Sono-electrochemical recovery of metal ions from their aqueous solutions

Journal of Hazardous Materials 318 (2016) 379–387

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Sono-electrochemical recovery of metal ions from their aqueous solutions Bingfeng Dong a , Asher Fishgold a , Paul Lee b , Keith Runge a , Pierre Deymier a , Manish Keswani a,∗ a b

Department of Materials Science and Engineering, The University of Arizona, 1235 E. James E. Rogers Way, Tucson, AZ 85721, USA Department of Chemistry and Biochemistry, The University of Arizona, 1306 E. University Blvd, Tucson, AZ 85721, USA

h i g h l i g h t s • Sono-electrochemical method is effective in the recovery of Pd, Ga and Pb. • Recovery efficiency depends on the type of metal ion and megasonic frequency used. • Pd is recovered mainly in metallic form while Ga and Pb show presence of oxide.

a r t i c l e

i n f o

Article history: Received 20 January 2016 Received in revised form 27 June 2016 Accepted 4 July 2016 Available online 5 July 2016 Keywords: Sonochemistry Electrochemistry Megasonic Metal recovery

a b s t r a c t Metal recovery from aqueous waste streams is an important goal for recycling, agriculture and mining industries. The development of more effective methods of recovery have been of increasing interest. The most common methods for metal recovery include precipitation, electrochemical, ion exchange, flocculation/coagulation and filtration. In the current work, a sono-electrochemical technique employing sound field at megasonic frequency (500 kHz or 1 MHz) in conjunction with electrochemistry is evaluated for enhanced recovery of selected metal ions (palladium, lead and gallium) with different redox potentials from their aqueous solutions. The surface morphology and elemental composition of the metal deposits were characterized using scanning electron microscopy, energy dispersive X-ray spectroscopy, and Xray photoelectron spectroscopy. The percent recovery was found to depend on the type of metal ion and the megasonic frequency used. Palladium was recovered in its metal form, while lead and gallium were oxidized during or after the recovery process. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Cost effective and energy efficient recovery of metals from aqueous waste streams is an important goal for recycling, agriculture, and mining industries [1,2]. Metals, such as lead, can contaminate the environment when mining operations are performed or electronic waste is processed [3]. Since rare earth and heavy metals are frequently toxic even at very low concentrations, and their refined forms are industrially valuable, the development of more effective methods of recovery have been of increasing interest [4]. The prevailing methods for metal recovery from aqueous sources include precipitation, electrochemical, ion exchange, flocculation/coagulation and filtration [4,5]. These tra-

∗ Corresponding author. E-mail address: [email protected] (M. Keswani). http://dx.doi.org/10.1016/j.jhazmat.2016.07.007 0304-3894/© 2016 Elsevier B.V. All rights reserved.

ditional methods suffer from drawbacks such as non-selectiveness in precipitation process, susceptibility to suspended solids during flocculation/coagulation, pH sensitivity during ion exchange, and poor removal rate in filtration [5,6]. Metal recovery methods can be categorized into two major groups based on the working mechanisms: physical and chemical processes [4–8]. Physical processes separate liquid from solid phase in a mixture, which can be accomplished in a variety of ways such as filtration and/or coagulation/flocculation [7]. Filtration is a physical remediation method that can remove heavy metal ions from solution using techniques like reverse osmosis (RO), although the filter can get quickly clogged, which leads to a decrease in membrane performance and higher costs due to eventual membrane replacement [5,8]. The other drawback with RO is that it is only effective with aqueous solutions containing low concentrations of contaminate [6]. Coagulation/flocculation in wastewater treatment process destabilize smaller particles by removing repulsive elec-

380

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

trostatic forces between them [9] causing them to agglomerate and settle faster. While useful at removing particles from aqueous solutions, coagulation/flocculation has its drawbacks since it is unable to remove solutes from solutions directly [10]. Compared with physical methods for recovering metals from aqueous solutions, chemical methods operate by causing chemical or electrochemical reactions of metals in solution or by changing the solubility limit of dissolved species [6,8]. Chemical precipitation is the most widely used method for heavy metal remediation, mainly due to its low cost and high removal efficiency [4,5], although it is plagued by the large amount of consumption of reagents to bring pH to the required level and environmental problems associated with its sludge production after the process [8,11]. Ion exchange is another option for metal recovery, which substitutes heavy metal ions with other ions [12]. High cost of installation and maintenance make it a limited usage, it has problems with suspended solids in aqueous solution as well [8]. An alternative chemical method is the electrochemical process, which has good efficiency and can be used with high concentrations of dissolved heavy metals [8]. It is also associated with lower cost compared to conventional chemical precipitation [13–19]. One advantage of this technique is that dissolved metal ions can be directly deposited in their metallic form [16]. Metals that have already been demonstrated to be removed by electrodeposition process include silver, copper, lead, gallium, nickel, zinc and cadmium [16,17,19]. Further the technique offers selectivity, which can greatly simplify the recycling process [20]. If the performance and efficiency of this technique can be further augmented through assistance of a low cost and a green technology, it will likely be of great value to industries interested in metal recovery. In this work, we employ a sono-electrochemical technique utilizing sound field at megasonic frequencies in conjunction with electrochemistry for enhanced recovery of selected metal ions, with widely differing redox potentials, from their aqueous solutions. Electrochemical potentials were applied in the range where hydrogen gas was produced, which was stabilized in the form of oscillating bubbles using the megasonic field. The extreme temperatures and highly reducing environment inside the bubbles were used for the reduction of metal ions to their zero state solid form. The metals chosen for this study were palladium, gallium and lead. Palladium is used for as a catalyst in automobile and other industries, and is critical for conservation of resources [21]. Lead is widely used in the industry, and it is an environmental contaminant above threshold level and can cause irreversible health effects to human beings [5]. Gallium is an important element for manufacturing electronic products, and its sources are limited in nature [22].The paper is organized as follows: in Section 2, we describe the experimental set-up and conditions and electrochemical, analytical and surface characterization techniques used in this study. In Section 3, we have presented and discussed the key results of this work. Finally, the main conclusions of this research are provided in Section 4. 2. Experimental 2.1. Chemicals used in this study Aqueous solutions were prepared using high resistivity deionized (DI) water (18 M-cm), and the chemicals used were of

