Unexpected reactivity between aromatic nitro compounds and PCl3/AlCl3. A new one-pot synthesis of phenazines

Article abstract of DOI:10.1016/S0040-4039(03)00373-3

The reaction between 4-nitroalkoxybenzenes 7 and PCl₃/AlCl₃, when carried out in appropriate molar ratio, gives a prevalent formation of diazenes 8 and 2,7-dialkoxyphenazines 9 with their new chlorinated derivatives 10-13. These compounds are obtained, in satisfactory yield, in a one-pot procedure, in mild conditions, from commercially available and safe starting materials. In this reaction both the reagents PCl₃ and AlCl₃ play a fundamental role in obtaining the products, and this method might be applicable to other 4-alkoxynitrobenzenes.

Full text of DOI:10.1016/S0040-4039(03)00373-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 2649–2653
Unexpected reactivity between aromatic nitro compounds and
PCl3/AlCl3. A new one-pot synthesis of phenazines
Rodolphe Abdayem,a Graziano Baccolini,a,* Carla Boga,a Magda Monarib and Simona Selvab
aDipartimento di Chimica Organica ‘A. Mangini’, Viale Risorgimento, 4-40136 Bologna, Italy
bDipartimento Di Chimica ‘G. Ciamician’, Via Selmi 2-40126 Bologna, Italy
Received 20 January 2003; revised 6 February 2003; accepted 7 February 2003
Abstract—The reaction between 4-nitroalkoxybenzenes 7 and PCl3/AlCl3, when carried out in appropriate molar ratio, gives a
prevalent formation of diazenes 8 and 2,7-dialkoxyphenazines 9 with their new chlorinated derivatives 10–13. These compounds
are obtained, in satisfactory yield, in a one-pot procedure, in mild conditions, from commercially available and safe starting
materials. In this reaction both the reagents PCl3 and AlCl3 play a fundamental role in obtaining the products, and this method
might be applicable to other 4-alkoxynitrobenzenes. © 2003 Elsevier Science Ltd. All rights reserved.
Since the last century the Friedel–Crafts-type reaction
using PCl3 and AlCl3 has been an excellent method for
the direct attachment of a phosphorus atom to an
aromatic ring to give aryldichlorophosphanes or
diarylchlorophosphanes. However, this reaction is very
sensitive to the type of substituents present on the
aromatic ring1,2 and in particular it fails completely
when thioanisoles are used.3
phosphole6 (5) or of methylphosphonates (6). The latter
arise from an unexpected insertion of phosphorus atom
in the OꢀCH3 bond7 (Scheme 2).
These findings prompted us to check whether the reac-
tion between 4-nitroanisole (7a) and PCl3/AlCl3,
reported to fail, might give 5- and 6-like compounds or
further unexpected results. However, the formation of
5-like compounds is improbable because of the presence
In the past we found4 that the phosphonation reaction
between thioanisoles (1) and PCl3/AlCl3 give, unexpect-
edly and in high yield, a new heterocyclic system,
namely fused [1,2,3]benzothiadiphosphole (2) (Scheme
1).
It addition, it is reported5 that the phosphonation
between anisoles 3 and PCl3/AlCl3 give conflictual
results and small amounts of phosphorylation products
4, together with a large amount of tars. We recently
repeated this reaction in order to understand why it was
unsuccessful and to explore the possibility to obtain
compounds related to 2. For this purpose we carried
out the reaction in different conditions and reagent
ratio, and we found a strong dependence of the reaction
products on the experimental conditions, it being possi-
ble to obtain, by varying the reagent ratio, the forma-
tion, as major products, of fused [1,2,3]benzooxadi-
Scheme 1.
Keywords: phosphorus trichloride; phenazines; diazenes; 4-nitroan-
isole.
* Corresponding author. Tel.: +39-051-2093616; fax: +39-051-
2093654; e-mail: baccolin@ms.fci.unibo.it
Scheme 2.
