The catalytic mechanism of diarylamine radical-trapping antioxidants

Article abstract of DOI:10.1021/ja509391u

Diarylamine radical-trapping antioxidants are important industrial additives, finding widespread use in petroleum-derived products. They are uniquely effective at elevated temperatures due to their ability to trap multiple radicals per molecule of diaryla

Full text of DOI:10.1021/ja509391u

Subscriber access provided by University of Kansas Libraries
Article
The Catalytic Mechanism of Diarylamine Radical-Trapping Antioxidants
Evan A Haidasz, Ron Shah, and Derek A. Pratt
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja509391u • Publication Date (Web): 30 Oct 2014
Downloaded from http://pubs.acs.org on October 31, 2014
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 9
Journal of the American Chemical Society
1
2
3
4
5
6
The Catalytic Mechanism of Diarylamine Radical-Trapping Antioxidants
7
8
Evan A. Haidasz, Ron Shah and Derek A. Pratt*
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
Department of Chemistry, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada
dpratt@uottawa.ca
Supporting Information Placeholder
ABSTRACT: Diarylamine radical-trapping antioxidants are important industrial additives, finding widespread use in petroleum-
derived products. They are uniquely effective at elevated temperatures due to their ability to trap multiple radicals per molecule of
diarylamine. Herein we report the results of computational and experimental studies designed to elucidate the mechanism of this
remarkable activity. We find that the key step in the proposed catalytic cycle – decomposition of the alkoxyamine derived from
capture of a substrate-derived alkyl radical with a diarylamine-derived nitroxide – proceeds by different mechanisms depending on
the structure of both the substrate and the diarylamine. N,N-Diarylalkoxyamines derived from saturated substrates and diphenyla-
mines decompose by N-O homolysis followed by disproportionation, whereas those derived from unsaturated substrates and diphe-
nylamines, or saturated substrates and N-phenyl-β-naphthylamine, decompose by an unprecedented concerted retro-carbonyl-ene
reaction. The alkoxyamines that decompose by the concerted process inhibit hexadecane autoxidations at 160°C to the same extent
as the corresponding diarylamine, whereas those alkoxyamines that decompose by the N-O homolysis/disproportionation pathway
are much less effective. This suggests that competing cage-escape of the alkoxyl radicals following N-O homolysis leads to signifi-
cantly less effective regeneration of diarylamines and imply that the catalytic efficiency of diarylamine antioxidants is substrate-
dependent. The results presented here have significant implications in the future design of antioxidant additives: diarylamines de-
signed to yield intermediate alkoxyamines that undergo the retro-carbonyl-ene reaction are likely to be more effective than existing
compounds, and will display catalytic radical-trapping activities at lower temperatures due to lower barriers to regeneration.
INTRODUCTION
mine upon in-cage disproportionation of the diarylaminyl and
alkoxyl radicals.7 This idea was supported by the observation
that heating of a mixture of alkoxyamines generated in situ from
decomposition of tert-butyl peroxy-2-ethylhexanoate in hexade-
cane in the presence of crude bis(p-octylphenyl)nitroxide gave
64% bis(p-octylphenyl)amine (among other products).5 No fur-
ther scrutiny of this mechanism is evident in the literature and
no significant improvement in the chemistry behind diaryla-
mine RTA technology has emerged since the 1950s.
Diarylamines are among the most commonly used radical-
trapping antioxidants (RTAs); compounds which inhibit the
free radical chain oxidation of hydrocarbons by trapping chain-
carrying peroxyl radicals (Eq. 1 and 2).1,2
Ar2NH + ROO• → Ar2N• + ROOH
Ar2N• + ROO• → non-radical products
(1)
(2)
While diarylamines display similar efficacies to more prolific
phenolic RTAs at ambient temperatures (typical k1 values of 104-
105 M-1s-1 and stoichiometric coefficients, n, of 2 based on Eq. 1
and 2),3 they have uniquely high efficacies at elevated tempera-
tures (>120°C), where one equivalent of diarylamine can trap
several chain-carrying peroxyl radicals. For example, as early as
1978 an unbelievable stoichiometry of n = 41 was reported from
a diphenylamine-inhibited autoxidation of paraffin oil at
130°C.4 Some time ago, Korcek and coworkers proposed a
mechanism accounting for these dramatic observations, shown
in Scheme 1.5 The large stoichiometric coefficients were as-
cribed to the observed regeneration of the diarylamine in situ
during hexadecane autoxidation at 160°C inhibited by the cor-
responding diarylnitroxide. Furthermore, the nitroxide proved
to be as powerful an RTA as its parent amine. The amine was
believed to be formed from a substrate-derived alkoxyamine –
therefore making use of the substrate as a stoichiometric reductant.6
Elevated temperatures were presumed necessary to cleave the N-
O bond of the alkoxyamine, which could lead to the diaryla-
Scheme 1. Proposed mechanism of catalytic activity of diaryla-
mine RTAs.
In recent years we have reported on the design and development
of novel heterocyclic diarylamines,8-10 some of which were found
to react up two orders of magnitude faster with peroxyl radicals
than conventional compounds at ambient temperatures (i.e. k1 >
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
107 M-1s-1) thanks to their comparatively weaker N-H bonds (ca.
from the π-system to the radical, consistent with its description
as a proton coupled electron transfer (PCET) reaction.8,12,13 The
TS structure is shown in Figure 2 along with the highest (dou-
bly) occupied molecular orbital, which clearly shows electron
delocalization from the π-HOMO of diphenylamine to the π*-
LUMO of the methylperoxyl radical. The enthalpic barrier to
the reaction was calculated to be 5.0 kcal/mol, and the corre-
sponding free energy barrier (ΔG‡ = 16.2 kcal/mol) yields a rate
constant of 1.8×104 M-1s-1 by application of transition state theo-
ry (at 25°C), in good agreement with the experimental result of
1.5×104 M-1s-1 (at 50°C).3 It should be pointed out that the TS is
preceded by a H-bonded pre-reaction complex, and followed by
a H-bonded post-reaction complex with enthalpies of -4.2 and
-9.0 kcal/mol, respectively, relative to the separated starting
materials (not shown in Figure 1). The phenyl rings in both the
TS and product diphenylaminyl radical are oriented to maxim-
ize the delocalization of the radical throughout the π-system,
although repulsion between the ortho hydrogens precludes com-
plete planarization of the two rings.14 Addition of p-methyl sub-
stituents to the rings drops the barrier 1.3 kcal/mol, corre-
sponding to a 10-fold increase in the rate constant at 25°C (k =
1.5×105 M-1s-1), in good agreement with experiment (k = 1.8×105
M-1s-1 for 4,4’-dioctyldiphenylamines at 37°C).8,9 The substituent
effect can be understood on thermodynamic grounds by the
greater stability of the diarylaminyl radical,15,16 and on kinetic
grounds by the reduced energy gap between the higher energy π-
HOMO of the substituted diphenylamine and π*-LUMO of the
peroxyl radical.
