Factors influencing the regioselectivity of the oxidation of asymmetric secondary amines with singlet oxygen

Article abstract of DOI:10.1002/chem.201500121

Aerobic amine oxidation is an attractive and elegant process for the α functionalization of amines. However, there are still several mechanistic uncertainties, particularly the factors governing the regioselectivity of the oxidation of asymmetric secondary amines and the oxidation rates of mixed primary amines. Herein, it is reported that singlet-oxygen-mediated oxidation of 1° and 2° amines is sensitive to the strength of the α-C-H bond and steric factors. Estimation of the relative bond dissociation energy by natural bond order analysis or by means of one-bond C-H coupling constants allowed the regioselectivity of secondary amine oxidations to be explained and predicted. In addition, the findings were utilized to synthesize highly regioselective substrates and perform selective amine cross-couplings to produce imines.

Full text of DOI:10.1002/chem.201500121

DOI: 10.1002/chem.201500121
Full Paper
&
Amine Oxidation
Factors Influencing the Regioselectivity of the Oxidation of
Asymmetric Secondary Amines with Singlet Oxygen
Dmitry B. Ushakov,[a] Matthew B. Plutschack,[a] Kerry Gilmore,*[a] and Peter H. Seeberger*[a, b]
Abstract: Aerobic amine oxidation is an attractive and ele-
gant process for the a functionalization of amines. However,
there are still several mechanistic uncertainties, particularly
the factors governing the regioselectivity of the oxidation of
asymmetric secondary amines and the oxidation rates of
mixed primary amines. Herein, it is reported that singlet-
oxygen-mediated oxidation of 18 and 28 amines is sensitive
to the strength of the a-CÀH bond and steric factors. Estima-
tion of the relative bond dissociation energy by natural
bond order analysis or by means of one-bond CÀH coupling
constants allowed the regioselectivity of secondary amine
oxidations to be explained and predicted. In addition, the
findings were utilized to synthesize highly regioselective
substrates and perform selective amine cross-couplings to
produce imines.
Introduction
Imines are valuable synthetic intermediates with a myriad of
applications, including the synthesis of pharmaceuticals, dyes,
as well as fine and agricultural chemicals.[1] Traditional methods
of imine preparation[2] (Scheme 1a) suffer from drawbacks re-
lated to the need to remove water and the high reactivity of
aldehydes, which often result in the formation of unwanted
byproducts. For these reasons, the development of an alterna-
tive strategy for the direct oxidation of amines to imines has
attracted much interest.[3] A range of methods for the prepara-
tion of secondary imines by catalytic condensation of primary
amines with alcohols (Scheme 1b)[4] or by homocondensation
of the corresponding primary amines (Scheme 1c) are
known.[5–9] The reported procedures for the cross-coupling of
amines are generally limited to the use of anilines, sterically
hindered amines, or alkyl amines in superstoichiometric
amounts (Scheme 1d).[10] The majority of procedures for the
oxidation of primary amines do not form aldimines, because
oxidative coupling with the starting material yields imines
bearing identical R groups (Scheme 1c).
Scheme 1. Methods for the synthesis of imines from amines.
we developed a fast and clean method for the oxidation of pri-
mary and secondary amines in a flow photoreactor.[14,15] Precise
control over reaction conditions allowed for divergence in
product formation in the oxidation of primary amines. The ex-
pected oxidative coupling (Scheme 1c) occurred in CH2Cl2
(0.5m) at room temperature; however, conditions were found
(THF, 0.1m, À508C) under which nucleophilic attack of the un-
oxidized amine on the in situ-generated primary aldimine did
not occur, providing the first example of unstable primary aldi-
mines being efficiently generated from their corresponding
amines.[14] Subsequent treatment of the solution exiting the
photoflow reactor with Me3SiCN/nBu4NF or methyl
cyanoacetate resulted in the formation of a-aminonitriles[14] or
a-cyano-a-ester epoxides,[16] respectively.
The use of singlet oxygen (1O2) is an elegant, metal-free
method for the oxidation of organic compounds.[11–13] Recently,
[a] Dr. D. B. Ushakov,+ M. B. Plutschack,+ Dr. K. Gilmore,
Prof. Dr. P. H. Seeberger
Max Planck Institute of Colloids and Interfaces
Am Mꢀhlenberg 1, 14476 Potsdam (Germany)
E-mail: kerry.gilmore@mpikg.mpg.de
peter.seeberger@mpikg.mpg.de
[b] Prof. Dr. P. H. Seeberger
Institute of Chemistry and Biochemistry, Freie Universitꢁt Berlin
Arnimallee 22, 14195 Berlin (Germany)
[+] These authors contributed equally to this work.
1
Although the oxidation of amines by O2 has been shown to
be a powerful tool for their rapid functionalization,[13,14,16] fac-
tors that influence the regioselectivity of 1O2 oxidation of
mixed primary or unsymmetrical secondary amines have not
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500121.
Chem. Eur. J. 2015, 21, 1 – 8
1
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
same benzylic radical (path II). This radical can then undergo
SET with a hydroperoxy radical to form an iminium ion, which
is quickly deprotonated to yield the desired imine.[21] Accord-
ing to both these mechanisms, hydrogen peroxide is formed
as a side product, the presence of which has been observed
and utilized previously.[16]
We first examined electronic control of secondary amine oxi-
dation in an attempt to determine the relevant effects and to
clarify the nature of the process (path I or II), which is beset by
inconsistencies in the prior literature. For example, Zhao
et al.[5k] reported that the stability of the radical formed on hy-
drogen-atom abstraction by 1O2 determines the selectivity
(path II), whereas Kçnig et al.[21] found that the presence of
electron-donating and electron-withdrawing groups on the
benzyl ring increased and decreased the rate of oxidation,
respectively, which indicates path I.
