Asuncion AcevedoEnvironmental Technologies Department,

Faculty of Environmental Sciences,Cadiz University, Avda. Republica Saharaui S/N,

11510-Puerto Real, Cadiz, Spain

Edward A. CarpioFaculty of Sciences,

National University of Engineering,P.O. Box 31-139,

Av. Tupac Amaru 210,Lima, Peru

Juan RodrıguezUniversity of Tarapaca,

Av. General Velazquez 1775,Arica, Chile;

Faculty of Sciences,National University of Engineering,

P.O. Box 31-139,Av. Tupac Amaru 210,

Lima, Peru

Manuel A. ManzanoEnvironmental Technologies Department,

Faculty of Environmental Sciences,Cadiz University, Avda. Republica Saharaui S/N,

11510-Puerto Real, Cadiz, Spaine-mail: manuel.manzano@uca.es

Disinfection of Natural Waterby Solar Photocatalysis UsingImmobilized TiO2 Devices:Efficiency in EliminatingIndicator Bacteria and OperatingLife of the SystemNatural water has been disinfected using TiO2 as the fixed catalyst incorporated in a home-made photoreactor, in which the dimensions and the design parameters are representativeof devices that are currently employed at larger scale. The catalyst was immobilized on theexternal surface of a cylinder of frosted glass situated in the longitudinal axis of a tubularglass reactor. Two alternative methods of immobilizing the catalyst on glass were studied:in the first, a commercial titanium oxide powder (AeroxideVR TiO2 P25) was mounted on apolymeric support; and in the second, it was applied by sol-gel deposition. Illuminationwas effected by installing the glass reactor in the irradiation chamber of a solar simulator.Under laboratory conditions, groundwater contaminated with cultured and wild bacteriawas treated photocatalytically, and the influence of the photolysis, the pumping, and thecatalysts was studied. The results obtained have demonstrated that the catalyst immobilizedin the interior of the photoreactor presents similar results, in the disinfection of E. coli, as0.5 g/l of TiO2 P25; and that, in 1.5 h approximately of simulated solar illumination (167kWUVA s/m2) on the sol-gel deposit of TiO2, it is possible to eliminate 100% of the bacteriacovered by international regulations in respect of water for human consumption. Withregard to the aging assay of the system, it was observed at 250 h of operation a reductionin the effectiveness of the disinfection process. At 0 and 250 h of operation, the percentagesof elimination of E. coli after 50 min of illumination were 100% and 99.5%, respectively.This reduction in the effectiveness of the process was due to the formation of a film ofcalcium carbonate adhering to the internal glass wall of the photoreactor, which is incontact with the liquid being treated, and to the presence of calcium carbonate precipitateson catalyst surface. [DOI: 10.1115/1.4005338]

Keywords: photoreactor, total coliform bacteria, Escherichia coli, intestinal enterococci,Clostridium perfringens, calcium carbonate

1 Introduction

One of the biggest problems with drinking water in the develop-ing countries is microbiological contamination. According toestimates made by the World Health Organization [1] there are1.1! 109 people in the world without access to water of drinkingquality. This leads to around 2.2! 106 deaths per year from diar-rhoea, most of these children less than 5 years old. The traditionalmethods of disinfection (ozonization and ultraviolet radiationusing lamps) are not applicable given their relatively high costs ofinitial installation and on-going operation [2]. In the case ofchlorination, the treatment produces adverse changes in the tasteand smell of the drinking water [3]; it is not effective in killingresistant microorganisms such as Cryptosporidium parvum [4];and undesirable by-products, trihalomethanes, are generated bythe disinfection process [5,6]. There is, therefore, considerablescientific and health/welfare interest in the development of alter-native methods for the disinfection of natural water using low costtechnologies that are simple and reliable.

Solar photocatalysis with TiO2 is one of the technologies thatmay be very useful in small or isolated rural communities whensolar illumination can be used. It has the considerable advantageof having a zero operating cost since it uses sunlight; also the pho-tocatalyst (TiO2) can be immobilized in the interior of the reactorand so does not need to be added or applied repeatedly. The basisof the process is that, when TiO2 is illuminated with radiation ofthe correct wavelength, that is, of energy higher than the band gapenergy of the semiconductor (h!>Ebg" 3.2 eV in the case of ana-tase TiO2), valence band holes (h#vb) and conduction band elec-trons (e$cb) are generated [7]:

TiO2 # hv%k<385nm&! e$cb # h#vb

The photogenerated valence band holes (h#vb) and conduction bandelectrons (e$cb) may recombine, and the absorbed energy is liber-ated in the form of heat; alternatively, they make their separateways to the surface of the TiO2, where both may react with thespecies adsorbed on the surface of the catalyst, to produce highlyoxidative species [8,9], i.e., the hydroxyl radical (HO'), superox-ide anion (O2

$), perhydroxyl radical (HOO'), and hydrogenperoxide (H2O2). Although the mechanism of heterogeneous pho-tocatalysis to produce reactive oxygen species (ROS) is well

Contributed by the Solar Energy Division of ASME for publication in theJOURNAL OF SOLAR ENERGY ENGINEERING. Manuscript received April 14, 2011; finalmanuscript received October 7, 2011; published online November 29, 2011. Assoc.Editor: Wojciech Lipinski.

Journal of Solar Energy Engineering FEBRUARY 2012, Vol. 134 / 011008-1Copyright VC 2012 by ASME

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

known, the mechanism of photocatalytic microbial inactivation isnot completely understood.

Basically, a microorganism consists of organic compounds,and exposure to ROS generated by photocatalysis results in oxi-dative damage to its cellular components [10]. It is normally andgenerally accepted that the 'OH radicals are the most active ROSfor both the oxidation of organic substances and the inactivationof bacteria and virus [9,11]. However, Kikuchi et al. [8] proposedthat the actual lethal agent is H2O2, thus a cooperative effect withother oxygen species, particularly, in the long-range bactericidaleffect. The attack begins on the cell membrane of the microor-ganisms adsorbed on the catalyst, specifically by the peroxy oxi-dation of the unsaturated phospholipids [12], with the subsequentloss of fluidity and increase in ion permeability [13]; this facili-tates the oxidative attack on the internal cellular components andresults in cell death [14]. According to Matsunaga et al. [15], theinhibition of respiratory activity caused by the oxidation of theCoenzyme A is the main cause of the death of the cells. The wayphotocatalysis TiO2/ultraviolet A radiation (UVA) works herestops further bacterial regrowing to disinfection in contrast toUVA light being used alone where the microorganisms may reac-tivate [11].

