Steric Demand and Rate-determining Step for Photoenolization of Di-ortho-substituted Acetophenone Derivatives

Article abstract of DOI:10.1111/php.12996

Laser flash photolysis of ketone 1 in argon-saturated methanol yields triplet biradical 1BR (τ = 63 ns) that intersystem crosses to form photoenols Z-1P (λ_max = 350 nm, τ ~ 10 μs) and E-1P (λ_max = 350 nm, τ > 6 ms). The activation barrier for Z-1P re-forming ketone 1 through a 1,5-H shift was determined as 7.7 ± 0.3 kcal mol^?1. In contrast, for ketone 2, which has a less sterically hindered carbonyl moiety, laser flash photolysis in argon-saturated methanol revealed the formation of biradical 2BR (λ_max = 330 nm, τ ~ 303 ns) that intersystem crosses to form photoenol E-2P (λ_max = 350 nm, τ > 42 μs), but photoenol Z-2P was not detected. However, in more viscous basic H-bond acceptor (BHA) solvent, such as hexamethylphosphoramide, triplet 2BR intersystem crosses to form both Z-2P (λ_max = 370 nm, τ ~ 1.5 μs) and E-2P. Thus, laser flash photolysis of ketone 2 in methanol reveals that intersystem crossing from 2BR to form Z-2P is slower than the 1,5-H shift of Z-2P, whereas in viscous BHA solvents, the 1,5-H shift becomes slower than the intersystem crossing from 2BR to Z-2P. Density functional theory and coupled cluster calculations were performed to support the reaction mechanisms for photoenolization of ketones 1 and 2.

Full text of DOI:10.1111/php.12996

Article type : Special Issue Research Article
Steric Demand and Rate-Determining Step for
Photoenolization of Di-ortho-Substituted Acetophenone
†
Derivatives
1
2
1
1
Anushree Das , Suma S. Thomas , August A. Garofoli , Kevin A. Chavez ,
1
2
1
Jeanette A. Krause , Cornelia Bohne , Anna D. Gudmundsdottir *
1Department of Chemistry, University of Cincinnati, Cincinnati, OH 45221
2Department of Chemistry and Centre for Advanced Materials and Related Technologies
CAMTEC), University of Victoria, PO Box 1700 STN CSC, Victoria, BC, Canada, V8W 2Y2
(
*
Corresponding author e-mail: anna.gudmundsdottir@uc.edu (Anna D.
Gudmundsdottir)
†This article is part of a Special Issue celebrating Photochemistry and Photobiology's 55th
Anniversary.
ABSTRACT
Laser flash photolysis of ketone 1 in argon-saturated methanol yields triplet biradical 1BR ( =
6
3 ns) that intersystem crosses to form photoenols Z-1P (max = 350 nm,  ~ 10 s) and E-1P
(
max = 350 nm,  > 6 ms). The activation barrier for Z-1P re-forming ketone 1 through a 1,5-H
-1
shift was determined as 7.7 ± 0.3 kcal mol . In contrast, for ketone 2, which has a less sterically
hindered carbonyl moiety, laser flash photolysis in argon-saturated methanol revealed the
formation of biradical 2BR (max = 330 nm,  ~ 303 ns) that intersystem crosses to form
photoenol E-2P (max = 350 nm,  > 42 s), but photoenol Z-2P was not detected. However, in
more viscous basic H-bond acceptor (BHA) solvents, such as hexamethylphosphoramide, triplet
2BR intersystem crosses to form both Z-2P (max = 370 nm,  ~ 1.5s) and E-2P. Thus, laser
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/php.12996
This article is protected by copyright. All rights reserved.

flash photolysis of 2 in methanol reveals that that intersystem crossing from 2BR to form Z-2P is
slower than the 1,5-H shift of Z-2P, whereas in viscous BHA solvents the 1,5-H shift becomes slower
than the intersystem crossing from 2BR to Z-2P. Density functional theory and coupled cluster
calculations were performed to support the reaction mechanisms for photoenolization of
ketones 1 and 2.
INTRODUCTION
Photoenolization is a light-initiated keto–enol tautomerization process, and the resulting
photoenols are high-energy ground-state intermediates that are highly reactive (1, 2). Efficient
photoenol reactivity has been captured in applications such as synthesizing complex natural
products and pharmaceutical drugs, and initiating release from photoremovable protecting
groups (3-14). The mechanism of photoenolization, which has been determined using transient
spectroscopy, can be outlined as follows: Upon excitation ortho-substituted arylketones form
the singlet excited state of the ketone chromophore, which undergoes efficient intersystem
crossing to form its triplet configuration (Scheme 1) (15-19). The triplet ketone decays by
intramolecular H-atom abstraction to form a triplet 1,4-biradical. Intersystem crossing of the
biradical results in the formation of both Z- and E-photoenols. Generally, E-photoenols are long-
lived intermediates that can re-form the starting material via a relatively slow solvent-mediated
proton transfer, and they are long lived enough to undergo electrocyclic ring closure or be
trapped in bimolecular reactions. In contrast, Z-photoenols are short-lived intermediates that
regenerate the starting material through efficient intramolecular 1,5-H shifts. It should be noted
that ortho-substituted arylketones that do not undergo efficient intersystem crossing can
undergo photoenolization from their singlet ketone to form Z-photoenols. Because the lifetimes
of Z-photoenols are generally on the order of a few nanoseconds, they are too short lived to
undergo electrocyclic ring closure or be trapped in bimolecular reactions. Thus, the formation of
Z-photoenols causes reactions that capture E-photoenols in synthetical applications to be less
effective, but does not affect reaction selectivity. Thus, a better understanding of the
reketonization of Z-photoenols has the potential to lead to more efficient use of E-photoenols in
various applications.
In this study, we compared the photoenolization process for ketones 1 and 2 (Scheme 2) to
determine how photoenolization is affected by substituents. Ketone 1 has two bulky isopropyl
groups, whereas ketone 2 has two ortho methyl groups. We investigated how the steric demand
of the ortho substituents affects the 1,5-H shifts in Z-photoenols Z-1P and Z-2P to re-form the
corresponding starting materials, as well as the rate of intersystem crossing of triplet 1,4-
biradicals 1BR and 2BR to form photoenols Z-1P and Z-2P, respectively.
This article is protected by copyright. All rights reserved.

