Regularities of decomposition of organic vapors using a photocatalytic air cleaner

Article abstract of DOI:10.1007/s11172-006-0050-4

The kinetics of oxidation of vapors of model air contaminators, viz., acetone, ethanol, and heptane, was studied using a photocatalytic air cleaner. The composition of the oxidation products was determined, and the rates of oxidation of the starting substances were measured. The deep oxidation of the starting substrates to CO₂ and H₂O occurs until their concentration achieves a limiting value. At higher concentrations a "breakthrough" of the starting substrate is observed. Ethanol is oxidized with the formation of intermediate products. The experimental data obtained were approximated by a kinetic model, which includes stages of formation of intermediates and their competitive adsorption. The results of the approximation agree well with the experimental data.

Full text of DOI:10.1007/s11172-006-0050-4

1866
Russian Chemical Bulletin, International Edition, Vol. 54, No. 8, pp. 1866—1873, August, 2005
Regularities of decomposition of organic vapors using
a photocatalytic air cleaner
I. A. Baturov,a,b A. V. Vorontsov,b and D. V. Kozlovb
ꢀ
aNovosibirsk State University,
2 ul. Pirogova, 630090 Novosibirsk, Russian Federation
bG. K. Boreskov Institute of Catalysis, Siberian Branch of the Russian Academy of Sciences,
5 prosp. Akad. Lavrent´eva, 630090 Novosibirsk, Russian Federation.
Fax: +7 (383 2) 33 1617. Eꢀmail: kdv@catalysis.ru
The kinetics of oxidation of vapors of model air contaminators, viz., acetone, ethanol, and
heptane, was studied using a photocatalytic air cleaner. The composition of the oxidation
products was determined, and the rates of oxidation of the starting substances were measured.
The deep oxidation of the starting substrates to CO2 and H2O occurs until their concentration
achieves a limiting value. At higher concentrations a "breakthrough" of the starting substrate is
observed. Ethanol is oxidized with the formation of intermediate products. The experimental
data obtained were approximated by a kinetic model, which includes stages of formation of
intermediates and their competitive adsorption. The results of the approximation agree well
with the experimental data.
Key words: photocatalysis, air purification, titanium dioxide, mineralization.
Photocatalytic methods for air purification have conꢀ
Experimental
siderable advantages over other methods. Unlike air puriꢀ
fication by adsorption, photocatalytic oxidation leads to
complete neutralization of air pollutants forming CO2
and H2O as final products, and no regular regeneration of
the photocatalyst is usually required. Almost any organic
substance even at low concentration can be oxidized unꢀ
der ambient conditions using air oxygen.
The photocatalytic air cleaner (PAC) (Fig. 1) was placed in
a sealed chamber with the volume 6 L. The chamber was inꢀ
cluded into a flow circuit (Fig. 2). Gas samplers used for chroꢀ
matographic analysis of the composition of the initial and final
reaction mixtures (IRM and FRM, respectively) were placed at
the inlet and outlet of the chamber.
5
The photocatalyst used in the air cleaner was Pt/TiO2
Photons with the energy exceeding the TiO2 band gap
(λ < 380 nm) can induce the transition of electrons from
the valence to the conduction band of a photocatalyst
resulting in the formation of electronꢀhole pairs. The latꢀ
ter are initiators of further redox transformations involvꢀ
ing adsorbed molecules of organic substances on the phoꢀ
tocatalyst surface.1,2 As a result of a series of these transꢀ
formations, almost all organic substances could comꢀ
pletely be oxidized to CO2 and H2O.
(0.2% Pt), which was uniformly deposited on a porous polyꢀ
4
3
5
The kinetics of photooxidation of organic substances
was studied in reactors of different types: batch, flowꢀ
circular, fluidizedꢀbed, and coil reactors.3,4 All reactors
described are small devices for laboratory use.
2
The purpose of the present work is to study the operaꢀ
tion of the industrial photocatalytic air cleaner (PAC)
designed for use in accomodation space and automobiles.
Acetone, ethanol, and heptane were chosen as test subꢀ
stances. These organic compounds are components of
solvents and can be found in the air of rooms and autoꢀ
mobile cabins.
6
1
Fig. 1. AerolaifꢀBN photocatalytic air cleaner produced at the
Luch NPO (Novosibirsk, Russia): 1, UV lamp; 2, contaminated
air; 3, ventilator; 4, purified air; 5, case; and 6, filter with photoꢀ
catalyst.
