Chemosphere 239 (2020) 124722

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Improved photocatalysis of perfluorooctanoic acid in water and

wastewater by Ga2O3/UV system assisted by peroxymonosulfate
Bentuo Xu a, b, John L. Zhou a, *, Ali Altaee a, Mohammad B. Ahmed a, Md Abu Hasan Johir a,
Jiawei Ren a, Xiaowei Li c
a
Centre for Green Technology, School of Civil and Environmental Engineering, University of Technology Sydney, 15 Broadway, NSW 2007, Australia
b
School of Life and Environmental Science, Wenzhou University, Wenzhou 325035, China
c
School of Environmental and Chemical Engineering, Institute for the Conservation of Cultural Heritage, Shanghai University, Shanghai 200444, China

h i g h l i g h t s g r a p h i c a l a b s t r a c t

100% PFOA was degraded within

60 min in Ga2O3/PMS/UV system.

Acidic condition at pH 3 was prefer-
able for PFOA degradation.

SO
e
e
4 , O2 and photogenerated elec-
trons were key active species.

PFOA was degraded to short chain
intermediates.

a r t i c l e i n f o a b s t r a c t

Article history: Perfluorooctanoic acid (PFOA) has attracted considerable attention worldwide due to its widespread
Received 17 June 2019 occurrence and environmental impacts. This research focused on the photocatalytic process for the
Received in revised form treatment of PFOA in water and wastewater. Gallium oxide (Ga2O3) and peroxymonosulfate (PMS) were
2 August 2019
mixed directly in PFOA solution, which was irradiated under different light sources. The treatment
Accepted 30 August 2019
system showed excellent performance that 100% PFOA was degraded within 90 min and 60 min under
Available online 2 September 2019
254 nm and 185 nm UV irradiation, respectively. Moreover, the degradation efficacy was unaffected by
Handling Editor: Jun Huang initial PFOA concentration from 50 ng L1 to 50 mg L1. Acidic solution (pH 3) improved the degradation
process. The quantum yield in the PMS/Ga2O3 system under UV light (254 nm) was estimated to be
Keywords: 0.009 mol E1. Scavengers such as tert-butanol (t-BuOH), disodium ethylenediaminetetraacetate (EDTA-
Perfluorooctanoic acid Na2) and benzoquinone (BQ) were added into PFOA solution to prove that sulfate radicals (SO
e 4 ), su-
Photocatalytic pathway peroxide radical ðO
e e
2 Þ and photogenerated electrons (e ) were the main active species with strong redox
PMS/Ga2O3/UV system ability for PFOA degradation in PMS/Ga2O3/UV system. Combined with the intermediates analysis, PFOA
Quantum yield
was degraded stepwise from long chain compound to shorter chain intermediates. In addition, PFOA in
Sulfate radicals
real wastewater exhibited similar degradation efficiency, together with 75e85% TOC removal by Ga2O3/
PMS under 254 nm UV irradiation. Therefore, Ga2O3/PMS system was highly effective for PFOA photo-
degradation under UV irradiation, which has potential to be applied for the perfluoroalkyl substances
(PFAS) treatment in water and wastewater.
© 2019 Elsevier Ltd. All rights reserved.

* Corresponding author.
E-mail address: junliang.zhou@uts.edu.au (J.L. Zhou).

https://doi.org/10.1016/j.chemosphere.2019.124722
0045-6535/© 2019 Elsevier Ltd. All rights reserved.
2 B. Xu et al. / Chemosphere 239 (2020) 124722