Table 1 Composition of the examined aqueous solutions. Metal

Electrolyte composition

pH

Palladium (Pd)

Pd(NO3 )2 HNO3 PbCl2 CH3 COONH4 GaCl3 NaCl

∼0.9

Lead (Pb) Gallium (Ga)

3.5 mM 1M 100 mM 0.5 M 10 mM 0.5 M

∼3.8 (adjusted using HAc) ∼2.5 (adjusted using HCl)

Fig. 1. Schematic of the three electrode system used in the study. WE, CE and RE refer to working, counter and reference electrodes.

analytical reagent (AR) grade. Palladium nitrate (Pd(NO3 )2 , 99.9%), lead chloride (PbCl2 , >99%), gallium chloride (GaCl3 , 99.9%), ammonium acetate (CH3 COONH4 , ≥97%), nitric acid (HNO3 , 70%), acetic acid (HAc, ≥99.7%), hydrochloric acid (HCl, 37%) and sodium chloride (NaCl, ≥99%) were purchased from Thermo Fisher Scientific (New Jersey). The chemical compositions and pH values of the solutions used in the study are listed in Table 1. 2.2. Voltammetry study of different metals The electrochemical setup consisted of a glass vessel (∼50 mL liquid volume) sealed with Parafilm with provision for inserting the electrodes. Three electrode system was used for both cyclic voltammetry (CV) and chronoamperometry studies, as illustrated in Fig. 1. During CV measurements, the potential of working electrode was scanned at 20 mV/s in the potential range of −2.0 to 1.0 V (cathodic first) (versus Ag/AgCl, sat. KCl), while in the chronoamperometry study, the potential was maintained at a constant value (versus Ag/AgCl, sat. KCl) by means of a Gamry Reference 3000TM Potentiostat (Warminster, PA). The electric current was measured as a function of voltage in the CV study, and as a function of time during the chronoamperometry study. Sonication experiments were performed by using: (a) Mini-meg® tank (PCT Systems Inc.) of volume ∼4.5 L and consisting of a 125 cm2 transducer affixed at the bottom with an operating frequency of 1 MHz and power density of 8 W/cm2 , and (b) a cubical reactor (12 L), custom made by Weber Ultrasonics Inc. (Clarkston, MI, USA), with 500 kHz transducer (73 cm2 ) on the side wall operating at 8 W/cm2 . In all the experiments performed, the aqueous solutions were saturated with ultrahigh purity argon (Ar, 99.999%) by bubbling

Table 2 Half reactions and details of three electrode electrochemical cell. Half reaction

E0 (V)

Working electrode

Counter electrode

Reference electrode

Pd2+ + 2e− → Pd (s) Pb2+ + 2e− → Pb (s) Ga3+ + 3e− → Ga (s)

+0.987 −0.126 −0.530

Pt mesh Copper sheet Vitreous carbon rod

Pt mesh Pt mesh Pt mesh

Ag/AgCl (sat. KCl) Ag/AgCl (sat. KCl) Ag/AgCl (sat. KCl)

