On the Reaction of Aminoxyls with Dioxiranes

Article abstract of DOI:10.1002/(SICI)1099-0690(199805)1998:5<871::AID-EJOC871>3.0.CO;2-F

In the reactions of dimethyldioxirane (1a) and methyl(trifluoromethyl)dioxirane (1b) with 2,2,6,6-tetramethylpiperidinyl-1-oxyl (2) (TEMPO) in acetone, the corresponding methoxyamine 1-methoxy-2,2,6,6-tetramethylpiperidine (5) is produced in ≥98% yield, b

Full text of DOI:10.1002/(SICI)1099-0690(199805)1998:5<871::AID-EJOC871>3.0.CO;2-F

FULL PAPER
On the Reaction of Aminoxyls with DioxiranesȠ
Anna Dinoia, Ruggero Curci*a, Patricia Carlonib, Elisabetta Damianib, Pierluigi Stipab, and Lucedio Greci*b
a
`
Centro C.N.R. “M.I.S.O.”, Dipartimento Chimica, Universita di Bari ,
via Amendola 173, I-70126 Bari, Italy
b
`
Dipartimento di Scienze dei Materiali e della Terra, Universita di Ancona ,
via Brecce Bianche, I-60131 Ancona, Italy
Received November 3, 1997
Keywords: Dioxirane / Homolysis, induced / Aminoxyl / Oxidation / Spin-trapping
In the reactions of dimethyldioxirane (1a) and methyl(trifluo-
romethyl)dioxirane (1b) with 2,2,6,6-tetramethylpiperidinyl-
2-phenyl-3H-indol-3-one-1-oxyl (4), with dioxiranes 1a and
1b in acetone have also been studied. In these cases, not only
1-oxyl (2) (TEMPO) in acetone, the corresponding methoxy- is the corresponding methoxyamine 8a produced (yield
amine 1-methoxy-2,2,6,6-tetramethylpiperidine (5) is produ- 12Ϫ16%), but quinoneimine-N-oxides 10 (yield 12Ϫ21%)
ced in Ն98% yield, both in air and under N2, and in the ab- and 11 (yield 18Ϫ19%) are also formed. Furthermore, signifi-
sence or presence of a hydrocarbon (adamantane). Kinetic
experiments show that aminoxyl 2 triggers the radical de-
composition of the dioxirane, in addition to scavenging
cant amounts (8Ϫ14%) of the amine 9 (the product of deoxy-
genation of 4) can be isolated. These observations provide
useful information concerning the free-radical mechanism
methyl radicals derived therefrom. The reactions of an ami- that follows the initial attack by the aminoxyl at the dioxi-
noxyl less prone to oxidation, namely 1,2-dihydro-2-methyl-
rane.
Introduction
Dioxiranes[1] have become well established as remarkable
oxidants, capable of highly selective oxidations, including
epoxidations and electrophilic O-insertion into non-acti-
vated alkane CϪH bonds[1]. The latter transformation[1][2e]
represents the highlight of dioxirane chemistry, since it al-
lows the selective oxyfunctionalization of cyclic, polycyclic,
and open-chain hydrocarbons[2e] with an efficiency match-
ing that of cytochrome P-450 enzymes[3]. Because of its rel-
evance, the mechanism of alkane oxidation by dioxiranes
has recently become a matter of much debate[4][5][6][7], and
thus reagents and conditions that may trigger the radical
decomposition of dioxiranes (hence their radical reactiv-
ity)[5][8][9] have come under close scrutiny. In this context,
Minisci et al. reported on an attempt to use the aminoxyl
TEMPO (2) as a trapping agent for radicals purportedly
formed during the oxidation of alkanes with dimethyldioxi-
rane (1a, DMD)[4b]. They also suggested that TEMPO
might induce the homolytic decomposition of the dioxirane.
Indeed, in a previous communication it was reported by
Adam, Bottle, and Mello[10] that photolysis of methyl(tri-
fluoromethyl)dioxirane (1b, TFD) in the presence of
2,2,5,5-tetramethylisoindolin-2-oxyl (3, TMIO) produces
methyl and trifluoromethyl radicals, which are trapped by
the radical scavenger.
Results and Discussion
Isolated dioxiranes generated in solution were employed
as reagents in reactions with aminoxyls 2 and 4. Solutions
of 0.06Ϫ0.10  dimethyldioxirane (1a)[2a] and 0.8Ϫ1.0 
methyl(trifluoromethyl)dioxirane (1b)[2d] in the parent ke-
tones were prepared using procedures and equipment, and
observing precautions, described in detail in previous re-
ports.
