Simultaneous photocatalytic reduction/degradation of divalent nickel/naphthalene pollutants in aqueous solutions

Pollution of water and wastewaters by toxic heavy metals and organic compounds has been a major concern during recent decades. Among heavy metals, nickel is one of the potentially toxic metals, which has many applications in important industries such as stainless steel, nickel alloy, storage battery and electroplating (Joshi et al. 2011). Although Ni(II) has been identified as a component for participating in important metabolic reactions such as ureolysis, acidogenesis and methane biogenesis (Akhtar et al. 2004), but human contact with excessive amount of this ion can cause problems from skin irritation to lung damage and mal effects on the nervous system and mucous membranes (Caicedo 2008). In aquatic environments, nickel appears as Ni(II) and zero-valent Ni⁰; however, the potential toxicity for Ni(II) is reported as high whereas Ni⁰ is only slightly toxic (Chen & Ray 2001). The World Health Organization (WHO) recommends maximum concentration of nickel not to exceed 0.5 mg/L in waters (WHO 2008). Thus, reduction of Ni(II) to less harmful species would be very beneficial. Further, wastewaters from industries like electroplating usually contain organic compounds, which are used in different stages of manufacturing processes as balms, ductile material, polisher and so on. Naphthalene (NA), for instance, is widely used in different metal industries as brightener (Doan & Saidi 2008).

Several methods have been reported to treat wastewaters containing heavy metals and other pollutants. Ion exchange, membrane separation, adsorption, coagulation, electrolysis, and chemical precipitation are noteworthy. At the same time, it has been pointed out that these methods are coincident with some disadvantages including secondary pollution, large quantities of chemical reagents, high cost, poor treatment efficiency at low pollutant concentration and high energy consumption (Akhtar et al. 2004; Alqadami et al. 2017; Naushad et al. 2017).

A promising alternative is using advanced oxidation processes (AOPs) for pollutants removal due to high efficiency, wide adaptability and potential for simultaneous treatment of different pollutants (Saravanan et al. 2018). Among AOPs, heterogeneous photocatalysis by using semiconductors has been demonstrated with great success for the reduction of toxic metals as well as oxidation of organic compounds (Wang et al. 2004).

Upon irradiation of the semiconductor–solution interface with light energies greater than the semiconductor band gap, electron–hole pairs are formed in the photocatalyst conduction and valence bands, respectively (Barakat 2004). The electron–hole pairs are then separated and the adsorbed species on the sites of the catalyst undergo photo-induced redox reactions. For this aim, titania has been known as a very efficient photocatalyst since its band gap is around 3.2 eV with conduction band (CB) energy of −0.3 eV and valence band of +2.9 eV at pH 5.6 (Blake et al. 1991). Thus, any metal ion with a reduction potential less negative than −0.3 eV will be potentially reduced by photo-generated electrons of TiO₂. Also, the photocatalytic process, under aerobic conditions, can generate several reactive oxygen species (ROS) including superoxide anion radical hydroxyl and hydrogen peroxide through reactions with dissolved oxygen (DO) (Nosaka & Nosaka 2017). The ROS contribute in the oxidation and reduction reactions.

A variety of methods for simultaneous photocatalytic reduction of heavy metals and degradation of organic compounds have been proposed. For instance, the photo-reduction of Cu(II) and Se(IV) in the presence of formic acid and ethylene diamine tetra-acetic acid (EDTA) was investigated by Aman et al. (2011). Simultaneous Cr(VI) reduction and NA degradation in aqueous solutions by UV/TiO₂ has been investigated in another work by Gutierrez et al. (2008). A synergic effect between Cr(VI) reduction and NA degradation by UV/TiO₂ process was reported. In our recent study, simultaneous photocatalytic reduction of Cr(VI) and Ni(II) ions, coupled with degradation of sodium dodecyl benzene sulfonate, was reported (Saien & Azizi 2015). Enhanced photocatalytic treatment of pollutants is of interest due to a low dosage of titania nanoparticles and mild operating conditions.