1
2
3
4
5
6
7
8
6 kcal/mol relative to diphenylamine). We anticipated that they
would also be more efficient than conventional compounds at
elevated temperatures since homolysis of the N-O bond in the
intermediate alkoxyamines would also be much more facile.
However, before setting out to investigate the properties of the
new compounds any further, we sought more information
about the mechanism proposed by Korcek and co-workers, and
present the results of our computational and experimental stud-
ies here.
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
RESULTS & DISCUSSION
Computations. The structures and energetics of the partici-
pants in the proposed catalytic cycle shown in Scheme 1 were
computed at the CBS-QB3 level of theory11 for the reaction
between
either
diphenylamine
(1)
or
4,4’-
dimethyldiphenylamine (2) and
a
model peroxyl radical,
methylperoxyl (MeOO•). The former was chosen as it is the
parent diarylamine, and the latter was chosen as a model for the
industry-standard dialkylated diphenylamines. The key station-
ary points and their relative enthalpies were determined at 25°C
and are shown in Figure 1.
The initial step, abstraction of the labile H-atom of diphenyla-
mine by a peroxyl radical (also Eq. 1) proceeds through a transi-
tion state characterized by proton transfer (roughly) within the
plane of the aromatic rings, and simultaneous electron transfer
Figure 1. CBS-QB3-calculated enthalpies (at 25°C) for relevant structures in Scheme 1.
2
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
N
1
+
MeOO
2
3
4
5
-28.6
kcal/mol
-33.0
kcal/mol
-32.0
kcal/mol
6
7
8
OCH3
OCH3
O
N
O
N
N
9
O
10
11
12
OCH3
14.7
kcal/mol
41.3
kcal/mol
41.1
kcal/mol
Figure 2. (A) Calculated TS structure for the reaction between
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
diphenylamine and a methylperoxyl radical, and (B) the highest
(doubly) occupied molecular orbital.
O
O
N
N
N
Coupling of the diphenylaminyl radical to a methylperoxyl to
yield an alkylperoxyamine is 32.0 kcal/mol downhill (enthalpi-
cally) and expected to be diffusion-controlled. N-O bond for-
mation must compete with addition of the peroxyl radical to the
aryl ring (the preferred pathway for reactions of peroxyl radicals
with phenoxyl radicals17) as is shown in Scheme 2. The para and
ortho adducts have C-O BDEs of 33.0 and 28.6 kcal/mol, re-
spectively. While, at first glance, it would seem that the pre-
ferred reaction of peroxyl radicals with diphenylaminyl radicals
should therefore follow the para-coupling pathway, the follow-up
reactions need to be considered. In the alkylperoxyamine inter-
mediate the O-O bond is much weaker than the N-O bond
(14.7 kcal/mol vs. 32.0 kcal/mol), cleaving to yield the nitroxide
and alkoxyl radical, whereas in the para-coupled adduct, the C-
O bond is much weaker than the O-O bond (33.0 kcal/mol vs.
41.3 kcal/mol), suggesting that at elevated temperatures, the
intermediates will be funneled towards the thermodynamic
product: the nitroxide. (Should the O-O bond in the para- or
ortho-coupled adduct be cleaved, in-cage disproportionation of
the two oxygen-centered radicals is likely to proceed to give an
N-phenyl iminoquinone. Alternatively, tautomerization of the
adduct may occur, followed by very rapid O-O dissociation and
in-cage disproportionation, to afford the same products.)30
O
Scheme 2. Competing pathways (and associated ΔH values) for
Ph2N• + •OOMe.
Combination of an alkyl (methyl) radical to the diphenylnitrox-
ide is exothermic (53.3 kcal/mol)19, and the N-O bond dissocia-
tion enthalpy in the resultant alkoxyamine is 42.1 kcal/mol (ΔG
= 29.5 kcal/mol),31 making it the rate-determining step in the
regeneration of the diarylamine from its diarylaminyl radical
upon disproportionation. Interestingly, we were able to locate a
transition state for concerted N-O dissociation and dispropor-
tionation. The enthalpic cost to reach the transition state (ΔH‡
= 39.4 kcal/mol) was slightly less than the N-O bond dissocia-
tion enthalpy, and given the small entropic cost associated with
the process (TΔS‡ ~ 0 kcal/mol at 298 K), suggested that the
concerted pathway may be accessible. However, we subsequently
found that this stationary point was unstable with respect to
RHF à UHF transformation,20 and while a stable UHF wave-
function could be obtained corresponding to the singlet biradi-
cal, no concerted TS structure could be located on this surface,
with the structure collapsing directly to products.
While struggling with the foregoing calculations, we wondered
whether another concerted process could intervene: a concerted
6-electron pericyclic process – an unprecedented retro-carbonyl-
ene (hereafter RCE) reaction – which would give the same car-
bonyl product and an imino-tautomer of diphenylamine:
The structure of the diphenylnitroxide is similar to that of the
diphenylaminyl radical. Nitroxides can have up to ca. 30% of
their spin density on the nitrogen atom,18 and the aromatic
rings in the diphenylnitroxide are oriented to maximize delocal-
ization of the radical while minimizing the repulsion associated
with planarizing the rings. Though the nitroxides derived from
1 and 2 are comparatively stable (the corresponding hydroxyla-
mines have O-H BDE’s of 71.4 and 70.8 kcal/mol, respectively),
they can potentially react with peroxyl radicals via their aryl
rings. (The para and ortho adducts from coupling diphe-
nylnitroxide and methylperoxyl have C-O BDEs of 20.9 and
19.6 kcal/mol, respectively. While they are expected to fragment
back to the nitroxide at elevated temperatures, competing tau-
tomerization and subsequent O-O bond cleavage will yield N-
oxides of the same N-phenyl iminoquinone products that arise
from addition of peroxyl radicals to the rings of diphenylaminyl
radicals, vide supra.)
