Transformation pathways of 2,4,6-trinitrobenzoic acid in the aqueous-phase hydrogenation over Pd/C catalyst

Article abstract of DOI:10.1016/j.molcata.2016.04.014

Results of the catalytic hydrogenation of 2,4,6-trinitrobenzoic acid (TNBA) in the presence of 1% Pd/C catalyst with analysis of the reaction products at the steps corresponding to the consumption of 1, 3, 5, 6, 8 and 9 mols of hydrogen per mol of TNBA are presented. Numerous reaction intermediates were identified using the ¹H and ¹³C NMR spectroscopy; the reaction mixture composition was assessed quantitatively at different steps of the hydrogenation. The data obtained were used to propose a scheme of TNBA transformations over the Pd/C catalyst in the aqueous-phase reaction conditions. The data on the component composition at different steps of TNBA hydrogenation are important to understand the process chemistry, to explore the possibility of selective formation of different types of intermediate nitroamino compounds, and to synthesize high-performance catalysts for the hydrogenation of aromatic polynitro compounds.

Full text of DOI:10.1016/j.molcata.2016.04.014

Accepted Manuscript
Title: Transformation pathways of 2,4,6-trinitrobenzoic acid in
the aqueous-phase hydrogenation over Pd/C catalyst
Author: Olga B. Belskaya Valentin P. Talsi Roman M.
Mironenko Vladimir A. Rodionov Sergey V. Sysolyatin
Vladimir A. Likholobov
PII:
S1381-1169(16)30145-5
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2016.04.014
MOLCAA 9858
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
11-3-2016
14-4-2016
15-4-2016
Please cite this article as: Olga B.Belskaya, Valentin P.Talsi, Roman
M.Mironenko, Vladimir A.Rodionov, Sergey V.Sysolyatin, Vladimir A.Likholobov,
Transformation pathways of 2,4,6-trinitrobenzoic acid in the aqueous-phase
hydrogenation over Pd/C catalyst, Journal of Molecular Catalysis A: Chemical
http://dx.doi.org/10.1016/j.molcata.2016.04.014
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Transformation pathways of 2,4,6-trinitrobenzoic acid in the aqueous-
phase hydrogenation over Pd/C catalyst
Olga B. Belskayaa,b*, Valentin P. Talsia, Roman M. Mironenkoa,c, Vladimir A.
Rodionova, Sergey V. Sysolyatind, Vladimir A. Likholobova,b,e
a Institute of Hydrocarbons Processing, Siberian Branch of the Russian Academy of
Sciences, Neftezavodskaya st., 54, 644040 Omsk, Russia
b Omsk State Technical University, pr. Mira, 11, 644050 Omsk, Russia
c Dostoevsky Omsk State University, pr. Mira, 55a, 644077 Omsk, Russia
d Institute for Problems of Chemical and Energetic Technologies, Siberian Branch of the
Russian Academy of Sciences, Socialisticheskaya st., 1, 659322, Biysk, Altai krai,
Russia
e Omsk Scientific Center, Siberian Branch of the Russian Academy of Sciences, pr.
Marksa, 15, 644024 Omsk, Russia
Corresponding author: Olga B. Belskaya
Present address: Institute of Hydrocarbons Processing, Siberian Branch of the Russian
Academy of Sciences, Neftezavodskaya st., 54, 644040 Omsk, Russia
E-mail: obelska@ihcp.ru
1

Tel.: +7 (3812) 670 474, fax: +7 (3812) 646 156
Graphical abstract
Highlights
 Hydrogenation of 2,4,6-trinitrobenzoic acid over Pd/Sibunit catalyst.
 Identification of numerous reaction intermediates using NMR spectroscopy.
 Scheme of transformations of substrate and intermediates in an aqueous medium.
Abstract:
Results of the catalytic hydrogenation of 2,4,6-trinitrobenzoic acid (TNBA) in the
presence of 1 % Pd/C catalyst with analysis of the reaction products at the steps
corresponding to the consumption of 1, 3, 5, 6, 8 and 9 moles of hydrogen per mole of
TNBA are presented. Numerous reaction intermediates were identified using the 1H and
13C NMR spectroscopy; the reaction mixture composition was assessed quantitatively at
different steps of the hydrogenation. The data obtained were used to propose a scheme
of TNBA transformations over the Pd/C catalyst in the aqueous-phase reaction
conditions. The data on the component composition at different steps of TNBA
hydrogenation are important to understand the process chemistry, to explore the
possibility of selective formation of different types of intermediate nitroamino
2

compounds, and to synthesize high-performance catalysts for the hydrogenation of
aromatic polynitro compounds.
Keywords: trinitrobenzoic acid, catalytic hydrogenation, aqueous-phase reaction,
palladium catalyst
1. Introduction
The catalytic hydrogenation of aromatic nitro compounds is of interest not only
as a model reaction in the development of theoretical bases for the liquid-phase
hydrogenation of organic compounds, it is also of practical importance for chemical
conversion of military high explosives into harmless products. Safe deactivation of
2,4,6-trinitrotoluene (TNT), which is among the most common explosives, by catalytic
hydrogenation has substantial environmental advantages over its utilization by
explosion, combustion, biological decomposition or chemical reduction [1-4]. So,
optimization of conditions of the catalytic process and development of more efficient
catalysts are the promising lines of investigation. The reduction of all nitro groups in
TNT can produce 2,4,6-triaminotoluene (TAT), which is an important component in the
synthesis of various commercial products [1-7]. The catalytic hydrogenation of aromatic
trinitro compounds including TNT is usually performed in organic solvents with the use
of bulk or supported catalysts containing nickel or platinum group metals [2-4, 7-33]. In
the general case, the process scheme includes the delivery of hydrogen to the catalyst
surface, its activation and interaction with nitro compound. Therewith, the hydrogen
activation step is considered as the limiting one [16, 18].
3