Reaction of TEMPO with Dimethyldioxirane
We found that either with the solution exposed to air or
in solvent purged with N2, dimethyldioxirane (1a) un-
We now report our observations concerning the reactions dergoes a rapid reaction with the aminoxyl TEMPO (2) in
of dimethyldioxirane (1a) and its trifluoro analogue 1b with acetone at 20°C, even in the absence of an alkane substrate.
two representative aminoxyls having varied characteristics, In fact, under conditions analogous to those adopted by
namely TEMPO (2) and 1,2-dihydro-2-methyl-2-phenyl- Minisci et al.[4b] (except that adamantane was omitted), the
3H-indol-3-one-1-oxyl (4, MPIO). We found that these rad- reaction affords the corresponding methoxyamine 5[11] in
ical species react readily with dioxiranes and promote their high yield, accompanied by only trace amounts of the ace-
radical decomposition.
tonoxy derivative 6[12] (Eq. 1).
Eur. J. Org. Chem. 1998, 871Ϫ876
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0505Ϫ0871 $ 17.50ϩ.50/0
871

A. Dinoi, R. Curci, P. Carloni, E. Damiani, P. Stipa, L. Greci
FULL PAPER
if alkane oxidation were to proceed via the formation of
radical pairs and out-of-cage diffusion[4b] occurs (Chart 1).
Furthermore, the data in Table 1 illustrate that, both in
air and under an inert atmosphere (nitrogen), the outcome
of the reaction is unaffected by the presence of adamantane;
indeed, the latter remains essentially unchanged. In all
cases, the yield of methoxyamine 5 is practically quantita-
The identities of products 5 and 6 were confirmed by tive, which is at variance with a report of a lower yield
comparison (GC/MS, 1H NMR) with authentic samples (ca. 60%)[4b]
and/or with literature data[11][12]. Products derived from de-
.
It is worthy of note that, both in air and under an inert
composition of dioxirane 1a, i.e. methyl acetate[8a] (N2) atmosphere, dioxirane decomposition induced by
CH3C(:O)OCH3 and acetyloxyacetone[9] CH3C(:O)OϪ TEMPO becomes much faster than non-radical hydro-
CH2C(:O)CH3 (Յ 1%) were also detected. Representative carbon hydroxylation[2e][5a]. For instance, under similar
data and reaction conditions are collected in Table 1 and conditions, a second-order constant k2 ϭ 0.31·10Ϫ2 Ϫ1sϪ1
Figure 1.
and an initial rate Ro ϭ 0.8·10Ϫ5  sϪ1 were estimated[5a]
Table 1. Reaction of TEMPO (2) with dimethyldioxirane (1a) in acetone at 20°C, in the absence and presence of adamantane (AdH)
Entry#
102•[DMD]o
102•[TEMPO]o
102•[AdH]o
Atmosphere
Reactn. time
(min)
TEMPO conv.
(%)[a]
105•Ro
(M sϪ1 [b]
(M)
(M)
(M)
)
1
2
3
4
5
6
7
2.93
2.93
2.80
2.80
2.84
2.90
2.90
0.55
0.55
2.94
2.93
14.0
2.80
2.80
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
2.80
2.80
air
N2
air
N2
air
air
N2
240
120
50
98
2.66
7.45
4.55
16.5
13.6
4.71
16.5
98
70
20
78
20
18
50
63[c]
82[c]
20
[a]
In all cases, the yield of methoxyamine 5 was Ն98%, as based on the amount of TEMPO consumed; conversions were determined
(±2%) by GC or GC/MS. Ϫ [b] Initial rates, estimated from the slopes of plots of dioxirane concentration vs. time over the initial 10Ϫ20%
[c]
of the reaction; the data reported are averages (±5%) from duplicate runs. Ϫ
Trace amounts of 1-adamantanol were also detected
(cf. refs.[4][5a]).
Figure 1. Decrease of dioxirane 1a concentration (cD) with time
from ca 0.03  initial concentration in acetone at 20.0°C: A: in air,
in the absence of TEMPO; B: solvent purged with dry N2, in the
absence of TEMPO; C: in air, in the presence of 0.005  TEMPO;
D: solvent purged with dry N2, in the presence of 0.005  TEMPO;
E: in air, in the presence of equimolar TEMPO; F: solvent purged
with dry N2, in the presence of equimolar TEMPO
Chart 1
for the hydroxylation of adamantane (AdH) by dioxirane
1a in air, in contrast to the much faster rates of entries 1,
3, and 5 (Table 1). The data in Table 1 and the kinetics of
dioxirane consumption with time depicted in Figure 1 indi-
cate that TEMPO, even at a molar ratio ca. 1:6 with respect
to the dioxirane, is quite effective in causing peroxide de-
composition. On the basis of Ro values collected for the
reaction in air (entries 1, 3, and 5), from a plot of log Ro
vs. log/[TEMPO]o, a fractional kinetic order of about 0.5
for the aminoxyl can be estimated.
