Enantioselective photocyclisation reactions of 2-aryloxycyclohex-2-enones mediated by a chiral copper-bisoxazoline complex

Article abstract of DOI:10.1016/j.tet.2017.05.005

The photocyclisation of the title compounds leads upon direct irradiation at λ = 366 nm in dichloromethane solution to racemic cis-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-ones (nine examples, 37–74% yield). Since it was found that the substrates show a sig

Full text of DOI:10.1016/j.tet.2017.05.005

Accepted Manuscript
Enantioselective photocyclisation reactions of 2-aryloxycyclohex-2-enones mediated
by a chiral copper-bisoxazoline complex
Verena Edtmüller, Alexander Pöthig, Thorsten Bach
PII:
S0040-4020(17)30472-6
10.1016/j.tet.2017.05.005
TET 28671
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 4 March 2017
Revised Date: 28 April 2017
Accepted Date: 1 May 2017
Please cite this article as: Edtmüller V, Pöthig A, Bach T, Enantioselective photocyclisation reactions of
2-aryloxycyclohex-2-enones mediated by a chiral copper-bisoxazoline complex, Tetrahedron (2017), doi:
10.1016/j.tet.2017.05.005.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Enantioselective photocyclisation reactions of 2-aryloxy-
cyclohex-2-enones mediated by a chiral copper-bisoxazoline
complex
Verena Edtmüller,a Alexander Pöthig,a Thorsten Bacha,*
a Department Chemie and Catalysis Research Center (CRC), Technische Universität München,
Lichtenbergstr. 4, 85747 Garching, Germany
Corresponding Author
Tel.: +49-89-289 13330; fax: +49-89-289 13330; e-mail: thorsten.bach@ch.tum.de
ABSTRACT
The photocyclisation of the title compounds leads upon direct irradiation at λ = 366 nm in
dichloromethane solution to racemic cis-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-ones (nine
examples, 37-74% yield). Since it was found that the substrates show a significant bathochromic
absorption shift upon treatment with EtAlCl2, it was attempted to perform the reactions
enantioselectively in the presence of a chiral Lewis acid. A complex of Cu(ClO4)2·6 H2O and a
bisoxazoline ligand gave the best enantioselectivities (up to 60% ee). Two procedures are
reported for the enantioselective photocyclisation. The first protocol is based on a direct
irradiation at λ = 368 nm (LED) with a catalyst loading of 50 mol% and it delivered the products
in 26-76% yield with 22-40% ee. The second protocol is applicable to electron rich 2-
aryloxycyclohex-2-enones (31-62% yield, 29-46% ee) and relies on sensitization by thioxanthone
(50 mol%) at λ = 419 nm.
1

ACCEPTED MANUSCRIPT
Keywords
Bisoxazoline ligand
Cu catalysis
Enantioselectivity
Photochemistry
Photocyclisation
Sensitization
Graphical Abstract
Cu(ClO4)2—6 H2O
O
O
O
N
N
O
H
O
Bn
Bn
O
r.t. (CH2Cl2)
X
X
26-76%
up to 60% ee
Conditions: h
(
= 368 nm) or h ( = 419 nm), 50 mol% thioxanthone
Introduction
In 1975, Schultz and Lucci reported on the first photocyclisation reaction of 2-aryloxycyclohex-
2-enones.1 Upon irradiation of substrate 1a for 23 hours in a solvent mixture of benzene,
methanol, and acidic acid, dihydrofuran rac-3a was formed (Scheme 1) and the compound was
isolated after re-crystallization in analytically pure form (80% yield). The photolysis was
2

ACCEPTED MANUSCRIPT
performed in pyrex glass which served as filter to cut-off short wavelength irradiation below λ =
280 nm. Still, a further conversion of the product was notable upon prolonged irradiation and
minor side products were isolated. The reaction was explored with a variety of different
substrates in the context of natural product synthesis as it promised an efficient access to the
morphine skeleton.2,3
SCHEME 1. Proposed reaction pathway for the photocyclisation of 2-phenyloxycyclohex-2-enone
1a to product rac-3a
O
O
*
h
O
O
(
280 nm)
1a
1a*
O
O
H
O
O
H
racꢀ2a
racꢀ3a
Mechanistic evidence suggested that the photocyclisation proceeds via a carbonyl ylide, e.g. rac-
2a, which is the formal product of a conrotatory ring closure. Dittami et al. trapped intermediates
of this type by an intramolecular 1,3-dipolar cycloaddition and established the relative
configuration of the products.4 To account for the relative cis-configuration of the
photocyclisation products (cf. rac-3a), it was assumed that protonation of the carbonyl ylides
occurs by the solvent and this notion was supported by deuteration experiments.2a Flash
photolysis studies by Wolff revealed that the reaction proceeds via the excited triplet state of the
aryloxyenone (1a*) but it was not possible to resolve the adiabatic reaction step from this state to
the triplet state of the carbonyl ylide.5 Additional evidence for the intermediacy of triplet
aryloxyenones was obtained from sensitization experiments and oxygen quenching studies.
3

ACCEPTED MANUSCRIPT
We became interested in the formal [6π]-photocyclisation6,7 of 2-aryloxycyclohex-2-enones8 in
the context of our work9 on Lewis-acid catalyzed enantioselective photochemical reactions.10 It
was found that Lewis acid coordination11 to cyclic alkenones leads to a bathochromic shift of the
intense (ε 15000 M−1 cm−1) ππ* absorption. Due to this shift, the strong absorption of the Lewis
acid complex overlaps the weak (ε ≤ 100 M−1 cm−1) nπ* absorption of the non-coordinated
enone, which is responsible for the [2+2] photocycloaddition chemistry of this class of
compounds. If irradiated at a suitable wavelength, only the Lewis acid complex is
photochemically excited due to its higher absorption cross-section and subsequent reactions can
proceed enantioselectively if a chiral Lewis acid is used.12 We speculated that compounds like
aryloxyenone 1a might show a similar behavior and might potentially undergo enantioselective
photocyclisation reactions in the presence of a chiral Lewis acid.13 Preliminary UV-Vis spectra
(Figure 1) suggested that the plan could be viable as compound 1a exhibited a significant
bathochromic shift upon addition of EtAlCl2 as the Lewis acid.
0,6
1 eq.
3 eq.
4 eq.
5 eq.
0,4
0,2
0,0
250
300
350
Wavelength (nm)
FIGURE 1. UV-Vis spectra of compound 1a in the presence of various amounts of EtAlCl2.
The strong absorption with maxima at λ = 223 nm (ε = 11450 M−1 cm−1) and at λ = 241 nm (ε =
10460 M−1 cm−1) vanished upon Lewis acid addition and a new band appeared with a maximum
4

ACCEPTED MANUSCRIPT
at λ = 292 nm (ε = 10840 M−1 cm−1). The strong band stretches into a wavelength region, in
which the weak nπ* absorption of the uncomplexed enone 1a occurs (λ = 320 nm, ε = 114 M−1
cm−1). Based on this result we started to search for chelating chiral Lewis acids which would
enable an enantioselective photocyclisation reaction of 2-aryloxycyclohex-2-enones. The results
of our experiments are summarized in this report.
RESULTS
Preparation of the starting materials and photocyclisation to racemic products. The synthesis of
the photochemical substrates was performed in analogy to the procedure of Schultz and co-
workers.1,2a Isophorone oxide (rac-4)14 served as the starting material and underwent a potassium
hydride-assisted ring opening/elimination sequence. A dipolar aprotic solvent additive is required
in this step and N,N-dimethylpropyleneurea (DMPU) was found to be a suitable alternative to the
toxic hexamethylphosphoramide. In addition, we found that DMSO was a superior solvent as
compared to THF, in particular if electron-deficient phenols (HOAr) were used as oxygen
nucleophiles (Table 1, entries 5-7). Yields were not further optimized as sufficient starting
material for the subsequent photochemical reaction could be readily secured.
Since we planned to perform all catalytic experiments in dichloromethane solution, the racemic
reactions were also run in this solvent. Employing fluorescent lamps,15 we found an optimum
yield at a wavelength of λ = 366 nm for the reaction 1a → rac-3a (Table 1, entry 1). At λ = 300
nm and λ = 350 nm, the yields were lower (35% and 68%, respectively) while no reaction was
observed at λ = 419 nm.16
5

