Intramolecular [2+2] photocycloaddition of 3- and 4-(but-3-enyl) oxyquinolones: Influence of the alkene substitution pattern, photophysical studies, and enantioselective catalysis by a chiral sensitizer

Article abstract of DOI:10.1002/chem.201300203

The intramolecular [2+2] photocycloaddition of four 4-(but-3-enyl) oxyquinolones (substitution pattern at the terminal alkene carbon atom: CH ₂, Z-CHEt, E-CHEt, CMe₂) and two 3-(but-3-enyl) oxyquinolones (substitution pattern: CH₂, CMe₂) was studied. Upon direct irradiation at λ=300 nm, the respective cyclobutane products were formed in high yields (83-95 %) and for symmetrically substituted substrates with complete diastereoselectivity. Substrates with a Z- or E-substituted terminal double bond showed a stereoconvergent reaction course leading to mixtures of regio- and diastereomers with almost identical composition. The mechanistic course of the photocycloaddition was elucidated by transient absorption spectroscopy. A triplet intermediate was detected for the title compounds, which-in contrast to simple alkoxyquinolones such as 3-butyloxyquinolone and 4-methoxyquinolone-decayed rapidly (τ≈1 ns) through cyclization to a triplet 1,4-diradical. The diradical can evolve through two reaction channels, one leading to the photoproduct and the other leading back to the starting material. When the photocycloaddition was performed in the presence of a chiral sensitizer (10 mol %) upon irradiation at λ=366 nm in trifluorotoluene as the solvent, moderate to high enantioselectivities were achieved. The two 3-(but-3-enyl)oxyquinolones gave enantiomeric excesses (ees) of 60 and 64 % at -25 °C, presumably because a significant racemic background reaction occurred. The 4-substituted quinolones showed higher enantioselectivities (92-96 % ee at -25 °C) and, for the terminally Z- and E-substituted substrates, an improved regio- and diastereoselectivity. Upon UV irradiation quinolones A undergo a rapid cyclization at the T₁ level (τ≈1 ns) through the respective 1,4-diradicals (e.g., 1), leading to the [2+2]-photocycloaddition products in high yields. Lifetimes and transient absorption spectra have been carefully measured. The high cyclization rate is one key element in the enantioselective formation of products, such as 3, when sensitized by catalyst (+)-2 (10 mol %). Copyright

Full text of DOI:10.1002/chem.201300203

FULL PAPER
DOI: 10.1002/chem.201300203
Intramolecular [2+2] Photocycloaddition of 3- and 4-(But-3-
enyl)oxyquinolones: Influence of the Alkene Substitution Pattern,
Photophysical Studies, and Enantioselective Catalysis by a Chiral Sensitizer
[
a]
[b]
[a]
[a]
Mark M. Maturi, Matthias Wenninger, Rafael Alonso, Andreas Bauer,
[
a]
[b]
[a]
Alexander Pçthig, Eberhard Riedle,* and Thorsten Bach*
Abstract: The intramolecular [2+2]
photocycloaddition of four 4-(but-3-
enyl)oxyquinolones (substitution pat-
tern at the terminal alkene carbon
atom: CH , Z-CHEt, E-CHEt, CMe )
cal composition. The mechanistic
course of the photocycloaddition was
elucidated by transient absorption
spectroscopy. A triplet intermediate
was detected for the title compounds,
which–in contrast to simple alkoxyqui-
nolones such as 3-butyloxyquinolone
and 4-methoxyquinolone–decayed rap-
idly (tꢁ1 ns) through cyclization to a
triplet 1,4-diradical. The diradical can
evolve through two reaction channels,
one leading to the photoproduct and
the other leading back to the starting
material. When the photocycloaddition
was performed in the presence of a
chiral sensitizer (10 mol%) upon irra-
diation at l=366 nm in trifluoroto-
luene as the solvent, moderate to high
enantioselectivities were achieved. The
two 3-(but-3-enyl)oxyquinolones gave
enantiomeric excesses (ees) of 60 and
64% at ꢀ258C, presumably because a
significant racemic background reac-
tion occurred. The 4-substituted quino-
lones showed higher enantioselectivi-
ties (92–96% ee at ꢀ258C) and, for the
terminally Z- and E-substituted sub-
strates, an improved regio- and diaster-
eoselectivity.
2
2
and two 3-(but-3-enyl)oxyquinolones
substitution pattern: CH , CMe ) was
(
2
2
studied. Upon direct irradiation at l=
00 nm, the respective cyclobutane
products were formed in high yields
83–95%) and for symmetrically substi-
3
(
tuted substrates with complete diaster-
eoselectivity. Substrates with a Z- or E-
substituted terminal double bond
Keywords: cycloaddition · enantio-
selectivity · photochemistry · sensi-
tizers · time-resolved spectroscopy
showed
a stereoconvergent reaction
course leading to mixtures of regio-
and diastereomers with almost identi-
[
2a]
Introduction
report by Evanega and Fabiny was followed in rapid suc-
cession by several significant contributions from other re-
search groups, most notably from the group of Naito and
[1]
Quinolone [2+2] photocycloaddition reactions are known
since the late 1960s when Evanega and Fabiny reported on
the intermolecular reaction of the parent compound (car-
[4]
Kaneko in the 1980s. Intramolecular [2+2] photocycload-
dition reactions were first reported in 1979 and a few re-
actions of this type were subsequently studied more close-
[4f]
[2]
bostyril) with various alkenes. These studies were preced-
[3a]
[3b]
[3c]
[4g,s,5]
ed by work of Taylor et al., Elliott, and Buchardt on
ly.
The synthetic importance of quinolone [2+2] photo-
the [2+2] photodimerization of quinolones. The initial
cycloaddition is mainly due to further ring-opening reactions
[1a,e,6]
at the cyclobutane ring,
which enable access to highly
substituted quinolone derivatives (see below).
+
+
Mechanistic work on the quinolone [2+2] photodimeriza-
tion and [2+2] photocycloaddition has revealed that the re-
actions proceed through the respective triplet excited state
[
a] M. M. Maturi, Dr. R. Alonso, Dr. A. Bauer, Dr. A. Pçthig,
Prof. Dr. T. Bach
Department Chemie and
Catalysis Research Center (CRC)
Technische Universitꢀt Mꢁnchen
Lichtenbergstrasse 4
(
T ), which is populated by direct excitation at l=300–
1
3
50 nm to the first excited singlet state (S ) and subsequent
1
[4d,7]
rapid intersystem crossing.
The triplet state can also be
8
5747 Garching (Germany)
Fax : (+49) 89-289-13315
E-mail: thorsten.bach@ch.tum.de
populated by sensitization employing compounds with high
triplet energy states, for example, benzophenone. The triplet
+
[
b] M. Wenninger, Prof. Dr. E. Riedle
Lehrstuhl fꢁr BioMolekulare Optik
Ludwig-Maximilians-Universitꢀt
state energy of the parent quinolone has been reported as
ꢀ1
E(T )=276 kJmol
4
corresponding to the energy of a
1
[7a]
34 nm photon. The photodimerization of quinolones has
been studied by laser flash photolysis,
Oettingenstrasse 67, 80538 Mꢁnchen (Germany)
E-mail: riedle@physik.uni-muenchen.de
[7a,b,8]
however, to the
+
best of our knowledge, the course of the intramolecular
2+2] photocycloaddition of quinolones has until recently
[
] These authors contributed equally to this work.
[
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201300203.
Chem. Eur. J. 2013, 19, 7461 – 7472
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7461