Most of the information available on the possibilities of usingsolar TiO2 photocatalysis for disinfection is limited to results insynthetic waters, distilled water, or buffer solutions [16], withTiO2 slurries and commercial strains [17]: there have been fewstudies with strains of wild bacteria [18,19]. Generally, a reduc-tion in efficiency due to degradation of the catalyst after a pro-longed period of use has been reported. This reduced efficiency isthe result mainly of the loss of catalyst, the adsorption of by-products of the reaction and/or of soluble species in the water,with the consequent reduction of the active surface [19–21].

The principal objective of this work is to make a systematicstudy of solar photocatalytic water disinfection with supportedTiO2. For the experiments in the laboratory, urban wastewaterwas added to natural water to simulate the characteristics of con-taminated drinking water which can be found in areas wherethere is not enough water supply and sanitation infrastructure.The wild microorganisms selected for monitoring the efficacy ofthe process are the principal groups of indicator bacteria consid-ered in the regulations covering water for human consumption[22]. Further, since most of the experimental results founddescribed in the bibliography have been obtained during the firsthours of operation of the immobilized catalyst [23], an additionalobjective has been to study the performance of the system andthe structure of the photoreactor during a prolonged period ofcontinuous operation. The dimensions (diameter and thickness ofthe glass tube, diameter of the catalyst support) and the designparameters (recirculation flow rate, illuminated volume-totalvolume ratio) of the photoreactor utilized in the experiments, inbatch mode, are representative of devices that are currentlyemployed at larger scale [24]. In which, the way to concentratesolar radiation and easily pumped of water is consolidate byusing compound parabolic collector reactors having glass tubesconnected in line.

2 Materials and Methods

2.1 Assay Water. The groundwater with which the experi-ments were performed (Table 1) was taken from the well (Santis-cal, 95 m deep) that supplies the population of Arcos de laFrontera (Cadiz, Spain).

Natural water from underground source was used becausephotocatalytic inactivation of microorganisms in deionized watermay significantly differ from that in real water containing variousorganic and inorganic compounds [25]. It can be observed that theexperiments will be done in rigorous conditions since the ground-water characteristics (relatively high values of colour, turbidity,hardness, and solids in suspension) cannot be considered espe-cially suitable for heterogeneous photocatalysis.

To perform the disinfection experiments, the groundwater wasinoculated, in function of the particular assay, with the commer-cial strain of Escherichia coli ATCCVR 11229

TM

supplied by LGCStandards (American Type Culture Collection, Manassas, VA) orelse with urban wastewaters obtained from the pretreatment outletof the urban wastewater treatment plant of Conil de la Frontera(Cadiz, Spain).

In the case of the assays performed with the commercial strain,0.1% (v/v) of a concentrated solution of Escherichia coli (100NTU) was added to the groundwater to produce an assay water withan initial concentration of the order of 107 CFU/100 ml. To preparethe mother solution, 1 ml of the Escherichia coli ATTCVR 1229strain was inoculated in 30 ml of tryptic soy broth (TSB) (ScharlabS.L. Barcelona, Spain) and incubated at 37 6 1 (C in constant agita-tion and aerobic conditions for 12 h. Then, 1 ml of this culture wasadded to 30 ml of fresh TSB Broth and incubated in agitation for 4 hat 37 6 1 (C to obtain the culture in exponential growth phase.Next, the suspension was centrifuged at 3000 rpm for 10 min andwas washed twice with saline solution (NaCl 0.9%), after which thecentrifugation was repeated. Lastly, the new pellet with bacteria insuspension was diluted in 100 ml of sterile water, and a bacterialconcentration of 1010 CFU/100 ml was obtained [26].

In the rest of the assays, 0.2% (v/v) of urban wastewater (Table1) was added to the groundwater. Consequently, the disinfectionprocess was studied in a mixture of many different microorgan-isms and in the presence of not only organic contaminants, whosepresence is known to have a negative effect on photocatalytic dis-infection [25], but also other interfering substances. These assayswere performed by determining indicator microorganisms (totalcoliform bacteria, Escherichia coli, intestinal enterococci andClostridium perfringens) included in the regulations relating thequality of water for human consumption [22].

2.2 Experimental Device. The experimental setup for thephotocatalytic reactions is represented in Fig. 1.

The experimental device employed in the assays consists of atubular reactor of borosilicate glass, 200 mm long, 44 mm diame-ter, and 1.6 mm thick, operating in a closed recirculating circuitdriven by a peristaltic pump (Percon N-M, JP Selecta), with a600 ml stirred reservoir tank of Pyrex glass. The catalyst, TiO2, isimmobilized on the external surface of a cylinder of frosted glass(30 mm external diameter and 250 mm length) situated in the

Table 1 Characteristics of the groundwater employed in theexperiments

GroundwaterWastewater

(average values)

Conductivity (lS/cm) 431 3780Colour (mg/L, Pt/Co scale) 26 350Turbidity (NTU) 0.06 3.89pH 7.79 7.70Chlorides (mg/l) 35.5 1092Sulphates (mg/l) 6 142.9Phosphates (mg/l) <0.01 4.91Total hardness (mg/l CaCO3) 246 267Dry residue (mg/l, 180 (C) 269 286Nitrate (mg/l) 16.28 10.75Ammonium (mg/l) 0.05 451.2Nitrites (mg/l) <0.01 9.11Oxidizability (mg O2/l) 2.8 —Chemical oxygen demand (mg O2/l) — 1128Calcium (mg/l) 85.3 87.5Magnesium (mg/l) 3.24 65.95Total suspended solid (mg/l) 6 860Total coliform bacteria (CFU/100 ml) 0 2.9 ) 108

Escherichia coli (CFU/100 ml) 0 1.3 ) 107

Intestinal enterococci (CFU/100 ml) 0 1.5 ) 106

Clostridium perfringens (CFU/100 ml) 0 8.1 ) 105

011008-2 / Vol. 134, FEBRUARY 2012 Transactions of the ASME

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

longitudinal axis of the glass reactor, which has a useful volumeof 97 ml. Experiments have been carried out using a total workingvolume of 440 ml and a recirculation flow rate of 600 (ml/min),which gives a residence time in the irradiated photoreactor of9.7 s. Illumination was carried out by introducing the glass reactorinto the irradiation chamber of Suntest CPS# (Atlas MaterialTesting Technology LLC) equipped with a xenon lamp (Atlas Xe1500 B.-56001794 E2) with wavelenght in the range 290–400 nmaccounting for 5–7% of total light emitted. As well as being al-ready equipped with the optical filter “coated quartz glass,” the so-lar simulator also used an additional filter known as “UV specialglass (Ref. 56052371, Atlas)” in order to filter the excess of radia-tion UVC and achieve the simulation of solar global radiation out-dors (daylight). The values for total and UVA irradiance in theirradiation chamber were 500 W/m2, corresponding [27] to 50%full solar irradiance (light intensity of the midday equatorial solarradiation) and 30 WUVA/m2 (320–400 nm, UV light meter PCE-UV34, PCE Group), respectively.

