Regeneration of Ce(IV) in simulated spent Cr-etching solutions using an undivided cell

Journal of Hazardous Materials 171 (2009) 755–760

Contents lists available at ScienceDirect

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

Regeneration of Ce(IV) in simulated spent Cr-etching solutions using an undivided cell Kuo-Lin Huang ∗ , Te-San Chen, Kuei-Jyum C. Yeh Department of Environmental Engineering and Science, National Pingtung University of Science and Technology, Pingtung 91201, Taiwan

a r t i c l e

i n f o

Article history: Received 22 December 2008 Received in revised form 21 April 2009 Accepted 13 June 2009 Available online 21 June 2009 Keywords: Spent Cr-etching solution Ce(IV) regeneration Anion effect Rate constant Ce(IV) yield

a b s t r a c t This study investigated regeneration of Ce(IV) from Ce(III) oxidation at 0.05–0.5 A cm−2 in 4 M HNO3 with/without anion impurities (SO4 2− (0.01–0.2 M), Cl− (0.01–0.08 M), and/or Cr2 O7 2− (0.005–0.016 M)) in an undivided cell. Both Ce(IV) yield and current efficiency (CE) were significantly lower in 0.1 M than in 1–4 M HNO3 and different on anode materials (in order Pt > IrO2 /Ti > glassy carbon). The apparent rate constants for Ce(III) oxidation on the Pt anode (k1 ) and for Ce(IV) reduction on a stainless steel cathode (k2 ) were (0.40–1.80) × 10−4 and (0.08–1.01) × 10−4 s−1 , respectively, corresponding to the apparent mass transfer coefficients of (2.0–9.0) × 10−3 and (0.4–5.1) × 10−3 cm s−1 , respectively. For Ce(III) oxidation at 0.3 A cm−2 in 4 M HNO3 containing multi-impurity (0.025 M SO4 2− + 0.08 M Cl− + 0.016 M Cr2 O7 2− ), the k1 was lowered by one order of magnitude although the k2 remained unchanged, and both Ce(IV) yield and CE were reduced by ∼89%. The decrease of Ce(IV) yield and CE by the uni-impurity was in order Cl− > Cr2 O7 2− > SO4 2− . The obtained parameters are useful to design undivided batch reactors for the Ce(IV) electro-regeneration in spent Cr-etching solutions. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Because of the expanded development of thin film transistor liquid crystal display (TFT-LCD) industry, the types and amounts of wastes generated from the TFT-LCD industry rapidly increase. In TFT-LCD manufacturing processes, Cr-etching is a common procedure and it is usually performed using cerium ammonium nitrate ((NH4 )2 Ce(NO3 )6 ) (CAN) dissolved in concentrated nitric acid (based on the redox reaction of 6Ce(IV) + Cr(0) → 6Ce(III) + Cr(VI)). After use, the spent Cr-etching solutions are regarded as hazardous wastes containing high concentrations of Cr(VI), protons, Ce(III), and anion impurities (e.g., SO4 2− , Cl− , and Cr2 O7 2− , depending on operating conditions). Hence, it is required to well handle spent Cretching solutions before discharging. Traditionally, spent Cr-etching solutions are treated with chemical coagulation followed by precipitation that produces considerable amounts of hazardous sludge with high leaching of total-Cr (Cr(III) + Cr(VI)) in Toxicity Characteristic Leaching Procedure (TCLP). Instead, this production of Cr- and Ce-abundant sludge can be avoided, if spent Cr-etching solutions can be regenerated and recycled. The regenerated Ce(IV)-abundant solutions may be recycled for the mentioned Cr-etching purpose, and also have a good potential to be applied for some electrochemical/environmental processes. For

∗ Corresponding author. Tel.: +886 8 770 3202x7092; fax: +886 8 774 0256. E-mail address: [email protected] (K.-L. Huang). 0304-3894/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jhazmat.2009.06.061