Along with the works dealing with the exhaustive hydrogenation of all nitro
groups in aromatic polynitro compounds, obtain of intermediate hydrogenation products
is also important. A study of the hydrogenation of TNT on a palladium or Raney Ni
catalysts in methanol and methanol/toluene medium demonstrated [4, 15, 19, 21] a
1
complex chain of consecutive and parallel reactions. Analysis of products by H NMR
spectroscopy and mass spectrometry for the hydrogenation of TNT on a Pd/C catalyst in
an alcohol solution made it possible to suggest scheme of transformations (Scheme 1)
[21].
Along with TNT, among the most accessible and promising trinitro compounds
is the product of TNT oxidation, 2,4,6-trinitrobenzoic acid (TNBA) [34-36]. Until now,
only few works on catalytic hydrogenation of TNBA have been published [12, 37-42].
In view of environmental and green chemistry, the using of non-toxic and inexpensive
solvent such as water is preferable, but the direct application of TNBA as the substrate
for aqueous-phase hydrogenation is limited by its low solubility in water [38, 43].
However, as a component of a salt with alkaline metals, this compound has a good
water solubility [38]. The literature provides no information about the effect of
conditions of the aqueous-phase hydrogenation and catalyst composition on the process
rate and depth and on the direction of TNBA transformations.
It should be noted that various methods were employed at different stages of
studying the transformations of aromatic polynitro compounds for identification of
hydrogenation products, in particular, the elemental analysis [7-10, 31], mass
spectrometry [14, 21, 31], gas-liquid chromatography [4, 7, 15-18, 25, 29], high
4

performance liquid chromatography [7, 20, 29] and spectral methods [2, 3, 11, 14, 19,
21, 26, 28, 31]. Among them, one of the most efficient techniques is NMR
spectroscopy. The non-destructive nature of the NMR is its doubtless advantage, which
is most topical for investigation of unstable aromatic polynitro compounds and
hydrogenation products.
In the present work, the catalytic hydrogenation of TNBA over Pd/C catalyst
was conducted in an aqueous medium, and the products were analyzed at individual
1
steps of the reaction. H and 13C NMR spectroscopy served as the main method of
investigation. The study was aimed to reveal the directions of TNBA transformations in
the aqueous-phase catalytic reaction. The choice of NMR is caused by not only a high
informativity of this method but also a low efficiency or inability to use other
techniques. So, we were not able to employ the capabilities of chromatography methods
with mass detection. Obviously, these methods should be found to be unsuitable for the
solution of our analytical task because of low stability of analyzed intermediates as well
as lack of published mass spectra for most of these compounds.
2. Experimental
2.1. Catalyst preparation
The 1 % Pd/С catalyst was synthesized by the hydrolytic precipitation of
palladium polyhydroxo complexes (PHC) on the surface of carbon material Sibunit [44-
46]. The synthesis includes (1) the preparation of palladium PHC by slow introduction
of the solution of KHCO3 to the H2PdCl4 solution under stirring and pH controlling, (2)
the impregnation of Sibunit by solution of PHC, (3) the reduction of palladium with
sodium formate, (4) the washing of the samples [47]. Aqueous solution of H2PdCl4 with
5

desired Pd concentration was prepared by dissolution of PdCl2 in concentrated
hydrochloric acid. Before further experiments, no any pretreatments of obtained catalyst
were performed and its moisture was about 50 %. Textural characteristics of the
support, according to the low-temperature adsorption of nitrogen, were as follows: the
specific surface area determined by the Brunauer-Emmett-Teller method was 422 m2/g;
the adsorption pore volume, 0.62 cm3/g; and the mean pore diameter, 5.9 nm.
Anchoring of palladium produced no changes in these parameters of the support.
2.2. Hydrogenation of TNBA
TNBA (98 % purity) was synthesized by oxidation of TNT as described by
Clarke and Hartman [34]. The sodium salt of TNBA was obtained by the addition of
sodium bicarbonate to the aqueous suspension of TNBA under careful stirring [38]. The
catalytic hydrogenation of the obtained salt (100 mL of 2 % aqueous solution) was
performed at a temperature of 323 K and pressure of 0.5 MPa in the presence of 1.00 g
catalyst (loading was about 400 g of TNBA per 1 g of Pd). The reaction was conducted
in a 180 mL steel autoclave equipped with a valve for hydrogen input and an external
thermostatted jacket. The reaction mixture was stirred by a magnetic stirrer at 1400 rpm
to prevent external diffusion limitations. The progress of the reaction was monitored by
measuring the volume of consumed hydrogen with a mass flow meter. The reaction was
stopped when the required amount of hydrogen was consumed. After cooling, the
sampling was carried out with a syringe.
2.3. Analysis of hydrogenation products by 1H and 13C NMR spectroscopy
6

