PHOTOCATALYTIC DESTRUCTION OF WATER POLLUTANTS USING A TIO2 FILM IN

PET BOTTLES

Manuel Heredia
John Duffy
Energy Engineering Program, University of Massachusetts Lowell
[email protected]
[email protected]

ABSTRACT its simplicity; the advanced photooxidation process provides

its effectiveness in the degradation of organic and chemical
The performance of the PET bottles coated with a TiO2 was pollutants.
studied in this work. These bottles can be used for solar
water disinfection because TiO2 combines the action of The approach of this technology consists in the use of
solar disinfection (SODIS) and advanced oxidation process recycled PET (Polyethylene Terephthalate) bottles that are
improving the effectiveness of these technologies (Litter et coated with a film of TiO2. These bottles are used as
al., 2005). The effect of different parameters such as initial photocatalytic batch reactor where the water can be treated
contamination, solar radiation, etc. is analyzed using in the sun for some hours to improve its quality (Litter et al.
kinetics models (Salih, 2003; Horie et. al, 2005) that are 2005).
validated by experiments. The contaminants studied are
bacteria (E. coli) and arsenic. The behavior of the PET Although this new method is an important contribution, the
bottles performance is simulated with actual weather data influence of different factors on its performance needed
from remote sites in Peru. some investigation (temperature, solar radiation, ambient
temperature), the PET leaching effect to the water after
some hours of exposition to the sun, the effect of this
1. INTRODUCTION technique on the resultant water (taste, smell, appearance,
etc.), the proper water storage time without recontamination,
Water, although vital for supporting life, is a major threat to and the actual feasibility of using and manufacturing the
human health. According to the World Health Organization, bottles in villages by local people.
1.8 million of people die because of diarrhea per year and
88% of these cases are related to the ingestion of In order to investigate these factors, the University of
contaminated water, lack of sanitation and bad hygiene Massachusetts Lowell, with its project Village
practice (World Health Organization [WHO], 2006.) Empowerment, has performed some experiments in isolated
Furthermore, almost 1/6 of World’s population do not have Peruvian villages in the Andes. Bottles coated with TiO2
access to safe drinking water, most of them in the third were tested in the project trip to Peru in June, 2006, and the
World. Meanwhile the construction of water purification results have been considered in the development of this
plants is not suitable sometimes in the case of developing work. In addition, mathematical models of water
countries; developed countries can suffer natural disasters disinfection methods that use TiO2 as catalyst (Chang et al.,
that can cut the drinking water supply to a lot of people. For 2000; Li et al., 1996.) are used in order to analyze the
these reasons, scientists and engineers have been individual effect of several factors and predict the
investigated several water purification methods. performance of this method under real conditions. Finally,
the mathematical model has been validated with
Litter et al. (2005) have developed a technique that experiments performed at the University.
combines both SODIS and photocatalyst methods in order
to use their advantages. Meanwhile SODIS contributes with
2. SOLAR DISINFECTION TECHNOLOGIES With quantum energy in Planck’s equation, the energy of a
photon is described by:
2.1.SODIS
c
Solar Disinfection (known as SODIS) is an ancient ε = h⋅ (1)
technology that uses the effect of solar UV radiation for the λ
inactivation and destruction of microorganisms in drinking
water. The technology is as simple as filling a plastic bottle Where ε is energy in Joules, c is speed of light (3 X 108
with water and exposing it to solar radiation for a m/s), λ is wavelength and h is Planck’s constant (6.625 X
determined time. 10-34 J-s). Using equation (1) and knowing that anatase
TiO2 has a bandgap of 3.25 eV (Syarif et al., 2002), we find
Other advantages of this technologies are the low cost (it that the maximum wavelength that would produce an effect
can use recycled bottles) and the avoidance of chemicals, on TiO2 is 382 nm.
such as chlorine. Although its effectiveness has been proved
in different places (EAWAG & SANDEC, 1998), this Now when the catalyst is activated by photons, negative
method is sensitive to the volume of water to be treated, electrons and positive holes are formed:
initial concentration of pollutants, water turbidity, solar
radiation, the amount of oxygen dissolved in the water and
the container used.
(
TiO2 + hv → TiO2 ⋅ ecb− + holevb+ ) (2)

Polyethylene terephtalate (know as PET) is a material used The most common reaction is the recombination of
in the manufacturing of plastic bottles and contains electrons and holes on the surface of the catalyst, reducing
stabilizers that protect the plastic against the UV effects. the quantum efficiency of the process (Bandala, Arancibia-
Due to its low weight, transmissivity of UV-A radiation and Bulnes, Orosio & Estrada, 2004).