R
R
R'
O
R'
O
H
H
(3)
N
N
Indeed, we were able to find a transition state structure corre-
sponding to this process (cf. Figure 3), and its enthalpic barrier,
ΔH‡ = 34.3 kcal/mol, was almost 8 kcal/mol lower than the N-
O BDE. The HOMO of the TS structure is a mixture of the π-
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
lowing the precedented diimide reduction of the 1,3-
HOMO of the imino-tautomer of diphenylamine and the π*-
LUMO of formaldehyde, the expected frontier MOs in the 6-
electron RCE pericyclic process. The low entropy of activation
of TΔS‡ = 1.4 kcal/mol at 298 K, suggests that this process could
be competitive with N-O bond homolysis.
cyclohexadiene-derived endoperoxide.23 With these compounds
in hand, we set out to determine the kinetics and mechanism of
their decomposition by HPLC.
1
2
3
4
5
6
7
8
O
OH
a
c
Ar
N
4
, Ar = Ph
O
5, Ar = p-tBuPh
Ar
Ar
O
O
b
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
O
O
OH
Ar
N
c
6
, Ar = Ph
7, Ar = p-tBuPh
Scheme 3. a. O2, hν, Rose Bengal, MeOH/CH2Cl2, 0°C, 6 hr,
46%; b. N2(CO2)2K2, AcOH, MeOH, 0°C, 2 hr, 46%; c. Ar2NH,
BuLi, Et2O, -78°C, 10 min, 30% (4), 22% (5), 20% (6), 15% (7).
In the event, the unsaturated alkoxyamine decomposed smooth-
ly (cf. Figure 4A), and roughly six times faster than the saturated
alkoxyamine at 120°C, as monitored directly by HPLC (k =
2.4×10-4 s-1 vs. k = 4.1×10-5 s-1). Product analysis revealed near
quantitative formation of diphenylamine in both cases. Howev-
er, subsequent experiments carried out at a variety of tempera-
tures between 75 and 165°C revealed markedly different Ar-
rhenius parameters for the two substrates (cf. Figure 4B). The
decomposition of the saturated alkoxyamine 6 was characterized
by Ea = 35.1±0.7 kcal/mol and logA = 15.1±0.4, whereas the
unsaturated alkoxyamine 4 decomposed with Ea = 30.2±0.7
kcal/mol and logA = 13.2±0.4. These results hinted at different
processes; the larger logA consistent with a simple (N-O) bond
dissociation in the transition state,24 and the smaller logA con-
sistent with more organization in the transition state, perhaps
due to concerted C-H cleavage such as in the computed RCE
reaction. The alkoxyamines derived from the 4,4’-di-tert-
butyldiphenylamine, 5 and 7, decomposed faster than their
diphenylamine-derived counterparts, 4 and 6, by 2.4- and 3.7-
fold at 120°C, respectively. The activation energies for the sub-
stituted compounds (30.3±0.3 kcal/mol for 5 and 33.5±0.2
kcal/mol for 7) were comparatively lower, and also showed that
the unsaturated alkoxyamine 5 had a significantly lower logA
(13.6±0.2) than the saturated compound alkoxyamine 7 (14.8±
0.1), indicating that regardless of the substitution of the diphe-
nylamine, both unsaturated alkoxyamines decomposed through
a pathway distinct from the saturated compounds.
Figure 3. (A) Calculated TS structure for the retro-carbonyl-ene
reaction of O-methyl N,N-diphenylhydroxylamine, and (B) its
HOMO.
Since the RCE mechanism involves concerted C-H cleavage, the
ease with which it occurs should be dependent on the identity
of the alkyl moiety. Indeed, when we replaced the methyl group
with an iso-propyl or allyl group, we found that the calculated N-
O BDEs increased slightly (44.5 and 42.6 kcal/mol, respectively,
compared to 42.1 kcal/mol), but the RCE reaction became
more facile by 1.8 (33.5) and 4.0 (31.3) kcal/mol, respectively.
While
the
alkoxyamine
derived
from
the
4,4’-
dimethyldiphenylamine had an expectedly weaker N-O bond
than in the diphenylamine-derived alkoxyamine owing to the
ability of the electron-donating methyl substituents to better
stabilize the electron-poor aminyl radical15,16 (N-O BDE = 41.3
kcal/mol), a concerted RCE TS was also found for this reaction,
and again, it was lower in enthalpy than that for bond dissocia-
tion (33.4 kcal/mol). These results suggest that the RCE reac-
tion may be an accessible pathway for decomposition of dia-
rylamine-derived alkoxyamines.32
Experiments. Despite the proposed central role of N,N-
diarylalkoxyamines in the catalytic activity of diarylamine radi-
cal-trapping antioxidants, there are only a handful of examples
of their preparation and characterization and only one study of
their reactivity – the seminal work of Korcek and co-workers
relayed above.5 We felt that our understanding of this reaction
would be greatly bolstered by experiments employing authentic
N,N-diarylalkoxyamines. As such, following Kelly’s work,21 we
were able to prepare saturated and unsaturated alkoxyamines
derived from diphenylamine and 4,4’-di-tert-butyldiphenylamine
(3) as shown in Scheme 3. Briefly, the endoperoxide formed
from photosensitized oxygenation of 1,3-cyclohexadiene22 was
treated with lithium diarylamide to prepare the unsaturated
alkoxyamines 4 and 5, and the corresponding saturated alkoxy-
amines 6 and 7, were obtained by the same chemistry, but fol-
To provide further insight on the lower logA value for 4 com-
pared to 6, we synthesized their octadeuterated counterparts 4-d8
and 6-d8 as in Scheme 3 (using ca. 94% perdeuterated 1,3-
cyclohexadiene25 as starting material) and measured their de-
composition kinetics at 120°C. These experiments revealed a
kinetic isotope effect of 1.8 for 4 and 1.2 for 6, which corre-
spond to values of 2.1 and 1.3 at 25°C – a primary kinetic iso-
tope effect for 4, suggesting some C-H bond cleavage in the
transition state of the decomposition reaction, and a secondary
kinetic isotope effect for 6.
