Adsorption Performance of Titanium Dioxide (TiO2) Coated
Air Filters for Volatile Organic Compounds
Lexuan Zhong, Chang-Seo Lee, Fariborz Haghighat*
Department of Building, Civil and Environmental Engineering
Concordia University, Montreal, Quebec H3G 1M8
Abstract
The photocatalytic oxidation (PCO) technology as an alternative method for air
purification has been studied for decades and a variety of PCO models indicate that the
adsorption of reactants on the catalyst surface is one of the major physical and chemical
processes occurring at a heterogeneous photocatalytic reaction. However, limited study
explored the adsorption effect of a photocatalyst. This study carried out a systematic
evaluation of adsorption performance of titanium dioxide (TiO2) coated fiberglass fibers
(FGFs), TiO2 coated carbon cloth fibers (CCFs), and original CCFs air filters at various
relative humidity conditions for nine volatile organic compounds. TiO2/FGFs, TiO2/CCFs,
and CCFs were characterized by SEM for morphology and N2 adsorption isotherm for
BET surface area and pore structure. A bench-scale adsorption test setup was constructed
and adsorption tests were performed at various relative humidity conditions and four
different injected concentrations for each compound. The isothermal adsorption curves at
low concentration levels were obtained and they were well described by Langmuir
isotherm model. It was noticed that there were significant differences between the
adsorption behaviours and photocatalytic activities of TiO2/FGFs and TiO2/CCFs. It was
concluded that adsorption performance is closely related to the characteristics of
substrates and therefore, the development of a substrate with high adsorption ability is a
promising trend for improving the performance of the UV-PCO technology.
Keywords: Filters, Titanium dioxide (TiO2), adsorption, relative humidity (RH),
photocatalytic reaction
*Corresponding Author:
Prof. Fariborz Haghighat
Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, Quebec, Canada.
Tel: 1-514-848-2424 (3192), Fax: 1-514-848-7965
E-mail: haghi@bcee.concordia.ca
1. Introduction
The indoor air quality (IAQ) has received enormous attention in the area of industrial,
educational and medical care sectors worldwide. It is anticipated that this concern will
continue in the near future, and it will be driven by a multitude of strategies to mitigate
IAQ problems. Due to the increased demand of energy savings, air filtration and
purification can be an alternative approach to improve IAQ, rather than solely relying on
the dilution ventilation. Photocatalytic oxidation (PCO), an innovative and developing air
purification technology, gradually becomes a promising approach for such application [1,
2].
Purification mechanism of PCO technology is that electron-hole pairs generated from the
surface of a semiconductor under UV light irradiation break down adsorbed organic
pollutants, theoretically, into water and carbon dioxide. In general, it is assumed that the
organic pollutant adsorbed on the substrate undergoes oxidation reactions through a
surface-bound hydroxyl radical formed from electrons transfer, rather than direct holes
transfer, although in some cases there is some evidence that it may occur [3]. Although it
is disputable that the PCO reaction may take place at fluid stream, most of researchers
propose that only those pollutant molecules in direct contact with catalyst surface
undergo PCO reaction [4-6]. In other words, adsorption process is an important
prerequisite for PCO reaction; it also affects the efficiency of PCO air cleaners.
As an alternative method for purification of air stream, the PCO technology has been
studied for decades, and numerous papers focusing on various aspects of PCO technology
have been reported. They include the improvement of photocatalysts, the evaluation of
PCO performance of different bench-scale reactors, potential intermediates, kinetic study
and PCO modeling development [7-11]. Although a variety of PCO models [11, 12]
indicate adsorption of reactants on the catalyst surface is one of physical and chemical
processes occurring at a heterogeneous photocatalytic reaction, literatures especially
aiming to explore the adsorption effect of a photocatalyst are limited. In addition,
different forms of Langmuir-Hinshelwood (L-H) models have been developed; a
Langmuir adsorption constant is a critical parameter in L-H models. Usually, the values
of adsorption coefficient were determined by fitting a kinetic PCO model with
experimental data [9, 13-14]. In fact the values derived from a kinetic study are usually
much higher than those obtained from the Langmuir adsorption isotherm in the absence
of light [3]. In few articles relevant to the adsorption isotherms [15-18], the adsorption
performance was examined with a limited number of compounds of interest, such as
toluene, acetone, and trichloroethylene. Moreover, the effect of relative humidity on
adsorption isotherms has not been fully examined. To the best of our knowledge, no
systematic studies have been conducted with respect to an adsorption phenomenon of a
photocatalyst for a wide range of VOCs. Also research on the adsorption properties of a
photocatalyst placed on various substrates is limited. Hence, this paper reports the
outcomes of a systematic approach for evaluation of the adsorption system performance
at various relative humidity conditions for various VOCs. It also provides explanations to
the differences of adsorption features observed between three air filters based on a
fundamental analysis of the mass transports, and how these differences can influence the
photocatalytic activity.
Experimental determination of adsorption coefficient is a basic research for the purpose
of widespread commercial utilization of UV-PCO technology since the adsorption
coefficient is a critical parameter that influences the surface coverage of adsorbed
compounds, thereby affecting the photocatalytic oxidation rate. In this study, the
adsorption performance of titanium dioxide (TiO2) loaded on two different substrates,
fiberglass fibers (FGFs) and carbon cloth fibers (CCFs) has been experimentally
investigated. For the FGFS, this paper demonstrates a systematic evaluation of adsorption
performance at various relative humidity conditions (9.6±0.6% - 70.2±2.7%) and at
room temperature of 22.8±0.5℃ for nine compounds: toluene, p-xylene, ethanol, 1butanol, methyl ethyl ketone (MEK), acetone, hexane, octane, and limonene. The
challenge air concentrations of each selected compound are from 0.5 ppm to 5 ppm. For
the CCFs with and without TiO2 coating, ethanol and hexane have been employed to
examine the adsorption performance, as a typical polar and a non-polar VOC,
respectively, and at various relative humidity conditions. This paper provides a profound
insight into the basic knowledge of TiO2 adsorption mechanism, helps to determine the
values of adsorption coefficients for a PCO modeling, and further reveals that the
adsorption behaviour is closely related to the characteristics of substrates.
