An unexpected copper catalyzed 'reduction' of an arylazide to amine through the formation of a nitrene intermediate

Article abstract of DOI:10.1016/j.tet.2013.04.091

Azido nitrobenzoxadiazole (NBD) was observed to undergo a 'reduction' reaction in the absence of an obvious reducing agent, leading to amine formation. In the presence of an excess amount of DMSO, a sulfoxide conjugate was also formed. The ratio of these two products was both temperature- and solvent-dependent, with the addition of water significantly enhancing the ratio of the 'reduction' product. Two intermediates of the azido-NBD reaction in DMSO were trapped and characterized by low-temperature EPR spectroscopy. One was an organic free radical (S=1/2) and another was a triplet nitrene (S=1) species. A mechanism was proposed based on the characterized free radical and triplet intermediates.

Full text of DOI:10.1016/j.tet.2013.04.091

Tetrahedron 69 (2013) 5079e5085
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
An unexpected copper catalyzed ‘reduction’ of an arylazide to amine
through the formation of a nitrene intermediate
,
,
Hanjing Peng a y, Kednerlin H. Dornevil a y, Alexander B. Draganova, Weixuan Chen a,
Chaofeng Dai a, William H. Nelson b, Aimin Liu a , Binghe Wang
,
a,
*
*
a Department of Chemistry, Center for Biotechnology and Drug Design, and Center for Diagnostics and Therapeutics, Georgia State University,
P.O. Box 3965, Atlanta, Georgia 30302-3965, USA
b Department of Physics and Astronomy, Georgia State University, P.O. Box 4106, Atlanta, Georgia 30302-4106, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Azido nitrobenzoxadiazole (NBD) was observed to undergo a ‘reduction’ reaction in the absence of an
obvious reducing agent, leading to amine formation. In the presence of an excess amount of DMSO,
a sulfoxide conjugate was also formed. The ratio of these two products was both temperature- and
solvent-dependent, with the addition of water significantly enhancing the ratio of the ‘reduction’
product. Two intermediates of the azido-NBD reaction in DMSO were trapped and characterized by low-
temperature EPR spectroscopy. One was an organic free radical (S¼1/2) and another was a triplet nitrene
Received 18 December 2012
Received in revised form 17 April 2013
Accepted 19 April 2013
Available online 25 April 2013
Dedicated to the memory of Professor Wil-
liam H. Nelson
(S¼1) species.
A
mechanism was proposed based on the characterized free radical and triplet
intermediates.
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
NBD
Azide
Reduction
Nitrene
EPR
1. Introduction
metals such as copper were also observed to facilitate the frag-
mentation.12 Understandably, reduction of the azido group nor-
Azido compounds are versatile molecules. They are being used
extensively in both organic reactions and chemical biology. For ex-
ample, the ‘click’ reaction between an azido group and an alkyne,1
including the copper catalyzed and copper-free ‘click’ reactions,2
are becoming a very important approach to DNA,3 protein,4 and
glycan labeling5 among other applications. The azido group is
known to undergo a variety of useful reactions, including the
Schmidt reaction, the Curtius reaction, and other important ring-
formation or ring-expansion reactions.6 Reduction of azides is not
only a very efficient method for the preparation of amines,7 amides,
and sulfonamides,8 but also a new approach for the quantitative
fluorescence detection of hydrogen sulfide in aqueous solutions.9
Azido compounds are known to fragment under light and/or at el-
evated temperature and to generate nitrenes,10 which can undergo
an insertion reaction with alkenes, yielding aziridines.11 Transition
mally requires a reducing agent. However, herein we report a case
where azido-NBD was ‘reduced’ to its corresponding amine in the
absence of an obvious reducing agent. Copper was observed to
catalyze the reaction. Electron paramagnetic resonance (EPR)
studies were also performed to gain initial insight into the reaction
process.
2. Results and discussion
In our effort aimed at modifying the fluorescent dye nitro-
benzoxadiazole (NBD), we synthesized azido-NBD (1) and conducted
copper-mediated Huisgen cycloaddition at room temperature in
a mixed solvent of water, DMF, and tert-butanol. Instead of the an-
ticipated [2þ3] cycloaddition product, we observed the formation of
the ‘reduction’ product, amino-NBD (3) in 61% yield, although the
reaction contained no obvious reducing agent. In the absence of any
reducing agent, reduction of the azido group is nearly impossible
unless the reaction went through a very reactive intermediate(s),
which could ‘rob’ electrons or hydrogen atoms from otherwise ‘inert’
compounds. The formation of a nitrene intermediate (compound 2)
* Corresponding authors. Tel.: þ1 404 413 5532 (A.L.); tel.: þ1 404 413 5544; fax:
þ1 404 413 5543 (B.W.); e-mail addresses: Feradical@gsu.edu (A. Liu), wang@
gsu.edu (B. Wang).
y
These two authors contributed equally to this work.
