Chemoselective amide formation using O-(4-nitrophenyl)hydroxylamines and pyruvic acid derivatives

Article abstract of DOI:10.1021/jo302175g

A series of O-(4-nitrophenyl)hydroxylamines were synthesized from their respective oximes using a pulsed addition of excess NaBH₃CN at pH 3 in 65a-75% yield. Steric hindrance near the oxime functional group played a key role in both the ease by which the oxime could be reduced and the subsequent reactivity of the respective hydroxylamine. Reaction of the respective hydroxylamines with pyruvic acid derivatives generated the desired amides in good yields. A comparison of phenethylamine systems bearing different leaving groups revealed significant differences in the rates of these systems and suggested that the leaving group ability of the Na-OR substituent plays an important role in determining their reactivity with pyruvic acid. Competition experiments (in 68% DMSO/phosphate buffered saline) using 1 equiv of N-phenethyl-O-(4-nitrophenyl)hydroxylamine and 2 equiv of pyruvic acid in the presence of other nucleophiles such as glycine, cysteine, phenol, hexanoic acid, and lysine demonstrated that significant chemoselectivity is present in this reaction. The results suggest that this chemoselective reaction can occur in the presence of excess α-amino acids, phenols, acids, thiols, and amines.

Full text of DOI:10.1021/jo302175g

Article
pubs.acs.org/joc
Chemoselective Amide Formation Using O‑(4-
Nitrophenyl)hydroxylamines and Pyruvic Acid Derivatives
Sonali Kumar, Rashi Sharma, Megan Garcia, Joseph Kamel, Caroline McCarthy, Aaron Muth,
and Otto Phanstiel, IV*
Department of Medical Education, College of Medicine, University of Central Florida, 6850 Lake Nona Boulevard, Orlando, Florida
32827, United States
S
* Supporting Information
ABSTRACT: A series of O-(4-nitrophenyl)hydroxylamines
were synthesized from their respective oximes using a pulsed
addition of excess NaBH3CN at pH 3 in 65−75% yield. Steric
hindrance near the oxime functional group played a key role in
both the ease by which the oxime could be reduced and the
subsequent reactivity of the respective hydroxylamine. Reaction
of the respective hydroxylamines with pyruvic acid derivatives
generated the desired amides in good yields. A comparison of phenethylamine systems bearing different leaving groups revealed
significant differences in the rates of these systems and suggested that the leaving group ability of the N−OR substituent plays an
important role in determining their reactivity with pyruvic acid. Competition experiments (in 68% DMSO/phosphate buffered
saline) using 1 equiv of N-phenethyl-O-(4-nitrophenyl)hydroxylamine and 2 equiv of pyruvic acid in the presence of other
nucleophiles such as glycine, cysteine, phenol, hexanoic acid, and lysine demonstrated that significant chemoselectivity is present
in this reaction. The results suggest that this chemoselective reaction can occur in the presence of excess α-amino acids, phenols,
acids, thiols, and amines.
traditional methods. In contrast, O-substituted hydroxylamines
are more robust to chromatography.5
INTRODUCTION
■
Working at the interface between chemistry and biology,
scientists rely heavily on labeling technologies, which tag and
identify specific biological targets in aqueous media.1 The
development of water-based chemical-coupling technologies is
important, as they can provide improved tools for probing
biological processes via efficient tagging reactions. Our
continued exploration of the novel amide-forming chemistry
between α-keto acids and hydroxylamine derivatives in water
led to the discovery of other reactive nitrogen sources for this
chemistry, specifically substituted O-phenylhydroxylamines. To
the best of our knowledge, this is the first comprehensive report
of the preparation, purification, relative rate comparisons,
solvent sensitivities, and chemoselectivity profile of this class of
reactive nitrogen reagents.
In an earlier report, we demonstrated that α-keto acids react
cleanly with the related N-benzoyloxyamines (e.g., 1b) and
found improved yields and reactivity compared to 1a (Scheme
1).5 The scope of these reactions was also extended to include
the related α-keto phosphonic acids (e.g., 3a,b), which provided
high yields of amides and exquisite chemoselectivity in their
reactions with N-(benzoyloxy)amines. The improved reactivity
was thought to be derived from the superior leaving group
ability of the benzoyloxy group.
There are two types of hydroxylamines which undergo this
amide-forming chemistry. Type I involves the O-unsubstituted
hydroxylamines such as 1a, whereas type II involves O-
substituted hydroxylamines such as 1b. Mechanistic studies by
Sanki et al. using 18O labeling revealed that type I hydroxyl-
amines (RNHOH) react via nitrone intermediates (5; Scheme
2) and suggested that these were converted to nitrilium (6) or
hydrated intermediates before generating amides.2,6 Later
experiments by Bode et al. ruled out the nitrilium
intermediate.4b In 2012, using selectively 18O-labeled substrates,
Bode et al. provided support for the involvement of α-lactones
(4a) and oxaziridine (4b) intermediates in the reactions of
these O-unsubstituted hydroxylamine systems (type I).4b In
contrast, no intermediates have been observed in the related O-
(benzoyloxy)hydroxylamine systems, presumably due to the
Our interest in this reaction was inspired by the findings of
Bode et al., who described a high-yield amide-forming reaction
in water via the condensation of N-substituted hydroxylamines
(1a; Scheme 1) and α-keto acids (e.g., 2).2−4
We were interested in extending this reaction beyond N-
substituted hydroxylamines (e.g., 1a) to include O-substituted
hydroxylamine starting materials. The rationale for this effort is
that unlike N-substituted hydroxylamines, which form nitrone
intermediates,2 the O-substituted systems may instead form the
more reactive oximinium ions and lead to improved rates of
reaction. Indeed, fast rates are necessary in bioorthogonal
tagging reactions, as they typically involve second-order
kinetics.1 Moreover, N-alkylhydroxylamines are not stable as
their free bases, which complicates their purification by
Received: October 3, 2012
Published: November 28, 2012
© 2012 American Chemical Society
10835
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
Scheme 1. Ligation of Hydroxylamine Derivatives and α-Keto Acid Motifs To Form Amides
Scheme 2. Proposed Mechanisms for These Reaction Types2,4b,6
high reactivity of their putative oximinium ion intermediates in
water (e.g., 7 in Scheme 2).5
The N-substituted O-phenylhydroxylamines (e.g., 1c in
Scheme 1) were attractive for numerous reasons. First, they
represented new type II hydroxylamine systems for this amide-
forming chemistry. Second, the O-phenyl motif was in principle
more stable to pH changes and hydrolysis than the benzoate
motif present in the N-(benzoyloxy)amines. Third, methods
exist to access the O-phenylhydroxylamines16−25 and a recent
patent detailed a new high-yield entry to O-phenylhydroxyl-
amine derivatives.16 Fourth, the N−OR bond is cleaved during
these reactions and alteration of the leaving group ability of the
−OR group via substituted phenoxy systems (e.g., RNH−
OC6H4−Y) could provide a way to modulate the chemical
reactivity and stability of these systems. Indeed, the pKa values
of phenol (pKa 9.95) and most substituted phenols lie between
those of water (pKa 15.74) and benzoic acid (pKa 4.2).26 In this
regard, the O-phenylhydroxylamines could in principle be
Even though we developed a direct entry to N-benzoyloxy-
amines from amines,7,8 these systems have caveats associated
with their use. While they show excellent utility in both
siderophore and N-hydroxy-containing peptides synthesis,9−12
they typically have high chemical reactivity.5,13 Even though
they can be stored for months in the freezer they slowly
degrade at room temperature,7 which could limit their use as
chemoselective reagents. Moreover, the ester motif can in
principle react with biological amines to give benzamides and
hydroxylamines. In addition, a potentially problematic O-to-N
acyl transfer reaction has been observed with adjacent
amines.12,14,15 In our efforts to develop more robust reagents,
we investigated the O-phenylhydroxylamines as an alternative
motif.
10836
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
a
Scheme 3.
a
Reagents: (a) DMF, solid NaOH at 42 °C; (b) 37% HCl at room temperature.
a
Table 1. Synthesis of O-Substituted Hydroxylamines
12
8
yield of oxime 13, %
yield of 14, %
12a (phenacetaldehyde)
12b (acetaldehyde)
8a (X = H)
8a (X = H)
8a (X = H)
8a (X = H)
8b (X = CF3)
8c (X = NO2)
8a (X = H)
8a (X = H)
8a (X = H)
82 (13a; R = Bn, R′ = X = H)
94 (13b; R = Me, R′ = X = H)
98 (13c; R = n-C6H13, R′ = X = H)
100 (13d; R = R′ = Me, X = H)
100 (13e; R = Bn, R′ = H, X = CF3)
60% (13f; R = n-C6H13, R′ = H, X = NO2
90 (13g; R = Ph, R′ = X = H
65 (14a)
75 (14b)
73 (14c)
65 (14d)
15 (14e)
5 (14f)
NA
12c (heptaldehyde)
12d (acetone)
b
12a (phenacetaldehyde)
12c (heptaldehyde)
12e (benzaldehyde)
12f (acetophenone)
12g (1H-indole-3-carbaldehyde)
100 (13h; R = Ph, R′ = Me, X = H)
NA
c
40 (13i; R = 1H-indol-3-yl, R′ = X = H)
NA
a
Reagents: (a) EtOH; (b) pulsed additions of NaBH3CN, pH 3 maintained via aliquots of CH3SO3H. NA = not applicable, as no reduction was
b
c
observed from the corresponding oxime. Used as crude oxime directly in the next step. Recrystallized yield.
tailored to have improved reactivity, chemoselectivity, and
stability depending upon their O-substituent.
using ethyl acetimidate (9) and 4-nitrochlorobenzene (10a).16
Using a modified method described in the Experimental
Section, we generated multigram quantities of O-(4-
nitrophenyl)hydroxylamine 8a (Scheme 3).