( )
chemical stability, it is preferred as the material for SODIS
treatment. PET can be used as bottles or as plastic bags. TiO2 ⋅ ecb− + holevb+ ⎯recomb
⎯⎯→ TiO2 + heat (3)

2.2. Advanced Oxidation Process-TiO2 Other electrons can react with reducible species such as
oxygen molecules. Positives holes can oxidize species
Titanium dioxide (TiO2), also know as titania, is a catalyst reducing the possibility of recombination (Son, Lee, Cho
used in the advanced oxidation process in water and Lu; 2004)
disinfection. It has been tested for the reduction of
hazardours organic chemicals and some inorganic pollutants The reactions that yield additional hydroxyl radicals may
(Bissen, Vieillard-Baron, Schindelin and Frimmel, 2001). It continue with other reactions. The hydroxyl radical is a very
is available in three different crystalline structures: anatase, powerful oxidizing agent which can oxidize organic
rutile and brokite. For photocatalytic purposes just anatase pollutants, directly or through intermediate compounds
and rutile are used because of their optical properties, (Goswami; 2001).
opacity and durability. Although rutile TiO2 has been
considered as photocatalytically inactive, it has been found 2.4. Factors Affecting UV Inactivation
that Degussa P-25, a commercial photocatalyst that has a
proportion anatase/rutile of 3-4, has the best performance in 2.4.1. pH and Temperature
photocatalysis (Sun and Smimioti,2003). In the case of
bacteria inactivation, It seems that the different For bacteria, the rate of disinfection is considered
characteristic of TiO2, such as the surface area and the independent from pH values, if it does not change
particle size, do not influence their bacterial inactivation drastically during the process (Rincon & Pulgarin; 2003).
efficiencies (Gummy, Rincon, Hadju and Pulgarin, 2005). Cho, Chung, Choi and Yoon (2004) established that when
pH varies from 5 to 9, reaction rates could be considered
2.3. TiO2 Photocatalytic Reactions independent of this characteristic. In the case of the effect of
temperature effects on bacteria disinfection, it was found
When a photon is absorbed by a material, an electron is that they are more susceptible to oxidative radicals at higher
released. In order to do that, the energy of the photon should temperatures (Cho et al.; 2004), where each microorganism
be greater than the bandgap of the material. Then, if the has its own resistance to the chemicals.
resultant electrons and positive holes react with the
surrounding molecules causing chemical reactions, a Temperature has also its own germicidal effect by itself.
photoreaction takes place (Goswami et al., 2000). Fujikawa, Ushioda and Kudo (1992) assumed that the effect
of temperature over bacteria can be described for a first oxygen concentration, temperature and any factor that could
order expression: interfere with the catalyst reaction.

dC 3.1.Langmuir-Hinshelwood Model
− = k (T ) ⋅ C (4)
dt The Langmuir-Hinshelwood model has been used to
describe several kinetic processes with good accuracy,
k (T ) is constant rate that is function of temperature. especially with chemical pollutants. This model has the
following relationship (Goswami, 2000):
2.4.2.Dissolved Oxygen
dC K 1 ⋅ K 2 ⋅ C
− = (6)
Bactericidal action of sunlight on water is dependent on the
amount of molecular oxygen (O2) dissolved in water. TiO2
dt (1 + K 2 C )
photoreacts with oxygen contented in O2 and H2O, therefore
the lack of sufficient supply of oxygen would reduce the Where C is the concentration of contaminant solutate in
reaction rates (Sagawe, Brandi, Bahnemann & water, K1 is the reaction constant, and K2 is the equilibrium
Cassano,2005); although photoreactions have been observed adsorption constant. When C tends to be 0 (low
under absence of dissolved oxygen . concentration), the expression (6) can be simplified as:

According with Herrera et al. (1999) to swirl or shake water dC

would be enough to provide enough oxygen to improve the
− = K1 ⋅ K 2 ⋅ C = K ⋅ C (7)
dt
efficiency of the reactions. After the reactions start and
dissolved oxygen is consumed, the needed molecular Where K can be understood as a total reaction constant that
oxygen can be absorbed from the surrounding air that is in is proportional to the intensity of the radiation
contact with the water.

