Catalytic oxidative cyclization of 2′-arylbenzaldehyde oxime ethers under photoinduced electron transfer conditions

Article abstract of DOI:10.1021/jo502324z

A series of 2′-arylbenzaldehyde oxime ethers were synthesized and shown to generate the corresponding phenanthridines upon irradiation in the presence of 9,10-dicyanoanthracene in acetonitrile. Mechanistic studies suggest that the oxidative cyclization reaction sequence is initiated by an electron transfer step followed by nucleophilic attack of the aryl ring onto the nitrogen of the oxime ether. A concave downward Hammett plot is presumably the result of a change in charge distribution in the radical cation species with strongly electron-donating substituents that yields a less electrophilic nitrogen atom and a decreased amount of cyclized product. The reaction is selective (no nitrile byproduct is formed unlike other photochemical reactions involving aldoxime ethers) as well as regiospecific when using 2′-aryl groups with metasubstituents, making this reaction a useful alternative for preparing substituted phenanthridines.

Full text of DOI:10.1021/jo502324z

Article
pubs.acs.org/joc
Catalytic Oxidative Cyclization of 2′-Arylbenzaldehyde Oxime Ethers
under Photoinduced Electron Transfer Conditions
Julie L. Hofstra, Brittany R. Grassbaugh, Quan M. Tran, Nicholas R. Armada, and H. J. Peter de Lijser*
Department of Chemistry and Biochemistry, California State University, Fullerton, California 92834-6866, United States
S
* Supporting Information
ABSTRACT: A series of 2′-arylbenzaldehyde oxime ethers
were synthesized and shown to generate the corresponding
phenanthridines upon irradiation in the presence of 9,10-
dicyanoanthracene in acetonitrile. Mechanistic studies suggest
that the oxidative cyclization reaction sequence is initiated by
an electron transfer step followed by nucleophilic attack of the
aryl ring onto the nitrogen of the oxime ether. A concave
downward Hammett plot is presumably the result of a change
in charge distribution in the radical cation species with strongly
electron-donating substituents that yields a less electrophilic
nitrogen atom and a decreased amount of cyclized product.
The reaction is selective (no nitrile byproduct is formed unlike
other photochemical reactions involving aldoxime ethers) as well as regiospecific when using 2′-aryl groups with meta-
substituents, making this reaction a useful alternative for preparing substituted phenanthridines.
Scheme 1. Proposed Mechanism for the Formation of
Carbonyl and Nitrile Compounds from Oximes under PET
Conditions
INTRODUCTION
■
Oximes and oxime ethers have found use in a number of
applications, many of which involve reactive intermediates such
as carbocations, radicals, and radical ions.1−3 Despite their
abundance and widespread use, little is known about the
reactivity of the intermediates formed in the oxidative processes
of oximes and oxime ethers. Work in our lab has focused on the
reactivity of oxime and oxime ether radical cations generated
under photoinduced electron transfer (PET) studies. Oximes
were found to react to give the parent carbonyl compounds via
an electron transfer pathway when irradiated in the presence of
a photosensitizer such as chloranil.4−6 Oximes derived from
ketones reacted under these conditions to regenerate the parent
ketone, whereas aldehyde oximes were found to react to give
both the corresponding aldehyde and nitrile products.4,6 The
important intermediate in the pathway that forms the parent
carbonyl compound is the iminoxyl radical, which is formed via
an electron transfer−proton transfer (ET−PT) sequence (for
oximes with low oxidation potentials) or via a hydrogen atom
transfer (HAT) pathway (for oximes with larger oxidation
potentials).6 The nitriles were proposed to result from
intermediate imidoyl radicals, which are formed via hydrogen
atom abstraction at the iminyl carbon (Scheme 1).7
a
a
S* is the excited state photosensitizer.4−7
oxygen) and subsequent β-scission (Scheme 2; top pathway).
In a nucleophilic solvent, the mechanism is most consistent
with a sequence involving electron transfer, followed by a
nucleophilic attack on the nitrogen, a methanol-assisted [1,3]-
proton transfer, and subsequent loss of an alcohol (Scheme 2,
bottom pathway).8 The latter mechanism presented itself as an
interesting possibility for developing a PET-induced cyclization
mechanism using oxime ethers with built-in nucleophiles. For
example, if aromatic rings are to act as the built-in nucleophile,
attack of the aryl ring on the nitrogen of the oxime ether moiety
could lead to the formation of substituted phenanthridines,
providing an alternative route to the limited existing method-
An important step in the formation of the iminoxyl radical is
the deprotonation of the acidic oxime radical cation as shown
by studies on a series of acetophenone8 and benzaldehyde
oxime ethers.9 Studies on the PET reactions of acetophenone
oxime ethers have shown that the reactivity of these substrates
is solvent-dependent.8 In a non-nucleophilic solvent (MeCN),
the product formation can be explained by a mechanism that
involves electron transfer followed by proton transfer (α to the
Received: October 10, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 2. Mechanistic Pathways of Oxime Ethers upon
One-Electron Oxidation8,9
Table 1. Initial Screening and Optimization of the
Conditions for PET Cyclization of 2′-Biaryl Oxime Ethers
λ
conc
(mM)
yield
c
b
entry
R1
R2
solvent
sensor (nm)
(%)
d
1
H
H
C6D6
CA
350
350
420
420
350
350
420
420
420
420
420
420
420
420
420
420
420
420
420
420
5
5
5
5
5
5
5
5
5
5
d
2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
CD3CN
C6D6
CA
d
3
H
DCA
DCA
CA
d
4
H
CD3CN
C6D6
d
5
tBu
6
tBu
CD3CN
C6D6
CA
19
ologies.10 We report here our studies on the oxidative
d
7
tBu
DCA
DCA
DCA
DCA
cyclization pathways of 2′-biaryl oxime ethers.
8
9
tBu
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
43
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
tBu
58
RESULTS AND DISCUSSION
e
■
10
50
Optimization of the Oxidative Cyclization Pathway.
To investigate the feasibility of the oxidative cyclization of 2′-
biaryl oxime ethers, a series of reactions testing different
conditions and substrates were carried out. The desired biaryl
aldehydes and ketones as well as their oxime and oxime ether
derivatives were prepared according to standard literature
methods with minor modifications.11,12 Our initial photo-
chemical experiments utilized the conditions established in
previous studies.4−9 Most reactions were carried out on a small
scale in NMR tubes using deuterated solvents in order to follow
the reactions by NMR, but in addition, several reactions were
carried out on a larger scale in order to isolate the product.
Samples were also analyzed by gas chromatography mass
spectroscopy (GC-MS) to verify product formation. As
expected, irradiation of an acetonitrile or benzene solution
containing the oxime and CA at 350 nm (Table 1, entries 1 and
2) resulted in the formation of the parent aldehyde and nitrile
only; no cyclization was observed. Similar results were obtained
with 9,10-dicyanoanthracene (DCA) as the sensitizer (entries 3
and 4), suggesting that under these conditions the pathways
that lead to the parent aldehyde and nitrile products are
predominant.4−6 In order for the oxidative cyclization to be
successful, the radical cation intermediate must have a
significant lifetime. Oxime radical cations are highly acidic
and will deprotonate quickly,13 prohibiting nucleophilic attack
by the aryl ring. Using the t-butyl oxime ether instead of the
oxime with CA as the sensitizer gave promising results in
acetonitrile as the solvent (entry 6). The same reaction in
benzene did not give the desired product, which is presumably
due to the nonpolar character of the solvent preventing
electron transfer from being effective (entry 5). The oxime
ethers were also found to absorb light at 300−350 nm, which
led to E/Z isomerization of the oxime ether as well as the
formation of undesirable side products, such as the parent
aldehyde and nitrile. To improve the yield of the desired
product, the reaction conditions were modified to use DCA as
the sensitizer and 420 nm light. Under these conditions, the
yield of the desired cyclic product (2a) increased to 43% when
dissolved in acetonitrile (entry 8). The t-butyl ether was chosen
initially because our previous work has shown that oxime ethers
with hydrogen atoms present on the carbon adjacent to the
oxygen can undergo deprotonation.8,9 However, irradiation of
the oxime methyl ether gave better results than the oxime t-
butyl ether (compare entries 8 and 9; entries 6 and 13),
11
trace
e
12
13
14
15
16
CA
5
28
61
46
18
40
41
49
19
DCA
DCA
DCA
DCA
DCA
DCA
DCA
0.34
0.17
0.02
5
f
17
f
18
1.5
0.76
5
f
19
20
a
Reaction conditions unless stated otherwise: concentration of oxime
ether = 15 mM; sensitizer = DCA, wavelength = 420 nm; photolysis
time = 4 h; solvent = CD3CN. Sensitizers have poor dissolution, thus
the effective concentration is actually less than 5 mM. Yield was
determined by H NMR spectroscopy using 4-nitrobenzaldehyde and
4-nitrobenzyl bromide as internal standards. Reaction was run for a
total of 5 h. Solution was purged with Ar prior to photolysis.
b
c
1
d
e
f
Photolyzed for 24 h.
suggesting that the deprotonation step is slow compared to
attack by the nucleophile. Deprotonation of the oxime methyl
ether leading to a high-energy primary radical may also play a
role in this selectivity.