A previous report by Sheradsky et al. demonstrated that the
N-alkyl-O-phenylhydroxylamines such as 1c were unstable
liquids, which could not be purified by distillation or
chromatography.27 We hypothesized that this was due to the
nucleophilic properties of the appended phenol ring.27 Since
the related N-benzoyloxyamine systems did not show the same
degree of instability, we reasoned that an electron-deficient
phenol could impart additional stability to the system. Indeed,
the 4-nitrophenol motif (e.g., 1c, where Y = NO2) solved this
issue and provided stable reagents for developing new methods.
Since the hydroxylamine N−O bond is cleaved as these systems
react with α-keto acids to form amides, they also have the
added advantage of releasing 4-nitrophenol, which provides a
convenient indicator of the reaction. This discovery was then
evaluated in other O-(nitrophenyl)hydroxylamine derivatives
and was found to be a general transformation for this
compound class. In addition, remarkable chemoselectivity was
observed for this amide-forming step in comparison to the
more classic approach using an amine and an acid chloride.
Coupling of this reagent 8a with a series of aldehydes and
ketones (12) provided the respective intermediate O-(4-
nitrophenyl) oximes (13), which could be isolated by column
chromatography or used directly in subsequent steps (see the
Supporting Information). Attempts to reduce these intermedi-
ate oximes to their related N-substituted O-(4-nitrophenyl)-
hydroxylamines 14, however, proved problematic. Sodium
borohydride, LiAlH4, NaBH(OAc)3,32 and H2 gas over Pd−C
all provided the over-reduced amine product, complex
mixtures, or no reaction. An extensive review of the literature
found limited reports associated with the reduction of N-alkyl
O-substituted oximes to their corresponding O-substituted
hydroxylamines.27−31 The consensus from these references
indicated that moderate yields could be obtained using
NaBH3CN at pH 3, when the pH is maintained by the
addition of inorganic acids dissolved in alcoholic solvents.27−31
NaBH3CN is a unique hydride source due to its stability to
hydrolysis at low pH.34 This feature is critical, as a competition
likely exists between hydrolysis of the reactive hydride reagent
at low pH34 and the desired hydride addition to the oxime
(which is presumably activated via protonation at low pH).
Our own investigation of this challenging selective oxime
reduction revealed that a two-stage pulsed approach worked
best at pH 3. Specifically, addition of NaBH3CN (3 equiv) to
the oxime 13a (R = Bn, R′ = X = H, Table 1) at pH 3
(monitored via a methyl orange indicator) gave approximately
50% conversion after 2 h. The pH was carefully maintained
over this 2 h period by addition of an organic acid,
RESULTS AND DISCUSSION
■
Synthesis. While there are several methods to generate the
O-phenylhydroxylamines, there are only limited synthetic
entries to the N-alkyl-O-phenylhydroxylamines, some of
which rely on organometallic reagents.22,27−31 After an
extensive screening of possible chemistries, we elected to
pursue the reductive amination approach using O-(4-
nitrophenyl)hydroxylamine (8a) and the desired ketone or
aldehyde. While the commercial cost of 8a is significant, a
recent patent described a high-yield entry to these systems
10837
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
a
Table 2. Amide Synthesis Reactions Using O-Substituted Hydroxylamines
O-arylhydroxylamine 14
reagent a or b
yield of amide 16, %
phenol 17
14a (R = Bn, R′ = H, X = H)
14b: (R = Me, R′ = H, X = H)
14c (R = n-C6H13, R′ = H, X = H)
14d (R = R′ = Me, X = H)
14e (R = Bn, R′ = H, X = CF3)
14f (R = n-C6H13, R′ = H, X = NO2)
14a (R = Bn, R′ = H, X = H)
a
a
a
a
a
a
b
99 (16a; R = Bn, R′ = H, R2 = Me)
17a (X = H)
17a (X = H)
17a (X = H)
17a (X = H)
17b (X = CF3)
17c (X = NO2)
95 (16b; R = R2 = Me, R′ = H)
71 (16c; R = n-C6H13, R′ = H, R2 = Me)
59 (16d; R = R′ = R2 = Me)
83 (16b; R = Bn, R′ = H, R2 = Me)
75 (16c; R = n-C6H13, R′ = H, R2 = Me)
100 (16e; R = R2 = Bn, R′ = H)
17a (X = H)
1
a
Reagents: (a) 75% t-BuOH in water, 65 °C, 18 h; (b) 68% DMSO in water, 65 °C, 15 h. Note: yields were determined by H NMR using an
internal standard.
methanesulfonic acid, in small increments (0.1 equiv). Typically
3 equiv of methanesulfonic acid was required to maintain the
pH over this 2 h period. However, the amount of CH3SO3H
needed was system dependent. Importantly, no further reaction
was observed upon continued stirring beyond 2 h, indicating
that the hydride reagent was consumed during this time period.
Additional NaBH3CN (3 equiv) and subsequent methanesul-
fonic acid additions were used to push the reaction closer to
completion and provided good conversions to the desired N-
substituted O-(4-nitrophenyl)hydroxylamine 14a.
Surprisingly, the related oximes generated from benzalde-
hyde, acetophenone, and 1H-indole-3-carbaldehyde did not
react under these optimized reduction conditions. Efforts to
reduce these bulky oximes with Pyr-BH3 gave no reaction, even
though the related O-methyloximes were converted in high
yield to O-methylhydroxylamines with this reagent by Kawase
et al.33
Using our pulsed-reduction approach, the desired O-(4-
nitrophenyl)hydroxylamines (14a−d) were synthesized in
good yields (65−75%, Table 1). These new materials 14a−d
were stable to column chromatography on silica gel and were
characterized by 1H and 13C NMR and mass spectrometry and
elemental analysis (see the Supporting Information). These N-
substituted hydroxylamines (14a−d) were then evaluated by
1H NMR for their ability to react with pyruvic acid in 75% t-
BuOH−water, and the results are shown in Table 2.2 The N-(2-
phenylethyl) system 14a was prepared in order to compare to
the prior system 1b.
was circumvented by using acidified water during the workup of
8c (see the Experimental Section) but still resulted in low yields
for 8c.
The substituted phenyl systems 8b,c were reacted with
phenacetaldehyde and heptaldehyde, respectively, to form their
crude respective oximes 13e,f, which were reduced with
NaBH3CN at pH 3 to the respective N-alkylhydroxylamines
14e (15%) and n-heptylhydroxylamine 14f (5%; see Table 1).
Again, the dinitro system was problematic in general and
decomposed during isolation using column chromatography at
both the oxime and the hydroxylamine stages. The 4-nitro-(3-
trifluoromethyl)phenyl system derived from 8b (14e) was
better behaved throughout its synthesis and was stable to silica
gel based chromatography.
Compounds 14e,f were also shown to react with pyruvic acid
(2 equiv) to give the desired acetamides 16a,c in 83% and 75%
yields, respectively (Table 2). In an effort to demonstrate the
general reactivity of α-keto acids to these systems, hydroxyl-
amine 14a was reacted with phenylpyruvic acid (2 equiv) and
gave a 100% yield of the desired 2-phenyl-acetamide 16e at
86% conversion in 68% DMSO−water after 15 h at 65 °C
(Table 2).
Next, we compared the relative reactivities of 1b, 14a, and
14e with pyruvic acid using pseudo-first-order reaction
conditions (Table 3) by HPLC. The rate constant for the
Table 3. Comparative Rates for 1b, 14e, and 14a with
Pyruvic Acid at 65 °C in 50% t-BuOH/Water As Determined
a
by HPLC
Yields of the acetamide products in Table 2 were determined
1
by H NMR using a dibenzyl ether internal standard (after an
aqueous workup to remove unreacted pyruvic acid). These
initial 1H NMR results (Table 2) suggested high conversion of
14a and high yield of the product acetamide 16a.