Data have been expressed to represent illumination time, t,which is defined as the amount of time each fluid element spendsin the illuminated portion of the Pyrex photoreactor. Illuminationtime (t) is calculated by the quotient (Villum texp)/Vtotal, whereVillum is the illuminated volume of the photoreactor (97 ml),Vtotal is the total volume (440 ml), and texp is the real experimentaltime.

2.3 The TiO2 Catalyst. The catalyst employed was Aero-xideVR TiO2 P25, supplied free of charge by Evonik DegussaCorporation-Aerosil (Frankfurt, Germany). The product, of 99.5%of purity, has an approximate composition of 82% anatase and18% rutile, a specific surface area of 50 m2/g, a mean particle sizeof 21 nm and a pH zero point of charge around 5.5.

To immobilize the titanium dioxide on the cylindrical supportof frosted glass (/" 30 mm) two different procedures were fol-lowed: polymeric support and deposition by sol-gel on glass. Theaim was to utilize simple and accessible materials and equipment.For the fixing of the TiO2 by means of polymeric support, a solu-tion was prepared comprising TiO2 in powder (AeroxideVR TiO2

P25) and expanded polystyrene from packaging 1:5 w/w in ethylacetate (99.7%, Merck KGaA) at 12% (weight/volume); this wasthen applied using a spray gun onto a polypropylene/polyethylene(PP/PE) membrane, with a ratio Vsprayed/Area of PP/PE mem-brane of about 1/15 ml cm$2. Once the film applied on the PP/PE

membrane was dry, the membrane was wrapped around the glasscylinder (30 mm E.D.) with Teflon tape to the tube ends. The sol-gel deposition was carried under an extractor hood by manual dip-coating (3 cycles of 30 s in, 30 s out) of the frosted glass cylinder,previously heated to 60 (C, in a 30/70 (vol/vol) mixture of tita-nium (IV) isopropoxide (97%, Sigma-Aldrich, Inc.) and Isopropa-nol (99.9%, Merck KGaA). The uncontrolled hydrolysis wascarried out using atmospheric moisture (temperature 24 (C, rela-tive humidity 75%). Coatings were dried at 60 (C for 10 h andthen calcined at 350 (C for 5 h. The oven temperature was raisedto the desired temperature at a heating rate of 5 (C min$1. Onceboth supports had been prepared, they were immersed in an ultra-sound bath to remove any titanium dioxide that was only weaklyfixed or deposited.

2.4 Analytical Methods. The membrane filtration methodwas employed for the detection of all the microorganisms studied;each sample was analyzed in triplicate or quadruplicate and thevalue assigned meet the requirement of a coefficient of variation(CV) of less than 30%; the results were expressed as CFU/100 ml.All the material and the solutions were sterilized before the analy-sis. In the initial moments of each assay serial dilutions of thesamples were performed. The assays finished when the result offiltering 20 ml of sample was below the limit of detection (2 CFU/100 ml). A chromogenic medium, Colinstant (Scharlab), based onthe detection of two enzymatic activities, b-D-Glucoronidase andb-D-Galactosidase, was utilised for the detection and quantifica-tion of the total coliform bacteria and Escherichia coli. The plateswere incubated at 37 (C for 24 before counting. The ISO 7899-2:2000 method of membrane filtration [28], which employs Sla-netz Bartley medium (Scharlab), was used in the detection andcounting of the intestinal enterococci; the method utilized todetermine the Clostridium perfringens was that described inDirective 98/2083/CE [29]: incubation in anaerobiosis in m-CPmedium (Scharlab) for 24 h at 44.0 (C and subsequent confirma-tion of the colonies grown on the membrane with vapours ofNH4OH.

The pH and temperature were measured using a portable Crison507 pH meter (Crison Instruments, S.A.), and dissolved oxygenusing a Crison OXI 45 P oxymeter (Crison Instruments, S.A). Forthe physicochemical characterization of the groundwater and theurban wastewater employed in the assays, standard methods(APHA-AWWA-WPCF, 1992) were applied. The analysis of themetals was performed after filtering the samples (0.45 lm) andacidifying them (pH 2) by means of inductively-coupled plasmaatomic absorption spectroscopy, ICP-AAS (Iris Intrepid, ThermoFisher Scientific, Inc). Scanning electron microscopic (SEM)analysis was performed with a FEI Quanta 200 SEM Microscopeequipped with an EDAX Genesis for energy dispersive X-raymicroanalyses (EDX). Scanning transmission electron microscopy(STEM) analysis was performed with a JEOL 2010 F STEMMicroscope equipped with a high angle angular dark field(HAADF) detector and INCA Link equipment for microanalysis.X-ray diffraction (XRD) patterns were obtained using a BrukerD8Advance diffractometer with a secondary monochromator ofgraphite and with a Cu Ka radiation. To determine the parametersof the specific surfaces, the BET method was used in an automaticanalyser (Micromeritics ASAP 2020 model) which works usingthe ASAP 2020 software.

3 Results and Discussion

Experiments with the commercial strain of Escherichia coliATCCVR 11229

TM

(Sec. 3.1) and with microorganism from urbanwastewaters (Sec. 3.2) were done by duplicate and each samplewas analyzed in triplicate to determine the concentration of micro-organisms studied. The results shown in the aging assay (Sec. 3.3)are the result of one experimental run and each sample was ana-lyzed in quadruplicates.

Fig. 1 Experimental device employed in the disinfectionexperiments

Journal of Solar Energy Engineering FEBRUARY 2012, Vol. 134 / 011008-3

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

3.1. Experiments With the Commercial Strain ofEscherichia coli ATCCVR 11229

TM

: Selection of the Support ofthe TiO2. Figure 2 shows the results of the disinfection assaysperformed on groundwater inoculated with the commercial strain.The objective of this series of experiments was to determine thedisinfection performance of the device under various conditions:in darkness with and without pumping; irradiating without cata-lyst; and applying the solar heterogeneous photocatalysis withTiO2 in suspension, and with TiO2 immobilised in the two differ-ent—ways described—by sol-gel deposition and by polymericsupport.

It was observed that, in the experiment conducted in darkness,with and without recirculation pumping and in absence of the cat-alyst (TiO2), the number of colonies remains approximately con-stant; this is indicative of the viability of the strain of E. coliemployed in the research.

