Multiplicity of reaction pathways in the processes of oxygen transfer to secondary amines by Mo(VI) and W(VI) peroxo complexes

Article abstract of DOI:10.1021/jo960544c

Oxidation of N,N-benzylmethylamine, N,N-benzylisopropylamine, and N,N-benzyl-tert-butylamine by both anionic and neutral Mo(VI) and W(VI) oxodiperoxo complexes yields the corresponding nitrones quantitatively. The oxidation reactions employing anionic oxidants were performed in CHCl₃ and follow second-order kinetics, first order with respect to the amine and to the oxidant. The data were rationalized on the basis of a rate-determining nucleophilic attack of the amine onto the peroxide oxygen of the oxidant, with a transition state in which N-O bond formation and O-O bond cleavage occur in a concerted way (electrophilic oxygen transfer mechanism). This attack yields the corresponding hydroxylamine, which then is furtherly oxidized to nitrone in a fast step. On the other hand, in the case of neutral oxidants ¹H-NMR data as well as kinetic data indicate that amine coordinates the metal center replacing the original ligand HMPA and yields a new peroxo complex. For N,N-benzyl-tert-butylamine such a complex was isolated and characterized. These new peroxo complexes can themselves behave as electrophilic oxidants, transferring oxygen to external amine molecules through the same pathway followed by anionic oxidants, or can yield the reaction product by intramolecular oxidation of the coordinate amine. Measurements of added HMPA effects on oxidation rate would seem more consistent with the electrophilic oxygen transfer mechanism.

Full text of DOI:10.1021/jo960544c

J . Org. Chem. 1996, 61, 6381-6387
6381
Mu ltip licity of Rea ction P a th w a ys in th e P r ocesses of Oxygen
Tr a n sfer to Secon d a r y Am in es by Mo(VI) a n d W(VI) P er oxo
Com p lexes†
Francesco P. Ballistreri,‡ Elena G. M. Barbuzzi,‡ Gaetano A. Tomaselli,*,§ and
Rosa M. Toscano‡
Dipartimento di Scienze Chimiche, University of Catania, V. le A. Doria 6, I-95125 Catania, Italy, and
Dipartimento di Chimica Organica, University of Palermo, Via Archirafi 26, I-90123 Palermo, Italy
Received March 21, 1996X
Oxidation of N,N-benzylmethylamine, N,N-benzylisopropylamine, and N,N-benzyl-tert-butylamine
by both anionic and neutral Mo(VI) and W(VI) oxodiperoxo complexes yields the corresponding
nitrones quantitatively. The oxidation reactions employing anionic oxidants were performed in
CHCl3 and follow second-order kinetics, first order with respect to the amine and to the oxidant.
The data were rationalized on the basis of a rate-determining nucleophilic attack of the amine
onto the peroxide oxygen of the oxidant, with a transition state in which N-O bond formation and
O-O bond cleavage occur in a concerted way (electrophilic oxygen transfer mechanism). This attack
yields the corresponding hydroxylamine, which then is furtherly oxidized to nitrone in a fast step.
1
On the other hand, in the case of neutral oxidants H-NMR data as well as kinetic data indicate
that amine coordinates the metal center replacing the original ligand HMPA and yields a new
peroxo complex. For N,N-benzyl-tert-butylamine such a complex was isolated and characterized.
These new peroxo complexes can themselves behave as electrophilic oxidants, transferring oxygen
to external amine molecules through the same pathway followed by anionic oxidants, or can yield
the reaction product by intramolecular oxidation of the coordinate amine. Measurements of added
HMPA effects on oxidation rate would seem more consistent with the electrophilic oxygen transfer
mechanism.
In tr od u ction
the electrophilic oxidations by peracids or by acidic
hydrogen peroxide.4
On the other hand, the peroxo complex bears a second
electrophilic site, i.e., the metal center. Therefore, a
nucleophilic substrate might attack the oxidant also or
alternatively on the metal and enter the coordination
sphere of the peroxo complex, thus behaving as a ligand
and yielding a substrate-oxidant association complex.
Such a complex might evolve into products through
intramolecular events (which in some documented cases
imply radical processes)5 or it might behave as a new
electrophilic oxidant transferring oxygen to a second
external molecule of substrate.6
The data obtained in this investigation disclose some
peculiar mechanistic aspects for the oxidation of second-
ary amines by Mo(VI) and W(VI) peroxo complexes. Such
substrates undergo different reaction pathways depend-
ing on the nature of the amine and of the ligand.