In order to analyze de photoreaction rate of dissolved

K = k⋅I (8)
oxygen, the Langmuir- Hinshelwood model can be used:
and finally the following expression is obtained:
a ⋅ b ⋅ C DO C
−r = (5)
= e − k ⋅D (9)
(1 + bC DO ) C0

Where k is the new reaction constants, I is UV radiation

intensity, t is time of exposition, D is UV dose and Co is the
where r is the reaction rate, a and b are equation constants initial solute concentration. The expression (9) has been
and C DO the concentration of dissolved oxygen used to describe disinfection processes in municipal
wastewater with good accuracy (Li et al., 1996 ).
2.4.3. Turbidity
3.2. Multitarget Single Hit (MSTH) Model
Turbid water attenuates more radiation, diminishes the
amount of dissolved oxygen, and increases the amount of The Langmuir-Hinshelwood model loses its accuracy when
organic elements that are affected by the oxidative radicals. deals with microorganisms such as E. coliform or S.
Temperature has the main germicidal effect on high cerevisiae. When the disinfection process starts,
turbidity water. EAWAG and SANDEC (1998) recommend microorganisms have an initial resistance against the
the use of sedimentation or filtration to separate solids from germicidal effect of UV; the higher the concentration, the
the water when turbidity is higher than 30 NTU. greater the resistance. In 1955 Douglas Lea develop a model
for cell killing by radiation called “target theory of cell
killing” that was experimentally validated for solar
3. MATHEMATICAL MODEL disinfection by Salih in 2000. Later, Salih (2003) came up
with an equation that would describe in a better way the
There have been several attempts to create a kinetics model behavior of microorganisms such as E. coli under the effects
of organic pollutants destruction. In order to be accurate and of UV radiation. MTSH can be expressed in the following
more general, models should take into consideration pH, way:
C
C0
(
= 1 − 1 − e − Kx1
'
)x2
(10) 4.1 Destruction of Bacteria

Initial bacteria (E. coli) concentration was set above 10

where x2 is the target number of viable cells (cells that cell/ml using Hach’s sulphur bacteria incubation test
'
should be inactivated) and x1 can be expressed as: (HACH Cat. 261076). The tests were performed at different
μ d
times at University of Massachusetts Lowell on the roof
− ⋅ where water samples were taken, and radiation, temperature
ρ ρ
x = D×e
'
1 (11) and time were registered. PET bottles of 600 ml were
μ d coated over half the area from top to bottom with TiO2
where D is UV dose and the expression ⋅ represents the
ρ ρ according to the procedure of Litter et al. (2005) wherein
attenuation of the environment that transmits UV radiation. HClO4 was used to acidify the coating and help it adhere to
EAWAG and SANDEC (1998) presents a table with the the bottle wall.
reduction of UV-A as function of turbidity.
In order to compare the performance between coated and
3.3.Series Event Model non-coated PET bottles, both bottles were tested at the same
time under the same conditions. To determine the water
This model has been developed to analyze the effect of sample recollection time, E. coli disinfection simulation was
TiO2 in water disinfection from bacterial contaminants. In used, selecting the time when the predicted bacteria
accordance with this model, the following assumptions were concentration reaches 10, 1 and 0.1 cells/ml. These bacteria
taken. concentrations were chosen because of the dilutions of
samples that were used (1, 1/10, 1/100, 1/1000) to test
The cell deactivation obeys a second-order reaction between presence of bacteria.
cells and oxidative radicals, and the death of a cell is caused
by n times reactions. The concentration of oxidative radicals 4.1.1. Effect of UV Radiation Intensity
is constant under quasi-steady state. Horie, David, Taya and
Tone (1996) proposed the resultant equation: Because the test was performed using natural UV radiation
and the data presented are average values of the data
C n −1
= exp(− k '⋅t ) ⋅ ∑
(k '⋅t ) i
(12)
obtained during September and October 2006 at the
University in Lowell from 12:00 p.m. to 4:00 p.m.
Co i =0 i! The following data was found when coated and non-coated
bottles were used in bacteria disinfection experiments.
where k ' can be expressed as follows

k ' = k r C OX (13)

where k r is a rate constant and C 0 X is concentration of

radicals. Furthermore, C 0 X can be expressed in the
following equation (Horie et al., 1996):

A× I
C OX = kφ ⋅ aT , 0 ⋅ (14)
k OX ⋅ VL

where kφ is a quantum yield per specific area, aT ,0 is

specific area, A is area exposed to radiation , I is
Fig. 1 E. coli concentration after using non coated bottles
radiation, k OX is the apparent decomposition rate constant
(*) data (-) model
of oxidative radicals and VL is volume of test solution.
From figures 1 it is obvious that the concentration of
bacteria has not decreased. It is important to point out that
4. RESULTS the UV radiation and the time of exposure are less than the
recommended values for solar disinfection.
 4.1.2.2 Afternoon Test

The tests and simulations were performed in the afternoon

from 12 p.m. to 4 p.m.