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

the gas for 20 min prior to the electrochemical studies and maintaining an Ar blanket on the surface of the solution for the entire duration of the experiments. Argon was applied in the aqueous solutions for deoxygenating the solution and preventing interference of oxygen with the electrochemical reactions. Standard reduction potential for each metal investigated, and the corresponding working, counter and reference electrodes used in the experiments for these metals are listed in Table 2. The magnitude of the electrical potential needed for recovery of the three metals increased in the following order: Ga > Pb > Pd, which also indicates the order of difficulty in recovering the metals. The concentrations of corresponding metal ions increased in the reverse order so that significant amount of metal can be recovered and quantified in each case under the experimental conditions. 2.3. Sample characterization In order to calculate the recovery efficiency for different metals in the solution, inductively coupled plasma mass spectrometry (ICP-MS) was used for quantitative determination of Pd2+ , Pb2+ and Ga3+ concentrations at time 0, 60 min and 120 min in the experiment. Liquid samples were filtered through 0.45 ␮m filters before appropriate dilution with ultrapure water for adjusting the concentration of the metal ions in the range suitable for ICP-MS measurements by the instrument. The background concentrations of Pd2+ , Pb2+ and Ga3+ in ultrapure water were 3.89, 0.26 and 6.26 ng/L respectively. An FEITM Inspec-S Scanning Electron Microscope (SEM) (Hillsboro, Oregon) equipped with standard imaging back scattered electron detector and an energy dispersive X-ray spectroscopy (EDX) system working at 30 kV was used for the examination of the surface morphology and elemental composition of the deposit on the working electrode after two-hour chronoamperometry study. In order to confirm the oxidation states of the elements deposited on the working electrode, X-ray photoelectron spectroscopy (XPS) method was applied to analyze the elemental composition of the deposit on the working electrode surface. XPS is a very surface sensitive technique that allows elements to be identified and quantified to 0.5–1% atomic composition in the top 1–15 nm of surface. XPS utilizes Al K␣ or Mg K␣ x-ray radiation to cause emission of core electrons from atoms in the surface due to the photoelectric effect. The electrons possess specific kinetic energies which, when analyzed and converted to binding energies, indicate the identity of the element it was emitted from and, in many cases, gives information on specific chemical or oxidation states for that element depending on the binding energy of the electron. X-ray photoelectron spectra were recorded by a Kratos Axis Ultra 165 DLD Hybrid using a delay line area detector. All elemental regions were collected using monochromatized Al K␣ radiation with the x-ray anode operated at 300 W (20 mA, 15 KeV). The maximum area was sampled using hybrid lens mode (a combination of electrostatic and magnetic immersion lenses) and the rectangular aperture was ∼300 ␮m × 700 ␮m. Survey spectra were collected using a pass energy of 160 eV. Elemental regions were collected using a pass energy of 20 eV. The FWHM of the Ag 3d peak for a clean silver foil under these conditions was 0.68 eV. The base pressure for the analysis was ∼5 × 10−9 torr. Pd, Pb and Ga deposits on various substrates were mounted on a standard Kratos sample puck. Samples exhibiting surface charging were neutralized by low energy electrons emitted from the in-situ charge neutralizer. All binding energies were corrected to the carbon 1s (adventitious carbon) peak referenced at 284.6 eV Binding Energy. Quantitative analysis was accomplished using the Kratos Vision Processing software by removing the background using either a standard linear or integral (Shirley) background sub-

381

traction routine to determine the area of each region. Any percent atomic composition information for these samples were calculated by the Kratos RSF (Relative Sensitivity Factors). Peak fitting was done by either asymmetric or Gaussion/Lorenztian peak model, included in the software, depending on the element or previous literature models used. Surface information for XPS is limited in most cases to a sampling depth of less than 20 nm because of the limited electron inelastic mean free path (IMFP). 2.4. Statistical method Student’s t-test statistical method was used for comparison between two sets of data and to determine if they were significantly different from each other. The difference was considered to be significant at 0.05 level (p < 0.05) in all cases. Statistical Package for the Social Sciences software (IBM SPSS, Version 20) was employed for the data analytical work in this study. 3. Results and discussion 3.1. Cyclic voltammetric study In the first step, the voltammetric technique was used to gain information about the general behaviors of the electrodeposition process of these three metals and to identify the range of potential where water/hydrogen ion reduction occurs to form hydrogen gas. Based on these measurements, a suitable potential condition was identified and used later in sono-electrochemical experiments to determine its effect on removal of metal ions from the soluble metal solutions. Electrodeposition of metals from different media was then studied at constant applied potential both in the presence and absence of sonication. Based on the results of our cyclic voltammetric study, working electrode potentials of −0.1 V, −0.5 V and −1.4 V (versus Ag/AgCl sat. KCl) for Pd, Pb and Ga, respectively, were selected for the recovery of metal species from different solutions. These potentials are in reasonable agreement with those used in previous studies [17,19,23]. From the standard electrode potentials of +0.915 V, −0.126 V and −0.53 V for Pd2+ , Pb2+ and Ga3+ respectively, it could be assumed that the sequence of difficulty level for reduction/recovery of these metals would be Pd < Pb < Ga, if only electrochemical method was applied to the aqueous solution without the assistance of acoustic field. 3.2. Comparison of metal ions reduction from aqueous solutions using the electrochemical and sono-electrochemical methods The concentrations of metal ions (Pd2+ , Pb2+ and Ga3+ ) from working solutions under different scenarios were analyzed by ICPMS. The results are expressed as percent reduction ratios (R) of these metal ions, which are defined as: R(%) = [(C0 − Ct )/C0 ] × 100 in the above equation, C0 and Ct are the concentrations of metal ions in solutions at the initial stage (t = 0) and at two different sampling times (t = 60 and 120 min), respectively. The changes and comparisons of reduction ratios under different conditions for these three metals are illustrated in Fig. 2. The results summarized in Fig. 2 suggest that at the end of 120 min of sono-electrochemical experiments, the concentrations of all three metal ions were reduced significantly, but different levels of reductions were observed depending on the type of metal. For the first 60 min, Fig. 2 shows that among the three metals, Pd2+ is an easy ion to be reduced, with or without the assistance of acoustic field. In the electrochemical reaction system alone, at

