Acid Is Key to the Radical-Trapping Antioxidant Activity of Nitroxides

Article abstract of DOI:10.1021/jacs.6b00677

Persistent dialkylnitroxides (e.g., 2,2,6,6-tetramethylpiperidin-1-oxyl, TEMPO) play a central role in the activity of hindered amine light stabilizers (HALS)-additives that inhibit the (photo)oxidative degradation of consumer and industrial products. The accepted mechanism of HALS comprises a catalytic cycle involving the rapid combination of a nitroxide with an alkyl radical to yield an alkoxyamine that subsequently reacts with a peroxyl radical to eventually re-form the nitroxide. Herein, we offer evidence in favor of an alternative reaction mechanism involving the acid-catalyzed reaction of a nitroxide with a peroxyl radical to yield an oxoammonium ion followed by electron transfer from an alkyl radical to the oxoammonium ion to re-form the nitroxide. In preliminary work, we showed that TEMPO reacts with peroxyl radicals at diffusion-controlled rates in the presence of acids. Now, we show that TEMPO can be regenerated from its oxoammonium ion by reaction with alkyl radicals. We have determined that this reaction, which has been proposed to be a key step in TEMPO-catalyzed synthetic transformations, occurs with k ～ 1-3 × 10¹⁰ M^-1 s^-1, thereby enabling it to compete with O₂ for alkyl radicals. The addition of weak acids facilitates this reaction, whereas the addition of strong acids slows it by enabling back electron transfer. The chemistry is shown to occur in hydrocarbon autoxidations at elevated temperatures without added acid due to the in situ formation of carboxylic acids, accounting for the long-known catalytic radical-trapping antioxidant activity of TEMPO that prompted the development of HALS.

Full text of DOI:10.1021/jacs.6b00677

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Acid is Key to the Radical-Trapping Antioxidant Activity of Nitroxides
Evan A Haidasz, Derek Meng, Riccardo Amorati, Andrea
Baschieri, Keith U. Ingold, Luca Valgimigli, and Derek A. Pratt
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b00677 • Publication Date (Web): 29 Mar 2016
Downloaded from http://pubs.acs.org on March 31, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Acid is Key to the Radical-Trapping Antioxidant Activity of
Nitroxides
Evan A. Haidasz,‡ Derek Meng,‡ Riccardo Amorati,¶ Andrea Baschieri,¶ Keith U. Ingold,§
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Luca Valgimigli¶,* and Derek A. Pratt‡,*
‡Department of Chemistry, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada. ¶Department of Chemistry “G.
Ciamician”, University of Bologna, Bologna, I-40126, Italy. §National Research Council of Canada, Ottawa, Ontario
K1A 0R6, Canada.
ABSTRACT: Persistent dialkylnitroxides (e.g. 2,2,6,6-tetramethylpiperidinoxyl, TEMPO) play a central role in the activity of hin-
dered amine light stabilizers (HALS) – additives that inhibit the (photo)oxidative degradation of consumer and industrial products.
The accepted mechanism of HALS comprises a catalytic cycle involving the rapid combination of a nitroxide with an alkyl radical
to yield an alkoxyamine that subsequently reacts with a peroxyl radical to eventually reform the nitroxide. Herein, we offer evi-
dence in favour of an alternative reaction mechanism involving the acid-catalyzed reaction of a nitroxide with a peroxyl radical to
yield an oxoammonium ion followed by electron transfer from an alkyl radical to the oxoammonium ion to reform the nitroxide. In
preliminary work we showed that TEMPO reacts with peroxyl radicals at diffusion-controlled rates in the presence of acids. Now
we show that TEMPO can be regenerated from its oxoammonium ion by reaction with alkyl radicals. We have determined that this
reaction, which has been proposed to be a key step in TEMPO-catalyzed synthetic transformations, occurs with k ~1-3×1010 M-1s-1 –
thereby enabling it to compete with O2 for alkyl radicals. The addition of weak acids facilitates this chemistry, whereas the addition
of strong acids slows the reaction by enabling back electron transfer. The chemistry is shown to occur in hydrocarbon autoxidations
at elevated temperatures without added acid due to the in situ formation of carboxylic acids, accounting for the long-known catalyt-
ic radical-trapping antioxidant activity of TEMPO that prompted the development of HALS.
INTRODUCTION
degenerative diseases in which lipid autoxidation (peroxida-
tion) have been implicated.7
Radical-trapping antioxidants (RTAs) inhibit the autoxida-
tion of hydrocarbons by trapping the peroxyl radicals (ROO•)
that propagate the chain reaction.1,2 Phenols and diphenyla-
mines are the archetype RTAs (represented by A-H in Eqs. 1
and 2); they undergo efficient formal H-atom transfer to per-
oxyl radicals to yield stable RTA-derived radicals that do not
continue the chain reaction (A•, Eq. 1). Instead, A• reacts with
another peroxyl to yield non-radical products (Eq. 2). Thus,
the overall stoichiometry of the ROO•:A-H reaction is 2.
Although nitroxides do not react with peroxyl radicals,3,8
they react very quickly with alkyl radicals (Eq. 3).9 At low
concentrations of O2 and sufficiently high concentrations of
nitroxide, this reaction competes with addition of O2 to alkyl
radicals (which propagates autoxidation) to break one chain
reaction. To account for the apparent catalytic RTA activity of
TEMPO (or HALS), it is widely accepted that TEMPO is re-
generated from the alkoxyamine.6 Among the various mecha-
nisms proposed for regeneration,10 the most reasonable starts
with H-atom abstraction from the alkoxyamine by a chain-
carrying peroxyl radical (Eq. 4, Scheme 1).10,11 This reaction
A-H + ROO• → A• + ROOH
(1)
A• + ROO• → non-radical products
(2)
Nitroxides, such as 2,2,6,6-tetramethylpiperidinoxyl
(TEMPO, Scheme 1), have long been of interest as RTAs.
Although early work by Brownlie and Ingold found TEMPO
to be unreactive to peroxyl radicals at 65°C,3 later work
showed that nitroxides were incredibly reactive at higher tem-
peratures.4,5 For example, a reaction stoichiometry of 510 was
reported for the TEMPO-inhibited autoxidation of paraffin oil
at 130°C, making TEMPO (and related hindered nitroxides)
the most efficacious RTA(s) ever reported.5 This remarkable
result coincided with the development of Hindered Amine
Light Stabilizers (HALS) – technology now universally used
to slow the (photo)oxidation of polymers and other petroleum-
derived products. HALS generally comprise secondary or ter-
tiary hindered amines that undergo in situ oxidation to nitrox-
ides.6 In addition to this application, nitroxides have been ex-
plored as potential therapeutic and preventive agents against
ROO
H
N
O
R''
R'
slow
R''
R'
ROOH
H
(3)
O
(4)
very fast
N
O
N
R''
TEMPO
R'
(6)
(5)
RO
very fast
R'
N
O
ROO
very fast
R''
Scheme 1. The proposed mechanism of hindered amine light
stabilizer (HALS) activity.