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.04.091

5080
H. Peng et al. / Tetrahedron 69 (2013) 5079e5085
can explain the product observed in this study. Specifically, nitrene
formation followed by hydrogen abstraction would explain the for-
mation of amino-NBD (3, Scheme 1). However, such fragmentations
normally require either light or heat.13 To the best of our knowledge,
room temperature azidoarene fragmentation in the absence of ir-
radiation to the extent that a side product is readily isolated in high
yield is considered unusual.
When comparing reaction outcomes in entries 1 and 5 as well as
3 and 15, one can see that the reaction was temperature-
dependent. In DMSO, the reaction gave a 16% conversion at room
temperature in 24 h without catalyst (Table 1, entry 1), whereas full
conversion was achieved in 24 h at 80 ꢀC (entry 2). It also seems
that ordinary room-light makes a difference. For example, at 80 ꢀC
the reaction achieved full conversion under light (entry 2), but only
about 50% conversion when the reaction was run in the dark (entry
5). Subsequent reactions in the presence of metal ions (entries 10
and 11, 16 and 17) or EDTA (entries 20 and 21, 22 and 23) also in-
dicate room-light affords reaction acceleration. It was found that
Cu(II) and Cu(I) (under N2 atmosphere) seem to facilitate the re-
action. Comparing the results from entries 7 and 9 to that of entry 5
in Table 1, one can see that the addition of copper significantly
accelerates the reaction (This was also confirmed in Fig. 1b.). In the
absence of any metal ions and in the dark, 48% of the starting
material remained at the end of the reaction (entry 5), whereas all
starting material disappeared when the same reaction was carried
out in the presence of Cu(II) (entry 6) or Cu(I) (entry 9). However,
addition of either Cu(II) or Cu(I) did not seem to affect the product
ratio significantly. For example, Cu(II)-facilitated reactions gave
a 3:1 ratio between 4 and 3 independent of whether the reaction
was run in the dark. This ratio is the same as the reaction under
light and in the absence of any metal ions (entry 2). We also tested
the effect of silver ions. It was surprising to find that Ag actually
inhibits the reaction. For example, heating 1 at 80 ꢀC in DMSO led to
the complete disappearance of the starting material after 24 h.
However, the same reaction in the presence of AgNO3 still had 11%
starting material after 24 h. The specific reason for this inhibition is
not clear.
O
S
X
NH
N
N
2
2
3
DMSO
Cu(II)
N
N
N
N
N
N
N
N
O
O
O
O
NO
NO
NO
NO
2
2
2
1
2
3
4
2S: X = N
2T: X = N
Scheme 1. The copper catalyzed NBDenitrene reaction in DMSO.
In order to eliminate the possibility that DMF was functioning as
the reducing agent, we conducted the reaction using DMSO as the
solvent. Interestingly, DMSO adduct 4 was obtained in addition to
the reduction product 3. The nitrene intermediate formed by the
fragmentation of an acyl azide group (e.g., Boc-N3) has been ob-
served to undergo a transfer reaction leading to conjugation with
sulfoxides and sulfides,14 but the aromatic nitrene transfer reaction
has not yet been reported. This reaction could provide an alterna-
tive approach for the synthesis of the sulfoximines, which are very
important intermediates in medicinal chemistry.15 This could also
offer the possibility of fluorescent labeling of the sulfoxide group.
To achieve a further understanding of this unique reaction,
a series of experiments were performed to study how several fac-
tors affect the reaction outcome. The reaction was carried out at
both room and elevated temperatures (80 ꢀC), with or without
a transition metal ion (Cu(I), Cu(II), Ag(I), Fe(II), or Fe(III)) as the
catalyst. The effects of EDTA, shielding the reaction from any light,
and added H2O were also examined. 1H NMR spectroscopy was
used to monitor the reaction progress. The results of the reactions
under different conditions are listed in Table 1.
1
(a)
0.8
0.6
0.4
0.2
0
rt, dark
80 C,
dark
0
10
20
30
Reaction Time (hour)
Table 1
1. DMSO-d6+Cu(I)
2. DMSO-d6/D2O+Cu(I)
3. DMSO-d6/D2O
Product analysis of the NBDenitrene reaction in DMSOa
1
0.8
0.6
0.4
0.2
0
(b)
Entry
Catalyst
Other
Solvent
Temp
1
3
4b
1
2
3
4
5
6
7
8
None
None
CuCl
None
None
CuCl2
CuCl2
CuCl
None
None
Dark
None
Dark
None
Dark
None
Dark
None
Dark
None
Dark
None
Dark
None
Dark
DMSO
DMSO
DMSO/H2O
DMSO/H2O
DMSO
DMSO
DMSO
DMSO
DMSO
rt
0.84
0
0.95
0.29
0.48
0
0
0
0
0.11
0.55
0
0
0
0.08
0.24
0.05
0.55
0.18
0.25
0.26
0.32
0.39
0.22
0.06
0.74
0.64
0.74
0.69
0.70
0.55
0.53
0.66
0.19
0.76
0.32
0.47
0.08
0.76
0
80 ꢀC
rt
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
0.16
0.34
0.75
0.74
0.68
0.61
0.66
0.39
0.26
0.36
0.26
0.24
0.30
0.26
0.17
0.11
0.10
0.24
0.29
0.43
0
10
Reaction Time (hour)
20
Fig. 1. Time dependent decrease of 1 as a percentage of the initial concentration under
different reaction conditions. (a) Reaction was performed in DMSO-d6/D2O in the
presence of Cu(I). (b) Reaction was performed at 80 ꢀC in the dark. DMSO-d6/D2O ratio
was 3:1. Cu(I) was 20 mmol %.