382

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

120

Reduction Ratio (%)

100

(a)

electrochemistry alone 500 kHz + electrochemistry 1 MHz + electrochemistry

80 60 40

n/a

20 0 Pd

Pb

Ga

120 electrochemistry alone

(b) Reduction Ratio (%)

100

500 kHz + electrochemistry 1 MHz + electrochemistry

80 60 40

n/a

20 0 Pd

Pb

Ga

Fig. 2. Comparison of (a) 60 min and (b) 120 min metal ion (Pd2+ , Pb2+ and Ga3+ ) reduction ratios in argon gas saturated solution (with or without acoustic field of 500 kHz or 1 MHz sonication).

t = 60 min, R was about 74% for Pd2+ , and with the assistance of the 1 MHz acoustic field, the reduction ratio increased to almost 100%. The color of the solution changed from dark orange to light yellow, accompanied by the dark color metal deposit appearing on the working electrode Pt mesh, which was silvery white in color originally. Statistical t-test results suggested that 500 kHz and 1 MHz sound fields can both improve the electrochemical removal efficiency of palladium ions in the solution significantly (p < 0.05), and reduction ratio of [Pd2+ ] in the 1 MHz acoustic field was significantly larger than that with 500 kHz field in the first hour experiment. At the end of the second hour in the palladium recovery experiments, R values were higher compared to those at 60 min, but were not significantly different from each other (R120min were about 100% in all cases). The results indicate that a complete reduction of Pd2+ ions in the solution could be achieved in the first hour with the application of potential in the presence of 1 MHz acoustic field, which was comparable with the removal ratios of two hours under no sonication (electrochemistry alone) and 500 kHz sonication with applied potential conditions. When comparing our results with those achieved with pure electrochemical method for full recovery of Pd as in the study by Jayakumar et al. [23], much less time is needed using our new sono-electrochemical method (1 h versus 8 h). For the removal/recovery of lead from aqueous acetate and chloride solution, the acoustic field played an important role as well. It appeared that among the three metals, lead was the most difficult one to remove/recover from the solution by using electrochemical method only, based on the lower percentage of the reduction ratio compared to those for Pd and Ga, both at 60 min and 120 min (Fig. 2). Application of an acoustic field at either 500 kHz or 1 MHz significantly increased the electrochemical reduction of Pb2+ in

the solution. It is apparent that the 1 MHz sonication doubled the reduction ratio of lead ions in the solution at 60 min and 120 min, suggesting that the highly reducing environment was formed near the surface of working electrode using high frequency megasonic field which caused more Pb2+ in the solution to be reduced [24]. Another reason is that sonication can play an important role in the mixing of solutions and reducing mass transfer limitations, by transporting more lead ions towards the working electrode. In the case of both Pd and Pb, use of 1 MHz frequency showed much better recovery compared to 500 kHz when used in combination with applied electrical potential. Therefore, in the removal/recovery of gallium from the aqueous solution, only electrochemical and 1 MHz acoustic field assisted sonoelectrochemical methods were used. The reduction ratio of gallium ion in the first hour was only about 35% using electrochemical method, and it was significantly improved to about 60% by applying the 1 MHz acoustic field. During the second hour of the experiment, the reduction ratios increased under both conditions, reaching 70 and 75%, for electrochemical and sono-electrochemical methods respectively. The fact that the reduction ratios for the three elements using electrochemical and sono-electrochemical methods increased in the following order Pd > Ga > Pb even though the initial concentration of the corresponding ions in the original solutions followed the reverse order (Pb > Ga > Pd) indicates that Pb is more difficult to recover than Ga, and Pd is the easiest to recover among the group. From the above results, it could be seen that the sonoelectrochemical method was found to be effective and efficient for removal and/or recovery of soluble metal/metalloid species from aqueous solutions. The method utilizes the principles of both electrochemistry and acoustics for the formation of zero valent metal particles from their soluble species. The use of electrochemistry is to generate hydrogen gas (from reduction of water or hydrogen ion) by applying a suitable potential to a substrate such as platinum, vitreous carbon or copper immersed in aqueous solution containing soluble metal/metalloid species. The half reaction on the working (cathodic) electrode can be expressed as: 2H+ (aq.) + 2e− ↔ H2 (g) or 2H2 O (l) + 2e− ↔ H2 (g) + 2OH− (aq.) The hydrogen gas generated is stabilized in the form of oscillating bubbles by acoustic and electrochemical parameters [24–26]. The soluble metal/metalloid ions reduce at the interface of the hydrogen filled bubbles under highly reducing environment (due to formation of hydrogen radicals) under extreme temperatures. High power density is essential for generating significant amplitude of oscillating stable hydrogen bubbles so that sufficient number of hydrogen radicals are created inside the bubbles through thermal decomposition of hydrogen gas. This results in a highly reducing environment needed for the reduction of metal ions. However, increase in the power density above a threshold may transform the stable bubbles to transient bubbles, which may negatively affect the recovery efficiency. Similarly, we expect that there will be an optimum in acoustic frequency for achieving maximum recovery. At megasonic frequencies, stable cavitation is dominant but the oscillation amplitude of bubbles decreases with increasing megasonic frequency due to smaller size of bubbles at higher frequency. Lower oscillation amplitude will reduce the efficiency of recovery. On the other hand, at ultrasonic frequencies, transient cavitation is prevalent and the effectiveness of the recovery will be lower. From the above discussion, it could be concluded that under certain conditions, metal reduction/recovery from their aqueous solutions is possible by electrochemical method alone. By placing the reaction in the acoustic field, the process could be accelerated, depending on the drive frequency of the transducer. The identi-

