Preparation of Pd–Ni Nanoparticles Supported on Activated Carbon for Efficient Removal of Basic Blue 3 from Water

Pd–Ni nanoparticles supported on activated carbon (Pd–Ni/AC) were prepared using a phase transfer method. The purpose of synthesizing ternary composites was to enhance the surface area of synthesized Pd–Ni nanoparticles, as they have a low surface area. The resulting composite was characterized by scanning electronic microscopy (SEM), X-ray diffraction (XRD) and energy-dispersive X-ray spectroscopy (EDX) for investigating its surface morphology, particle size, percentage of crystallinity and elemental composition, respectively. The XRD data and EDX analysis revealed the presence of Pd–Ni alloys impregnated on the AC. Pd–Ni/AC was used as an adsorbent for the removal of the azo dye basic blue 3 from an aqueous medium. Kinetic and isotherm models were used to calculate the adsorption parameters. The most suitable kinetic model amongst the applied models was the pseudo-second-order model, confirming the chemisorption characteristics of the process, and the most suitable isotherm model was the Langmuir model, with a maximum adsorption capacity of 333 mg/g at 333 K. Different experimental parameters, such as the adsorbent dosage, pH, temperature and contact time, were optimized. The optimum parameters reached were: a pH of 12, temperature of 333 K, adsorbent dosage of 0.01 g and optimum contact time of 30 min. Moreover, the thermodynamics parameters of adsorption, such as Gibbs free energy (ΔG°), enthalpy (ΔH°) and entropy (ΔS°), showed the adsorption processes being exothermic with values of ΔH° equal to −6.206 kJ/mol and being spontaneous with ΔG° values of −13.297, −13.780 and −14.264 kJ/mol, respectively at 293, 313 and 333 K. An increase in entropy change (ΔS°) with a value of 0.0242 kJ/mol K, indicated the enhanced disorder at a solid–solution interface during the adsorption process. Recycling the adsorbent for six cycles with sodium hydroxide and ethanol showed a decline in the efficiency of the selected azo dye basic blue 3 up to 79%. The prepared ternary composite was found effective in the removal of the selected dye. The removal of other pollutants represents one of the possible future uses of the prepared adsorbent, but further experiments are required.


Introduction
Despite the rapid development in water reclamation technologies, water pollution is still a major global problem [1]. Bodies of water have been polluted with harmful chemicals due to rapid industrialization, leading to potable water scarcity [2]. Many pollutants, such as heavy metals, radionuclides, phenols, pesticides, herbicides, antibiotics, dyes, etc. [2][3][4][5][6][7][8][9], are constantly released from industry into the environment. Amongst the mentioned pollutants, dyes are the most hazardous, containing complex unsaturated organic compounds absorbing light in the visible region, thus limiting photosynthesis in aquatic environments [10,11]. Synthetic dyes are widely used in the clothing, paper, leather, oil, pharmaceutical and food industries as coloring agents [12]. In the textile industry, over 10,000 tons of dye are used per year, and about 10-15% of these dyes are in the form of dye residues being released into bodies of water [13]. According to a World Health Organization report, dye contributes about 17-20% to different pollution sources, and of this percentage, 10-15% are azo dyes, which, along with other hazardous effects (carcinogenic), have caused hepatotoxicity in animal testing [2][3][4][5][6][7][8][9].
The discharge of colour products into bodies of water adversely affects the aquatic environment due to their non-biodegradable complex molecular structures and their byproducts, especially in azo dyes [14]. Colour byproducts in wastewater effluents can cause toxicity depending on their exposure period and the dye concentration. Dyes absorb and reflect sunlight entering the water and can thus interfere with bacterial growth and inhibit photosynthesis in aquatic plants [15,16]. The toxicity of dyes may occur either due to the direct action of the original compound or through its intermediate metabolites, such as naphthalene, benzidine and other aromatic amines. These compounds are byproducts of micro-organisms, and the azo dyes' metabolites are also carcinogenic and mutagenic [17][18][19]. In human beings, these compounds can cause allergic dermatosis, respiratory diseases, contact dermatitis, asthma, changes in immunoglobulin level and colon and rectum cancer [20][21][22][23][24]. These secondary metabolites can also cause severe genotoxic and phytotoxic effects in plants [25].
Bimetallic nanoparticles (BNPs) that are composed of two different metals, as opposed to monometallic NPs, have drawn the attention of scientists [73]. Comparatively, bimetallic NPs have a large surface area, thus serving as effective catalysts compared with monometallic nanoparticles [74]. Their properties can be further improved to a great extent when these bimetallic nanoparticles are loaded on a high surface area material, such as activated carbon, graphene oxide, etc. [61].
In this paper, we have synthesized bimetallic Pd-Ni nanoparticles by a phase transfer method; NPs were then impregnated on activated carbon to enhance the surface area of the overall ternary complex. The morphology (shape and size), elemental composition and percentage of crystallinity were analyzed through different instrumental techniques. The supported bimetallic NPs were used for adsorption of basic blue 3 dye from an aqueous medium. The mechanisms and rate of the adsorption process were investigated by applying different kinetic and isotherm models. Furthermore, the adsorbent material used was regenerated by treatment with suitable eluents to check its reproducibility.

