Materials Research Bulletin 61 (2015) 259269

Contents lists available at ScienceDirect

Materials Research Bulletin

journal homepage: www.elsevier.com/locate/matresbu

Chemical degradation of trimethyl phosphate as surrogate for organophosporus pesticides on nanostructured metal oxides
Vclav tengl *, Jir Henych, Tom Grygar, Ral Prez

Materials Chemistry Department, Institute of Inorganic Chemistry AS CR v.v.i., Re,
250 68, Czech Republic

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 27 December 2013
Received in revised form 26 September 2014
Accepted 6 October 2014
Available online 12 October 2014

Nanostructured TiO2 and mixed oxides of Ti and Fe, Hf, In, Mn or Zr -were prepared by homogeneous
hydrolysis of aqueous solution of metal sulphates with urea. The oxides were characterised by X-ray
powder diffraction (XRD), scanning electron microscopy, particle size distribution, surface area and
porosity. The oxide materials consists of a few nanometre primary crystals (mainly anatase) arranged in a
few micrometre regular spherical agglomerates with specic surface area 133511 m2 g
1. The FTIR
diffuse spectroscopy was used for monitoring chemical degradation of trimethylphosphate (TMP) as a
surrogate for organo-phosphorus pesticides under ambient and higher temperatures. Undoped TiO2 and
Ti,Mn-mixed oxide were most active in cleavage (hydrolysis) of CH3O from TMP at room temperature and
100  C. Cleavage of CH3O in the other studied mixed oxides was not complete until temperature exceeds
the boiling point of TMP.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Nanostructured oxides
Stoichiometric degradation
Trimethyl phosphate

1. Introduction
Laboratory research on degradation of chemical warfare agents
(CWAs) as sarin and soman or organophosphate-based pesticides
as malathion is complicated by their high toxicity. Therefore, their
surrogates, containing similar functional groups, are used [1].
Dimethylmethyl phosphate (DMMP) and trimethyl phosphate
(TMP) are the most common surrogates used in this eld [2].
Interactions between DMMP and metal oxides, which includes
chemisorption and surface chemical decay, have been studied
thoroughly, especially between DMMP and TiO2 [211]. Other
studies were dedicated to interactions of DMMP and aluminium
oxides [12], as well as TMP sorption and decomposition on
hematite, maghemite and goethite nanoparticles [13,14] and
zeolites [15]. Also a computational research was conducted on
DMMP and TMP interactions with CaO [16]. Apart from adsorption
process, many of mentioned studies were focused on successive
decomposition processes, both stoichiometric [314] and photocatalytic [2,9,10,14,17]. It should be mentioned, that most of the
studies used a procedure with solid oxide exposed to gaseous
phosphates, while liquid-phase TMP was employed in this work.
Photocatalytic decomposition enhancement by TiO2 doping
was studied already in two last decades of the 20th century [18].
Positive effect of Au-doped TiO2 on DMP decomposition have been

* Corresponding author. Tel.: +420 2 6617 2193; fax: +420 2 2094 0157.
E-mail addresses: stengl@iic.cas.cz, stengl@me.com (V. tengl).
http://dx.doi.org/10.1016/j.materresbull.2014.10.030
0025-5408/ 2014 Elsevier Ltd. All rights reserved.

proven later [5]. With regard to CWAs, TiO2 doping by Zn, Mn, In
and Ge also resulted in better decomposition rates of sulphur
mustard [1923] and soman [1,19,21,23]. Zirconium doped titania
resulted in increased specic surface area, decreased particle size
and enhanced surface hydroxylation leading to better sorption and
decomposition of toxic compounds (or their surrogates) [19,22].
Hirakawa et al. [11] observed that decomposition rate of sarin on
TiO2 was 3.3 times higher than the DMMP. Still, bonding energies of
DMMP sorbed on CaO showed to be closer to the values of
adsorbed tabun bonds, that those in case of TMP [16].
1.1. TMP/DMMP adsorption on TiO2
DMMP adsorption on TiO2 (and iron oxides and hydroxides
[13]) surface can have both molecular and reactive form. While the
rst mentioned occurs mainly by hydrogen bonding between
oxygen atom in phosphoryl group (PO) and surface hydroxyl
group [3,4], the latter takes place on Lewis acid sites (Ti4+) and
active oxygen species of initial surface [3]. Higher surface
concentration of OH group lead to lower TMP stability compared
to dehydroxylated surface [16]. DMMP decomposition on dehydroxylated TiO2 surface seems unlikely [7].
1.2. TMP/DMMP decomposition on TiO2
The decomposition itself is mostly considered as a CH3O group
cleavage from DMMP/TMP molecule at low temperature such as
214 K (
59  C) [8], room temperature [4], or after heating at 200  C

