Mechanism of photocatalytic disinfection using titania-graphene

composites under UV and visible irradiation

Brenda R. Cruz-Ortiza,e*, Jeremy W.J. Hamiltonb, Cristina Pablosd, Lourdes

Díaz-Jiméneza, Dora A. Cortés-Hernándeza, Preetam K. Sharmab, María

Castro-Alférezc, Pilar Fernández-Ibañezc, Patrick S.M. Dunlopb, John A. Byrneb

a CINVESTAV-Unidad Saltillo, Ave. Industria Metalúrgica No. 1062, Parque

Industrial Saltillo-Ramos Arizpe, C.P. 25900, Ramos Arizpe, Coahuila, México.b NIBEC, Ulster University, Newtownabbey, Northern Ireland, BT37 0QB, United

Kingdom.c Plataforma Solar de Almería – CIEMAT, PO Box 22, 04200 Tabernas, Almería,

Spain.d Department of Chemical and Environmental Technology, ESCET, Universidad

Rey Juan Carlos, c/ Tulipán s/n, 28933, Móstoles, Madrid, Spain.e Autonomous University of Coahuila. Blvd. V. Carranza s/n Col. República Ote.,

C.P. 25280, Saltillo Coahuila, México.

*Corresponding author: Brenda R. Cruz-Ortiz. Autonomous University of

Coahuila. Blvd. V. Carranza s/n Col. República Ote., C.P. 25280, Saltillo

Coahuila, México. Phone: +52 844 4159534 ext. 220.

E-mail address: b.cruz@uadec.edu.mx

Abstract

Photocatalysis has been shown to be effective for the disinfection of water

contaminated with pathogenic microorganisms. In order to increase the solar

efficiency of photocatalysis on titanium dioxide (TiO2) it is necessary to modify

the TiO2 so that visible photons may be utilised in addition to the UV. TiO2 -

reduced graphene oxide composites (TiO2-rGO) were prepared by the

photocatalytic reduction of exfoliated graphene oxide (GO) using P25 (Evonik-

Aeroxide) as the photocatalyst. The composites were tested for the inactivation

1

of E.coli as the model microorganism under UV-Vis and visible only irradiation

at relatively low light intensities to help elucidate the mechanism of disinfection.

The results showed a 6 log inactivation of E. coli after 120 min of treatment with

unmodified TiO2-P25 and the same level of inactivation was achieved after 90

min with TiO2-rGO under UV-Vis irradiation. Under visible irradiation only, the

TiO2-rGO gave a 5.3 log inactivation of E.coli following 180 min of treatment

whereas the unmodified P25 gave only a 1.7 log-reduction in the same time,

similar to that observed in the light control. Using probes, the main reactive

oxygen species involved in the disinfection process were determined to be

hydrogen peroxide, hydroxyl radicals, and singlet oxygen under UV-Vis

irradiation; and only singlet oxygen under visible only irradiation. Scavenger

studies were also performed to further elucidate the mechanism of disinfection.

Keywords

Disinfection; TiO2-rGO; Visible light activity; Reactive oxygen species;

Photocatalytic mechanism; Scavenger studies

1 Introduction

According to WHO and UNICEF in 2015, 663 million people lacked access to

improved drinking water sources yet access to safe water is considered a

human right [1]. The proxy indicator used in the global survey methodology, i.e.

“use of improved drinking water sources”, does not necessarily mean that the

water from these sources is safe to drink. Many people are forced to rely on

sources that are microbiologically unsafe, leading to a higher risk of contracting

waterborne diseases, including typhoid, hepatitis A and E, polio, and cholera. It

2

was estimated that diarrhoeal disease claimed the lives of 2.5 million people in

2008 [2]. This is greater than the combined burden of HIV/AIDS and malaria, for

children under five [3]. There is a need for simple and effective household water

treatment interventions that can be used to help reduce the incidence of water-

borne disease in developing regions. The efficacy of solar disinfection may be

improved by the addition of a photocatalyst. Titanium dioxide (TiO2)

photocatalysis (under UV irradiation) has been shown to be effective for the

inactivation of a wide range of microorganisms [4–6]. When TiO2 is irradiated

with UV photons, valence band electrons are excited to the conduction band,

creating electron-hole (e--h+) pairs. This e--h+ pair can react at the particle

interface with water and oxygen to produce reactive oxygen species (ROS)

such as hydroxyl radical (•OH), superoxide radical anion (O2∙ˉ), singlet oxygen

(1O2), and hydrogen peroxide (H2O2). ROS can be generated by both oxidation

reactions involving valence band holes, or by reduction reactions involving

conduction band electrons (Table 1).

TABLE 1

Singlet oxygen may be produced by photocatalysis by the oxidation of the

superoxide radical anion (eq. 10) [7]. Doping of TiO2 has been used to achieve

visible light activity, however the electrochemical reduction potential of the mid-

gap state is not positive enough to generate hydroxyl radical, but there is

evidence that singlet oxygen might be formed by the oxidation of superoxide at

the mid-gap state [8]. This may be a potential mechanism for disinfection for

visible light active TiO2. Furthermore, the mid gap states have also been

3

reported to directly oxidise organic matter (eq. 14) or anions in solution to

produce radicals such as persulphate [9] or chlorine (see table 2) [10].