Instrumentation
The synthesized adsorbent was characterized by using a 30 KV scanning electron microscope (JSM5910, JEOL, Tokyo, Japan) with an EDX detector (INCA100/Oxford Instruments, Buckinghamshire, UK), X-ray diffractometer (JDX 3532, JEOL, Tokyo, Japan) with SEI and EDX detectors (INCA200/Oxford Instruments, Buckinghamshire, UK), and a CuKα source. The dye BB-3 concentration was quantified by measuring absorbance with a double beam UV-vis spectrophotometer (UV-1800, Shimadzu Scientific Instruments Inc., Kyoto, Japan) at a wavelength of 654 nm. The solution pH was determined by using a pH meter. Adsorption tests were conducted in a thermostatic water-bath shaker.

Synthesis of Supported Bimetallic Palladium-Nickel Nanoparticles
Pd-Ni nanoparticles were synthesized in a two-phase liquid system with slight modification in the reported procedure by adding capping agents, such as oleylamine and oleic acid [75]. About 5.6 mL of aqueous solution of PdCl 2 (0.022 M) and 5.6 mL of aqueous solution of NiCl 2 (0.033 M) were added to a solution of tetraoctylammonium bromide (ToABr) as a phase-transfer agent in toluene (0.538 g of TOABr in 37 mL of toluene). Then, the solutions were stirred for 2 h. The toluene phase was thereafter separated from the aqueous phase and discarded. The capping agent (23.4 µL of oleic acid and 46 µL of oleylamine) was then added to the mixture and was stirred for another 30 min. To reduce the Pd 2+ to Pd 0 and Ni 2+ to Ni 0 , 35 mL of NaBH 4 solution (0.2 M) was drop wise added into the reaction mixture and stirred for 12 h. After completing the process, the Pd-Ni colloidal particles were obtained, which were then washed with a dilute solution of KOH (5 mM) and distilled water. To support the Pd-Ni particles on the activated carbon, 80 mg of powdered activated carbon was added. After the addition of activated carbon, the mixture was stirred for 12 h. Then, the Pd-Ni powder supported on the activated carbon (Pd-Ni/AC) was separated and purified with chloroform, acetone, toluene and ethanol [75]. The prepared powders were kept in an oven at 70 • C for 2 h and stored in bottles. The summary of the entire procedure is shown in Figure 2.