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

260

[3]. CH3O group cleavage occurs at room temperature on zeolites

[15] and aluminium oxides [12]. For decomposition process,
surface hydroxyl groups seems to be essential, thus, it strongly
depends on prior oxide treatment, especially on the annealing [10].
Main product of CH3O group hydrolysis is methanol [4,6,15].
Other products formed on metal oxides surfaces include carboxylates, formates, formaldehyde, CO2 and H2O [4].
In FTIR spectrum, red shift in v(PO) vibration band (at
1237 cm
1 [3] or 1234 cm
1[4]) compared to gas-phase frequency
is an evidence of molecule adsorption on the surface through
phosphoryl group [3,13]. Its decrease is accompanied by rising of v
(PO) band at 11701090 cm
1 [4].
Evidence of CH3O group cleavage can be observed as an
intensity increase of na(Ti(OCH3)) band at 2927 cm
1 (2942 cm
1
[4]) and ns(Ti(OCH3)) at 2827 cm
1 (2834 cm
1 [4]) and also
n(CO) bands at 1115 cm
1 and 1058 cm
1 [8]. More distinct is a
decrease of n(PO) band located between 1170 and 1090 cm
1 [4].
As CH3O groups undergo decomposition at higher temperatures,
corresponding bands d(CH3O) at 1459 cm-1 and n(CO) at
1064 cm-1 decrease in intensity [3]. Some bands may not appear
in the spectrum when gaseous methanol is formed [13,24].
In this study we used nanostructured mixed metal oxides of Ti,
Zr, Fe, In and Mn prepared by homogeneous hydrolysis for
stoichiometric degradation of TMP and DMP. The methods of
operando diffuse FTIR spectroscopy was used for degradation
experiments. The aim was to compare a series of oxides and reveal
chemical modications and modes of application for enhancement
of decontamination of organo-phosphorus agents. In our previous
papers [1,20,21] we described the kinetics of the removal of the
organophospshorus agents from solution into products, not
desorbed by solvents. In this study we focused on chemical forms
of those rmly sorbed products and on their fates at elevated
temperatures.

of dissolving dened amount of TiOSO4 in distilled water acidied

with 10 ml of 98% H2SO4. Then, the specic dopant was added into
solution together and then dened amount of urea was added. The
mixture was heated for approximately 6 h. The formed precipitates
were decanted, ltered off, dried at 105  C and disintegrated by
manual crushing in a mortar.
2.2. Characterization methods
Diffraction patterns were collected with diffractometre Bruker
D2 equipped with conventional X-ray tube (CuKa radiation, 30 kV,
10 mA). The primary divergence slit module width 0.6 mm, Soller
Module 2.5, Airscatter screen module 2 mm, Ni Kbeta-lter
0.5 mm, step 0.00405 , a counting time per a step 3 s and the
LYNXEYE 1-dimensional detector were used. Qualitative analysis
was performed with the DiffracPlus Eva software package (Bruker
AXS, Germany) using the JCPDS PDF-2 database [27].
The specic surface area of samples were determined from
nitrogen adsorptiondesorption isotherms at liquid nitrogen
temperature using a Coulter SA3100 instrument with 15 min
outgas at 150  C. The BrunauerEmmettTeller (BET) method
was used for surface area calculation [28], the pore size
distribution (pore diameter, pore volume and micropore surface
area of the samples) was determined by the BarrettJoyner
Halenda (BJH) method [29].
The particle size distribution was determined by laser scattering using ZEN 1600 equipment (Malvern Co.). The sample was
tested in a square glass cuvette with round aperture (PCS8501)
lled with ethylene glycol.
Scanning electron microscopy (SEM) studies were performed
using a Philips XL30CP microscope equipped with EDX (energy
dispersive X-ray), Robinson, SE (secondary electron), and BSE
(backscattered electron) detectors. The sample was placed on
adhesive C slice and coated with 10 nm thick AuPd alloy layer.