Oxidation of amines is a reaction of relevant interest
both for biological implications (e.g., the in vivo metabo-
lism) and for synthetic applications aimed to introduce
oxygen into nitrogen-containing compounds.1 It has been
reported that primary and secondary amines are ef-
ficiently oxidized by Mo(VI) and W(VI) peroxo complexes
and particularly that benzylalkylamines are selectively
and quantitatively transformed into nitrones.2
Despite the interest and the relevance of such reactions
no kinetic study has been undertaken in order to collect
some piece of information on the possible reaction
pathways. This prompted us to mechanistically inves-
tigate oxidation reactions of model benzylalkylamines to
nitrones by Mo(VI) and W(VI) peroxo complexes. Previ-
ous investigations indicated that organic substrates with
reaction sites carrying lone pairs or π electrons, such as
thioethers and alkenes, react as nucleophiles toward Mo-
(VI) and W(VI) peroxo complexes, through a clean bimo-
lecular process, attacking the peroxide oxygen along the
O-O bond (electrophilic oxygen transfer).3 This mech-
anism is very similar to that suggested by Bartlett for
Resu lts a n d Discu ssion
Oxid a tion Rea ction s w ith P ICO-M, P IC-M, a n d
BENZ-Mo (M ) MoVI, WVI). As shown in Chart 1, these
peroxo complexes are ammonium salts, whose anionic
parts bear the transferable oxygen (for this reason they
are called “anionic oxidants”). PIC and PICO are coor-
dinatively saturated peroxometal complexes, containing
* To whom correspondence should be addressed. Fax: 0039 95
580138. E-mail: gtomaselli@dipchi.unict.it.
† Dedicated to Professor Giorgio Modena on the occasion of his 70th
birthday.
‡ University of Catania.
(3) (a) Arcoria, A.; Ballistreri, F. P.; Spina, E.; Tomaselli, G. A.;
Toscano, R. M. Gazz. Chim. Ital. 1990, 120, 309-313. (b) Ballistreri,
F. P.; Tomaselli, G. A.; Toscano, R. M. J . Mol. Catal. 1991, 68, 269-
275. (c) Ballistreri, F. P.; Bazzo, A.; Tomaselli, G. A.; Toscano, R. M.
J . Org. Chem. 1992, 57, 7074-7077.
(4) Bartlett, P. D. Record. Chem. Progr. 1950, 11, 47-56. See also:
Curci, R.; Edwards, J . O. In Organic Peroxides; Swern, D., Ed.; Wiley
Interscience: New York, 1970; Vol. 1, pp 199-264 and references
therein.
§ University of Palermo.
X Abstract published in Advance ACS Abstracts, August 1, 1996.
(1) (a) Poulsen, L. L.; Kadlubar, F. F.; Ziegler, D. M. Arch. Biochem.
Biophys. 1974, 164, 774-775. (b) Sheldon, B.; Bruice, T. C. J . Am.
Chem. Soc. 1980, 102, 6498-6503.
(2) (a) Murahashi, S. I.; Mitsui, H.; Shiota, T.; Tsuda, T.; Watanabe,
S. J . Org. Chem. 1990, 55, 1736-1744. (b) Ballistreri, F. P.; Chiacchio,
U.; Rescifina, A.; Tomaselli, G. A.; Toscano, R. M. Tetrahedron 1992,
48, 8677-8684. (c) Tollari, S.; Bruni, S.; Bianchi, C. L.; Rainoni, M.;
Porta, F. J . Mol. Catal. 1993, 83, 311-322.
(5) Campestrini, S.; Di Furia, F. Tetrahedron 1994, 50, 5119-5131.
(6) Di Furia, F.; Modena, G. Pure Appl. Chem. 1982, 54, 1853-1866.
S0022-3263(96)00544-0 CCC: $12.00 © 1996 American Chemical Society

6382 J . Org. Chem., Vol. 61, No. 18, 1996
Ballistreri et al.
Ta ble 1. Secon d -Or d er Ra te Con sta n ts for th e
Stoich iom etr ic Oxid a tion of N,N-Ben zyla lk yla m in es
C6H5CH2NHR to Nitr on es by An ion ic P er oxo Com p lexes
in CHCl3
Ch a r t 1
R )
-CH3 104k2 -CH(CH3)2 104k2 -C(CH3)3 104k2
a
a
b
A
oxidant
(M-1 s-1
)
(M-1 s-1
)
(M-1 s-1
)
1. PICO-Mo
2. PIC-Mo
3. BENZ-Mo
4. PIC-W
8.80
7.93
10.9
22.4
1.35
0.980
1.19
4.08
1.44
3.46
15.9
20.2 (2.75a)
36.6
5. PICO-W
a
b
T ) 10 °C. T ) 40 °C.
strong nucleophiles such as thioethers to sulfoxides.9
Consequently, the results can be rationalized on the basis
of the external electrophilic oxygen transfer mechanism,
which involves a rate-determining nucleophilic attack of
the amine onto the peroxide oxygen of the oxidant, with
a transition state in which N-O bond formation and
O-O bond cleavage occur in a concerted way. This attack
yields the corresponding hydroxylamine, which then in
a faster way is further oxidized to nitrone.10
the bidentate picolinate and picolinate N-oxide ligands,
respectively.7 Therefore, coordination of a substrate to
the metal would require the opening of a tooth of the
bidentate ligand in order to make available a coordination
site. On the other hand, BENZ-Mo contains the mono-
dentate carboxylate ligand and has a free coordination
site. However, X-ray data indicate that the apical
position is not completely free and some hindrance to the
entering of a ligand may occur, due to the short distance
between the molybdenum atom and the carbonyl oxygen
of the carboxylate group.8
The amine reactivity sequence Me > Pri > But might
reflect the presence of steric effects during the nucleo-
philic attack, which favor the less hindered nucleophile,
i.e., C6H5CH2NHCH3.