Scheme 2. Selected examples of aerobic oxidation of asymmetric secondary
amines and competing oxidation between electron-rich and electron-
deficient benzylamines.
Evaluation of mechanistic path I
been studied.[17] Reports of aerobic oxidations of asymmetric
para-substituted secondary benzylamines[18] did not describe
any selectivity (Scheme 2).[13,19]
First, we explored the oxidation of N-benzyl-4-methoxybenzyl-
amine (1, Figure 1): a 5 mgmLÀ1 solution of amine 1 and
Herein, we present our findings suggesting that the rate of
1O2-mediated oxidation of primary and secondary amines is di-
rectly proportional to the bond dissociation energy (BDE) of
the CÀH bonds adjacent to the nitrogen atom. The relative
BDEs can be estimated theoretically by natural bond order
1
(NBO) analysis or experimentally by means of JCH coupling
constants obtained by NMR spectroscopy.[20] These methods
provided an estimation of CÀH bond hybridization, and with
these data we could predict the selectivity prior to oxidation.
In addition, we show that the inherent selectivity can be over-
ridden by significant steric hindrance (Scheme 1e).
Results and Discussion
Figure 1. Photooxidation of N-benzyl-4-methoxybenzylamine (1) and
relevant (CH=N) H NMR peaks.
1
The commonly accepted mechanism for the oxidation of
1
amines by O2, generated by energy transfer from an excited
1
dye, begins with the formation of an exciplex between O2 and
tetraphenylporphyrin (TPP) (0.02 mgmLÀ1, 1.5 mol%)[22] in THF
(2 mLminÀ1) was mixed with oxygen gas (2 mLminÀ1) via a T-
mixer prior to entering the 7.5 mL flow photooxidation
module at 258C.[23] A back-pressure regulator (7 bar) was
mounted at the end of the system to increase the solubility of
the lone pair of the amine (Scheme 3). Subsequent single elec-
tron transfer (SET) followed by deprotonation of the methylene
proton leads to the benzylic radical shown in Scheme 3, path I.
Alternatively, hydrogen-atom abstraction also results in the
1
oxygen gas. The reaction mixtures were analyzed by H NMR
spectroscopy immediately following solvent evaporation, as
imines are known to be readily hydrolyzed and yields of the
isolated products are strongly dependent on the purification
process.[24] The selectivity was determined by integration of
the imine protons (CH=N) to give the relative ratio of the four
oxidation products (2:3:4:5 1.00:1.14:0.94:1.04, Figure 1).[25]
Four products were observed as opposed to two as a result of
hydrolysis and cross-condensation of the initial pair of imines
(3/4) upon rotary evaporation of the crude reaction mixture in
the presence of trace amounts of water.[26]
Imines 2 and 3, formed by oxidation of the benzylamine
side, do not isomerize to imines 4 and 5,[24,27] and thus the ini-
Scheme 3. Proposed mechanisms for singlet-oxygen-mediated oxidation of
secondary amines.
&
&
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
tial ratio of photooxidation products, and hence the selectivity,
can be quantified by the sum of the integrals for the respec-
tive peaks of the product pairs (2+3 and 4+5; 1.08:1.00).[28]
Other solvents such as ethyl acetate, toluene, and CH2Cl2 af-
forded the same oxidation ratio. To determine whether the oxi-
dation proceeds via a charged benzylic intermediate (path I),
several para-substituted secondary amines were investigated
(Table 1, entries 1–4).
6) and led us to investigate the influence of BDE on oxidation
selectivity.
Ortho-functionalized asymmetric dibenzylamines: a
combination of electronic and steric effects
Further investigation of substitution of the aromatic ring
revealed a good degree of selectivity for ortho-substituted
asymmetric dibenzylamines (62–69%; Table 1, entries 7–9).[31]
Moreover, the photooxidation of ortho-disubstituted aromatics
resulted in almost complete selectivity for oxidation of the un-
substituted benzyl side (90–95%; Table 1, entries 10–12).
Table 1. Correlation of A-side oxidation with the difference in %s
character.[a]
Since inductive effects are only observed over short distan-
ces and electronegative substituents increase neighboring
BDEs,[32] we hypothesized that the BDE of the CÀH bond is the
main determinant of selectivity. An NBO analysis of the sub-
strate scope was undertaken in an attempt to quantify this se-
lectivity trend.[33] A general trend was observed in which the
CÀH bonds adjacent to the nitrogen atom in the most selec-
tive substrates had higher differences in hybridization (%s
character) than less selective substrates (Table 1). Hybrid orbital
theory explains that an increase in hybridization of the carbon-
centered orbital leads to a higher BDE (sp>sp2 >sp3).[34] The
calculated differences in the %s character of the two CÀH
bonds a to the nitrogen atom are summarized in Table 1.