Because of this improved reactivity and because the methyl
ethers are easier to prepare than the t-butyl ethers, all further
experiments were done exclusively with the methyl derivative.
The effect of dissolved oxygen on the product formation was
found to be negligible (entry 10) suggesting that any radical
species formed in these reactions reacts very rapidly and
reaction with oxygen is not competitive. Although unusual, this
is similar to previous observations from our laboratory on
different systems involving the formation of highly reactive
radical species.7 To further tune the reaction, the role of DCA
was examined in more detail. As expected, in the absence of the
photosensitizer, virtually no reaction occurs and none of the
cyclized product is formed (entries 11 and 12). The standard
reaction conditions use a catalytic amount of DCA (33 mol %
photosensitizer loading); however, in these experiments, we
observed that a significant amount of DCA remained
undissolved throughout the reaction. Integration of the
appropriate peaks in the NMR spectrum of the reaction
mixture suggested that the dissolved amount of DCA was
approximately 0.3 mM (2 mol %), which is significantly lower
than the standard amount used in these reactions. As such, the
B
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 2. Summary of Results for the PET Cyclization Reactions of 2′-Arylbenzaldehyde Oxime Ethers in Acetonitrile in the
Presence of 9,10-Dicyanoanthracene
d
e
yield (%)
isomer ratio
a
b
c
oxime ether (Y)
Eox (V)
ΔGET (kcal mol−1
)
IP (kcal mol−1
)
2
3
DCA blank
H (1a)
1.79
1.89
1.64
1.78
1.92
1.93
1.88
1.54
1.74
1.89
1.93
1.67
1.88
−3.8
−1.6
−7.3
−4.1
−0.9
−0.6
−1.8
−9.6
−5.0
−1.5
−0.7
−6.5
−1.8
201.0
203.0
200.2
200.4
205.1
208.5
203.2
189.1
200.2
206.2
209.8
191.8
58
2
3
0
1
2
1
1
1
1
5
1
1
4
57:43
71:29
88:12
54:46
75:25
70:30
40:60
82:18
59:41
89:11
49:51
66:34
70:30
98:2
3-Cl (1b)
33
55
63
24
<2
56
12
94
22
94:6
3-OCH3 (1c)
3-CH3 (1d)
3-CF3 (1e)
3-NO2 (1f)
4-Cl (1g)
93:7
94:6
95:5
64:36
40:60
85:15
88:12
92:8
4-OCH3 (1h)
4-CH3 (1i)
4-CF3 (1j)
4-NO2 (1k)
4-OPh (1l)
3-F (1m)
trace
59
50:50
93:7
203.4
b
36
98:2
a
Oxidation potentials (half-wave potentials) obtained by cyclic voltammetry. Calculated free energy changes for the one-electron transfer process
1
from the compounds to DCA* in acetonitrile by the application of the Rehm−Weller equation: the excited singlet energy of DCA, 2.90 eV, and
reduction potential of DCA, −0.91 V vs SCE.17,18 Calculated (AM1) ionization potential.21,22 Absolute yields determined by 1H NMR using two
c
d
e
internal standards. Isomer ratios (E/Z) determined by 1H NMR after photolysis (most oxime ethers had an isomer ratio of >95:5 at the onset of the
experiments).
amount of DCA was varied to determine the catalytic nature of
the sensitizer. These experiments show that the same yield of
2a is obtained when using a catalyst (photosensitizer) loading
of 2 mol % or higher (entries 14−16). Even at 0.1 mol % DCA,
a reasonable 18% yield of 2a is obtained after 4 h of photolysis.
These results suggest that the sensitizer acts as a catalyst,
making this reaction attractive because of its catalytic efficiency.
Although the original reaction time (4 h) was chosen arbitrarily
at the onset of these experiments, studies on the time
dependence of the product formation show that, in fact, this
is the optimum time for the reactions using the standard
amount of catalyst. The product formation reaches a maximum
between 4 and 6 h of irradiation and then stabilizes. After
prolonged irradiation (24 h), the product yield was found to
decrease slightly (entries 17−19).
It should be noted that, in these experiments with the oxime
ethers and DCA as the sensitizer, no nitrile formation is
observed. Previous work in our laboratory on the PET reactions
of oximes and oxime ethers derived from aldehydes typically
resulted in the formation of both the parent aldehyde and the
corresponding nitrile products. In many cases, the nitrile
product was the major product.4 Existing methods for
preparing phenanthridines from aldoximes or related com-
pounds all generate significant amounts of the nitrile,10,14,15 and
therefore, the oxidative cyclization pathway presented here is
unique in that it can be used to prepare aldehyde-based
phenanthridine derivatives without also producing the undesir-
able nitrile side products. In addition, our method is catalytic
with respect to the concentration of the photosensitizer, which
is a major advantage over the photosensitized cyclization
method described by Walton.15 Finally, an experiment with the
methyl ketone oxime ether shows that our method can also be
used to prepare ketone-based phenanthridines (entry 20),
although no effort has yet been made to further optimize this
reaction.
Mechanistic Studies. With the optimized conditions
available, we carried out a number of experiments to investigate
the mechanistic aspects of this reaction. Initial experiments
involved m- and p-substituted derivatives 1a−1k.16 The data
obtained from this set of compounds indicate that, in general,
both oxime ethers with electron-donating and electron-
withdrawing substituents are reactive in the oxidative
cyclization pathway, although oxime ethers with electron-
donating substituents tended to give higher yields than those
with electron-withdrawing substituents (Table 2). An obvious
exception to this trend was 1h (4-OCH3). Using the percent
yield of the substituted phenanthridines as a measure of the
equilibrium constant for the reaction (KX), analysis of this data
set in a Hammett plot of the relative product formation (log
KX/KH) against σ+ values19,20 gives a straight line with a slope
(ρ+) of −0.7 if the 4-OCH3 (1h) derivative is excluded (Figure
1). Plotting the data against σ gave similar results, although
there was more scatter in the data. To determine whether the
large deviation of 1h (σ+ = −0.778; Figure 1) was part of a
concave downward correlation, a separate experiment with the
4-OPh derivative (1l; σ+ = −0.50) was carried out to
corroborate this observation. Compound 1l reacted to give
the expected phenanthridine product 2l but in lower yield than
compound 1i, confirming the downward trend.
The right-hand side of the Hammett plot (negative slope
with ρ+ = −0.7) clearly suggests the involvement of a species
with a developing positive charge, which is consistent with a
pathway involving reactive intermediates generated by an
oxidative PET process.5,6 Previously, the photochemical
C
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
oxidation potentials for the oxime ethers indicate that the free
energy of electron transfer (ΔGET) is negative in all cases,
which makes the PET pathway favorable, although none of the
reactions are expected to be diffusion-controlled (Table 2).17
Plotting the peak potentials (Ep) of the oxime ethers against σ+
gives a reasonable correlation (Figure 3), although there is
Figure 1. Hammett plot of PET cyclization reaction of substituted 2′-
arylbenzaldehyde oxime ethers.
reactions of oxime esters studied by Rodriguez14 and Walton16
were proposed to proceed via iminyl radical intermediates. A
plot of the data points from Walton’s study against σ+ shows a
line with a slope that is close to zero, consistent with a neutral
radical intermediate such as the iminyl radical.15 Deb and
Yoshikai10 studied a transition-metal-catalyzed cyclization
reaction of O-acetyl oximes and proposed a Friedel−Crafts-
type mechanism. The slope of the data points is in agreement
with the formation of an intermediate with a developing
positive charge. However, comparing the data sets (Figure 2), it
is obvious that in our system a different mechanism operates.