1. Introduction C6F13COOH, 99%), perfluorohexanoic acid (PFHxA, C5F11COOH, 

97%), perfluoropentanoic acid (PFPeA, C4F9COOH, 97%), per-
Perfluoroalkyl substances (PFAS), anthropogenic compounds fluorobutanoic acid (PFBA, C3F7COOH, 98%), pentafluoropropionic
characterised by their fully fluorinated hydrophobic carbon and a acid (PFPA, C2F5COOH, 97%) and trifluoroacetic acid (TFA, CF3COOH,
hydrophilic end group, have been applied in many industrial,  99%) were obtained from Sigma-Aldrich, Australia. Catalyst ma-
commercial and domestic products and as the key ingredient in terials of gallium oxide (Ga2O3) and Oxone® were directly used
aqueous film-forming foams since the 1950s (Braunig et al., 2019). without further treatment. The scavengers tert-butanol (t-BuOH),
Due to the highly stable carbon-fluorine bond with a bond energy of disodium ethylenediaminetetraacetate (EDTA-Na2) and benzoqui-
544 kJ mol1 (Saleh et al., 2019), PFAS are ubiquitous in the envi- none (BQ) were also obtained from Sigma-Aldrich. All solutions
ronment and globally detected in many environmental matrices and were prepared using ultra-pure water, obtained from distil cham-
even in human tissues (Jian et al., 2018; Mazzoni et al., 2019). For ber and a Milli-Q system.
example, the Williamtown site from Sydney, Australia was found to
contain up to mg L1 level of perfluorooctanoic acid (PFOA) and
perfluorooctane sulfonate (PFOS), two kinds of well-known PFAS,
over the past decades (www.health.nsw.gov.au/environment). Both 2.2. UV irradiation experiments
substances worked their way through the surface water or soil to
the groundwater underneath the site, leading to the land contam- A cylindrical reactor vessel was filled with PFOA aqueous solu-
ination close to the site (Anderson et al., 2018; Janda et al., 2019). tion of 200 mL and different initial PFOA concentrations (50 ng L1,
Until now, these chemicals have been reported to be linked to many 50 mg L1, 50 mg L1) at room temperature. The concentration of
human diseases such as kidney cancer (Mulabagal et al., 2018), 50 mg L1 was comparatively high, compared with the concentra-
thyroid disease (Blake et al., 2018), testicular cancer (Mastrantonio tions of PFOA occurring in the most contaminated water; it was also
et al., 2017) and pregnancy-induced hypertension (Apelberg et al., used as the initial concentration in a previous study (Chen et al.,
2007; Holtcamp, 2012). Therefore, as a precautionary approach, 2015). The solid catalyst of Ga2O3 was dispersed in PFOA solution,
new technology with strong degradation ability for PFAS removal is to which PMS was then added to initiate the reaction with
urgently needed to mitigate PFAS contamination. continuous stirring for 30 min under darkness. The effect of solu-
As the most representative class of PFAS, PFOA has been drawing tion pH on PFOA photodegradation was assessed by repeating ex-
increasing worldwide attention for treatment. Recent studies tar- periments at pH 3, 5, 7 and 10. Light irradiation was provided by
geting on PFOA removal used different treatment such as ultrasonic three separate lamps: a 32-W low-pressure UV lamp at 254 nm
treatment (Lin et al., 2016), sonochemical degradation with peri- wavelength (Cnlight Co. Ltd., Shanghai, China), a 32-W low-
odate (Lee et al., 2016), and microwave enhanced oxidation and pressure UV lamp at 185 nm wavelength (Cnlight Co. Ltd.,
reduction (Li et al., 2017), but they all involve complex operation Shanghai, China), and a 50-W xenon lamp at 400e800 nm wave-
and high energy cost. Significantly, heterogeneous photocatalysis is length (Nbet Co. Ltd., Beijing, China). The irradiation intensity of UV
attractive due to the low energy consumption and high photo- light (254 nm) was 1.71 mW cm2 as measured by an UV intensity
catalytic ability for PFOA degradation (Chen et al., 2015; Xu et al., meter (ST-512). During photodegradation, the lamps were placed
2018). A variety of materials used as the catalysts also provide the 5 cm above the liquid surface of the reactor and the solution was
possibility for the enhancement of this technology. Notably, Ga2O3, well mixed by magnetic stirring (Li et al., 2016; Zhao et al., 2015). At
an environmentally friendly material, was found to be superior to regular time intervals, aliquots of the sample were taken using a
TiO2 when treating PFOA in solution (da Silva et al., 2017), and has syringe and filtered through a filter (Puradisc syringe filter, 0.2 mm,
several advantages such as chemical and biological stability, easy Whatman) before analysis. For comparison, control experiments by
availability and nontoxicity. These characteristics enable it to be a Ga2O3, or PMS, or Ga2O3/PMS in darkness were conducted. In
frequently used catalyst in both research and industry applications addition, photodegradation experiments by Ga2O3 or PMS under
(Lukic et al., 2017). UV were conducted. All experiments were conducted in triplicate.
Furthermore, peroxymonosulfate (PMS, commercial name of Fig. A1 represents the overall process of the experiments.
Oxone®, 2KHSO5$KHSO4$K2SO4) is an environmentally friendly
oxidant (Khan et al., 2017). It can be activated by UV light to pro-
duce sulfate radicals (SO
e4 ) which have strong oxidizing ability
(Gao et al., 2016). A recent study (Wu et al., 2018) used persulfate- 2.3. Analytical methods
assisted photocatalytic ozonation, and their results revealed 99.2%
removal of PFOA. Nevertheless, such significant research is limited, The characterization methods for the catalysts were provided in
and still there is a research gap in the kinetics and mechanistic appendix. A triple quadrupole ultra-high-performance liquid
study of PFOA degradation in complex systems. This study aimed to chromatograph tandem mass spectrometer (UHPLC-MS/MS) from
investigate, for the first time, the photocatalytic ability of Ga2O3 Shimadzu (model 8060) equipped with a binary pump and a Shim-
assisted with PMS used as the catalysts for PFOA photodegradation pack column (1.6 mm, 2.0 mm  50 mm) was used for the quanti-
under UV light in water and wastewater. Experimental conditions tative analysis of PFOA and its degradation products (PFHpA,
such as catalyst amount and solution pH were evaluated by PFHxA, PFPeA, PFBA, PFPrA, TFA). The mobile phase A was Milli-Q
comparative experiments to explore the optimum conditions. Be- water and mobile phase B was methanol. The flow rate was
sides, the intermediates generated during the process and active 0.4 mL min1 and the injection volume was 1 mL. The elution
species in charge of PFOA degrading were analysed to derive the gradient of PFOA analysis method was initiated with 50% B for
possible photocatalytic mechanism in PMS/Ga2O3/UV system. 2.5 min, then 100% B for the next 1 min followed by 50% B for
another 1.5 min. The total run time of this method was 5 min. The
2. Materials and methods mass spectrometer was operated in multiple reaction monitoring
(MRM) mode. For PFOA, m/z 169.1 and m/z 219.0 are used as the
2.1. Chemicals and materials qualitative ions and m/z 369.0 is used as the quantitation ion to
avoid mass interference. The tandem mass spectrometry operating
PFOA (C7F15COOH, 95%), perfluoroheptanoic acid (PFHpA, conditions of the target compounds are listed in Table A1.
 B. Xu et al. / Chemosphere 239 (2020) 124722 3

3. Results and discussion respectively. Based on the results in Fig. 1A, the rate constant for
PFOA degradation according to system configuration or the type of
3.1. Photocatalytic ability of Ga2O3/PMS catalyst used in the reactor (single or mixed catalyst) was found to
be 0.0041 min1, 0.0013 min1 and 0.0156 min1, for Ga2O3, PMS
The powder X-ray diffraction (XRD) pattern of Ga2O3 used in the and Ga2O3/PMS, respectively. A catalyst combination of PMS/Ga2O3
study is shown in Fig. A2, which indicates that all of the diffraction significantly promoted the photocatalytic ability, resulting in 3.8
peaks are in agreement with those of monoclinic phase of beta- and 12 times higher rate constant than that of single Ga2O3 and
Ga2O3 (JCPDS No. 41e1103) and the sharp diffraction peaks reveal PMS, respectively. Therefore, the results imply that sulfate radicals
that the samples has a high crystalline quality (Wang et al., 2015). (SO
e
4 ) generated by PMS significantly enhanced the catalytic
The morphologies of these commercial catalysts were investigated ability of Ga2O3 targeting on the PFOA degradation under UV irra-
by scanning electron microscope (SEM) and Ga2O3 shows to be rods diation. As the PMS/Ga2O3 system for PFOA photodegradation has
with clear boundaries in Fig. A3. not been studied before, their photocatalytic kinetics and mecha-
In order to improve the catalytic efficacy, Ga2O3 was mixed with nism were further investigated.
PMS in the PFOA solution as the catalysts for the PFOA degradation
under UV light (254 nm). For strict comparison, the initial con-
3.2. Factors influencing photodegradation efficiency
centrations of Ga2O3 and PMS were 0.25 g L1 and 0.41 g L1,
respectively, so that the molar ratio between Ga2O3 and PMS was
3.2.1. Effect of catalyst loading
1:1. Fig. 1A shows that PFOA was completely degraded within
Many previous studies have shown that the amount of catalyst
120 min by the catalysts of Ga2O3 mixed with PMS. However, when
has a positive effect on photocatalytic efficacy (Carbajo et al., 2018;
only PMS or Ga2O3 was used as the control group under the same
Xu et al., 2017). Because the number of reactive sites increase with
condition, only 18% and 58.3% of PFOA was degraded, respectively
the increase of the catalyst amount that promote the rate of
within 180 min. Thus, the photocatalytic ability decreased as
oxidation. However, a high concentration of catalyst would make
Ga2O3/PMS > Ga2O3 > PMS. The kinetics of the photodegradation of
the solution turbid, and reduce the photo permeability of UV light,
PFOA fitted well to the pseudo-first-order model (R2 > 0.90)
with a negative effect on the photocatalytic degradation rate. In this
described by eq. (1):
study, under the 1:1 molar ratio of PMS and Ga2O3, different
amounts of PMS/Ga2O3 (PMS at 4.35, 0.90, 0.41 and 0.15 g L1,
lnðC0 =Ct Þ ¼ kt (1) Ga2O3 at 2.65, 0.55, 0.25 and 0.10 g L1) were used to assess the
performance in PFOA removal. Fig. 1B shows that PMS ranging from
where k is the rate constant (min1), C0 and Ct are the concentra- 0.41 to 4.35 g L1 and Ga2O3 ranging from 0.25 to 2.65 g L1
tions of PFOA (mg L1) in solution at irradiation time 0 and t, generated identical results that 100% PFOA was degraded within