2. Experimental
2.3. Stoichiometric degradation of methyl phosphate
2.1. Preparation of nanostructured metal oxides
All used chemicals, titanium oxo-sulphate, indium(III) sulphate,
zirconium chloride, iron(III) sulphate, hafnium chloride, potassium
permanganate, urea and trimethylphosphate (>99,5% purity) were
supplied by Aldrich. ZrOSO4 and HfSO4 were prepared by a reaction
of stoichiometric mount of ZrCl4 or HfCl4 and sulphuric acid.
Chlorides were dissolved in a 98% sulphuric acid solution in a
porcelain crystallization dish, heated at 100  C until HCl escaped
from the reaction mixture. Then, the solution was heated to
crystallization.
Stoichiometric degradation experiments were run on ve
different mixed oxides based on titania, included non doped
TiO2 and mixed oxides of Ti with In, Zr, Fe/Zr, Mn and Hf. All studied
oxides were prepared by homogeneous hydrolysis of TiOSO4 in
aqueous solutions using urea as the precipitation agent as
described in previous studies [20,23,25,26]. The process consisted

Degradation process has been studied on FTIR Spectrometer

Nicolet Impact 400D equipped with the Praying MantisTM
(Harrick) for diffuse reection measurement (DRIFTS). Heating
of the cell was controlled by ATC-024-3 equipment (Harrick).
The amount of each oxide inserted into the cell was different as
the cell had to be completely lled and the oxide specimens had
different densities. One drop of TMP (approximately 9.23 ml  78.9
mmol) was dosed onto the oxide surface using a syringe. It resulted
in different TMP concentration in the oxide. Also, different specic
surface area and weight of the samples lead to different surface
area available for adsorption process. It is clear, that higher specic
surface area leads to lower coverage of TMP on the oxide surface (in
the case of equal dosage of the phosphates) and conversely.
TMP dosage was probably higher than the amount of
organophosphate which could be sorbed and hydrolysed on the
oxide surface. In case of DMMP, 0.68 mmol/mg of TiO2 (with

Table 1
Properties of the prepared samples.
Sample

Total pore
Cryst. Specic
surface area volume
size
2
1
[nm] [m g ]
[cm3 g
1]

Average
agglomerate
size
[nm]

Average weight of Average surface

oxide [mg]
area of oxide
[m2]

Average TMP amount per

1 mg of oxide [mmol]

Average TMP amount per 1 m2

of oxide [mmol]

TiO2
TiIn
TiZr
TiFeZr
TiMn
TiHf

3.9
8.9
5.6

7.7
5.5

2188
1501
1497
1218
689
660

73.7
64.2
47.6
44.2
86.3
72,5

1,071
1,229
1,658
1,785
0,914
1,088

8,051
4,680
3,246
4,756
4,065
3,050

133.1
262.6
510.7
375.4
224.9
356,8

0.0834
0.2343
0.2908
0.3558
0.3005
0,2555

9.8
16.86
24.31
16.59
19.41
25,868

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

Fig. 1. Temperature programme for in situ operando FTIR.

specic surface area 340 m2/g) can be hydrolysed stoichiometrically [6]. It would mean over-dosage in our case (Table 1).
Prior to the FTIR measurements all samples were annealed in
the cell for 30 min at 300  C in order to remove most of the
humidity and organic species adsorbed on the oxide surface. Before
TMP was dosed onto the oxide surface, all samples were allowed to
cool to room temperature. Consequent measurements were
carried out at ambient temperature, or using temperature
programme shown in Fig. 1.
3. Results and discussion
3.1. Structure and texture of prepared oxides
The identity of synthesized ZrOSO4 and HfSO4 was veried by
XRD (it corresponded to ICDD PDF 01-0366 and HfSO4 to ICDD PDF
24-0467, respectively).
The powder XRD patterns of the prepared TiO2 and mixed
oxides of Ti and Zr, Hf, Fe/Zr, Mn and In are shown in Fig. 2. From
the XRD patterns and the corresponding characteristic 2u values of
the diffraction lines, it can be conrmed that TiO2 has anatase
structure (ICDD PDF 21-1272), no other polymorphs of titania were
formed. The broadening of diffraction lines indicates a small size of
primary titania crystals. The average size t of primary crystals was
calculated from the peakwidth at half maximum B (FWHM), using
the Scherrer equation [30],
t

kl
Bcosu

(1)

where k is a shape factor of the particle (it is 1 if the spherical shape

is assumed), l and u are the wavelength and the incident angle of
the X-rays, respectively. The crystallite size was calculated from a
diffraction plane (1 0 1) of anatase (see Table 1); the crystallites of
doped anatase specimens are larger than of non-doped TiO2.
However, the average sizes of agglomerated particles decreases (as
apparent from Table 1).
Only two of the studied mixed oxides contained a phase
admixture according to XRD (detection limit about 1%) MnO2 (in Ti,
Mn-oxide) and Zr4+ doped TiFe oxides, anatase TiO2 and goethite
FeOOH [22].
Sorption of organophosphates is an essential prerequisite for
their subsequent decomposition, therefore, the specic surface
area of the oxides were analysed. The surface area and total pore
volume of as-prepared samples is listed in Table 1. Doping has a
positive effect on the specic surface area. Even Ti,Mn mixed oxide,
with the lowest surface area among all studied mixed oxides, has
almost twice specic surface areas than non-doped TiO2. The most
efcient in this respect was Zr addition and the second best was codoping by Zr and Fe. Also Hf doping proved to considerably
enhance the specic surface area of the resulting mixed oxide.