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 8, pp. 1812—1818, August, 2005.
1066ꢀ5285/05/5408ꢀ1866 © 2005 Springer Science+Business Media, Inc.

Air purification with photocatalytic cleaner
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
1867
A (%)
80
60
40
20
Fig. 2. Scheme of the flowꢀtype system used for experiment:
1, air flow purified from СО2, moisture content 50% at 22 °С;
2, syringeꢀtype pump; 3, samplers for IRM and FRM; 4, chamꢀ
ber with PAC; and 5, foam flowmeter.
300
350
400
λ/nm
Fig. 3. Emission spectrum of a TL´E 22W/10 lamp (Philips)
used in the AerolaifꢀBN PAC (data from the manufacturer).
meric support (catalyst amount ∼3 g). The previous studies
showed6 that dopping TiO2 with platinum enhances a total rate
of the photocatalytic reaction and redistributed intermediate
reaction products. When ethanol vapors are photooxidized on
the Pt/TiO2 photocatalyst, carbon monoxide is not formed in
the gas phase, although for the use of pure TiO2 noticeable
amounts of CO are observed among the reaction products.6
To prepare the IRM, the starting substrate was injected with
a syringeꢀtype pump into an air flow with the 50% relative moisꢀ
ture content at room temperature. The device is a platform with
a mounted microsyringe. A stop, which pushes the microsyringe
plunger, moves slowly along the platform. A volatile component
is introduced into the air flow where it evaporates rapidly. Deꢀ
pending on the velocity of syringe pushing, different concentraꢀ
tions of reactants can be created: from zero to those correspondꢀ
ing to the saturated vapor. The concentrations of the reactants
in air in a range of 0—10 000 ppm were used. The temperature
inside the chamber (320 K) exceeds the ambient temperature by
20 K due to the heat evolved by the working PAC.
The switchedꢀon PAC stirred air inside the chamber with a
velocity of 10 m3 h–1 at flow rates of the IRM and FRM of
0.012 m3 h–1. The ratio of volume rates of stirring inside the
reactor and passing of a mixture through the reactor was 830,
which enabled the chamber with the PAC inside to be considꢀ
ered as an ideal stirred reactor. This assumption was used in the
mathematical description of the processes under study.
After a new reactant concentration in the IRM was speciꢀ
fied, a new stationary state was established within 6—8 h, which
approximately tenfold greater than the contact time. The conꢀ
centrations of the reactants and products in the FRM changed
with establishing the stationary state. The concentrations were
measured at an interval of 1 h. For analysis of the results, the
stationary concentrations of the reactants and products in the
FRM were used.
It was assumed in the calculations of the quantum efficiency
of photocatalytic processes that the UV radiation of the lamp
corresponded to the wavelength 365 nm (Fig. 3) and the fraction
of UV radiation was 25% of the electric power of the lamp
(22 W), which is typical of the most part of class UVꢀA commerꢀ
cial lowꢀpressure Hg lamps.
Results and Discussion
Oxidation of acetone. Experimental data. On passing
through the reactor, the acetone vapors undergo deep
oxidation according to the following equation
Me2CO + 4 O2
3 CO2 + 3 H2O.
The kinetic curves of СО2 and acetone vapors at the
outlet of the chamber vs. acetone concentration in the
IRM are shown in Fig. 4. No formation of gaseous interꢀ
mediate products of acetone oxidation was observed in a
C, ppm
6000
5000
4000
1
3000
2000
1000
2
The oxidation rate was calculated by the formula
Wox = (CIRM — CFRM)•10–6•V/(RT ) (mol min–1),
where CIRM and CFRM are the substrate concentrations in the
IRM and FRM, respectively (ppm); V is the gas flow rate
(L min–1); R = 0.082 L atm mol–1 K–1 is the universal gas conꢀ
stant; and T is temperature (K).
Gas samples were analyzed on an LKhMꢀ8 chromatograph
equipped with a flameꢀionization detector and a ParopakꢀT colꢀ
umn (L = 1 m, d = 3 mm). When studying the CO2 content, a
sample was separated on a carbon column and then passed
through a methanation reactor and analyzed using a flameꢀ
ionization detector.