2. Methodology
2.1 Materials
Two commercially available PCO filters, TiO2 coated on fiberglass fibers and TiO2
coated on carbon cloth fibers, were examined in this study. The exact properties of filters
and catalyst preparation cannot be given due to the proprietary of these TiO2 filters. The
morphology of these materials was evaluated by scanning electron microscope (SEM,
Hitachi S-4700 Model). Brunauer-Emmett-Tele (BET) surface area and pore parameters
were determined by nitrogen adsorption-desorption isotherm measurements
(Micromeritics ASAP 2000).
Nine reagent grade chemicals were selected as representative of indoor air contaminants
[19], and they cover major chemical categories and have a wide range of different
physical properties like molecule weight and polarity, which are shown in Table 1. They
included toluene (99.9%), p-xylene (99.9%), 1-butanol (99.9%), n-hexane (96%), octane
(95%), MEK (99.9%), acetone (99.5%) and d-limonene (97%) from Fisher Scientific Inc.
(Canada), and ethanol (99%) from SAQ (Société des alcools du Québec - Québec
Alcohol Board).
Table 1
Physical Properties of the Selected VOCs [20]
BPb
VOCs
Molecular
MWa
VPc
Chemical
DCd
Class
Formula
(g/mol)
(℃)
(mmHg)
Ethanol
C2H6O
46.1
78.4
44
24.3
Alcohols
1-butanol
C4H10O
74.1
117.2
6
16.68
Acetone
C3H6O
58.1
56.1
180
20.7
Ketones
MEK
C4H8O
72.1
79.4
78
18.51
Toluene
C7H8
92.1
111.1
21
2.38
Aromatic
p-Xylene
C8H10
106.2
138.3
9
2.2
n-Hexane
C6H14
86.2
68.9
124
1.88
Alkane
Octane
C8H18
114.2
125.6
10
1.94
d-Limonene Terpenes
C10H16
136.2
176
2
2.3
Note: (a) MW stands for molecular weight, (b) BP stands for boiling point, (c) VP stands
for vapor pressure, (d) DC stands for dielectric constant at 20-25℃. Water (very polar)
has a dielectric constant of 80.10 at 20 ℃.
2.2 Adsorption experimental setup
To quantitatively study equilibrium adsorption and to examine the impact of relative
humidity, the bench-scale adsorption test setup complying with ASHRAE Standard 145.1
[21] was used, see Figure 1. The laboratory compressed air was used as the carrier gas, its
flow rate was controlled by a mass flow meter, and it was purified by passing it through a
granular activated carbon (GAC) filter to remove potential contaminants. The relative
humidity of the mixed stream was varied from (9.6%±0.6%) to (70.2%±2.7%) by
adjusting the flow rate of the compressed air into the distilled water bottle. In addition,
the temperature of the distilled water bottle was maintained constant through a water bath
so that it could provide water vapor with a steady concentration. The selected chemicals
were in liquid state at room temperature; they were automatically injected through a
syringe pump (KD Scientific). The injected concentrations of selected chemicals were
calculated on the basis of airflow rate, chemical injection rate and the chemical properties.
Before connecting to a media column, a PTFE tube (Figure 1- (1)) was first connected to
the test system and was used to transport gaseous pollutant to the calibrated multi-gas
analyzer (Innova AirTech Instruments 1312). After the inlet concentration reached to a
steady state, the PTFE tube was replaced by a media column filled with TiO2/FGFs,
TiO2/CCFs, or CCFs (Figure 1- (2)). The concentration of challenge VOCs after the
media column was measured continuously till it was equal to the previous stabilized
concentration. For each compound, the adsorption test was performed at T= 22.8±0.5℃
and four different injected concentrations (0.496±0.013 ppm, 0.990±0.010 ppm, 1.971
±0.023 ppm, and 4.955±0.063 ppm). Before performing each adsorption experiment,
the air filter was conditioned overnight under the corresponding humid conditions to be
tested; after each adsorption test, the air filter was conditioned again for regeneration by
passing through humidified compressed air.
Flow Controller
••••
••••
GAC
filter
(2)
Humidity
PCO
filter
(1)
Compressed
air
Mass flow
meter
Figure 1: Adsorption test setup
2.3 Adsorption isotherm
Many different expressions that describe dynamic equilibrium of sorbed-phase and
gaseous phase have been proposed, among which the Langmuir isotherm is the most
widely applied in the field of surface kinetics. The Langmuir isotherm model describes
adsorbate-adsorbent systems in which the extent of adsorbate coverage is limited to
monolayer coverage of the surface, especially at low loadings. The sorbed-phase
concentration of the VOC at the catalyst fibers surface,CS , can also be estimated by this
model.
C0 KC
S
CS = f(C) = 1+KC
(1)
where CS0 is the maximum sorbed-phase concentration corresponding to monolayer
complete coverage, C is the gaseous phase concentration, K is an equilibrium constant,
which is the adsorption constant divided by the desorption constant. When the
concentration of challenge gases is very low, that is KC<<1, Eq. (1) can be simplified to;
Cs = f(C) = CS0 KC = K ′𝐿 C
(2)
where K ′𝐿 is the synthetic Langmuir parameter that embeds both the saturation
capacity, CS0 , and the equilibrium constant, K. Eq. (2) can be changed into Eq. (3) by
converting the sorbed-phase concentration, Cs, to the mass of VOCs adsorbed in the
surface of catalyst, ms, and introducing the mass of air filters, mf, to compensate its
impact to the adsorption behavior.
𝑚𝑠
𝑚𝑓
= 𝐾𝐿 𝐶
(3)
where ms is the mass of adsorbed VOCs, mf is the mass of air filters, and KL is the
adsorption coefficient which is defined as the ratio between the mass of adsorbed VOCs
in solid phase and the concentration of VOCs in air phase at adsorption equilibrium per
unit weight of air filters.