Here, an attempt was made for the first time, for the photocatalytic reduction of nickel ions originating from nitrate or sulfate salts and in the presence of NA. The nitrate and sulfate salts are often present in the wastewaters of some industries like petroleum refining, electroplating, petrochemical, dyeing, textile and steel manufacturing (Ashour et al. 2008; Besharat 2010). NA, on the other hand, is extensively used as brightener and ductile agent in the nickel electroplating industries (Doan & Saidi 2008; Besharat 2010). Experiments were conducted in an efficient photo-sono reactor working with UV irradiation. For practical applications, the process performance was evaluated in terms of reaction kinetics, energy consumption and a specified process efficiency.

All the used chemicals, including nickel nitrate hexa-aqua (98.5%), nickel sulfate hexa-aqua (99%), 1-(2-pyridylazo)-2-naphthol (PAN) (>99%), ethanol (99.9%) and Triton X-100, naphthalene (99%), sulfuric acid (98%), sodium nitrate (98.9%), sodium sulfate (99%), potassium iodine (99.8%), sodium acetate (99%), ammonium dimolybdate (>99%) and sodium hydroxide (>97%), were purchased from Merck. Titanium dioxide nanosize particles, P-25 (>99.5%), was an Evonik product (anatase to rutile weight ratio of about 80/20) with Brunauer-Emmett-Teller (BET) surface area of 50 ± 15 m²/g and average particle diameter of 21 nm. All aqueous solutions were prepared using deionized water with conductivity of less than 0.08 μS/cm.

A 1.25 L cylindrical photo-sono reactor made of glossy stainless steel with 90 mm diameter and 200 mm height was used for the experiments. The light source was a 250 W mercury lamp (165 mm body length and 80 mm arc length) with wavelength range of 280–400 nm and the maximum emission of 365 nm. The lamp was installed centrally and immersed in the solutions. This configuration allows a homogeneous irradiation and perfect reflection for the beams contacting the reactor wall. At the beginning of each experiment, the ultrasound source (28 kHz, 60 W), located outside the bottom of the reactor, was working for 5 min, propagating waves in the reaction media in order to break up any particle cluster and providing homogeneous TiO₂ nanoparticle suspensions. During experiments, temperature was maintained constant by means of a stainless steel water-flow jacket from a thermostat bath.

To run experiments, a solution (1 L) containing the specified concentration of a Ni(II) salt and NA was prepared and was transferred into the reactor after pH adjustment by using either sulfuric acid or sodium hydroxide dilute solutions. Temperature was then set to the desired value. An amount of the catalyst particles (100 mg/L) was added, and prior to light irradiation, the suspension was sonicated and then mixed for 30 min in the dark to ensure adsorption–desorption equilibria of the pollutants on TiO₂ particles. All the experiments were performed while the content was continuously mixed with a simple agitator. Initial pollutant concentration in the mixed solutions was 5 mg/L of Ni(II) and 10 mg/L of NA (unless mentioned otherwise). The selection of nickel and NA initial concentration was with respect to the feasible final concentration at the end of treatments being close to the permissible limit for discharge into surface water: 2 and 0.06 mg/L for Ni(II) and NA, respectively (WHO 2008). According to a report, the NA concentration frequently found in electroplating, dyeing and textile wastewaters is within 0.1–300 mg/L (Nesterenko-Malkovskaya et al. 2012).

To follow the reaction progress, 2 mL samples (at least two samples) were withdrawn each time, and after vigorous centrifuging to separate nanoparticles, were divided into 1 mL samples for analyzing Ni(II) and naphthalene content.

As a representative of DO, the amount of H₂O₂ present was measured by colorimetry method with iodide. Briefly, 0.2 mL of diluted solution was mixed with 1.6 mL deionized water, 0.1 mL 1 M potassium iodine, and 0.1 mL 1 M sodium acetate buffer containing a few drops of ammonium dimolybdate as catalyst for the oxidation of by H₂O₂ to ⁠. The absorbance of solutions was measured by the UV-vis spectrophotometer at λ_max = 360 nm (Diesen & Jonsson 2014).

