Photoreactions with a Twist: Atropisomerism-Driven Divergent Reactivity of Enones with UV and Visible Light

Article abstract of DOI:10.1002/chem.201601509

Light-induced transformation of atropisomeric and achiral enones displays divergent reactivity. Photocyclization leading to 3,4-dihydroquinolin-2-one was observed with achiral enone carboxamide, whereas the atropisomeric enone carboxamides underwent hydro

Full text of DOI:10.1002/chem.201601509

DOI: 10.1002/chem.201601509
Full Paper
&
Enone Photoreactivity
Photoreactions with a Twist: Atropisomerism-Driven Divergent
Reactivity of Enones with UV and Visible Light
Nandini Vallavoju,[a] Avadakkam Sreenithya,[b] Anoklase J.-L. Ayitou,[a] Steffen Jockusch,[c]
Raghavan B. Sunoj,*[b] and Jayaraman Sivaguru*[a]
Abstract: Light-induced transformation of atropisomeric and
achiral enones displays divergent reactivity. Photocyclization
leading to 3,4-dihydroquinolin-2-one was observed with
achiral enone carboxamide, whereas the atropisomeric
enone carboxamides underwent hydrogen abstraction lead-
ing to spiro-b-lactams. Detailed photochemical, photophysi-
cal, and theoretical investigations have provided insight into
this divergent reactivity and selectivity.
Introduction
photochemical transformation, but also allow us to investigate
the photochemical reactivities of the atropisomeric substrates
1a–i and to compare these with that of the corresponding
achiral substrate 1j (Scheme 1).
Controlling the course of photochemical reactions leading to
differential reactivity is often achieved by altering the reaction
medium.[1] Some of the classical examples include the use of
constrained media, such as crystals, cyclodextrins, dendrimers,
cucurbiturils, or zeolites, which alter the excited-state reactivity,
leading to products that are not typically observed in
solution.[1a,2] We were interested in exploring substrates that
might display contrasting photochemical reactivity due to
atropisomerism in solution.[3] Such divergent photoreactive
pathways dictated by atropisomeric chromophores might
enable us to understand the intricacies involved in chiral in-
duction/transfer during such photochemical reactions.[4] In this
report, we detail the contrasting photochemical reactivity be-
tween atropisomeric enones 1a–i and the corresponding achi-
ral enone 1j,[5,6] highlighting the role of restricted bond rota-
tion(s) in dictating the reaction pathway.
Results and Discussion
We synthesized atropisomeric enone carboxamides 1a–i to
evaluate the influence of axial chirality in determining the
stereoselectivity in forming the photoproduct during photo-
chemical transformation. In addition, to elucidate the influence
of an ortho-tert-butyl group on the chemical reactivity, we syn-
thesized the corresponding achiral analogue 1j. The individual
atropisomers (P and M isomers) of 1a–i were easily separable
on a chiral stationary phase and were characterized by NMR
spectroscopy, optical rotation, HRMS, and single-crystal XRD.[8]
As the optically pure atropisomeric enones were stable at
room temperature, racemization kinetic measurements
(Table 1) were performed at elevated temperatures (458C and
758C) in different solvents (acetonitrile or toluene) to deter-
mine the racemization rate constant (krac), the half-life of race-
mization (t1/2-rac), and the activation energy of racemization
(DG°rac). Inspection of Table 1 reveals a high barrier for CÀNaryl
One of the characteristic features of a,b-enones is their abili-
ty to undergo photochemical transformations from both the
np* and pp* excited states.[7] This provided us with an oppor-
tunity to explore the photochemical reactions of enone car-
boxamides and to fine-tune the photoreactivity based on sub-
stitution. In addition, we envisaged that ortho-substitution on
the N-phenyl ring would not only provide us with an atropiso-
meric platform to study axial to point chiral transfer during
bond rotation, which is reflected in the high values of t1/2-rac
.
For example, in the case of 1a in MeCN at 458C (Table 1,
entry 1), the t1/2-rac value was 16 days.
The high racemization barriers for atropisomeric enones 1a–
i provided us with the ideal platform to evaluate them for
atropselective phototransformation in solution under ambient
conditions. Irradiation of enone carboxamides 1a–j (Scheme 1)
were performed under various conditions (Table 2), and the
photoproducts were characterized by NMR spectroscopy and
single-crystal XRD.[8] In the case of achiral enone carboxamide
1j, we observed a cyclized photoproduct 4. On the other
hand, atropisomeric enone carboxamides 1a–h underwent
photochemical hydrogen abstraction leading to the corre-
sponding spiro-b-lactam photoproduct 2. This provided us
with an opportunity to investigate UV- and/or visible light-
[a] N. Vallavoju, A. J.-L. Ayitou, Prof. Dr. J. Sivaguru
Department of Chemistry and Biochemistry
North Dakota State University
1231 Albrecht Blvd., Fargo, ND, 58108-6050 (USA)
E-mail: jayaraman.sivaguru@ndsu.edu
[b] A. Sreenithya, Prof. Dr. R. B. Sunoj
Department of Chemistry
Indian Institute of Technology Bombay, Mumbai 400076 (India)
[c] Dr. S. Jockusch
Department of Chemistry
Columbia University, New York, NY, 10027 (USA)
Supporting information for this article and ORCID number(s) for the au-
thor(s) are available under http://dx.doi.org/10.1002/chem.201601509.
Chem. Eur. J. 2016, 22, 11339 – 11348
11339
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. Dictating photochemical reactivity through atropisomerism of enone carboxamides.