The %s character can also be determined experimentally
from one-bond CÀH coupling constants (1JCH =5ꢁ%s).[35,36]
Entry
R
D%s character (BÀA)
% of A side [O]
NBO
1JCH/5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
4-OMePh
4-NO2Ph
4-MePh
3,4,5-OMePh
1-naphthyl
2-naphthyl
2-OMePh
2-ClPh
2-FPh
2,6-ClPh
2,6-FPh
2,3,4,5,6-FPh
Mes
tBu
0.01
0.01
0.08
52
55
52
57
52
À0.11
0.01
0.03
À0.19
0.34
0.26
0.36
0.41
0.70
0.66
0.75
0.36
À0.07
0.00
0.02
0.00
À0.01
0.36
0.44
0.42
0.76
0.88
1.14
0.01
À0.41
À0.34
À0.51
55
71[c]
62
62
>95
90
>95
81
43
1
Coupled 13C NMR spectra were used to determine the JCH
values of the two sets of CÀH bonds a to the nitrogen atom in
each substrate (Table 1). The differences in hybridization of the
CÀH bonds in the substrates indicate a strong inductive effect
on hybridization. Incorporation of electronegative groups at
the ortho position increases the electronegativity of the ipso
carbon atom, and thereby p character is withdrawn from the
adjacent benzylic carbon atom.[37] As a result, less p character
is available for the benzylic CÀH bonds, hence the observed in-
crease in s character. As the substrates examined only include
aromatic rings substituted with groups more electronegative
than hydrogen,[38] we did not observe any preferential oxida-
tion of the substituted side. Oxidation of the unsubstituted
benzyl side was always preferred, that is, the difference in %s
character is directly related to the rate of oxidation. Although
this difference was higher for the difluoro- than for the respec-
tive dichloro-substituted compound (Table 1, entries 11 and 10,
respectively), the observed ratio of imine products revealed
a lower selectivity. Che et al.[13] previously noted the impor-
tance of steric hindrance. In our case, the greater steric bulk of
chlorine in comparison with fluorine (A values: 0.43 versus
0.15)[39] can explain the observed selectivities.
[b]
leelamine
1-adamantyl
–
36
38
À0.34
[a] Full conversion was observed for all substrates. For full experimental
details, see the Supporting Information. [b] NBO analysis of this substrate
was not performed. [c] Average yield over four experiments with a range
of 4%.
The ratio of imine products obtained on photooxidation of
the p-nitro-monosubstituted dibenzylamine derivative (1.2:1,
Table 1, entry 1) was similar to the results obtained for the oxi-
dation of the mono p-methoxy derivative. In fact, the introduc-
tion of substituents in either the para or the meta position did
not lead to any significant differences in the oxidation selectivi-
ty, and oxidation of the unsubstituted benzyl side was slightly
favored. Based on the relative acidities of p-NO2- and p-OMe-
substituted compounds (pKa =10.8 and 19.1 for phenols; 20.9
and 27 for anilines),[29] we expected there to be a significant
difference, as well as a switch in selectivity, if the reaction
proceeded via path I. However, this was not observed, which
suggests that the acidity of this proton does not significantly
influence the outcome of the oxidation.
From what is known computationally of benzyl radicals,[30]
extension of the conjugated systems to naphthyl groups
makes radical delocalization less efficient, and the calculated
BDEs of 1-naphthyl and 2-naphthyl CÀH bonds are higher than
those of benzylic CÀH bonds by about 1 kcalmolÀ1. This is re-
flected in the poor selectivity observed (Table 1, entries 5 and
For a more detailed investigation of steric effects, a mesityl
derivative (Table 1, entry 13) was prepared. For this substrate,
each of the benzyl CÀH bonds has nearly the same %s charac-
ter and should therefore result in a poorly selective oxidation.
However, analysis of the photooxidation products revealed
that oxidation occurred predominantly (81%) at the less hin-
dered side (A value of CH3 is 1.7),[39] demonstrating that steric
effects play an important role in selectivity as well.
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Changing the regioselectivity
tylamine with a mesityl group (Table 2, entry 4). Photooxidation
now occurred predominantly on the alkyl side with 88% selec-
tivity. To the best of our knowledge, these results are the first
examples of selective imine formation as the result of amine
oxidation away from the benzylic position of a disubstituted
amine.
On the basis of the above results, we hypothesized that by ex-
changing one of the aromatic rings with a group that is less in-
ductively withdrawing, such as an sp3-hybridized alkyl group,
the regioselectivity of oxidation could be tuned away from the
benzylic position. However, efforts to study aliphatic amines
were challenging due to the instability of the alkyl-substituted
imines formed after oxidation,[24] which prevented quantifica-
tion of the reaction selectivity.[40]
A plot of the oxidation selectivity versus the difference in %s
character is shown in Figure 2.[43] Substrates with little influ-
ence over the selectivity occupy the middle of the graph,
whereas those with groups that exhibit a slow rate of oxida-
These issues are avoided when a quaternary carbon atom is
in the b position of the alkyl chain.[41] Oxidation of secondary
amines bearing benzyl and neopentyl residues (Table 1, en-
tries 14–16) did not result in high selectivity towards the
benzyl side, as has been reported in other oxidation systems.[42]
Conversely, we observed an inversion in selectivity, with
a slight preference for oxidation of the alkyl side (57–64%).
These results were supported by the estimated differences in
%s character, obtained by NBO analysis or 13C NMR spectrosco-
py (Table 1). The negative differences for these substrates re-
flects the stronger (higher %s character) CÀH bonds of the
benzylic side than the alkyl side, which leads to the predicted
preference for oxidation away from the benzyl group.
Design of highly selective substrates
The differences in %s character could be used to design even
more selective substrates. For example, an increase in the per-
centage of oxidation at the neopentyl side could be obtained
by exchanging the benzyl group (D%s character: À0.07;
Table 1, entry 14) with a more inductively withdrawing group
such as 2-fluorobenzyl or 2,6-diflurobenzyl (D%s character:
À0.44 and À0.72, respectively, Table 2, entries 1 and 2). The
latter (Table 2, entry 2) resulted in almost complete selectivity
(92%) for oxidation on the alkyl side. Moreover, photooxida-
tion of the dichloro derivative (Table 2, entry 3) resulted in se-
lective oxidation of the alkyl side, with only trace amounts of
benzyl-side oxidation. These results are in complete agreement
with the described stereoelectronic effects (Table 1).