The same mechanism suggested for molybdenum de-
rivatives still keeps for tungsten oxidants. The higher
reaction rate displayed by these oxidants follows the
observed general trend that W(VI)-peroxo complexes are
nearly 10-30 times more reactive than Mo(VI)-peroxo
complexes in reactions involving an external attack to
the peroxide oxygens by a nucleophilic substrate.11 This
behavior probably can be ascribable to a Lewis acidity of
tungsten peroxo complexes higher than that of molyb-
denum derivatives, which might be responsible of a
higher electrophilicity of peroxide oxygens.12
N,N-Benzylmethylamine (Me), N,N-benzylisopropyl-
amine (Pri), and N,N-benzyltertbutylamine (But) are
transformed quantitatively into nitrones by these peroxo
complexes in CHCl3 according to the following equation:
In order to have further support for the mechanistic
picture that envisages a simple bimolecular oxidation
reaction and check whether a coordination process is or
(9) Campestrini, S.; Conte, V.; Di Furia, F.; Modena, G. J . Org.
Chem. 1988, 53, 5721-5724.
(10) In order to check the intermediacy of hydroxylamine in the
formation of nitrone, we investigated the oxidation of N,N-dibenzy-
lamine and of the corresponding hydroxylamine, i.e., N,N-dibenzylhy-
droxylamine, by PIC-W in CHCl3. Both substrates yield the same
nitrone quantitatively, and the oxidation rate of N,N-dibenzylhydroxyl-
amine is nearly 10 times larger than that of N,N-dibenzylamine. Thus,
these findings seem to give support to the probable involvement of
hydroxylamine as an intermediate in these oxidation processes.
(11) Arcoria, A.; Ballistreri, F. P.; Tomaselli, G. A.; Di Furia, F.;
Modena, G. J . Mol. Cat. 1984, 24, 189-196.
(12) Cyclic voltammetry measurements, performed by Di Furia et
al. (Bonchio, M.; Conte, V.; Di Furia, F.; Modena, G.; Moro, S.;
Carofiglio, T.; Magno, F.; Pastore, P. Inorg. Chem. 1993, 32, 5797-
5799), indicated for W peroxo complexes reduction peak potentials less
negative than those observed for the corresponding Mo derivatives.
Furthermore, the authors reported a good correlation between these
reduction peak potentials and IR O-O bond frequencies. On the basis
of these observations, it might be conceivable to assume that the LUMO
of W peroxo complexes, involving the peroxide oxygens, is lower in
energy than that of Mo peroxo complexes and, therefore, that W peroxo
complexes have a Lewis acidity higher than that of Mo oxidants.
The reactions follow second-order kinetics, first order
with respect to each reactant. Pertinent results are
displayed in Table 1. For the molybdenum series the
data suggest that the coordination sphere of the metal
does not play any appreciable role. In fact, the observed
reactivity sequence PICO-Mo = BENZ-Mo = PIC-Mo is
particularly telling in showing the scarce sensitivity of
the reaction rate upon the identity of the coordination
sphere of the oxidant, excluding therefore the possibility
of an attack of the substrate to the metal center. This
behavior is similar to that observed in the oxidation of
(7) Bortolini, O; Campestrini, S.; Di Furia, F.; Modena, G. J . Org.
Chem. 1987, 52, 5467-5469.
(8) Campestrini, S.; Di Furia, F.; Rossi, P.; Torboli, A. J . Mol. Catal.
1993, 83, 95-105.

Oxygen Transfer to Secondary Amines
J . Org. Chem., Vol. 61, No. 18, 1996 6383
Ta ble 2. Oxid a tion of C6H5CH2NHBu t by
Sch em e 1
MoO5C6H5CH2NHBu t a n d by MoO5HMP A in CH3CN a t 40
°C
[amine] 104‚k1
103‚k1/[amine]
(M-1 s-1
(M)
(s-1
)
)
oxidant
0.124
0.224
0.303
0.406
0.604
1.76 1.42
2.99 1.33
MoO5C6H5CH2NHBut
3.37 1.11 k2 ) 1.23 × 10-3
4.81 1.18
6.80 1.13
0.123
0.300
0.409
0.508
0.602
1.00
1.74 1.41
3.41 1.14
MoO5HMPA
5.15 1.26
6.30 1.24 k2 )1.23 × 10-3
7.00 1.16
11.8
1.18
Ta ble 3. Oxid a tion of C6H5CH2NHR (R ) Me, P r i) by
MoO5HMP A in CH3CN a t 40 °C
R
[amine] (M)
103‚k1 (s-1
)
103‚k1/[amine](M-1 s-1
)
Me
0.089
0.153
0.400
0.788
1.01
1.40
1.60
1.91
2.52
3.97
4.66
5.80
5.80
7.45
6.70
6.45
28.3
25.9
11.7
7.36
5.74
5.32
4.19
3.38
Pri
0.202
0.802
1.30
1.32
3.95
5.30
6.90
6.53
4.93
4.08
3.43
is not occurring, we performed a competitive experiment
with two different amines, i.e., Pri and But. These two
amines were pitted together to react with PIC-W at 10
°C in CHCl3. If the two substrates were competing to
form an appreciable concentration of oxidant-amine
complexes, the observed ratio of the two corresponding
nitrones should be different from that calculated from
the rate constants measured in separate experiments.13
What comes out from the pertinent results (Scheme 1)
indicates that the observed ratio is, within the experi-
mental error, that one expected from the rate constant
ratio and suggests the absence of prior coordination
equilibria. Furthermore, the activation parameters mea-
sured for the oxidation of Pri with PIC-W, ∆Hq293K ) 51
kJ mol-1, and ∆Sq293K ) -110 J K-1 mol-1, are those
expected for a simple bimolecular mechanistic scheme.