2.4 Adsorption analysis method
Based on the experimental data obtained from both upstream and downstream air
concentrations measurements, adsorption capacity of the fiberglass PCO filter for a
certain VOC gas can be evaluated. This capacity is expressed as the ratio of the adsorbed
mass of contaminant gas over the removal media weight [22]:
RC


ads
0


(t )
(t )
Q Cup
 Cdown
dt
M PCO
(4)
where R C is the filter capacity when it reaches equilibrium; Tads is the elapsed time of
(t)
(t)
adsorption test (min); Q is the airflow rate (m3/min); Cup and Cdown are the upstream and
downstream concentration (g/m3) as a function of elapsed time, respectively; MPCO is the
mass of removal media (g): for TiO2/FGFs, it was 2.715 g; for TiO2/CCFs with TiO2, it
was 0.890 g; and for original CCFs, it was 0.386 g. It should be noted that in this study
(t)
Cup is computed as the average upstream concentration (before adsorption occurs).
2.5 Photocatalysis experimental setup
A single pass mode test rig was built to evaluate the PCO effectiveness for treating
ethanol and hexane with inlet concentrations of 0.25 ppm, 0.5 ppm, and 1 ppm. The test
rig was composed of four parallel test ducts with 0.3 m by 0.3 m of cross-sectional area
each. UV-PCO section was designed to be versatile so that different UV-PCO systems
with various geometries can be installed. Each duct has a fan with a variable speed
control so that the airflow rate can be controlled regardless of the flow resistances of
different UV-PCO systems. Since the length of the duct before the fan is long enough, it
is reasonable to assume the airflow as ideal plug flow. This was verified through
measurement at various locations both at the upstream and downstream. The geometry
dimensions of test rig on the elevation view are presented in Figure 2, and detailed setup
can be found in [23]. In this study, the UV-PCO reactor contained three 0.3 m by 0.3 m
TiO2/FGFs air filters or TiO2/CCFs air filters arranged in two banks, and in each bank
two UV lamps (2╳18.4 W UVC lamps or 2╳18.4 W VUV lamps) were employed (total
four) for providing irradiation. The nominal 254nm UV output power at 100 hours and
26.7℃ was 5.8 W, and 254nm UV output at one meter from a lamp was 59 mW/cm2. The
vertical distance between the surfaces of UV lamps and the PCO filter was approximate 5
cm.
76.20 cm
(30.00’’)
60.96 cm
(24.00’’)
355.60 cm
(140.00’’)
76.20 cm
(30.00’’)
60.96 cm
(24.00’’)
35.56 cm
(14.00’’)
45.72 cm
(18.00’’)
PCO reactor
30.99 cm
(12.20’’)
38.61 cm
(15.20’’)
30.99 cm
(12.20’’)
38.61 cm
(15.20’’)
59.28 cm
(23.34’’)
128.90 cm 121.28 cm
(50.75’’) (47.75’’)
PCO reactor
60.96 cm
(24.00’’)
76.20 cm
(30.00’’)
60.96 cm
(24.00’’)
35.56 cm
(14.00’’)
45.72 cm
(18.00’’)
Figure 2: Dimensions of the test rig on elevation view
The target concentration of the selected challenge gases was at sub-ppm level (0.25-1
ppm) which represented indoor air pollution. All PCO experiments were carried out at
room conditions. Table 2 summarizes the detailed experimental conditions on which the
experiments were performed. The experimental data collected from upstream and
downstream measurement ports were employed to calculate the effectiveness of a UVPCO air cleaner. Single-pass removal efficiency, which was determined by the amount of
the removed pollutant from the air stream after it went through the air cleaner, is widely
used to evaluate the performance of an air cleaner.
𝜂𝑡 (%) =
𝐶𝑢𝑝,𝑡 −𝐶𝑑𝑜𝑤𝑛,𝑡
𝐶𝑢𝑝,𝑡
× 100
Table 2: Photocatalysis experimental conditions
Parameter
Experimental Conditions
Single VOC
ethanol and hexane
Inlet concentration (ppm)
0.25-1
(5)
Volumetric flow rate (m3/h)
Face velocity (m/s)
Relative humidity (%)
Light intensity (W/m2)
Temperature (℃)
170 (100 cfm)
0.5
15-30
29-36
23
3 Results
Figure 3 presents the equilibrium adsorption isotherms for the selected challenge VOCs
under different RH conditions. The dimensionless mass ratio (g/g) was used to facilitate
the comparison. This figure shows, at four RH levels, all isotherms are linear with respect
to the equilibrium concentration. This indicates that the adsorption behaviours of
TiO2/FGFs, TiO2/CCFs, and CCFs follow ideal monolayer adsorption at low ppm
concentrations. It should be noted that the slope of each regression line is the adsorption
coefficient of individual VOC for various RH. The results clearly indicate that the
adsorption performance of the TiO2 catalyst varies for different compounds and
substrates. And adsorption coefficient is significantly affected by the presence of water
vapor.