4
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
from the computations, our view is that ‘unactivated’ alkoxy-
A
1
2
3
4
5
6
7
amines undergo N-O homolysis and disproportionation to yield
the diarylamines,33 whereas suitably ‘activated’ substrates can
access the RCE pathway to yield the same products. Further-
more, a switch from the high-enthalpy/low-entropy N-O homol-
ysis to the low-enthalpy/high-entropy RCE pathway with chang-
ing temperature is evident from the changing deuterium incor-
poration in the diphenylamine product.
8
9
With the authentic alkoxyamines in-hand, we also evaluated
their ability to inhibit the autoxidation of a hydrocarbon at
elevated temperature relative to the corresponding diarylamines.
We selected hexadecane because Korcek’s suggested mechanism
in Scheme 1 was derived largely from experiments carried out
with this hydrocarbon.5,26,27 Moreover, we employed the same
temperature (160°C) and the same type of stirred flow reactor as
originally used by Korcek.26 The stirred flow reactor makes use
of a constant flow of O2 to stir the reaction mixture and ensure
that it is continuously oxygenated – a necessity at this tempera-
ture, otherwise mass transfer of O2 becomes rate-limiting.26,27 In
principle, since the half-lives of the unsaturated and saturated
alkoxyamines at 160°C (derived from the Arrhenius parameters
given above) are only ca. 100 and 200 seconds, respectively, they
should decompose largely to the diphenylamines in the first few
minutes of the reaction, giving rise to inhibited autoxidations
indistinguishable from those inhibited by the corresponding
diphenylamines. The data, presented as the concentration of
hydroperoxide determined as a function of time (using our pre-
viously described pro-fluorescent phosphine28), are shown in
Figure 5. Interestingly, while the data from autoxidations inhib-
ited by the unsaturated alkoxyamines 4 and 5 were fully con-
sistent with our expectations, the saturated alkoxyamines 6 and
7 were found to be less effective than either the unsaturated
alkoxyamines or authentic diphenylamines.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
B
Figure 4. (A) Decomposition of 4 monitored by HPLC between
105 and 150°C. (B) Temperature dependence of the decompo-
sition of 4 (n), 5 (l), 6 (p), and 7 (u).
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
These results imply that while the unsaturated alkoxyamines
decompose to give diphenylamine quantitatively in hexadecane
at 160°C, the same cannot be said of the saturated alkoxy-
amines. This was hinted at in Korcek’s original experiment,
where he found that heating of a mixture of alkoxyamines gen-
erated in situ from decomposition of tert-butyl peroxy-2-
ethylhexanoate in hexadecane in the presence of bis(p-
octylphenyl)nitroxide gave 64% bis(p-octylphenyl)amine.5 If we
consider this in the context of the foregoing mechanistic stud-
ies, the results highlight the competition that exists between
disproportionation of the initially formed diphenylaminyl-
alkoxyl radical pairs and other possible reactions, e.g. H-atom
abstraction from hexadecane by the alkoxyl radical, which can
initiate autoxidation. In contrast, the unsaturated alkoxyamines
yield the same extent of inhibition as the corresponding diphe-
nylamines because no radicals are formed in the RCE reaction,
and the diphenylamines are produced quantitatively. This leads
to the somewhat surprising conclusion that the catalytic radical-
trapping antioxidant activities of diarylamines are likely to be highly
substrate-dependent. That is, while unsaturated hydrocarbons may
be more oxidizable than saturated ones, when oxidized at ele-
vated temperatures in the presence of diarylamines, they should
give rise to alkoxyamines which are more efficiently converted
D
D
D
D
D
D
OH
D
OH
D
Ph
N
Ph
N
D
D
Ph
O
Ph
O
H
D
D
D
D
H
4-d8
6-d8
Direct evidence for the RCE mechanism was obtained upon
examination of the products derived from the decomposition of
4-d8 and 6-d8; deuterium incorporation was observed in the ortho
position of one of the phenyl rings of the diphenylamine de-
rived from 4 (49% at 90°C and decreasing with increasing tem-
perature to 28% at 150°C, as judged by the diminished integra-
1
tion of the signal in the H-NMR spectra relative to the meta
and para proton signals), whereas the diphenylamine derived
from decomposition of 6 contained no deuterium (see Support-
ing Information). These results, taken alongside the kinetic
isotope effects and the Arrhenius parameters suggest a change in
mechanism upon going from an ‘unactivated’ (i.e. saturated) to
‘activated’ (i.e. unsaturated) substrate. Incorporating the insights
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
ΔH‡ for the RCE reaction of the O-methyl alkoxyamine derived
back to the starting diarylamines, which will be characterized by
higher stoichimetric numbers.
from N-phenyl-β-naphthylamine (8) 35 would be 6 kcal/mol low-
er than the corresponding alkoxyamine derived from diphenyl-
amine. Gratifyingly, we were able to prepare the two corre-
1
2
3
4
5
6
A
sponding N-phenyl-β-naphthylamine-derived alkoxyamines
and 10 as in Scheme 3.
9
7
8
9
OH
OH
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
N
N
NH
O
O
9
10
8
The β-naphthyl-containing alkoxyamines 9 and 10 decomposed
rapidly compared with the corresponding diphenylamine-
derived compounds (at 120°C, k = 2.1×10-3 and 6.4×10-4 s-1,
respectively). More interestingly, the temperature dependence of
the rate constants for their disappearance yielded Arrhenius
parameters of logA = 13.0±0.3 and Ea = 28.2±0.4 kcal/mol for 9
and logA = 12.9±0.2 and Ea = 29.0±0.4 kcal/mol for 10, imply-
ing that both compounds decompose via the RCE pathway.
Moreover, hexadecane autoxidations at 160°C inhibited by 9
and 10 were indistinguishable from each other, and from hexa-
decane autoxidations inhibited by N-phenyl-β-naphthylamine35
(cf. Figure 5). Thus, by lessening the energetic penalty paid by
the group on the amine to undergo the concerted process (the
aromatic stabilization of the ‘second’ ring of naphthalene is
estimated to be 10-12 kcal/mol less than that in benzene),29 we
can favour this pathway even for ‘unactivated’ (i.e. saturated)
alkoxyamines.