Further support for an electron transfer mechanism comes
from electrochemical and computational data. The measured
Figure 3. Correlation of measured peak potentials (Ep) and calculated
ionization potentials (IP) with σ+ for a series of substituted 2′-
arylbenzaldehyde O-methyl oxime ethers.
more scatter as compared to the data plotted in Figure 1.
Previously, we have used calculated ionization potentials (IP) as
an alternative method for correlating the reactivity of
substituted oximes in PET reactions.5,6 Plotting the calculated
ionization potentials (AM1, kcal/mol)21,22 against σ+ also
shows a reasonable correlation (Figure 3), supporting the
electron transfer mechanism.
On the basis of these results, we propose a mechanism for
the conversion of 2′-arylbenzaldehyde oxime ethers to the
corresponding phenanthridines in which the initial step
involves oxidation of the oxime ether to the corresponding
radical cation intermediate (Scheme 3). Nucleophilic attack by
the aromatic ring onto the nitrogen of the oxime ether radical
cation leads to the formation of a resonance-stabilized distonic
radical cation, which can be represented as a structure
containing a Wheland-type cation and an α-amino radical.
Work by Gould and co-workers has shown that these types of
α-amino radicals undergo very rapid N−O bond homolysis.23
This reaction would produce a cyclized cation and a methoxy
radical, which is reduced by the DCA radical anion to
regenerate the photosensitizer. The methoxy anion is converted
to methanol by abstraction of a proton from the cyclized cation
intermediate, which generates the final phenanthridine product.
This proposed mechanism is consistent with the catalytic role
of DCA as it is regenerated in the reaction. Furthermore, NMR
spectra of the crude product mixtures clearly show the
formation of methanol as a product in the reaction.
The Hammett data (Figure 1) show a concave downward
trend for the oxime ethers with strong electron-donating
substituents (1h, 1l). This type of trend has traditionally been
associated with mechanistic pathways where there is a change in
the rate-determining step.20 This explanation is consistent with
Figure 2. Hammett plots of data from our study compared to data
from previous studies.10,15
D
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 3. Proposed Mechanism for the Formation of Phenanthridines from the 2′-Arylbenzaldehyde O-Methyl Oxime Ether
Radical Cation Intermediate
Scheme 4. Change in Rate-Limiting Step as a Function of Substituent in the PET Reactions of 2′-Arylbenzaldehyde O-Methyl
Oxime Ethers
our proposed mechanism (Scheme 3). Assuming that N−O
homolysis (loss of the methoxy radical) and loss of the proton
in the last (re-aromatization) step are both fast, we can simplify
the mechanism as shown in Scheme 4. Step 1 is the rate-
limiting step for compounds appearing on the right side of
Figure 1 (σ+ > −0.3), whereas step 2 is rate-limiting for
compounds with strong electron-donating substituents on the
left side of Figure 1 (σ+ < −0.3). For the oxime ethers with
electron-withdrawing and weakly electron-donating substitu-
ents, step 1 (electron transfer) is rate-limiting. Electron transfer
is much less favorable for compounds with electron-with-
drawing substituents (as seen by the higher Ep and IP values)
and raises the energy of the radical cation intermediate. The
high-energy intermediate then reacts rapidly via an internal
nucleophilic attack by the aromatic ring on the nitrogen atom
of the oxime ether. When the substituents are strongly electron-
donating (4-OCH3 and 4-OPh), significantly lower oxidation
and ionization potentials are observed and step 1 becomes
more favorable. Additionally, these substituents stabilize the
radical ion intermediate, making nucleophilic attack by the
aromatic ring (cyclization) less favorable, and as a result, step 2
becomes rate-limiting.
An alternative explanation would be that the decreased
reactivity is a result of changes in the charge distributions in the
radical cation as a function of the substituents. For the oxidative
cyclization reaction to occur, a significant positive charge
development is expected on the nitrogen atom. Therefore, if
the charge development on the nitrogen decreases as a result of
substituent effects elsewhere in the molecule, the aryl ring is
less likely to attack the nitrogen and form the cyclized product.
To investigate this possibility, we focused on the calculated
(AM1) charge densities in the neutral and radical cation species
of oxime ethers 1h and 1i. The AM1 program calculates atomic
electrostatic charges (ESP) and natural atomic charges; both
were used for this analysis and were found to give similar
results. The data presented and discussed here are an average of
the differences in charge distribution between the neutral and
the radical cation as calculated by the two methods. A summary
of these data is listed in Table 3; the complete data set can be
found in the Supporting Information. For compound 1i, it can
Table 3. Summary of Calculated (AM1) Changes in Atomic
Charge Distributions for Non-Hydrogen Atoms between the
Neutral and Radical Cation Structures of Oxime Ethers 1h
and 1i (Atom Labels Refer to Figure 4)
oxime ether 1h
oxime ether 1i
a
b
a
b
%Δ
%Δ
avg
1%
%Δ
%Δ
avg
2%
C1
1%
3%
5%
7%
1%
1%
3%
2%
C4
2%
3%
−4%
14%
5%
−2%
13%
4%
−3%
14%
4%
C2
2%
3%
C6
4%
5%
C5
2%
1%
20%
17%
4%
14%
11%
1%
17%
14%
3%
C3
−17%
0%
−10%
1%
−14%
1%
C7
N1
O1
C8
−5%
3%
−6%
3%
−5%
3%
9%
11%
8%
10%
9%
9%
0%
−1%
20%
12%
−1%
4%
−1%
22%
16%
0%
−1%
−13%
5%
−3%
−9%
4%
−2%
−11%
4%
C9
23%
20%
2%
C10
C11
C12
C13
C14
O2
C15
2%
1%
1%
9%
7%
2%
1%
1%
0%
3%
1%
2%
1%
1%
7%
7%
7%
2%
1%
1%
10%
2%
13%
−4%
12%
−1%
−1%
−1%
−1%
a
b
Atomic electrostatic charges (ESP). Natural atomic charges.
E
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
be seen that, in going from the neutral oxime ether to the
radical cation, the positive charge development increases on
carbon atoms 2 (14%), 3 (14%), and 5 (17%) as well as on the
oxime ether nitrogen atom (10%) (Table 3 and Figure 4). The
studied (1b−1e) favored cyclization to form the corresponding
2′-phenanthridine, as shown by comparison of the reaction
products (2b−2e) to published NMR spectra of both 2′- and
4′-phenanthridines.24 According to the proposed mechanism,
the cyclic carbocation intermediates leading to the two
regioisomers have the same degree of stabilization (by
resonance), and therefore, we speculate that the observed
regiochemistry is due to steric effects in the transition state; the
regioisomer that gives the lowest energy conformation to create
minimal steric hindrance effects is favored. To test this
hypothesis, compound 1m (3-F) was prepared and studied.
The van der Waals radius of fluorine (147 pm) is smaller and
closer in size to hydrogen (120 pm) than most other groups or
elements,29 and if steric effects are important in this reaction, a
smaller substituent would be expected to result in the formation
of both regioisomers. Analysis of the NMR spectrum of the
cyclic product obtained from the PET reaction of 1m indicates
the presence of both isomers in approximately a 9:1 ratio
(2′:4′). The fact that a smaller substituent leads to the
formation of both isomers (albeit in different amounts) lends
some support to the steric argument. It should be noted that
the transition-metal-catalyzed cyclization reactions of O-acetyl
oximes derived from 2′-arylacetophenones studied by Yoshikai
did not display this type of regioselectivity when using meta-
substituted derivatives but rather gave 1:1 mixtures of the two
isomers.10
Figure 4. Optimized (AM1) structure of the radical cation of oxime
ether 1h with atom numbers for calculated atomic charge distributions
listed in Table 3. Atom numbers for 1i are identical with the exception
of oxygen atom 2.
In addition to the pathways that lead to product formation
(phenanthridine, aldehyde, or nitrile), several substrates,
including 3- and 4-NO2 (1f and 1k), showed E/Z isomerization
as a dominant pathway. The extent of isomerization was
determined in the presence and in the absence of the sensitizer.
From Table 2, it can be seen that the parent oxime ether (1a)
shows little isomerization in the absence of DCA at 420 nm
(blank). At shorter wavelengths (254, 300, or 350 nm),
significant isomerization occurred within 1 h of photolysis. The
new isomer did not convert back upon prolonged exposure to
light or upon standing for several days. To investigate whether
the cyclization was preceded by isomerization, a sample of 1a
was exposed to 300 nm wavelength light for 1 h in the absence
of a sensitizer, followed by exposure to 420 nm wavelength
light in the presence of DCA for 4 h. When 1a was photolyzed
at 300 nm in the absence of a photosensitizer, the isomer ratio
changed from 99:1 to 30:70. Sensitized (DCA) photolysis of
this mixture for 4 h at 420 nm gave a 29% yield of 2a. The
isomer ratio after the second photolysis stage was further
reduced to 3:97.