When reactions of TEMPO with the dioxirane were per-
A further observation is that, in agreement with previous
formed in the presence of an equimolar amount of ada- literature[1][8a], i.e. similar to other radical species such as
mantane (e.g. entries 6 and 7), comparison with GC and O2•Ϫ[8a], the aminoxyls themselves induce the radical de-
GC/MS data of authentic adamanthoxyamine 7[13] allowed composition of the peroxide, in addition to trapping radical
us to establish that the latter is not formed in any detectable species derived therefrom. Consistent with this view, both
amount. However, this trapping product would be expected in air and under N2, the rates are found to depend markedly
872
Eur. J. Org. Chem. 1998, 871Ϫ876

Reaction of Aminoxyls with Dioxiranes
FULL PAPER
Reaction of MPIO with Dimethyldioxirane and
Methyl(trifluoromethyl)dioxirane
upon the aminoxyl concentration, as shown in Table 1 (en-
tries 1Ϫ5) and Figure 1.
It is of interest to compare the reactivity of dioxiranes
towards TEMPO (2) with that towards 1,2-dihydro-2-
methyl-2-phenyl-3H-indol-3-one-1-oxyl (4). In fact, the for-
mer should be more prone to oxidation, considering the
different anodic potentials [Epa ϭ ϩ0.62 and ϩ1.10 V vs.
Scheme 1
SCE (in DMF/water, 6:4) for TEMPO[18][19] and MPIO[19]
,
respectively]. This difference is sufficient to make other re-
action pathways available. The outcome of the reactions of
aminoxyl 4 (MPIO) with dioxiranes 1a and 1b in acetone is
shown in Eq. 2.
As outlined in Scheme 1, by analogy with the reaction of
TEMPO with acyl peroxides[14], one might envisage an
SH2-type attack by the aminoxyl at the dioxirane OϪO
bond, yielding the very labile covalent adduct I. Concurrent
with, or alternatively to this, an outer-sphere electron-trans-
fer might give rise to the caged pair II containing the
oxoammonium cation[15]; the fact that the characteristic in-
tense red color[14a][15] of the oxoammonium species is not
observed could be due to fast in-cage back electron transfer,
thereby leading to generation of the crucial bis(oxy)methyl-
ene biradical 1Ј (Scheme 1). It is worthy of note that, at
variance with findings concerning the reaction of acyl per-
oxides with TEMPO[14], in our case no products derived
from oxidative degradation[14a] of the nitroxide were de-
tected. Then, due to β-scission of the bis(oxyl) diradical 1Ј
to yield methyl radicals, radical-chain dioxirane decompo-
sition would ensue, as shown in Scheme 1. The inhibition
of radical decomposition of dioxirane by dissolved oxygen
gas has been reported[4][5a], and is well understood[8a] in
terms of the trapping of diffusing methyl radicals. On the
other hand, formation of the methoxyamine by the alterna-
Representative product distributions are presented in
Table 2. Compounds 8a[20], 9[21], 10[20], and 11[22] could be
isolated by semi-preparative TLC; they were identified by
their spectral characteristics and/or by comparison with
authentic samples.
In accord with the aforementioned higher oxidation po-
tential of MPIO compared to that of TEMPO, we found
that a moderate excess of DMD (1a) was necessary in order
to achieve high aminoxyl conversion; this might be ascribed
to dioxirane chain decomposition competing with methyl
radical trapping by the aminoxyl 4 (cf. Scheme 1). However,
the formation of indoxyl 9 (the deoxygenation product of
4) and of quinoneimine N-oxides 10 and 11 suggests that
other reaction pathways also become viable with this ami-
noxyl. The latter products are clearly derived from oxi-
dation of the aromatic ring in aminoxyl 4, for which several
pathways might be envisaged. One possibility involves the
process depicted in Scheme 2.
•
tive trapping of the reactive CH3 by TEMPO should occur
The dioxirane radical anion/oxoammoniun cation pair
might arise by an inner-sphere ET, following the formation
of the initial labile adduct I (Scheme 1)[23]. Then, oxy-
functionalization at the aromatic ring of 4 might occur larg-
at a rate of ca. 108  sϪ1 [16]
.
Reaction of TEMPO with Methyl(trifluoromethyl)dioxirane
The reaction of TEMPO with the trifluoromethyl an- ely by an in-cage process, for instance as depicted in Scheme
alogue (1b) gave similar results as those described above, 2 (R ϭ CH3)[24]. Here, labile intermediates such as 12 and
but proceeded at a remarkably faster rate. Typically, with 13 could be formed following attack at the aromatic ring;
initial TFD (1b) and TEMPO concentrations both of ca. indeed, it is well documented[25] that oxoammonium ions
0.1  in acetone at 0°C, 55% conversion of the aminoxyl (arising from oxidation of aromatic nitroxides such as
was achieved within just 20 min. and methoxyamine 5 was MPIO) sustain nucleophilic attack at conjugated positions
obtained in high yield (Ն96% by GC). In accordance with of their benzene ring[26]. Then, in line with the general reac-
the non-photolytic (i.e. thermal) reactivity of dioxirane 1b tivity of dioxiranes towards aminoxyls outlined in Scheme
•
towards TMIO[10], formation of the CF3 trapping product 1, in the reactions of MPIO (4) with dioxirane 1a and 1b,
(i.e. the trifluoro analogue of 5) was not detected to any formation of the trapping product methoxyamine 8a is still
appreciable amount; this might be due to the much faster prominent (cf. Eq. 2 and Table 2).
