Homolytic Reactivity of Group 14 Organometallic Hydrides toward Nitroxides

Article abstract of DOI:10.1021/jo972178i

The spontaneous reactions of Et₃SiH, Ph₃GeH, Bu₃SnH, and (Me₃Si)₃SiH with 2,2,6,6-tetramethyl1-piperidinoxyl (TEMPO) and related nitroxides have been investigated. Metal hydrides characterized by relatively weak metal-hydrogen bonds, like Ph₃GeH and Bu₃SnH, reduce TEMPO to the corresponding hydroxylamine while no reaction is observed with Et₃SiH and Ph₃SiH. Tris(trimethylsilyDsilane, on the other hand, is able to reduce the nitroxide to the corresponding hydroxylamine and amine in a ca. 1:1 ratio. By repeating the above reactions in the presence of thermal or photochemical radical initiators, deoxygenation of TEMPO was obtained in high yield with (Me₃SD)₃SiH and (Me₃Si)₂Si(H)Me, but not with other silanes, germanes, and stannanes. A mechanism for the reduction of TEMPO by the two trimethylsilyl-substituted silanes is proposed, and kinetic data for the various steps of the overall reaction are reported. In particular rate constants and activation parameters have been measured for the hydrogen transfer reaction from several silanes to the hindered aminyl radical 2,2,6,6-tetramethylpiperidinyl.

Full text of DOI:10.1021/jo972178i

J . Org. Chem. 1998, 63, 1687-1693
1687
Hom olytic Rea ctivity of Gr ou p 14 Or ga n om eta llic Hyd r id es
tow a r d Nitr oxid es
Marco Lucarini,* Emanuela Marchesi, and Gian Franco Pedulli*
Dipartimento di Chimica Organica “A. Mangini” Universit a` di Bologna Via S. Donato 15,
I-40127 Bologna, Italy
Chryssostomos Chatgilialoglu
I.Co.C.E.A. Consiglio Nazionale delle Ricerche Via P. Gobetti 101, I-40129 Bologna, Italy
Received December 2, 1997
The spontaneous reactions of Et
-piperidinoxyl (TEMPO) and related nitroxides have been investigated. Metal hydrides character-
ized by relatively weak metal-hydrogen bonds, like Ph GeH and Bu SnH, reduce TEMPO to the
corresponding hydroxylamine while no reaction is observed with Et SiH and Ph SiH. Tris(tri-
3 3 3 3 3
SiH, Ph GeH, Bu SnH, and (Me Si) SiH with 2,2,6,6-tetramethyl-
1
3
3
3
3
methylsilyl)silane, on the other hand, is able to reduce the nitroxide to the corresponding
hydroxylamine and amine in a ca. 1:1 ratio. By repeating the above reactions in the presence of
thermal or photochemical radical initiators, deoxygenation of TEMPO was obtained in high yield
3 3 3 2
with (Me Si) SiH and (Me Si) Si(H)Me, but not with other silanes, germanes, and stannanes. A
mechanism for the reduction of TEMPO by the two trimethylsilyl-substituted silanes is proposed,
and kinetic data for the various steps of the overall reaction are reported. In particular rate
constants and activation parameters have been measured for the hydrogen transfer reaction from
several silanes to the hindered aminyl radical 2,2,6,6-tetramethylpiperidinyl.
In tr od u ction
portance of nitroxides has also increased enormously in
the field of polymer stabilization since it became evident
4
Nitroxides represent the most important class of long-
lived free radicals.1 Strong persistency especially char-
acterizes aliphatic nitroxides, R NO, having a radical
2
that the mechanism of inhibition of polymer photooxi-
dation by hindered amine light stabilizers (HALS) in-
volves oxidation of these compounds to nitroxide radicals.
During polymer exposure to radiation, HALS give rise
to nitroxides which can intercept alkyl radicals from the
polymer to yield aminoethers or hydroxylamines which,
in turn, react with peroxyl radicals to regenerate nitrox-
ides. The high effectiveness of hindered amines as
polymer stabilizers is therefore essentially due to their
ability to give rise to the cyclic regeneration of nitroxides.
It is conceivable that the ability of HALS to protect
polymers from photodegradation could be substantially
increased if regeneration not only of the nitroxide but
also of the amine, i.e., the other active species in the
polymer stabilization process, could be achieved by using
suitable additives.
center sterically hindered by the presence of bulky alkyl
groups. Thus, 2,2,6,6-tetramethyl-1-piperidinoxyl (TEM-
PO) and several related derivatives have lifetimes of
years when kept in the dark at ca. 0 °C. The persistency
of hindered aliphatic nitroxides is essentially due to the
fact that they cannot react with themselves by dispro-
portionation, i.e., the reaction mainly responsible for the
decay of these species. Nitroxides containing hydrogens
â to the nitrogen atom have instead lifetimes of only few
seconds at room temperature because of the ease by
which they undergo disproportionation.