Spectra were recorded on an Avance-400 (Bruker) NMR spectrometer using
standard ampoules 5 mm in diameter at the Larmor frequency of 400 MHz (1H) or
100.6 MHz (13C) at a temperature of 303 K in a pulsed mode. The signals from residual
protons (H of 3.30 ppm, a quintet) or carbon nuclei in methyl groups (C of 49.0 ppm, a
septet) of methanol-d4 were used as the internal standard for the chemical shifts. A
1
signal from water at ~4.9 ppm in the H NMR spectra was suppressed by presaturation
method using the pulse sequence program (zgpr) provided with the Bruker software.
The J-modulated (JMOD) pulse sequence [48] was employed for recording all of the
13C NMR spectra. In the JMOD 13C NMR spectra, the methine and methyl signals have
opposite phase to those of quaternary and methylene resonances. NMR spectra were
simulated with the ACD/Labs 6.00 (Advanced Chemistry Development Inc.) software.
Simulation was especially useful in the absence of published spectra. The quantitative
1
composition of reaction mixture was assessed by H NMR from the integral intensity
corresponding to the signals of individual components non-overlapping with other
resonances. Bearing in mind that none of the signals of studied compounds was
saturated in the NMR registration conditions, it was admitted that the integral intensity
of signals is directly proportional to the number of molecules of a component in the
solution and to the number of protons in the molecules generating these signals.
3. Results and discussion
3.1. Obtaining of TNBA hydrogenation products
The hydrogenation of the sodium salt of TNBA under the chosen conditions
resulted in the consumption of 1500 mL hydrogen (reduced to normal conditions),
which is close to the stoichiometric value (9 mole per mole of TNBA) at a complete
7

hydrogenation of all nitro groups in trinitro compound. The obtained hydrogen
consumption curve (Fig. 1a) does not have the pronounced steps, which testifies to the
absence of selective consecutive reduction of nitro groups with a strongly different rate
and suggests that the reduction involves all nitro groups. Such a process leads to a great
number of the reaction products and complicates their identification [4, 19, 21].
To examine the composition of products resulting from the hydrogenation of
TNBA, points ii-vii, corresponding to the consumption of 1, 3, 5, 6, 8 and 9 moles of
hydrogen per mole of TNBA, were distinguished on the curve of consumption rate
versus the amount of consumed hydrogen (Fig. 1b). Further experiments consisted in
the hydrogenation of six identical solutions of the substrate with termination of the
reaction in one of the chosen points. This gave the subjects of inquiry represented by
aqueous solutions of the sodium salt of TNBA (sample i), products of complete
transformation (sample vii), and products of consecutive incomplete hydrogenation of
TNBA (samples ii-vi).
3.2. NMR study of the composition of TNBA hydrogenation products
1
The H NMR spectra of the starting compound, final and intermediate
hydrogenation products of TNBA corresponding to samples i-vii (Fig. 1), which were
obtained by the stepwise hydrogenation, are depicted in Fig. 2. It should be noted that in
some cases intermediate compounds were identified using the 13C and 1H NMR spectra
that were obtained in our similar studies on the hydrogenation of more concentrated
solutions of the substrate (10 wt. %) with the catalysts of different activity.
1
The H NMR spectrum of the starting solution (sample i) in addition to the
signal from the solvent shows only a singlet at 9.25 ppm corresponding to the aromatic
8

protons of TNBA (Fig. 2a). At the initial step of hydrogenation (sample ii) with the
consumption of 1 mole of H2 per mole of TNBA, three main compounds were identified
in the solution (Fig. 2b). Along with the signal from unreacted TNBA, the spectrum has
a singlet at 7.87 ppm and two doublets at 8.27 and 8.33 ppm with long range couplings
of 2.2 Hz (the splitting typical of meta protons), which can be assigned to aromatic
protons of 4-(hydroxyamino)-2,6-dinitrobenzoic acid (1) and 6-(hydroxyamino)-2,4-
dinitrobenzoic acid (2), respectively (Table 1). As a single alternative to compounds 1
and 2 we considered their analogs containing the amino groups instead of
hydroxyamino ones. However, such compounds give signals of their aromatic protons
in a noticeably stronger field ( 0.4-0.7 ppm) [49, 54]. In addition, a weak signal
possibly corresponding to 4-amino-2,6-dinitrobenzoic acid (3) is observed at the next
hydrogenation steps (Fig. 2с and d) along with the signal from compound 1. The
chemical shifts of the signals that were attributed to TNBA and compounds 1 and 2 in
the 13C NMR spectrum of sample ii (Fig. S1 in the Supplementary materials) agree
satisfactorily with the similar values in the simulated spectra of individual compounds
(Table 2).
In the solution obtained by the consumption of 3 moles of H2 per mole of TNBA
(sample iii), the starting compound is absent, and compound 1 gives the most intense
signal in the 1H NMR spectrum (Fig. 2с). This spectrum has also the next high-intensity
group of signals in the region of 7.0-7.1 ppm, which is a overlapping of the singlet at
7.09 ppm with two doublets at ~7.06 and ~7.10 ppm. The validity of this fact was
confirmed by the analysis of spectra of the samples in which only a singlet or only two
doublets are observed in the region under consideration (Figs. 2d and 2f; Fig. S2). The
singlet at 7.09 ppm can be assigned to aromatic protons of 2,6-bis(hydroxyamino)-4-
9