ACCEPTED MANUSCRIPT
Table 1. Formation of substrates 1 by epoxide ring opening/elimination from isophorone oxide (rac-4) and
subsequent photocyclisation to racemic products rac-3
entry
Ar
1
yielda
[%]
rac-3 t [h]b yieldc
[%]
1
2
3
1a
1b
1c
72
77
83
rac-3a 24
rac-3b 24
rac-3c 22
72
54
54
4
5
6
7
8
9
1d
1e
1f
39d
80
37
61
87
80
rac-3d
9
50
44
37
40
49e
57
rac-3e 14
rac-3f
rac-3g
rac-3h
rac-3i
12
5
1g
1h
1i
4
15
a Yield of isolated product 1. b Reaction time until full conversion was reached. c Yield of isolated product rac-3a. d The reaction was performed in
THF as the solvent. e The yield was 81% at c = 20 mM.
The high yield achieved for product rac-3a could not be reproduced for all other substrates but
again it was not attempted to optimize the procedure. In general, reactions were performed at a
concentration of c = 10 mM under deaerated conditions and were stopped if no starting material
6

ACCEPTED MANUSCRIPT
could be detected by TLC. The relative configuration of the products rac-3 was assigned based
on the reported NMR shift data of known compounds of this class. The constitution of product
rac-3i (C-C bond formation at carbon atom C1 of the naphthalene) had been previously
established5 and was confirmed.
SCHEME 2. Influence of the water content of the solvent on the relative configuration of
photocyclisation products from substrate 1a
h
( = 368 nm)
O
O
H
c = 10 mM
O
O
r.t. (CH2Cl2)
+
3a
racꢀ
5a
racꢀ
1a
c (H2O) < 20 ppm
c (H2O) > 50 ppm
10%
61%
44%
—
As mentioned above, the cis configuration is assumed to be the result of the intermolecular
protonation of the intermediate carbonyl ylide. Scheme 2 illustrates the results obtained by
irradiation with a light-emitting diode (LED) at λ = 368 nm17 in dichloromethane solutions with
varying water contents. If distilled dichloromethane was used which was not further dried the
only product was cis-product rac-3a. Under strictly anhydrous conditions, trans-diastereoisomer
rac-5a was obtained as the major product that was configurationally stable upon chromatography.
Its formation can be explained by an intramolecular suprafacial 1,4-hydrogen shift to occur in
carbonyl ylide rac-2a (cf. Scheme 1). When treated with base (e.g. K2CO3 in wet CH2Cl2),
compound rac-5a was quantitatively transformed into cis-diastereoisomer rac-3a.
Search for an appropriate chiral Lewis acid. Substrate 1a exhibits two Lewis basic oxygen atoms
which could potentially form a five-membered chelate complex with an appropriate Lewis acid.
Consequently, the use of a chiral chelating Lewis acid with a C2-symmetric ligand seemed to be a
good starting point to identify suitable catalysts to promote the reaction of 1a to product 3a
enantioselectively.
7

ACCEPTED MANUSCRIPT
Table 2. Ratio of enantiomers as observed in the Cu(OTf)2-promoted photocyclisation 1a → 3a/ent-3a: Influence of
the chiral ligand
h
( = 366 nm)
O
O
50 mol% Cu(OTf)2
60 mol% ligand
r.t. (CH2Cl2)
H
H
O
O
+
1a
3a
ent-3a
eed
[%]
entrya
ligand
yieldb [%]
e.r.c
O
O
1
2
44
55/45
43/57
10
14
N
N
tBu
6a
tBu
52
47
52
3
4
60/40
56/44
20
12
5
20e
50/50
6
7
35f
44
50/50
50/50
a The reaction was run to completion (20-24 h reaction time) unless indicated otherweise. b Yield of isolated product. c The enantiomric ratio (e.r.)
was determined by chiral HPLC analysis. d The enantiomeric excess (ee) was calculated from the e.r.. e 50% conversion after 24h. f 65%
conversion after 24 h.
8

ACCEPTED MANUSCRIPT
The ligand screen was performed with Cu(OTf)218 and chiral oxazoline ligands19 in deaerated
dichloromethane solution and a few selected results are summarized in Table 2. As expected from
previous work9 the presence of a Lewis acid did not lead to a rate acceleration but rather did the
reaction rate decrease. The only meaningful enantioselectivities were recorded with bidentate
bisoxazoline (box)20 ligands 6a-6d derived from 2,2-dimethylmalonic acid (entries 1-4). An
asymmetric induction by the other ligands 6e-6g (entries 5-7) could not be detected and racemic
product rac-3a was obtained. Among the box ligands, the respective dibenzylated ligand 6c was
the superior choice although the enantioselectivity was far from optimal (20% ee, entry 3). A
more electron rich benzyl group (PBB = para-benzyloxybenzyl, ligand 6d, entry 4) delivered a
lower enantiomeric excess (ee). Although the absolute configuration at the stereogenic centers of
ligands 6b and 6c were identical the preference for one product enantiomer was opposite (entries
2 and 3, vide infra).
With box ligand 6c providing the highest enantioselectivity in the screening, this ligand was used
in combination with different metal salts (see Supplementary Material). Typical conditions (see
Table 2) included the use of 0.5 equiv. of the metal salt and 0.6 equiv. of ligand 6c in CH2Cl2
solution (c = 10 mM, λ = 366 nm, t = 24 h). Among the various metal salts, there was only a
single beneficial effect to be observed when Cu(ClO4)2·6 H2O was employed as the source of the
copper ion. Although the conversion was even slower (31% yield after 24 h) than with Cu(OTf)2,
the enantiomeric excess doubled to 40% ee. Since these reactions were performed at room
temperature it was hoped that a lower reaction temperature might lead to an improved
enantioselectivity. Disappointingly, the photocyclisation did not proceed at −65 °C and was
sluggish at 0 °C. A lower catalyst loading led to a decreased enantioselectivity if the reaction was
performed at ambient temperature. If the amount of Cu(ClO4)2·6 H2O was increased from 0.5 to
1.0 equivalents, the yield increased but the enantioselectivity decreased (53%, 36% ee). If the
9

ACCEPTED MANUSCRIPT
ligand loading was increased to 1.0 equiv. and the loading of Cu(ClO4)2·6 H2O was kept at 0.5
equiv. there was a significant increase in enantioselectivity but the yield did not improve (34%,
60% ee).
Screening with various substrates and absolute configuration. Two light sources which emit at
366 and 368 nm, respectively, were evaluated to perform the enantioselective reactions with all
available substrates (Table 3). The 366 nm light source consists of a set of 16 fluorescent lamps
with a broad emission spectrum and with a significant heat evolution.15 The 368 nm light source
is a LED with a narrow emission spectrum, which is immersed in a flask21 and which evolves no
heat. Reactions with the latter light source turned out to be more reproducible at ambient
temperature. In addition, the reaction with parent substrate 1a was faster and higher yielding with
the LED than with the fluorescent lamp. Under standard conditions (Table 2, entry 1), the yield
was 53% as compared to 31% while the enantioselectivity remained identical (40% ee). Donor
substitution in para-position of the aryl group was inconsequential to the enantioselectivity
(entries 2-4) while acceptor substitution led to a decrease in enantioselectivity (entries 5-7). The
relatively high yields in the reactions of compounds 1e, 1g, and 1h (entries 5, 7, 8) indicate that
racemic background reactions which occur upon direct excitation are significant. The
naphthyloxy-substituted substrate 1i gave product 3i in a modest yield and with low
enantioselectivity (entry 9). In general, there was no substrate 1b-1i which showed a better
performance than parent compound 1a. Yields of isolated products 3 remained on average
moderate and varied between 26-76%. The enantioselectivity was also variable and ee values
between 22% and 40% ee were recorded. All products were levorotatory indicating that their
absolute configuration was identical irrespective of the aromatic substituent.
10