Experiments carried out for several of the substituted oxime
ethers gave similar results. These results show that photo-
isomerization is a competing pathway that may prevent the
desired oxidative cyclization from taking place. Oxime ethers 1f,
1g, and 1k showed significant isomerization when irradiated in
the absence of a photosensitizer. Of these three compounds,
only 1g gave a reasonable amount of the desired cyclized
product. Isomerization can also occur at 420 nm wavelength in
the presence of DCA, which in some cases is thought to be
responsible for the poor cyclization yields. These results suggest
that in some cases isomerization leads to an unreactive oxime
ether isomer that cannot participate in the desired cyclization
reaction.30,31
majority of the positive charge development is found to occur
on the benzaldehyde ring and the oxime ether moiety, which is
consistent with the observation that the aryl ring acts as a
nucleophile and forms the cyclized product in good yield. A
similar analysis of compound 1h shows a different charge
distribution upon loss of an electron in the oxime ether. The
majority of the positive charge development occurs on carbon
atoms 9 (22%) and 10 (16%), as well as oxygen atom 2 (12%).
The nitrogen atom of the oxime ether moiety has only 2% of
the total charge development. These results clearly show that
para-substituents such as OCH3 and OPh attached to the aryl
ring cause a shift in the oxidation “location” away from the
oxime ether moiety. As a result, nucleophilic attack by the aryl
ring onto the nitrogen atom is less likely to happen and the
product yield decreases.
The oxidative cyclization pathway of 2′- arylbenzaldehyde O-
methyl oximes shows an interesting regioselectivity. Biaryl
oxime ethers with meta-substituted phenyl rings have the ability
to cyclize in two different positions, thus creating regioisomers
(2′- and 4′-phenanthridines; Scheme 5). In all cases where a
reasonable amount of product was formed, the NMR spectra of
the photolysis mixtures were very clean and it was possible to
determine peak splitting patterns in the crude reaction
mixtures. With the exception of compound 1f, which only
showed E/Z isomerization, each of the meta-compounds
Scheme 5. Potential Regiochemistry in the Formation of
Phenanthridines from 2′-Arylbenzaldehyde O-Methyl Oxime
Ethers
To determine the synthetic utility of the reaction, several
large-scale experiments (10−100 mL) were carried out in order
to isolate and quantify the product. The large-scale reaction of
1a gave an isolated yield of 60% of 2a, consistent with the yield
F
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
1
methanol and infused directly into the ESI ion source which was
interfaced with the time-of-flight mass analyzer in reflectron mode to
obtain high-resolution mass measurement. An internal calibrant was
used for accurate mass measurement. For GC/MS measurements, the
samples were dissolved in dichloromethane (DCM) and injected into
the GC column (DB-5) for separation. As the samples elute out of the
GC, chemical ionization with ammonia reagent gas was used for
generating the ions. These ions were measured with a time-of-flight
instrument in reflectron mode for high resolution. During the
ionization step, calibrant reagents were added to be used as internal
calibrants for the accurate mass measurement.
Oxidation Potentials. Cyclic voltammetry was used to determine
oxidation potentials of the oxime ethers dissolved in acetonitrile with
tetraethylammonium perchlorate (0.1 M) as the electrolyte. Solutions
were degassed under an argon atmosphere and scanned at a rate of
100 mV/s against a Ag/AgCl reference electrode.
determined by H NMR (Table 2). Large-scale reactions of
oxime ethers 1d and 1g gave yields of 49 and 82% for 2d and
2g, respectively, in reasonable agreement with the yields
1
determined by H NMR. The chemical yields for the reactions
described here vary from 12 to 94%, and although these
numbers leave much to be desired, it is worth pointing out that
these reactions are useful for a variety of reasons. Specifically,
the synthetic utility of these reactions arises from the fact that
under these conditions (a) aldehyde-based oxime ethers do not
generate the undesirable nitrile byproducts; (b) the reaction
can be accomplished using only a catalytic amount (2 mol %)
of the photosensitizer; and (c) the reaction with meta-
substituents on the aryl ring is regiospecific. None of these
characteristics is available in similar reactions currently used to
prepare phenanthridines.
General Procedure for Suzuki−Miyaura Cross-Coupling
Reactions.11,12 An oven-dried 40 mL vial fitted with a screw-top
septa was charged with the boronic acid (1 equiv), PdCl2 (0.05 equiv),
PPh3 (0.15 equiv), and a stirring magnet. The system was evacuated
and backfilled with argon three times prior to the addition of DMF
(5−10 mL), the aryl bromide (1 equiv), and degassed 2 M Na2CO3 (2
equiv) via syringe while stirring. Vials were placed in an oil bath and
stirred at 110 °C for >1 h. Reaction progress was monitored by TLC,
and the heating ceased once conversion to product had occurred. The
mixture was cooled to room temperature before deionized H2O was
added and filtered through Celite, after which it was rinsed with ether.
Additional water and diethyl ether were added to separate the aqueous
and organic layers. The organic layer was washed with water (3 × 40
mL), brine (3 × 40 mL), and 5% LiCl (2 × 40 mL), while the aqueous
layer was washed with diethyl ether (2 × 40 mL). Solvent was
removed by concentrating in vacuo, and the crude product was
purified via dry column chromatography (ether/hexane). The purified
product was then concentrated by removing solvent in vacuo and dried
In summary, photolysis of 2′-arylbenzaldehyde oxime ethers
in the presence of DCA resulted in a regioselective oxidative
cyclization reaction. The method uses only a catalytic amount
of the photosensitizer, and unlike other similar methods, it did
not generate the undesirable nitrile product. A Hammett study
showed the involvement of a species with a significant amount
of positive charge, and the reaction is proposed to proceed via
an oxime ether radical cation intermediate, which undergoes
nucleophilic attack by the built-in nucleophile (aromatic ring).
A mechanistic analysis revealed a concave downward Hammett
plot, which can be explained in terms of a change in the rate-
limiting step upon changing the substituents on the 2′-aryl
group from electron-withdrawing to electron-donating. Alter-
natively, computational data suggest that strong electron-
donating groups such as OCH3 and OPh cause a shift in the
positive charge distribution away from the oxime ether moiety,
decreasing the electrophilic character of the nitrogen atom as
well as the nucleophilic character of the aryl ring, leading to a
lower yield of the cyclized product. The mechanistic
interpretation is consistent with the substituent effect, electro-
chemical data, and computational ionization potentials. The
reaction is regiospecific for meta-substituted oxime ethers,
favoring cyclization to form only the corresponding 2′-
phenanthridine, which is believed to be due to a steric effect.
E/Z isomerization of the oxime ethers was found in some cases
to generate unreactive isomers, limiting the cyclization pathway.
Further studies to explore the scope and limitations of systems
like these are currently underway.
1
using a high vacuum equipped with a solvent trap. H NMR spectra
were consistent with those reported in literature.12,32−39
Biphenyl-2-carbaldehyde O-tert-Butyl Oxime. N-tert-Butoxyph-
thalimide (364.8 mg, 1.6 mmol, 1 equiv) and methylhydrazine (86.2
mg, 1.6 mmol, 1 equiv) reacted in 7 mL DCM for a total of 15 h.
Biphenyl-2-carbaldehyde (234.0 mg, 1.6 mmol, 1 equiv) was dissolved
in 4 mL ethanol. This solution and glacial acetic acid (101.8 mg, 1.6
mmol, 1 equiv) were added to reaction mixture according to the above
O-tert-butyl oximation procedure. The reaction was stirred at 70 °C for
a total of 24 h to afford the desired product as a light yellow oil (297.9
1
mg, 92%): H NMR (400 MHz, CDCl3) δ 1.34 (s, 9 H), 7.29−7.48
(m, 8 H), 8.01−8.08 (m, 1 H), 8.04 (s, 1 H); 13C NMR (101 MHz,
CDCl3) δ 27.6, 79.1, 125.9, 127.3, 127.4, 128.3, 129.0, 129.7, 130.3,
130. 7, 139.9, 141.9, 146.2; HRMS (CI) m/z calcd for C17H19NO [M
+ 1] 254.1545, found 254.1540.