Table 1. Racemization of atropisomeric enones under different
conditions.[a]
Table 3. Photochemical reactivity of 1a in various solvents.[a]
Entry
Solvent
Direct irradiation
Sensitized irradiation
Entry Cmpd. Solvent/T[8C] Racemization parameters
% Conv. (Yield [%])
% Conv. (Yield [%])
krac [dÀ1
]
DG° [kcalmolÀ1
]
t1/2-rac [d]
rac
1
2
3
4
5
6
7
8
MeCN
MeOH
EtOAc
THF
toluene
MCH
77 (58)
63 (40)
[b]
[b]
1
2
3
4
5
6
7
1a
MeCN/45
toluene/75
toluene/75
MeCN/45
toluene/75
MeCN/45
toluene/75
0.044
0.50
0.36
0.026
0.84
0.047
1.3
27.8
28.7
29.1
28.1
28.5
27.8
28.1
16
–
–
1.2
2.0
27
0.8
15
54 (41)
12 (10)
[b]
[b]
1b
1e
–
–
57 (19)
45 (16)
53 (27)
18 (9)
18 (4)
8 (5)
36 (26)
25 (11)
1 f
benzene
CH2Cl2
0.5
[a] The reported values are subject to an error of Æ3%. Optically pure
isomers were employed for racemization kinetic measurements in a given
solvent at a set temperature. Racemization was monitored by HPLC anal-
ysis on a chiral stationary phase.
[a] Irradiations performed using Rayonet reactor at ca. 350 nm (direct irra-
diation) and ca. 420 nm (thioxanthone-sensitized irradiation). The report-
ed values are subject to an error of Æ5% and are an average of a mini-
mum of three runs. [1a]ꢀ2.3 mm. Conversion (% conv.) and % yield were
ascertained by 1H NMR spectroscopy using triphenylmethane as an
internal standard. [b] Product decomposition was observed.
Table 2. Direct and sensitized irradiations of 1a in acetonitrile.[a]
Entry
Irradiation conditions
2/% Conversion (Yield [%])
system to optimize our reaction conditions to maximize con-
version, yield, and mass balance (Tables 2 and 3). Irradiation of
enone 1a (Scheme 1) was carried out under both direct irradia-
tion (ca. 300 or 350 nm) and sensitized irradiation (using visible
light with thioxanthone acting as a triplet sensitizer/catalyst).
After the photoreaction, the reaction mixture was concentrat-
ed under reduced pressure and the photoproduct(s) was puri-
fied by column chromatography and characterized by NMR
spectroscopy, optical rotation, and single-crystal XRD (Tables 2
and 3).[8] The conversion, yield, and mass balance were calcu-
lated by 1H NMR spectroscopy using triphenylmethane as as
internal standard.[8]
1
2
3
4
5
6
7
8
9
bb/Pyrex cutoff, 3 h
bb/Pyrex cutoff, 6 h
ca. 300 nm, N2, 3 h
ca. 350 nm, N2, 3 h
ca. 350 nm, N2, 20 mol% BP, 3 h
ca. 350 nm, N2, 8 h
ca. 420 nm, 20 mol% TX, N2, 3 h
ca. 420 nm, 20 mol% TX, N2, 10 h
bb/Pyrex cutoff, À208C, 3 h[b]
62 (37)
89 (35)
77 (36)
26 (26)
61 (43)
77 (57)
20 (17)
63 (48)
3 (2)
[a] Irradiations performed using a Rayonet reactor at room temperature
unless otherwise specified. bb/Pyrex cutoff=broadband irradiation per-
formed using a 450 W Hg lamp with a Pyrex cutoff filter. The reported
values are subject to an error of Æ5% and are an average of a minimum
of three runs. Conversion (% convn.) and % yield were ascertained by
1H NMR spectroscopy using triphenylmethane as an internal standard.
[1a] ꢀ 2.3 mm. BP=benzophenone; TX=thioxanthone. [b] Low temper-
ature maintained using a cryocool.
Direct irradiation of 1a in MeCN showed moderate to high
conversions (Table 2, entries 1–4 and 7). Decreasing the temp-
erature to À208C gave very low conversion (Table 2, entry 9).
Moderate conversions were observed in sensitized irradiations.
However, we observed an increase in conversion when em-
ploying benzophenone (ET ꢀ69 kcalmolÀ1; Table 2, entry 5) as
the triplet sensitizer. The reaction proceeded efficiently under
visible-light-sensitized irradiation with thioxanthone as a sensi-
tizer (ET ꢀ63 kcalmolÀ1; Table 2, entries 7 and 8). To study the
induced atropselective hydrogen abstraction of enones 1a–h
leading to spiro-lactams 2 and 3 (Scheme 1).
To appreciate the dichotomy in the reactivity between achi-
ral and atropisomeric enones 1, we selected 1a as a model
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11340
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
effect of solvent on the photoreaction, we employed different
solvents for direct (ca. 350 nm) and triplet-sensitized visible
light (ca. 420 nm) irradiations (Table 3). Optimal conversions
and yields were observed, except in methanol and tetrahydro-
furan (THF), in which decomposition was observed. Based on
the results, acetonitrile was chosen as the solvent of choice for
evaluating the atropselective photoreaction of 1 (Table 4).[8]
Irradiation of optically pure atropisomeric 1a–h resulted in
spiro-b-lactam photoproducts 2 and 3. The diastereomeric
ratio (d.r.) in the photoproduct, that is, the diastereoselectivity
Table 4. Photochemical reactivities of enones 1a–j.[a]
Entry Substrate
Major
Isolated yield [%]/
Selectivity (2)
photoproduct
45 (2a)
2a/3a 95:5
1
98% ee (2a)
98% ee (3S,4R-2a)
1
between 2 and 3, was determined by H NMR spectroscopy.[8]
In all of the investigated substrates, the spiro-b-lactam product
was observed with diastereomeric ratio 2:3=95:5. The enan-
tiomeric excess (ee) in both 2 and 3 was determined by HPLC
separation on a chiral stationary phase (Table 4).[8] Based on lit-
erature precedence, a CÀNaryl bond with an o-tert-butyl sub-
stituent can rotate freely in solution due to the reduced CÀNÀ
C bond angle in the b-lactam ring.[9] Single-crystal X-ray analy-
sis of 2a and 2b indeed revealed a reduced CÀNÀC bond
angle in the spiro-b-lactam.[8] Thus, facile CÀNaryl bond rotation
in the spiro-b-lactam could be expected at room temperature,
and this was confirmed by opposite optical rotation values in
the enantiomeric photoproducts. The enantioselectivities of all
of the investigated atropisomeric enone carboxamides were >
95% (Table 4). This can be attributed to the high energy barrier
to rotation about the C-Naryl chiral axis (Table 1), which pre-
vents racemization during the photoreaction, thereby allowing
transfer of axial chirality to point chirality in the photo-
product(s) with high enantiomeric excess.