261

Particle size (or size of particles agglomerates) strongly

depends on the preparation method. In this study, all samples
were prepared in the same way, thus the agglomerate size was
controlled by chemical composition of the synthesis mixture and
also on additional mechanical disintegration. The size of the
agglomerates has been evaluated; the results are shown in Fig. 3.
Results revealed that addition of dopants has a positive effect on
the agglomerate size as it was reduced in all samples. Adsorption
process on such particles is supposed to be more efcient because
adsorbed molecules show to be more stable than on larger
particles [31].
BET surface area of of the prepared TiO2 and mixed oxides
depends on the amount of dopant in TiO2. The largest specic
surface area (511357 m2g
1) have mixed oxides of Ti with Zr and
Hf, respectively. The effect of Zr doping was described in previous
studies [17,26], Zr doping hindered the particle growth and cause
amorphisation of the titania, which consequently leads to
increased surface area.
According to the shape, the isotherm may be divided into six
basic types: type I for the microporous materials, type II and III for
the nonporous or purely macroporous adsorbents and type IV and
V for the mesoporous material. Type VI is a rare case, where
absorption takes place in several separate steps at different
intervals of pressure adsorbate.
All prepared samples displayed a type I isotherm with
desorption hysteresis loop A [33]. A hysteresis type is due to
cylindrical pores opened at both ends and the microporosity of
pore size distribution is under pore diameter 6 nm [29]. Because
primary crystal size decreases on doping, the growth of the specic
surface area of the prepared oxides must be attributed to the
growth of the pores between the crystals and decreased size of the
agglomerates.
As it follows from the SEM micrographs, (see Fig. 4), the
synthesised samples of the TiO2 and mixed oxides of Ti and Zr, Hf
and In consist of spherical agglomerates (clusters) of narrow size
distribution: the mixed oxides clusters are 12 mm in diameter.
These agglomerates are obviously the result of the synthesis
pathway (urea-hydrolysis), for which such regular spherical
bundles of primary nanocrystals are typical. Hence, the agglomerate size is the feature of synthesis, not mechanical disintegration of
the synthesis products. Based on the HRTEM results [19,20], these
spherical agglomerates are formed by primary 45 nm anatase
nanocrystals (consistently with the line broadening analysis of
XRD), interlayered by a small fraction of amorphous material.
3.2. Interaction of TMP and oxides at room temperature
FTIR intensities of the TMP on oxides differ, as expected,
according to the specic surface areas of the oxides (Fig. 1). The
highest intensities of the spectral bands were found in mixed Ti,Zroxide sample, followed by Ti,Fe,Zr-mixed oxide, Ti,Hf- Ti,In- and Ti,
Mn-mixed oxides and non-doped TiO2. At room temperature,
methyl bands can be observed in CH region in all spectra around
2958 cm
1 and 2854 cm
1, but these frequencies slightly varied in
individual oxides, as seen in Fig. 5. In lower frequency region
(>1600 cm
1) CH3O groups vibrations are represented by wide
band centred at 1458 cm
1. This band has a key importance for TMP
decomposition observation, because it is associated with CH3O
vibrations (similar frequencies were observed also in DMMP on
TiO2 [4,32], thus any change in this region is a clear proof of CH3O
group transformation. In contrast, CH region is less conclusive as
obviously also decomposition products contain CH groups.
Phosphoryl group (PO) can be found in all spectra around
1284 cm
1, which is a higher wavenumber than reported [32].
Another important vibration is that of PO bond located at
1188 cm
1 (not observed in non-doped TiO2 and Ti,Mn-mixed

262

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

Fig. 2. XRD patterns of prepared oxides: (a) TiO2, Ti,Zr-, Ti,Hf-, Ti,In-mixed oxides, (b)Ti,Mn mixed oxide, (c) and Ti,Fe,Zr-mixed oxide.