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00373-3

2650
R. Abdayem et al. / Tetrahedron Letters 44 (2003) 2649–2653
of a nitro group on the aromatic ring, which disfavours
the electrophilic attack occurring in the last step of the
formation pathway of 5.6
The unexpected formation, in a one-pot reaction, of
phenazines 9–13, prompted us to carry out the reaction
with AlCl3 alone, or with PCl3 alone, in order to
investigate the role played by each reagent. We found
that the mixing of 4-nitroanisole (7a) with AlCl3 at
50°C does not produce new products, but probably
only the reversible formation of a complex due to the
coordination of AlCl3 with the oxygen atom, while an
increase of the reaction temperature causes a demethyl-
ation reaction with formation of 4-nitrophenol. Even
the reaction between 7a and PCl3 does not allow
detectable amounts of products. In conclusion, one can
evince that the formation of compounds 8–13 is due to
the concomitant action of the AlCl3/PCl3 couple.
After repeated attempts, varying the reagent ratios and
reaction conditions and monitoring the outcome of the
reaction by GC–MS analysis, we observed that the use
of a reagent ratio 7a:PCl3:AlCl3 of 1:3:1 at 50–55°C
gives a complex mixture of products (Scheme 3) in
which trans-1,2-bis(4-methoxyphenyl)diazene (8a, 30%
yield), 2,7-dimethoxyphenazine (9a, 15%) and its chlori-
nated derivatives (10a–13a, 30% overall yield) were
predominant.
The results obtained with 4-nitroanisole (7a) suggested
that we should extend the study to other nitro aromatic
compounds. In particular, in the same experimental
conditions, we carried out the reaction with 4-ethoxy-
The fact that we observed the formation of traces of
anilino derivatives 14 with all the substrates suggests
the presence of a nitrene as intermediate. The recover-
ing of phenazines and diazenes can also agree with a
mechanism involving nitrenic species,13 from which par-
allel pathways can depart to give the observed products
(Scheme 4).
nitrobenzene,
4-phenoxy-nitrobenzene,
4-bromo-
nitrobenzene, 3-methoxy-nitrobenzene, 3-chloro-nitro-
benzene, and 4-methoxy-2-methyl-1-nitrobenzene: in all
cases traces of anilino derivatives 14-like were observed
in GC–MS, but compounds 8b–12b were otained (8b,
7%, 9b–12b, 34% overall) only using 4-ethoxy-nitrobenz-
ene (7b). It is noteworthy that 10–13 are new com-
pounds and that chlorophenazines are substrates partic-
ularly prone to SNAr and, for this reason, they could be
used as versatile starting materials to obtain other
phenazinic derivatives.
However, the fact that compounds 8–13 are obtained
only when nitro derivatives 7a and 7b are used, could
be explained hypothesizing that the presence in para to
The major reaction products 8–13 were isolated by
chromatography on silica gel and fully characterized.8
In particular, chemico-physical data of compounds
8a,9a 8b,9b 9a10 and 9b11 were in agreement with pub-
lished data. Since the formation of phenazinic nucleus
was unpredictable, the X-ray diffraction analysis12 of a
single crystal of 10a unequivocally confirmed the struc-
ture and the relative position of the substituents. In
addition, the structure of new dichloro isomers 11–13
1
was determined from the analysis of their H and 13C
NMR spectral data.
Figure 1. ORTEP drawing of 10a. Thermal ellipsoids are
drawn at the 30% probability level. The crystallographically
,
independent bond distances (A) and angles (°) are as follows
Figure 1 shows the ORTEP drawing of 10a, whose
molecular symmetry is CS. A second image of the
chlorine atom bonded to C(3%) was found in the crystal;
therefore the molecular model possesses a statistical
inversion centre due to a random packing of molecules
related by a pseudo twofold axis to which all the atoms,
except chlorine, conform.