40 (2b)
2b/3b 95:5
2
97% ee (3R,4S-2b)
97% ee (ent-2b)
3
4
no reaction
Inspection of Table 4 reveals a clear difference in the ob-
served products from photoreactions of atropisomeric enones
1a–h and the achiral enone 1j. In the case of 1j, the expected
6p-photocyclization leading to the cyclized photoproduct 4
(isolated yields cis-4: 50% and trans-4=25%; Table 4, entry 10)
was observed.[6] On the other hand, atropisomeric enone car-
boxamide 1 underwent photochemical hydrogen abstraction
leading to the corresponding spiro-b-lactam photoproduct 2.
To further evaluate this unusual hydrogen abstraction process,
we systematically varied the N-benzyl substituent with various
electron-withdrawing and electron-donating groups (Table 4,
entries 1, 4–7) to understand their influence in changing the
reactivity and selectivity. We replaced the benzyl group with
a methyl group (Table 4, entry 3) as well as with different aryl
units, namely an electron-deficient pyridyl unit (Table 4,
entry 8) and an electron-rich furan unit (Table 4, entry 9) to
assess whether the hydrogen abstraction process would occur
efficiently, as these substituents will influence the CÀH bond
dissociation energies in the NÀCH2 functionality. Conversions
and isolated yields were dependent on the N-CH2-aryl/N-alkyl
groups. With a simple N-benzyl substituent (Table 4, entry 1) as
a reference, the conversions and isolated yields were found to
increase with electron-withdrawing groups (EWG) (Table 4, en-
tries 4 and 5), whereas they decreased with electron-donating
groups (EDG; Table 4, entries 6 and 7). Electron-rich furan
substitution (as in 1i) did not result in any product formation
(Table 4, entry 9). The electron-deficient pyridyl derivative 1h
gave low yields due to the products being unstable under our
52
50
5
>98% ee (2e)
>98% ee (ent-2e)
31
6
7
>95% ee (2 f)
>98% ee (ent-2 f)
20
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11341
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
For photophysical investigations, we employed atropisomer-
ic enone 1b, the unreactive atropisomeric N-methyl enone 1c,
and the achiral enone 1j as exemplars to understand the diff-
erences in their excited-state reactivities. Neither the achiral
nor the atropisomeric substrates showed any noticeable room
temperature fluorescence in methylcyclohexane (MCH) or etha-
nol (EtOH). On the other hand, at 77 K in EtOH glass, there was
observable phosphorescence (Figure 1A, B) with triplet ener-
gies of 62 and 61 kcalmolÀ1 for 1b and 1j, respectively. We be-
lieve that, upon photoexcitation, enone carboxamide 1 under-
goes an efficient intersystem crossing to the triplet-excited
manifold, which is reflected in a lack of fluorescence in these
systems. Lifetime measurements of the phosphorescence
signal revealed a bi-exponential decay with lifetimes of around
0.15 ms and 0.1 s for both 1b and 1j at 77 K in EtOH glass, in-
dicating the presence of both np* and pp* triplet excited
states (Figure 1).[7] The above study implies that both atropiso-
meric and achiral enones react from the triplet excited state.
Having ascertained the involvement of a triplet excited
state, we turned to triplet-triplet absorption studies to under-
stand the difference in the reactivity mode between the atropi-
someric enone 1b, the non-reactive atropisomeric enone 1c,
and the achiral enone 1j. Pulsed laser excitation (lex =308 nm,
pulse width 15 ns) of an argon-saturated solution of atropiso-
meric substrates 1b and 1c in acetonitrile generated transient
absorption spectra (Figure 1C) centered around 330 nm, with
lifetimes of 1.7 ms and 1.9 ms, respectively (Figure 1C, inset).
The transients were efficiently quenched by oxygen (e.g., for
1c with a bimolecular quenching rate constant of 1.8 (Æ0.1)”
Table 4. (Continued)
Entry Substrate
Major
photoproduct
Isolated yield [%]/
Selectivity (2)
8
15[b]
9
no reaction
50 (cis-4)
25 (trans-4)
10
[a] All irradiations were performed in acetonitrile as solvent at room
temperature at ca. 350 nm (direct irradiation) using a Rayonet reactor.
The reported values are subject to an error of Æ3%. The ee values were
obtained from HPLC analysis on a chiral stationary phase and are an aver-
age of a minimum of three runs. Sign of optical rotation in CHCl3. Abso-
lute configurations of 1b and 2b were determined by single-crystal XRD
using the Flack parameter. Diastereomeric ratios were determined by
1H NMR spectroscopic analysis of the crude samples. [b] The isolated
yield was low as we observed decomposition during purification.
conditions. Changing the N-benzyl substituent in 1a to an N-
methyl substituent (as in 1c) resulted in no observable product
formation.