MATERIALS AND METHODS
Laser flash photolysis. Transient UV-Vis spectra and corresponding kinetic traces were acquired
using an excimer laser (308 nm, 17 ns) (20). The stock solutions of ketones 1 and 2 were
prepared in spectroscopic-grade solvents, such as methanol, acetonitrile,
hexamethylphosphoramide (HMPA) and dimethyl sulfoxide (DMSO), so that the absorbance was
between 0.2 and 0.8 at 308 nm. Typically, ~1.5 mL of stock solution was placed in a quartz
cuvette (10 mm × 10 mm cross-section and 48 mm length). As required, the solution was then
purged with argon or oxygen for 5 min. The reaction rates were obtained by fitting an average
of 3–8 kinetic traces.
Arrhenius plots. The laser flash photolysis system (YAG laser, 266 nm, 10 ns) has been
described in detail elsewhere (21). Methanol solutions of ketones 1 and 2 were prepared in
spectroscopic-grade methanol so that the absorbance was ~0.5 at 266 nm. All solutions were
purged with N gas before the decays were recorded. The decay rate constants were collected
2
between 313.15 and 253.15 K for 1 and 313.15 to 233.15 K for 2. For 1 in methanol, the laser
flash photolysis experiment presented some uncertainty as to the baseline. Decays measured
with and without the collection of a baseline yielded similar lifetimes. The data obtained from
both these experiments were fit simultaneously to yield the parameters reported for the
analysis of Arrhenius plots. The kinetic traces obtained from the 266 nm and 308 nm laser
irradiation gave similar results.
Calculations. All geometries were optimized using density functional theory (DFT) calculations
in Gaussian09 at the B3LYP level of theory with the 6-31+G(d) basis set (22-24), M062X (25),
and coupled cluster (26, 27). Time-dependent density functional theory (TD-DFT) calculations
were performed to locate the energies of the excited singlet and triplet states of the optimized
ground states (28-32). Analysis of the second derivative of the energy with respect to the
internal coordinates confirmed that all transition states had one imaginary vibrational
frequency. The intrinsic reaction coordinate (IRC) was calculated to verify that the located
transition states corresponded to the respective reactant and products (33, 34).
Phosphorescence. Phosphorescence spectra of ketones 1 and 2 were obtained in frozen ethanol
matrices at 77 K using a spectrophotometer that has been described in the literature (35).
Synthesis of starting materials. Ketone 2 was purchased from Sigma-Aldrich. Ketone 1 was
synthesized as described below.
Ketone 1 was obtained using a procedure similar to that reported by Murphy and Prager (36).
1
,3,5-Triisopropylbenzene (1.0 mg, 4.8 mmol) was mixed with acetic anhydride (50 mg, 4.9
mmol) in 10 mL of dichloromethane. After cooling the mixture to 0 °C, AlCl (1.3 g, 10 mmol)
3
was added in four portions. The mixture was gradually warmed to room temperature while
stirring, and then refluxed for 4 h. Three cubes of ice (~30 mL) were added to the cooled
reaction mixture. Once the ice melted, 20 drops of 1 M HCl was added, and the resulting mixture
was extracted three times with dichloromethane (20 mL). The organic layer was washed with
brine and dried over anhydrous MgSO , followed by solvent removal under vacuum. The residue
4
was purified by column chromatography using a silica column eluted with a mixture of ethyl
acetate and n-hexane (20:80, v/v), and ketone 1 was obtained after crystallization from diethyl
This article is protected by copyright. All rights reserved.