The separated TiO₂ particles, after process utilization, were washed with deionized water and then were dried at room temperature in a dark chamber. These particles as well as pure TiO₂ particles (before usage) were characterized in different ways.

In preliminary experiments, the reduction of sole Ni(II) ions, originating from nitrate or sulfate salts, within concentration of 5–20 mg/L and with the conventional pH range of 7.5–9.5 (Saien et al. 2014; Saien & Azizi 2015), was examined. Titania photocatalyst concentration was fixed at 100 mg/L and temperature (T) was set within 20 to 40 °C. Under conditions of pH = 9.5 and T= 35 °C, removal efficiencies of 71.4% and 77.2% were achieved with 5 mg/L initial concentration of Ni(II), originating from nitrate and sulfate salts, respectively (Figure 1).

Thus, the capture of holes by counter-anions causes the photo-promoted electrons on the CB to become more readily available for nickel reduction, inhibiting electron–hole recombination. The interaction between counter anions and titania particles leads to the easier adsorption of nickel cations and therefore more ion reduction. It is noteworthy that the influence of ion strength on the electrostatic interaction between ions and the catalyst surface follows the charge-density order, presented in the Hofmeister series. Since anion has a higher ionic strength compared to ⁠, higher ion reduction will correspond to nickel sulfate. In a previous study on the photocatalytic treatment of metal ions and organic pollution by sulfuric acid-modified titania particles, it was reported that sulfate ions improve the photocatalytic activity (Samantaray et al. 2003). Meanwhile, ions from nitrate salt may convert to nitrite ions, ⁠, due to photolysis, and in turn generate H₂O₂ molecules, which is a disturbing factor in the reduction of Ni(II) ions (Tzou et al. 2008).

Figure 2(a) shows significant enhancement in the nickel RE with pH variation from 7.5 to 9.5. It is well known that TiO₂ surface is negatively charged at pHs more than pH_PZC (point of zero charge) which is within 6.7–7.5 for P-25 TiO₂ (Wang et al. 2004). Increase in the solution pH causes the TiO₂ surface to find negative charge, leading to higher Ni(II) ion adsorption. Thus, the nickel reduction increased from 23.6 to 69.8% for nickel sulfate and from 17.6 to 62.1% for nickel nitrate when pH was increased from 7.5 to 9.5, respectively. In addition, Ni(OH)₂ formation and precipitation occurs under high alkaline pHs (Lin & Rajeshwar 1997; Shirzad-Siboni et al. 2012; Gutha et al. 2015). Accordingly, for treatments with just the photocatalytic process, the pH of solutions was limited to maximum 9.5.

Figure 2(a) also demonstrates that temperature enhances the nickel RE; however, amounts more than 35 °C, do not provide much variation, which may be due to the low level activation energy in the photocatalytic reactions (Saien & Azizi 2015).

The influence of nickel initial concentrations is shown in Figure 2(b). Obviously, RE decreases with nickel concentration. This is because the catalyst surface becomes highly covered at high nickel concentrations and the light reaching the catalyst surface is diminished. Therefore, electron–hole formation becomes severe.

Moreover, kinetic studies revealed a pseudo-first-order kinetics for photocatalytic reduction of Ni(II) under conditions of [Ni(II)]₀ = 5 mg/L, pH = 9.5 and T= 35 °C, for both used salts. The rate constants were obtained as 1.81 × 10⁻³ and 1.63 × 10⁻³ 1/min for sulfate and nitrate salts, respectively.