35.2%RH
50.8%RH
16
69.8%RH
y = 0.1775x
R² = 0.9830
12
8
y = 0.0139x
R² = 0.9817
y = 0.0275x
R² = 0.9198
4
y = 0.0054x
R² = 0.9162
0
0
2
4
6
1-butanol
9.3%RH
8
Adsorbed mass/PCO filter mass×10-3
Adsorbed mass/ PCO filter mass×10-4
Ethanol
20
10
35.0%RH
4
y = 0.1357x
R² = 0.9993
16
50.8%RH
y = 0.1523x
R² = 0.9813
71.4%RH
12
8
y = 0.0209x
R² = 0.9901
4
0
y = 0.0081x
R² = 0.9903 y = 0.0021x
R² = 0.9459
0
5
10
Concentration (g/m3)×10-3
15
y = 0.0546x
R² = 0.9966
0
5
10
Concentration (g/m3)×10-3
15
MEK
9.3%RH
34.7%RH
69.8%RH
8
0
Adsorbed mass/PCO filter mass×10-3
Adsorbed mass/PCO filter mass×10-4
Acetone
50.8%RH
y = 0.8868x
R² = 0.9447
12
Concentration (g/m3)×10-3
20
10.0%RH
16
y = 0.0127x
R² = 0.9978
9.0%RH
6
34.0%RH
5
48.7%RH
y = 0.3573x
R² = 0.9418
70.4%RH
4
3
2
y = 0.0520x
R² = 0.9875
1
y = 0.0246x
R² = 0.9965
0
0
5
10
Concentration (g/m3)×10-3
15
y = 0.0030x
R² = 0.9562
9.9%RH
Adsorbed mass/ PCO filter mass×10-4
Adsorbed mass/ PCO filter mass×10-4
Toluene
5
36.3%RH
50.2%RH
y = 0.2392x
R² = 0.9907
4
72.5%RH
3
2
y = 0.0157x
R² = 0.9729
y = 0.0249x
R² = 0.9969
1
y = 0.0006x
R² = 0.9952
0
0
5
10
15
p-Xylene
30.7%RH
47.1%RH
y = 0.7853x
R² = 0.9852
16
8
0
0
5
y = 0.0112x
R² = 0.9989
34.0%RH
20
15
10
y = 0.0009x
R² = 0.9933
5
0
10
20
25
15
20
25
25
10.0%RH
y = 0.0012x
R² = 0.9927
20
15
10
5
0
0
5
10
Concentration (g/m3)×10-3
15
20
Concentration (g/m3)×10-3
10.6%RH
Limonene
Adsorbed mass/ PCO filter mass×10-3
15
Hexane
9.4%RH
Adsorbed mass/PCO filter mass×10-6
Adsorbed mass/PCO filter mass×10-5
10
Concentration (g/m3)×10-3
Octane
5
y = 0.0358x
R² = 0.9998
y = 0.0011x
R² = 0.9831
y = 0.0825x
R² = 0.9972
4
20
30
0
62.4%RH
12
Concentration (g/m3)×10-3
25
9.3%RH
20
7
34.6%RH
y = 0.2298x
R² = 0.9371
6
50.4%RH
70.0%RH
5
4
3
2
y = 0.0229x
R² = 0.9737
1
y = 0.0038x
R² = 0.9465
y = 0.0007x
R² = 0.9656
0
0
5
10
15
20
25
Concentration (g/m3)×10-3
30
(a) TiO2/FGFs
30.0%RH
46.5%RH
5
y = 0.5398x
R² = 0.9898
70.9%RH
4
y = 0.4125x
R² = 0.9988
3
y = 0.3057x
R² = 0.9859
2
y = 0.2291x
R² = 0.9752
1
0
0
2
4
6
8
10
12
9.63%RH
Hexane
8.8%RH
Adsorbed mass/PCO filter mass×10-3
Adsorbed mass/ PCO filter mass×10-3
Ethanol
6
35
32.11%RH
30
47.44%RH
y = 1.7882x
R² = 0.9780
25
73.79%RH
20
y = 0.9035x
R² = 0.9991
15
y = 0.4382x
R² = 0.9733
10
5
y = 0.2876x
R² = 0.9779
0
0
5
10
15
Concentration (g/m3)×10-3
Concentration (g/m3)×10-3
(b) CCFs
20
8.8%RH
28.1%RH
30
45.9%RH
y = 0.3315x
R² = 0.9548
25
70.7%RH
y = 0.1811x
R² = 0.9846
20
15
y = 0.0886x
R² = 0.9748
10
5
y = 0.0378x
R² = 0.9819
0
0
2
4
6
8
10
Concentration (g/m3)×10-3
12
Hexane
Adsorbed mass/PCO filter mass×10-3
Adsorbed mass/PCO filter mass×10-4
Ethanol
35
10.9%RH
25
30.6%RH
46.0%RH
y = 1.1098x
R² = 0.9728
20
71.3%RH
15
10
y = 0.3358x
R² = 0.9970
5
y = 0.0981x
R² = 0.9829
y = 0.0555x
R² = 0.9546
0
0
5
10
15
20
Concentration (g/m3)×10-3
(c) TiO2/CCFs
Figure 3: Adsorption isotherms of the selected challenge gases at various RH conditions
(9.6% ± 0.6% - 70.2% ± 2.7%) and at 22.8±0.5 ℃ for (a) TiO2/FGFs (b) CCFs (c)
TiO2/CCFs
4 Discussions
4.1 Characterization
Figure 4 shows the scanning electron microscopy (SEM) images of TiO2/FGFs and
TiO2/CCFs which shows fibers are randomly oriented for both media and TiO2/CCFs
displays fracture surfaces. Nano-TiO2 particles were coated on the substrates, and the
magnified SEM images in Figure 4b and 4d demonstrate nano-TiO2 powders were more
uniformly distributed on the FGFs’ surface than those loaded on the CCFs due to the
CCFs’ fiber roughness. Figure 5 presents nitrogen adsorption isotherms for CCFs,
TiO2/CCFs and TiO2/FGFs and their pore size distributions which were obtained from
desorption branch of nitrogen isotherm by the BJH (Barrett-Joyner-halenda) method [24].
The adsorption isotherms (plotted in Figure 3) indicate the order of adsorption capacity is
CCFs > TiO2/CCFs > TiO2/FGFs, which is also further verified by the measured BET
surface areas (Table 3). The BET surface area of TiO2/FGFs was much smaller than that
of TiO2/CCFs. It is worth mentioning that BET surface area decreased from 1490.8605
m2/g to 887.6638 m2/g when TiO2 was loaded on the surface of CCFs. Additionally, the
pore size distribution curve manifests that the pore volume of TiO2/CCFs decreased to
some extent in the range of mesopores and formation of maropores resulting from TiO2
particle agglomeration were observed compared with the original CCFs, see Figure 5.
This observation indicates the process of TiO2 coating did not destroy the pore structure,
and the presence of the agglomeration of catalyst was at the external surface of fibers.
This is consistent with the observation reported by Guo et al. [25] when they examined
similar materials.
Figure 5 also shows that pore diameter distributions were narrow for CCFs and
TiO2/FGFs, and pores of three materials (CCFS, TiO2/FGFs, TiO2/CCFS) were mainly
mesopores. It can also be found from Table 3 that the pore volume and pore size of
TiO2/CCFs are larger than those of TiO2/FGFs.