ether (300 mg, 1.2 mmol, 25% yield). The obtained NMR spectrum was consistent with the
reported one in the literature (37).
1H NMR (CDCl
.25–1.23 (m, 18H) ppm.
3
, 400 MHz): δ 7.0 (s, 2H), 2.94–2.83 (m, 1H), 2.77–2.67 (m, 2H), 2.49 (s, 3H),
1
Product studies. A solution of ketone 2 in isopropanol (15 mL, 0.2 mmol, 0.01 M) in a Pyrex
round-bottom flask was purged with argon for 20 min, and a rubber cap was fitted to the flask
and sealed with parafilm. The resulting solution was irradiated with stirring for ~38 h using a
medium-pressure mercury arc lamp. Cyclobutanol 5 was formed in 92% yield.
1H NMR (400 MHz, CDCl
): 7.03 (s, 2H), 3.31–3.27 (d, J = 16 Hz, 1H), 3.16–3.12 (d, J = 16 Hz,
3
13
1
1
H), 2.29 (s, 3H), 1.69 (s, 3H), 1.30 (s, 9H) ppm. C NMR (75 MHz, CDCl
31.5, 125.5, 118.0, 78.2, 48.1, 35.1, 31.7, 24.8, 16.8 ppm.
3
): 152.8, 145.9, 140.5,
RESULTS
Laser flash photolysis in solution
We performed laser flash photolysis (excimer laser, λ = 308 nm, 17 ns) (20) experiments to
detect and measure the lifetimes of the 1,4-biradicals and Z-photoenols formed by
intramolecular H-atom abstraction in ketones 1 and 2. Laser flash photolysis of ketone 1 in
argon-saturated methanol showed broad transient absorption between 310 and 390 nm (Figure
1
A). This absorption was not quenched in oxygen-saturated methanol and thus, it is assigned to
photoenols Z-1P and E-1P. This assignment is further supported by comparison to the TD-DFT
calculated spectra (Figure 1b), which show major electronic transitions at 382 nm (f = 0.1453)
and 371 nm (f = 0.1665) for Z-1P and E-1P, respectively. Analysis of the kinetics at 350 nm
7
-1
revealed that the transient is formed with a rate constant of 1.58 × 10 s (τ = 63 ns, Figure 2a)
on a shorter timescale, whereas on a longer timescale the decay can be fitted as a mono-
5
-1
exponential function to yield a rate constant of 1 × 10 s (τ ~ 10 µs, Figure 2b). The decay rate
constant was not affected by oxygen, which further supports the assignment of the transient
absorption to photoenol Z-1P. On the shorter timescale, the absorption did not decay fully. The
small amount of residual absorption decayed on a millisecond timescale and hence, it is
assigned to photoenol E-1P, which has a lifetime of >6 ms. Furthermore, the rate of forming the
7
-1
transient at 350 nm increased in oxygen-saturated methanol to 1.91 × 10 s (τ = 52 ns), which
indicates that the precursor of Z-1P decays faster in oxygen-saturated methanol. Thus, the
precursor of Z-1P and E-1P, triplet 1,4-biradical 1BR, has a lifetime of ~63 ns in argon-
saturated methanol and ~52 ns in oxygen-saturated methanol.
Laser flash photolysis of ketone 2 in argon-saturated methanol produced a transient spectrum
with broad absorption between ~330 and 420 nm (Figure 3a). The two major bands, located at
3
30 and 350 nm, each exhibited a different kinetic profile. TD-DFT calculations showed that the
major electronic transitions for 2BR are at 413 nm (f = 0.0292), 354 nm (f = 0.0312), and 327
nm (f = 0.0353 nm, Figure 3b), which correlate well with the transient absorbance with λmax at
3
30 nm. Kinetic analysis of the decay at 330 nm showed that biradical 2BR decays with a rate
6
-1
constant of 3.4 × 10 s (τ = 303 ns) in argon-saturated methanol. In comparison, the transient
This article is protected by copyright. All rights reserved.

absorption with λmax at 390 nm can be attributed to either photoenol Z-2P or E-2P, or both, as
TD-DFT calculations revealed that their major electronic transitions are at 403 nm (f = 0.1189)
and 385 nm (f = 0.1359, Figure 3b) for Z-2P and E-2P, respectively. Kinetic analysis of the
decays at 370 and 390 nm on shorter timescales showed a mono-exponential decay attributed
to 2BR. In contrast, on longer timescales, the decay rate constant was determined to be slower
4
-1
than 2.4 × 10 s (τ >42 s), which is assigned to E-2P. No transient absorption was observed
that can be assigned to Z-2P in methanol. Hence, we theorize that biradical 2BR must be longer
lived than Z-2P; which explains why the only photoenol observed is E-2P. This notion that the
shorter decay rate constant is due to 2BR and not Z-2P is further supported by the analysis of
the decay in air- and oxygen-saturated methanol. The decay rate constant of 2BR increased to
7
-1
7
-1
1
.03 × 10 s (τ ~ 97 ns) and 2.5 × 10 s (τ ~ 40 ns) in air- and oxygen-saturated methanol,
respectively (Figure 3c).
On the contrary, in viscous BHA- solvents, such as DMSO and HMPA, laser flash photolysis of
ketone 2 allowed direct observation of both photoenols Z-2P and E-2P. Laser flash photolysis of
ketone 2 in DMSO and HMPA resulted in transient spectra with λmax at 390 nm (Figure 4a).
6
5
Kinetic analysis at shorter timescales yielded decay rate constants of 3.8 × 10 and 7.8 × 10
~300 ns and ~1.5 µs, Figure 4b) for Z-2P in DMSO and HMPA, respectively. On longer
timescales, E-2P was also detected, decaying on the millisecond timescale. The rate constants
(
7
-1
for forming Z-2P and E-2P were 1.69 × 10 s ( = 52 ns) in argon-saturated HMPA and 2.17 ×
7
-1
1
0 s ( = 46 ns) in oxygen-saturated HMPA, which are attributed to the decay rate of 2BR.
Thus, viscous BHA solvents have the ability to reduce the rate of the 1,5-H shift in Z-2P, making
this process slower than intersystem crossing, and thus, Z-2P can be detected directly. For laser
flash photolysis data summary see Table S3.
The formation of E-2P was further supported by quenching studies in the presence of NaN
which assists its reketonization to ketone 2. Kinetic analysis revealed that the lifetime of E-2P
gradually decreases with the addition of NaN (Figure 5), confirming the formation of E-2P and
3
,(38)
3
its reketonization to 2 can be increased by base in the solvent.
Arrhenius plot
Using the Arrhenius equation, ln(k
measured as a function of temperature, were used to determine the activation barriers for the
,5-H shift in Z-1P and for 2BR to twist and intersystem cross to form Z-2P. The decay rate
constant (k ) of Z-1P in nitrogen-saturated methanol was monitored at 350 nm. From the slope
d
) = -E /RT + ln(A), the decay rate constants of Z-1P and 2BR,
a
1
d
-1
of the Arrhenius plot (Figure 6), we obtained a transition state barrier of 7.7. ± 0.3 kcal mol .
Similarly, the decay rate constant of 2BR in nitrogen-saturated methanol was monitored at 330
nm as a function of temperature. From the slope of the Arrhenius plot (Figure 6), we obtained a
-1
transition state barrier of 4.7 ± 0.1 kcal mol , which reflects the barrier for 2BR to achieve the
correct conformation to intersystem cross to photoenols E-2P and Z-2P.
These results demonstrate that in nitrogen-saturated methanol, the transition state barrier for
the 1,5-H shift in Z-1P is significantly larger than the barrier for intersystem crossing in 1BR,
whereas the 1,5-H shift in Z-2P must has a smaller transition state barrier than the barrier for
intersystem crossing of 2BR.
This article is protected by copyright. All rights reserved.