The behavior of nitroxide radicals has been extensively
investigated over the last 40 years in different fields
2
ranging from medicinal chemistry to living free radical
polymerization.3 The latter application is gaining a
In a recent paper it has been shown that regeneration
of diarylamines, Ar
temperatures (>120 °C) by thermal decomposition of the
related aminoethers, Ar NOR, formed during the inhibi-
2
NH, can take place at elevated
growing interest since it has been shown that the
polymerization of styrene-based monomers, mediated by
TEMPO, can be controlled to levels previously obtained
only by using anionic or cationic procedures. The im-
2
5
tion process of the autoxidation of hexadecane. Deoxy-
genation of hindered aliphatic nitroxides by thiophenols
to give the corresponding amines has been also de-
scribed.6
(
1) Rozantsev, E. G. Free Nitroxyl Radicals; Plenum Press: New
York, 1970 and references therein.
2) See for examples: Zhdanov, R. I. Bioactive Spin Label; Springer-
(
Verlag: Heidelberg, 1992 and references therein. Samuni, A.; Krishna,
C. M.; Riez, P.; Finkelstein, E.; Russo, A. Free Radicals Biol. Med. 1989,
(4) Gugumus, F. Polym. Deg. Stab. 1993, 40, 167. Denisov, E. T.
Polym. Deg. Stab. 1991, 34, 325. Gugumus, F. Polym. Deg. Stab. 1989,
24, 289. Allen, N. S. Chem. Soc. Rev. 1986, 15, 373. Sedlar, J .; Marchal,
J .; Petruj, J .; J . Polym. Photochem. 1982, 2, 175. Shilov, Yu B.;
Battalova, R. M.; Denisov, E. T. Doklady Akdademii nauk, SSSR 1972,
207, 288.
6
, 141.
(3) Hawker, C. J .; Barclay, G. G.; Dao, J . J . Am. Chem. Soc. 1996,
1
18, 11467. Hawker, C. J . Trends Polym. Sci. 1996, 4, 183. Moad, G.
Rizzardo, E. Macromolecules 1995, 28, 8722. Kazmaier, P. M.; Moffat,
K. A.; Georges, M. K.; Veregin, R. P.N.; Hamer, G. K. Macromolecules
1
995, 28, 1841. Li, I.; Howell, B. A.; Matyjaszewski, K.; Shigemoto,
(5) J ensen, R. K.; Korcek, S.; Zimbo, M.; Gerlock, J . L. J . Org. Chem.
1995, 60, 5396.
(6) Carloni, P.; Damiani, E.; Iacussi, M.; Greci, L.; Stipa, P.; Cauzi,
D.; Rizzoli, C.; Sgarabotto, P. Tetrahedron 1995, 51, 12445.
T.; Smith, P. B.; Priddy, D. B. Macromolecules 1995, 28, 6692. Georges,
M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer, G. K. Trends Polym.
Sci. 1994, 2, 66.
S0022-3263(97)02178-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/14/1998

1
688 J . Org. Chem., Vol. 63, No. 5, 1998
Lucarini et al.
Ta ble 1. Con ver sion s for th e Sp on ta n eou s Rea ction s of
TEMP O (0.2 M) w ith Or ga n om eta llic Hyd r id es (0.4 M)
a n d Yield s of th e Cor r esp on d in g Hyd r oxyla m in e a n d
Am in e in Ben zen e a t 80 °C
yield of the
hydroxylamine
1 (%)
yield of the
amine
conversion
(%)
hydride
time
3 (%)
Et3SiH
Ph3GeH
Bu3SnH
<3
85
98
67
3 h
3 h
10 m
3 h
-
-
3
3
75.5
96
(TMS)3SiH
41
56
We have recently found that polyhydrosilanes, under
free radical conditions in the absence of oxygen, are
capable of reducing nitroxides to the corresponding
amines.7 Given the importance of this reaction for its
possible application in the above-mentioned field, we
have undertaken an investigation on the reactivity of
hindered aliphatic nitroxides in liquid solution toward
simpler silanes, as well as other Group 14 organometallic
compounds such as tin and germanium hydrides.
-
5
F igu r e 1. Decay of TEMPO (8.2 × 10 M) in the presence of
Bu SnH (0.017 M) at 60 °C. Insert: first-order plot of the
3
experimental data.
Resu lts a n d Discu ssion
Ta ble 2. Ra te Con sta n ts for th e Rea ction of TEMP O
w ith Or ga n om eta llic Hyd r id es
Sp on t a n eou s R ea ct ion s of TE MP O w it h Met a l
Hyd r id es. The room-temperature reaction of Group 14
metal hydrides with persistent nitroxides in oxygen free
solutions afforded different results depending on the
nature of the organometallic derivative employed. Thus,
when using triethylsilane or triphenylsilane as the
reducing agent, no appreciable reaction was observed
even hours after mixing the reactants.
On the other hand, when reacting TEMPO with
tributyltin hydride or triphenylgermanium hydride in
deoxygenated benzene, the nitroxide concentration was
observed to decrease gradually with time. GC/MS analy-
ses (see Table 1) of the reaction mixtures showed that
hydride
T/K
[hydride]/Ma
k1/M-1 s-1
-
-
2
4
Bu3SnH
Ph3GeH
Et3SiH
333
333
333
0.034-0.062
0.38-0.68
0.5
2.5 × 10
1.8 × 10
b
a
Range of concentrations employed. b Too low to be measurable.
give stable recombination products (eq 3), as suggested
by the detection of hexabutylditin in the reaction mixture
by GC/MS.
