Mechanisms of nucleophilic substitution reactions of methylated hydroxylamines with bis(2,4-dinitrophenyl)phosphate. Mass spectrometric identification of key intermediates

Article abstract of DOI:10.1021/jo0494735

Mono- and dimethylation of hydroxylamine on nitrogen does not significantly affect rates of initial attack of NHMeOH and NMe₂OH on bis(2,4-dinitrophenyl)phosphate (BDNPP), which is largely by oxygen phosphorylation. O-Methylation, however, blocks this reaction and NH ₂OMe then slowly reacts with BDNPP via N-attack at phosphorus and at the aryl group. With NHMeOH, the initial product of O-attack at phosphorus reacts further, either by reaction with a second NHMeOH or by a spontaneous shift of NHMe to the aryl group via a transient cyclic intermediate. There is a minor N-attack of NHMeOH on BDNPP in an S_N2(Ar) reaction. Reactions occurring via N-attack are blocked by N-dimethylation, and reaction of NMe ₂OH with BDNPP occurs via O-attack, generating a long-lived product. Reaction mechanisms have been probed, and intermediates identified, by using both NMR and MS spectroscopy, with the novel interception of key reaction intermediates in the course of reaction by electrospray ionization mass and tandem mass spectrometry.

Full text of DOI:10.1021/jo0494735

Mech a n ism s of Nu cleop h ilic Su bstitu tion Rea ction s of Meth yla ted
Hyd r oxyla m in es w ith Bis(2,4-d in itr op h en yl)p h osp h a te. Ma ss
Sp ectr om etr ic Id en tifica tion of Key In ter m ed ia tes
†
†
†
‡
J osiel B. Domingos, Elisane Longhinotti, Tiago A. S. Brand a˜ o, Clifford A. Bunton,
§
§
,†
Leonardo S. Santos, Marcos N. Eberlin, and Faruk Nome*
Chemistry Department, Federal University of Santa Catarina, Florian o´ polis, SC, 88040-900, Brazil,
Department of Chemistry and Biochemistry, University of California,
Santa Barbara, California 93106-9510, and Thomson Mass Spectrometry Laboratory,
Institute of Chemistry, State University of Campinas, 13083-970, Brazil
faruk@qmc.ufsc.br
Received March 31, 2004
Mono- and dimethylation of hydroxylamine on nitrogen does not significantly affect rates of initial
attack of NHMeOH and NMe
by oxygen phosphorylation. O-Methylation, however, blocks this reaction and NH
2
OH on bis(2,4-dinitrophenyl)phosphate (BDNPP), which is largely
OMe then slowly
2
reacts with BDNPP via N-attack at phosphorus and at the aryl group. With NHMeOH, the initial
product of O-attack at phosphorus reacts further, either by reaction with a second NHMeOH or by
a spontaneous shift of NHMe to the aryl group via a transient cyclic intermediate. There is a minor
N-attack of NHMeOH on BDNPP in an S
blocked by N-dimethylation, and reaction of NMe
N
2(Ar) reaction. Reactions occurring via N-attack are
OH with BDNPP occurs via O-attack, generating
2
a long-lived product. Reaction mechanisms have been probed, and intermediates identified, by using
both NMR and MS spectroscopy, with the novel interception of key reaction intermediates in the
course of reaction by electrospray ionization mass and tandem mass spectrometry.
1
In tr od u ction
strongly inhibited reaction. However, reactions with Me
2
-
NOH did not liberate more than 1 equiv of DNP and that
with NH OMe liberated less than 1 equiv of DNP, but
As an extension of earlier experiments,1 we have
compared nucleophilic substitutions of bis(2,4-dinitro-
2
more than 1.5 equiv were apparently liberated in reaction
phenyl)phosphate (BDNPP) by NH
O-methyl derivatives. The initial reaction of monoanionic
BDNPP, an activated bisaryl phosphate, with 0.1 M NH
2
OH and its N- and
1
with MeNHOH. Other minor products were not exam-
ined, but in the present work we discuss products and
kinetics in detail, particularly as regards the extent to
which methylation on nitrogen affects both the initial
O-attack on phosphorus and subsequent reactions of the
phosphorylated species, and confirm that O-methylation
blocks reactions on oxygen. Hydroxylamine and its N-
methyl derivatives are R-effect nucleophiles that react
much more readily than expected from relationships
2
-
OH at pH 9 releases ca. 1.7 equiv of 2,4-dinitrophenoxide
3
ion, DNP, and is approximately 10 fold faster than the
spontaneous hydrolysis. We showed that the initial
reaction with nonionic NH OH generates a short-lived
2
O-phosphorylated hydroxylamine, detected by NMR spec-
troscopy, and is followed by three reactions: (i) attack of
hydroxylamine on this intermediate, generating 2,4-
dinitrophenyl phosphate dianion (DNPP), which subse-
quently forms 2,4-dinitrophenoxide anion, DNP; (ii) in-
tramolecular displacement of the second DNP anion and
rapid decomposition of the cyclic intermediate to form
phosphorylhydroxylamine and eventually inorganic phos-
phate; (iii) a novel rearrangement with intramolecular
aromatic nucleophilic substitution involving a cyclic
intermediate and O to N migration of the aromatic group.