Product studies
Photolysis of ketones 1 yields the corresponding cyclobutanol 4, as reported in previous
product studies (39), owing to electrocyclic ring closure of E-1P (Scheme 3). Similarly,
photolysis of ketone 2 in argon-saturated isopropanol forms cyclobutanol 5 via electrocyclic
ring closure on E-2P (Scheme 3).
Calculations
To compare and explain the relative stabilities of the Z-enols obtained from ketones 1 and 2, we
performed calculations in Gaussian09 at the B3LYP, M062X, and CBS-QB3 levels of theory using
the 6-31+G(d) basis set (22-27). For B3LYP calculations, DFT was used to optimize the ground
states (S ) of ketones 1 and 2 and their respective triplet excited states, biradicals, and
0
photoenols.
The optimized geometry of the ground state (S ) of 1 reveals that the carbonyl group is
0
orthogonal to the benzene ring plane owing to steric hindrance created by the two bulky ortho
isopropyl groups (Figure 7). This configuration leads to a torsional angle of 88° between the
C=O and the phenyl ring, revealing that the ketone and phenyl are not conjugated with each
other. The optimized geometry is in agreement with the X-ray structure of ketone 1,(40) which
has a torsional angle of 80° between the C=O and the phenyl ring, highlighting the out-of-plane
geometry of the carbonyl group. TD-DFT calculations on the optimized S of 1 located its first
0
-1
and second triplet excited states (T1K and T2K of 1) at 83 and 84 kcal mol above S
0
of 1. These
values are considerably higher than the reported energy of the triplet ketone of acetophenone
-
1
(
74 kcal mol ) (41), but more similar to the energies of the triplet ketones of aliphatic ketones
-1
such as acetone (79 kcal mol ) (42), as the C=O chromophore is not conjugated to the phenyl
ring.
-1
In comparison, the optimized structure of T1K of 1 is located 72 kcal mol above its S , which is
0
considerably lower than the energy obtained from the TD-DFT calculations. The torsional angle
between the C=O and the phenyl ring is 66° in the optimized structure of T1K of 1, demonstrating
that the C=O and the phenyl group are also not fully conjugated in the T1K of 1. The C=O bond
length in the optimized structure of T1K of 1 is 1.33 Å and the carbon–carbon bonds in the
phenyl ring have lengths between 1.39 and 1.43 Å, which suggest a (n,π*) configuration. This
configuration is confirmed by spin-density calculations (Scheme 4), which show that the
unpaired spin is mainly located on the carbonyl carbon and oxygen atoms.
-1
The optimized structure of 1BR is located 62 kcal mol above the S of 1, and spin-density
0
calculations show that the unpaired electron density is mainly localized on the ortho-carbon and
the ketyl carbon atoms, indicating that the radical has a localized 1,4-biradical character. The
OH–C–C–CH
2
and (CH
3
)
2
C–C–C–C(OH)(CH
3
) torsional angles in 1BR are 42° and 10°,
and T of 1.
respectively, and thus 1BR is less sterically congested than S
0
K
-1
The optimized structures of Z-1P and E-1P are located 38 and 40 kcal mol , respectively, above
the S of 1. Z-1P and E-1P are twisted (CH C=C–C=C(OH)(CH ) torsional angles of 56° and 55°,
0
)
3 2
3
respectively) owing to the steric demand of the isopropyl group. Photoenols E-1P and Z-1P are
more twisted than their precursor 1BR.
This article is protected by copyright. All rights reserved.