nitrobenzoic acid (4), while the two doublets (7.06, 7.10 ppm) to 2-amino-6-
(hydroxyamino)-4-nitrobenzoic acid (5). The correctness of identification of compound
5 is undeniable. Since this compound is the main component of sample iv (the
consumption of 5 moles of H2 per mole of TNBA) (Figs. 2d and 3), the corresponding
13
signals in the C NMR spectrum can clearly be determined (Fig. 4). The signals that
1
correspond to compound 5 in the H and 13C NMR spectra can belong only to the
reduction product of TNBA that contains one nitro group and two nonequivalent meta
13
protons (H-3 and H-5) in the benzene ring. As shown by the analysis of simulated C
NMR spectra for all possible locations of amino, hydroxyamino and nitro groups in the
ortho and para positions of benzoic acid, the 13C NMR spectrum depicted in Fig. 4 can
belong only to compound 5 in which the nitro group places in the para position to the
carboxyl group. After identification of compound 5, the structure of compound 4 also
becomes well established, although it could be validated only by the simulated 1H NMR
spectrum of the mixture of compounds 4 and 5, which demonstrates the experimentally
observed overlapping of the signals.
1
The H NMR spectrum of sample iii (Fig. 2c) also has proton signals in the
saturated region. The intensity of these signals monotonically increases in the spectra of
samples iii-vi (Figs. 2с-f; see also Fig. 3). The compounds that give singlets in the
1
region of 3.2-3.7 ppm were assigned, regarding the published H NMR spectra of
individual compounds [50, 55, 56], to methylene protons of (1Z)-cyclohexane-1,3,5-
trione trioxime (6) (three singlets at 3.22, 3.46, 3.70 ppm) and (1E,3E,5E)-cyclohexane-
1,3,5-trione trioxime (7) (one singlet at 3.50 ppm). Additionally, the identification of
these compounds was based on the spectra of the samples obtained by hydrogenation of
more concentrated (10 wt. %) solutions of TNBA where the content of compounds 6
10

1
and 7 was ~50 %. The H NMR spectrum of one of such samples is depicted in Fig. 5,
13
13
and the JMOD C NMR spectrum – in Fig. S3. The experimental and simulated C
NMR spectra of trioximes 6 and 7 are virtually identical (Table 2).
It should be noted that a singlet at 5.75 ppm assigned to aromatic protons of
1,3,5-triaminobenzene (8), which is the product of complete hydrogenation of TNBA,
was observed for the first time in the spectrum of sample iv (Fig. 2d). Later, upon
consumption of 8 and 9 moles of H2 per mole of TNBA (Figs. 2f and 2g), compound 8
became the main product (Fig. 3).
Before starting the identification of minor reduction products of TNBA, let us
consider the main route of the reaction, which can be derived from quantitative
assessment of the composition of reaction mixture at different hydrogenation steps by
1H NMR spectroscopy (Figs. 2 and 3). First, TNBA is reduced to 1 and 2, a further
hydrogenation of which leads to 4 and then to 5. The resulting compound 5 transforms
into trioximes 6 and 7 which, in their turn, are hydrogenated to 8 (Scheme 2).
It should be noted that the formation of compound 5 cannot be attributed only to
the consecutive hydrogenation of TNBA through the formation of compounds 2 and 4.
Even at the first step upon consumption of 1 mole of H2 per mole of TNBA (Figs. 2b
and 3), compound 1 predominates in the mixture of hydrogenation products, and at the
next steps (Figs. 2с, d and 3) the observed substantial increase in the content of 5 can
occur mainly via the reduction of 1. Hence, along with consecutive hydrogenation, the
formation of compound 5 should include a step at which the carboxyl group migrates to
form a symmetric molecule of intermediate 4. Perhaps, such migration occurs via
decarboxylation of 9 and subsequent carboxylation of formed carbanion, i.e. in the same
11

way as described for the Henkel process (also referred to as the Raecke rearrangement)
which is the isomerization of potassium phthalate to terephthalate [57, 58].
In addition, the main route of the reaction under consideration implies the
transformation of compound 5 into trioximes 6 and 7, which cannot result from the
direct reduction. Most likely, there occurs the intramolecular redox transformation of
compound 5 into symmetric 2,4,6-tris(hydroxyamino)benzoic acid (10) whose
decarboxylation gives 1,3,5-tris(hydroxyamino)benzene (11), which is able to isomerize
into oximes 6 and 7 (Scheme 2). Although the identification of intermediates 10 and 11
has failed, some of the 1H NMR spectra of reaction mixtures (samples iv and v) showed
a singlet with the chemical shift at about 6.30 ppm, which can belong, according to the
simulated spectrum (Table 1), to aromatic protons of compound 11. Earlier it was
assumed that trioxime 6 or 7 exists in tautomeric equilibrium with 11 [59]. However,
more recent investigations by NMR have been established that compound 11 always
transforms into trioximes, among of which unsymmetric isomer 6 prevails [50, 55]. The
1
ratio of isomers 6 and 7 estimated from H NMR data (about 90 % of 6, see Fig. 3) is
close to those determined in [50]. It is important that upon catalytic hydrogenation of
trioxime, aromatic structure is restored [55].
To determine the composition of components of samples ii-vi, which are
1
characterized by less intense signals in the H NMR spectra (Figs. 2b-f), we analyzed
1
13
also the H and C NMR spectra of the solutions obtained earlier in the experiments
with a higher content of the substrate (10 wt. %) and shown in Figs. 5, 6 and S4.
Therewith, Fig. 6 shows the spectra of the same solution with different holding time,
12