To determine the rate constants for hydrogen transfer
1
k , the reactions of TEMPO with a large excess of an
organometallic hydride were carried out inside an EPR
tube by recording the EPR spectra of the mixture at
different times. The time dependence of the TEMPO
concentration showed clean first-order decays with both
2
,2,6,6-tetramethyl-1-piperidinol (1) was the major prod-
uct derived from the nitroxide together with a very small
amount of 2,2,6,6-tetramethylpiperidine (3). This pro-
vides evidence that TEMPO reacts with the hydride by
hydrogen transfer to give the corresponding hydroxyl-
amine and tin or germyl radicals (eqs 1 and 2).
3 3
Bu SnH and Ph GeH in benzene at 60 °C (Figure 1).
Deviation from first order was observed after ca. 80% of
the nitroxide was consumed, and a small residual signal
also remained at very long reaction times. The measured
rate constants, kEPR, were independent of the concentra-
tion of TEMPO and directly proportional to the concen-
tration of the metal hydride, thus indicating that the
reaction is a pseudo first-order process involving hydro-
gen atom abstraction from the hydride.
The bimolecular rate constants obtained from the kEPR
values determined in the initial part of the kinetics and
from the hydride concentrations, in the assumption that
each molecule of hydride traps two nitroxide radicals (see
note 8), are listed in Table 2.
A still different behavior was observed when studying
the reaction of TEMPO with tris(trimethylsilyl)silane
(
3
TMS) SiH since, in this case, no reproducible decay rates
The yield of hydroxylamine based on nitroxide con-
could be obtained. GC/MS analyses of the reaction
mixtures provided evidence for the formation of another
major product derived from TEMPO besides 2,2,6,6-
tetramethyl-1-piperidinol. This was identified as 2,2,6,6-
tetramethylpiperidine (3) (Table 1), similarly to what
sumption was 96% with Bu
Although a 50% yield is expected on the basis of reactions
and 2, the excess of hydroxylamine obtained may be
3 3
SnH and 73% with Ph GeH.
1
due to the reversibility at 80 °C of the reaction 2 giving
back the tin or germyl radicals and the starting nitroxide.
The latter may undergo hydrogen transfer from the
unreacted metal hydride, and the Goup 14 radicals will
7
obtained in the reaction of nitroxides with polysilanes.
This result prompted us to more thoroughly investigate
the reduction of TEMPO by using several silanes and
different reaction conditions.
Th er m a lly or P h otoch em ica lly In itia ted Rea c-
tion s. The reactions of TEMPO with a variety of silanes
(7) Chatgilialoglu, C.; Guerrini, A.; Lucarini, M.; Pedulli, G. F.;
Borzatta, V.; Carrozza, P.; Da Roit G.; Lucchini, V. Organometallics,
in press.

Homolytic Reactivity of Group 14 Organometallic Hydrides
J . Org. Chem., Vol. 63, No. 5, 1998 1689
the reducing agent is (TMS)
SiH. When the reaction is
Ta ble 3. Rea ction of TEMP O w ith Silicon Hyd r id es
3
initiated with UV light (entries 12-16), at least 20-25%
of amine is formed even in the absence of metal hydrides
and of radical initiators (entry 12). The amine is obtained
conver- yield of
initi- meth-
sion
(%)
amine
3 (%)
a
entry
silane
ator
od
1
2
3
4
5
6
7
8
9
Et3SiH (1:5)b
PhSiH3 (1:5)
Ph3SiH (1:5)
(TMS)3SiH (1:2)
Et3SiH (1:5)
PhSiH3 (1:5)
(TMS)3SiH (1:2)
(TMS)3SiH (1:2)
Et3SiH (1:5)
Ph3SiH (1:5)
(TMS)3SiH (1:2)
none
DBP
DBP
DBP
DBP
TBPB
TBPB
TBPB
AIBN
TBHP
TBHP
TBHP
none
TBP
c
c
c
c
c
c
c
d
e
e
e
f
<5
<5
<5
96
<5
<5
97
69
52
43
95
13
48
47
100
57
-
-
in a high yield with (TMS)
observed when using thermal initiation, and with
TMS) Si(H)Me.
Mech a n ism of F or m a tion of th e Am in e. The
formation of 2,2,6,6-tetramethylpiperidine (3) in the
reactions of TEMPO with (TMS) SiH and (TMS) Si(H)Me
can be explained as follows. If we consider the reaction
with (TMS) SiH, the first step involves generation of
tris(trimethyl)silyl radicals by hydrogen transfer to the
initiating radicals followed by addition of (Me Si) Si to
3
SiH, similarly to what was
-
(
2
93
-
-
85
43
<5
<5
97
19
24
33
98
74
3
2
3
1
1
1
1
1
1
1
0
1
2
3
4
5
6
3
3
Et3SiH (1:2)
Ph3SiH (1:2)
(TMS)3SiH (1:2)
f
the oxygen atom of the nitroxide (eqs 4-6). The resulting
adduct, 4, will then fragment by cleavage of the nitrogen-
oxygen bond to give an aminyl radical 5, which by
abstracting a hydrogen atom (eq 9) from the starting
silane affords the amine 3.