EXPERIMENTAL SECTION
4′-Phenoxybiphenyl-2-carbaldehyde. 4-Phenoxyphenylboronic
acid (1.2865 g, 6 mmol, 1 equiv), 2-bromobenzaldehyde (700 μL, 6
mmol, 1 equiv), and aqueous 2 M Na2CO3 (6 mL, 12 mmol, 2 equiv)
reacted in the presence of PdCl2 (118.4 mg, 0.60 mmol, 0.10 equiv)
and PPh3 (480.6 mg, 1.80 mmol, 0.30 equiv) while dissolved in 30 mL
of DMF. The reaction was stirred at 110 °C for 5 min before
purification via column chromatography (ether/hexane) to afford the
■
General Information. All chemicals were obtained from
commercial sources and used as supplied without further purification
unless otherwise noted. 9,10-Dicyanoanthracene was recrystallized
from pyridine at 80 °C. All aqueous bases were prepared by dissolving
the appropriate amount of solid by mass into a volumetric flask before
being diluted with distilled water and degassed with heat and agitation
under vacuum. All GC/MS measurements were performed with a DB-
5MS column (30 m × 0.25 mm × 0.25 μm) coupled to a quadrupole
mass selective detector. Nuclear magnetic resonance spectra were
recorded on a 400 MHz instrument and obtained with 1H decoupling.
The data are reported as follows: chemical shift (multiplicity, coupling
constant in Hz, integration). Multiplicities are abbreviated as follows: s
= singlet, d = doublet, t = triplet, q = quartet, quint = quintet, sept =
septet, m = multiplet, and br = broad. All 1H NMR spectra are
reported in δ (ppm) units and are relative to residual solvent signals of
CDCl3 (7.26 ppm) or CD3CN (1.94 ppm). All 13C NMR spectra are
reported in δ (ppm) units and are relative to residual CDCl3 (77.0
ppm) solvent signals. HRMS measurements were done on one of two
different instruments. ESI-MS measurements were carried out using
direct infusion electrospray ionization. The samples were dissolved in
1
desired product as a white solid (1.5236 g, 93%): mp 69−71 °C; H
NMR (400 MHz, CDCl3) δ 7.07−7.14 (m, 4 H), 7.17 (tt, J = 7.40,
1.13 Hz, 1 H), 7.32−7.44 (m, 4 H), 7.44−7.54 (m, 2 H), 7.65 (td, J =
7.53, 1.51 Hz, 1 H), 8.03 (ddd, J = 7.78, 1.52, 0.75 Hz, 1 H), 10.04 (d,
J = 0.75 Hz, 1 H); 13C NMR (101 MHz, CDCl3) δ 118.3, 119.4, 123.9,
127.7, 127.7, 129.9, 130.8, 131.5, 132.4, 133.6, 133.8, 145.3, 156.6,
157.8, 192.4; HRMS (CI) m/z calcd for C19H14O2 [M + NH4]
292.1338, found 292.1344.
O-tert-Butyl Oximation. A 100 mL round-bottom flask equipped
with a stirring magnet was charged with N-tert-butoxyphthalimide (1
equiv) and 10 mL of dichloromethane. After dissolution, methylhy-
drazine (1 equiv) was added and a white precipitate formed. The
mixture was stirred at room temperature for >12 h. The aldehyde or
ketone (1 equiv) was dissolved in 4 mL of ethanol and was added to
G
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
the reaction mixture, followed by the addition of glacial acetic acid (1
equiv). The mixture was stirred and heated to reflux for >12 h.
Reaction progress was monitored by TLC, and the heating was ceased
once conversion to product occurred. The mixture was cooled to room
temperature, and 5 mL of 1 M aq HCl was added. The aqueous layer
was extracted with dichloromethane (3 × 20 mL). The combined
organic layers were washed with saturated 20 mL aq sodium
bicarbonate (NaHCO3). The aqueous layer was re-extracted with 20
mL of DCM. Organic extracts were dried using anhydrous sodium
sulfate (Na2SO4), and solvent was removed by concentrating in vacuo.
The crude product was dried using a high vacuum equipped with a
solvent trap. In some cases, products were purified by dry column
chromatography (EtOAc/hexane eluent). Products were characterized
by gas chromatography mass spectrometry (GC-MS) and nuclear
magnetic resonance spectroscopy (1H NMR and 13C NMR).
at 85 °C for a total of 1 h to afford the desired product as a white solid
(145.0 mg, 92%): mp 58−59 °C; 1H NMR (400 MHz, CDCl3) δ 3.84
(s, 3 H), 3.96 (s, 3 H), 6.83−6.98 (m, 3 H), 7.31−7.46 (m, 4 H), 7.97
(dd, J = 7.78, 1.51 Hz, 1 H), 8.06 (s, 1 H); 13C NMR (101 MHz,
CDCl3) δ 55.2, 61.9, 113.0, 115.3, 122.2, 125.9, 127.6, 129.3, 129.5,
129.7, 130.0, 140.9, 141.9, 147.8, 159.4; HRMS (CI) m/z calcd for
C15H15NO2 [M + Na] 264.1000, found 264.0998.
3′-Methylbiphenyl-2-carbaldehyde O-Methyl Oxime (1d). 3′-
Methylbiphenyl-2-carbaldehyde (390.4 mg, 2 mmol, 1 equiv) and
methoxylamine hydrochloride (334.1 mg, 4 mmol, 2 equiv) reacted in
the presence of NaOAc (736.4 mg, 9 mmol, 4.5 equiv) while dissolved
in 40 mL of 50% ethanol in water. The reaction was stirred at 85 °C
for a total of 1 h to afford the desired product as a light yellow oil
1
(291.2 mg, 65%): H NMR (400 MHz, CDCl3) δ 2.04 (s, 3 H), 3.90
(s, 3 H), 7.10 (dd, J = 7.28, 1.00 Hz, 1 H), 7.14−7.31 (m, 4 H), 7.32−
7.43 (m, 2 H), 7.73 (s, 6 H), 7.96−8.01 (m, 1 H); 13C NMR (101
MHz, CDCl3) δ 21.4, 61.9, 125.9, 126.8, 127.4, 128.1, 128.2, 129.4,
129.6, 130.2, 130.3, 137.9, 139.5, 142.3, 147.9; HRMS (CI) m/z calcd
for C15H15NO [M + Na] 248.1051, found 248.1049.
1-([1,1′-Biphenyl]-2-yl)ethanone O-tert-Butyl Oxime. N-tert-Bu-
toxyphthalimide (453.7 mg, 2 mmol, 1 equiv) and methylhydrazine
(182.2 mg, 2 mmol, 1 equiv) reacted in 5 mL of DCM for a total of 15
h. 1-([1,1′-Biphenyl]-2-yl)ethanone (406.8 mg, 2 mmol, 1 equiv) was
dissolved in 4 mL of ethanol. This solution and glacial acetic acid
(131.3 mg, 2 mmol, 1 equiv) were added to reaction mixture according
to the above O-tert-butyl oximation procedure. The reaction was
stirred at 70 °C for a total of 24 h to afford the desired product as a
3′-(Trifluoromethyl)biphenyl-2-carbaldehyde O-Methyl Oxime
(1e). 3′-(Trifluoromethyl)biphenyl-2-carbaldehyde (427.4 mg, 1.71
mmol, 1 equiv) and methoxylamine hydrochloride (288.3 mg, 3.42
mmol, 2 equiv) reacted in the presence of NaOAc (631.4 mg, 7.7
mmol, 4.5 equiv) while dissolved in 20 mL of 50% ethanol in water.
The reaction was stirred at 85 °C for a total of 7 h to afford the desired
1
clear oil (495.9 mg, 92%): H NMR (400 MHz, CDCl3) δ 1.30 (s, 9
H), 1.70 (s, 3 H), 7.29−7.48 (m, 9 H); 13C NMR (101 MHz, CDCl3)
δ ppm 16.4, 27.7, 78.1, 127.0, 127.2, 128.2, 128.4, 129.0, 129.3, 130.2,
138.0, 140.5, 141.5, 156.0; HRMS (CI) m/z calcd for C18H21NO [M +
Na] 290.1521, found 290.1528.