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

383

fication of metals and their oxidation states are discussed in the following sections. 3.3. Characterization of metal deposits using SEM and EDX measurements Surface morphology and elemental composition of the metal deposits after 2 h of sono-electrochemical studies at 1 MHz were examined using SEM with EDX techniques. For each sample, three representative spots on the surface of the working electrode covered with metal/metalloid were scanned under SEM. EDX was used for those specific spots to get the elemental composition information on the electrode surface. On platinum mesh potentially covered with palladium, the deposit, as shown in Fig. 3a, appears to consist of spherical particles all over the surface with a dendrite growth observed in the center of the SEM image. These morphological characteristics are consistent with observations in other studies where only electrochemical method was applied for the recovery of Pd [23]. As for the EDX results shown in Fig. 3b, the dominant element in the spectrum is Pd, which is ∼80% in atomic percentage. There is no evidence of oxygen appearance on the surface of the metal, suggesting that Pd detected on the surface of Pt is in its pure metal form, and not oxidized. This is representative of the chemical characteristics of Pd, which being a noble metal does not easily oxidize. In the sono-electrochemical process of lead deposition on a copper electrode, the metal appeared gradually on the surface of the copper. The dendrites of metal (lead) formed on the border of the electrodes and grew with additional sonication time, while a lead film was observed with various geometric forms, such as plates and globules, which totally covered the copper substrate in the solution (Fig. 3c). This is consistent with the results of a study by Agulló et al. [27], where an ultrasonic frequency of 30 kHz at 150 W was used for electrodeposition of lead on copper using a fluorobath. A metallic gray colored metal appeared on the copper surface when the electrode was still immersed in the solution, although the layer turned into dark metal deposit when it was exposed to air after the experiment, because the deposit was oxidized easily due to the properties of metallic lead, which is that it forms a protective oxide layer. In the following EDX characterization of the covered surface areas (Fig. 3d), the oxygen peak was clearly observed, suggesting that the top layer of lead was oxidized. It is possible that under the oxidized layer of the metal deposit, there was still metallic lead. Thus, XPS was also employed in our study to understand the oxidation state of the metals and to obtain the information about the elemental composition of the oxidized layer on the top of the surface. Fig. 3e is the SEM image of Ga deposited on vitreous carbon electrode. In the center there is a large flake of metal/metalloid covering the carbon substrate. During the 2 h sono-electrochemical reduction, it was observed that some white flakes forming on the working electrode were intermittently being dislodged from the electrode surface and falling down in the solution (picture not shown here). This may be due to strong agitation from sonication that could cause some of the deposits to fall off the working electrode if they were not adhered tightly onto the vitreous carbon rod. There were oxygen peaks observed in the EDX spectrum of the Ga deposits (Fig. 3f), which indicates that Ga may have been oxidized during exposure to air just prior to EDX analysis or during the sono-electrochemical reduction treatment. 3.4. XPS results The survey spectrum of the palladium deposited on a platinum mesh shows mostly Pd with some C, O and a small amount of Pb

Fig. 3. SEM images and EDX spectra of metals formed on working electrodes. (a) SEM image and (b) EDX spectrum of palladium formed on platinum mesh from aqueous solution. (Magnification 20,000× for image). (c) SEM image and (d) EDX spectrum of lead formed on copper sheet from aqueous solution. (Magnification 10,000×for image). (e) SEM image and (f) EDX spectrum of gallium formed on vitreous carbon rod from aqueous solution. (Magnification 10,000×for image).

Pd 3pPd 3p O 1s

Pd 3s

O (Auger)

(a)

1200

1000

800

600

400

Pd 4s Pd 4p

Pb 4f

C 1s

Pb 4d

Intensity,cps

Pd 3d

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

Pd (Auger)

384

200

0

Binding Energy, eV

Intensity(cps)

Pd 3d 5/2

Background(cps) Fit Envelope Pd 3d metal fit

(b)

Pd 3d 3/2

Intensity, cps

Pd 3d oxide fit

346

344

342

340

338

336

334

332

Binding Energy, eV Fig. 4. Palladium XPS results. (a) Survey of Pd on Pt mesh. Survey spectrum of Pd deposited on Pt mesh showing predominantly Pd with some C (285 eV) and Pb (140 eV) impurities. (b)Spectra fit of the Pd 3d region using Shirley (integral) background, asymmetric metal lineshapes and a symmetric Gaussian-Lorentzian model for the oxide peak. The fit clearly shows only native oxide.