Oxid a tion Rea ction s w ith MO(O2)2HMP A (M )
MoVI, WVI). P r eequ ilibr ia In volvin g Coor d in a tion
of Am in e to th e Oxid a n t. The peroxo derivatives of
the title, differently from the anionic oxidants of the
previous section, are neutral peroxometal complexes with
a free coordination site. Thus, there is a higher prob-
ability that with these oxidants the amine might attack
the metal center entering the coordination sphere of the
metal itself.
2.01
in Table 2. It can be noted that the process obeys a
second-order rate law, first order in each reactant. This
observation seems consistent with an external nucleo-
philic attack of the amine onto the peroxide oxygen, as
discussed in the previous section.
Oxidation reactions of amines with the two neutral
peroxo complexes MoO(O2)2HMPA and WO(O2)2HMPA,
respectively, were also performed in CH3CN.
Under these conditions nitrones are again the oxida-
tion products. The pertinent kinetic data are displayed
in Tables 2-5.
For the reaction of MoO5HMPA with But (Table 2) we
observed a linear dependence of the pseudo-first-order
rate constant kobs on the amine concentration (a second-
order overall reaction), which indicates that a simple
bimolecular process is occurring. Furthermore, the
observed k2 value is equal, within the experimental error,
to that observed employing 5a as an oxidant. Therefore,
these findings strongly suggest that the oxidation of But
in these two cases is carried on by the same species, i.e.,
MoO5C6H5NHBut. This implies that when the starting
oxidant used is MoO5HMPA a faster prior step is occur-
ring, in which a molecule of amine coordinates the metal
center replacing the original ligand HMPA and trans-
forming readily and quantitatively MoO5HMPA into the
new peroxo complex 5a , which then is attacked externally
by a second amine molecule onto the peroxide oxygen to
lead to the formation of products (Scheme 2).
Indeed, we observed that on adding N,N-benzyl-tert-
butylamine into a CHCl3 solution of MoO5HMPA a
precipitate is readily formed. The precipitate was iso-
1
lated and analyzed. Elemental analysis, H-NMR, and
iodometric titer indicated the formation of a new peroxo
complex, MoO5C6H5CH2NHBut (5a ), which is obtained
from MoO5HMPA by substitution of the HMPA ligand
with a molecule of amine. Since such a complex is soluble
in CH3CN we decided to employ it to oxidize But in such
solvent. The reaction affords the corresponding nitrone
quantitatively, and the pertinent kinetic data are shown
Sch em e 2
K
MoO5L + R2NH
\
(1)
(2)
MoO5R2NH + R2NH
(13) Webb, J . L. In Enzymes and Metabolic Inhibitors; Academic
Press: New York, 1963; pp 52-68.
This scheme yields a rate law, v1 ) k2[MoO5L]0[R2NH]0

6384 J . Org. Chem., Vol. 61, No. 18, 1996
Ballistreri et al.
(iii) MO5L‚R2NH (ox3) + R2NH h
MO5‚2R2NH (ox4) + L
The species ox2 in tha case of N,N-benzyl-tert-butyl-
amine corresponds to the peroxo complex MoO5C6H5CH2-
NHBut, isolated as reported in the previous section.
Since the nucleophilicity of R2NH is higher than that
of HMPA the electrophilicity of the peroxide oxygens is
expected to decrease on increasing the molecule number
of amines in the coordination sphere of the metal, i.e.,
MO5L (ox1) > MO5R2NH (ox2) > MO5L‚R2NH (ox3) >
1
MO5‚2R2NH (ox4). By H-NMR we observed Lfree/Lbound
ratio values of 2.9, 7.5, and 175 for R ) Me, Pri, and But,
respectively, (amine/MoO5L ) 1).
The practically neglectable amount of Lbound observed
in the case of But indicates the absence in solution of
1
species bearing the ligand HMPA. Thus, H-NMR ob-
servations also suggest a large value for K (eq 1),
confirming the reaction mechanism reported in Scheme
2.