TABLE 2

The ROS and other radicals such as chlorine are effective for the inactivation of

microorganisms through damage to the cell wall and cell membrane resulting in

a loss of respiratory activity and leakage of intracellular components [11,12].

TiO2-based composites prepared with graphene (G), graphene oxide (GO) and

reduce graphene (rGO) have emerged as promising materials for photocatalysis

[13]. Titania-rGO composites can show improved solar photocatalytic efficiency

due different possible mechanisms including; improving charge separation and

reducing electron hole recombination rates; changing the ROS distribution; and

inducing visible light activity due to the presence of Ti-O-C bonds giving mid

gap states above the valence band gap [14,15]. Other reports suggest that

TiO2-rGO may improve the photocatalytic performance of the TiO2, because the

rGO may act as an electron bridge to promote the electron transfer between

TiO2 particles, which retards the recombination and prolongs the carrier lifetime

[16,17]. We previously reported rapid disinfection of water under real sun

irradiation with TiO2-rGO composites and there was some evidence of singlet

oxygen production under visible excitation [18]. In that study, the timescales for

disinfection under real sun were must too short to elucidate the full

mechanism/s involved.

4

The aim of this work was to investigate the mechanism of disinfection using

titania-graphene composites under UV-Vis and visible only irradiation under

relatively low light intensities (as compared to our previous work). In this work

we used a range of chemical probes and scavengers for ROS to interrogate the

mechanism photocatalytic disinfection.

2 Materials and methods

2.1 Synthesis and characterization of TiO2-rGO

The TiO2-rGO composite was synthesised by UV photocatalytic reduction of

exfoliated GO (Nanoinnova) in the presence of TiO2 particles (Evonik Aeroxide

P25) with methanol as a hole scavenger. This method and the characterisation

of the TiO2-rGO composites was described previously [18]. Briefly, GO and

TiO2 particles (2% w/v) were suspended separately each in 100 mL absolute

methanol and sonicated using a tip sonication for 10 min. The two suspensions

were mixed to obtain 5% w/v GO concentration with respect to TiO2. The

resulting suspension was sonicated for 1 h and then irradiated with UV-B for 6 h

in oxygen free nitrogen (OFN) atmosphere. The product was then filtered and

dried in air.

2.2 Photocatalytic disinfection of E. coli in stirred tank reactor

2.2.1 Stirred tank reactor

A custom-built stirred tank reactor (Fig. 1S) was employed for the disinfection

experiments as this reactor has been well characterised in previous work [19].

The cylindrical reactor has an inner diameter of 11 cm, and the suspension

volume was 200 mL. Mixing of the suspension was achieved using a stainless

5

steel propeller with a rotation speed of 500 rpm (IKA Eurostar digital,

homogenator motor). The suspension was irradiated from below using a xenon

lamp (Horiba, Jobin Yvon FL-1039/40) with cut off filters > 285 nm for UV-Vis

light and > 420 nm for visible light. The UV irradiance (0.47 mW/cm2,

integrated in the range of 285-400 nm) was measured using a spectral

radiometer (Horiba, Jobin Yvon Gemini 180 monochromator and SpectrAcq2

detector) at the reactor base. The reactor was kept at a constant temperature

(20 ± 2°C) throughout the experiments.

2.2.2. Preparation of E. coli

E. coli K-12 (ATCC 23631) was used as the model bacterium for evaluating the

disinfection kinetics. E. coli was cultured overnight in Luria Bertani broth

(tryptone 10 g/L, NaCl 10 g/L and yeast extract 5 g/L, Sigma Aldrich) and

incubated at 37 ºC under aerobic conditions. Bacteria were collected after 18 h

of incubation, which corresponds to the initial bacterial stationary phase,

yielding a concentration of 108 CFU/mL. E. coli suspensions were centrifuged at

4200 rpm for 10 min, washed three times with ¼ strength Ringer’s solution, the

pellet was re-suspended in Ringer’s solution and diluted to the required cell

density (106 CFU/mL).

2.2.3 Photocatalitic disinfection experiments

A volume of 0.2 L of saline solution (9 wt.% NaCl) with E. coli an initial

concentration of 106 CFU/mL was added into the reactor and the suspension

was air sparged prior to and during the reaction. The rate of inactivation of E.

coli was determined under UV-Vis and visible only irradiation, in the presence

6

and absence of TiO2-P25 (Aeroxide, Evonick, Degussa Corporation), graphene

oxide (GO, Nanoinnova) and TiO2-rGO (TiO2 was P25). Table 3 summarizes the

concentrations of each material employed in the disinfection experiments under

UV-Vis light irradiation.

TABLE 3

For photocatalytic disinfection under visible only irradiation, the catalyst loading

of 18 mg/L gave the best disinfection results under UV-Vis light for TiO2-P25

and TiO2-rGO. This loading was used to compare the disinfection efficiency for

GO and rGO (Nanoinnova). During the experiments, samples were collected at

0, 15, 30, 60, 120, 150 and 180 min, and colony forming units were enumerated

using the standard plated counting method trough a serial 10-fold dilution in

Ringer’s solution. Aliquots of 100 µL were spread in triplicate on Luria Bertani

agar (tryptone 10 g/L, NaCl 10 g/L, yeast extract 5 g/L and agar 10 g/L, Sigma

Aldrich). When very low concentrations of E. coli were expected to be found in

samples, 500 μL aliquots were plated to reach a detection limit of 2 CFU/mL.