2000
4000
6000
Cac, ppm
Fig. 4. Change in the concentration of СО2 and acetone vapors
(C) at the outlet of the chamber with the photocatalytic air
cleaner at different concentrations of acetone vapors (Cac) in the
IRM. Components of the FRM: 1, СО2; and 2, acetone. Solid
lines are the approximating curves by the simulation results, and
dotted lines are the asymptotic straight lines to which the experiꢀ
mental data approach with an increase in the acetone concenꢀ
tration in the IRM.

1868
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
Baturov et al.
wide concentration range, which agrees with our earlier
results5 and data of other authors.7
catalyst surface (N0/V )•(dθA/dt). This change is caused
by the introduction and removal of the reactant, as well as
its conversion. In other words, the following expression is
valid:
The concentrations of acetone (Cac) and СО2 in the
FRM increase gradually with an increase in the acetone
concentration in the IRM from 0 to 2000 ppm. For the
further increase in Cac in the IRM, the СО2 concentration
in the FRM remains unchanged, being ~6000 ppm. This
implies that the maximum efficiency of the PAC is
achieved when Cac in the IRM is 2000 ppm. If Cac in the
IRM continues to increase, Cac in the FRM increases by
an equivalent value; the rate of complete acetone oxidaꢀ
tion remains at the former level. At any acetone concenꢀ
tration in the IRM, the carbon content in the IRM and
FRM is the same. This implies that the reactor operates,
in fact, under the steadyꢀstate conditions and no intermeꢀ
diate oxidation products are accumulated on the photoꢀ
catalyst surface.
,
(1)
where θA = (KaCA)/(1 + KaCA).
Under the steadyꢀstate conditions, the left part of
Eq. (1) is equal to zero, which enables one to derive the
following dependence of the concentration of reactant A
in the FRM on CA0 in the IRM:
CA(CA0) = (M + N)/(2Ka),
(2)
,
The limiting rate of acetone oxidation, which
was achieved in experiment (see Fig. 4), is 1.8•10–5
mol min–1. This allows one to estimate the quantum effiꢀ
ciency of acetone oxidation assuming that 16 photons
are needed for the complete oxidation of one acetone
molecule.8 The resulting quantum efficiency is ϕ =
16•(Wox/Wq)•100% = 34% (Wq is the quantum flux from
the lamp, mol min–1).
.
At high CA0 the plot CA(CA0) becomes linear
0
CA(CA0) = CA – N0k/(2u).
This equation is described by a straight line to which
the plot of the acetone concentration in the FRM vs. its
starting concentration in the IRM (see Fig. 4, dotted
lines) tends asymptotically. The N0k/(2u) value correꢀ
sponds to the maximum acetone oxidation rate in the
reactor.
Using Eq. (2), we performed the numerical approxiꢀ
mation of the experimental data (see Fig. 4, solid lines).
The СО2 concentrations were approximated by the funcꢀ
tion CCO = 3(Cac – Cac). The calculated adsorption
constant of acetone (Ka = 0.06 0.02 (ppm)–1) differs
from the value obtained for other TiO2 samples (Ka
0.0026 (ppm)–1) by an order of magnitude.5 Perhaps, the
adsorption constants can substantially change under
photoirradiation and steadyꢀstate reaction conditions.
The typical number of active sites N0 on the surface of
the TiO2 photocatalyst is known11: 5•1014 cm–2. Knowꢀ
ing this value, one can estimate the apparent rate constant
of acetone oxidation assuming that the whole photocataꢀ
Simulation of the kinetics of acetone oxidation. To deꢀ
scribe the kinetics, we considered the twoꢀstep model of
acetone oxidation
A + Z
AZ
AZ,
P + Z,
0
2
where A is an acetone molecule, Z is an adsorption site,
AZ is an acetone molecule adsorbed on the surface, P are
the oxidation products (carbon dioxide and water), Ka is
the adsorption equilibrium constant of acetone, and k is
the rate constant of acetone oxidation. At least two known
facts are described in this scheme.
1. At low concentrations the adsorption isotherm of
acetone is described by the Langmuir model.5,9
2. No gaseous intermediate products are formed in
substantial concentrations upon the photocatalytic oxiꢀ
dation of acetone.5,10,11
=
lyst is irradiated: k = (2.8 1)•10–7 s–1
.