example, Ce(IV) is a common mediator (oxidant) for electrosynthesis, oxidation, or destruction of organics and it can be regenerated from Ce(III) electro-oxidation in mediated electrochemical oxidation (MEO) processes [1–10]. Electrochemical treatments have various advantages such as wide application, simple equipment, easy operation, low temperature requirement, and no sludge formation [11,12]. Electrolysis may be cathodic for metal deposition recovery [13] or anodic for regeneration of spent oxidants or destruction of toxic species [14]. Therefore, it is attractive to electrochemically regenerate Ce(IV) from Ce(III) oxidation in spent Cr-etching solutions. However, little attention has been paid to the electrolytic regeneration of Ce(IV) from the spent Cr-etching solutions containing anion impurities. The electrochemical behaviors of the Ce(III)/Ce(IV) couple on planar and rotating glassy carbon electrodes in nitric acid solutions with/without anion impurities were investigated using cyclic voltammetry (CV) and linear sweep voltammetry (LSV) in our previous works [15,16]. In this study, therefore, we further explored the regeneration of Ce(IV) from Ce(III) electro-oxidation using different electrodes from simulated spent Cr-etching solutions. An undivided cell was used in this study because it is simple in design and easy to scale up further [2]. It also benefits from low internal cell resistances. Some operational variables such as current density, electrode area/electrolyte volume ratio, anode materials, nitric acid concentration, initial Ce(III) concentration, and impurity were tested to evaluate their effects on the Ce(III) oxidation in terms of current efficiency and Ce(IV) yield.

756

K.-L. Huang et al. / Journal of Hazardous Materials 171 (2009) 755–760

These parameters are useful for the design of the electrolytic Ce(IV) regeneration system. 2. Experimental 0.2 M cerium solutions were prepared by dissolving equivalent amount of solid Ce(NO3 )3 ·6H2 O (analytical reagent grade, Research Chemicals Ltd.) in nitric acid solutions of desired molarities. Along with the Ce(NO3 )3 ·6H2 O chemical, highly pure NaCl (Showa), K2 Cr2 O7 (Merck), and Na2 SO4 (Merck) were added as the sources of Cl− , Cr2 O7 2− , and SO4 2− , respectively, into 4.0 M HNO3 to prepare Ce(III) solutions with the anion impurities. Both types of solutions (with/without anion impurities) were used to investigate the effects of current density, anode material, nitric acid concentration, ratio of anode area to electrolyte volume, and anion impurity on the electro-oxidation of Ce(III) to Ce(IV). The anode material was platinum (1 or 2.25 cm2 ), glassy carbon (1 cm2 ), or DSA (dimensional stable anode, IrO2 on Ti, 1 cm2 ) whereas a stainless steel plate was used as the cathode (1 cm2 ) due to its commercial availability, low cost, high mechanical property, and good long-term behavior [17]. (Note that stainless steel cathodes have been coupled with some anodes used for Ce(IV) regeneration in methanesulfonic acid [1,2].) The Ce(III) electro-oxidation of each prepared solution (volume = 50 mL) was performed in a single batch chamber using constant-current operation controlled by a DC power supply (Good Will Instrument CO., LTD GPS-2303). The cell voltage and current were monitored with time. All the experiments were performed at room temperature (25 ± 2 ◦ C). A redox potential titrator (Metrohm 702 SM Titrino) was used to quantitatively determine the Ce(IV) concentrations in samples using ferrous ammonium sulfate dissolved in 1 M nitric acid [18]. The concentrations of total cerium were analyzed by inductively couple plasma atomic emission spectrometry (ICP-AES) (Perkin Elmer optima 2100 DV). According to Faraday’s laws of electrolysis, the theoretical production of Ce(IV) amount (mt ) at electrolysis time t can be calculated as below [2] mt =

MIt zF

(1)

where M: molecular weight of Ce, I: current, z: electrons transferred per Ce(III) ion (z = 1 in Ce(III) → Ce(IV) + e− ), and F: the Faraday’s constant. The current efficiency is calculated based on the following equation Current efficiency (CE) =

me × 100% mt

(2)

where me is the Ce(IV) amount obtained experimentally. The Ce(IV) percent yield is calculated using Yield (%) =

 Ce(IV)  a

Ce(IV)t

Fig. 1. Variations of (a) current efficiency and (b) Ce(IV) yield with time (with fitting curves) for the Ce(III) oxidation in 4 M HNO3 ([Ce(III)]initial = 0.2 M) with/without multiple anion impurities (SO4 2− , Cl− , and Cr2 O7 2− = 0.025, 0.08, and 0.016 M, respectively) (Pt anode: 1 cm2 ).

to oxidize (0.2 M) Ce(III) to Ce(IV) in 4 M HNO3 on a 1 cm2 (projected geometric area) Pt anode. As expected, the current efficiency (CE) decreased with increasing electrolysis time regardless of the difference in current density (Fig. 1a); additionally, increasing current density increased the cell voltage (2.1–2.8 V) and energy consumption (EC) (0.6–3.6 Wh g−1 ) but decreased the current efficiency (Table 1). For 120 min electrolysis, the CE was 64% at 0.05 A cm−2 , about fourfold that at 0.5 A cm−2 . This phenomenon is associated with the main electrochemical reactions (at the Pt anode in 4 M HNO3 ) as follows. Ce3+ → Ce4+ + e− +

