Destructive adsorption of carbon tetrachloride on nanometer titanium dioxide

Article abstract of DOI:10.1039/b312082a

The reaction between nanometer TiO₂ and CCl₄ in the presence of oxygen has been studied in order to investigate the potential of nanometer TiO₂ as a destructive reagent for chlorinated hydrocarbons. Two kinds of nanometer TiO₂ of different size were synthesized. The properties of the destructive adsorption of CCl₄ over nanometer TiO₂ of about 40 nm in size were compared with that over nanometer TiO₂ of about 80 nm in size. The reactivity toward CCl₄ of nanometer TiO₂ of about 40 nm in size was remarkably higher than that of nanometer TiO₂ of about 80 nm in size. The products produced included CO₂, COCl₂, Cl₂, titanium oxychloride, TiCl₄, carbon, CO and HCl. HCl was a major gaseous product. Hydrogen of HCl came from traces of water in the carrier gas and surface OH groups of TiO₂. The formation of TiCl₄ made the interaction surface renew so that CCl₄ could further interact with the bulk of the particles. TiO₂ was regenerated through the exchange of chlorine with oxygen at 550°C in the carrier gas so that larger quantities of CCl₄ are decomposed over nanometer TiO₂ in the presence of oxygen. A mechanism has been proposed for the destructive adsorption and desorption of CCl₄ over nanometer TiO₂ in the presence of oxygen.

Full text of DOI:10.1039/b312082a

View Article Online / Journal Homepage / Table of Contents for this issue
R E S E A R C H P A P E R
Destructive adsorption of carbon tetrachloride on nanometer
titanium dioxide
a
Guohong Liu, Jinglin Wang, Yongfa Zhu and Xinrong Zhang*
ab
a
a
a
Department of Chemistry, Tsinghua University, 100084, Beijing, P. R. China.
E-mail: xrzhang@chem.tsinghua.edu.cn; Fax: þ86-10-6277-2485; Tel: þ86-10-6278-7678
Department of Chemistry, Jindongnan Teachers college, 046011, Changzhi, P. R. China
b
Received 29th September 2003, Accepted19th December 2003
F|rst published as an AdvanceArticle on the web 27th January 2004
The reaction between nanometer TiO and CCl in the presence of oxygen has been studied in order to
2
4
investigate the potential of nanometer TiO
nanometer TiO of different size were synthesized. The properties of the destructive adsorption of CCl over
2
as a destructive reagent for chlorinated hydrocarbons. Two kinds of
2
4
nanometer TiO
The reactivity toward CCl of nanometer TiO of about 40 nm in size was remarkably higher than that of
2
of about 40 nm in size were compared with that over nanometer TiO
2
of about 80 nm in size.
4
2
nanometer TiO
oxychloride, TiCl , carbon, CO and HCl. HCl was a major gaseous product. Hydrogen of HCl came from
2 2 2 2
of about 80 nm in size. The products produced included CO , COCl , Cl , titanium
4
traces of water in the carrier gas and surface OH groups of TiO
surface renew so that CCl could further interact with the bulk of the particles. TiO was regenerated through
2 4
. The formation of TiCl made the interaction
4
2
ꢀ
the exchange of chlorine with oxygen at 550 C in the carrier gas so that larger quantities of CCl
decomposed over nanometer TiO in the presence of oxygen. A mechanism has been proposed for the
4
are
2
4 2
destructive adsorption and desorption of CCl over nanometer TiO in the presence of oxygen.
An alternative and cleaner destruction method than incinera-
tion is oxidative catalysis over metal oxides as proposed in
the literature.
Introduction
1
13–17
Carbon tetrachloride (CCl
and has been widely used as a chemical intermediate for
producing chlorofluorocarbons (CFCs) such as CCl and
CCl F, which can continuously destroy the stratospheric
4
) is a volatile organic compound,
Recently, Klabunde and co-workers have shown that CaO
and MgO are able to destroy chlorinated hydrocarbons,
yielding CO traces of COCl , and the corresponding metal
2 2
F
3
2
2
2
18
ozone. CCl
vent. The increasing amounts of CCl released in the environ-
4
is also extensively applied as an industrial sol-
chlorides. They also showed that complete transformation
of the oxide into the chloride was possible by depositing
3
4
1
9
ment, together with its suspected toxicity and carcinogenic,
hepatotoxic, and nephrotoxic properties,
researchers worldwide to find clean and effective methods for
2 3
transition-metal oxides (e.g., Fe O ) onto the oxide.
4
,5
20
Weckhuysen et al. indicated that the activities of the alka-
have prompted
line earth metal oxides for decomposing CCl paralleled their
4
its destruction.