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

263

Fig. 3. Particle size distribution of the prepared oxides.

oxide with TMP). Missing band can indicate a very fast

decomposition of the corresponding groups or their involvement
in chemisorption. Very broad band containing several overlapping
bands is between 1130 and 1030 cm
1; it is associated mainly with
CO vibrations [32].
Already under room temperature, distinct changes in IR spectra
were observed. In TiO2 sample, bands in CH region increased in
intensity during the rst hour (Fig. 6). Similar progress have been
observed at 1458 cm
1 and 1288 cm
1. Compared to ATR-FTIR
spectrum of liquid TMP, vibrations of PO bonding at 1185 cm
1
have not been observed indicating CH3O group cleavage. It should

be mentioned, that vibrations in this area can be associated to

POH vibrations also, as spectral changes indicates.
Spectral changes in CH region of Ti,In-mixed oxide were
different from those of TiO2 (Fig. 6). For the rst 40 min, CH bands
were increasing in intensity, followed by decrease for next 20 min.
As d(CH3O) at 1458 cm
1 underwent similar changes, we can
conclude that, even at room temperature, some CH3O groups are
eliminated from TMP molecule, consequently transformed, probably into methanol that is subsequently desorbed. The intensity of
the broad n(PO) band decreased similarly and due to simultaneous increase of PO vibrations we assume, that some PO

264

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

Fig. 4. SEM images of the studied oxide samples. (a) TiO2, (b) Zr-doped, (c) In-doped, (d) Fe/Zr doped, (e) Mn-doped and f) Hf-doped Ti oxides.

bonds were transformed into PO bonds. This would only be

possible by chemisorption of the phosphate products to the oxide
via the PO group.
Different reaction course was observed with Ti,Zr- (Fig. 6), Ti,Fe,
Zr- and Ti,Hf-mixed oxides. CH bands started to decrease after 20
(Ti,Zr-mixed oxide) and 30 min (the other oxides). Similarly,
decrease was observed at 1460 cm
1, in contrast with no decrease
of n(PO) band, thus, the transformation of the CH3O group
probably occurred without previous cleavage from the molecule.
Band of PO bonding decreased signicantly in intensity indicating its conversion, probably into PO as in previous cases. It should
be mentioned, that n(PO) decrease caused by CH3O group
cleavage could be masked by intensity increase due to simultaneous PO bond transformation, and vice versa.
In the case of Ti,Mn-mixed oxide, no changes in CH region were
observed (if some bands occurred, they were indistinguishable due
to a high spectral noise), nor for d(CH3O) vibrations at 1463 cm
1.
Similarly, as in the case of non-doped TiO2, n(PO) band is
missing in the spectrum, a clear vidence
of fast CH3O group cleavage. Therefore, bands in CH region and
that at 1463 cm
1 can be attributed to adsorbed CH3O species.
Slightly different frequencies of these bands, compared to the
previous samples, could indicate that. No IR spectral changes, i.e.
no further decomposition occurred at room temperature.

3.3. Interaction of TMP and oxides under elevated temperatures

In non-doped titania and Ti,Mn-mixed oxide annealing at
100  C lead to signicant d(CH3O) bands reduction while CH bands
changed only slightly indicating decomposition of CH3O groups
leading to formation of more stable CH containing products. As
n(PO) band at 1282 cm
1 disappeared, n(PO) band at 1188 cm
1
increased, indicating bond transformation (Fig. 7) similarly as it
was discussed above. Another band, located at 1300 cm
1, can be
observed in the spectrum, probably due to some reversible
structural change of titania (titania obtained by urea-hydrolysis
has a large amount of structural defects). Its origin was conrmed
by additional FTIR measurements of titania without TMP: the same
spectral change was observed. All the changes described above
were observed both on TiO2 and Ti,Mn-mixed oxide.
Spectral changes in Ti,In-, Ti,Zr-, Ti, Fe, Zr- and Ti, Hf-mixed
oxides were quite similar to pure and Mn-modied oxides. CH
bands were reduced in intensities more than on TiO2 sample. This
could have two reasons. Either CH3O groups (on the oxide surface
or still in the TMP molecules) were transformed by oxidative
reactions into different compounds with no CH bonds, or volatile
compounds were desorbed (CH3OH has a boiling point of 65  C).
Simultaneously, d(CH3O) vibrations decreased in intensity, but
remaining bands after 30 min at 100  C showed, that some residual

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

265

Fig. 5. Position of bands in CH region (left), differences in spectra intensities (right) (TMP measurement with oxide as a background).