F igu r e 1. Oxidation of C6H5CH2NHR (R ) Me, Pri, But) by
MoO5HMPA in CH3CN at 40 °C.
However, since under the pseudo-first-order conditions
amine/oxidant ratios in the range 6-100 were adopted,
we performed 1H-NMR measurements employing an
amine/oxidant ratio ) 10. Under such conditions, in all
cases we observed only the signals of Lfree. Therefore, it
seems reasonable to consider equilibrium 1 quantitatively
shifted to the right side also for Me and Pri and so we
are left with the possibility that the oxidant present in
solution should be ox2, regardless of amine or of metal
identity. But, due to the large excess of amine, a second
equilibrium, i.e., the equilibrium forming ox4, should also
be taken in consideration.
(if K[R2NH] . [L]), in agreement with the observed rate
law and with all the experimental findings.
1H-NMR Mea su r em en ts. On the other hand, the
data concerning the oxidation of Me and Pri (Table 3) and
the relative plots, reporting the observed pseudo-first-
order rate constants kobs versus the concentration of
amine, show a nonlinear dependence of the oxidation rate
on the amine concentration (Figure 1), at variance with
the behavior of But. In order to get information on
possible coordination processes of amines with the oxi-
dants, which might be responsible of these observed
K′
1
kinetics, we performed H-NMR spectra of CD3CN solu-
MO5R2NH + R2NH y
\
z MO5‚2R2NH
(3)
tions containing a 1:1 ratio amine/oxidant. The spectra
were recorded at -20 °C, because at such a temperature
the oxidation process does not occur during the recording
time interval (no signals due to the formation of nitrones
are observed). Under these experimental conditions an
0.058 M solution of MO5HMPA (M ) Mo, W) in CD3CN
exhibits a doublet centered at 2.78 ppm due to the CH3
resonance of the ligand (Lbound). Upon addition of 0.058
M of amine a decrease in the signals at 2.78 ppm and
the appearance of a new doublet centered at 2.55 ppm is
In this case product formation also occurs through an
additional route (eq 4), under the assumption that both
MO5R2NH and MO5‚2R2NH oxidize the amine electro-
philically, the latter being less effective than the former.
MO5‚2R2NH + R2NH
9
k2′8 P
(4)
A more complete reaction scheme, including eqs 2-4,
should be considered, and the following rate law can be
derived.
1
observed. By comparison with the H-NMR spectrum of
HMPA in the same solvent, the signals centered at 2.55
ppm are attributed to the CH3 resonance of the free
ligand (Lfree). Furthermore, an increase in the amount
of the added amine causes an increase in the signals at
2.55 ppm (Lfree) at the expense of those at 2.78 ppm
(Lbound). Meanwhile, the benzylic protons of amine, which
lie at 3.68 ppm, are moved downfield (4.15 ppm). 1H-
NMR experiments suggest that a coordination process
involving the amine is occurring. Unfortunately, we were
not able to get any precipitate from the solution (even in
CHCl3, addition of C6H5CH2NHCH3 or C6H5CH2NHPri
to a solution of MoO5HMPA does not yield any precipi-
tate). On the other hand, due to these coordination
2
k2[R2NH]0[MO5L]0 + k′2K′[R2NH]0 [MO5L]0
v2 )
(5)
1 + K′[R2NH]0
In order to check the reliability of the kinetic model
we processed the data concerning the oxidation process
of C6H5CH2NHMe with MO5HMPA (Table 3) by a com-
puter iterative procedure. The outcome (Figure 2) indi-
cates a fair agreement between the calculated (curve) and
the observed (full triangles) values and allows us to
obtain the following values k2 ) (5.26 ( 0.516) × 10-2
,
k′2 ) (5.68 ( 1.27) × 10-4 and K′ ) 8.40 ( 0.48. The k2
and k′2 values, reckoned by eq 5, are in accord with the
expectation that MoO5R2NH is an electrophilic oxidant
more reactive than MoO5‚2R2NH.
1
equilibria and compatibly to H-NMR observations, dif-
ferent species might be formed as shown in the following
equations:
Equation 5 becomes equal to v1 for K′ = 0. This is the
case for But, probably because the size of the But group
makes the entrance of a second molecule in the coordina-
tion sphere of Mo difficult, at least in the explored range
of amine concentrations.
(i) MO5L (ox1) + R2NH h MO5‚R2NH (ox2) + L
(ii) MO5L (ox1) + R2NH h MO5L‚R2NH (ox3)

Oxygen Transfer to Secondary Amines
J . Org. Chem., Vol. 61, No. 18, 1996 6385
Ta ble 5. Oxid a tion of C6H5CH2NHR (R ) Me, P r i, Bu t) by
WO5HMP A in CH3CN a t 40 °C
R
[amine] (M)
103‚k1 (s-1
)
103‚k1/[amine] (M-1 s-1
)
Mea
0.394
0.795
1.20
1.60
2.01
0.660
1.26
1.76
1.67
1.60
1.68
1.58
1.47
1.02
0.796
Pri
0.210
0.799
1.60
1.11
1.92
2.64
3.34
5.29
2.40
1.65
1.58
2.12
But
0.0507
0.200
0.800
1.30
1.42
2.27
2.93
2.73
28.0
11.4
3.66
2.11
a
Values at 10 °C.