9
CuCl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
AgNO3
AgNO3
CuCl2
CuCl2
CuCl
DMSO
DMSO
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO/H2O
DMSO
In all the above reactions, the DMSO adduct 4 was the major
product. However, upon addition of water, reaction seems to slow
down (entry 2 vs 4) leading to the reduction product 3 being the
major component. For example, when the reaction was performed
in a solvent system of DMSO/H2O (3:1), the ratio of the two isolated
products 3 and 4 was found to be around 3:1, while the reaction in
dry DMSO gave a 1:3 ratio. Again, addition of either Cu(I) or Cu(II)
accelerated the reaction but did not seem to alter the ratio of the
reduction product and the DMSO adduct. For example, in the
presence of water, the ratio between 3 and 4 was approximately 3:1
with or without added Cu (at the valence state of I or II) (entries 12
and 14). When carried out in the dark, this ratio decreased slightly.
CuCl
0.07
0
AgNO3
AgNO3
FeCl3
FeCl2
None
None
None
None
0.17
0.29
0.23
0.71
0
None
None
EDTA, Dark
EDTA
EDTA, Dark
EDTA
0.39
0.1
DMSO
a
Unless otherwise indicated, reactions were proceeded for 24 h with 10 mg of 1,
0.8 mL of DMSO-d6 or DMSO-d6/D2O 3:1, 20 mmol % of catalyst. ‘Dark’ refers to
protecting from light using foil.
b
Ratio is determined by the crude 1H NMR spectra.

H. Peng et al. / Tetrahedron 69 (2013) 5079e5085
5081
Table 2
Such results suggest that copper ion addition did not fundamen-
Product analysis of the NBDenitrene reaction in DMFa
tally alter the reaction pathway. Interestingly, Ag(I) addition ac-
celerated the reaction when water was present. For example, in
entry 2, 29% of starting material remained after 24 h of reaction;
while the same reaction in the presence of Ag(I) resulted in com-
plete conversion of the starting material to the product. In the
meantime, addition of Ag(I) also led to a decrease of the ratio be-
tween 3 and 4 from 3:1 to about 2:1. For the reactions in the
presence of water, we also studied the effect of Fe(III). It was found
that Fe(III) played very little role in the reaction (entries 19 and 20).
In order to examine the catalytic effects by trace amount of
metal ions present in water, we also performed experiments by
adding EDTA (entries 20e23). The reaction was found to be in-
complete (29% conversion) in the presence of EDTA while in the
dark (entry 20) at 80 ꢀC. A comparison of entries 20 and 21 or 22
and 23 also suggest that light can accelerate the reaction. This series
of studies suggest that (1) light and heat can accelerate the re-
action; (2) Cu(I) and Cu(II) can also accelerate the reaction; (3)
water can slow down the reaction rate, while altering the ratio of
the product by facilitating the formation of reduction product; (4)
Ag(I) and Fe(II or III) has little or mixed effect on the reaction, (5)
trace amount of ions in the buffer and/or DMSO plays an important
role in accelerating the reaction; and (6) in the absence of light and
metal ion catalyst (EDTA addition), the reaction is slow even at an
elevated temperature (80 ꢀC).
In order to achieve a more in depth understanding of the re-
action profiles, the effects of temperature, catalyst, and solvent on
the reaction rates were examined by monitoring the disappearance
of the starting material using 1H NMR every 2 h for 8 h, and then at
24 h or 28 h point (Fig. 1). In Fig. 1a, the results confirmed the ac-
celeration effects by both heat and light, as one would expect. As is
shown in Fig. 1b, the highest reaction rate was achieved in DMSO in
the presence of Cu(I), while the reaction slowed down when carried
out in a mixed solvent of DMSO-d6/D2O (3:1). When the reaction
was carried out in DMSO-d6/D2O without a catalyst, the reaction
went slowly and was not complete within 24 h. These results fur-
ther support that heat, light or catalyst can indeed accelerate the
reaction. At the same time, water addition obviously reduced the
reaction rate. It is interesting to note that although water decreases
the overall reaction rate, its presence led to an increased proportion
of the reduction product 3.