Other substituted O-(aryl)hydroxylamines were also synthe-
sized to investigate the relative reactivities of substituted O-
phenyl systems in this reaction (Scheme 3). It was expected
that electron-deficient systems would enhance the leaving
group ability of the N-phenoxy substituent and provide systems
with increased reactivity. Thus, the related O-(4-nitro-3-
(trifluoromethyl)phenyl)hydroxylamine (8b; 72%) and O-
(3,4-dinitrophenyl)hydroxylamine (8c; 15%) were synthesized
using modifications of the methods used to synthesize the 4-
nitrophenyl system 8a. Although the trifluoromethyl system 8b
was well-behaved and stable to the workup, the related dinitro
system 8c was highly reactive and decomposed when exposed
to aqueous base (aqueous Na2CO3) during its isolation. This
compd
rate (min‑1)
pKa
1b (R = benzoyl)
0.0184
0.0122
0.0047
4.20
6.07
7.08
14e (R = 3-(trifluoromethyl)-4-nitrophenyl)
14a (R = 4-nitrophenyl)
a
Pseudo-first-order conditions (10 equiv of pyruvic acid) were used. A
plot of pKa (x axis) vs rate (min−1, y axis) gave a line defined by y =
−0.0048x + 0.0395; r2 = 0.96.
10838
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
respective reactions between pyruvic acid (100 mM) and 1b,
14a, or 14e (10 mM) were calculated using diphenylmethane
(DPM, 10 mM) as an internal standard to determine the
amount of starting material consumed per unit time (min). As
shown in Table 3, the rate for the benzoyloxy derivative 1b was
approximately 1.5 times faster than that for 14e and nearly 4
times faster than that for 14a. This is consistent with the
relative leaving group ability of these systems: e.g., the pKa
value of benzoic acid (4.2), 3-trifluoromethyl-4-nitrophenol
(6.07), and 4-nitrophenol (7.08). This relationship was
approximated via linear regression (r2 = 0.96) to give a line
defined as
Table 5. Solvent Study in t-BuOH−Water Mixtures for N-4-
Nitrophenoxy Derivative 14a at 65 °C
rate (in min−1) = −0.0048 (pKa value of the substituted
yield of
amide 16a, %
est yield of 16a based upon
b
a
t-BuOH, % conversn,
%
conversn, %
50
60
34
43
42
49
56
62
15
20
22
26
32
22
44
47
52
53
57
36
phenol) + 0.0395
This equation describes the relationship between rate and pKa
and can be used to estimate the rates of future O-phenyl
systems via their respective phenol’s pKa value.
Using the HPLC analysis method, these reactions were
further optimized for solvent composition. As shown in Table
4, a 70% solution (by volume) of t-BuOH in water at pH 7.4
(1)
70
80
90
100
a
b
Based on starting material consumed as measured by HPLC. Taking
into account the percent conversion: e.g., for the 90% reaction 32/0.56
= 57.
Table 4. Solvent Study in t-BuOH−Water Mixtures for N-
Benzoyloxy Derivative 1b at 65 °C
Table 6. Solvent System Study for 14a in DMSO−Water
Mixtures at 65 °C and pH 7.4
amt of
t-BuOH, % conversn,
yield of amide est yield of 16a based upon
b
a
%
16a, %
conversn, %
amt of
yield of amide est yield of 16a based
b
50
60
83
93
94
94
98
89
43
79
80
78
74
26
52
85
85
83
76
29
a
entry DMSO, % conversn,
%
16a, %
upon conversn, %
c
1
2
10
20
30
40
50
32
32
64
76
76
86
80
73
67
63
8
11
20
43
73
78
67
58
38
30
23
34
31
56
96
91
84
79
57
70
c
c
c
c
80
3
90
4
100
5
a
b
Based on starting material consumed. Taking into account the
percent conversion: e.g., for the 70% reaction 80/0.94 = 85.
6
60
70
7
8
80
using 2 equiv of pyruvic acid and 1 equiv of the N-benzoyloxy
system 1b gave the highest yield of 16a (85%). In contrast, the
N-(4-nitrophenoxy)amine 14a (Table 5) derivative gave the
highest yield of 16a (57%) in 90% t-BuOH in water at the same
pH and molarity of reagents.
9
90
10
100
48
a
b
Based on starting material consumed as measured by HPLC. Taking
into account the percent conversion: e.g., for the 60% reaction 78/0.86
= 91. Reaction mixtures appeared turbid.
c
The reaction between 14a and pyruvic acid was also
performed in other solvents (i.e., EtOAc, DMF, acetonitrile,
and DMSO) in an effort to increase the yield of 16a. Of these
solvents, DMSO had the fewest byproducts formed and the
highest yields and was chosen for further optimization
experiments by the addition of water at pH 7.4.
As shown in Table 6, a 60% solution of DMSO in water (by
volume at pH 7.4) gave high conversion (86%) and 78% yield.
This yield corresponds to a high yield of 16a (91% yield) when
one corrects for the unreacted starting material. Note:
experiments performed at higher water loadings (e.g., 10−
50% DMSO in water) provided lower percent conversions and
turbid mixtures and introduced significant solubility constraints
of the internal standard due to the higher water content.
Nevertheless, this survey of reaction conditions allowed us to
identify optimal conditions for this transformation (60%
DMSO/water mixture).
The dramatic solvent effect on this reaction, where a 60%
DMSO/water mixture resulted in significantly higher reac-
tivities, yields, and conversions for the N-(4-nitrophenoxy)-
amine system 14a, can be explained by two factors. First, water
plays a functional role in the mechanism of this reaction and is
10839
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
needed for hydration of the putative oximinium ion
intermediate 7 to form the product amide. We have shown
that mixed solvent systems containing water further facilitate
conversion to the product (Table 6: entry 6 vs entry 10).
Second, solvents with high dielectric constants such as water (ε
= 78) and DMSO (ε = 47) greatly facilitate this reaction.
Although dielectric constants are a bulk property of the solvent,
they are a good indicator of the solvent’s ability to
accommodate separation of charge.35a Indeed, mixtures
containing solvents with high dielectric constants (DMSO
and water) were more optimal than the related t-BuOH (ε =
12.5)/water mixtures. We speculate that these mixed aqueous
solvents containing polar aprotic solvents such as DMSO may
help lower the energy of the transition state, leading to ionic
intermediates such as 7.35a In general, solvents with lower
dielectric constants resulted in lower yields (e.g., solvent/
dielectric constant ε: DMF/38; CH3CN/37, EtOAc/6).
The fact that the chemical reactivity of these systems can be
modulated via changes in the steric demands of the N-alkyl
substituent, the electron richness of the phenoxy substituent,
and the reaction solvent composition provides additional
flexibility in developing new biological applications for these
self-reporting systems. Indeed, a preliminary assessment in
Chinese hamster ovary (CHO) cells demonstrated that both
14a and 14c were nontoxic at 100 μM after a 48 h incubation at
37 °C.
in the presence of NAD+ and pyruvate dehydrogenase.35b Even
so, in the presence of 2 equiv of cysteine hydrochloride
monohydrate, 1 equiv of 14a still provided a 36% yield of 16a.
In sharp contrast, control experiments using the more classic
reagents phenethylamine and acetyl chloride gave trace yields
of 16a (≤1% yield) after identical challenges and workup.
These experiments demonstrated that significant chemo-
selectivity is present in the α-keto acid reaction with 14a.
Conclusion. Using a pulsed addition of NaBH3CN at pH 3,
a series of O-(4-nitrophenyl)hydroxylamines were synthesized
from their respective oximes. Careful pH control was necessary
and was maintained by a methyl orange indicator and
incremental additions of methanesulfonic acid. Yields of the
desired O-substituted hydroxylamines ranged from 5 to 75%.
Steric hindrance near the oxime functional group was shown to
play a key role in both the ease by which the oxime could be
reduced and the subsequent reactivity of the respective
hydroxylamine. As expected, the less hindered systems were
more reactive.
Reaction of the respective hydroxylamines with pyruvic acid
generated the desired amide derivatives in good yields. A
solvent composition study was conducted and revealed that
high yields could be obtained with 14a in 60% DMSO/water at
pH 7.4. A rate comparison of phenethylamine systems bearing
N-OBz, N-3-trifluoro(4-nitrophenoxy), and N-(4-nitrophe-
noxy) groups revealed significant differences in the rates of
these systems and suggested that the leaving group ability of the
N substituent plays a defining role in determining the reactivity
of these systems. This insight allows one to further refine the
selectivity and reactivity of these systems by modulating the
electron richness of the appended O-phenyl substituent.
Alterations in the size of the N-alkyl substituent and the
solvent system also were shown to influence the rate of these
reactions.
An assessment of the chemoselectivity of this reaction was
conducted, and the results are shown in Table 7. Compound
a
Table 7. Chemoselectivity Experiments
Many preparation methods exist to provide amides from
amine starting materials using reactive electrophiles such as acid
chlorides. These reactive electrophiles can react with numerous
competing nucleophiles, including thiols, phenols, and amines
(e.g., Table 7, entry 6). In contrast, our results suggest that high
chemoselectivity is present in the coupling of N,O-substituted
hydroxylamines and α-keto acids in aqueous solvent mixtures
(e.g., Table 7, entry 1).5 Other type II hydroxylamine systems
have recently been shown to have high chemoselectivity in their
reactions with α-keto acids. For example, unprotected peptides
containing a terminal cyclic hydroxylamine (5-oxaproline) were
shown to couple with unprotected α-keto acid containing
peptides in 46% yield (in 90% DMSO/water at 60 °C for 20 h)
in the presence of 0.1 M oxalic acid).4c In this regard, it may be
possible to further increase reaction efficiencies by the addition
of specific additives such as oxalic acid. We also note that a
related chemoselective reaction between α-keto acids and
oximes has been described by Bode et al.36a Since these
reactions work particularly well at pH 7.4, future work will
explore the biological applications of N,O-substituted hydrox-
ylamine−α-keto acid couplings.36b Indeed, the high chemo-
selectivity observed here can be used to develop future
reagents, which perform selective acylation reactions in the
presence of biological nucleophiles.
starting
compd
yield of 16a,
entry
Y
additive
%
1
2
3
4
14a
14a
14a
14a
COOH lysine and phenol
COOH glycine
98
93
36
39
COOH cysteine
COOH cysteine and hexanoic
acid
5
6
18
18
Cl
cysteine and hexanoic
acid
0
1
Cl
lysine and phenol
a
Reagents: 68% by volume DMSO in phosphate buffered saline at 65
°C for 18 h in the presence of a diphenylmethane internal standard
(see the Experimental Section for more details).