ACS Paragon Plus Environment

Journal of the American Chemical Society
produces an alkyl radical that rapidly undergoes β-
Page 2 of 10
+
R
1O2
+
X
1
2
3
4
5
6
7
8
R'
fragmentation to yield a carbonyl compound and aminyl radi-
cal (Eq. 5), from which TEMPO can be regenerated following
reaction with a peroxyl radical (Eq. 6). This series of reactions
completes a catalytic cycle wherein the substrate undergoing
autoxidation serves as the stoichiometric reductant, and each
pass through the cycle traps two chain-carrying radicals.
O
(10)
+
R
OH
N
O
R'
+
N
OH
HX
The resultant hydroxylamine could then react with a peroxyl
radical to regenerate TEMPO – known to proceed readily with
k = 5×105 M-1s-1.3 No precedent exists in the literature for this
mechanism. However, the reaction of hydrogen peroxide with
TEMPO+ was suggested some time ago by Denisov as a key
step in the TEMPO-inhibited autoxidation of ethylbenzene.17
More recently, Goldstein et al. investigated the kinetics of this
reaction in water.7b They found that the reaction was only par-
ticularly facile at alkaline pH, where electron transfer from the
peroxide anion was possible, cf. k = 5×104 M-1s-1 at pH 8 and
5.1 M-1s-1 at pH 4.7b Given that the latter value must represent
the upper bounds in acidified organic solvent, this mechanism
would appear to be irrelevant in the current context. Moreover,
the reductant in the current context must be a hydroperoxide,
not H2O2, which is less acidic and therefore less reactive as an
ET reagent. Indeed, when we monitored the kinetics of the
reaction of the tetrafluoroborate salt of the oxoammonium ion
A short time ago we found that the long-held dogma that ni-
troxides do not react with oxygen-centered radicals12 must be
applied with care. When autoxidations were carried out at
ambient temperatures in the presence of added acid (e.g. acetic
acid or trifluoroacetic acid), nitroxides were excellent radical-
trapping antioxidants.13 In our preliminary report we showed
that TEMPO, when protonated (Eq. 7), reacts with peroxyl
radicals at rates approaching the diffusion-controlled limit
(e.g. kinh ~ 1×108 M-1s-1 with 10 mM p-TsOH) as in Eq. 8.14
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H+
N
OH
(7)
N
O
+
+
ROO
N
OH
N
O
ROOH (8)
derived from TEMPO (hereafter TEMPO+BF4 ) with t-
–
BuOOH in acetonitrile, we obtained a rate constant of only
(4.5±0.5)×10-3 M-1s-1 (see Supporting Information for details).
This value, in light of the modest amount of hydroperoxides
that are formed in the early stages of an (inhibited) autoxida-
tion, suggests that this cannot be the mechanism of TEMPO
regeneration.
Furthermore, we showed that in the presence of weak car-
boxylic acids (e.g. acetic or benzoic acid), very long inhibited
periods are observed, during which the nitroxide is not con-
sumed – corresponding to massive stoichiometric numbers13
–
such as those observed in high temperature autoxidations.5
Thus, under weakly acidic conditions, TEMPO must be regen-
erated from its corresponding oxoammonium ion, TEMPO+.13
In sharp contrast, in the presence of strong acids (e.g. tri-
fluoroacetic acid), only one peroxyl radical was trapped, sug-
gesting that TEMPO could not be regenerated from TEMPO+.
At the time of our preliminary communication, the mechanism
by which TEMPO was regenerated was elusive, as was the
reason it occurred in the presence of weak acids, but not
stronger ones. Herein we offer an explanation for these results
and demonstrate that it accounts for the remarkable inhibition
of autoxidation by nitroxides, and by extension, HALS.
Further consideration of all the species present in the inhib-
ited autoxidations left only one possible candidate to reduce
the oxoammonium ion: the chain-carrying alkyl radical, R•
(Eq. 11):18
+
+
(11)
R
R
N
O
N
O
Computations carried out at the CBS-QB3 level of theory19
suggest that this electron transfer reaction is exergonic, with
ΔG° = -3.9 and -6.1 kcal/mol in the gas phase and acetonitrile,
respectively, for R = cumyl (cumene was one of the substrates
that we autoxidized in our previous studies).13
RESULTS
Mechanistic Possibilities. The oxoammonium ion derived
from TEMPO is best known for its role as the catalytic oxidant
in the oxidation of alcohols to carbonyl compounds.15,16 This is
believed to occur via a hydride transfer in a non-covalent pre-
reaction complex (Eq. 9):
Product Studies in a Model Reaction. Since inhibited au-
toxidations are inherently complex, evidence for the reaction
in Eq. 11 was initially sought in a model system. Cumyl radi-
cals were generated from the thermal decomposition of
azocumene in acetonitrile at 50°C in the presence of
X
–
TEMPO+BF4 and Me4NOAc. The latter was included to pro-
+
+
(9)
O
N
OH
R
OH
N
R
duce the same TEMPO+AcO– species that is expected to form
in a TEMPO-inhibited autoxidation in the presence of acetic
acid. The reaction products were analyzed by HPLC with UV
detection. The results are shown in Figure 1.
O
+
HX
As such, at the conclusion of our preliminary work,13 we
wondered whether a (somewhat) similar reaction could occur
with hydroperoxide products formed during an autoxidation.
This reaction would require hydride transfer from the β-carbon
with concomitant formation of 1O2 (Eq. 10):
In the absence of O2 (via sparging the solutions extensively
with argon), the cumyl-TEMPO adduct was the sole product
formed and appeared at a rate coinciding with the rate at
which azocumene decomposed.20 This is consistent with the
expected reaction sequence shown in Scheme 2. Thus, decom-
2
ACS Paragon Plus Environment

Page 3 of 10
Journal of the American Chemical Society
position of azocumene proceeds to yield two cumyl radicals – Inhibited Autoxidations. If TEMPO is formed from
TEMPO+ in an autoxidation, the oxoammonium ion should
inhibit the autoxidation similarly to the nitroxide from the
outset. Thus, matched styrene autoxidations were carried out
where either TEMPO or TEMPO+ were added with or without
acetic acid. The results are shown in Figure 2. Similar results
were obtained in autoxidations of cumene (see Supporting
Information).