TBP
TBP
g
g
g
(TMS)2Si(H)Me (1:2) TBP
a
TBHP ≡ di-tert-butyl hyponitrite, AIBN ≡ azobis(isobutyroni-
trile), DBP ≡ dibenzoyl peroxide, TBPB ≡ tert-butyl peroxyben-
b
zoate, TBP ≡ di-tert-butyl peroxide. The ratios in parentheses
indicate the molar ratio between TEMPO and the silicon hydride.
c
d
Toluene at 110 °C for 1 h (DBP 15% or TBPB 15%). Benzene
e
at 80 °C for 1 h (AIBN 15%). Benzene at 60 °C for 1 h (TBHP
1
0%). f Toluene at 80 °C under UV irradiation for 2 h (TBP 10%
g
if present). Benzene at 80 °C under UV irradiation for 2 h (TBP
1
0%).
were carried out under nitrogen atmosphere for 1-2 h
in benzene or toluene and in the presence of a thermal
or a photochemical radical initiator. As thermal initia-
tors we used di-tert-butyl hyponitrite (TBHP), azobis-
(
isobutyronitrile) (AIBN), dibenzoyl peroxide (DBP) or
tert-butyl peroxybenzoate (TBPB), and di-tert-butyl per-
oxide (TBP) as photochemical initiator. The reaction
mixtures were analyzed by GC/MS using tert-butylben-
zene as internal standard. The results of this study,
collected in Table 3, can be summarized as follows; when
the reaction was initiated thermally at 60 °C with TBHP,
quantitative reduction of TEMPO to the corresponding
amine 3 was achieved in the presence of (TMS)
3
SiH
The apparently unexpected detection of the siloxane
Me Si )Si(H)OSiMe can be accounted by two different
3 2 3
reaction pathways, one implying the formation (eq 7) of
the tris(trimethylsilyl)silyloxy radical (6) which has been
(
entry 11), while in the presence of Et SiH or Ph SiH less
3
3
(
than 5% yield of amine was obtained (entries 9, 10). In
all cases no hydroxylamine formation was detected by
GC/MS.
reported to undergo easy migration of a Me
3
Si group from
3
In the case of the reaction of TEMPO with (TMS)
one of the major reaction products was identified as the
siloxane (Me Si) (H)OSiMe from its mass spectrum.
3
SiH
8
silicon to oxygen to give (Me
3
Si
2
)Si(•)OSiMe (7) (eq 8);
this, under the present experimental conditions, may
abstract a hydrogen atom from the starting silane af-
fording the siloxane (Me Si )Si(H)OSiMe (eq 10).
3 2 3
2
2
3
When the reaction was photolytically initiated at 80 °C
with TBP the formation of the amine was almost quan-
titative with (TMS)
silyl)silane, (TMS) Si(H)Me, the nitroxide was also con-
verted to the amine, but less efficiently (entry 16). With
Et SiH or Ph SiH a ca. 30% yield of amine was obtained
entries 13 and 14), a value similar to the yield obtained
by photolyzing the reaction mixture in the absence of the
silane (entry 12).
Initiators other than TBHP and TBP were also inves-
tigated. AIBN behaved as a less effective initiator (entry
3
SiH (entry 15); with bis(trimethyl-
2
3
3
(
The other possibility is that the N-silyloxy amine
adduct 4 will fragment in a concerted way to give the
silyl radical 7 directly in addition to the aminyl radical
5
(eq 7′).
The various reaction steps will be discussed below in
detail, and the experiments carried out to discriminate
the two mechanisms will be described.
The initial steps of the reaction involve the formation
of silyl radicals and their combination with nitroxides (eq
8
), due to the fact that the alkyl radicals resulting from
its decomposition may react not only with the silane but
also with TEMPO, while DBP and TBPB were both
efficient initiators (entries 4, 7). When using Et
3
SiH and
Ph SiH no reaction was obtained with the last two
3
6
) for which the rate constants are unknown. Since it
initiators (entries 1-3 and 5, 6). These results indicate
that the reaction of nitroxides with metal hydrides in a
deoxygenated medium and in the presence of thermal
initiators leads to the corresponding amines only when
has been shown that TEMPO reacts with several alkyl
(8) Ballestri, M.; Chatgilialoglu, C.; Lucarini, M.; Pedulli, G. F. J .
Org. Chem. 1992, 57, 948.

1
690 J . Org. Chem., Vol. 63, No. 5, 1998
Lucarini et al.
9
radicals with a very high rate, it is also presumable that
the rate constants for the combination of nitroxides with
silyl radicals are large. To verify if this is the case,
competitive kinetic studies for the reaction of triethylsilyl
3
radicals, Et Si, with TEMPO and with molecular oxygen
were carried out.
The experiments were performed by introducing into
a thin EPR tube an air-saturated solution of the reaction
3
mixture containing Et SiH (2 M), the thermal initiator
tert-butyl hyponitrite (TBHP), the perdeuterated ali-
phatic nitroxide 2,2,6,6-tetramethyl-1-piperidinyloxyl-d17
1
1
(
TEMPOL-d17), and a small amount of R-tocopherol (R-
F igu r e 2. The circles show the experimental time dependence
of the height of the central line of the EPR spectrum of
TEMPOL-d17 observed during the decomposition of TBHP (43
-3
TOH, 1.510 M). The tube was then sealed and put into
the EPR cavity kept at constant temperature (50 °C), and
the spectra were recorded at constant time intervals. The
reactions which take place in this system are exemplified
below (eqs 11-16).