(the molecular model possesses a statistical C2V symmetry in
the crystal): N(10)ꢀC(6A) 1.312(7), N(10)ꢀC(1A) 1.372(7),
C(1A)ꢀC(1) 1.432(7), C(1)ꢀC(2) 1.340(8), C(2)ꢀC(3) 1.425(9),
C(3)ꢀC(4) 1.330(8), Cl(1)ꢀC(3) 1.550(7), O(1)ꢀC(2) 1.376(6),
O(1)ꢀC(11)
1.418(7);
O(1)ꢀC(2)ꢀC(3)
115.0(6),
Cl(1)ꢀC(3)ꢀC(2) 119.2(5), C(1A)ꢀN(10)ꢀC(6A) 116.9(5),
C(2)ꢀO(1)ꢀC(11) 118.0(5).
Scheme 3.

R. Abdayem et al. / Tetrahedron Letters 44 (2003) 2649–2653
2651
Scheme 4.
It is interesting to note that many syntheses of phenazi-
nes are reported in the literature,16 but they often occur
in very low yields, from starting materials difficult to
procure or by pyrolysis of azides. Up to now, the best
synthesis for the building of phenazinic intermediates
seems to be the so-called ‘Beirut reaction’,17 in which
the reaction between benzofuroxane and hydro-
quinones allows the formation, in good yields and in
mild conditions, of 2-hydroxyphenazines 5,10-dioxide,
versatile substrates for easy functional group manipula-
tion.
the nitro group of strong electron-releasing sub-
stituents, such as alkoxy groups, stabilizes the nitrene,
thus permitting its decay from singlet to triplet state,
with consequent formation of intra- and intermolecular
coupling products.13,14 4-Phenoxy-nitrobenzene and 4-
bromo-nitrobenzene bear substituents with an electron-
releasing power lower than that of alkoxy groups, while
4-methoxy-2-methyl-1-nitrobenzene could probably
have negative steric effects in the cyclization pathway.
It is interesting to note that Cadogan reported15 that
the reaction between nitro- and nitroso compounds and
trialkylphosphites produces nitrenic species, but the
latter cannot be formed using PCl3 alone. In addition, it
is reported15b that the reaction of p-nitroanisole with
triethyl phosphite does not provide diazenes. This
emphasizes the importance, in our case, of the concomi-
tant presence of both, PCl3 and AlCl3, in the reaction
mixture for the formation of diazenes and phenazines.
In conclusion, we found that the reaction between
4-alkoxy-nitrobenzenes (7a,b) and PCl3/AlCl3 gives,
when carried out in appropriate molar ratio, prevalent
formation of 2,7-dialkoxyphenazines and their new
chlorinated derivatives. The importance of our method
lies in the fact that such heterocycles can be obtained,
in satisfactory yield, in a one-pot procedure, in mild
conditions and from commercially available and safe
starting materials. In this reaction both the reagents,
PCl3 and AlCl3 play a fundamental role in obtaining
the products and this method might be applicable to
other 4-alkoxynitrobenzenes.
Furthermore, the finding of chlorinated compounds
reveals that the system 7/PCl3/AlCl3 is extremely com-
plex and might produce molecular chlorine, probably
through oxido-reductive processes. This is supported by
the results obtained carrying out reactions between
compound 9a and both PCl3 and PCl3/AlCl3: in these
cases no chlorinated product was observed, whereas
treatment of a solution of 9a with chlorine gave a
mixture of compounds 10a–13a. The yields of
chloroderivatives 10–13 can thus be increased by simple
bubbling of chlorine in the reaction mixture.
Acknowledgements
Work supported by the University of Bologna (funds
for selected research topics A. A. 2000–2002) and the
Ministero dell’Universita` e della Ricerca Scientifica e

2652
R. Abdayem et al. / Tetrahedron Letters 44 (2003) 2649–2653
Tecnologica. We thank Professor Piero Spagnolo of the
Dipartimento ‘A. Mangini’ of the University of
Bologna for helpful discussion.