[
9]
(
see below) not been investigated by transient absorption
spectroscopy.
[2+2] Photocycloaddition reactions of quinolones were in-
different behavior in the catalytic reaction seemed to arise
from the different kinetic parameters involved in the respec-
tive [2+2] photocycloaddition. It was speculated that upon
triplet sensitization within a substrate–catalyst complex the
subsequent cyclization to an intermediate 1,4-biradical
should be faster for substrate 4 (five-membered ring forma-
tion) than for substrate 2 (six-membered ring formation).
The rate of dissociation from the chiral lactam (+)-1 com-
petes efficiently with the cyclization rate in the latter case,
causing this reaction to occur with lower enantioselectivity.
Indeed, it was experimentally shown that the triplet lifetime
of compound 4 is significantly shorter than the triplet life-
time of compound 2. The lifetime of compound 2 in deaerat-
ed trifluorotoluene at ambient temperature could be deter-
ACHTUNGTRENNUNG
[5b,10]
itially studied in our group
lective photochemical reactions mediated by a chiral tem-
Two-point hydrogen-bonding of the quinolone
lactam to the lactam part of a chiral bicyclic skeleton de-
rived from Kempꢃs triacid led to an efficient enantioface
differentiation. An intermolecular reaction of this type was
implemented as key step in the first enantioselective total
synthesis of the Melodinus alkaloid (+)-meloscine. In this
case, a ring expansion of the cyclobutane to a cyclopenta-
none was achieved by a retro-benzilic acid rearrange-
ment.
When searching for catalytic enantioselective photochemi-
cal reactions, we discovered that dextrorotatory xanthone
in the context of enantiose-
[11]
plate.
[12]
[13]
[14]
mined by laser flash photolysis as t [T (2)]=57 ns. However,
1
[
15]
the detection limit of the laser setup did not allow for an ac-
(
+)-1 and its levorotatory enantiomer (ꢀ)-1 exhibit superior
curate determination of t [T (4)], which was estimated to be
at least by a factor of 0.4 shorter than t [T (2)].
1
1
[9]
catalytic activity in intramolecular quinolone [2+2] photocy-
cloaddition reactions provided that the reacting quinolone
In the present study we have largely broadened the scope
of quinolone substrates and investigated the reaction behav-
ior of differently substituted 3- and 4-(but-3-enyl)oxyquino-
lones 6–9 (Figure 1) in the absence and the presence of a
chiral sensitizer.
[9,16]
was appropriately chosen (Scheme 1).
It was found that
Figure 1. Quinolone substrates 6–9 employed in this study.
The key objective of the study was to prove for 4-(but-3-
enyl)oxyquinolones that a higher cyclization rate consistent-
ly leads to high enantioselectivity. By using the diastereo-
merically pure substrates (Z)-6 and (E)-6, conclusions re-
garding the stereospecificity of the reaction were expected
to be drawn. Data for the enantioselective reactions were
collected at three different temperatures (ꢀ258C, 08C, and
RT). The yet completely unexplored 3-(but-3-enyl)oxyqui-
nolones were studied for comparison. The reaction kinetics
of the photocycloaddition was followed by transient absorp-
tion spectroscopy both on the fs–ps and on the ns–ms time-
scale to provide a comprehensive picture of the reaction
course through the respective singlet- and triplet state life-
times.
Scheme 1. Major differences in the enantioselectivity (ee) of the intramo-
lecular [2+2] photocycloaddition of 4-(pent-4-enyl)oxyquinolone (2) and
4
-(but-3-enyl)oxyquinolone (4) upon sensitization with xanthone (+)-1;
structure of crossed regioisomers crossed-3 and crossed-5 (r.r.=regioiso-
meric ratio).
the reaction of 4-(but-3-enyl)oxyquinolone (4) proceeded
under identical conditions with a significantly higher enan-
tioselectivity than the reaction of 4-(pent-4-enyl)oxyquino-
lone (2), the alkyl tether of which is longer by one methyl-
ene group than the tether of the former substrate.
The binding properties of both substrates 2 versus 4 and
products 3 versus 5 to lactam (+)-1 should be almost identi-
cal. Indeed, it had been previously shown that the reactions
2
!3 and 4!5 proceed with high enantioselectivity in the
Results and Discussion
[9,10c]
presence of a stoichiometrically employed template,
which bears the same 1,5,7-trimethyl-3-azabicyclo-
ACHTUNGTRENNUNG[ 3.3.1]nonan-2-one skeleton as catalyst 1. Consequently, the
Synthesis of the quinolone substrates: The synthesis of 4-
(but-3-enyl)oxyquinolones 6 and 7 (Scheme 2) commenced
7462
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7461 – 7472

Photocycloaddition of 3- and 4-(But-3-enyl)oxyquinolones
FULL PAPER
(
light source: RPR-3000 ꢄ, emission spectrum, see the Sup-
porting Information; c=5 mm). Since the quinolone absorp-
tion is relatively high at this wavelength, [2+2] photocy-
cloaddition reactions tend to proceed smoothly without ad-
dition of a sensitizer. Indeed, 4-(hex-3-enyl)oxyquinolones
(
6) were converted within 40 min into the respective prod-
ucts. Surprisingly, the regioselectivity was relatively low and,
besides the straight products rac-14, significant amounts of
the crossed product rac-15 were obtained (Scheme 4). The
Scheme 2. Synthesis of 4-(alk-3-enyl)oxyquinolones 6 and 7 from 4-chlor-
oquinolin-N-oxide (10) by nucleophilic substitution and subsequent pho-
tochemical rearrangement.
with an aromatic nucleophilic substitution reaction at 4-
[17]
chloroquinolin-N-oxide
(10). The respective alcohols,
which were required to produce the substitution products 11
and 12, are commercially available. The subsequent rear-
rangement of the N-oxides to the desired products was per-
[18]
Scheme 4. Intramolecular [2+2] photocycloaddition of 4-(hex-3-enyl)oxy-
quinolones (6) to straight products rac-14 and crossed product rac-15
formed according to a previously reported method, which
[19]
we have optimized in recent years. Two modifications are
crucial. Firstly, the reaction, which proceeds at the singlet
hypersurface, is performed in an oxygen-saturated methanol
solution, which avoids reactions of the products from the
triplet state (see below). Secondly, a continuous flow system
is being used for irradiation, which allows one to optimize
the exposure time of the substrates to the light source. It
was gratifying to note that this optimization has led to a
high level of reproducibility and generality. Product forma-
tion was essentially quantitative in all cases and side prod-
ucts were not detected (Scheme 2).
(
r.r.=14:15; d.r.=14a:14b); light source: RPR-3000 ꢄ (emission spec-
trum, see the Supporting Information).
straight products rac-14 were formed as two diastereoisom-
ers with rac-14a prevailing over rac-14b. A second crossed
diastereoisomer apart from rac-15 was not observed. The
relative configuration of the products was proven by exten-
1
sive one- and two-dimensional H NMR studies (see the
Supporting Information for further details). Although it is
not apparent from the two-dimensional drawings, the prefer-
ence for the respective diastereoisomer rac-14a in the
straight series and rac-15 in the crossed series can be readily
understood from three-dimensional molecular models. In
the minor straight diastereoisomer rac-14b, the ethyl group
shows a major steric interaction with the dihydroquinolone
ring, which makes the formation of this diastereoisomer less
favorable. The situation, that is, the interaction of the ethyl
group with the dihydroquinolone ring, would be even worse
in the diastereoisomer of rac-15, which is consequently not
formed at all. The photocycloaddition was not stereospecific
as expected from a reaction occurring mainly in the triplet
The preparation of 3-(but-3-enyl)oxyquinolones 8 and 9
was straightforward based on literature precedence
[20]
(
Scheme 3). Known 3-hydroxyquinolone (13) was treated
Scheme 3. Synthesis of 3-(alk-3-enyl)oxyquinolones 8 and 9 from the re-
spective bromides by nucleophilic substitution with 3-hydroxyquinolone
[22,23]
manifold.
Even the E isomer (E)-6, with the ethyl group and the
alkyl chain in a relative trans-configuration, gave predomi-
nantly products rac-14a and rac-15, in which these substitu-
ents are cis-positioned. The fact that the relative product
ratios (regioisomeric ratio (r.r.) and diastereomeric ratio
(d.r.)) are not completely identical for (Z)-6 and (E)-6
seems to indicate a minor contribution of the singlet excited
(
13).
with the respective alkenyl bromides and 1,8-diazabicyclo-
[21]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
[5.4.0]undec-7-ene (DBU) in isopropanol
to deliver the
desired products by nucleophilic substitution. In an analo-
gous fashion, 3-butoxyquinolone was prepared from 13 and
butyl bromide.
state (S ) to the net reaction. Based on the transient spectro-
1
scopy data (see below), however, it is more likely that differ-
ences in the first bond formation and in the subsequent re-
action branching within the triplet manifold are responsible
for this result. The Z isomer (Z)-6 exhibits more than a two-
fold product efficiency relative to (E)-6. If the irradiation
ACHTUNGTRENNUNG[ 2+2] Photocycloaddition reactions in the absence of a sen-
sitizer at l=300 nm: To check the general feasibility of the
photocycloaddition and to secure racemic material for ana-
lytical purposes, reactions were performed at l=300 nm
Chem. Eur. J. 2013, 19, 7461 – 7472
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7463