Figure 8 displays the calculated stationary points on the energy surface of ketone 1 to form
photoenols E-1P and Z-1P on the triplet surface. The calculated transition state for hydrogen
-1
atom abstraction to form 1BR is located less than 1 kcal mol above T1K of 1, which is consistent
with the spectroscopic observation that intramolecular H-atom abstraction is efficient in 1.
In the optimized structure of ketone 2, the carbonyl group is not fully conjugated with the
phenyl ring owing to presence of two ortho methyl substituents (Figure 7). The dihedral angle
between the C=O and the phenyl ring is only 51° in S of 2, confirming that the carbonyl moiety
0
is not fully conjugated with the phenyl ring. However, because the steric demand of the two
ortho methyl groups in ketone 2 is less than that of the isopropyl groups in ketone 1, the extent
of conjugation between the carbonyl group and the phenyl ring is greater in ketone 2. The X-ray
structure of ketone 2 also demonstrates that the carbonyl group is not conjugated with the
phenyl ring, as the torsional angle between these groups is 80(43). However, the optimized
minimal energy structure of ketone 2 is different from the structure adopted in the crystal
lattice, presumably because the best packing arrangement of ketone 2 is achieved with a
conformer rather than the minimal energy structure.
-
1
TD-DFT calculations estimate that T1K of 2 is located 75 kcal mol above its S . In comparison,
0
-1
the optimized structure of T1K of 2 is located 70 kcal mol above its ground state (Figure 8).
Further, the optimized structure of T1K of 2 possesses a torsional angle of only 47° between the
C=O and the phenyl ring, and spin-density calculations support that T1K of 2 has a (n,π*)
configuration (Scheme 4).
-1
The optimized structure of 2BR is located 63 kcal mol above S
mainly localized on the ortho and C–OH carbon atoms, with a small contribution on the phenyl
ring. As the OH–C–C–CH torsional angle is 48°, the extent of twisting in 2BR and T of 2 is
similar. The optimized structures of photoenols Z-2P and E-2P are located 40 and 38 kcal mol ,
0
of 2, and the spin density is
2
K
-1
respectively, above S of 2. The dienyl moieties are twisted ~40° from planar in both Z-2P and
0
E-2P.
The transition state for ɣ-H atom abstraction by T1K of 2 to form 1,4-biradical 2BR is located 7
-1
kcal mol above T1K of 2. The calculations show that the steric demand of the ortho isopropyl
group is significantly larger than that of the ortho methyl groups; therefore, 1, T of 1, 1BR, Z-
P, and E-1P are less planar than 2, T of 2, 2BR, Z-2P, and E-2P, respectively.
K
1
K
The calculated transition state barriers for 1,5-H shifts in Z-1P and Z-2P to re-form ketones 1
and 2, respectively, using various levels of theory are shown in Table 1, along with the
measured barriers obtained from the Arrhenius plots. It should be noted that the energies were
calculated with respect to the Z-enol conformers, in which the O–H bond points towards the
ortho methylene. The transition state energies obtained from the B3LYP calculations are not
corrected for the zero-point energy, whereas those obtained from the CBS-QB3 calculations are
corrected for the zero-point energy. The CBS-QB3 and B3LYP calculated transition state barriers
for the 1,5-H shift in Z-1P correspond well to the experimental value, whereas M062X over
estimates it. In comparison, the calculated transition state barrier for the 1,5-H shift in Z-2P to
re-form ketone 2 is smaller than that for Z-1P re-form ketone 1, with B3LYP yielding the
smallest barrier and M062X the largest. The measured barrier for intersystem crossing (4.8 kcal
-1
mol ) is similar, within experimental error, to the calculated barrier for the 1,5-H shift.
In addition, we calculated the transition state barrier for photoenols E-1P and E-2P to undergo
electrocyclic ring closure to form products 4 and 5, respectively. These transition states were
This article is protected by copyright. All rights reserved.

-1
located 19 and 20 kcal mol above photoenols E-1P and E-2P, respectively. Thus, the ortho
substituents do not strongly affect the electrocyclic ring closure reactions of E-1P and E-2P (44,
4
5).
Phosphorescence
The phosphorescence of ketones 1 and 2 was investigated in frozen ethanol glass at 77 K. The
phosphorescence spectra of ketones 1 and 2 show vibrational bands characteristic of triplet
ketones with (n,π*) configurations, indicating that T1K of 1 and 2 have (n,π*) configurations.
However, the vibrational bands are not as well resolved as those for acetophenone derivatives
without ortho substituents (Figure 9). Thus, we assigned the [0,0] emission band to the onset of
the emission for each ketone to calculate the energies of T1K of 1 and 2 (Table 2).
We compared the energies of T1K of 1 and 2 obtained from the phosphorescence spectra to
those obtained from calculations using B3LYP and M062X hybrid functions. The energies from
the TD-DFT calculations are in excellent agreement with the energies obtained from the
phosphorescence spectra. In contrast, the energies obtained from the optimized structures of T
K
of 1 and 2 using both M062X and B3LYP are somewhat lower than the experimentally obtained
values.
DISCUSSION
Ketone 1 undergoes efficient H-atom abstraction on the triplet surface and forms both Z- and E-
photoenols. The measured transition state barrier for Z-1P to re-form ketone 1 through a 1,5-H
-1
shift is 7.7 ± 0.3 kcal mol , and this step is slower than intersystem crossing to form Z-1P.
Because the CH C=C–C=COHCH torsional angle in photoenol Z-1P significantly deviates from
3
3
planarity to accommodate the steric demand of the isopropyl group, Z-1P must untwist to
generate a planar conformer with good orbital alignment for the 1,5-H shift. In contrast, there is
less steric congestion around the carbonyl moiety of ketone 2, and therefore the rate-
determining step for its photoenolization is the intersystem crossing of 2BR to form photoenols
2P. Photoenol Z-2P, which is well aligned for undergoing a 1,5-H shift to re-form ketone 2,
reacts faster than it is formed and therefore, it is not observed. The calculated transition state
barrier for the 1,5-H shift in Z-2P is considerably less than that for Z-1P. However, the measured
barrier for intersystem crossing of 2BR is on the same order as the calculated transition state
barrier for the 1,5-H shift in 2BR, and therefore, it is possible to influence the rate of
photoenolization and re-ketonization, so in viscous BHA solvents the 1,5-H shift of becomes the
slowest step.
Haag et al. previously demonstrated that in the photoenolization of o-methylacetophenone,
intersystem crossing to form photoenols is the slowest step in solvents, such as cyclohexane,
methanol, and isopropanol. whereas in more viscous BHA solvents, such as HMPA and DMSO,
the 1,5-H shift is sufficiently restricted to become slower than intersystem crossing of the triplet
biradical precursor(1). Thus, our results mirror those of Haag and co-workers for the
photoenolization of o-methylacetophenone. In comparison, we have shown that Z-photoenols
from o-methyl substituted valerophenone and butyrophenone ester derivatives re-ketonize
This article is protected by copyright. All rights reserved.