We found no evidence for initial N-phosphorylation or
aromatic nucleophilic substitution. With NHMeOH and
2
between nucleophilicity and Br o¨ nsted basicity. We also
used electrospray ionization mass and tandem mass
spectrometry to monitor the reaction by “fishing” ionic
intermediates and products directly from solution into
the gas phase and to probe the mechanism in solution
via the novel interception and structural characterization
of key intermediates. This technique complements NMR
spectroscopy in the identification of intermediates of
reactions in solution.
Resu lts a n d Discu ssion
2
NMe OH, we saw only small rate effects of methylation
on nitrogen, whereas methylation on oxygen, as expected,
Kin etics a n d P r od u cts. Reactions of BDNPP with
0
2
.1 M hydroxylamine derivatives (NR OR′, R, R′) H, Me)
†
Federal University of Santa Catarina.
University of California, Santa Barbara.
S a˜ o Paulo State University of Campinas.
were followed by the increasing absorbance at 400 nm
‡
§
(
1) Domingos, J . B.; Longhinotti, E.; Bunton, C. A.; Nome, F. J . Org.
(2) J encks, W. P.; Carriuolo, J . J . Am. Chem. Soc. 1960, 82, 1778-
1786.
Chem. 2003, 68, 7051-7058.
1
0.1021/jo0494735 CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/03/2004
6
024
J . Org. Chem. 2004, 69, 6024-6033

Substitution Reactions of Methylated Hydroxylamines
TABLE 1. Over a ll Ra te a n d Dissocia tion Con sta n ts for
F ittin g th e p H-Ra te P r ofiles in F igu r e 1a
NHMeOH
NMe2OH
NH2OMe
SN(P)
HNMeOH
, M-1 s-1
-3
-3
-5
k
6.59 × 10
3.00 × 10
2.64 × 10
yield, mol %
74
100
40
SNAr
, M-1 s-1
-3
-5
k
2.31 × 10
3.95 × 10
HNMeOH
yield, mol %
26
60
kNR O-, M s-1
-1
9.62
6.18
2.38
5.20
2
pKa1
4.62
a
Dissociation constants are from the NIST Standard Reference
Database 46, Version 6.0, NIST Standard Reference Data, Gaith-
-
7
-1
ersburg, MD 20899, and k0 and kOH ) 1.90 × 10
s
and 2.92 ×
-
3
-1 -1
1
0
M
s , respectively.
F IGURE 1. Plots of log kobs versus pH for nucleophilic attack
2 2
on BDNPP by MeNHOH (O), (Me) NOH (2), NH OMe (3), and
H
2
O ([), at 25 °C.
SCHEME 1
in the pH range from 4 to 12. We had shown that reaction
of BDNPP with NH OH, which releases ca. 1.7 equiv of
2
dinitrophenoxide ion, DNP, is biphasic in some condi-
tions. Some methyl hydroxylamine reactions show similar
kinetic behavior, but with different yields of DNP. The
pH-rate profiles for the initial formation of DNP are
shown in Figure 1 and were fitted by using eq 1 or its
F IGURE 2. Plots of k
1
(9) and k
2
(O) as a function of
[NHMeOH] for nucleophilic attack on BDNPP, at pH 9 and
2
0
5 °C. Insert: kinetic plots for [NHMeOH] ) (a) 0.1 and (b)
.6 M.
probably related to the increased basicity2,3 and is
understandable in terms of the evidence of Hengge et al.
modified form for NH
2
OMe, where there is no reaction
of a hydroxyl group or alkoxide ion.
regarding the role of the NH
NH
OH with 4-nitrophenyl acetate.4 Kinetics of the
reaction of NH OH with BDNPP approximate to first
2
group in the reaction of
2
k
[NR OR′]
2
NR OR′
-
2
2
kobs ) k + k [OH ] +
+
o
OH
+
1
[
H ]
K
a2
order, but the reaction is biphasic. The reaction kinetics
with MeNHOH are typical of consecutive reactions at all
pH and [MeNHOH] examined (curves a and b in Figure
2). Values of the observed rate constants (kobs) plotted in
Figure 1 give k , for the initial attack in a biphasic
1
reaction and release of DNP, which is obtained by fitting
its formation to eq 2. Figure 2 also shows the slower
1
+
+
+
Ka1
[H ]
k
-
[NR OR′]
NR O
2
2
(1)
+
[
H ]
1
+
Ka2
2
reaction, step k , which includes hydrolysis of the initially
formed monoester 2,4-dinitrophenyl phosphate (DNPP),
o
In eq 1, R and R’ can be H or Me, k is the rate constant
possibly with catalysis by NHMeOH.5
for the spontaneous reaction, kOH is that for reaction with
OH , kNR OR′, kNR O-, Ka1, and Ka2 are rate and equilibrium
-
2
2
k1
-
k1t
(e-k1t
-
constants for the nonionic (A) and deprotonated (B)
species of NR OH, respectively (Scheme 1); values for
[DNP] ) [BDNPP]0 1 - e
-
{
k2 - k
1
2
initial attack on phosphorus and the aromatic group are
in Table 1. The spontaneous hydrolysis of BDNPP and
e- ) + [DNPP] (e-k2t) + [DNP]0 (2)
k2t
0
}
-
reaction with OH generate monoester and dinitrophen-
In eq 2 k
1
and k
2
are overall first-order constants with
respect to BDNPP and DNPP, respectively. The initial
oxide ion. These reactions are relatively unimportant at
pH 9-10, where products were examined, and their
contribution to product formation can be neglected.