3
mM) at 50 °C in benzene in the presence of Et SiH (2.0 M),
TEMPOL-d17 (0.150 mM) and R-tocopherol (1.5 mM). Lines a
and b are the corresponding simulations using k13/k14 ratios
of 32 and 6.1, respectively.
The change of both nitroxide and oxygen concentration
during the course of the reaction within the closed tube
was determined by following the evolution of the three
lines EPR spectrum of the nitroxide. The integrated
intensity of these signals is proportional to the nitroxide
concentration while their width is related to the concen-
tration of oxygen since the effect of the latter on the EPR
spectra of radicals is that of producing a strong broaden-
ing of the spectral lines due to Heisenberg spin ex-
change.1
5,16
In oxygen-containing media the measured
line width can be expressed as the sum of two terms (eq
9) one (Wint) being the intrinsic width and the other the
1
oxygen-induced broadening. The latter one is propor-
tional to the oxygen concentration through a constant b
which is a function of the interaction radius between
oxygen and the radical and of the diffusion coefficient of
Under the usual steady-state approximation for the
concentration of the transient species, the kinetic equa-
tions describing the decay of oxygen and of the nitroxide
17
oxygen in solution.
W ) Wint + b[O2]
(19)
(
NOX) with time during the decomposition of TBHP can
1
2
be worked out as in eqs 17 and 18.
An other spectral parameter, sensitive to the oxygen
concentration, but much easier to measure than the
width, is the peak to peak height of the EPR line, I. This
is proportional to both the radical concentration and the
reciprocal of W.2 In the present case where the radical
is a nitroxide, the EPR signal height is expressed by eq
20 indicating that, when carrying out the reaction in a
closed system, I depends not only on the nitroxide
concentration but also on the amount of oxygen dissolved
in solution.
d[NOX]
k
14[NOX]
-
) Ri
(17)
(18)
dt
k [NOX] + k [O ]
1
4
13
2
d[O2]
k [O ]
13 2
-
) Ri
dt
k [NOX] + k [O ]
14 13 2
(
9) tert-Butyl, benzyl and cumyl radicals react with TEMPO at 293
8
8
8
-1 -1
K with rate constants of 7.6 × 10 , 4.9 × 10 , and 1.8 × 10
M
s ,
respectively.10
As a typical example, Figure 2 shows the dependence
(
10) Chateauneuf, J .; Lusztyk J .; Ingold, K. U. J . Org. Chem. 1988,
on time of the intensity of the central line of TEMPOL-
5
1
3, 1629. Beckwith, A. L. J .; Bowry, V. W.; Moad, G. J . Org. Chem.
988, 53, 1632.
-5
d
17 (2.6 × 10 M). At the beginning of the reaction the
(
11) Perdeuterated TEMPOL was used since the contribution to the
line width from unresolved proton hyperfine splitting is negligible.
12) R-Tocopherol has been added to the solution as scavenger of
the R SiOO peroxyl radicals. The reaction of tert-butoxyl radicals with
R-TOH (whose rate constant is 3.1 × 10 M
since the ratio between the concentrations of Et
(13) Valgimigli, L.; Banks, J . T.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1995, 117, 9966.
(
3
(14) Chatgilialoglu, C.; Ingold, K. U.; Lustzyk, J .; Nazran, A. S.;
Scaiano, J . C. Organometallics 1983, 2, 1332.
(15) Eastman, M. P.; Kooster, R. G.; Das M. P.; Freed, J . H. J . Chem.
Phys. 1969, 51, 2690.
(16) Backer, J . M.; Budker, V. G.; Eremenko S. I.; Molin, Yu. N.
Biochim. Biophys. Acta 1977, 460, 152.
9
-1 -1 13
s
)
has been neglected
3
SiH and R-TOH is
1
330 while the ratio between the rate constants for the hydrogen
323 6
-1 -1 14
abstraction from R-TOH and the silane (k12 ) 8.7 × 10
M
s
)
is only 350.

Homolytic Reactivity of Group 14 Organometallic Hydrides
J . Org. Chem., Vol. 63, No. 5, 1998 1691
[
NOX]
I ∝
(20)
2
(
Wint + b[O ])
2
intensity of the EPR lines increases since the silyl
radicals produced in reaction 12 react with oxygen rather
than with the nitroxide. While oxygen is consumed in
the reaction vessel, the width of the EPR lines decreases
and their height increases. Only when most of the
oxygen has been consumed do the silyl radicals begin to
react with TEMPOL-d17 (eq 14) and the EPR signals of
the latter start to decrease.