Colonies were counted after incubation of 24 h at 37 ºC. At the beginning of

each experiment, a control sample was kept in dark at room temperature, and

plated in triplicate again at the end of the experiment. Re-growth of was

determined for all experiments by leaving the last samples (150 and 180 min) at

room temperature for 48 h followed by plating using the method described.

2.3 Singlet oxygen (1O2) detection

7

Singlet oxygen detection was performed by measuring the fluorescence

intensity of a singlet oxygen specific fluorophore (Singlet oxygen sensor green

(SOSG), S-36002 Invitrogen). In the presence of singlet oxygen, the fluorophore

emits a green fluorescence (525 nm). TiO2-P25, TiO2-rGO or rGO were

suspended at a concentration of 18 mg/L in 20 mL of SOSG 2 μM and

irradiated under continuous stirring with UV-Vis (cut off filter < 285 nm) or

visible light (cut off filter < 420 nm) (Horiba, Jobin Yvon FL-1039/40). Samples

(0.2 mL) were taken from the reactor at 2 min intervals and the fluorescent

intensity was measured with excitation and emission at 504 and 525 nm

respectively with a luminescence spectrometer (Tecan Genios FL). Control

experiments were performed in water to identify the background fluorescence

due to self-sensitisation by the probe.

2.4 Superoxide radical (O2●-) detection

XTT (2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide)

is a water soluble compound used as superoxide probe by means of the XTT

reduction by O2●- leading to the formation of XTT-formazan with an absorption

peak at 470 nm [20]. A concentration of 18 mg/L of TiO2-P25, TiO2-rGO or rGO

was suspended in 20 mL of XTT 100 μM prepared in distilled water. The

suspension was irradiated under continuous stirring UV-Vis or visible light.

Samples were taken at 2 min intervals, centrifuged at 11560 rpm during 15 min

at 20°C, in order to separate the particles. The optical absorbance was

determined at 470 nm in a spectrometer (Perkin Elmer Lambda). Superoxide

radical detection was also investigated using benzoquinone as a probe. 1,4-

benzoquinone is a compound, slightly soluble in water, that reacts with

8

superoxide radical leading to the formation of a semiquinone with an absorption

peak at 430 nm [21]. A 300 µM solution of 1,4 benzoquinone was used for

detection, prepared by mixing a 600 µm solution of 1,4 benzoquinone dissolved

in acetone with equal volumes of used photocatalyst supernatant or with equal

volumes of materials utilised for calibration of this probe. Potassium superoxide

(KO2), a stable salt of superoxide anion, and hydrogen peroxide were used to

calibrate the sensitivity of the benzoquinone assay. The photocatalyst

suspensions, 18 mg/L of TiO2-P25, TiO2-rGO were irradiated with UV-Vis light

under continuous stirring. Samples were taken at different intervals during 30

min and centrifuged to separate the catalyst before mixing with equal volume of

reagent. The optical absorbance was measured at 430 nm using a 3 cm path

length quartz cell.

2.5 Hydroxyl radical (OH) detection

P-nitrosodimethylaniline (RNO) is an organic dye molecule used as a probe

compound and a spin trap for detection of hydroxyl radicals and is easily

detectable using UV-Vis spectrometry. The bleaching of RNO has been

reported to be very selective to oxidation by hydroxyl radicals [22,23]. TiO2-P25,

TiO2-rGO or rGO were suspended in a solution of RNO 17 μM and irradiated

with UV-Vis or visible light under continuous stirring. Samples were taken at 2

min intervals, centrifuged, and the optical absorbance was recorded at 440 nm

in a UV-Vis spectrometer (as above). Controls were performed in the absence

of catalyst to determine any photolytic effect.

9

2.6 Hydrogen peroxide detection

Hydrogen peroxide reacts rapidly with Ti(IV) ions in solution to produce a deep

yellow titanium peroxo complex. This complex is commonly produced by

reaction with acidified TiOSO4 with trace peroxide in solution. Quantification of

peroxide produced under UV-Vis and visible irradiation of RGO-TiO2 catalysts

was achieved by sampling 2 mL from the reactor and mixing with 2 mL of

TiOSO4 (280mM), and measuring the optical absorbance in a 3 cm quartz cell

at 410 nm in a UV-Vis spectrometer (as above). Calibration of absorbance

response as a function of hydrogen peroxide content is given in the

supplementary information.

2.7 Scavenger studies

Sacrificial substrates that have high affinity for particular ROS’s have been

reported as a method to differentiate the contribution of each ROS to

photocatalytic activity of novel materials [24–27]. In this study bacterial

disinfection studies were repeated in the stirred tank reactor in the presence

and absence of ROS scavengers. Isopropanol (10 mM) was used to scavenge

hydroxyl radicals, oxalate (10 mM) was used to scavenge valence band holes,

and Fe(II)-EDTA (10 mM) was used a as scavenger for peroxide. Nitrogen

sparging of the reactor to remove oxygen in combination with isopropanol and

oxalate was used to estimate determine the contribution to disinfection by ROS

generated through an oxygen reduction route utilising conduction band

electrons. Iodide was used as a scavenger for valence band holes. Control

measurements were additionally performed in presence of each scavenger but

in the absence of irradiation.