slower than their biradical precursors.(18,19) Similarly, 2,5-dimethylphenacyl ester derivatives
also form Z-photoenols that are longer-lived than their triplet biradical precursors.(46) Thus, it
is not well understood what factors control the intersystem crossing rates of triplet biradicals
involved in photoenolization of o-methylacetophenone derivatives.
Interestingly, it has been proposed that the formation of cyclobutanols from ortho-substituted
arylketones in the solid state do not occur through photoenolization, but rather through direct
intersystem crossing of the biradicals to form cyclobutanols (47, 48). This notion was supported
by laser flash photolysis of nanocrystals, which revealed that E-photoenols are not observed.
Because the energies and the structures of biradicals 1BR and 2BR are similar to the energies
and structures of the calculated transition state barrier for forming cyclobutanols 4 and 5,
respectively, it is possible that in some media, it is favorable to cross over from the triplet
surface of the biradical to that of the cyclobutanol through a conical intersection (Figure 10). As
the energy gap between biradicals 1BR and 2BR and the calculated transition state barriers for
the 1,5-H shifts is larger, it is unlikely that the biradicals can intersystem cross to re-form the
corresponding ketones directly, without first forming Z-1P and Z-2P.
CONCLUSION
In this work, we demonstrated that steric crowding around the carbonyl group in ortho-
substituted aryl ketones affects the photoenolization process. For photoenol Z-2P, the transition
state barrier for the 1,5-H shift is on the same order as the rotation barrier required for
intersystem crossing from its precursor 2BR. Therefore, the reaction environment controls
whether intersystem crossing of 2BR or the 1,5-H shift of Z-2P is the slowest process. In
contrast, as photoenol Z-1P has more steric crowding, the transition state barrier to re-form
ketone 1 through the 1,5-H shift is significantly larger than the rate for intersystem crossing
from 1BR.
ACKNOWLEDGEMENTS: We acknowledge funding from NSF (CHE-1464694) and
the Ohio Supercomputer Center for supporting this work. AD is grateful for generous support
from the Chemistry Department at the University of Cincinnati including a RITE fellowship.
Researchers at UVic thank NSERC (RGPIN-121389-2012) for funding and CAMTEC for the use of
shared facilities.
SUPPORTING INFORMATION
Additional Supporting Information is available in the online version of this article:
Arrhenius plot (Figures S1 and S2 and Tables S1 and S2), summary of transient spectroscopy
data (Table S3) and calculations (remaining schemes and tables).
This article is protected by copyright. All rights reserved.

REFERENCES
1
. Haag, R., J. Wirz and P. J. Wagner (1977) The photoenolization of 2-methylacetophenone and
related compounds. Helv. Chim. Acta 60, 2595-2607.
2
. Yang, N. C. and C. Rivas (1961) A new photochemical primary process , the photochemical
enolization of o-substituted benzophenones. J. Am. Chem. Soc. 83, 2213-2213.
3
. Klán, P., T. Š. Solomek, C. G. Bochet, A. l. Blanc, R. Givens, M. Rubina, V. Popik, A. Kostikov and J.
Wirz (2013) Photoremovable protecting groups in chemistry and biology: Reaction mechanisms
and efficacy. Chem. Rev. 113, 119-191.
4
. Klan, P., J. Wirz and A. D. Gudmundsdottir (2012) Photoenolization and its applications. In In
rd
CRC Handbook of Organic Photochemistry & Photobiology, 3 Edition, Vol. 1. (Edited by A.
Griesbeck, M. Oelgemoller and F. Chetti), pp. 627-652. CRC Press, Boca Raton, FL.
5
. Sankaranarayanan, J., S. Muthukrishnan and A. D. Gudmundsdottir (2009) Photoremovable
protecting groups based on photoenolization. Adv. Phys. Org. Chem. 43, 39-77.
6
. Mayer, G. and A. Heckel (2006) Biologically active molecules with a “light switch”. Angew.
Chem., Int. Ed. 45, 4900-4921.
7
. Herrmann, A. Using photolabile protecting groups for the controlled release of bioactive
volatiles. Photochem. Photobio. Sci. 11, 446-459.
8
. Dell'Amico, L., A. Vega-Penaloza, S. Cuadros and P. Melchiorre (2016) Enantioselective
organocatalytic Diels-Alder trapping of photochemically generated hydroxy-o-
quinodimethanes. Angew. Chem., Int. Ed. 55, 3313-3317.
9
. Moorthy, J. N., P. Mal, N. Singhal, P. Venkatakrishnan, R. Malik and P. Venugopalan (2004)
Highly diastereoselective tandem photoenolization-hetero-diels-alder cycloaddition reactions of
o-tolualdehydes in the solid state. J. Org. Chem. 69, 8459-8466.
1
0. Nicolaou, K. C., D. L. F. Gray and J. Tae (2004) Total synthesis of hamigerans and analogues
thereof. Photochemical generation and Diels-Alder trapping of hydroxy-o-quinodimethanes. J.
Am. Chem. Soc. 126, 613-627.
1
1. Nicolaou, K. C. and D. L. F. Gray (2004) Total synthesis of hybocarpone and analogues
thereof. A facile dimerization of naphthazarins to pentacyclic systems. J. Am. Chem. Soc. 126,
6
1
07-612.
2. Pelliccioli, A. P. and J. Wirz (2002) Photoremovable protecting groups: Reaction mechanisms
and applications. Photochem. Photobio. Sci. 1, 441-458.
1
3. Pika, J., A. Konosonoks, R. M. Robinson, P. N. D. Singh and A. D. Gudmundsdottir (2003)
Photoenolization as a means to release alcohols. J. Org. Chem. 68, 1964-1972.
1
4. Konosonoks, A., P. J. Wright, M.-L. Tsao, J. Pika, K. Novak, S. M. Mandel, J. A. Krause Bauer, C.
Bohne and A. D. Gudmundsdóttir (2005) Photoenolization of 2-(2-methyl benzoyl) benzoic acid,
methyl ester:ꢀ Effect of E-photoenol lifetime on the photochemistry. J. Org. Chem. 70, 2763-2770.
This article is protected by copyright. All rights reserved.

