300 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVESPhysical properties
Reactions with acids and alkalisLike other aromatic nitro compounds, trinitrotoluene is resistant to the action
of acids. Only concentrated nitric acid at a temperature over 110°C can oxidize
it to 2,4,6-trinitrobenzoic acid, and at 200°C trinitrobenzene is formed, as a result
of decarboxylation (p. 254).
However, trinitrotoluene reacts with alkalis, yielding organo-metallic products.
The readiness of trinitrotoluene to react with alkalis has suggested the idea that
it is an acid. This problem was studied by Farmer [75] in 1901. He applied a method,
based on determining the partition coefficient K for trinitrotoluene between two
liquid phases: water and benzene or water and ligroin phase, with addition of
sodium hydroxide to the water phase.
While for picric acid a value K = 0.164 had been found by this method (Roth-
mund and Drucker [76]), for trinitrotoluene Farmer found K = 2.0 x 10-14, which
means that trinitrotoluene is an acid 1013 times weaker than picric acid. In effectit has virtually no acidic properties.
Similarly Brand and his co-workers [77] inferred from the results of spectro-
photometric studies that 2,4,6-trinitrotoluene, in a solution of 100% sulphuric
acid, was not ionized, and could undergo partial ionization only in very con-
centrated oleum. These investigations are discussed more fully on p. 219.
On the other hand, attention should be drawn to the cryometric investigations
of nitro compounds, already referred to (p. 218), which made Hantzsch express
the view that the nitro group may confer a basic character to the compound. He
gave only p- nitrotoluene as an example. Gillespie [78] has suggested that 2,4,6-
trinitrotoluene in sulphuric acid solution also behaves as a base.
Wyatt and Brayford [79] have tried to explain the inconsistency of the spectro-
graphic and cryometric results. On the basis of their cryometric measurements
on solutions of 2,4,6-trinitrotoluene, 1,3,5-trinitrobenzene and picric acid in
sulphuric acid in the presence of compounds interferring with the dissociation of
the solvent, they finally concluded, on the basis of the spectrophotometric meas-
urements, that polynitro compounds should be regarded as non-electrolytes.
Alkalis, when reacted with trinitrotoluene, very easily effect a considerable
change in the substance, yielding red or brown coloured addition products contain-
ing metal. Inorganic acids separate from these products an organic substance,
which is no longer trinitrotoluene. Numerous investigations carried out to elucidate
the structure of this substance have given no definite answer as yet.
Several workers investigating trinitrotoluene many years ago took an interest
in the nature of these substances (e.g. Wilbrand [34] in 1863; Hepp [35] in 1882).
Korezynski [80] was one of the first who investigated the reaction quantitatively.
In 1908 he introduced dry ammonia to various nitro compounds, including trini-
trotoluene and trinitrobenzene. When the reaction was carried out at a low tem-