Adsorption Experiments
The adsorption of basic blue 3 dye on the prepared adsorbent was performed at a pH of 12. About 0.01 g Pd-Ni/AC was mixed with a 10 mL solution of BB-3 in reagent bottles. The pH of the solutions was adjusted by the addition of a few drops of HCl (0.1 M) or NaOH (0.1 M) based on the requirement of the experiments. The solutions were mixed in a thermostatic water bath shaker for different time intervals. The reagent bottles were removed from the shaker at a predetermined time, the adsorbent was separated and the concentration of BB-3 dye in the filtrate was measured by a UV-vis spectrophotometer at a wavelength of λ max 654 nm. The dye's concentration was determined based on the calibration curve drawn for a concentration range. The adsorption capacity of adsorbent q t (mg g −1 ) and the removal efficiency were calculated by using Equations (1) and (2), respectively, according to a comparative study performed [76,77]: where q t is the amount of dye adsorbed (mg/g), %R is the percentage of dye removal efficiency, C 0 is the initial concentration of dye (mg/L), C t is the concentration of dye after adsorption, V is the volume of dye solution (mL) and W is the weight of the adsorbent (g). Various kinetics models were used to enumerate the kinetic constants of adsorption.
Adsorption isotherm experiments were performed on different concentrations, which ranged from 50 mg/L to 500 mg/L with 0.01 g of the prepared adsorbent, while the pH kept was 12. After the addition of the adsorbent, the solutions were stirred for 1 h. Thereafter, the adsorbent was separated from the solution through filtration, and at 654 nm the remaining concentrations were determined using a UV-vis spectrophotometer.
Similar effects of adsorbent dosage, pH and temperature were determined for 100 mg/L solutions, keeping the other parameters the same as the aforementioned.

Characterization of Pd-Ni/AC Morphology and Elemental Composition
The surface morphology of Pd-Ni/AC was determined by SEM analysis. The SEM images are given in Figure 3a-d. The micrographs show rough interfaces along with many holes and nanoparticles spreading over the material. Overall, the surface shows irregular particle channels for the adsorption of the specified dye.  Figure 4a shows the EDX elemental analysis of the prepared adsorbent. The results show that elemental atomic weight percentages of Pd-Ni/AC were found to be: C (56.63%), O (13.56%), Pd (29.30%) and Ni (0.51%). Similarly, Figure 4b presents the XRD pattern for the Pd-Ni/AC adsorbent. According to the Scherer formula, the calculated crystallite size is 7.4 nm while the crystallinity is determined to be 83%. Moreover, the characteristic peaks confirm the existence of the Pd-Ni alloy formation.

Adsorbent Dosage Effect on Adsorption
Different adsorbent dosages (0.005-0.03 g) were used to find the optimum amount of adsorbent used for BB-3 removal from the aqueous media in the adsorption experiments. The results are shown in Figure 5a, emphasizing that a linear increase occurred in the removal efficiency of dye with an increase in adsorbent dosage up to 0.01 g of the adsorbent, which was due to the increase of adsorbent surface area and the greater availability of the adsorption active sites. Hence, about 0.01 g of adsorbent dosage was used as the optimal dosage in the subsequent experiments.

Effect of pH on Adsorption
The acidic or basic media had a major effect on the nature and charge of the adsorbent, as well as on BB-3 adsorption. BB-3 adsorption was investigated at different pH levels (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14). According to Figure 5b, the maximum dye removal was obtained at pH (8)(9)(10)(11)(12). As the dye is cationic in nature, high pH favorability was encountered and optimum adsorption values were achieved. The second most contributing factor was soft-soft interaction between the functional groups that were present at the adsorbate and the metal centers at the adsorbent. At high pH (8)(9)(10)(11)(12), electrostatic interaction between the adsorbent and the dye, as well as the dye's metal center interaction, enhanced the BB-3 removal efficiency. At lower pH, the decrease in the removal efficiency of BB-3 might be due to H + ion competition with BB-3 molecules for the adsorption sites at the adsorbent.

Effect of Contact Time on Adsorption of BB-3
It was found that the extent of dye removal was improved by increasing contact time, and a maximum value was reached after 30 min. Figure 5c demonstrates that after 30 min contact time, there was no significant effect on dye adsorption observed. Thus, a 30 min reaction time was considered to be the optimum contact time.