T. Bach, E. Riedle, et al.
experiments were stopped before the conversion was com-
plete, the re-isolated starting material 6 showed epimeriza-
tion around the double bond. At 08C the reactions (c=
rapid five-membered ring cyclization it was expected that
these lifetimes are significantly shorter than the 57 ns deter-
[9]
mined for t [T (2)]. A second goal was to substantiate the
1
5
mm in PhCF ) were interrupted after five minutes of irra-
mechanistic course of the intramolecular quinolone [2+2]
photocycloaddition in an extended time interval (fs–ms) with
detection of all relevant intermediates. Time-resolved spec-
troscopic data covering a sufficient temporal and spectral
range for quinolone [2+2] photocycloaddition reaction are
3
diation. If Z-6 was employed as starting material, the con-
version was 60% and the re-isolated substrate showed an E/
Z ratio of 31/69. If E-6 was used, the conversion was 44%
and starting material 6 was recovered in an E/Z ratio of 51/
[7–9]
4
9.
The outcome of the [2+2] photocycloaddition of substrate
was less complex than that of substrates 6. Indeed, a single
scarce
and it was required to obtain a solid database for
future studies.
7
All decay rates and species-associated spectra (SAS) of
the occurring intermediate states discussed below were de-
termined by applying a global data analysis on the measured
data matrices (see the Experimental Section). 4-Methoxy-
quinolone (19) and 3-butoxyquinolone (20), which cannot
react in an intramolecular [2+2] photocycloaddition, served
as reference to study the intrinsic photophysical properties
reaction product rac-16 was obtained in high yield
(
(
Scheme 5). Complete conversion was achieved after 3.5 h
c=5 mm).
such as lifetime and spectral properties of the S and T1
1
state (Figure 2).
Scheme 5. Selective intramolecular [2+2] photocycloaddition of 4-(4-
methylpent-3-enyl)oxyquinolone (7) to straight product rac-16.
Figure 2. Structures of 4-methoxyquinolone (19) and 3-butoxyquinolone
(
20), which were used in the present study as reference compounds.
The excitation wavelength was l=270–275 nm in all tran-
sient spectroscopy measurements. A reference measurement
for 4 at l=330 nm showed that the observed dynamics are
Scheme 6. Selective intramolecular [2+2] photocycloaddition of 3-(alk-3-
enyl)oxyquinolones 8 and 9 to straight products rac-17 and rac-18.
not changed by the excitation to the S state instead of the
2
S1 state. Since the influence of pump stray light can be dra-
matically reduced by the shorter wavelength, this choice was
preferred. To secure optical transparency in the UV range
and due to solubility reasons, acetonitrile was employed as
the solvent in all experiments ensuring a homogeneous solu-
tion of equal concentration (cꢁ0.5 mm, optical density
[318 nm, 1 mm]ꢁ0.5) for all substrates. The studies were
performed under aerobic conditions at room temperature.
The solubility of oxygen in air-equilibrated acetonitrile is
Similarly, the reaction of 3-(alk-3-enyl)oxyquinolones 8
and 9 (c=2.5 mm) proceeded smoothly providing single
products rac-16 and rac-17 in high yields (Scheme 6). It is
notable that the crossed regioisomer was not observed in
the photocycloaddition of 3-(but-3-enyl)oxyquinolone (8).
The reaction of the 4-substituted analogue 4 delivers under
identical conditions the cycloaddition products in a regioiso-
meric ratio of 86:14. The regioselective outcome of the re-
action of substrate 9 is in accord with the above-mentioned
reaction of the 4-substituted analogue 7. The reaction times
required for complete conversion were relatively short for
the 3-alkenyloxy-substituted quinolones 8 and 9 reflecting in
part their higher absorption at l=300 nm (see below).
[9]
[24]
known to be 2.4 mm and at this oxygen concentration an
interference with intramolecular reaction processes (display-
7
ꢀ1
ing rate constants larger than 10 s ) was not to be expect-
[25]
ed.
For the time-resolved measurements and their interpreta-
tion the steady-state absorption and emission spectra of all
compounds were determined. As expected, the spectral
properties were equal within the measurement error for all
4-substituted quinolones (19, 4, (Z)-6, (E)-6, 7) and for all 3-
substituted quinolones (20, 8, 9), since the electronic proper-
ties of the absorbing chromophore remain constant within
each substrate series. The measured spectra are exemplarily
shown for both chromophores 19 and 20 in Figure 3. Al-
though the shape of the spectra is very similar for the two
compounds, it should be noted that the absorption of the 3-
Transient absorption spectroscopy on multiple timescales:
Time-resolved spectroscopic studies, both on the fs–ps and
on the ns–ms timescale, were performed with all newly pre-
pared quinolones 6–9 as well as with the previously reported
substrate 4. Two major goals were associated with this study.
The first goal was to obtain precise data about the liftetimes
t [T ] of the triplet intermediates derived from substrates 4
1
and 6–9. Given the previously mentioned assumption of a
7464
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7461 – 7472

Photocycloaddition of 3- and 4-(But-3-enyl)oxyquinolones
FULL PAPER
Table 1. Parametrization of the excited state dynamics of all synthesized
quinolones: The S
,4-diradical lifetime t [T-DR] were directly obtained from transient ab-
sorption spectroscopy. The quantum yield for product formation F was
determined by actinometric measurements. The quantum yields for trip-
let formation FT1, population of the triplet 1,4-diradical F and the ratio
clos/kretro of the rates for second bond-formation and retrosynthetic dis-
1 1 1
lifetime t [S ], the triplet lifetime t [T ] and the triplet
1
P
D
k
[
a]
connection were calculated on the basis of a sequential rate model.
Entry
t (S
ps]
1
)
t (T
[ns]
1
)
t (T-DR)
[ns]
F
P
F
T1
F
D
k
clos/kretro
[
1
2
3
4
5
6
7
8
19
4
229
237
116
–
–
0.98
–
–
0.651 29
1.319 30
0.934 32
0.647 34
0.30 0.98 0.98 31:69
0.72 0.99 0.97 74:26
0.32 0.99 0.98 33:67
0.26 0.99 0.98 26:74
(Z)-6 199
(E)-6 202
7
20
8
9
176
296
362
301
81
–
23
–
0.97
–
–
[
b]
ꢁ1.5
0.73 0.96 0.95 77:23
0.54 0.97 0.96 57:43
0.957 25
Figure 3. UV/Vis absorption spectra (cꢁ0.1 mm) and normalized emis-
sion (cꢁ0.01 mm) after 315 nm excitation of a) 4-methoxyquinolone (19)
and b) 3-butoxyquinolone (20) in MeCN, which are typical for all other
studied quinolones.
[
a] The excitation wavelength was l=270–275 nm in all transient spectro-
scopy measurements. Acetonitrile was employed as the solvent ensuring
a homogeneous solution of equal concentration (cꢁ0.5 mm, optical den-
sity [318 nm, 1 mm]ꢁ0.5) for all substrates. [b] The lifetime is in a time
window that is not well-covered with the fs–ps (too slow) or with the ns–
ms (too fast) setup.
substituted quinolone, that is, compound 20, is by a factor 2
more intense than the absorption of 19. The steady-state
spectra allow estimating the radiative decay rate k of the
rad
[26]
S state according to Strickler and Berg.
We calculated
the 10 ns timescale (see above), a triplet yield close to unity
can be assumed (see Table 1). A direct return to the elec-
tronic ground state by internal conversion can be excluded
due to the lack of any significant ground state bleach com-
ponent in the spectrum associated with the 250 ps decay
(see the Supporting Information). The lowest triplet state
exhibits a clear absorption band centered at 450 (19) and
400 nm (20), respectively, which is in agreement with report-
1
ꢀ
1
ꢀ1
the values of krad =(14 ns) for 19 and krad =(9 ns) for 20.
The photophysical properties of reference compounds 19
and 20 were studied by transient absorption spectroscopy. A
very broad, weakly structured excited state absorption of
the S1 state was found, which is overlaid by signals of
ground state bleach (310–340 nm) and stimulated emission
(
370–450 nm) both for 19 (compare Dt=1 ps in Figure 4a)
[7a,b]
and 20. The lifetime of the S state is dominated by a rela-
ed triplet spectra of related compounds.
According to
1
[9]
tively fast intersystem crossing (ISC) into the triplet mani-
fold on the 250 ps timescale. Exact values for the S lifetime
of the various compounds are given in Table 1. Since the
competing radiative deactivation of the S state occurs on
previous data, the triplet state lifetime of 19 in degassed
trifluorotoluene is in the ms range. Due to oxygen-mediated
triplet quenching we observe a significantly reduced lifetime
in the 100 ns domain both for 19 and 20 in air-equilibrated
MeCN. The exact values are summarized in Table 1.
1
1
Since all synthesized quinolones 4 and 6–9, which can un-
dergo intramolecular [2+2] photocycloaddition, qualitatively
show the same photochemical behavior and only differ
quantitatively, the following discussion of the reaction kinet-
ics is based on quinolone 4 (Figure 4). Compound 4 exhibits
after excitation the same spectral properties of the S state
1
as already observed for the reference compound 19 (see Fig-
ure 3a). ISC occurs on the 100 ps timescale, whereas the
lifetime t [T ] of the triplet state is significantly reduced
1
compared with reference 19. We find a lifetime of t
[
T (4)]=651 ps as compared to t [T (19)]=116 ns (Table 1).
1 1
However, in contrast to 19, the T state of 4 does not decay
1
to the ground state, but a third state with a significant re-
maining ground state bleach signature is populated. Since
the competing decay of the T1 state by oxygen-mediated
triplet quenching occurs on a much longer timescale (as
measured for 19) a close to unity quantum yield for the pop-
ulation of this newly formed state can be safely assumed.
This transient state decays with a time constant of t [T-
DR(4)]=29 ns. The remaining spectral signature is perma-
Figure 4. Transient absorption from measurements on the ps (a) and on
the ns (b) timescale of quinolone 4 in MeCN at selected pump-probe
delays Dt after 270 nm excitation.
Chem. Eur. J. 2013, 19, 7461 – 7472
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7465