(
3) Bissot, T. C.; Parry, R. W.; Campbell, D. H. J . Am. Chem. Soc.
1957, 79, 796-800.
4) Hess, R. A.; Hengge, A. C.; Cleland, W. W. J . Am. Chem. Soc.
997, 119, 6980-6983.
5) Kirby, A. J ., J encks, W. P. J . Am. Chem. Soc. 1965, 87, 3209-
3216.
Rea ction w ith MeNHOH. The overall initial reaction
(
2
with MeNHOH is ca. 3-fold faster than that with NH -
1
OH,1 in the pH-independent region (Figure 1), and
releases ca. 1.6 equiv of DNP. This rate difference is
(
J . Org. Chem, Vol. 69, No. 18, 2004 6025

Domingos et al.
F IGURE 3. 1H NMR spectra of the reaction mixture of 0.01 M BDNPP with 0.1 M NHMeOH, in D
for identification of signals.
O, pD 9, and 25 °C. See Table
2
2
SCHEME 2
reaction of BDNPP with NH
2
OH gives both stable
by attack of the OH group at phosphorus. Signals and
their assignments are shown in Table 2 and Figures 3
and 4. Concentrations shown in Table 2 were estimated
products and intermediates, some of which are suf-
ficiently long-lived to be identified, generally by NMR
1
1
spectroscopy. On the basis of this earlier evidence we
after H signals of BDNPP, 1, were no longer detected.
postulate that a series of mechanistically reasonable
reactions of MeNHOH are involved (Scheme 2), and we
show that this Scheme is consistent with all of the
evidence on kinetics and products. In Scheme 2, the state
of ionization of species shown is that at pH 9-10, with
Figure 5 shows relative concentrations of species that
form or decompose in the course of the reaction. Concen-
trations are based on the H NMR signals in the aromatic
1
region, and some minor products that give signals
overlapping those of the major products are not shown.
Nonetheless, Figure 5 shows that formation of 4, which
does not contains phosphorus (Table 2), is related to
disappearance of intermediate 2 and also to the initial
reaction of BDNPP, 1 (Scheme 2).
Product 4, N-methyl-N-(2,4-dinitrophenyl) hydroxyl-
amine, is formed by two distinct reactions, viz., aromatic
nucleophilic substitution at C-1 of both BDNPP (in an
initial reaction) and 2 by the amino group of MeNHOH
in a subsequent reaction, and for purpose of identification
1
rapid acid-base equilibration. Figures 3 and 4 show H
and 31P NMR spectra taken at intervals, and the long-
lived intermediates and products are characterized by
this NMR spectral data. The NMR spectra detected four
species in the course of the reaction, which were identi-
fied chemically as specified, viz., DNP, the transient
species, 2, the monoester DNPP, 3, and a long-lived
product, 5. Intermediate 2, which slowly decomposes, is
formed in the initial reaction of MeNHOH with BDNPP
6
026 J . Org. Chem., Vol. 69, No. 18, 2004

Substitution Reactions of Methylated Hydroxylamines
F IGURE 4. 31P NMR spectra of the reaction mixture of 0.01 M BDNPP with 0.1 M NHMeOH, in D
O, pD 9, and 25 °C. See Table
2
2
for identification of signals.
TABLE 2. NMR Da ta of Rea ction P r od u cts of 0.01 M BDNP P w ith 0.1 M NHMeOH, in D2O, p D 9, 25 °C
compd
DNP
1H NMR δ (ppm)
31P NMR δ (ppm)/(yield, mol %)a
δ 6.74 (d, 1H, J ) 9.6 Hz, Ar), 8.11 (dd, 1H, J ab)9.6 Hz and J bx)3.0 Hz, Ar),
8
.90 (d, 1H, J ) 3.0 Hz, Ar)
1
2
3
4
5
δ 7.80 (d, 1H, J ) 9.1 Hz, Ar), 8.58 (dd, 1H, J ab)9.1 Hz and J bx)2.9 Hz, Ar),
.94 (d, 1H, J ) 2.9 Hz, Ar)
δ 7.75 (d, 1H, J ) 9.2 Hz, Ar), 8.55 (dd, 1H, J ab)9.2 Hz and J bx)2.7 Hz, Ar),
.62 (d, 1H, J ) 2.7 Hz, Ar)
δ 7.86 (d, 1H, J ) 9.4 Hz, Ar), 8.48 (dd, 1H, J ab)9.4 Hz and J bx)2.9 Hz, Ar),
.84 (d, 1H, J ) 2.9 Hz, Ar)
δ 7.22 (d, 1H, J ) 9.6 Hz, Ar), 8.24 (dd, 1H, J ab)9.6 Hz and J bx)2.6 Hz, Ar),
.62 (d, 1H, J ) 2.6 Hz, Ar)
δ 7.86 (d, 1H, J ) 9.4 Hz, Ar), 8.48 (dd, 1H, J ab)9.4 Hz and J bx)2.6 Hz, Ar),
.86 (d, 1H, J ) 2.6 Hz, Ar)
-13.17 (0%)
-3.21 (36%)
0.04 (17%)
8
8
8
8
7.56 (13%)
8
6
Pi
5.43 (25%)
1.86 (9%)
a
Relative areas of signals of identified compounds from 31P spectra.