1
which allows observing transformation of the solution components with time. The H
NMR spectrum displayed in Fig. 5 shows not only the resonances relating to
compounds 5-8 but also two relatively intense doublet-triplet pairs of signals at 6.10
(doublet), 6.14 (triplet) and 5.88 (triplet), 5.94 (doublet) ppm, which are present in Figs.
2d-f, too. These signals can be assigned to aromatic protons of 3,5-
bis(hydroxyamino)aniline (12) and 5-(hydroxyamino)benzene-1,3-diamine (13),
respectively. Correctness of the attribution is confirmed by good agreement between
1
parameters of the experimental and simulated H NMR spectra of these compounds
(Table 1) and by the results of analysis of the corresponding JMOD 13C NMR spectrum
(Table 2, Fig. S3). Compounds 12 and 13 are present in samples iv-vi, reach a
maximum content at the consumption of 8 moles of H2 per mole of TNBA, and are
absent in sample vii (Fig. 3). As shown in Scheme 2, compounds 12 and 13 are the
intermediate reduction products of oximes 6 and 7; however, other routes of their
formation (e.g. from compound 11) also cannot be ruled out.
A singlet with the chemical shift at ~5.86 ppm, which is seen in Figs. 2d-g, can
be assigned to aromatic protons of 2,4,6-triaminobenzoic acid (14). In the spectra
depicted in Fig. 6, this signal is relatively intense (second in the intensity after the signal
from compound 8), and its assignment to 14 is unambiguous, taking into account the
simulated spectrum (Table 1) and the analysis of the JMOD 13C NMR spectrum (Table
2, Fig. S4). Other signals in the spectra shown in Fig. 6 are the two doublets at 6.64,
6.88 ppm and two doublets at 6.39, 6.63 ppm, which were possibly attributed to
aromatic protons of 4-amino-2-(hydroxyamino)-6-nitrobenzoic acid (15) and 2,4-
diamino-6-nitrobenzoic acid (16), respectively. Therewith, a less stable 15 converts into
16 upon holding of the sample.
13

The spectra in Figs. 2d-f, in addition to the indicated signals of compounds 15
and 16, show another two doublets at about 6.77 and 6.87 ppm, which was probably
attributed to aromatic protons of 2-amino-4-(hydroxyamino)-6-nitrobenzoic acid (17).
One can see in a detailed fragment of the spectrum in Fig. 2f (Fig. S5) that all the pairs
of doublets for compounds 15-17 are well enough resolved and do not overlap, in
distinction to the spectra in Figs. 2d and 2e. Parameters of the experimental and
simulated spectra of compounds 15-17, especially the relative positions of the
corresponding pairs of doublets, are consistent with each other. Correctness of the
identification of compound 16 is confirmed also by JMOD 13C NMR spectrum (Table 2,
13
Fig. S4). Signals corresponding to compounds 15 and 17 were not detected in the C
NMR spectra, probably due to their insufficient stability during the long-term (12 h)
recording of the 13C NMR spectrum.
Thus, the identification of compounds and the assessment of their contents at
consecutive hydrogenation steps give grounds to suggest that the reduction of TNBA
produces not only the main intermediate 5 but also its isomers 15 and 17. Similar to
compound 5, isomers 15 and 17 seem to be capable of transforming into the final
triamine 8 through oximes 6 and 7 and compounds 12 and 13 according to Scheme 2. At
the same time, isomers 5, 15 and 17 can be reduced to compound 16, which then lead to
the formation of triamine 8, probably via the decarboxylation of intermediate 14
(Scheme 3).
1
Some very weak and poorly resolved signals in the H NMR spectra under
consideration can be attributed to aromatic protons of 3,5-diaminophenol (18), 5-
aminobenzene-1,3-diol (19) and benzene-1,3,5-triol (20), which can be produced by the
14