1
5. Pelliccioli, A. P., P. Klán, M. Zabadal and J. Wirz (2001) Photorelease of HCl from o-
methylphenacyl chloride proceeds through the Z-xylylenol. J. Am. Chem. Soc. 123, 7931-7932.
1
6. Guerin, B. and L. J. Johnston (1989) Laser flash photolysis studies of 2,4,6-trialkylphenyl
ketones. Can. J. Chem. 67, 473-480.
1
7. Small, R. D. and J. C. Scaiano (1977) Role of biradical intermediates in the photochemistry of
o-methylacetophenone. J. Am. Chem. Soc. 99, 7713-7714.
1
8. Das, A., E. A. Lao and A. D. Gudmundsdottir (2016) Photoenolization of o-
methylvalerophenone ester derivative. Photochem. Photobio. 92, 388-398.
1
9. Muthukrishnan, S., J. Sankaranarayanan, T. C. S. Pace, A. Konosonoks, M. E. DeMichiei, M. J.
Meese, C. Bohne and A. D. Gudmundsdottir (2010) Effect of alkyl substituents on photorelease
from butyrophenone derivatives. J. Org. Chem. 75, 1393-1401.
2
0. Muthukrishnan, S., J. Sankaranarayanan, R. F. Klima, T. C. S. Pace, C. Bohne and A. D.
Gudmundsdottir (2009) Intramolecular H-atom abstraction in -azido-butyrophenones:
Formation of 1,5 ketyl iminyl radicals. Org. Lett. 11, 2345-2348.
2
1. Liao, Y. and C. Bohne (1996) Alcohol effect on equilibrium constants and dissociation
dynamics of xanthone−cyclodextrin complexes. J. Phys. Chem. 100, 734-743.
2
2. Frisch, M. J., G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G.
Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr, J.
E. Peralta, F. o. Ogliaro, M. J. Bearpark, J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, R.
Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M.
Cossi, N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L.
Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D.
Daniels, Ã. d. n. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox (2009) Gaussian 09.
Gaussian, Inc., Wallingford, CT, USA.
2
3. Becke, A. D. (1993) Density functional thermochemistry. III. The role of exact exchange. J.
Phys. Chem. 98, 5648-5652.
2
4. Lee, C., W. Yang and R. G. Parr (1988) Development of the colle-salvetti correlation-energy
formula into a functional of the electron density. Phys. Rev. B 37, 785-789.
2
5. Zhao, Y. and D. G. Truhlar (2008) The M06 suite of density functionals for main group
thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and
transition elements: Two new functionals and systematic testing of four M06-class functionals
and 12 other functionals. Theor. Chem. Acc. 120, 215-241.
2
6. Montgomery, J. A., M. J. Frisch, J. W. Ochterski and G. A. Petersson (2000) A complete basis
set model chemistry. VII. Use of the minimum population localization method. J. Phys. Chem.
12, 6532-6542.
1
This article is protected by copyright. All rights reserved.

2
7. Montgomery, J. A., M. J. Frisch, J. W. Ochterski and G. A. Petersson (1999) A complete basis
set model chemistry. VI. Use of density functional geometries and frequencies. J. Phys. Chem.
10, 2822-2827.
1
2
8. Labanowski, J. K. and J. W. Andzelm (1991) Density functional methods in chemistry. (Edited
by K. L. Jan and W. A. Jan), pp. 443. Springer-Verlag New York, Inc.
2
9. Parr, R. G. and Y. Weitao (1989) Density functional theory in atoms and molecules. Oxford
University Press: Oxford.
3
0. Stratmann, R. E., G. E. Scuseria and M. J. Frisch (1998) An efficient implementation of time-
dependent density-functional theory for the calculation of excitation energies of large
molecules. J. Phys. Chem. 109, 8218-8224.
3
1. Bauernschmitt, R. D. and R. Ahlrichs (1996) Treatment of electronic excitations within the
adiabatic approximation of time dependent density functional theory. Chem. Phys. Lett. 256,
4
3
54-464.
2. Foresman, J. B., M. Head-Gordon, J. A. Pople and M. J. Frisch (1992) Toward a systematic
molecular orbital theory for excited states. J. Phys. Chem. 96, 135-149.
3
3. Gonzalez, C. and H. B. Schlegel (1990) Reaction path following in mass-weighted internal
coordinates. J. Phys. Chem. 94, 5523-5527.
3
4. Gonzalez, C. and H. B. Schlegel (1989) An improved algorithm for reaction path following. J.
Chem. Phys. 90, 2154-2161.
3
5. Ranaweera, R. A. A. U., T. Scott, Q. Li, S. Rajam, A. Duncan, R. Li, A. Evans, C. Bohne, J. P.
Toscano, B. S. Ault and A. D. Gudmundsdottir (2014) Trans–cis isomerization of vinylketones
through triplet 1,2-biradicals. J. Phys. Chem. A 118, 10433-10447.
3
6. Murphy, R. and R. H. Prager (1978) Organoboranes in organic synthesis: Ix. Carbonylation
products of organoboranes derived from myrcene. J. Organom. Chem. 156, 133-147.
3
7. Hog Daniel, T. and M. Oestreich (2009) B(C
using silicon‐stereogenic silanes: Stereoinduction by single‐point binding. Eur. J. Org. Chem.
009, 5047-5056.
6
F
5
) ‐catalyzed reduction of ketones and imines
3
2
3
8. Scaiano, J. C., V. Wintgens and J. C. Netto-Ferreira (1992) Scavenging of photoenols by acids
and bases. Tetrahedron Lett. 33, 5905-5908.
3
9. Kitaura, Y. and T. Matsuura (1971) Photoinduced reactions—XLVIII: Steric and substituent
effects on photoreactions of 2,4,6-trialkylphenyl ketones. Tetrahedron 27, 1597-1606.
4
0. Blair, A. D., A. D. Hendsbee and J. D. Masuda (2011) 1-(2,4,6-Triisopropylphenyl)ethanone.
Acta Crystallogr. Sect. E 67, o2731.
4
1. Ghoshal, S. K., S. K. Sarkar and G. S. Kastha (1981) Effects of intermolecular hydrogen-
bonding on the luminescence properties of acetophenone, characterization of emission states.
Bull. Chem. Soc. Jpn. 54, 3556-3561.
This article is protected by copyright. All rights reserved.