F igu r e 2. Oxidation rates of Me by MoO5HMPA in CH3CN.
Calculated values (curve) and observed values (full triangles).
Ta ble 4. Oxid a tion of C6H5CH2NHR (R ) Me, P r i, Bu t) by
MoO5HMP A in CH3CN in th e P r esen ce of HMP A a t 40 °C
[amine] (M)
[HMPT] (M)
103‚k1 (s-1
)
Me
3.97
(0.153)
0.149
0.793
1.59
2.71
0.725
0.439
Pri
1.32
(0.202)
0.117
0.325
0.577
1.49
1.11
0.595
0.422
0.450
But
(0.200)
0.246
0.271
0.334
0.408
0.399
0.500
1.00
2.00
3.00
F igu r e 3. Oxidation of C6H5CH2NHR (R ) Me, Pri, But) by
WO5HMPA in CH3CN at 40 °C.
observed for But itself with MoO5HMPA. The difference
might be due to the fact that for a second amine molecule
to enter the coordination sphere of the metal center is
easier with W than with Mo.
Mich a elis-Men ten Kin etic Mod el. A priori we can-
not exclude the possibility that the formation of an
amine-oxidant association complex is followed by the
oxidation reaction of the coordinate amine within the
coordination sphere of the metal, according to Scheme
3:
This reasoning seems supported by the data reported
in Table 4, which show the effects of externally added
HMPA on the oxidation rate of C6H5CH2NHR by MoO5-
HMPA. The addition of HMPA should give rise to the
formation of the oxidant species MO5HMPA‚R2NH,
through the equilibrium 6,
K′′
MO5R2NH + HMPA \y z MO5HMPA‚R2NH (6)
Sch em e 3
which is less reactive than MO5R2NH, and therefore, a
decrease in rate should be observed. This expectation is
fulfilled for R ) Me and Pri, which show a rate depression
about 10 and 3 times, respectively, when HMPA = 1.6
M is added, whereas for R ) But essentially no rate
decrease is observed on adding HMPA up to = 3.0 M,
indicating the difficulty of a second ligand to enter the
coordination sphere of MoO5C6H5CH2NHBut.
MO5L + R2NH y\
Kz MO5R2NH + L
K′
MO5R2NH + R2NH y
\z MO5‚2R2NH
k1
MO5R2NH
9
8 P
(7)
(8)
Reactivity data concerning the oxidation reactions with
WO5HMPA (Table 5) find a similar rationale, i.e., the
reactivity curve profiles, obtained plotting kobs values
versus amine concentrations, depend on the simultaneous
presence of the two electrophilic oxidizing species WO5R2-
NH and WO52R2NH, the latter less reactive than the
former. Obviously, WO52R2NH concentration increases
on increasing the initial concentration of the amine
employed.
MO52R2NH
9
k1′8 P
The kinetic expression related to Scheme 2 is
k1[MO5L]0 + k′1K′[R2NH]0[MO5L]0
v3 )
(9)
1 + K′[R2NH]0
In the case of WO5HMPA the behavior of But is similar
to that of Me and Pri (Figure 3) but different from that
assuming a large value for K and thus [MO5L]0 ) [MO5R2-
NH]0.

6386 J . Org. Chem., Vol. 61, No. 18, 1996
Ballistreri et al.
F igu r e 5. Effects of externally added HMPA on the oxidation
rate of Me with MoO5HMPA in CH3CN at 40 °C. [Me] ) 0.16M;
[MoO5HMPA] ) 0.016M.
F igu r e 4. Oxidation rates of Me by MoO5HMPA in CH3CN.
Calculated values (curve) and observed values (full squares).
value (1/[HMPA] f 0, [HMPA] f ∞), kobs ) 4.98 × 10-4
s-1, which is significantly different from zero. Such a
value is equal, within the experimental error, to that
observed (kobs ) 4.46 × 10-4 s-1) performing the same
reaction in HMPA itself as a solvent.
Therefore, such evidence seems consistent with an
electrophilic behavior toward amines by these oxidants
in analogy to that previously observed toward thioethers
and alkenes.
We processed again by iterative procedure the data
concerning the oxidation reaction of C6H5CH2NHCH3
with MO5HMPA according to eq 9, and the outcome,
obtained for k1 ) (1.55 ( 1.17) × 10-3, k′1 ) (7.82 ( 1.1)
× 10-3, and K′ ) 2.85 ( 1.08, indicates a fair agreement
(Figure 4) also in this case between the experiments and
the kinetic model of Scheme 3.