consecutive hydrolysis of compound 8 (Scheme 4) when the reaction is carried out in an
aqueous medium [42, 60]. A singlet at 5.89 ppm corresponding to 20 overlaps with the
signals from other compounds in most of the spectra. In some spectra, pairs of the
signals (doublet-triplet) attributed to compounds 18 and 19 are not clearly seen due to
1
overlapping with other resonances. The H NMR spectrum of the sample of reaction
mixture with an increased content of one of the tested compounds, which gives doublet
at 5.78 and triplet at 5.83 ppm (Fig. S6), showed that the indicated signals belong to
aromatic protons of 18 (Table 1). In addition, identification of 18 was validated by
JMOD 13C NMR (Fig. S7, Table 2). Compound 19 was identified less reliably.
1
Fig. 2g depicts the H NMR spectrum of sample vii that was obtained by the
exhaustive hydrogenation of all nitro groups of TNBA (about 9 mole of H2 per mole of
TNBA was consumed). The content of 8 in the reaction mixture, according to 1H NMR,
was above 95 % (see Fig. 3). To provide a correct identification of low intensity signals,
more concentrated solutions (10 wt. %) with a close composition of products were also
1
examined. In the H NMR spectrum displayed in Fig. S8, very weak signals
corresponding to aromatic protons of compounds 18-20 are observed.
Probably, the side hydrolysis reactions enhance the polycondensation of
1
13
hydrogenation products. In many H and JMOD C NMR spectra, a number of weak
1
13
signals may be observed. Taking into account the simulated H and C NMR spectra
(Tables 1 and 2) and the literature data [61, 62], these signals were assigned to the
oxidative polycondensation product that may be represented by the structural fragment
shown below.
15

NH2
N
N
NH2
H2N
NH2
H2N
NH2
H
N
HN
H2N
NH
O
NH2
NH2 OH
21
1
There is a satisfactory agreement between parameters of the H NMR spectrum
(Fig. S8) that were attributed to the aromatic protons of oxidative polycondensation
1
product of 8 and the corresponding parameters of the simulated H NMR spectrum of
structure 21 (Table 1). Generally speaking, the respective parameters of experimental
and simulated 13C NMR spectra (Fig. S9, Table 2) are also close to each other.
4. Conclusions
The suggested conditions of the hydrogenation of TNBA as a component of the
sodium salt in an aqueous medium in the presence of 1 % Pd/C catalyst made it possible
to perform a complete hydrogenation of all nitro groups with simultaneous
decarboxylation and to selectively produce 1,3,5-triaminobenzene at a stoichiometric
hydrogen consumption.
At the same time, the NMR analysis of reaction mixtures at consecutive steps of
the reaction allowed us to obtain unique information on the composition of intermediate
transformation products and to propose a scheme of consecutive and parallel reactions
16

occurring in the hydrogenation. Systematization of the large body of experimental data
1
acquired by the H and 13C NMR investigation of the solutions with different and
sometimes predominant content of individual components allowed for a reliable
identification of numerous intermediates, although experimental spectral information
reported in the literature is quite scarce.
The data on the component composition at different steps of TNBA
hydrogenation, which were obtained in the study, are important to understand the
process chemistry, to explore the possibility of selective formation of different types of
intermediate nitroamino compounds, and to synthesize high-performance catalysts for
the hydrogenation of aromatic trinitro and in general polynitro compounds.
Acknowledgements
The authors are grateful to S. Vysotskii for participation in the catalyst
preparation and testing. In the work, the equipment of the Omsk Regional Center of
Collective Usage, Siberian Branch of the Russian Academy of Sciences was used. The
support from the Russian Foundation for Basic Research (Project No. 16-03-00601) is
gratefully acknowledged.
References
[1] V.A. Tartakiovsky, S.A. Shevelev, M.D. Dutov, A.Kh. Shakhnes, A.L. Rusanov,
L.G. Komarova, A.M. Andrievsky, in: H. Krause (Ed.), Conversion Concepts for
Commercial Applications and Disposal Technologies of Energetic Systems, Kluwer
Academic Publishers, Dordrecht, 1997, pp. 137-149.
[2] P. Wanninger, Chem. Unserer Zeit 29 (1995) 135-140.
17

[3] P. Wanninger, in: K.K. Kuo (Ed.), Challenges in Propellants and Combustion, 100
Years after Nobel, Begell House, New York, 1997, pp. 155-166.
[4] E.F. Litvin, V.Z. Sharf, Ross. Khim. Zhurnal 44 (2000) 90-98.
[5] A.L. Rusanov, L.G. Komarova, D.Yu. Likhatchev, S.A. Shevelev, V.A.
Tartakovsky, Russ. Chem. Rev. 72 (2003) 899-911.
[6] V.I. Ushkarov, K.I. Kobrakov, A.I. Alafinov, S.A. Shevelev, A.Kh. Shakhnes,
Theor. Found. Chem. Eng. 41 (2007) 671-674.
[7] V.A. Kashaev, G.Kh. Khisamutdinov, S.A. Shevelev, S.I. Valeshnii, A.Kh.
Shakhnes, A.P. Bavrina, Theor. Found. Chem. Eng. 42 (2008) 650-656.
[8] F. Hein, F. Wagner, Ber. Dtsch. Chem. Ges. 68 (1935) 856-864.
[9] S.S. Voris, P.E. Spoerri, J. Am. Chem. Soc. 60 (1938) 935-936.
[10] J.E. Gill, R. MacGillivray, J. Munro, J. Chem. Soc. (1949) 1753-1754.
[11] T. Yamaoka, H. Hosoya, S. Nagakura, Tetrahedron 24 (1968) 6203-6213.
[12] E. Vero, J.N. Vickers, GB Patent 1 106 088 (1968), to Fisons Industrial Chemicals
Ltd.
[13] W.A. Freeman, C.F. Liu, Inorg. Chem. 14 (1975) 2120-2124.
[14] B. Kohne, K. Praefcke, T. Derz, T. Gondro, F. Frolow, Angew. Chem. Int. Ed.
Engl. 25 (1986) 650-651.
[15] K.B. van Gelder, J.K. Damhof, P.J. Kroijenga, K.R. Westerterp, Chem. Eng. Sci.
45 (1990) 3159-3170.
[16] U. Kürschner, H. Ehwald, G. Alscher, Сhem. Tech. 45 (1993) 448-452.
[17] H. Ehwald, G. Alscher, Сhem. Tech. 47 (1995) 297-301.
[18] U. Kürschner, H. Ehwald, G. Alscher, Catal. Lett. 34 (1995) 191-199.
18