Photocatalytic oxidation of heptane vapors. Experimenꢀ
tal results. The complete photocatalytic oxidation of the
heptane vapors corresponds to the following empirical
equation:
Let us assume that the reaction occurs in an ideal
stirred reactor (chamber) with the volume V (L) at the
flow rate of the reaction mixture u (L s–1) and content of
catalyst active sites N0 (mol). The concentration of subꢀ
0
stance A at the inlet is equal to CA (mol L–1). Since the
C7H16 + 11 O2
7 CO2 + 8 H2O.
concentrations are rather low, both flow rates at the inlet
and outlet of the reactor are uqual to u. Let θA be the
fraction of sites on the surface occupied by acetone. The
total change in the concentration of reactant A in the
chamber involves a change in the concentration of A both
in the gas phase dCA/dt and in the layer adsorbed on the
In our experiments the only oxidation products deꢀ
tected were СО2 and water. The plots of the concentraꢀ
tions of heptane and СО2 at the outlet of the chamber vs.
heptane concentration at the inlet of the chamber are
shown in Fig. 5.

Air purification with photocatalytic cleaner
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
1869
C, ppm
tion of the kinetics of heptane oxidation. In this case, A is
a heptane molecule, AZ is a heptane molecule adsorbed
on the surface, Ka is the adsorption constant of heptane,
and k is the firstꢀorder rate constant of heptane oxidation
on the photocatalyst surface. It can be assumed that the
dependence of the heptane concentration at the outlet of
the chamber on the concentration at the inlet of the chamꢀ
ber is also described by Eq. (2).
10000
8000
6000
4000
2000
0
The numerical approximation of the experimental data
on heptane oxidation was performed using Eq. (2) (see
Fig. 5, solid lines). The СО2 concentration was approxiꢀ
1
2
0
mated by the function CCO2 = 7(Chep – Chep), where
Chep is the heptane concentration. The calculated
values of the adsorption constant of heptane is Ka
=
0
2000
4000
6000
Chep, ppm
(4.1 1.5)•10–4 (ppm)–1. Assuming that the whole cataꢀ
lyst is irradiated, the apparent rate constant (k) of photoꢀ
Fig. 5. Change in the concentration of СО2 and heptane vapors
(C) at the outlet of the chamber with the photocatalytic air
cleaner at different concentrations of heptane vapors (Chep) in
the IRM. Components of the FRM: 1, heptane; and 2, СО2.
Solid lines are the approximating curves by the simulation reꢀ
sults, and dotted lines are the asymptotic straight lines to which
the experimental data approach with an increase in the acetone
concentration in the IRM.
catalytic heptane oxidation is (1.1 0.37)•10–6 s–1
.
Photocatalytic oxidation of ethanol vapors. Experimenꢀ
tal data. Unlike acetone and heptane, ethanol vapors proꢀ
duces intermediate oxidation products of the photocataꢀ
lytic oxidation, which were detected in the gas phase of
the reactor: acetaldehyde, acetic acid, and ethyl acetate,
as well as the deep oxidation products (CO2 and H2O).
Based on these results, the scheme of ethanol oxidation
was proposed (Scheme 1). Each step designated by arrow
can consist of several successive elementary steps.
The heptane concentration in the FRM increases
gradually as the heptane concentration in the IRM inꢀ
creases from ~1000 to 4000 ppm. The СО2 concentration
also increases gradually and reaches a constant level of
~10000 ppm. This means that the maximum productivity
of the PAC is achieved at a heptane concentration in the
IRM of 4000 ppm and more. The further increase in the
heptane content in the IRM results in an increase in the
heptane concentration in the FRM by an equivalent value,
while the rate of heptane oxidation remains at the former
level, which is similar to the situation for acetone oxiꢀ
dation.
Scheme 1
As in the case of acetone oxidation, the carbon conꢀ
tent in the IRM and FRM is the same for any heptane
concentration in the IRM. This indicates that the reactor
operates under the steadyꢀstate conditions and no interꢀ
mediate oxidation products are accumulated on the phoꢀ
tocatalyst surface.
Similar intermediates of the photocatalytic oxidation
of ethanol were described previously.12—14
The limiting rate of deep heptane oxidation achieved
in experiments is 1.3•10–5 mol min–1. This makes it posꢀ
sible to estimate the quantum efficiency of the process
using assumptions on the photon flux, which are similar
to those for acetone. In addition, when estimating the
quantum efficiency, we suppose that 44 photons are
needed for the complete oxidation of one heptane molꢀ
ecule. This number is equal to the change in the sum of
oxidation states of carbon atoms in the empirical equaꢀ
tion. The resulting quantum efficiency of photocatalytic
heptane oxidation is ϕ = 44•(Wox/Wq)•100% = 67%.