1
product as a light yellow oil (305.5 mg, 64%): H NMR (400 MHz,
CDCl3) δ 3.96 (s, 3 H), 7.29−7.35 (m, 1 H), 7.40−7.49 (m, 4 H),
7.70 (d, J = 8.28 Hz, 2 H), 7.97 (s, 1 H), 7.99 (dd, J = 6.78, 1.25 Hz, 1
1
H); 13C NMR (101 MHz, CDCl3) δ 61.9, 124.0 (q, JCF = 272 Hz),
O-Methyl Oximation. A 40 mL vial fitted with a screw-top septa
and equipped with a stirring magnet was charged with the aldehyde (1
equiv), methoxylamine hydrochloride (CH3ONH2·HCl, 98%, 2
equiv), and solid sodium acetate (NaOAc, 4.5 equiv) or sodium
hydroxide (NaOH, 4.5 equiv) before dissolution in 20 mL of 50%
ethanol in water. The mixture was stirred and heated at 85 °C.
Reaction progress was monitored by TLC, and the heating was ceased
once conversion to product occurred. The mixture was cooled to room
temperature and extracted with dichloromethane (3 × 20 mL).
Solvent was removed by concentrating in vacuo, and the crude product
was dried using a high vacuum equipped with a solvent trap. In some
cases, products were purified by dry column chromatography (ether/
hexane). Products were characterized by GC-MS and nuclear magnetic
resonance spectroscopy (1H NMR and 13C NMR).
Biphenyl-2-carbaldehyde O-Methyl Oxime (1a). Biphenyl-2-
carbaldehyde (352.3 mg, 1.9 mmol, 1 equiv) and methoxylamine
hydrochloride (579.3 mg, 6.9 mmol, 3.5 equiv) reacted in the presence
of NaOH (9 mmol, 4.5 equiv) while dissolved in 13 mL of 38%
ethanol in water. The reaction was stirred at 85 °C for a total of 1.5 h
to afford the desired product as a light yellow oil (404.3 mg, 99%): 1H
NMR (400 MHz, CDCl3) δ 3.96 (s, 3 H), 7.29−7.47 (m, 8 H), 7.98
(dd, J = 7.53, 1.51 Hz, 1 H), 8.05 (s, 1 H); 13C NMR (101 MHz,
CDCl3) δ 61.9, 126.0, 127.5, 127.5, 128.3, 129.5, 129.7, 129.7, 130.2,
139.5, 142.1, 147.9; HRMS (CI) m/z calcd for C14H13NO [M + Na]
234.0895, found 234.0903.
3
3
124.3 (q, JCF = 3.67 Hz), 125.4, 126.2 (q, JCF = 3.67 Hz), 126.4,
128.2, 128.7, 129.7, 129.8, 130.1, 130.9 (q, 2JCF = 32.28 Hz), 133.1 (q,
5JCF = 1.47 Hz), 140.4 (m), 147.0; HRMS (CI) m/z calcd for
C15H12F3NO [M + 1] 280.0949, found 280.0957.
3′-Nitrobiphenyl-2-carbaldehyde O-Methyl Oxime (1f). 3′-Nitro-
biphenyl-2-carbaldehyde (681.0 mg, 3 mmol, 1 equiv) and methoxyl-
amine hydrochloride (501.1 mg, 6 mmol, 2 equiv) reacted in the
presence of NaOAc (1.1074 g, 13.5 mmol, 4.5 equiv) while dissolved
in 40 mL of 50% ethanol in water. The reaction was stirred at 85 °C
for a total of 20 h to afford the desired product as a tan solid (406.3
mg, 53%): mp 114−116 °C; 1H NMR (400 MHz, CDCl3) δ 3.96 (s, 3
H) 7.31−7.35 (m, 1 H), 7.42−7.52 (m, 2 H), 7.59−7.68 (m, 2 H),
7.95 (s, 1 H), 7.97−8.03 (m, 1 H), 8.22 (dt, J = 1.69, 1.04 Hz, 1 H),
8.26 (dt, J = 7.28, 2.13 Hz, 1 H); 13C NMR (101 MHz, CDCl3) δ 62.0,
122.3, 124.2, 126.5, 128.6, 129.2, 129.7, 129.8, 130.0, 135.6, 139.1,
141.2, 146.5, 148.1; HRMS (CI) m/z calcd for C14H12N2O3 [M + Na]
279.0746, found 279.0742.
4′-Chlorobiphenyl-2-carbaldehyde O-Methyl Oxime (1g). 4′-
Chlorobiphenyl-2-carbaldehyde (145.1 mg, 0.66 mmol, 1 equiv) and
methoxylamine hydrochloride (111.0 mg, 1.31 mmol, 2 equiv) reacted
in the presence of NaOAc (241.8 mg, 2.95 mmol, 4.5 equiv) while
dissolved in 20 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 1 h to afford the desired product as a yellow oil
(128.9 mg, 78%): 1H NMR (400 MHz, CDCl3) δ 3.94 (s, 3 H), 7.21−
7.25 (m, 2 H), 7.26−7.30 (m, 1 H), 7.31−7.43 (m, 5 H), 7.94−7.97
(m, 1 H), 7.98 (s, 1 H); 13C NMR (101 MHz, CDCl3) δ 62.0, 126.2,
127.9, 128.5, 129.6, 129.7, 130.0, 131.0, 133.7, 138.0, 140.7, 147.4;
HRMS (CI) m/z calcd for C14H12ClNO [M + H] 246.0686, found
246.0686.
4′-Methoxybiphenyl-2-carbaldehyde O-Methyl Oxime (1h). 4′-
Methoxybiphenyl-2-carbaldehyde (128.3 mg, 0.6 mmol, 1 equiv) and
methoxylamine hydrochloride (104.3 mg, 1.2 mmol, 2 equiv) reacted
in the presence of NaOAc (224.9 mg, 2.7 mmol, 4.5 equiv) while
dissolved in 20 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 2 h to afford the desired product as a white solid
(144.3 mg, 99%): mp 49−51 °C; 1H NMR (400 MHz, CDCl3) δ 3.87
(s, 3 H), 3.96 (s, 3 H), 6.95−7.00 (m, 2 H), 7.22−7.27 (m, 2 H),
7.30−7.38 (m, 2 H), 7.39−7.45 (m, J = 1.51 Hz, 1 H), 7.95 (dd, J =
7.78, 1.51 Hz, 1 H), 8.06 (s, 1 H); 13C NMR (101 MHz, CDCl3) δ
55.3, 61.9, 113.8, 126.1, 127.2, 129.5, 129.7, 130.2, 130.8, 131.9, 141.8,
3′-Chlorobiphenyl-2-carbaldehyde O-Methyl Oxime (1b). 3′-
Chlorobiphenyl-2-carbaldehyde (147.8 mg, 0.66 mmol, 1 equiv) and
methoxylamine hydrochloride (110.2 mg, 1.31 mmol, 2 equiv) reacted
in the presence of NaOAc (241.7 mg, 2.95 mmol, 4.5 equiv) while
dissolved in 20 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 1 h to afford the desired product as a light yellow
1
oil (137.2 mg, 82%): H NMR (400 MHz, CDCl3) δ 3.95 (s, 3 H),
7.16−7.20 (m, 1 H), 7.27−7.46 (m, 6 H), 7.95−7.98 (m, 1 H), 7.98
(s, 2 H), 7.99 (s, 1 H); 13C NMR (101 MHz, CDCl3) δ 62.0, 126.2,
127.7, 128.0, 128.1, 129.5, 129.6, 129.6, 129.7, 130.1, 134.3, 140.5,
141.4, 147.2; HRMS (CI) m/z calcd for C14H12ClNO [M + Na]
268.0505, found 268.0499.
3′-Methoxybiphenyl-2-carbaldehyde O-Methyl Oxime (1c). 3′-
Methoxybiphenyl-2-carbaldehyde (139.3 mg, 0.66 mmol, 1 equiv) and
methoxylamine hydrochloride (112.2 mg, 1.3 mmol, 2 equiv) reacted
in the presence of NaOAc (252.4 mg, 3.1 mmol, 4.5 equiv) while
dissolved in 20 mL of 50% ethanol in water. The reaction was stirred
H
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
148.1, 159.1; HRMS (CI) m/z calcd for C15H15NO2 [M + Na]
264.1000, found 264.1003.
