Diazaphenoxazines and Diazaphenothiazines: Synthesis of the "correct" Isomers Reveals They Are Highly Reactive Radical-Trapping Antioxidants

Article abstract of DOI:10.1021/acs.orglett.7b00615

The preparation of 2,4-diazaphenothiazines and 2,4-diazaphenoxazines via a copper-catalyzed intramolecular amination is described. Literature approaches which utilize easily accessed (2′-aminophenyl) 4-pyri(mi)dyl sulfides undergo a Smiles rearrangement that gives rise to the 1,3-diaza derivatives instead, confirmed by X-ray crystallography. Inversion of the polarity of the cyclization avoids the rearrangement and affords the desired products. Preliminary kinetic studies suggest that 2,4-diazaphenothiazines and diazaphenoxazines, but not the 1,3-diaza isomers, are remarkably potent radical-trapping antioxidants.

Full text of DOI:10.1021/acs.orglett.7b00615

Letter
pubs.acs.org/OrgLett
Diazaphenoxazines and Diazaphenothiazines: Synthesis of the
“
Correct” Isomers Reveals They Are Highly Reactive
Radical‑Trapping Antioxidants
Evan A. Haidasz and Derek A. Pratt*
Department of Chemistry and Biomolecular Sciences, University of Ottawa, Ottawa, Ontario, Canada K1N 6N5
*
S Supporting Information
ABSTRACT: The preparation of 2,4-diazaphenothiazines and 2,4-diazaphenoxazines via a copper-catalyzed intramolecular
amination is described. Literature approaches which utilize easily accessed (2′-aminophenyl) 4-pyri(mi)dyl sulfides undergo a
Smiles rearrangement that gives rise to the 1,3-diaza derivatives instead, confirmed by X-ray crystallography. Inversion of the
polarity of the cyclization avoids the rearrangement and affords the desired products. Preliminary kinetic studies suggest that 2,4-
diazaphenothiazines and diazaphenoxazines, but not the 1,3-diaza isomers, are remarkably potent radical-trapping antioxidants.
henoxazines and phenothiazines (i.e., 1 and 2, Figure 1)
than the industry-standard alkylated DPA at ambient temper-
1
−5
18,19
P
figure prominently in medicinal chemistry,
photophysical properties have enabled their use in dye-sensitized
and their
atures,
a difference that translates to the inhibition of high
2
0
temperature autoxidations of heavy hydrocarbons by 5.
6
−8
9−11
solar cells,
organic light emitting diodes,
and as photo-
Reasoning that phenoxazine and phenothiazine would present a
better scaffold for further optimization, we sought to incorporate
oxygen and sulfur into the structures of 4, 5, and related
heterocyclic diarylamines.
1
2
redox catalysts. Phenothiazines have also long been known to
13
possess potent antioxidant activity and have been pursued as
additives to protect petroleum-derived products, including
lubricants, rubber, and fuels from (peroxyl) radical-mediated
autoxidation. At ambient temperatures, phenothiazine is 440-fold
14
more reactive to peroxyl radicals than diphenylamine (3), the
base structure of the industry-standard alkylated diphenylamine
1
5,16
(
DPA) radical-trapping antioxidants (RTAs).
Phenoxazine is
1
4,17
Although there has been some interest in aza-analogues of
a further 3-fold more reactive under the same conditions.
2
1−29
phenoxazine and phenothiazine,
2-aza, 4-aza-, and 2,4-diaza
30,31
derivatives have received little attention.
Of particular
relevance, Bakavoli and co-workers reported the preparation of
phenothiazine analogs of 4/5 (Scheme 1) and their study as
31−34
inhibitors of 15-lipoxygenase catalysis.
When we attempted
to reproduce Bakavoli’s synthesis of a representative 2,4-
diazaphenothiazine (the 3-pyrrolidine-substituted derivative 6),
the crude material contained only trace amounts of the product.
Figure 1. General structures of phenoxazine (1), phenothiazine (2), and
diphenylamine (3) and the rate constants for their reactions with peroxyl
33
radicals (k , in chlorobenzene).
inh
However, upon recrystallization (as reported), a significant
quantity of the major product was converted to the initially trace
(
reported) product. We surmised that our crude material may
Previous work by our group has shown that incorporation of
heteroatoms at the 3- and/or 5-position of one or both of the
aromatic rings of diarylamines greatly stabilizes them toward one
have undergone oxidation during recrystallization. Indeed, when
the reaction was repeated and the crude material was treated with
a basic methanolic solution of iodine, the reported product was
obtained. This product was characterized by us to be disulfide 7
18,19
electron oxidation.