In this step, preliminary experiments on adsorption of Ni(II) and NA on TiO₂ particles (100 mg/L) were performed in darkness. Adsorption of Ni(II) ions, originating from both the salts, on TiO₂ nanoparticles was measured in the presence and absence of NA (10 mg/L). The percentage of adsorption after a typical 60 min time is shown in Figure 3. The surface charge of TiO₂ particles in the used solution is slightly negative under natural solution pH of 7.5, which favors the electrostatic attraction between the opposite charged TiO₂ particles and Ni(II) cations. This adsorption is decreased in the presence of NA due to competition between Ni(II) and NA species. NA is rather a neutral compound and its adsorption is favored around neutral pHs (Lair et al. 2008). On the other hand, NA adsorption is enhanced in the presence of Ni(II), so that 13.2% NA adsorption increases to 15.4% when accompanied by either of the nickel salts. The ‘salting-out’ effect can be the major reason for this observation in the mixture (Lair et al. 2008). An increase in the ionic strength due to the presence of nickel salts causes a decrease in the solubility of NA and the resulting adsorption of hydrophobic NA species on the surface of TiO₂ particles.

In the photocatalytic process with light emission, Figure 4 shows that, in contrast with darkness, the RE for Ni(II) reduction, from either of the salts, is enhanced upon coupling with NA degradation, and reaches 52.7% and 47.8% at pH 7.5. REs for Ni(II) reduction, in the absence of NA and under the same conditions were 25.8% and 22.3% for sulfate and nitrate salts respectively. This positive performance can be attributed to the consumption of photo-generated holes in NA degradation and inhibition of the electron–hole recombination. The photo-promoted CB electrons, generated by TiO₂ particles under UV irradiation, become more readily available for Ni(II) ions and hence an easier reduction will be established. The NA degradation is also enhanced in the presence of nickel ions. The RE of NA was raised from 36.5% to 50.6% with nickel sulfate and to 48.3% with nickel nitrate. The stronger NA adsorption in the presence of nickel salts can be considered to be due to the salting-out effect. The enhancement of Ni(II) reduction and NA degradation implies a synergism for the process.

It is worth noting that similar to a previous report on photocatalytic degradation of aniline with TiO₂ particles (Wang et al. 2009), a surface charge transfer complex can form between opposite charges of Ti⁴⁺ and O²⁻ from the TiO₂ structure and C⁻ and H⁺ from NA, respectively (Rei et al. 2002). Based on these assumptions, for which a simple scheme is illustrated in Figure 5, the formed NA-TiO₂ surface species can be photo-excited upon light absorption and initiate a direct electron transfer into the CB of TiO₂ particles. The photo-induced electrons, injected into CB, can easily be trapped by Ni(II) ions which act as electron acceptor and accelerate NA conversion to a radical cation (⁠⁺). Subsequently, intermediates such as 1-naphthol, 1,2-benzenedicarboxaldehyde, 1,4-naphthoquinone, alkyl phthalates and some others are formed by a series of electron transfer mineralization mechanisms (Wang et al. 2009). The in situ formation of Ni/TiO₂ and NiO/TiO₂ nanocomposites is one advantage of this process. This product can be used as an efficient photocatalyst in degradation of pollutants under UV or solar irradiations (Sharma & Lee 2015).

The Fourier transform infrared (FTIR, Perkin-Elmer, Spectrum 65, USA) spectra of TiO₂ powders, before and after usage, are shown in Figure 6(a). The broad peaks centered between 3,600 and 3,400 cm⁻¹ and at 1,630 cm⁻¹ correspond to the stretching vibrations of O-H groups and the bending vibrations of the adsorbed water molecules, respectively. The peak appearing below 1,200 cm⁻¹ corresponds to Ti-O-Ti vibrations. Further, in Ni/TiO₂ composite, a weak band is observed at 2,925 cm⁻¹, which can be due to the presence of Ni over the TiO₂ surface (Begum et al. 2008). Other bands at 1,713.8 cm⁻¹ and 1,387.5 cm⁻¹ are assigned to N-O in nitrate ions (Hankare et al. 2011) and the band between 1,450 and 1,350 cm⁻¹ is assigned to S=O in sulfate ions (Samantaray et al. 2003). The other appeared peaks correspond to the adsorption of trace amounts of sulfate or nitrate anions on the surface of titania nanoparticles. Further, the peak at about 1,000 cm⁻¹ can be assigned to Ti-O-C stretching bond (Pavia et al. 2009). It is significant that there is no strong peak, relevant to aromatic C-H and C-C stretching vibrations, within 2,800–3,000 cm⁻¹ and 1,672–1,696 cm⁻¹ regions, respectively. These confirm that adsorbed NA species on the TiO₂ surface are completely oxidized.