1
2
3
4
5
6
7
one of which reduces the oxoammonium ion to TEMPO, and
the other reacts with TEMPO to give the observed alkoxy-
amine product. The same profile was observed in chloroben-
zene (see Supporting Information). Similar results were ob-
tained in experiments where azo(phenylethane), which yields
secondary 1-phenylethyl radicals, was used in lieu of azocu-
mene (see Supporting Information).
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. O2 consumption during the autoxidation of styrene (4.3
M) in acetonitrile initiated by AIBN (0.05 M) at 30°C (a, dotted
line), and corresponding experiments carried out in the presence
of 13 µM of either TEMPO+BF4– (b, d) or TEMPO (c, e) without
(b, c) or with (d, e) acetic acid (43 mM).
Figure 1. Decomposition of azocumene (n) and formation of the
cumyl-TEMPO adduct in the presence of TEMPO+BF4– (25 mM)
and Me4N+AcO– (25 mM) in acetonitrile at 50°C in the presence
of argon (l), air (p) or O2 (q).
As we reported in our preliminary communication,13
TEMPO + AcOH gave rise to a lengthy inhibited period corre-
sponding to a reaction stoichiometry of n >> 1. When
TEMPO+BF4– was used in place of TEMPO, the initial rate of
autoxidation was not suppressed to the same extent as with
TEMPO, but significant inhibition was still observed. Each of
the TEMPO-inhibited and TEMPO+-inhibited autoxidations
could be modeled using the standard kinetic scheme for the
inhibited autoxidation of styrene,23 but including the proposed
inhibition reactions in Eq. (7), (8) and (11), and fit by numeri-
N
Ph
2
Ph
N
+
Ph
O
O
N
+
N
+
Ph
Ph
O
N
cal integration to yield the unknown rate constant k11
=
+
N
(2.4±0.4)×1010 M-1s-1 (see Supporting Information for the de-
tails). The magnitude of this rate constant accounts for the
successful competition of TEMPO+ with O2 for reaction with
alkyl radicals necessary for this mechanism to be viable.
O
Ph
Ph
Scheme 2. Formation of the cumyl-TEMPO adduct from
–
azocumene and TEMPO+BF4 .
The combination of TEMPO and AcOH is also effective at
inhibiting the autoxidation of styrene in chlorobenzene, but no
significant inhibitory activity is observed past n = 1, implying
that oxidation by TEMPO+ no longer effectively competes
with O2 for alkyl radicals. Indeed, when used directly,
In the presence of air, the cumyl-TEMPO adduct was also
formed, but at a much lower level, whereas when the reaction
was carried out under an atmosphere of O2, no adduct was
observed. The lower yields of alkoxyamine in the presence of
air were expected given the rapid reaction of cumyl radicals
with O2 (k ~ 3×109 M-1s-1),21 which outcompetes the reaction
of cumyl radicals with TEMPO (k = 1.2×108 M-1s-1).9 This is
underscored by the complete lack of alkoxyamine formed
when the reaction was carried out in the presence of O2. The
addition of acetic acid (25 mM) had no effect on the reaction
(see Supporting Information).22
–
TEMPO+BF4 had only a small effect on the autoxidation
(Figure 3.). Modelling the data as above yields the rate con-
stant k11 = 1.5 ×109 M-1s-1, which is similar to the rate constant
for the competing reaction of the styryl radicals with O2 (~3
×109 M-1s-1). However, since [O2] = 2 mM >> [TEMPO+] = 13
µM, this reaction is not competitive and no catalytic activity is
observed.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. O2 consumption during the autoxidation of styrene (4.3
M) in chlorobenzene initiated by AIBN (0.05 M) at 30°C (a, dot-
ted line) and corresponding experiments in the presence of 13 µM
of either TEMPO+BF4 (b) or TEMPO (c) and AcOH (44 mM).
EPR Spectroscopy. To corroborate the foregoing product
studies and inhibited autoxidation data, we sought to directly
demonstrate that TEMPO is produced from TEMPO+ under
the conditions of a typical autoxidation. In fact, upon addition
–
of TEMPO+BF4 to an air-saturated solution of either cumene
or styrene in acetonitrile containing AIBN as radical initiator,
the formation of TEMPO could be observed directly by EPR
(Figure 4A).
The evolution of the spectra revealed a rate of TEMPO for-
mation (3.5×10-9 Ms-1) that was only slightly lower than the
rate of radical initiation from AIBN decomposition (Ri =
6.1×10-9 Ms-1).24 (The rate determined in a cumene autoxida-
tion was 5.0×10-9 Ms-1 – see Supporting Information.) As a
result, since the rate constant for O2 addition to the alkyl radi-
cals derived from AIBN and styrene (or cumene) is k ~ 3×109
Figure 4. TEMPO formation during the autoxidation of styrene
(4.3 M) in acetonitrile initiated by AIBN (50 mM) at 30°C in the
–
presence of TEMPO+BF4 (0.66 mM). Top: representative EPR
spectra. Bottom: rate with (l) and without (n) added acetic acid
(35 mM).
–
M-1s-1, [O2] ~ 2 mM and [TEMPO+ BF4 ] = 0.66 mM, we can
On the Addition of Strong Acids. Each of the three forego-
ing experiments were also carried out with TFA in place of
AcOH. The data are presented in the Supporting Information.
To summarize the results: 1) significantly lower yields of the
cumyl-TEMPO adduct were observed when azocumene was
decomposed in the presence of TEMPO+BF4– and TFA instead
estimate that under these conditions k11 ~ 1×1010 M-1s-1. Inter-
estingly, when acetic acid (35 mM) was added to the reaction
mixture, the formation of TEMPO was accelerated (compared
data sets in Figure 4B), yielding an observed rate that was
indistinguishable from the rate of radical initiation and imply-
ing that the rate constant for electron transfer had increased to
k11 ~ 3×1010 M-1s-1 – in excellent agreement with the rate con-
stant derived from fitting the inhibited autoxidation data (vide
supra). It is important to note that no TEMPO was observed in
the absence of AIBN. Moreover, in the absence of either cu-
mene or styrene, the formation of TEMPO due to reduction of
TEMPO+ by AIBN-derived radicals was determined to be
much slower (k11 ~ 1×108 M-1s-1 – see Supporting Infor-
mation).
–
of AcOH, 2) the combination of TEMPO+BF4 and TFA did
not inhibit the autoxidation of styrene in acetonitrile, and 3) no
TEMPO was observed by EPR spectroscopy when AIBN was
decomposed in the presence of TEMPO+BF4– and TFA instead
of AcOH.