10

3 Results and discussion

3.1 Photocatalytic disinfection under UV-Vis irradiation

The inactivation of E. coli with TiO2-P25 at different loadings under UV-Vis

irradiation is shown in Figure 1. A 6-log reduction in E. coli viability was

observed following 120 min of irradiation with a catalyst loading of 18 mg/L. For

a catalyst loading of 9 mg/L a 6 log reduction took 150 min. In addition, there

was no evidence of bacterial re-growth up to 48 h following the photocatalytic

treatment. TiO2-P25 at 36 and 72 mg/L loadings achieved log-reductions of 4.8

and 3.6 at 180 min, respectively. With 500 mg/L of TiO2-P25 a lower rate of E.

coli inactivaction was observed due to the photocatalyst screening the

irradiation from the reactor bulk volume. The E. coli viability was not affected

following 180 min in the dark at room temperature. In the light control

experiments, the photolytic disinfection gave a 3.6 log-reduction at 180 min.

During the experiments the temperature remained constant at 20 °C, which is

below the optimal E. coli growth temperature, and had no effect on the

decrease of cell viability. The accumulated UV dose was calculated using

Equation 15.

DOSEUV (J m−2 )=∫ IrradianceUV (W m−2 )×dt (s) (15)

The data were analysed using the Geeraerd and Van Impe Inactivation Model

Fitting Tool (GInaFit, version 1.6) [28]. The results were fitted using the log

linear with shoulder and tail model, the kinetic results using the time as

independient variable are shown in Table 4.

11

FIGURE 1

The TiO2-rGO was found to be more efficient for the inactivation of E. coli under

UV-Vis irradiation as compared to the unmodified TiO2. With the TiO2-rGO at a

catalyst loading of 18 mg/L, the detection limit was reached after only 90 min of

irradiation (Fig. 2). The characteristic shoulder of photocatalytic disinfection was

observed in the inactivation curve obtained with TiO2-rGO at 18 mg/L. At higher

TiO2-rGO catalyst loadings, the disinfection efficiency decreased. During the

experiments, no significant change in pH was observed. The kinetic

parameters determined using the Gina-fit tool are given in table 4.

FIGURE 2

TABLE 4

3.2 E. coli disinfection under visible irradiation

Figure 3 shows E. coli disinfection under visible light irradiation ( > 420 nm)

with TiO2-P25, TiO2-rGO, GO and rGO at 18 mg/L. The best inactivation rate

was obtained with TiO2-rGO, which gave a 5.3 log-reduction after 180 min of

photocatalytic treatment. These results demonstrate that TiO2-rGO is active

under visible only irradiation. Long et al. [29] reported that visible light activity of

TiO2-rGO composites may be due to oxygen vacancy states created by

12

chemical interactions between GO and Ti-O-Ti bonds. No significant difference

was observed for the inactivation of E.coli with TiO2-P25, GO or rGO under

visible light irradiation, as compared to the light control. The latter showed

approximately a 1-log decrease in the colony count following 180 min

irradiation.

FIGURE 3

3.3 ROS production under UV-Vis irradiation

ROS production was investigated in distilled water using specific detection

methods for singlet oxygen, hydrogen peroxide, superoxide, and hydroxyl

radicals. For 1O2 detection, the fluorescence intensity of a singlet oxygen

specific fluorophore (SOSG) was measured. TiO2-P25 and TiO2-rGO were able

to produce 1O2 under UV-Vis irradiation, while rGO did not show activity (Fig. 4).

TiO2-rGO showed higher 1O2 production than TiO2-P25 demonstrating the

interaction between TiO2 and rGO (Ti-O-C) which enhanced the photocatalytic

performance for E. coli disinfection.

FIGURE 4

Hydroxyl radical (OH) detection under UV-Vis irradiation (Fig. 5) in presence of

TiO2-P25, TiO2-rGO and rGO was assessed by means of RNO bleaching (Eq.

16).

13

RNO+OH ●→RNO●OH (16)

TiO2-rGO showed slightly higher production than TiO2-P25, suggesting that the

rGO acts as a sink for conduction band electrons, reducing the rate of charge

carrier recombination and leading to greater OH production for the TiO2-rGO

composites as compared to the TiO2 alone. In contrast, rGO alone did not

produce OH radicals.

The XTT reduction test showed no significant O2- formation under UV-Vis and

visible irradiation. As superoxide formation has been previously reported with

UV activated TiO2 further testing was undertaken using benzoquinone as a

probe for O2-. Again no evidence of superoxide formation was observed

during 30 min of irradiation under UV-Vis irradiation. Retesting under different

catalyst concentrations and higher irradiation intensities gave similar results, in

that superoxide production was not observed (data not shown).