•
•
generation of CH3 compared to CF3 following β-cleavage
As in the case of TEMPO (see above), the corresponding
labile trifluoromethoxyamine (from CF3 trapping) could
•
of bis(oxy)methylene biradical 1Ј[17]
Eur. J. Org. Chem. 1998, 871Ϫ876
.
873

A. Dinoi, R. Curci, P. Carloni, E. Damiani, P. Stipa, L. Greci
Table 2. Product distribution from the reaction of indolinoneoxyl (4) with dioxiranes in acetone at 0°C
FULL PAPER
Products (% yield) [c]
[a]
Dioxirane
(D/A) ratio
Reactn. time (h) Conv.[b] (%)
alkoxyamine
amine
quinoneimine N-oxide
DMD (1a)
TFD (1b)
3
1
3
0.5
80
75
8a (16)
8a (12)
9 (8.5)
9 (14)
10 (12), 11 (18)
10 (21), 11 (19)[d]
[a]
Ratio of initial dioxirane and aminoxyl concentrations; initial concentrations of dioxiranes 1a and 1b were in the ranges 0.02Ϫ0.04
[b]
and 0.20Ϫ0.40 , respectively. Ϫ
Degree of aminoxyl conversion, as determined from substrate recovery (semi-preparative TLC). Ϫ
[c]
[d]
Yield of product isolated (±3%), based on the amount of aminoxyl reacted. Ϫ Other products also detected (see text).
Scheme 2
Figure 2. Experimental (A) and simulated (B) EPR spectrum of 5-
(trifluoromethyl)-1,2-dihydro-2-methyl-2-phenyl-3H-indol-3-one-1-
oxyl (14) (CHCl3): aN 8.85 (1 N), aF 3.93 (3 F), aH-7 2.83 (1 H),
aH-4,6 1.00 (2 H), aH(CH3) 0.19 G (3 H)
not be detected. However, in one case, a trace amount of a
by-product was isolated, the EPR spectrum of which was
consistent with that simulated for the 5-trifluoromethyl
trapping product 14 (Figure 2); it exhibited many similarit-
ies [e.g. aF ca. 3.9 (3 F) G] with EPR spectral data recorded
for analogous 2,2-diphenylindolinone-1-oxyls in which a
CF3 substituent is attached at the aromatic nucleus[22]
Further work is in order to confirm this parallel.
.
One conspicuous difference between the reactivity
towards dioxiranes of TEMPO (or TMIO) and that of
MPIO is that the latter aminoxyl yields significant amounts
of the corresponding indoxyl 9 (Eq. 2, Table 2). This prod-
uct is derived from deoxygenation of the parent aminoxyl
4. By analogy to the reported[27][28] deoxygenation of elec-
tron-rich N-oxides by dioxiranes, this finding can most sim-
ply be explained by invoking fragmentation of the initially
formed labile adduct IЈ (cf. Scheme 1) according to the pat-
tern outlined in Eq. 3.
might present several facets and mechanistic features that
are arduous to unravel. However, besides the mechanistic
details, the results reported herein show conclusively that
aminoxyls readily undergo direct reaction with dioxiranes
and trigger the radical decomposition of these powerful oxi-
dants. Thus, it can be concluded that aminoxyls are not
suitable traps for radical intermediates formed in the course
of DMD oxidations.
The reason for this pathway becoming practicable in the
reaction of MPIO with dioxiranes might have its origin in
the extra resonance stabilization available to the aminyl rad-
ical intermediate 9Ј[29]
.
Taken as a whole, the aforementioned observations indi-
cate that the reactivity of dioxiranes toward aminoxyls
874
Eur. J. Org. Chem. 1998, 871Ϫ876

Reaction of Aminoxyls with Dioxiranes
FULL PAPER
We thank the Ministry of University, Scientific, and Technological
Research of Italy (MURST 40) and the CNR - Progetto Strategico
“Tecnologie Chimiche Innovative” (Rome, Italy) for partial support.
Kinetic Measurements: Runs were performed by following the
decrease of dioxirane concentration (by iodometry) with time, ac-
cording to previously described analytical techniques[2e][5]. The fol-
Work in Ancona was also supported through a contract by the lowing is representative of the simple procedure adopted in these
Italian Research Council (CNR, Rome).
experiments: at zero time an aliquot of a thermostatted (20.00 ±
0.05°C) solution of TEMPO (2) (0.15 mmol) in acetone (10 ml) was
added in one portion to a 0.08  dimethyldioxirane (1a) solution in
acetone (16 ml, 0.64 mmol) (also thermostatted). Aliquots (0.2 ml)
of the reaction solution were then withdrawn at timed intervals,
quenched with excess KI/EtOH, and the liberated I2 was deter-
mined by iodometry. Runs were performed under various con-
ditions, keeping the ratio of initial dioxirane 1a and TEMPO (2)
concentrations in the range 0.1Ϫ5. Initial rates Ro ( sϪ1) were
Experimental Section
Instrumentation: M.p.: Electrothermal melting point apparatus.