The simplest explanation for the lack of regeneration of
TEMPO from TEMPO+ in the presence of a strong acid is that
it is destroyed under these conditions. Indeed, Ma and
coworkers have suggested that TEMPO+ is unstable in the
presence of H2SO4 at elevated temperatures (100°C, monitored
by UV-Vis spectroscopy),26 where it is believed to eliminate
HNO, eventually yielding unsaturated aliphatic products
In chlorobenzene, the rate of formation of TEMPO from
–
TEMPO+BF4 in the presence of the same amount of acetic
acid enables derivation of a rate constant of k11 = 2.6×109 M-1
s-1 for the electron transfer – just over an order of magnitude
slower than in acetonitrile and in excellent agreement with the
kinetic data from the inhibited autoxidations.25
–
(Scheme 3). However, when we subjected TEMPO+BF4 (400
µM) to 10 mM TFA in acetonitrile at 50°C, no change was
observed in the UV-Vis spectrum over the course of 1 hour.27
4
ACS Paragon Plus Environment

Page 5 of 10
Journal of the American Chemical Society
at 30°C (a) and corresponding experiments in the presence of
H+
-H+
1
2
3
4
5
6
7
8
9
10
TEMPO (13 µM) and acetic acid (43 mM) (b), acetic acid (43
mM) and Oct4NBF4 (0.13 M) (c), acetic acid (43 mM) and HBF4
(43 mM) (d), and acetic acid (43 mM) with HBF4 (43 mM) added
at the time indicated by the arrow (e).
N
O
N
N
OH
OH
-HNO
-H+
Autoxidations at Elevated Temperatures. Given the effect
of weak acids on the RTA activity of nitroxides at ambient
temperatures, we wondered whether the addition of a weak
acid could bolster the activity of nitroxides at elevated temper-
atures as well. Consequently, we examined the effect of a ni-
troxide/acid combination on the autoxidation of paraffin oil in
a stirred flow reactor at 160°C.29,30 The reaction was initiated
with a small amount of tetralin hydroperoxide (5 mM) and a
dispersed stream of O2 was continuously passed through the
solution to stir the medium and prevent mass transfer of O2
from becoming rate limiting. Aliquots were removed at regu-
lar time intervals and the hydroperoxide concentration in each
was determined using our previously described pro-fluorescent
phosphine (1).31
Scheme 3. Proposed mechanism for the decomposition of
TEMPO+ in the presence of strong acid.
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Since the oxoammonium ion was stable under the reaction
conditions, we next considered the possibility that the key
electron transfer reaction (Eq. 11) was reversible. If so, the
strength of the acid and/or the nucleophilicity of its conjugate
base, are expected to influence the position of the key redox
equilibrium in Scheme 4.
RX
ROH
ROOH
X-
p-Tol
p-Tol
p-Tol
P
P
X-
p-Tol
O
+
(11)
R
N
O
Φf, ox/Φf = 10.2
+
N
O
R
O
O
O
O
O
O
1
-H+
O2
Since TEMPO is somewhat volatile at 160°C, a higher mo-
lecular weight analogue (2) was investigated, along with pal-
mitic acid in lieu of the acetic acid used at ambient tempera-
tures.32 The results are shown in Figure 6A.
R-H
ROO
Scheme 4. Reactions influencing the reversibility of the elec-
tron transfer between TEMPO+ and an alkyl radical.
Thus, the more acidic the medium, the less favourable will
be loss of a proton from the carbocation and/or the trapping of
the carbocation by the less nucleophilic conjugate base. In-
2
deed, if
a strong acid (e.g. HBF4) is added to a
Addition of the bis-nitroxide 2 to the paraffin autoxidation
yielded the expected sigmoidal profile of hydroperoxide pro-
duction in an inhibited autoxidation at elevated temperatures.33
Although inclusion of 4 mM palmitic acid to the reaction mix-
ture did not significantly improve the inhibitory activity of 2, a
higher loading (40 mM) produced a significant increase in the
inhibited period. At first glance, the lack of effect of the lesser
amount of acid and the fact that 2 is capable of inhibiting the
autoxidation in the absence of added acid, suggested that the
acid-catalyzed reaction mechanism accounts for only a frac-
tion of the activity of nitroxides at elevated temperatures.
However, the autoxidation of paraffinic hydrocarbons is
known to produce carboxylic acids (in addition to hydroperox-
ides) from the very early stages of the reaction.34,35 Therefore,
acids formed in situ may activate the nitroxide to reaction with
peroxyl radicals. To probe this possibility, we added 4 and 40
mM of the non-volatile, non-nucleophilic base 2,4,6-tri-tert-
butylpyridine (TTBP) to autoxidations inhibited by 2, and
found that it progressively diminished the RTA activity of the
nitroxide. Importantly, the addition of either acid or base had
no effect on the rate of the uninhibited autoxidation, suggest-
ing that these additives play no role other than modulating the
pH of the reaction (see Supporting Information for more data).
TEMPO/AcOH-inhibited autoxidation, the catalytic effect
disappears (Figure 5).28 Likewise, if an excess of the conjugate
base of the strong acid (e.g. Oct4NBF4) is added, the catalytic
effect decreases substantially (Figure 5). The latter result sug-
gests that the oxidation potential of TEMPO+ is also dependent
on the identity of the counterion.
Figure 5. O2 consumption during the autoxidation of styrene (4.3
M) in acetonitrile containing 1% H2O initiated by AIBN (0.05 M)
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
To confirm that carboxylic acids were present in the paraf- DISCUSSION
Page 6 of 10
1
2
3
4
5
6
7
8
fin autoxidations, they were determined using the aminated
coumarin derivative 3, whose fluorescence is greatly enhanced
(38-fold) upon protonation (see Supporting Information for
details).36 The results are shown in Figure 6B.
Despite the widespread use of HALS, their mechanism has
remained a mystery since the suggestion that nitroxides must
be the key intermediate in a catalytic radical-trapping cycle.6
The commonly presented mechanism (Scheme 1) requires the
nitroxide to compete with O2 for alkyl radicals, but it is well
known that the R• + O2 reaction is 20- to 30-fold faster than
the reaction of R• with TEMPO.9,21 This is irrelevant if the
TEMPO/O2 concentration ratio is large, but in many of the
experiments that established the unique reactivity of HALS
this condition is not fulfilled (likewise for the conditions under
which HALS-containing materials are manipulated and/or
used). For example, in the pioneering experiments carried out
by Bolsman, Blok and Frijns,5 which showed a stoichiometric
number of 510 for the inhibition of paraffin oil autoxidation,
only 1-4 µM TEMPO was used. Given that [O2] ~2 mM,38 the
rate of alkoxyamine formation must be at least 10,000-times
slower than the rate of peroxyl radical formation.