Fig. 1. (A) Comparison of photocatalytic performance for PFOA removal among Ga2O3, PMS and Ga2O3/PMS under 254 nm UV. (B) Degradation rate of PFOA by different amounts of
PMS (0.16, 0.41, 0.90, 4.35 g L1) and Ga2O3 (0.10, 0.25, 0.55, 2.65 g L1). (C) Photocatalytic performance by different PMS/Ga2O3 ratios (1:2, 1:1, 2:1, 3:1) under 254 nm UV. (D) The
fitting of degradation curves and rate constant (k) derived.
4 B. Xu et al. / Chemosphere 239 (2020) 124722

180 min. However, when the amount of PMS and Ga2O3 was (Burns et al., 2008), PFOA in the aqueous solution is mainly in the
decreased to 0.15 and 0.10 g L1, respectively, the degradation protonated form (C7F15COOH) at pH < 2.8 and the anionic form
process was significantly inhibited, as only about 33% PFOA was (C7F15COO) at pH > 2.8 based on eqs (4) and (5):
removed during 180 min. The results may be attributed to the
insufficient active groups produced in the system with a low con- C7 F15 COOH ¼ Hþ þ C7 F15 COO (4)
centration of Ga2O3 and PMS. The results are consistent with results
by Carbajo et al. (2018). In comparison, 0.5 g L1 TiO2 doped with Pb C7 H15
could degrade almost 100% PFOA (50 mg L1) within 480 min with
pKa ¼ pH  log (5)
C7 H15 COOH
a rate constant of 0.0086 min1 (Chen et al., 2016), which is lower
than 0.0156 min1 using Ga2O3 (0.25 g L1) and PMS (0.41 g L1) in The zeta potential of Ga2O3 continuously decreased with the
this study. Following the principle of low cost and high efficacy, the increased solution pH and the points of zero charge (pHpzc) was 8.6
dosages of 0.41 g L1 of PMS and 0.25 g L1 of Ga2O3 were chosen as as measured (Fig. A4), which is in accordance with the research by
the optimum combination under 1:1 molar ratio between PMS and Zhao et al. (2015). This suggests that when the solution pH was
Ga2O3 to obtain the most effective photodegradation. lower than 8.6, the surface of Ga2O3 was positively charged due to
Fig. 1C depicts the photocatalytic degradation process of the protonation; while over pH 8.6, the surface of Ga2O3 was
aqueous PFOA in the presence of PMS/Ga2O3 with different PMS/ negatively charged. In addition, compounds adsorption on catalyst
Ga2O3 molar ratios such as 1:2, 1:1, 2:1 and 3:1 within 90 min. The is a significant step during the photocatalytic process. The lower the
dose of Ga2O3 was kept constant at 0.25 g L1 while the amount of pH value is, the more positive the Ga2O3 surface will be. Therefore,
PMS was varied at 0.21, 0.41, 0.82, and 1.23 g L1. When the ratio comparatively high amount of PFOA was adsorbed on the surface of
was 3:1, the results showed that 100% PFOA was removed within Ga2O3 at pH 3 due to the Colombian attraction, promoting the
90 min. While lower ratio of PMS and Ga2O3 (i.e. 2:1, 1:1 and 1:2) photodegradation efficacy. Zhao et al. (2015) proposed a similar
resulted in lower PFOA degradation (i.e. 82%, 79% and 47%). explanation in their study.
Furthermore, the pseudo-first-order kinetics were used to evaluate
the degradation of PFOA with different ratio of persulfate addition 3.2.3. Effect of initial PFOA concentration
as shown in Fig. 1D. The rate constant increased with the increasing Initial PFOA concentration might affect the degradation rate
ratio of PMS/Ga2O3 in the system as k 3:1 ¼ 0.0211 min1 > k (Hamid and Li, 2018; Tian and Gu, 2018). This study investigated
2:1 ¼ 0.0173 min
1
> k 1:1 ¼ 0.0156 min1 > k 1:2 ¼ 0.0061 min1. three initial PFOA concentrations: 50 mg L1, 50 mg L1 and
The findings therefore suggest that SO
e 4 played a critical role in
50 ng L1, to cover its likely levels in the aquatic environment. In the
PMS/Ga2O3/UV system for PFOA degradation and a sufficient dose solution, PMS and Ga2O3 concentrations were set at 0.41 g L1 and
of PMS could boost the degradation efficacy during the photo- 0.25 g L1, respectively and the solution pH was adjusted to 3 to
catalytic process. Thus, increasing the molar ratio between PMS and achieve the best performance of PFOA degradation. Fig. 2C shows
Ga2O3 from 1:2 to 3:1 would be better than increasing their indi- that all three levels of concentration have similar degradation ef-
vidual amounts for PFOA removal. ficacy with similar rate constant at 0.0144e0.0156 min1 (Fig. 2D),
demonstrating that the photodegradation performance was rela-
3.2.2. Effect of solution pH tively independent of initial PFOA concentrations. The results are
The photodegradation of PFOA was evaluated at various initial due to the fact that 0.41 g L1 of PMS and 0.25 g L1 of Ga2O3 pro-
solution pH (pH 3, 5, 7 and 10) as shown in Fig. 2A. Under UV vided sufficient active sites and radicals, which completely
irradiation for 120 min, PFOA exhibited almost 100% removal at pH degraded PFOA ranging from 50 ng L1 to 50 mg L1 at the similar
3 which was decreased to 86% and 78% when pH was increased to 5 degradation rate. Therefore, PMS/Ga2O3/UV system provided very
and 7, respectively. The removal of PFOA further dropped signifi- effective treatment of PFOA whether at trace level (ng L1) or high
cantly to 27% under the basic condition of pH 10. Moreover, the level (mg L1), with great potential for decontaminating water and
degradation results followed the pseudo-first-order kinetic profile wastewater with varying degrees of pollution.
during the photocatalytic process under UV irradiation, with the
rate constant decreasing as kpH3 ¼ 0.0156 min1 > kpH5 ¼ 0.0099 3.2.4. Effect of light sources
min1 > kpH7 ¼ 0.0084 min1 > kpH10 ¼ 0.0025 min1 (Fig. 2B). To explore the effect of light sources on the efficacy, photo-
Hence, the degradation efficiency increased with lower solution degradation of PFOA (50 mg L1) under the optimum conditions
pH. This could be attributed to the reaction between sulfate radicals (1.23 g L1 PMS, 0.25 g L1 Ga2O3, pH 3) were conducted with both
(SO
e 
4 Þ and OH ions to form hydroxyl radicals (
OH) as shown in
UV light (185 nm and 254 nm) and visible light (400e800 nm).
eq. (2) (Liang et al., 2007): Fig. 3 presents that Ga2O3 with PMS under 185 nm UV light has
higher degradation efficacy, as 100% PFOA was removed within
SO$  2
4 þ OH þ SO4 þ ,OH (2) 60 min. While under 254 nm UV light, the time spent for complete
removal increased to 120 min. Furthermore, under visible light
As previously reported, the hydroxyl radicals have poorer (400e800 nm), PFOA could only be degraded to a smaller extent.
reactivity with PFOA in aqueous solution compared with SO
e 4 or The results show that PMS/Ga2O3 absorption of 185 nm UV light
photogenerated hole and electron pairs (Hori et al., 2004); thus, was higher than that of 254 nm UV light, which in turn was
hydroxyl radicals production in the reaction system could slow significantly higher than that of visible light.
down the PFOA degradation. Such a conclusion was also drawn Notably, Li et al. (2012; 2013a; 2013b) synthesized nano-
from the research by Lee et al. (2009). In addition, in basic solution, structured In2O3 such as nanocubes, nanoplates and porous
abundant OH ions would react with
OH, which further inhibited microsphere. Until now, these synthesized catalysts showed better
the reaction of PFOA degradation following eq. (3): performance of PFOA degradation under UV light (15 W, 254 nm)
compared with other synthesized catalysts (i.e. CueTiO2, TiO2
,OH þ OH /O, þ H2 O (3)
nanotube arrays, PbeBiFeO3/RGO) in previous studies (Chen et al.,
On the other hand, the pH also affects the dissociation of 2015; Shang et al., 2018; Wu et al., 2016), which could remove all
ionizable chemicals such as PFOA in solution and consequently, PFOA (30 mg L1) within 30, 60 and 120 min, respectively. However,
influences their photocatalytic performance. As the pKa is about 2.8 such catalysts need many precursors and solvothermal process for
 B. Xu et al. / Chemosphere 239 (2020) 124722 5

