C. P. Frazier et al. / Tetrahedron Letters 56 (2015) 3353–3357
3355
While the Diels–Alder reaction with nitrosoformate intermedi-
ates is known to be a highly efficient reaction,18 the corresponding
ene reaction of nitrosoformate intermediates often suffers from
poor yields due to the over-oxidation of the allylic hydroxylamine
products.21 Recent efforts have been undertaken by a number of
research groups to circumvent this known decomposition reac-
tion,22 and thus the ene reaction has proven to be an effective tool
for evaluating the mildness of oxidation reaction conditions.7 In
this context, we elected to test our newly developed oxidation con-
ditions with a subset of olefin partners. We were delighted to find
that the ene adducts were readily formed under our optimized
conditions with no observable product decomposition (Table 3).
As with other studies, the efficiency of the ene reaction is directly
correlated to the substitution of the olefin; fully substituted olefins,
such as 2,3-dimethylbutene, form the corresponding allylic
hydroxylamine in quantitative yield (13). In contrast, mono-substi-
tuted olefins such as 1-octene react with nitrosocarbonyl com-
pounds with reduced yields (17). In each case, the photoredox
conditions compared favorably with previous oxidative nitrosocar-
bonyl ene methodologies (13–17).1 Based on these results, we
envision that these conditions can be utilized for a wide variety
of reaction platforms and will facilitate the expansion of nitroso-
carbonyl chemistry into previously unexplored areas of organic
synthesis.
the synthesis of a nitroso-Diels –Alder block copolymer 21.24
Hydroxamic acid terminated PEG polymer (2k mPEG-HA) 19 was
allowed to react under slightly modified reaction conditions with
cyclopentadiene end-capped polystyrene polymer (12k PS-Cp)
20. After 24 h, complete conversion of 20 was observed and block
copolymer 21 was successfully isolated by trituration in methanol–
water (1:1) (Scheme 1).
O
H
O
O
N
O
OH
+
n
O
m
O
Ph
19
20
2k mPEG-HA
12k PS-Cp
8 mol% Ru(bpy)3Cl2,
6 mol eq 2,6-lutidine
DCE, air, rt, hν, 24 h
O
O
O
N
O
n
O
O
m
O
Ph
21
14k mPEG-b-PS
Scheme 1. Photoredox catalyzed synthesis of nitroso-Diels–Alder block copolymer
21.
Encouraged by the mildness of our reaction conditions, we next
elected to explore the efficiency of temporal control for our opti-
mized conditions. While on–off studies are not normally per-
formed for photoredox reactions in small molecule chemistry,
this type of experiment is regularly conducted with photoredox
polymerizations,12,23 and is a prerequisite for transitioning a meth-
odology into a number of materials science applications such as
the patterning of functionalized surfaces.13 Therefore, we were
delighted to find that the reaction could be effectively turned on
and off based on the presence or absence of light. The reaction
showed no reduction in efficiency even after iteratively turning
the light on and off. Moreover, the reaction could even be shut
off for extended periods of time (e.g., 18 h) and started again once
the light was reintroduced, highlighting a unique element of con-
trol never exploited for the generation of nitrosocarbonyl interme-
diates (Fig. 2).
We next explored the distance dependence of our newly devel-
oped reaction conditions. For the design of this experiment, the
lights were placed at a variable distance away from the sidewall
of the reaction flask and the reaction was monitored periodically
for conversion. As expected, when the lights were touching the
sidewall of the flask, the reaction proceeded the fastest, with a rate
paralleling our on–off study. As the lights were moved further from
the flask, the reaction rate dropped precipitously. Based on these
results, there is a clear spatial dependence of the reaction rate:
the intensity of the light has a direct effect on the reaction
(Fig. 3). Based on the distance dependence and the on–off studies,
we believe that this reaction has the potential to find widespread
use in applications where temporal and spatial control are desired,
and our studies in this area will be reported in due time.
Eager to demonstrate application of these conditions in a mate-
rials setting, we decided to employ our photoredox conditions for
Based on the trends we observed during the reaction develop-
ment, we were interested in further exploring the mechanism. In
particular, we wanted to distinguish between a photoredox mech-
anism and a mechanism involving singlet oxygen. In order to probe
this, we turned to Stern–Volmer studies to confirm the direct inter-
action between the catalyst and the Cbz-protected hydroxyl-
amine.25 As expected, the magnitude of fluorescence of the
catalyst had a linear dependence on the concentration of the
Cbz-protected hydroxylamine,26 indicating an interaction between
the excited state of the catalyst and the Cbz-protected hydroxyl-
amine (Fig. 4). In contrast, no fluorescence quenching was
observed in the presence of varying concentrations of the diene
(See SI for more information). Based on these results, we propose
that the excited state of the catalyst (Ru(bpy)3Cl⁄2) is directly oxi-
dizing the Cbz-protected hydroxylamine to the corresponding rad-
ical cation, consistent with the known redox potentials of the
catalyst excited state (E1/2 = 1.08 V vs Ag/AgNO3)8 and the Cbz-pro-
tected hydroxylamine (Ep/2 = 0.95 V vs Ag/AgNO3, see Supporting
information). The subsequent reduction of molecular oxygen forms
the corresponding superoxide radical, which facilitates the conver-
sion of the Cbz-hydroxylamine radical cation to the nitrosocarbon-
yl species (Fig. 5A).
To explore the mechanism further, we tested the reaction effi-
ciency in the presence of a known singlet oxygen sensitizer, tetra-
phenylporphyrin (TPP).27 To our surprise, upon irradiation of the
Figure 2. On–off studies for the developed photoredox oxidation. aReactions
conducted using dimethyl terephthalate as the internal standard. bAliquots were
removed at the indicated times and conversion was determined by 1H NMR
analysis.
reaction,
a small amount of hetero-Diels–Alder adduct was