When it is illuminated with simulated solar radiation in the ab-sence of catalyst, the elimination of 99.5% of the microorganismsin 53 min (95.2 kWUVAs/m2) is detected; the concentrationdecreases from 1.6 107 to 8.7 104 CFU/100 ml, due to the jointaction of the bactericidal effect of the UVA radiation present inthe artificial sunlight employed and of the mechanical stresscaused by the recirculation pumping. These disinfection percen-tages obtained employing only solar radiation are similar thanthose found by Ubomba-Jaswa et al. [30]; in compatible condi-tions (recirculating batch, 108 CFU/100 ml of E. coli K-12 inwell-water using a compound parabolic collector with a concen-tration factor of 1), these authors calculated a value of UVA dosenecessary to achieve 99.9% disinfection of >108 kWUVAs/m2.

Despite the high percentage of disinfection observed, the resultcannot be considered satisfactory since the final concentration ofmicroorganisms exceeds that stipulated in the regulations fordrinking water, 0 CFU/100 ml [22]; moreover, it has been demon-strated that, when the disinfection is carried out by solar radiationonly (photolysis), the microorganisms are later reactivated [31].

The best results are obtained employing the catalyst, 0.5 g/l ofTiO2 in suspension, a concentration within the range consideredideal [14]. In these conditions, the 100% elimination of E. coli isobtained in 53 min. (95.2 kWUVAs/m2), its concentration decreas-ing from 3.8 106 to 0 CFU/100 ml. These results are similar tothose obtained by Rincon and Pulgarın [25]; these authors experi-mented in Pyrex glass bottles (50 ml) with drinking water (SuntestCPS# irradiation chamber, 400 W/m2) and obtained a 99.9%elimination of E. coli K12 (initial concentration 107 CFU/100 ml)in 45 min with 0.5 g/l TiO2 P25. Despite the good results, thisoption had to be discounted because the TiO2 in powder needs tobe removed from the water prior to consumption; also the results

were considered to represent the maximum degree of efficacy thatcan be achieved with the catalyst immobilised.

Comparing the results obtained with the catalyst in suspensionand immobilised, the same result was obtained with all these after53 min, CFU/100 ml< detection limit (2 CFU/100 ml), although aslightly faster rate of removal was observed with the catalyst insuspension. Therefore, to obtain a 5-log decrease of E. coli ATCCVR

11229TM

23 min (41.4 kWUVAs/m2), 28 min (50.4 kWUVAs/m2)y 32 min (57.6 kWUVAs/m2) were necessary with TiO2 in suspen-sion, immobilised by sol-gel and by polymeric support, respec-tively. This similarity in results between TiO2 in suspension andsol-gel, despite having much greater contact in suspension, couldbe associated with the difference between the value of BET surfacearea for the TiO2 fixed (95.03 m2/g for sol-gel procedure) and Aero-xideVR TiO2 P25 (50 m2/g).

The sol-gel deposition was selected, from the two ways studiedof immobilizing the catalyst; the three reasons for this preferenceare the simplicity of the fixing procedure, the total absence of sub-stances originating from the support released to the water (whichmight occur in the case of the polymeric support), and its mechan-ical resistance to the flow of water. Supported material presents agood adhesion to the glass substrate, which is very important forlong term applications.

3.2 Experiments With Microorganism From UrbanWastewaters: Effectiveness of the Treatment for RegulatedBacteria. These assays were performed employing groundwatercontaminated with urban wastewater, and their objective was todetermine the performance in the elimination of the variousmicroorganisms regulated under the standards, in the presence ofinterfering substances such as organic matter, the negative effectof which has been demonstrated [32]. The initial concentrations ofeach group of microorganisms were: 5.5 105 CFU/100 ml for totalColiforms; 3.3 104 CFU/100 ml for E. coli; 3.7 103 CFU/100 mlfor Intestinal enterococci; and 2.1 103 CFU/100 ml for Clostrid-ium perfringens. Figure 3 presents the results obtained, for thevarious bacteria studied, using immobilized titanium dioxide bysol-gel deposition.

First it is observed that, in the control assays, conducted indarkness, without agitation and in absence of the catalyst (TiO2),the concentration of all the microorganisms studied in totalremains approximately constant, which is an indicative of thegood state of the microorganisms.

With respect to the evolution of the concentration of Esche-richia coli, the elimination of 100%, 0 CFU/100 ml, is observedin 53 min; this is equivalent to a cumulative ultraviolet dose of

Fig. 2 Results of the disinfection assays with the strain E. coliATCCVR 11229

TM

Fig. 3 Results of the disinfection assays with wild strains ofE. coli, total coliform bacteria, intestinal enterococci andC. perfringens

011008-4 / Vol. 134, FEBRUARY 2012 Transactions of the ASME

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

95.2 kWUVAs/m2. These results for the wild E. coli (0.2% v/v ofurban residual water in groundwater) coincide completely withthe ones obtained in the current investigation for the commercialstrain of Escherichia coli ATCCVR 11229

TM

in groundwater, whichshows that the presence of small quantities of interfering substan-ces in natural water to be disinfected, will not affect nor stopexecution of the process.

These values for the time of illumination and the cumulativeUVA dose needed to reach the value of CFU/100 ml below thedetection limit (D.L.) are higher (79 min and 142.9 kWUVAs/m2)for total coliform bacteria and intestinal enterococci, which areinitially present at concentrations one order of magnitude higherand lower, respectively, than the initial concentration of E. coli.

The results for total disinfection obtained in the current investi-gation (500 W/m2) for E. coli (53 min, 441 Wh/m2) and total coli-forms (79 min, 655 Wh/m2) differ from those reported by Geloveret al. [23], who found that, on a sunny day (irradiance of morethan 1000 W/m2), it took only 15 min (250 Wh/m2) and 30 min(500 Wh/m2) of irradiation to inactivate E. coli and total coli-forms, respectively, applying disinfection with immobilized TiO2.The difference in results could be explained mainly because inour essays, the disinfection team works in recirculating batches(interrupted solar dose), to simulate a centralised disinfection fa-cility for a small comunnity, whereas Gelover et al.’s essays, [23]were undertaken in 2 l polyethylene terephthalate (PET) bottles(uninterrupted solar dose) to address the problem of disinfectionin an individual level (2 l capacity). It is a proven fact thatthe speed of disinfection is higher when the solar radiation isdelivered in an uninterrupted manner [30]. Also, the higher initialconcentration of microorganisms employed in our study (2 ordersof magnitude higher) may explain this difference; the reason isthat a longer time is required for bacterial inactivation when theinitial concentration of bacteria is higher [33].

Lastly, the bacteria found to be most resistant to the processof disinfection was Clostridium perfringens, for which 93 min(167 kWUVAs/m2) were needed to eliminate a concentration 1order of magnitude less than that of E. coli. Less effectiveness ofthe treatment to inactive C. perfringens is caused by its ability toform and create spores which make them much stronger due tothe fact that they can prevent permeability created by the afore-mentioned treatment [34].