Methods for the disposal of CCl
basicity; CO2 was the only gas-product formed when CCl
ꢀ
4
6
include incineration and
4
was reacted with BaO and SrO at 200–300 C. At 450–
ꢀ
7
photodecomposition. Incineration is presently the most
widely used method. Unfortunately, high temperatures are
required, and toxic byproducts, such as dioxins and furans
600 C, CaO and MgO were very active, and COCl was found
2
2
in the destruction of CCl . Weckhuysen et al. also studied
1
4
the destructive adsorption of CCl4 on La O and CeO2 in
2
3
8
are frequently formed. Therefore, there appears to be a need
for the development of simpler and cleaner methods for safely
the absence of oxygen by X-ray photoelectron spectroscopy
and in-situ Raman spectroscopy; La O was much more reac-
2
3
destroying CCl
4
.
tive than CeO . Recently, the Weckhuysen group has deve-
2
During the past decade, nanoparticle research has become
quite popular in various fields of chemistry, physics and mate-
rials. Nanomaterials are clusters of atoms or molecules of
loped novel and stable catalyst systems based on supported
lanthanide oxides for the hydrolysis of CHCs into CO and
2
9
HCl at relatively low temperatures in the presence of steam.
The active phase consists of a lanthanide oxide chloride, and
COCl2 is a reaction intermediate involved in this catalytic
materials, ranging in size from <1 nm to almost 100 nm, fall-
ing between single atoms or molecules and bulk materials.
1
0
2
2,23
Since the grain sizes are so small, a significant volume of the
microstructure in nanometer materials is composed of inter-
faces, mainly grain boundaries, i.e., a large volume fraction
of the atoms resides in grain boundaries. Consequently, nano-
meter materials display novel and often-enhanced properties
hydrolysis process.
According to the literature, the destructive adsorption of
2
1
21
CCl on metal oxides such as La O , CeO , alkaline earth
4
2
19,27–29
3
2
30,31
1
8,20,24–29
31
and NiO was
metal oxides,
performed in the absence of any oxidant, and therefore only
Fe O ,
CuO
2
3
1
1
compared to traditional materials. There has been a conti-
nued increase in the number of research investigations in recent
years on the potential chemical applications of these novel
materials. Nanometer metal oxides, an example of important
nanomaterials, possess very high lattice energies and melting
points due to their high ionicity. Furthermore, these solids
can exist with numerous surface sites with enhanced surface
N , He or Ar were used as carrier gases. However, the amount
2
of carbon tetrachloride removed in the presence of oxygen was
3
2
increased slightly although the influence of oxygen on lantha-
nide oxide materials brought a detrimental effect on the cataly-
2
2,23
tic properties of the materials.
samples, CCl always coexists with oxygen (O ). Hence, the
Moreover, for atmospheric
4
2
destructive adsorption of CCl in the presence of O is more
4
2
1
2
reactivity, such as crystal corners, edges, or ion vacancies.
relevent than in the absence of O2 .
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
985

View Article Online
ꢁ
1
reactor at a rate of 12 mL min . Measured amounts of
Another disadvantage in the current reports is that the life-
time of the reactor used for the destructive adsorption of CCl4
is very limited, because of the formation of metal chlorides and
CCl were injected through a septum inlet positioned at the
4
beginning of the reactor so that the CCl
evaporated immediately in the reactor. CCl
4
in the flow stream
was introduced
metal oxychlorides, e.g., BaCl
2
, etc. To regenerate the reactor,
4
a series of processes including dissolution, precipitation, filtra-
tion, and calcination was needed, which limits the application
of the reactor to real environmental treatment of pollution by
in the reactor as a pulse of 2 mL, and the time period between
the pulses was about 5 min. The contact time was 15 s.
The gaseous products, eluted by the carrier gas, were col-
lected in an airtight syringe. The volume of the gaseous pro-
ducts collected was 40 mL, and then injected into an IR cell
equipped with two KBr windows. FTIR spectra were recorded
on a Perkin-Elmer FTIR spectrometer (1760X) for the gaseous
2
0
CCl
The aim of this work was to investigate the potential of nano-
meter TiO for the destructive adsorption of CCl in the pre-
sence of oxygen. The formation of TiCl during the destructive
adsorption of CCl makes the interaction surface renew so that
CCl can further interact with the bulk of the particles. TiO
4
in air.
2
4
4
4
4
products after each injection of 2 mL CCl by comparing a
4
2
background under the same conditions. The gaseous products
after chemical desorption were also monitored in the same
manner.
was regenerated through the exchange of chlorine with oxygen
ꢀ
in the carrier gas at 550 C so that larger quantities of CCl
can be decomposed over nanometer TiO
oxygen.
4
2
in the presence of
The solid products were monitored using X-ray photo-
electron spectroscopy (XPS). The XPS spectra were acquired
using a Perkin-Elmer (PHI) model 5300 ESCA spectrometer.
XPS spectra were typically recorded using a pass energy of
89.45 eV, a step increment of 0.8 eV, and an Al anode power
Experimental
ꢁ
10
of 250 W at a base pressure of 10
Torr.
Reagents
Quantitative analysis of Ti was carried out using an induc-
tively coupled plasma-mass spectrometer (ICP-MS) (Elan
6000 (PE Sciex, Toronto, Canada)). This system consists of a
cross flow nebuliser and a Scott double pass Ryton spray
chamber. A Ti standard was purchased from the National
Research Center of Certified Reference Materials (China).