it was also prepared by reacting 2,4-dinitrochlorobenzene
becoming major techniques for mechanistic studies of
solution reactions involving ionic reactants, intermedi-
ates, or products.7 On the basis of our background in
with NHMeOH. The initial S
N
2(Ar) reaction of MeNHOH
,8
with BDNPP also generates DNPP, 3 (Scheme 2). Detec-
2
-
8
tion of the phosphorylated product, MeNHOPO
5
3
, 6, at
reaction mechanisms by using these techniques, we used
3
1
.43 ppm in the P NMR spectrum, shows that there is
both ESI-MS and ESI-MS/MS to probe the mechanism
of the reactions shown in Scheme 2. We used ESI to “fish”
the anionic intermediates and products directly out of
solution into the gas phase, to monitor intermediates and
products as a function of time and to characterize
structures by CID via MS/MS. The spectrometer was
a second aromatic nucleophilic substitution product, 4,
derived from intermediate 2, as well as in the initial
reaction of BDNPP. These results fit the reactions in
1
Scheme 2 where, unlike for reaction with NH
2
OH, there
are independent competing nucleophilic substitutions on
BDNPP, with attack on phosphorus and carbon.
Mass spectrometry with electrospray ionization (ESI)6
allows transfer of ions directly from solution into the gas
phase and is characterized by the gentleness of formation
of gaseous ions, and ESI-MS and ESI-MS/MS are rapidly
(6) (a) Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J .
B. Anal. Chem. 1985, 57, 675-679. (b) Fenn, J . B.; Mann, M.; Meng,
C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (c)
Cole, R. B. Electrospray Ionization Mass Spectroscopy; Wiley: New
York, 1997.
J . Org. Chem, Vol. 69, No. 18, 2004 6027

Domingos et al.
respectively), m/z 263 (3), and m/z 212 (deprotonated 4).
The same ions were observed after 30 min. These ions
show dissociation behaviors that fully match proposed
structures of intermediates, as indicated by structural
assignments of fragment ions displayed in each respec-
tive mass spectrum, and are consistent with the results
from NMR spectroscopy. In Figure 7 ions shown as a solid
circle are decomposed in a CID to products, shown as an
open circle. As expected, the fates of the two ions with
m/z of 292 depend on the time of reaction in solution
(Figure 7b,c), but CID products of the other ions do not
change with the time of reaction in solution.
Ionic intermediates 2 and 5 (Scheme 2) are isomers,
and their relative amounts and therefore the ESI-MS/
MS spectrum of the ions of m/z 292 change drastically
with time. Samples taken after 10 min of reaction in
solution (Figure 7b) show that these ions dissociate in
the mass spectrometer by CID mainly into four fragment
ions of m/z 263, 183, 97, and 79. The fragment ions, of
m/z 263 and 183, are formed in the gas phase from 2
F IGURE 5. Apparent concentrations of main products of the
reactions of 0.01 M BDNPP with 0.1 M NHMeOH, in D O, pD
, and 25 °C: (9) 1, (4) DNP , (1) 2, and (O) 4.
2
9
(Scheme 3), whereas dissociation to ions of m/z 97 and
operated in the negative-ion mode so that only anions
were monitored, which simplifies treatment of reactions
of BDNPP. Within the range of m/z examined only
monoanions are observed (Figure 6).
Figure 6 shows the ESI-MS spectrum after 10 min of
reaction of nonionic NHMeOH with BDNPP in aqueous
methanol (50% v/v) at pH 10 and 25 °C. Major ions clearly
detected in the mass spectrum correspond to the reactant
anion 1 (BDNPP), m/z 429, intermediates and products
79 is evidence for the presence of 5 (Scheme 4). Therefore,
both 2 and 5 are present in the reaction mixture at an
early stage of the reaction. After 30 or 60 min of reaction
(Figure 7c), however, the ion of m/z 292 generates only
the two fragment ions of m/z 97 and 79. This temporal
change in product ion distribution from the CID of ion of
m/z 292 confirms that in a later stage of reaction
intermediate 5, rather than 2, become a dominant
species, as shown in Scheme 2, where rearrangement of
2, in solution, generates 5. Although electrospray is a
gentle method we did not detect 6, m/z of 125, because it
probably rapidly generates metaphosphate ion.
