Phenoxazine: A privileged scaffold for radical-trapping antioxidants

Article abstract of DOI:10.1021/acs.joc.7b02025

Diphenylamines are widely used to protect petroleum-derived products from autoxidation. Their efficacy as radical-trapping antioxidants (RTAs) relies on a balance of fast H-atom transfer kinetics and stability to one-electron oxidation by peroxidic specie

Full text of DOI:10.1021/acs.joc.7b02025

Subscriber access provided by GRIFFITH UNIVERSITY
Article
Phenoxazine: A Privileged Scaffold for Radical-Trapping Antioxidants
Luke A. Farmer, Evan A Haidasz, Markus Griesser, and Derek A. Pratt
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.7b02025 • Publication Date (Web): 08 Sep 2017
Downloaded from http://pubs.acs.org on September 8, 2017
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.
The Journal of Organic Chemistry 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 42
The Journal of Organic Chemistry
1
2
3
4
Phenoxazine: A Privileged Scaffold for Radical-Trapping Antioxidants
5
6
7
8
Luke A. Farmer, Evan A. Haidasz, Markus Griesser 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 and Biomolecular Sciences, University of Ottawa, 10 Marie Curie
Pvt., Ottawa, Canada.
*Corresponding author: Derek A. Pratt
Phone: 613-562-5800
Fax: 613-562-5170
E-mail: dpratt@uottawa.ca
TOC Graphic:
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 42
1
2
3
4
Abstract
5
6
7
8
9
Diphenylamines are widely used to protect petroleum-derived products from autoxidation. Their
efficacy as radical-trapping antioxidants (RTAs) relies on a balance of fast H-atom transfer kinetics
and stability to one electron oxidation by peroxidic species. Both H-atom transfer and one-electron
oxidation are enhanced by substitution with electron-donating substituents, such as the S-atom in
phenothiazines, another important class of RTA. Herein we report the results of our investigations
of the RTA activity of the structurally-related, but essentially ignored, phenoxazines. We find that
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
the H-atom transfer reactivity of substituted phenoxazines follows an excellent Evans-Polanyi
ꢄꢅ°ꢆ
correlation spanning ꢀ
= 4.5 × 106 M-1 s-1 and N-H BDE = 77.4 kcal mol-1 for 3-CN,7-NO2-
ꢁꢂꢃ
ꢄꢅ°ꢆ
phenoxazine to ꢀ
= 6.6 × 108 M-1 s-1 and N-H BDE = 71.8 kcal mol-1 for 3,7-(OMe)2-
ꢁꢂꢃ
phenoxazine. The reactivity of the latter compound is the greatest of any RTA ever reported, and
is likely to represent a reaction without enthalpic barrier since log A for this reaction is likely ∼8.5.
The very high reactivity of most of the phenoxazines studied required the determination of their
kinetic parameters by inhibited autoxidations in the presence of a very strong H-bonding co-
solvent (DMSO), which slowed the observed rates by up to two orders of magnitude by
dynamically reducing the equilibrium concentration of (free) phenoxazine as an H-atom donor.
Despite their remarkably high reactivity toward peroxyl radicals, the phenoxazines were found to
be comparatively stable to one-electron oxidation relative to diphenylamines and phenothiazines
(E° ranging from 0.59 V to 1.38 V vs. NHE). Thus, phenoxazines with comparable oxidative
stability to commonly used diphenylamine and phenothiazine RTAs had significantly greater
reactivity (by up to 2 orders of magnitude). Computations suggest that this remarkable balance in
H-atom transfer kinetics and stability to one electron oxidation results from the ability of the
bridging oxygen atom in phenoxazine to serve as both a p-electron donor to stabilize the aminyl
radical and s-electron acceptor to destabilize the aminyl radical cation. Perhaps most excitingly,
phenoxazines have “non-classical” RTA activity – where they trap >2 peroxyl radicals each – at
ambient temperatures.
2
ACS Paragon Plus Environment

Page 3 of 42
The Journal of Organic Chemistry
1
2
3
4
Introduction.
5
6
7
8
9
Autoxidation, the radical-mediated chain reaction responsible for the spontaneous oxidation of
hydrocarbons, is a ubiquitous process limiting the longevity of all living things and petroleum-
derived materials. Radical trapping antioxidants (RTAs) compete with propagation of the
autoxidation chain reaction by H-atom transfer to a chain-carrying peroxyl radical (Eq. 1).1,2
Subsequent reaction of the RTA-derived radical with another chain-carrying peroxyl radical can
inhibit a second chain reaction (Eq. 2). The efficacy of RTAs is largely governed by two metrics:
the rate constant of the initial H-atom transfer reaction (the inhibition rate constant, kinh) and the
number of chains that are broken (the inhibition reaction stoichiometry, n).
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
A-H + ROO• à A• + ROOH
(1)
(2)
A• + ROO• à non-radical products
The among most widely-used RTAs are substituted phenols (e.g. butylated hydroxytoluene,
BHT) and diphenylamines (e.g. alkylated diphenylamines, ADPAs). The reactivity of both classes
of RTA obey good Evan-Polanyi relationships, where decreases in the phenolic O-H and aminic
N-H bond dissociation enthalpies (BDEs) brought about by substitution of the aryl ring(s) with
electron-donating groups (EDGs) give rise to larger kinh.3,4 The drop in BDE associated with the
introduction of EDGs is accompanied by a drop in E° and concomitant stability to one electron
oxidation (by product hydroperoxides or O2), which can lead to production of radical species rather
than their trapping. Thus, achieving a balance of reactivity to H-atom transfer and stability to one-
electron oxidation must be at the center of any RTA development strategy.
Although the structure-reactivity relationships in phenolic RTAs have been extensively studied,
diphenylamines have been comparatively less well-investigated.2,5–8 Of the existing work in this
area, the careful thermochemical and kinetic studies of the reactivity of phenothiazines –
diphenylamines (1) fused together by a central thiazine ring (2) – by Lucarini et al is particularly
noteworthy.6 They showed that phenothiazine has an N-H bond that is 6.5 kcal/mol weaker than
in diphenylamine (78.2 vs. 84.7 kcal mol-1)9 and a ~600-fold greater reactivity toward peroxyl
radicals (kinh = 8.8 ´ 106 vs. 1.5 ´ 104 M-1 s-1 at 50 °C). Introduction of electron-donating methoxy
substituents para relative to the aminic N-H further decreased the N-H BDE by 3.1 kcal mol-1 and
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 42
1
2
3
4
further increased kinh by ~6-fold to 5.0 ´ 107 M-1 s-1 – the largest kinh ever reported for a
diarylamine.10
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
In the same manuscript, Lucarini et al. presented corresponding data for phenoxazine (3),
revealing that it possessed an even weaker N-H bond than phenothiazine (76.1 vs. 78.2 kcal mol-
1)9 and a correspondingly greater kinh of 2.9 ´ 107 M-1 s-1. No substituted phenoxazines were
investigated. Perhaps most interestingly, while phenothiazine and the substituted phenothiazines
that were investigated trapped ∼2 peroxyl radicals each in inhibited autoxidations (i.e. n ∼ 2),
phenoxazine apparently trapped five. This was not explained. Moreover, although not pointed out
at the time, despite its greater reactivity and stoichiometry, phenoxazine is (marginally) more
stable to one-electron oxidation than phenothiazine (vide infra).11 Hence, it would appear that
phenoxazine is a privileged scaffold for RTA design and development. Herein we describe our
efforts to characterize the structure-reactivity relationships in phenoxazines as RTAs, from which
it is revealed that they are privileged structures with compelling prospects for real-world
applications.
Results.
I. Exploiting Kinetic Solvent Effects to Clock the Fastest RTAs
The very high reactivity of phenoxazine to peroxyl radicals makes precise quantitation of the
kinetics of the reaction challenging. Determination of inhibition rate constants of this magnitude
(mid-107 M-1s-1) requires pushing the venerable inhibited autoxidation approach,12,13 by which
reaction progress is generally monitored by O2 consumption, to its limit. Our own variation of the
inhibited autoxidation approach,14 in which a small amount of a highly absorbing co-substrate
(PBD-BODIPY, Figure 1A) is added to the reaction mixture as a signal carrier, has an even lower
limit (kinh ≤ 107 M-1 s-1) due to the inherent reactivity of the co-substrate to peroxyl radicals (i.e.
kBODIPY= 2720 M-1 s-1 in styrene/chlorobenzene) and the maximum PBD-BODIPY concentration
4
ACS Paragon Plus Environment

