3.1 Catalyst characterization
3.1.1 XRD Analysis
The XRD patterns of the TiO2 photocatalyst before and after treatment are given in Fig. 2(a) and Fig. 2(b).
As can be seen, the catalyst consists of a bi-crystalline structure consistent with that of mixed-phase TiO2-(A/R). The diffraction peaks correspond to the planes assigned to the anatase phase (Reference code: 04-016-2837) and rutile phase (Reference code: 04-007-5403), as marked in Fig. 2. The similarity of the graphs (Fig. 2a, Fig. 2b) shows that there was no significant change in the structure of TiO2 before and after treatment, hence implying good catalyst stability. The peaks appear sharp which implies a high degree of crystallinity and subsequently, the higher photocatalytic activity of the TiO2 used in this study. Furthermore, the lack of additional peaks in the spectrum indicates the absence of impurities in the sample. The phase composition obtained in this study was 77.3: 22.7% (A/R) which closely approximates the commercial P25 (80A:20% R). The average TiO2 crystallite size can be calculated using the Scherrer equation:
\(DP=\frac{0.89ʎ}{Bcos\theta }\) (2)
where ʎ = X-ray wavelength (nm), θ = angle of diffraction and β = half-width of the diffraction peak (FWHM) of the (101) and (110) reflections of anatase and rutile, respectively. In the current work, the average crystallite size obtained was ≈ 12.70 nm and ≈ 42.13 nm for anatase and rutile respectively. A smaller particle size increases the number of active sites and consequently the transfer rate of charge carriers. This boosts the photocatalytic activity of the photocatalyst and subsequently, the degradation process.
3.1.2 SEM Analysis
To further verify the morphology of TiO2, SEM analysis was done as shown in Fig. 3 (a).
It appears that the photocatalyst is composed of clusters of nanoparticles that are roughly spherical in shape, which agrees with other authors (Dissanayake et al. (2020). The surface chemical composition was determined by SEM-EDX, and the spectrum is depicted in Fig. 3 (b). The elemental analysis suggests that the nanoparticles are composed of O and Ti atoms, with a weight ratio of 53:48% Ti/O. The appearance of C can be ascribed to contamination during the preparation of samples.
3.1.3 TEM Analysis
The TEM image (Fig. 4) depicts a relatively uniform polygonal shape with good particle dispersion. There is apparent aggregation of nano particles in small clusters. The pattern resembles that of mixed-phase TiO2-P25 reported in other studies (Lei et al., 2021).
The UV-Visible spectrum of TiO2 (SI file, Fig. S1) shows maximum absorption at ≈ 300 nm, which extends well into the Visible region (> 400 nm). This was an expected observation for mixed-phase TiO2 as already highlighted in section 1.0. The result signifies the possibility of utilizing a larger portion of solar radiation in photocatalysis, using mixed-phase TiO2. Similar findings were reported by other authors (Q. Wang et al., 2018).
3.2 Model Fitting and statistical analysis
The optimization results were treated with CCD to analyze the factorial design with experimental factors A, B, C, D, and E converted to coded levels, as displayed in Table 2.
Table 2 Factors and coded levels in CCD
Factor Coded factor rankings
|
|
Min (-1)
|
Moderate (0)
|
Max (1)
|
A
|
3.0
|
6.5
|
10.0
|
B
|
5.0
|
27.5
|
50.0
|
C
|
0.2
|
0.85
|
1.5
|
D
|
5.0
|
27.5
|
50.0
|
E
|
5.0
|
32.5
|
60.0
|
Min-minimum, Max-maximum
Table 3 represents the CCD matrix in coded levels alongside the actual and predicted removals.