14a (1 equiv) and pyruvic acid (2 equiv) were challenged to
react in the presence of competing nucleophiles (2 equiv each)
such as cysteine, hexanoic acid, lysine, glycine, and phenol in
68% DMSO/water at 65 °C for 18 h. Competition experiments
involving glycine, lysine, and phenol each provided high yields
of the desired acetamide (≥93% yield), indicating that the
presence of these functional groups was not detrimental to the
coupling chemistry to form 16a.
In contrast, cysteine (Table 7, entry 3) was shown to
compete with this reaction at pH 7.4, whereas glycine did not
(entry 2). This suggested that thiol groups can compete with
14a for pyruvic acid. It is well-known that, during aerobic
cellular respiration, thiols can react with pyruvic acid to form
thioesters. For example, acetyl-coA is formed by this chemistry
EXPERIMENTAL SECTION
■
General Considerations. All solvents and most reagents were
used as received from the vendor, and column chromatography was
run on silica gel (40−63 μm) with the indicated solvent systems. High-
10840
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
resolution mass spectra (HRMS) were acquired via electrospray
ionization and an Agilent 6224 time-of-flight mass spectrometer.
O-(4-Nitrophenyl)hydroxylamine (8a).16 A solution of ethyl
acetimidate 9 (7.48 g: 72.5 mmol) and 4-chloronitrobenzene 10a
(9.13 g: 58 mmol) in DMF (15 mL) was stirred mechanically at room
temperature. The yellow solution was warmed to 30 °C, and solid
NaOH (2.83 g: 70.7 mmol) was added in six portions. The reaction
mixture was sonicated to facilitate reaction. The reaction mixture was
then warmed to 42 °C and stirred for 1.5 h. The solid NaOH pellets
slowly dissolved and formed a yellow suspension over time
(presumably containing a NaCl precipitate). The reaction mixture
from 11b), δ 2.16 (s from 11b), and δ 1.38 (t from 11b). The reaction
mixture was cooled to room temperature, 5 mL of acidified water (at
pH 2) was added to the reaction flask, and the solution was extracted
three times with CH2Cl2. The organic layer was separated, dried over
anhydrous Na2SO4, filtered, and concentrated to give 8b (530 mg,
72%) as a solid.38 Significantly higher yields were obtained when these
compounds were isolated under acidic conditions.
1
Compound 8b: H NMR (CDCl3) δ 8.00 (d, 1H), 7.61 (d, 1H),
7.41 (d, 1H); 13C NMR (CDCl3) δ 163.4, 140.1, 126.8, 124.8 (q, 2JC−F
1
3
= 34 Hz), 120.9 (q, JC−F = 272 Hz), 115.2, 112.1 (q, JC−F = 6 Hz).
O-(3,4-Dinitrophenyl)hydroxylamine (8c). A solution of ethyl
acetimidate (9; 5.0 g, 48 mmol) was stirred with 3,4-dinitrochlor-
obenzene (10c; 7.85 g, 38 mmol) in DMF (10 mL) at room
temperature. The yellow solution was warmed to 30 °C (via an oil
bath), and solid NaOH (1.92 g, 48 mmol) was added in four to five
portions. A mechanical stirrer was used for good stirring. The reaction
mixture was then sonicated for 1−1.5 h to ensure that all the solid
dissolved into the solution and then warmed to 40 °C for 2 h. The
solid NaOH pellets began to dissolve and formed a yellow suspension.
The reaction was completed in 3 h and was monitored by TLC and 1H
1
was stirred overnight. H NMR (CDCl3) provided a convenient way
to monitor the reaction using the following signals: δ 7.53 (d, from
chloronitrobenzene 10a), 7.23 (d, product 11a), 4.21 (q, 11a), 3.99
(q, starting ester 9). The aromatic signal integration ratio of δ 7.53
(10a) vs δ 7.23 (11a) or TLC (33% hexane in CHCl3) with Rf = 0.62
for 10a and Rf = 0.41 for 11a was convenient to monitor percent
conversion over time. Water (24 mL) was then added at 40 °C, the
slurry that formed was cooled to 15 °C and filtered, and the solid was
washed with water. The yellow solid was air-dried and gave the
intermediate 11a (14.24 g, 88% yield). The subsequent hydrolysis step
involved adding 11a (14.24 g) to 37% HCl (11.9 g solution, 10 mL,
4.4 g of HCl: 0.12 mol) in portions over 20 min at room temperature.
The white suspension was vigorously sonicated in order to ensure
complete conversion to 8a. After 2 h of sonication, the white
suspension was neutralized with 33% NaOH solution (55 mL), cooled
to room temperature, and filtered and the yellow solid washed with
1
NMR (CDCl3). H NMR provided a better way of monitoring using
the disappearance of the singlet at δ 7.89 (from 10c) and appearance
of signals at δ 8.00 (d, from 11c), δ 7.94 (d, from 11c), δ 7.71 (s, from
11c), and δ 4.2 (q, from 11c). On TLC two spots were observed,
whose intensities corresponded to the integration ratio in 1H NMR for
the E and Z isomers. TLC (33% hexane/CH2Cl2) gave Rf = 0.37 for
10c and Rf = 0.53 and 0.28 for the two isomers of 11c. Water (20 mL)
was added at 40 °C, and the reaction mixture was cooled to 15 °C and
stirred for 15 min. A yellow solid appeared, which was filtered and
washed with fresh water to give the intermediate 11c (9.0 g, 68%
yield). The subsequent hydrolysis step involves adding 11c (9.0 g)
into 37% HCl solution (23.3 mL, 280 mmol, 12 M) in portions over a
period of 20 min at room temperature. A mechanical stirrer was used
for complete conversion to 8c. The reaction was then sonicated for 2.5
h and monitored by TLC (40% hexane/CH2Cl2): Rf = 0.55 and 0.34
(E and Z isomers of 11c) and Rf = 0.25 for 8c. The reaction was
cooled to room temperature, 50 mL of acidified water (at pH 2, see
note below) was added to the reaction flask, and the solution was
extracted three times with CH2Cl2. The organic layer was dried over
anhydrous Na2SO4, filtered, and concentrated to give 8c21a as a crude
unstable solid (1.0 g, yield 15%). (Note: this yield was later improved
to 52% by not adding acidified water and simply extracting the material
with CH2Cl2 from the concentrated HCl layer as its HCl salt. Excess
CH2Cl2 was needed, as the HCl salt is only partially soluble in
CH2Cl2.)
1
water and air-dried (11.82 g). H NMR of the crude solid showed
complete conversion (e.g., loss of ethyl pattern). Efforts to further
purify this material by column chromatography (100% CHCl3) gave
significant reductions in yield and provided the pure O-(4-
nitrophenyl)hydroxylamine product 8a (4.5 g, 50% yield, Rf(100%
CHCl3) = 0.38) as a solid. Later experiments demonstrated that the air-
dried crude solid could be used without further purification for oxime 13
formation in the next step. We noted that the main impurity was 4-
nitrophenol ( 17a).
Caution! A series of related O-phenylhydroxylamines have been
screened by differential scanning calorimetry (DSC) and accelerating
rate calorimetry (ARC), and the O-(4-nitrophenyl)hydroxylamine 8a
had an onset temperature for its decomposition of 120 °C (vs 90 °C
for the related O-(2,4-dinitrophenyl)hydroxylamine).37 Although no
violent decompositions were observed during our investigations,
caution should be taken not to heat these samples near their
decomposition temperature, as a violent explosion could result.37
1
Compound 8a: H NMR (CDCl3) δ 8.20 (d, 2H), 7.24 (d, 2H),
6.10 (br s, 2H, O-NH2); 13C NMR (CDCl3) δ 166.4, 141.8, 125.7,
113.3.
1
Compound 8c: H NMR (CDCl3) δ 7.901 (d, 1H, J = 8.8 Hz),
7.900 (d, 1H, J = 2.2 Hz), 7.02 (dd, 1H, J = 8.8 and 2.2 Hz), 6.40 (s,
2H, O-NH2) (note that the signal at δ 7.90 appears as dd but is
actually two separate doublets, as is evident by their respective
coupling constants at 7.02 ppm); 13C NMR (CDCl3) δ 156.0, 141.7,
135.7, 126.7, 121.2, 116.2.