CH3O groups were still present on the solid. The most visible
change in the spectra was the disappearance of n(PO) band at
1280 cm
1. Despite n(PO) band increase has not been so
signicant, we suppose that PO bond were converted into P
O bonds. Small intensity change of this band could be masked by a
cleavage of methoxy-groups from the molecule. Simultaneously,
many bands appeared in the region centred at 1120 cm
1. We
associate them to some decomposition products strongly bonded
to the oxide surface, as they do not disappear even at higher
temperatures (see below). Note, that boiling point temperature of
TMP is 197  C, DMP 174  C and MMP 250  C. When temperature
was increased to 200  C, remaining bands in CH region on TiO2
sample can be attributed to decomposition products of CH3O
groups, which are not present in the sample anymore, as conrmed
by disappearance of the band at 1452 cm
1. On the contrary, CH
vibrations bands in Ti,In- and Ti,Zr increased and decreased in Ti,
Mn- and Ti,Fe,Zr-mixed oxides. As d(CH3O) band in these samples
almost disappeared, we can assume that the part of the CH3O
groups not escaped as gaseous products as was transformed into
some other, non-volatile CH species. In Ti,Hf-oxide blue shift of
bands in CH region is clearly visible, thus, in conjunction with n(P
O) band we can assume, that both changes are interrelated and
refer to methoxy group cleavage. Moreover, decrease of the n(PO)
band intensity was observed in mixed oxides of Ti with In, Zr and
Fe,Zr only at that increased temperature (not at room temperature).
The intensity of the broad band between 1170 cm
1 and
1000 cm
1 further increased on heating (except for Ti,Mn-oxide),
but rather due to the oxide structural rearrangement and not as a
consequence of TMP reactions, because these bands signicantly
weaken when the oxides were cooled to room temperature. The
decomposition of CH3O group further continues at 200  C. This
temperatures is sufcient for TMP destruction on non-doped
titania or TMP evaporation (TMP boiling point is 197  C). If any CH
groups are present in a sample at 200  C, those cannot be
associated to free TMP molecules, but to either chemisorbed TMP
or to some of its decomposition products. On the contrary, decrease
of CH bands in the spectra at temperatures below 200  C should
indicate chemical, not thermal decomposition of the molecule.

At 300  C, the CH bands are still present in the spectra of Ti,In-, Ti,
Zr-, Ti,Fe,Zr- and Ti,Hf-oxides. Small frequency shifts have been
observed indicating some transformation of sorbed species. Broad
band with a centre at 1340 cm
1 observed only in TiO2 and Ti,Mnmixed oxide (the wavenumber of the band is lower than on TiO2)
continued to rise. We can say, that 300  C is a sufcient temperature
for TMP destruction or desorption from Mn doped titania. At 400  C,
the CH vibrational bands can be seen only in Ti,Fe,Zr-mixed oxide
(Fig. 8) and Ti,Hf mixed oxide spectra (Fig. 7). Despite very signicant
intensity decrease, the presence of these bands denotes, that
decomposition products are different (and more stable) from those
in other oxides. The rest of the spectra just replicates the progress
observed at 300  C. At the temperature 500  C and higher, no CH
species can be seen in the spectra of any oxide. This temperature is
sufcient for destruction of all organic residues irrespective of their
form. There were only phosphorous species located on the oxide
surface, probably forming Ti-O-P-O-Ti network [4], with some
adsorbed simple CO species. TiO2 and Ti,Mn-oxide spectra show
increase of the broad band at 1360 cm
1, respectively 1300 cm
1
under heating up to 600  C. As stressed before, this is probably
related to the heating as there were no remaining species which
could be transformed into products with vibrations in this region.
Some changes in TiO2 and Ti,Mn-oxide specic band mentioned
above occurred, but their origin remains unclear for us.
After 30 min at 600  C, all samples were cooled to a room
temperature and then the spectra were acquired again. Surface rehydroxylation is clearly visible as broad OH negative bands
disappeared. Simultaneously, all bands in 13001000 cm
1 region
decreased signicantly in intensity. As mentioned above, unidentied wide bands at 1360 cm
1 in TiO2 and 1320 cm
1 in Ti,Mnmixed oxide nearly disappeared (Fig. 9).
Previous studies performed on chemical warfare agents revealed,
that Zr-doped titania increased the removal of CWA [26], while Mndoped titania had only half such activity towards soman as compared
to non-doped titania [1]. Based on our different conclusions we must
assume, that chemical interactions between titania and TMP (or
organophosphorus CWAs) that we must distinguish removal of those
agents (deactivation, decontamination) and chemical decomposition, as mentioned before [11].

266

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

Fig. 6. CH and ngerprint region of IR spectra of TiO2, Ti,In- and Ti,Zr-mixed oxides at different temperatures (TMP measurement with oxide as a background).

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

Fig. 7. CH and ngerprint region of the IR spectra of TiO2, Ti,Zr-and and Ti,Hf-mixed oxides (TMP measurement with oxide as a background).

267

268

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

Fig. 8. CH and ngerprint region of IR spectra of Ti,In-, Ti,Fe,Zr- and Ti,Mn-mixed oxides at different temperatures (TMP measurement with oxide as a background).