For C6H5CH2NHCH(CH3)2 we obtained k1 ) (1.61 (
1.76) × 10-4, k′1 ) (1.44 ( 0.140) × 10-2, and K′ ) 0.441
( 0.083.
Con clu sion s
Therefore, the kinetic treatment does not allow us to
discriminate between the external electrophilic pathway
and the Michaelis-Menten-type behavior.
HMP A Effects. On the other hand, it might be
helpful in such systems to add a strong, nonoxidizable
ligand, typically HMPA itself, and to study the effect on
oxidation rate.
Oxidation of N,N-benzylalkylamines by Mo(VI) and
W(VI) peroxo complexes yield quantitatively nitrones
through a reaction mechanism that depends on the ligand
of the complex. When the ligand is strongly bound to
the metal and it is not easy to make available a coordina-
tion site (anionic oxidants) the amine attacks externally
the peroxide oxygen and, through a Bartlett-type transi-
tion state, yields the corresponding hydroxylamine, which
is rapidly converted to nitrone. On the other hand, when
the metal presents a coordination site free (neutral
oxidants) the amine attacks the metal center and, under
the experimental conditions adopted, removes the pre-
existing ligand HMPA and yields a new oxidant species,
containing substrate molecules in the coordination sphere,
which is itself able to act as an electrophilic oxidant
toward the amine molecules.
Alternatively, the amine-oxidant association complex
might evolve to products through an oxidation process
that involves the coordinate substrate molecules (internal
oxidation) and that presumably requires radical events.
However, a study of oxidation rate effects on adding
increasing amounts of HMPA seems to be consistent with
the external electrophilic oxygen transfer mechanism.
Within the framework of the mechanism that involves
intramolecular oxidation of the coordinate substrate, we
would expect that increasing amounts of added HMPA
will continously inhibit the reaction rate until it will drop
to zero, since the displacement of the coordinate amine
by HMPA will give firstly MO5HMPA‚R2NH, which will
be then converted into MO5(HMPA)2 on increasing the
concentration of HMPA.
Alternatively, on the basis of the external electrophilic
oxygen transfer mechanism, we would expect a continous
slowdown of the reaction rate until a minimum value of
the oxidation rate is reached, which corresponds to the
oxidation rate performed by the species MO5(HMPA)2.
This species, in fact, which is presumably the only
oxidant present in solution at high concentration of
HMPA, is still able to act as an electrophilic, even if poor,
oxidant. The absence, under such conditions, of oxidant
species bearing amine molecules in the coordination
sphere of the metal, does not make operative the internal
mechanism.
Exp er im en ta l Section
Ma ter ia ls. Chloroform (Carlo Erba, RPE) and acetonitrile
(Merck, LiChrosolv) were distilled before use over P4O10 and
over calcium hydride, respectively. N,N-Benzylmethylamine,
N,N-benzylisopropylamine, and N,N-benzyl-tert-butylamine
(Aldrich) were purified by distillation over calcium hydride.
Tungsten(VI) and molybdenum(VI) oxodiperoxo complexes
MO(O2)2HMPA,14 molybdenum(VI) oxodiperoxopicolinate [MO-
Therefore, a plot of kobs vs 1/[HMPA] should display a
straight line, whose intercept is zero in the case of the
internal oxidation, whereas it is different from zero in
the case of the electrophilic oxygen transfer mechanism.
The plot obtained from the data concerning the oxida-
tion of C6H5CH2NHCH3 by MoO5HMPT in CH3CN (Fig-
ure 5), employing a low concentration of amine and
adding increasing amounts of HMPA, shows an intercept
(14) Mimoun, H.; Seree De Roche, I.; Sajus, L. Bull. Soc. Chim. Fr.
1969, 1481-1492.

Oxygen Transfer to Secondary Amines
J . Org. Chem., Vol. 61, No. 18, 1996 6387
(O2)2O2CNC5H4]- [(C4H9)4N+] (PIC),15 Mo(VI) oxodiperoxo pi-
colinate N-oxide [MoO(O2)2O2C(O)NC5H4]- [(C4H9)N+] (PI-
CO),15 and oxodiperoxo benzoate [MoO(O2)2O2CC6H5]-
[(C4H9)4N+] (BENZ)16 were prepared by the previously reported
original procedures. (Ca u tion , exp losion h a za r d !)
P r ep a r a tion of MoO(O2)2C6H5CH2NHBu t. To a 5 mL
chloroform solution, containing 0.200 mmol of C6H5CH2NHBut,
was added at room temperature a 5 mL chloroform solution
containing 0.203 mmol of MoO(O2)2HMPA. The precipitate
readily formed was filtered, washed with CHCl3, and dried.