FIGURE 5

Detection of hydrogen peroxide (H2O2) was performed in distilled water and in

aqueous NaCl as used for the disinfection experiments. Figure 6 shows the

concentration of H2O2 produced over time with the TiO2-rGO composites under

UV-Vis irradiation. The concentration of H2O2 measured was greatly reduced in

the NaCl solution as compared to distilled water. Under visible irradiation ( >

420 nm) the quantity of peroxide produced was comparable with the limit of

14

detection and could not be confirmed over the time scale of these

measurements.

FIGURE 6

3.4 ROS production under visible irradiation

TiO2-P25 and TiO2-rGO were found to produce 1O2 under visible light irradiation

(Fig. 7). The reaction rate constants of TiO2 and TiO2-rGO (k= 0.028 and 0.037,

respectively) were lower than that obtained for the same materials under UV-vis

irradiation (k= 0.066 for TiO2-P25 and 0.074 for TiO2-rGO). These results

suggest that 1O2 is the main ROS involved in the inactivation of E. coli with the

TiO2-rGO under visible light irradiation. The rGO alone did not show any 1O2

production. Unmodified TiO2-P25 showed 1O2 production probably as a result of

the light absorption by rutile (~20 wt.% in P25, Ebg = 3.0 eV ≡ 413 nm) although

the rate of 1O2 production was lower than that observed with the TiO2-rGO

composites. In the present study a UV cut off filter ( < 420 nm) was used for

experiments under visible light irradiation and according to its transmittance

spectrum (Fig. 2S) the filter still transmits down to 400 nm. Studies made by

Magalhães et al. [30] also showed a slight antibacterial activity against E. coli

using TiO2-P25 under visible light. However, they did not perform ROS

detection to elucidate the mechanism. The disinfection results under visible

only irradiation showed that the unmodified TiO2-P25 did not show any

significant difference to photolytic disinfection alone (Fig. 3). These results

15

suggest that there is a required threshold concentration of singlet oxygen

required for disinfection to be effective.

FIGURE 7

TiO2-P25, TiO2-rGO and rGO did not produce OH radicals under visible light

irradiation (Fig. 8) which is in agreement with those results obtained by Renfigo

et al. [7] using N,S co-doped TiO2 and electron spin resonance (ESR) spin

trapping with 5,5-dimethyl-pyrroline-1-oxide (DMPO) as hydroxyl radical

scavenger. This is expected since the required electrochemical reduction

potential for OH is very positive (2.38 V (SHE)) and holes in mid gap states

created by Ti-C bonds do not have a positive enough electrochemical potential

to generate OH. It is possible that superoxide radical anion produced by

reduction of oxygen by conduction band electrons can be re-oxidised by these

mid-gap states to yield singlet oxygen [8].

FIGURE 8

3.5 Scavenger study

The scavenger study showed that isopropyl and oxalate scavengers associated

with hydroxyl radicals and direct hole oxidation had little effect on kill rate (Fig.

9). In the presence of isopropyl alcohol and oxalate, with N2 sparging, there was

a significant reduction in disinfection rate was observed suggestive that ROS

are derived from oxygen reduction reaction. The Fe-EDTA scavenger used to

16

eliminate hydrogen peroxide showed the most significant reduction in

disinfection rate. The results of scavenger studies, particularly investigating

disinfection, should be interpreted with care as some of the scavengers used

are not specific for ROS or may take part in side reactions. KI is normally used

as a hole scavenger [24]. In this work, it was observed that the addition of KI

resulted in a significant increase in the rate of disinfection (Fig. 9). The

oxidation of iodide could lead to the formation of iodine which is a well-known

disinfectant.

FIGURE 9

In this study, the key ROS detected were hydroxyl radicals, hydrogen peroxide

and singlet oxygen with both the unmodified TiO2-P25 and with the TiO2-rGO

composites, under UV-Vis irradiation. Surprisingly the formation of O2●- was not

detected using two different probes. Under visible only irradiation, only singlet

oxygen formation was detected for both the unmodified TiO2-P25 and the TiO2-

rGO composites, with the latter showing a greater rate of singlet oxygen

formation. The rapid oxidation of superoxide to form singlet oxygen could

explain the non-detection of superoxide. Also, singlet oxygen may also be

formed by the reaction of H2O2 (one of the ROS observed in our study) with

chlorine [31,32]. The chlorine radical has been observed as a by-product of

photocatalysis under visible light irradiation when chloride ions are present

[10,33].

FIGURE 10

Figure 10 shows the various potential mechanisms in operation to generate

ROS on TiO2-rGO under UV-Vis or visible only irradiation. In the presence of

17

chloride, chlorine may be produced by oxidation of the Cl- by mid-gap states.

The chlorine may subsequently react with H2O2 generating singlet oxygen.

Alternatively, any superoxide formed by the one electron reduction of molecular

oxygen, maybe be oxidised by the mid-gap states to form singlet oxygen.

Conclusions

Photocatalytic disinfection experiments were undertaken with TiO2-rGO and

unmodified TiO2-P25 under UV-Vis and visible only irradiation. ROS probes

and scavengers were employed to interrogate the mechanism of disinfection.