2
of m/z 292, DNP of m/z 183, 3 of m/z 263, and the final
product 5 of m/z 292. The overall appearance of the
spectrum and the relative intensities of these ions were
shown, by continuous ESI-MS monitoring, to change
modestly from 1 to 60 min of reaction in solution. Each
of these anions was then mass-selected and structurally
characterized via collision-induced dissociation (CID)
with nitrogen in ESI-MS/MS measurements. Figure 7a-e
show the ESI-MS/MS spectra for ions of m/z 429 (1), 292
Based on the ESI-MS(/MS) results, NMR and kinetic
data, and consistent with results on the reaction of NH -
2
OH with BDNPP, we conclude that there is both aromatic
substitution, generating DNPP, 3, and 4, and initial
phosphorylation of the OH group of NHMeOH by BDNPP
(Scheme 2), which generates 0.7 equiv of DNP, forming
intermediate 2, which breaks down slowly by two distinct
(2 and/or 5, after 10 and 60 min of reaction, b and c,
F IGURE 6. ESI-MS spectrum of the reaction mixture of 0.01 M BDNPP with 0.1 M NHMeOH, in aqueous methanol (50% v/v)
at pH 10.
6
028 J . Org. Chem., Vol. 69, No. 18, 2004

Substitution Reactions of Methylated Hydroxylamines
1
overall reaction of BDNPP with NH
2
OH, except that
there was no aromatic nucleophilic substitution on BD-
NPP, and with NHMeOH spontaneous conversion of the
phosphorylated hydroxylamine, analogous to 2, into DNP
is not observed and is blocked by N-methylation. We note
that NMR and MS spectroscopy are complementary
methods leading to similar conclusions, despite differ-
ences in solvents, which were D
MeOH (MS).
2
O (NMR) and aqueous
Rea ction w ith NMe
2
OH. The reaction of BDNPP
with NMe OH shows the usual first-order kinetic profile
2
at all the pH examined and releases only 1 equiv of DNP.
The first-order rate constant in the pH-independent
region is lower than that with NH
Figure 1). Dimethylation on nitrogen simplifies the
reaction in that only 1 equiv of DNP is liberated by
nucleophilic O-attack of Me NOH at phosphorus, as
shown by the 1H and P NMR spectra after approxi-
mately 3 half-lives (Table 3). Therefore, with 0.1 M NMe
2
OH by about 20%
(
2
31
2
-
OH, the only products of reaction with BDNPP are the
dinitrophenoxide ion, DNP, and the long-lived phospho-
rylated derivative, 7 (Scheme 6). Dimethylation on
nitrogen blocks reactions of NMe
the first-formed intermediates in reactions with NH
2
OH similar to those of
1
2
OH
and NHMeOH and also aromatic nucleophilic substitu-
tion by an amino group.
Rea ction w ith NH
2
OMe. The reaction of BDNPP
with NH OMe is much slower than those with the
2
N-methylated hydroxylamines. Values of kobs are from
initial rates, because subsequent hydrolysis of monoester
DNPP, 3, perturbs the first-order kinetics, and reaction
gives less than 1 equiv of DNP (Figure 1 and Table 1).
This consequence of blocking the OH group confirms that
attack of the other hydroxylamine derivatives on BDNPP
is preferably by oxygen rather than nitrogen. The sub-
sequent decomposition of the initial products makes
F IGURE 7. ESI-MS/MS spectra after CID of anionic inter-
mediates observed in the reaction of 0.01 M BDNPP with 0.1
M MeNHOH in aqueous methanol (50% v/v) at pH 10 for ions
of m/z 429, 292, 263, and 212.
2
reaction with NH OMe less simple than expected for a
nucleophile with only one reactive center. The half-life
in the pH-independent region (Figure 1) is ca. 13.7 h, and
reaction is not much faster than the spontaneous hy-
9
drolysis and is similar to that for reaction of NH
2
OMe
pathways, (a) aromatic nucleophilic substitution de-
scribed above, giving 4 and 6, or (b) spontaneous rear-
rangement where the terminal NHMe group attacks the
dinitrophenyl moeity to form a transient cyclic Meisen-
with the monoester, DNPP (t1/2 ) 15.5 h). Because these
reactions are slow, other side reactions may be involved
1
31
but were not identified. We detected by H and P NMR
spectroscopy product 9 (Table 4), from aromatic nucleo-
philic substitution on the substrate (Scheme 7), which is
heimer complex, as for reaction with NH
2
OH.1 This
complex rapidly ring opens, giving 5 as a long-lived
product (Scheme 5). This Meisenheimer complex must
be short-lived, because we saw no buildup of color or
NMR signals typical of it in the course of reaction. The
methyl group on nitrogen sterically hinders this intramo-
lecular attack, and only a small amount of 5 is therefore
formed. As noted, aromatic substitution on 1 also occurs,
generating 4 and the monoester DNPP, 3, although it is
the minor reaction (Figure 5), and as noted, the spectral
data for 4 are identical to those of an authentic sample
(
7) (a) Koch, K. J .; Gozzo, F. C.; Nanita, S. C.; Takats, Z.; Eberlin,
M. N.; Cooks, R. G. Angew. Chem. 2002, 114, 1797-1800; Angew.
Chem., Int. Ed. 2002, 41, 1721-1724. (b) Meurer, E. C.; Sabino, A. A.;
Eberlin, M. N. Anal. Chem. 2003, 75, 4701-4709. (c) Sabino, A. A.;
Machado, A. H. L.; Correia, C. R. D.; Eberlin, M. N. Angew. Chem.,
Int. Ed. 2004, 43, 2514-2518. (d) Hilderling, C.; Adlhart, C.; Chen, P.