As can be appreciated from the data reported here, 93 min ofillumination are needed in order to comply with the requirementset in the regulations covering water for human consumption, of0 CFU/100 ml for each of the microorganisms studied [22]; thistime of illumination is equivalent to a cumulative UVA dose andoverall dose of 167 kWUVAs/m2 and 2783 kWs/m2, respectively.

Disinfection rate constants were calculated based on first-orderkinetic model in which - ln (C )C0

$1) was plotted as a function ofillumination time; the slope of the line was the k, the first-orderinactivation rate constant. t90 is defined as the illumination timefor bacterial concentration to decrease by 90% and it is given by -(ln 0,1)/k. Dose90 is the irradiation dose required for the bacterialconcentration to decrease by 90% and it is the product of irradi-ance (500 kW/m2 or 30 kWUVA/m2) and illumination time (s).

The choice of first order reaction kinetics is based on the sim-plicity of this model and for being the most accurate one in termsof experimental results. Additionally, it is one of the most used

and widely recognised model in the photocatalytic disinfectionfield [35,36].

Table 2 gives the values of the first order kinetic constant foreach of the experiments performed with groundwater contami-nated with urban wastewater, together with the time and dose nec-essary to achieve 90% disinfection.

The values presented in the above table corroborate the findingthat the process described here is more effective, notably forE. coli, but also for total coliform bacteria, intestinal enterococciand C. perfringens, in that order. In other words, the kinetic con-stant is higher; shorter times are required to achieve 90% disinfec-tion; and the UVA dose required to achieve that result is lower.These findings are in accordance with the previously reportedresults [32] regarding the greater suceptibility of E. coli to photo-catalytic treatment, in comparison with the Enterococcus species;and, as this study has shown, the susceptibility of E. coli to thistreatment is considerably greater than that of C. perfringens.

In all these experiments, with both the commercial strain andbacteria originating from urban wastewaters, the initial tempera-ture of the assay water was 25 (C, and after 30 min. it reached34 6 1 (C; these values coincide with those reported in other stud-ies [25]. The pH at all times remained at 7.5 6 0.2; this range isadequate for the development of the microorganisms, and theeffectiveness of the process was not found to be altered withinthis range [25]. The concentration of oxygen, on average, droppedfrom 8.74 mg O2/l initially, to 7.02 mg O2/l at the end of theexperiment; this drop can be attributed to the increase of tempera-ture of assay water, since the saturation at all times was 100%; itcan be deduced from this that the agitation at the mixing unit com-pensates for the consumption of oxygen by the microorganismsand by the photochemical reaction itself.

3.3 Test of the Operating Life of the System. With theobject of determining the period of time for which the disinfectionequipment can be used without requiring maintenance, the equip-ment was operated continuously for 400 h at 30 WUVA/m2. Thistime would be equivalent in real conditions to 2 months disinfect-ing natural waters every day with the solar plant operated in thetested conditions (1.5 h of illumination time, or 6.8 h of real timeconsidering a quotient Villum/Vtotal" 0.22, is required diary toeliminate 100% of the microorganisms at 30 WUVA/m2). Thewater used in this operating life test was the same water as thatemployed in the first experiments, urban waste water diluted withgroundwater (Table 3). To assure that the microorganisms used in

Table 2 Kinetic constant, t90 and Dose90 for the assays performed with bacteria originating from urban waste waters

Escherichia coli Total coliform bacteria Intestinal enterococci Clostridium perfringens

Co (CFU/100 ml) 3.3! 10#04 5.5! 10#05 3.8! 10#03 2.1! 10#03

K (min$1) 0.566 0.237 0.165 0.102r2 0.999 0.976 0.981 0.980t90 (min) 4.1 9.7 13.9 22.6UVA Dose 90% (kWUVA s/m2) 7.3 17.5 25.1 40.7Overall dose 90% (MW s/m2) 0.12 0.52 0.75 1.22

Table 3 Characteristics of the assay water in the test of operat-ing life

t (h) COD (mg/l) TSS (mg/l) VSS (mg/l) E. coli (CFU/100 ml)

0 5 8 5 3.3 104

50 9 11 6 2.1 104

100 8 15 10 2.6 104

150 14 26 18 1.5 104

250 15 22 15 4.6 103

350 7 11 7 1.2 104

400 18 22 15 8.6 104

Journal of Solar Energy Engineering FEBRUARY 2012, Vol. 134 / 011008-5

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

the aging investigation were in perfect conditions, residual waterfrom the effluent of the wastewater treatment plant were takendaily to be used as inoculum, which explains the variety of resultsshown on Table 3 below.

This would be a rigorous test since its characteristics (relativelyhigh values of colour, turbidity, hardness, and solids in suspen-sion) cannot be considered especially suitable for heterogeneousphotocatalysis.

It can be observed (Table 3) that there are differences in thecharacteristics of the water tested, which must be due to the vari-ability of the urban wastewaters that were added daily to thegroundwater, since they were taken from the wastewater treatmentplant the same day that they were added. Figure 4 shows theresults obtained in the aging assay.

In the first place, it can be observed that when the equipmentwas utilized for the first time (neither the glass of the photoreactornor the film of TiO2 had previously been in contact with the waterto be disinfected) 10 min of illumination was required to eliminate99.2% of the E. coli present; 50 min of illumination was requiredto reach 100% elimination.

The process continued with this level of performanceunchanged until the equipment reached 150 h of continuous opera-tion; from that time until the completion of 350 h of operation

there was a decline in effectiveness. At 250 h of operation, thepercentages of elimination of E. coli after 10 and 50 min of illumi-nation were 92.2 and 99.5%, respectively, but at 350 h the percen-tages were 87.4 and 99.5%.

This reduction in the effectiveness of the process may be due tothe formation, clearly observable to the eye, of a whitish filmadhering to the internal glass wall of the photoreactor, which is incontact with the liquid being treated.

Figure 5(a) is a micrograph obtained by SEM of the materialdeposited, after the glass has been scratched, and supported on anadhesive strip of graphite. Examination by EDX of the elementalcomposition of the material deposited, Fig. 5(b), showed anatomic Ca/O proportion of 1.00/2.97, which confirmed the exis-tence of a film of calcium carbonate on the glass of thephotoreactor.

It is probable that this film of calcium carbonate is the cause ofreduced radiation, quantified at 17% less (by Quantitherm LightMeter, Hansatech), reaching the interior of the photoreactor; thiswould account for the observed decrease of the efficacy of the dis-infection process. To check this assumption, after 399 h of opera-tion in the test of operating life, the glass of the photoreactor wasreplaced by another clean glass with identical characteristics. Af-ter this replacement, the results that are shown for 400 h of

Fig. 4 Percentages of disinfection of E. coli after 10 and 50 min of illumination, in func-tion of the operating time of the equipment

Fig. 5 Deposit on the internal glass wall of the photo-reactor. (a) SEM micrograph of the deposit after being scraped off theglass; (b) general spectrum.