The scanning electron microscopy (SEM, Tescan, Mira 3, Czech Republic) was used to determine the morphology of the TiO₂ nanoparticles before usage and of collected and dried particles after usage. The SEM image (Figure 6(b)) reveals that the formed nanocomposite particles have regular and spherical shape after TiO₂ doping with nickel species (Ni⁰ and NiO). The bigger found particle size (25.6 nm compared to 20.2 nm) is similar to a report for synthesizing mesoporous (Ni²⁺/Ni³⁺)-TiO₂ particles with combined sol-gel and thermal decomposition methods (Rajendran et al. 2018). In a previous study using X-ray photoelectron spectroscopy to analyze used photocatalyst particles (Saien et al. 2014), the presence of nickel species on the titania surface (Ni⁰ and NiO) was confirmed, because of immediate oxidation of metallic nickel.

In addition, the TiO₂ particles, before and after usage, were characterized by N₂ adsorption–desorption porosimetry using a BET analyzer (BELSORP, Mini II, Japan). The results, presented in Table 1, indicate that the average surface area and the average diameter of TiO₂ particles before usage agree with the manufacturer's reported values (‘Materials and methods' section). Meanwhile, the collected TiO₂ particles after usage consistently have less surface area and large pores, compared with initial particles, respectively. These kinds of changes have also been reported for nanocomposites such as Ag/TiO₂ and NiO/TiO₂, synthesized by methods other than photocatalytic reduction (Rajendran et al. 2018; Saravanan et al. 2018).

The respective average Ni(II) and NA REs show an ascending and then a flat variation with pH. The highest average RE was found at pH around 7.5 for both the sulfate and nitrate nickel salts.

Investigation on temperature effect (Figure 7(b)) revealed that nickel reduction and NA degradation increase with temperature increase from 20 to about 35 °C and then remain almost constant perhaps due to a temperature disfavor effect on their adsorption. An average enhancement of about 12.4% was achieved for treatments below pH 7.5 after 90 min. Temperature helps the photocatalytic reactions to compete with the non-favorable electron–hole pair recombination; however, the low level activation energy of photocatalytic reactions gives a small temperature effect (Saien & Azizi 2015). Temperature of 35 °C can suit operations, giving 58.4% average RE (59.2% for Ni(II) and 57.5% for NA) for sulfate and 55.5% average RE (54.5% for Ni(II) and 56.6% for NA) for nitrate salts. In addition to the reasons given in the section on just Ni(II) reduction, more electron–hole recombination is expected under a low level of metal reduction, which in turn gives lower NA degradation.

The role of ROS in the process was investigated by continuous nitrogen gas purging into solutions. Figure 7(c) shows that 61.5% average RE (74.3% for Ni(II) and 48.8% for NA) and 54.2% average RE (65.9% for Ni(II) and 42.5% for NA) were obtained for sulfate and nitrate salts, respectively, and a higher reduction of Ni(II) ion and a lower oxidation of NA occurred. It has been reported that reduction of metals, with low reduction potential, is difficult with DO due to competition between metal ions and DO to attract electrons in the CB of TiO₂ (Yang et al. 2012). Meanwhile the efficiency of NA oxidation is diminished under these conditions. DO acts as an effective electron acceptor to extend the hole's lifetime and also may be involved in the formation of ROS, such as H₂O₂ molecule, which contribute in photocatalyst reactions of NA degradation (Nosaka & Nosaka 2017).