Fig. 2. (A) Photocatalytic performance under different initial pH conditions (i.e. pH 3, pH 5, pH 7 and pH 10) activated in PMS/Ga2O3/UV system within 120 min ([PFOA] ¼ 50 mg L1,
[PMS] ¼ 0.41 g L1, [Ga2O3] ¼ 0.25 g L1). (B) The fitting of degradation curves within 90 min and the rate constant (k) derived. (C) Photocatalytic performance with initial PFOA
concentration of 50 mg L1, 50 mg L1 and 50 ng L1 in PMS/Ga2O3/UV system within 120 min ([PMS] ¼ 0.41 g L1, [Ga2O3] ¼ 0.25 g L1). (D) The fitting of degradation curves within
90 min and the rate constant (k) derived.

their synthesis, hence involving high cost. In this study, PMS/Ga2O3/ cost and high degradation efficiency, hence demonstrating signif-
UV system prepared by mixing the commercial Oxone and Ga2O3 at icant potential to be applied for PFAS removal in the aquatic
certain molar ratio (e.g. 3:1) under UV irradiation (32 W, 254 nm, environment.
1.71 mW cm2) achieved comparatively high efficacy with com-
plete removal of 50 mg L1 PFOA in 90 min. Moreover, when the
wavelength of UV light was changed to 185 nm, the degradation 3.3. Calculation of quantum yield
time was shortened to 60 min by PMS/Ga2O3. Therefore, PMS/
Ga2O3/UV system has the advantages of being easily prepared, low The results were analysed in terms of not only the first-order
kinetic model, but also the explicit effect of photon absorption.
The quantum yield (Ф), defined as moles of PFOA degraded per
Einstein of photons absorbed by certain amount of catalysts, was
estimated by eq. (6):

Total number ðmolesÞof PFOA degraded

Ф¼
Total number ðmolesÞ of photons absorbed in the system
(6)
The photocatalytic rate constant of the PFOA, k, under UV light
(254 nm) can be used for the calculation of the reaction quantum
yield as shown in eq. (7):

k
Ф¼ D (7)
Ia

where kD ¼ d(C0-Ct)/dt (0.43 mg L1 min1), and Ia is the photon

numbers absorbed by the catalysts, which can be calculated by eq.
(8):
 