Angew. Chem. 1998, 110, 2831-2835; Angew. Chem., Int. Ed. 1998,
3
7, 2685-2689. (e) Chen, P. Angew. Chem. 2003, 115, 2938-2954;
Angew. Chem., Int. Ed. 2003, 42, 2832-2847. (f) Meyer, S.; Metzger,
J . O. Anal. Bional. Chem. 2003, 377, 1108. (g) Meyer, S.; Koch, R.;
Metzger, J . O. Angew. Chem., Int. Ed. 2003, 42, 4700-4703. (h) Santos,
L. S.; Pavam, C. H.; Almeida, W. P.; Coelho, F.; Eberlin, M. N. Angew.
Chem., Int. Ed. 2004, in press.
(
Table 2). The contribution of kHNMeOH in eq 1 for the
initial reaction of BDNPP can be separated into two rate
(
8) (a) Meurer, E. C.; Santos, L. S.; Pilli, R. A.; Eberlin, M. N. Org.
3
1
constants according to initial product yields ( P NMR),
k
Lett. 2003, 5, 1391-1394. (b) D’Oca, M. G. M.; Moraes, L. A. B.; Pilli,
R. A.; Eberlin M. N. J . Org. Chem. 2001, 66, 3854-3864. (c) Tomazela,
D. M.; Moraes, L. A. B.; Pilli, R. A.; D’Oca, M. G. M.; Eberlin, M. N. J .
Org. Chem. 2002, 67, 4652-4658. (d) Carvalho, M.; Gozzo, F. C.;
Mendes, M. A.; Sparrapan, R.; Kascheres, C.; Eberlin, M. N. Chem.
Eur. J . 1998, 4, 1161-1168. (e) Meurer, E. C.; Sabino, A. A.; Eberlin,
M. N. Anal. Chem. 2003, 75, 4701-4709.
SNAr
HNMeOH
SN(P)
HNMeOH
(carbon attack) and k
(phosphorus at-
-
3
-3
-1 -1
tack), which are 2.31 × 10 and 6.59 × 10
M
s ,
respectively (Table 1).
Some of these reactions in solution are similar to those
identified kinetically and by NMR spectroscopy in the
(9) Bunton, C. A.; Farber, S. J . J . Org. Chem. 1969, 34, 767-772.
J . Org. Chem, Vol. 69, No. 18, 2004 6029

Domingos et al.
SCHEME 3
SCHEME 4
SCHEME 5
TABLE 3. NMR Sp ectr a of Rea ction P r od u cts of 0.01 M BDNP P w ith 0.1 M NMe2OH, in D2O, p D 9, 25 °C
compd
DNP
1H NMR δ (ppm)
31P NMR δ (ppm)/(yield, mol %)a
δ 6.74 (d, 1H, J ) 9.6 Hz, Ar), 8.11 (dd, 1H, J ab)9.6 Hz and J bx)3.0 Hz, Ar),
8
.90 (d, 1H, J ) 3.0 Hz, Ar)
1
7
δ 7.80 (d, 1H, J ) 9.1 Hz, Ar), 8.58 (dd, 1H, J ab)9.1 Hz and J bx)2.9 Hz, Ar),
.94 (d, 1H, J ) 2.9 Hz, Ar)
δ 7.75 (d, 1H, J ) 9.2 Hz, Ar), 8.55 (dd, 1H, J ab)9.2 Hz and J bx)2.7 Hz, Ar),
.62 (d, 1H, J ) 2.7 Hz, Ar)
-13.17 (12%)
-3.21 (88%)
8
8
a
Relative areas of signals of identified compounds.
SCHEME 6
1
much slower than the corresponding reaction of NHMeOH.
Monoester DNPP, 3, is the other product of this S Ar
the first-formed intermediates. The strong inhibition of
N
the initial reactions by O-methylation confirms the
original assumption that attack on phosphorus is pref-
erentially by the OH group, although there is a slow
reaction with BDNPP but was only detected in low
concentration because its hydrolysis, probably with amine
catalysis, is not much slower than the initial attack on
BDNPP. The other major product of this overall reaction
is formed by attack of amino nitrogen at the phosphorus
atom of BDNPP, giving the long-lived phosphorylated
product, 8, and DNP. This is the only example of nitrogen
phosphorylation in solution reactions of the hydroxyl-
amines with BDNPP. The NMR spectra were taken after
approximately 48 h. Reactions shown in Scheme 7 are
slow, due in part to the low basicity of this hydroxyl-
amine.
1
reaction with NH
derivative 8 and DNP (Scheme 7). Deprotonation of
NHMeOH and NMe OH at high pH generates oxide
anions, which like NH
2
OMe forming the N-phosphorylated
2
-
2
O should be very effective de-
phosphorylating agents,1 and we assume that the in-
creases in observed first-order rate constants with in-
creasing pH involve this intervention. Values of the first
0
pK
a
are known for the methylated hydroxylamines (Table
1
1
1), and the second pK
a
2
of NH OH is 13.74. On the
Effects of Meth yla tion of NH
2
OH. To a first ap-
(10) Simanenko, Y. S.; Popov, A. F.; Prokop’eva, T. M.; Savyolova,
V. A.; Belousova, I. A., Zubareva, T. M. Mendeleev Commun. 1994,
proximation mono- and dimethylation of hydroxylamine
on nitrogen do not have major kinetic effects on the initial
O-attack on BDNPP, but they greatly affect reactions of
2
10-212.