Table 3
Coded experimental schedule with actual and predicted removals
% NVP removal (Y1) % TOC removal (Y2)
|
Run A B C D E Actual Predicted Actual Predicted
|
1
|
1
|
-1
|
1
|
1
|
-1
|
38.42
|
38.41
|
30.38
|
30.00
|
2
|
-1
|
0
|
0
|
0
|
0
|
88.69
|
88.65
|
83.69
|
84.00
|
3
|
-1
|
-1
|
1
|
-1
|
-1
|
71.74
|
71.86
|
66.75
|
66.88
|
4
|
1
|
-1
|
-1
|
1
|
1
|
56.22
|
56.1
|
34.29
|
34.39
|
5
|
-1
|
-1
|
1
|
1
|
1
|
76.11
|
75.99
|
73.91
|
73.82
|
6
|
1
|
0
|
0
|
0
|
0
|
59.65
|
59.74
|
49.36
|
49.57
|
7
|
0
|
0
|
0
|
0
|
1
|
67.42
|
67.94
|
62.76
|
61.79
|
8
|
1
|
1
|
-1
|
1
|
-1
|
50.14
|
50.14
|
34.51
|
34.01
|
9
|
-1
|
1
|
-1
|
-1
|
-1
|
77.72
|
77.84
|
62.14
|
62.15
|
10
|
0
|
0
|
0
|
0
|
0
|
67.14
|
67.36
|
59.1
|
58.78
|
11
|
1
|
-1
|
-1
|
-1
|
-1
|
47.05
|
47.16
|
29.78
|
30.09
|
12
|
0
|
0
|
0
|
0
|
0
|
67.75
|
67.36
|
57.92
|
58.78
|
13
|
1
|
-1
|
1
|
-1
|
1
|
31.22
|
31.2
|
34.29
|
34.73
|
14
|
0
|
0
|
0
|
0
|
0
|
66.57
|
67.36
|
61.27
|
58.78
|
15
|
1
|
1
|
-1
|
1
|
1
|
81.39
|
81.28
|
76.92
|
76.72
|
16
|
0
|
0
|
0
|
0
|
0
|
67.86
|
67.36
|
57.17
|
58.78
|
17
|
-1
|
-1
|
-1
|
-1
|
1
|
67.09
|
67.09
|
71.68
|
72.29
|
18
|
0
|
1
|
0
|
0
|
0
|
65.53
|
65.54
|
55.68
|
57.10
|
19
|
0
|
0
|
-1
|
0
|
0
|
61.96
|
61.97
|
50.29
|
49.84
|
20
|
0
|
0
|
0
|
0
|
-1
|
64.11
|
63.64
|
54.33
|
55.81
|
21
|
0
|
0
|
0
|
0
|
0
|
67.05
|
67.36
|
60.17
|
58.78
|
22
|
1
|
1
|
1
|
-1
|
-1
|
33.90
|
34.01
|
43.66
|
43.51
|
23
|
0
|
0
|
1
|
0
|
0
|
58.36
|
58.4
|
52.89
|
53.85
|
24
|
0
|
0
|
0
|
0
|
0
|
67.99
|
67.36
|
59.1
|
58.78
|
25
|
-1
|
1
|
1
|
-1
|
1
|
78.15
|
78.15
|
76.29
|
76.43
|
26
|
-1
|
-1
|
-1
|
1
|
-1
|
51.13
|
51.14
|
66.56
|
66.34
|
27
|
0
|
-1
|
0
|
0
|
0
|
60.72
|
60.75
|
51.35
|
50.45
|
28
|
-1
|
1
|
1
|
1
|
-1
|
70.37
|
70.38
|
78.92
|
78.23
|
29
|
1
|
1
|
1
|
1
|
1
|
43.97
|
43.84
|
47.96
|
47.59
|
30
|
1
|
1
|
-1
|
-1
|
1
|
41.65
|
41.64
|
42.75
|
43.08
|
31
|
0
|
0
|
0
|
-1
|
0
|
62.65
|
62.23
|
59.38
|
57.56
|
32
|
0
|
0
|
0
|
1
|
0
|
64.05
|
64.52
|
56.71
|
59.05
|
Table 3 shows that the NVP and TOC removals depended on significant variation in process parameters or combinations. The CCD empirical model describing the relationship between Y1, Y2, and the five process factors (excluding insignificant factors), is represented in Eqs. (3) and (4) respectively:
Y1 = 77.17–9.362 A + 0.8670 B + 40.02 C − 0.0382 D + 0.2070 E + 0.5581 A2 − 0.008315 B2 16.978 C2 − 0.007864 D2 − 0.002074 E2 − 0.03556 AB − 1.8297 AC + 0.04018 AD – 01833 AE − 0.08759 BC + 0.001195 BD − 0.000937 BE + 0.03627 CD − 0.01853 CE + 0.006055DE (3)
Y2 = 89.09–13.30 A + 0.414 B + 30.15 C + 0.169 E + 0.6534 A2 − 0.00989 B2 − 16.41 C2 + 0.01966 AB − 0.01809 AD + 0.1175BC − 0.0696 CE (4)
The Pareto chart, Fig. S5 (SI file) shows that all the studied factors (A, B, C, D, and E) were significant for NVP removal, which strongly agrees with most previous findings in photocatalytic degradation studies (Moradi et al., 2016). H2O2 had a negligible effect on TOC removal (SI file, Fig. S6). pH was the most dominant variable in NVP and TOC removals (Fig. S5 and Fig. S6), probably signifying the importance of NVP speciation in the aqueous environment.
The sufficiency of the proposed model in representing the experimental data was evaluated using regression statistics and ANOVA as shown in Table S4 and Table S5 (SI file). The model was suitable to represent the experimental data as indicated by the p-value of 0.000. The lack-of-fit (LOF)p > 0.05 for NVP and TOC removals, indicating good model predictability. The R2 and adj. R2 were close to each other and ≈ 1 for both NVP and TOC removals (Table S4, Table S5), demonstrating good model fit. The actual vs predicted (SI file, Fig. S7 and Fig. S8), as well as the residual vs. fitted response plots (SI file, Fig. S9 and Fig. S10) further emphasized the fitness of the model to describe NVP and TOC removals, respectively. Refer to the SI file, section S4 for further details of the ANOVA and statistical analysis.
3.3 Effect of operating parameters
3.3.1 Effect of pH
Figure S11 and Fig. S12 (SI file) show similar trends in the mutual interactions of experimental factors. The plots show that NVP degradation was most favourable under acidic conditions, reaching maximum NVP and TOC removals at pH 3. The findings resonate with Bhembe et al. (2020). To explain this, it is noteworthy to mention that pH poses a direct effect on the PZC of the catalyst, pKa of the compounds, acid/base speciation, pollutant hydrolysis, reactivity rate of radical species/oxidant, as well as the degradation pathway. In this study, the high NVP removal at pH 3 could be attributed to the low pKa value of NVP (2.8) (PubMed), which rendered the compound weakly basic and therefore susceptible to protonation at acidic pH. The enhanced removals could also be due to the proportionate increase of •OH radicals (with increased H+ concentration), minimum e− CB/h+ VB recombination, and smaller agglomerated catalyst particles prevalent in acidic compared to basic conditions (Rani & Karthikeyan, 2020). Furthermore, NVP exists mainly in its anionic form at pH > pKa, while TiO2 is positively charged at pH < PZC. The surface charge analysis (SI file, Fig. S2) shows that the PZC of TiO2 used in this work is ≈ 5.9 which lies within the 5.6–6.4 range widely reported in the literature (Rostami et al., 2021). Eqs. (5) and (6) detail the surface charge of TiO2 at PZC < pH < PZC respectively.
Thus, at pH < PZC, NVP degradation is supported by electrostatic interactions between NVP and the TiO2 surface. The positive surface charge of TiO2 decreased with increasing pH towards PZC, while the anionic form of NVP increased. This resulted in weakened NVP-TiO2 electrostatic interactions. Above PZC, NVP-TiO2 repulsion became dominant due to like charges (negative) on both the catalyst and NVP molecules further lowering the removals.
3.3.2 Effect of concentration
The synergistic interaction of NVP with other experimental factors was beneficial at low concentrations (< 27.5µg/l) (SI file, Fig. S11 and Fig. S12). A result that closely resonates with previous studies (Rostami et al., 2021). At higher concentrations, excess NVP molecules could have absorbed UV photons or shielded the catalyst surface from UV penetration. Thus, hindering catalyst activation and subsequently suppressing the production of reactive entities such as e− CB/ h+ VB, •OH and O2•−. The UV-Visible absorption spectra (SI file, Fig. S1) shows that NVP absorbs at ≈ the same wavelength range as TiO2. The deactivation of excited species by collision with ground-state substrates may have further inhibited radical production. Furthermore, excess NVP molecules may have competed with the generated intermediates for the reactive species. Thus, lowering the degradation rate, as well as suppressing the mineralization process.
3.3.3 Effect of catalyst
A similar trend as described for NVP concentration was observed for the catalyst in the dosage range 0.2–0.85 mg/L. A further increase in catalyst dosage achieved no enhancement in the removals (Fig. S11, Fig S12). The results agree with other studies (Wang et al., 2015). At lower dosages, increasing catalyst loading elevated the number of available active sites, producing more e− CB/ h+ VB upon illumination. This may have boosted ROS generation, resulting in higher degradation rates. Above 0.85 mg/L, aggregation and sedimentation may have set in shielding the active sites from direct irradiation. The excess catalyst could have also caused light scattering, reducing the photon flux at the catalyst surface. The overall effect of which, was diminished catalyst activation. Generally, high catalyst dosage is not desirable in water treatment or industrial processes, as it adds to the operational cost and increases sludge production. The latter may require a post-treatment step, which further increases the treatment cost.
3.3.4 Effect of H2O2
As already shown in section 3.2, the effect of H2O2 was negligible for TOC removal (Fig. S12) On the contrary, NVP removal was enhanced with increasing H2O2 concentration (Fig. S11), reaching a maximum at 50 µg/L. These findings are comparable with others (Lumbaque et al., 2020). In photocatalytic reactions, H2O2 may generate •OH according to Eq. (7–9). However, the results of the scavenger experiments (section 3.6) showed that O2•−and e− CB were irrelevant in this work. Therefore, only Eq. (7) will be discussed further. From Eq. (7), it is evident that •OH was produced by the photolytic decomposition of H2O2. It appears that increasing H2O2 concentration favored the decomposition process. Thus, more •OH was produced, resulting in higher degradation rates.
However, other studies (Shokri et al.,
2019) have reported that H
2O
2 acts as a scavenger of
•OH(Eq. (10)) at high concentrations and undergoes self-decomposition (Eq. (11)) into H
2O and O
2 molecules under alkaline conditions. Thus, reducing its availability as an oxidant and, most important, as a source of
•OH radicals. This could also explain the lowest removals observed at alkaline pH (pH = 10) in this work.
It, therefore becomes crucial to determine the accurate optimal pH and [H2O2] in photodegradation studies. In addition, high concentrations could result in increased treatment costs.
3.3.5 Effect of time
There was a relative improvement in NVP and TOC removals with increasing time (Fig. S11 and Fig. S12). The results confirm the recalcitrant nature of the compound, which required prolonged exposure to UV/Visible irradiation for effective degradation. The findings agree with other authors (Salarian et al., 2016).
Though possible explanations of observed trends in photocatalytic degradation studies are always given, it is generally challenging to match photocatalytic efficiency with Physio-chemical properties of the compounds, because of their complex nature and existence of both acidic and basic groups and/or atoms in their structures.
The proposed model predicted maximum NVP removal of 92.60% under optimum conditions of pH = 3.0, NVP concentration = 40 µg/L, TiO2 dosage = 0.92 mg/l, H2O2 concentration = 33.64 mg/l and time = 60 min. For TOC removal, a maximum of 89.40% was predicted under similar optimized conditions; pH = 3.0, NVP concentration = 34.09 µg/L, TiO2 dosage = 0.94 mg/l, H2O2 concentration = 50 mg/l and time = 60 min. The model validation studies showed
NVP and TOC removals of 89.23% and 85.71%, which closely approximated the predicted responses. Thus, substantiating the model fitness to represent the elimination of NVP from wastewater using the UV/TiO2/H2O2 process.
3.5 Photocatalytic degradation Kinetics
Preliminary work (S1 file, section S5) was conducted and revealed that the removals of NVP and TOC by adsorption, chemical oxidation, and catalytic oxidation were less than 10% after 60 min (Table S6). The results imply very slow reactivity of these processes and consequently negligible contribution to NVP degradation. The slight improvement in removal by photolysis (21.78%) (Table S6) was probably due to the presence of the aromatic rings and conjugated π system in NVP, which are good UV chromophores and therefore undergo direct photolysis (Fig. S1). The results agree with (Ngumba et al. (2020). With the addition of H2O2, NVP/TOC removals increased in the order UV/H2O2 (53.60/33.41%) < UV/TiO2 (67.14/49.72%) < UV/TiO2/H2O2 (89.23/85.71) (Table S6). This demonstrates the positive effect of H2O2 on the degradation process. Furthermore, the results suggest increased ROS production due to synergistic interactions of the involved processes. Similar observations have been made by other authors (Phan Nguyen et al., 2020)
The degradation kinetics was evaluated at different pH and H2O2 concentrations while keeping the other factors at optimum levels. The L-H model (Eq. (12)) was used to represent the reaction kinetics:
$$\text{ln}\left(\frac{Co}{C}\right)=k1+constant$$
12
Where k1 is the pseudo-first-order rate constant. Figure 5 shows the degradation kinetics of NVP.
The insert of Fig. 5 represents a plot of ln (Co/C) against time. R2 ≈ 0.999 (Table S7) for all pH levels. Thus, indicating good conformity to the pseudo-first-order kinetics. The rate constants for pH 3, pH 6.5 and pH 10 were 0.03676 min− 1, 0.01855 min− 1 and 0.00905 min− 1 respectively. The results translate to a 75.38% drop from pH 3 to pH 10, suggesting a strong degradation dependence on pH. This could be attributed to the acidic/basic speciation of NVP in wastewater that subsequently determined its protonation as well as affinity/repulsion tendency for the catalyst (section 3.3.1). Figure 6 depicts the effect of H2O2 on the reaction kinetics.
The effect of H2O2 on the rate constant (insert of Fig. 6) was quite opposite to that of pH (Fig. 5). The rate constant increased by 43.55% from 0.02075 min− 1 (5 mg/l) to 0.03676 min− 1(50 mg/l) (Table S7), demonstrating increased •OH production (Section 3.3.4).
The findings of the current work have been compared to previous studies as shown in Table 4.
Table 4
Comparison of current work with previous studies.
Pollutant/Source
|
Treatment method
|
Irradiation source
|
Catalyst /adsorbent
|
Conditions
|
Removal (%)
|
k1/min
|
Reference
|
Nevirapine
|
UV/TiO2/H2O2
|
UV-Visible
|
TiO2 (A-R)
|
NVP-40 µg/l
|
89.23
|
0.0367
|
This study
|
SWW
|
|
xenon lamp (250 W)
|
|
catalyst- 0.92 mg/l
|
|
|
|
|
|
|
|
H2O2- 33.64 mg/l
|
|
|
|
|
|
|
|
pH- 3.0
|
|
|
|
|
|
|
|
Time- 1 h
|
|
|
|
Nevirapine
|
UV/H2O2
|
Low pressure
|
|
NVP-20 µg/l
|
52.9
|
0.0792
|
a
|
SWW
|
|
Hg lamp (40 W)
|
|
H2O2- 20.4 mg/l
|
|
|
|
|
|
|
|
Time-0.5 h
|
|
|
|
Nevirapine
|
UV/(FL-BP@Nb2O5)
|
UV-Visible
|
FL-BP@Nb2O5
|
NVP- 5 mg/l
|
68.5
|
0.0152
|
b
|
SWW
|
|
Xenon lamp (900 W)
|
|
catalyst: 30 mg/l
|
|
|
|
|
|
|
|
pH: 3
|
|
|
|
|
|
|
|
time- 3 h
|
|
|
|
Nevirapine
|
adsorption
|
|
Graphene wool (GW)
|
NVP- 5 mg/l
|
84
|
|
c
|
SWW
|
|
|
|
GW: 10 mg/5ml
|
|
|
|
|
|
|
|
time- 72 h
|
|
|
|
para-chlorophenol
(p-CP)
|
UV/TIO2/H2O2
|
High pressure
|
TiO2-Degussa P25
|
p-CP-64.275 mg/l
|
98
|
0.0307
|
d
|
SWW
|
|
Hg lamp(100W)
|
|
catalyst- 1000 mg/l
|
|
|
|
|
|
|
|
H2O2-64.6 mg/l
|
|
|
|
|
|
|
|
pH-6.7
|
|
|
|
|
|
|
|
time- 2.5 h
|
|
|
|
Meropenem (MERO)
|
UV/TiO2/H2O2
|
2 UV lamp
(15 W)
|
TiO2-Degussa P25
|
MERO- 0.5 mg/l
|
99.1
|
0.1182
|
e
|
SWW
|
|
|
(immobilized)
|
H2O2-100 mg/l
|
|
|
|
|
|
|
|
time- 1h
|
|
|
|
2, 4 Dichlorophenol
(2,4-DCP)
|
UV/TiO2/H2O2
|
UV lamp
(125 W)
|
TiO2 anatase
|
2,4-DCP-100 mg/l
|
83.5
|
|
f
|
(wastewater)
|
& O2
|
|
|
catalyst- 1000 mg/l
|
|
|
|
|
|
|
|
H2O2-100 mg/l
|
|
|
|
|
|
|
|
pH-4
|
|
|
|
|
|
|
|
time- 1.5 h
|
|
|
|
Yellow procion (EFA)
|
UV/TiO2/H2O2
|
3 high pressure
|
TiO2-Degussa P25
|
RTE- 500ml
|
95.51
|
0.0078
|
g
|
real textile effluents (RTE)
|
|
mercury lamps
(250 W)
|
|
catalyst- 250 mg/l
|
|
|
|
|
|
|
|
H2O2-340 mg/l
|
|
|
|
|
|
|
|
pH-7.5
|
|
|
|
|
|
|
|
time- 6 h
|
|
|
|
Amoxicillin (AMX)
|
UV/TiO2/H2O2
|
UV lamp (365nm)
|
TiO2-Degussa P25
|
AMX-104mg/l
|
COD-74.62
|
|
h
|
Cloxacillin (CLX)
|
|
|
|
CLX-103 mg/l
|
|
|
|
Ampicillin (AMP)
|
|
|
|
AMP-105 mg/l
|
|
|
|
SWW
|
|
|
|
catalyst- 1000 mg/l
|
|
|
|
|
|
|
|
H2O2-100 mg/l
|
|
|
|
|
|
|
|
pH-5
|
|
|
|
|
|
|
|
time- 24 h
|
|
|
|
Detergent anionic surfactant
|
UV/TiO2/H2O2
|
UV lamp
|
TiO2
|
LWW- 100 ml
|
75.24
|
|
i
|
Laundry wastewater (LWW)
|
|
|
|
TiO2-40 mg/100ml
|
|
|
|
|
|
|
|
H2O2- 30 mM
|
|
|
|
|
|
|
|
Time − 24 h
|
|
|
|
Ngumba et al., 2020a, Bhembe et al., 2020b, Phan Nguyen et al., 2020c,Thu & Juang, 2015d, Verma & Kumar, 2014e, Dixit et al., 2011f, Garcia et al., 2007g, Elmolla & Chaudhuri, 2010h, Wahyuni et al., 2017i
As shown in Table 4, the rate constant obtained in this work was much higher than the 0.0152 min− 1 previously reported by Bhembe et al. (2020). This suggests that TiO2 supported the generation of reactive species better than the FL-BP@Nb2O5 photocatalyst used by Bhembe and colleagues (2020). The higher NVP concentration (5000 µg/L) used by the authors compared to the 40 µg/L used in this work (Table 4), may have impeded the catalyst activation (section 3.3.2). Ngumba and colleauges, (2020) obtained a much higher rate constant, 0.0792 min− 1 (Table 4) in an earlier study using the UV/H2O2 system. However, the overall NVP removal was lower (52.9%) compared to 89.23% obtained in the current work. The authors revealed that direct photolysis and chemical oxidation (H2O2, only) were less important in the degradation of NVP, which correlates with the current work. These results imply that the photolytic decomposition of H2O2 (Eq. 7) was the sole source of •OH in their work. Therefore, the higher rate constant obtained could be attributed to the rapid decomposition of H2O2 under the low-pressure Hg lamp (ʎ ≈ 254 nm) which provided a higher photon flux than the UV/Visible light used in this study. However, the lower overall NVP removal (52.9%) (Ngumba et al., 2020) suggests that the total number of generated •OH radicals was not sufficient to sustain the complete degradation process. These findings demonstrate that the synergistic effects of the UV/H2O2 and UV/TiO2 systems (UV/TiO2/H2O2) provided more •OH in this work. Furthermore, section 3.6 shows that additional •OH was supplied by the reactions of h+ VB and H2O (Eq. 15, 16). The longer irradiation time (60 min) used in this work compared to the 30 min used by Ngumba and colleagues (2020) may have further enabled the degradation of more NVP molecules. In another study, Adeola & Forbes. (2022) reported 84% NVP removal (Table 4). Although the results are comparable to the current study, the UV/TiO2/H2O2 was more efficient than the adsorption system. It is important to mention that the CCD optimized conditions in this work might have enhanced the production of ROS leading to higher NVP degradation. The other studies employed the OFAT optimization method, which is presumed to be less accurate. Table 4 also indicates that the UV/TiO2/H2O2 process was highly successful in the remediation of a diverse range of organic pollutants, with removals in the range of 75–99%. The discrepancy of results could be assigned to the pH-related factors highlighted in section 3.3.1, as well as other experimental conditions used.
3.6 Contributions of different reactive species.
Commonly used radical scavengers such as EDTA, potassium bromate, isopropanol and benzoquinone were used to investigate the contributions of h+ VB, e− CB, •OH and O2•− to NVP degradation, respectively. Figure 7 displays the effect of different scavengers on the degradation kinetics. The relevant rate constants and regression coefficients are represented in Table S8
There was a remarkable drop in the rate constant (Fig. 7) from 0.03676 min− 1, without a scavenger to 0.00299 min− 1 and 0.00406 min− 1 in the presence of isopropanol and EDTA respectively (Table S8). Thus, indicating significant inhibition of the degradation process, which translates to 91.87% and 88.96% degradation contributions by the •OH and h+ VB respectively. The slight drop in the rate constant for e− CB (15.61%) and O2−• (27.09%) (Table S8) imply their minimum participation and can be neglected. Therefore, it is plausible to conclude that NVP degradation was predominantly driven by the •OH and h+ VB. Other authors obtained similar results (Tafreshi et al., 2019. The findings give an insight into the degradation pathway. Thus, Fig. 8 illustrates the possible degradation mechanism for NVP. Eq. (13–19) represent the detailed proposed reaction pathway.
As Fig. 8 shows, the irradiation of TiO2 with a photon ≥ band gap facilitates the transition of a photoelectron (e− CB) from the full valence band (VB) to the empty conduction band (CB) leaving electron-deficient holes (h+ VB) in the VB. This process generates (e− CB/ h+ VB) pairs as shown in Eq. (13). Since charge recombination was offset by the mixed-phase TiO2(A/R) used in this study, it is assumed that the generated h+ VB were available in abundance. The h+ VB either degraded NVP directly (Eq. (18)), or indirectly via the formation of •OH (Eqs. (14), (15) and (20)). Additional •OH was contributed by the direct photolysis of H2O2 (Eq. (16)). The generated intermediates were possibly further oxidized into mineralization products such as CO2, and H2O (Eq. (19)).
3.7 Photocatalyst reusability
The ability to recycle the catalyst and use it multiple times is a desirable feature as it reduces the total wastewater treatment cost as well as minimizing water pollution due to residual catalyst content in the effluent. In this work, the reusability and stability of the TiO2 photocatalyst was investigated by recycling experiments under optimum conditions. Figure 9 shows the recycling results.
There was ≈ 11% NVP loss from cycle 1 to cycle 4 (Fig. 9). The results indicate good catalyst stability and the possibility of multiple re-uses without drastic deactivation of active sites. Figure 2(b) supports the findings. The slight decrease in NVP removal may be attributed to catalyst fouling, due to backward reactions and compounds chemical bound/adsorbed onto the catalyst surface. This deactivates the catalyst active sites, hindering the production of reactive species. The re-use of TiO2 photocatalyst in wastewater treatment has also been detailed by other authors (Tafreshi et al., 2019; Alvarez-Corena et al., 2016).