Melting points are not corrected. Ϫ FT-IR: Nicolet 20SX. Ϫ 1H
NMR: Varian Gemini 200 or Varian XL 200 (200 MHz). Spectra
were referenced to residual isotopic impurity CHCl3 (δ ϭ 7.24) of
the solvent CDCl3 and/or to TMS. Ϫ MS: Carlo Erba QMD 1000
GLC/MS spectrometer with a direct probe apparatus (EI, 70 eV). estimated from the slopes of plots of dioxirane concentration vs.
Ϫ EPR: Varian E4 spectrometer. Hyperfine coupling constants are
given in Gauss. Ϫ GC: Perkin-Elmer 8420 capillary gas chromatog-
time, based on the initial 10Ϫ20% of the reaction. In each case, at
least two independent runs were performed and the Ro values aver-
raph with FID and VOCOL fused silica capillary column (3.0 mm aged (estimated error Յ ±5%) (Table 1).
film thickness, 60 m ϫ 0.53 mm ID). Ϫ GC/MS: Hewlett-Packard
Reaction of MPIO (4) with Dimethyldioxirane (1a): To a stirred
model 5890 gas chromatograph with Hewlett-Packard model 5970
mass selective detector (EI, 70 eV) and cross-linked methylsilicone
capillary column (HP1, 0.33 mm film thickness, 25 m ϫ 0.22 mm
ID).
solution of MPIO (4) (190 mg, 0.8 mmol) in CH2Cl2 (5 ml) kept
at 0°C, a standardized solution of dioxirane 1a (3 ml, 0.08 , 0.240
mmol) in acetone was added in three portions. According to TLC
analysis, after 3 h total substrate conversion had not been achieved.
Removal of the solvent in vacuo and semi-preparative TLC (cyclo-
hexane/ethyl acetate, 8:2) resulted in partial recovery of the starting
material (37 mg, 0.156 mmol) and isolation of the reaction prod-
Materials: High purity commercial solvents (Fluka) dichloro-
methane, acetone, cyclohexane, ethyl acetate, and 1,1,1-trifluoro-
2-propanone (TFP) (b.p. 22°C) were further purified by standard
˚
methods, stored over 5-A molecular sieves at 2Ϫ5°C, and routinely
ucts (individual yields are reported in Table 2). Compounds 8a[20]
,
redistilled prior to use. For semi-preparative TLC, silica gel plates
9
[21], 10[20], and 11[22] were identified by comparison of their spec-
˚
(0.25 mm layer thickness, 60 A medium pore diameter) (Fluka)
tral characteristics with those of authentic samples obtained inde-
pendently and/or literature data. Very small amounts of the follow-
ing labile intermediates (see Results and Discussion) could be de-
tected and quickly isolated by TLC; the labile nature of radical
intermediates 12a and 13a prevented their further purification by
crystallization or by common chromatographic techniques:
were employed. High purity commercial 2,2,6,6-tetramethylpiperi-
dinyl-1-oxyl (2, TEMPO) (Fluka) was further purified by subli-
mation (m.p. 38Ϫ39°C). The synthesis and spectral characteri-
stics of 1,2-dihydro-2-methyl-2-phenyl-3H-indol-3-one-1-oxyl (4,
MPIO) (m.p. 167°C) have been reported elsewhere[30]. Previously
described protocols were employed in order to obtain dioxiranes
1a and 1b as solutions in the parent ketones[1][2]. We used the “Cu-
rox” triple salt 2KHSO5 ·KHSO4 ·K2SO4 (a gift from Peroxid-
Chemie GmbH, Munich, Germany) as a source of potassium per-
oxymonosulfate, which was employed in the synthesis of the dioxi-
ranes.
5-(2-Hydroxy-2-propyloxy)-1,2-dihydro-2-methyl-2-phenyl-3H-
indol-3-one-1-oxyl (12a): IR (KBr): ν˜ ϭ 3358, 1719, 1607 cmϪ1. Ϫ
MS (70 eV); m/z (%): 311 (5) [Mϩ Ϫ 1], 281 (7), 222 (58), 77 (100).
Ϫ EPR (C6H6): aN 9.18 (1 N), aH-7 2.83 (1 H), aH-4,6 1.00 G (2
H)[31]
.