Fig. 2 E. coli concentration after using coated bottles with

TiO2 film (*)data (-) model

Bottles that were coated with TiO2 film present better

performance destroying bacteria than ones that are not
coated. The final concentration of bacteria that was found is Fig 4 Destruction of bacteria during afternoon (*) data (-)
10 cell/100ml, although it does not imply that all bacteria simulation
cells were destroyed, this value is used as disinfection lower
limit because of the threshold of the bacteria presence test The dose accumulation and the disinfection rate are slower
that were used (test strips). in the morning as a consequence of the lower radiation that
reaches the bottle; after that, when it is 12:00 p.m., the
4.1.2 Effect of Initial Time of Exposure radiation increases, helping the final stage of the
disinfection. In the other hand, the dose accumulation has
To analyze the effect of increasing and decreasing radiation, the inverse behavior during the afternoon, so the rate of
water disinfection tests were performed at different starting disinfection gets slower which makes the time that is
hours, testing some bottles in the morning and some in the required for disinfection longer
afternoon. The analyzes was performed only for bottles that
were coated with 0.04 g of TiO2 and that were filled with 4.1.3 Effect of Initial Bacteria Concentration
water with initial E. coli concentration of 10 cell/ml.
The difficulty of experimentally obtaining high bacteria
4.1.2.1 Morning Test concentrations was overcome using mathematical
simulation.
The tests and simulations were performed in the morning
from 10 a.m. to 12 p.m.

Fig. 5 Required time for disinfection of water contaminated

Fig. 3 Destruction of bacteria in the morning (*) data (-) with E. coli at different UV radiation levels
simulation
According to figure 5, effect of initial bacteria concentration a consequence of the way that TiO2 is used. Because TiO2 is
is greater when the irradiation is low, because bacteria have fixed to the bottle’s wall and the photocatalytic oxidation
higher resistance at lower radiation. At the same time, the occurs near to the TiO2 particles, the increment of TiO2
required time of exposure has a logarithmic dependency of concentration only affects the portion of contaminated water
the initial bacteria concentration. that is in contact with the TiO2 film.

The curve that represents time of disinfection at 56 W/m2 of 4.1.5 Storage Time and Recontamination
UV radiation, which is the one that correspond to 1000
W/m2 of total solar radiation, shows that the time needed to In order to test the storage time for this technology, the
disinfect water that contains 1 cell of E. coli per milliliter is same water sample with high bacteria concentration (+
approximately 30 minutes, which is still a long period if we 1000/100ml) was tested in two different bottles: one was
consider the low bacteria concentration and the high UV coated with TiO2 and the other one was not coated. The
radiation. bottles were exposed to low solar radiation (winter season –
cloudy day) for few hours. After both bottles were tested, it
4.1.4 Effect of TiO2 Concentration in Coating Film was found that both sample contaminations were below the
threshold of the test (1000/100ml).After two days of dark
period (no exposure to sun), the results showed that the
bottle with TiO2 was keeping the low bacteria concentration
meanwhile the concentration of bacteria in the non-coated
bottle went back over the threshold (+ 1000/100ml).

4.2 Simulation of Bacteria Destruction in Cochapeti

The Village Empowerment Peru Project (U Mass Lowell)

has worked in Cochapeti since 1999, in the Peruvian
Andean Mountains at 3400 meter above sea level at 9.98° S.
latitude and longitude 77.65 ° W.

Radiation data was collected by the Peru Project team and it

Fig. 6 Required time for total water disinfection at different was used in the following simulation. This data was
TiO2 concentration obtained by data logger system that saved hourly average
values of total radiation (measured on the roof of the
medical clinic in Cochapeti), ambient temperature, and
information of the solar systems installed in the clinic. This
information corresponds to a period of two years.

The following assumptions were taken:

• The contaminant to be simulated is E coli.