3-Chloro-2,7-diethoxyphenazine (10b): yellow solid, 16%
yield mp: 146–147°C (from dichloromethane) Rf=0.60
(n-hexane:diethyl ether, 1:1); lH (300 MHz, CDCl3): 8.18
(s, 1H), 8.01 (d, 1H, J=9.5 Hz), 7.50 (dd, 1H, J=9.5 Hz,
J=2.5 Hz), 7.44 (s, 1H), 7.34 (d, 1H, J=2.5 Hz), 4.30 (q,
2H, J=7.0 Hz), 4.23 (q, 2H, J=7.0 Hz), 1.60 (t, 3H,
J=7.0 Hz), 1.54 (t, 3H, J=7.0 Hz); lC (75.56 MHz,
CDCl3): 160.1, 155.0, 143.6, 141.7, 140.2, 139.3, 131.2,
129.7, 128.5, 126.8, 106.4, 104.9, 65.3, 64.3, 14.5 14.4; MS
(m/z, %): 302 (M+, 72), 274 (20), 246 (100), 217 (33);
HRMS calcd for C16H15ClN2O2: 302.0822, found:
302.0802.
2,7-Dichloro-3,8-dimethoxyphenazine (11a): yellow solid,
3% yield Rf=0.64 (n-hexane:diethyl ether, 1:1); lH (300
MHz, CDCl3): 8.18 (s, 2H,), 7.47 (s, 2H), 4.11 (s, 6H); lC
(75.56 MHz, CDCl3): 156.0, 142.5, 139.8, 131.0, 129.0,
106.1, 56.7; MS (m/z, %): 310 (M++2, 66), 308 (M+, 100),
293 (35), 265 (39); HRMS calcd for C14H10Cl2N2O2:
308.0119, found: 308.0109.
2,7-Dichloro-3,8-diethoxyphenazine (11b): yellow solid, 3%
yield, Rf=0.80 (n-hexane:diethyl ether, 1:1); lH (300
MHz, CDCl3): 8.16 (s, 2H,), 7.42 (s, 2H), 4.32 (q, 4H,
J=7.0 Hz), 1.61 (t, 6H, J=7.0 Hz); lC (75.56 MHz,
CDCl3): 155.3, 142.5, 139.6, 131.3, 128.9, 106.5, 65.3,
14.4; MS (m/z, %): 336 (M+, 40), 280 (100), 251 (23), 216
(13); HRMS calcd for C16H14Cl2N2O2: 336.0432, found:
336.0439.
1,6-Dichloro-3,8-dimethoxyphenazine (12a): yellow solid,
3% yield, Rf=0.63 (n-hexane:diethyl ether, 1:1); lH (300
MHz, CDCl3): 7.58 (d, 2H, J=2.7 Hz), 7.26 (d, 2H,
J=2.7 Hz), 3.93 (s, 6H); lC (75.56 MHz, CDCl3): 159.7,
143.9, 141.0, 133.3, 125.4, 106.7, 56.9; MS (m/z, %): 310
(M++2, 64), 308 (M+, 100), 293 (34), 265 (45); HRMS
calcd for C14H10Cl2N2O2: 308.0119, found: 308.0122.
1,6-Dichloro-3,8-diethoxyphenazine (12b): yellow solid, 3%
yield, Rf=0.70 (n-hexane:diethyl ether, 1:1); lH (300
MHz, CDCl3): 7.54 (d, 2H, J=2.7 Hz), 7.32 (d, 2H,
J=2.7 Hz), 4.26 (q, 4H, J=7.0 Hz), 1.62 (t, 6H, J=7.0
Hz); MS (m/z, %): 336 (M+, 37), 280 (100), 251 (25);
HRMS calcd for C16H14Cl2N2O2: 336.0432, found:
336.0436.
References
1. Michaelis, A. Ber. Dtsch. Chem. Ges. 1879, 12, 1009.
2. Fild, M.; Schmutzler, R. In Organic Phosphorus Com-
pounds; Kosolapoff, G.; Maier, L., Eds.; Wiley-Inter-
science, 1972; Vol. 4, p. 79.