This modification enables ring sub-
stitution with strongly electron-donating groups, which weaken
the N−H bond and increases the rate of H atom transfer to
peroxyl radicals without compromising stability to one-electron
oxidation. For example, diarylamine 4 is ∼20-fold more reactive
Received: February 28, 2017
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b00615
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Reported Synthesis of 3-Amino-2,4-
diazaphenothiazines Analogous to 4/5
To suppress the Smiles rearrangement, the polarity of the
reaction was reversed; that is, the positions of the amine and
bromide substituents on the pyrimidyl and phenyl rings were
37
exchanged (i.e., 11 in Scheme 4). Following acetylation to give
Scheme 4. Preparation of 2,4-Diazaphenothiazine 6 and Its
Structure As Determined by Single Crystal X-ray Diffraction
(see Supporting Information (SI)), implying that the initial
product was the corresponding thiol, and the difference between
our experiment and that of Bakavoli was simply that in situ
oxidation of the thiol(ate) was minimized in our hands. Thus, 8
35
undergoes a Smiles rearrangement under these conditions, with
the amide displacing the thiolate as in Scheme 2.
Scheme 2. Literature Conditions for Cyclization of 8 to 6
Yields the Rearranged Thiol and Corresponding Disulfide 7
1
2, the cyclization proceeded with excellent conversion, and
subsequent deprotection provided the desired 2,4-diazapheno-
thiazine 6 in 86% (isolated) yield. Given that our H and
1
13
C
NMR spectral data were significantly different than those
reportedbyBakavolietal. (whichwehavenowshowncorrespond
33
to disulfide 7), we determined the structure of our product
unambiguously by single-crystal X-ray diffraction in order to
confirm that it was indeed 6.
The synthesis of the corresponding 2,4-diazaphenoxazine
derivative was similarly problematic. The preparation of 13, the
diaryl ether analogous to diaryl thioether 8 (Scheme 2), required
explicit prior formation of the phenolate, and despite the poorer
leaving group, the Smiles rearrangement proceeded readily upon
installation of the diethylamino substitutent to give an ∼1:1
mixture of 13 and 14 (Scheme 5). Subsequent cyclization of 14
In an attempt to favor cyclization over the Smiles rearrange-
ment, we converted the amine to the less nucleophilic acetamide
Scheme 5. Preparation of 1,3-Diazaphenoxazine 15
(9)andsubjectedittoUllmann-typeconditions(Scheme3). This
afforded a new product whose structure was determined
unambiguously by single-crystal X-ray diffraction to be the 1,3-
diazaphenothiazine 10. Thus, although the cyclization proceeded
(
in good conversion, but poor yield after chromatography), it
follows the Smiles rearrangement. Unfortunately, the same was
true when the bromide was exchanged with an iodide. Evidently,
these substrates are too activated to the Smiles rearrangement,
36
which copper may even promote.
Scheme 3. Preparation of 1,3-Diazaphenothiazine 10 and Its
Structure As Determined by Single Crystal X-ray Diffraction
gave the 1,3-diazaphenoxazine 15. However, as was the case with
the sulfur analog, exchanging the amine and halide substituents
provided the desired 2,4-diazaphenoxazine 16 (Scheme 6). To
the best of our knowledge, 16 is the first reported 2,4-
diazaphenoxazine.
With the two isomers of representative diazaphenothiazines
and diazaphenoxazines in hand, the RTA activity of each was
measured by the inhibited autoxidation of 1,4-dioxane. The
autoxidations were monitored by following the consumption of
B
DOI: 10.1021/acs.orglett.7b00615
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
1
8
Scheme 6. Preparation of 2,4-Diazaphenoxazine 16
positions is highly detrimental. The inhibited periods (of ca.
600 s) observed for the 2,4-diaza isomers correspond to the
trapping of two chain-carrying peroxyl radicals−as is typical for
1
16,18,19
diarylamineRTAsatambienttemperatures.
Theveryslight
initial rate of the autoxidation inhibited by the 2,4-diazapheno-
thiazine 6 corresponds to a rate constant for its reaction with
dioxane
6
−1 −1
chain-carrying peroxyl radicals of kinh = 5.0 × 10 M s ;
while the 2,4-diazaphenoxazine 16 is so reactive that it completely
suppresses the autoxidation, precluding determination of a rate
constant. To enable comparison of the reactivity of 6 with other
diarylamines, its inhibition rate constant was corrected for the
retarding effect of H-bonding of the aminic H-atom to dioxane
H
using the Ingold-Abraham relationship (eq 4, where Δβ = 0.26 is
PBD-BODIPY, a highly oxidizable and highly absorbing substrate
which is added to the autoxidation to enable reaction monitoring
2
H
the difference in the H-bond basicity of the solvents and α = 0.44
2
4
0,41
3
8,39
is the H-bond acidity of the H-atom donor; see SI).