O-(4-Nitrobenzotrifluoride)hydroxylamine (8b). A solution of
ethyl acetimidate (9; 0.5 g, 4.8 mmol) was stirred with 5-chloro-2-
nitrobenzotrifluoride (10b; 0.85 g, 3.8 mmol) in DMF (2 mL) at room
temperature. The yellow solution was warmed to 30 °C (via an oil
bath), and solid NaOH (0.192 g; 4.8 mmol) was added in four to five
portions. A mechanical stirrer was used to maintain good stirring. The
reaction mixture was then sonicated for 1.5 h to ensure that all the
solid dissolved into the solution. The reaction was completed after 1.5
h of sonication (color changed from yellow to yellowish orange). The
reaction was monitored by TLC (40% CH2Cl2/hexane; Rf = 0.6 for
General Method for Preparation of O-(4-Nitrophenyl) N-
Substituted Oximes 13. The O-4-nitrophenylhydroxylamine 8a (1
mmol) was dissolved in absolute EtOH (8 mL) and the respective
aldehyde or ketone 12 (1 mmol) was added dropwise. Note: a drop of
trifluoroacetic acid was added to facilitate some of the reactions (i.e.,
with acetaldehyde, acetone, and acetophenone). The reaction mixture
was stirred at room temperature for 24 h and monitored by TLC (e.g.,
50% hexane in CHCl3: Rf = 0.35 for 14a). Note: the reaction rates
varied and the phenacetaldehyde reaction was complete within 2 h.
After TLC indicated high conversion, the solution was concentrated to
give the crude oxime, which could either be purified by column
1
10b and Rf = 0.4 for 11b) and H NMR (CDCl3) by following the
appearance of δ 4.21 (q from 11b), δ 2.16 (s from 11b), and δ 1.38 (t
from 11b). After all the starting material was consumed as indicated by
TLC, water (2 mL) was added at 40 °C and the reaction mixture was
cooled to 15 °C and stirred for 15 min. A yellow solid appeared, which
was filtered, washed with water, and air-dried to give the intermediate
11b (900 mg, 86% yield). The subsequent hydrolysis step involved
adding 11b (100 mg, 0.3 mmol) to 37% HCl (0.225 mL, 2.7 mmol, 12
M) in portions over a period of 20 min. A mechanical stirrer was used
for proper stirring and to facilitate conversion to 8b. The reaction was
then sonicated for 2.5 h at room temperature and monitored by TLC
1
chromatography or used directly in the next step. The H NMR data
and isolated yields for the oxime intermediates are listed to show that
the first step (oxime formation) proceeded generally in high yield.
2-Phenylacetaldehyde O-(4-Nitrophenyl) Oxime (13a): 210 mg,
82% yield as a mixture of E/Z oximes in ∼2:1 ratio was observed,
1
which complicated the spectra; H NMR (CDCl3) δ 8.35−8.20 (two
1
(40% hexane/CH2Cl2; Rf = 0.5 for 11b and Rf = 0.3 for 8b). The H
doublets, 2H), 7.90 (t, 0.66 H), 7.47−7.20 (m, 7.3 H), 3.93 and 3.70
NMR (CDCl3) spectra showed the disappearance of peaks δ 4.21 (q
(two doublets, 2H).
10841
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
Acetaldehyde O-(4-Nitrophenyl) Oxime (13b): 169 mg, 94% yield
as a mixture of E/Z oximes in a 3/2 ratio was observed, which
complicated the spectra; 1H NMR (CDCl3) δ 8.25−8.10 (two
doublets, 2H), 7.85 (q, 0.6H), 7.30−7.15 (m, 2.4H), 2.10−2.00 (two
doublets, 3H).
N-Ethyl-O-(4-nitrophenyl)hydroxylamine (14b): oil, 136 mg, 75%
1
yield; TLC 50% CHCl3/hexane, Rf = 0.4; H NMR (CDCl3) δ 8.17
(d, 2H, 3JH−H = 9.34 Hz), 7.21 (d, 2H, 3JH−H = 9.34 Hz), 6.10 (t, 1H),
3.19 (m, 2H), 1.20 (t, 3H); 13C NMR (CDCl3) δ 166.1, 141.6, 125.7,
113.6, 46.8, 12.2, HRMS calcd for C8H11N2O3 (M + H) 183.0764,
found 183.0765. Anal. Calcd for C8H10N2O3: C, 52.74; H, 5.53; N,
15.38. Found: C, 52.86; H, 5.59; N, 15.22.
Heptanal O-(4-Nitrophenyl) Oxime (13c): 245 mg, 98% yield as a
mixture of E/Z oximes in a ∼2.3:1 ratio was observed, which
complicated the spectra; 1H NMR (CDCl3) δ 8.25−8.20 (merged
doublets, 2H), 7.81 and 7.07 (two triplets in a 2.3:1 ratio, total 1H),
7.23 (d, 2H), 2.56 and 2.38 (two multiplets, total 2H), 1.60 (m, 2H),
1.45−1.28 (m, 6H), 0.91 (m, 3H).
N-Heptyl-O-(4-nitrophenyl)hydroxylamine (14c): oil, 184 mg, 73%
1
yield; H NMR (CDCl3) δ 8.23 (d, 2H), 7.20 (d, 2H), 6.20 (t, 1H),
3.15 (m, 2H), 1.53 (m, 2H), 1.30 (m, 8H), 0.88 (m, 3H); 13C NMR
(CDCl3) δ 166.0, 141.6, 125.8, 113.6, 52.4, 31.7, 29.1, 27.1, 27.0, 22.6,
14.1. Anal. Calcd for C13H20N2O3: C, 61.88; H, 7.99; N, 11.10. Found:
C, 62.15; H, 8.02; N, 11.07.
Propan-2-one O-(4-Nitrophenyl) Oxime (13d): 194 mg, 100%
1
yield; H NMR (CDCl3) δ 8.20 (d, 2H), 7.25 (d, 2H), 2.08 (s, 3H),
N-Isopropyl-O-(4-nitrophenyl)hydroxylamine (14d): oil, 127 mg,
65% yield; 1H NMR (CDCl3) δ 8.17 (d, 2H, J = 9.2 Hz), 7.23 (d, 2H,
J = 9.2 Hz), 5.86 (d, 1H, NH), 3.43 (m, 1H), 1.13 (d, 6H, J = 6.2 Hz);
13C NMR (CDCl3) δ 166.5, 141.0, 125.7, 113.7, 52.3, 19.8; HRMS
calcd for C9H13N2O3 (M + H) 197.0921, found 197.0916. Anal. Calcd
for C9H12N2O3: C, 55.04; H, 6.16; N, 14.28. Found: C, 55.36; H, 6.20;
N, 14.28.
2.07 (s, 3H).
Note: oximes 13e,f are described separately in the syntheses of
14e,f.
Benzaldehyde O-(4-Nitrophenyl) Oxime (13g): 217 mg, 90% yield,
presumably the all-E isomer; H NMR (CDCl3) δ 8.48 (s, 1H), 8.26
(d, 2H), 7.75 (d, 2H), 7.49 (m, 3H), 7.37 (d, 2H).
1
Acetophenone O-(4-Nitrophenyl) Oxime (13h): 256 mg, 100%
yield; TLC (50% hexane/50% CHCl3) Rf = 0.35; 1H NMR (CDCl3) δ
8.25 (d, 2H, J = 8.7 Hz), 7.79 (d, 2H), 7.47 (br s, 3H), 7.40 (d, 2H, J =
8.7 Hz), 2.50 (s, 3H); 13C NMR (CDCl3) δ 164.2, 160.2, 142.4, 135.1,
130.5, 128.7, 126.7, 125.8, 114.5, 13.8.
1H-Indole-3-Carbaldehyde O-(4-Nitrophenyl) Oxime (13i). The
O-(4-nitrophenyl)hydroxylamine 8a (150 mg, 0.97 mmol) was
dissolved in absolute EtOH (7 mL) and TFA (7.22 μL). The
indole-3-carboxaldehyde 12g (141 mg, 0.97 mmol) was added
dropwise over 1 min. The mixture was heated to 80 °C and stirred
overnight. The solution was concentrated under vacuum to give oxime
13i as a bright yellow solid. The solid was recrystallized with ethyl
acetate/hexane to yield 95 mg (40%) of the product. 1H NMR
(acetone-d6): δ 8.88 (s, 1H) 8.32 (d, 2H) 8.28 (d, 1H) 7.95 (d, 1H)
7.55 (m, 3H), 7.30 (m, 2H).
General Method for Reduction of O-(4-Nitrophenyl) N-
Substituted Oximes to Hydroxylamines. The oxime (1 mmol)
was dissolved in MeOH (3 mL) and treated with sodium
cyanoborohydride NaBH3CN (3 mmol). Using a trace of methyl
orange as an in situ indicator, pH 3 was attained by adding
methanesulfonic acid (1 mmol) to maintain a consistent red color.