4
2. Borkman, R. F. and D. R. Kearns (1966) Electronic‐relaxation processes in acetone. J. Phys.
Chem. 44, 945-949.
4
3. van Koningsveld, H., J. J. Scheele and J. C. Jansen (1987) Structure of 4-tert-butyl-2,6-
dimethylacetophenone and comparison with its FeCl complex. Acta Crystallogr. Sect. C 43, 294-
3
2
4
96.
4. Muthukrishnan, S., T. C. S. Pace, Q. Li, B. Seok, G. de Jong, C. Bohne and A. D. Gudmundsdottir
(
2011) Comparison of photoenolization and alcohol release from alkyl-substituted benzoyl
benzoic esters. Can. J. Chem. 89, 331-338.
4
5. Li, Q., J. Sankaranarayanan, M. Hawk, V. T. Tran, J. L. Brown and A. D. Gudmundsdottir (2012)
The effects of h-bonding and sterics on the photoreactivity of a trimethyl butyrophenone
derivative. Photochem. Photobio. Sci. 11, 744-751.
4
6. Zabadal, M., A. P. Pelliccioli, P. Klán and J. Wirz (2001) 2,5-dimethylphenacyl esters:ꢀ A
photoremovable protecting group for carboxylic acids. J. Phys. Chem. A 105, 10329-10333.
4
(
7. Kuzmanich, G., J. Xue, J.-C. Netto-Ferreira, J. C. Scaiano, M. Platz and M. A. Garcia-Garibay
2011) Steady state and transient kinetics in crystalline solids: The photochemistry of
nanocrystalline 1,1,3-triphenyl-3-hydroxy-2-indanone. Chemi. Sci. 2, 1497-1501.
4
8. Ito, Y., H. Takahashi, J.-Y. Hasegawa and N. J. Turro (2009) Photocyclization of 2,4,6-
triethylbenzophenones in the solid state. Tetrahedron 65, 677-689.
Table 1. Calculated transition state barriers for 1,5-H shifts in Z-photoenols and experimental
transition state barriers.
Calculated transition state barrier Experimentally measured transition state
-1
barrier (kcal mol-1)
for 1,5-H shift (kcal mol )
B3LYP
M062X
CBS-QB3
1,5-H shift
Intersystem
crossing
Z-1P
Z-2P
7.60
4.79
10.57
7.22
8.35
7.7 ± 0.3
6.299
4.7 ± 0.1
This article is protected by copyright. All rights reserved.

-1
Table 2. Experimental and calculated energies (kcal mol ) of T1K of 1 and 2.
Ketone
Phosphorescence B3LYP
TD-DFT
M062X
Optimization TD-DFT
Optimization
1
2
82 (350 nm)
77 (373 nm)
83
75
72
70
81
77
73
74
FIGURE CAPTIONS
Figure 1. A) Transient spectra obtained by laser flash photolysis of 1 in argon-saturated
methanol. B) TD-DFT calculated spectra of 1BR, Z-1P, and E-1P.
Figure 2. A) Kinetic trace obtained from laser flash photolysis of ketone 1 in argon-saturated
methanol at 350 nm. B) Kinetic traces for the decay of Z-1P in argon- (red) and oxygen-
saturated methanol (black) at 350 nm.
Figure 3. Transient spectra obtained by laser flash photolysis of 2 in argon-saturated
methanol. B) TD-DFT calculated spectra of 2BR, Z-2P, and E-2P. C) Kinetic trace obtained by
laser flash photolysis of ketone 2 in argon- (red) and oxygen-saturated methanol (black) at 350
nm.
Figure 4. A) Transient spectra obtained by laser flash photolysis of 2 in argon-saturated
HMPA. B) Kinetic traces obtained at 330 nm in argon- and oxygen-saturated HMPA.
Figure 5. Kinetic traces at 330 nm obtained by laser flash photolysis of 1 in argon-saturated
methanol as a function of added of NaN (1 M).
3
Figure 6. Arrhenius plots of the decay rate constants obtained by laser flash photolysis of
ketone 1 at 350 nm and ketone 2 at 330 nm in nitrogen-saturated methanol as a function of
temperature.
Figure 7. Optimized structures of 1, 1BR, E-1P, and Z-1P and 2, 2BR, E-2P, and Z-2P using
B3LYP/6-31+G(d).
Figure 8. Calculated stationary points on the energy surfaces of ketones A) 1 and B) 2.
-1
Energies are in kcal mol .
Figure 9. Phosphorescence spectra of ketones A) 1 and B) 2 obtained in ethanol glass at 77 K.
Figure 10. Calculated IRC graphs for the transitions state for forming cyclobutanol 4 from E-
1P (black dots) and ketone 1 from Z-1P (red dots).
This article is protected by copyright. All rights reserved.

Scheme 1.
Scheme 2.
Scheme 3.
Scheme 4. Calculated spin densities for T1K of 1 and T1K of 2.
This article is protected by copyright. All rights reserved.

Scheme 1
This article is protected by copyright. All rights reserved.

Scheme 2
Scheme 3
This article is protected by copyright. All rights reserved.

Scheme 4
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.