The
by conventional spectrophotometry (eq 1).
The rate
PhCl
7
−1 −1
corresponding value of kinh = 4.5 × 10 M
s
exceeds that of
1
9
any diarylamine RTA ever reported. Given that the 2,4-
diazaphenoxazine 16 is aneven better inhibitor than 6 (consistent
with the difference between phenoxazine and phenothiazine, see
Figure 1), its rate constant must be >10 M s , suggesting E
0 since log A for this H-atom transfer is likely to be around 8.
14
8
−1 −1
a
4
2
≈
log(ki Pn hh Cl) = log(kinh
dioxane
2
H
H
2
) + 8.3α Δβ
(4)
constants for the reaction of the amines with chain-carrying
peroxyl radicals and the reaction stoichiometry (kinh and n,
respectively) can be calculated from the initial rate and inhibited
period of PBD-BODIPY consumption through the use of eqs 2
and 3, respectively. Representative autoxidation traces for 6, 10,
The 2-aza and 4-aza phenothiazines and phenoxazines can also
be prepared using the approach described above.
15, and 16 are shown in Figure 2.
−
d[PBD‐BODIPY]
dt
= kROO
[PBD‐BODIPY]Ri
Although compounds of these types have previously been
21,43
nk [RTA]
(2)
(3)
reported,
the conditions under which they have been
inh
prepared are often significantly harsher than those described
n[RTA]
21,44
above (e.g., 10 equiv of activated Cu, 160 °C).
Moreover,
tinh
=
Ri
many of the reported routes are vulnerable to the Smiles
rearrangement, leading to the incorrect (undesired) isomer. For
example, we have found that recent reports of the syntheses of
45
46
substituted 4-azaphenothiazines and 4-azaphenoxazines lead
to 1-azaphenothiazines and 1-azaphenoxazines, respectively (see
SI for complete details). Not unexpectedly, it is difficult to assign
the structures of these molecules simply from their NMR spectra,
1
3
examples of which are included in Figure 3 (see the SI for
C
47
NMRspectra). Thisunderscorestheneedformultiplemodesof
characterization for structural elucidation.
Figure 2. Co-autoxidation of 1,4-dioxane (2.9 M) and PBD-BODIPY
10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C (black) and
(
1
inhibited by 2 μM of 6 (green), 10 (red), 15 (gray), and 16 (blue).
Reaction progress was monitored by absorbance at 587 nm (ε = 123 023
Figure 3. H NMRspectra of 1-azaphenoxazine and 4-azaphenoxazine in
d
-DMSO at 25 °C.
6
−1
−1
M
cm ).
In summary, we have developed a successful strategy to
Although the 1,3-diazaphenothiazine 10 barely inhibits the
synthesize 2,4-diazaphenothiazine and 2,4-diazaphenoxazine
derivatives. Some previously reported syntheses of these
compounds have been mischaracterized because of a favorable
Smiles rearrangement. Reversing the polarity of the intra-
molecular cross-coupling reaction suppresses the rearrangement
and allows successful construction of the central oxazine or
autoxidation and the 1,3-diazaphenoxazine 15 only retards it, the
corresponding 2,4-diaza isomers 6 and 16 are remarkably
effective. The stark difference in reactivity between the isomers
is consistent with the trends we have observed in the simpler
diarylamines, where nitrogen incorporation at tautomerizable
C
DOI: 10.1021/acs.orglett.7b00615
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
thiazine ring. 1,3-Diazaphenoxazine and diazaphenothiazine
derivatives have moderate reactivity as radical-trapping anti-
oxidants, while the corresponding 2,4-diaza derivatives are some
of the fastest RTAs ever reported, trapping peroxyl radicals at
rates approaching diffusion. Given these results, we anticipate
more detailed studies on the synthesis and properties of
azaphenoxazine and phenothiazine derivatives as RTAs.
(16) Ingold, K. U.; Pratt, D. A. Chem. Rev. 2014, 114, 9022−9046.
17) Azaphenothiazine/phenoxazine nomenclature has evolved over
(
the years; historically, 1-azaphenothiazine in the American/British
nomenclature was 4-azaphenothiazine in the German system. Thus,
great care must be taken upon consideration of literature data where
structures are not explicitly shown. Current IUPAC nomenclature is
based on the American/British system (see Figure 1).