7-(2-Hydroxy-2-propyloxy)-1,2-dihydro-2-methyl-2-phenyl-3H-
indol-3-one-1-oxyl (13a): IR (KBr): ν˜ ϭ 3414, 1721, 1606 cmϪ1. Ϫ
MS (70 eV); m/z (%): 311 (6) [Mϩ Ϫ 1], 238 (100), 222 (18), 77
(89). Ϫ EPR (C6H6): aN 9.03 (1 N), aH-5 3.20 (1 H), aH-4,6 1.00 (2
Reaction of TEMPO (2) with Dimethyldioxirane (1a): To a stirred
solution of TEMPO (2) (0.060Ϫ0.32 ) in acetone (10 ml) kept at
20°C, a standardized solution of dioxirane 1a (16 ml, 0.08 , 0.64
mmol) in acetone was added in one portion. The reaction mixture
was monitored by iodometry until the peroxide had been com-
pletely consumed. The reaction mixture was then analyzed by GC
and/or GC/MS in order to determine the substrate conversion and
product yields (Table 1). After addition of a small amount of 1,2-
diphenylhydrazine (in order to quench any unreacted TEMPO by
converting it into the corresponding hydroxylamine), and removal
of the solvent in vacuo, the identities of the products were con-
H), aH (CH3) 0.18 G (3 H)[31]
.
Reaction of MPIO (4) with Methyl(trifluoromethyl)dioxirane
(1b): To a stirred solution of MPIO (4) (190 mg, 0.8 mmol) in
CH2Cl2 (5 ml) kept at 0°C, a standardized solution of dioxirane 1b
(1 ml, 0.8 , 0.8 mmol) in 1,1,1-trifluoropropanone was added in
one portion. After 0.5 h, TLC analysis revealed that total conver-
sion of the substrate had not been achieved. According to the pro-
cedure described above, some starting material (45 mg, 0.19 mmol)
was recovered and compounds 8a, 9, 10, and 11 were isolated in
the yields shown in Table 2. A further reaction product was isolated
as a deliquescent solid in trace amounts (< 5 mg), which gave: EPR
(CHCl3): aN 8.85 (1 N), aF 3.93 (3 F), aH-7 2.83 (1 H), aH-4,6 1.00
(2 H), aH (CH3) 0.19 G (3 H). This intermediate was tentatively
identified as 5-(trifluoromethyl)-1,2-dihydro-2-methyl-2-phenyl-
3H-indol-3-one-1-oxyl (14) by comparison with simulated EPR
spectra (Figure 2).
1
firmed by H-NMR analysis.
1-Methoxy-2,2,6,6-tetramethylpiperidine (5)[11]
: b.p. 60°C (8
1
Torr). Ϫ H NMR (CDCl3): δ ϭ 3.60 (s, 3 H), 1.44 (m, 6 H), 1.15
(s, 6 H), 1.09 (s, 6 H). Ϫ MS (70 eV); m/z (%): 171 (6) [Mϩ], 157
(10), 156 (100) [Mϩ Ϫ CH3], 125 (5), 109 (20), 100 (12), 97 (10),
88 (36), 87 (10), 83 (11), 69 (44), 58 (14), 56 (44), 55 (57), 43 (16),
42 (56), 41 (93), 40 (10), 39 (36).
1-Acetonoxy-2,2,6,6-tetramethylpiperidine (6)[12]: Oil; 1H NMR
(CDCl3): δ ϭ 4.40 (s, 2 H), 2.17 (s, 3 H). Ϫ MS (70 eV); m/z (%):
213 (7) [Mϩ], 198 (52) [Mϩ Ϫ CH3], 156 (29) [Mϩ Ϫ 57], 126 (19),
123 (12), 109 (10), 83 (42), 82 (10), 81 (10), 70 (18), 69 (50), 58 (33),
56 (64), 55 (100), 43 (54), 42 (35), 41 (59), 39 (16).
Ƞ
Dedicated to Prof. Fulvio Di Furia (University of Padova,
Italy), deceased October 1997. Presented in part at the 7th In-
ternational Symposium on Organic Free Radicals, Bardolino,
Italy, June 16Ϫ21, 1996; Abstracts, 83.
Eur. J. Org. Chem. 1998, 871Ϫ876
875

A. Dinoi, R. Curci, P. Carloni, E. Damiani, P. Stipa, L. Greci
FULL PAPER
[1]
[1a]
For recent reviews, see:
Pure Appl. Chem. 1995, 67, 811Ϫ822. Ϫ
Hadjiarapoglou, R. Curci, R. Mello in Organic Peroxides (Ed.:
W. Ando), Wiley, New York, 1992, chapter 4, p. 195. See also
references therein.
R. Curci, A. Dinoi, M. F. Rubino, [17] [17a] R. D. Dolbier, Chem. Rev. 1996, 96, 1557Ϫ1584. Ϫ [17b] X.-
[1b]
W. Adam, L. P.
K. Jiang, X.-Y. Li, K.-Y. Wang, J. Org. Chem. 1989, 54,
5648Ϫ5650.