(b)
(a)
(d)
(c)
Figure 4: a) SEM images of TiO2/FGFs and b) magnified image of TiO2/FGFs c) SEM
images of TiO2/CCFs and d) magnified image of TiO2/CCFs
Isotherm Plot
CCFs
CCFs
600
Pore volume (cm 3/g)
Volume adsorbed (cm3/g)
700
TiO2/CCFs
500
TiO2/FGFs
400
300
200
100
TiO2/CCFs
TiO2/FGFs
1.2
1
0.8
0.6
0.4
0.2
0
0
0
0.2
0.4
0.6
Relative pressure (P/P0)
0.8
1
1
10
Pore diameter (nm)
(a)
(b)
Figure 5. (a) N2 adsorption isotherms (b) Pore size distributions
100
Table 3: BET surface areas and pore structure parameters for three PCO filters
Parameter
CCFs
TiO2/CCFs TiO2/FGFs
2
BET surface area (m /g)
1490.8605
887.6638
105.7063
BJH cumulative desorption pore volume
0.260631
0.638990
0.103594
(cm3/g)
BJH desorption average pore diameter (nm)
2.75819
9.58866
3.59489
TiO2 loading (wt%)
--14.32
4.63
4.2 Adsorption coefficients verification
Demeestere et al. [15] presented the linear adsorption curves for trichloroethylene,
toluene, and chlorobezene on the photocatalyst TiO2 Degussa P25 at 25.0℃. They found
that the adsorption coefficient of toluene was 0.00253 m3/g and 0.000103 m3/g for RH=0%
and 57.8%, respectively. Coronado et al. [26] reported the adsorption constant of acetone
for the weak sites of TiO2 thin film was 0.2 m3/g and 0.046 m3/g for RH=0% and 25%,
respectively. Tomida et al. [16] investigated the adsorption isotherms of acetone on the
photocatalyst of TiO2 coated on silica beads and obtained similar results. All the reported
adsorption coefficient values are at the same order of magnitude as those obtained in this
study.
4.3 Correlation between adsorption coefficient and properties of VOC
Results of TiO2/ FGFs indicate that for the polar VOCs, such as ethanol, isobutanol,
acetone, and MEK, the adsorption coefficients are roughly one order of magnitude higher
than those of non-polar VOCs for a given RH level. Also, the results further indicate the
TiO2/ FGFs filter, a polar substance, has a higher adsorption preference for polar VOCs.
The order of adsorption capacity for the selected chemical classes follows the sequence of
alcohols > ketones and terpenes > aromatics > alkanes. This feature is attributed to the
strength of the corresponding intermolecular forces between adsorbed VOCs and the
catalyst surface. Dispersion forces are the main intermolecular forces holding non-polar
alkanes in the solid phase, which are weaker than van der Waals interactions for aromatic
hydrocarbons. Due to the high dipole moment of the carbonyl group, dipole-dipole
interactions for ketones are stronger than van der Waals attractions for hydrocarbons. In
addition to van der Waals interactions, hydrogen bonding plays a greater role for the
attraction between alcohols and hydrated catalyst surface. This order agrees with the
photocatalytic oxidation rates reported by Hodgson et al. [27] and Obee and Hay [28].
Thus, it is inferred that the PCO reaction rate is closely related to the adsorption process.
Another obvious trend has been observed from the results of TiO2/ FGFs. That is, for the
compounds in the same chemical class, the adsorption coefficient increases with the
increase of molecular weight and boiling point. This agrees with the adsorption
characteristics for sorption-based media. For the structurally similar compounds, the
higher the boiling point is the greater the intermolecular forces. Hence, the van der Waals
forces of the heavier compounds make them more likely to be adsorbed to the TiO2
catalyst.
Adsorption results of CCFs and TiO2/CCFs demonstrate that CCFs belongs to a nonpolar material due to the higher adsorption capacity of hexane compared with that of
ethanol, see Figure 3. Moreover, adsorption capacity of original CCFs is higher than that
of TiO2/CCFs for both hexane and ethanol since the BET surface area of CCFs is larger
than that of TiO2/CCFs, which results in the fact that adsorption sites on CCFs are
obviously more than that those on TiO2/CCFs.
4.4 Effect of relative humidity
The effect of RH on the adsorptive performance of three substrates was investigated in
the range of 9% (2300 ppm)-70% (19600 ppm) at room temperature of 22.8±0.5℃. From
Figure 6, it can be noticed that the adsorption coefficient of each tested compound
decreases dramatically as RH rises from 9% to 70% for three substrates. Obviously, the
presence of water vapor plays an important role on the adsorption behaviour of VOCs on
the catalyst.
1.8
0.8
Adsorption coefficienct (m3/g)
Adsorption coefficienct (m 3/g)
2
TiO2/FGFs
0.9
1-Butanol
MEK
d-Limonene
Ethanol
Acetone
p-Xylene
Toluene
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Hexane(CCFs)
Hexane(TiO2/CCFs)
Ethanol(CCFs)
Ethanol(TiO2/CCFs)
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0
0%
20%
(a) TiO2/FGFs
40%
RH
60%
80%
0%
20%
40%
RH
60%
80%
(b) TiO2/ CCFs and CCFs
Figure 6: Relation between adsorption coefficient and RH for different VOCs at 22.8±
0.5℃: (a) TiO2/FGFs, and (b) TiO2/ CCFs and CCFs
Figure 7 shows the variation of KL as a function of RH. This figure indicates that K L
varies exponentially: an increasing amount of physically adsorbed water vapour
significantly results to reducing the adsorbed organic molecules at the surface of catalyst.
This phenomenon can be attributed to the competition of water molecules and VOCs
molecules at the surface sites. Hydrogen bonding for water is stronger than van der Waals
interactions for most tested VOCs so that water is more easily adsorbed on the surface of
photocatalysts. According to Maudhuit et al. [18], the formation of one or several layers
of water clusters at the surface of TiO2 in humid air conditions decreases accessibility of
the pollutants to active sites. In addition, decreasing trends of adsorption capacity with
RH also indicates the tested three PCO air filters in this study are hydrophilic. Figure 7
shows there is a linear relationship between ln(KL) and RH.