H+
H
N
N
Φf,H+/Φf = 38
O
O
O
O
O
O
3
9
The commercially-sourced light paraffin oil contained 0.14
mM carboxylic acid, and upon equilibrating the sample under
a N2 atmosphere and then heating to 160°C this increased to
0.5 mM.37 Addition of the initiator and changing the atmos-
phere to O2 lead to a rapid accumulatation of acid, reaching 15
mM within 1500 s. In the inhibited autoxidations, the produc-
tion of carboxylic acids mirrored the formation of hydroperox-
ides (albeit at lower absolute levels) - consistent with the po-
tent antioxidant activity of 2 and its diminution upon addition
of base to the autoxidation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nevertheless, if alkoxyamines are indeed formed in these
reactions, they must decompose to regenerate the nitroxide to
complete the catalytic cycle. Recent computations and a rein-
terpretation of literature data, led to the conclusion10 that it
must be via the sequence of Eq. (4), (5) and (6) in Scheme 1:
H-atom abstraction from the position adjacent the O-atom in
the alkoxyamine, followed by β-fragmentation of the resultant
alkyl radical to give an aminyl radical, and then oxidation to
give the nitroxide – as suggested earlier by Bolsman, Blok and
Frijns.11 Whilst it seems quite clear that this is the mechanism
of the reaction of an alkoxyamine with peroxyl radicals, it is
unlikely to be kinetically relevant in an inhibited autoxidation.
In fact, we have attempted to measure the rate constant for the
reaction of a model alkoxyamine (O-allyl-TEMPO) with per-
oxyl radicals using the radical clock approach and found that it
proceeds with k ≤ 1 M-1s-1 at 37°C, expectedly similar to H-
atom abstraction from the α-CH2 group of an ether.39 Assum-
ing logA = 8 for this H-atom transfer49 suggests that the rate
constant could be as big as ~70 M-1s-1 at 130°C, but given that
the alkoxyamine can be present at only 1-4 µM at most (as-
suming all of the TEMPO added in the first place is converted
to the alkoxyamine), this means that this reaction must be ca. 6
orders of magnitude too slow to compete with chain propaga-
tion (which has k[substrate] ~ 90 s-1)11,40 in Bolsman, Blok and
Frijns’ pioneering experiment.
Given our observation that TEMPO becomes an effectively
catalytic RTA in the presence of weak acid at room tempera-
ture, it seemed reasonable to consider this as an alternative to
the currently accepted mechanism. Indeed, the addition of acid
to a paraffin oxidation at 160°C improved the already catalytic
RTA activity of bis(nitroxide) 2, but more importantly, the
addition of base was able to wipe it out in a dose-dependent
manner (i.e. the activity of the nitroxide correlated with the
amount of acid present). This must rule out the mechanism in
Scheme 1, or any mechanism which invokes H-atom abstrac-
tion from the alkoxyamine in the turnover limiting step, since
it cannot account for these observations. Moreover, the
bis(oxoammonium) derived from 2 inhibited the paraffin oxi-
dation similarly to 2 itself. Therefore, we suggest that the
unique reactivity of nitroxides as RTAs (and HALS, in gen-
eral) is better accounted for by the mechanism shown in
Scheme 5.
Figure 6. Hydroperoxide (top) and acid (bottom) formation in
the autoxidation of light paraffin oil at 160°C initiated by 5
mM tetralin hydroperoxide (n) and inhibited by 40 µM 2 (‚)
with 4 mM palmitic acid („), 40 mM palmitic acid (¿), 4
mM 2,4,6-(t-Bu)3pyridine () or 40 mM
2,4,6-(t-
Bu)3pyridine (l).
6
ACS Paragon Plus Environment

Page 7 of 10
Journal of the American Chemical Society
Although the secondary amines found in HALS are far less
1
2
3
4
5
6
7
8
reactive toward peroxyl radicals than diarylamines, HALS
have the distinct advantage that the nitroxide intermediate in
their catalytic cycle is persistent and does not undergo delete-
rious reactions with peroxyl radicals. The same cannot be said
for the nitroxides derived from diarylamines, which undergo
competing reactions with peroxyl radicals at ring carbon at-
oms,3 leading to complex mixtures of products,44 and preclud-
ing continuation of the cycle in Scheme 5.
9
O
N
O
N
10
11
12
13
14
15
16
(12)
+
ROO
ROO
H
This fact underscores the unique reactivity of HALS, but al-
so prompts the question: what eventually leads to exit from the
nitroxide’s catalytic cycle? Why are nitroxides not infinitely
reactive? The answer to this question may provide the key to
the design of the ultimate hindered amine light stabilizers.
Scheme 5. Proposed mechanism of acid-catalyzed RTA ac-
tivity of nitroxides.
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although this mechanism still requires the participation of
an alkyl radical in the catalytic cycle, the reaction it undergoes
is an electron transfer (to regenerate the nitroxide from the
oxoammonium ion), which has been shown to compete effec-
tively with its combination with O2 (vide supra).