× 100%

(3)

where Ce(IV)a is the actual amount of Ce(IV) generated at the end of electrolysis and Ce(IV)t is the theoretical amount of Ce(IV) that can be generated (Ce(IV)t = Electrolyte volume × [Ce(III) + Ce(IV)]initial , [Ce(IV)]initial = 0 M in this study). The energy consumption (EC) may be calculated using EC = EIt/me , where E is the cell voltage. 3. Results and discussion 3.1. Effects of current density and electrolysis time on Ce(III) oxidation Current density is an important parameter that may influence electrochemical oxidation of Ce(III). In the absence of anion impurities, constant-current electrolysis at 0.05–0.5 A cm−2 was adopted

E ◦ = 1.61 V

2H2 O → O2 + 4H + 4e

−

◦

E = 1.229 V

(4) (5)

The former (Ce(III) oxidation) was less enhanced than the latter (anodic water electrolysis) with increasing overpotential (due to the increase of current density), and thus the CE was lowered. On the other hand, the Ce(IV) yield increased with increasing electrolysis time using the tested current densities; however, the increase in Ce(IV) yield after 120 min electrolysis was slight when the current density reached 0.3 A cm−2 or higher (Fig. 1b). At 0.3 A cm−2 , the CE and yield were (24 ± 0.63)% and (54 ± 1.42)%, respectively (Table 1). Within the electrolysis time of 120–240 min, the Ce(IV) yields obtained were similar using 0.3 and 0.5 A cm−2 , but the CE was greater at the former current density than at the latter. In addition, higher anodic potentials resulted from higher current densities favor the formation of platinum oxides that are detrimental for the reaction kinetics [19]. Therefore, the current density of 0.3 A cm−2 and the electrolysis time of 120 min were

K.-L. Huang et al. / Journal of Hazardous Materials 171 (2009) 755–760

757

Table 1 Parameters obtained from 0.2 M Ce(III) oxidation on Pt (1 cm2 ) in 4 M HNO3 with/without anion impurities in the undivided cell at different current densities (constant-current electrolysis). J, A cm−2 0.05 0.1 0.3 0.5 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3

Impurity

SO4 2− SO4 2− SO4 2− Cl− Cl− Cl− Cr2 O7 2− Cr2 O7 2− Cr2 O7 2− Multi*

Conc., M

E, V

EC†, Wh g−1

CE†, %

Y†, %

104 k1 , s−1

104 k2 , s−1

103 kL1 , cm s−1

103 kL2 , cm s−1

0.6 1.1 2.1 3.6 2.0 2.1 2.1 3.2 8.8 19.5 2.6 2.9 3.0 19.9

64 42 24 15 24 24 23 16 6 3 19 17 17 3

24 32 54 56 55 54 52 35 13 6 43 38 37 6

0.40 0.57 1.32 1.80

0.08 0.25 0.71 1.01

2.0 2.9 6.6 9.0

0.4 1.3 3.6 5.1

0.01 0.1 0.2 0.01 0.04 0.08 0.005 0.01 0.016 Multi*

2.1 2.4 2.6 2.8 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6

0.10

0.72

0.5

3.6

J: current density; E: cell voltage; EC: energy consumption; CE: current efficiency; Y: Ce(IV) yield; k1 and k2 : apparent rate constants for Ce(III) oxidation at anode and Ce(IV) reduction at cathode, respectively; kL1 and kL2 : apparent mass transfer coefficients for Ce(III) oxidation at anode and Ce(IV) reduction at cathode, respectively; *: 0.025 M SO4 2− + 0.08 M Cl− + 0.016 M Cr2 O7 2− ; and †: data at 120 min.