0.33 equiv), the oxime ether (15 μmol, 1 equiv), and 1 mL of
acetonitrile-d3 (CD3CN) or benzene-d6 (C6D6). The samples were
photolyzed at 420 nm using a photoreactor equipped with nine
mercury vapor lamps for a total of 4 h, unless otherwise noted. An
internal standard of either 4-nitrobenzylbromide or 4-nitrobenzalde-
hyde was added following photolysis to determine product yield by 1H
NMR. GC-MS was also used to verify product formation. Whenever
available, the product NMR spectra were compared to those available
in the literature: 2a,39 2b,25 2c,25 2d,40 2e,26 2f,27 2g,25 2h,41,42 2i,26
2j,42 and 2m.42,43
4′-Methylbiphenyl-2-carbaldehyde O-Methyl Oxime (1i). 4′-
Methylbiphenyl-2-carbaldehyde (551.8 mg, 2.8 mmol, 1 equiv) and
methoxylamine hydrochloride (460.8 mg, 5.6 mmol, 2 equiv) reacted
in the presence of NaOAc (1.0411 g, 12.6 mmol, 4.5 equiv) while
dissolved in 40 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 46 h to afford the desired product as a dark
orange oil (440.7 mg, 70%): 1H NMR (400 MHz, CDCl3) δ 2.41 (s, 3
H), 3.96 (s, 3 H), 7.19−7.26 (m, 2 H), 7.23 (d, J = 7.53 Hz, 2 H),
7.30−7.47 (m, 4 H), 7.94−7.98 (m, 1 H), 8.06 (s, 1 H); 13C NMR
(101 MHz, CDCl3) δ 21.1, 61.9, 126.0, 127.3, 129.0, 129.5, 129.6,
129.7, 130.2, 136.6, 137.2, 142.1, 148.0; HRMS (CI) m/z calcd for
C15H15NO [M + H] 226.1232, found 226.1231.
ASSOCIATED CONTENT
■
S
* Supporting Information
NMR spectra for new compounds; descriptions and results
from photolysis experiments; and tabulated data for calculated
charge distributions. This material is available free of charge via
the Internet at http://pubs.acs.org.
4′-(Trifluoromethyl)biphenyl-2-carbaldehyde O-Methyl Oxime
(1j). 4′-(Trifluoromethyl)biphenyl-2-carbaldehyde (957.0 mg, 3.83
mmol, 1 equiv) and methoxylamine hydrochloride (646.1 mg, 7.67
mmol, 2 equiv) reacted in the presence of NaOAc (1.4138 g, 17.235
mmol, 4.5 equiv) while dissolved in 40 mL of 50% ethanol in water.
The reaction was stirred at 85 °C for a total of 1 h to afford the desired
product as a tan solid (808.0 mg, 76%): mp 66−68 °C; 1H NMR (400
MHz, CDCl3) δ 3.96 (s, 3 H), 7.30−7.35 (m, 1 H), 7.40−7.52 (m, 3
H), 7.56 (t, J = 7.53 Hz, 1 H), 7.60 (s, 1 H), 7.63−7.69 (m, 1 H), 7.97
(s, 1 H), 8.00 (dd, J = 7.53, 1.76 Hz, 1 H); 13C NMR (101 MHz,
AUTHOR INFORMATION
Corresponding Author
*E-mail: pdelijser@fullerton.edu.
■
Notes
1
3
CDCl3) δ61.9, 124.1 (q, JCF = 272 Hz), 125.3 (q, JCF = 3.67 Hz),
The authors declare no competing financial interest.
2
126.3, 128.3, 129.7, 129.7 (q, JCF = 33 Hz), 129.8, 130.0, 143.2 (q,
5JCF = 1.47 Hz), 147.0; HRMS (CI) m/z calcd for C15H12F3NO [M +
1] 280.0949, found 280.0960.
ACKNOWLEDGMENTS
■
4′-Nitrobiphenyl-2-carbaldehyde O-Methyl Oxime (1k). 4′-Nitro-
biphenyl-2-carbaldehyde (733.5 mg, 3.23 mmol, 1 equiv) and
methoxylamine hydrochloride (547.0 mg, 6.46 mmol, 2 equiv) reacted
in the presence of NaOAc (1.1945 g, 14.54 mmol, 4.5 equiv) while
dissolved in 32 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 2.5 h to afford the desired product as a tan solid
This material is based upon work supported by the National
Science Foundation under Grant No. CHE-0844110.
REFERENCES
■
(1) (a) Beckmann, E. Ber. Bunsen-Ges. Phys. Chem. 1886, 19, 988.
(b) Gawley, R. E. Org. React. 1988, 35, 14−24. (c) Gregory, B. J.;
Moodie, R. B.; Schofield, K. J. Chem. Soc. (B) 1970, 338.
1
(437.3 mg, 53%): mp 113−116 °C; H NMR (400 MHz, CDCl3) δ
3.96 (s, 3 H), 7.31−7.35 (m, 1 H), 7.45−7.53 (m, 4 H), 7.95 (s, 1 H),
7.98−8.02 (m, 1 H), 8.28−8.34 (m, 2 H); 13C NMR (101 MHz,
CDCl3) δ 61.9, 123.4, 126.5, 128.6, 129.6, 129.7, 129.7, 130.1, 130.4,
139.3, 144.5, 146.2, 146.4, 147.1; HRMS (CI) m/z calcd for
C14H12N2O3 [M + H] 257.0926, found 257.0922.
(2) (a) Bieleman, J. H.; Bolle, T.; Braig, A.; Glaser, J. K.; Spang, R.;
Kohler, M.; Valet, A. In Additives for Coatings; Bieleman, J. H., Ed.;
̈
Wiley-VCH: Weinheim, Germany, 2000; pp 257−348. (b) van
Gorkum, R.; Bouwman, E. Coord. Chem. Rev. 2005, 249, 1709.
(3) Tanase, S.; Hierso, J.-C.; Bouwman, E.; Reedijk, J.; ter Borg, J.;
Bieleman, J. H.; Schut, A. New J. Chem. 2003, 27, 854.
(4) de Lijser, H. J. P.; Fardoun, F. H.; Sawyer, J. R.; Quant, M. Org.
Lett. 2002, 4, 2325.
4′-Phenoxybiphenyl-2-carbaldehyde O-Methyl Oxime (1l). 4′-
Phenoxybiphenyl-2-carbaldehyde (190.2 mg, 0.66 mmol, 1 equiv) and
methoxylamine hydrochloride (111.8 mg, 1.31 mmol, 2 equiv) reacted
in the presence of NaOAc (279.7 mg, 2.95 mmol, 4.5 equiv) while
dissolved in 20 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 1 h to afford the desired product as a white solid
(200.8 mg, 95%): mp 67−68 °C; 1H NMR (400 MHz, CDCl3) δ 3.95
(s, 3 H), 7.02−7.06 (m, J = 8.78 Hz, 2 H), 7.06−7.10 (m, 2 H), 7.14
(tt, J = 7.40, 1.13 Hz, 1 H), 7.23−7.28 (m, J = 8.78 Hz, 2 H), 7.29−
7.45 (m, 5 H), 7.95 (dd, J = 7.53, 1.51 Hz, 1 H), 8.07 (s, 1 H); 13C
NMR (101 MHz, CDCl3) δ 61.9, 118.3, 119.4, 123.6, 126.2, 127.5,
129.6, 129.7, 129.8, 130.2, 131.0, 134.3, 141.4, 147.8, 156.8, 157.1;
HRMS (CI) m/z calcd for C20H17NO2 [M + H] 304.1338, found
304.1330.
(5) (a) de Lijser, H. J. P.; Kim, J. S.; McGrorty, S. M.; Ulloa, E. M.
Can. J. Chem. 2003, 81, 575. (b) Park, A.; Kosareff, N. M.; Kim, J. S.;
de Lijser, H. J. P. Photochem. Photobiol. 2006, 82, 110.
(6) de Lijser, H. J. P.; Hsu, S.; Marquez, B. V.; Park, A.;
Sanguantrakun, N.; Sawyer, J. R. J. Org. Chem. 2006, 71, 7785.
(7) de Lijser, H. J. P.; Burke, C. R.; Rosenberg, J.; Hunter, J. J. Org.
Chem. 2009, 74, 1679.
(8) de Lijser, H. J. P.; Tsai, C. K. J. Org. Chem. 2004, 69, 3057.
(9) de Lijser, H. J. P.; Rangel, N. A.; Tetalman, M. A.; Tsai, C. K. J.
Org. Chem. 2007, 72, 4126.