[19] W. Chuntiag, V.G. Dorokhov, G.A. Boiko, B.S. Bal’zhinimaev, V.V. Barelko,
Dokl. Chem. 402 (2005) 111-113.
[20] M.E. Fuller, C.E. Schaefer, J.M. Lowey, Chemosph. 67 (2007) 419-427.
[21] V.G. Dorokhov, V.I. Savchenko, in: A.A. Berlin, N.A. Khalturinskiy, A.A.
Voshkin (Eds.), International Conference on Chemical Technology, LENAND,
Moscow, 2007, pp. 383-385.
[22] A.I. Kozlov, V.L. Zbarskij, N.V. Khodov A.F. Kuimov, RU Patent 2 333 795
(2008), to ZAO Russkij katalizator.
[23] A.V. Gavrikov, V.G. Dorokhov, A.S. Ivanov, A.F. Pal, O.F. Petrov, A.N.
Ryabinkin, V.I. Savchenko, A.O. Serov, E.A. Skryleva, A.N. Starostin, V.E. Fortov,
Yu.M. Shul’ga, Dokl. Phys. 55 (2010) 55-57.
[24] Z.-Y. Xiao, X. Zhao, X.-K. Jiang, Z.-T. Li, Langmuir 26 (2010) 13048-13051.
[25] F. Cárdenas-Lizana, S. Gómez-Quero, G. Jacobs, Y. Ji, B.H. Davis, L. Kiwi-
Minsker, M.A. Keane, J. Phys. Chem. C 116 (2012) 11166-11180.
[26] G.J. Merten, C. Neis, S. Stucky, V. Huch, E. Rentschler, H. Natter, R.
Hempelmann, K. Stöwe, K. Hegetschweiler, Eur. J. Inorg. Chem. (2012) 31-35.
[27] Z.-G. Tao, X. Zhao, X.-K. Jiang, Z.-T. Li, Tetrahedron Lett. 53 (2012) 1840-1842.
[28] A.D. Pomogailo, G.I. Dzhardimalieva, J. Catalysts (2013) 276210.
[29] G. Dormán, L. Kocsis, R. Jones, F. Darvas, J. Chem. Health Saf. 20 (2013) 3-8.
[30] D.V. LaBarbera, E.B. Skibo, J. Org. Chem. 78 (2013) 11887-11895.
[31] J. Mahmood, D. Kim, I.-Y. Jeon, M.S. Lah, J.-B. Baek, Synlett 24 (2013) 246-248.
[32] K. Hermann, M. Nakhla, J. Gallucci, E. Dalkilic, A. Dastan, J.D. Badjić, Angew.
Chem. 125 (2013) 11523-11526.
19

[33] T. Trzeciak, L. Niedzicki, G. Groszek, P. Wieczorek, M. Marcinek, W Wieczorek,
J. Power Sources 252 (2014) 229-234.
[34] H.T. Clarke, W.W. Hartman, Org. Synth. 2 (1922) 95-97.
[35] M.L. Kastens, J.F. Kaplan, Ind. Eng. Chem. 42 (1950) 402-413.
[36] B.J. Kim, J.M. Choi, J.G. Cho, E.Y. Lee, S.Y. Kim, D.H. Lee, KR Patent 10-2011-
0123222 (2011), to Korea Institute of Industrial Technology.
[37] G.S. Samuelsen, V.L. Garik, G.B.L. Smith, J. Am. Chem. Soc. 72 (1950) 3872-
3874.
[38] J. Miller, M. Searle, US Patent 3 678 114 (1972), to Fisons Ltd.
[39] Y. Chen, D. Zhao, Ranliao Gongye 38 (2001) 24-25.
[40] Q.-H. Yang, Y.-S. Feng, L. Chen, Zhongguo Yiyao Gongye Zazhi 36 (2005) 735-
736.
[41] V.A. Kuz’mina, Yu.T. Lapina, S.V. Sysolyatin, Polzunovsky Vestnik (2008) 129-
131.
[42] Z.-L. Yao, R.-B. Wang, H.-L. Liu, Yingyong Huagong 38 (2009) 1153-1155.
[43] L. Desvergnes, Monit. Sci. 16 (1926) 201-208.
[44] V.A. Semikolenov, Russ. Chem. Rev. 61 (1992) 168-174.
[45] V.F. Surovikin, G.V. Plaxin, V.A. Semikolenov, V.A. Likholobov, I.J. Tiunova,
US Patent 4 978 649 (1990).
[46] V.F. Surovikin, Yu.V. Surovikin, M.S. Tsekhanovich, Russ. J. Gen. Chem. 77
(2007) 2301-2310.
[47] P.A. Simonov, S.Yu. Troitskii, V.A. Likholobov, Kinet. Catal. 41 (2000) 255-269.
[48] C. Le Cocq, J.-Y. Lallemand, J. Chem. Soc. Chem. Commun. (1981) 150-152.
20