used to test effects of other mentioned parameters on the Ce(III) oxidation. In the undivided cell, the electro-reduction of Ce(IV) to Ce(III) on cathode also influenced the variations of Ce(IV) yield and CE with time. At the cathode, the primary and secondary reactions were hydrogen evolution (2H+ + 2e− → H2 , E◦ = 0.0 V) and Ce(IV) reduction (Ce4+ + e− → Ce3+ , E◦ = 1.61 V), respectively [20]. In addition, the cathodic reduction of nitrate ions might also occur (i.e., NO3 − + 4H+ + 3e− → NO + 2H2 O, E◦ = 0.957 V). It is common that hydrogen evolution is kinetically controlled while Ce(IV) reduction is mass transfer controlled at high current densities [19]. Therefore, the cathode current efficiency for hydrogen evolution was promoted by high current densities whereas the anode current efficiency for Ce(III) oxidation was raised by low current densities. At high current densities with sufficient overpotentials, the Ce(III) oxidation (primary reaction) and oxygen evolution (secondary reaction) on anode were also mass transfer and kinetically controlled, respectively. 3.2. Apparent rate constants/mass transfer coefficients and modeling of Ce(III) oxidation The rate constants experimentally determined can be plugged into numerical models to predict the Ce(IV) yield along with time in the system. If the forward (Ce(III) oxidation) and backward (Ce(IV) k1

Incorporating the concentrations of Ce(III) and Ce(IV) at 240 min as the CAe and CBe respectively, the ln[(CA0 − CAe )/(CA − CAe )] vs. time plots for the constant-current electrolysis cases were linear with R2 = 0.97–0.99 (Fig. 2), verifying the first-order rate dependence of reaction. At 0.05–0.5 A cm−2 , the k1 values obtained from the plots were (0.69–2.06) × 10−4 s−1 , corresponding to k2 values of (0.99–1.24) × 10−4 s−1 . However, if plugging the determined k1 and k2 values into the following equation obtained form the integration of Eq. (6) CA =

CA0 k2 (1 − e−(k1 +k2 )t ) + CA0 e−(k1 +k2 )t k1 + k2

(9)

the modeled Ce(IV) yield curves for the 0.3 and 0.5 A cm−2 cases fairly matched the experimental data in 120–240 min but the experimental data before 120 min were lower than the modeled curves; at 0.05 and 0.1 A cm−2 , most of experimental data were significantly smaller than modeled values, implying that the overestimation of k1 and k2 . This is because a simple batch reactor, by its nature, operates in an unsteady state, and the reactant and product concentrations are time-dependent [21]. Therefore, the deviation between modeled curves and experimental data is mainly attributed to the CAe values taken at 240 min. The Ce(IV) yields of 0.3 and 0.5 A cm−2 cases approached plateaus (pseudo steady-state) but those of 0.1 and 0.05 A cm−2 cases did not at 210–240 min (Fig. 1b), reflecting the overestimation of CAe and underestimation of CBe .

k2

reduction) reactions are expressed as A−→B and B−→A, respectively, the rate equation may be written as −

dCB dCA = = k1 CA − k2 CB dt dt

(6)

where CA and CB are the concentrations of Ce(III) and Ce(IV) at time t, respectively, whereas k1 and k2 are the apparent rate constants for Ce(III) oxidation at anode and Ce(IV) reduction at cathode in 4 M HNO3 , respectively. The k1 can be obtained from the slope of ln[(CA0 − CAe )/(CA − CAe )] vs. time plots [6], ln

 C − C  A0 Ae CA − CAe

=

 1  XAe

k1 t

(7)

where CA0 and CAe are the initial and equilibrium concentrations of Ce(III), respectively, and XAe (=1 − (CAe /CA0 )) is the fractional conversion at equilibrium, while the k2 can be calculated using the equilibrium constant (Kc ) Kc =

CBe k1 = CAe k2

where CBe is the concentration of Ce(IV) at equilibrium.

(8)

Fig. 2. Plots of ln[(CA0 − CAe )/(CA − CAe )] vs. time (with regression lines) for the Ce(III) oxidation in 4 M HNO3 at different constant-current electrolysis with/without anion impurities (Pt anode: 1 cm2 ).