109 mÀ1 sÀ1 [8]
and were assigned to triplet states. On the other
,
hand, pulsed laser excitation (lex =308 nm) of an argon-satu-
rated solution of the achiral enone 1j in acetonitrile generated
Based on our observed reactivity of enone carboxamides 1,
three important questions need to be addressed: a) What is
the nature of the reactive excited state? b) Why is there a diff-
erence in reactivity between achiral enone carboxamide 1j
and atropisomeric enone carboxamides 1a–i? c) What is the
origin of the stereoselectivity with complete enantio- and dia-
stereo-control in the bicyclo-b-lactam photoproduct? To
address these questions, we carried out detailed photophysical
investigations and computational analyses on our substrates.
a
transient absorption spectrum (Figure 2A) centered at
around 380 nm with a lifetime of about 15 ns (Figure 2B),
which was also assigned to the triplet (31j*) state. This transient
gave rise to a new transient centered at around 330 nm as
well as a broad absorbance at 750 nm (Figure 2C). This new
transient showed a lifetime of about 6 ms (Figure 2D, E) and
was not quenched by molecular oxygen. Based on literature
precedence and the characteristic features of this transient
Figure 1. Phosphorescence decay kinetics of atropisomeric enone 1b (A) and achiral enone 1j (B) at 77 K in EtOH glass. The respective phosphorescence
spectra are given in the insets. C) Transient absorption spectra after pulsed laser excitation (lex =308 nm, 15 ns pulse width) of argon-saturated solutions of
atropisomeric enones 1b and 1c in acetonitrile. Absorbance kinetic traces monitored at 330 nm (inset).
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11342
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. Transient absorption spectra monitored at 0–40 ns (A) and 0.1–3.1 ms (C) after pulsed laser excitation (lex =308 nm, 15 ns pulse width) of an argon-
saturated solution of 1j in acetonitrile. Absorbance kinetic traces monitored at different wavelengths and time scales are shown in (B), (D), and (E).
(Figure 2 and Figure S3),[8] we assign it to a zwitterionic (ZI-1j)
species involved in the reaction pathway.[8,10]
To address the dichotomy of differential reactivity between
the achiral and atropisomeric enones, we examined the single-
1
As the reaction was efficient with thioxanthone (TX) acting
as a visible-light photocatalyst (sensitizer), we investigated its
role in promoting the reaction through photophysical meas-
urements. Laser excitation (lex =355 nm; 7 ns pulse width) of
TX in the presence of varying concentrations of atropisomeric
enones 1b and 1c as well as achiral enone 1j in argon-saturat-
ed acetonitrile solution gave the bimolecular quenching rate
constant.[8] Very high quenching rate constants close to the di-
ffusion-limited values were observed for all three substrates.
The triplet excited states of TX were efficiently quenched by
1b, 1c, and 1j with bimolecular quenching rate constants (kq)
crystal XRD structures of the reactants as well as their H NMR
spectra in CD3CN.[8] Based on analysis of the single-crystal XRD
1
data and H NMR spectra of the enones, we surmised that the
origin of the difference in reactivity likely arises from the differ-
ent conformations in solution and the divergent reactivities of
their excited states. From inspection of the crystal structure of
1j, it is clear that the distance from the a-enone-carbon to the
N-phenyl ortho-carbon is about 3.617 Š (Scheme 2, right). This
distance is optimal for photocyclization.[11] We believe that the
cyclization occurs from the triplet excited state of 1j
(Scheme 2, right).[6d] The involvement of the triplet excited
state was hypothesized based on photophysical evidence (Fig-
ures 1 and 2). We were also successful in obtaining single crys-
of 1.13 (Æ0.05)”109 mÀ1
s
À1, 1.49 (Æ0.05)”109 mÀ1 sÀ1, and 1.32
(Æ0.05)”109 mÀ1
s
À1, respectively (Figure S4).[8]
Scheme 2. Understanding the difference in reactivity between atropisomeric enones 1a–h and achiral enone 1j.
Chem. Eur. J. 2016, 22, 11339 – 11348 www.chemeurj.org ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11343

Full Paper
tals of atropisomeric enones and their photoproducts. For
some of the spiro-b-lactam photoproducts 2, the absolute con-
figuration was determined by single-crystal XRD using the
Flack parameter.[8] Analysis of the single-crystal X-ray structure
of 1b revealed that the N-aryl substituent was perpendicular
to the plane formed by the CH2-N-C=O substituents, likely re-
flecting a minimized steric interaction between the nitrogen
substituents and the o-tert-butyl group (Scheme 2). Additional-
ly, the distance from the a-enone-carbon to the N-phenyl
ortho-carbon bearing the tert-butyl group was evaluated as
about 5.047 Š, and the distance from the a-enone-carbon to
the N-phenyl ortho-carbon without a tert-butyl group was eval-
uated as about 3.965 Š. A similar scenario was observed for
the other derivatives.[8] Although the XRD distances in 1b re-
vealed that cyclization at the N-phenyl ortho-carbon without
a tert-butyl group is quite feasible, the developing 1,3-strain
(A-strain) between the N-(CH2-aryl) substituent and the ortho-
tert-butyl group in the photoproduct will hinder formation of
the 6p-photocyclization product (Scheme 2, left). Whereas our
hypothesis on the developing 1,3-strain likely accounts for the
divergent reactivity between atropisomeric and achiral enones,
the lack of reactivity of the methyl derivative 1c and the furan
derivative 1i is likely due to the difference in bond dissociation
energies of the methyl hydrogen atoms in the N-CH2 substitu-
ent.
cal species cyclizes via a low-energy singlet transition state
1[TS2-1a]‡ to the photoproduct 2a.
Next, we examined the 6p-photocyclization of achiral car-
3
boxamide 1j from its triplet excited state 1j* (Figure 4). The
relative energies of different intermediates and transition
states involved in this pathway were computed with respect to
31j*. As one of the possible pathways, we examined the feasi-
bility of abstraction of the benzylic hydrogen leading to a trip-
let 1,4-diradical TDR1-1j via the transition state 3[TS1-1j]‡.
However, this process is uphill by 14.9 kcalmolÀ1. On the other
hand, two distinct modes of cyclization were noted as likely on
the basis of the computed energetics. For instance, the ring
3
closure through transition state [TS2-1j]‡ is an uphill process
cis
by 15.8 kcalmolÀ1, leading to a 1,4-diradical TDR2-1jcis, in
which the hydrogen atoms at the newly formed ring junction
are in a cis orientation (Figure 4, top; pathway a). On the other
3
hand, trans-cyclization through [TS2-1j]‡
exhibited a lower
trans
activation barrier of 6.7 kcalmolÀ1 (Figure 4, top; pathway b)
opening up a lower energy pathway leading to the triplet 1,4-
diradical TDR2-1jtrans
.