(
18) Hanthorn, J.; Valgimigli, L.; Pratt, D. J. Am. Chem. Soc. 2012, 134,
306−8309.
19) Hanthorn, J.; Amorati, R.; Valgimigli, L.; Pratt, D. J. Org. Chem.
012, 77, 6895−6907.
20)Shah, R.;Haidasz, E. A.;Valgimigli, L.;Pratt, D. A. J. Am. Chem. Soc.
015, 137, 2440−2443.
21) Ito, Y.; Hamada, Y. Chem. Pharm. Bull. 1978, 26, 1375−1383.
(22) Okafor, C. O. Int. J. Sulfur Chem. Part B Q. Reports Sulfur Chem.
971, 6, 237−265.
23) Okafor, C. O. Int. J. Sulfur Chem. Part B Q. Reports Sulfur Chem.
971, 6, 345−360.
8
(
2
(
2
(
ASSOCIATED CONTENT
Supporting Information
■
*
S
TheSupportingInformationisavailablefreeofchargeontheACS
Publications website at DOI: 10.1021/acs.orglett.7b00615.
Crystallographic information (CIF, CIF)
Synthetic schemes and procedures, characterization data,
inhibited autoxidation procedures (PDF)
1
(
1
(
24) Takahashi, T.; Yoneda, F. Chem. Pharm. Bull. 1958, 6, 46−49.
25) Takahashi, T.; Yoneda, F. Chem. Pharm. Bull. 1958, 6, 378−381.
26) Petrow, V. A.; Rewald, E. L. J. Chem. Soc. 1945, 313−315.
27) Schuler, W. A.; Klebe, H. Justus Liebigs Ann. Chem. 1962, 653, 172.
28) Pluta, K.; Morak-Młodawska, B.; Jelen, M. J. Heterocycl. Chem.
́
2009, 46, 355−391.
AUTHOR INFORMATION
Corresponding Author
(
(
(
(
■
*E-mail: dpratt@uottawa.ca.
ORCID
(
29) Most of the previous reports of azaphenothiazine and
Derek A. Pratt: 0000-0002-7305-745X
Notes
azaphenoxazine derivatives are described in the patent literature. See
the SI for examples where their preparation is described.
(
1
30) Saggiomo, A. J.; Craig, P. N.; Gordon, M. J. Org. Chem. 1958, 23,
906−1909.
31) Nikpour, M.; Mousavian, M.; Davoodnejad, M.; Alimardani, M.;
Sadeghian, H. Med. Chem. Res. 2013, 22, 5036−5043.
32) Bakavoli, M.; Nikpour, M.; Rahimizadeh, M.; Saberi, M. R.;
The authors declare no competing financial interest.
(
ACKNOWLEDGMENTS
■
This work was supported by grants from the Natural Sciences and
Engineering Research Council of Canada and the Canada
Foundation for Innovation. D.A.P. and E.A.H. also acknowledge
the support of the Canada Research Chairs and Ontario Graduate
Scholarships programs, respectively.
(
Sadeghian, H. Bioorg. Med. Chem. 2007, 15, 2120−2126.
(33) Bakavoli, M.; Sadeghian, H.; Tabatabaei, Z.; Rezaei, E.;
Rahimizadeh, M.; Nikpour, M. J. Mol. Model. 2008, 14, 471−478.
(34) Pooryaghoobi, N.; Bakavoli, M.; Alimardani, M.; Bazzazan, T.;
Sadeghian, H. Iran. J. Basic Med. Sci. 2013, 16, 784−789.
35) Bonvicino, G. E.; Yogodzinski, L. H.; Hardy, R. A. J. Org. Chem.
962, 27, 4272−4280.
36) Truce, W. E.; Kreider, E. M.; Brand, W. W.; Truce, W. E.; Kreider,
(
1
(
REFERENCES
■
(
1) Dumitriu, G.-M.; Bîcu, E.; Belei, D.; Rigo, B.; Dubois, J.; Farce, A.;
Ghinet, A. Bioorg. Med. Chem. Lett. 2015, 25, 4447−4452.
E. M.; Brand, W. W. Organic Reactions; John Wiley & Sons, Inc.:
Hoboken, NJ, USA, 2011; pp 99−215.
(
2) Prinz, H.; Chamasmani, B.; Vogel, K.; Bo
Gerlach,M.;Gunther, E.G.;Amon,P.;Ivanov, I.;Mu
011, 54, 4247−4263.
3) Rajagopalan, R.; Lin, T.-S.; Karwa, A. S.; Poreddy, A. R.; Asmelash,
B.; Dorshow, R. B. ACS Med. Chem. Lett. 2012, 3, 284−288.