ACCEPTED MANUSCRIPT
Table 3. Cu-Mediated enantioselective photocyclisation of 2-aryloxy¬cyclohex-2-enones 1 to products 3 upon direct
irradiation at λ = 368 nm
entrya
product
t [h]b r.s.m.c yieldd eee
[%]
[%]
[%]
1
20
20
24
24
17
24
23
23
24
53
40
2
3
4
5
6
7
8
9
45
32
26
76
51
52
65
36
39
39
39
22
27
30
26
28
17
17
O
H
O
OMe
3d
O
H
O
23
COOMe
3f
O
H
O
3i
a All reactions were performed on a scale of 0.1 mmol (c= 10 mM) with a LED lamp (3 W power output)17 as the light source. b Irradiation time. c
Yield of recovered starting material. d Yield of isolated product. e The enantiomeric excess (ee) was determined by chiral HPLC analysis.
11

ACCEPTED MANUSCRIPT
The absolute configuration of the major enantiomer in the reaction 1h → 3h could be elucidated
by anomalous X-ray diffraction (Figure 2). In order to obtain a configurationally homogenous
sample the ee of the compound was enriched by chiral semipreparative HPLC to >99%. The
identity of the enantiomer, of which the crystal structure was determined, was confirmed by
subsequent HPLC analysis.
FIGURE 2. Absolute configuration of product 3h as determined by anomalous X-ray diffraction.
As in the solid-state reaction of 2-arylthiocyclohex-2-enones,8 the enantioface differentiation in
the current photocyclisation reaction is likely due to a helical conformation. In the present case
the helicity must be induced by the chiral Cu complex. Attack at the β-carbon atom of the enone
occurs from the respective Re face as depicted for substrate 1a in Figure 3.
FIGURE 3. Twisted conformation 1a’ of substrate 1a, possible coordination of 1a to the Cu
bisoxazoline complex Cu(ClO4)2·6c, and structure of Cu(H2O)2(SbF6)2·6a
12

ACCEPTED MANUSCRIPT
Coordination of compound 1a to the CuII bisoxazoline complex is expected to occur in a more or
less square-planar fashion with the two Lewis-basic oxygen atoms of the substrate binding to the
central metal atom. However, it is known that the oxygen atoms and the nitrogen atoms of the
bisoxazoline ligand are not located in a single plane but that the square planar arrangement is
somewhat distorted.22 X-Ray crystallographic data for the complex Cu(H2O)2(SbF6)2·6a for
example revealed that there is a positive dihedral angle between the marked atoms O1-Cu-N1-C1
of ca. +30 °.22a A similar dihedral angle is observed for O2-Cu-N2-C2 leading to a twist in the
coordination of the water atoms with one water molecule positioned below but the other
positioned above the plane. If one assumes that the oxygen atoms of compound 1a follow the
same binding pattern as the oxygen atoms of the water molecules, it can be readily explained why
conformation 1a’ is preferred and why enantiomer 3a is the major product of the
photocyclisation. Moreover, the hypothesis also explains the reversal of enantioselectivity with
ligand 6b. In the Cu(H2O)2(SbF6)2·6b the above-mentioned dihedral angles are negative21,22b
inducing a twist of substrate 1a in the opposite direction with the Si face now being more readily
accessible. Another aspect deserves to be mentioned. If one assumes an initial coordination as
shown in Figure 3, the ether oxygen atom of the substrate will progressively decomplex from the
copper center when approaching the transition state to the ylide intermediate 2a (Scheme 1).
Among other factors (vide supra), the insufficient chelation may be a reason for the only
moderate enantioselectivity of the photocyclisation.
Reaction mechanism and visible-light induced photocyclisation in the presence of a sensitizer.
Although it was established earlier that the photocyclisation of substrates 1 proceeds via a triplet
intermediate,5 it was not clear whether the Cu-catalyzed reaction was also a triplet process.
Preliminary experiments with O2 as putative quencher of a triplet intermediate revealed that the
13

ACCEPTED MANUSCRIPT
photocyclisation of 1a was indeed slower than under exclusion of oxygen (see Supplementary
Data) but the rate decrease was less significant than observed for the non-catalyzed reaction.5
When irradiating a substrate with options for competing pathways, i.e. photocyclisation vs. [2+2]
photocycloaddition, it had been previously found4c that the photocycloaddition which is a fast
triplet process23 prevails. In the present study, 2-phenyloxyenones 7 were compared under the
conditions of the Cu-catalyzed process and it was found that substrate 7a (R = Me) expectedly
yields the photocyclisation product 8 (Scheme 3). The enantioselectivity determination suffered
from insufficient baseline separation but the determined ee was in the range which was
previously observed for products 3. Product 8 was also levorotatory. Substrate 7b (R = pent-4-
enyl) gave upon irradiation under the Cu-catalyzed conditions almost exclusively the [2+2]
photocycloaddition product rac-9. The same observation has been previously made upon direct
excitation of 7b.4c The results illustrate that the [2+2] photocycloaddition is significantly faster
than the photocyclisation and it adds another piece of evidence that also the Cu-promoted
reaction proceeds via the aryloxyenone triplet state.
SCHEME 3. Photocyclisation vs. [2+2] photocycloaddition in the reaction of substrates 7
In a final set of experiments it was probed whether the excitation of enones 1 could be achieved
by sensitization. Thioxanthone (TXT) seemed a suitable sensitizer which would allow the
14

ACCEPTED MANUSCRIPT
reaction to be performed with visible light. Gratifyingly, it was found that the addition of 50
mol% of the sensitizer enabled a conversion of several substrates upon irradiation at λ = 419 nm
for 24 hours (Table 4). Complete reactions were found for substrates 1a, 1d, 1h, and 1i (39-62%
yield, entries 1, 4, 8, and 9). Aryloxyenones 1e-1g with an electron deficient aryl group showed a
very slow conversion (entries 5-7). The reaction did not proceed or remained incomplete after 24
h. Presumably, the triplet energy of the sensitizer24 is too low to promote these substrates into the
excited state. Notable enantioselectivites (46% and 47% ee) were observed in two cases (entries 3
and 6).
Table 4. Cu-Mediated enantioselective photocyclisation of 2-aryloxy¬cyclohex-2-enones 1 to products 3 upon
sensitized excitation at λ = 419 nm
entrya
product
r.s.m.b
[%]
yieldc eed [%]
[%]
1
2
3
4
5
6
7
8
9
3a
3b
3c
3d
3e
3f
57
31
32
62
11
17
30
35
46
29
27
47
19
32
23
66
75
e
3g
3h
3i
39
57
30
9
a All reactions were performed on a scale of 0.1 mmol (c= 10 mM) with a set of fluorescence lamps (RPR-4190 Å)15d as the light source. The
reaction time was in all cases 24 hours. b Yield of recovered starting material. c Yield of isolated product. d The enantiomeric excess (ee) was
determined by chiral HPLC analysis. e No reaction product could be isolated.
15

