A facile two-step synthesis of 1,3,5-trinitrobenzene

Article abstract of DOI:10.1055/s-2004-815958

Synthesis of 1,3,5-trinitrobenzene was achieved by converting phloroglucinol into its trioximes by reaction with aqueous hydroxylamine followed by oxidation of the trioxime product with 90% nitric acid.

Full text of DOI:10.1055/s-2004-815958

SHORT PAPER
499
A Facile Two-Step Synthesis of 1,3,5-Trinitrobenzene
Synthesis
e
of 1,3,5-T
r
f
initroben
f
zene rey C. Bottaro,* Ripudaman Malhotra, Allen Dodge
Chemical Science and Technology Laboratory, SRI International, 333 Ravenswood Avenue, Menlo Park, CA 94025, USA
Fax +1(650)8594321; E-mail: Jeffrey.Bottaro@sri.com
Received 2 October 2003
This work is dedicated to the cherished memory of Tom Highsmith.
OH
NO2
NOH
Abstract: Synthesis of 1,3,5-trinitrobenzene was achieved by con-
NH2OH
HNO3
verting phloroglucinol into its trioximes by reaction with aqueous
hydroxylamine followed by oxidation of the trioxime product with
90% nitric acid.
HO
OH
O2N
NO2
HON
NOH
Key words: 1,3,5-trintirobenzene, cyclohexanetrione trioximes,
phloroglucinol
1
2
3
Scheme 1 Synthesis of 1,3,5-trinitrobenzene
The little-known explosive 1,3,5-trinitrobenzene (TNB)
has not replaced TNT (2,4,6-trinitrotoluene) in spite of su-
perior detonation velocity and pressure of TNB.1–3 The
main reason for this is the economics; the most commonly
cited synthesis of TNB involves oxidation of TNT to tri-
nitrobenzoic acid followed by decarboxylation.3 Direct
nitration of benzene requires such harsh conditions that
yields are poor and purification of TNB is very difficult.
Even the use of modern catalysts such as lanthanide ni-
trates does not alleviate this problem.4 Thus it stands that
TNT successfully competes with TNB in what is a billion
dollar annual market. We wish to report here a facile two-
step synthesis of TNB starting from phloroglucinol. The
only remaining technological impediment to its large-
scale application is the absence of an economical dehydra-
tive trimerization of HOAc or a trimerization of ketene to
produce pholoroglucinol. This challenge has a high prob-
ability of being met once it is defined in the appropriate
setting.
as EtOAc or CH2Cl2 resulted in intractable emulsions.
However, when phloroglucinol was added directly to a
large excess of aqueous hydroxylamine a clear light yel-
low solution was obtained, from which a readily filterable
solid separated out over the course of three hours.
The crude brown trioxime product was not soluble in wa-
ter, even upon boiling, nor was it soluble to any apprecia-
ble extent in pure THF. It is soluble in HOAc, and also in
strong acids, but its dissolution in sulfuric and methane-
sulfonic acids is very exothermic and can lead to charred
products. However, we found trioximes 2, to be complete-
ly soluble in mixtures of THF and water. Purification of
the trioxime was achieved by dissolving the crude product
in a mixture of THF and water, and then evaporating the
THF at reduced pressure in a rotary evaporator. The solid
thus obtained was markedly lighter in color and the re-
maining solution was dark brown. Infrared analysis (KBr
pellet) showed bands at 3450 and 2900 cm–1 correspond-
ing to O–H and. C–H stretches, and at 1640 cm–1 for the
Given the difficulties with the previous syntheses of tri-
nitrobenzene, we examined an alternate approach; one in
which a heteroatom is already at the 1-, 3-, and 5-posi-
tions. Phloroglucinol (1, 1,3,5, trihydroxybenzene)
seemed to be a reasonable starting point. Furthermore, the
fact that a small fraction of pholoroglucinol exists as the
keto tautomer, 1,3,5-cyclohexanetrione, means that it
should be possible to introduce nitrogen functions at those
positions by making oxime derivatives. Indeed, Tokura et
al. describe synthesis of trioximes, 2, by reaction of phlo-
roglucinol with hydroxylamine hydrochloride in the pres-
ence of K2CO3.5
1
C=N stretch. The H NMR of the product in DMSO
showed three peaks between 10.7 and 10.8 ppm. These
strongly deshielded peaks were not visible when D2O was
added to the sample, and thus they can be assigned to the
oxime hydroxyls. The three peaks are not equal in intensi-
ty; the central peak is somewhat more intense. The spec-
trum also contains three almost equally intense peaks at
3.1, 3.3 and 3.5 ppm corresponding to the aliphatic pro-
tons. There is a weaker fourth peak at 3.33 ppm about 15%
as intense as the others. There are two isomers of the tri-
oxime, symmetric and unsymmetric. The symmetric iso-
mer should show a single peak for the hydroxy and the
aliphatic protons, whereas the unsymmetric isomer is ex-
pected to give three peaks for each of these environments.
The NMR spectrum is consistent with a mixture contain-
ing 95% of the unsymmetric isomer. Tokura et al. raised
the question of isomerization of cyclohexanetrione
oximes to 1,3,5-benzenetrihydroxylaminol. The NMR
data presented here, in particular, the results of exchange
We have found isolation of the product trioximes to be
somewhat difficult. The solid product often separated
slowly as a very fine powder that was difficult to filter. At-
tempts to extract the oximes with an organic solvent such
SYNTHESIS 2004, No. 4, pp 0499–0500
x
x
.
x
x
.2
0
0
4
Advanced online publication: 09.02.2004
DOI: 10.1055/s-2004-815958; Art ID: M04303SS
© Georg Thieme Verlag Stuttgart · New York