In another experiment the triethylsilyl adduct to
TEMPO was prepared by reacting the nitroxide with
Et Si radicals in triethylsilane containing an equimo-
3
lecular amount of the TBHP initiator at 50 °C. This
adduct (identified by its mass spectrum) was kept at 80
Simulation of the time dependence of the intensity of
the EPR spectrum of TEMPOL-d17 during the course of
the reaction provides the ratio of the rate constants for
the combination of silyl radicals with oxygen and with
the nitroxide k13/k14
simultaneous differential eqs 17 and 18 gave the best fit
to the experimental points by using k13/k14 ) 6.1, i.e., the
3
Et Si radicals react with oxygen six times faster than
with the nitroxide. The absolute value of k13 has not been
measured in solution but only in the gas phase (ca. 1.5
°
C for 12 h without observing any appreciable decomposi-
tion.
On this basis it can be concluded that only the adducts
of the polysilanes, (TMS) SiH and (TMS) Si(H)Me, frag-
3
2
.
Numerical integration18 of the
ment at the oxygen-nitrogen bond via a concerted
mechanism to afford aminyl radicals which give the
observed amine. This reaction can take place only if the
3 2
hydrogen transfer from the (TMS) SiH or (TMS) Si(H)Me
to the aminyl radical (eq 9) is characterized by a rate
constant large enough to compete with other possible
pathways for the decay of the aminyl radical. Despite
the great number of rate constants for the reaction of
simple alkyl radicals with different substrates, kinetic
data for nitrogen-centered radicals are substantially more
1
0
-1 -1 19
×
10
.
M
s
)
for the reaction of Me
3
Si radicals with
O
2
By assuming this value also in benzene, k14 is
9
-1 -1
estimated to be ca. 2.5 × 10 M
s
, thus implying that
20
the reaction of silyl radicals with nitroxides is so fast that,
in the absence of oxygen, all other reactions of silyl
radicals can be safely disregarded.
21
limited. Newcomb and co-workers have measured rate
constants for the bimolecular reactions of thiols and tin
hydride with dialkylaminyl radical and Roberts and
Ingold22 have found that the 2,2,6,6-tetramethylpiperi-
dinyl radical decays, following pseudo first-order kinetics,
by abstracting a hydrogen atom from most of the solvents
used.
To measure the rate constants for hydrogen abstraction
from silanes by hindered aminyl radicals, 2,2,6,6-tetra-
methylpiperidine was reacted in the EPR cavity with tert-
butoxy radicals, photolytically produced from di-tert-butyl
peroxide in deoxygenated benzene or tert-butylbenzene,
in the presence of a given silane. UV irradiation of the
solutions resulted in the immediate formation of a radical
whose EPR spectrum showed nitrogen and proton split-
The following step in the proposed reaction mechanism
for the deoxygenation of the nitroxide to give the amine
(eqs 4-9) is represented by the cleavage of the nitrogen-
oxygen bond of the silylaminoether to give an aminyl
radical via either a two-step pathway implying the
formation of a silanoxyl radical (6) which undergoes fast
rearrangement to 7 (eqs 7, 8) or a concerted path leading
directly to 7 (eq 7′). In the case of the reaction with
(
TMS)
since the final products are identical. On the other hand,
the fact that neither amine, nor silanol (Et SiOH or
Ph SiOH), formation was observed when reacting TEMPO
with Et SiH or Ph SiH suggests that the tris(trimethyl-
3
SiH the two mechanisms cannot be differentiated
3
3
3
3
silyl)silyl adduct to TEMPO (4) undergoes a concerted
fragmentation. It would be hard to justify why cleavage
of the nitrogen-oxygen bond to give aminyl and silyloxy
radicals (eq 7) should occur only in the case of 4 and not
tings and a g-value (aN 14.66 G, a (18H) 0.82 G, g 2.0046)
characteristic of the 2,2,6,6-tetramethylpiperidinyl radi-
cal (5). When the light was shut off, the EPR signals
disappeared following good pseudo first-order kinetics
with decay times dependent on the nature and concen-
tration of the silane present in the system. In the case
H
in that of the Et
3 3
Si or Ph Si adducts.
To further investigate this point the silane was reacted,
under free radical conditions, with nitromethane, MeNO
In this case the silyl radicals are known to add to MeNO
to give the corresponding silyloxy-nitroxides (eq 21) which
2
.
2
3
of (TMS) SiH also the EPR spectrum of the corresponding
silyl radical was visible; however, its decay was much
faster than that of the aminyl radical.
8
decay by fragmentation at the N-O bond (eq 22) to afford
The room-temperature rate constants, reported in
Table 4, were obtained from the measured kEPR values
and the concentration of the silanes in the assumption
that each molecule of silane traps two aminyl radicals.
The Arrhenius parameters for reaction 9 are also listed
in Table 4 while the corresponding plots are shown in
Figure 3.
3
MeNdO and R SiO. The latter radical is expected to
abstract a hydrogen atom from the starting silane, thus
giving the silanols Et SiOH and Ph SiOH (eq 23) for R
Et and Ph, or to undergo reactions 8 and 10 to afford
the siloxane (Me Si) Si(H)OSiMe for R ) Me Si.
3
3
)
3
2
3
3
GC/MS analysis of the final mixtures of the reactions
of nitromethane with the three silanes revealed, in all
cases, the major products as those resulting from nitro-
gen-oxygen cleavage of the intermediate silyloxy nitrox-
A remarkable feature of the data in Table 4 is the
unusually low values of log A which for hydrogen
2
4
abstraction reactions are normally in the range 8.5-9.0.