contamination. The O 1s peak is strongly overlapped by the Pd 3p 3/2 peak and cannot be used for quantification on this sample. The Pd 3d doublet was fit to a combination of an asymmetric metallic peak and a symmetric Gaussian-Lorentzian oxide peak. The metallic peak appears at 335.5 eV and the oxide peak shows up at 336.5 eV. References put the metallic peak at 335.0 eV and the native oxide at 336.7 eV [28]. This film exhibits a mostly metallic surface (Fig. 4). The survey spectrum of lead deposited on a copper electrode shows there are lead, oxygen and carbon, no other elements were detected. All of the Cu substrate was coated in the area of analysis. It was expected that this sample should have an oxide layer due to exposure to ambient atmosphere. Spectrum of Pb 4f was fitted to a set of two states with comparable areas. The fit of Pb 4f spectrum shows two contributions, one at 137.5 eV and the other at 138.5 eV. References put the Pb metal at 136.9 eV and the oxides PbO2 and Pb3 O4 at 137.8 eV and 138.4 eV, respectively. Pb native oxide also appears at 138.4 eV [29]. Because

of the peak widths and known profiles, there are a minimum of two contributors, and it is probably a mixed oxide. An abraded and sputtered Pb standard shows the metallic Pb 4f 5/2 at a binding energy of 136.3 eV on this spectrometer. As shown in the figure, O to Pb ratio for this sample is 1.35, and it is corresponded to a Pb3 O4 composition, at least in the top 20 nm (Fig. 5). The survey spectrum of gallium on glassy carbon shows that the surface is composed of gallium oxide. The C1s signal indicates that a very small amount of carbon is present and that the substrate is not exposed. The Ga sample, deposited on the glassy carbon electrode, showed oxygen and a small amount of carbon. The fit of the Ga 2p shows two different forms of Ga at binding energies of 1118.7 eV and 1120.9 eV. Several references report a native gallium oxide at 1118.75 eV and Ga2 O3 at 1118.25 eV [30]. Metallic Ga is reported at 1116.7 eV. For this sample, Ga within the surface sampling depth appears to be oxidized in the Ga 2p spectrum (Fig. 6).

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

Fig. 5. Lead XPS results. (a) Survey of Pb on Cu. The survey shows complete coverage of Pb on top of the oxygen free, high conductivity, (OFHC) copper electrode. The surface has a native oxide and a small concentration of carbon impurities. (b) Pb 4f regions for Pb deposited on an oxygen free, high conductivity, (OFHC) copper electrode. The region was fit using a Shirley background and each peak was fit to 2, primarily, gaussian peaks. The oxide could be a mix of oxides under such a broad envelope. (c)Standard Pb foil, Argon ion sputtered, showing the metallic Pb and some remaining native oxide.

Spectrum of Ga 3d appears to have three contributors compared with that of Ga 2p. The low binding energy component was modeled as a metallic peak with a 3d doublet at 19.8 eV with a peak splitting of 20.5 eV and the other two contributors were modeled as individual single peaks at 20.4 eV and 21.4 eV, respectively. There is a contribution of the O 2s at 22.8 eV. The presence of the metallic form of Ga can be due to the higher inelastic mean free path for

385

Fig. 6. Gallium XPS results. (a) Survey of Ga on Glassy Carbon. The survey shows predominantly Ga with some Oxygen and Carbon. Carbon is most likely from exposed areas of the glassy carbon electrode or from ambient exposure from atmosphere. (b)Ga 2p 3/2 spectral region, fit using a Shirley background and symmetric GaussianLorentzian lineshapes. This clearly shows the presence of only the native oxide. (c)Ga 3d spectral region, fit using a Shirley background and symmetric GaussianLorentzian lineshapes. The two low BE peaks are the metallic 3d5/2 and 3d3/2 contributions for the Ga metal using peak separation of 0.46 eV. For all gallium oxides in this region, the 3d5/2 and 3d3/2 contributions are unresolvable. The largest peak represents the native oxide and the highest BE peak is the O 2s.

the photoelectrons due to higher kinetic energy. The composition shows approximately a 1:3 ratio of Ga to O on the surface (Table 3). It is unclear why it is so high. Comparing the Ga 2p and Ga 3d peaks show a contribution to the 3d at low binding energy that is due to the presence of metallic Ga,