ACCEPTED MANUSCRIPT
CONCLUSION
In summary, we have discovered the first enantioselective photocyclisation reactions of 2-
aryloxycyclohex-2-enones in solution and we have proven that a chiral Lewis acid approach is
applicable to this reaction class. A complex of Cu(ClO4)2·6 H2O and bisoxazoline ligand 6c was
employed as the Lewis acid in most of the reported transformations. Evidence was collected that
the Cu-promoted reactions follow – like the uncatalyzed reactions – a triplet mechanism. In the
dynamic catalyst-substrate system there is an equilibrium between the non-complexed substrate
and the substrate in the complex. For the non-complexed substrate excitation occurs at long
wavelength via its weak nπ* absorption, e.g. for 1a at λ = 320 nm (ε = 114 M−1 cm−1). The long-
wavelength absorption of the Lewis acid complex has ππ* character. Upon direct excitation, the
fact that population of the ππ* triplet state can only occur via the ππ* singlet state hampers the
catalysis because intersystem crossing (ISC) is likely slow.25 This issue was already discussed in
the context of enantioselective Lewis-acid promoted [2+2] photocycloaddition reactions.26 The
uncatalyzed reaction can proceed via rapid ISC from the nπ* singlet state and thus acts as a
significant racemic background reaction.27 Upon sensitization, the chosen triplet sensitizer
apparently does not allow for a perfect discrimination between complexed and non-complexed
substrate. As a result, the racemic background reaction remains viable and dilutes the asymmetric
induction of the chiral Lewis acid. Based on this analysis, it should be possible to further improve
the enantioselectivity by judicious choice of the sensitizer. Promising results along these lines
have been recently achieved by the Yoon group in the context of [2+2] photocycloaddition
reactions.12b
16

ACCEPTED MANUSCRIPT
EXPERIMENTAL SECTION
General Methods:
All reactions sensitive to air or moisture were carried out in flame-dried glassware under a
positive pressure of argon using standard Schlenk techniques. Dry tetrahydrofuran (THF) and
dichloromethane (CH2Cl2) were obtained from an MBRAUN MB-SPS 800 solvent purification
system. Other dry solvents and chemicals were obtained from commercial suppliers in the highest
purity available and were used without further purification. Technical solvents used for aqueous
workup and for column chromatography [n-pentane (pentane), ethyl acetate (EtOAc), diethyl
ether (Et2O), dichloromethane (CH2Cl2), methanol (MeOH)] were distilled prior to use. The
following compounds were prepared according to published procedures: rac-4,14 6a,22a 6b,28 6c,29
6e.30 Photochemical experiments at λ = 366 nm and λ = 419 nm were performed in Duran tubes
(volume 10 mL) in an RPR-100 photochemical reactor (Southern New England Ultra Violet
Company, Branford, CT, USA) equipped with fluorescence lamps (λ = 366 nm, λ = 419 nm).15
Photochemical experiments using a LED (λ = 368 nm)17 were carried out in a Schlenk tube
(diameter = 1 cm) with a polished quartz rod as an optical light guide, which was roughened by
sandblasting at one end.21 The roughed end has to be completely submerged in the solvent during
the reaction, in order to guarantee optimal and reproducible irradiation conditions. Prior to
irradiation, the dichloromethane was deoxygenated by purging with argon in an ultrasonicating
bath for 15 minutes. Column chromatography was performed on silica gel 60 (Merck, 230-240
mesh) with the eluent mixtures given for the corresponding procedures. Thin-layer
Chromatography (TLC) was performed on silica-coated glass plates (silica gel 60 F 254).
Compounds were detected by UV (λ = 254 nm, 366 nm), KMnO4 and CAM solution (cerium
ammonium molybdate). Analytical HPLC was performed using a chiral stationary phase (Daicel
ChiralCell, Chemical Industries, flow rate: 1.0 mL/min, type and eluent is given for the
17

ACCEPTED MANUSCRIPT
corresponding compounds) and UV detection (λ = 210 nm or 254 nm) at 20 °C. IR spectra were
recorded on a JASCO IR-4100 (ATR) or a Perkin Elmer Frontier IR-FTR spectrometer by ATR
technique. The signal intensity is assigned using the following abbreviations: s (strong), m
(medium), w (weak). MS and HRMS measurements were performed on a Thermo Scientific DFS
instrument (EI) or a Thermo Scientific LTQ-FT Ultra (ESI). 1H and 13C spectra were recorded at
300 K either on a Bruker AV-360, a Bruker AVHD-400, or a Bruker AVHD-500 spectrometer.
Chemical shifts are reported as parts per million (ppm) relative to chloroform [δ (1H) = 7.26 ppm,
δ (13C) = 77.16 ppm]. All coupling constants (J) are reported in Hertz (Hz). The relative
configuration of chiral products and the multiplicity of the 13C-NMR signals were determined by
two-dimensional NMR experiments (COSY, NOESY, HSQC, HMBC). X-ray crystallography
was performed on a Bruker D8 Venture Duo IMS system equipped with a Helios optic
monochromator and a Mo IMS microsource (λ = 0.71073 Å). The data was analyzed using a
Bruker SAINT software package using a narrow-frame algorithm. UV/Vis spectra were recorded
on a Perkin Elmer Lambda 35 UV/Vis spektometer using a Hellma precision cell made of quartz
SUPRASIL® with a pathway of 1 mm. Optical rotations were determined using a
Bellingham+Stanley ADP440+ polarimeter.
General procedure for the synthesis of the irradiation precursor. To a solution of the
appropriate phenol (1.0 equiv.) in dry DMSO (0.3 mL/mmol), KH in mineral oil (30%,
0.1 equiv.) was added and the mixture was stirred at room temperature for 10 min. After addition
of isophorone oxide14 (rac-4, 1.05 equiv.), DMPU (0.82 equiv.) was added. The reaction mixture
was stirred at 100 °C for 24 h. After cooling to room temperature, the solution was extracted
three times with Et2O (5 mL/mmol). The combined organic layers were washed with brine (10
18