758

K.-L. Huang et al. / Journal of Hazardous Materials 171 (2009) 755–760

Instead, the k1 and k2 values were also determined using data fitting (R2 > 0.99). (The Ce(IV) yield (0 ≤ Y ≤ 1, not in %) = 1 − (CA /CA0 ) and the CA /CA0 can be obtained from Eq. (9), so the equation Y = a(1 − e−bt ) (where a = k1 /b and b = k1 + k2 ) was used to fit experimental data and determine the values of a, b, k1 , and k2 .) The fitted k1 and k2 values were (0.40–1.80) × 10−4 and (0.08–1.01) × 10−4 s−1 , respectively (increasing with increasing current density) (Table 1). Plugging the fitted k1 and k2 into Eq. (9), one can find that the fitting curves well match the experimental data (Fig. 1b). The corresponding apparent mass transfer coefficients (kLX ) were calculated by kLX = kx (V/Ae ) where x = 1 or 2; V and Ae : volume of electrolyte and projected geometric area of electrode, respectively. Accordingly, the calculated kL1 ((2.0–9.0) × 10−3 cm s−1 ) and kL2 ((0.4–5.1) × 10−3 cm s−1 ) also increased with increasing current density (Table 1). Although operating conditions (e.g., anode type, electrolyte, and cell configuration) may affect the magnitude of apparent mass transfer coefficient, the determined kL1 values were comparable to that (1.4 × 10−3 cm s−1 ) for 0.42 M Ce(III) oxidation (methanesulfonate in methanesulfonic acid (MSA)) on a flat plate in an MP (multipurpose) cell [1] and those ((4.0–8.3) × 10−3 cm s−1 ) for 0.46 M Ce(III) oxidation (methanesulfonate in MSA) on (plate and mesh) DSA-O2 anodes in batch/batch re-circulation/MP cells [2]. The kL1 (2.0 × 10−3 cm s−1 ) at 0.05 A cm−2 was similar to a value (2.2 × 10−3 cm s−1 ) for 0.035 M Ce(III) oxidation (in 5 M HNO3 ) at 0.04 A cm−2 in literature which shows that the apparent mass transfer coefficient obtained using a platinized titanium DSA is independent of anolyte supply flowrate (0.15–2 L h−1 ) in a divided cell [22]. Juang and Wang [13] also indicated that CE strongly depended on mass transfer in addition to electrolysis time, current density, and cell geometry. 3.3. Effect of Aa /V ratio on Ce(III) oxidation Three ratios of anode area to electrolyte volume (Aa /V) (1/100 (0.01), 1/50 (0.02), and 2.25/50 (0.045)) were tested for the Ce(III) electro-oxidation. The three Aa /V cases exhibited statistically different means of Ce(IV) yields or CEs in triplicate measurements (n = 3) at a significant level of ˛ = 0.05. The Ce(IV) yield along with CE in 4 M HNO3 increased with increasing Aa /V ratio (Fig. 3). This increase of Ce(IV) yield was much higher for increasing Aa /V from 0.01 to 0.02 than for increasing Aa /V from 0.02 to 0.045. A similar tendency was also observed for the CE. At Aa /V = 0.045 ([Ce(III)]initial = 0.2 M, electrolysis time = 120 min, and I = 0.3 A), the maximum Ce(IV) yield and current efficiency were (58 ± 1.21)% and (26 ± 0.54)%, respectively, lower than those reported in literature using greater Aa /V ratios for different electrodes and electrolytes [2–5,18]. Although

Fig. 3. Effect of anode area/electrolyte volume ratio (A/V = 1/100, 1/50, and 2.25/50 in cm−1 ) on current efficiency and Ce(IV) yield (n = 3, with one standard deviation in each error bar) ([Ce(III)]initial = 0.2 M, electrolysis time = 120 min, and I = 0.3 A).

the cost-effectiveness needs to be considered, the increase of Ce(IV) yield and current efficiency may be further tested via using larger Aa /V or Ac /V ratios (Ac : cathode area). The Ce(IV) yield and current efficiency are also influenced by anode material, electrolyte concentration, or initial Ce(III) concentration (see more discussion in Sections 3.4–3.6). 3.4. Effect of anode material on Ce(III) oxidation In this study, platinum plate, DSA (IrO2 /Ti), and glassy carbon (GC) materials were tested and compared for their performance on Ce(III) oxidation. At 0.3 A cm−2 , the Ce(IV) yields or CEs of the three different anode cases were significantly different (˛ = 0.05), and the performance of anodes was in the order Pt > DSA > GC for both Ce(IV) yield and CE in 4 M HNO3 (Fig. 4), inconsistent with that of their oxygen overpotentials because their Ce(III) oxidation overpotentials also need to be considered. Spotnitz et al. [1] reported a similar observation for Ce(IV) generation in 2 M methanesulfonic acid in a cell using a Nafion separator. After the electrolysis in 4 M HNO3 , the surface of GC electrode was damaged and blisters on the electrode surface were observed, which might be resulted from the oxidation of carbon on electrode surface by nitric acid. Paulenova et al. [23] also observed that the surface of a GC electrode was damaged after the voltammetry in 4 M H2 SO4 at 40 ◦ C or in 2 M H2 SO4 at 60 ◦ C. Therefore, the GC anode is not suitable for the Ce(IV) electro-regenerating system. 3.5. Effect of electrolyte (nitric acid) concentration on Ce(III) oxidation Fig. 5 shows that Ce(IV) yield and CE increase with the increasing electrolyte (nitric acid) concentration for Ce(III) oxidation at 0.3 A cm−2 . At ˛ = 0.05, the comparison of 1 vs. 4 M HNO3 or 0.1 vs. 1–4 M HNO3 was statistically different in Ce(IV) yield (or CE) while that of 1 vs. 2 M HNO3 or 2 vs. 4 M HNO3 was not. Both the Ce(IV) yield and CE were significantly lowered when the HNO3 concentration decreased from 1 to 0.1 M, which is mainly related to the decrease in kinetics of Ce(III) oxidation. Wei et al. [18] indicated that a higher proton concentration is electrochemically favorable for the electron transfer of the Ce(III)/Ce(IV) couple (on a glassy carbon electrode) although they used 0.02 M Ce(III) in 1–6 M HNO3 in experiments. Balaji et al. [4] also observed that Ce(IV) yield increased with increasing HNO3 concentration and the production of Ce(IV) with a better current efficiency occurred at 10 A and 80 ◦ C on an IrO2 /Ti DSA mesh (140 cm2 ) in 2–3.5 M HNO3 , but they used