011008-6 / Vol. 134, FEBRUARY 2012 Transactions of the ASME

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

operation, in Fig. 4, were obtained: there was an almost total re-covery of the effectiveness of the process, to 99.1 and 99.9% dis-infection at 10 and 50 min of illumination, respectively. Despitethis, it was considered useful to study the formation of deposits onthe film of TiO2 with the object of weighting the contribution of

this factor to the problem of decreasing efficacy of disinfectionwith prolonged time of operation. Figure 6 shows the results ofthe EDX analysis by SEM of the clean frosted glass (support ofthe catalyst) and of this same support once the TiO2 has been fixedby deposition using the sol-gel procedure, before the test.

Fig. 6 (a) and (c) SEM micrographs and (b) and (d) EDX elemental composition obtained by SEM of the ground glass usedas support for the catalyst (1), and of this same support after the TiO2 has been fixed by deposition using the sol-gel proce-dure, before the test of operating life

Fig. 7 STEM Micrograph (a) and DRX diffractogram (b) of a fragment of the TiO2 deposited onthe surface of the catalyst

Journal of Solar Energy Engineering FEBRUARY 2012, Vol. 134 / 011008-7

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

The irregular aspect presented by the frosted glass surfaceemployed as the support for the TiO2, and the presence of clearlydefined angle edges, can be observed in Fig. 6(a). These featuresare the result of mechanically altering the surface of the glass inthe grinding process, with the object of generating increased sur-face area to which the layer of catalyst is applied. Figure 6(b)shows the general spectrum of the frosted glass support beforeapplying the sol-gel deposition of the TiO2; a typical compositionof glass, silicates, and the presence of minor components such assodium and aluminium, are observed.

Figure 6(c) shows the sol-gel deposition of the TiO2 on glasssurface before the operating life test is performed. The presenceof an extensive network of fused crystallites of TiO2 is observed,with the presence of irregular-shaped titania aggregates arrangedscattered over the surface or incorporated within the film. Themost distinctive irregular-shaped titania aggregates have a diame-ter between 0.2 y 0.5 lm and are capable of forming much biggerparticles with heterogeneous distribution (Leica Quantimet 500.Leica, Hamburg, Germany).

The existence of this film of TiO2 is confirmed by smootherappearance presented by the angle edges in Fig. 6(c), comparedwith the sharper aspect they present in Fig. 6(a), and by the ele-mental composition by EDX performed at different points on thesurface, which showed the presence of titanium at all the points.The EDX analysis showed the following atomic percentage ratiosfor Ti/Si: 0.025 for the overall surface of the catalyst; 0.018 in thefilm of TiO2, at those points where the aggregates are notobserved; 0.029 where the aggregates of titanium are observedincorporated in the film; and 0.051 where the aggregates of tita-nium are disseminated on the film surface.

The structure of the extensive network of fused crystallitescan be observed in greater detail in the micrograph shown asFig. 7(a), obtained by STEM of a fragment of the TiO2 depositedon the surface of the catalyst; here the continuity of the crystalsand the existence of zones of varying thickness can be observed.

Figure 7(b) is the diffractogram (XRD) of the TiO2 depositedon the surface of the catalyst, which showed the presence of ana-tase only, anatase is the metastable crystal form of TiO2 expectedto be formed after sintering at 350 (C and normal pressure. Thediffractograms display a prominent peak due to the anatase (101)reflections, as well as small peaks assigned to anatase (004),(200), (105), (211) and (204) [37]. The crystalline domain wasestimated for each sample from the full width at a half maximum(FWHM) of (101) peak from ScherrerC

¸s equation, D" 0.9 k/b

cos h, with k" 1.540598 A the X-ray wavelength of the CuKaradiation, h the diffraction angle, and b the FWHM of the diffrac-tion peak. A crystalline grain size of 25 nm was calculated. Grain

domains are also presented in Fig. 7(a), which are part of the filmpresented in Fig. 6(c).

In Fig. 8(a), showing the surface of the catalyst after the agingassay, it can be observed the presence of precipitates on its sur-face; these could be responsible, in the long term (>400 h), for theloss of effectiveness in the disinfection process. These precipitatesoccupy approximately 13.42% of the surface of the support (LeicaQM500) and the mean particle size ranges from 2 to 30 lm. EDXanalysis of this surface (Fig. 8(b)) showed its elemental composi-tion of Ca/O/C to be in the proportion: 1.00/3.24/9.96; the sameCa/O ratio as on the internal surface of the photoreactor. This iscompatible with the presence of CaCO3 and with the existence ofcarbonated remains of the process of oxidation of the microorgan-isms and of the organic matter present in the assay water.

These results demonstrate that the presence of precipitates ofcalcium carbonate on the internal wall of the photoreactor and onthe catalyst is the principal factor involved in the aging of the dis-infection equipment. A way or suggestion to get rid of these pre-cipitates of calcium carbonate is the thorough cleaning of thephotoreactor using an acid dissolution made from the juice of spe-cific fruits which contain citric acid. In fact, citric acid dissolu-tions are used (1–2% w) to eliminate calcium carbonate fouling inindustrial applications [38], hence specific fruits can be used astheir juices contain high concentration of citric acid [39]. Never-theless, after 400 h of operation, not evidence of the catalysis leak-ing in the solution has been observed, ensuring the well use of thecatalyst for long periods at rural areas, characterized for the lackof technical facilities.

4 Conclusions

The results obtained at laboratory scale have demonstrated that,in the experimental conditions tested (30 WUVA/m2), it is possibleto eliminate 100% of the bacteria covered by international regula-tions in respect of water for human consumption, in approxi-mately 1.5 h of solar illumination, with the catalyst immobilisedin the interior of the photoreactor by sol-gel deposit.

The good results obtained on changing the glass of the solarphotoreactor, after a prolonged period of continuous operation,indicate the benefits of maintaining the equipment to keep theglass clean. Instead of changing the glass, the most appropriatemaintenance would be cleaning with an acid solution, from juiceof fruits such as lemon or lime, to remove the film of calciumcarbonate that can reduce its performance. In any case, this tech-nology may be of special interest in developing countries since, inaddition to its effectiveness, it does not involve consumption ofenergy or of reagents and therefore no operating costs for these

Fig. 8 SEM micrographs (a) and EDX elemental composition (b) obtained by SEM of the catalyst supported on ground glassafter the operating life test

011008-8 / Vol. 134, FEBRUARY 2012 Transactions of the ASME

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

items are incurred. Furthermore the equipment required is eco-nomic, simple to construct and operate and should have a longoperating life with a minimum of maintenance.