Entry
Catalyst
Other
Solvent
Temp
1
3
5b
1
2
3
4
5
6
None
None
CuCl
CuCl2
CuCl2
AgNO3
None
None
None
None
Dark
None
DMF
rt
0.64
0.18
0
0
0
0.30
0.59
0.93
0.92
0.91
0.83
0.06
0.24
0.07
0.08
0.09
0.17
DMF/H2O
DMF/H2O
DMF/H2O
DMF/H2O
DMF/H2O
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
80 ꢀC
0
a
Reactions were carried out for 24 h with 10 mg of 1, 0.8 mL of DMF or DMF/H2O
3:1, 20 mmol % of catalyst.
b
Ratio is determined by the crude 1H NMR spectra.
even when the reaction was performed in dark (entry 5). It is worth
noting that percentage of compound 5 was higher in the absence of
an added catalyst (entries 1 and 2). This is easy to understand since
the relative reaction rates of the substitution reaction and nitrene
formation process were altered with the addition of a catalytic
metal ion. Comparison of entries 1 and 2 also suggest that H2O
slowed down the reaction in DMF as well.
In order to verify the reactive species involved in the reaction,
we also conducted EPR studies. Since azido groups are known to
fragment and generate nitrenes, it is reasonable to assume that
a nitrene intermediate could be possibly involved in this re-
action. The observation of the DMSO adduct of NBD also supports
nitrene formation from azido-NBD. After heating for 20 min at
80 ꢀC, the azido-NBD solution (80 mM) in DMSO and water (3:1
ratio) yielded a broad signal covering 5360 Gauss range with
three well-resolved resonance components and an anisotropic S
¼ 1/2 EPR signal at g¼2.0046 (Fig. 2A). The broad EPR signal can
be simulated by an S¼1 triplet species (Fig. 2B) with a D value of
0.6263ꢁ0.0010 cmꢂ1 and three principal g values of 1.99, 2.033,
and 2.033. The large zero-field splitting (ZFS) parameter D is
When the reaction was performed in DMF, no corresponding
solvent adduct was observed. The ‘reduction’ product 3 was the
major component (ca. 90% yield) (Scheme 2). Compound 5 was
observed as a minor product, which was presumably formed from
the substitution reaction between compound 1 and the free
dimethylamine present in DMF. Control reactions using dimethyl-
amine confirmed this aspect. Product analysis results are shown in
Table 2.
Scheme 2. The NBDenitrene reaction in DMF.
Fig. 2. Low-temperature (10 K) continuous-wave X-band EPR spectrum of azido-NBD
(80 mM) in DMSO and H2O. (A) EPR spectrum of azido-NBD obtained at 10 K, mi-
crowave frequency 9.64 GHz, microwave power 1 mW, modulation frequency 100 kHz,
modulation 5 G, (B) simulation of the triplet state, (C) simulation of the free radical,
and (D) simulated EPR spectrum by combining B and C. The inset is a magnification of
the g¼2 region, which showed a free-radical species with partially resolved hyperfine
structure.
When compared with the reactions in DMSO, the reaction went
slower in DMF with a conversion of 38% at room temperature after
24 h (Table 2, entry 1). When heated to 80 ꢀC, 82% conversion was
achieved (entry 2). The presence of transition metal catalysts in
addition to heating drove the reaction to completion (entries 3e6),

5082
H. Peng et al. / Tetrahedron 69 (2013) 5079e5085
characteristic of triplet nitrenes.16 However, the observed D value
of the azido-NBD reaction was close to but also noticeably
smaller than that of the triplet ground state of p-nitrophenyl
nitrene (0.9833 cmꢂ1).17 It is known that
p
delocalization re-
duces the density of the
p-electron on the nitrogen atom and,
thus, reduces the value of D.18 The D value of nitrenes was noted
to be proportional to the inverse cube of the average distance
between the two unpaired electrons; the smaller D value in-
dicated greater delocalization of the spin density away from the
nitrogen atom.19
The S¼1/2 EPR signal at g¼2.0046 had additional hyperfine
structures (Fig. 2), consistent with an organic species with free-
radical characters. Simulation of the EPR spectrum confirmed the
presence of multiple species and enabled us to estimate the
relative ratio between the two EPR-active species (Fig. 2). The
free-radical signal accounted for about 5e8% of the total spin,
whereas that the triplet EPR signal accounted for the majority of
the total spin (92e95%). Because the signal was extremely broad
and there is no spin standard available, it was not possible to
precisely determine the absolute concentration of the triplet in-
termediate based on the EPR signal strength. Integration of the
simulated EPR signals suggested that the free-radical signal in-
tensity was about 1/12 to 1/15 of the triplet signal. Thus, the total
EPR-active species was a few percent of the total azido-NBD. If
these were the reactive species, they would explain the slow
reaction rate.
In the absence of water, the triplet EPR signal was more axial
and less rhombic (Fig. 3). A notable difference was that the ratio of
S¼1/2 and S¼1 species had increased by one order of magnitude
when water was absent from the reaction mixture. The presence of
water caused about 10-fold increase of the ratio between reduction
product 3 and DMSO adduct 4 in the same time period (Table 1).
Since the ‘reduction’ of the nitrene has to go through a radical in-
termediate 6 (Scheme 3), the suppression of radical species by
water might be due to the accelerated conversion of the radical into
the reduction product 3.