H₂O₂ is a rather stable molecule of the ROS, and the amount of this species was measured easily with iodide method (Nosaka & Nosaka 2017) either with or without nitrogen purging. The results in Figure 7(d) show that H₂O₂ formation highly increases at initial times and then tends to decrease due to its consumption in NA oxidation. As expected, H₂O₂ formation is very low in continuous nitrogen purging due to the absence of DO.

Owing to practical applications, the kinetics of photocatalytic reactions was investigated under mild mentioned operating conditions and during 90 min treatment. Results show that Ni(II) photocatalytic reduction and NA degradation in the mixed solution are well described by pseudo-first-order kinetic models as where ⁠, and k are the initial concentration and at any time, t, of Ni(II) or NA, and the kinetic rate constant, respectively. The rate constants as well as the coefficient of determination (R²) for different investigated cases are listed in Table 2.

Based on obtained rate constants, the electrical energy related to the photocatalytic removal of each pollutant under dominant conditions was obtained as 1,116.3 and 1,199.8 kWh/m³ for Ni(II) and 1,170.7 and 1,280.1 kWh/m³ for NA, when sulfate and nitrate nickel salts were used in the mixtures, respectively. The calculated values for simultaneous treatments are less than the electrical quantities required for treatment of each pollutant individually (Table 3). Considering the current electrical energy (industrial sector) price in the US market as 0.0702 US$/kWh in 2018 (US EIA 2018), the electrical energy cost relating to the above four electrical energy amounts is estimated as 78.4, 84.2 and 82.2, 89.9 US$/m³, respectively.

This criterion reflects the efficiency, achieved with respect to the level of energy consumption as well as the catalyst dosage. Accordingly, the effect of temperature and the reaction progress are involved in PE via the E_EC parameter.

The removal efficiencies as well as other described criteria are listed in Table 2 in comparison with other previously reported studies on photocatalytic reduction of Ni(II). The pseudo-first-order reactions are relevant under corresponding operating conditions. It is seen that despite nearly the same order of ⁠, the PE values for nickel ion, in this work, are about one order of magnitude higher than those of referenced works. Effective light in the used photoreactor and the synergism action of the reduction/degradation process are the main reasons for this very significant preference.

In this study the enhanced photocatalytic reduction of Ni(II) was demonstrated when simultaneously performed with naphthalene degradation and using a low dosage of titania nanoparticles. The photocatalyst hole consumption by naphthalene causes lowering of the electron–hole recombination and acceleration of the electron trapping by Ni(II) ions. Further, the formed naphthalene-TiO₂ species can be photo-excited upon light absorption and provide a direct electron transfer into Ni(II) ions. Confirming the results, the TiO₂ particles before and after usage were characterized by means of several techniques and it was found that average particles size, after usage, was increased due to nickel doping while the pore size was increased. The other important findings were: (i) the optimum mild operating conditions were pH 7.5 and 35 °C; (ii) under pertinent conditions, the removal efficiency of Ni(II) ions, for both the used salts, was about twice that in the absence of naphthalene; (iii) there appears to be more opportunity for the reduction of Ni(II) ions when the sulfate salt is used, compared with nitrate salt, individually or simultaneously; (iv) nitrogen purging in the samples confirmed the favorable DO role in naphthalene degradation and negative role in nickel reduction; (v) the Ni(II) ion reduction and naphthalene degradation obey pseudo-first-order kinetic reactions; (vi) electrical energy consumption analysis showed a relatively low required energy for the simultaneous treatment systems in comparison with individual treatments under the same conditions; and (vii) based on different criteria, including economic parameters, the significant preference of the process was revealed.

The authors wish to thank the university authorities for providing the financial support to carry out this project. The authors also acknowledge Evonik Industries for supplying TiO₂ (P-25) as a gift to our research group.