Page 5 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
that can be achieved without saturating the photomultipliers in a conventional spectrophotometer
(10 µM). Indeed, phenoxazine completely inhibits PBD-BODIPY consumption in styrene
autoxidations in chlorobenzene under standard conditions (black trace, Figure 1B), precluding
determination of kinh from the initial rate (Eq. 3) as in Figure 1C.14
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 1. (A) PBD-BODIPY is used as a signal carrier in the autoxidation of styrene or 1,4-
dioxane. (B) Co-autoxidation of styrene (4.3 M) and PBD-BODIPY (10 μM) in PhCl (black), 1,4-
dioxane (2.9 M) in PhCl, and 1,4-dioxane (2.9 M) in 2:1 PhCl:DMSO initiated with 6 mM AIBN
at 37°C. Reaction progress was monitored at 591 nm (ε = 139 000 M-1 cm-1), 587 nm (ε = 123 000
M-1 cm-1), and 587 nm (ε = 118 200 M-1 cm-1), respectively. (C) Rate constants (kinh) and
stoichiometries (n) for the reactions of inhibitors (RTAs) with chain-carrying peroxyl radicals can
be determined from the initial rates and the duration of the inhibited periods as in Eqs. 3 and 4,
respectively. (D/E) The equilibrium between ‘free’ RTA (illustrated by phenoxazine) and RTA
participating in a 1:1 hydrogen-bond complex with arbitrary solvent S. Assuming the 1:1 complex
does not react with peroxyl radicals, the kinetics of radical-trapping can be described as in Eq. 5
or Eq. 6.
In an attempt to circumvent this limitation, we exchanged styrene for an oxidizable substrate that
would make H-bonds with phenoxazine (Figure 1D). This would lower the (free) concentration of
phenoxazine in solution, thereby increasing the rate of the inhibited autoxidation such that it could
be reliably measured. Ingold has shown that the kinetics of H-atom transfer reactions are retarded
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 42
1
2
3
4
5
6
7
8
9
in a wholly predictable fashion in H-bond accepting media, wherein the kinetics obey a pre-
dissociation model as in Eq. 5 in Figure 1E.15,16 Thus, if the equilibrium constant for H-bond
formation between phenoxazine and the substrate (KHB) is known, the rate constant for its reaction
in the absence of strong H-bonding interactions (ꢀꢁꢇꢂꢃ) can be derived from the observed rate
constant (ꢀꢁꢈꢂꢃ). This expression can be recast using Abraham’s H-bond acidity (ꢉꢊꢋ)17 and H-bond
basicity (ꢌꢊꢋ)18 parameters – which generally range from 0 to 1 – as in Eq. 6, a relationship that
has been shown to hold for a variety of H-atom transfer reactions under a variety of conditions.
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
Thus, styrene was exchanged for dioxane as the oxidizable substrate. Dioxane is relatively easily
oxidized (kp = 0.5 M-1 s-1)19 – ensuring that a free radical chain reaction is maintained in the
autoxidation – and is also a good H-bond acceptor (ꢌꢊꢋ = 0.41).18 However, using dioxane as the
substrate was insufficient to increase the rate of the phenoxazine-inhibited autoxidation to a
measurable level (Figure 1D). Hence, an even stronger H-bond acceptor was added as a co-solvent:
DMSO (ꢌꢊꢋ = 0.78).18 Upon doing so, the inhibited autoxidation clearly proceeded with a non-zero
slope. Derivation of kinh from the initial rate required determination of kBODIPY in this solvent system
-1 -1
(5900 M s , see Supporting Information), from which ꢀ
5 -1 -1
= (5.3 ± 0.5) × 10 M s could be
ꢍꢎꢈꢏ
ꢁꢂꢃ
derived using Eq. 3.
To correct for the H-bonding interaction of phenoxazine with DMSO, thereby enabling
derivation of the inherent reactivity of phenoxazine to peroxyl radicals, required that we determine
ꢉꢊꢋ of phenoxazine and ꢌꢊꢋ of the solvent system. While exhaustive compilations of ꢌꢊꢋ parameters
exist for a wide array of HBAs, our systems of interest consist of an amalgam of 2 or 3 potential
HBDs. A direct approach to ꢌꢊꢋdetermination of a HBA would be to determine KHB using
quantitative IR spectroscopy in CCl4.20,21 We, however, opted for the 19F NMR method devised by
Taft et. al.,22 where differences in the chemical shift of 4-fluorophenol and 4-fluoroanisole are
related directly to KHB (see the Supporting Information for further details). Using this method, an
effective ꢌꢊꢋ of 0.35 was determined in 25% v/v dioxane in PhCl and ꢌꢊꢋ of 0.60 in 25% v/v
dioxane with 25% v/v DMSO in PhCl.23
Similarly, the ꢉꢊꢋ of phenoxazine was determined by 1H NMR. Abraham and co-workers have
established that the chemical shift difference of the acidic proton of an HBD in CDCl3 and d6-
DMSO obeys an excellent linear relationship with ꢉꢊꢋ (R2 = 0.938 for 54 different species with a
6
ACS Paragon Plus Environment

Page 7 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
total of 72 different protic functionalities).24 Unfortunately, we found that phenoxazines are
generally characterized by poorly resolved 1H NMR spectra in CDCl3. However, 1H NMR spectra
in benzene-d6 are very well resolved and further, the chemical shift difference of the acidic protons
of a variety of HBDs in this solvent relative to d6-DMSO were also found to correlate very nicely
to ꢉꢊꢋ according to the expression in Eq. 7 (R2 = 0.972 for 11 different species ranging from α2H =
0.08 to 0.82, see ESI for the correlation).
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
ꢉꢊꢋ = 0.134 ꢐꢑꢍꢎꢈꢏꢒꢓꢃꢋ − 0.1087
(7)
The determination of ꢐꢑꢍꢎꢈꢏꢒꢓꢃꢋ for phenoxazine affords ꢉꢊꢋ = 0.44 from this correlation.
Using this value of ꢉꢊꢋ, ꢌꢊꢋ = 0.60 for the autoxidation solvent system and kinh = (5.3 ± 0.5) × 105
M-1 s-1 determined from the inhibited autoxidation enables the derivation of kinh = (3.9 ± 0.4) × 107
M-1 s-1 in chlorobenzene (ꢌꢊꢋ = 0.09),18,25 which is in excellent agreement with the value obtained
by Lucarini et al. in that solvent (2.9 × 107 M-1 s-1).6
Figure 2. Inhibition rate constants derived from dioxane autoxidations in PhCl (black squares) or
PhCl/DMSO (red circles) correlate very well with inhibition rate constants directly determined
from styrene or 1-hexadecene autoxidations in PhCl. Representative inhibitors: PMC (A), 5-
hydroxy-2-dimethylaminopyrimidine (B), 5-hydroxy-4,6-dimethyl-2-dimethylaminopyrimidine
(C), 2-methylundecan-2-persulfide (D), phenothiazine (2), 3,7-di-tert-butylphenothiazine (E), 3,7-
dimethoxyphenothiazine (F), phenoxazine (3). See ESI for full details.
In order to further validate the use of kinetic solvent effects to enable the determination of rate
constants for very fast reactions of RTAs with peroxyl radicals, we carried out additional
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 42
1
2
3
4
5
6
7
8
9
autoxidations inhibited by phenothiazine, 3,7-dimethoxyphenothiazine and 3,7-di-tert-
butylphenothiazine, as studied by Lucarini et al. by O2 consumption in the inhibited autoxidation
of styrene.6 Moreover, since these compounds have largely similar HBD ability (ꢉꢊꢋ = 0.34-0.38),
we also studied a few examples that were either excellent H-bond donors (two substituted
pyrimidinols, ꢉꢊꢋ = 0.56/0.65)26 or very poor ones (a hydropersulfide, ꢉꢊꢋ = 0.09).27 The data is
shown in Figure 2, along with the data for 2,2,5,7,8-pentamethylchroman-6-ol (PMC), a truncated
form of a-tocopherol – Nature’s premier RTA and an important benchmark compound in any RTA
study.28 Overall, the agreement with literature values determined directly in a non-H-bonding
solvent (chlorobenzene) is excellent.
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
II. Structure-Activity Relationships in Phenoxazines as RTAs
With a methodology established to enable quantitative studies of very fast RTAs, the synthesis of
a small library of substituted phenoxazines was carried out in order to evaluate the substituent
effects on the parent structure’s reactivity. 2-Trifluoromethylphenoxazine 4, 3-
trifluoromethylphenoxazine 5, 2-tert-butylphenoxazine 6, 3-tert-butylphenoxazine 7 and 3,7-di-
tert-butylphenoxazine 8, were each synthesized via intramolecular Cu-catalyzed Ullman-type
couplings of the acetamido and bromo/iodo functionalities of corresponding diaryl ethers as in
Scheme 1A.
2,4,6,8-Tetra-tert-butylphenoxazine 9 was synthesized by reduction of 2,4-di-tert-butyl-6-
nitrophenol followed by self-condensation in acetic acid with zinc dust in a manner previously
described.29 2,8-Di-tert-butylphenoxazine 10 was synthesized by the same method (Scheme 1B).
N,N-Diethyl-3-sulfonamidephenoxazine 11, 3-cyanophenoxazine 12, and 3-methoxy-7-
nitrophenoxazine 13 were obtained by SNAr of the appropriate o-aminophenol on a o-haloaryl/2-
nitro-haloaryl (Scheme 1C). The formation of the diaryl ether followed by a Smiles rearrangement
brokers an exclusive formation of the phenoxazine derivative wherein the substituent is located
para to the reactive N-H.30,31 3-Nitrophenoxazine 14 and 3-cyano-7-nitrophenoxazine 15 were
obtained by electrophilic nitration of 10-acetylphenoxazine32 (with subsequent hydrolysis) and 3-
cyanophenoxazine, respectively (Scheme 1D). 3-Methoxyphenoxazine 16 and 3,7-
dimethoxyphenoxazine 17 were obtained by alkylation and then hydrolysis of hydroxy-N-
8
ACS Paragon Plus Environment