ACCEPTED MANUSCRIPT
mL/mmol), dried over Na2SO4 and filtered. After evaporation the crude material was purified by
column chromatography.
2-(4-tert-Butylphenoxy)-3,5,5-trimethyl-2-cyclohexen-1-one (1c). According to the general
procedure, compound 1c was synthesized starting from 4-tert-butylphenol (925 mg, 6.16 mmol,
1.0 equiv.). Purification by column chromatography (pentane/Et2O 10:1, UV, CAM) gave the
product as a light yellow solid (1.46 g, 5.10 mmol, 83%). m.p.: 98 °C. TLC: Rf = 0.54
(pentane/Et2O 2:1) [UV, CAM]. IR (ATR): ν̃
(cm-1) = 2954 (m, C-H), 1676 (C=O), 1508 (s),
1230 (s), 1181 (s), 835 (s), 827 (s). 1H NMR (400 MHz, CDCl3): δ (ppm) = 1.14 [s, 6H,
C(CH3)2], 1.28 [s, 9H, C(CH3)3], 1.89 (s, 3H, CH3), 2.41 (s, 2H, H-6), 2.42 (s, 2H, H-4), 6.74-
6.77 (m, 2H, Har), 7.24-7.27 (m, 2H, Har). 13C {1H} NMR (101 MHz, CDCl3): δ (ppm) = 18.3 (q,
CH3), 28.6 [q, C(CH3)2], 31.7 [q, C(CH3)3], 33.4 (s, C-5), 34.2 [s, C(CH3)3], 45.8 (t, C-4), 52.1 (t,
C-6), 114.3 (d, Car), 126.5 (d, Car), 143.9 (s, CtBu), 144.4 (s, C=C), 145.9 (s, C=C), 155.5 (s, Car),
193.2 (s, C-1). MS (EI, 70 EV): m/z (%) = 150 (16), 135 (100), 107 (28). HRMS (EI, 70 eV):
Calculated for C19H26O2 [M+] = 286.1927. Found = 286.1925.
2-(4-Cyanophenoxy)-3,5,5-trimethyl-2-cyclohexen-1-one (1g). According to the general
procedure, compound 1g was synthesized starting from 4-cyanophenol (734 mg, 6.16 mmol,
1.0 equiv.) Purification by column chromatography (pentane/Et2O 3:2, UV, CAM) gave the
product as a light yellow solid (965 mg, 3.78 mmol, 61%). m.p.: 62 °C. TLC: Rf = 0.60
(pentane/Et2O 1:2) [UV, CAM]. IR (ATR): ν̃
(cm-1) = 3270 (m), 2957 (w, C-H), 2231 (s, CN),
1674 (s, C=O), 1602 (s, C=C), 1505 (s), 1239 (s), 1166 (s), 836 (s, C-H). 1H NMR (360 MHz,
CDCl3): δ (ppm) = 1.15 [s, 6H, C(CH3)2], 1.89 (s, 3H, CH3), 2.42 (s, 2H, H-6), 2.45 (s, 2H, H-4),
6.88 (d, J = 8.1 Hz, 2H, Har), 7.56 (d, J = 8.1 Hz, 2H, Har).13C {1H} NMR (90 MHz, CDCl3): δ
19

ACCEPTED MANUSCRIPT
(ppm) = 18.2 [q, C(CH3)2], 28.6 (q, CH3), 33.4 (s, C-5), 45.7 (t, C-4), 51.8 (t, C-6), 105.4 (s, Car),
115.9 (d, Car), 119.1 (s, CN), 134.3 (d, Car), 143.1 (s, C=C), 146.7 (s, C=C), 161.1 (s, Car), 192.1
(s, C-1). MS (EI, 70 eV): m/z (%) = 256 (100) [M+], 240 (13) [(M-CH3)+], 227 (28), 199 (35),
143 (24), 130 (36), 109 (35), 69 (84). HRMS (EI): Calculated for C16H17NO2 [M+] = 255.1258.
Found = 255.1254.
2-(4-Bromophenoxy)-3,5,5-trimethyl-2-cyclohexen-1-one (1h). According to the general
procedure, compound 1h was synthesized starting from 4-bromophenol (533 mg, 3.08 mmol, 1.0
equiv.). Purification by column chromatography (pentane/Et2O 10:1, UV, CAM) gave the
product as a yellow solid (832 mg, 2.69 mmol, 87%). m.p.: 46 °C. TLC: Rf = 0.60 (pentane/Et2O
2:1) [UV, CAM]. IR (ATR): ν̃
(cm-1) = 2957 (m, C-H), 1668 (s, C=C), 1479 (s), 1227 (s), 823 (s).
1H NMR (400 MHz, CDCl3): δ (ppm) = 1.14 [s, 6H, C(CH3)2], 1.88 (s, 3H, CH3), 2.40 (s, 2H, H-
6), 2.43 (s, 2H, H-4), 6.67-6.73 (m, 2H, Har), 7.31-7.35 (m,2H, Har). 13C {1H} NMR (101 MHz,
CDCl3): δ (ppm) = 18.2 (q, CH3), 28.6 [q, C(CH3)2], 33.4 (s, C-5), 45.8 (t, C-4), 51.9 (t, C-6),
114.1 (s, C-10), 116.8 (d, C-8), 132.5 (d, C-9), 143.6 (s, C=C), 146.3 (s, C=C), 156.9 (s, C-7),
192.7 (s, C-1). MS (EI, 70 eV): m/z (%) = 308 (46) [M+], 280 (8), 183 (15), 172 (100), 93 (31),
79
62 (20). HRMS (EI, 70 eV): Calculated for C15H17O2 Br [M+] = 308.0414. Found = 308.0406.
81
Calculated for C15H17O2 Br [M+] = 310.0386. Found = 310.0393.
2-(2-Naphthoxy)-3,5,5-trimethyl-2-cyclohexen-1-one (1i). According to the general procedure,
compound 1i was synthesized starting from 2-naphthol (444 mg, 3.08 mmol, 1.0 equiv.).
Purification by column chromatography (pentane/Et2O 10:1, UV, CAM) gave the product as a
light yellow solid (695 mg, 2.48 mmol, 80%). m.p.: 87 °C. TLC (pentane/Et2O 2:1): Rf = 0.40
[UV, CAM]. IR (ATR): ν̃
(cm-1) = 2956 (m, C-H), 1678 (s, C=O), 1629 (C=C), 1249 (s), 1179
20

ACCEPTED MANUSCRIPT
(s), 807 (s). 1H NMR (400 MHz, CDCl3): δ (ppm) = 1.19 [s, 6H, C(CH3)2], 1.92 (s, 3H, CH3),
2.47 (s, 2H, H-6), 2.49 (s, 2H, H-4), 6.99 (d, J = 2.5 Hz, 1H, Har), 7.22 (dd, J = 2.5 Hz, J = 9.0
Hz, 1H, Har), 7.30-7.35 (m, 1H, Har), 7.38-7.43 (m, 1H, Har), 7.63-7.65 (m, 1H, Har), 7.75-7.78
(m, 2H, Har). 13C {1H} NMR (101 MHz, CDCl3): δ (ppm) = 18.3 (q, CH3), 28.6 [q, C(CH3)2],
33.5 (s, C-5), 45.9 (t, C-4), 52.1 (t, C-6), 108.8 (d, Car), 117.8 (d, Car), 124.1 (d, Car), 126.5 (d,
Car), 127.0 (d, Car), 127.8 (d, Car), 129.7 (d, Car), 129.9 (s, Car), 134.4 (s, Car), 143.9 (s, Car), 146.2
(s, C=C), 155.7 (s, C=C), 192.9 (s, C-1). MS (EI, 70 EV): m/z (%) = 280 (6) [M+], 237 (1)
[(C16H13O2)+], 205 (12), 144 (35) [(C10H7O)+], 115 (50), 82 (13). HRMS (EI, 70 eV): Calculated
for C19H20O2 [M+] = 280.1458. Found = 280.1453.
General procedure for racemic photoreactions. The irradiation precursor 1 (0.10 mmol) was
transferred into a Duran tube and was dissolved in 10 mL deaerated dichloromethane. The
mixture was irradiated at room temperature at λ = 366 nm until no starting material was detected
by TLC. The solvent was removed under reduced pressure and the crude product was purified by
column chromatography.
General procedure for enantioselective photoreactions at λ= 368 nm. Cu(ClO4)2·6H2O (18.5
mg, 50.0 ꢀmol, 0.50 equiv.) and bisoxazoline ligand 6c (21.8 mg, 60.0 ꢀmol, 0.60 equiv.) were
dissolved in 2 mL deaerated dichloromethane and stirred at room temperature for three hours.
The irradiation precursor 1 (0.10 mmol, 1.00 equiv.) was dissolved in 5 mL deaerated
dichloromethane and transferred by syringe into a Schlenk tube. The catalyst solution was
transferred by syringe into the same Schlenk tube and the residual catalyst was washed with 2 ×
1.5 mL of deaerated dichloromethane into the tube. The reaction mixture was irradiated at room
temperature for the indicated period of time. The reaction mixture was diluted with 10 mL
21