11) NIST Standard Reference Database 46, Version 6.0; distributed
by NIST Standard Reference Data, Gaithersburg, MD 20899.
(
6
030 J . Org. Chem., Vol. 69, No. 18, 2004

Substitution Reactions of Methylated Hydroxylamines
TABLE 4. NMR Sp ectr a l Da ta for Rea ction P r od u cts of 0.01 M BDNP P w ith 0.1 M NH2OMe, in D2O a t p D 9, 25 °C
compd
DNP
1H NMR δ (ppm)
31P NMR δ (ppm) (yield, mol %)a
δ 6.74 (d, 1H, J ) 9.6 Hz, Ar), 8.11 (dd, 1H, J ab)9.6 Hz and J bx)3.0 Hz, Ar),
8
.90 (d, 1H, J ) 3.0 Hz, Ar)
1
δ 7.80 (d, 1H, J ) 9.1 Hz, Ar), 8.58 (dd, 1H, J ab)9.1 Hz and J bx)2.9 Hz, Ar),
-13.17 (14%)
0.36 (34%)
8
.94 (d, 1H, J ) 2.9 Hz, Ar)
8
δ 3.63 (s, 3H, CH3), 7.74 (d, 1H, J ) 9.1 Hz, Ar), 8.54 (dd, 1H, J ab)9.1 Hz and
J bx)2.7 Hz, Ar), 8.90 (d, 1H, J ) 2.7 Hz, Ar)
3
δ 7.86 (d, 1H, J ) 9.4 Hz, Ar), 8.48 (dd, 1H, J ab)9.4 Hz and J bx)2.8 Hz, Ar),
0.04 (ca. 0%)
8
.84 (d, 1H, J ) 2.8 Hz, Ar)
9
δ 3.89 (s, 3H, CH3), 7.52 (d, 1H, J ) 9.6 Hz, Ar), 8.39 (dd, 1H, J ab)9.6 Hz and
J bx)2.4 Hz, Ar), 9.08 (d, 1H, J ) 2.4 Hz, Ar)
Pi
2.17 (52%)
a
Relative areas of signals of identified compounds.
SCHEME 7
1
assumption that methylation has little effect on values
of the first and second pK , we estimate approximate
2
NH OH, but monomethylation in NHMeOH allows it to
a
compete with dephosphorylation and effectively inhibits
any reactions at phosphorus in the first-formed phos-
phorylated hydroxylamine, which would otherwise be
similar to those observed earlier.1
values of the second-order rate constants for attack in
-
-
-1 -1
BDNPP by NHMeO and NMe
2
O as 9.6 and 2.4 M
s
,
respectively.
Dimethylation on nitrogen does not significantly affect
O-attack on phosphorus (Scheme 4 and Table 1), but it
blocks subsequent reactions. Aromatic substitution by an
amino group is precluded by dimethylation, which also
modestly hinders O-attack, probably sterically.
Formation of DNP in the reaction of NHMeOH with
BDNPP is complex because it is formed directly by
dephosphoryation and more slowly from DNPP, the
product of aromatic nucleophilic substitution on BDNPP.
We had seen no evidence for aromatic substitution with
Con clu sion s
Scheme 8 summarizes our conclusions from kinetic
NMR spectroscopy and ESI-MS regarding the individual
reactions of BDNPP and intermediates with hydroxyl-
amine and its methylated derivatives and the extent to
which methylation controls initial reactions (top section
of Scheme 8) and those of first formed intermediates
(bottom section in Scheme 8). N-Monomethylation of NH
2
-
J . Org. Chem, Vol. 69, No. 18, 2004 6031

Domingos et al.
SCHEME 8
OH does not significantly affect the rate of O-phospho-
rylation of NHMeOH by BDNPP, but it affects subse-
quent reactions of the first-formed intermediate. N-Di-
methylation also has minor effects on the initial reaction
first-order kinetics for the initial nucleophilic attack upon the
substrate. Solutions were self-buffered by the amine/amine
hydrochloride at pH 4.0-7.0, prepared by addition of aqueous
standard NaOH (0.1 M; Merck) to aqueous amine hydrochlo-
ride, and by borate buffer (0.005 M) from pH 8 to 10, and
of NMe
product. O-Methylation strongly inhibits the initial at-
tack of NH OMe on phosphorus and then aromatic
2
OH that gives a long-lived O-phosphorylated
NaOH was used at higher pH.1
2 2
Reactions were in H O (or D O) at 25.0 °C, followed by
2
appearance of DNP at 400 nm on a diode-array spectropho-
tometer with a thermostated cell holder. The pH of each
reaction mixture was measured at the end of each run.
