trans isomer was proposed, the time evolution of the photoproducts establishes similar reactivity
for the two isomers and strongly implicates the common twisted intermediate. Other interesting
examples are the photohydrations of trans-stilbene, trans-3-methoxystilbene and trans-3-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
84
methoxyanethole in α-, β- and γ-cyclodextrins. Protonation of the excited state by one of the
three OH substituents on the rim of the cyclodextrins followed by reaction with H O was
2
proposed. The essential role of the cyclodextrin OH was established by the observation that
84
photohydration is completely suppressed when trimethyl-β-cyclodextrin was employed. The
8
5
photoaddition of CH OH to 2-(3-aminostyryl)naphthalene and the photohydrations of 1- and 2-
3
8
6
1
(3-hydroxystyryl)naphthalenes are other potential candidates for p* involvement in adduct
formation. Photochemical addition of the NH bond of secondary alkylamines across the central
8
7
88
bond of stilbenes also occurs both inter- and intra-molecularly. Protonation by the neutral
amine is not an attractive initial step and a strong case has been made for its occurrence as a
second step following formation of the stilbene anion radical by electron transfer. However, the
88
1
fact that t- and c-St undergo intramolecular photoaddition at similar rates does suggest that p*
can serve as a stilbene partner in these reactions also.
SUPPORTING INFORMATION
Table S1 listing all quantum yields measured in this work. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*
Email: saltiel@chem.fsu.edu
ACS Paragon Plus Environment
2
8

Page 29 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
The authors declare no competing financial interests.
ACKNOWLEDGMENT
This research was supported in part by the National Science Foundation, most recently by Grant
No. CHE-1361962. It was also supported by the Florida State University. We thank Mr. David
Boose for technical assistance in initial experiments.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
REFERENCES
(
1) 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.
(2) Saltiel, J.; Sun, Y.-P. Cis-Trans Isomerization of C=C Double Bonds. In Photochromism,
Molecules and Systems; Dürr, H. and Bouas-Laurent, H., Eds.; Elsevier: Amsterdam, 1990; pp 64-
1
64.
(3) Waldeck, D. H. Photoisomerization Dynamics of Stilbenes. Chem. Rev.1991, 91, 415-436.
(
4) Saltiel, J.; Sun, Y.-P. The Intrinsic Potential Energy Barrier for Twisting in the trans-Stilbene
S
1
State in Hydrocarbon Solvents. J. Phys. Chem. 1989, 93, 6246-6250.
(
5) Saltiel, J.; Waller, A. S.; Sears, Jr., D. F.; Garrett, C. Z. Fluorescence Quantum Yields of
trans-Stilbene-d
0
and -d in n-Hexane and n-Tetradecane. Medium and Deuterium Isotope Effects
2
on Decay Processes. J. Phys. Chem. 1993, 97, 2516-2522.
(
6) Saltiel, J.; Waller, A. S.; Sears, Jr., D. F.; Hoburg, E. A.; Zeglinski, D. M.; Waldeck, D. H.
Fluorescence Quantum Yields and Lifetimes of Substituted Stilbenes in n-Alkanes. A
ACS Paragon Plus Environment
2
9