Figure 2. Percentage A-side oxidation (see Tables 1 and 2) plotted versus the
difference in %s character between the CÀH bonds on the A and B sides, as
determined experimentally. Labels for the data points x.y represent Table x,
entry y. The plot (only diamonds) was fitted by means of a linear fit with the
equation y=34.3x+53.6. The uncertainty due to NMR integration is repre-
sented by the error bars (see Supporting Information for equation). Entries
for which steric hindrance has a significant impact are shown in circles (see
Figure 3).
tion relative to benzyl lie in the upper right corner. Those with
increased rates[44] resulting in selectivity towards the alkyl side
lie in the bottom left. We propose that the plot in Figure 2 can
be used as a predictive model for asymmetric secondary
amines bearing substituents that lack significant steric bulk. In
addition, we believe that relative steric effects can be account-
ed for in structurally similar groups (2,6-F/Cl/Me versus H)
according to their respective A values.[39]
To further illustrate the influence of steric hindrance on
selectivity, we replaced the benzyl group of N-benzylneopen-
Table 2. Electronic and steric effects enhance oxidation selectivity away
from an aromatic ring.[a]
The data points shown in circles (Figure 2) were not used for
the linear fit due to the influence of steric bulk on the reaction
selectivity. Figure 3 depicts the percentage of deviation of
ortho-disubstituted examples from the predictive model versus
their respective A values.[39,45] For dibenzyl amines, positive A
values are used, placing the data point in the upper right
quadrant of the plot, whereas for benzyl alkyl amines, the op-
posite of the respective A value is used, placing the data point
in the lower left quadrant. The observed linear dependence
nicely shows that steric effects complement the inherent elec-
Entry
R
D %s character (BÀA)
% of A side [O]
NBO
1JCH/5
1
2
3
4
2-FPh
À0.44
À0.72
À0.76
À0.42
À0.81
À1.35
À1.17
À0.43
29
8
<5
12
2,6-FPh
2,6-ClPh
Mes
[a] Full conversion was observed for all substrates. For full experimental
details, see the Supporting Information.
&
&
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
4
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
in 85% yield,[46] corresponding to oxidation of the unsubstitut-
ed benzyl amine (Scheme 4b).
For two amines bearing CÀH bonds with similar %s charac-
ter adjacent to the nitrogen atom, such as benzylamine (8)
and p-methoxybenzylamine (12), similar rates of oxidation
would be expected, as is evidenced by the lack of selectivity in
the corresponding secondary amine (Table 1, entry 1). Accord-
ingly, when a 1:1 mixture of 8 and 12 (each with 26.90 %s) un-
derwent photooxidation under oxidative coupling conditions
(CH2Cl2, RT) four imine products were obtained in a ratio of
1.2:1 favoring the unsubstituted side. These results were simi-
lar to those reported in Table 1 (entry 1) and were consistent
with the literature reports on oxidations of amine mixtures.[19c]
The desired secondary imine 13 could, however, be efficiently
generated from amines with similar benzylic CÀH bonds by
first oxidizing one of them (8, À508C, THF) and treating the re-
sulting primary aldimine 10 with amine 12, as shown in
Scheme 4c, which gave 92% of 13 as a pure compound fol-
lowing evaporation. Simply exchanging the substrates results
in the formation of the opposite imine 15 in 94% yield.
Figure 3. Percentage deviation from the predictive model plotted versus the
A values. Positive A values correspond to R2 groups containing sterically
bulky substituents and negative A values to R1 groups with sterically bulky
substituents. Positive deviation means the points lie above the fitted line
and negative values correspond to points that lie below the line. Labels for
the data points x.y represent Table x, entry y.
tronic influence (Figure 2), which allows for the prediction of
the regioselective outcome for secondary amine oxidations.
Selective cross-condensation of amines
Conclusion
We expected the selectivities observed above to translate into
the rates of oxidation for primary amines as well. For example,
due to the difference in their %s character (0.88), the photooxi-
dation of an equimolar mixture of 2,6-difluorobenzylamine (9,
27.78 %s) and benzylamine (8, 26.90 %s) was expected to ex-
hibit a similar degree of selectivity to that observed for the re-
spective secondary amine (90%, Scheme 4a).[46] To avoid for-
mation of N-benzylidenebenzylamine (4), which results from
the condensation between benzylamine and benzaldimine, the
photooxidation was performed at À508C in THF (conditions
for generating primary imine).[14] Imine 7 was indeed obtained
We found that the selectivity observed in singlet-oxygen-medi-
ated oxidations of primary and secondary amines is deter-
mined by the BDE of the competing CÀH bonds adjacent to
the nitrogen atom. This suggests that oxygen-mediated oxida-
tions likely proceed via a hydrogen-atom abstraction pathway
(Scheme 3, path II). NBO analysis and NMR spectroscopy can
be used as predictive tools to estimate the differences in %s
character. Good and excellent selectivity could be observed
with differences greater than 0.3 and 0.7%, respectively. In ad-
dition, it was shown that steric hindrance plays an important
role in the regioselective outcome and can overpower the in-
trinsic preferences. The predictive power of these differences
can be utilized not only for photooxidation of asymmetrical
secondary amines, but also for primary amines. Finally, to the
best of our knowledge, we have given the first accounts both
of selective imine formation away from a benzylic position as
well as the selective equimolar cross-coupling of two different
amines, without the limitation of using aromatic or sterically
hindered amines.