The procedure for synthesis of TiO nanoparticles was as fol-
2
lows: 25 mL of tetrabutyl titanate was slowly added to 50 mL
ethanol solution at room temperature and 0.1 mL water was
2 5 2
added to this mixture. Then, 3 mL NH(OC H ) was added
dropwise to the mixed solution under stir. A light yellow solu-
tion was obtained and allowed to stand for 3 days. Then 95%
ethanol solution was added into the solution. A transparent
primrose sol–gel was obtained after the mixed solution was
hermetically sealed for 1 day. After being allowed to dry, it
was divided into two portions and calcined in air. The tem-
Analysis of Cl
2
was performed by bubbling the effluent
ꢁ1
stream through a 0.1 mol L NaOH solution. The Cl concen-
tration was determined by titration with sodium thiosulfate
using starch as an indicator after the collected solution was
2
3
3
reacted with about 1 g KI.
ꢀ
ꢀ
ꢁ1
perature was raised to either 650 and 800 C at 6–8 C min
and maintained for 2 h. Then, the TiO powder was cooled
to room temperature in air. Two TiO powders were obtained
2
Results and discussion
2
ꢀ
ꢀ
(
Ti-650 (650 C) and Ti-800 (800 C)). Carbon tetrachloride
In order to compare the destructive adsorption of different
sizes of nanometer TiO , two nanometer TiO samples were
synthesized as described above. Ti-650 and Ti-800 were about
0 and 80 nm in size, respectively, as seen from Fig. 1. As
was of analytical reagent grade purchased from Beijing
Chemical Factory.
2
2
4
Characterization of materials
shown in Fig. 2, both anatase and rutile crystals were present
in Ti-650 and Ti-800, most of the crystals were anatase in
Ti-650 while most of crystals were rutile in Ti-800.
Materials were characterized by transmission electron
microscopy (TEM), X-ray powder diffraction (XRD), and
Brunauer-Emmet-Teller (BET) surface area analysis. The
topography and particle size of the materials were measured
using a Hitachi H-800 transmission electron microscope. The
accelerating voltage of the electron beam was 200 kV. The
samples for TEM experiments were dispersed in water with
ultrasonic treatment for approximately 10 min and then trans-
ferred to a copper grid with a polymer membrane in order to
disperse the nanoparticles. XRD measurements were obtained
using a Bruker D8 Advance diffractometer with Cu-Ka radia-
tion. BET surface areas were measured on a Micromeritics
ASAP 2010 instrument using nitrogen gas adsorption at
liquid-nitrogen temperature.
The results obtained from BET measurements showed that
the surface area and average pore diameter of Ti-650 were
2
.76 m g and 6.74 nm, respectively, while the surface area
ꢁ1
9
and average pore diameter of Ti-800 were 8.60 m g and
2
ꢁ1
1
3.3 nm, respectively.
The gaseous products evolving from the titanium dioxide
surface during the destructive adsorption process were mainly
monitored by FTIR spectroscopy. Results obtained for Ti-650
Studies of destruction of CCl4
The adsorption and chemical desorption of carbon tetrachlor-
ide were performed with a TiO
of nanometer TiO was introduced into a quartz microcolumn
2
packed quartz column. 100 mg
2
(
50 mm ꢂ 4 mm i.d.) plugged with a small portion of glass
wool at both ends. One end of this column was connected to
the sample inlet. Before use, the quartz column was condi-
tioned by programming temperature from room temperature
to 600 C at a rate of 20 C min to eliminate the influence
of previous absorbates. The temperature of the reactor was
monitored and controlled by a temperature controller. The
ꢀ
ꢀ
ꢁ1
carrier gas was composed of 78% N
(
2 2
, 21% O and 1% Ar
by volume). The carrier gas was allowed to flow through the
Fig. 1 TEM images (a) Ti-650 and (b) Ti-800.
9
86
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4

View Article Online
In the case of Ti-800, carbon tetrachloride was completely
removed in each injection until a total of 50 mL was injected.
After that, a very small portion of CCl
and released from the reactor with the reacted products. The
undecomposed CCl increased as more carbon tetrachloride
was introduced. The amount of phosgene also increased with
increasing the amount of CCl undecomposed, but the amount
4
was undecomposed
4
4
of phosgene almost did not change until a total of 100 mL pos-
sibly because active oxygen atoms over the surface of titania
were replaced by a certain number of chlorine atoms, and
led to homeostasis. The sample weighed 0.1047 g before
reaction and 0.0877 g after reaction, indicating a weight loss
of 16.2%.
To investigate the effect of the temperature on the decompo-
sition efficiency the destruction of CCl over Ti-650 was stu-
4
ꢀ
died at different temperatures (250, 350, 450 and 550 C) by
FTIR. The results obtained are shown in Table 1 and in Fig.
(a)–(e). At 250 C traces of undecomposed CCl
observed in the releases from the reactor even during the initial
ꢀ
3
4
were
injections, suggesting that a portion of CCl
ꢀ
4
was decomposed
at 250 C. The overall ability of Ti-650 to decompose carbon
tetrachloride increased with increasing the temperature from
ꢀ ꢀ
2
50 to 550 C. At 550 C no carbon tetrachloride was observed
with release from the reactor during the destructive adsorption
until a total of 180 mL was injected. Carbon monoxide was
apparently observed in the gaseous products at 550 C, but
ꢀ
Fig. 2 XRD patterns of (a) Ti-650 and (b) Ti-800.