The pH was controlled by adding incremental aliquots of
methanesulfonic acid (0.1 mmol) to the reaction mixture every time
it turned from red to orange, which indicated pH >3. At the end of 2 h,
no further color changes were observed (and no additional conversion
was noted if stirred beyond 2 h). At this point, NaBH3CN (3 mmol)
was again added to drive the reaction to completion and the pH was
again controlled via incremental methanesulfonic acid addition. The
reaction was stirred for 2 h at pH 3 after the second pulse of
NaBH3CN was added. The reaction was monitored by TLC (e.g., 50%
hexane in CHCl3 for 14a) for high conversion. The solution was
concentrated, dissolved in dichloromethane, and washed three times
with aqueous Na2CO3. The organic layer was separated, dried over
anhydrous Na2SO4, filtered, and concentrated to give the crude
hydroxylamine derivative, which was purified by column chromatog-
raphy on silica gel. Isolated yields ranged from 5 to 75%, depending
upon the substrate. It was noted that continued pulsed additions of
reducing agent (i.e., beyond two 3 equiv NaBH3CN pulses) resulted in
the generation of increasing amounts of the corresponding amine (e.g.,
phenethylamine), presumably via reduction of the desired hydroxyl-
amine product. In general, two pulses seemed best at balancing over-
reduction vs product yield, but this trend was also substrate
dependent.
O-(4-Nitrobenzotrifluoride)-N-phenethylhydroxylamine (14e):
The O-(4-nitrobenzotrifluoride)hydroxylamine 8b (500 mg, 2.25
mmol) was dissolved in absolute EtOH (20 mL), and phenyl-
acetaldehyde (12a; 270 mg, 2.25 mmol) was added dropwise at room
temperature. The mixture was stirred at room temperature and
monitored by TLC (40% CH2Cl2/hexane; Rf = 0.45 for oxime 13e)
1
and H NMR (CDCl3). The reaction was completed in 4 h. The
solution was concentrated under vacuum. A crude oil (990 mg) was
obtained, which was used directly in the next reaction. Crude 13e: 1H
NMR (CDCl3) δ 8.1−8.0 (d, 2H), 7.90 (t, 1H), 7.5−7.3 (m, 5H), 3.9
(d, 2H), 3.4 (d, 1H). The crude oxime 13e (990 mg) was dissolved in
methanol (20 mL), and NaBH3CN (188 mg, 3.0 mmol) was added. A
trace of methyl orange was added as an indicator to monitor the pH
needed for the reduction. Methanesulfonic acid (28 mg, 0.3 mmol)
was added, and the pH was controlled by adding aliquots (0.3 mmol)
of methanesulfonic acid to the reaction mixture every time it turned
orange, indicating a higher pH. The reaction was monitored by TLC
(40% CH2Cl2 in hexane). After each addition of hydride, the reaction
mixture was stirred for 2 h and the pH adjusted accordingly. Four
additional rounds of NaBH3CN addition were performed to facilitate
conversion (total NaBH3CN added 412 mg, 6.56 mmol) and the
reaction mixture was stirred at room temperature overnight. After 15 h
at room temperature, the reaction was approximately 60% complete by
TLC. The solution was concentrated, dissolved in dichloromethane,
and washed three times with water (pH 2). Note: exposure to aqueous
Na2CO3 decomposed the product. The organic layer was separated, dried
over anhydrous Na2SO4, filtered, and concentrated to give the crude
product 14e. The product mixture was purified by flash column
chromatography at Rf = 0.45 with 10% CH2Cl2/hexane to give product
1
14e as a yellow oil (105 mg, 15% yield). Compound 14e: H NMR
(CDCl3) δ 7.90 (d, 2H), 7.60 (s, 1H), 7.3−7.45 (m, 5H), 6.25 (t, 1H,
NH), 3.40 (t, 2H), 2.9 (t, 2H); 13C NMR (CDCl3) δ 163.7, 138.2,
128.8, 128.7, 127.8, 126.7, 116.5, 113.5, 113.4, 53.2, 33.2; HRMS calcd
for C15H14F3N2O3 (M + H) 327.0951, found 327.0958.
N-Heptyl-O-(3,4-dinitrophenyl)hydroxylamine (14f): The O-(3,4-
dinitrophenyl)hydroxylamine 8c (100 mg, 0.5 mmol) was dissolved in
absolute EtOH (5 mL), and heptaldehyde (12c; 57 mg, 0.5 mmol)
was added dropwise. The mixture was stirred at room temperature and
monitored by TLC (40% CH2Cl2/hexane), showing two spots
corresponding to the two isomers of oxime 13f at Rf = 0.70 and
0.65. The reaction was completed in 4 h. The solution was
concentrated under vacuum. and the product obtained (oil, 110 mg,
60%) was used directly in the next step. Crude 13f: 1H NMR (CDCl3;
a 3:1 mixture of E/Z isomers was obtained, which complicated the
spectra) δ 7.92 (m, 2H), 7.75 (s, 0.25H), 7.72 (s, 0.75H), 7.13 (t,
0.25H), 7.05 (t, 0.75H), 2.62 (q, 0.5H), 2.39 (q, 1.5H), 1.57 (m, 4H),
1.39 (m, 2H), 1.33 (m, 4H), 0.91 (t, 3H). The oxime 13f (110 mg, 0.3
mmol) was dissolved in methanol (2 mL), and sodium cyanoborohy-
dride (56 mg, 0.9 mmol) was added. A trace of methyl orange was
added as an indicator to monitor the pH during the reduction.
O-(4-Nitrophenyl)-N-phenethylhydroxylamine (14a): oil, 167 mg,
65% yield; 1H NMR (CDCl3) δ 8.17 (d, 2H; J = 9.4 Hz), 7.33 (t, 2H),
7.25−7.18 (m, 5H), 6.16 (t, 1H), 3.39 (m, 2H), 2.92 (t, 2H); 13C
NMR (CDCl3) δ 165.8, 138.6, 128.9, 128.7, 126.6, 125.8, 113.6, 53.2,
33.2; MS calcd for C14H15N2O3 (M + H) 259.1077, found 259.1073.
Anal. Calcd for C14H14N2O3·0.1H2O: C, 64.65; H, 5.50, N, 10.77.
Found: C, 64.88; H, 5.45; N, 10.42.
10842
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
Methanesulfonic acid (28 mg, 0.3 mmol) was added, and the solution
turned red. The pH was carefully controlled by adding methane-
sulfonic acid aliquots (0.3 mmol) to the reaction mixture every time it
turned orange, which indicated a higher pH. The reaction was
monitored by TLC (40% CH2Cl2/hexane). Four successive aliquots of
sodium cyanoborohydride (1.2 eq.; 224 mg) were added to the
reaction mixture, and the pH was adjusted after each addition. The
reaction mixture was then stirred overnight at room temperature. The
reaction was approximately 60% complete by TLC. The solution was
concentrated, dissolved in dichloromethane, and washed three times
with pH 7 water (note: exposure to an aqueous Na2CO3 wash
decomposed the compound). The organic layer was separated, dried
over anhydrous Na2SO4, filtered, and concentrated to give the crude
product 14f. The product mixture was purified by flash column
chromatography with 10% CH2Cl2/hexane to give 14f as an unstable
yellow oil (5 mg, 5% yield). Three bands eluted in the following order:
3,4-dinitrophenol (a decomposition product), unreduced oxime 13f,
HPLC Method for Rate Constant Determination. A gradient
elution using 0.1% TFA in acetonitrile and water was developed which
provided baseline separation between the starting materials, product
amide 16a, and the diphenylmethane standard. Using this method, the
respective starting materials (1 equiv of 1b and 14a) were dosed into
respective vials along with 10 equiv of pyruvic acid and heated in an
incubator-shaker at 65 °C in 50% t-BuOH/pH 7.4 buffer. Vials were
taken out at specific time points and cooled to 0 °C to stop the
reaction. Then a sample from each vial (10 μL) was injected into the
HPLC for analysis. The ratio of the area percent of the starting
material vs the internal standard was determined for each time point. A
plot of ln([starting material]/[std]) vs time (min) gave a line with
negative slope of −k for this pseudo-first-order condition. The
respective values of k are given in Table 3. Note: additional HPLC
experiments were performed to determine yields in the different
solvent systems and can be found in the Supporting Information.