CONCLUSIONS
For the past 40 years there has been widespread agreement
that nitroxides are key intermediates in the catalytic activity of
hindered amine light stabilizers, but the precise mechanism
has been elusive. The formation of an alkoxyamine, and its
subsequent decomposition to reform a nitroxide, was believed
to comprise the catalytic cycle. However, for the reasons dis-
cussed above, this cannot account for the reactivity of nitrox-
ides as autoxidation inhibitors under most conditions. Instead,
we have advanced the more plausible proposal that 1) protona-
tion of the nitroxide creates a potent formal H-atom donor
which reacts with a peroxyl radical,14 and 2) the resultant oxo-
ammonium ion oxidizes an alkyl radical, regenerating the ni-
troxide. The latter electron transfer has been shown to be
competitive with O2 addition in both model systems and au-
thentic autoxidations (k11 = 1-3 ×1010 M-1s-1). This pathway
provides a more plausible explanation for the catalytic activity
of nitroxides as RTAs and also supports a role for this electron
transfer reaction in TEMPO-mediated synthetic transfor-
mations. The ability of nitroxides to inhibit the autoxidation of
hydrocarbons at high temperature is shown to depend on acid
formation in situ. We have shown that the proposed catalytic
cycle is also fully operative at ambient temperatures, suggest-
ing that it may be of biological relevance and may contribute
to the biological activity of TEMPO and related nitroxides.7
Diarylamines are also common additives to petroleum-
derived products, but because their oxidation products are
highly coloured, they are generally not used in applications
where colour changes are undesirable (e.g. in lightly coloured
plastics). Given that diarylamines are secondary amines whose
catalytic activity also depends on the formation of a nitroxide
intermediate (cf. Scheme 6),41 it is possible that the mechanism
in Scheme 5 could also contribute to their activity.42,43 Howev-
er, it is expected that: 1) alkoxyamine formation will be slight-
ly faster for Ar2NO• + R• than for TEMPO• + R• due to re-
duced steric hindrance, 2) unimolecular decomposition of the
resultant Ar2NOR to regenerate Ar2NH has a lower barrier
than the return of the nitroxide,43 3) the diaryl nitroxide will be
less basic than TEMPO, and 4) the diaryloxoammonium ion
will be a weaker oxidant than the TEMPO+. Moreover, in re-
lated work, we found that acid formation diminishes the activi-
ty of some Ar2NH.30
EXPERIMENTAL SECTION
Reagents and solvents were purchased from commercial
sources and used without further purification unless otherwise
indicated. TEMPO+BF4 , azocumene,46 bis(nitroxide) 1,47
– 45
phosphine 231 and BODIPY 336 were synthesized by literature
precedent. Oxidizable substrates (cumene, styrene, paraffin
oil) were carefully purified from stabilizers and hydroperox-
ides before use and were stored refrigerated under nitrogen.
Scheme 6. Proposed mechanism of catalytic activity of dia-
rylamine RTAs.
Model Reactions. TEMPO+BF4– (12 mg, 50 µmol) and tet-
ramethylammonium acetate (6.7 mg, 50 µmol) were added to a
3 mL vial. The contents were dissolved in acetonitrile (sparged
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 10
with either argon or oxygen if necessary) and placed under the the plate was stirred for 28 s, after which a solution of fluoro-
1
2
3
4
5
6
7
8
appropriate atmosphere. To this, azocumene (13 mg, 50 µmol)
was added. The contents were heated and stirred at 50°C. Ali-
quots (100 µL) were removed every 40 minutes and diluted to
1 mL with methanol containing 2.8 mM hexylbenzene. Sam-
ples were analyzed on a Waters Acquity H-Class UPLC
equipped with a XTerra® RP-18 Column (5 µM × 4.6 mm ×
150 mm) using 20% H2O in methanol at a flow rate of 0.4
mL/min and detection at 216 nm.
genic phosphine 1 (20 µL of a 250 µM stock solution in ace-
tonitrile) was injected and the plate stirred for an additional 8
seconds and then allowed to rest for 2 more seconds. The fluo-
rescence of each well was measured every 1.3 seconds for 30
seconds (excitation = 340 nm; emission = 425 nm). The con-
centration of hydroperoxide was determined from the rate of
phosphine oxidation using the rate constant for the reaction of
the dye with secondary hydroperoxides in tert-amyl alcohol
(1.2 M-1 s-1) assuming pseudo-first-order kinetics.31
Inhibited Autoxidations. Experiments were performed in a
two-channel oxygen uptake apparatus, based on a Validyne
DP 15 differential pressure transducer.48 In a typical experi-
ment, an air-saturated solution of either styrene or cumene
containing AIBN was equilibrated with an identical reference
solution containing also excess of 2,2,5,7,8-pentamethyl-6-
hydroxychromane (PMHC) (25 mM). After equilibration, and
when a constant O2 consumption was reached, a stock solution
9
In tandem, four duplicates (1.2 µL) of each sample were
loaded into separate wells of a 96-well microplate and the
automated reagent dispenser of the microplate reader used to
dilute each sample with 25% isopropyl alcohol in methanol
(280 µL) and the plate was stirred for 15 seconds. Afterward, a
solution of fluorogenic amine 3 (20 µL of a 630 µM stock in
methanol) was added. The plate was stirred for 8 seconds and
the fluorescence of each well was measured every second for
10 s (excitation = 315 nm; emission = 395 nm).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-
of TEMPO or TEMPO+BF4 (final concentration: 5-50 µM)
and a stock solution of the desired acid were injected in the
sample flask. The oxygen consumption in the sample was
measured, after calibration of the apparatus, from the differen-
tial pressure recorded with time between the two channels.
Rates of initiation, Ri, were determined for each conditions in
preliminary experiments by the inhibitor method using PMHC
as reference antioxidant: Ri = 2 [PMHC] / τ, where τ is the
length of the induction period. Kinetic traces were analyzed by
numerical fitting using GEPASI software.
ASSOCIATED CONTENT
Supporting Information
Synthetic details and results of model studies, inhibited autoxida-
tions, EPR experiments, and paraffin autoxidations referred to, but
not included in, the manuscript. The Supporting Information is
available free of charge on the ACS Publications website.
EPR measurements. EPR spectra were recorded in a X-
band spectrometer equipped with temperature control, with the
following settings: microwave power 6.3 mW, time constant
and conversion time 20.48 ms, modulation amplitude 1.0 G.
Typical samples contained TEMPO+BF4 (0.5 - 5 mM), AIBN
(0.05 M) and styrene or cumene (40-50% v/v) in MeCN or
PhCl and were air-saturated. Samples were transferred in a
capillary glass tube and were placed in the thermostatted in-
strument cavity at 303 K, so to cover the entire sensitive area.
CH3COOH and CF3COOH (30-50 mM), where added to the
sample as indicated. Absolute concentration of TEMPO
formed in the sample was determined upon calibration with
standard solutions in the corresponding solvent mixture. Time
evolution of the EPR signal was monitored by an automated
acquisition routine and analyzed using either the intensity of
the first spectral line or the double integration of the whole
spectrum.
AUTHOR INFORMATION
Corresponding Authors
dpratt@uottawa.ca
luca.valgimigli@unibo.it
ACKNOWLEDGEMENTS
This work was supported by grants from the Natural Sciences and
Engineering Research Council of Canada, the Canada Foundation
for Innovation, the Italian Ministry of Education (MIUR, Rome;
project 2010PFLRJR, PROxi) and the University of Bologna
(FARB project FFBO123154). DAP acknowledges support from
the University of Ottawa and the Canada Research Chairs pro-
gram, respectively. EAH acknowledges the support of an Ontario
Graduate Scholarship. We thank Prof. Claudio Zannoni and Prof.
Alberto Arcioni for access to EPR equipment, and Dr. Isabella
Miglioli for assistance with EPR experiments.