Fig. 4. X-band EPR spectra of Cu(II) (1 mM) and azido-NBD (80 mM) in DMSO during
the Cu(II)-mediated catalytic reaction (A) and Cu(II) in DMSO in the absence of azido-
NBD (B). The inset compares the resonances at the g¼2 region.
Cu(II) ion showed an axial EPR spectrum with copper hyperfine
splitting in the parallel region similar to that observed in regular
copper coordination compounds (Fig. 4A). The A// value was mea-
sured to be 130 G, which was in the typical type II copper range
with N or N/O ligands and much larger than that of type I copper
with sulfur ligands. In contrast, in the absence of azido-NBD, the
EPR spectrum did not show a resolved hyperfine splitting in the
parallel region at various temperatures (5e30 K) regardless of
strong and weak microwave power applied (Fig. 4B). The copper
EPR signal was more difficult to saturate by applied microwave
power when azido-NBD was present, suggesting that NBD, or an
intermediate derived from it, binds to Cu(II) during the catalytic
reaction in DMSO.
The divalent metal ion caused a significant increase of both the
S¼1/2 and S¼1 species, thereby providing further evidence that
these EPR-active species were the reactive intermediates. The EPR
spectra in the absence and presence of water (Fig. S1) both showed
resolved hyperfine splitting and were distinct from the spectrum
without azido-NBD. Adding H2O to the azido-NBD/DMSO/Cu(II)
mixture caused the formation of a Cu(II) complex with a spec-
trum reminiscent of that from of CuðH2OÞ26þ complex. This may
result in the deactivation of copper catalyst, which may explain
why the addition of H2O slows down the reaction. The free radical
to triplet ratio was significantly reduced in the presence of water.
Although the g¼2.0046 EPR species partially overlapped the copper
signal, it was still possible to estimate its intensity from the signal
amplitude (Fig. 5).
Fig. 3. X-band EPR spectrum of azido-NBD (80 mM) in DMSO. Instrument conditions
were the same as those in Fig. 2. The inset magnifies the g¼2 region.
We have also characterized the Cu(II)-mediated catalytic pro-
cess by EPR spectroscopy. Fig. 4 shows EPR spectra of CuCl2 (1 mM)
in the presence and absence of azido-NBD in DMSO (heated for
20 min at 80 ꢀC), respectively. In the presence of azido-NBD, the

H. Peng et al. / Tetrahedron 69 (2013) 5079e5085
5083
absent, showed only one type II copper. Nevertheless, it is rea-
sonable to believe that the contribution of the two possible reso-
nance structures of amide in DMF has caused two distinct forms of
the Cu(II)/azido-NBD complex. The free-radical signal was almost
completely suppressed, indicating water-induced fast consumption
of the free radical.
As to the detailed mechanism of formation for the two
products, one can postulate the formation of the NBD-nitrene 2
after the loss of a nitrogen molecule, thermally, photochemically,
and/or upon metal ion catalysis (Scheme 3). Given the nature of
the two isoforms of nitrene 2 (singlet 2S and triplet 2T), it is
proposed that the DMSO adduct 4 was produced from the addi-
tional reaction between the singlet nitrene 2S and DMSO. It is
also proposed that the formation of product 3 was from the
triplet state, through a radical-type mechanism by hydrogen
abstraction. A copperenitrene radical intermediate 6, which is
evidenced by EPR spectrum, can be generated as a stabilized
nitrene.
Scheme 3. Proposed mechanisms of the azido-NBD ‘reduction’ reaction leading to the
amine product via a triplet intermediate and DMSO conjugate formation.
Fig. 5. EPR spectra of Cu(II) and azido-NBD in DMSO in the absence (A) and presence of
water (B).
In the presence of water, compound 3 was the major product.
One possible way to explain the enhanced ‘reduction’ product
ratio in the presence of water is that H2O can donate a H atom and
thus accelerates the proton abstraction pathway. However, given
the reactive nature of hydroxy radicals, we thought that this was
an unlikely scenario. We conducted quantum mechanics (QM)
calculation to understand the relative energetics in a proton ab-
straction process. We chose to perform the calculation at the
When the reaction was carried out in DMF, copper exhibited
a complex EPR signal consistent with a metal-bound triplet with
hyperfine splitting patterns in both the parallel and perpendicular
regions (Fig. 6). The control sample, in which azido-NBD was
B3LYP/6-31þG
*
level because it is a cost-effective method and was
already used to model the free-radical propagation reaction
steps.20 Due to the extremely high energy of the hydroxyl radical,
there would be an energy difference of about 20e25 kcal/mol
between the starting materials and the products if H2O indeed
acted as an H-donor (Fig. 7). The energy difference would be much
smaller at 8e13 kcal/mol if DMSO acted as a H-donor. Thus in
a mix-solvent system, one would expect that DMSO be the
dominant donor in a hydrogen-abstraction process and water
addition would not be expected to perturb the production dis-
tribution ratio by simply providing a large quantity of a hydrogen
donor (water). In order to further examine the possibility for the
solvents to act as hydrogen donors, a H NMR experiment was also
performed to test the kinetic isotope effect (KIE) using mixed
solvent systems DMSO-d6/D2O, DMSO-d6/H2O and DMSO/D2O.