ꢀ
no carbon monoxide was observed at 250 C. However, when
ꢀ
the reactor was heated to 550 C after a total of 120 mL CCl
4
ꢀ
was injected at 250 C, the releases from the reactor contained
carbon monoxide, indicating that carbon was produced during
the destructive adsorption at lower temperature, which was
further proved by the XPS experiments, and carbon reacted
with oxygen in the carrier gas to produce carbon monoxide
at higher temperature. Tetrachloroethylene was not detected
ꢀ
was compared with those for Ti-800 at 450 C as shown in
Table 1 and in Fig. 3(a)–(e). The observed main products from
the decomposition of carbon tetrachloride over Ti-650 and
ꢀ
Ti800 at 450 C included carbon dioxide as indicated by the
ꢁ
1
vibration at 2340 cm , phosgene as shown by the
nC
=
O
O
ꢁ1
n
C
=
vibrations at 1827 and 1812 cm and by the nC–Cl vibra-
ꢀ
ꢁ
1
at temperatures in the range from 250 to 550 C, because the
ꢁ1
tions at 863 and 849 cm , hydrogen chloride as revealed by
ꢁ
1
n
literature reported that tetrachloroethylene was formed during
C–Cl vibration at 920 cm was not observed, although some
the nH–Cl vibration at 3100–2640 cm and a small amount
of carbon monoxide as shown by the n vibrations at 2170
CO
19,32
ꢀ
ꢁ
1
the destructive adsorption of CCl
4
.
At 350 C the change
and 2110 cm . However, the two different titania samples
exhibited significant quantitative and qualitative differences
in their behavior toward carbon tetrachloride.
of the amount of phosgene formed was apparently observed.
First, the amount of phosgene increased with increasing the
total amount of CCl
that, the amount of phosgene decreased with continuing to
increase the total amount of CCl injected, because at the
4
injected until a total of 32 mL, but after
In the case of Ti-650, carbon tetrachloride was completely
removed in each injection until a total of 80 mL was injected.
4
After that, a very small portion of CCl
4
as indicated by the
C–Cl vibration at 794 cm was undecomposed and released
from the reactor with the produced products. The amount of
phosgene increased with increasing the amount of CCl unde-
ꢁ
1
initial reaction stage the chlorination of the surface of titanium
dioxide was relatively complete, then the ability of chlorination
of the surface of titanium dioxide decreased as a result of the
replacement of oxygen on the surface of titanium dioxide by
the chlorine atoms. Therefore, the amount of phosgene
produced increased. However, when the capacity of titanium
dioxide to decompose carbon tetrachloride started to decline,
the amount of phosgene produced decreased as the amount
of carbon tetrachloride eluted from the reactor further
n
4
composed, but before that the amount of phosgene almost did
not change, which suggested that phosgene was an intermedi-
ate product in the decomposition of CCl . The sample weighed
4
0.1050 g before reaction and 0.1010 g after reaction, indicating
a weight loss of 3.8% because a portion of titania was
converted to titanium tetrachloride as illustrated later by
ICP-MS in this paper.
ꢀ
increased. At 550 C no phosgene was observed, which showed
that the phosgene produced further reacted with the surface.
The results above proved that phosgene was an intermediate
product in the decomposition of CCl . The weight losses at
4
Table 1 Destructive adsorption of CCl
4
on nanometer TiO
2
ꢀ
2
50, 350, 450, 550 C were 5.1, 10.0, 3.8 and 0.0%, respectively.
The weight loss mainly resulted from formation of titanium
tetrachloride as illustrated by ICP-MS. The reason why the
weight loss increased and then decreased with the increase of
temperature was that at lower temperature the chlorine in
the oxychloride produced during the chlorination was not easy
to be replace by oxygen atoms in the carrier gas, and oxychlor-
ide was further chlorinated to form titanium tetrachloride;
however, at higher temperature the chlorine in the oxychloride
produced during the chlorination readily reacted with oxygen
atoms in the carrier gas to form titanium dioxide, and titanium
tetrachloride was no longer formed. The weight loss was less at
Temp./ Initial
ꢀ
Final Weight
a
a
a
b
Sample
Ti-650
C
weight /g weight /g loss (%) BT /mL
250
0.1072
0.1040
0.1050
0.1050
0.1047
0.1017
0.9360
0.1010
0.1050
0.0877
5.1
10.0
3.8
2
4
3
4
5
50
50
50
80
> 180
50
0.0
16.2
Ti-800
450
a
Data refer to the weight of the sample before adsorption and after
b
adsorption of 180 mL of CCl
amount of CCl completely removed before any undecomposed CCl
is observed in the effluent.
4
.
BT, break through volume, the
4
4
ꢀ ꢀ
2
decomposed at 250 C. At 550 C the weight loss was 0.0%
50 C than at 350 C because less of carbon tetrachloride was
ꢀ
ꢀ
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
987

View Article Online
ꢀ
ꢀ
Fig. 3 FTIR spectra of the gaseous products of CCl
amount of CCl (in mL) introduced.