Simulation of heptane oxidation. The twoꢀstep model
of heptane oxidation, which is completely equivalent to
the scheme of acetone oxidation, was used for the simulaꢀ
We identified ethyl acetate as an intermediate oxidaꢀ
tion product only under the steadyꢀstate reaction condiꢀ
tions. No formation of ethyl acetate was observed for the
photocatalytic oxidation of ethanol under the static conꢀ
ditions. Probably, under the steadyꢀstate conditions, the
acetic acid that formed interacts with excess ethanol to
form ethyl acetate, which is displaced from the photoꢀ
catalyst surface by other adsorbed compounds.
The plots of the stationary concentrations of ethanol,
acetaldehyde, acetic acid, ethyl acetate, and carbon dioxꢀ
ide in the FRM of the chamber vs. ethanol concentration
in the IRM are shown in Fig. 6. It can be seen that only
CO2 is present in the FRM when the concentration of
ethanol in the IRM does not exceed 1000 ppm. In the

1870
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
Baturov et al.
C, ppm
Simulation of the kinetics of ethanol oxidation in an
ideal stirred reactor. To reveal all regularities of the reacꢀ
tion, we studied the twoꢀstep model of photooxidation of
the ethanol vapors with competitive adsorption of the
intermediate products and reactants in an ideal stirred
reactor (Scheme 2).
1000
Scheme 2
1
2
A + Z
AZ
AZ
BZ
500
BZ
B + Z
P + Z
3
BZ
2000
4000
6000 [EtOH], ppm
A is the oxidized substrate; B and P are the intermediate and final
oxidation products, respectively;
C, ppm
K
A and KB are the adsorption constants of substances A and B,
respectively, on the photocatalyst surface (L mol–1);
k
1 and k2 are the heterogeneous rate constants of the second and
fourth steps, respectively (s–1).
4000
3000
2000
1000
The model is based on an assumption that adsorption
sites on the photocatalyst surface are simultaneously the
reaction centers.
The proposed scheme does not describe the mechaꢀ
nism of ethanol photooxidation in detail, because it does
not reflect all intermediates observed. Since the acetaldeꢀ
hyde concentration exceeds considerably the concentraꢀ
tions of two other intermediates, this kinetic scheme qualiꢀ
tatively corresponds to the main regularities of the ethaꢀ
nol photooxidation.
4
5
Let the reaction to occur in an ideal stirred reactor
with the volume V (L) at the flow rate of the mixꢀ
ture u (L s–1) involving N0 moles of active sites of the
catalyst and the concentration of substance A at the inlet
2000
4000
6000 [EtOH], ppm
Fig. 6. Concentrations of СО2 (1), AcОН (2), AcОEt (3),
EtОН (4), and MeСНО (5) at the outlet of the ideal stirred
photocatalytic reactor (C) vs. concentration of ethanol in
the IRM.
0
is CA (mol L–1). Let us accept that the initial reaction
mixture is strongly dilute (in our case, with air) and, hence,
the flow rates of the mixture at the outlet and inlet of the
reactor are equal to u.
At any moment, the concentrations of all participants
of the reaction obey the following system of equations:
case of full ethanol oxidation, the theoretical concentraꢀ
tion of CO2 almost coincided with the measured value,
which indicates the complete conversion of ethanol to
СО2 and H2O without accumulation of intermediate
products.
With the further increase in the ethanol concentration
at the inlet of the reactor, ethanol, acetic acid, acetaldeꢀ
hyde, and ethyl acetate were observed at the outlet of the
reactor, and the CO2 concentration decreased. This imꢀ
plies the competitive adsorption of ethanol and products
of its oxidation. The latter are displaced from the surface
without further oxidation and, thus, the rate of ethanol
conversion to the deep oxidation products (СО2) deꢀ
creases.
,
(3)
(4)
,
θA = (KACA)/(1 + KACA + KBCB),
B = (KBCB)/(1 + KACA + KBCB),
(5)
(6)
θ
where θA and θB are the fractions of sites on the photoꢀ
catalyst surface occupied by molecules A and B, respecꢀ
tively.