However, no obvious KIE was observed for either DMSO or H2O
from the formation of NBD-NH2 (Fig. 8). Such results are consis-
tent with the formation of the reactive nitrene intermediate(s)
being the rate-limiting step. Although we could not rule out the
possibility of water acting as H-donors, the effect of H2O is more
likely to be due to its ability to affect the distribution of the two
nitrene species.
Fig. 6. EPR spectra of Cu(II) and azido-NBD in DMF (A) and Cu(II) in DMF without
azido-NBD (B).

5084
H. Peng et al. / Tetrahedron 69 (2013) 5079e5085
catalyst, and elevated temperature. In addition, solvents signifi-
-471000
-471050
-471100
(a)
cantly affect product distribution with the addition of water sub-
stantially enhancing the proportion of the ‘reduction’ product 3.
EPR spectroscopy provides solid evidence for the presence of
a triplet nitrene 2T and a radical intermediate, which become more
pronounced during reaction at higher temperature or in the pres-
ence of copper catalyst. Based on the results from 1H NMR and EPR
spectroscopic studies, a possible mechanism has been proposed.
Considering the significant change of the product distribution in
different solvents, one can envision that the ratio of the reactive
intermediates might be affected by the solvents. The results pre-
sented offer insight into possible reactions that arylazido com-
pounds can undergo, provide an alternative preparation method for
sulfoximines, and offer the possibility of fluorescent labeling of the
sulfoxide group.
24.8 kcal/mol
-471450
19.5 kcal/mol
-471500
-770200
4. Experimental section
4.1. General experimental methods
(b)
13.2 kcal/mol
Unless otherwise noted, all reagents were purchased from
Aldrich and used without further purification. All reactions were
performed under the protection of N2, except for the NMR experi-
ments. Thermometers were not calibrated. 1H NMR and 13C NMR
spectra were recorded on a Bruker 400 NMR spectrometer. IR was
recorded on a PerkinElmer Spectrum One FT-IR Spectrometer. Mass
spectral analyses were performed on an ABI API 3200 (ESI-Triple
Quadruple). Low-temperature X-band EPR first derivative spectra
were recorded in perpendicular mode (TE102) on a Bruker ER200D
spectrometer at 100-kHz modulation frequency using a 4116DM
resonator. Sample temperature was maintained with an ITC503S
temperature controller, an ESR910 liquid helium cryostat, and
LLT650/13 liquid helium transfer tube (Oxford Instruments, Concord,
MA). EPR spectral simulations were performed by using a Windows-
based spin simulation program and a dedicated triplet simulation
program, both of which were written by Dr. Andrew Ozarowski.
-770300
-770600
-770700
8.3 kcal/mol
Fig. 7. Quantum mechanics calculation on the reaction between nitrene and H2O (a)
and DMSO (b).
4.2. Preparation of azido-NBD (1)
Sodium azide (584 mg, 9.04 mmol, 3.0 equiv) was dissolved in
56 mL of H2O/acetone (1:1, v/v) in a 250 mL round bottom flask.
Then a solution of NBD chloride (600 mg, 3.01 mmol, 1.0 equiv) in
acetone (28 mL) was added into the flask dropwise at room tem-
perature with stirring. The reaction mixture turned from light
yellow to yellow-brown within 5 min. After 10 min, the reaction
mixture was concentrated under vacuum to remove acetone. The
yellow precipitates were filtered out and washed with water
(20 mLꢃ3) to give 1 as yellow crystals (576 mg, 93%), which were
used for the next step without purification. 1H NMR (400 MHz,
CDCl3):
d
¼8.49 (d, J¼8.0 Hz, 1H), 7.04 ppm (d, J¼8.0 Hz, 1H); 13C
NMR (100 MHz, DMSO-d6):
d
¼145.9, 143.7, 137.8, 133.9, 131.8,
116.8 ppm; IR 3093, 2113, 1631, 1527, 1444, 1279, 996 cmꢂ1; mp
100e101 ꢀC (H2O/acetone); HRMS (ESþ), m/z calculated for
C6H2N6O3Na: 229.0086, found: 229.0079 [MþNa]þ.
4.3. General procedure of the NBDenitrene reaction in the
presence of copper
Fig. 8. Kinetic isotope effect on formation of NBDeNH2.
In an oven-dried 5-mL round bottom flask equipped with
a magnetic stir bar, we added azido-NBD (1, 10 mg, 0.05 mmol,
1.0 equiv) and CuCl2 (1.2 mg, 0.01 mmol, 0.2 equiv). The system was
degassed using a vacuum pump and backfilled with Ar. Then sol-
vent (DMSO/H2O 0.6 mL/0.2 mL) (or DMSO (0.8 mL)) was added
into the system. The reaction mixture was heated at 80 ꢀC with
stirring for 24 h. Then the reaction solution was diluted with ethyl
acetate (30 mL) and washed with H2O (10 mLꢃ3) and brine (10 mL).