Ep  1  102$ðaðlÞ$CÞ$z A1  A0
Ia ¼  (8)
z A1
Fig. 3. Photocatalytic performance by PMS/Ga2O3 under UV light (185 nm, 254 nm)
and visible light (400e800 nm), [PMS] ¼ 1.23 g L1, [Ga2O3] ¼ 0.25 g L1. where Ep is the average photon intensity (1.71 mW cm2), C is the
6 B. Xu et al. / Chemosphere 239 (2020) 124722

concentration of the catalyst solution (2.8 mM) and z is the mixed findings indicated that photocatalysis by Ga2O3/PMS under UV light
depth of the water column (1.3 cm). In addition, A is the amount of irradiation (254 and 185 nm) was highly stable and not easily
light absorbed by the catalyst at wavelength 254 nm, which was disturbed by other organic compounds in the real wastewater.
detected by the spectrophotometer. a(l) is the light attenuation Besides, total organic carbon (TOC) was also analysed before and
coefficient of the catalysts which is calculated as: after the treatment. As shown in Fig. 4B, in the effluent sample, 75%
TOC was removed under 254 UV light whereas 32% TOC was
A removed under 185 nm UV light. Similarly, in the influent sample,
aðlÞ ¼ (9)
Cl 85% TOC removal was observed under 254 nm UV, which was
higher than 54% TOC removal under 185 nm UV. Thus, TOC could be
where l is the distance that light travels through the solution (1 cm).
significantly removed after the treatment under 254 nm UV light,
Based on the data summarized in Table A2, the quantum yield by
which was better than that under 185 nm UV light. Such different
the catalysts of PMS/Ga2O3 (0.41 g L1 and 0.25 g L1, respectively)
level of TOC removal might be related to fact that 185-nm UV light
under UV light (254 nm) was 0.009 mol E1. The details of calcu-
does not propagate through water as well as 254-nm UV light
lation are presented in the supplemental material.
(Imoberdorf and Mohseni, 2011). They suggested that 99% of the
185 nm radiation was absorbed by raw water molecules in the
3.4. Application in wastewater treatment reactor, whereas only 48% of the 254-nm radiation was absorbed.
Thus, although being more powerful than 254 nm UV light, the 185
Wastewater samples were taken from the influent and effluent nm UV light is more absorbed by water and less absorbed by TOC
of a municipal sewage treatment plant in Sydney, Australia, to molecules, hence generating lower TOC degradation removal in
prepare 50 mg L1 of PFOA solution. The characteristics of the water than by 254 nm light.
influent and effluent samples are listed in Table A3. When the PFOA
and catalysts was added, the pH was changed to 3.0 ± 0.2 without 3.5. Photocatalysis mechanism
pH adjustment. Hereupon, photodegradation of PFOA under the
optimum conditions (1.23 g L1 PMS, 0.25 g L1 Ga2O3, pH 3) were 3.5.1. Intermediates analysis
conducted with both UV light (185 nm and 254 nm). As shown in Intermediates during PFOA photocatalytic process were identi-
Fig. 4A, under 254 nm UV irradiation, 100% PFOA in wastewater was fied and quantified with UHPLC-MS/MS, which were the shorter-
degraded by Ga2O3/PMS within 120 min with the rate constant of chain PFAS including PFHpA, PFHxA, PFPeA, PFBA, PFPA and TFA
0.012 min1, which was similar to that obtained in pure water. as shown in Fig. A5. Fig. 4C presents the time-dependence of the
Under 185 nm UV irradiation, 60e75% PFOA in wastewater was intermediates during the photodegradation process over PMS/
removal at 30 min, and totally degraded within 60 min, which was Ga2O3 under 254 nm UV light irradiation. PFHpA increased slowly
also equal to the efficacy during treatment in pure water. Such to about 18.77 mg L1 within 120 min, with the target pollutant

Fig. 4. (A) Degradation of PFOA in wastewater samples including influent and effluent from a wastewater treatment plant under UV 254 nm and 185 nm. (B) The rate of TOC removal
by Ga2O3/PMS under 254 nm and 185 nm UV light, with [PFOA] ¼ 50 mg L1, [PMS] ¼ 0.45 g L1, [Ga2O3] ¼ 0.25 g L1. (C) Evolution of PFOA and shorter-chain intermediates during
PFOA degradation in PMS/Ga2O3/UV system within 120 min. (D) Effects of different scavengers (i.e. Pure Ga2O3, EDTA-Na2, BQ, t-BuOH, no scavengers) on the PFOA degradation in
PMS/Ga2O3/UV system within 120 min. Insert showing the fitting of degradation curves within 90 min and degradation rate constant (k) derived.
 B. Xu et al. / Chemosphere 239 (2020) 124722 7

PFOA decreasing from 50 mg L1 to nearly zero. The amount of constant was 0.0037 min1, indicating that O
e 2 radicals generated
PFHxA was gradually increased to 7.98 mg L1 within 120 min and by electrons also had high oxidative power for PFOA degradation. In
PFPeA was firstly detected at 90 min and slightly increased to addition, t-BuOH added in the solution resulted in 63% PFOA
0.175 mg L1 at 120 min. While species with shorter chains (i.e. removal with the rate constant of 0.0068 min1. The findings
PFBA PFPA and TFA) had the low concentrations (below limit of indicate that
OH radicals should be a secondary active species
quantification). This phenomenon suggests that PFOA was compared with SO
e
e
4 and O2 radicals reacting with PFOA under UV
degraded stepwise into shorter chain intermediates such as PFHpA, irradiation. Additionally, when EDTA-Na2 was added in the system,
PFHxA and PFPeA, which was consistent with the study by Lee et al. only a slight decrease was observed on the degradation rate
(2009). compared with that of no scavenger addition, as 80% PFOA was
degraded within 120 min and the rate constant could reach 0.0087
min1. Such findings suggest that photogenerated electrons rather
3.5.2. Analysis of active species than photogenerated holes were the main factor promoting the
In PMS/Ga2O3/UV system, five types of active species including photocatalytic efficacy, which are consistent with the study con-
sulfate radicals (SO
e þ e
4 ), photoinduced holes (h ) and electrons (e ), ducted by Zhao et al. (2015). Conclusively, the results prove that the

e
hydroxyl radicals (
OH), and superoxide radical anions (O2 ), could SO
e
e
4 radicals produced by PMS, O2 radicals, and photogenerated
be produced during the photocatalytic process as shown in eqs e
electrons (e ) played major roles in degrading PFOA in the PMS/
(10)e(13) (Chen et al., 2016): Ga2O3/UV system. The
OH radicals are of secondary importance
while photogenerated holes have little effect on the degradation
HSO ,
5 þ hv/SO4 þ ,OH (10)
efficacy.