Chemoselectivity Test with 14a. (a). Glycine Challenge. Into a
glass vial equipped with a phenolic screw-capped lid, glycine (11.6 mg,
155 μmol), N-phenethyl-O-(4-nitrophenyl)hydroxylamine 14a (10.6
mg, 41 μmol), and diphenylmethane (DPM; 5.0 mg, 29.7 μmol) were
dissolved in DMSO (0.65 mL) and freshly distilled pyruvic acid (10.9
μL, 155 μmol) was added immediately followed by the addition of
phosphate-buffered saline (PBS, 0.3 mL). The vial was then capped
and shaken, and the 68% DMSO/water solution was swirled overnight
in an orbital shaker at 65 °C. The transparent solution turned dark
green. Workup involved diluting the sample with CH2Cl2 (4 mL),
separating the organic layer, and washing it twice with saturated
aqueous Na2CO3. After base addition, the aqueous phase was bright
yellow due to the generation of sodium p-nitrophenolate. The organic
layer was separated, dried over anhydrous sodium sulfate, filtered, and
1
and then 14f. Compound 14f: H NMR (CDCl3) δ 7.87 (d, 1H, J =
8.7 Hz), 7.81 (d, 1H, J = 2 Hz), 6.97 (dd, 1H, J = 8.7 and 2 Hz), 6.37
(br s, 1H, NH), 3.16 (br t, 2H), 1.58 (m, 2H), 1.31 (m, 6H), 0.88 (t,
3H, J = 6.8 Hz); 13C NMR (CDCl3) δ 155.6, 140.9, 135.7, 126.8,
120.8, 116.7, 52.5, 31.7, 29.1, 27.0, 26.9, 22.6, 14.1. HRMS for M =
C13H19N3O5 (297.1325) was not observed, likely due to the high
temperature used for ionization. Instead, the HRMS provided a signal
consistent with facile N−O bond cleavage, loss of the 3,4-
dinitrophenoxy fragment (C6H3N2O5), and observation of a C7H16N
fragment. HRMS (M − C6H3N2O5): calcd for C7H16N 114.1283,
found 114.1282.
General Method for the Synthesis of Amides 16. Each of the
4-nitrophenylhydroxylamines 14a−d (25 mM solution in 75% t-
BuOH in a 7.4 pH buffer) was combined with an equivalent volume of
a pyruvic acid solution (50 mM in the same solvent system) to give a
final concentration of the hydroxylamine (12.5 mM) and pyruvic acid
(25 mM), respectively. Dibenzyl ether (DBE) was added as an internal
standard. The reactions were swirled in an incubator-shaker at 65 °C
overnight. The volatiles were removed under reduced pressure, and
dichloromethane was added to the residue. The organic layer was
washed with aqueous Na2CO3 and the organic layer separated, dried
over anhydrous Na2SO4, filtered, and concentrated. Yields were
determined via integration of the DBE standard versus a diagnostic
chemical shift for each acetamide derivative. Typically, either the
acetamide Me group near 2.0 ppm or the α proton on the carbon
attached to the amide nitrogen was used for quantification. NMR
spectra are provided for the relevant compounds in the Supporting
Information. The results are given in Table 2.
1
concentrated to give a colorless oil. H NMR (CDCl3) was used to
estimate the yield using the DPM singlet at 3.98 ppm and the
acetamide singlet at 1.93 ppm. The yield of acetamide 16a was 93%.
(b). Cysteine Challenge. Into a vial equipped with a phenolic screw-
capped lid, cysteine hydrochloride hydrate (27 mg, 154 μmol), N-
phenethyl-O-(4-nitrophenyl)hydroxylamine 14a (9.0 mg, 34.8 μmol),
and diphenylmethane (12.3 mg, 73.1 μmol) were dissolved in DMSO
(0.65 mL) and freshly distilled pyruvic acid (10.9 μL, 13.9 mg, 155
μmol) was added immediately followed by the addition of phosphate-
buffered saline (0.3 mL). The vial was then capped and shaken, and
the 68% DMSO/water solution was swirled overnight in an orbital
shaker at 65 °C. The heterogeneous mixture was dark brown. Workup
involved diluting the sample with CH2Cl2 (4 mL), separating the
organic layer, and washing it twice with saturated aqueous Na2CO3.
The aqueous phase was dark brown. The organic layer was separated,
dried over anhydrous sodium sulfate, filtered, and concentrated to give
N-Phenethylacetamide (16a): see percent yield given in Table 2;
1H NMR (CDCl3) δ 7.4−7.1 (m, 5H), 5.40 (s, 1H), 3.51 (q, 2H), 2.76
(t, 2H), 1.96 (s, 3H); 13C NMR (CDCl3) δ 170.2, 138.9, 128.7, 128.6,
126.5, 40.7, 35.6, 23.3.39
1
a dark yellow oil. H NMR (CDCl3) was used to estimate the yield
using the DPM singlet at 3.98 ppm and the acetamide singlet at 1.93
ppm. The yield of acetamide 16a was 36.4%.
40
The H NMR spectra of 16b and 16d40 matched their literature
1
(c). Cysteine and Hexanoic Acid Challenge. Into a vial equipped
with a phenolic screw-capped lid, cysteine hydrochloride hydrate (27.4
mg, 156 μmol), hexanoic acid (18 mg, 19.4 μL, 156 μmol), N-
phenethyl-O-(4-nitrophenyl)hydroxylamine 14a (20.9 mg, 80.9 μmol),
and diphenylmethane (6.0 mg, 35.6 μmol) were dissolved in DMSO
(1.3 mL) and freshly distilled pyruvic acid (13.6 mg, 10.9 μL, 155
μmol) was added immediately followed by the addition of phosphate-
buffered saline (0.6 mL). The vial was then capped and shaken, and
the 68% DMSO/water solution was swirled overnight in an orbital
shaker at 65 °C. Workup involved diluting the sample with CH2Cl2 (4
mL), separating the organic layer, and washing it twice with saturated
aqueous Na2CO3. The aqueous phase was dark brown. The organic
layer was separated, dried over anhydrous sodium sulfate, filtered, and
concentrated to give a dark yellow oil. 1H NMR (CDCl3) was used to
estimate the yield of 16a using the DPM singlet at 3.98 ppm and the
acetamide singlet at 1.93 ppm. The yield of 16a was 39%.
spectra. Compound 16c has been synthesized previously.41
N-Phenethyl-2-phenylacetamide (16e).5 The O-(4-nitrophenyl)-
N-phenethylhydroxylamine 14a (25 mg, 0.096 mmol) and phenyl-
pyruvic acid (15, where R2 = Bn; 31 mg, 0.19 mmol) were dissolved in
DMSO (1.8 mL). Diphenylmethane (DPM; 6.76 mg, 0.04 mmol) was
added into the vial as an internal standard followed by addition of H2O
(pH 7.4, 1.2 mL) to form a 60% DMSO/H2O mixture for the reaction.
The addition of H2O in the vial resulted in formation of a white
precipitate, which was dissolved by adding some more DMSO (0.8
mL). The reaction (now in a 68% DMSO/H2O solution) was then
swirled in an incubator shaker at 65 °C overnight. The reaction mass
was then dissolved in dichloromethane (50 mL) and extracted with
water (four times) to remove the DMSO present in the reaction mass.
The organic layer was separated, dried over Na2SO4, filtered and
concentrated. The product was confirmed by 1H NMR (CDCl3),
looking at the distinctive CH2 (s, δ 3.4 ppm) from the amide 16e. The
DPM standard gives a singlet at 3.9 ppm. The relative integrations
were used to calculate the yield of 16e: 22 mg (86% conversion, 100%
yield). The 1H NMR matched the literature spectrum obtained
previously for 16e.5
(d). Lysine and Phenol Challenge. Into a vial equipped with a
phenolic screw-capped lid, lysine hydrochloride (16.7 mg, 91 μmol),
phenol (8.6 mg, 91 μmol), N-phenethyl-O-(4-nitrophenyl)-
hydroxylamine 14a (11.8 mg, 45.7 μmol), and diphenylmethane
(12.4 mg, 73.7 μmol) were dissolved in DMSO (0.65 mL) and freshly
10843
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
distilled pyruvic acid (6.43 μL, 91 μmol) was added immediately
followed by the addition of water (0.3 mL). The vial was then capped
and shaken, and the 68% DMSO/water solution was swirled overnight
in an orbital shaker at 65 °C. The transparent solution turned yellow
with no precipitate noted. Workup involved diluting the sample with
CH2Cl2 (4 mL), separating the organic layer, and washing it twice with
saturated aqueous Na2CO3. The aqueous phase was bright yellow. The
organic layer was separated, dried over anhydrous sodium sulfate,
filtered, and concentrated to give a light yellow oil. 1H NMR (CDCl3)
was used to estimate the yield using the DPM singlet at 3.98 ppm
(2H) and the quartet at 3.44 ppm (2H). The yield of acetamide was
98.1%.
Control Experiments with Acetyl Chloride and Phenethyl-
amine. (a). In the Presence of Cysteine and Hexanoic Acid. In a
glass vial equipped with a phenolic screw cap, phenethylamine (17.6
mg, 0.145 mmol), cysteine hydrochloride hydrate (52 mg, 0.29 mmol),
hexanoic acid (34 mg, 36 μL, 0.29 mmol), and diphenylmethane (13.4
mg, 0.08 mmol) were combined in DMSO (2.6 mL). Deionized water
(1.2 mL) was then added to provide a translucent 68% DMSO/H2O
solution. Acetyl chloride (22.8 mg, 20.7 μL, 0.29 mmol) was added,
and the vial was capped and shaken vigorously for 10 s to provide a
transparent solution. The reaction vial was attached to an orbital
shaker equilibrated at 65 °C and swirled overnight. The solution was
cooled to room temperature and then diluted in dichloromethane (50
mL) and washed with water (four times) to remove the DMSO
present in the reaction mass. The organic layer was separated, dried
over anhydrous Na2SO4, filtered, and concentrated to give 33 mg of
the crude product as a colorless oil. Using the relative integrations
compared to the DPM standard (δ 3.93, s, 2H), the 1H NMR
(CDCl3) of the oil revealed a 66% recovery of hexanoic acid (δ 2.34, t,
2H) and no trace of 16a (0% yield). No unreacted phenethylamine
was detected in the organic layer. Investigation of the water phase
revealed the presence of phenethylamine salts.