3. Conclusion
In summary, an unusual ‘reduction’ of an arylazide was observed
and examined in detail. The arylazide was reduced to the corre-
sponding amine 3 in the absence of an obvious reducing agent. In
the presence of DMSO as a solvent, a DMSO adduct 4 was also
formed. This reaction was facilitated by light, the addition of copper

H. Peng et al. / Tetrahedron 69 (2013) 5079e5085
5085
The organic layer was dried over anhydrous Na2SO4. Solvent was
References and notes
evaporated to yield the crude product, which was purified by flash
chromatography (silica gel, eluent: hexane/ethyl acetate 2:1) to
give pure NBDeNH2 (3) as orange-red needles and NBDeDMSO (4)
as orange crystals. NBDeNH2 (3): 1H NMR (400 MHz, DMSO-d6):
1. (a) Huisgen, R. Angew. Chem. 1963, 75, 604e637; (b) Rostovtsev, V. V.; Green, L.
G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41, 2596e2599; (c)
Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40,
2004e2021.
2. Baskin, J. M.; Prescher, J. A.; Laughlin, S. T.; Agard, N. J.; Chang, P. V.; Miller, I. A.;
Lo, A.; Codelli, J. A.; Bertozzi, C. R. Proc. Natl. Acad. Sci. U.S.A. 2007, 104,
16793e16797.
3. (a) Salic, A.; Mitchison, T. J. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 2415e2420; (b)
Dai, C. F.; Wang, L. F.; Sheng, J.; Peng, H. J.; Draganov, A. B.; Huang, Z.; Wang, B.
H. Chem. Commun. 2011, 3598e3600; (c) Lin, N.; Yan, J.; Huang, Z.; Altier, C.;
Yan, J.; Carrasco, N.; Suyemoto, M.; Johnson, L.; Fang, H.; Wang, Q.; Wang, S.;
Wang, B. Nucleic Acids Res. 2007, 35, 1222e1229; (d) El-Sagheer, A. H.; Brown, T.
Chem. Soc. Rev. 2010, 39, 1388e1405.
d
¼8.87 (br, 1H), 8.48 (d, J¼8.8 Hz, 1H), 6.39 ppm (d, J¼8.8 Hz, 1H);
13C NMR (100 MHz, CD3OD):
d
¼147.1, 144.24, 144.20, 137.0, 121.8,
101.8 ppm; mp 238e239 ꢀC (methanol); IR 3424, 3334, 2922, 2496,
1640, 1250 cmꢂ1; HRMS (ESþ), m/z calculated for C6H5N4O3:
181.0362, found: 181.0360 [MþH]þ. NBDeDMSO (4): 1H NMR
(400 MHz, acetone-d6):
d
¼8.54 (d, J¼8.4 Hz, 1H), 7.06 (d, J¼8.4 Hz,
1H), 3.65 ppm (s, 6H); 13C NMR (100 MHz, acetone-d6):
d¼150.1,
4. Vila, A.; Tallman, K. A.; Jacobs, A. T.; Liebler, D. C.; Porter, N. A.; Marnett, L. J.
Chem. Res. Toxicol. 2008, 21, 432e444.
148.4, 145.3, 136.1, 128.1, 111.5, 42.5 ppm; mp 212e214 ꢀC (ethyl
acetate), IR 3005, 2923, 1613, 1515, 1294, 1108, 830 cmꢂ1; HRMS
(ESþ); m/z calculated for C8H8N4O4S: 257.0345, found: 257.0348
[MþH]þ. In DMF as the solvent, NBDeNMe2 (5) 1H NMR (400 MHz,
5. (a) Laughlin, S. T.; Baskin, J. M.; Amacher, S. L.; Bertozzi, C. R. Science 2008, 320,
664e667; (b) Chang, P. V.; Prescher, J. A.; Sletten, E. M.; Baskin, J. M.; Miller, I.
A.; Agard, N. J.; Lo, A.; Bertozzi, C. R. Proc. Natl. Acad. Sci. U.S.A. 2010, 107,
1821e1826; (c) Sawa, M.; Hsu, T. L.; Itoh, T.; Sugiyama, M.; Hanson, S. R.; Vogt,
P. K.; Wong, C. H. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 12371e12376.
6. (a) Lang, S.; Murphy, J. A. Chem. Soc. Rev. 2006, 35, 146e156; (b) Benati, L.;
Bencivenni, G.; Leardini, R.; Nanni, D.; Minozzi, M.; Spagnolo, P.; Scialpi, R.;
Zanardi, G. Org. Lett. 2006, 8, 2499e2502.
7. (a) Corey, E. J.; Link, J. O. J. Am. Chem. Soc. 1992, 114, 1906e1908; (b) Kazemi, F.;
Kiasat, A. R.; Sayyahi, S. Phosphorus Sulfur 2004, 179, 1813e1817.