Page 9 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
acetylphenoxazines derived from the reduction and acylation of phenoxazin-3-one and 7-hydroxy-
phenoxazin-3-one, respectively (Scheme 1E).
7
8
Scheme 1. Preparation of Substituted Phenoxazines.
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
Inhibited autoxidations of dioxane in DMSO/PhCl were carried out as described above, with the
choice of conditions dictated largely by the ꢉꢊꢋvalues determined by NMR. The data is assembled
in Table 1. As expected, phenoxazines with electron-withdrawing groups (EWGs) had
ꢓꢃꢔꢕ
substantially higher ꢉꢊꢋ parameters than phenoxazine, such that their ꢀ
could be determined
ꢁꢂꢃ
directly from inhibited autoxidations of dioxane sans DMSO whereas the phenoxazines substituted
with electron-donating groups (EDGs) had marginally lower ꢉꢊꢋ – and were substantially more
reactive than phenoxazine – so DMSO was added to the inhibited autoxidations. The quality of the
ꢉꢊꢋ determinations for the substituted phenoxazines is reflected in the excellent correlation of these
values with the corresponding ꢖꢗꢒ substituent constants33 (Figure 3A); expected since they are
essentially partially dissociated Brønsted acids in the H-bonded complexes.
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 42
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. (A) Linear free energy relationship of ꢘꢖꢗꢒ and ꢉꢊꢋ for 3-substituted, or 3,7-
2
ꢓꢃꢔꢕ
disubstituted phenoxazines (R = 0.955) (B) Linear free energy relationship of ꢘꢖꢗꢙ and logꢀ
ꢁꢂꢃ
for 3-substituted and 3,7-disubstituted phenoxazines ( r+ = -0.68, R2 = 0.982) (C) Linear free
energy relationship of ꢘꢖꢗꢙ and N-H BDE (black squares, R2 = 0.959) and ꢘꢖꢗꢙand Eo (red circles,
R2 = 0.944) for 3-substituted and 3,7-disubstituted phenoxazines.
10
ACS Paragon Plus Environment

Page 11 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The rate constants derived from the inhibited autoxidations span more than two orders of
magnitude; the lowest determined for 3-CN,7-NO2-phenoxazine (4.5 ± 0.1 × 106 M-1 s-1) and the
highest determined for 3,7-(OMe)2-phenoxazine (6.6 ± 1.1 × 108 M-1 s-1). In general, and as
expected based on the kinetics of the reactions of other diarylamines with peroxyl radicals,5,6,8
substitution with EDGs increase the rate and EWGs decrease the rate. In fact, the values of kinh for
the substituted phenoxazines correlate very well with the corresponding ꢖꢗꢙ substituent constants33
– or sum of the substituent constants in the case of multiple substitutions – as is evident in Figure
3B. Also given in Table 1 are N-H BDEs that were calculated by the high-accuracy CBS-QB3
complete basis set methodology34 and the standard potentials determined by cyclic voltammetry.
As expected, both properties correlate well with Σꢖꢗꢙ (Figure 3C).35
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
III. Understanding the Superiority of Phenoxazine as an RTA
The CBS-QB3 calculations used to predict the N-H BDEs given in Table 1 were expanded to
provide a view of the transition state structure for H-atom transfer from phenoxazine to peroxyl
radicals (Figure 4A). The structure greatly resembles that of the reaction between diphenylamine
and a peroxyl radical (Figure 4B) in that the proton is transferred roughly in the plane of the
molecular framework, while the electron is delocalized in approximately perpendicular orbitals
(i.e. the p-HOMO of phenoxazine and p*-SOMO of the peroxyl radical).8 Thus, the reaction can
be described as a proton-coupled electron transfer, where the proton and electron are exchanged
between different pairs of orbitals.7,8,27,36–39 The greater reactivity of phenoxazine is based on both
enthalpic and entropic factors: the N-H bond is weaker, decreasing DH‡ (0.8 vs. 5.1 kcal mol-1),
and the aryl rings are already fixed in optimal positions to maximize delocalization of the unpaired
electron in the aminyl radical, minimizing the entropic cost (DS‡ = -36.4 vs. -37.7 cal mol-1 for
phenoxazine and diarylamine respectively).
In carrying out these calculations we also sought some insight into the origin of the relatively
high E° of phenoxazine (0.87 V vs. NHE) despite its low N-H BDE (76.1 kcal mol-1) and
corresponding high reactivity (kinh = 3 ´ 107 M-1 s-1). To put this into context, diarylamines of
similar or lower reactivity are generally characterized by E° values which are ∼0.4 V lower.8 Of
course, a fundamental difference between phenoxazine and diphenylamine is the planar structure
of the latter – in both the starting material and the incipient radical and radical cation – which
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 42
1
2
3
4
5
6
7
8
9
should improve spin delocalization. Indeed, the N-H BDE and IP in 9,10-dihydroacridine are
predicted to be 2.4 and 4.3 kcal mol-1 lower, respectively, than in diphenylamine. However, when
the CH2 bridge is replaced with an O-atom the predicted N-H BDE decreases by 11.2 kcal mol-1,
but this is accompanied by a decrease of only 8.7 kcal mol-1 in IP. The decrease in BDE upon
inclusion of a heteroatom bridge in lieu of a methylene unit was anticipated, given that electron-
donation from the heteroatom can better stabilize the electron poor aminyl radical by resonance.
However, an even greater effect was expected on the ionization potential, given that the radical
cation is inherently more electron poor than the amine. In order to eliminate the resonance
contribution of the oxygen atom on the stability of the aminyl radical and/or cation, and focus only
its inductive/field effect, we also calculated the N-H BDEs and IPs in piperidine and morpholine.
Although the N-H BDEs are essentially identical, the oxygen atom clearly raises the IP of the
amine. The inductive/field effect is even more obvious when the conformation of the piperidine
and morpholine rings are constrained to be planar (as opposed to the minimum energy chair
conformer); the IPs now differ by 8.3 kcal mol-1, while the N-H BDEs remain essentially the same.
A similar, but attenuated, trend exists in the corresponding sulfur compounds. These data are
summarized in Figure 4E.
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
12
ACS Paragon Plus Environment

Page 13 of 42
The Journal of Organic Chemistry
1
2
3
4
5
Table 1. Inhibition Rate Constants, ꢉꢊꢋ H-Bond Donor Parameters, Inhibition Stoichiometries, N-
H Bond Dissociation Enthalpies and Oxidation Potentials for Substituted Phenoxazines.
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
H b
ꢚꢞꢛꢜꢝ
ꢟꢝꢠꢡ
ꢛꢜꢝ
RTA
n
α2
BDE
(kcal mol-1) c
Eo d
ꢚ
(M-1 s-1) a
(M-1 s-1)
15e
14e
12e
5e
1.5 ± 0.1 × 105
2.4 ± 0.1
2.5 ± 0.1
2.2 ± 0.2
2.3 ± 0.1
2.2 ± 0.1
2.3 ± 0.1
1.5 ± 0.1
2.2 ± 0.1
2.0 ± 0.1
2.1 ± 0.1
1.8 ± 0.1
1.9 ± 0.1
1.8 ± 0.1
1.7 ± 0.1
1.4± 0.1
0.69
0.62
0.56
0.53
0.53
0.51
0.60
0.44
0.43
0.43
0.43
0.41
0.42
0.39
0.40
4.5± 0.1 × 106
77.5
76.5
76.2
76.1
76.3
76.2
74.0
75.2
74.6
74.5
74.0
73.9
72.8
-
1.38
1.15
1.13
1.06
1.06
1.05
0.97
0.87
0.83
0.81
0.78
0.75
0.72
0.67
0.59
7.3 ± 0.2 × 105
1.1 ± 0.1 × 106
1.5 ± 0.2 × 106
1.5 ± 0.1 × 106
2.0 ± 0.1 × 106
2.8 ± 0.2 × 106
5.3 ± 0.5 × 105
9.0 ± 0.3 × 105
1.1 ± 0.1 × 106
1.5 ± 0.1 × 106
2.6 ± 0.2 × 106
3.1 ± 0.2 × 106
4.7 ± 0.8 × 106
(1.3 ± 0.2 × 107)g
1.6 ± 0.1 × 107
1.8 ± 0.1 × 107
2.0 ± 0.3 × 107
2.1 ± 0.2 × 107
2.5 ± 0.2 × 107
5.6 ± 0.5 × 107
3.9 ± 0.4 × 107
6.0 ± 0.2 × 107
7.4 ± 0.6 × 107
1.0 ± 0.1 × 108
1.4 ± 0.1 × 108
1.8 ± 0.1 × 108
2.1 ± 0.4 × 108
(6.6 ± 1.1 × 108)
11e
4e
13e
3f
6f
7f
10f
8f
16f
9f
17f
70.7
aRates of inhibition determined for 2 µM of inhibitor a minimum of three times. bCalculated from ΔδDMSO-PhH of the relevant proton
in the 1H NMR spectrum in conjunction with Eq. 7. Spectra can be found in the ESI. cCalculated by CBS-QB3. dValues in V vs. NHE
e
f
ꢥꢤꢆꢦ
ꢢꢣꢤ
ꢥꢤꢆꢦ
ꢢꢣꢤ
ꢧꢪꢫꢬ
.
ꢢꢣꢤ
determined by cyclic voltammetry in acetonitrile.
ꢀ
was calculated from ꢀꢢꢧꢣꢢꢤꢨꢩ. ꢀ
was calculated from ꢀ
gꢀꢢꢧꢣꢪꢤꢫꢬwas determined via kinetic modeling as described previously, see ref. 14.
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 42
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 4. (A/B) Calculated TS structures of the reaction between a peroxyl radical (exemplified
by methylperoxyl) and phenoxazine and diphenylamine. (C/D) Calculated singly occupied
molecular orbitals of the TS. (E) CBS-QB3 calculated N-H BDEs and IPs of diphenylamine (top),
9,10-dihydroacridine, phenothiazine and phenoxazine (middle) and piperidine, thiomorpholine
and morpholine (bottom). For piperidine, thiomorpholine and morpholine, data are presented for
both the minimum energy conformation, as well as a planarized (C2v) structure. (F) Calculated
minimum energy and planarized structures of morpholine radical cation and molecular dipole
moment vector.
Discussion
Phenothiazines, which are synthesized simply by treating diphenylamines with elemental sulfur
in the presence of a catalytic oxidant (e.g. I2), have been used as antioxidants since the 1950s.40
Phenoxazines, despite being known to possess a weaker N-H bond and greater reactivity toward
peroxyl radicals,6 are not widely used – if at all. One consideration may be that they are not as
conveniently prepared on industrial scale.41 Another consideration may be that early reports
suggested that phenoxazines were no more reactive than phenothiazines – at least at elevated
temperatures – discouraging further investigation.42 The more recent results of Lucarini et al.6
strongly suggested to us that phenoxazines merit further investigation as RTAs; in particular, since
they possess greater reactivity toward peroxyl radicals and greater stability to one-electron
oxidation compared to phenothiazines.
14
ACS Paragon Plus Environment