T. Bach, E. Riedle, et al.
nent within the observed temporal window of several ms and
significantly lower in ground state bleach intensity than its
parent state. This observation strongly indicates that the
parent state is depopulated by two competing processes:
one leading back to the starting material and another one
leading to the formation of a stable product, which does not
absorb in the spectral range above 300 nm (see Figure 4b).
The transient absorption spectra are fully in line with the
mechanistic course previously suggested for related enone
(kclos) and the retrocleavage (kretro) are likely to be responsi-
ble for the branching leading to the product in one reaction
channel and to the starting material in the other reaction
channel. A putative singlet 1,4-diradical (S-DR; not shown
in Scheme 7) is spectroscopically inaccessible, since its popu-
lation rate is presumably much lower than the sum of de-
population rates kretro and kclos. Still, since the lifetime t [T-
DR] of the triplet 1,4-diradical stays constant within the
measurement error for both chromophores, the branching
between product formation and retrocleavage likely occurs
not in the triplet 1,4-diradical state but happens in the sin-
glet 1,4-diradical.
[23,27]
[2+2] photocycloaddition reactions,
in which a triplet
1,4-diradical state (T-DR) is proposed to be the key inter-
mediate of the reaction. Adapted to the reaction of quino-
lones, a picture evolves as drawn in Scheme 7 for the proto-
The ratio of kclos and kretro is to some extent influenced by
the substitution pattern of the side chain. The cis configura-
tion of the ethyl group in (Z)-6 for instance favors the
second bond formation (Table 1) leading to a higher product
quantum yield F for (Z)-6 (0.72) than for (E)-6 (0.32). The
P
higher quantum yield is qualitatively reflected by the previ-
ously mentioned higher conversion of (Z)-6 (60%) as com-
pared with (E)-6 (44%) after five minutes of irradiation. A
terminal dimethyl substitution favors the retrocleavage over
the ring-closure as evident when comparing these substrates
to the respective unsubstituted alkenyloxyquinolones (cf., 4
vs. 7 and 8 vs. 9 in Table 1).
Based on a sequential rate model, taking the determined
rates, spectra, and states into account, the species-associated
absorption spectra (SAS) of the T and the triplet 1,4-diradi-
1
cal state can be extracted from the transient data matrices.
Scheme 7. Mechanistic scheme for the intramolecular [2+2] photocy-
cloaddition of substrate 4 serving as a general scheme for the discussion
of all other quinolone substrates (for nomenclature and abbreviations,
see narrative). For the structure of the crossed regioisomer, see
Scheme 1.
The obtained spectrum for the T state of 4 is in agreement
1
with the previously published T absorption spectra of relat-
1
[7a,b,8]
ed compounds
and shown in Figure 5. The maximum of
typical substrate 4. The triplet 1,4-diradical is formed from
the T state, which in turn is populated from the respective
1
singlet excited state with the ISC rate constant kISC. For all
substrates 4, 6–9, which can undergo an intramolecular
[
2+2] photocycloaddition, the triplet state lifetime t [T ] is
1
limited by the cyclization to the triplet 1,4-diradical T-DR.
The cyclization occurs between the photoactivated double
bond and the adjacent carbon–carbon double bond of the
side chain, which is not part of the absorbing chromophore.
In agreement with previously reported rate constants for
1
Figure 5. Species-associated absorption spectra (SAS) of the T state of
the representative quinolones 4-(but-3-enyl)oxyquinolone (4) and 3-(but-
3-enyl)oxyquinolone (8) after 270 nm excitation in MeCN.
[28]
kcycl in enone [2+2] photocycloaddition reactions
this
process is rapid and occurs on the timescale of t [T ]ꢁ1 ns
1
(
Table 1).
The observed lifetimes t [T-DR], which fit very well to
[
29]
previously reported triplet 1,4-diradical lifetimes, strongly
support the assignment of the spectral signatures of the oc-
curring intermediates to the respective triplet 1,4-diradicals.
Although the determined reaction kinetics suggest a close to
unity quantum yield for the population of the triplet 1,4-dir-
adical state, the actinometrically determined quantum yields
FP for product formation are clearly below unity (see
Table 1). Therefore branching reaction kinetics, in which
one state is depopulated by two competing processes with
comparable probabilities, can be assumed. The ring-closure
the T–T absorption is observed at lꢁ450 nm for 4, whereas
the T–T absorption of the 3-substitued quinolone 8 is blue-
shifted to lꢁ400 nm.
The calculated SAS of the triplet 1,4-diradical state of
both chromophores exhibit a strong, characteristic absorp-
tion band centered at around 300 nm (4, (Z)-6, (E)-6, 7) and
at approximately 350 nm (8, 9) as shown for the 1,4-diradi-
cals derived from 7 and 9 in Figure 6. Interestingly, the trip-
let 1,4-diradicals of 4, (Z)-6, (E)-6 and 7 show an additional,
less-intense absorption band centered around 380 nm
7466
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7461 – 7472

Photocycloaddition of 3- and 4-(But-3-enyl)oxyquinolones
FULL PAPER
Table 2. Enantioselective intramolecular [2+2] photocycloaddition of
quinolones 6 at l=366 nm to provide enantiomerically enriched products
[
a]
1
4 and 15.
[
b]
[c]
[c]
[d]
Entry
T
Yield
r.r.
d.r.
ee [%] ee [%] ee [%]
(14a) (14b) (15)
Figure 6. Species-associated absorption spectra (SAS) of the triplet 1,4-
diradical (T-DR) states of selected quinolones 7 and 9.
[
8C] [%]
14:15 14a:14b
A
T
N
T
E
N
N
ACHTUNGTRENNUNG
1
2
3
4
5
6
(E)-6 RT 93
(E)-6 89
76/24 76/24
82/18 82/18
82/18 85/15
86/14 85/15
88/12 88/12
88/12 88/12
78
90
92
80
88
92
72
82
84
70
78
84
68
74
74
66
68
70
0
[
[
e]
e]
(
dotted blue line, Figure 6), which exhibits a Franck–
(E)-6 ꢀ25 91
!
(Z)-6 RT 91
Condon progression, which is very similar to the S1 S tran-
0
(Z)-6
0 89
sition (Figure 3a). Also, the triplet 1,4-diradicals derived
from compounds 8 and 9 exhibit an additional, redshifted,
less-intense absorption band, which could not be well re-
solved due to a low signal intensity (arrow, Figure 6). Since
the spectra of the 1,4-diradical states are well separated
from the triplet spectra, transient absorption spectroscopy
can be used to differentiate between the two species and
follow the course of the [2+2] photocycloaddition reaction.
(Z)-6 ꢀ25 90
[
a] All reactions were conducted at the indicated temperature T by using
a RPR-100 reactor with sixteen 366 nm, 8 W fluorescent lamps (e.g., Phi-
lips black light blue; emission spectrum, see the Supporting Information)
as the irradiation source in dry, deaerated trifluorotoluene (c=5 mm) as
the solvent. [b] Yield of isolated product. [c] The ratio of regioisomers
(
r.r.) and diastereoisomers (d.r.) ratio was determined by HPLC. [d] The
enantiomeric excess (ee) was determined by chiral HPLC after separa-
tion of the individual isomers 14a, 14b, and 15. [e] The irradiation time
was 45 min.
Enantioselective [2+2] photocycloaddition reactions in the
presence of a sensitzer at l=366 nm: Enantioselective reac-
tions were performed in the presence of the chiral sensitizer
prehensively provided in Table 2. The reaction yields were
high for either starting diastereoisomer (E)-6 (Table 2, en-
tries 1–3) or (Z)-6 (Table 2, entries 4–6). As observed in the
racemic series, the influence of the starting olefin configura-
tion was minimal. The Z isomer delivered slightly higher se-
lectivities regarding the formation of regioisomers and dia-
stereoisomers, the effect being most pronounced at room
temperature (Table 2, entries 1 and 4). In this category, how-
ever, the most remarkable effect is exerted by the catalyst.
If compared under otherwise identical conditions to the rac-
emic series (Scheme 4), the r.r. increased from 60:40 to
76:24 (Table 2, entry 1) for (E)-6 and from 65:35 to 86:14
(Table 2, entry 4) for (Z)-6. The d.r increased from 51:49 to
76:24 (entry 1) for (E)-6 and from 77:23 to 85:15 (entry 4)
for (Z)-6.
(
+)-1 or its enantiomer (ꢀ)-1. The sensitizer is complexed
to the substrate with a high association constant and serves
[16]
two purposes. First, it provides enantioselectivity due to
steric hindrance (see below). Second, it ensures at the
chosen wavelength a selective excitation only within the
substrate-sensitizer complex. The excitation of uncomplexed
sensitizer leads to a loss of photons as diffusional activation
is kept low by the choice of concentrations. If optical excita-
tion of the substrate is possible within the chosen wave-
length a racemic background reaction must be expected due
to the large number of uncomplexed substrate molecules.
The reactions were performed at three different tempera-
tures, that is, ꢀ258C, 08C and ambient temperature (RT)
employing 10 mol% of the catalyst. For the reactions of the
4
-alkenyloxyquinolones the substrate concentration was
The enantiomeric excess was measured for the individual
products after having completely separated the three iso-
mers 14a, 14b, and 15 by preparative HPLC. The absolute
configuration was assigned based on analogy to the known
configuration of compound 3 and of a crossed product relat-
ed to 15. The assignment is supported by specific rotation
data. In general, the enantiomeric excesses decreased in
the order 14a>14b>15. By comparing Table 2 entries 1
versus 4, 2 versus 5, and 3 versus 6 it is obvious that the in-
fluence of the starting olefin geometry does not significantly
influence the ee value. However, there is a temperature in-
fluence, which has also been observed for substrate 4.
Lower reaction temperatures favor an enantioselective
product formation.
consistently 5 mm in trifluorotoluene as the solvent. Since
the enantioselective reactions of substrate 4 (Scheme 1) had
so far been only performed at ꢀ258C, the reactions were
also conducted at higher temperature. Complete conversion
was achieved after a reaction time of 50 min. At 08C the
yield was 84% with a regioselectivity of r.r.=76:24, the
major regioisomer 5 showing an enantiomeric excess of
[9]
91%. At ambient temperature the total yield was 82% with
an r.r. of 81:19 and 88% ee for the straight product 5. In
previous work the enantioselectivity for the reaction of sub-
strate 2 (Scheme 1) had been determined to be 11% ee at
3
08C and 15% ee at 08C.
Due to the occurrence of three isomers, the [2+2] photo-
cycloaddition of substrates 6 was (among all reactions stud-
ied) the most complex to analyze and the results are com-
The temperature effect was slightly more pronounced for
substrate 7 compared with substrate 4. The former quino-
Chem. Eur. J. 2013, 19, 7461 – 7472
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7467