Adsorption Kinetics of BB-3 Adsorption
Adsorption kinetics must be considered for understanding the precise adsorption mechanism. Adsorption kinetics were applied for the assessment of adsorbate-adsorbent interaction and adsorption parameters [78]. Different kinetics models were applied to assess the kinetics of BB-3 adsorption on the prepared adsorbent.
The linear form of the pseudo-first-order kinetic equation can be given as follows [76,77]: log q e − q t = logq e − k 1 2.303 (3) In this equation, k 1 (min −1 ) is the equilibrium rate constant, q e and q t are the amounts of dye adsorbed at equilibrium and t is time. The values of the constants k 1 and q e were determined from the slope and intercept of the log (q e − q t ) versus t [79] plot, which is shown in Figure 6a ( Table 2).
The correlation coefficient R 2 was less than 0.9, which confirmed that this model did not show the best fit and proved that the sorption process was not a physisorption process.
The pseudo-second-order kinetic equation was expressed as [76,77]: where k 2 (g/mg/min) is the pseudo-second-order rate constant. The slopes and intercepts of the plot of t/q t versus t (Figure 6b) were used to calculate the q e (mg/g) and k 2 values [80]. The pseudo-second-order constant k 2 (g/mg/min) and correlation coefficient R 2 values for the adsorption process of BB-3 on the Pd-Ni/AC were calculated and are listed in Table 2.
To understand the mechanism of the adsorption, an intraparticle diffusion model was used [76,77]: where k id is the intraparticle diffusion constant, and C shows the thickness of boundary layer. The values are shown in Table 3. Figure 6c showed that the correlations do not pass through the origin, from which it can be concluded that the intraparticle and film diffusion are dominant. The Elovich equation was applied to explain some aspects of the adsorption process. A mathematical form of the model can be expressed as [76,77]: where α is the initial sorption rate (mg/g/min), and the parameter β is related to the extent of surface coverage and activation energy for chemisorption (g/mg) [81][82][83]. The q t vs. lnt was plotted and is shown in Figure 6d, and the value of α and β were determined, which are listed in Table 3. The Natarajan and Khalaf equation gives the relationship between the initial dye concentration C 0 and the concertation at a certain time C t [76,77]: where k is the Khalaf constant, its value calculated by plotting log C 0 C t against time t (Figure 6e). The values of constants are given in Table 3.    Natarajan and Khalaf equation

Isotherm Study
In order to identify the superficial properties and affinity of the adsorbent for adsorbate, isotherm models were used. The isotherm study signified how the adsorbate interacted with the adsorbent and helped in enumerating adsorption capacities. Langmuir, Freundlich and Temkin models were applied to explain the experimental data [84].

Langmuir Isotherm Model
This model can be successfully applied to the adsorption of different organic and inorganic pollutants on various adsorbents. According to this model, the adsorption occurs in monolayers, and the process takes place at certain homogenous sites contained by the adsorbent [76,77]. The model can be expressed as C e q e = C e Q m + 1 K L Q m (8) where C e , q e , K L and Q m are the concentration of dye at equilibrium (mg/L), amount adsorbed at equilibrium (mg/g), Langmuir constant (L/g), and maximum adsorption capacity (mg/g), respectively [85]. Plotting C e /q e , versus C e enables us to determine the value of K L and Q m (from slope and intercept, Figure 7a-c). Their values, along with correlation coefficient R 2 , are given in Table 4. The Langmuir maximum adsorption capacities for sorption of BB-3 on Pd-Ni/AC were found to be 238.95, 277.78 and 333.3 mg/g at 293, 313 and 333 K, respectively. Langmuir isotherm provided high correlation coefficient values and fit well with the experimental data.

Freundlich Isotherm Model
This model concerns non ideal and reversible adsorption processes, which are not limited to the formation of monolayer [76,77]. This model can be expressed as lnq e = lnKf + 1 n ln C e (9) where C e , q e , Kf and 1/n are the equilibrium dye concentrations, amount of dye adsorbed at equilibrium, adsorption capacity at unit concentration and adsorption intensity, respectively. When 1/n = 0, the process will be irreversible; when 0 < 1/n < 1, the process will be favorable and when 1/n > 1, the process will be unfavorable [86].
A plot of lnq e versus C e enables us to determine the values of the above parameters from the slope and intercept (Figure 7d-f), the values of which are given in Table 4. The values of 1/n were: 0.3181, 0.3139 and 0.3379 at 293, 313 and 333 K, respectively, which indicates that the adsorption process was energetically favorable.