Ga2 O3 þ hv/Ga2 O3 hþ þ e (11) 3.5.3. Photocatalytic process
According to the intermediates presence and main active spe-
hþ þ H2 O/,OH þ Hþ (12) cies photogenerated electrons (ee), SO
e
e
4 and O2 radicals during
the photocatalytic process as mentioned above, the PFOA degra-
dation process can be described as follows. First, the fluorine atom
e þ O2 /O,
2 (13)
was more preferable than the carbon atom to react with photo-
To confirm which active species take the leading place in the induced electron (e‒), leading to the cleavage of CeF bonds,
photocatalytic reaction, t-BuOH, EDTA-Na2 and BQ were added to similar to previous study on perfluorocarboxylates photo-
the system, which were used as scavengers of
OH, holes and O
e 2 , degradation (Qu et al., 2010; Song et al., 2013). Moreover, the a-
respectively. Besides, degradation by Ga2O3 only was used to position CeF bond was more easily attacked by e‒ due to the
simulate the condition of adding the scavenger of SO
e 4 radicals. inductive effect of the carboxyl in PFOA. Thus, fluorine was elimi-
Fig. 4D shows these scavengers effect on the photocatalysis in PMS/ nated from PFOA to form C7F13H2COOH:
Ga2O3/UV system and the rate constants were provided during
90 min of the photocatalytic process. When no scavengers were C7 F15 COOH þ e /C7 F14 COOH þ F (14)
added, 100% PFOA could be degraded within 120 min with the rate
constant of 0.156 min1. However, during the degradation by Ga2O3 $C7 F14 COOH þ H2 O/C7 F14 COOH þ $OH (15)
only (equivalent to the scavenging of SO
e4 ), the degradation rate
dropped to 47%, and the rate constant decreased to 0.0081 min1. C7 F14 COOH þ e /$C7 F13 COOH þ F (16)
The results suggest that SO
e4 produced by PMS was a critically
important oxidant for PFOA degradation in the photocatalytic sys-
$C7 F13 COOH þ H2 O/C7 F13 COOH þ $OH (17)
tem. Similarly, when BQ was added in the PFOA solution, the
degradation rate was significantly reduced to 38% and the rate C7F13H2COOH was excited by UV irradiation to generate C6F13

Fig. 5. The mechanism of PFOA degradation during the photocatalytic process in PMS/Ga2O3/UV system. The active species are marked in red, and the ion or compounds removed
during degradation are marked in blue. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
8 B. Xu et al. / Chemosphere 239 (2020) 124722

radical, COOH radical and CH2 carbene. Then C6F13 and COOH initial concentration of PFOA ranged from 50 ng L1 to 50 mg L1.
radical could recombine to form PFHpA (C6F13COOH): The quantum yield during photocatalysis by PMS/Ga2O3 under UV
light (254 nm) was 0.009 mol E1. Through the analysis of in-
C7 F13 H2 COOH / $C6 F13 þ : CH2 þ $COOH (18) termediates, PFOA was gradually degraded from long chian species
into short chain intermediates. Scavenger experiments proved that
$C6 F13 þ $COOH/C6 F13 COOHðPFHpAÞ (19) SO
e
e
4 radicals, O2 radicals and photogenerated electrons were the
most important active species contributing to the PFOA photo-
Consequently, PFHpA was decomposed into PFHxA and PFPeA,
degradation. PFOA in real wastewater was also treated by Ga2O3/
and theoretically, it would continue degradation to form PFBA,
PMS under UV 254 nm irradiation with similar degradation per-
PFPA and TPA in the same way and finally become mineralized to
formance to pure water, in addition to 75e85% TOC removal
CO2 and fluoride ions in the stepwise manner as reported by Li et al.
simultaneously. Overall, the PMS/Ga2O3/UV system has great po-
(2013b).
tential to be applied in the photodegradation of PFOA and other
In the Ga2O3/UV system, when the adsorption of photon energy
similar organic pollutants in water and wastewater.
is equal or more than the band gap energy of Ga2O3, holes and
electrons are generated on the surface of Ga2O3 particles due to the
Acknowledgements
irradiation of UV light. Notably, the photo-induced electrons were
the main active species for PFOA degradation as mentioned in the
The authors would like to thank the China Scholarship Council
previous section, which were also reported by Trojanowicz et al.
(CSC) for financial support (Grant No. 201606890028).
(2018). However, the recombination of the photo-induced hole
and electron could seriously reduce the photocatalytic efficiency (Li
Appendix A. Supplementary data
et al., 2018). In the PMS/Ga2O3/UV system, sulfate radical anions
(SO
e
e
4 ) and O2 can oxidize PFOA into perfluorinated alkyl radicals
Supplementary data to this article can be found online at
(
C7F15):
https://doi.org/10.1016/j.chemosphere.2019.124722.
SO$ 2
4 þ C7 F15 COOH/SO4 þ $C7 F15 COOH
þ
(20)
References
þ þ
$C7 F15 COOH /$C7 F15 þ CO2 þ H (21) Anderson, R.H., Adamson, D.T., Stroo, H.F., 2018. Partitioning of poly-and per-
fluoroalkyl substances from soil to groundwater within aqueous film-forming
foam source zones. J. Contam. Hydrol. 220, 59e65.
C7 F15 COOH þ O$
2 /$C7 F15 þ CO2 [ (22) Apelberg, B.J., Witter, F.R., Herbstman, J.B., Calafat, A.M., Halden, R.U., Needham, L.L.,
Goldman, L.R., 2007. Cord serum concentrations of perfluorooctane sulfonate
Then formed C7F15 radical react with water to form C6F13COF (PFOS) and perfluorooctanoate (PFOA) in relation to weight and size at birth.
after Hþ and F elimination: Health Perspect. 115, 1670e1676.
Blake, B.E., Pinney, S.M., Hines, E.P., Fenton, S.E., Ferguson, K.K., 2018. Associations
between longitudinal serum perfluoroalkyl substance (PFAS) levels and mea-
$C7 F15 þ H2 O/C7 F15 OH þ Hþ (23) sures of thyroid hormone, kidney function, and body mass index in the Fernald
Community Cohort. Environ. Pollut. 242, 894e904.
Braunig, J., Baduel, C., Barnes, C.M., Mueller, J.F., 2019. Leaching and bioavailability of
C7 F15 OH / C6 F13 COF þ HF (24) selected perfluoroalkyl acids (PFAAs) from soil contaminated by firefighting
activities. Sci. Total Environ. 646, 471e479.
By hydrolysis, C6F13COF is converted into PFHpA (C6F13COOH) Burns, D.C., Ellis, D.A., Li, H., McMurdo, C.J., Webster, E., 2008. Experimental pKa
with CF2 units reduced and followed by a stepwise degradation of determination for perfluorooctanoic acid (PFOA) and the potential impact of
PFOA to form shorter chain PFAS: pKa concentration dependence on laboratory-measured partitioning phenom-
ena and environmental modeling. Environ. Sci. Technol. 42, 9283e9288.
Carbajo, J., Tolosana-Moranchel, A., Casas, J., Faraldos, M., Bahamonde, A., 2018.
C6 F13 COF þ H2 O/C6 F13 COOHðPFHpAÞ þ HF (25) Analysis of photoefficiency in TiO2 aqueous suspensions: effect of titania hy-
drodynamic particle size and catalyst loading on their optical properties. Appl.
Furthermore, the excited e could activate MPS to produce SOe 4 Catal. B 221, 1e8.
radicals, which have been demonstrated in a previous study (Liu Chen, M.J., Lo, S.L., Lee, Y.C., Huang, C.C., 2015. Photocatalytic decomposition of
et al., 2017). These processes could explain that photodegradation perfluorooctanoic acid by transition-metal modified titanium dioxide. J. Hazard
Mater. 288, 168e175.
efficiency was much better in PMS/Ga2O3/UV system than by sole Chen, M.J., Lo, S.L., Lee, Y.C., Kuo, J., Wu, C.H., 2016. Decomposition of per-
Ga2O3 or PMS: fluorooctanoic acid by ultraviolet light irradiation with Pb-modified titanium
dioxide. J. Hazard Mater. 303, 111e118.
€, M., Keiski, R.L., Ojala, S., 2017. Photocatalytic
SO$ $ 2
4 þ SO4 /S2 O8 (26) da Silva, F.L., Laitinen, T., Pirila
degradation of perfluorooctanoic acid (PFOA) from wastewaters by TiO2, In2O3
and Ga2O3 catalysts. Top. Catal. 60, 1345e1358.
 $ Gao, H., Chen, J., Zhang, Y., Zhou, X., 2016. Sulfate radicals induced degradation of
S2 O2
8 þ e /SO4 (27) Triclosan in thermally activated persulfate system. Chem. Eng. J. 306, 522e530.
Hamid, H., Li, L.Y., 2018. Fate of perfluorooctanoic acid (PFOA) in sewage sludge
Overall, the mechanism during the photocatalysis process of during microwave-assisted persulfate oxidation treatment. Environ. Sci. Pollut.
PFOA degradation in the PMS/Ga2O3/UV system is postulated in Res. 25, 10126e10134.
Fig. 5. Holtcamp, W., 2012. Pregnancy-induced hypertension “probably linked” to PFOA
contamination. Environ. Health Perspect. 120, A55eA59.
Hori, H., Hayakawa, E., Einaga, H., Kutsuna, S., Koike, K., Ibusuki, T., Kiatagawa, H.,
Arakawa, R., 2004. Decomposition of environmentally persistent per-
4. Conclusions fluorooctanoic acid in water by photochemical approaches. Environ. Sci. Tech-
nol. 38, 6118e6124.
PFOA removal from aqueous solution was studied by UV pho- Imoberdorf, G., Mohseni, M., 2011. Degradation of natural organic matter in surface
water using vacuum-UV irradiation. J. Hazard Mater. 186, 240e246.
tocatalysis using Ga2O3 and further enhanced in the presence of €dler, K., Brauch, H.J., Zwiener, C., Lange, F.T., 2019. Robust trace analysis
Janda, J., No
PMS. Specifically, the application of 1.23 g L1 of PMS and 0.25 g L1 of polar (C2-C8) perfluorinated carboxylic acids by liquid chromatography-
of Ga2O3 was able to degrade 100% PFOA in aqueous solution within tandem mass spectrometry: method development and application to surface
90 min under UV 254 nm and 60 min under UV 185 nm. Further- water, groundwater and drinking water. Environ. Sci. Pollut. Res. 26,
7326e7336.
more, acidic condition at pH 3 was favourable for PFOA degradation Jian, J.M., Chen, D., Han, F.J., Guo, Y., Zeng, L., Lu, X., Wang, F., 2018. A short review on
in PMS/Ga2O3/UV system. Such degradation was not affected by the human exposure to and tissue distribution of per-and polyfluoroalkyl
 B. Xu et al. / Chemosphere 239 (2020) 124722 9

substances (PFASs). Sci. Total Environ. 636, 1058e1069. 180e185.

Khan, S., He, X., Khan, J.A., Khan, H.M., Boccelli, D.L., Dionysiou, D.D., 2017. Kinetics Mazzoni, M., Buffo, A., Cappelli, F., Pascariello, S., Polesello, S., Valsecchi, S., Volta, P.,
and mechanism of sulfate radical- and hydroxyl radical-induced degradation of Bettinetti, R., 2019. Perfluoroalkyl acids in fish of Italian deep lakes: environ-
highly chlorinated pesticide lindane in UV/peroxymonosulfate system. Chem. mental and human risk assessment. Sci. Total Environ. 653, 351e358.
Eng. J. 318, 135e142. Mulabagal, V., Liu, L., Qi, J., Wilson, C., Hayworth, J.S., 2018. A rapid UHPLC-MS/MS
Lee, Y.C., Chen, M.J., Huang, C.P., Kuo, J., Lo, S.L., 2016. Efficient sonochemical method for simultaneous quantitation of 23 perfluoroalkyl substances (PFAS) in
degradation of perfluorooctanoic acid using periodate. Ultrason. Sonochem. 31, estuarine water. Talanta 190, 95e102.
499e505. Qu, Y., Zhang, C., Li, F., Chen, J., Zhou, Q., 2010. Photo-reductive defluorination of
Lee, Y.C., Lo, S.L., Chiueh, P.T., Chang, D.G., 2009. Efficient decomposition of per- perfluorooctanoic acid in water. Water Res. 44, 2939e2947.
fluorocarboxylic acids in aqueous solution using microwave-induced persulfate. Saleh, N.B., Khalid, A., Tian, Y., Ayres, C., Sabaraya, I.V., Pietari, J., Hanigan, D.,
Water Res. 43, 2811e2816. Chowdhury, I., Apul, O.G., 2019. Removal of poly-and per-fluoroalkyl substances
Li, D., Yu, J.C.C., Nguyen, V.H., Wu, J.C., Wang, X., 2018. A dual-function photo- from aqueous systems by nano-enabled water treatment strategies. Environ.
catalytic system for simultaneous separating hydrogen from water splitting and Sci. Water Res. Technol. 5, 198e208.
photocatalytic degradation of phenol in a twin-reactor. Appl. Catal. B 239, Shang, E., Li, Y., Niu, J., Li, S., Zhang, G., Wang, X., 2018. Photocatalytic degradation of
268e279. perfluorooctanoic acid over Pb-BiFeO3/rGO catalyst: kinetics and mechanism.
Li, M., Yu, Z., Liu, Q., Sun, L., Huang, W., 2016. Photocatalytic decomposition of Chemosphere 211, 34e43.
perfluorooctanoic acid by noble metallic nanoparticles modified TiO2. Chem. Song, Z., Tang, H., Wang, N., Zhu, L., 2013. Reductive defluorination of per-
Eng. J. 286, 232e238. fluorooctanoic acid by hydrated electrons in a sulfite-mediated UV photo-
Li, S., Zhang, G., Zhang, W., Zheng, H., Zhu, W., Sun, N., Zheng, Y., Wang, P., 2017. chemical system. J. Hazard Mater. 262, 332e338.
Microwave enhanced Fenton-like process for degradation of perfluorooctanoic Tian, H., Gu, C., 2018. Effects of different factors on photodefluorination of per-
acid (PFOA) using Pb-BiFeO3/rGO as heterogeneous catalyst. Chem. Eng. J. 326, fluorinated compounds by hydrated electrons in organo-montmorillonite sys-
756e764. tem. Chemosphere 191, 280e287.
Li, Z., Zhang, P., Li, J., Shao, T., Jin, L., 2013a. Synthesis of In2O3-graphene composites Trojanowicz, M., Bojanowska-Czajka, A., Bartosiewicz, I., Kulisa, K., 2018. Advanced
and their photocatalytic performance towards perfluorooctanoic acid decom- oxidation/reduction processes treatment for aqueous perfluorooctanoate
position. J. Photochem. Photobiol. A 271, 111e116. (PFOA) and perfluorooctanesulfonate (PFOS)ea review of recent advances.
Li, Z., Zhang, P., Shao, T., Li, X., 2012. In2O3 nanoporous nanosphere: a highly effi- Chem. Eng. J. 336, 170e199.
cient photocatalyst for decomposition of perfluorooctanoic acid. Appl. Catal. B Wang, Y., Li, N., Duan, P., Sun, X., Chu, B., He, Q., 2015. Properties and photocatalytic
125, 350e357. activity of b-Ga2O3 nanorods under simulated solar irradiation. J. Nanomater.
Li, Z., Zhang, P., Shao, T., Wang, J., Jin, L., Li, X., 2013b. Different nanostructured In2O3 16, 126.
for photocatalytic decomposition of perfluorooctanoic acid (PFOA). J. Hazard Wu, D., Li, X., Zhang, J., Chen, W., Lu, P., Tang, Y., Li, L., 2018. Efficient PFOA degra-
Mater. 260, 40e46. dation by persulfate-assisted photocatalytic ozonation. Sep. Purif. Technol. 207,
Liang, C., Wang, Z.-S., Bruell, C.J., 2007. Influence of pH on persulfate oxidation of 255e261.
TCE at ambient temperatures. Chemosphere 66, 106e113. Wu, Y., Li, Y., Tian, A., Mao, K., Liu, J., 2016. Selective removal of perfluorooctanoic
Lin, J.C., Hu, C.Y., Lo, S.L., 2016. Effect of surfactants on the degradation of per- acid using molecularly imprinted polymer-modified TiO2 nanotube arrays. Int. J.
fluorooctanoic acid (PFOA) by ultrasonic (US) treatment. Ultrason. Sonochem. Photoenergy 2016, 1e10.
28, 130e135. Xu, B., Ahmed, M.B., Zhou, J.L., Altaee, A., Wu, M., Xu, G., 2017. Photocatalytic
Liu, Y., Zhang, Y., Guo, H., Cheng, X., Liu, H., Tang, W., 2017. Persulfate-assisted removal of perfluoroalkyl substances from water and wastewater: mechanism,
photodegradation of diethylstilbestrol using monoclinic BiVO4 under visible- kinetics and controlling factors. Chemosphere 189, 717e729.
light irradiation. Environ. Sci. Pollut. Res. 24, 3739e3747. Xu, B., Ahmed, M.B., Zhou, J.L., Altaee, A., Xu, G., Wu, M., 2018. Graphitic carbon
Lukic, S., Menze, J., Weide, P., Busser, G.W., Winterer, M., Muhler, M., 2017. Decou- nitride based nanocomposites for the photocatalysis of organic contaminants
pling the effects of high crystallinity and surface area on the photocatalytic under visible irradiation: progress, limitations and future directions. Sci. Total
overall water splitting over b-Ga2O3 nanoparticles by chemical vapor synthesis. Environ. 633, 546e559.
ChemSusChem 10, 4190e4197. Zhao, B., Li, X., Yang, L., Wang, F., Li, J., Xia, W., Li, W., Zhou, L., Zhao, C., 2015. b-Ga2O3
Mastrantonio, M., Bai, E., Uccelli, R., Cordiano, V., Screpanti, A., Crosignani, P., 2017. nanorod synthesis with a one-step microwave irradiation hydrothermal
Drinking water contamination from perfluoroalkyl substances (PFAS): an method and its efficient photocatalytic degradation for perfluorooctanoic acid.
ecological mortality study in the Veneto Region. Italy Eur. J. Public Health 28, Photochem. Photobiol. 91, 42e47.