4
over Ti-650 at 250 (a), 350 (b), 450 (c) and 550 C (d) and over Ti-800 at 450 C (e): t ¼ total
4
during the destructive adsorption of a total of 180 mL injected,
indicating that Ti-650 had remarkable potential to decompose
4
CCl over Ti-650 was 100% from the first run to the fortieth
run. Then, the decomposition efficiency of CCl over Ti-650
4
CCl
4
.
decreased slightly with an increase of the run number. Even so,
the decomposition efficiency was 99.00% for the ninetieth run.
An experiment was carried out to examine the possibility of
reusing the catalyst. The column packed with Ti-650 was con-
ditioned to the adsorption temperatures (250, 350, 450 and
ꢀ
5
50 C). 2 mL CCl
4
was injected as a pulse into the carrier
adsorption on the catalyst. The temperature
was held for 5 min until all CCl molecules adsorbed on the
gas flow for CCl
4
4
catalyst were decomposed. This procedure was repeated 90
times at the same working conditions. The peak height of
ꢁ
1
the nC–Cl vibration at 794 cm was used as a quantitative
standard for CCl
sition efficiency at different temperatures. At 250 C, the max-
imum decomposition efficiency of CCl4 over Ti-650 was
4
. Fig. 4 shows the run number vs. decompo-
ꢀ
7
CCl4 over Ti-650 decreased rapidly with an increase of the
4.77% of the first run. Then, the decomposition efficiency of
ꢀ
run number. At 350 C, the maximum decomposition efficiency
of CCl over Ti-650 was 99.96% for the first run. Then, the
4
4
decomposition efficiency of CCl over Ti-650 decreased with
the increase of the run number with a value of 87.45% for
ꢀ
Fig. 4 Run number dependence of decomposition efficiency of CCl
ꢀ
4
the ninetieth run. At 450 C, the decomposition efficiency of
over Ti-650 at different temperatures (250, 350, 450 and 550 C).
9
88
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4

View Article Online
ꢀ
At 550 C, the decomposition efficiency of CCl
was 100% from the first run to the ninetieth run. The results
4
over Ti-650
indicated a satisfactory efficiency of decomposition of CCl
ꢀ
4
over TiO
2
for reuse at temperatures higher than 450 C.
In the literature it was reported that the thermal decomposi-
ꢀ
tion of CCl
4
over a-Fe
2
O
3
at temperatures of 400–600 C in a
1
9
fixed-bed pulse reactor resulted in the formation of Cl
Therefore we investigated whether Cl was formed during
the destructive adsorption and during the desorption over
Ti-650 using an iodometric method. 10 mL of 0.1 mol L
NaOH solution was used as the absorption solution. The col-
lection times during the destructive adsorption at different tem-
2
.
2
ꢁ
1
ꢀ
peratures (250, 350, 450 and 550 C) and during the desorption
ꢀ
at 550 C were 1 and 10 min, respectively. The concentration of
Cl in the absorption solution during the chlorination were 0
2
Fig. 5 ICP-MS analysis of TiCl
4
evaporation corresponding to the
ꢁ
1
ꢀ
ꢀ
ꢁ4
ꢁ1
ꢀ
ꢁ4
mol L at 250 C, 0.8 ꢂ 10 mol L at 350 C, 1.2 ꢂ 10
concentration of the CCl injection.
4
ꢁ
1
ꢁ4
ꢁ1
ꢀ
mol L at 450 C, and 1.5 ꢂ 10 mol L at 550 C, respec-
tively, indicating that Cl
absorption except at temperatures lower than 250 C and the
2
was produced during the destructive
ꢀ
corresponded to hydroxyl species on the surface because the
surface of TiO exposed in air before the measurement
absorbed some water. No apparent differences in the Ti 2p
and the O 1s spectra were observed for samples before destruc-
tive adsorption, after destructive adsorption and after regen-
eration, indicating that the solid sample was mainly
amount of Cl
in the range from 250 to 550 C. The concentration of Cl in
2
increased with an increase of the temperature
ꢀ
2
3
6
2
ꢀ
the absorption solution during the desorption at 550 C after
the adsorption at different temperatures (250, 350, 450 and
ꢀ
ꢀ
ꢁ1
ꢁ1
ꢀ
ꢁ4
ꢁ1
5
3
L
50 C) were 3.4 mol L at 250 C, 4.6 ꢂ 10 mol L at
ꢁ4
ꢀ
ꢁ4
presented in the form of TiO whether before or after reaction.
2
50 C, 6.4 ꢂ 10
mol L
at 550 C, respectively, indicating that Cl was produced
at 450 C, and 6.5 ꢂ 10
mol
ꢁ
1
ꢀ
Two carbon C 1s peaks at a binding energy of 284.57 and
88.37 eV were observed on all the samples. The low-
2
2
during the destructive absorption and the amount of Cl2
increased with the increase of the temperature. Cl produced
binding-energy region was associated with surface pollution
which corresponded to the fact that the samples are exposed
to air before the XPS experiments while the high-binding-
energy region was due to carbonate species on the surface
2
during the destructive adsorption in the presence of oxygen,
also, resulted not from decomposition of phosgene but from
decomposition of oxychloride produced during the chlorina-
tion, because no chlorine gas formed although phosgene was
3
6
formed during synthesis of TiO
2
.