0
20
RH (%)
40
60
80
0
Acetone
lnKL = -6.8496RH - 1.3540 (R² = 0.9983)
MEK
lnKL = -7.5066RH - 0.4227 (R² = 0.9836)
Ethanol
lnKL = -5.7982RH - 1.3543 (R² = 0.9815)
1-Butanol
lnKL = -7.0135RH + 0.5550 (R² = 0.9977)
-4
Toluene
lnKL = -5.8133RH - 3.4396 (R² = 0.9541)
-5
p-Xylene
lnKL = -7.8335RH - 2.0199 (R² = 0.9790)
-1
lnKL(m3/g)
-2
-3
d-Limonene lnKL = -9.8801RH - 0.4312 (R² = 0.9978)
-6
-7
-8
(a) TiO2/FGFs
RH (%)
0
20
40
60
80
1
lnKL(m3/g)
0
-1
-2
-3
Hexane(CCFs)
lnKL = -2.9326RH + 0.7969 (R² = 0.9619)
-4
Hexane(TiO2/CCFs) lnKL = -5.0844RH + 0.4676 (R² = 0.9416)
-5
Ethanol(CCFs)
-6
lnKL = -1.4064RH - 0.4912 (R² = 0.9937)
Ethanol(TiO2/CCFs) lnKL = -3.5504RH - 0.7666 (R² = 0.9983)
(b) TiO2/ CCFs and CCFs
Figure 7: Adsorption profiles of various compounds with relative humidity range from
9.6% to 70.2% at 22.8±0.5℃ for (a) TiO2/FGFs, and (b) TiO2/CCFs and CCFs
4.5 Comparison adsorption performance of TiO2/FGFs and TiO2/CCFs
Table 3 shows that the BET surface area of TiO2/ CCFs is higher than that of TiO2/FGFs
results in TiO2/ CCFs having a greater adsorption capacity. This is ascribed to the fact
that a large number of adsorption sites on the TiO2/ CCFs surface are beneficial to rapid
molecular diffusion of VOCs, thereby increasing the adsorption capacity. Moreover,
CCFs as a supporting substrate have high adsorption ability. When the challenge VOCs
come in contact with the surface of PCO filter, part of the molecules are directly adsorbed
by the TiO2, while the rest may be physically captured by CCFs. Test results of hexane
shown in Fig. 6 (b) are an excellent interpretation of the significant impact of the
substrate on the adsorption performance. For TiO2/FGFs, the adsorption capacity of
hexane was very low. Even when RH was greater than 9.6%±0.6%, the adsorption
phenomena were not clearly observed (Fig. 3 (a)). While for TiO2/ CCFs, the adsorption
of hexane was obviously observed at various RH values (Fig. 3 (c)). The adsorption
capacity was found as 1.1098 m3/g (9.6%±0.6%), 0.3358 m3/g (32.9%±2.5%), 0.0981
m3/g (48.3%± 1.8%) and 0.0555 m3/g (70.2% ±2.7%), which was greater than the
adsorption capacity of ethanol under the same RH conditions, respectively (Fig. 3(c)).
This is consistent with results reported in previous studies [29] that removal performance
of the granular activated carbon filters is positively correlated to the VOCs molecular
weight.
Figure 7 shows that the presence of water vapor has less influence on the adsorption
behavior of TiO2/ CCFs than that of TiO2/FGFs. The strength of hydrophilicity of a
substrate determines the extent of interactions of adsorbed water film on the TiO2 surface.
Hence, FGFs is more hydrophilic compared with CCFs so that affinity of water for FGFs
is stronger than that of CCFs. This can be interpreted with the fact that adsorption energy
of water vapor for fiberglass is 0.57 J/mol [30], which is much smaller than that of 27.2
kJ/mol for activated carbon [31]. Cao et al. [32] reported that TiO2 by itself is a kind of
strongly hydrophilic substance. Therefore, RH parameter, to some extent, influences the
adsorption property of air filters with surfaces coated with TiO2.
4.6 photocatalytic activity of TiO2/FGFs and TiO2/CCFs
Since adsorption process is one of important steps involved in the UV-PCO technology,
the study of photocatalytic activity of TiO2/FGFs and TiO2/CCFs is necessary for better
understanding of the influence of adsorption process on the UV-PCO. Figure 8 shows
that single-pass UV-PCO removal efficiency of TiO2/CCFs is distinctly higher than that
of TiO2/FGFs for both compounds at various concentrations. Generally, the larger
specific surface area helps to increase photocatalytic activity, which can be interpreted
from three aspects. First, more active sites are provided through coating of TiO2 nanoparticles on larger BET surface (887.6638 m2/g) of TiO2/CCFs (Table 3). Second, under
the same level of illumination, extremely dispersed TiO2 particles on the surface of
TiO2/CCFs generate more negative electron-positive hole pairs which may reacts with
water vapor for the formation of hydroxyl radicals, thus reducing the recombination of
electron-hole pairs. Third, large specific surface area is conducive to the rapid molecular
diffusion of VOCs. Matos et al. [33] also observed the enhanced photoactivity through
combination of photocatalysis and activated carbon for treatment of toluene. They
demonstrated that the contact interface between TiO2 and activated carbon permits
adsorbed pollutants diffuse from activated carbon to TiO2 for photooxidation. Then
activated carbon was able to adsorb more pollutants from gas stream, and passed them to
TiO2 again. It is this pollutant transfer cycle, as a driven force, that promotes the
photocatalytic conversion of pollutants.
For TiO2/FGFs air filter, the ordering of photooxidation activity is consistent with the
rank of adsorption characteristics discussed above, that is, adsorbed ethanol at the surface
of TiO2/FGFs reacts more readily with hydroxyl radicals than adsorbed hexane does.
This conclusion is in agreement with the results reported by Hodgson et al. [27]. For
TiO2/CCFs air filter, although the adsorption capacity of ethanol is lower than that of
hexane, the PCO removal efficiency of ethanol is still higher than that of hexane,
indicating in addition to adsorption process, UV-PCO performance is also affected by the
photochemical reactions of VOCs. When heterogeneous mass transfer of pollutants from
gas phase to solid phase does not limit the PCO process, surface PCO reaction is the
controlling step during UV-PCO technology.