T. Bach, E. Riedle, et al.
Table 3. Enantioselective intramolecular [2+2] photocycloaddition of
quinolones 8 and 9 at l=366 nm to provide enantiomerically enriched
[
a]
products ent-17 and ent-18, respectively.
Scheme 8. Highly enantioselective intramolecular [2+2] photocycloaddi-
tion of 4-(4-methylpent-3-enyl)oxyquinolone (7) to product 16.
lone reacted with excellent, so far unprecedented enantiose-
lectivity at ꢀ258C (96% ee) and at 08C (92% ee) but
showed a lower selectivity (84% ee) at room temperature
[
b]
[c]
Entry
R
T
Yield
[%]
ee
[
8C]
[%]
(
1
Scheme 8). Yields of the [2+2] photocycloaddition product
6 were consistently high and the enantioselectivities are re-
markable given that only 10 mol% of the catalyst was used
1
2
3
4
5
6
8
8
8
9
9
9
H
H
H
Me
Me
Me
RT
0
ꢀ25
RT
0
85
83
85
86
88
85
46
54
64
58
60
60
[
d]
(
(
Scheme 8). In particular, the reaction outcome at ꢀ258C
90% yield, 96% ee) powerfully demonstrates that enantio-
ꢀ25
selective catalytic photochemistry is getting competive with
other areas of enantioselective organocatalysis.
[
a] All reactions were conducted at the indicated temperature T by using
a RPR-100 reactor with sixteen 366 nm, 8 W fluorescent lamps (e.g., Phi-
lips black light blue; emission spectrum, see the Supporting Information)
as the irradiation source in dry, deaerated trifluorotoluene (c=2.5 mm) as
the solvent. [b] Yield of isolated product. [c] The enantiomeric excess
was determined by chiral HPLC. [d] The irradiation time was 15 min.
Compared with the high enantioselectivities achieved with
-alkenyloxyquinolones 4 and 7, the results obtained with
4
their 3-substituted analogues 8 and 9 were somewhat disap-
pointing. Due to a lower solubility, these reactions were per-
formed at a substrate concentration of c=2.5 mm in trifluor-
otoluene. The enantiomer (ꢀ)-1 of catalyst (+)-1 was em-
ployed for reasons of supply. Compound 1 is prepared race-
mically and subsequently separated by chiral HPLC deliv-
ering equal amounts of (ꢀ)-1 and (+)-1. To keep the
nomenclature consistent the major enantiomers obtained
from (ꢀ)-1, which is the enantiomer ent-1 of (+)-1, are
named ent-17 and ent-18. The [2+2] photocycloaddition re-
actions proceeded rapidly and delivered the respective prod-
ucts in high yields of 83–88%. The enantioselectivity in-
Factors governing the stereoselectivity in the [2+2] photocy-
cloaddition reaction: A major outcome of the photophysical
studies is the fact that they substantiate the previous as-
sumption of a high cyclization rate kcycl (i.e., a short triplet
lifetime) favoring high enantioselectivity. Substrates 4, 6,
and 7 show triplet state lifetimes in the order of 1 ns, where-
as the triplet state lifetime of 2 is much higher. In the latter
case the dissociation rate from catalyst (+)-1 is apparently
in the order of the cyclization rate (10 –10 s ) accounting
for a large fraction of reactions to occur after dissociation
from the catalyst. The enantioselectivity of the reaction 2!
3 is consequently low (see above). As shown in Scheme 7,
the cyclization step determines the absolute configuration of
the product irrespective of its regioselectivity. An enantio-
face differentiation occurs in a 1,5,7-trimethyl-3-azabicyclo-
[16]
7
8
ꢀ1
creased with
Table 3; entries 1–3, entries 4–6) but did not exceed
4% ee.
Since enantioselective [2+2] photocycloaddition reactions
a
decrease of the reaction temperature
(
6
of 3-substituted quinolones have not yet been studied, it was
desirable to prove the absolute configuration of the prod-
ucts, when produced in the presence of a chiral 1,5,7-tri-
methyl-3-azabicyclo
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
[3.3.1]nonan-2-one. To this end, the
ACHTUNGTRENNUN[G 3.3.1]nonan-2-one with benzoxazolyl substituents of any
minor enantiomer 17 of the reactions discussed above was
separated by chiral HPLC and converted into the respective
bromo compound 21 (Scheme 9) by treatment with N-bro-
mosuccinimide (NBS). The absolute configuration of this
product was unambiguously determined by anomalous X-
ray diffraction (see the Supporting Information for further
details).
kind in position 7 because the substituent shields effectively
one face of the quinolone double bond from an attack. The
fidelity of this shielding is high even at ambient temperature
as clearly seen by the enantioselectivities achieved with sub-
strates 4 (88% ee) and 7 (84% ee) at ambient temperature.
The slightly lower enantioselectivitiy in the reactions of 4-
(hex-3-enyl)oxyquinolones (6!14a) (Table 2, entries 1 and
4
) could be accounted for by their slightly lower cyclization
rate. The decreased enantioselectivity observed for the
crossed product 15 is likely due to the slower rate of six-
membered ring formation.
The approach of a tethered olefin occurs in quinolones
complexed to the enantiomer (ꢀ)-1 of sensitizer (+)-1 of
course from the opposite face explaining the formation of
ent-17 and ent-18 in significant enantiomeric excesses
(Figure 7). To explain the lower enantioselectivity achieved
Scheme 9. Preparation of the bromo compound 21, the absolute configu-
ration of which was elucidated by anomalous X-ray diffraction.
7468
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7461 – 7472

Photocycloaddition of 3- and 4-(But-3-enyl)oxyquinolones
FULL PAPER
deterioration of enantioselectivity, as observed for substrates
8
and 9.
The transient absorption spectroscopy results not only
corroborate this analysis but they also provide valuable data
for further investigations. Although the enantioselective
process within the substrate complexes depicted in Figure 7
appears understood, it is not yet clear how the energy trans-
fer within this complex is accomplished on a molecular
level. Operationally more difficult studies are required to in-
vestigate the substrate–photocatalyst complex by transient
absorption spectroscopy. The data and spectra now obtained
Figure 7. Preferred approach (!) of the tethered olefin in the complexes
of (+)-1 and (ꢀ)-1 with the respective 4-alkenyloxyquinolones 4, 6, 7
(
left) and 3-alkenyloxyquinolones 8, 9 (right).
(
Figures 4–6) present an excellent starting point for further
[30]
with substrates 8 and 9 (relative to substrates 4, 6, and 7),
the absorption properties of these compounds must be
taken into account. Both quinolones 8 and 9 show a signifi-
cantly stronger absorption at lꢁ350 nm in the UV/Vis spec-
trum, which allows for a relatively rapid background reac-
tion under the irradiation conditions. When reactions were
performed at room temperature in the absence of sensitizer
work. Dexter energy transfer is certainly conceivable and
has been previously discussed
[9,16]
but requires to be proven
in the context of further experiments. Work along these
lines is ongoing in our laboratories.
Experimental Section
(
ꢀ)-1 under conditions otherwise identical to the conditions
of Table 3, yields of 21% (for rac-17) and 17% (for rac-18)
were obtained after an irradiation time of ten minutes. In
other words, the background reactions account for an ee loss
of at least 20 percentage points.
General: All reactions sensitive to air or moisture, were carried out in
flame-dried glassware under argon pressure using standard Schlenk tech-
niques. Photochemical experiments were performed in flame-dried pyrex
irradiation tubes (diameter: 1.0 cm, volume: 15 mL for low-temperature
irradiation) under argon pressure. Irradiation experiments were per-
formed in a Rayonet RPR-100 photochemical reactor (Southern New
England Ultra Violet Company, Branford, CT, USA) equipped with fluo-
rescence lamps (Philips black light blue 8 W, lmax =366 nm; Rayonet
RPR-3000 ꢄ, lmax =300 nm). For the photochemical reactions dry a,a,a-
trifluorotoluene was used and degassed by purging with argon in an ultra-
sonicating bath for 15 min. Flash chromatography was performed on
silica gel 60 (Merck, 230–400 mesh) with the eluent mixtures given in the
corresponding procedures. Thin layer chromatography (TLC) was per-
formed on silica-coated glass plates (silca gel 60 F 254). Compounds
were detected by UV (l=254 nm, 366 nm) fluorescence. All solvents for
chromatography were distilled prior to use. Preparative HPLC separa-
From the lifetimes for the triplet intermediates T-DR
(
(
Scheme 7, Table 1) and the dissociation rate from templates
+)-1 and (ꢀ)-1 (see above) it is likely that a non-negligible
fraction of these intermediates is still bound to the sensitizer
before cyclobutane ring formation. This assumption is in
agreement with the fact that the sensitizer influences the
diastereo- and regioselectivity in the reaction of 4-(hex-3-
enyl)oxyquinolones (6) (Scheme 4 vs. Table 2). Diaster-
eoisomers (14b) and constitutional isomers (15), which lead
to an increased steric repulsion with the xanthone moiety,
are disfavored (Table 2).
ꢀ1
tions were performed on a Waters HPLC system (flow rate: 20 mLmin
achiral stationary phase: XBridge C18 column, 5 mm, 30ꢅ150 mm;
eluent: H O/MeCN=70:30, 1% trifluoroacetic acid) with a flame ioniza-
;
2
tion detector. The temperature method is given for the corresponding
compounds. Analytical HPLC was performed using a chiral stationary
Conclusion
ꢀ
1
phase (Daicel ChiralCell, Chemical Industries, flow rate: 1.0 mLmin
,
type and eluent are given for the corresponding compounds) and UV de-
tection. IR: JASCO IR-4100. MS/HRMS: Thermo Scientific DFS
In this study a key principle for the design of chiral photoca-
talysts has been confirmed. The enantioselectivity determin-
ing step of the desired reaction must occur fast enough to
compete with substrate dissociation from the photocatalyst.
For the currently used xanthone 1, this requirement asks for
CꢀC bond-forming reactions on the ns timescale. As proven
1
13
HRMS. H and C NMR: Bruker AV-360 and AV-500 recorded at 300 K.
1
Chemical shifts are reported relative to the solvent [CHCl : d( H)=
3
1
3
1
13
7.26 ppm,
5
d AHCTUNGTRENNUNG( C)=77.0 ppm, [D ]DMSO: d( H)=2.50 ppm, d ACHTUNGTRENNUNG( C)=
3
9.5 ppm] as reference. Apparent multiplets that occur as a result of acci-
dental equality of coupling constants to those of magnetically non-equiv-
alent protons are marked as virtual (virt.). The relative configuration of
by transient absorption spectroscopy, the cyclization of sub-
strates 4, 6–9 does occur within this time window guarantee-
ing excellent selectivity for the formation of products 5, 14,
and 16. If the cyclization is slower, for example, for the for-
mation of product 3 from 2 or of product 15 from 6, the
enantioselectivity decreases.
13
chiral products and the multiplicity of the C NMR signals were deter-
mined by two-dimensional NMR spectra (COSY, NOESY, HSQC,
HMBC). Specific rotations were determined by using a PerkinElmer 241
MC polarimeter (sodium vapor lamp). Melting points were measured on
a Bꢁchi 510 of the company Bꢁchi and are not corrected.
Transient absorption spectroscopy on multiple timescales: The femtosec-
ond broadband pump-probe setup has been described in detail else-
[
31]
In addition, it became again evident in this study that the
success of the enantioselective catalysis depends on an ap-
propriate choice of excitation wavelength and on the ab-
sorption properties of the substrate and the respective sub-
strate–photocatalyst complex. If excitation can occur in the
absence of the photocatalyst, background reactions lead to a
where. Briefly, a Ti/sapphire amplifier system (CPA 2001, Clark-MXR)
with a repetition rate of 1 kHz was used to pump a noncollinear optical
parametric amplifier (NOPA) which provides pulses with a center wave-
length of 540–550 nm. After compression and frequency doubling in a
BBO-crystal (100 mm thickness) we achieved pump pulses with a center
wavelength of 270–275 nm and sub-ps duration. The beam with a pulse
energy of about 400 nJ was focused with a spot size of 100–150 mm
Chem. Eur. J. 2013, 19, 7461 – 7472
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7469