An intersystem crossing in TDR2-1jtrans can give rise to a zwit-
terionic intermediate ZI-1jtrans. From this zwitterionic (ZI) inter-
mediate, two different transformations can be envisaged. A
direct proton transfer can lead to the cyclized product 4, as de-
picted in Figure 4. This kind of direct proton transfer from C7
To gain further insights into mechanistic aspects of the
atropselective photoreaction (as our photophysical studies re-
vealed that the reaction originates from a triplet excited state),
computational investigations were carried out on substrates
1a and 1j. Time-dependent density functional theory
(TDDFT)[12] calculations on both 1a and 1j showed that the
lowest energy triplet excitation is from HOMOÀ1 to LUMO. Ex-
amination of these orbitals indicated a pp* excited state (Fig-
ure S5). To comprehend the intriguing differential reactivity
and to offer a plausible mechanistic rationale for both hydro-
gen abstraction (for 1a–i) and photocyclization (for 1j) path-
ways, we computed all of the stationary points potentially in-
volved in the reaction pathway.
to C4 in ZI-1jtrans via [TS3-1j]‡
is found to be uphill by
trans
15.5 kcalmolÀ1. On the other hand, inclusion of two water mol-
ecules (moisture present under our reaction conditions in ace-
tonitrile, substantiated by deuterium incorporation from the
solvent,[8] see below) through a specific hydrogen-bonding in-
teraction with the negatively charged oxygen atom of ZI-1jtrans
resulted in an additional stabilization by around 12 kcalmolÀ1
.
For the alternative pathway involving two water-assisted enoli-
zations from ZI-1jtrans-sol via [TS4-1j]‡enol, a barrier of 16.5 kcal
molÀ1 was calculated.[14] Two water-assisted tautomerizations
of the resulting enol 1jenol can provide access to cis-4 and
trans-4 photoproducts through [TS5-1j]‡ and [TS5-1j]‡trans, re-
cis
spectively. The elementary step barriers of 29.4 kcalmolÀ1 and
28.9 kcalmolÀ1 for the formation of cis-4 and trans-4 products,
respectively, are quite comparable. However, the barrier for the
one water-assisted tautomerization in the cis product is
3.9 kcalmolÀ1 lower than that for trans product formation,
which is consistent with the experimental observation of cis-4
as the major product (Figure S10). To verify the above reason-
ing of solvent-assisted product formation, we carried out the
experiment in a mixture of CD3CN/D2O as the solvent. Analysis
of the reaction mixture revealed deuterium incorporation in
the product.[8]
To understand the mechanism[13] and stereochemical out-
come of hydrogen abstraction from atropisomeric enone car-
boxamides 1a–h, we performed DFT calculations on M-1a
(Figure 3) as a representative example leading to the corre-
sponding spirolactam photoproduct 2a. Due to free rotation
about the C1ÀN2 and C1ÀC4 bonds (Figure 3), the rotamer M-
1a’ was found to be 0.8 kcalmolÀ1 higher in energy than M-
1a, and these are expected to be in equilibrium. The triplet ex-
cited-state geometry 31a* has conjugated radical character
with maximum spin densities on carbon atoms C4 and C5 (Fig-
ure S6).[8] Taking the energy of 31a* as a reference, the energet-
ics of the hydrogen abstraction pathway leading to the spiro-
lactam product 2a was computed (Figure 3). The reactive trip-
let 31a* undergoes a hydrogen abstraction, resulting in the for-
mation of a triplet 1,4-diradical TDR-1a that is lower in energy
by 12.3 kcalmolÀ1. The formation of this TDR-1a is an uphill
As computational insights were useful in delineating the di-
vergent reactivity of achiral and atropisomeric enones, we fur-
ther probed the atropselectivity observed in 1a–h leading to
spiro-lactams 2. Once again, we utilized M-1a as the model
3
system. Two key transition states, that is, triplet [TS1-1a]‡ and
1
singlet [TS2-1a]‡, were found to play a crucial role in deter-
3
process that occurs through a triplet transition state [TS1-1a]‡.
mining the stereochemistry of the spirolactam product
(Figure 3). For abstraction of the benzylic hydrogen atom
through triplet 3[TS1-1a]‡, two distinct hydrogen abstraction
Intersystem crossing of triplet 1,4-diradical TDR-1a leads to
a lower energy singlet 1,4-biradical SDR-1a. This singlet diradi-
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11344
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 3. Energy profile diagram obtained at the SMDMeCN/UBH and HLYP/6-31+G**//SMDMeCN/UBH and HLYP/6-31+G* level of theory for the photochemical
reaction of atropisomeric 1a leading to spiro-b-lactam 2a.
modes were considered (Figure 5). The transition state
3[TS1-1a]‡anti, wherein the benzylic phenyl ring is orientated
away from the ortho-tert-butyl substituent, was found to be
lower in energy by 2.6 kcalmolÀ1 as compared to the transition
3
state [TS1-1a]‡syn, in which the benzylic phenyl ring and ortho-
tert-butyl substituent are closer to each other, resulting in in-
creased steric crowding.
The next key transition state (Figure 3) that determines the
stereochemistry during the CÀC bond formation leading to the
Scheme 3. Atropselective photoreaction of an axially chiral enone.