Observed first-order rate constants (kobs) were calculated with
a nonlinear least-squares fitting of the absorbance vs time
nucleophilic substitution contributes significantly, al-
though it is minor for the other hydroxylamines, which
are much more reactive nucleophiles than NH OMe.
2
Eletrospray MS and NMR spectroscopy are comple-
mentary techniques for identification of reaction inter-
mediates and products. The former allows sampling and
identification, but NMR provides structural information,
although data acquisition takes time, especially for nuclei
that relax slowly.
profile.
NMR Sp ectr oscop y. All 1H and 31P NMR spectra were
1
monitored on a spectrometer (400 MHz for H) in D
2
O at 25
°C, generally with a delay time of 1s. Relaxation is slow at
some nitroarene positions, and there were increases of 15-
1
2
0% in areas of some of the H signals when the delay time
was increased to 20 s at the end of reaction, but these long
delays create problems in examining spectra of intermediates
Exp er im en ta l Section
in the course of reaction.1 Concentrations estimated by
1
H
NMR spectroscopy were based on signals for which relaxation
is relatively fast. Most of the 31P NMR measurements during
reaction were made with a delay time of 1 s, but the areas did
not change when the delay time was increased to 3 s in
measurements at the end of reaction. The sodium salt of
Ma ter ia ls. BDNPP as the pyridinium salt was prepared
9
as described. The pyridinium ion was exchanged for sodium
+
ion on cation-exchange resin (Dowex 50W X8) in the Na form.
DNPP, as the pyridinium salt, was prepared by the procedure
1
2
of Rawji and Milburn. The hydroxylamines, as their hydro-
chlorides, and 2,4-dinitrophenol were of the highest purity
available and were used as purchased.
1
BDNPP was used in the NMR work, because H signals of the
1
1
pyridinium ion complicate the H NMR spectra. Aliphatic H
NMR signals were not useful because they are obscured by
signals of the excess methylated hydroxylamines and H-D
Kin etics. Reactions followed spectrophotometrically were
started by adding 30 µL of stock solution of the substrate (1 ×
1
exchange precludes examination of NH signals. The H and
C chemical shifts are referred to internal sodium 3-(tri-
methylsilyl)propionate (TSP), and those of P are referred to
-3
1
0
M) in water to 3 mL of reaction mixture, which contained
13
a large excess of the nucleophile (g0.1 M), ensuring strictly
31
external 85% phosphoric acid. The value of pD was obtained
(12) Rawji, G.; Milburn, R. M. J . Org. Chem. 1981, 46, 1205-1206.
2
by adding 0.4 to the observed pH of solutions in D O at 25
6
032 J . Org. Chem., Vol. 69, No. 18, 2004

Substitution Reactions of Methylated Hydroxylamines
C.13 Concentrations monitored by H and 31P NMR spectro-
1
spectroscopy, with comparisons with the spectra of authentic
material of BDNPP and DNPP, which were prepared as
described above; 2,4-dinitrophenol was used at the appropriate
°
scopy were generally taken after the initial reaction was
complete, so that compositions did not change significantly
during the measurements.
2-
pH,andO-phosphorylatedN-methylhydroxylamine,HNMeOPO
3
ESI-MS a n d ESI-MS/MS. All experiments were performed
on a hydrid triple quadrupole linear ion-trap mass spectrom-
eter. For typical electrospray ionization (ESI) conditions 1 mL
(designated 6), was prepared by reaction of potassium phos-
phoramidate with N-methyl hydroxylamine.14 Spectra were
monitored for reactions at pH where the nonionic nucleophile
is dominant. Products were examined with the hydroxylamines
in significant excess over BDNPP, and their concentrations
changed little in the course of reaction.
-
6
of 1 × 10 M BDNPP, in aqueous methanol (50% v/v) at pH
1
0, was mixed with 100 µL of 0.1 M aqueous MeNHOH. A
microsyringe pump delivered the reagent solution in to the
ESI source at a flow rate of 10 µL/min. ESI and the QqQ (linear
trap) mass spectrometer was operated in the negative-ion
mode. Main conditions were curtain gas nitrogen flow, 20 mL
Ack n ow led gm en t. We are indebted to PRONEX,
CAPES, CNPq, CNPq/NSF, FAPESP, Instituto do Mile-
nio de Materiais Complexos (FINEP) and the U.S. Army
Office of Research for financial support of this work.
Collaboration between the Brazilian and U.S. institu-
tions was made possible by the NSF-CNPq Cooperative
program.
-1
min ; ion spray voltage, -4500 eV; declustering potential, -21
eV; entrance potential, -10 eV; collision cell exit potential, -12
eV. The anionic species were subjected to collision-induced
dissociation (CID) with nitrogen by using collision energies
ranging from 5 to 45 eV.
P r od u cts. Most reaction products and intermediates were
identified by ESI-MS and ESI-MS/MS, absorption and NMR
J O0494735
(
14) J encks, W. P.; Gilchrist, M. J . Am. Chem. Soc. 1965, 87, 3199-
(13) Fife, T. H.; Bruice, T. C. J . Phys. Chem. 1961, 65, 1079-1080.
3209.
J . Org. Chem, Vol. 69, No. 18, 2004 6033