The intensity of the C 1s
ꢀ
peak at the binding energy of 284.57 eV after destructive
adsorption was higher than before destructive adsorption,
which indicated that deposited coke was formed during
destructive adsorption. No differences of the C 1s spectra
were observed between samples before destructive adsorption
and after regeneration, indicating that deposited coke was
decomposed at relatively high temperature.
produced at 250 C.
Our preliminary experiment has shown that when 50 mL
CCl was injected as a pulse into the reactor, a yellow liquid
4
which produced a white fog when exposed to air was effluent
from the reactor, suggesting that the yellow liquid produced
ꢀ
34
(boiling point ca. 136 C) evaporated over
may be TiCl
the quartz column at 250 C and liquefied in the vessel behind
4
ꢀ
The Cl 2p spectrum at a binding energy of 197.97 eV was
observed after destructive adsorption, but no chlorine atoms
could be detected before destructive adsorption, which proved
that the chlorinated product was produced after destructive
adsorption. The chlorinated product was a partially chlori-
the reactor. To investigate whether TiCl
porated during destructive adsorption at 250 C, the absorp-
tion solutions of gas-phase products after decomposition of
4
was formed and eva-
ꢀ
CCl
sensitive technique for the detection of Ti species. CCl
4
on nanometer TiO
2
were detected using ICP-MS, a very
was
4
4
nated product, not TiCl , since the concentration of the Cl
added in one batch of 600 mL instead of pulses. The results
are shown in Fig. 5. The titanium species was not detected
by ICP-MS when a portion of CCl below 10 000 ppm was
4
species was extremely low in comparison with the Ti and O
ꢀ
species, and since the boiling point of TiCl
4
ꢀ
was 136.4 C.
No chlorine was detected after heating to 550 C, which indi-
cated that the partially chlorinated product was regenerated
due to the exchange of oxygen with chlorine.
injected. For higher concentrations of CCl , however, the eva-
4
porated Ti species was detected by ICP-MS. It seems that
TiCl could only be produced from the reaction of TiO and
CCl above 10 000 ppm CCl , which suggested that the solid
4
2
4
4
product obtained from the chlorination was titanium oxy-
chloride as a result of partial chlorination of nanometer TiO
According to the literature, metal oxides were transformed
to M Cl (such as BaCl ) by the injection of CCl into the col-
umn during the destructive adsorption of CCl4 .
order to monitor the chlorination process of destructive
adsorption in the present study, XPS was employed for the
characterization of the solid products on nanometer titanium
2
.
Discussion
x
y
2
4
The thermal decomposition of CCl over nanometer TiO at a
4
2
1
8–20,24–29
ꢀ
In
temperature of 250–550 C in the quartz column resulted in
HCl, CO , COCl , Cl , titanium oxychloride, TiCl , CO,
2
2
2
4
and carbonaceous deposits. HCl was shown as a major gas-
eous product in all IR spectra. Hydrogen of HCl came from
traces of water in the carrier gas and surface OH groups of
dioxide particles when 50 mL CCl was injected as a pulse.
4
ꢀ
The sample was heated to 550 C and held for 15 min after
destructive adsorption. Fig. 6 shows the acquired Ti 2p, O 1s,
TiO . The above results obtained suggest a possible reaction
2
pathway for the destructive adsorption of CCl4 over nan-
ometer TiO2 as shown in Scheme 1. The first step involves
the physisorption of CCl4 on the nanometer TiO2 surface.
The relatively positive carbon atom interacts with a negative
oxygen site and the relatively negative chlorine atoms interact
with positive metal ion sites to form an intermediate. The
covalent C–Cl is then polarized and broken by the TiO lattice.
The overall result of this process leads to a dissociative adsorp-
3
5
Cl 2p and C 1s spectra. The Ti 2p spectra included two lines:
Ti 2p3/2 and Ti 2p1/2 . The Ti 2p3/2 line was identified by a sin-
gle peak at a binding energy of 458.67 eV. The separation
between the Ti 2p3/2 and Ti 2p1/2 peaks was 5.7 eV. The mea-
sured binding energy of the Ti 2p3/2 and Ti 2p1/2 peaks indi-
cated that titanium mainly existed as TiO . The O 1s spectra
2
showed a main peak at a binding energy of 529.27 eV with a
shoulder at a binding energy of 530.97 eV. This shoulder
4 2
tion of CCl to form COCl as a result of oxygen abstraction
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
989

View Article Online
Fig. 6 XPS spectra: (a) Ti 2p, (b) O 1s, (c) Cl 2p, (d) C 1s.
1
9,20,32
from the TiO
as a result of chlorine abstraction from the CCl molecules.