3. Sime´on, F.; Jaffre`s, P.; Villemin, D. Tetrahedron 1998,
54, 10111–10118.
4. (a) Baccolini, G.; Mezzina, E.; Todesco, P. E.; Foresti, E.
J. Chem. Soc., Chem. Commun. 1988, 304; (b) Baccolini,
G.; Mezzina, E.; Todesco, P. E. J. Chem. Soc., Perkin
Trans. 1 1988, 3281.
5. Miles, J. A.; Beeny, M. T.; Ratts, K. W. J. Org. Chem.
1975, 40, 343.
6. Baccolini, G.; Bazzocchi, M.; Boga, C. Eur. J. Org.
Chem. 2001, 2229.
7. Baccolini, G.; Boga, C. Synlett 1999, 822.
8. Typical procedure: 4-nitroanisole (7a) (4.99 g, 0.033 mol),
PCl3 (8.64 mL, 0.099 mol), and AlCl3 (4.41 g, 0.033 mol)
were mixed under nitrogen atmosphere into a dried
apparatus equipped with a mechanical stirrer and kept
at 0°C with an ice bath. After 1 h, the temperature was
increased to 50–55°C and the reaction monitored by
GC–MS analysis. After about 24 h, the reaction mixture
was treated with water and dried over anhydrous MgSO4.
The residue was mixed with silica gel and extracted in
continuous with CHCl3 in a Soxhlet apparatus. After
concentration ‘in vacuo’, the flash chromatography of the
residue gave compounds 8–13. Compounds 8a, 8b, 9a, 9b
were isolated in 30, 7, 15 and 12% yield, respectively.
The overall yield of compounds 9a–13a was 45% and of
9b–12b was 34%. Compounds 11b–13 were contaminated
with traces of other isomers. Compounds obtained in
traces were identified by comparison with authentic com-
mercial samples or by GC–MS analysis. Even if com-
pound 10a was cited in literature,18 the unique data
reported are the molecular peak and the melting point of
a sample contaminated with 9a. Here we report other
data on this.
1,7-Dichloro-3,8-dimethoxyphenazine (13a): yellow solid,
2% yield, Rf=0.60 (n-hexane:diethyl ether 1:1); lH (300
MHz, CDCl3): 8.20 (s, 1H), 7.61 (d, 1H, J=2.6 Hz), 7.45
(s, 1H), 7.23 (d, 1H, J=2.76 Hz), 4.10(s, 3H), 3.91 (s,
3H); lC (75.56 MHz, CDCl3): 157.85, 143.2, 140.4, 132.2,
127.2, 108.4, 56.8; MS (m/z,%): 310 (M++2, 63), 308 (M+,
100), 293 (27), 265 (47); HRMS calcd. for
C14H10Cl2N2O2: 308.0129, found: 308.0118.
trans-1,2 Bis(4-methoxyphenyl) diazene (8a): yellow solid,
30% yield, impure of traces of cis-isomer, Rf=0.76 (n-
hexane:diethyl ether, 1:1); lH (300 MHz, CDCl3): 7.27 (d,
4H, J=9.1 Hz), 6.95 (d, 4H, J=9.1 Hz), 3.88 (s, 6H,
OCH3), lC (75.56 MHz, CDCl3): 161.5, 147.1, 124.3,
114.1, 55.5; MS (m/z,%): 242 (M+, 87), 135 (35), 107
(100), 92 (37), 77 (42).
9. (a) Barbero, M.; Degani, I.; Dughera, S.; Fochi, R.;
Perracino, P. Synthesis 1988, 1235; (b) Jaffari, G. A.;
Nunn, A. J. J. Chem. Soc (C) 1971, 823.
10. Zingaro, R. A.; Herrera, C. Bull. Chem. Soc. Jpn. 1989,
62, 1382.
11. Sayo, H.; Mori, K.; Michida, T. Chem. Pharm. Bull.
1979, 27, 2316.