3 3 3 2 3
ides, i.e., Et SiOH, Ph SiOH, and (Me Si) Si(H)OSiMe .
(
20) Newcomb, M. Tetrahedron 1993, 49, 1151, and references
therein.
(21) Newcomb, M.; Horner, J . H.; Shahin, H. Tetrahedron Lett. 1993,
(
(
17) Hyde J . S.; Subczynski, W. K. J . Magn. Reson. 1984, 56, 125.
18) Cheney W.; Kincaid, D. Numerical Mathematics and Comput-
34, 5523. Musa, O. M.; Horner, J . H.; Shahin, H.; Newcomb, M. J . Am.
Chem. Soc. 1996, 118, 3862.
(22) Roberts, J . R.; Ingold, K. U. J . Am. Chem. Soc. 1973, 95, 3228.
ing; Wadsworth: Belmont, 1985.
19) Niiranem, J . T.; Gutman, J . J . Phys. Chem. 1993, 97, 4106.
(

1
692 J . Org. Chem., Vol. 63, No. 5, 1998
Lucarini et al.
F igu r e 3. Arrhenius plots for the reaction of the aminyl radical 5 with silicon hydrides in tert-butylbenzene solutions.
Ta ble 4. Kin etic P a r a m eter s for th e Rea ction of th e
to the hydroxylamine, a major amount of amine is
formed. When the same reaction is carried out in the
presence of radical initiators the reduction of TEMPO to
the corresponding amine occurs with an almost quantita-
2
,2,6,6-Tetr a m eth ylp ip er id in yl Ra d ica l w ith Silicon
Hyd r id es
a
log A/
Ea/
kcal mol
kH298(R2N•)/
kH (RCH2•)/
-
1
-1
-1
-1 -1
-1 -1
hydride
M
s
M
s
M
s
tive yield with (TMS)
conspicuous yield with (TMS)
conditions, i.e., in the presence of radicals able to abstract
a hydrogen atom from the metal hydride, Et SiH and
Ph SiH afford the silyl adducts to the nitroxide and no
3
SiH, and with a lower although
2
Et3SiH
5.2
5.7
5.8
6.9
4.8
4.8
6.5
6.7
5.8
6.6
4.4
5.2
2.7
5.3
32.1
119.2
34.4
10.0
6.4 × 10 (298 K)
Si(H)Me. Under the same
2
4
Ph3SiH
Ph2SiH2
PhSiH3
TMS)3SiH
TMSSi(H)Me2
4.6 × 10 (413 K)
4
5.6 × 10 (413 K)
4
3
2.9 × 10 (413 K)
5
3
(
1.4 × 10 (298 K)
3
amine or hydroxylamine are obtained. Therefore, only
polysilanes seem to be able to perform the deoxygenation
of nitroxides. This has been interpreted by suggesting
that this reaction takes place by a concerted fragmenta-
tion of the aminosilyl ether formed by addition of the silyl
3.5 × 10 (298 K)
a
kH (RCH2•) is the rate constant for hydrogen abstraction from
the silane by primary alkyl radicals.23
These low values can be explained in terms of the high
steric demand for a reaction in which both the radical
3 2
radical from (TMS) SiH and (TMS) Si(H)Me to the
oxygen atom of the nitroxide. The driving force of this
reaction appears to be in the initial stage the formation
of a strong oxygen-silicon bond at the expense of the
center in the abstracting aminyl radical and the hydro-
gen-silicon bond in the silane are sterically hindered.25
Due to the importance of steric factors in these reactions,
the order of reactivity does not follow the strength of the
metal-hydrogen bond as for the reactions of silanes with
alkyl radicals (see the last column of Table 4) or other
2
9
much weaker nitrogen-oxygen and silicon-silicon bonds.
Once the fragmentation has occurred, recombination of
the silyl radical 7 with the aminyl 5 does not take place
because of the steric crowding around the radical centers
of both species, so that evolution to the amine and to the
2
3
nonhindered radicals.
3 2 3
siloxane (Me Si) Si(H)OSiMe is preferred.
Con clu sion s
Finally, a noteworthy point is the formation under UV
irradiation in benzene of ca. 20% amine from the nitrox-
ide in the absence of any other reactants. Although
nitroxides have been reported to react photochemically
with aliphatic solvents32 by hydrogen abstraction to
afford hydroxylamines and aminoethers, it seems difficult
to envisage a reasonable reaction path leading to its
3
TEMPO reacts spontaneously at 80 °C with Bu SnH
and Ph GeH by hydrogen atom abstraction from the
3
hydride to give the corresponding hydroxylamine char-
acterized by an O-H bond dissociation energy (BDE) of
-
1
26
6
)
(
9 kcal mol in benzene. Bu
3
SnH, BDE (Me
reacts faster than Ph GeH, BDE
SiH and Ph SiH
3
Sn-H)
-
1 27
77 kcal mol
Ge-H) ) 82.6 kcal mol
do not react at 333 K due to the large endothermicity of
,
3
-
1 28
33
Bu
3
,
while Et
3
3
deoxygenation.