2
surface and a partially chlorinated TiO surface
literature.
the surface to form CO after abstraction of another oxygen
2
COCl is readsorbed, and further reacts with
4
2
COCl
2
is an intermediate product in the decomposition
of CCl4 as proved above, which is consistent with the
from the surface. Further chlorination results in the formation
of a fully chlorinated material.
The formation of hydrogen chloride during the destructive
adsorption of CCl4 could result from COCl2 reaction with
traces of water in the carrier gas, because the amount of hydro-
gen chloride scarcely changed with the increase of the total
ꢀ
amount of CCl
4
at the same temperature, such as at 250 C,
and the fact that traces of water existed in the carrier gas
was proved by measurement of the carrier gas using FTIR
which showed the n
to 3000 cm
vibration as a broad band from 3650
O–H
ꢁ
1
The formation of carbon in the destructive adsorption of
ꢀ
CCl
4
at 350 C is proved by XPS. When CCl
4
was adsorbed
on the partially chlorinated TiO surface, the relatively nega-
2
tive chlorine atoms interact with positive metal ion sites, but
the relatively positive carbon atom can not interact with a
negative oxygen site, thus leading to an unstable intermediate.
The interaction of chlorine with the metal weakens the cova-
lent C–Cl in this intermediate. The overall result of this process
leads to a dissociative adsorption of CCl
further chlorinated TiO surface as a result of chlorine abstrac-
tion from the CCl molecules. Carbon formed could react with
oxygen gas at relatively high temperature in the carrier gas to
produce CO and CO
4
to form carbon and a
4
2
.
The partially chlorinated TiO could still exchange chlorine
2
atoms with oxygen atoms of oxygen in the carrier gas. When
oxygen was adsorbed on the partially chlorinated TiO
2
ꢀ
heated to certain temperature, such as 550 C, the interaction
of the relatively negative oxygen with the relatively positive
Ti led to disconnection of chlorine with the Ti center. As a
result, TiO was regenerated, and Cl was formed due to the
2
2
recombination of chlorine cleaved from the partially chlori-
nated TiO , in agreement with our previous work. The
3
7
2
2 2
regeneration of TiO makes us believe that nanometer TiO
could be potentially applied to deal with chlorinated
hydrocarbons.
Scheme 1 Scheme of the destructive adsorption of CCl
nanometer TiO in the presence of oxygen.
2
4
over
9
90
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4

View Article Online
The further chlorinated TiO surface could be fully chlori-
nated with CCl to form TiCl . Since the boiling point of TiCl
4
4
5
6
J. J. Kaufman, W. S. Koski, S. Roszak and K. Balasubramanian,
Chem. Phys., 1996, 204, 233.
A. H. Gilani, K. H. Janbaz and B. H. Shah, Pharmacol. Res.,
4
4
was lower than the temperature the reactor was heated to, the
formation of TiCl resulted in the weight loss of the reactor,
and oxide surface regeneration for further adsorption. The
smaller sized nanometer TiO sample was more reactive than
1
998, 37, 31.
J. V. Michael and S. S. Kumaran, Combust. Sci. Technol., 1998,
34, 31.
4
1
2
7
8
J. Stark and J. Rabani, J. Phys. Chem. B, 1999, 103, 8524.
S. L. Huang and L. D. Pfefferle, Environ. Sci. Technol., 1989,
the larger sized sample, possibly because first, the anatase/
rutile ratio was different for those two samples that can
account for variation in activities; second, the difference in
apparent densities of powders resulted in variation of contact
2
3, 1085.
9
0
1
2
E. A. Chandross and R. D. Miller, Chem. Rev., 1999, 99, 1641.
C. Suryanarayana, Int. Mater. Rev., 1995, 40, 41.
A. Henglein, Chem. Rev., 1989, 89, 1861.
E. Lucas, S. Decker, A. Khaleel, A. Seitz, S. Fultz, A. Ponce,
W. F. Li, C. Carnes and K. J. Klabunde, Chem. Eur. J., 2001,
7, 2505.
1
1
1
times and should influence of the degree of CCl
further, the concentration of surface OH groups should be dif-
ferent for Ti-650 and Ti-800. The partially chlorinated TiO
derived from 40 nm TiO also could be more reactive so that
the exchange of chlorine atoms with oxygen atoms of oxygen
gas in the carrier gas was faster. Therefore, the 40 nm TiO
was regenerated more rapidly than the 80 nm TiO . Hence,
4
conversion;
2
1
3
4
S. Krishnamoorthy, J. A. Rivas and M. D. Amiridis, J. Catal.,
2
2
000, 193, 264.
S. C. Petrosius, R. S. Drago, V. Young and G. C. Grunewald,
1
2
J. Am. Chem. Soc., 1993, 115, 6131.
2
15 E. J. Shin and M. A. Keane, Appl. Catal. B, 1998, 18, 241.
16 G. Sinquin, C. Petit, J. P. Hindermann and A. Kiennemann,
Catal. Today, 2001, 70, 183.
2
the weight loss of the 40 nm TiO was lower than the 80 nm
TiO . Similarly, the partially chlorinated TiO derived from
2
2
1
1
1
7
8
9
R. Schneider, D. Kiessling and G. Wendt, Appl. Catal. B, 2000,
8, 187.