Experimental Section
General procedure for the photooxidation of secondary
amines[47]
A solution of amine (50 mg) and TPP (1 mg) in THF (10 mL) was
pumped by a Vapourtec R2+ pump with a flow rate of 2 mLminÀ1
.
In an ETFE T-mixer (IDEX Health and Science) the substrate solution
was mixed with oxygen (99.995%, <3.0 ppmmolÀ1 H2O; ALPHA-
GAZ 1 O2, Werk DEF 2 Krefeld-Gellep), which was delivered from an
oxygen gas tank. Gas pressure was regulated to 20 bar and the
flow adjusted to 5 mLminÀ1 with a gas-flow controller (Influx,
SV1B5-AI05). This solution was then pumped through the 7.5 mL
photoreactor consisting of fluorinated ethylene–propylene copoly-
mer (FEP) tubing (IDEX Health and Science, natural color, 1.57 mm
Scheme 4. Strategies for selective oxidative imine synthesis.[46]
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Liu, Z. Wang, X. Fu, C.-H. Yan, Org. Lett. 2012, 14, 5692–5695; m) H.
Yuan, W.-J. Yoo, H. Miyamura, S. Kobayashi, J. Am. Chem. Soc. 2012, 134,
13970–13973; n) W. P. To, G. S. M. Tong, W. Lu, C. Ma, J. Liu, A. L. F.
Chow, C. M. Che, Angew. Chem. Int. Ed. 2012, 51, 2654–2657; Angew.
Chem. 2012, 124, 2708–2711; o) S. Furukawa, Y. Ohno, T. Shishido, K.
Teramura, T. Tanaka, ACS Catal. 2011, 1, 1150–1153; p) L. Aschwanden,
T. Mallat, M. Maciejewski, F. Krumeich, A. Baiker, ChemCatChem 2010, 2,
666–673; q) A. Grirrane, A. Corma, H. Garcia, J. Catal. 2009, 264, 138–
144; r) J. Wang, S. Lu, X. Cao, H. Gu, Chem. Commun. 2014, 50, 5637–
5640; s) M. H. So, Y. Liu, C. M. Ho, C. M. Che, Chem. Asian J. 2009, 4,
1551–1561; t) X. H. Li, M. Antonietti, Angew. Chem. Int. Ed. 2013, 52,
4572–4576; Angew. Chem. 2013, 125, 4670–4674; u) N. Kang, J. H. Park,
K. C. Ko, J. Chun, E. Kim, H.-W. Shin, S. M. Lee, H. J. Kim, T. K. Ahn, J. Y.
Lee, S. U. Son, Angew. Chem. Int. Ed. 2013, 52, 6228–6232; Angew.
Chem. 2013, 125, 6348–6352; v) B. Yuan, R. Chong, B. Zhang, J. Li, Y.
Liu, C. Li, Chem. Commun. 2014, 50, 15593–15596.
outer diameter, 0.76 mm inner diameter) wrapped around a glass
plate (7ꢁ9 cm2) in two layers. A 30 cm piece of tubing was used
for connection of the T-mixer with a photoreactor. A 0.5 mL pre-
cooling loop covered with aluminum foil was placed additionally
in the bath before the photoreactor. An LED module (OSA Opto
Light, OLM-018 B, 420 nm emission wavelength, 72 W; Manson
HCS-3202 power supply) was mounted in front of this plate at
a distance of 3 cm. The photoreactor was suspended in a water
bath. A piece of FEP tubing (30 cm) was used for connection of
the outlet of photoreactor to a 7 bar back-pressure regulator (aver-
age pressure in the system was around 9 bar). The residence time
in the photoreactor was measured to be 2.5 min. Concentration on
the rotary evaporator was performed at 308C. Analysis of the prod-
uct mixture was performed by relative integration of characteristic
and non-overlapping signals of imines in 1H NMR spectra.[48]
[6] For examples using stoichiometric oxidants, see: a) Q.-L. Yuan, X.-T.
Zhou, H.-B. Ji, Catal. Commun. 2010, 12, 202–206; b) G. Chu, C. Li, Org.
Biomol. Chem. 2010, 8, 4716–4719; c) H. Choi, M. P. Doyle, Chem.
Commun. 2007, 745–747; d) X.-F. Wu, A. Petrosyan, T. V. Ghochikyan,
A. S. Saghyan, P. Langer, Tetrahedron Lett. 2013, 54, 3158–3159; e) T. K.
Achar, S. Maiti, P. Mal, RSC Adv. 2014, 4, 12834–12839.
[7] For examples of the oxidation of asymmetric secondary amines with
stoichiometric oxidants, see: a) K. Nicolaou, C. J. Mathison, T. Montag-
non, J. Am. Chem. Soc. 2004, 126, 5192–5201; b) T. Mukaiyama, A.
Kawana, Y. Fukuda, J. Matsuo, Chem. Lett. 2001, 390–391.
Acknowledgements
We gratefully acknowledge financial support from the Max
´
Planck Society and the DAAD. We thank Stella Vukelic for the
help with the synthesis of amines in entries 9 and 12 (Table 1).
We also thank Thomas Kolrep and Fabian Klautzsch (all Free
University Berlin) for assistance in preparing high-resolution
mass spectra.
[8] For a recent example of enzymatic amine oxidation, see: D. Ghislieri,
A. P. Green, M. Pontini, S. C. Willies, I. Rowles, A. Frank, G. Grogan, N. J.
Turner, J. Am. Chem. Soc. 2013, 135, 10863–10869.
[9] For recent examples of dehydrogenative imine formation, see: a) C.