(b). In the Presence of Lysine and Phenol. Into a vial equipped
with a phenolic screw-capped lid, lysine hydrochloride (16.6 mg, 0.091
mmol), phenethylamine (5.5 mg, 5.85 μL, 0.04568 mmol), phenol
(8.56 mg, 0.091 mmol), and diphenylmethane (12.7 mg, 0.073 mmol)
were dissolved in DMSO (0.65 mL) and acetyl chloride (7.14 mg, 6.49
μL, 0.091 mmol) was added immediately followed by the addition of
PBS (0.3 mL). The vial was then shaken and capped and the 68%
DMSO/PBS solution swirled overnight in an orbital shaker at 65 °C.
The transparent solution was then diluted with CH2Cl2 (4 mL) and
washed twice with a 10% saturated Na2CO3 solution. The organic
layer was separated, dried over anhydrous Na2SO4, filtered, and
concentrated to give a colorless oil. 1H NMR (CDCl3) revealed
unreacted phenethylamine (75% recovered) and a trace of the
acetamide product (1% yield).
ACKNOWLEDGMENTS
■
We thank the UCF College of Medicine and the Department of
Medical Education for their support of this work in the form of
startup funds for O.P.
REFERENCES
■
(1) Sletten, E. M.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2009, 48,
6974−6998.
(2) Bode, J. W.; Fox, R. M.; Baucom, K. D. Angew. Chem., Int. Ed.
2006, 45, 1248−1252.
(3) Carrillo, N.; Davalos, E. A.; Russak, J. A.; Bode, J. W. J. Am. Chem.
Soc. 2006, 128, 1452−1453.
(4) (a) Ju, L.; Lippert, A. R.; Bode, J. W. J. Am. Chem. Soc. 2008, 130,
4253−4255. (b) Pusterla, I.; Bode, J. W. Angew. Chem., Int. Ed. 2012,
51, 513−516. (c) Ogunkoya, A. O.; Pattabiraman, V. R.; Bode, J. W.
Angew. Chem., Int. Ed. 2012, 51, 9693−9697.
(5) Arora, J. S.; Kaur, N.; Phanstiel, O., IV J. Org. Chem. 2008, 73,
6182−6186.
(6) Sanki, A. K.; Talan, R. S.; Sucheck, S. J. J. Org. Chem. 2009, 74,
1886−1896.
(7) Phanstiel, O.; Wang, Q. X.; Powell, D. H.; Ospina, M. P.; Leeson,
B. A. J. Org. Chem. 1999, 64, 803−806.
(8) Wang, Q. X.; King, J.; Phanstiel, O., IV J. Org. Chem. 1997, 62,
8104−8108.
(9) Bergeron, R. J.; Phanstiel, O., IV J. Org. Chem. 1992, 57, 7140−
7143.
(10) Guo, H.; Naser, S. A.; Ghobrial, G.; Phanstiel, O., IV J. Med.
Chem. 2002, 45, 2056−2063.
(11) Gardner, R. A.; Ghobrial, G.; Naser, S. A.; Phanstiel, O., IV J.
Med. Chem. 2004, 47, 4933−4940.
(12) Wang, L.; Phanstiel, O. J. Org. Chem. 2000, 65, 1442−1447.
(13) Nemchik, A.; Badescu, V.; Phanstiel, O. Tetrahedron 2003, 59,
4315−4325.
(14) Braslau, R.; Axon, J. R.; Lee, B. Org. Lett. 2000, 2, 1399−1401.
(15) Skwarczynski, M.; Kiso, Y. Curr. Med. Chem. 2007, 14, 2813−
2823.
(16) Hansson, L. O.; Bergh, A.; Elkund, L. New Process for
Preparing Hydroxylamines and Medicaments WO/2010/116140,
World Patent, 2010.
(17) Miyazawa, E.; Sakamoto, T.; Kikugawa, Y. Org. Prep. Proc. Int.
1997, 29, 594−600.
(18) Endo, Y.; Shudo, K.; Okamoto, T. J. Am. Chem. Soc. 1982, 104,
6393−6397.
(19) First report of the synthesis of nitrophenoxyamine via
rearrangement of p-nitrophenyl-N-hydroxycarbamate in 60% yield:
Blakeley, R. L.; Zerner, B. Chem. Ind. 1973, 3, 133−134.
(20) Blakeley, R. C.; Hinds, J. A.; Kuntz, H. E.; Webb, E. C.; Zerner,
B. Biochemistry 1969, 8, 1991−2000.
ASSOCIATED CONTENT
■
S
* Supporting Information
(21) (a) Castellino, A. J.; Rapoport, H. J. Org. Chem. 1984, 49, 1348−
1352. (b) Bumgardner, C. L.; Lilly, R. L. Chem. Ind. 1962, 559−560.
(22) Maimone, T. J.; Buchwald, S. L. J. Am. Chem. Soc. 2010, 132,
9990−9991.
1
Figures, tables, and text giving H NMR spectra for 8a−c,
oxime intermediates 13a−d,g−i, compounds 14a−f, and a
crude mixture containing 16e, 13C NMR spectra for
compounds 8a−c, 13h, 14a−f, 16a, details of the kinetic
studies of 1b and 14a with 10 equiv of pyruvic acid, details of
the solvent studies of 1b and 14a in t-BuOH/pH 7.4 buffer
mixtures and of 14a in DMSO/pH 7.4 buffer mixtures. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(23) Sheradsky, T.; Salemnick, G.; Nir, Z. Tetrahedron 1972, 28,
3833−3843.
(24) Ilvespaa, A. O.; Marxer, A. Helv. Chim. Acta 1963, 46, 2009−
̈
̈
2020.
(25) Liu, Y.; Qian, J.; Lou, S.; Xu, Z. J. Org. Chem. 2010, 75, 6300−
6303.
(26) Donarski, W. J.; Dumas, D. P.; Heitmeyer, D. P.; Lewis, V. E.;
Raushel, F. M. Biochemistry 1989, 28, 4650−4655.
(27) Sheradsky, T.; Nov, E. J. Chem. Soc., Perkin Trans. 1 1980,
2781−2786.
AUTHOR INFORMATION
■
Corresponding Author
*Tel: 407-823-6545. E-mail: Otto.Phanstiel@ucf.edu.
(28) Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am. Chem. Soc.
1971, 93, 2897−2904.
Notes
(29) Bernhart, C.; Wermuth, C.-G. Tetrahedron Lett. 1974, 2493−
2496.
The authors declare no competing financial interest.
10844
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845

The Journal of Organic Chemistry
Article
(30) Example using NaBH3CN with methyl orange and inorganic 6
M HCl in MeOH: House, H. O.; Lee, L. F. J. Org. Chem. 1976, 41,
863−869.
(31) Fleming, I. Compr. Org. Synth. 1992, 8, 60−77.
(32) Abdel-Majid, A. F.; Carson, K. G.; Harris, B. D.; Maryanoff, C.
A.; Shah, R. D. J. Org. Chem. 1996, 61, 3849−3862.
(33) Kawase, M.; Kikugawa, Y. J. Chem. Soc., Perkin Trans. 1 1979,
643−645.
(34) Lane, C. F. Synthesis 1975, 135−146.
(35) (a) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry,
4th ed.; Kluwer Academic/Plenum Publishers: New York, 2000; Part
A; Structure and Mechanisms, pp 237−241. (b) Nelson, D. L.; Cox,
M. M. Lehninger Principles of Biochemistry, 4th ed.; W. H. Freeman:
New York: 2005; pp 604−606.
(36) (a) Flores, M. A.; Bode, J. W. Org. Lett. 2010, 12, 1924−1927.
(b) Key, J. A.; Li, C.; Cairo, C. W. Bioconjugate Chem. 2012, 23, 363−
371.
(37) Thermal analysis allowed the substituted O-(nitrophenyl)
hydroxyamines to be used safely as aminating agents: Boyles, D. C.;
Curran, T. T.; Parlett, R. V., IV Org. Process Res. Dev. 2002, 6, 230−
233.
(38) Ishida, T.; Hashimoto, I.; Tsuru, K., Production of O-Aryl
Hydroxylamine. Jpn. Kokai Tokkyo Koho (1985) JP 60-169447,
February 09, 1985.
́
(39) Llinares, J.; Elguero, J.; Faure, R.; Vincent, E. J. Org. Magn.
Reson. 1980, 14, 20−24.
(40) Avalos, M.; Babiano, R.; Barneto, J. L.; Bravo, J. L.; Cintas, P.;
́
Jimenez, J. L.; Palacios, J. C. J. Org. Chem. 2001, 66, 7275−7282.
(41) Rosen, T.; Lico, I. M.; Chu, D. T. W. J. Org. Chem. 1988, 53,
1580−1582.
10845
dx.doi.org/10.1021/jo302175g | J. Org. Chem. 2012, 77, 10835−10845