Y. X. Li, H. Li and K. J. Klabunde, Environ. Sci. Technol., 1994,
8, 1248.
the 40 nm TiO
2
at higher temperature could be more reactive
2
so that the exchange of chlorine atoms with oxygen atoms of
oxygen gas in the carrier gas was faster. TiO was regenerated
more rapidly at higher temperature. Therefore, the weight loss
of nanometer TiO
lower temperature.
In conclusion, nanometer TiO
method could have a remarkable potential to remove chlori-
nated hydrocarbons in the environment. The smaller 40 nm
TiO2 sample adsorbed and decomposed more carbon tetra-
2
2
P. D. Hooker and K. J. Klabunde, Environ. Sci. Technol., 1994,
28, 1243.
20 B. M. Weckhuysen, G. Mestl, M. P. Rosynek, T. R. Krawietz,
2
was less at higher temperature than at
J. F. Haw and J. H. Lunsford, J. Phys. Chem. B, 1998, 102, 3773.
B. M. Weckhuysen, M. P. Rosynek and J. H. Lunsford, Phys.
Chem. Chem. Phys., 1999, 1, 3157.
P. Van der Avert and B. M. Weckhuysen, Angew. Chem., Int. Ed.,
2002, 41, 4730.
2
synthesized by the sol–gel
2
1
2
2
chloride than the larger 80 nm TiO
of TiCl made the interaction surface renew so that CCl could
2
sample. The formation
23 B. M. Weckhuysen, Phys. Chem. Chem. Phys., 2003, 5, 4531.
24 O. B. Koper, E. A. Wovchko, J. A. Glass, J. T. Yates and K. J.
Klabunde, Langmuir, 1995, 11, 2054.
4
4
further interact with the bulk of the particles. TiO was regen-
erated through the exchange of chlorine with oxygen in the
carrier gas so that large quantities of CCl would be decom-
2
2
5
K. J. Klabunde, J. Stark, O. Koper, C. Mohs, D. G. Park, S.
Decker, Y. Jiang, I. Lagadic and D. J. Zhang, J. Phys. Chem.,
4
1
996, 100, 12 142.
posed over nanometer TiO2 in the presence of oxygen. A
mechanism has been proposed for the destructive adsorption
26
O. Koper, I. Lagadic and K. J. Klabunde, Chem. Mater., 1997,
9, 838.
and desorption of CCl
4
over nanometer TiO
2
.
27 Y. Jiang, S. Decker, C. Mohs and K. J. Klabunde, J. Catal., 1998,
80, 24.
1
2
2
3
3
8
9
0
1
S. P. Decker, J. S. Klabunde, A. Khaleel and K. J. Klabunde,
Environ. Sci. Technol., 2002, 36, 762.
S. Decker, I. Lagadic, K. J. Klabunde, J. Moscovici and A.
Michalowicz, Chem. Mater., 1998, 10, 674.
Y. C. Chien, H. P. Wang and Y. W. Yang, Environ. Sci. Technol.,
Acknowledgements
We gratefully acknowledged financial support of the work by
National Natural Science Foundation of China (No:
2
001, 35, 3259.
C. L. Carnes, J. Stipp and K. J. Klabunde, Langmuir, 2002, 18,
352.
32 A. Khaleel and B. Dellinger, Environ. Sci. Technol., 2002, 36,
620.
20075014), National Science and Technology Committee of
China (No: GN-99-4), and the Doctoral Research Foundation
of Chinese Education Ministry (No. 2000000303).
1
1
3
3
Standard Methods for the Examination of Water and Waste-
water, American Public Health Association, Washington, DC,
References
1
989, pp. 4–51.
1
2
R. M. Lago, M. L. H. Green, S. C. Tsang and M. Odlyha, Appl.
Catal. B, 1996, 8, 107.
34 Beijing Chemial Reagent Corporation, Catalog of Chemical
Reagents, 1999.
35 J. F. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben,
Handbook of X-ray Photoelectron Spectroscopy, Perkin-Elmer
Corp., Eden Prairie, MN, 1992.
36 P. Babelon, A. S. Dequiedt, H. Mostefa-Sba, S. Bourgeois, P.
Sibillot and M. Sacilotti, Thin Solid Films, 1998, 322, 63.
37 G. H. Liu, Y. F. Zhu, X. R. Zhang and B. Q. Xu, Anal. Chem.,
2002, 74, 6279.
R. G. Prinn, R. F. Weiss, P. J. Fraser, P. G. Simmonds, D. M.
Cunnold, F. N. Alyea, S. O’Doherty, P. Salameh, B. R. Miller,
J. Huang, R. H. J. Wang, D. E. Hartley, C. Harth, L. P. Steele,
G. Sturrock, P. M. Midgley and A. McCulloch, J. Geophys.
Res., 2000, 105, 17 751.
3
S. N. Pentyal, P. J. S. Vig, B. S. Sekhon and D. Desaiah, Cell.
Signall., 1994, 6, 561.
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 9 8 5 – 9 9 1
991