ACCEPTED MANUSCRIPT
dichloromethane and washed with 20 mL ethylenediaminetetraacetic acid (EDTA) solution. The
aqueous phase was extracted with dichloromethane (3 × 15 mL), dried over Na2SO4, filtered and
the solvent was removed under reduced pressure. The crude product was purified by column
chromatography to afford the corresponding photocyclisation product. If starting material was
recovered the yield of recovered starting material (r.s.m.) is provided.
General procedure for sensitized enantioselective photoreactions at λ = 419 nm.
Cu(ClO4)2·6H2O (18.5 mg, 50.0 ꢀmol, 0.5 equiv.) and the bisoxazoline ligand 6c (21.8 mg, 6.00
ꢀmol, 0.6 equiv.) were dissolved in 2 mL deaerated dichloromethane and stirred at room
temperature for three hours. The irradiation precursor 1 (0.10 mmol, 1.0 equiv.) and thioxanthone
(10.6 mg, 50.0 ꢀmol, 0.5 equiv.) was dissolved in 5 mL deaerated dichloromethane and
transferred by syringe into a Duran tube. The catalyst solution was transferred by syringe into the
same Duran tube and the residual catalyst was washed with 2 × 1.5 mL of deaerated
dichloromethane into the tube. The reaction mixture was irradiated at room temperature for 24
hours. The reaction mixture was diluted with 10 mL dichloromethane and washed with 20 mL
EDTA solution. The aqueous phase was extracted with dichloromethane (3 × 15 mL), dried over
Na2SO4, filtered and the solvent was removed under reduced pressure. The crude product was
purified by column chromatography to afford the corresponding photocyclisation product. If
starting material was recovered the yield of recovered starting material (r.s.m.) is provided.
2,2,9b-Trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3a). According to the
general procedure, compound 1a (23.0 mg, 100 ꢀmol) was irradiated (irradiation time: 22 h).
Purification by column chromatography (pentane/Et2O 8:1, CAM) gave the product as a colorless
22

ACCEPTED MANUSCRIPT
oil (16.6 mg, 72 ꢀmol, 72%). TLC: Rf = 0.37 (P/Et2O 2:1) [CAM]. The spectroscopic data
matched the literature values.2a
The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1a (irradiation time
20 h) and gave product 3a (12.2 mg) in 53% yield with 40% ee. [α]D20 = −66.7 (c = 0.51, CH2Cl2)
[40% ee]. Chiral HPLC (OJ-RH, 150 × 4.6 mm, MeCN (A)/H2O = 20% (A) → 100% (A), 1
mL/min, λ = 210 nm, 254 nm): tR [racemate] = 13.8 min (3a), 16.3 min (ent-3a). The
enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1a and gave product 3a
(13.2 mg) in 57% yield with 30% ee.
2,3,4a,9b-Tetrahydro-2,2,8,9b-tetramethyl-1H-dibenzofuran-4-one (rac-3b). According to
the general procedure, compound 1b (24.4 mg, 100 ꢀmol) was irradiated (irradiation time: 24 h).
Purification by column chromatography (pentane/Et2O 8:1, CAM) gave the product as a colorless
oil (13.2 mg, 54 ꢀmol, 54%). TLC: Rf = 0.33 (P/Et2O 2:1) [CAM]. The spectroscopic data
matched the literature values.2a
The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1b (irradiation time:
20 h) and gave product 3b (11.0 mg) in 45% yield with 39% ee. [α]D20 = −35.4 (c = 0.57,
CH2Cl2) [39% ee]. Chiral HPLC (OJ-RH, 150 × 4.6 mm, MeCN (A)/H2O = 20% (A) → 100%
(A), 1 mL/min, λ = 210 nm, 254 nm): tR [racemate] = 14.0 min (3b), 15.1 min (ent-3b). The
enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1b and gave product 3b
(7.5 mg) in 31% yield with 35% ee, 19% r.s.m.
8-tert-Butyl-2,2,9b-trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3c).
According to the general procedure, compound 1c (28.6 mg 100 ꢀmol) was irradiated (irradiation
time: 22 h). Purification by column chromatography (pentane/Et2O 8:1, CAM) gave the product
23

ACCEPTED MANUSCRIPT
as an orange oil (15.5 mg, 54 ꢀmol, 54%). TLC: Rf = 0.33 (P/Et2O 2:1) [CAM]. IR (ATR): ν
1) = 2956 (m, C-H), 1727 (C=O), 1485 (s), 1026 (s), 817 (s). 1H NMR (400 MHz, CDCl3): δ
̃
(cm-
(ppm) = 0.56 [s, 3H, C-2(CH3)], 1.11 [s, 3H, C-2(CH3)], 1.27 [s, 9H, C(CH3)3], 1.40 [s, 3H, C-
9b(CH3)], 1.94 (d, J = 14.6 Hz, 1H, CHH-1), 2.20 (dd, J = 12.8 Hz, J = 2.3 Hz, 1H, CHH-3), 2.26
(dd, J = 14.6 Hz, J = 2.3 Hz, 1H, CHH-1), 2.36 (d, J = 12.8 Hz, 1H, CHH-3), 4.50 (s, 1H, H-4a),
6.86 (d, J = 8.4 Hz, 1H, Har), 7.03 (d, J = 2.1 Hz, 1H, Har), 7.14 (dd, J = 8.4 Hz, J = 2.1 Hz, 1H,
Har). 13C {1H} NMR (101 MHz, CDCl3): δ (ppm) = 27.2 [q, C-2(CH3)], 31.8 [q, C(CH3)3], 32.2
[q, C-9b(CH3)], 32.5 [q, C-2(CH3)], 34.5 [s, C(CH3)3], 36.1 [s, C-2], 46.1 (t, C-1), 49.6 (s, C-9b),
51.9 (t, C-3), 91.1 (d, C-4a), 109.8 (d, Car), 118.8 (d, Car), 125.1 (d, Car), 133.8 (s, Car), 144.6 (s,
Car), 155.7 (s, Car), 208.7 (s, C-4). MS (EI, 70 eV): m/z (%) = 286 (23) [M+], 271 (100)
[(C18H23O2)+], 173 (24), 153 (53), 57 (48). HRMS (EI, 70 eV): Calculated for C19H26O2 [M+] =
286.1927. Found = 286.1923.
The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1c (irradiation time:
24 h) and gave product 3c (9.1 mg) in 32% yield with 39% ee, 17% r.s.m. [α]D20 = −7.4 (c = 0.55,
CH2Cl2) [46% ee]. Chiral HPLC (AS-RH, 150 × 4.6 mm, MeCN (A)/H2O = 20% (A) → 100%
(A), 1 mL/min, λ = 210 nm, 254 nm): tR [racemate] = 16.9 min (ent-3c), 18.2 min (3c).
The enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1c and gave product
3c (9.2 mg) in 32% yield with 46% ee, 32% r.s.m.
8-Methoxy-2,2,9b-trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3d).
According to the general procedure, compound 1d (26.0 mg, 100 ꢀmol) was irradiated
(irradiation time: 9 h). Purification by column chromatography (pentane/Et2O 8:1, CAM) gave
the product as a yellow solid (13.0 mg, 50 ꢀmol, 50%). TLC: Rf = 0.21 (P/Et2O 2:1) [CAM]. The
spectroscopic data matched the literature values.2a
24