Hammond, M. T. Schꢂmperli, I. Hermans, Chem. Eur. J. 2013, 19, 13193–
13198; b) L.-P. He, T. Chen, D. Gong, Z. Lai, K.-W. Huang, Organometallics
2012, 31, 5208–5211.
Keywords: imines
regioselectivity · singlet oxygen
·
oxidation
·
photochemistry
·
[10] a) A. Monopoli, P. Cotugno, F. Iannone, F. Ciminale, M. M. Dell’Anna, P.
Mastrorilli, A. Nacci, Eur. J. Org. Chem. 2014, 5925–5931; b) E. Zhang, H.
Tian, S. Xu, X. Yu, Q. Xu, Org. Lett. 2013, 15, 2704–2707; c) J. H. Park,
K. C. Ko, E. Kim, N. Park, J. H. Ko, D. H. Ryu, T. K. Ahn, J. Y. Lee, S. U. Son,
Org. Lett. 2012, 14, 5502–5505; d) M. Largeron, M. B. Fleury, Angew.
Chem. Int. Ed. 2012, 51, 5409–5412; Angew. Chem. 2012, 124, 5505–
5508; e) A. E. Wendlandt, S. S. Stahl, Org. Lett. 2012, 14, 2850–2853; f) Z.
Hu, F. M. Kerton, Org. Biomol. Chem. 2012, 10, 1618–1624; g) H. Huang,
J. Huang, Y.-M. Liu, H.-Y. He, Y. Cao, K.-N. Fan, Green Chem. 2012, 14,
930–934; h) L. Liu, S. Zhang, X. Fu, C.-H. Yan, Chem. Commun. 2011, 47,
10148–10150; i) F. Su, S. C. Mathew, L. Mçhlmann, M. Antonietti, X.
Wang, S. Blechert, Angew. Chem. Int. Ed. 2011, 50, 657–660; Angew.
Chem. 2011, 123, 683–686; j) X. Jin, Y. Liu, Q. Lu, D. Yang, J. Sun, S. Qin,
J. Zhang, J. Shen, C. Chu, R. Liu, Org. Biomol. Chem. 2013, 11, 3776–
3780; k) R. D. Patil, S. Adimurthy, Adv. Synth. Catal. 2011, 353, 1695–
1700; l) D. V. Jawale, E. Gravel, E. Villemin, N. Shah, V. Geertsen, I. N.
Namboothiri, E. Doris, Chem. Commun. 2014, 50, 15251–15254.
[11] a) E. L. Clennan, A. Pace, Tetrahedron 2005, 61, 6665–6691; b) E. L.
Clennan, Tetrahedron 2000, 56, 9151–9179.
[12] a) F. Lꢃvesque, P. H. Seeberger, Org. Lett. 2011, 13, 5008–5011; b) D. Ko-
petzki, F. Lꢃvesque, P. H. Seeberger, Chem. Eur. J. 2013, 19, 5450–5456;
c) K. Gilmore, D. Kopetzki, J. W. Lee, Z. Horvath, D. T. McQuade, A.
Seidel-Morgenstern, P. H. Seeberger, Chem. Commun. 2014, 50, 12652–
12655; d) A. Talla, B. Driessen, N. J. W. Straathof, L.-G. Milroy, L. Bruns-
veld, V. Hessel, T. Noꢄl, Adv. Synth. Catal. 2015, DOI: 10.1002/
adsc.201401010.
[13] G. Jiang, J. Chen, J.-S. Huang, C.-M. Che, Org. Lett. 2009, 11, 4568–4571.
[14] D. B. Ushakov, K. Gilmore, D. Kopetzki, D. T. McQuade, P. H. Seeberger,
Angew. Chem. Int. Ed. 2014, 53, 557–561; Angew. Chem. 2014, 126,
568–572.
[1] a) S. Kobayashi, Y. Mori, J. S. Fossey, M. M. Salter, Chem. Rev. 2011, 111,
2626–2704; b) K.-i. Yamada, K. Tomioka, Chem. Rev. 2008, 108, 2874–
2886; c) J.-H. Xie, S.-F. Zhu, Q.-L. Zhou, Chem. Rev. 2011, 111, 1713–
1760; d) J. P. Adams, J. Chem. Soc. Perkin Trans. 1 2000, 125–139.
[2] a) S. Nieto, J. M. Dragna, E. V. Anslyn, Chem. Eur. J. 2010, 16, 227–232;
b) L. R. Reddy, M. Prashad, Chem. Commun. 2010, 46, 222–224; c) R. S.
Varma, R. Dahiya, S. Kumar, Tetrahedron Lett. 1997, 38, 2039–2042; d) B.
Miriyala, S. Bhattacharyya, J. S. Williamson, Tetrahedron 2004, 60, 1463–
1471.
[3] For recent reviews on amine oxidation, see: a) M. T. Schꢂmperli, C. Ham-
mond, I. Hermans, ACS Catal. 2012, 2, 1108–1117; b) M. Largeron, Eur. J.
Org. Chem. 2013, 5225–5235; c) R. D. Patil, S. Adimurthy, Asian J. Org.
Chem. 2013, 2, 726–744; d) B. L. Ryland, S. S. Stahl, Angew. Chem. Int.
Ed. 2014, 53, 8824–8838; Angew. Chem. 2014, 126, 8968–8983.
[4] a) M. S. Kwon, S. Kim, S. Park, W. Bosco, R. K. Chidrala, J. Park, J. Org.