Fig. 4. Effect of anode material on current efficiency and Ce(IV) yield (n = 3, with one standard deviation in each error bar) ([Ce(III)]initial = 0.2 M, Pt cathode: 1 cm2 , electrolysis time = 120 min, and I = 0.3 A).

K.-L. Huang et al. / Journal of Hazardous Materials 171 (2009) 755–760

759

120 min electrolysis, the case of [Ce(III)]initial = 0.05 M only achieved negligible CE and Ce(IV) yield ((0.19 ± 0.27)% and (0.38 ± 0.54)%, respectively). 3.7. Effects of anion impurities on Ce(III) oxidation

Fig. 5. Effect of nitric acid concentration on current efficiency and Ce(IV) yield (n = 3, with one standard deviation in each error bar) ([Ce(III)]initial = 0.2 M, Pt anode:1 cm2 , electrolysis time = 120 min, and I = 0.3 A).

a more complicated system in which anolyte and catholyte solutions were circulated through a electrochemical cell installed with a Nafion 324 separator. Side reactions at the anode and cathode also influenced the Ce(IV) yield and CE. The anodic water electrolysis was much enhanced at the lowest HNO3 concentration, resulting in more evolution of O2 (visual observation) to hinder Ce(III) diffusion from bulk to anode surface. At t = 120 min, the total supplied charge was 2.24 × 10−2 F (Faraday) and the corresponding current efficiencies of Ce(III) oxidation ranged from (1.2 ± 0.28)% to (24 ± 0.63)% in 0.1–4 M HNO3 in which ∼(4.3–5.5) × 10−3 moles of O2 evolved if the left supplied charge was used mainly for water electrolysis. On the other hand, at 1–4 M HNO3 , the cathode side reaction (2H+ + 2e− → H2 ) was much improved to inhibit the Ce(IV) reduction. 3.6. Effect of Ce(III) initial concentration on Ce(III) oxidation The three cases with different Ce(III) concentrations in 4 M HNO3 differed significantly in Ce(IV) yield (or CE) (˛ = 0.05) that increased with increasing initial Ce(III) concentration (0.05–0.20 M) at 0.3 A cm−2 (Fig. 6). (In fact, the Ce(IV) yield = fractional conversion × 100% = [1 − (CA /CA0 )] × 100%.) At higher Ce(III) concentrations, the increase of driving force for Ce(III) mass transfer and electrolyte conductivity also improved the Ce(IV) yield. After

Fig. 6. Effect of initial Ce(III) concentration on current efficiency and Ce(IV) yield (n = 3, with one standard deviation in each error bar). (Pt anode: 1 cm2 , electrolysis time = 120 min, and I = 0.3 A.)