1
spirolactam photoproduct is [TS2-1a]‡. We refined the ener-
gies at the CASSCF(8,6)/ANO-RCC level of theory by making
use of the optimized geometries of the four lowest energy dia-
stereomeric transition states obtained at the UBH and HLYP/6-
31+G* level of theory. As illustrated in Figure 6, the relative
energy of the transition state for the formation of 3S,4R-2a is
2.6 kcalmolÀ1 lower than that for the formation 3S,4S-2a. The
greater distortion in the latter transition state can be attributed
to the energy difference. This corresponds to a diastereoselec-
tivity of 97% and compares well with our observed experimen-
tal selectivity (d.r. 95:5; Table 4). The formation of other stereo-
isomers required higher activation energies (Figure 6). Thus,
atropselective spirolactam formation from axially chiral enone
(Scheme 3) is determined by the relative orientation of the
benzylic aryl group and the ortho-tert-butyl substituent during
the hydrogen-abstraction step (via 3[TS1-1]‡) as well as the
ring-closure step (1[TS2-1]‡).
strain in the atropisomeric system. This forces the system to
deactivate from the excited state through a different pathway.
When a methylene hydrogen atom with appropriate bond dis-
sociation energy is available for abstraction, the reaction is
channeled towards a pathway that results in the spiro-b-
lactam photoproduct, as observed with atropisomeric enones.
Conclusion
In conclusion, we have uncovered divergent reactivity based
on restricted bond rotation(s), which is a feature of atropiso-
meric substrates when compared to their achiral counterparts.
The restricted bond rotation(s) in atropisomeric systems can
be fine-tuned to arrest the photochemical reactivity that is typ-
ically observed in the corresponding achiral system. This re-
striction of reactivity from the excited state opens up a new
pathway leading to unexpected reactivity.[2a,15] Through photo-
chemical, photophysical, and computational investigations, the
origin of this unusual reactivity has been delineated, opening
Our study has revealed that upon photoexcitation atropiso-
meric enones attain the triplet excited state. The excited-state
energy that is typically directed towards 6p-photocyclization
(as in achiral enone 1j) is hindered due to the developing 1,3-
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11345
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 4. Energy profile diagram at the SMDMeCN/UBH and HLYP/6-31+G**//SMDMeCN/UBH and HLYP/6-31+G* level of theory for the photocyclization of 1j.
Experimental Section
General irradiation procedures
Photoirradiations were carried out at room temperature either
under constant bubbling of N2 or with an N2-degassed solution
using one of the following light sources: a) broad band (450 W
medium pressure mercury lamp), b) Rayonet reactor equipped
with either approximately 300, 350, or 420 nm lamps. All reported
photoreactions were carried out for a minimum of three trials. In
a typical experiment, enone carboxamide (e.g., 1a; c ꢀ 2.3–
2.3 mm or 1 mg of the reactant in 1 mL of a given solvent in
a Pyrex test tube) was dissolved and the reaction mixture was de-
gassed with N2 for 10–15 min. The reaction mixture was then irra-
diated with a given lamp source (broad band or Rayonet reactor,
Figure 5. Optimized geometries of the hydrogen abstraction transition
ca. 350 nm for direct irradiation or ca. 420 nm for sensitized irradia-
3
states [TS1-1a]‡ for 1a computed at the SMD(CH3CN)/UBH and HLYP/6-
tion with thioxanthone). After photoirradiation, a known amount
31+G**//SMD(CH3CN)/UBH and HLYP/6-31+G* level of theory. The total elec-
of internal standard (triphenylmethane) was added and the solu-
tronic energies with respect to the lowest energy transition state are provid-
tion was concentrated under reduced pressure to obtain the crude
ed in parentheses.
residue; % conversion, % mass balance, and % yield were obtained
1
from H NMR spectroscopy. Large-scale reactions were carried out
in batches (4”20 mL Pyrex test tubes per batch). After the irradia-
up new and interesting excited-state chemistry in light-
induced transformations.
tion, the reaction mixtures were combined and excess solvent was
evaporated under reduced pressure. The resultant crude residue
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11346
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
CASSCF(8,6) level using the optimized geometries of the stereo-
controlling CÀC bond formation transition states. CASSCF calcula-
tions were performed using the MOLCAS 7.8 program.[24] Basis sets
for all atoms were employed from the ANO-RCC[25] library available
in the MOLCAS package. Further details of CASSCF calculations are
given in Figure S11.
Supporting information for this article is given via a link at the end
of the document. Supporting information contains experimental
procedures, single-crystal XRD (CIF format; CCDC-1469514,
1469515, 1469516, 1469517, 1469518, 1469519, 1469520, 1469521,
and 1469522), characterization data and analysis conditions, and
optimized computational coordinates. Figures S5–S12 cited in the
text are available.
Acknowledgements
The authors thank the National Science Foundation (NSF) for
generous support for their research (CHE-1213880) and CHE-
1465075 for the purchase of a drysolv system and Combiflash.
A.J.A. thanks the NSF for a graduate research fellowship. The
authors thank the NDSU Center for Computationally Assisted
Science and Technology (CCAST) for providing computational
resources through the U.S. Department of Energy Grant No.
DE-SC0001717. The authors thank Ms. Ramya Raghunathan
and Ravichandranath Singathi for their help during the prepar-
ation of this manuscript and Dr. Angel Ugrinov for solving the
XRD structures.
Figure 6. Optimized geometries of the CÀC bond formation transition states
1[TS2-1a]‡ obtained at the SMD(CH3CN)/UBH and HLYP/6-31+G* level of
theory. Relative energies at the CASSCF(8,6)/ANO-RCC//SMD(CH3CN)/UBH and
HLYP/6-31+G* level are given in parentheses. Bottom: Single-crystal XRD
structure of 3R,4S-2b obtained from photoreaction of (P)-1b.
was purified by Combiflash using a hexanes/ethyl acetate mixture
as the eluent.