8. Omura, K.; Uchida, T.; Irie, R.; Katsuki, T. Chem. Commun. 2004, 2060e2061.
9. (a) Peng, H. J.; Cheng, Y. F.; Dai, C. F.; King, A. L.; Predmore, B. L.; Lefer, D. J.;
Wang, B. H. Angew. Chem., Int. Ed. 2011, 50, 9672e9675; (b) Lippert, A. R.; New,
E. J.; Chang, C. J. J. Am. Chem. Soc. 2011, 133, 10078e10080; (c) Peng, H.; Chen,
W.; Wang, B. In Gasotransmitters: Physiology and Pathophysiology; Hermann, A.,
Sitdikova, G. F., Weiger, T. M., Eds.; Springer: Berlin, Heidelberg, Germany, 2012;
pp 99e137; (d) Peng, H.; Chen, W.; Cheng, Y.; Wang, B. Sensors 2012, 12,
15907e15946.
acetone-d6):
d
¼8.49 (d, J¼9.2 Hz, 1H), 6.39 (d, J¼9.2 Hz, 1H),
3.71 ppm (s, 6H); 13C NMR (100 MHz, DMSO-d6):
d
¼147.1, 145.3,
145.2, 136.7, 102.7 ppm; mp 214e216 ꢀC (ethyl acetate); HRMS
(ESþ); m/z calculated: 209.0675, found: 209.0670 [MþH]þ.
4.4. Reaction monitoring using 1H NMR spectroscopy
Azido-NBD (1, 7 mg, 0.034 mmol, 1.0 equiv) (and CuBr (1.0 mg,
0.007 mmol, 0.2 equiv)) was added into an NMR tube. Then solvent
(0.4 mL, DMSO-d6 or DMSO-d6/D2O 3:1) was added into the tube in
dark, and the reagents were shaken well to mix in dark for less than
30 s. Then the 1H NMR spectra of reaction mixture were recorded
every 2 h for 8 h, and after 24 h.
10. (a) Horner, L.; Christmann, A. Angew. Chem. 1963, 2, 599e608; (b) Moss, R. A.;
Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry; Wiley-Interscience:
New Brunswick, NJ, 2004; (c) Scriven, E. F. V.; Turnbull, K. Chem. Rev. 1988, 88,
297e368.
11. Muller, P.; Fruit, C. Chem. Rev. 2003, 103, 2905e2919.
12. Kwart, H.; Kahn, A. A. J. Am. Chem. Soc. 1967, 89, 1950e1951.
13. (a) Lord, S. J.; Lee, H. L.; Samuel, R.; Weber, R.; Liu, N.; Conley, N. R.; Thompson,
M. A.; Twieg, R. J.; Moerner, W. E. J. Phys. Chem. B 2010, 114, 14157e14167; (b)
He, J.; Chang, J.; Chen, R.; Wang, Q. Chin. J. Pharm. 2002, 33, 425e426.
14. (a) Bergmeier, S. C.; Stanchina, D. M. J. Org. Chem. 1997, 62, 4449e4456; (b)
Bach, T.; KSrber, C. Tetrahedron Lett. 1998, 39, 5015e5016; (c) Bach, T.; Korber, C.
Eur. J. Org. Chem. 1999, 1033e1039.
15. Moessner, C.; Bolm, C. Org. Lett. 2005, 7, 2667e2669.
4.5. Computational studies
All calculations were performed using the Gaussian 03 pro-
gram.21 Initial geometry optimizations and DFT calculations were
€
carried out using the B3LYP22 with the standard 6-31þG
*
basis set.
16. Sander, W.; Grote, D.; Kossmann, S.; Neese, F. J. Am. Chem. Soc. 2008, 130,
4396e4403.
Acknowledgements
17. Hebden, J. A.; McDowell, C. A. J. Magn. Reson. 1971, 5, 115e133.
18. Smolinsky, G.; Snyder, L. C.; Wasserman, E. Rev. Mod. Phys. 1963, 35, 576e577.
19. Weil, J. A.; Bolton, J. R. Electron Spin Resonance: Elementary Theory and Practical
Applications, 2nd ed.; Wiley-Interscience: Hoboken, New Jersey, 2007.
20. Yu, X. R.; Pfaendtner, J.; Broadbelt, L. J. J. Phys. Chem. A 2008, 112, 6772e6782.
21. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken,
V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G.
A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.;
Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.;
Laham, A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02;
Gaussian Inc.: Pittsburgh PA, 2003.
We thank Drs. Andrew Ozarowski and Jerzy Krzystek for pro-
viding us with a triplet simulation program and for helping with
spectral analyses. This work was supported by the National In-
stitutes of Health grants (GM084933 and GM086925 to B.W.) and
the National Science Foundation (MCB0843537 to A.L.). K.H.D. ac-
knowledges the fellowship support from Southern Regional Edu-
cation Board (SREB). W.C. acknowledges an MBDAF fellowship from
Georgia State University; and H.P. also acknowledges the financial
support from GSU University Fellowship for Center for Diagnostics
and Therapeutics and an MBDAF fellowship.
Supplementary data
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.tet.2013.04.091.
22. (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648e5652; (b) Lee, C.; Yang, W.; Parr, R.
G. Phys. Rev. B 1988, 37, 785e789.