Effects of SO4 2− , Cl− , and Cr2 O7 2− impurities on Ce(III) oxidation were also examined at 0.3 A cm−2 constant-current electrolysis in 4 M HNO3 . With increasing SO4 2− concentration (0.01–0.2 M), the cell voltage and energy consumption were found to be almost unchanged whereas both Ce(IV) yield and CE slightly decreased (Table 1), although the presence of SO4 2− in nitric acid may change the dominant Ce(III) active species and is electrochemically unfavorable for the Ce(III) oxidation [15,16]. This result is also attributed to the shift of formal potential to less positive values [15,16,18] and the species dominance of Ce3+ > Ce(III)-sulfate complexes [24]. Dissimilarly, Wei et al. [18] found that Ce(IV) yield (%) increased with H2 SO4 concentration (0.5–2 M) at potentiostatic electrolysis of 0.1 M Ce(III). At this point, if comparing their and our observations, it is inferred that both sulfate and proton influence the Ce(IV) yield. The increase of chloride concentration (0.01–0.08 M) in 4 M HNO3 significantly increased the energy consumption but decreased the Ce(IV) yield and CE, although the cell voltages were the same (Table 1). Chloride ions were oxidized via 2Cl− → Cl2 + 2e− (E◦ = 1.358 V) and this reaction competed with the Ce(III) oxidation on anode, leading to the decline of Ce(IV) yield by ∼35–89%, in comparison to that in the absence of Cl− . Therefore, the presence of Cl− in 4 M HNO3 is also electrochemically unfavorable for the Ce(III) oxidation, consistent with the previous observation on glassy carbon electrodes [15,16]. Also, a PtCl2 film was possibly formed on the electrode surface to obstruct the Ce(III) oxidation. Some earlier studies reported that the rate of electrochemical oxidation of H2 O2 at platinum also decreased markedly in the presence of chloride, because of the formation of a PtCl2 barrier film on electrode surface [25,26]. The variation of energy consumption, CE, and Ce(IV) yield with increasing Cr2 O7 2− concentration (0.005–0.016 M) was less significant in comparison to that of Cl− case, although both the cases exhibited a similar variation trend of these parameters (Table 1). The decrease of Ce(IV) yield (or CE) in the presence of anion impurities was in sequence Cl− > Cr2 O7 2− > SO4 2− . At 0.016 M Cr2 O7 2− , both Ce(IV) yield and CE were lowered by ∼31%, when compared to those in the absence of dichromate. This inhibition of Ce(III) oxidation might be associated with the passivation of anode surface by Cr(VI) [27]. When the anion impurities were co-present (0.025 M SO4 2− , 0.08 M Cl− , and 0.016 M Cr2 O7 2− ) in 4 M HNO3 , the CE and Ce(IV) yield were only 3% and 6%, respectively (Fig. 1a and b, respectively), at 0.3 A cm−2 for 120 min electrolysis and they were both lowered by 89%, if compared to those in the absence of the impurities. Furthermore, the energy consumption, CE, or Ce(IV) yield was close to that of Cl− case. The k1 and k2 obtained from data fitting (R2 = 0.99) were 0.10 × 10−4 and 0.72 × 10−4 s−1 , respectively (with corresponding kL1 and kL2 of 0.5 × 10−3 and 3.6 × 10−3 cm s−1 , respectively) (Table 1). (Note that the fitting curve for the multianion case was also in good agreement with the experimental data (Fig. 1b).) The k1 was approximately 1/13 that of the withoutimpurity case, although the k2 was almost unchanged. Therefore, the removal of these anion impurities is crucial for the electrolytic recovery of Ce(IV) from Ce(III) oxidation and the removal of Cl− is more critical than that of the other two impurities. Although the simulated solutions were similar to real ones, before practical application, further study is necessary to evaluate effect of NH4 + (another abundant impurity in real spent Cr-etching solutions) on the Ce(IV) regeneration because the cation species may compete with Ce(III) for oxidation on anode.