T. Bach, E. Riedle, et al.
(
FWHM) into a 1 mm flow cell containing the sample solution with an
removed in vacuo and the resulting oil was purified by flash column chro-
optical density below 1 within the spectral probe range. A supercontin-
uum (290–750 nm) was generated and used as probe by focusing another
matography (50 g silica, pentane/EtOAc=2:1). The racemic compound
f
rac-18 was obtained as a colorless solid (57.7 mg, 0.24 mmol, 95%). R =
1
part of the Ti/sapphire laser into a rotating CaF
2
(4 mm thickness) disk.
3
0.62 (EtOAc [UV]); m.p.=158–1608C; H NMR (500 MHz, CDCl ,
3
4
The relative polarization of pump and probe pulses was set to the magic
angle (54.78) to avoid contributions from the rotational relaxation to the
observed kinetics. A computer controlled delay line was used to set
pump–probe delays up to 2 ns. After the interaction in the sample, the
probe light was dispersed with a fused silica prism and detected with a
303 K, TMS): d=8.15 (br s, 1H; NH), 7.15 (virt. td, Jꢁ7.8 Hz, J=
1.3 Hz, 1H; H-8), 6.98 (virt. td, Jꢁ7.6 Hz, J=1.1 Hz, 1H; H-9), 6.90
(dd, J=7.5 Hz, J=1.3 Hz, 1H; H-10), 6.74 (dd, J=7.9 Hz, J=1.1 Hz,
3
4
3
4
3
4
2
3
3
1H; H-7), 4.45 (ddd, J=9.0 Hz, J=7.7 Hz, J=3.1 Hz, 1H; H-3a), 4.15
2
3
3
(virt.td, Jꢁ J=9.3 Hz, J=6.5 Hz, 1H; H-3b), 3.29 (s, 1H; H-10b), 2.94
(virt. dt, J=8.9 Hz, Jꢁ J=1.8 Hz, 1H; H-1a), 2.27–1.90 (m, 2H; H-2),
1.16 (s, 3H; CH ), 0.81 ppm (s, 3H; CH ); C NMR (250 MHz, CDCl
03 K, TMS): d=170.4 (s, C-5), 136.5 (s, C-6a), 128.5 (d, C-10), 127.8 (d,
3
3
4
5
24 pixel CCD. The chirp of the white light was corrected for prior to
1
3
data analysis. The resulting temporal resolution was better than 100 fs.
3
3
3
,
3
For transient absorption measurements on the ns timescale we used a
pulsed ns laser system (NT242, Ekspla) with 1 kHz repetition rate and in-
tegrated OPO for spectral tunability, to pump the sample in a 1 mm flow
cell (see above). The ns laser is externally triggered and is electronically
synchronized with the Ti/sapphire amplifier system (see above), running
at 1 kHz repetition rate as well. We still use the supercontinuum, generat-
C-8), 123.5 (d, C-9), 121.6 (s, C-10a), 115.8 (d, C-7), 78.3 (s, C-4), 71.0 (t,
C-3), 56.6 (d, C-1a), 50.5 (d, C-10b), 34.9 (s, C-1), 27.8 (t, C-2), 26.4 (q,
CH ), 24.8 ppm (q, CH ); IR (ATR): n˜ =2942, 2356, 1682, 1594, 1495,
3 3
ꢀ
1
1
[
389, 1143, 1076, 1048, 973, 827 cm ; MS (EI, 70 eV): m/z (%)=243 (1)
+
+
+
M ], 212 (1), 198 (2), 174 (3) [M ꢀC
5
H
9
], 161 (100) [M ꢀC
6
H10], 143
+
+
+
(
ꢀ
5) [M ꢀC
6
H
12O], 133 (3) [M ꢀC
7
H10O], 115 (5), 104 (3) [M
2
ed by focusing the fs pulses into a CaF disk for probing the sample. The
+
+
C
8
H
11
O
2
], 89 (4) [C
7
H
5
], 83 (7), 77 (3) [C
6
H
5
], 67 (2), 55 (15);
pump–probe delay Dt is set by a computer controlled delay generator,
which introduces a specified temporal delay Dt in the electronic trigger
pulse from the master oscillator before the pulse triggers the release of
the ns excitation pulse. This setup for transient absorption spectroscopy,
combining two electronically connected laser sources, provides a tempo-
ral resolution better than 2.5 ns and allows recording transient absorption
spectra up to temporal delays in the 100 ms regime.
HRMS (EI): m/z: calcd for C15
17 2
H O N: 243.1254; found: 243.1254.
Representative procedure for the enantioselective [2+2] photocycloaddi-
tion under sensitized conditions at lꢁ366 nm: (3aS,4aS,10bS)-4,4-di-
methyl-2,3,4,4a-tetrahydro-furo-[2’,3’:2,3]-cyclobuta ACHUTNGRNEUNG[ 1,2-c]-quinolin-5-
AHCTUNGTRENNUNG
ꢀ
1
c=5 mmolL ), cooled to ꢀ258C and irradiated at l=366 nm (light
source: Philips black light blue; emission spectrum, see the Supporting
Information) for 20 min. Silica (500 mg!dry load) was added and the
solvent was evaporated under reduced pressure. Following flash column
chromatography (5 g silica, pentane/EtOAc=1:1), photoproduct 16 was
To obtain kinetics from the TA data we used a global data analysis that
[
32]
is based on a formalism developed by Fita et al. and extracts decay as-
[
31]
sociated difference spectra (DADS) from the transient data.
Before
applying the global fit routine to our TA data, the wavelength scale was
rescaled to a scale linear in DE to avoid he enhancement of the weight of
the long-wave components. The global fit analysis provides a global set
obtained as a colorless solid (11.0 mg, 45.2 mmol, 90%, 96% ee). R
f
=0.64
(
1
EtOAc [UV]); HPLC (chiral phase AD-H, n-hexane/iPrOH=90:10,
i i
of time constants t and the corresponding DADS A (l).
ꢀ
1
R
mLmin , l=210 and 254 nm): t =12.6 min [ent-16], 13.6 min [16];
1
ꢀ
ꢁ
2
0
X
m.p.=122–1268C; [a] = +11.2 (c=0.9, MeOH) [96% ee]; H NMR
(500 MHz, CDCl , 303 K, TMS): d=8.35 (brs, 1H; NH), 7.30 (d, J=
7
ꢀ
t
D
DODðl; tÞ ¼
A
i
ðlÞ exp
=
ð1Þ
3
i
t
i
3
3
3
.7 Hz, 1H; H-10), 7.21 (virt. t, Jꢁ7.7 Hz, 1H; H-8), 7.05 (virt. t,
J
3
Further data processing and the calculation of the species associated ab-
sorption spectra (SAS) were based on inserting the obtained DADS
A (l) into a sequential rate model.
i
ꢁ
7.5 Hz, 1H; H-9), 6.73 (d, J=8.0 Hz, 1H; H-7), 4.42–4.34 (m, 1H; H-
3
2
a), 4.09–3.99 (m, 1H; H-2b), 3.02 (s, 1H; H-4a), 2.57 (d, J=9.2 Hz,
1
CH
H; H-3a), 2.10–2.04 (m, 2H; H-3), 1.21 (s, 3H; CH
3
), 1.12 ppm (s, 3H;
1
3
Actinometric measurements: We used the previously mentioned ns laser
system with integrated OPO, which provides a spatially, spectrally and
energetically well-defined light beam, for photo-excitation of the sample
solution in a transparent cuvette. A very similar approach, but using
LEDs as light source for the determination of product quantum yields
3
); C NMR (126 MHz, CDCl
3
, 303 K, TMS): d=168.5 (s, C-5), 136.3
(s, C-6a), 129.5 (d, C-8), 127.7 (d, C-10), 124.2 (d, C-9), 124.0 (s, C-10a),
115.5 (d, C-7), 79.4 (s, C-10b), 68.9 (t, C-2), 58.1 (d, C-3a), 51.8 (d, C-4a),
34.8 (s, C-4), 27.8 (t, C-3), 26.1 (q, CH ), 24.8 ppm (q, CH ); IR (ATR):
3 3
n˜ =2925 (w, ArH), 2861 (w, CalH), 1662 (s, C=O), 1594 (m, NH), 1381 (s),
[
33]
ꢀ1
has been reported. The irradiation was performed in a standard 10 mm
cuvette containing about 3.8 mL sample solution, exhibiting an optical
density of about 1.2 at the absorption maximum at 317 nm while stirring
with a magnetic stirring bar. This ensures that more than 90% of the inci-
dent photons at l=317 nm are absorbed. We measured Ptrans, the light in-
tensity transmitted by the sample solution, and Pref, the light intensity
transmitted by a reference solution comprising just solvent in a similar
cuvette.
1274 (s), 1139 (s), 1063 (s), 1039 cm (vs, C-O); MS (EI, 70 eV): m/z
+
+
(%): 243 (70) [M ], 161 (100) [M ꢀC
(44); HRMS (EI): m/z: calcd for C15
Single-crystal X-ray absolute structure determination of compound 21:
Formula: C13 12BrNO ; M =294.14; crystal color and shape: pale-yellow
6
H
10], 82 (40), 67 (23), 55 (69), 44
H
17NO : 243.1254; found: 243.1254.
2
H
2
r
fragment, crystal dimensions=0.15ꢅ0.15ꢅ0.20 mm; crystal system: tri-
clinic; space group P1 (no. 1); a=6.4063(3), b=9.8661(5), c=
1
5
0.3432(5) ꢄ; a=114.362(2), b=103.133(2), g=95.359(2)8, V=
3
ꢀ1
ꢀ3
66.91(5) ꢄ ; Z=2;
000
m ACHTUNGTRENUNNG( MoKa)=3.613 mm ; 1calcd =1.723 gcm ; F =
ðPref ꢀ PtransÞDt
296; T=123 K; q range=2.3–25.48; data collected: 18004; independent
data [I >2s(I )/all data/Rint]: 3951/3951/0.048; data/restraints/parameters:
N
abs
¼
ð2Þ
E
g
o
o
3
o o o o
951/0/307; R1 [I >2s(I )/all data]: 0.0244/0.0247; wR2 [I >2s(I )/all
ꢀ
3
The decrease in absorption intensity of the starting material at 317 nm
was used for the calculation of the amount of converted starting material
within the irradiation time. In each case, we measured 3 data points for
an over-all conversion of typically below 20% of the total amount of
sample in the cuvette. The quantum yield then was the ratio of formed
data]: 0.0658/0.0663; GOF=1.098; D1max/min: 0.448/ꢀ0.329 eꢄ . Flack
parameter x=0.045(7). Hooft parameter y=0.092(6). For more details
see the Supporting Information. CCDC-919729 (21) contains the supple-
mentary crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
product molecules Nprod and absorbed photons Nabs
.
Representative procedure for the [2+2] photocycloaddition at short
wavelength (lꢁ300 nm) leading to racemic products: 1,1-dimethyl-
1
,2,3,10b-tetrahydro-1H-furo[2’,3’:1,4]cyclobuta AHCUTNGTRENNUNG[ 1,2-c]quinolin-5 ACHTUNGTRNENUNG( 6H)-one
(
rac-18): Compound 9 (61.0 mg, 0.25 mmol) was dissolved in a,a,a-tri-
ꢀ
1
fluorotoluene (100 mL, c=2.5 mmolL ). The solution was irradiated at
l=300 nm (light source: RPR-3000 ꢄ, emission spectrum, see the Sup-
porting Information) at ambient temperature for 10 min. The solvent was
7470
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7461 – 7472