12. Crystallographic studies of 10a: C14H10Cl1N2O2, M=
273.69, monoclinic, space group P21/n, a=3.9575(5), b=
Selected data for new chloro compounds:
3-Chloro-2,7-dimethoxyphenazine (10a): yellow solid, 22%
yield mp: 236–238°C (from dichloromethane) Lit.18: 234–
237°C, Rf=0.54 (n-hexane:diethyl ether, 1:1); lH (300
MHz, CDCl3): 8.19 (s, 1H,), 8.03 (d, 1H, J=9.6 Hz), 7.51
(dd, 1H, J=9.6 Hz, J=2.8 Hz), 7.49 (s, 1H), 7.38 (d, 1H,
J=2.8 Hz), 4.11 (s, 3H, OCH3), 4.02 (s, 3H, OCH3); lC
(75.56 MHz, CDCl3): 160.7, 155.6, 143.9, 141.9, 140.5,
139.7, 130.6, 130.0, 128.9, 126.3, 106.2, 104.8, 56.7, 55.9:
MS (m/z, %): 276 (M++2, 35), 274 (M+, 100), 259 (18),
231 (53), 216 (7), 188 (14); HRMS calcd for
C14H11ClN2O2: 274.0509, found: 274.0512.
,
18.188(2), c=8.7141(9) A, i=102.316(4)°, V=612.80(12)
,
3
A , T=293(2) K, Z=2, Dcalcd=1.483 mg/m3, v(Mo
Ka)=0.310 mm−1, 6036 reflections collected, 1284 unique
which were used in all calculations. The value of the

R. Abdayem et al. / Tetrahedron Letters 44 (2003) 2649–2653
2653
goodness-of-fit indicator was 0.965. Final R1(F)=0.0904
[I>2|(I)] and wR2(F2)=0.2641 (all data). The crystallo-
graphic data for 10a have been deposited with the Cam-
bridge Crystallographic Data Centre as supplementary
publication numbers CCDC 200015. Copies of the data can
be obtained, free of charge, on application to CCDC, 12
Union Road, Cambridge, CB2 1EZ, UK [fax: +44(0)-1223-
336033 or e-mail: deposit@ccdc.cam.ac.uk].
Sons: New York, 1970; (c) Iddon, B.; Meth-Cohn, O.;
Scriven, E. F. V.; Suschitzky, H.; Gallagher, P. T. Angew.
Chem., Int. Ed. Engl. 1979, 18, 900 and references cited
therein.
15. (a) Cadogan, J. I. G.; Todd, M. J. J. Chem. Soc. (C) 1969,
2808; (b) Cadogan, J. I. G.; Sears, D. J.; Smith, D. M.;
Todd, M. J. J. Chem. Soc. (C) 1969, 2813; (c) Cadogan,
I. J. Acc. Chem. Res. 1972, 5, 303.
16. Porter, A. E. A. In Comprehensive Heterocyclic Chemistry;
Katritzky, A. R.; Rees, C. W.; Lwowski, W., Eds. Pyrazi-
nes and their Benzo Derivatives; Pergamon Press: New
York, 1984; Vol. 3, Part 2, pp. 157–197.
17. Haddadin, M. J.; Issidorides, C. H. J. Org. Chem. 1966,
31, 4067.
13. (a) Walker, P.; Waters, W. A. J. Chem. Soc. 1962, 1632;
(b) Nay, B.; Scriven, E. F. V.; Suschitzky, H.; Thomas, D.
R. J. Chem. Soc., Perkin Trans. 1 1980, 611.
14. (a) Smith, P. A. In Azides and Nitrenes-Reactivity and
Utility; Scriven, E. F. V., Ed. Aryl and Heteroaryl Azides
and Nitrenes; Academic Press: New York, 1984; Chapter
3, pp. 95–197; (b) Nitrenes; Lwowski, W., Ed.; J. Wiley &
18. Cheng, J. D.; Shine, H. J. J. Org. Chem. 1975, 40, 703.