• The bottle will be placed in a horizontal surface that
will not receive shadows from the surroundings.
• The effect of solar radiation is accumulative on the
process of killing bacteria.
• Initial bacteria concentration of 1000 cells/ml.
• The water temperature will be approximately the same
as the water temperature obtained under the same
Fig. 7 Disinfection of E. coli with different TiO2 ambient conditions. This assumption allows using the
concentrations (-) 0.02 g/bottle (-) 0.04 g/bottle (--) 0.06 data obtained in Lowell during fall (10-17 °C ambient
g/bottle (..) 0.12 g/bottle. temperature) in Cochapeti simulation during daytime
(8-20 °C ambient temperature). Therefore, during
The effect of TiO2 concentration is presented in the Figs. 26 warmer seasons, the model is conservative but it would
and 27. If TiO2 concentration is increased from 0.02 g/bottle be able to determine the minimum time to obtain safe
to 0.12 g/bottle, the time of exposure is reduced by 24 water at given conditions
minutes at initial bacteria concentration of 1000 cell/ml.
This little increment in the efficiency of the coated bottles is
According with the data, the month with lowest radiation is purification methods. However, other contaminants can be
March. The month with the highest radiation is October. removed with the photocatalyst such as arsenic.

We obtain the following simulation results for March: The simulations were performed using 8.00 a.m. as the
initial time for exposure, which means that during the first
hours, the accumulation of UV dose was slow. If the
exposure starts at noon, the time will be reduced to an hour
because of disinfection will take the advantages of the
higher radiation and the consequent faster UV dose
accumulation.

5. OTHER RESULTS AND DISCUSSION

The coating process used is very simple and results in TiO2

coated bottles that can be used several times (at least for six
times without obvious degradation of TiO2 film, although
more tests should be performed to study just this
characteristic). The performance and the quality of the TiO2
film obtained depend on the care shown in the coating
process. For that reason care is needed. Locally made
Fig. 8. Destruction of bacteria in Cochapeti in March with vinegar is not a substitute for the HClO4 acid.
(-) coated bottles (-) non coated bottles
The simulation results establish that under low radiation in
Fig. 8 shows that simple solar disinfection takes more than cloudy conditions (in Cochapeti), the time for water
one day in order to reach contamination levels of E. coli 10 disinfection using TiO2 is still very long (7 hours). The
cells/100ml. In contrast the bottle coated with TiO2 takes 7 increment of disinfection rate depends mainly on TiO2
hours if the exposure begins at 8:00 a.m. concentration and initial contaminant concentration.
Theoretically, the concentration of TiO2 can be increased as
much as it is needed, but this concentration would depend
on the solution used to create the film (% W/V and
repetitions of coating) and how much TiO2 can be attached
to the PET walls.

The time of exposure needed for purification depends on the

initial time of exposure, and it is recommended to start in
the morning when the advantages of increasing radiation
can be used.

The use of TiO2 film has advantages over SODIS although

still it may be less efficient than suspended TiO2 because
when we fix the TiO2, we reduce the area of catalyst that is
in contact to the contaminants.

TiO2 shows advantages in reducing the disinfection time for

Fig. 9 Destruction of bacteria in Cochapeti in October (-) bacterial contaminants and eliminating chemicals that
TiO2 film (-) SODIS simple solar disinfection does not degrade.