500
J. C. Bottaro et al.
SHORT PAPER
was detected upon exposure to I2 vapors at an Rf of 0.85. This spot
was close to that of phloroglucinol (Rf 0.9), but it was not UV active
like the starting material. TLC also showed a few other spots near
the origin. Purification was achieved by dissolving about 2.0 g of
the crude product in THF–water (1:1, 50 mL) mixture, and then
evaporating THF in a rotary evaporator. The resulting solid (1.65 g)
was light brown, and gave a single spot by TLC.
with D2O, argue against the presence of the isomerized
structure.
Conversion of oximes into gem-dinitro products by
Ponzio reactions and its variants has been reviewed by
Honey et al.6Axenrod et al. have reported on the oxidative
nitrolysis of oximes to give gem-dinitro products by nitric
acid.7 However, instead of producing 1,1,3,3,5,5-hexa-
nitrocyclohexane, we found that reaction of trioximes 2 in
90% nitric acid produced 1,3,5-trinitrobenzene (3)
(Scheme 1). Evidently, the loss of HONO from adjacent
carbons to give the aromatic product was extremely favor-
able. The reactions were very vigorous and accompanied
by evolution of gases. For safe operation it was necessary
to cool the reactions in an ice bath and also moderate them
with a solvent. Addition of small aliquots of the trioximes
to cooled nitric acid was preferred over the addition of ni-
tric acid to the trioximes. The former mode limits the
amount of material that can be oxidized. When dry pow-
der of the trioxime is added, a small burst of reaction can
be observed when the powder hits the liquid. Addition of
a suspension of the trioximes in CH2Cl2 to nitric acid pro-
ceeded more smoothly, and was the preferred method.
1,3,5-Trinitrobenzene could be detected in the reaction
mixture by TLC almost immediately after completion of
the addition. A second product of slightly lower polarity
was also observed. At longer reaction times or upon
warming the reaction mixture to about 55 °C, the second
product was converted to TNB. By using preparative
TLC, we were able to isolate a small amount (about 30
mg) of this second product. Whereas in dilute solution this
second product was colorless, it developed a green tint
upon concentrating. Its IR spectrum was very similar to
that of TNB, except for an additional band at 1325 cm–1.
The development of green tint and the IR data are consis-
tent with the presence of nitroso group, and we suspect
that this compound has one or two nitroso groups, which
are subsequently oxidized to the nitro groups.
1,3,5-Trinitrobenzene; Typical Procedure
Approximately 8.0 g concd HNO3 (90%) under Ar were taken in a
25 mL 3-necked, round-bottom flask and cooled in an ice bath. In a
separate flask, the trioximes of 1,3,5-cyclohexanetrione (171 mg,
1.0 mmol) were suspended in about CH2Cl2 (2 mL). About 100
L
aliquots of the suspension were slowly added to the cooled stirring
HNO3. Immediately following each addition there was evolution of
fumes of nitrogen oxides, which were swept away by Ar. Sufficient