Chem. 2009, 74, 2877–2879; b) J. W. Rigoli, S. A. Moyer, S. D. Pearce,
J. M. Schomaker, Org. Biomol. Chem. 2012, 10, 1746–1749; c) J.-F. Soulꢃ,
H. Miyamura, S. Kobayashi, Chem. Commun. 2013, 49, 355–357; d) B.
Gnanaprakasam, J. Zhang, D. Milstein, Angew. Chem. Int. Ed. 2010, 49,
1468–1471; Angew. Chem. 2010, 122, 1510–1513; e) H. Tian, X. Yu, Q.
Li, J. Wang, Q. Xu, Adv. Synth. Catal. 2012, 354, 2671–2677; f) L. Han, P.
Xing, B. Jiang, Org. Lett. 2014, 16, 3428–3431.
[5] For recent examples of aerobic amine oxidations, see: a) A. E. Wend-
landt, S. S. Stahl, J. Am. Chem. Soc. 2014, 136, 11910–11913; b) L. Ye, Z.
Li, ChemCatChem 2014, 6, 2540–2543; c) X.-J. Yang, B. Chen, X.-B. Li, L.-
Q. Zheng, L.-Z. Wu, C.-H. Tung, Chem. Commun. 2014, 50, 6664–6667;
d) S. Zhao, C. Liu, Y. Guo, J.-C. Xiao, Q.-Y. Chen, J. Org. Chem. 2014, 79,
8926–8931; e) S. Furukawa, A. Suga, T. Komatsu, Chem. Commun. 2014,
50, 3277–3280; f) P. K. Chow, C. Ma, W. P. To, G. S. M. Tong, S. L. Lai, S. C.
Kui, W. M. Kwok, C. M. Che, Angew. Chem. Int. Ed. 2013, 52, 11775–
11779; Angew. Chem. 2013, 125, 11991–11995; g) N. Li, X. Lang, W. Ma,
H. Ji, C. Chen, J. Zhao, Chem. Commun. 2013, 49, 5034–5036; h) J. Jin,
H.-W. Shin, J. H. Park, J. H. Park, E. Kim, T. K. Ahn, D. H. Ryu, S. U. Son, Or-
ganometallics 2013, 32, 3954–3959; i) W. P. To, Y. Liu, T. C. Lau, C. M.
Che, Chem. Eur. J. 2013, 19, 5654–5664; j) B. Huang, H. Tian, S. Lin, M.
Xie, X. Yu, Q. Xu, Tetrahedron Lett. 2013, 54, 2861–2864; k) L. Huang, J.
Zhao, S. Guo, C. Zhang, J. Ma, J. Org. Chem. 2013, 78, 5627–5637; l) L.
[15] For reviews on flow photochemistry, see: a) J. P. Knowles, L. D. Elliott,
K. I. Booker-Milburn, Beilstein J. Org. Chem. 2012, 8, 2025–2052; b) Y. Su,
N. J. Straathof, V. Hessel, T. Noꢄl, Chem. Eur. J. 2014, 20, 10562–10589;
c) K. Gilmore, P. H. Seeberger, Chem. Rec. 2014, 14, 410–418.
[16] D. B. Ushakov, K. Gilmore, P. H. Seeberger, Chem. Commun. 2014, 50,
12649–12651.
[17] For the mechanism of singlet-oxygen-mediated oxidation of tertiary
amines, see: a) E. Baciocchi, T. Del Giacco, A. Lapi, Org. Lett. 2004, 6,
&
&
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
4791–4794; b) E. Baciocchi, T. Del Giacco, A. Lapi, Org. Lett. 2006, 8,
1783–1786; c) E. Baciocchi, T. Del Giacco, O. Lanzalunga, A. Lapi, J. Org.
Chem. 2007, 72, 9582–9589; d) C. Ferroud, P. Rool, J. Santamaria, Tetra-
hedron Lett. 1998, 39, 9423–9426; e) M. Matsumoto, Y. Kitano, H. Ko-
bayashi, H. Ikawa, Tetrahedron Lett. 1996, 37, 8191–8194.
[30] For computational studies on the differences in radical delocalization
for naphthyl and benzyl groups, see: Y. Sui, R. Glaser, U. Sarkar, K. Gates,
J. Chem. Theory Comput. 2007, 3, 1091–1099.
[31] When tris(bipyridine)ruthenium(II) chloride was used as a photocatalyst
for the oxidation of N-benzyl-2-methoxybezylamine (Table 1, entry 7) in
acetonitrile, the same ratio of imine products was detected. This obser-
vation suggests that our findings on the selectivity of singlet-oxygen-
mediated oxidations of secondary amines are common for other cata-
lytic systems in which oxygen anion radicals are the active species.
[32] M. J. Dewar, P. J. Grisdale, J. Am. Chem. Soc. 1962, 84, 3539–3541.
[33] The calculated hybridization values were taken as the average of the
two CÀH bonds adjacent to the nitrogen from NBO analysis of the
B3LYP/6-311+ +g(d,p) optimized structures by using Gaussian NBO
Version 3.1.
[18] In addition to the singlet-oxygen-mediated pathway, aerobic oxidations
can also occur by a different mechanism in which an excited dye (or
photocatalyst) directly oxidizes the amine to the radical cation. The
photocatalyst is regenerated by single electron transfer to 3O2, provid-
ing the same radical-ion pair as proposed in Scheme 3, see ref. [21].
[19] a) X. Lang, H. Ji, C. Chen, W. Ma, J. Zhao, Angew. Chem. Int. Ed. 2011, 50,
3934–3937; Angew. Chem. 2011, 123, 4020–4023; b) X. Lang, W. Ma, Y.