Air purification with photocatalytic cleaner
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
1871
The left parts of Eqs (3) and (4) correspond to the
overall change in the amount of oxidized substrate A and
intermediate B in the reactor caused by chemical transꢀ
formations on the surface and introduction and removal
of the substance from the reactor by the flow. Equations (5)
and (6) describe competitive adsorption of the starting
substrate A and intermediate B. When the reaction proꢀ
ceeds under the steadyꢀstate conditions, the left parts of
Eqs (3) and (4) are equal to zero. Substituting Eqs (5)
and (6) into Eqs (3) and (4), we have the following system
of nonlinear equations:
The concentration of substance B can be expressed
through the concentration of substance A as follows:
.
(12)
a
[A], ppm
8000
1
6000
4000
2000
2
3
,
(7)
k1´
.
(8)
Let us introduce designations k1´ = N0k1/u and
k2´ = N0k2/u and rewrite the equations in the final form
tgα = 1
4000
2000
6000
8000 [A]0, ppm
C, ppm
,
(9)
b
2000
1500
1000
500
B3
B2
.
(10)
B1
In essence, Eqs (9) and (10) are the material balance
of the system with respect to the starting substrate A and
intermediate B.
It seems difficult to solve the system of Eqs (9) and (10)
in the general form. Therefore, we restrict our considerꢀ
ation by a particular case when B is adsorbed more weakly
than A (KACA >> KBCB). In this case, the term KBCB in
the denominators of Eqs (9) and (10) can be neglected.
Then two ultimate cases follow from Eq. (9)
P1
P2
P3
2000
CA
0 and CA
CA0 – k1´. In practice,
4000
6000
8000 [A]0, ppm
0
this means that at low concentrations CA substance A is
completely transformed into B and P. At very high conꢀ
centrations when the whole surface is occupied by molꢀ
ecules A, the rate of decomposition of substance A beꢀ
comes constant.
The general expression is obtained by the solution of
the quadratic equation (9) with respect to the concentraꢀ
tion CA
Fig. 7. a. Concentration of component A at the outlet of the
reactor at different initial concentrations CA calculated by
0
Eq. (11). Three different curves were plotted at the same conꢀ
version rate constant k1´ = 2000 (ppm)–1 but different adsorpꢀ
tion constants (KA) of the oxidized substrate. The region marked
by rectangle in the main plot is shown in more detail in insert:
KA = 0.001 (1), 0.01 (2), and 0.1 (ppm)–1 (3). b. Concentrations
of intermediate B and product P (C) at the outlet of the reactor
at different initial concentrations of the substrate CA0 calculated
by Eq. (12). When plotting the curves, we used the constant
values k1´ = 2000 (ppm)–1 and КА = 0.1 (ppm)–1 corresponding
to the curve in Fig. 7, a; k2´•KB = 100 (P1—B1), 10 (P2—B2),
and 1 (P3—B3).
.
(11)

1872
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
Baturov et al.
It can be seen that the product k2´KB completely deꢀ
termines a change in the concentration of substance B at
the outlet of the reactor. When this value is small, for
example, due to a low rate of transformation of B into P,
substance B leaves the reactor virtually without further
transformation, and its concentration is close to the limꢀ
stant of photocatalytic oxidation k1 and k2 are 5•10–6 and
1.7•10–7 s–1, respectively.
The proposed kinetic scheme qualitatively describes
the experimentally observed change in the concentration
of the ethanol vapors. In addition, the obtained adsorpꢀ
tion rate constants for ethanol agree well with pubꢀ
lished data.12
0
iting value CB = CA – CA = k1´. In the case when the
value of the product k2´KB is high, substance B can be
transformed into P almost completely at the initial step.
Then the concentration of B at the outlet is close to zero.
This model also predicts that the concentration of the
final product P passes through a maximum and the conꢀ
version tends to zero in the limit, at a high concentration
of the starting substance.
*
*
*
Thus, the following conclusions can be drawn.
1. Reactions of photocatalytic oxidation of acetone
and heptane vapors in a flowꢀtype photocatalytic ideal
stirred reactor proceed completely to form CO2 and H2O
without formation and desorption of intermediate oxidaꢀ
tion products to the gas phase. The quantum efficiencies
of acetone and heptane oxidation in the reactor are apꢀ
proximately 34 and 67%, respectively.
2. The twoꢀstep kinetic model of photocatalytic oxiꢀ
dation of acetone and heptane, including the step of reꢀ
versible substrate adsorption and irreversible photocataꢀ
lytic oxidation to the final products, well describes the
experimental data. The values of adsorption constants
obtained by the nonlinear approximation of the experiꢀ
mental data are close to published values.