Page 15 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The application of the spectrophotometric inhibited co-autoxidation approach14 to the
determination of the peroxyl radical-trapping kinetics of phenoxazines required slowing the
inhibition reaction sufficiently to obtain a measurable rate for the inhibited part of the autoxidation.
This was accomplished by the addition of a very strong H-bond accepting co-solvent (DMSO) to
the reaction medium, precluding H-atom transfer from all but the tiny amount of non-H-bonded
phenoxazine present at equilibrium. Although there are many examples of the utilization of the
empirical equation derived by Ingold to estimate H-atom transfer rate constants in solvents of
differing H-bond accepting ability (to which we refer as the Ingold-Abraham relationship, see Eq.
6),21,43,44 we believe this to be the most extreme, given that DMSO is such an excellent H-bond
acceptor (ꢌꢊꢋ = 0.78) and the very high inherent reactivity of some of the H-atom donors (i.e. in
the absence of H-bonding interactions) are the fastest ever observed (kinh up to mid 108 M-1 s-1). As
such, prior to our investigations of substituted phenoxazines, we first vetted the use of DMSO in
combination with the Ingold-Abraham relationship by deriving the inhibition rate constants for
other highly reactive RTAs whose kinetics had already been determined by oxygen consumption.
The excellent agreement instilled confidence that data derived from PBD-BODIPY/dioxane co-
autoxidations in DMSO are reliable.
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
The reactivity of the substituted phenoxazines was, as expected, impressive. At the low end,
substitution at the 3- and 7- positions with cyano and nitro groups yields kinh = 4.5 × 106 M-1 s-1;
roughly one order of magnitude lower than phenoxazine. At the high end, substitution at the 3-
and 7-positions with methoxy groups yields kinh = 6.6 ± 1.1 × 108 M-1 s-1; roughly one order of
magnitude greater than phenoxazine. Phenoxazines substituted with single substituents were
characterized by intermediate reactivity which correlated very well with the s+ Hammett-type
electrophilic substituent constant to yield a strong linear free energy relationship (Figure 3B)
whose negative reaction constant (r+ = -0.68) is consistent with increasing electron-poorness of
the aryl rings upon H-atom transfer from the phenoxazines to peroxyl radicals.
As far as we are aware, 3,7-dimethoxyphenoxazine is the most reactive radical-trapping
antioxidant ever reported. In fact, it would appear that this reactivity is essentially the limit for
these types of reactions. Previous work from our group had suggested that log A ∼7.5 for the
reactions of diarylamines with peroxyl radicals,45 but the fact that the two aryl rings are fixed in
phenoxazines reduces the entropy requirement to achieve the transition state (wherein optimal
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 42
1
2
3
4
5
6
7
8
9
delocalization of the unpaired electron in the incipient aminyl radical can occur). To a first
approximation, these two bond rotations may contribute at most (since they are coupled) ½ log
unit each to log A, suggesting that a value of ∼8.5 is reasonable for phenoxazines. The only
similarly reactive RTA reported to date is the 5-hydroxy-7-aza-2,3-dihydroindole shown in Chart
1, which was also believed to react with peroxyl radicals with essentially no enthalpic barrier (i.e.
Ea ∼ 0).46
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
Chart 1. Radical-trapping antioxidants with log kinh ∼ log A ∼ 8.5.
Although they react more slowly, phenoxazines substituted with electron-withdrawing groups
are arguably of greater practical interest. For example, the slowest phenoxazine we studied (3-CN,
7-NO2) is ∼300 fold more reactive than diphenylamine (compare ꢀꢁꢓꢂꢃꢃꢔꢕ = 4.5 ´ 106 and 1.5 ´ 104
M-1 s-1, respectively)6 yet possesses greater stability to one-electron oxidation (compare E° = 1.38
and 1.24 V vs. NHE, respectively).47 Moreover, phenoxazines substituted with a single strong
electron-withdrawing group (CN, NO2, CF3, SO2NEt2) were roughly 100-fold more reactive than
ꢓꢃꢔꢕ
the ADPAs used commercially (compare ꢀ
∼2´107 versus 1.8 ´ 105 M-1 s-1,8 respectively)
ꢁꢂꢃ
while also being at least as stable to one electron oxidation (E° >1 V vs. NHE). A particularly
interesting compound is 3-methoxy-7-nitrophenoxazine. Since methoxy and nitro groups have the
same magnitude of s+ constants, but of opposite sign (-0.78 and 0.79, respectively),33 we expected
their effects on the reactivity of phenoxazine to negate one another. However, this compound
reacted 50% more quickly with peroxyl radicals than phenoxazine – and was characterized by a
E° value that was 100 mV greater – hence, the best of each substituent is contributed to these two
key RTA parameters. The greater reactivity is consistent with the lower calculated N-H BDE in
this compound compared to phenoxazine (74.4 vs. 75.2 kcal mol-1), perhaps suggesting the role of
captodative stabilization of the aminyl due to polarization across the two rings.
The unique combination of increased oxidative stability and increased H-atom transfer reactivity
of phenoxazines relative to phenothiazines and diphenylamines is most clearly illustrated in Figure
16
ACS Paragon Plus Environment