ACCEPTED MANUSCRIPT
The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1d (irradiation time:
24 h) and gave product 3d (6.0 mg) in 26% yield with 34% ee, 17% r.s.m. [α]D20 = −19.0 (c =
0.53, CH2Cl2) [34% ee]. Chiral HPLC (AS-RH, 150 × 4.6 mm, MeCN (A)/H2O = 20% (A) →
100% (A), 1 mL/min, λ = 210 nm, 254 nm): tR [racemate] = 12.7 min (ent-3d), 14.3 min (3d).
The enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1d and gave product
3d (16.0 mg) in 62% yield with 29% ee.
8-Acetyl-2,2,9b-trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3e). According
to the general procedure, compound 1e (27.2 mg, 100 ꢀmol) was irradiated (irradiation time: 14
h). Purification by column chromatography (pentane/Et2O 2:1, CAM) gave the product as a
colorless oil (12.0 mg, 44 ꢀmol, 44%). TLC: Rf = 0.34 (P/Et2O 1:2) [CAM]. The spectroscopic
data matched the literature values.2a
The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1e (irradiation time:
17 h) and gave product 3e (20.7 mg) in 76% yield with 22% ee. [α]D20 = −20.7 (c = 0.87, CH2Cl2)
[27% ee]. Chiral HPLC (AD-H, 250 × 4.6 mm, n-heptane/i-PrOH = 90:10, 1 mL/min, λ = 210
nm, 254 nm): tR [racemate] = 11.9 min (ent-3e), 14.3 min (3e). The enantioselective reaction at
λ = 419 nm was performed with 100 ꢀmol 1e and gave product 3e (3.0 mg) in 11% yield with
27% ee, 23% r.s.m.
8-Carbomethoxy-2,2,9b-trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3f).
According to the general procedure, compound 1f (28.8 mg, 100 ꢀmol) was irradiated (irradiation
time: 12 h). Purification by column chromatography (pentane/Et2O 2:1, CAM) gave the product
as a colorless oil (10.7 mg, 37 ꢀmol, 37%). TLC: Rf = 0.31 (P/Et2O 1:2) [CAM]. The
spectroscopic data matched the literature values.2a
25

ACCEPTED MANUSCRIPT
The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1f (irradiation time:
24 h) and gave product 3f (14.6 mg) in 51% yield with 27% ee, 23% r.s.m. [α]D20 = −31.0 (c =
0.65, CH2Cl2) [27% ee]. Chiral HPLC (AD-H, 250 × 4.6 mm, n-heptane/i-PrOH = 90:10, 1
mL/min, λ = 210 nm, 254 nm): tR [racemate] = 9.0 min (ent-3f), 10.2 min (3f). The
enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1f and gave product 3f
(5.0 mg) in 17% yield with 47% ee, 66% r.s.m.
8-Cyano-2,2,9b-trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3g). According
to the general procedure, compound 1g (25.5 mg, 100 ꢀmol) was irradiated (irradiation time: 5
h). Purification by column chromatography (pentane/Et2O 3:2, CAM) gave the product as a white
solid (10.3 mg, 40 ꢀmol, 40%). m.p.: 121 °C. TLC: Rf = 0.33 (P/Et2O 1:2) [CAM]. 1H NMR (400
MHz, CDCl3): δ (ppm) = 0.57 [s, 3H, C-2(CH3)], 1.13 [s, 3H, C-2(CH3)], 1.41 [s, 3H C-
9b(CH3)], 2.00 (d, J = 14.9 Hz, 1H, CHH-1), 2.22-2.26 (m, 2H, CHH-1, CHH-3), 2.40 (d,
J = 13.1 Hz, 1H, CHH-3), 4.65 (s, 1H, H-4a), 7.02 (d, J = 8.2 Hz, 1H, Har), 7.31 (s, 1H, Har), 7.48
(d, J = 8.2 Hz, 1H, Har). 13C {1H} NMR (101 MHz, CDCl3): δ (ppm) = 27.0 [q, C-2(CH3)], 32.4
[q, C-2(CH3)], 32.6 [q, C-9b(CH3)], 36.1 (s, C-2), 45.9 (t, C-1), 49.1 (s, C-9b), 51.8 (t, C-3), 91.5
(d, C-4a), 105.1 (s, Car), 111.8 (d, Car), 119.3 (s, CN), 126.1 (d, Car), 134.0 (d, Car), 136.3 (s, Car),
161.4 (s, Car), 206.2 (s, C-4). MS (EI, 70 eV): m/z (%) = 255 (61) [M+], 240 (34) [(C15H14NO2)+],
198 (46), 156 (100), 83 (78). HRMS (EI, 70 eV): Calculated for (C16H17NO2) [M+] = 255.1254.
Found = 255.1272. The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1g
(irradiation time: 23 h) and gave product 3g (13.3 mg) in 52% yield with 30% ee. [α]D20 = −25.9
(c = 0.54, CH2Cl2) [30% ee]. Chiral HPLC (AD-H, 250 × 4.6 mm, n-heptane/i-PrOH = 90:10, 1
mL/min, λ = 210 nm, 254 nm): tR [racemate] = 13.7 min (ent-3g), 16.7 min (3g). The
enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1g and gave no product.
26

ACCEPTED MANUSCRIPT
8-Bromo-2,2,9b-trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-3h). According
to the general procedure, compound 1h (30.9 mg, 100 ꢀmol) was irradiated (irradiation time: 4
h). Purification by column chromatography (pentane/Et2O 8:1, CAM) gave the product as a
colorless solid (15.0 mg, 49 ꢀmol, 49%). m.p.: 86 °C. TLC: Rf = 0.33 (P/Et2O 2:1) [CAM]. IR
(ATR): ν̃
(cm-1) = 2958 (m, C-H), 1722 (s, C=O), 1456 (s), 1180 (s), 1017 (s), 809 (s), 641 (s). 1H
NMR (400 MHz, CDCl3): δ (ppm) = 0.60 [s, 3H, C-2(CH3)], 1.12 [s, 3H, C-2(CH3)], 1.40 (s, 3H,
C-9b(CH3)], 1.94 (d, J = 14.8 Hz, 1H, CHH-1), 2.20-2.24 (m, 2H, CHH-1, CHH-3), 2.37 (d,
J = 12.7 Hz, 1H, CHH-3), 4.54 (s, 1H, H-4a), 6.83 (d, J = 8.5 Hz, 1H, Har), 7.13 (d, J = 2.1 Hz,
1H, Har), 7.24 (dd, J = 8.5 Hz, J = 2.1 Hz, 1H, Har), 13C {1H} NMR (101 MHz, CDCl3): δ (ppm)
= 27.1 [q, C-2(CH3)], 32.4 [q, C-2(CH3)], 32.4 [q, C-9b(CH3)], 36.1 (s, C-2), 46.0 (t, C-1), 49.6
(s, C-9b), 51.8 (t, C-3), 91.1 (d, C-4a), 112.4 (d, Car), 113.5 (s, Car), 125.1 (d, Car), 131.4 (d, Car),
136.9 (s, Car), 157.0 (s, Car), 207.2 (s, C-4). MS (EI, 70 eV): m/z (%) = 308 (61) [M+], 293 (31)
[(C14H14BrO2)+], 251 (22), 210 (96), 83 (100). HRMS (EI, 70 eV): Calculated for (C15H1779BrO2)
[M+] = 308.0406. Found = 308.0404. Calculated for (C15H1781BrO2) [M+] = 310.0386. Found =
310.0383. The enantioselective reaction at λ = 368 nm was performed with 100 ꢀmol 1h
(irradiation time: 23 h) and gave product 3h (20.1 mg) in 65% yield with 26% ee. [α]D20 = −15.7
(c = 0.51, CH2Cl2) [30% ee]. Chiral HPLC (AD-H, 250 × 4.6 mm, n-heptane/i-PrOH = 90:10, 1
mL/min, λ = 210 nm, 254 nm): tR [racemate] = 10.3 min (ent-3h), 13.2 min (3h). The
enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol 1h and gave product 3h
(12.0 mg) in 39% yield with 30% ee.
10,10,11a-Trimethyl-9,10,11,11a-tetrahydronaphtho[2,1-b]benzofuran-8(7aH)-one (rac-3i).
According to the general procedure, compound 1i (28.0 mg, 100 ꢀmol) was irradiated (irradiation
27