[18]
[19]
L. B. Volodarsky, in Imidazoline Nitroxides (Ed.: CRC Press
Inc.), Boca Raton, Florida, 1988, vol 1, p. 175.
P. Carloni, E. Damiani, L. Greci, P. Stipa, G. Marrosu, R. Pe-
trucci, A. Trazza, Tetrahedron 1996, 52, 11257Ϫ11264.
[2] [2a]
R. W. Murray, R. Jeyaraman, J. Org. Chem. 1985, 50,
2847Ϫ2853. Ϫ [2b] L. Cassidei, M. Fiorentino, R. Mello, O. Sci-
[20] [20a] L. Greci, Tetrahedron 1982, 38, 2435Ϫ2439. Ϫ [20b] R. And-
ruzzi, C. Berti, L. Greci, A. Trazza, J. Chem. Res. 1982,
1840Ϫ1863.
[2c]
acovelli, R. Curci, J. Org. Chem. 1987, 52, 699Ϫ700. Ϫ
W.
Adam, Y.-Y Chan, D. Cremer, J. Gauss, D. Scheutzow, M.
[2d]
Schindler, J. Org. Chem. 1987, 52, 2800Ϫ2803. Ϫ
R. Mello,
[21]
C. Berti, L. Greci, L. Marchetti, J. Chem. Soc., Perkin Trans. 2
M. Fiorentino, O. Sciacovelli, R. Curci, J. Org. Chem. 1988, 53,
3890Ϫ3891. Ϫ [2e] R. Mello, M. Fiorentino, C. Fusco, R. Curci,
J. Am. Chem. Soc. 1989, 111, 6749Ϫ6757.
1979, 233Ϫ236.
[22]
C. Berti, M. Colonna, L. Greci, L. Marchetti, Tetrahedron 1977,
[3]
For some recent references, see e.g.: [3a] S.-Y. Choi, P. E. Eaton,
P. F. Hollenberg, K. E. Liu, S. J. Lippard, M. Newcomb, D. A.
Putt, S. P. Upadhyaya, Y. Xiong, J. Am. Chem. Soc. 1996, 118,
6547Ϫ6555. Ϫ [3b] Cytochrome P-450. Structure, Mechanism and
Biochemistry, 2nd ed. (Ed.: P. R. Ortiz de Montellano), Plenum,
New York, 1995. See references therein.
33, 2321Ϫ2327.
[23]
It should be noted that in this case the intense red color of the
aminoxyl 4 starting material would preclude direct observation
of the rise of the red-colored oxoammonium species.
Based on the available literature (refs.[8b][8c]), dioxirane radical
[24]
anions freely diffusing out of the cage would be expected
to react readily with the dioxirane, yielding the dimer
[4a] F. Minisci, L. Zhao, F. Fontana, A. Bravo, Tetrahedron Lett.
[4]
[5]
ϪOϪCR(CH3)ϪOϪOϪCR(CH3)ϪO•, and hence R(CH3)Cϭ
[4b]
1995, 36, 1697Ϫ1700. Ϫ
A. Bravo, F. Fontana, G. Fronza,
Ϫ•
O and superoxide O2
.
A. Mele, F. Minisci, J. Chem. Soc., Chem. Comm. 1995,
1573Ϫ1574.
[25] [25a]
C. Berti, L. Greci, J. Org. Chem. 1981, 46, 3060Ϫ3063. Ϫ
[25b]
[5a] R. Curci, A. Dinoi, C. Fusco, M. A. Lillo, Tetrahedron Lett.
1996, 37, 249Ϫ252. Ϫ [5b] W. Adam, R. Curci, L. DЈAccolti, A.
Dinoi, C. Fusco, F. Gasparrini, R. Kluge, R. Paredes, M.
Schulz, A. K. Smerz, L. A. Veloza, S. Weinkötz, R. Winde,
Chem. Eur. J. 1997, 3, 105Ϫ109.
A. Alberti, R. Andruzzi, L. Greci, P. Stipa, G. Marrosu,
A. Trazza, M. Poloni Tetrahedron 1988, 44, 1503Ϫ1510.
[26]
Actually, by carrying out the reaction of DMD with 4 at inter-
mediate conversions (30Ϫ50%), two new aminoxyls could be
isolated in tiny amounts (Յ 3%) by TLC, one having a greenish-
and the other a reddish-brown color in solution; this is typical
of indolinone-1-oxyls substituted with electron-donor groups at
C-5 and at C-7, respectively[25]. The IR spectra of both com-
pounds displayed an OϪH absorption near 3400 cmϪ1 besides
the expected carbonyl stretching (ca. 1720 cmϪ1), and a band at
[6]
R. Vanni, S. J. Garden, J. T. Banks, K. U. Ingold, Tetrahedron
Lett. 1995, 36, 7999Ϫ8002.