Computational details
Keywords: asymmetric photochemistry
·
atropisomerism
·
Calculations were performed using the Gaussian 09 suite of quan-
tum chemical programs.[16] All geometry optimizations were initial-
ly performed at the UBH and HLYP[17] level of theory, which has
been successfully employed in the study of radical systems[18] with
the 6–31+G* basis set.[19] All stationary points were verified by vi-
brational frequency analysis. Further verification of transition states
was performed using intrinsic reaction coordinate (IRC) calcula-
tions.[20] Solvent effects were incorporated within the continuum
solvation model using the Truhlar–Cramer universal solvation
model.[21] Since the reactions were performed in acetonitrile, a con-
tinuum solvent dielectric constant of acetonitrile (e=35.688) was
employed for the calculations (SMDCH3CN). Singlet diradical species
were studied using the broken-symmetry approach.[22] The wave-
function resulting from this approach is affected by spin contami-
nation. Spin-pure energies were calculated using the spin-
correction method proposed by Yamaguchi, Houk, and co-work-
ers[23] as given below:
atropselective photoreactions
·
enones
·
hydrogen
abstraction · photochirogenesis
[1] a) V. Ramamurthy, Photochemistry in Organized and Constrained Media,
Wiley-VCH, New York, 1991, 429–493; b) T. Mori, R. G. Weiss, Y. Inoue, J.
Am. Chem. Soc. 2004, 126, 8961–8975; c) M. A. Garcia-Garibay, Acc.
Chem. Res. 2003, 36, 491–498; d) J. N. Gamlin, R. Jones, M. Leibovitch,
B. Patrick, J. R. Scheffer, J. Trotter, Acc. Chem. Res. 1996, 29, 203–209;
e) M. Veerman, M. J. E. Resendiz, M. A. Garcia-Garibay, Org. Lett. 2006, 8,
2615–2617; f) J. Sivaguru, A. Natarajan, L. S. Kaanumalle, J. Shailaja, S.
Uppili, A. Joy, V. Ramamurthy, Acc. Chem. Res. 2003, 36, 509–521; g) C.
Yang, T. Mori, Y. Origane, Y. H. Ko, N. Selvapalam, K. Kim, Y. Inoue, J. Am.
Chem. Soc. 2008, 130, 8574–8575; h) T. Bach, H. Bergmann, B. Grosch,
K. Harms, J. Am. Chem. Soc. 2002, 124, 7982–7990.
[2] a) N. Vallavoju, J. Sivaguru, Chem. Soc. Rev. 2014, 43, 4084–4101; b) B. C.
Pemberton, R. Raghunathan, S. Volla, J. Sivaguru, Chem. Eur. J. 2012, 18,
12178–12190.
[3] a) A. Ates, D. P. Curran, J. Am. Chem. Soc. 2001, 123, 5130–5131; b) A.
Honda, K. M. Waltz, P. J. Carroll, P. J. Walsh, Chirality 2003, 15, 615–621;
c) J. Clayden, Chem. Commun. 2004, 127–135; d) E. Kumarasamy, R.
Raghunathan, M. P. Sibi, J. Sivaguru, Chem. Rev. 2015, 115, 11239–
11300.
[4] a) A. J.-L. Ayitou, A. Clay, E. Kumarasamy, S. Jockusch, J. Sivaguru, Photo-
chem. Photobiol. Sci. 2014, 13, 141–144; b) A. Clay, E. Kumarasamy,
A. J. L. Ayitou, J. Sivaguru, Chem. Lett. 2014, 43, 1816–1825; c) E.
Kumarasamy, R. Raghunathan, S. Jockusch, A. Ugrinov, J. Sivaguru, J.
Am. Chem. Soc. 2014, 136, 8729–8737; d) A. Iyer, S. Jockusch, J.
Sivaguru, J. Phys. Chem. A 2014, 118, 10596–10602; e) E. Kumarasamy,
J. Sivaguru, Chem. Commun. 2013, 49, 4346–4348; f) E. Kumarasamy,
J. L. Jesuraj, J. N. Omlid, A. Ugrinov, J. Sivaguru, J. Am. Chem. Soc. 2011,
133, 17106–17109; g) J. L. Jesuraj, J. Sivaguru, Chem. Commun. 2010,
46, 4791–4793; h) A. J.-L. Ayitou, J. Sivaguru, J. Am. Chem. Soc. 2009,
131, 5036–5037; i) A. J.-L. Ayitou, J. L. Jesuraj, N. Barooah, A. Ugrinov, J.
Sivaguru, J. Am. Chem. Soc. 2009, 131, 11314–11315.
1
3
1ESC ¼ EUB þ fSC½1EUB À EUB
Š
ð1Þ
ð2Þ
1
CT2
hs2i
fSC
¼
¼
1 À CS2
hs2À hs2ii
3
1
1
where ESC =spin-corrected energy, EUB =spin-contaminated energy,
SC =fraction of spin contamination, CS and CT are the coefficients
of the pure singlet and triplet components, and S2 =total spin an-
f
gular momentum operator.
The discussions are presented on the basis of the total electronic
energies obtained at the SMD(CH3CN)/UBH and HLYP/6-31+G**//
SMD(CH3CN)/UBH and HLYP/6–31+G* level of theory. Single-point
computations were carried out by computing the energetics at the
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11347
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[5] a) C. Meyer, O. Piva, J.-P. Pete, Tetrahedron 2000, 56, 4479–4489; b) S.
Le Blanc, J.-P. Pete, O. Piva, Tetrahedron Lett. 1992, 33, 1993–1996; c) T.
Shintani, K. Kusabiraki, A. Hattori, A. Furutani, K. Tsutsumi, T. Morimoto,
K. Kakiuchi, Tetrahedron Lett. 2004, 45, 1849–1851.
[6] a) T. Naito, Y. Tada, I. Ninomiya, Heterocycles 1984, 22, 237–240; b) Y.
Ogata, K. Takagi, I. Ishino, J. Org. Chem. 1971, 36, 3975–3979; c) T. Bach,
B. Grosch, T. Strassner, E. Herdtweck, J. Org. Chem. 2003, 68, 1107–1116;
d) A. J.-L. Ayitou, J. Sivaguru, Chem. Commun. 2011, 47, 2568–2570.