Paraffin Autoxidations. Paraffin autoxidations were carried
out using an analogous procedure to hexadecane autoxidations
previously carried out by our group.30,31 Light paraffin oil (100
mL) was thoroughly degassed with N2 and then heated to
160°C in a stirred flow reactor. Once the temperature stabi-
lized, 0.04 mmol of bis(2,2,6,6-tetramethylpiperidin-1-oxyl-4-
yl) decanedioate (2), the appropriate amount of either palmitic
acid or 2,4,6-tri-tert-butyl pyridine, and 82 mg (0.5 mmol) of
tetralin hydroperoxide49 were added to the solution. The flow
of N2 was replaced with 1 L/min O2. Aliquots (0.5 mL) were
removed every 5 minutes and allowed to cool to room temper-
ature for analysis.
REFERENCES
1 Ingold, K. Chem. Rev. 1961, 61, 563–589.
2
Ingold, K. U.; Pratt, D. A. Chem. Rev. 2014, 114, 9022–9046.
3
Brownlie, I. T.; Ingold, K. U. Can. J. Chem. 1967, 45, 2427–2432.
4
Sudnik, M. V.; Romantsev, M. F.; Shapiro, A. B.; Rozantsev, E. G.
Bull. Acad. Sci. USSR 1976, 24, 2702-2704.
5
Bolsman, T.; Blok, A.; Frijns, J. Recl. des Trav. Chim. des Pays-Bas
1978, 97, 310–312.
6
See the following reviews on the subject: (a) Allen, N. S. Chem. Soc.
Rev. 1986, 15, 373-404; (b) Gugumus, F. Polym. Degrad. Stab. 1993, 40,
167−215; (c) Pospisil, J. Adv. Polym. Sci. 1995, 124, 87−189.
Four duplicates (30 µL) of each sample were loaded into a
96-well microplate. Each well was read sequentially following
manipulation as follows: an automated reagent dispenser was
used to dilute the sample with tert-amyl alcohol (200 µL) and
8
ACS Paragon Plus Environment

Page 9 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
7
20
See for example: (a) Samuni, A. M.; Goldstein, S.; Russo, A.;
The induction period observed in the alkoxyamine formation is
due to the necessary accumulation of TEMPO to outcompete the oxo-
ammonium ion for the cumyl radicals.
Mitchell, J. B.; Krishna, M. C.; Neta, P. J. Am. Chem. Soc. 2002, 124,
8719-8724; (b) Goldstein, S.; Merenyi, G.; Russo, A.; Samuni, A. J. Am.
Chem. Soc. 2003, 125, 789–795; (c) Borisenko G. G.; Martin I.; Zhao Q.;
Amoscato A. A.; Kagan, V. E. J. Am. Chem. Soc. 2004, 126, 9221–9232;
(d) Wipf, P.; Xiao, J.; Jiang, J.; Belikova, N. A.; Tyurin, V.; Fink, M. P.;
Kagan, V. E. J. Am. Chem. Soc. 2005, 127, 12460-12461; (e) Dikalova, A.
E., Bikineyeva, A. T., Budzyn, K., Nazarewicz, R. R., McCann, L., Lewis,
W., Harrison, D. G., and Dikalov, S. I. Circ. Res. 2010, 107, 106−116; (f)
Ji, J., Kline, A. E., Amoscato, A., Samhan-Arias, A. K., Sparvero, L. J.,
Tyurin, V. A., Tyurina, Y. Y., Fink, B., Manole, M. D., Puccio, A. M.,
Okonkwo, D. O., Cheng, J. P., Alexander, H., Clark, R. S. B., Kochanek,
P. M., Wipf, P., Kagan, V. E., and Bayir, H. Nature Neurosci. 2012, 15,
1407−1413; (g) Canistro, D.; Boccia, C.; Falconi, R.; Bonamassa, B.;
Valgimigli, L.; Vivarelli, F.; Soleti, A.; Genova, M. L.; Lenaz, G.; Sapone,
A.; Zaccanti, F.; Abdel-Rahman, S. Z.; Paolini, M. J. Gerontol. A Biol. Sci.
Med. Sci. 2015, 70, 936-943.
21
Maillard, B.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc. 1983,
105, 5095–5099.
22
Interestingly, corresponding reactions carried out in the absence of
Me4NAcO yielded lower amounts of alkoxyamine were observed. The
effect of couterions to the oxoammonuim ion will be discussed below.
23
Burton, G. W.; Doba, T.; Gabe, E.; Hughes, L.; Lee, F. L.; Prasad,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
L.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 7053-7065.
24
Calculated from the inhibition time in O2 uptake auotxidations in-
hibited by PMHC.
25
In the absence of acid the initial rate is noticeably slower and non-
linear – increasing with time, perhaps due to depletion of O2 over the
longer period of time and improving the competition with alkyl radicals
– see Figure S16 in the Supporting Information.
8
26
Kovtun, G. A.; Aleksandrov, A. L.; Golubev, V. A. Bull. Acad. Sci.
Ma, Y.; Loyns, C.; Price, B.; Chechik, V. Org. Biomol. Chem., 2011, 9,
USSR 1974, 22, 2115-2121.
9
5573-5578.
27
(a) Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1992, 114, 4992–
It should be pointed out that the spectral characteristics ascribed to
TEMPO+ by Ma et al.26 do not correspond to those reported by other
investigators, e.g. Goldstein et al.7b or those determined in our own
laboratory. See ref. 7b for a discussion.
4996; (b) Sobek, J.; Martschke, R.; Fischer, H. J. Am. Chem. Soc. 2001,
123, 2849–2857.
10
Gryn’ova, G.; Ingold, K. U.; Coote, M. L. J. Am. Chem. Soc. 2012,
28
Numerical fitting of styrene autoxidations in MeCN (30°C) inhibit-
134, 12979–12988.
11
ed by TEMPO (or TEMPO+) in the presence of TFA (4.3 mM) and
HBF4 (4.3 mM) afforded k11 ~ 1×109 M-1s-1 and < 109 M-1s-1, respective-
Bolsman, T.; Blok, A.; Frijns, J. Recl. des Trav. Chim. des Pays-Bas
1978, 97, 313–319.
12
ly.
This can be found throughout the literature, in particular the two
29
Jensen, R. K.; Korcek, S.; Mahoney, L. R.; Zinbo, M. J. Am. Chem.
areas where nitroxide radical chemistry is particularly important: radical
polymerization (Matyjaszewski, K.; Davis, T. P. Handbook of Radical
Polymerization; Wiley: Hoboken, NJ, 2002) and spin-labeling (Kochergin-
sky, N.; Swartz, H. M. Nitroxide Spin Labels: Reactions in Biology and Chemis-
try; CRC Press: Boca Raton, FL, 1995).