Zhao, C. Chen, H. Ji, J. Zhao, Chem. Eur. J. 2012, 18, 2624–2631; c) M.
Rueping, C. Vila, A. Szadkowska, R. M. Koenigs, J. Fronert, ACS Catal.
2012, 2, 2810–2815; d) K. N. Tayade, M. Mishra, J. Mol. Catal. A 2014,
382, 114–125.
[20] For studies on the mechanism of singlet-oxygen-mediated sulfide oxi-
dation using %s character estimated by NMR spectroscopy, see: T. Aka-
saka, M. Haranaka, W. Ando, J. Am. Chem. Soc. 1991, 113, 9898–9900.
[21] A. Berlicka, B. Kçnig, Photochem. Photobiol. Sci. 2010, 9, 1359–1366.
[22] The oxidation selectivity was found to be completely independent of
the amount of photocatalyst and solution concentration of the secon-
dary amine (see Table S3 in the Supporting Information).
[34] S. J. Blanksby, G. B. Ellison, Acc. Chem. Res. 2003, 36, 255–263.
[35] a) N. Muller, D. E. Pritchard, J. Chem. Phys. 1959, 31, 1471–1476; b) N.
Muller, D. E. Pritchard, J. Chem. Phys. 1959, 31, 768–771.
[36] For the prediction of regioselectivity of the radical halogenation using
1
ˇ
%s character estimated via JCH, see: C. Mazal, O. Skarka, J. Kaleta, J.
Michl, Org. Lett. 2006, 8, 749–752.
[37] H. Bent, Chem. Rev. 1961, 61, 275.
[38] L. Pauling, J. Am. Chem. Soc. 1932, 54, 3570–3582.
[39] E. L. Eliel, S. H. Wilen, Stereochemistry of organic compounds, Wiley, New
York, 2008.
[23] See the Supporting Information for details.
[40] Formation of 2,4-dinitrophenyl hydrazones could be used for trapping
aliphatic imines,[10d] but since the yields of the reaction with hydrazine
were not quantitative for all substrates, analysis of photooxidation
products was inaccurate and thus was not utilized in this study.
[41] The inability to form an enamine from the resulting imine presumably
prevents the side reactions that were observed when methyl, ethyl,
butyl, or phenylethyl derivatives were subjected to photooxidation.
[42] For examples of preferential oxidation of the benzylic position versus
the aliphatic position, see refs. [5e], [5k], and [21].
[24] R. W. Layer, Chem. Rev. 1963, 63, 489–510.
[25] For the chemical shifts of the formed imines utilized to determine the
product distribution, see Table S2 in the Supporting Information.
[26] Although we did not face any difficulties in analyzing the mixture of
four imines, hydrolysis of the initial pair of imines could be prevented
by using freshly prepared CaH2 powder. In this case, only traces of
cross-coupling products were detected. For the recent application of
CaH2 for gentle and efficient water removal, see: A. Zhdanko, M. E.
Maier, Organometallics 2013, 32, 2000–2006.
[27] Imine 2, prepared by condensation of p-anisaldehyde and benzylamine,
was subjected to photooxidation conditions. Subsequent concentration
under reduced pressure afforded the identical compound without any
traces of other imines. Indeed, for migration of the carbon–nitrogen
double bond (azomethine–azomethine isomerization), strongly basic
conditions (for example KOH, tBuOK or reflux with Et3N) are required;
see: B. Gangadasu, P. Narender, B. C. Raju, V. J. Rao, Indian J. Chem. Sect.
B 2005, 44, 2598–2600.
[28] When the photoreactor temperature was decreased to À858C, a slight
increase of selectivity towards the unsubstituted side (1.3:1) was ob-
served.
[29] a) F. G. Bordwell, D. Algrim, N. R. Vanier, J. Org. Chem. 1977, 42, 1817–
1819; b) F. G. Bordwell, D. J. Algrim, J. Am. Chem. Soc. 1988, 110, 2964–
2968; c) F. G. Bordwell, R. J. McCallum, W. N. Olmstead, J. Org. Chem.
1984, 49, 1424–1427.
[43] Experimentally derived values for the differences in %s character were
used. NBO-derived values can also be used (see Figure S1 in the Sup-
porting Information).
[44] Relative to benzyl.
[45] For the optimized three-dimensional structures of secondary amines
bearing sterically bulky groups, see the figures in Table S4 in the Sup-
porting Information.
[46] Yields were determined by 1H NMR spectroscopy with 1,2,4,5-tetrame-
thylbenzene as an internal standard.
[47] For the graphical scheme, see the Supporting Information.
[48] For the list of the peaks used for determining the product ratio by
1H NMR specrtroscopy, see the Supporting Information.
Received: January 12, 2015
Published online on && &&, 0000
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
7
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
Selective imine synthesis: The influ-
ence of electronic and steric factors on
the singlet-oxygen-mediated oxidation
of 18 and 28 amines was investigated.
Estimation of the relative bond dissocia-
tion energy (BDE) by natural bond order
(NBO) analysis or from one-bond CÀH
coupling constants allowed the regiose-
lectivity of secondary amine oxidations
to be explained and predicted, and the
findings were utilized to design a selec-
tive approach towards 28 imines by oxi-
dative cross-couplings of 18 imines (see
scheme).
&
Amine Oxidation
D. B. Ushakov, M. B. Plutschack,
K. Gilmore,* P. H. Seeberger*
&& – &&
Factors Influencing the
Regioselectivity of the Oxidation of
Asymmetric Secondary Amines with
Singlet Oxygen
&
&
Chem. Eur. J. 2015, 21, 1 – 8
www.chemeurj.org
8
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!