386

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387

Table 3 Quantitative results. Element

Region

Atomic%

Carbon Gallium Oxygen

C 1s Ga 2p O 1s

11.16 23.10 63.52

while the Ga 2p doesn’t show any metallic Ga. The Ga 2p peaks that are fitted in the figure are in the range of 1116 eV binding energy. This is a relatively low kinetic energy of ∼370 eV compared to the Ga 3d electrons at 20–22 eV binding energy or ∼1466 eV kinetic energy. Because of this difference in kinetic energies, the Ga 3d peak has a longer inelastic mean free path or escape depth. If the oxide layer is on top with metallic Ga underneath, the fit should show some metallic Ga. Compared to the Ga 2p region, which has a lower kinetic energy and lower escape depth for the electrons, it will show no metallic Ga since the only electrons coming out of the surface with no loss in energy are from the layer of oxidized Ga. The XPS results for Pd, Pb and Ga show a good correlation with those obtained from SEM/EDX analysis. For XPS, the information is in more details, since the composition of the top layer of the samples could be derived from the matching spectra of the elemental species. The oxidation state of each metal was derived from the results as well. 3.5. Comparison of sono-electrochemical method with traditional metal recovery methods Traditional chemical precipitation is widely used for metal removal from water/wastewater treatment process [31]. After pH is adjusted to basic conditions (which is the pH of minimum solubility of metal compound), the dissolved metal ions are gradually converted to the insoluble solid phase by chemical reactions with a precipitant agent such as lime. Usually, the metal precipitated from the solution is in the form of hydroxide. The mechanism of metal being removed by chemical precipitation is presented in the following equation: Mn+ + nOH− ↔ M(OH)n ↓ where Mn+ and OH− represent the dissolved metal ions and the precipitant agent, respectively, while M(OH)n is the insoluble metal hydroxide formed as the final product. In the above precipitation process, different kind of metal ions will be converted to their metal hydroxide forms and will be settled to the bottom of the solution. It is hard to separate the metals from one another in the mixture, and there will be a following dewatering process in order to dry these compounds before they were converted to their metal forms. So the process will not be cost effective for the recovery of the metals investigated in the present study. For sono-electrochemical metal recovery process, the major advantage would be selective deposition of pure metal/metalloid on the working electrode when suitable potential is applied using a potentiostat. Sonication can improve the efficiency of metal recovery, as our research results have indicated. The final product could be much more easily collected in comparison with the precipitation process, and the metals will be ready to be reused depending on their purity. The sono-electrochemical metal recovery process employed in our work utilizes acoustic and electrical energy. To calculate the power (P) at 1 MHz, the power density of 8 W/cm2 is multiplied by the area of the transducer (exposed to the solution). This gives, P = 8 W/cm2 × 125 cm2 = 1 kW

The power is then converted to energy (E) by multiplying with time (t = 1 h). E = P × t = 1 kW × 1 h = 1 kWh (corresponding to energy per unit volume of 1 kWh/0.2 L or 5 kWh/L) For electrical energy applied to the solution during the reaction, the power is calculated based on the current (0.01 A) times the potential (V) between the WE and the CE (assuming a maximum of 10 V). P = V × I = 10 V × 0.01A = 0.1 W The power is then converted to energy by multiplying with time (1 h). E = P × t = 0.1 W × 1 h = 0.1 × 10−3 kWh (corresponding to energy per unit volume of 0.1 × 10−3 kWh/0.2 L or 5 × 10−4 kWh/L) Compared with the megasonic energy, the electric energy is much smaller and can be neglected for the sono-electrochemical process. As of December 2015, the industrial cost of electricity is $0.0642/kWh, so the cost of acoustic energy (operated for 1 h) is: Cost = $0.0642/kWh × 1kWh = $0.0642 Based on these calculations, $7.12, $0.15 and $6.11 will be required for the recovery of one gram of Pd, Pb and Ga in our lab scale experiments. Comparing the cost of the metal recovery processes, the traditional metal precipitation process would include: the cost of precipitant agent, dewatering and separation of these metals from each other, as well as the cost for converting the metals to their zero valent forms. In the case of the sono-electrochemical method, the electric cost for both sonication and electrochemical process would be the major cost. 4. Conclusions Palladium, lead and gallium were effectively recovered using the sono-electrochemical method by applying combination of suitable sonication frequency and electric potential to the working electrode immersed in the experimental solution. SEM/EDX and XPS analyses indicate that Pd can be recovered mainly in its metal form, while Ga and Pb were oxidized during or after the recovery process. The reduction of metal ions likely occurs on the hydrogen bubble/liquid interface, where extreme temperature and reducing environment are generated during bubble oscillations. Although, the metal recovery experiments were conducted with one metal at a time, we expect that by using suitable range of electrical potentials (identified using cyclic voltammetry measurements), we can selectively recover metals from the mixture of metal ions in aqueous solution. When comparing this metal recovery technique with traditional water/wastewater metal recovery methods, it appears to be more economical and effective since metals or metal oxides could be derived directly from the process without any further treatment. Acknowledgements The authors acknowledge Tech Launch Arizona and Water, Environmental, and Energy Solutions (WEES) for their partial financial support of this work. We would also like to thank Dr. Reyes Sierra for early discussions in the project. References [1] J. Cui, L. Zhang, Metallurgical recovery of metals from electronic waste: a review, J. Hazard. Mater. 158 (2008) 228–256.