[7] [7a]
Y. Angelis, X. Zhang, M. Orfanopulos, Tetrahedron Lett.
[7b]
1996, 37, 5991Ϫ5994. Ϫ
G. Asensio, R. Mello, M. E. Gon-
`
zales-Nun˜ez, C. Boix, J. Royo, Tetrahedron Lett. 1997, 38,
ca. 1600 cmϪ1, which is typical of the indoline framework[20][21]
.
2373Ϫ2376.
Based on their MS m/z ϭ 311 [Mϩ Ϫ 1] peaks and fragmen-
tation patterns (see Experimental Section), we tentatively assign
to these intermediates the isomeric structures of adducts 12a
and 13a (Scheme 2, 12 and 13: R ϭ CH3). The EPR spectra
are consistent with these assignments; in fact, besides yielding
different aN (1 N) and similar aH-4,6 (2 H) values, one com-
pound gives aH-7 ϭ 2.83 G and the other aH-5 ϭ 3.20 G, which
is in agreement with isomeric 12a and 13a, respectively (see Ex-
perimental Section; cf. also C. Berti, M. Colonna, L. Greci,
L. Marchetti, Tetrahedron 1977, 33, 3149Ϫ3154). It should be
mentioned that, in the reaction of 4 with TFD (1b), the anal-
ogous adducts 12b and/or 13b (Scheme 2, 12 and 13: R ϭ CF3)
could not be detected; this might be ascribed to their faster
consecutive conversion into quinoneimine N-oxides by TFD, an
oxidant which is more powerful than DMD.
[8] [8a]
`
W. Adam, R. Curci, M. E. Gonzales-Nun˜ez, R. Mello, J.
[8b]
Am. Chem. Soc. 1991, 113, 7654Ϫ7658. Ϫ
W. Adam, G.
`
Asensio, R. Curci, M. E. Gonzalez-Nun˜ez, R. Mello, J. Am.
Chem. Soc. 1992, 114, 8345Ϫ8349. Ϫ [8c] W. Adam, G. Asensio,
`
R. Curci, A. Marco, M. E. Gonzales-Nun˜ez, R. Mello, Tetra-
hedron Lett. 1992, 33, 5833Ϫ5836.
[9]
M. Singh, R. W. Murray, J. Org. Chem. 1992, 57, 4263Ϫ4270.
W. Adam, S. E. Bottle, R. Mello, J. Chem. Soc., Chem. Comm.
1991, 771Ϫ773.
[10]
[11]
G. W. Whitesides, T. L. Newirth, J. Org. Chem. 1975, 40,
3448Ϫ3450.
[12] [12a] D. H. Hunter, D. H. Barton, W. J. Motherwell, Tetrahedron
[12b]
Lett. 1984, 25, 603Ϫ606. Ϫ
V. A. Golubev, R. I. Zhdanov,
I. T. Protsishin, E. G. Rozantsev, Dokl. Phys. Chem. 1970,
[27]
[28]
195, 952Ϫ955.
R. W. Murray, Chem. Rev. 1989, 89, 1187Ϫ1201.
[13]
H. A. Togo, M. Aoki, T. Kuramochi, M. Yokoyama, J. Chem.
Soc., Perkin Trans. 1 1993, 2417Ϫ2427.
W. Adam, K. Brivida, F. Duschek, D. Golsch, W. Kiefer, M.
Sies, J. Chem. Soc., Chem. Comm. 1995, 1831Ϫ1832.
It is worthy of mention that interception of the aminyl radical
9Ј by oxygen would be expected to give back the initial amin-
oxyl 4: L. Greci, A. MarЈin, P. Stipa, P. Carloni, Polym. Deg.
Stab. 1995, 50, 305Ϫ312.
[14] [14a]
[29]
. Moad, E. Rizzardo, D. H. Solomon, Tetrahedron Lett.
[14b]
1981, 22, 1165Ϫ1168. Ϫ
V. W. Bowry, J. Lusztyk, K. U.
Ingold, J. Am. Chem. Soc. 1989, 111, 1927Ϫ1928.
[15] [15a]
J. A. Cella, J. A. Kelley, E. F. Kenehan, Tetrahedron Lett.
[15b]
[30]
[31]
1975, 2869Ϫ2872. Ϫ
1988, 27, 509Ϫ533. Ϫ
J. Bobbitt, M. C. Flores, Heterocycles
C. Berti, M. Colonna, L. Greci, L. Marchetti, Tetrahedron 1975,
31, 1745Ϫ1753.
[15c]
P. L. Anelli, F. Montanari, S. Quici,
Org. Synth. 1990, 69, 212Ϫ219. See also references therein.
J. Chateauneuf, J. Lusztyk, K. U. Ingold, J. Org. Chem. 1988,
53, 1629Ϫ1632.
The original EPR spectrum is available on request by e-mail to:
<Ruggero Curci> csmirc17@area.ba.cnr.it.
[16]
[97333]
876
Eur. J. Org. Chem. 1998, 871Ϫ876