B
Figure 5. Hydroperoxide formation in the autoxidation of hex-
adecane at 160°C initiated by 10 mM tetralin hydroperoxide
(n) and inhibited by 100 µM of (A): diphenylamine 1 (l) n =
8.4, unsaturated alkoxyamine 4 (q) n = 8.2 or saturated alkoxy-
amine 6 (u) n = 5.8; (B): 4,4’-di-tert-butyldiphenylamine 3 (l) n
= 9.5, unsaturated alkoxyamine 5 (q) n = 9.1 or saturated
alkoxyamine 7 (u) n = 6.3.
While the foregoing results certainly provide insight into the
mechanism of diarylamine regeneration, and show that the effi-
cacy of the diarylamine as a RTA is likely to be dependent on
the substrate that is undergoing oxidation, it would arguably be
more useful to know if the structure of the diarylamine could
instead be altered to favour decomposition by the RCE reaction
in lieu of N-O homolysis/disproportionation. We surmised that
replacement of one of the phenyl rings in the diphenylamine-
derived alkoxyamine with a naphthyl ring may decrease the en-
thalpic barrier for the RCE reaction sufficiently that an activat-
ed (i.e. unsaturated) alkyl element would no longer be required
for the reaction to proceed in this way. CBS-QB3 calculations
suggested that this would indeed be the case, predicting that the
Figure 6. Hydroperoxide formation in the autoxidation of hex-
adecane at 160°C initiated by 10 mM tetralin hydroperoxide
(n), and inhibited by 100 µM of either N-phenyl-β-
naphthylamine 8 (l) n = 5.2, unsaturated alkoxyamine 9 (q) n
= 5.2 or saturated alkoxyamine 10 (u) n = 5.2.
Despite the greater reactivity of N-phenyl-β-naphthylamine35
toward peroxyl radicals compared to diphenylamine (k =
6
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
from diphenylamines and saturated substrates decompose by N-
1.1×105 M-1s-1 (8) vs. 4.2×104 M-1s-1 (1) at 65°C),34 and significant-
1
O homolysis followed by disproportionation, whereas those
derived from either diphenylamines and unsaturated substrates
or N-phenyl-β-naphthylamine35 and saturated substrates decom-
pose by a concerted retro-carbonyl-ene reaction. This change in
mechanism has significant implications in the use of diaryla-
mine radical-trapping antioxidants, since they are expected to be
comparatively better inhibitors of the autoxidation of unsatu-
rated substrates owing to the more efficient regeneration of the
diarylamine from its corresponding alkoxyamine. Moreover, it
provides important insight into the design of the next genera-
tion of diarylamine radical-trapping antioxidants: compounds
which turnover via the retro-carbonyl ene pathway will be far
more effective than those that turnover via N-O homolysis and
in-cage disproportionation, as do the industry-standard alkylated
diphenylamines. While substitution of one of the phenyl rings
for a naphthyl ring achieves this, it also makes deleterious (off-
cycle) reaction pathways more accessible. Limiting these reac-
tions, while preserving the RCE pathway for diarylamine regen-
eration, should be the focus of future research.
ly more favorable regeneration from its saturated alkoxyamine
(vide supra), it is not as effective an inhibitor of hexadecane au-
toxidation (compare Figures 5A and 6 and stoichiometric fac-
tors derived from the data of n = 8.4 for the former and 5.2 for
the latter). The likely explanation is that addition of peroxyl
radicals to the α-position of the naphthyl ring of the N-phenyl-β-
naphthylaminyl radical intermediate will also be made easier
compared to the diphenylaminyl radical due to the reduced
aromaticity of the naphthyl ring. Indeed, the CBS-QB3-
calculated C-O BDE of the adduct is 41.7 kcal/mol, which is
13.1 kcal/mol stronger than the analogous adduct of the diphe-
nylaminyl radical (see Scheme 1), which makes peroxyl radical
addition to the ring 9.7 kcal/mol more favourable than N-O
bond formation (see Scheme 4). Although this reaction is re-
versible, it must compete with subsequent O-O bond homolysis
(or tautomerization followed by homolysis), which requires
roughly the same amount of energy as the reverse reaction (41.1
kcal/mol). Moreover, O-O bond homolysis is likely to be irre-
versible under the reaction conditions as the resultant radicals
can disproportionate to the iminoquinone and alcohol; prod-
ucts from which the diarylamine cannot be regenerated.
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
EXPERIMENTAL SECTION
General. Reagents were purchased from commercial suppliers
N
+
MeOO
and
used
without
further
purification.
2,3-
Dioxabicyclo[2.2.2]oct-5-ene was synthesized from the reaction
of 1,3-cyclohexadiene with singlet oxygen according to the pro-
cedure of Ziegert and Brase.22 The diimide reduction of the
unsaturated endoperoxide was carried out according to the pro-
cedure of Coughlin, Brown, and Salomon.23 Cyclohexadiene-d8
was produced as a mixture of isomers via the procedure of
Mugridge, Bergman, and Raymond,25 and the crude mixture
was subjected directly to singlet oxidation. Column chromatog-
raphy was carried out using flash silica gel (40-63 µm, 230-400
-41.7
-32.0
kcal/mol
N
kcal/mol
OOCH3
N
OOCH3
Scheme 4. Key competing pathways (and associated ΔH values)
for phenyl β-napthylaminyl + •OOMe.
1
mesh). H and 13C NMR were recorded on a Bruker AVANCE
spectrometer at 400 MHz and 100 MHz, respectively, unless
specified otherwise. High-resolution mass spectra were obtained
on a Kratos Concept Tandem mass spectrometer.
The foregoing results have important implications with respect
to the design of novel diarylamine radical-trapping antioxidants.
They suggest that while optimization of the stability of the dia-
rylaminyl radical may promote the initial H-atom transfer (pro-
ton-coupled electron transfer, Eq. 1) from the diarylamine to
the peroxyl radical, this may not be helpful for promoting re-
generation of the amine, since it will promote N-O bond ho-
molysis and may facilitate cage escape. Instead, it would appear
that diarylamines which are good H-atom donors and whose
corresponding alkoxyamines preferentially follow the retro-
carbonyl-ene route to regenerate the diarylamine would be ideal
radical-trapping antioxidants. We are currently exploring these
ideas.