Temkin Isotherm Model
This model has two basic assumptions: (i) with the coverage of the surface, the heat of adsorption increases due to a positive interaction between adsorbate and adsorbent, and (ii) adsorption is described as a homogeneous distribution of binding sites of uniform energies up to a certain limit [76,77]. This model can be expressed as q e = B 1 lnK T + B 1 lnC e (10) where B 1 and K T are the heat of adsorption and the equilibrium binding constant (L/mg) respectively [87]. Plotting q e versus lnC e enabled us to find out the values of these constants (Figure 7g-i). Their values are given in Table 4. According to Table 4, the R 2 values show that the adsorption of BB-3 Pd-Ni/AC isotherm followed the Langmuir isotherm, as its value was greater than that of the Freundlich and Temkin isotherms. This indicates that the present adsorption processes can be better described by the Langmuir than the Freundlich and Temkin models, suggesting that adsorption occurs as the monolayer on the homogenous adsorbent surface.

Thermodynamic Study
The thermodynamic parameters attained for the adsorption of BB-3 onto Pd-Ni/AC were calculated using the following equations [76,77]: (11) where K c = Q e /C e .
∆G • , ∆H • , ∆S • , K c and K a are the changes in Gibbs free energy, enthalpy, and entropy; the ratio of equilibrium adsorption; equilibrium concentration and the Langmuir constant, respectively. These values are given in Table 5, while the Van't Hoff plot is given in Figure 8. At different temperatures, the Gibbs free energy change (∆G • ) was found to be negative, indicating that the adsorption process was feasible and spontaneous. The decline in ∆G • values was found as temperature increased, indicating that the adsorption process occurs more favorably at lower temperatures [88]. The increased randomness at the dye/adsorbent interfaces and a strong affinity for the adsorption were verified by the positive value of change in enthalpy (∆H • ) [89]. During adsorption, a positive entropy change (∆S • ) indicated increased disorder at the solid-solution interface.

Regeneration Study
To determine the regeneration efficiency of Pd-Ni/AC, the sample was washed with suitable eluents (dilute sodium hydroxide and ethanol) and then used several times, and the removal efficiency was monitored. From Figure 9 it was found that the removal efficiency was reduced up to 79% after six rounds of regeneration, showing that Pd-Ni/AC can be used with a sufficient efficiency up to several cycles.

Comparing Adsorption Capacities of Reported Adsorbents with Current Adsorbent
Pd-Ni nanoparticles as an adsorbent showed maximum adsorption capacity compared with previously reported results [90][91][92][93][94] (Table 6). Hence, Pd-Ni nanoparticles may be an effective adsorbent for the removal of BB-3 dye and other pollutants. Previous studies have shown the assessment of the effects of pH and other factors on the solubility and performance of nanomaterials, which could be applied similarly in future studies [95][96][97].

Conclusions
The aim of this study was to synthesize the Pd-Ni nanomaterials and to impregnate them onto activated carbon to enhance the surface of prepared nanoparticles. The prepared ternary adsorbent was used as an efficient adsorbent for the removal of basic blue 3 (BB-3) from water. The bifunctional alloy formation was confirmed from the metal analyses by XRD and the surface morphology by SEM analysis. About 94% removal of the BB-3 was achieved through the prepared adsorbent, indicating its high efficiency to remove the selected dye from water. Various equilibrium and kinetics models were used to calculate different parameters of the adsorption, where pseudo-second-order kinetics and Langmuir models were found to be the best models that accommodated the experimental data with high R 2 values. The calculated thermodynamic parameters showed the process to be spontaneous (∆S • = 0.0242 kJ/mol K and ∆G • = −13.297, −13.780 and −14.264 kJ/mol, respectively at 293, 313 and 333 K) and exothermic (∆H • = −6.206 kJ/mol) in nature. The prepared adsorbent was effective in the removal of the selected dye and further tests are needed to evaluate its applicability for other pollutants as well. Funding: This research was funded by project number T190087MIMV and the European Commission, MLTKT19481R "Identifying best available technologies for decentralized wastewater treatment and resource recovery for India" and SLTKT20427 "Sewage sludge treatment from heavy metals, emerging pollutants and recovery of metals by fungi", and by projects KIK 15392 and 15401 by the European Commission.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.