Soc. 1979, 101, 7574–7584.
30
Shah, R.; Haidasz, E. A.; Valgimigli, L.; Pratt, D. A. J. Am. Chem.
Soc. 2015, 137, 2440–2443.
31
Hanthorn, J.; Haidasz, E.; Gebhardt, P.; Pratt, D. Chem. Commun.
13
2012, 48, 10141–10143.
32
Amorati, R.; Pedulli, G. F.; Pratt, D. A.; Valgimigli, L. Chem.
Commun. (Camb). 2010, 46, 5139–5141.
14
A sample of TEMPO heated in a vial to 160°C lost ca. 60% of its
mass over the course of 1 hour.
33
The role of an acid-mediated disproportionation of TEMPO to
produce the corresponding hydroxylamine was ruled out on kinetic
grounds (see reference 13). However, it is also possible that the reaction
takes place by a coupled proton-electron transfer wherein a proton is
transferred from the acid to the peroxyl radical concerted with the
movement of an electron from TEMPO to the peroxyl. Since we found
that there is a primary kinetic isotope effect on these reactions, a H-
atom must be in flight in the RDS of the reaction, precluding any rapid
pre-equilibrium between peroxyl and protonated peroxyl, such as that
which is involved when phenols react with peroxyl radicals under acidic
conditions, see: Valgimigli, L.; Amorati, R.; Petrucci, S.; Pedulli, G. F.;
Hu, D.; Hanthorn, J. J.; Pratt, D. A. Angew. Chem. Intl. Ed. 2009, 48,
8348-8351.
This arises due to the three distinct phases observable in the initial
part of the reaction; the inhibited autoxidation, the uninhibited autoxida-
tion where propagation proceeds at a constant rate and the slowing of
propagation in favour of termination with increasing concentration of
hydroperoxides, which increase radical concentration upon thermolysis.
Similar data was obtained when the corresponding bis-oxoammonium
ion was added in lieu of 2 (see Supporting Information).
34
Jensen, R. K.; Korcek, S.; Mahoney, L. R.; Zinbo, M. J. Am. Chem.
Soc. 1981, 103, 1742–1749.
35
Jalan, A.; Alecu, I. M.; Meana-Pañeda, R.; Aguilera-Iparraguirre, J.;
Yang, K. R.; Merchant, S. S.; Truhlar, D. G.; Green, W. H. J. Am. Chem.
Soc. 2013, 135, 11100–11114.
36
15
Sheldon, R. A.; Arends, I. W. C. E.; Brink, G.-J.; Dijksman, A. Acc.
Shah, R.; Pratt, D. A. In preparation.
Chem. Res. 2002, 35, 774–781.
16
37
This increase is likely the result of thermal decomposition of γ-
(a) Bailey, W. F.; Bobbitt, J. M.; Wiberg, K. B. J. Org. Chem. 2007,
dihydroperoxides present at concentration below those reliably deter-
72, 4504–4509; (b) Hoover J. M.; Ryland B. L.; Stahl S. S.; J. Am. Chem.
Soc., 2013, 135, 2357–2367; (c) Ryland B. L.; McCann S. D.; Brunold, T.
C.; Stahl S. S. J. Am. Chem. Soc., 2014, 136, 12166–12173.
mined by our method (ca. 1 mM).
38 Bolsma, Blok and Frijns suggest [O2]~10-3 mM in paraffin at 130°C
under an atmosphere of O2, but do not provide a reference or details of
the measurement.11 Korcek and co-workers report [O2] = 2.65 mM in
17
Denisov, E. T. Polym. Degrad. Stab. 1991, 34, 325–332.
18
Baran has invoked this electron transfer in a mechanistic proposal
hexadecane at 160°C under an atmosphere of O2, which they deter-
mined directly, see: Jensen, R. K.; Korcek, S.; Zinbo, M. Oxid. Commun.
1990, 13, 258-262.
for the guided desaturation of unactivated alkanes. See: Voica, A.-F.;
Mendozza, A.; Gutekunst, W. R.; Fraga, J. O.; Baran, P. S. Nature Chem.
2012, 4, 629-635.
39
Up to 2.7 M of 2,2,6,6-tetramethyl-1-(2-propenyloxy)-piperidine
19
Montgomery, J. A.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A.
had no detectable influence on the product distribution in a methyl
linoleate clocking experiment (carried out as described in Roschek, B.;
Tallman, K. A.; Rector, C. L.; Gillmore, J. G.; Pratt, D. A.; Punta, C.;
Porter, N. A. J. Org. Chem. 2006, 71, 3527–3532). The maximum rate
J. Chem. Phys. 1999, 110, 2822-2827.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 10
1
2
3
4
5
6
7
8
9
constant for transfer from the H-atom donor under these conditions is 1
M-1s-1. For comparison, the propagation rate constant for H-atom trans-
fer from isopropyl tert-butyl ether (a reasonably similarly hindered ether)
has been measured as kp = 0.02 M-1s-1, see: Howard, J. A.; Ingold, K. U.
Can. J. Chem. 1970, 48, 873–880.
40
Bennett, J. E.; Eyre, J. A.; Summers, R. J. Chem. Soc. Perkin Trans. 2
1974, 797–800.
41
Adamic, K.; Dunn, M.; Ingold, K. U. Can. J. Chem. 1969, 47, 287–
294.
42
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jensen, R. K.; Korcek, S.; Zinbo, M.; Gerlock, J. L. J. Org. Chem.
1995, 60, 5396–5400.
43
Haidasz, E. A.; Shah, R.; Pratt, D. A. J. Am. Chem. Soc. 2014, 136,
16643–16650.
44 Pospisil, J. Polymer Degrad. Stab. 1991, 34, 85-109.
45 Holan, M.; Jahn, U. Org. Lett. 2014, 16, 58–61.
46
Ikeda, H.; Hoshi, Y.; Namai, H.; Tanaka, F.; Goodman, J. L.;
Mizuno, K. Chem. - A Eur. J. 2007, 13, 9207–9215.
47
Valgimigli, L.; Pedulli, G. F.; Paolini, M. Free Radic. Biol. Med. 2001,
31, 708-716.
48 Amorati, R.; Pedulli, G. F.; Valgimigli, L. Org. Biomol. Chem., 2011, 9,
3792–3800
49 Knight, H. B.; Swern, D. Org. Synth. 1954, 34, 90-91.
50
Benson, S. W. Thermochemical Kinetics, 2nd Ed.; Wiley: New York,
NY, 1976.
TOC Graphic:
10
ACS Paragon Plus Environment