V. tengl et al. / Materials Research Bulletin 61 (2015) 259269

269

References

Fig. 9. Effect of cooling in the spectra of non-doped TiO2 spectrum after interaction
with TMP and heating according to programme in Fig. 1 (TMP measurement with
oxide as a background).

The DRIFTS study revealed that the improved performance of

doped TiO2 is due to larger specic surface areas and perhaps
stronger interactions with the organophosphorus decay product
rather than their enhanced complete chemical decay.
4. Conclusions
TiO2 doping lead to different mechanism of TMP decomposition
depending on the used modifying elements. Mn-doping favoured
fast decomposition of the TMP molecule (CH3O group cleavage and
further oxide surface methoxylation) already at room temperature,
similarly as what was observed with non-doped titania also.
Contrarily to these two oxides mixed oxides of Ti with In, Zr, Fe, Zr
and Hf caused CH3O group hydrolysis without previous cleavage.
Temperature necessary for vanishing IR bands of TMP decay
products from most mixed oxides was higher than in non-doped
TiO2 and Ti,Mn-mixed oxides. There are at least two possible
explanations. Firstly, the TMP decomposition products on TiO2 and
Ti,Mn-mixed oxide are different and less stable than those
observed on other mixed oxides, or secondly, TiO2 and Ti,Mn
mixed oxide are more efcient in decomposition of any CHcontaining species.
PO bond of TMP showed signicant changes after interaction
with the oxides: its vibrational bands started to decrease already at
room temperature in all mixed oxides. We interpret that change as
PO bond conversion into single or some intermediate kind of P
O bond, as described in previous studies [4,8]. All prepared
samples, hence, started to deactivate the TMP molecule already
at the room temperature, but the way of deactivation is elementspecic. Organic (CH-bearing) products of the reactions had a
different character and are completely decomposed at 300  C (nondoped TiO2 and Ti,Mn-mixed oxide), 400  C (Ti,In- and Ti,Zr-mixed
oxides) and 500  C (Ti,Fe,Zr- and Ti,Hf-mixed oxides).
Acknowledgement
This work was supported by RVO 61388980 and Czech Science
Foundation (Project No. P106/12/1116).