Single-pass removal efficiency
70%
60%
TiO2/FGFs(UVC)
50%
TiO2/FGFs(VUV)
40%
TiO2/CCFs(VUV)
30%
20%
10%
0%
250ppb
500ppb 1000ppb 250ppb
500ppb 1000ppb
Ethanol
Hexane
Figure 8: Single-pass UV-PCO removal efficiency of ethanol and hexane at three initial
concentration levels: 0.25 ppm, 0.5 ppm, and 1 ppm. (Flow rate= 170 m3/h, RH=15%30%, irradiance= 29-36 W/m2)
Therefore, the fundamental mechanisms involved in whole PCO process are divided into
several elemental mass transfer processes occurring in series, namely 1) convection,
diffusion and boundary transfer of contaminants in the air-phase; 2) inter-phase mass
transfer of reactant species; 3) adsorption and UV-PCO reaction in the solid-phase [34].
The above mentioned UV-PCO test results further verify that the adsorption process is an
important prerequisite for PCO reaction. And the large specific surface area could
increase the photocatalytic reaction rate as a local high pollutant concentration can be
formed on the surface of TiO2 by the adsorption of CCFs. It is worthwhile to mention
PCO reaction occurs on the surface of catalyst fibers illuminated with UV lights. That is,
only adsorbed VOCs on the solid surface that is exposed to UV irradiation effectively
participate in UV-PCO reactions. As for the synergistic effect of adsorption and UV-PCO
on the removal of indoor VOCs, it is necessary to conduct more tests in the future to
further the knowledge of two processes affecting the conversion efficiency.
5. Conclusions
In this study, the adsorption performance of TiO2/FGFs, TiO2/CCFs, and CCFs air filters
for several compounds at various RH conditions was tested using adsorption isotherm
method. This method can be employed to separately examine the adsorption behaviours
of different air filters when the PCO technology works in the absence of illumination.
The isothermal adsorption curves at low concentration levels were well described by
Langmuir isotherm model. TiO2/FGFs, TiO2/CCFs, and CCFs were characterized by
SEM for morphology and N2 adsorption isotherm for BET surface area and pore structure.
The following conclusions can be drawn from this study:
(1) Adsorption capacity can be ranked as: CCFs > TiO2/CCFs > TiO2/FGFs, which are
consistent with the order of measured BET surface areas of three air filters. This
shows the fact that the adsorption performance is affected by the properties of
substrates.
(2) The test results of TiO2/FGFs demonstrate that TiO2/FGFs air filter presents
hydrophilic property, and compounds with high polarity show higher affinity to the
surface of TiO2/FGFs air filter due to the strong intermolecular forces. The adsorption
capacity of the selected chemical classes ranks as follows: alcohols > ketones and
terpenes > aromatics > alkanes. Moreover, a larger compound in the same chemical
class has a greater adsorption coefficient because of van der Waals interactions. The
test results of TiO2/CCFs show CCFs belongs to a non-polar substrate which prefers
to adsorb non-polar compounds. Therefore, the adsorptive performance is the sum of
interactions between surface constitutes and a specific VOC.
(3) An increase of RH decreases the adsorption capacity for three tested media
attributable to the strong hydrogen bonding for water. In addition, the influence of RH
on the adsorption behaviour of TiO2/CCFs is less important than that on TiO2/FGFs.
(4) Due to the differences in adsorption performance, photocatalytic activity of the
TiO2/CCFs air filter is obviously higher than that of TiO2/FGFs. High performance of
TiO2/FGFs provides a promising direction to explore other supporting substrates with
high adsorption ability.
Acknowledgements
The authors would like to express their gratitude to the Natural Science and Engineering
Research Council Canada (NSERC) for the financial support through a CRD grant and
Circul-Aire, Inc. for the support in design and construction of the experimental facility at
Concordia. Electron microscopy research group of McGill University is acknowledged
for the help of SEM analysis. The authors thank Dr. Raymond Le Van Mao and Mr. Hai
Tao Yan for technical assistance of BET analysis.
References
[1] B. Wang, R. Mortazavi, F. Haghighat, Evaluation of modeling and measurement
techniques of ultraviolet germicidal irradiation effectiveness e towards design of
immune building, Indoor Built Environ. 18 (2009) 101-113.
[2] L. Zhong, F. Haghighat, P. Blondeau, and J. Kozinski, Modeling and physical
interpretation of photocatalytic oxidation efficiency in indoor air applications, Build
Environ. 45 (2010) 2689-2697.
[3] A. Mills, S. L. Hunte, An overview of semiconductor photocatalysis, J. Photochem.
Photobiol. A Chem. 108 (1997) 1-35.
[4] T.L. Thompson, J.T. Yates, Surface science studies of the photoactivation of TiO2new photochemical process, Chem. Rev. 106 (2006) 4428-4453.
[5] V.I. Gaya, A.H. Abdullah, Heterogeneous Photocatalytic degradation of organic
contaminants over titanium dioxide: A review of fundamentals, progress and
problems. J. Photochem. Photobiol. C 9 (2008) 1-12.
[6] O. Debono, F. Thevenet, P. Gravejat, V. Hequet, C. Raillard, L. Lecoq, N. Locoge,
Toluene Photocatalytic oxidation at ppb levels: kinetic investigation and carcon
balance determination, Appl. Catal. B Environ. 106 (2011) 600-608.
[7] J. Krysa, M. Keppert, J. Jirkovsky, V. Stengl, J. Subrt, The effect of thermal treatment
on the properties of TiO2 photocatalyst, Mater. Chem. Phys. 86 (2004) 333-339.
[8] H. Liu, X. Ye, Z. Lian, Y. Wen, W. Shangguan, Experimental study on photocatalytic
oxidation of formaldehyde and its by-products, Res. Chem. Intermediat. 32 (2006) 916.
[9] M. Sleiman, P. Conchon, C. Ferronato, J.M. Chovelon, Photocatalytic oxidation of
toluene at indoor air levels (ppbv): towards a better assessment of conversion,
reaction intermediates and mineralization, Appl. Catal. B Environ. 86 (2009) 159–
165.
[10] Q.L. Yu, M.M. Ballari, H.J.H. Brouwers, Indoor air purification using heterogeneous
Photocatalytic oxidation. Part II: Kinetic study, Appl. Catal. B Environ. 99 (2010)
58-65.
[11] V. Tomasic, F. Jovic, Z. Gomzi, Photocatalytic oxidation of toluene in the gas phase:
modeling an annular photocatalytic reactor, Catal. Today 137 (2008) 350–356.
[12] R.G. Changrani, G.B. Raupp, Two-dimentional heterogeneous model for a
reticulated-foam photocatalytic reactor, AIChE. J. 46 (2000) 829–842.
[13] A. Bouzaza, C. Vallet, A. Laplanche, Photocatalytic degradation of some VOCs in
the gas phase using an annular flow reactor—determination of the contribution of
mass transfer and chemical reaction steps in the photo-degradation process, J.
Photochem. Photobiol. A Chem. 177 (2006) 212–217.
[14] F. Shiraishi, T. Ishimatsu, Toluene removal from indoor air using a miniaturized
photocatalytic air purifier including a preceding adsorption/ desorption unit. Chem.
Eng. Sci. 64 (2009) 2466–2472.
[15] K. Demeestere, J. Dewulf, H.V. Langenhove, B. Sercu, Gas-solid adsorption of
selected volatile organic compounds on titanium dioxide Degussa P25, Chem. Eng.
Sci. 58 (2003) 2255-2267.
[16] T. Tomida, N. Okada, M. Katoh, Adsorption and photocatalytic decomposition of
volatile organic compounds on photocatalyst of TiO2- Silica Beads, Adsorption 11
(2005) 865-869.
[17] J. Chen, G. Li, Z. He, T. An, Adsorption and degradation of model volatile organic
compounds by a combined titania-montmorillonite-silica photocatalyst, J. Hazard.
Mater. 190 (2011) 416-423.
[18] A. Maudhuit, C. Raillard, V. Hequet, L.L. Coq, J. Sablayrolles, L. Molins,
Adsorption phenomena in photocatalytic reactions: The case of toluene, acetone and
heptane. Chem. Eng. J. 170 (2011) 464-470.
[19] D.W. VanOsdell, Evaluation of test methods for determining the effectiveness and
capability of gas-phase air filtration equipment for indoor air applications-phase I:
literature review and test recommendations, ASHRAE Trans. 100 (1994) 511-523.
[20] NIOSH Pocket Guide to Hazardous Chemicals, National Institute of Occupational
Safety and Health, 2010 http://www.cdc.gov/niosh/npg/, March 5, 2012 last
accessed.
[21] ASHRAE, ANSI/ASHRAE Standard 145.1: Laboratory Test Method for Assessing
the Performance of Gas-Phase Air-Cleaning Systems: Loose Granular Media,
Atlanta, Georgia: American Society of Heating, Refrigerating and Air-Conditioning
Engineers, Inc., 2008.
[22] F. Haghighat, C.S. Lee, B. Pant, G. Bolourani, N. Lakdawala, A. Bastani, Evaluation
of various activated carbons for air cleaning: towards the design of immune and
sustainable buildings, Atmos. Environ. 42 (2008) 8176-8184.
[23] C.S. Lee, L. Zhong, D. Farhanian, C. Flaherty, F. Haghighat, Development of a
parallel test system for the evaluation of UV-PCO systems, Cold Climate
Conference, Calgary, Alberta, Canada, 2012.
[24] E. P. Barrett, L.G. Joyner, P.P. Halenda, The determination of pore volume and area
distributions in porous substances. I. Computations from nitrogen isotherms, J. Am.
Chem. Soc. 73 (1951) 373–380.
[25] T. Guo, Z. Bai, C. Wu, T. Zhu, Influence of relative humidity on the photocatalytic
oxidation (PCO) of toluene by TiO2 loaded on activated carbon fibers: PCO rate and
inermediates accumulation, Appl. Catal. B Environ. 79 (2008) 171-178.
[26] J.M. Coronado, M.E. Zorn, I.T. Tejedor, M.A. Anderson, Photocatalytic oxidation of
ketones in the gas phase over TiO2 thin films: a kinetic study on the influence of
water vapor, Appl. Catal. B Environ. 43 (2003) 329-344.
[27] A. T. Hodgson, H. Destaillats, D. P. Sullivan, W. J. Fisk, Performance of ultraviolet
photocatalytic oxidation for indoor air cleaning applications, Indoor Air 17 (2007)
305-316.
[28] T.N. Obee, S.O. Hay, The estimation of photocatalytic rate constants based on
molecular structure: extending to multi-component systems, J. Adv. Oxid. Technol.
4 (1999) 1–6.
[29] H. Kholafaei, C.S. Lee, F. Haghighat, Full-Scale test method for the performance
evaluation of in-duct filters with mixture of VOC, Clima Conference, Turkey, 2010.
[30] R.J. Huang, T. Demirel, T.D. Mcgee, Calculation and interpretation of surface free
energy of wetting of E-glass by vapors, J. Am. Ceram. Soc. 56 (1973) 87-91.
[31] B.F. Delage, P. Pre, P.L. Cloirec, Effects of moisture on warning of activated carbon
bed during VOC adsorption, J. Environ. Eng. 125 (1999) 1160-1167.
[32] L. Cao, Z. Gao., S.L. Suib, T.N. Obee, S.O. Hay, J.D. Freihaut, Photocatalytic
oxidation of toluene on nanoscale TiO2 catalysts: studies of deactivation and
regeneration, J. Catal. 196 (2000) 253–261.
[33] J. Matos, A. Garcia, J.M. Chovelon, C. Ferronato, Combination of adsorption on
activated carbon and oxidative photocatalysis on TiO2 for gaseous toluene
remediation, Open Mater. Sci. J. 4 (2010) 23-25.
[34] L. Zhong and F. Haghighat, Modeling and validation of a photocatalytic oxidation
reactor for indoor environment applications, Int. J. of Chem. Eng. Sci. 66 (2011)
5945-5954