Acknowledgment

The authors are grateful to the Spanish Agency for InternationalCooperation for Development (Ministry of External Affairs andCooperation, Spain) for funding this research; the Central Servicesof Science and Technology of the University of Cadiz for thescientific collaboration provided; and the Andalusian Centre forMarine Science and Technology (Cadiz, Spain) where the experi-mental part of this research was carried out. J.R. wishes to thankthe UTA-MIDEDUC Convenio de Desempeno for financialsupport.

Nomenclature

ATCC " American Type Culture CollectionBET " a nitrogen adsorption technique used to measure the

specific surface area of a solid. Named after itsinventors Brunauer, S., Emmett, P.H., and Teller, E.

BOD5 " biologycal oxygen demand (mg O2/l)CFU " colony forming unitCOD " chemical oxygen demand (mg O2/l)CPC " compound parabolic collectorCV " coefficient of variation, the ratio of the standard

deviation to the mean multiplied by 100%D " the averaged dimension of crystallites, (A)

D.L. " detection limitDSMZ " German collection of microorganisms and cell

culturesE.D. " external diameter (mm)EDX " energy dispersive X-ray microanalysesEPS " expanded polystyrene

FWHM " the full width at half maximum intensity of the peakHAADF " high angle angular dark field detector

ICP-AAS " inductively-coupled plasma atomic absorptionsectroscopy

m-CP " Clostridium Perfringens Agar BasePE " polyethylenePP " polypropylene

ROS " reactive oxygen speciesSEM " scanning electron microscopic

STEM " scanning transmission electron microscopyTSB " tryptic soy brothTSS " total suspended solids (mg/l)

UVA " ultraviolet radiation with wavelengths between 320and 400 nm

VSS " volatile suspended solids (mg/l)XRD " X-ray diffractionZPC " zero Point of Charge, the pH at which adsorption of

the potential determining ions on the oxide (H# andOH$) is balanced

C " concentration of bacteria at instant t.C0 " initial concentration of bacteria Dose (kWs/m2),

Product of irradiance (kW/m2) and illumination time(s) Dose90, Dose required for the bacterial concentra-tion to decrease by 90% (kWs/m2) E. coli, Esche-richia coli

Ebg " band gap energy of the semiconductore$cb " conduction band electrons of the semiconductorh#vb " valence band holes of the semiconductor

Irradiance (kW/m2), the amount of radiation (kW)striking a given area of a surface (m2)

k " the first-order inactivation rate constant (min$1)r2 " correlation coefficient

texp " real experimental time (min)t90 " the time for bacterial concentration to decrease by

90% (min) t, Illumination time, amount of time each

fluid element spends in the illuminated portion of thephotoreactor (min)

Villum " illuminated volume of the photoreactor (ml)Vtotal " total volume of the experimental device (photoreac-

tor, tubes and mixing tank) (ml)b " the full width at half maximum (FWHM) of the dif-

fraction peak/ " diameter (mm)k " radiation wavelength (nm). The Scherrer formula:

the X-ray wavelength of the CuKa radiation (A)h " the diffraction angle (rad)

References[1] UN, 2003, United Nations World Water Development Report: Water for Peo-

ple, Water for Life, www.unesco.org/water/wwap/wwdr/wwdr1/table_contents/index.shtml.

[2] Munoz, I., Rieradevall, J., Torrades, F., Peral, J., and Domenech, X., 2005,“Environmental Assessment of Different Solar Driven Advanced OxidationProcesses,” Sol. Energy, 79(4), pp. 369–375.

[3] DWI, 2000, “DWI Consumer Market Research: Drinking Water Quality Reportof Public Perceptions,”www.dwi.gov.uk/consumer/marketr/cr2000.htm.

[4] Driedger, A. M., Rennecker, J. L., and Marinas, B. J., 2000, “Sequential Inacti-vation of Cryptosporidium parvum Oocysts and Ozone and Free Chlorine,”Water Res., 34(14), pp. 3591–3597.

[5] Bull, R. J., Krasner, S. W., Daniel, P., and Bull, R. D., 2001, “Health Effectsand Occurrence of Disinfection By-Products,” AWWA Research Foundationand American Water Works Association, Denver, CO.

[6] Schiliro, T., Pignata, C., Rovere, R., Fea, E., and Gilli, G., 2009, “The Endo-crine Disrupting Activity of Surface Waters and of Wastewater Treatment PlantEffluents in Relation to Chlorination,” Chemosphere, 75, pp. 335–340.

[7] Serpone, N., and Pellizzetti, E., 1989, Photocatalysis: Fundamentals and Appli-cations, Wiley, New York.

[8] Kikuchi, Y., Sunada, K., Iyoda, T., Hashimoto, K., and Fujishima, A., 1997,“Photocatalytic Bactericidal Effect of TiO2 Thin Films: Dynamic View of theActive Oxygen Species Responsible for the Effect,” J. Photochem. Photobiol.A, 106, pp. 51–56.

[9] Rincon, A. G., and Pulgarin, C., 2003, “Photocatalytical Inactivation of E. coli:Effect of (Continuous–Intermittent) Light Intensity and of (Suspended–Fixed)TiO2 Concentration,” Appl. Catal. B, 44, pp. 263–284.

[10] Huang, Z., Maness, P., Blake, D. M., Wolfrum, E. J., Smolinski, S. L., andJacoby, W. A., 2000, “Bactericidal Mode of Titanium Dioxide Photocatalysis,”J. Photochem. Photobiol., A, 130(2–3), pp. 163–170.

[11] Robertson, J. M. C., Robertson, P. K. J., and Lawton, L. A., 2005, “A Compari-son of the Effectiveness of TiO2 Photocatalysis and UVA Photolysis for theDestruction of Three Pathogenic Micro-Organisms,” J. Photochem. Photobiol.,A, 175, pp. 51–56.

[12] Sunada, K., Watanabe, T., and Hashimoto, K., 2003, “Studies on Photokillingof Bacteria on TiO2 Thin Film,” J. Photochem. Photobiol., A. 156 (1–3), pp.227–233.

[13] Wainwright, M., 2000, “Methylene Blue Derivatives-Suitable Photoantimicro-bials For Blood Products Disinfection?,” Int. J. Antimicrob. Agents, 16(4), pp.381–394.

[14] Rincon, A. G., Pulgarın, C., Adler, N., and Peringer, P., 2001, “InteractionBetween E. coli Inactivation and DBP-Precursors—Dihydroxybenzene Iso-mers— in the Photocatalytic Process of Drinking-Water Disinfection WithTiO2,” J. Photochem. Photobiol., A, 139(2–3), pp. 233–241.