During the high radiation season, simple solar disinfection According to the results obtained in the storage time test,
takes more 4 and a half hours, while disinfection using recontamination in simple SODIS bottles was observed
coated bottles takes almost 3 hours making the process during dark conditions when water disinfection occurs under
faster by more than 90 minutes. Although it means a saving low radiation. It means that effective bacterial inactivation
of time, the required time for disinfection is still long if it is was not reached. In the other hand, TiO2 coated bottles do
compared with the use of chlorine or other water not present recontamination signal as stable bacteria
concentrations were observed during more than a week.
Safe handling of HClO4 and the tedious work required to (5)Malato, S., Cassady, J., Reed, R. H. and McGuigan,
coat the bottles (in order to avoid the formation of bubbles K.G.(2004), A novel TiO2-assisted solar photocatalytic
on the film surface) makes this process, although simple, batch-process disinfection reactor for the treatment of
not recommended for non trained people. This process could biological and chemical contaminants in domestic drinking
be recreated in the villages by medical technicians who have water in developing countries, Solar Energy,77, pp.649-675
some chemical handling training or by the villagers, after (6) Duffie, J. and Beckman, W.,(1991), Solar engineering of
safety rules training. Futures research should contemplate thermal processes, New York, NY: John Wiley & Sons
the use of other acids in the coating process that are less 1991
dangerous to handle. (7) Goswami, Y., Kreith F. and Kreider J.,(2000), Principles
of solar engineering, Philadelphia, PA: Taylor& Francis
Before the use of SODIS or TiO2 in any location, 2000
measurement of contaminants concentrations is needed to (8) Gumy, D., Rincon, A. G., Hajdu, R. and Pulgarin,
ensure that they can be removed with the process to be C.,(2005), Solar photocatalysis for detoxification and
utilized. In the case of arsenic, it is important to know if the disinfection of water: Different types of suspended and
water contains enough Iron (III) to react with arsenate. fixed TiO2 catalysts study, Sciencedirect. Retrieved
November 15, 2005, from http://www.sciencedirect.com
The quality and durability of the TiO2 on the bottles depend (9) Fujikawa H., Ushioda H. and Kudo Y., (1992), Kinetics
on quality of the work performed in the coating process. of escheria coli destruction by microwave irradiation,
Some bottles have not shown signs of TiO2 film Applied and Enviromental Microbiology, Mar.1992, pp.
deterioration after several uses (more than six times), others 920-924
after three uses. In general, TiO2 coated bottles can be used (10)Li, X. Z., ZHANG, M. and CHUA, H.,(1996),
until they lose their film. Disinfection of municipal wastewater by sensitized
photooxidation , Water Science Tech., 33(3, pp. 111-118)
Although water purified by TiO2 has not shown signs of (11)Litter, M. I., Blesa, M. A., Hidalgo, M., Apella, M. C.,
recontamination in more than a week, it is recommended to Jardim, W. F., Guimarães, J.R., Mansilla, H. D., Cornejo,
drink water soon anyway. Bad handling of the bottles (open L., Leal, M. T., Jiménez, A. E., Rodríguez, J. and Saunders,
the bottles and drink some water or touch the interior of the R., (2005), Solar energy based water potabilizationa: Low-
bottle with hands) can definitely change this time. As a cost technologies for isolated regions of Latin America and
general recommendation, if the water is not going to be the Caribbean. Unpublished manuscript, Comisión
consumed immediately, keep the exposure to solar radiation Nacional de Energía Atómica, Argentina
until the day or night it is going to be consumed. Further (12)Meichtry, J. M., Lin, H. J., de la Fuente, L., Levy, I. K.,
research is needed on safe storage times with this method. Gautier, E. A., Blesa, M. A. and Litter, M.A (2005), Low-
cost TiO2 photocatalytic technology for water potabilization
In general, the TiO2 approach is very promising for remote in plastic bottles for isolated regions. photocatalyst fixation,
areas for destruction of a variety of contaminants in, and Retrieved September 05, 2005, from Comisión Nacional de
storage of, drinking water. It appears that the coating Energía Atómica, Argentina Web site:
process can be undertaken locally with some training and http://www.cnea.gov.ar/xxi/ambiental/agua-
that recycled bottles can be used. Additional research is pura/publicaciones/sol-05-1097.pdf
needed to refine the process and to test for maximum (13)Rincon, A. and Pulgarin, C., (2004) Field solar E.coli
storage times. inactivation in the absence and presence of TiO2: Is UV
solar dose an appropriate parameter for standardization of
6. REFERENCES water solar disinfection? , Solar Energy, 77, pp.635-648
(14)Salih, F. M. (2000), Practical proof of the validity of
(1) Chang, H., Wu, N. and Zhua, F., (2000), Kinetic model target theory by simulating cellular targets, Medical
for photocatalytic degradation of organic contaminants in a Sciences, 2, pp. 81-86
thin-film TiO2 catalyst, Water Research, 34 (2, pp. 407-416) (15)Salih, F. M. (2003), Formulation of a mathematical
(2) Cho, M. , Chung, H. , Choi, W. and Yoon, J., (2004), model to predict solar water disinfection, Water Research,
Linear correlation between inactivation of E. coli and OH 37, pp. 3921-3927
radical concentration in TiO2 photocatalytic disinfection, (16)Swiss Federal Institute for Environmental Science and
Water Research, 38 , pp. 1069-1077 Technology and EAWAG's Department of Water and
(3) Diffey, L.,(2002), Sources and measurement of Sanitation in Developing Countries (1998), SODIS
ultraviolet radiation, Sciencedirect. Retrieved July 15, 2006, Technical notes, (No. 1-17) Retrieved December 16, 2005
from http://www.sciencedirect.com from http://www.sodis.ch/files/notes.pdf
(4) Duffy, E. F. , Al Touati, F., Kehoe, S. C., McLoughlin,
O. A., Gill, L. W., Gernjak, W., Oller, I., Maldonado, M. I.,