[7] a) N. J. Turro, V. Ramamurthy, J. C. Scaiano, Modern Molecular Photo-
chemistry of Organic Molecules, University Science Books, Sausalito, CA,
2010, 629–704; b) D. I. Schuster, G. Lem, N. A. Kaprinidis, Chem. Rev.
1993, 93, 3–22.
J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. Heyd,
E. N. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C.
Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J.
Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma,
V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich,
A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox,
Gaussian, Inc., Wallingford, CT, USA, 2009.
[17] A. D. Becke, J. Chem. Phys. 1993, 98, 1372–1377.
[18] a) F. Godin, M. PrØvost, F. Viens, P. Mochirian, J.-F. Brazeau, S. I. Gorelsky,
Y. Guindon, J. Org. Chem. 2013, 78, 6075–6103; b) I. Ryu, T. Fukuyama,
M. Tojino, Y. Uenoyama, Y. Yonamine, N. Terasoma, H.
Matsubara, Org. Biomol. Chem. 2011, 9, 3780–3786; c) S. H. Kyne, C. H.
Schiesser, H. Matsubara, J. Org. Chem. 2008, 73, 427–434.
[8] Refer to Supporting Information.
[9] O. Kitagawa, M. Fujita, M. Kohriyama, H. Hasegawa, T. Taguchi, Tetrahe-
dron Lett. 2000, 41, 8539–8544.
[10] J. Jia, M. G. Steinmetz, R. Shukla, R. Rathore, Tetrahedron Lett. 2008, 49,
4621–4623.
[11] A. J.-L. Ayitou, N. Vallavoju, A. Ugrinov, J. Sivaguru, Photochem. Photo-
biol. Sci. 2011, 10, 1380–1383.
[19] a) P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–222;
b) V. A. Rassolov, J. A. Pople, M. A. Ratner, T. L. Windus, J. Chem. Phys.
1998, 109, 1223–1229.
[20] a) C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154–2161; b) C.
Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523–5527.
[21] A. V. Marenich, C. J. Cramer, D. G. Truhlar, J. Phys. Chem. A 2009, 113,
6378–6396.
[22] a) L. Noodleman, E. R. Davidson, Chem. Phys. 1986, 109, 131–143; b) L.
Noodleman, D. A. Case, A. Aizman, J. Am. Chem. Soc. 1988, 110, 1001–
1005.
[23] a) K. Yamaguchi, F. Jensen, A. Dorigo, K. N. Houk, Chem. Phys. Lett. 1988,
149, 537–542; b) N. C. James, J. M. Um, A. B. Padias, H. K. Hall, K. N.
Houk, J. Org. Chem. 2013, 78, 6582–6592.
[24] a) G. Karlstrçm, R. Lindh, P.-Š. Malmqvist, B. O. Roos, U. Ryde, V.
Veryazov, P.-O. Widmark, M. Cossi, B. Schimmelpfennig, P. Neogrady, L.
Seijo, Comput. Mater. Sci. 2003, 28, 222–239; b) V. Veryazov, P.-O.
Widmark, L. Serrano-AndrØs, R. Lindh, B. O. Roos, Int. J. Quantum Chem.
2004, 100, 626–635; c) J. A. Duncan, J. Am. Chem. Soc. 2009, 131,
2416–2416; d) F. Aquilante, L. De Vico, N. FerrØ, G. Ghigo, P.-Š.
[12] a) R. E. Stratmann, G. E. Scuseria, M. J. Frisch, J. Chem. Phys. 1998, 109,
8218–8224; b) R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 1996,
256, 454–464; c) M. E. Casida, C. Jamorski, K. C. Casida, D. R. Salahub, J.
Chem. Phys. 1998, 108, 4439–4449.
[13] a) R. Rajeev, R. B. Sunoj, J. Org. Chem. 2012, 77, 2474–2485; b) R. Rajeev,
R. B. Sunoj, J. Org. Chem. 2013, 78, 7023–7029; c) M. Abe, Chem. Lett.
2013, 113, 7011–7088; d) R. K. Mohamed, P. W. Peterson, I. V. Alabugin,
Chem. Lett. 2013, 113, 7089–7129; e) W. T. Borden, J. Org. Chem. 2011,
76, 2943–2964; f) A. Sinicropi, F. Barbosa, R. Basosi, B. Giese, M. Olivucci,
Angew. Chem. Int. Ed. 2005, 44, 2390–2393; Angew. Chem. 2005, 117,
2442–2445.
[14] a) All attempts to locate one water-assisted enolization transition state
resulted in direct product formation through [TS3–1j]trans (Figure S8).
[15] a) N. Vallavoju, S. Selvakumar, S. Jockusch, M. T. Prabhakaran, M. P. Sibi,
J. Sivaguru, Adv. Synth. Catal. 2014, 356, 2763–2768; b) N. Vallavoju, S.
Selvakumar, S. Jockusch, M. P. Sibi, J. Sivaguru, Angew. Chem. Int. Ed.
2014, 53, 5604–5608; Angew. Chem. 2014, 126, 5710–5714; c) N.
Vallavoju, S. Selvakumar, B. C. Pemberton, S. Jockusch, M. P. Sibi, J.
Sivaguru, Angew. Chem. Int. Ed. 2016, 55, 5446–5451.
[16] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.
Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
ˇ
Malmqvist, P. Neogrµdy, T. B. Pedersen, M. Pitonµk, M. Reiher, B. O. Roos,
L. Serrano-AndrØs, M. Urban, V. Veryazov, R. Lindh, J. Comput. Chem.
2010, 31, 224–227.
[25] B. O. Roos, R. Lindh, P.-Š. Malmqvist, V. Veryazov, P.-O. Widmark, A. C.
Borin, J. Phys. Chem. A 2008, 112, 11431–11435.
Received: March 31, 2016
Published online on July 6, 2016
Chem. Eur. J. 2016, 22, 11339 – 11348
www.chemeurj.org
11348
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim