3.1. X-ray Diffraction (XRD) Characterization
The structures and phases of the composite observed from the diffraction pattern are shown in
Figure 2. The standard atomic parameters of rutile TiO
2 which is tetragonal with a space group of
P42/
mnm were taken from the Inorganic Crystal Structure Database (ICSD) 98-016-5920; ICSD 98-016-3491 for the Na
2Ti
6O
13 (NTO) structure was monoclinic with space group
C12/
m1, and ICSD 98-008-2085 for the gold structure was cubic with a space group of
Fm–3
m [
42,
43,
44]. The structure of the rutile/NTO composite showed a pattern with peaks at 2θ = 27.5°(110), 36.2°(011), 39.3°(111), 41.2°(111), 54.3°(121), 56.6°(220), 64.0°(002), 69.0°(031), and 69.8°(112) which were derived from the rutile phase structure. Additionally, peaks at 2θ = 11.8°(200), 14.1°(20-1), 24.5°(110), 30.1°(310), 33.4°(402), 43.2°(40-4), 44.1°(602), and 48.6°(020) were obtained from the structure of the Na
2Ti
6O
13 phase. For the Au/TiO
2/NTO composites, there were additional peaks at 2θ = 38.3°(111), 64.6°(022), and 77.7°(113) originating from the gold phase structure.
The percentage of crystal structure was calculated with the Rietveld and refined method using the HighScore Plus software (PANalytical 3.0.5), while scale factors, zero shift, and coefficients of shifted polynomial functions were used to match the background [
45]. The goodness of fit (GoF) match value for the experimental XRD pattern was calculated to confirm the accuracy of the Rietveld refinement. GoF is a statistical model which describes how well the experimental results are obtained with a series of observations and is calculated using Equation (6).
where
Rwp (weighted profile R-factor) is the simplest difference index and
Rexp (expected R-factor) is the expected “best
Rwp” quantity.
Table 1 shows that the GoF value in the sample ranged from 3 to 5, proving the fair purity and good crystallization for our samples [
46].
Table 2 presents the percentage of the Rietveld refinement phase. The results show that before the addition of gold, two phases were formed, namely rutile TiO
2 by 43.4% and NTO by 56.6%. NTO was formed because the synthesis process carried out using the sol-gel method with a strong alkali NaOH intercalated Na
+ ions into the TiO
2 structure which then form sodium titanate with a monoclinic structure [
21]. The crystal structure of NTO was rutile as shown in
Figure 3. After composition was performed, the percentage of gold increased with the higher addition of the HAuCl
4 precursor. The gold percentages in a row were 0.2, 3.8, and 6.0% for 1, 2, and 3% Au/TiO
2/NTO composites, respectively.
Table 2 shows the percentage crystallinity of the samples. Based on the results, the crystallinity of the composite increased with the addition of gold. The crystallinity of the TiO
2/NTO composite was 76.05%, while that of Au/TiO
2/NTO increased to 83–84%, and the highest with 3% increase was found in Au/TiO
2/NTO of 84.48%. Large crystallinity is needed in photocatalysis to avoid the possibility of electron–hole recombination and increase photocatalytic activity [
36].
The unit cell parameters of the sample are shown in
Table 3. The rutile crystal of the synthesized composite had a smaller lattice volume compared to the standard, namely ICSD 98-016-5920 [
42]. Based on the results, the rutile lattice volume of the TiO
2/NTO composite was 62.414 Å
3, while in the 1% Au/TiO
2/NTO composite, the volume decreased to 62.371 Å
3. Moreover, at higher concentrations of gold composite, the volume also decreased. The rutile lattice volumes for 2 and 3% Au/TiO
2/NTO were 62.436 Å
3 and 62.416 Å
3, respectively. The results also showed that the synthesized NTO had a larger lattice volume than the standard, namely ICSD 98-016-3491 [
43]. With the addition of gold, the volume of the crystal lattice decreased at 1 and 2% Au/TiO
2/NTO, but increased at 3%. Meanwhile, in gold crystals, the synthesized product had a different lattice volume, which was larger or smaller than the standard, namely ICSD 98-008-2085 [
44]. The trend of the gold concentration effect on decreasing or increasing lattice volumes for rutile, NTO, and gold were further examined.
3.6. Photocatalytic Activity Test on Phenol Degradation
Several previous studies have reported a composite of NTO as a photocatalyst. Examples of studies that used NTO composites with various compounds are presented in
Table 4. NTO used in this study was synthesized from the addition of a NaOH base to TiO
2. However, studies on TiO
2/NTO composites with Au have not been previously reported.
The photocatalysis activity was tested against a decrease in the concentration of phenol to determine the performance of the synthesized photocatalyst compared to P25 TiO
2. Phenol is found in industrial waste with different concentrations depending on the type of industry. For example, the pulp and paper industry produces 20–80 mg/L of phenol, the rubber industry 3–10 mg/L, and the textile industry 100–150 mg/L [
57,
58,
59]. This study used a concentration of 20 mg/L because it is the value that exists in industrial waste. The test was carried out using a simulated sample of phenol standard solution which was analyzed using high-performance liquid chromatography (HPLC)
Figure 8a shows that the decrease in concentration until the 60th minute is an adsorption effect of the composite and the test was carried out without UV irradiation. Meanwhile, the test with UV irradiation was conducted at 60 to 180 min and the results showed a significant decrease in concentration after photocatalyst activation by light. The efficiency of reducing phenol concentration is demonstrated in
Table 5. Based on the results, the 3% Au/TiO
2/NTO composite had the largest degradation percentage of 82.94% with an adsorption percentage of 3.39%, while TiO
2/NTO in the absence of gold showed a small percentage of degradation, namely 57.17%. This proves that the presence of gold enhances the photocatalyst to efficiently degrade phenol. In addition, the best photocatalysis was shown with 3% gold composite, while at 1 and 2% there was a decrease in the efficiency. However, the 3% gold composite also showed the best phenol adsorption of 3.39% compared to the 2% composite with 2.88% and the lowest was found in the 1% composite of 1.81%. This shows that the higher the gold composited, the better the photocatalyst adsorption capacity, leading to more active sites for better photocatalytic activity.
Another factor that enhanced photocatalysis in the presence of gold was the higher crystallinity of the Au/TiO
2/NTO composite which ranged from 83 to 85% compared to the composite without gold, namely 76.05%. Crystallinity reduces the occurrence of recombination [
36], thereby increasing photocatalysis efficiency. In addition, the presence of heterophase also enhances the formation of a synergistic effect of electrons, which effectively stimulates their transfer from one phase to another [
60,
61].
The reaction between the photocatalyst and phenol corresponds to a multiphase surface reaction. The Langmuir–Hinshelwood first-order kinetic model was used to evaluate the reaction kinetics [
62,
63].
where
r0 is the initial photocatalytic efficiency (mg/L·min), t is reaction time (min),
k is reaction efficiency constant (min
−1), K is reaction equilibrium constant, and C
s is reactant concentration (mg/L);
k and K are determined by several factors in the reaction system, including the catalyst concentration, light intensity, initial concentration of reactants, reaction temperature, physical properties of reactants, and amount of oxygen available.
When the substrate reaction concentration is low, KC ≥ 1; this can be simplified to a first-order equation, namely [
39]:
where
k1 is the reaction constant for the first-order reaction. At the start of the reaction, t = 0 and C
t = C
0; the equation can be obtained after deformation as follows [
64]:
where C
t is the concentration of phenol in solution at t minutes, C
0 is the initial concentration of phenol, and b represents the constant.
Figure 7b shows that ln (C
t/C
0) and t are in a good linear relationship and their performance is consistent with the first-order reaction. The kinetic constant of the reaction can be determined to estimate the total reaction rate, and then compare the photocatalytic efficiency under different conditions. The first-order kinetic equations, reaction rate constants, and correlation coefficient (R
2) of the photocatalytic reactions on different catalysts are shown in
Table 5. Based on the results, the reaction constant
k1 increases in the presence of the gold composite, indicating that the composite promotes the photocatalytic reaction process. In addition,
k1 showed the highest value in 3% gold composite, namely 0.0146 min
−1, which was higher than P25 TiO
2 at 0.0108 min
−1. These results indicate that the photocatalytic reaction efficiency of phenol degradation with a 3% Au/TiO
2/NTO catalyst is 200% and 140% more effective than without gold composite and commercial P25 TiO
2, respectively. This implies that the appropriate amount of gold on the TiO
2/Na
2Ti
6O
13 composite surface will significantly increase its photocatalytic ability.
Under visible light irradiation, the light-generated hole–electron pair appeared in the gold electron state due to surface plasmon resonance (SPR) [
65]. The conduction band energy of TiO
2 and NTO was lower than the Fermi energy level of gold, but the electrons generated from gold can move to the conduction band of TiO
2 or NTO. Furthermore, the reduction of O
2 by light-induced electrons on the rutile surface was not efficient, but in the presence of NTO, the electrons were more active for O
2 reduction. Therefore, electron-preferential transfer from TiO
2 to NTO can effectively suppress photogenerated electron–hole pair recombination and accelerate the photodegradation process [
66]. The excited electrons were transferred to rutile TiO
2 and then NTO to initiate the reaction with dissolved oxygen, while the holes reacted with H
2O or OH
–, thereby avoiding electron–hole recombination, which can enhance the photocatalytic effect. In other words, the combination of gold and TiO
2/NTO is expected to produce charge separation conditions with relatively low oxidation potential (Au
+) and with the same reduction potential for the NTO conduction band as TiO
2/NTO.
A schematic representation of the mechanism of phenol degradation by Au/TiO
2/NTO is shown in
Figure 9. In the literature, it has been reported that the conduction and valence band potentials in NTO are more negative than in rutile [
67]. The hole charge transfer process will lead from the valence band of the rutile phase to the valence band of the NTO phase because it has a higher potential, which will further oxidize the compounds present in the medium. Meanwhile, the photo-induced electrons from the conduction band of the NTO phase will migrate to the conduction band of the rutile phase because of the lower potential. Moreover, the presence of Au can trap the resulting photo-induced electrons in the conduction band and prevent them from returning through the Schottky barrier formed between Au and TiO
2/NTO. This interface is beneficial because it contributes to reducing the electron–hole recombination rate and improving photocatalytic performance [
68].
3.8. Large-Scale Economic Evaluation
In order to evaluate the possible use of this technology further in real industry, it is important to carry out a cost analysis of each process to determine whether it can be applied in real case scenarios [
72,
73,
74]. A similar methodology was carried out by Durán et al. [
75] to determine the chemicals and energy cost. The engineering design of the phenol removal unit using the Au/TiO
2/NTO catalyst was proposed based on the experimental results of the batch obtained, catalyst dose of 0.15 g/L, contact time of 2 h, and light intensity of 300 watts. In this case, the reactor is designed to be able to treat phenol waste with a phenol effluent flow rate of 3 L/min, based on previous research [
76]. In large-scale photocatalysis, reactor design will determine the effectiveness and efficiency of photocatalysis [
77], one of which is reactor geometry [
78]. In this study, the reactor to be used is cylindrical, with the volume calculated based on Equations (11) and (12) [
79].
where
Q is the volume flow rate of phenol waste (m
3/day),
V is the required volume of the photocatalysis unit (m
3), t is the contact time (h),
n is the number of tanks,
D is the tank diameter (cm), and
L is the tank length (cm).
The designed reactor volume is 0.36 m
3 with a diameter and tube length of 68 and 100 cm, respectively. Wastewater containing phenol with a flow rate of 3 L/min is assisted by using two pumps. Then, in the wastewater reactor, it was stirred at a speed of 300 rpm while UV light was directed using a 300 W UV lamp. Then, an economic feasibility assessment is carried out which is presented in
Table 6. The factory operating hours are 24 h a day and five days a week (288 days/year). The total cost is calculated based on the aspects of fabrication, mechanics, operations, and energy. The fabrication cost for a photocatalytic reactor with a volume of 0.36 m
3 is USD 1440 and for spare parts including electrical and mechanical equipment, it is USD 900 [
79]. Furthermore, the total operating cost is determined as the sum of the material, energy, and labor costs [
80]. The cost of preparing the Au/TiO
2/NTO composite catalytic material is calculated to be 0.961 USD/m
3 for TTIP, 0.370 USD/m
3 for HPC, 0.057 USD/m
3 for NaOH, 0.011 USD/m
3 for HCl, 0.097 USD/m
3 for NaBH
4, and 269.4 USD/m
3 for HAuCl
4 based on the chemical cost at the local supplier. The cost of energy consumption is calculated for a 300 W lamp, pump, and photocatalytic reactor stirrer with a total of 0.493 USD/m
3 based on the tariff in Indonesia (0.074 USD/kWh) [
81]. The cost of the catalyst material treatment is 5% of the material cost, which is USD 13.545 [
79]. The total salary of a technician working in Indonesia for three people is around USD 1500 gross per month (14.467 USD/m
3).
The reuse of catalytic materials can reduce operating costs. In addition, conducting research in sunlight significantly lowers research costs. In this case, a UV light simulator was not used and could reduce the total cost by up to 76% based on previous studies [
82]. However, without gold, the total operational cost without technician cost can be calculated to be 2.064 USD/m
3 cheaper than the previous study for the degradation of 2,4-dicholorophenoxyacetic acid of 11.71 USD/m
3 [
83]. In the presence of gold, the cost of the catalyst increases considerably to 284.934 USD/m
3, which is a limitation in this study to be applied in industry.
Break-even analysis is determined to obtain the minimum requirement of the production capacity and profitability index to obtain information about the profit [
84]. In this study, the processing cost of the treatment outside the factory is estimated at 75 USD/m
3 [
79]. In this study, profit can be achieved if the catalyst material design is calculated without using gold (TiO
2/Na
2Ti
6O
13). Therefore, in this analysis, the cost of expenditure is calculated without using gold. To determine the number of break-even points/
QBE (m
3) and profit/
P (USD), Equations (13) and (14) [
79] are used.
where FC is fixed cost (USD), r is revenue (USD), v is variable cost (USD), and
Q is the volume flow rate of phenol waste (m
3/year).
The fixed cost used is USD 2340 which comes from the photo-catalysis reactor fabrication and the required spare parts, while the operating cost is taken as a variable cost of USD 16.53 (cost without gold). The break-even analysis is shown in
Figure 11a. It is found that the break-even quantity is 40.02 m
3. After that, a payback analysis was carried out to estimate the possibility for the year of profit [
85]. The payback period for 5 years of operation is shown in
Figure 11b. The benefit from treated waste is 1244.16 m
3/year for USD 70,405. Thus, it is obtained that the return can be determined as 0.46 years. These results indicate that the pilot-scale design proposed for phenol wastewater treatment has the great advantage of using composites without gold. However, future research should find alternative materials to replace gold so that costs can be reduced.