3′-Fluorobiphenyl-2-carbaldehyde O-Methyl Oxime (1m). 2′-
Fluorobiphenyl-2-carbaldehyde (169.5 mg, 0.85 mmol, 1 equiv) and
methoxylamine hydrochloride (142.6 mg, 1.7 mmol, 2 equiv) reacted
in the presence of NaOAc (319.9 mg, 3.83 mmol, 4.5 equiv) while
dissolved in 20 mL of 50% ethanol in water. The reaction was stirred
at 85 °C for a total of 1.5 h to afford the desired product as a light
yellow oil (196.5 mg, 99%): 1H NMR (400 MHz, CDCl3) δ 3.97 (s, 3
H), 7.01−7.12 (m, 3 H), 7.29−7.34 (m, 1 H), 7.35−7.48 (m, 3 H),
7.95−8.01 (m, 1 H), 8.02 (s, 1 H); 13C NMR (101 MHz, CDCl3) δ
62.0, 114.4 (d, 2JCF = 20.54 Hz), 116.6 (d, 2JCF = 21.27 Hz), 125.5 (d,
(10) Deb, I.; Yoshikai, N. Org. Lett. 2013, 15, 4254 and references
therein.
(11) Miyaura, N.; Yamada, K.; Suzuki, A. Tetrahedron Lett. 1979, 20,
3437.
(12) Ye, F.; Shi, Y.; Zhou, L.; Xiao, Q.; Zhang, Y.; Wang, J. Org. Lett.
2011, 13, 5020.
(13) (a) Rhodes, C. J. J. Chem. Soc., Faraday Trans. 1990, 86, 3303.
(b) Bordwell, F. G.; Guo, Z. J. J. Org. Chem. 1992, 57, 3019.
(14) (a) Alonso, R.; Campos, P. J.; García, B.; Rodríguez, M. A. Org.
Lett. 2006, 8, 3521. (b) Alonso, R.; Caballero, A.; Campos, P. J.;
Rodríguez, M. A. Tetrahedron 2010, 66, 8828.
(15) McBurney, R. T.; Slawin, A. M. Z.; Smart, L. A.; Yu, Y.; Walton,
J. C. Chem. Commun. 2011, 47, 7974.
(16) Several ortho-substituted 2′-arylbenzaldehyde oxime ethers were
prepared and studied, as well; however, they did not produce the
desired phenanthridine products in good yields. Instead, in some cases,
3
4JCF = 2.93 Hz), 126.2, 128.0, 129.6, 129.7 (d, JCF = 6.60 Hz), 129.9,
4
3
130.0, 140.7 (d, JCF = 2.20 Hz), 141.8 (d, JCF = 7.34 Hz), 147.4,
162.6 (d, 1JCF = 247 Hz); HRMS (CI) m/z calcd for C14H12FNO [M
+ H] 230.0981, found 230.0970.
Photolysis of Oxime Ethers. Three nuclear magnetic resonance
(NMR) tubes were each charged with DCA (0.015−5 μmol, 0.001−
I
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
significant amounts of 2a were formed, suggesting an ipso-substitution
reaction. This pathway is currently being investigated.
(43) Liu, Y.-Y.; Song, R.-J.; Wu, C.-Y.; Gong, L.-B.; Hu, M.; Wang,
Z.-Q.; Xie, Y.-X.; Li, J.-H. Adv. Synth. Catal. 2012, 354, 347.
(17) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259.
(18) Gould, I. R.; Ege, D.; Moser, J. E.; Faird, S. J. Am. Chem. Soc.
1990, 112, 4290.
(19) Brown, H. C.; Okamoto, Y. J. Am. Chem. Soc. 1958, 80, 4979.
(20) (a) Leffler, J. E.; Grunwald, E. Rates and Equilibria of Organic
Reactions; John Wiley and Sons, Inc.: New York, 1963; p 189.
(b) Hart, H.; Sedor, E. A. J. Am. Chem. Soc. 1967, 89, 2342.
(21) Ionization potentials were calculated at the AM122 level of
theory using Spartan: Spartan ’08, Wave function, Inc., 18401 Von
Karman Avenue, Suite 370, Irvine, CA 92715 USA.
(22) AM1 calculations: Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.;
Stewart, J. J. P. J. Am. Chem. Soc. 1985, 107, 3902.
(23) (a) Lorrance, E. D.; Kramer, W. H.; Gould, I. R. J. Am. Chem.
Soc. 2002, 124, 15225. (b) Lorrance, E. D.; Kramer, W. H.; Gould, I.
R. J. Am. Chem. Soc. 2004, 126, 14071.
(24) Comparison of the reaction products (2b−2e) to published
NMR spectra of both 2′- and 4′-phenanthridines reveals similarities
only with 2′-phenanthridine isomers.25−28 For example, the H NMR
1
spectrum for meta-Cl-substituted phenanthridine (2b) contains a
doublet at 8.75 ppm with a 4J = 2.3 Hz, indicating that this singlet was
split by a meta-proton. A doublet of doublets at 7.78 ppm contains
coupling constants of 3J = 8.6 Hz and 4J = 2.3 Hz, suggesting that this
proton is adjacent to the C−Cl group and was split not only by its
ortho-proton but also by the meta-proton at 8.75 ppm. Other doublets
3
at 8.72, 8.20, and 8.15 ppm all displayed J values of 6.4, 7.9, and 8.8
Hz, respectively. The two triplets at 7.99 and 7.86 ppm also displayed
3
4
both J and J coupling (Figure S1).
(25) Buden, M. E.; Dorn, V. B.; Gamba, M.; Pierini, A. B.; Rossi, R.
́
A. J. Org. Chem. 2010, 75, 2206.
(26) Intrieri, D.; Mariani, M.; Caselli, A.; Ragaini, F.; Gallo, E.
Chem.Eur. J. 2012, 18, 10487.
(27) Linsenmeier, A. M.; Williams, C. M.; Brase, S. J. Org. Chem.
̈
2011, 76, 9127.
(28) Maestri, G.; Larraufie, M. H.; Derat, E.; Ollivier, C.;
Fensterbank, L.; Lacote, E.; Malacria, M. Org. Lett. 2010, 12, 5692.
̂
(29) Bondi, A. J. Phys. Chem. 1964, 68, 441.
(30) It has been reported that the photosensitized isomerization of
the carbon−nitrogen double bond may be an oxygen-sensitive
process.31 To determine whether the presence of oxygen negatively
affected the desired oxidative cyclization pathway (by promoting the
isomerization pathway), experiments were also carried out in the
absence of oxygen. An argon-purged solution of the 3-NO2 (1f)
derivative was found to give the same yield of the phenanthridine
product 2f upon photolysis in the presence of DCA as the air-saturated
solution. The amount of isomerization was found to be slightly greater
in the absence of air (60:40 instead of 70:30).
(31) Kawamura, Y.; Takayama, R.; Nishiuchi, M.; Tsukayama, M.
Tetrahedron Lett. 2000, 41, 8101.
(32) Chosson, E.; Rochais, C.; Legay, R.; Sopkova de Oliveira Santos,
J.; Rault, S.; Dallemagne, P. Tetrahedron 2011, 67, 2548.
(33) Iuliano, A.; Piccioli, P.; Fabbri, D. Org. Lett. 2004, 6, 3711.
(34) Motti, E.; Della Ca’, N.; Xu, D.; Piersimoni, A.; Bedogni, E.;
Zhou, Z.-M.; Catellani, M. Org. Lett. 2012, 14, 5792.
(35) Chamoin, S.; Houldsworth, S.; Kruse, C. G.; Iwema Bakker, W.;
Snieckus, V. Tetrahedron Lett. 1998, 39, 4179.
(36) Dupuis, C.; Adiey, K.; Charruault, L.; Michelet, V.; Savignac, M.;
̂
Genet, J.-P. Tetrahedron Lett. 2001, 42, 6523.
(37) Zhao, J.; Yue, D.; Campo, M. A.; Larock, R. C. J. Am. Chem. Soc.
2007, 129, 5288.
(38) Sindelar, M.; Lutz, T. A.; Petrera, M.; Wanner, K. T. J. Med.
Chem. 2013, 56, 1323.
(39) Read, M. L.; Gundersen, L.-L. J. Org. Chem. 2013, 78, 1311.
(40) Li, Z.; Fan, F.; Yang, J.; Liu, Z.-Q. Org. Lett. 2014, 16, 3396.
(41) Li, D.; Zhao, B.; LaVoie, E. J. J. Org. Chem. 2000, 65, 2802.
(42) Portela-Cubillo, F.; Scott, J. S.; Walton, J. C. J. Org. Chem. 2008,
73, 5558.
J
dx.doi.org/10.1021/jo502324z | J. Org. Chem. XXXX, XXX, XXX−XXX