The TiO2-rGO composites showed better UV-Vis photocatalytic disinfection

rates as compared to the unmodified TiO2. It was found that the main ROS

produced under UV-Vis irradiation were hydroxyl radical, hydrogen peroxide

and singlet oxygen. The superoxide radical anion was not detected under UV-

Vis or visible light irradiation despite attempts to observe this radical with two

different assays. Scavenger studies showed a large reduction in the

disinfection rate when Fe-EDTA was used as a scavenger for hydrogen

peroxide. A reduction in disinfection efficiency was also observed when hole

scavengers were added under anoxic conditions (N2 sparging). These results

suggest that H2O2 has a dominant role in the disinfection mechanism, however,

the photocatalytic oxidation of chloride may generate chlorine which can react

with peroxide to yield singlet oxygen. The TiO2-rGO composites were found to

show visible light activity for the inactivation of E.coli and the dominant ROS

was determined to be singlet oxygen.

Acknowledgements

18

The authors acknowledge the economic support provided by the National

Council of Science and Technology (CONACyT-Mexico) for the financial

support (project 138286) and the scholarship for doctoral studies and

sponsoring a research visit to NIBEC for BRCO.

References

[1] WHO/UNICEF, Progress on Sanitation and Drinking-water: 2015 Update

and MDG Assessment Update, (2015).

[2] WHO, Cause-specific mortality: regional estimates for 2008. Geneva,

World Health Organization., (n.d.).

[3] L. Liu, H.L. Johnson, S. Cousens, J. Perin, S. Scott, J.E. Lawn, et al.,

Global, regional, and national causes of child mortality: an updated

systematic analysis for 2010 with time trends since 2000, Lancet. 379

(2012) 2151–2161.

[4] C. Pulgarin, Fe vs. TiO2 Photo-assisted Processes for Enhancing the

Solar Inactivation of Bacteria in Water, Chim. Int. J. Chem. 69 (2015) 7–9.

[5] C. Adán, J. Marugán, S. Mesones, C. Casado, R. van Grieken, Bacterial

inactivation and degradation of organic molecules by titanium dioxide

supported on porous stainless steel photocatalytic membranes, Chem.

Eng. J. (2016).

[6] S.M. Zacarías, M.L. Satuf, M.C. Vaccari, O.M. Alfano, Photocatalytic

inactivation of bacterial spores using TiO2 films with silver deposits,

Chem. Eng. J. 266 (2015) 133–140.

19

[7] J.A. Rengifo-Herrera, K. Pierzchała, A. Sienkiewicz, L. Forró, J. Kiwi, C.

Pulgarin, Abatement of organics and Escherichia coli by N, S co-doped

TiO2 under UV and visible light. Implications of the formation of singlet

oxygen (1O2) under visible light, Appl. Catal. B Environ. 88 (2009) 398–

406.

[8] J.W.J. Hamilton, J.A. Byrne, P.S.M. Dunlop, D.D. Dionysiou, M. Pelaez,

K.O. Shea, et al., Evaluating the Mechanism of Visible Light Activity for

N,F-TiO2 Using Photoelectrochemistry, J. Phys. Chem. C. 118 (2014)

12206–12215.

[9] Q. Mi, A. Zhanaidarova, B.S. Brunschwig, H.B. Gray, N.S. Lewis, A

quantitative assessment of the competition between water and anion

oxidation at WO3 photoanodes in acidic aqueous electrolytes, Energy

Environ. Sci. 5 (2012) 5694–5700.

[10] R. Yuan, S. Fan, H. Zhou, Z. Ding, S. Lin, Z. Li, et al., Chlorine-radical-

mediated photocatalytic activation of C-H bonds with visible light, Angew.

Chemie. 125 (2013) 1069–1073.

[11] V.A. Nadtochenko, A.G. Rincon, S.E. Stanca, J. Kiwi, Dynamics of E.

coli membrane cell peroxidation during TiO2 photocatalysis studied by

ATR-FTIR spectroscopy and AFM microscopy, J. Photochem. Photobiol.

A Chem. 169 (2005) 131–137.

[12] K. Sunada, T. Watanabe, K. Hashimoto, Studies on photokilling of

bacteria on TiO2 thin film, J. Photochem. Photobiol. A Chem. 156 (2003)

227–233.

20

[13] S. Morales-Torres, L.M. Pastrana-Martínez, J.L. Figueiredo, J.L. Faria,

A.M.T. Silva, Design of graphene-based TiO2 photocatalysts--a review.,

Environ. Sci. Pollut. Res. Int. 19 (2012) 3676–87.

[14] H. Zhang, X. Lv, Y. Li, Y. Wang, J. Li, P25-graphene composite as a high

performance photocatalyst., ACS Nano. 4 (2010) 380–6.

[15] Y. Zhang, C. Pan, TiO2/graphene composite from thermal reaction of

graphene oxide and its photocatalytic activity in visible light, J. Mater. Sci.

46 (2011) 2622–2626.

[16] A.W.X. W.S. Wang, D.-H. Wang, W.G. Qu, L.Q. Lu, Large ultrathin

anatase TiO2 nano-sheets with exposed {001} facets on graphene for

enhanced visible light photo-catalytic activity., J. Phys. Chem. C,. 116

(2012) 19893–19901.