Page 17 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
5, where log kinh is plotted versus E° for the three types of arylamines. The three essentially parallel
correlations reflect the fact that in compounds with a similar reactivity toward peroxyl radicals,
the phenoxazine is considerably more stable to one-electron oxidation than phenothiazine, which
is then more stable than the diphenylamine. Likewise, for compounds of the same oxidative
stability, the phenoxazine is 5 to 10-fold more reactive than the phenothiazine, and even more so
than the diphenylamine. The calculations described above provide good rationale for this; the p-
donor ability of the oxygen atom in phenoxazine is responsible for the weaker N-H bond and larger
kinh compared to diphenylamine, but its s-acceptor ability (electronegativity) minimizes the drop
in E°. Although the sulfur atom in phenothiazine is also a good p-donor, such that it also possesses
a weak N-H bond and large kinh, since sulfur is less electronegative than oxygen it has a smaller
effect on E°.
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 5. Inhibition rate constants of substituted phenoxazines (black), phenothiazines (red),
and diarylamines (blue) plotted as a function of their oxidation potentials (E°). Representative
inhibitors: 4,4'-dioctyldiphenylamine (G), 4-methoxydiphenylamine (H), N2,N2-dimethyl-N5-
phenylpyrimidine-2,5-diamine (I), 4,4'-dimethoxydiphenylamine (J), N1,N1-dibutyl-N4-
(pyrimidin-5-yl)benzene-1,4-diamine(K),
4-dimethylaminodiphenylamine
(L),
N5-(6-
(dimethylamino)pyridin-3-yl)-N2,N2-dimethylpyridine-2,5-diamine (M).8 See ESI for structures.
One of the most interesting observations made by Lucarini et al. in their original work6 was that
styrene autoxidations inhibited by phenoxazine had inhibited periods that corresponded to the
trapping of five (5!) peroxyl radicals. The foregoing dioxane/PBD-BODIPY co-autoxidations
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 42
1
2
3
4
were characterized by inhibited periods that correspond to n ∼ 2, similar to phenols and other
5
6
aminic antioxidants. However, when phenoxazine-inhibited co-autoxidations were carried out in
7
8
9
styrene, a dramatic increase in the stoichiometric factor was observed (Figure 6A). Interestingly,
(ꢮ)
the kinetic traces featured two distinct inhibited periods: an initial fully inhibited period (ꢭ
∼
ꢢꢣꢤ
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
(ꢊ)
ꢢꢣꢤ
1700 s) corresponding to n ∼ 2 and a secondary inhibited period (ꢭ ∼ 6700 s) corresponding to
an additional n ∼ 8, for a total of n ∼ 10. Moreover, the rate of the autoxidation beyond the second
inhibited period remained retarded relative to the uninhibited rate. As far as we are aware, this
level of radical-trapping capacity at ambient temperatures is unprecedented. It is noteworthy that
two distinct inhibited phases (and subsequent retarded phase) in the phenoxazine-inhibited
autoxidation of styrene were not reported by Lucarini et al.6 This may be due to the significantly
greater rate of initiation that they used to produce a non-zero inhibited autoxidation rate (necessary
for them to derive kinh), which would lead to a much less pronounced inflection point at ꢭꢢ(ꢣꢮꢤ). It
should also be pointed out that the previous measurements were made at 50°C.
18
ACS Paragon Plus Environment

Page 19 of 42
The Journal of Organic Chemistry
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 6. (A) Co-autoxidation of styrene (4.3 M) and PBD-BODIPY (10 uM) initiated by AIBN
(6 mM) in PhCl at 37 oC (black) and inhibited by 2 μM phenoxazine 3 (red, n(1) = 2.2 ± 0.1, n(2) =
8.0 ± 1.9). (B) Corresponding autoxidations inhibited by 2 μM of 3-cyanophenoxazine 12 (blue,
n(1) = 2.4 ± 0.1, n(2) = 9.4 ± 0.5), 3-(N,N-diethylsulfonamide)-phenoxazine 11 (green, n(1) = 2.3 ±
0.1, n(2) = 8.1 ± 0.6), 3-tert-butylphenoxazine 7 (orange, n(1) = 1.9 ± 0.3, n(2) = 6.7 ± 1.4) and 3-
methoxyphenoxazine 16 (violet, n(1) = 1.8 ± 0.3, n(2) = 3.9 ± 0.5).
In addition to the striking substrate-dependence of this “non-classical” RTA behavior, we also
found that the rate of the secondary inhibited period is largely independent of the structure of the
phenoxazine (kinhsecondary = 5.1 ± 0.1 × 105 M-1s-1 for 3, 5.3 ± 0.4 × 105 M-1s-1 for 7, 6.2 ± 0.1 × 105
M-1s-1 for 11, 7.2 ± 0.3 × 105 M-1s-1 for 12, and 4.3 ± 0.5 × 105 M-1s-1 for 16). Representative
examples are shown in Figure 6B, which features data for styrene autoxidations inhibited by both
electron-rich (methoxy- or tert-butyl-substituted) and electron-poor (sulfonamide- or cyano-
substituted) phenoxazines. The major difference between these data sets is the difference in the
lengths of both the primary and secondary inhibited periods, with the more electron-rich
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 42
1
2
3
4
5
6
7
8
9
phenoxazines giving rise to noticeably shorter overall inhibited periods, such that the overall n
decreases upon increasing electron-richness: 11.6 (3-CN), 10.4 (3-Et2NSO2), 10.2 (3-H), 8.6 (3-t-
Bu), 5.7 (3-OMe). This is presumably due to the increased oxidative lability of the phenoxazines
along this series, which depletes the RTA and generates radicals rather than traps them.
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
Non-classical (n > 2) RTA activity of diarylamines is well known, but only at elevated
temperatures (i.e. >100 °C)2 – such as those experienced by engine oil lubricants, rubbers and
plastics to which ADPAs are added. The need for elevated temperatures is easily understood upon
consideration of the accepted mechanism for the catalytic activity of ADPAs, shown in Scheme
2.48 Following the initial H-atom transfer to a peroxyl radical, the diphenylaminyl radical reacts
with a second peroxyl radical to form a nitroxide. The nitroxide then combines with an alkyl radical
to form an alkoxyamine, that can undergo rate-limiting fragmentation to reform the ADPA. The
fragmentation requires cleavage of the N-O bond either by homolysis/disproportionation or a
concerted retro-carbonyl-ene pericyclic process,49 both of which do not operate on the timescale
of the inhibited autoxidations shown in Fig. 6 at 37 °C (k ~ 3 ´ 10-9 s-1).49 Of course, the rate may
increase if the aminyl radical formed upon N-O homolysis is significantly more stable, as is the
aminyl derived from phenoxazine. However, the difference in the aminyl radical stabilities – which
can be estimated from the difference in the N-H BDEs between ADPAs (∼82 kcal mol-1) and
phenoxazine (∼76 kcal mol-1) – is too little to increase the rate of fragmentation of the putative
alkoxyamine sufficiently to account for the secondary inhibition observed in the styrene
autoxidations (Figure 6).50 Moreover, the rates during the secondary inhibited periods to which the
“non-classical” RTA activity is attributed are almost independent of the structure of the
phenoxazine – inconsistent with N-O homolysis to yield aminyl radicals of significantly different
stability being rate-determining (i.e. the difference in N-H BDEs of the cyano and methoxy
substituted phenoxazines is 3.2 kcal/mol). Clearly, another mechanism must operate – and must
do so in styrene, but not dioxane. Efforts to elucidate this mechanism continue in our laboratories.
20
ACS Paragon Plus Environment

Page 21 of 42
The Journal of Organic Chemistry
1
2
3
4
Scheme 2. Mechanism of Diarylamine RTAs at Elevated Temperatures.
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
Given the higher reactivity, radical-trapping capacity and oxidative stability of phenoxazines
compared to other aminic RTAs, it will be interesting to see how they compare in industrially-
relevant contexts. Moreover, given that even the least reactive phenoxazine studied here is still
more reactive than a-TOH, the most biologically-active form of Vitamin and Nature’s premier
RTA (kinh = 3.2 × 106 M-1 s-1),28 it would be interesting to see if these compounds have any
biological activity. Again, this is particularly interesting given that a-TOH is comparatively easy
to oxidize (E° = 0.93 V).51 Indeed, we recently found that phenoxazine is the most potent inhibitor
of lipid autoxidation driven cell death (also called ferroptosis) ever reported,52 corroborating an
earlier study that suggested that phenoxazine was a potent neuroprotectant.53 Thus, some of the
compounds described here may provide leads toward potential preventive and/or therapeutic
agents for degenerative conditions wherein lipid autoxidation has been implicated.
Conclusions
Application of the kinetic solvent effect model of Ingold has enabled quantitation of the kinetics
of the reactions of substituted phenoxazines with peroxyl radicals (kinh). The addition of DMSO as
a co-solvent to dioxane autoxidations suppressed the reactivity of phenoxazines by up to two
orders of magnitude, thereby enabling derivation of their inhibition rate constants under more
relevant (i.e. non-H-bonding) conditions via the Ingold-Abraham relationship. The reliability of
this approach can be appreciated explicitly by the accuracy of the derived rate constants with
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 42
1
2
3
4
respect to previously characterized RTAs, as well as implicitly based on the strong linear
ꢓꢃꢔꢕ
ꢢꢣꢤ
correlation between ꢖꢗꢙ and log ꢀ
for the newly studied substituted phenoxazines.
5
6
7
8
9
The reactivity of the phenoxazines spanned 4.5 × 106 M-1 s-1 for 3-CN,7-NO2-phenoxazine to 6.6
× 108 M-1 s-1 for 3,7-(OMe)2-phenoxazine. The reactivity of the latter compound is the greatest of
any RTA ever reported, and is likely to represent a reaction without enthalpic barrier since log A
for this reaction is likely ∼8.5. Linear free energy relationships observed between the inhibition
rate constants, standard potentials and N-H bond strengths of phenoxazines parallel those of
substituted diarylamines and phenothiazines. However, with respect to the other classes of aminic
RTAs, the substituted phenoxazines are decisively privileged since compounds of similar
oxidative stability (quantified by Eo) are orders of magnitude more reactive to peroxyl radicals.
The results of computations account for this, by indicating that the heteroatom bridge in
phenoxazines (and to a lesser extent, phenothiazines) stabilizes the aminyl radical by electron
donation resonance while destabilizing the aminyl radical cation by electron withdrawal by
induction. Perhaps most interestingly, we confirm the once-reported non-classical RTA behavior
of phenoxazine at ambient temperatures, and show that 1) it is dependent on the rate of initiation
of autoxidation, 2) is only weakly dependent on the electronic characteristics of the phenoxazine,
and 3) is dependent on the substrate undergoing autoxidation. The precise mechanism of this non-
classical RTA behavior remains unknown.
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
General Methods All chemicals and solvents were purchased from commercial suppliers unless
1
13
19
otherwise indicated. H, C, and F NMR spectra were recorded on either a 300 or 400 MHz
instrument. UV-visible spectra and associated kinetics were measured with a spectrophotometer
equipped with a temperature controller unit and a thermostated 6 × 6 multicell holder. Styrene was
washed thrice with 1 M NaOH, once with water and then dried with MgSO4 prior to distillation
under reduced pressure at 50 °C. Thereafter, it was passed through a column of silica and stored
at -20 °C for up to 4 days. Immediately prior to use, it was passed through a column filled with 2/3
basic alumina (bottom) and 1/3 silica (top). 3-Cyanophenoxazine,30 10-acetylphenoxazine,54 3,7-
dihydroxy-10-acetylphenoxazine,55
phenoxazone,56
N-[2-(4-(tert-butyl)-2-
22
ACS Paragon Plus Environment

Page 23 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
iodophenoxy)phenyl]acetamide,57 5-hydroxy-2-dimethylaminopyrimidine,26b 5-hydroxy-4,6-
dimethyl-2-dimethylaminopyrimidine,26b
2-methylundecan-2-persulfide,27
3,7-di-tert-
7
butylphenothiazine,58 and 3,7-dimethoxy-phenothiazine59 were prepared according to previously
8
9
reported procedures.
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
2-Iodo-5-tert-butylphenol. 3-tert-Butylphenol (2.55 g, 17.0 mmol), potassium iodide (2.78 g,
16.8 mmol), and NaOH (0.68 g, 17.0 mmol) were dissolved in MeOH (46 mL) under stirring,
cooled with an ice bath. Approximately 22 mL of 6% sodium hypochlorite solution (bleach) was
added drop wise over an hour until the potassium iodide was consumed. On completion, the
reaction mixture was poured into water, the aqueous mixture extracted thrice with EtOAc. The
organic phase washed twice with water, dried over MgSO4, filtered and the solvent removed under
reduced pressure. The residue was purified by column chromatography (9:1, hexanes:EtOAc) to
render the product as a clear oil, which quickly changed to a reddish color and under cooling in a
freezer, formed prismatic crystals (3.27 g, 69%). 1H NMR (400 MHz; CDCl3): δ 7.56 (d, J = 8.3
Hz, 1H), 7.05 (d, J = 2.3 Hz, 1H), 6.74 (dd, J = 2.3, 8.3 Hz, 1H), 5.20 (br, 1H), 1.30 (s, 9H). Spectra
were consistent with previous reports.60
4-(tert-Butyl)-2-chloronitrobenzene. 5-(tert-Butyl)-2-nitroaniline (0.98 g, 5.0 mmol) was
dissolved in 20 mL concentrated HCl at room temperature under vigorous stirring, to which
sodium nitrite (0.42 g, 6.1 mmol) in a minimum of water was added slowly. The suspension was
allowed to stir for 1 hour at room temperature. Copper(I) chloride (50 mg, 0.5 mmol) dissolved in
a small amount of HCl was added slowly and the mixture heated to 100 °C for 1 hour. The reaction
mixture was poured into water and the aqueous mixture extracted thrice with Et2O. The organic
phase was washed twice with water, dried over MgSO4, filtered and the solvent removed under
reduced pressure. The orange residue (0.33 g) was largely the desired product based on crude
spectra (see ESI), though attempts to completely isolate the product from the side products by
column chromatography were unsuccessful, so the crude product was used without further
purification.
4-(tert-Butyl)-1-iodo-2-(2-nitrophenoxy)benzene. 2-Iodo-5-tert-butylphenol (2.79 g, 10.0
mmol) and 2-chloronitrobenzene (1.58 g, 10.0 mmol) were dissolved in DMSO (15 mL) to which
K2CO3 was added (2.81 g, 20.3 mmol) and the reaction mixture was placed under a N2 atmosphere.
With vigorous stirring, the reaction was heated to 100 °C overnight. The reaction mixture was
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 42
1
2
3
4
5
6
7
8
9
poured into water and extracted thrice with EtOAc. The organic phase was washed twice with
water, dried over MgSO4, filtered and the solvent removed under reduced pressure. The product
was purified by column chromatography (9:1, pet. ether:Et2O) to render a clear oil (2.92 g, 73%).
1H NMR (300 MHz; CDCl3): δ 8.01 (dd, J = 1.7, 8.0 Hz, 1H), 7.79 (d, J = 8.3 Hz, 1H), 7.51-7.45
(m, 1H), 7.21-7.16 (m, 1H), 7.09 (d, J = 2.2 Hz, 1H), 7.03 (dd, J = 2.2, 8.3 Hz, 1H), 6.78 (dd, J =
1.1, 8.3 Hz, 1H), 1.28 (s, 9H). 13C NMR (75 MHz; CDCl3): δ 154.4, 154.2, 150.7, 139.6, 134.1,
125.9, 124.5, 122.6, 118.6, 118.2, 85.4, 34.8, 31.1. HRMS (EI, magnetic sector) m/z: M+ Calcd
for C16H16INO3 397.01749, found 397.01553.
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
1-(2-Bromophenoxy)-2-nitro-4-(trifluoromethyl)benzene. Same general procedure where 2-
bromophenol (1.75 g, 10.0 mmol) and 4-chloro-3-nitrobenzotrifluoride (2.26 g, 10.0 mmol) were
heated in DMSO with K2CO3. The resulting oil (3.49 g; 97%) was used in the subsequent reaction
1
without further purification. H NMR (400 MHz; CDCl3): δ 8.28 (d, 2.3 Hz, 1H), 7.72-7.69 (m,
2H), 7.45-7.40 (m, 1H), 7.24-7.18 (m, 2H), 6.87 (d, 8.8Hz, 1H). 13C NMR (101 MHz; CDCl3): δ
153.0, 150.7, 139.6, 134.5, 130.9 (q, JC-F = 3.3Hz), 129.4, 127.7, 125.1 (q, JC-F = 34.9 Hz), 123.7
(q, JC-F = 4.0 Hz), 122.8 (q, JC-F = 272.2 Hz), 122.6, 118.2, 115.8. 19F NMR (376 MHz, CDCl3): δ
-63.40. HRMS (EI, magnetic sector) m/z: M+ Calcd for C13H7BrF3NO3 360.95614, found
360.95252.
1-Iodo-2-(2-nitrophenoxy)-4-(trifluoromethyl)benzene. Same general procedure where 2-iodo-
5-trifluoromethylphenol (1.62 g, 5.6 mmol) and 2-chloronitrobenzene (0.66 g, 4.2 mmol) were
heated in DMSO with K2CO3. Attempts to completely isolate the product from the side products
by column chromatography were unsuccessful, so the crude product was used without further
1
13
purification. See ESI for crude H and C NMR spectra. HRMS (EI, magnetic sector) m/z: M+
Calcd for C13H7IF3NO3 408.94227, found 408.94262.
4-(tert-Butyl)-2-(5-(tert-butyl)-2-iodophenoxy)-1-nitrobenzene. Same general procedure where
2-iodo-5-tert-butylphenol (0.42 g, 1.5 mmol) and 4-(tert-butyl)-2-chloronitrobenzene (0.32 g, 1.5
mmol) were heated in DMSO with K2CO3. The product was purified by column chromatography
1
(9:1, pet. ether:Et2O) to afford a white solid (0.34 g, 50%). H NMR (400 MHz; CDCl3): δ 7.98
(d, J = 8.6 Hz, 1H), 7.79 (dd, J = 0.6, 7.9 Hz, 1H), 7.22 (dd, J = 2.0, 8.6 Hz, 1H), 7.00-6.97 (m,
2H), 6.86 (d, J = 2.0 Hz, 1H), 1.26 (s, 9H), 1.24 (s, 9H). 13C NMR (101 MHz; CDCl3): δ 159.0,
24
ACS Paragon Plus Environment

Page 25 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
154.7, 154.1, 150.0, 139.6, 138.1, 125.9, 123.6, 120.2, 117.1, 116.5, 84.6, 35.4, 34.8, 31.0, 30.8.
HRMS (EI, magnetic sector) m/z: M+ Calcd for C20H24INO3 453.08009, found 453.08007.
7
8
N-(2-(5-(tert-butyl)-2-iodophenoxy)phenyl)acetamide.
4-(tert-Butyl)-1-iodo-2-(2-
9
nitrophenoxy)benzene (2.88 g, 7.2 mmol) and 5 equivalents of SnCl2 (6.83 g, 36.0 mmol) were
dissolved in EtOAc along with 10 equivalents of H2O (1.3 mL). The reaction stirred at room
temperature overnight, then was poured into water and extracted thrice with EtOAc. The organic
phase was washed twice with water, dried over MgSO4, filtered and the solvent removed under
reduced pressure. The residue was dissolved in neat acetic anhydride (3.4 mL) and allowed to stir
overnight. The reaction was quenched with saturated Na2CO3 solution and extracted thrice with
EtOAc. The organic phase was washed once with water, dried over MgSO4, filtered and the solvent
removed under reduced pressure. The resulting product (2.65 g, 90%) was further purified by
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
1
column chromatography (4:1, pet. ether:Et2O) to afford a colorless oil. H NMR (300 MHz;
CDCl3): δ 8.45 (dd, J = 1.3, 8.1 Hz, 1H), 7.84 (br, 1H), 7.79 (d, J = 8.3 Hz, 1H), 7.15-7.09 (m,
13
1H), 7.03-6.96 (m, 3H), 6.69 (dd, J = 1.4, 8.1 Hz, 1H), 2.23 (s, 3H), 1.27 (s, 9H). C NMR (75
MHz; CDCl3): δ 168.3, 154.7, 154.3, 145.3, 139.3, 129.0, 123.73, 123.70, 121.0, 117.5, 115.9,
104.5, 85.0, 34.8, 31.1, 25.0. HRMS (EI, magnetic sector) m/z: M+ Calcd for C18H20INO2
409.05387, found 409.05556.
N-(2-(2-bromophenoxy)-5-(trifluoromethyl)phenyl)acetamide. Same general procedure where
1-(2-bromophenoxy)-2-nitro-4-(trifluoromethyl)benzene (3.47 g, 9.6 mmol) is reduced with
SnCl2, then acetylated in neat acetic anhydride. The resulting solid (3.20 g, 90%) was used in the
1
subsequent reaction without further purification. H NMR (400 MHz; CDCl3): δ 8.82 (br, 1H),
7.94 (br, 1H), 7.70 (dd, J = 1.6, 8.0 Hz, 1H), 7.39 (ddd, J = 1.6, 8.0, 8.0 Hz, 1H), 7.24-7.21 (m,
1H), 7.19-7.15 (m, 1H), 7.12 (dd, J = 1.5, 8.1 Hz, 1H), 6.68 (d, J = 8.5 Hz, 1H), 2.27 (s, 3H). 13C
NMR (101 MHz; CDCl3): δ 168.5, 151.4, 147.7, 134.3, 129.2, 129.0, 126.9, 125.8 (q, JC-F = 33.0
Hz), 123.88 (q, 272.2 Hz), 122.05, 120.69 (q, JC-F = 3.7 Hz), 117.94 (q, JC-F = 3.7 Hz), 115.54,
114.90, 24.89. 19F NMR (376 MHz, CDCl3): δ -63.11. HRMS (EI, magnetic sector) m/z: M+ Calcd
for C15H11BrF3NO2 372.99253, found 372.99317.
N-(2-(2-iodo-5-(trifluoromethyl)phenoxy)phenyl)acetamide. Same general procedure where 1-
iodo-2-(2-nitrophenoxy)-4-(trifluoromethyl)benzene (0.41 g, 1.0 mmol) is reduced with SnCl2,
then acetylated in neat acetic anhydride. The isolated crude solid was recrystallized from hexanes
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 42
1
2
3
4
5
6
7
8
9
to afford pure white needles (0.21 g, 50%); mp 114-115 °C; 1H NMR (400 MHz; CDCl3): δ 8.48
(d, J = 8.4 Hz, 1H), 8.03 (d, J = 8.2 Hz, 1H), 7.69 (b, 1H), 7.23-7.08 (m, 4H), 6.80 (dd, J = 1.0,
8.2 Hz, 1H), 2.22 (s, 3H). 13C NMR (101 MHz; CDCl3): δ 168.3, 155.9, 144.3, 140.7, 132.6 (q, JC-
F = 33.4 Hz), 129.5, 125.2, 124.1, 123.2 (q, JC-F = 272.5 Hz), 122.3 (q, JC-F = 3.7 Hz), 121.5, 117.3,
115.3 (q, JC-F = 3.7 Hz), 92.8, 24.9. 19F NMR (376 MHz, CDCl3): δ -64.09. HRMS (EI, magnetic
sector) m/z: M+ Calcd for C15H11F3INO2 420.97866, found 420.97881.
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-(4-(tert-butyl)-2-(5-(tert-butyl)-2-iodophenoxy)phenyl)acetamide. 4-(tert-Butyl)-2-(5-(tert-
butyl)-2-iodophenoxy)-1-nitrobenzene (0.71 g, 1.6 mmol) and 5 equivalents of SnCl2 dihydrate
(1.52 g, 8.0 mmol) were dissolved in EtOAc. The reaction was stirred at room temperature
overnight, then poured into water and extracted thrice with Et2O. The organic phase was washed
twice with water, dried over MgSO4, filtered and the solvent removed under reduced pressure. The
residue was dissolved in THF (10 mL) to which were added triethylamine (0.30 mL, 2.1 mmol)
and acetyl chloride (0.13 mL, 1.8 mmol). The mixture was stirred overnight, then poured into
water and extracted twice with Et2O. The organic phase was washed twice with water, dried over
MgSO4, filtered and the solvent removed under reduced pressure. The resulting product (0.68 g,
91%) was further purified by column chromatography (9:1 hexanes:EtOAc) to afford a colorless
1
oil. H NMR (300 MHz; CDCl3): δ 8.32 (d, J = 8.6 Hz, 1H), 7.78 (d, J = 8.6 Hz, 1H), 7.72 (br,
1H), 7.16 (dd, J = 2.2, 8.6 Hz, 1H), 6.97-6.93 (m, 2H), 6.86 (d, J = 2.2 Hz, 1H), 2.19 (s, 3H), 1.24
13
(s, 18H). C NMR (75 MHz; CDCl3): δ 168.2, 155.0, 154.0, 147.3, 144.4, 139.2, 126.8, 122.9,
120.9, 120.8, 115.7, 114.6, 84.0, 34.8, 34.5, 31.2, 31.0, 24.8. HRMS (EI, magnetic sector) m/z:
M+ Calcd for C22H28INO2 465.11647, found 465.11865.
2-tert-Butyl-10-acetylphenoxazine. Into a N2 flushed sealed tube were added N-(2-(4-(tert-
butyl)-2-iodophenoxy)phenyl)acetamide (3.42 g, 8.3 mmol), K2CO3 (2.32 g, 16.8 mmol),
copper(I) iodide (86 mg, 0.05 eq.), N,N'-dimethylethylenediamine (90 μL, 0.10 eq.), and toluene
(35 mL). The reaction mixture was heated to 120 °C for at least 10 hours under vigorous stirring,
After completion, the reaction mixture was diluted with dichloromethane and passed through a
plug of silica, after which the solvent was removed under reduced pressure. The crude product was
purified by column chromatography (3:1, hexanes:EtOAc) to render 2-tert-butyl-10-
acetylphenoxazine as an off-white solid (1.65 g, 71%). 1H NMR (400 MHz; CDCl3): δ 7.53 (d, J
26
ACS Paragon Plus Environment

Page 27 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
= 7.9 Hz, 1H), 7.47 (d, J = 2.2 Hz, 1H), 7.23 (dd, J = 2.4, 8.6 Hz, 1H), 7.21-7.10 (m, 3H), 7.06 (d,
J = 8.5 Hz, 1H), 2.34 (s, 3H), 1.34 (s, 9H). Spectra were consistent with previous reports.57
7
8
9
3-tert-Butyl-10-acetylphenoxazine. Same general procedure using N-(2-(5-(tert-butyl)-2-
iodophenoxy)phenyl)acetamide (2.63 g, 6.4 mmol) for the CuI catalyzed intramolecular Ullmann
coupling. The product (1.26 g, 70%) was recrystallized from hexanes to render 3-tert-butyl-10-
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
1
acetylphenoxazine as white needles; mp 139-140 °C; H NMR (400 MHz; CDCl3): δ 7.50-7.47
(m, 1H), 7.42-7.39 (m, 1H), 7.22-7.11 (m, 5H), 2.34 (s, 3H), 1.33 (s, 9H). 13C NMR (101 MHz;
CDCl3): δ 169.3, 151.1, 150.7, 150.6, 129.6, 126.8, 125.1, 124.4, 123.2, 120.3, 116.9, 114.0, 34.7,
31.2, 23.0. HRMS (EI, magnetic sector) m/z: M+ Calcd for C18H19NO2 281.14158, found
281.14291.
2-Trifluoromethyl-10-acetylphenoxazine. Same general procedure using N-(2-(2-
bromophenoxy)-5-(trifluoromethyl)phenyl)acetamide (3.18 g, 8.5 mmol) for the CuI catalyzed
intramolecular Ullmann coupling. The crude product was purified by column chromatography
(3:1, hexanes:EtOAc) to render 2-trifluoromethyl-10-acetylphenoxazine as an off-white solid
(2.12 g, 85%). 1H NMR (400 MHz; CDCl3): δ 7.87 (d, J = 1.8 Hz, 1H), 7.48-7.46 (m, 1H), 7.42-
7.39 (s, 1H), 7.26-7.15 (m, 4H), 2.35 (s, 3H). Spectra were consistent with previous reports.57
O-acetyl-3-hydroxy-10-acetylphenoxazine. Phenoxazone (0.80 g, 4.1 mmol) and anhydrous
SnCl2 (2.40 g, 12.7 mmol) were suspended in acetic anhydride (10 mL) with triethylamine (110
μL, 0.2 eq.). The reaction was heated to 120 °C under vigorous stirring and reaction progress
monitored by TLC. After completion, the reaction was quenched in saturated Na2CO3 and the
aqueous mixture extracted thrice with EtOAc. The organic phase was washed twice with water,
dried over MgSO4, filtered and the solvent removed under reduced pressure. This afforded the
crude product as a light yellow solid (0.64 g, 56%) which was further purified by column
1
chromatography (9:1, hexanes:EtOAc). H NMR (400 MHz; benzene-d6): δ 7.26-7.21 (m, 2H),
6.93 (d, J = 2.6 Hz, 1H), 6.89-6.87 (m, 1H), 6.79-6.72 (m, 2H), 6.69 (dd, J = 2.6, 8.7 Hz, 1H), 1.75
(s, 3H), 1.68 (s, 3H). 13C NMR (101 MHz; benzene-d6): δ 168.6, 168.5, 152.0, 151.4, 149.7, 130.4,
128.0, 127.1, 126.1, 125.9, 124.0, 117.3, 117.0, 111.2, 22.8, 20.7. HRMS (EI, magnetic sector)
m/z: M+ Calcd for C16H13NO4 283.08446, found 283.08662.
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 42
1
2
3
4
5
6
7
8
9
3-Methoxy-10-acetylphenoxazine. O-acetyl-3-hydroxy-10-acetylphenoxazine (0.59 g, 2.1
mmol) was dissolved in methanol (20 mL) under a N2 atmosphere. Stirring at room temperature,
4 mL of aqueous 10% K2CO3 solution was added. The reaction was monitored by TLC, with the
starting material (1:1 Hexane:EtOAc, Rf = 0.50) giving way to the hydrolyzed intermediate (Rf =
0.40). The reaction mixture was quenched with 1 M HCl and extracted thrice with EtOAc. The
organic phase was dried over MgSO4, filtered and the solvent removed under reduced pressure.
The green residue was immediately dissolved into DMF (15 mL) to which K2CO3 was added (0.75
g, 5.4 mmol) and the mixture was placed under a N2 atmosphere. Under vigorous stirring, methyl
iodide was added (170 μL, 2.7 mmol) and the reaction left to stir overnight. On completion, the
reaction mixture was poured into water, and the aqueous mixture extracted thrice with EtOAc. The
organic phase was washed twice with water, dried over MgSO4, filtered and the solvent removed
under reduced pressure. The residue was purified by column chromatography (3:1,
hexanes:EtOAc) to render 3-methoxy-10-acetylphenoxazine as an amber coloured transparent
residue (0.38 g, 71%). 1H NMR (400 MHz; benzene-d6): δ 7.37 (br, 1H), 7.20 (br, 1H), 6.98-6.96
(m, 1H), 6.83-6.76 (m, 2H), 6.63 (d, J = 2.7 Hz, 1H), 6.44 (dd, J = 2.8, 8.8 Hz, 1H), 3.18 (s, 3H),
1.83 (s, 3H). 13C NMR (101 MHz; benzene-d6): δ 168.8, 159.1, 152.6, 151.6, 130.9, 127.0, 126.3,
126.0, 123.8, 123.5, 117.2, 109.7, 102.9, 55.4, 22.8. HRMS (EI, magnetic sector) m/z: M+ Calcd
for C15H13NO3 255.08954, found 255.09214.
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
3,7-Dimethoxy-10-acetylphenoxazine. 3,7-Dihydroxy-10-acetylphenoxazine (0.69 g, 2.7 mmol)
and methyl iodide (340 μL, 5.5 mmol) were dissolved in DMF (15 mL) to which K2CO3 was added
(0.79 g, 5.7 mmol). The mixture was stirred overnight at room temperature, then poured into water
and the aqueous mixture extracted thrice with EtOAc. The organic phase was washed twice with
water, dried over MgSO4, filtered and the solvent removed under reduced pressure. An orange
residue was afforded which solidified overnight on standing in a freezer. The solid was
recyrstallized twice from ethanol to afford orange needles (0.42 g, 55%); mp 118-120 °C; 1H NMR
(400 MHz; CDCl3): δ 7.38-7.36 (m, 2H), 6.70-6.67 (m, 4H), 3.81 (s, 6H), 2.29 (s, 3H). 13C NMR
(101 MHz; CDCl3): δ 169.6, 158.3, 151.7, 125.5, 122.6, 109.0, 102.5, 55.7, 22.8. HRMS (EI,
magnetic sector) m/z: M+ Calcd for C16H15NO4 285.10011, found 285.09708.
3-Nitro-10-acetylphenoxazine. In glacial acetic acid (12 mL) was suspended 10-
acetylphenoxazine (1.12 g, 5.0 mmol) and solution of HNO3 diluted in acetic acid (3:4,
28
ACS Paragon Plus Environment

Page 29 of 42
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
HNO3:AcOH, 1.4 mL) was added with swirling for ~3 minutes where the suspension dissolves to
give a single red phase. The reaction mixture is allowed to stand (without stirring) for ~15 minutes
after which the reaction was poured into 50 mL of water. The resulting precipitate was filtered and
recrystallized from ethanol to afford the 3-nitro-10-acetylphenoxazine as orange needles (1.11 g,
82%); mp 136-138 °C; 1H NMR (400 MHz; DMSO-d6): δ 8.08 (dd, J = 2.7, 8.9 Hz, 1H), 8.00 (d,
J = 2.6 Hz, 1H), 7.89 (d, J = 8.9 Hz, 1H), 7.65 (dd, J = 1.9, 7.9 Hz, 1H), 7.36-7.23 (m, 3H), 2.32
(s, 3H). 13C NMR (101 MHz; DMSO-d6): δ 168.9, 150.2, 149.4, 145.3, 135.2, 128.3, 127.6, 125.9,
125.2, 124.4, 119.0, 116.8, 111.9, 22.9. HRMS (EI, magnetic sector) m/z: M+ Calcd for
C14H10N2O4 270.06406, found 270.06128.
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
2-(2,4-Dinitrophenyl)amino-5-methoxyphenol. 2-Amino-5-methoxyphenol (0.32 g, 2.3 mmol)
and 1-chloro-2,4-dinitrobenzene (0.47 g, 2.3 mmol) were dissolved in a mixture of ethanol (25
mL) and water (5 mL) to which sodium acetate (0.76 g, 9.3 mmol) was added. The solution was
heated to reflux, under vigorous stirring for 4 hours, then allowed to cool. The reddish precipitate
was filtered and purified by column chromatography (3:1, hexanes:EtOAc) to afford the product
as red needles (0.59 g, 84% yield); mp 168-170 °C; 1H NMR (400 MHz; DMSO-d6): δ 9.96 (br,
1H), 9.81 (br, 1H), 8.89 (d, J = 2.7 Hz, 1H), 8.21 (dd, J = 2.7, 9.6 Hz, 1H), 7.17 (d, J = 8.6 Hz,
1H), 6.78 (d, J = 9.6 Hz, 1H), 6.57 (d, J = 2.7 Hz, 1H) 6.52 (dd, J = 2.7, 8.6 Hz, 1H), 3.75 (s, 3H).
13C NMR (101 MHz; DMSO-d6): δ 159.5, 153.5, 147.7, 135.7, 130.4, 129.6, 128.7, 123.3, 117.1,
117.0, 105.3, 102.2, 55.2. HRMS (EI, magnetic sector) m/z: M+ Calcd for C13H11N3O6 305.06479,
found 305.06249.
3-Methoxy-7-nitrophenoxazine (13). 2-(2,4-Dinitrophenyl)amino-5-methoxyphenol (0.22 g, 0.7
mmol) was dissolved in dry DMF (4 mL) and heated to 100 °C with stirring under a N2 atmosphere.
At 100 °C, crushed NaOH was added (70 mg, 1.75 mmol) and the mixture was heated to 120 °C
for 2 hours. The reaction mixture was poured into water and extracted 3 times with Et2O. The
organic solution was washed twice with water, dried over MgSO4, filtered and the solvent removed
under reduced pressure. The product was purified by column chromatography (3:1,
1
hexanes:EtOAc) to afford a deep red solid (0.10 g, 55%) H NMR (400 MHz; DMSO-d6): δ 9.27
(br, 1H), 7.69 (dd, J = 2.6, 8.8 Hz, 1H), 7.31 (d, J = 2.6 Hz), 6.48 (d, J = 8.5 Hz, 1H), 6.47 (d, J =
8.8 Hz, 1H), 6.40 (dd, J = 2.7, 8.5 Hz, 1H), 6.36 (d, J = 2.7 Hz, 1H), 3.65 (s, 3H). 13C NMR (75
MHz; DMSO-d6): δ 155.3, 142.8, 141.4, 139.9, 138.8, 122.5, 122.2, 114.6, 111.6, 110.0, 108.7,
29
ACS Paragon Plus Environment