ACCEPTED MANUSCRIPT
time: 15 h). Purification by column chromatography (pentane/Et2O 8:1, CAM) gave the product
as an orange solid (16.0 mg, 57 ꢀmol, 57%). m.p.: 119-121 °C. TLC: Rf = 0.41 (P/Et2O 2:1)
[CAM]. IR (ATR): ν̃
(cm-1) = 2924 (m, C-H), 1715 (s, C=O), 1263 (m), 1028 (s), 804 (s), 744
(m). 1H NMR (400 MHz, CDCl3): δ (ppm) = 0.64 [s, 3 H, C-10(CH3)], 1.17 [s, 3H, C-10(CH3)],
1.67 [s, 3H, C-11a(CH3)], 2.15 (d, J = 14.8 Hz, 1H, CHH-9), 2.31 (dd, J = 13.6 Hz, J = 1.8 Hz
1H, CHH-11), 2.39 (d, J = 13.6 Hz, 1H, CHH-11), 2.78 (dd, J = 14.8 Hz, J = 1.8 Hz, 1H, CHH-
9), 4.59 (s, 1H, H-7a), 7.23 (d, J = 8.8 Hz, 1H, Har), 7.31 (ddd, J = 8.1, J = 6.9, J = 1.1 Hz, 1H,
Har), 7.47 (ddd, J = 8.4, J = 6.9, J = 1.4 Hz, 1H, Har), 7.69 (d, J = 8.8 Hz, 1H, Har), 7.80-7.82 (m,
1H, Har), 7.92-7.95 (m, 1H, Har). 13C {1H} NMR (101 MHz, CDCl3): δ (ppm) = 27.0 [q, C-
10(CH3)], 30.4 [q, C-11a(CH3)], 32.1 [q, C-10(CH3)], 35.6 (s, C-10), 46.6 (t, C-11), 51.3 (s, C-
11a), 51.5 (t, C-9), 91.2 (d, C-7a), 113.0 (d, Car), 121.6 (d, Car), 123.1 (d, Car), 124.6 (s, Car),
126.8 (d, Car), 129.8 (d, Car), 130.1 (s, Car), 130.3 (d, Car), 130.6 (s, Car), 156.0 (s, Car), 208.2 (s,
C-8). MS (EI, 70 eV): m/z (%) = 280 (39) [M+], 265 (36) [(C18H17O2)+], 237 (3) [(C16H13O2)+],
220 (25), 205 (100), 182 (60), 153 (25), 83 (25). HRMS (EI, 70 eV): Calculated for (C19H20O2)
[M+] = 280.1458. Found = 280.1456. The enantioselective reaction at λ = 368 nm was performed
with 100 ꢀmol 1i (irradiation time: 24 h) and gave product 3i (10.0 mg) in 36% yield with 28%
ee. [α]D20 = −43.1 (c = 0.51, CH2Cl2) [9% ee]. Chiral HPLC (OJ-RH, 150 × 4.6 mm, MeCN
(A)/H2O = 20% (A) → 100% (A), 1 mL/min, λ = 210 nm, 254 nm): tR [racemate] = 16.0 min
(3i), 16.5 min (ent-3i). The enantioselective reaction at λ = 419 nm was performed with 100 ꢀmol
1i and gave product 3i (16.0 mg) in 57% yield with 9% ee.
trans-2,2,9b-Trimethyl-2,3,4a,9b-tetrahydro-1H-dibenzofuran-4-one (rac-5a). According to
the general procedure, compound 1a (23.0 mg, 100 ꢀmol) was irradiated (irradiation time: 22 h)
in dry CH2Cl2. Purification by column chromatography (pentane/Et2O 8:1, CAM) gave cis-
28

ACCEPTED MANUSCRIPT
product rac-3a as a colorless oil (2.3 mg, 10 ꢀmol, 10%) and trans-product rac-5a as a colorless
oil (10.1 mg, 44ꢀmol, 44%). TLC: Rf = 0.61 (P/Et2O 2:1) [CAM]. IR (ATR): ν̃
(cm-1) = 2928 (m,
C-H), 1738 (s, C=O), 1459 (s), 1203 (s), 1044 (s), 748 (s). 1H NMR (500 MHz, CDCl3): δ (ppm)
= 1.14 (s, 3H, C-9b(CH3), 1.27 [s, 3H, C-2(CH3)], 1.28 [s, 3H, C-2(CH3)], 2.21-2.27 (m, 2H, H-
3), 2.46 (dd, J = 18.1 Hz, J = 1.0 Hz, 1H, CHH-1), 2.55 (d, J = 18.1 Hz, 1H, CHH-1), 4.94 (s,
1H, H-4a), 6.94-6.97 (m, 2H, Har), 7.08-7.10 (m, 1H, Har), 7.14-7.18 (m, 2H, Har). 13C {1H}
NMR (126 MHz, CDCl3): δ (ppm) = 23.0 (q, CH3), 34.6 [q, C-2(CH3)], 34.7 (s, C-2), 36.3 [q, C-
2(CH3)], 46.1 (t, C-3), 48.4 (s, C-9b), 53.0 (t, C-1), 92.1 (d, C-4a), 111.8 (d, Car), 121.7 (d, Car),
122.3 (d, Car), 128.5 (d, Car), 138.0 (s, Car), 158.4 (s, Car), 203.4 (s, C-4). MS (EI, 70 eV): m/z
(%) = 230 (43) [M+], 215 (66) [(C14H15O2)+], 173 (42) [(C11H19O2)+], 145 (40), 131 (99), 83
(100). HRMS (EI, 70 eV): Calculated for (C15H18O2) [M+] = 230.1301. Found: 230.1296.
2,2'-(Propane-2,2-diyl)bis[4-(4-(benzyloxy)benzyl)-4,5-dihydrooxazole] (6d). To a solution of
2,2-dimethylmalononitrile (128 mg, 1.36 mmol, 1.00 equiv.) in dry toluene was added Zn(OTf)2
(989 mg, 2.72 mmol, 2.00 equiv.). The solution was stirred for 5 min at room temperature and
(S)-2-amino-3-(4-benzyloxyphenyl)-1-propanol31 (700 mg, 2.72 mmol, 2.00 equiv.) was added.
The solution was heated under reflux for 72 h. After cooling to room temperature, the solution
was washed with saturated NaHCO3-solution and brine, dried over Na2SO4 and filtered. After
evaporation of the solvent, the crude material was purified by column chromatography (EtOAc).
The product was obtained as a white solid (366 mg, 0.64 mmol, 47%). m.p.: 67 °C. TLC:
Rf = 0.18 (EtOAc) [KMnO4]. IR (ATR): ν̃
(cm-1) = 2894 (w, C-H), 1650 (s, C=N), 1510 (s), 1240
(s), 730 (s). 1H NMR (400 MHz, CDCl3): δ (ppm) = 1.45 [s, 6H, (CH3)2], 2.61 (dd, J = 13.8 Hz,
J = 8.4 Hz, 2H, 2 × CHHPh), 3.01 (dd, J = 13.8 Hz, J = 4.7 Hz, 2H, 2 × CHHPh), 3.99 (dd,
J = 8.4 Hz, J = 6.9 Hz, 2H, 2 × OCHH), 4.16 (dd, J = 9.2 Hz, J = 8.4 Hz, 2H, 2 × OCHH), 4.32-
29