[1] V. tengl, et al., Mesoporous titaniummanganese dioxide for sulphur mustard

and soman decontamination, Mat. Res. Bull. 46 (11) (2011) 20502056.
[2] E.A. Kozlova, P.G. Smirniotis, A.V. Vorontsov, Comparative study on photocatalytic oxidation of four organophosphorus simulants of chemical warfare
agents in aqueous suspension of titanium dioxide, J. Photochem. Photobiol. A:
Chem. 162 (23) (2004) 503511.
[3] C.S. Kim, R.J. Lad, C.P. Tripp, Interaction of organophosphorous compounds
with TiO2 and WO3 surfaces probed by vibrational spectroscopy, Sens.
Actuators B: Chem. 76 (13) (2001) 442448.
[4] D.A. Panayotov, J.R. Morris, Uptake of a chemical warfare agent simulant
(DMMP) on TiO2: reactive adsorption and active site psoisoning, Langmuir 25
(6) (2009) 36523658.
[5] D.A. Panayotov, J.R. Morris, Catalytic degradation of a chemical warfare agent
simulant: reaction mechanisms on TiO2-supported au nanoparticles, J. Phys.
Chem. C 112 (19) (2008) 74967502.
[6] D.A. Trubitsyn, A.V. Vorontsov, Experimental study of dimethyl methylphosphonate decomposition over anatase TiO2,, J. Phys. Chem. B 109 (46)
(2005) 2188421892.
[7] D.A. Trubitsyn, A.V. Vorontsov, Molecular and reactive adsorption of dimethyl
methylphosphonate over (0 0 1) and (1 0 0) anatase clusters, Comput. Theor.
Chem. 1020 (Sept 15) (2013) 6371.
[8] C.N. Rusu, J.T. Yates, Adsorption and decomposition of dimethyl methylphosphonate on TiO2, J. Phys. Chem. B 104 (51) (2000) 1229212298.
[9] A. Kiselev, et al., Adsorption and photocatalytic degradation of diisopropyl
uorophosphate and dimethyl methylphosphonate over dry and wet rutile
TiO2, J. Photochem. Photobiol. A: Chem. 184 (12) (2006) 125134.
[10] J.A. Moss, et al., Adsorption and photodegradation of dimethyl methylphosphonate vapor at TiO2 surfaces, J. Phys. Chem. B 109 (42) (2005) 1977919785.
[11] T. Hirakawa, et al., Specic properties on TiO2 photocatalysis to decompose
isopropyl methylphosphonouoridate and dimethyl methylphosphonate in
gas phase, J. Photochem. Photobiol. A: Chem. 264 (July 15) (2013) 1217.
[12] V.N. Sheinker, M.B. Mitchell, Quantitative study of the decomposition of
dimethyl methylphosphonate (DMMP) on metal oxides at room temperature
and above, Chem. Mater. 14 (3) (2002) 12571268.
[13] P. Mkie, P. Persson, L. sterlund, Adsorption of trimethyl phosphate and
triethyl phosphate on dry and water pre-covered hematite: maghemite, and
goethite nanoparticles, J. Colloid Interface Sci. 392 (Feb 15) (2013) 349358.
[14] P. Mkie, P. Persson, L. sterlund, Solar light degradation of trimethyl
phosphate and triethyl phosphate on dry and water-precovered hematite and
goethite nanoparticles, J. Phys. Chem. C 116 (28) (2012) 1491714929.
[15] Q. Meng, et al., Adsorption of organophosphates into microporous and
mesoporous NaX zeolites and subsequent chemistry, Environ. Sci. Technol. 45
(7) (2011) 30003005.
[16] Y. Paukku, A. Michalkova, J. Leszczynski, Adsorption of dimethyl methylphosphonate and trimethyl phosphate on calcium oxide: an ab initio study,
Struct. Chem. 19 (2) (2008) 307320.
[17] A. Mattsson, et al., Photodegradation of DMMP and CEES on zirconium doped
titania nanoparticles, Appl. Catal. B: Environ. 92 (34) (2009) 401410.
[18] C.K. Grtzel, M. Jirousek, M. Grtzel, Accelerated decomposition of active
phosphates on TiO2 surfaces, J. Mol. Catal. 39 (3) (1987) 347353.
[19] V. tengl, et al., Zirconium doped nano-dispersed oxides of Fe: Al and Zn for
destruction of warfare agents, Mater. Charact. 61 (11) (2010) 10801088.
mec, In3+-doped TiO2 and TiO2/In2S3 nanocomposite
[20] V. tengl, F. Oplutil, T. Ne
for photocatalytic and stoichiometric degradations, Photochem. Photobiol. 88
(2) (2012) 265276.
[21] V. tengl, et al., Mesoporous manganese oxide for warfare agents degradation,
Micropor. Mesopor. Mater. 156 (July 1) (2012) 224232.
[22] V. tengl, et al., Sulphur mustard degradation on zirconium doped TiFe
oxides, J. Hazard. Mater. 192 (3) (2011) 14911504.
[23] V. tengl, et al., Ge4+ doped TiO2 for stoichiometric degradation of warfare
agents, J. Hazard. Mater. 227228 (Aug 15) (2012) 6267.
[24] K. Varazo, et al., Methanol chemistry on Cu and oxygen-covered Cu nanoclusters
supported on TiO2(110), J. Phys. Chem. B 108 (47) (2004) 1827418283.
[25] V. tengl, et al., Impact of Ge4+ ion as structural dopant of Ti4+ in anatase:
crystallographic translation, photocatalytic behavior, and efciency under UV
and VIS irradiation, J. Nanomater. 2012 (2012) 11.
[26] V. tengl, et al., Zirconium doped titania: destruction of warfare agents and
photocatalytic degradation of orange 2 dye, Open Process Chem. J. 1 (2008) 17.
[27] J. CPDS, PDF 2 database, Release 50, International Centre for Diffraction Data,
Newtown Square. 2000.
[28] S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multimolecular
layers, J. Am. Chem. Soc. 60 (1938) 309319.
[29] E.P. Barrett, L.G. Joyner, P.P. Halenda, The determination of pore volume and
area distributions in porous substances. 1. Computations from nitrogen
isotherms, J. Am. Chem. Soc. 73 (1) (1951) 373380.
[30] P. Scherrer, Bestimmung der Gre und der inneren Struktur von Kolloidteilchen mittels Rntgenstrahlen, Gttinger Nachrichten Gesell 2 (1918) 98.
[31] H. Wang, F. Shadman, Effect of particle size on the adsorption and desorption
properties of oxide nanoparticles, AIChE J. 59 (5) (2013) 15021510.
[32] C.S. Kim, R.J. Lad, C.P. Tripp, Interaction of organophosphorous compounds
with TiO2 and WO3 surfaces probed by vibrational spectroscopy, Sens.
Actuators B: Chem. 76 (13) (2001) 442448.
[33] S. Lowell, J.E. Shields, Powder Surface Area and Porosity, Springer, 1991.