Photocycloaddition of 3- and 4-(But-3-enyl)oxyquinolones
FULL PAPER
Acknowledgements
Yagishita, M. Sakamoto, T. Mino, T. Fujita, Org. Lett. 2011, 13,
6
168–6171.
This project was supported by the Deutsche Forschungsgemeinschaft
[6] Review: J. C. Namyslo, D. E. Kaufmann, Chem. Rev. 2003, 103,
1485–1537.
(
DFG) in the framework of the DFG Research Training Group “Chemi-
cal Photocatalysis” (GRK 1626). M.M.M. and M.W. acknowledge fellow-
ship support by the GRK. R.A. is the recipient of an Alexander von
Humboldt research fellowship. We thank Olaf Ackermann and Georg
Rudolf (AK Sieber) for help with the HPLC analysis.
[7] a) T. Yamamuro, I. Tanaka, N. Hata, Bull. Chem. Soc. Jpn. 1971, 44,
667–671; b) T. Yamamuro, N. Hata, I. Tanaka, Bull. Chem. Soc. Jpn.
1973, 46, 29–34; c) F. D. Lewis, G. D. Reddy, J. E. Elbert, B. E. Till-
berg, J. A. Meltzer, M. Kojima, J. Org. Chem. 1991, 56, 5311–5318.
[
[
8] H. Gçrner, T. Wolff, Photochem. Photobiol. 2008, 84, 1224–1230.
9] C. Mꢁller, M. M. Maturi, A. Bauer, M. C. Cuquerella, M. A. Miran-
da, T. Bach, J. Am. Chem. Soc. 2011, 133, 16689–16697.
[
1] General reviews on [2+2] photocycloaddition reactions: a) J. P.
Hehn, C. Mꢁller, T. Bach in Handbook of Synthetic Photochemistry
[
10] a) T. Bach, H. Bergmann, K. Harms, Angew. Chem. 2000, 112,
391–2393; Angew. Chem. Int. Ed. 2000, 39, 2302–2304; b) T. Bach,
2
(Eds.: A. Albini, M. Fagnoni), Wiley-VCH, Weinheim, 2009,
H. Bergmann, J. Am. Chem. Soc. 2000, 122, 11525–11526; c) T.
Bach, H. Bergmann, B. Grosch, K. Harms, J. Am. Chem. Soc. 2002,
pp. 171–215; b) S. A. Fleming in Molecular and Supramolecular
Photochemistry, Vol. 12 (Eds.: A. G. Griesbeck, J. Mattay), Marcel
Dekker, New York, 2005, pp. 141–160; c) P. Margaretha in Molecu-
lar and Supramolecular Photochemistry, Vol. 12 (Eds.: A. G. Gries-
beck, J. Mattay), Marcel Dekker, New York, 2005, pp. 211–237;
d) J. P. Pete in CRC Handbook of Organic Photochemistry and Pho-
tobiology, 2nd ed. (Eds.: W. Horspool, F. Lenci), CRC, Boca Raton,
1
1
24, 7982–7990; d) S. Breitenlechner, T. Bach, Angew. Chem. 2008,
20, 8075–8077; Angew. Chem. Int. Ed. 2008, 47, 7957–7959.
[
11] Reviews: a) J. Svoboda, B. Kçnig, Chem. Rev. 2006, 106, 5413–5430;
b) D. M. Bassani in Supramolecular Photochemistry (Eds.: V. Rama-
murthy, Y. Inoue), Wiley, Hoboken, 2011, pp. 53–86.
2
004, pp. 71/1–71/14; e) T. Bach, Synthesis 1998, 683–703; f) J.-P.
[12] T. Bach, H. Bergmann, B. Grosch, K. Harms, E. Herdtweck, Synthe-
sis 2001, 1395–1405.
[13] D. S. Kemp, K. S. Petrakis, J. Org. Chem. 1981, 46, 5140–5143.
[14] a) P. Selig, T. Bach, Angew. Chem. 2008, 120, 5160–5162; Angew.
Chem. Int. Ed. 2008, 47, 5082–5084; b) P. Selig, E. Herdtweck, T.
Bach, Chem. Eur. J. 2009, 15, 3509–3525.
Pete, Adv. Photochem. 1996, 21, 135–216; g) J. Mattay, R. Conrads,
R. Hoffmann in Methoden der Organischen Chemie (Houben-Weyl),
Vol. E21c, 4th ed. (Eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer,
E. Schaumann), Thieme, Stuttgart 1995, pp. 3085–3132; h) M. T.
Crimmins, T. L. Reinhold, Org. React. 1993, 44, 297–588.
[
2] a) G. R. Evanega, D. L. Fabiny, Tetrahedron Lett. 1968, 9, 2241–
[15] Reviews: a) C. Mꢁller, T. Bach, Aust. J. Chem. 2008, 61, 557–564;
b) C. Yang, Y. Inoue in Supramolecular Photochemistry (Eds.: V.
Ramamurthy, Y. Inoue), Wiley, Hoboken, 2011, pp. 115–153.
[16] C. Mꢁller, A. Bauer, T. Bach, Angew. Chem. 2009, 121, 6767–6769;
Angew. Chem. Int. Ed. 2009, 48, 6640–6642.
2246; b) G. R. Evanega, D. L. Fabiny, J. Org. Chem. 1970, 35, 1757–
1761; c) G. R. Evanega, D. L. Fabiny, Tetrahedron Lett. 1971, 12,
1749–1752.
[
[
3] a) E. C. Taylor, W. W. Paudler, Tetrahedron Lett. 1960, 1, 1–3;
b) I. W. Elliott, J. Org. Chem. 1964, 29, 305–307; c) O Buchardt,
Acta Chem. Scand. 1964, 18, 1389–1396.
[
[
17] E. Ochiai, J. Org. Chem. 1953, 18, 534–551.
18] a) O. Buchardt, Acta Chem. Scand. 1963, 17, 1461–1462; b) A.
Albini, E. Fasani, L. M. Dacrema, J. Chem. Soc. Perkin Trans. 1
4] a) B. Loev, M. M. Goodman, K. M. Snader, Tetrahedron Lett. 1968,
9
, 5401–5404; b) T. S. Cantrell, J. Org. Chem. 1974, 39, 3063–3070;
1
980, 2738–2742; c) A. Albini, M. Alpegiani, Chem. Rev. 1984, 84,
c) R. G. Hunt, C. J. Potter, S. T. Reid, M. L. Roantree, Tetrahedron
Lett. 1975, 16, 2327–2330; d) O. Buchardt, J. J. Christensen, N.
Harrit, Acta Chem. Scand. B 1976, 30, 189–192; e) D. J. Haywood,
S. T. Reid, J. Chem. Soc. Perkin Trans. 1 1977, 2457–2458; f) C.
Kaneko, T. Naito, M. Somei, J. Chem. Soc. Chem. Commun. 1979,
4
3–71.
[
[
[
[
19] A. Bakowski, M. Dressel, A. Bauer, T. Bach, Org. Biomol. Chem.
011, 9, 3516–3529.
20] F. Marsais, A. Godard, G. Queguiner, J. Heterocycl. Chem. 1989, 26,
589–1594.
2
1
804–805; g) C. Kaneko, T. Naito, Chem. Pharm. Bull. 1979, 27,
2254–2256; h) T. Naito, C. Kaneko, Chem. Pharm. Bull. 1980, 28,
3150–3152; i) C. Kaneko, T. Naito, N. Nakayama, Chem. Pharm.
21] T. Nishi, F. Tabusa, T. Tanaka, H. Ueda, T. Shimizu, T. Kanbe, Y.
Kimura, K. Nakagawa, Chem. Pharm. Bull. 1983, 31, 852–860.
22] a) D. Becker, M. Nagler, S. Hirsh, J. Ramun, J. Chem. Soc. Chem.
Commun. 1983, 371–373; b) D. Becker, M. Nagler, Y. Sahali, N.
Haddad, J. Org. Chem. 1991, 56, 4537–4543; c) D. Becker, N. KIi-
movich, Tetrahedron Lett. 1994, 35, 261–264.
23] Review on the mechanism of the [2+2] photocycloaddition, see:
D. I. Schuster, G. Lem, N. A. Kaprinidis, Chem. Rev. 1993, 93, 3–22.
24] C. Franco, J. Olmsted III, Talanta 1990, 37, 905–909.
Bull. 1981, 29, 593–595; j) T. Chiba, M. Okada, T. Kato, J. Hetero-
cycl. Chem. 1982, 19, 1521–1525; k) C. Kaneko, T. Naito, Y.
Momose, H. Fujii, N. Nakayama, I. Koizumi, Chem. Pharm. Bull.
1
982, 30, 519–525; l) T. Naito, Y. Momose, C. Kaneko, Chem.
[
Pharm. Bull. 1982, 30, 1531–1534; m) C. Kaneko, T. Naito, Y.
Momose, N. Shimomura, T. Ohashi, M. Somei, Chem. Pharm. Bull.
[
[
1
983, 31, 2168–2171; n) C. Kaneko, N. Shimomura, Y. Momose, T.
25] Based on previous work (e.g., A. A. Abdel-Shafi, F. Wilkinson, J.
Phys. Chem. A 2000, 104, 5747–5757), a bimolecular rate constant
Naito, Chem. Lett. 1983, 1239–1242; o) T. Naito, C. Kaneko, Chem.
Pharm. Bull. 1983, 31, 366–369; p) T. Chiba, T. Kato, A. Yoshida,
R. Moroi, N. Shimomura, Y. Momose, T. Naito, C. Kaneko, Chem.
Pharm. Bull. 1984, 32, 4707–4720; q) S. Nonoyama, N. Yonezawa,
K. Saigo, T. Hirano, M. Hasegawa, Chem. Lett. 1987, 487–490; r) M.
Sato, K. Kawakami, C. Kaneko, Chem. Pharm. Bull. 1987, 35, 1319–
for triplet state quenching by oxygen can be estimated as k
1
q
ꢁ3ꢅ
9
ꢀ1 ꢀ1
0 m s . The minimum lifetime of a triplet species in air-equili-
ꢀ
7
brated acetonitrile is consequently tmin ꢂ1/[k ꢅ2.4 mm]ꢁ1.3ꢅ10 s.
q
Oxygen quenching should not significantly influence the lifetime of
a species with an intrinsically shorter lifetime than tmin. Indeed, as
seen from Table 1, all intermediates relevant for the [2+2] photocy-
1
321; s) C. Kaneko, T. Suzuki, M. Sato, T. Naito, Chem. Pharm.
Bull. 1987, 35, 112–123; t) S. Nonoyama, N. Yonezawa, K. Saigo, M.
Hasegawa, T. Hirano, Bull. Chem. Soc. Jpn. 1988, 61, 2387–2391;
u) E. Sato, Y. Ikeda, Y. Kanaoka, Liebigs Ann. Chem. 1989, 781–
ꢀ
8
cloaddition processes show lifetimes below 30 ns (3ꢅ10 s).
26] S. J. Strickler, R. A. Berg, J. Chem. Phys. 1962, 37, 814–822.
27] For recent theoretical studies on intramolecular and intermolecular
[
[
7
88; v) H. Suginome, K. Kobayashi, M. Itoh, S. Seko, A. Furusaki, J.
[
2+2] photocycloaddition reactions, see: a) S. A. Bradley, B. J.
Org. Chem. 1990, 55, 4933–4943; w) J. Kurita, T. Yamanaka, T. Tsu-
chiya, Heterocycles 1991, 32, 2089–2092; x) K. Kobayashi, M.
Suzuki, H. Suginome, J. Org. Chem. 1992, 57, 599–606; y) M. Saka-
moto, N. Sato, T. Mino, Y. Kasashima, T. Fujita, Org. Biomol.
Chem. 2008, 6, 848–850; z) F. Yagishita, T. Mino, T. Fujita, M. Saka-
moto, Org. Lett. 2012, 14, 2638–2641.
Bresnan, S. M. Draper, N. W. A. Geraghty, M. Jeffares, T. McCabe,
T. B. H. McMurry, J. E. OꢃBrien, Org. Biomol. Chem. 2011, 9, 2959–
2968; b) J. R. Cucarull-Gonzꢆlez, J. Hernando, R. Alibꢇs, M. Figuer-
edo, J. Font, L. Rodrꢈguez-Santiago, M. Sodupe, J. Org. Chem. 2010,
75, 4392–4401; c) K. Somekawa, Y. Oda, T. Shimo, Bull. Chem. Soc.
Jpn. 2009, 82, 1447–1469; d) P. Jaque, A. Toro-Labbꢇ, P. Geerlings,
F. De Proft, J. Phys. Chem. A 2009, 113, 332–344.
[
5] a) D. F. Cauble, V. Lynch, M. J. Krische, J. Org. Chem. 2003, 68, 15–
2
1; b) S. Brandes, P. Selig, T. Bach, Synlett 2004, 2588–2590; c) F.
Chem. Eur. J. 2013, 19, 7461 – 7472
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7471