[15] Matsunaga, T., Tomoda, R., Nakajima, T., and Wake, H., 1985,“Photoelectrochemical Sterilization of Microbial Cells by SemiconductorPowders,” FEMS Microbiol. Lett., 29(1–2), pp. 211–214.

[16] McCullagh, C., Robertson, J. M. C., Bahnemann, D. W., and Robertson, P. K.J., 2007, “The Application of TiO2 Photocatalysis for Disinfection of WaterContaminated With Pathogenic Microorganims: A Review,” Res. Chem.Intermed., 33(3–5), pp. 359–375.

[17] Fernandez, P., Blanco, J., Sichel, C., and Malato, S., 2005, “Water Disinfectionby Solar Photocatalysis Using Compound Parabolic Collectors,” Catal. Today,101(3–4), pp. 345–352.

[18] Navntoft, C., Araujo, P., Litter, M. I., Apella, M. C., Fernandez, D., Puchulu,M. E., Hidalgo, M. V., and Blesa, M. A., 2007, “Field Tests of the Solar WaterDetoxification SOLWATER Reactor in Los Pereyra, Tucuman, Argentina,” J.Sol. Energy Eng., 129(1), pp. 127–134.

[19] Rodrıguez, J., Jorge, C., Zuniga, P., Palomino, J., Zanabria, P., Solıs, J. L., andEstrada, W., 2010, “Solar Water Disinfection Studies With Supported TiO2 andPolymer-Supported Ru(II) Sensitizers in a Compound Parabolic Collector,” J.Sol. Energy Eng., 132, pp. 1–5.

[20] Ponce, S., Carpio, E., Venero, J., Estrada, W., and Rodrıguez, J., 2009,“Titanium Dioxide onto Polyethylene for Water Decontamination,” J. Adv.Oxid. Technol., 12(1), pp. 81–86.

[21] Rao, K. V. S., Subrahmanyamb, M., and Boule, P., 2004, “Immobilized TiO2

Photocatalyst During Long-Term Use: Decrease of its Activity,” Appl. Catal.,B, 49, pp. 239–249.

[22] WHO, 2008, World Health Organization Guidelines for Drinking-Water Qual-ity, 3rd ed., Vol. 1, Geneve, Switzerland.

Journal of Solar Energy Engineering FEBRUARY 2012, Vol. 134 / 011008-9

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

[23] Gelover, S., Gomez, L. A., Reyes, K., and Leal, M. T., 2006, “A Practical Dem-onstration of Water Disinfection Using TiO2 Films and Sunlight,” Water Res.,40(17), pp. 3274–3280.

[24] Malato, S., Blanco, J., Alarcon, D. C., Maldonado, M. I., Fernandez-Ibanez, P.,and Gernjak, W., 2007, “Photocatalytic Decontamination and Disinfection ofWater With Solar Collectors,” Catal. Today, 122(1–2), pp. 137–149.

[25] Rincon, A. G., and Pulgarın, C., 2004, “Effect of pH, Inorganic Ions, OrganicMatter and H2O2 on E. coli K12 Photocatalytic Inactivation by TiO2. Implica-tions in Solar Water Disinfection,” Appl. Catal., B, 51(4), pp. 283–302.

[26] Quek, P. H., and Hu, J., 2008, “Indicators for Photoreactivation and DarkRepair Studies Following Ultraviolet Disinfection,” J. Ind. Microbiol. Biotech-nol., 35(6), pp. 533–541.

[27] Bozzi, a., Yuranova, T., Guasaquillo, I., Laub, D., and Kiwi, J., 2005, “Self-cleaning of Modified Cotton Textiles by TiO2 at Low Temperatures Under Day-light Irradiation,” J. Photochem. Photobiol., A, 174, pp. 156–164.

[28] ISO 7899-2, 2000, “Water Quality—Detection and Enumeration of IntestinalEnterococci. Part 2: Membrane Filtration Method.”

[29] Directive 98/83/CE, Council Directive of 3 November on the quality of waterintended for consumption, DOCE no. L 330, of 5 December 1998.

[30] Ubomba-Jaswa, E., Navntoft, C., Polo-Lopez, I., Fernandez-Ibanez, P., andMcGuigan, K. G., 2009, “Solar Disinfection of Drinking Water (SODIS): AnInvestigation of the Effect of UVA Dose on Inactivation Efficiency,” Photo-chem. Photobiol. Sci., 8 (5), pp. 587-595.

[31] Rincon, A. G., and Pulgarin, C., 2004, “Field Solar E. coli Inactivation in theAbsence and Presence of TiO2: Is UV Solar Dose an Appropriate Parameter

for Standardization of Water Solar Disinfection?,” Sol. Energy, 77(5),pp. 635–648.

[32] Alrousan, D. M. A., Dunlop, P. S. M., McMurray, T. A., and Byrne, J. A.,2009, “Photocatalytic Inactivation of E. coli in Surface Water Using Immobi-lised Nanoparticle TiO2 Films,” Water Res., 43(1), pp. 47–54.

[33] Rincon, A. G., and Pulgarin, C., 2004, “Bactericidal Action of Illuminated TiO2

on Pure Escherichia coli and Natural Bacterial Consortia: Post-IrradiationEvents in the Dark and Assessment of the Effective Disinfection Time,” Appl.Catal., B, 49(2), pp. 99–112.

[34] Alvarez, I., Condon, S., and Raso, J., 2006, Pulsed Electric Fields Techno-logy for the Food Industry, Science and Business Media, New York, NY,pp. 97–129.

[35] Gumy, D., Rincon, A. G., Hajdu, R., and Pulgarin, C., 2006, “Solar Photocatal-ysis for Detoxification and Disinfection of Water: Different Types of Suspendedand Fixed TiO2 Catalysts Study,” Sol. Energy, 80, pp. 1376–1381.

[36] Marugaan, J., van Grieken, R., Sordo, C., and Cruz, C., 2008, “Kinetics of thePhotocatalytic Disinfection of Escherichia coli Suspensions,” Appl. Catal., B,82, pp. 27–36.

[37] Powder Diffraction Files (Int. Center for Diffraction Data); files 21-1272.[38] Al-Amoudi, A., and Lovitt, R. W., 2007, “Fouling Strategies and the Cleaning

System of NF Membranes and Factors Affecting Cleaning Efficiency,” J. Mem-brane Sci., 303, pp. 4–28.

[39] Campo, G., Berregi, I., Caracena, R., and Santos, I., 2006, “Quantitative Analy-sis of Malic and Citric Acids in Fruit Juices Using Proton Nuclear MagneticResonance Spectroscopy,” Anal. Chim. Acta, 556, pp. 462–468.

011008-10 / Vol. 134, FEBRUARY 2012 Transactions of the ASME

Downloaded 13 Feb 2013 to 70.194.130.178. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm