Journal of the American Chemical Society
Communication
Scheme 3. Mechanistic Features of Photocycloaddition of 1,3-Dicarbonyl Compounds with Amino-Alkenes
note that irradiation of 4a in methanol gave 76% isolated yield
Scheme 2) and irradiation of the mixture of 1a and 2a under
similar conditions gave 80% isolated yield (Scheme 1). This
once again reinforced our conjecture that enaminone 4a was
likely responsible for the observed reactivity upon irradiation
of 1,3-dicarbonyl compound 1a and amino-styrene 2a
To understand the observed reactivity, preliminary photo-
physical studies were performed on enaminones to understand
their excited state properties. We utilized enaminone 4a as a
model compound. There was no observable fluorescence of 4a
at room temperature (fluorescence quantum yield <0.001),
that indicated fast excited state deactivation processes. Time
resolved luminescence measurements of 4a at 77 K in ethanol
glass revealed a weak phosphorescence centered around 490
nm (Figure 1, right; red). This weak phosphorescence limited
the ability to ascertain the excited state lifetime of the triplet
state. To overcome this limitation, we synthesized the control
substate 5a (lacking the alkenyl substituent on the phenyl ring
with N-Boc protection) that displayed phosphorescence
similar to 4a albeit with higher intensity with a lifetime of
∼75 ms (Figure 1, right). This showed that the triplet was
localized on the enaminone functionality. We believe that the
fast relaxation of the excited state of 4a can have its origin in
distinct deactivation modes viz., (a) isomerization of the
double bond, and/or (b) excited state intramolecular proton
transfer (ESIPT), and/or (c) charge transfer in the excited
state (as it is a push−pull system). Our photophysical studies
revealed that the triplet excited state energy of 4 to be around
58 kcal/mol above the ground state. This enabled us to utilize
(
(
Scheme 1). Control studies in the absence of light (thermal
10
control) led to the recovery of enaminone 4a. To further
understand the reactivity patterns, the photoreactivity of
enaminone 4a leading to the photoproduct 3a was investigated
under various conditions (Scheme 2). Two distinct aspects
were specifically evaluated viz., role of the solvent and
irradiation wavelength. As expected, the photoreaction was
faster at ∼350 nm, compared to purple LED irradiation
10
(
Scheme 2) due to the difference in the optical density of the
substrate at a given concentration. Consequently, longer
reaction times were employed for purple LED irradiations
for achieving similar conversions as ∼350 nm irradiation. The
phototransformation was clean and efficient in methanol at
∼
350 nm with 76% isolated yield of photoproduct 3a (3a
1
10
observed exclusively in crude H NMR spectroscopy). The
reaction was also observed to be clean and efficient with purple
LED albeit with longer irradiation times. Moderate yield of 3a
was observed in other solvents viz. 37% yield in acetonitrile,
thioxanthone (E ∼ 64 kcal/mol) as sensitizer/photocatalyst
T
4
2
3% yield in ethyl acetate and 27% yield in toluene (Scheme
to carry out the transformations. Irradiation of enaminones
4a−c in the presence of thioxanthone at ∼420 nm for 48 h
resulted in the dihydropyran photoproducts with yields of 29%
of 3a, 20% for 3b, and 43% for 3c (Scheme 2). In the absence
of thioxanthone, irradiation of 4b at ∼420 nm for 44 h did not
).
To generalize the observed reactivity leading to photo-
product 3, we investigated the reactivity of enaminones 4b−h
which were independently synthesized from the corresponding
1
0
10
diketones 1b−f and 2a,b (Scheme 2), respectively.
show any appreciable photoproducts (<2% conversions).
Irradiation at ∼350 nm in acetonitrile resulted in isolated
yield of 40% for 4b and 60% for 4c. Changing the solvent to
methanol resulted in isolated yield of 30% for 4b, 80% for 4c,
and 60% for 4d. Purple LED irradiation of enaminone 4b−f in
methanol gave the photoproduct 3b−h in yields varying from
This indicated that the reaction can also be performed under
1
0
photocatalytic conditions. As electron-transfer initiated
10
reactivity was endergonic based on the redox potentials of
13,14
enaminone 4 and thioxanthone,
the photoreactivity under
sensitized/photocatalytic conditions occurs likely via an energy
transfer process (Scheme 2).
9
to 78% (9% for 4b, 78% for 4c, 49% for 4d, 71% for 4e, 10%
for 4f). The dialkyl substituted enaminone 4b gave lower
conversions because its absorptivity was weak in the visible
region. Irradiation of enaminones 4g and 4h featuring α-
methyl substituted-amino styrene unit gave the corresponding
photoproducts 3g and 3h in 50% and 35% yields, respectively
On the basis of our photochemical and photophysical
investigations, we propose a preliminary mechanistic model for
the observed reactivity (Scheme 3). Irradiation of diketone 1
and amino-styrene 2, results in the excitation of in situ
generated enaminone 4. The photoexcited enaminone can
react either through a singlet or triplet manifold via four
distinct pathways (Scheme 3) viz., (i) a diradical pathway; (ii)
an ionic pathway; and (iii) an electron transfer pathway or (iv)
an excited state intramolecular proton transfer (ESIPT)
(
Scheme 2). A point to note is the reaction efficiency with
enaminones featuring α-methyl substituted-amino styrene (30
min to 1 h irradiation with purple LED) was more efficient
than enaminones derived from 2a (24−48 h irradiation with
purple LED) highlighting the role of the substitution on the
styrenyl unit.
15−17
pathway
leading to 3. The reaction pathway depends on
the substrate(s) and the employed conditions for the
3
679
J. Am. Chem. Soc. 2021, 143, 3677−3681