[49] A. Preiss, M. Elend, S. Gerling, S. Tränckner, Magn. Reson. Chem. 43 (2005) 736-
746.
[50] J.C. Bottaro, R. Malhotra, A. Dodge, Synth. (2004) 499-500.
[51] B.-S. Liao, S.-T. Liu, J. Org. Chem. 77 (2012) 6653-6656.
[52] M. Peterca, M.R. Imam, C.-H. Ahn, V.S.K. Balagurusamy, D.A. Wilson, B.M.
Rosen, V. Percec, J. Am. Chem. Soc. 133 (2011) 2311-2328.
[53] J.-Y. Park, J.H. Kim, J.M. Kwon, H.-J. Kwon, H.J. Jeong, Y.M. Kim, D. Kim,
W.S. Lee, Y.B. Ryu, Bioorg. Med. Chem. 21 (2013) 3730-3737.
[54] R. Bhadra, R.J. Spanggord, D.G. Wayment, J.B. Hughes, J.V. Shanks, Environ.
Sci. Technol. 33 (1999) 3354-3361.
[55] I. Arai, Y. Sei, I. Muramatsu, J. Org. Chem. 46 (1981) 4597-4599.
[56] A. Calvo Mateo, M. Esteban Morales, T. Gallego Camara, M. Muñoz Mouton, Int.
Patent WO 91/12231 (1991), to Rhone-Poulenc Farma, S.A.E.
[57] Y. Ogata, M. Hojo, M. Morikawa, J. Org. Chem. 25 (1960) 2082-2087.
[58] Z. Wang, Comprehensive Organic Name Reactions and Reagents, John Wiley &
Sons, New York, 2010.
[59] N. Tokura, I. Shirai, T. Sugahara, Bull. Chem. Soc. Jpn. 35 (1962) 722-725.
[60] W.H. Lee, J.S. Kim, S.C. Kim, Bull. Korean Chem. Soc. 30 (2009) 3105-3106.
[61] J.B. Torrance, S. Oostra, A. Nazzal, Synth. Met. 19 (1987) 709-714.
[62] I. Johannsen, J.B. Torrance, A. Nazzal, Macromol. 22 (1989) 566-570.
21

Figure captions
Fig. 1. Hydrogen consumption curve (a) and hydrogen consumption rate versus the
amount of consumed hydrogen (b) for the hydrogenation of TNBA (in the form of
sodium salt) in aqueous solution at a temperature of 323 K and pressure of 0.5 MPa
over the 1 % Pd/C catalyst (400 g of TNBA per 1 g of Pd).
1
Fig. 2. The H NMR spectra (303 K, 400 MHz, H2O) of samples i (a), ii (b), iii (c), iv
(d), v (e), vi (f), and vii (g).
Fig. 3. (a) The change of composition of reaction mixture during the hydrogenation of
TNBA (in the form of sodium salt) in aqueous solution at a temperature of 323 K and
pressure of 0.5 MPa over the 1 % Pd/C catalyst (400 g of TNBA per 1 g of Pd). The
consumption of 0, 1, 3, 5, 6, 8 and 9 moles of hydrogen per mole of TNBA corresponds
to samples i-vii. The main components (TNBA, compounds 1, 2, 5, 6, 8 and 17)
detected by 1H NMR in samples i-vii are indicated; (b) The isolated fragment of Fig. 3a
for components with a lower concentration: 6, 7, 12-17.
13
Fig. 4. The C{1H} JMOD NMR spectrum (303 K, 100.6 MHz, H2O) of sample iv,
whose 1H NMR spectrum is depicted in Fig. 2d. The assignment of signals to compound
5 is shown in Table 2.
Fig. 5. The 1H NMR spectrum (303 K, 400 MHz, H2O) of reaction mixture obtained by
hydrogenation of TNBA (in the form of sodium salt) in aqueous solution (concentration
22

of TNBA is 10 wt. %) at a temperature of 343 K and pressure of 0.5 MPa. The content
of compounds 6 and 7 is 48 %. The other signals in the spectrum belong to compounds
13
5, 8, 12, 13, 15 and 21 (see Table 1). The JMOD C NMR spectrum of this solution,
which was used for identification of compounds 6-8 and 14, is displayed in Fig. S3 (in
the Supplementary materials).
1
Fig. 6. (a) The H NMR spectrum (303 K, 400 MHz, H2O) of the reaction mixture
obtained by hydrogenation of TNBA (in the form of sodium salt) in aqueous solution
(concentration of TNBA is 10 wt. %) at a temperature of 343 K and pressure of 0.5
1
MPa. The spectrum was recorded immediately after sampling. (b) The H NMR
spectrum of the same reaction mixture after holding in an NMR ampoule for 12 h. The
attribution of signals can be found in Table 1.
23