-
1 23
this reaction, BDE (Me
3
Si-H) ) 95 kcal mol
Si-H) ) 83
as metal hydride the spontaneous reaction
with TEMPO follows a different path since, in addition
.
When
Exp er im en ta l Section
using tris(trimethylsilyl)silane, BDE ((TMS)
3
-
1 23
Mater ials. Bis(trimethylsilyl)methylsilane (TMS)
2
Si(H)Me,34
kcal mol
,
pentamethyldisilane,34 and tert-butyl hyponitrite were pre-
35
(
23) Chatgilialoglu, C. Chem. Rev. 1995, 95, 1229.
(29) The BDE of N-O bond in aminoethers having a structure
-1 31
(24) Benson, S. W. Thermochemical Kinetics, 2nd ed.; Wiley-Inter-
similar to 4 has been estimated to be ca. 33 kcal mol
;
that of an
-1 30
science: New York, 1976.
Si-O bond is 128 kcal mol
and that of a Si-Si bond is ca. 75 kcal
-1 30
(25) Howard, J . A. Adv. Free-radical Chem. 1971, 4, 49.
mol .
(
26) Mahoney, L. R.; Mendenhall, G. D.; Ingold, K. U. J . Am. Chem.
(30) Walsh, R. Acc. Chem. Res. 1981, 14, 246.
(31) Step, E. N.; Turro, N. J .; Klemchuk, P. P.; Grande, M. E. Die
Angew. Makro. Chem. 1995, 232, 65.
(32) Keana, J . F. W.; Dinerstein, R. J .; Baitis, F. J . Org. Chem. 1971,
36, 209.
Soc. 1973, 95, 8610.
27) Zavitsas, A. A.; Chatgilialoglu, C. J . Am. Chem. Soc. 1995, 117,
0645.
28) Clark, K. B.; Griller, D. Organometallics 1991, 10, 746.
(
1
(

Homolytic Reactivity of Group 14 Organometallic Hydrides
J . Org. Chem., Vol. 63, No. 5, 1998 1693
pared following literature procedure. All other materials were
commercially available.
saturated at room temperature and introduced (ca. 200 L) into
a capillary tube with the internal diameter of ca. 1.85 mm. A
second capillary tube (external diameter of 1.60 mm) sealed
at one end was introduced into the sample tube so as to leave
very little dead volume space. The decrease of the nitroxide
concentration and the oxygen uptake was followed by EPR
spectroscopy as described in the paper.
EP R Mea su r em en ts. The EPR spectra were recorded on
a Bruker ESP300 spectrometer equipped with an NMR gauss-
meter for field calibration, a Bruker ER033M field-frequency
lock, and a Hewlett-Packard 5350B microwave frequency
counter for the determination of the g-factors, which were
corrected with respect to that of perylene radical cation in
Gen er a l P r oced u r e for th e Rea ction of TEMP O w ith
Sila n es. A solution of TEMPO (0.1 M), the silane (0.1-0.5
M), a radical initiator, and an internal GC standard, in
benzene or toluene, was degassed and sealed under argon in
Pyrex ampules. The reaction mixture was thermolyzed or
photolyzed in the temperature range 333-373 K. The prod-
ucts of the reactions were analyzed by a HP GC5890 Series II
chromatograph coupled to an HP mass selective detector Model
5
971 using a 30 m × 0.25 mm HP-5MS cross-linked 5%
phenylsilicone capillary column with temperature program-
ming from 50 to 250 °C. The products of interest were
identified by comparison of their retention times with authen-
tic materials.
2 4
concentrated H SO (g 2.00258). Photolysis was carried out
by focusing the unfiltered light from a 500 W high-pressure
mercury lamp on the EPR cavity. Digitized decay traces of
the EPR signals necessary for kinetic measurements were
acquired by means of a dedicated computer and were trans-
ferred to an AT-486 (33 MHz) personal computer in order to
be analyzed. The temperature was controlled with a standard
variable temperature accessory and was monitored before and
after each run with a copper-constantan thermocouple.
Kin etic Stu d ies for th e Rea ction of Tr ieth ylsilyl Ra d i-
ca l w ith TEMP OL-d 17 a n d Molecu la r Oxygen . A solution
of triethylsilane (2.0 M), tert-butyl hyponitrite (0.1 M), 2,2,6,6-
-
4
tetramethyl-1-piperidinyloxyl-d17 (1.5 × 10 M), and a small
-
3
amount of R-tocopherol (1.5 × 10 M) in benzene was air-
(33) One of the referees suggested that the formation of the amine
could be explained by admitting that the photoexcited nitroxide
dimerize to give two aminyl radicals and molecular oxygen similarly
to the well-known dimerization of peroxyl radicals followed by frag-
mentation of tetraoxide intermediates which break down to yield
diatomic oxygen and two alkoxy radicals.
Ack n ow led gm en t. Financial support from MURST
and CNR is gratefully acknowledged. We also thank
Dr. S. Bernardoni for assistance in some kinetic
measurements.
(
34) Kumada, M.; Ishikawa, M.; Maeda, S. J . Organomet. Chem.
964, 2, 478.
35) Mendenhall, G. D. Tetrahedron Lett. 1983, 24, 451.
1
(
J O972178I