time was allowed for the fuming to subside (about 15 s) before the
next aliquot was added. After all of the trioximes had been added,
the flask was fitted with a reflux condenser and transferred to a bath
held at 55 °C. After 3 h the contents of the flask were poured into
water (100 mL), and extracted with EtOAc (50 mL). The EtOAc ex-
tract was washed sequentially with water, NaHCO3 solution, and
brine, dried over anhyd Na2SO4, and then flash chromatographed
over silica gel; yield: 109 mg (50%). The IR spectrum of the product
was identical with that of authentic 1,3,5-trinitrobenzene and
showed the band due to aromatic C–H stretch at 3150 cm–1 and a
pair of bands due to the NO2 group at 1340 and 1560 cm–1. The 1H
NMR spectrum showed a single peak at 9.2 ppm.
The progress of the oxidation reaction was monitored by TLC on
silica, with toluene as the eluant. A small aliquot of the reaction
mixture was diluted with water and extracted with EtOAc. TLC of
the EtOAc extract showed two fast moving spots (Rf 0.95, 0.88),
both UV active, along with a spot near the origin. TLC of an authen-
tic sample of 1,3,5-trinitrobenzene showed a spot with an Rf 0.88.
References
(1) Kohler, J.; Meyer, R. In Explosives, 4th ed., revised and
extended; VCH Publishers: USA, 1994, 387–388.
(2) For a brief review on explosive materials see: Kirk Othmer,
Concise Encyclopedia of Chemical Technology, 4th ed.;
Wiley Interscience: New York, 1999, 787.
(3) See: The Merck Index, 12th ed.; Budavari, S., Ed.; Merck
and Co. Inc.: New Jersey, 1996, 1657.
(4) Forsberg, J. H.; Spaziano, V. T.; Balasubramanian, T. M.;
Liu, G. K.; Kinsley, S. A.; Duckworth, C. A.; Poteruca, J. J.;
Brown, P. S.; Miller, J. L. J. Org. Chem. 1987, 52, 1017.
(5) Tokura, N.; Shirai, I.; Sugahara, T. Bull. Chem. Soc. Jpn.
1962, 35, 722.
(6) Honey, P. J.; Millar, R. W.; Coombes, R. G. ACS Symposium
series 623, In Nitration: Recent Laboratory and Industrial
Developments; Albright, L. F.; Carr, R. V. C.; Schmitt, R. J.,
Eds.; American Chemical Society: Washington DC, 1996,
134–150.
1,3,5-Cyclohexanetrione Trioximes; General Procedure
Phloroglucinol dihydrate (8.11 g; 0.050 mol) was taken in a 500 mL
Erlenmeyer flask. To it was added a 50% by weight aq solution of
hydroxylamine (29.2 g; 0.44 mol). Phloroglucinol dissolved in the
aqueous medium and over a period of about 3 min to give a clear
yellow solution. The solution was let stand at r.t. After 1 h a thin lay-
er of solid had separated and the solution had turned noticeably
darker. The solution was gently agitated to break up the crystals and
allowed to stand for 2 h. The solution turned dark brown and a co-
pious quantity of solids had separated. The solids were filtered,
washed with water and dried (yield: 7.59 g; 88%). The product was
analyzed by TLC on silica gel (toluene–i-PrOH, 1:1). A strong spot
(7) Axenrod, T.; Watnick, C.; Yazdekhast, H.; Dave, P. R. J.
Org. Chem. 1995, 60, 1959.
Synthesis 2004, No. 4, 499–500 © Thieme Stuttgart · New York