[17] C.G. G.D. Sharma, M.L. Keshtov, A.R. Khokhlov, D. Tasis, Improved

power conversion efficiency by insertion of RGO–TiO2 composite layer as

optical spacer in polymer bulk heterojunction solar cells., Org. Electron.

15 (2014) 348–355.

[18] P. Fernández-Ibáñez, M.I. Polo-López, S. Malato, S. Wadhwa, J.W.J.

Hamilton, P.S.M. Dunlop, et al., Solar photocatalytic disinfection of water

using titanium dioxide graphene composites, Chem. Eng. J. 261 (2015)

36–44.

[19] T.A. McMurray, J.A. Byrne, P.S.M. Dunlop, J.G.M. Winkelman, B.R.

Eggins, E.T. McAdams, Intrinsic kinetics of photocatalytic oxidation of

formic and oxalic acid on immobilised TiO2 films, Appl. Catal. A Gen. 262

21

(2004) 105–110.

[20] L. Brunet, D.Y. Lyon, E.M. Hotze, P.J.J. Alvarez, M.R. Wiesner,

Comparative photoactivity and antibacterial properties of C60 fullerenes

and titanium dioxide nanoparticles., Environ. Sci. Technol. 43 (2009)

4355–60.

[21] J.M. Burns, W.J. Cooper, J.L. Ferry, D.W. King, B.P. DiMento, K.

McNeill, et al., Methods for reactive oxygen species (ROS) detection in

aqueous environments, Aquat. Sci. 74 (2012) 683–734.

[22] M.E. Simonsen, J. Muff, L.R. Bennedsen, K.P. Kowalski, E.G. Søgaard,

Photocatalytic bleaching of p-nitrosodimethylaniline and a comparison to

the performance of other AOP technologies, J. Photochem. Photobiol. A

Chem. 216 (2010) 244–249.

[23] C. Kim, H. Park, S. Cha, J. Yoon, Facile detection of photogenerated

reactive oxygen species in TiO2 nanoparticles suspension using

colorimetric probe-assisted spectrometric method., Chemosphere. 93

(2013) 2011–5.

[24] W. Wang, L. Zhang, T. An, G. Li, H.-Y. Yip, P.-K. Wong, Comparative

study of visible-light-driven photocatalytic mechanisms of dye

decolorization and bacterial disinfection by B–Ni-codoped TiO2

microspheres: The role of different reactive species, Appl. Catal. B

Environ. 108 (2011) 108–116.

[25] G. Li, X. Nie, J. Chen, Q. Jiang, T. An, P.K. Wong, et al., Enhanced

visible-light-driven photocatalytic inactivation of Escherichia coli using g-

22

C3N4/TiO2 hybrid photocatalyst synthesized using a hydrothermal-

calcination approach, Water Res. 86 (2015) 17–24.

[26] G. Li, X. Nie, Y. Gao, T. An, Can environmental pharmaceuticals be

photocatalytically degraded and completely mineralized in water using g-

C3N4/TiO2 under visible light irradiation?—Implications of persistent toxic

intermediates, Appl. Catal. B Environ. 180 (2016) 726–732.

[27] G. Li, X. Nie, J. Chen, P.K. Wong, T. An, H. Yamashita, et al., Enhanced

simultaneous PEC eradication of bacteria and antibiotics by facilely

fabricated high-activity {001} facets TiO2 mounted onto TiO2 nanotubular

photoanode, Water Res. 101 (2016) 597–605.

[28] A.H. Geeraerd, V.P. Valdramidis, J.F. Van Impe, GInaFiT, a freeware

tool to assess non-log-linear microbial survivor curves., Int. J. Food

Microbiol. 102 (2005) 95–105.

[29] M. Long, Y. Qin, C. Chen, X. Guo, B. Tan, W. Cai, Origin of Visible Light

Photoactivity of Reduced Graphene Oxide / TiO2 by in Situ Hydrothermal

Growth of Undergrown TiO2 with Graphene Oxide, J. Phys. Chem. C. 117

(2013) 16734–16741.

[30] P. Magalhães, J. Ângelo, V.M. Sousa, O.C. Nunes, L. Andrade, A.

Mendes, Synthesis and assessment of a graphene-based composite

photocatalyst, Biochem. Eng. J. 104 (2015) 20–26.

[31] W. Tian, W. Shi, H. Yang, R. Cui, L. Deng, Production of singlet oxygen

by the reaction of non-basic hydrogen peroxide with chlorine gas, Phys.

Chem. Chem. Phys. 14 (2012) 13344–13349.

23

[32] M. V. Zagidullin, V.D. Nikolaev, M.I. Svistun, N. a. Khvatov, Singlet

oxygen generator based on the reaction of chlorine with filament-guided

jets of basic hydrogen peroxide, Theor. Found. Chem. Eng. 41 (2007)

164–170.

[33] L. Ye, J. Liu, C. Gong, L. Tian, T. Peng, L. Zan, Two different roles of

metallic Ag on Ag/AgX/BiOX (X = Cl, Br) visible light photocatalysts:

Surface plasmon resonance and Z-Scheme bridge, ACS Catal. 2 (2012)

1677–1683.