B. Dong et al. / Journal of Hazardous Materials 318 (2016) 379–387 ˜ V.F. [2] A. Fuentes, M. Lloréns, J. Saez, M.I. Aguilar, A.B. Pérez-Marín, J.F. Ortuno, Meseguer, Ecotoxicity, phytotoxicity and extractability of heavy metals from different stabilised sewage sludges, Environ. Pollut. 143 (2006) 355–360. [3] J. Duruibe, M. Ogwuegbu, J. Egwurugwu, Heavy metal pollution and human biotoxic effects, Int. J. Phys. Sci. 2 (2007) 112–118. [4] F. Fu, Q. Wang, Removal of heavy metal ions from wastewaters: a review, J. Environ. Manage. 92 (2011) 407–418. [5] M.A. Barakat, New trends in removing heavy metals from industrial wastewater, Arabian J. Chem. 4 (2011) 361–377. [6] H. Eccles, Treatment of metal-contaminated wastes: why select a biological process? Trends Biotechnol. 17 (1999) 462–465. [7] E. Blondeel, V. Depuydt, J. Cornelis, M. Chys, A. Verliefde, S.W.H.V. Hulle, Physical-chemical treatment of rainwater runoff in recovery and recycling companies: Pilot-scale optimization, J. Environ. Sci. Health Part A 50 (2015) 1083–1098. [8] T.A. Kurniawan, G.Y. Chan, W.-H. Lo, S. Babel, Physico–chemical treatment techniques for wastewater laden with heavy metals, Chem. Eng. J. 118 (2006) 83–98. [9] N.K. Shammas, Coagulation and flocculation, in: Physicochemical Treatment Processes, Springer, 2005, pp. 103–139. [10] Q. Chang, G. Wang, Study on the macromolecular coagulant PEX which traps heavy metals, Chem. Eng. Sci. 62 (2007) 4636–4643. [11] P. Bose, M.A. Bose, S. Kumar, Critical evaluation of treatment strategies involving adsorption and chelation for wastewater containing copper, zinc and cyanide, Adv. Environ. Res. 7 (2002) 179–195. [12] A. Da¸browski, Z. Hubicki, P. Podko´scielny, E. Robens, Selective removal of the heavy metal ions from waters and industrial wastewaters by ion-exchange method, Chemosphere 56 (2004) 91–106. [13] N. Kongsricharoern, C. Polprasert, Electrochemical precipitation of chromium (Cr 6+) from an electroplating wastewater, Water Sci. Technol. 31 (1995) 109–117. [14] K. Jüttner, U. Galla, H. Schmieder, Electrochemical approaches to environmental problems in the process industry, Electrochim. Acta 45 (2000) 2575–2594. [15] G. Chen, Electrochemical technologies in wastewater treatment, Sep. Purif. Technol. 38 (2004) 11–41. [16] C. Ascensão, L. Ciríaco, M. Pacheco, A. Lopes, Metal recovery from aqueous solutions, Portugaliae Electrochim. Acta 29 (2011) 349–359.

387

[17] E. Ghali, M. Girgis, Electrodeposition of lead from aqueous acetate and chloride solutions, Metall. Trans. B 16 (1985) 489–496. [18] N. Kaminari, D. Schultz, M. Ponte, H. Ponte, C. Marino, A. Neto, Heavy metals recovery from industrial wastewater using Taguchi method, Chem. Eng. J. 126 (2007) 139–146. [19] D.O. Flamini, S.B. Saidman, J.B. Bessone, Electrodeposition of gallium onto vitreous carbon, JApEl 37 (2007) 467–471. [20] T.E. Lister, P. Wang, A. Anderko, Recovery of critical and value metals from mobile electronics enabled by electrochemical processing, Hydrometallurgy 149 (2014) 228–237. [21] H. Umeda, A. Sasaki, K. Takahashi, K. Haga, Y. Takasaki, A. Shibayama, Recovery and concentration of precious metals from strong acidic wastewater, Materi. Trans. 52 (2011) 1462–1470. [22] G. Phipps, C. Mikolajczak, T. Guckes, Indium and gallium: long-term supply, Renew. Energy Focus 9 (2008) 56–59. [23] M. Jayakumar, K.A. Venkatesan, T.G. Srinivasan, P.R.V. Rao, Studies on the feasibility of electrochemical recovery of palladium from high-level liquid waste, Electrochim. Acta 54 (2009) 1083–1088. [24] M. Keswani, S. Raghavan, P. Deymier, Electrochemical investigations of stable cavitation from bubbles generated during reduction of water, Ultrason. Sonochem. 21 (2014) 1893–1899. [25] M. Keswani, S. Raghavan, P. Deymier, S. Verhaverbeke, Megasonic cleaning of wafers in electrolyte solutions: possible role of electro-acoustic and cavitation effects, MiEng 86 (2009) 132–139. [26] R. Balachandran, P. Deymier, S. Raghavan, M. Keswani, A sono-electrochemical technique for enhanced particle removal from tantalum surfaces, ECS Solid State Lett. 3 (2014) P49–P52. [27] E. Agulló, E. Expósito, V. Montiel, A. Aldaz, Influence of an ultrasonic field on lead electrodeposition on copper using a fluoroboric bath, NJCh 23 (1999) 95–101. [28] K. Noack, H. Zbinden, R. Schlögl, Identification of the state of palladium in various hydrogenation catalysts by XPS, CatL 4 (1990) 145–155. [29] D.L. Cocke, M.Y.A. Mollah, J.R. Parga, T.R. Hess, J.D. Ortego, An XPS and SEM/EDS characterization of leaching effects on lead- and zinc-doped portland cement, J. Hazard. Mater. 30 (1992) 83–95. [30] A.I. Serykh, M.D. Amiridis, In-situ X-ray photoelectron spectroscopy study of supported gallium oxide, SurSc 604 (2010) 1002–1005. [31] Metcalf & Eddy, Wastewater Engineering: Treatment and Reuse, 4th ed., McGraw-Hill Science/Engineering/Math, 2003.