General Procedure for the synthesis of Diarylalkoxy-
amines. In a dry Schlenk flask, under an inert atmosphere, 2.2
mmol of the appropriate diarylamine was dissolved in 15 ml dry
Et2O (0.13 M) and cooled to -78°C. Once cooled, 0.90 ml of n-
BuLi in hexanes (2.5 M, 2.2 mmol) was added dropwise to the
diarylamine solution, after which it was allowed to stir for 15
min at -78°C. 2.0 mmol of the endoperoxide was dissolved in 1-
2ml of dry Et2O and was added rapidly to the cooled reaction
(over ca. 20 seconds; the colour of the solution typically became
a dark green). The reaction was allowed to stir for 10 minutes
before an excess of water was added to quench the reaction and
the solution was allowed to warm to room temperature. The
organic phase was separated and the aqueous phase extracted
with Et2O two additional times. The combined organic phases
were washed with brine, then dried with MgSO4, filtered, and
concentrated in vacuo. The alkoxyamines were purified by silica
gel chromatography with 40% to 70% Et2O/petroleum ether,
typically yielding a viscous oil which was crystalized from a small
amount of Et2O in hexanes at -20°C overnight.
CONCLUSIONS
The synthesis of a series of N,N-diarylalkoxyamines has enabled
kinetic and mechanistic experiments that clarify the mechanism
of the catalytic behavior of diarylamine radical-trapping antioxi-
dants. The determination of Arrhenius parameters, deuterium
kinetic isotope effects, and isotope incorporation experiments
confirm computational predictions that alkoxyamines derived
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
4-[N,N-Diphenyl(aminoxy)]-2-cyclohexen-1-ol (4). Yield: 136.9, 133.2, 130.4, 129.0, 128.8, 127.39, 127.37, 126.4, 125.8,
30% Off-white solid. 1H NMR (400 MHz; CDCl3): δ 7.31 (dd,
J = 8.5, 7.4, 4H), 7.17-7.14 (m, 4H), 7.13-7.08 (m, 2H), 6.01-
5.94 (m, 2H), 4.34 (m, 1H), 4.17 (br, 1H), 2.09-2.02 (m, 1H),
1.87-1.75 (m, 3H). 13C NMR (100 MHz; CDCl3): δ 149.4,
134.3, 128.9, 128.8, 124.4, 121.1, 74.8, 66.0, 28.4, 24.1.
HRMS: m/z Calc: C18H19NO2: 281.1416 Found: 281.1427.
125.0, 124.4, 121.1, 120.7, 116.8, 74.4, 64.3, 27.7, 24.1,
HRMS: m/z Calc: C26H37NO2: 331.1572 Found: 331.1555.
1
2
3
4
5
6
7
8
4-[N-2-Naphthyl-N-phenyl(aminooxy)]cyclohexanol
(10).
1
Yield: 11% Off-white solid. H NMR (300 MHz; DMSO-d6): δ
7.87 (m, 3H), 7.60 (s, 1H), 7.51-7.33 (m, 4H), 7.26 (d, J = 8.9
Hz, 1H), 7.18-7.11 (m, 3H), 4.44 (d, J = 3.6 Hz, 1H), 3.91 (m,
1H), 3.57 (m, 1H), 1.90 (m, 2H), 1.67-1.43 (m, 6H). 13C NMR
(76 MHz; DMSO-d6): δ 149.2, 146.6, 133.2, 130.3, 128.9,
128.7, 127.4, 127.3, 126.4, 124.9, 124.2, 121.1, 120.7, 116.6,
77.9, 65.5, 30.4, 26.2, HRMS: m/z Calc: C26H37NO2: 333.1729
Found: 333.1731.
4-[N,N-Di(4-tert-butylphenyl)aminooxy]-2-cyclohexen-1-ol
1
(5). Yield: 22% Off-white solid. H NMR (400 MHz; DMSO-
9
d6): δ 7.34 (d, J = 8.8 Hz, 4H), 7.01 (d, J = 8.8 Hz, 4H), 5.86
(dd, J = 10.2, 1.7 Hz, 1H), 5.79 (dd, J = 10.5, 2.9 Hz, 1H), 4.86
(d, J = 5.5 Hz, 1H), 4.19 (m, 1H), 3.97 (m, 1H), 1.97-1.91 (m,
1H), 1.68-1.56 (m, 3H), 1.26 (s, 18H). 13C NMR (100 MHz;
CDCl3): δ 147.04, 146.95, 134.0, 129.2, 125.6, 120.7, 74.4,
66.0, 34.3, 31.4, 28.5, 24.1. HRMS: m/z Calc: C26H35NO2:
393.2668 Found: 393.2667.
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
Alkoxyamine Decomposition Experiments. 1.0 ml of hexade-
cane was added to 0.015 mmol of alkoxyamine that was weighed
into a glass vial. A magnetic Teflon coated stir bar was added
and the vial was capped with a rubber septum. The suspension
was stirred for a few minutes (gentle heating was usually re-
quired) to dissolve the solid. The vial was then placed in a heat-
ing block at the appropriate temperature were allowed to stir for
30s before the first sample was taken. 100µL samples were taken
at regular intervals and added to 900µL of 2.2 mM solution of
benzyl alcohol in 2% isopropanol/hexanes (HPLC grade). Sam-
ples were analysed using a Waters 2695 Alliance HPLC (1-4:
1.2% iPrOH/hexanes, 0.8 ml min-1; 5min, 1.4%
iPrOH/hexanes, 0.8 ml min-1; 10 min 1.6% iPrOH/hexanes,
1.2 ml min-1; 5-6: 2.0% iPrOH/hexanes, 0.6 ml min-1; 30min;
Sunfire Silica column (5µm, 4.6 x 250 mm)) and analysed by
UV absorbance at 215nm.
4-[N,N-Diphenyl(aminooxy)]cyclohexanol (6). Yield: 20%
White solid. 1H NMR (400 MHz; CDCl3): δ 7.35-7.30 (m, 4H),
7.20-7.17 (m, 4H), 7.14-7.09 (m, 2H), 3.97 (m, 1H), 3.83-3.78
(m, 1H), 2.07-1.99 (m, 2H), 1.76-1.62 (m, 6H), 1.36 (br, 1H).
13C NMR (100 MHz; CDCl3): δ 149.7, 128.8, 124.2, 121.0,
77.5, 68.1, 30.8, 26.7. HRMS: m/z Calc: C18H21NO2: 283.1572
Found: 283.1551.
4-[N,N-Di(4-tert-butylphenyl)aminooxy]-cyclohexanol (7).
1
Yield: 15% White solid. H NMR (400 MHz; CDCl3): δ 7.30
(d, J = 8.8 Hz, 4H), 7.08 (d, J = 8.8 Hz, 4H), 3.91 (m, 1H),
3.80-3.75 (m, 1H), 2.01 (m, 2H), 1.72-1.62 (m, 6H), 1.30 (s,
18H). 13C NMR (100 MHz, CDCl3) δ 147.3, 146.8, 125.6,
120.5, 68.2, 34.3, 31.4, 30.8, 26.8. HRMS: m/z Calc:
C26H37NO2: 395.2824 Found: 395.2849.
Hexadecane Autoxidations. n-Hexadecane (100 mL) was
thoroughly degassed with argon and then heated to 160°C while
argon was continuously bubbled through the liquid. Once the
temperature stabilized, 0.10 mmol of inhibitor (1-10) and 164
mg (1.0 mmol) of tetralin hydroperoxide were added to the
solution and the flow of argon was replaced with O2. Aliquots
(1.5 mL) were removed every 5 minutes, and allowed to cool to
room temperature for analysis. Four duplicates (30 µL) of each
sample were loaded into separate wells of a 96-well microplate
and the automated reagent dispenser of the microplate reader
was used to dilute each sample with tert-amyl alcohol (200 µL)
and a solution of a fluorgenic phosphine dye solution (20 µL of
a 250 µM stock solution in acetonitrile, see Eq. 4) immediately
before reading. The plate was stirred for 8 seconds, and allowed
to rest for 2 seconds more, and the fluorescence of each well was
measured every second for 60 seconds (absorption 340 nm;
emission 425 nm). The concentration of hydroperoxide was
determined from the rate of phosphine oxidation using the rate
constant for the reaction of the dye with secondary hydroperox-
ides in tert-amyl alcohol (k = 1.2 M-1s-1) assuming pseudo-first-
order kinetics.28
4-[N,N-Diphenyl(aminooxy)]-2-(1,2,3,4,5,5,6,6-d8)-
cyclohexen-1-ol (4-d8). Yield: 29% Off-white solid. ca. 94% D.
1H NMR (300 MHz; CDCl3): δ 7.31 (dd, J = 8.4, 7.2 Hz, 4H),
7.16 (dd, J = 8.7, 1.2 Hz, 4H), 7.13-7.08 (m, 2H), 5.97 (d, J =
9.7 Hz, 0.15H), 4.33 (s, 0.07H), 4.18-4.15 (m, 0.06H), 2.01 (s,
0.07H), 1.78 (d, J = 12.7 Hz, 0.21H), 1.51 (s, 1H). 13C NMR
(76 MHz, CDCl3) δ 149.3, 128.9, 124.3, 121.1 HRMS: m/z
Calc: C18H19NO2: 289.1924 Found: 289.1918.
4-[N,N-Diphenyl(aminooxy)]-2-(1,2,3,4,5,5,6,6-d8)-
1
cyclohexanol (6-d8). Yield: 14% White solid. ca. 93% D. H
1
NMR H-NMR (300 MHz; CDCl3): δ 7.30 (dd, J = 8.5, 7.2
Hz, 4H), 7.16 (d, J = 8.7 Hz, 4H), 7.09 (t, J = 7.3 Hz, 2H),
3.94-3.92 (m, 0.07H), 3.78-3.76 (m, 0.07H), 3.40 (t, J = 6.6
Hz, 0.06H), 1.97 (br, 1H), 1.66 (br, 1H), 1.31 (s, 1H). 13C
NMR (76 MHz; CDCl3): δ 149.7, 128.8, 124.1, 121.0. HRMS:
m/z Calc: C18H21NO2: 291.2074 Found: 291.2096.
4-[N-2-Naphthyl-N-phenyl(aminooxy)]-2-cyclohexen-1-ol
1
(9). Yield: 13% Off-white solid. H NMR (400 MHz; DMSO-
ACKNOWLEDGEMENTS
d6): δ 7.86 (m, 3H), 7.61 (d, J = 2.0 Hz, 1H), 7.47 (td, J =
13.6, 5.8 Hz, 2H), 7.37 (t, J = 7.9 Hz, 2H), 7.27 (dd, J = 8.8,
2.2 Hz, 1H), 7.16 (m, 3H), 5.92-5.84 (m, 2H), 4.89 (d, J = 5.6
Hz, 1H), 4.33 (m, 1H), 3.99 (m, 1H), 2.04-1.98 (m, 1H), 1.75-
1.56 (m, 3H).. 13C NMR (100 MHz; DMSO-d6): δ 149.0, 146.3,
This work was supposed by grants from the Natural Sciences
and Engineering Research Council of Canada and the Canada
Foundation for Innovation. DAP acknowledges an excellent
8
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
(15)
(16)
(17)
(18)
(19)
Pratt, D. A.; DiLabio, G. A.; Valgimigli, L.; Pedulli, G. F.;
discussion with Jonathan Burton (University of Oxford) and
support from the University of Ottawa and the Canada Re-
search Chairs program, respectively.
Ingold, K. U. J. Am. Chem. Soc. 2002, 124, 11085.
Pratt, D. A.; DiLabio, G. A.; Mulder, P.; Ingold, K. U. Acc.
Chem. Res. 2004, 37, 334.
Boozer, C. E.; Hammond, G. S.; Hamilton, C. E.; Sen, J. N. J.
Am. Chem. Soc. 1955, 77, 3233.
Stable Radicals: Fundamentals and Applied Aspects of Odd-Electron
Compounds; Hicks, R. G., Ed.; John Wiley & Sons, Ltd., 2010.
Beckwith, A.; Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc.
1992, 114, 4983.
Bauernschmitt, R.; Ahlrichs, R. J. Chem. Phys. 1996, 104, 9047.
Kelly, D. R.; Bansal, H.; Morgan, J. J. Tetrahedron Letters 2002,
43, 9331.
1
2
3
4
5
6
7
8
Supporting Information Available: Complete kinetic data,
NMR spectra for isotope labeling studies, NMR spectra of syn-
thesized alkoxyamines and Cartesian coordinates and energetics
of computed structures. This information is available free of
charge via the internet at http://pubs.acs.org.
(20)
(21)
9
REFERENCES
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
(22)
(23)
Ziegert, R.; Bräse, S. Synlett 2006, 2006, 2119.
Coughlin, D. J.; Brown, R. S.; Salomon, R. G. J. Am. Chem.
Soc. 1979, 101, 1533.
Benson, S. W. Thermochemical Kinetics; John Wiley & Sons,
Ltd., 1976.
Mugridge, J. S.; Bergman, R. G.; Raymond, K. N. Angew. Chem.
Int. Ed. 2010, 49, 3635.
Jensen, R. K.; Korcek, S.; Mahoney, L. R.; Zinbo, M. J. Am.
Chem. Soc. 1979, 101, 7574.
Igarashi, J.; Jensen, R. K.; Lusztyk, J.; Korcek, S.; Ingold, K. U.
J. Am. Chem. Soc. 1992, 114, 7727.
Hanthorn, J. J.; Haidasz, E.; Gebhardt, P.; Pratt, D. A. Chem.
Commun. (Camb.) 2012, 48, 10141.
Schleyer, P. V. R.; Pühlhofer, F. Org. Lett. 2002, 4, 2873.
These are expected to hydrolyze to yield quinones, which are
observed products in diphenylamine-inhibited autoxidations,
see ref. 4.
(1)
(2)
(3)
Ingold, K. U. Chem. Rev. 1961, 61, 563.
Ingold, K. U.; Pratt, D. A. Chem. Rev. 2014, 114, 9022.
Lucarini, M.; Pedrielli, P.; Pedulli, G. F.; Valgimigli, L.;
Gigmes, D.; Tordo, P. J. Am. Chem. Soc. 1999, 121, 11546-
11553.
Bolsman, T.; Blok, A. P.; Frijns, J. Recl. Trav. Chim. Pays-Bas
1978, 97, 310.
(24)
(25)
(26)
(27)
(28)
(4)
(5)
(6)
Jensen, R. K.; Korcek, S.; Zinbo, M.; Gerlock, J. L. J. Org.
Chem. 1995, 60, 5396.
It has also been suggested that nitroxides can be catalytic RTAs
because of reduction of their oxoammonium ions by substrate-
derived alkyl radicals (see ref. 2). The nitroxides are excellent
RTAs when protonated by carboxylic acids (see: Amorati, R.;
Pedulli, G. F.; Pratt, D. A.; Valgimigli, L. Chem. Commun.
(Camb.) 2010, 46, 5139), and carboxylic acids are known to be
formed in autoxidations (see: Jensen, R. K.; Korcek, S.; Ma-
honey, L. R.; Zinbo, M. J. Am. Chem. Soc. 1981, 103, 1742 and
Jalan, A.; Alecu, I. M.; Meana-Paneda, R.; Aguilera-
Iparraguirre, J.; Yang, K. R.; Merchant, S. S.; Truhlar, D. G.;
Green, W. H. J. Am. Chem. Soc. 2013, 135, 11100).
In contrast, the reaction of N,N-dialkylalkoxyamines with
peroxyl radicals is believed to be the reaction responsible for
the catalytic activities of the significantly less reactive hindered
amine light stabilizers, see: Gryn’ova, G.; Ingold, K. U.; Coote,
M. L. J. Am. Chem. Soc. 2012, 134, 12979.
Hanthorn, J. J.; Valgimigli, L.; Pratt, D. A. J. Am. Chem. Soc.
2012, 134, 8306.
Hanthorn, J. J.; Amorati, R.; Valgimigli, L.; Pratt, D. A. J. Org.
Chem. 2012, 77, 6895.
Hanthorn, J. J.; Valgimigli, L.; Pratt, D. A. J. Org. Chem. 2012,
77, 6895.
Montgomery, J. A.; Frisch, M. J.; Ochterski, J. W.; Petersson,
G. A. J. Chem. Phys. 1999, 110, 2822.
(29)
(30)
(31)
(32)
In contrast, the O-C bond is significantly weaker than the N-O
bond in N,N-dialkylalkoxyamines, see ref. 7.
It is difficult to determine ΔS‡ for the N-O bond dissocia-
tion/in-cage disproportionation pathway for direct compari-
sons to the concerted process, or for comparison to our solu-
tion-phase experiments (vide infra).
(7)
(33)
(34)
The small ‘secondary’ KIE observed for the decomposition of
the saturated alkoxyamine likely reflects the commitment fac-
tor for the N-O homolysis given the subsequent in-cage dispro-
portionation has to compete with radical recombination.
Brownlie, I. T.; Ingold, K. U. Can. J. Chem. 1966, 44, 861-868.
The rate constants reported in this paper for N-phenyl-β-
naphthylamine and diphenylamine were 5.0×104 and 2.0×104
M-1s-1, respectively at 65°C, but these should be revised upward
by a factor of 2.1 given the revised value of 2kt for styrylperoxyl
radicals, from which they are derived. See: Howard, J. A. in
"Free Radicals"; Kochi, J. K., Ed.; Wiley: New Work, 1973;
Vol. 2, 3-62.
(8)
(9)
(10)
(11)
(12)
(13)
(14)
Isborn, C.; Hrovat, D. A.; Borden, W. T.; Mayer, J. M.; Car-
penter, B. K. J. Am. Chem. Soc. 2005, 127, 5794.
DiLabio, G. A.; Johnson, E. R. J. Am. Chem. Soc. 2007, 129,
6199.
DiLabio, G. A.; Litwinienko, G.; Lin, S.; Pratt, D. A.; Ingold,
K. U. J. Phys. Chem. A 2002, 106, 11719.
(35)
N-phenyl-β-naphthylamine was, at one time, among the most
widely used diarylamine RTAs, but is now used sparingly ow-
ing to its carcinogenic potential. See: Weiss, T.; Brüning, T.;
Bolt, H. M. Crit. Rev. Toxicol. 2007, 37, 553.
ROOH
ROO
when R = saturated, RDS
is N-O homolysis, then
in-cage disproportionation
N
Ar
Ar
ROO
R
O
H
when R = unsaturated
and/or Ar = naphthyl, RDS is
concerted retro-carbonyl-ene
reaction:
N
O
N
Ar
Ar
Ar
Ar
R-H=O
RDS
R
O
N
H
OR
R
O
O
N
Ph
N
Ar
Ar
Ar
Ar
R
9
ACS Paragon Plus Environment