FIGURE CAPTIONS

Fig. 1. E. coli disinfection at several TiO2-P25 concentrations under UV-Vis irradiation: () 9 mg/L; () 18 mg/L; () 36 mg/L; () 72 mg/L; () 500 mg/L; light control (); E. coli viability at 180 min (, ). DL (detection limit) = 2 CFU/mL.

Fig. 2. E. coli disinfection at several TiO2-rGO concentrations under UV-Vis irradiation: () 18 mg/L; () 36 mg/L; () 72 mg/L; light control (); E. coli viability at 180 min (, , ). DL (detection limit) = 2 CFU/mL.

Fig. 3. E. coli disinfection at several TiO2-rGO concentrations under visible irradiation: () 18 mg/L TiO2-P25; () 18 mg/L TiO2-rGO; () 18 mg/L GO; () 18 mg/L rGO; light control (); E. coli viability at 180 min (, , ). DL (detection limit) = 2 CFU/mL.

Fig. 4. Singlet oxygen production under UV-Vis irradiation with () TiO2-P25, () TiO2-rGO and () rGO at 18 mg/L. () Control (water).

Fig. 5. Hydroxyl radical detection by means of RNO bleaching n under UV-Vis irradiation with () TiO2-P25, () TiO2-rGO, and () rGO at 18 mg/L. () Control (water).

Fig. 6. Hydrogen peroxide detection using TiOSO4, conditions: UV-Vis irradiation with () TiO2-rGO in distilled water or () TiO2-rGO, in 1:2 saline; and visible irradiation of () TiO2-rGO in distilled water or () TiO2-rGO, in 1:2 saline.

Fig. 7. Singlet oxygen production under visible light irradiation with () TiO2-P25, () TiO2-rGO and () rGO at 18 mg/L. () Control (water).

Fig. 8. Hydroxyl radical detection by means of RNO bleaching under visible light irradiation with () TiO2-P25, () TiO2-rGO and () rGO at 18 mg/L. () Control (water).

24

Fig. 9. E. coli disinfection by 18 mg/L of TiO2-rGO under UV-Vis irradiation in the presence and absence of scavengers: (◀) no scavenger; () 10 mM KI; () 10 mM isopropanol; () 10 mM Na2C2O4; () 10 mM Na2C2O4, 10 mM isopropanol and N2; (■) 0.1 mM Fe-EDTA. DL (detection limit) = 2 CFU/mL.

Figure 10: Mechanism of photocatalytic generation of ROS on TiO2-rGO under UV and visible irradiation, and in the presence of chloride.

TABLES

Table 1. ROS generation via oxidation and reduction pathways.

ROS formed by CB electrons ROS formed by VB holesO2+e

−¿→O2.−¿¿ ¿ Eq. 1 H 2O+h+¿→HO.+H +¿ ¿¿ Eq. 6

O2.−¿+H +¿→HOO .¿ ¿ Eq. 2 HO .+OH−¿→H 2O2¿ Eq. 7

O2.−¿+2 H +¿+e

−¿→H 2O2¿¿¿ Eq. 3 H 2O 2+h

+¿→HOO.+H+¿¿ ¿ Eq. 8

HOO.+e−¿+H +¿→H 2O 2¿¿ Eq. 4 HOO.+OH−¿→O2.−¿+H 2O ¿ ¿ Eq. 9

H 2O2+e−¿+H +¿→HO .+H2 O ¿¿ Eq. 5 O2

.−¿+h+¿→❑1O 2¿ ¿ Eq. 10

Table 2. Oxidation reactions for mid gap holes.

Mid gap state oxidative species

25

Cl−¿+h+¿→Cl. ¿¿ Eq. 11

SO42−¿+h+¿→SO4

−. ¿¿ Eq. 12

SO4−.+SO4

−.→S2O82−¿¿ Eq. 13

RH+h+¿→R+H+¿¿ ¿ Eq. 14

Table 3. Catalyst loading for the UV-Vis irradiation experiments

TiO2-P25 (mg/L) TiO2-rGO (mg/L)9 1818 3636 7272500

Table 4. Effect of TiO2-P25 and TiO2-rGO at different concentration on E. coli disinfection under UV-Vis light: kinetic parameters found using the Gina-fit tool with the log linear + shoulder + tail model.

Catalyst (mg/L)ParameterSl (shoulder length, min)

kmax (1/min) Log Nres (residualbacterial concentration)

TiO2-P25 (9) 22.42 ± 4.83 0.13 ± 0.01 0.32 ± 0.11TiO2-P25 (18)** 45.28 ± 5.13 0.18 ± 0.02 0.27 ± 0.18TiO2-P25 (36) 23.10 ± 7.25 0.09 ± 0.01 1.62 ± 0.15TiO2-P25 (72) 44.86 ± 7.13 0.09 ± 0.01 2.78 ± 0.12TiO2-P25 (500)*** - 0.04 ± 0.01 4.56 ± 0.10TiO2-rGO (18) 6.24 ± 3.17 0.20 ± 0.01 0.29 ± 0.08TiO2-rGO (36)*** - 0.11 ± 0.01 0.26 ± 0.14TiO2-rGO (72)* - 0.11 ± 0.02 1.42 ± 0.20* No fits on shoulder model** No fits on tail model

26

*** No fits on shoulder + tail model

27

28

29

30

31

32

33

34

35

36