3. Intermediate oxidation products, viz., acetaldehyde,
acetic acid, and ethyl acetate, were found to evolve to
the gas phase during the photocatalytic oxidation of the
ethanol vapors. For a small ethanol concentration
(<1000 ppm) at the inlet of the reactor, no evolution of
intermediates is observed.
4. The fourꢀstep kinetic scheme for the photocatalytic
oxidation of the ethanol vapors, including the steps of
formation of the intermediate and final reaction products
and the reversible competitive adsorption of the substrate
and intermediate, was proposed. The results obtained usꢀ
ing this model agree qualitatively with the experimenꢀ
tal data.
All described regularities, which were obtained by
Eqs (11) and (12), are plotted in Fig. 7.
The experimentally measured dependence of the СО2
concentration in the FRM on the ethanol concentration
in the IRM (see Fig. 6, a) shows that the conversion of
alcohol to the final product СО2 decreases indeed with an
increase in the concentration of the starting substrate in
the IRM.
The results of the nonlinear approximation of the exꢀ
perimental data by the theoretical curves calculated from
Eqs. (11) and (12) for ethanol and acetaldehyde are shown
in Fig. 8.
Under an assumption that the whole photocatalyst
surface is irradiated in the reactor, the adsorption conꢀ
stants of the reactant (A) and intermediate product (B)
are 0.5 and 0.32 (ppm)–1, respectively, and the rate conꢀ
C, ppm
1
2
4000
3000
2000
1000
5. It is shown that at high concentrations of ethanol
(>1000 ppm) the products of incomplete oxidation are
displaced from the TiO2 surface due to competitive adꢀ
sorption.
This work was financially supported by the Russian
Science Support Foundation, the Finnish Academy of
Sciences (Program "Russia in Flux," Grant 208134), and
the CRDF (Grant NOꢀ008ꢀX1).
References
2000
4000
6000
[EtOH], ppm
Fig. 8. Concentrations of ethanol (1) and acetaldehyde (2) at
the outlet of the ideal stirred photocatalytic reactor (C) vs. conꢀ
centration of ethanol in the initial reaction mixture approxiꢀ
mated by the theoretical curve.
1. M. A. Fox and M. T. Dulay, Chem. Rev., 1993, 93, 341.
2. M. R. Hoffman, S. T. Martin, W. Choi, and D. W.
Bahnemann, Chem. Rev., 1995, 95, 96.

Air purification with photocatalytic cleaner
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 8, August, 2005
1873
3. Photocatalytic Purification and Treatment of Water and Air,
Eds D. F. Ollis and H. AlꢀEkabi, Elsevier, Amsterdam,
1993, p. 481.
4. A. V. Vorontsov, D. V. Kozlov, P. G. Smirniotis, and V. N.
Parmon, Kinet. Katal., 2005, 46, 466 [Kinet. Catal., 2005,
No. 3 (Engl. Transl.)].
5. A. V. Vorontsov, I. V. Stoyanova, D. V. Kozlov, V. I.
Simagina, and E. N. Savinov, J. Catal., 2000, 189, 360.
6. A. V. Vorontsov and V. P. Dubovitskaia, J. Catal., 2004,
221, 102.
7. M. L. Sauer and D. F. Ollis, J. Catal., 1994, 149, 81.
8. D. V. Kozlov, A. A. Panchenko, D. V. Bavykin, E. N.
Savinov, and P. G. Smirniotis, Izv. Akad. Nauk, Ser. Khim.,
2003, 1041 [Russ. Chem. Bull., Int. Ed., 2003, 52, 1100].
9. J. M. Coronado, M. E. Zorn, I. TejedorꢀTejedor, and M. A.
Anderson, Appl. Catal. В: Env., 2003, 43, 333.
10. A. V. Vorontsov, G. B. Barannik, O. I. Snegurenko, E. N.
Savinov, and V. N. Parmon, Kinet. Katal., 1997, 38, 97
[Kinet. Catal., 1997, 38 (Engl. Transl.)].
11. M. L. Sauer and D. F. Ollis, J. Catal., 1996, 163, 215.
12. M. R. Nimlos, E. J. Wolfrum, M. L. Brewer, J. A. Fennell,
and G. Bintner, Environ. Sci. Technol., 1996, 30, 3102.
13. M. L. Sauer and D. F. Ollis, J. Catal., 1996, 158, 570.
14. D. S. Muggli, J. T. McCue, and J. L. Falconer, J. Catal.,
1998, 173, 470.
Received April 27, 2004