Elemental analysis (C, 38.99; H, 5.31; N, 4.07, respectively)
is in agreement with the calculated values (C, 38.92; H, 5.05;
N, 4.13) referred to the formula MoO(O2)2C6H5CH2NHBut. An
aliquot of the sample dissolved in CD3CN gave the following
1H-NMR spectrum: δ (CD3CN) 7.44-7.60 (m, 5H, aromatic
protons), 4.20 (s, 2H, benzylic protons), 1.47 (s, 9H, CH3),
iodometric titer of 0.0689 mmol of compound yielded 0.1322
mmol of active oxygen (96%), as expected for a diperoxo
complex of general formula MO(O2)2L.
due to the starting amine). Usually, the agreement between
the two sets of rate constants, obtained by titration and by
UV measurements, respectively, is within (5% (rate data,
obtained following the consumption of the oxidant by titration,
must be divided by 2 to match the corresponding rate data
obtained by UV since only half-oxidant is involved in the rate-
determining formation of hydroxylamine).
Reaction products were isolated and characterized as re-
ported previously.2b
For the study of the stoichiometry of the oxidation reactions
an appropriate internal GC standard was added in the reaction
mixture. Aliquots of the mixture were withdrawn at the
desired time interval, and the amount of the products formed
was determined by GLC, whereas in the same aliquot the
content of the oxygen not reacted was determined by iodomet-
ric titration. Usually, conversion values are in the range 90-
95%, whereas the yields in nitrone are 80-90% with respect
to the oxidant reacted. For the reaction of C6H5CH2NHMe
with MoO5HMPA (conversion 98%), we obtained 82% of the
corresponding nitrone and 12% of C6H5CHdNOH; in such a
case the nitrone was determined by 1H-NMR (internal stan-
dard (C6H5)3CH) monitoring the CH3 resonance of nitrone (3.80
ppm), whereas the oxime was determined by HPLC, employing
a Hypersil silica column (0.5 µm, 250 × 4.6 mm), flow rate 1.0
mL/min, detector set at λ ) 275 nm, standard ) biphenyl;
mobile phase, (A) hexane, (B) chloroform; gradient, 0-10 min
100% A, 10-15 min 30% B, 15-20 min 30% B, 20-25 100%
A.
In str u m en ta tion . Gas chromatographic analyses of the
reaction mixtures were carried out on a Perkin-Elmer 8420
gas chromatograph equipped with a flame ionization detector
and program capability; a 25-m SE-30 capilllary column was
used for product determinations.
HPLC analyses were performed on a Varian 5000 instru-
ment equipped with a UV-vis detector (J ASCO UV DEC-100)
and a LCI-100 Perkin-Elmer integrator.
GC-MS analyses of reaction mixtures were performed on
a Hewlett-Packard Model 5890 gas chromatograph (using an
HP-1 dimethylpolysiloxane 25 capillary column) equipped with
a Hewlett-Packard MS computerized system Model 5971A.
UV measurements were performed by a UV/vis Perkin-
Elmer LAMBDA 2S spectrometer, using 1 cm thermostatic
quartz cells.
1H-NMR measurements were obtained on a Bruker AC 200
MHz spectrometer.
C, H, N, analyzer Fisons Model EA 1108 was used for
elemental analyses.
For comparison, C6H5CH2NHBut gave the following 1H-
NMR spectrum: δ (CD3CN) 7.20-7.32 (m, 5H, aromatic
protons), 3.72 (s, 2H, benzylic protons), 1.17 (s, 9H, CH3).
Thus, the downfield shift of benzylic protons (0.48 ppm) and
CH3 protons (0.30 ppm), respectively, is consistent with an
amine molecule bound to the metal.
The same compound is obtained starting with 1.56 mmol of
amine and 0.159 mmol of oxidant.
Kin etics. All kinetic runs were carried out under
a
nitrogen atmosphere. In a typical run, 5 mL of CHCl3 (CH3-
CN) containing 0.15 mmol of peroxo complex was added to a
CHCl3 (CH3CN) solution (5 mL) containing 1-15 mmol of
amine in
a glass reactor maintained at the appropriate
temperature. Aliquots of the reaction mixture were withdrawn
at various time intervals, and the disappearance of the oxidant
was followed by iodometric titration, under pseudo-first-order
conditions, employing a 10-100 times excess of substrate over
the oxidant. Blank experiments indicated that the decomposi-
tion of the oxidant is a negligible process in the time scale of
the oxidation reactions. The reactions followed pseudo-first-
order kinetics for 2 half-lives; pseudo-first-order rate constants,
obtained as slopes from conventional plots of ln [O]act vs time,
were evaluated by using a linear least-squares computer
program (Plot it 3.0) and were reproducible within (5%.
Kinetics were followed also by UV technique monitoring the
formation of the nitrone; thus, in the case of C6H5CH2NHPri
the formation of the nitrone was monitored at λ ) 297 (
)
2.03 × 104) and the pseudo-first-order rate constants were
obtained from plots of ln ([product]∞ - [product]t) vs time (after
correcting the o.d. values for a small and constant absorption
Ack n ow led gm en t. Financial support from CNR
and from MURST is gratefully acknowledged.
(15) Bortolini, O.; Campestrini, S; Di Furia, F.; Modena,G. J . Org.
Chem. 1987, 52, 5093-5095.
(16) Campestrini, S. Ph.D. Thesis, Padova University, 1988.
J O960544C