T. Bach, E. Riedle, et al.
[
28] a) G. Jones II, B. R. Ramachandran, J. Photochem. 1976, 5, 341–
44; b) R. Gleiter, E. Fischer, Chem. Ber. 1992, 125, 1899–1911;
c) D. Becker, Y. Cohen-Arazi, J. Am. Chem. Soc. 1996, 118, 8278–
284.
C. A. Kenesky, J. P. Toscano, Phys. Chem. Chem. Phys. 2003, 5,
1059–1063; i) X. Cai, P. Cygon, B. Goldfuss, A. G. Griesbeck, H.
Heckroth, M. Fujitsuka, T. Majima, Chem. Eur. J. 2006, 12, 4662–
4667.
3
8
[
29] a) R. D. Small, J. C. Scaiano, J. Phys. Chem. 1977, 81, 2126–2131;
b) J. C. Scaiano, Tetrahedron 1982, 38, 819–824; c) D. Becker, N.
Haddad, Y. Sahali, Tetrahedron Lett. 1989, 30, 2661–2664; d) A. Ru-
dolph, A. C. Weedon, Can. J. Chem. 1990, 68, 1590–1597; e) N. A.
Kaprinidis, G. Lem, S. H. Courtney, D. I. Schuster, J. Am. Chem.
Soc. 1993, 115, 3324–3325; f) D. Andrew, D. J. Hastings, A. C.
Weedon, J. Am. Chem. Soc. 1994, 116, 10870–10882; g) J. F. D.
Kelly, J. M. Kelly, T. B. H. McMurry, J. Chem. Soc. Perkin Trans. 2
[30] D. L. Dexter, J. Chem. Phys. 1953, 21, 836–850.
[31] U. Megerle, I. Pugliesi, C. Schriever, C. F. Sailer, E. Riedle, Appl.
Phys. B 2009, 96, 215–231.
[32] P. Fita, E. Luzina, T. Dziembowska, C. Radzewicz, A. A. Grabow-
ska, J. Chem. Phys. 2006, 125, 184508.
[33] U. Megerle, R. Lechner, B. Kçnig, E. Riedle, Photochem. Photobiol.
Sci. 2010, 9, 1400–1406.
Received: February 27, 2013
1
999, 1933–1941; h) A. D. Cohen, B. M. Showalter, D. A. Brady,
Published online: April 10, 2013
7472
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7461 – 7472