4) Ramprasad, J.; Nayak, N.; Dalimba, U. Eur. J. Med. Chem. 2015, 106,
5−84.
5) Marcu, A.; Schurigt, U.; Mu
Prinz, H. Eur. J. Med. Chem. 2016, 108, 436−443.
6) Karlsson, K. M.; Jiang, X.; Eriksson, S. K.; Gabrielsson, E.; Rensmo,
H.; Hagfeldt, A.; Sun, L. Chem. - Eur. J. 2011, 17, 6415−6424.
7)Tian,H.;Yang, X.;Chen, R.;Hagfeldt, A.;Sun,L. EnergyEnviron. Sci.
009, 2, 674−677.
8) Tian, H.; Yang, X.; Chen, R.; Pan, Y.; Li, L.; Hagfeldt, A.; Sun, L.
Chem. Commun. 2007, 353, 3741.
9) Salunke, J. K.; Wong, F. L.; Feron, K.; Manzhos, S.; Lo, M. F.;
̈
hm, K. J.; Aicher, B.;
̈
̈
ller,K.J. Med.Chem.
(
37) A similar strategy has been employed in the synthesis of some
2
(
21
monoazaphenoxazines; although conditions for the cross-coupling are
considerably harsher, they do not report isomerization in the final
product.
(
7
(
(
38) Haidasz, E. A.;VanKessel, A. T. M.;Pratt, D. A. J. Org. Chem. 2016,
1, 737−744.
39) PBD-BODIPY consumption is too slow in autoxidations of more
8
̈
ller, K.; Moll, H.; Krauth-Siegel, R. L.;
(
conventional substrates (e.g., styrene, cumene) that are inhibited by very
reactive RTAs (e.g., phenoxazine), prompting us to select dioxane as a
substrate since it also slows H-atom transfer from good H-bond donors
(
(
2
(
(e.g., amines) because of complex formation; see refs 40 and 41.
(40) Snelgrove, D. W.; Lusztyk, J.; Banks, J. T.; Mulder, P.; Ingold, K. U.
J. Am. Chem. Soc. 2001, 123, 469−477.
(
(
41) Litwinienko, G.; Ingold, K. U. Acc. Chem. Res. 2007, 40, 222−230.
(
19
42) H-atom transfer from diarylamines takes place with log A ≈ 7.
Shinde, D.; Patil, A.;Lee, C. S.; Roy, V. A. L.;Sonar, P.; Wadgaonkar, P. P.
J. Mater. Chem. C 2016, 4, 1009−1018.
10) Data, P.; Pander, P.; Okazaki, M.; Takeda, Y.; Minakata, S.;
Since achieving the transition state for H-atom transfer from the tricyclc
compounds does not require restricting the rotation of the two N-aryl
bonds, the log A is expected to be ∼1 log unit greater.
(
Monkman, A. P. Angew. Chem., Int. Ed. 2016, 55, 5739−5744.
11) Wu, K.; Zhang, T.; Zhan, L.; Zhong, C.; Gong, S.; Jiang, N.; Lu, Z.-
H.; Yang, C. Chem. - Eur. J. 2016, 22, 10860−10866.
12) Pearson, R. M.; Lim, C.-H.; McCarthy, B. G.; Musgrave, C. B.;
Miyake, G. M. J. Am. Chem. Soc. 2016, 138, 11399−11407.
13) Murphy, C. M.; Ravner, H.; Smith, N. L. Ind. Eng. Chem. 1950, 42,
479−2489.
14) Lucarini, M.; Pedrielli, P.; Pedulli, G. F.; Valgimigli, L.; Gigmes, D.;
Tordo, P. J. Am. Chem. Soc. 1999, 121, 11546−11553.
15) Ingold, K. U. Chem. Rev. 1961, 61, 563−589.
(
(
43) Poulenc, R. Azaphen
44) The characterization of initially reported derivatives was generally
́
othiazines. FRD1170119, 1955.
(
21,43
based on IR, elemental analysis, and melting point determination.
(
(45) Hu, W.; Zhang, S. J. Org. Chem. 2015, 80, 6128−6132.
(46) Shen, C.; Wu, X.-F. Catal. Sci. Technol. 2015, 5, 4433−4443.
(47)ItshouldbenotedthatalthoughtheoriginalNMRcharacterization
(
2
(
for these compounds is often incomplete, melting points for 1- and 4-
azaphenoxazine and azaphenothiazine have been reported and are
consistent with those measured by our group.
(
D
DOI: 10.1021/acs.orglett.7b00615
Org. Lett. XXXX, XXX, XXX−XXX