760

K.-L. Huang et al. / Journal of Hazardous Materials 171 (2009) 755–760

4. Conclusions For Ce(III) oxidation at different current densities in 4 M HNO3 without anion impurities, it was found that current efficiency decreased but Ce(IV) yield increased with increasing electrolysis time. The average CE at 0.05 A cm−2 was about fourfold that at 0.5 A cm−2 . At 0.05–0.5 A cm−2 , the fitted apparent rate constants for Ce(III) oxidation on the Pt anode (k1 ) and for Ce(IV) reduction on the stainless steel cathode (k2 ) ranged in (0.40–1.80) × 10−4 and (0.08–1.01) × 10−4 s−1 , respectively, corresponding to the apparent mass transfer coefficients (kL1 and kL2 ) of (2.0–9.0) × 10−3 and (0.4–5.1) × 10−3 cm s−1 , respectively. At 0.3 A cm−2 , both Ce(IV) yield and CE increased with increasing initial Ce(III) concentration and ratio of anode area to electrolyte volume, exhibited the order of Pt > DSA > GC, and were similar in 1–4 M HNO3 but significantly lower in 0.1 M HNO3 . In the presence of impurities, cell voltages were stable at 0.3 A cm−2 . The energy consumption almost unchanged whereas both Ce(IV) yield and CE slightly decreased with increasing SO4 2− concentration (0.01–0.2 M). The increase of energy consumption and decrease of Ce(IV) yield and CE with increasing impurity concentration were significantly higher for the Cl− case (0.01–0.08 M) than for the Cr2 O7 2− (0.005–0.016 M) case. The decrease of Ce(IV) yield and CE displayed the order of Cl− > Cr2 O7 2− > SO4 2− . When the 4 M HNO3 contained 0.025 M SO4 2− + 0.08 M Cl− + 0.016 M Cr2 O7 2− , the k1 was one order of magnitude lowered, and the CE and Ce(IV) yield were both reduced by ∼89%, in comparison to those of the without-impurity case. Therefore, the removal of these anion impurities before electrolysis is favored for the regeneration of Ce(IV) from Ce(III) oxidation, and the removal of Cl− is more critical than that of the other two impurities. 5. Acknowledgement The authors would like to thank the National Science Council of the Republic of China, Taiwan, for partially financially supporting this research under Contract No. NSC-96-2221-E-020-011-MY2. References [1] R.M. Spotnitz, R.P. Kreh, J.T. Lundquist, P.J. Press, Mediated electrosynthesis with cerium(IV) in methanesulphonic acid, J. Appl. Electrochem. 20 (1990) 209–215. [2] T. Raju, C.A. Basha, Electrochemical cell design and development for mediated electrochemical oxidation Ce(III)/Ce(IV) system, Chem. Eng. J. 114 (2005) 55–65. [3] M. Matheswaran, S. Balaji, S.J. Chung, I.S. Moon, Mineralization of phenol by Ce(IV)-mediated electrochemical oxidation in methanesulphonic acid medium: a preliminary study, Chemosphere 69 (2007) 325–331. [4] S. Balaji, S.J. Chung, R. Thiruvenkatachari, I.S. Moon, Mediated electrochemical oxidation process: electro-oxidation of cerium(III) to cerium(IV) in nitric acid medium and a study on phenol degradation by cerium(IV) oxidant, Chem. Eng. J. 126 (2007) 51–57.

[5] V.V. Kokovkin, S.J. Chung, S. Balaji, M. Matheswaran, I.S. Moon, Electrochemical cell current requirements for toxic organic waste destruction in Ce(IV)-mediated electrochemical oxidation process, Korean J. Chem. Eng. 24 (2007) 749–756. [6] M.E. Armenta-Armenta, A.F. Diaz, Oxidation of benzoic acid by electrochemically generated Ce(IV), Environ. Sci. Technol. 39 (2005) 5872–5877. [7] S. Balaji, V.V. Kokovkin, S.J. Chung, I.S. Moon, Destruction of EDTA by mediated electrochemical oxidation process: monitoring by continuous CO2 measurements, Water Res. 41 (2007) 1423–1432. [8] J.W. Lee, S.J. Chung, S. Balaji, V.V. Kokovkin, I.S. Moon, Destruction of EDTA using Ce(IV) mediated electrochemical oxidation: a simple modeling study and experimental verification, Chemosphere 68 (2007) 1067–1073. [9] M. Matheswaran, S. Balaji, S.J. Chung, I.S. Moon, Mediated electrochemical oxidation of phenol in continuous feeding mode using Ag (II) and Ce(IV) mediator ions in nitric acid: a comparative study, Chem. Eng. J. 144 (2008) 28–34. [10] S. Balaji, S.J. Chung, M. Matheswaran, V.V. Kokovkin, I.S. Moon, Destruction of organic pollutants by cerium(IV) MEO process: a study on the influence of process conditions for EDTA mineralization, J. Hazard. Mater. 150 (2008) 596–603. [11] G. Chen, Electrochemical technologies in wastewater treatment, Sep. Purif. Technol. 38 (2004) 11–41. [12] C.A. Martinez-Huitle, S. Ferro, Electrochemical oxidation of organic pollutants for the wastewater treatment: direct and indirect processes, Chem. Soc. Rev. 35 (2006) 1324–1340. [13] R.J. Juang, S.W. Wang, Metal recovery and EDTA recycling from simulated washing effluents of metal contaminated soils, Water Res. 34 (2000) 3795–3803. [14] T. Tzedakis, A. Savall, Electrochemical regeneration of Ce(IV) for oxidation of p-methoxytoluene, J. Appl. Electrochem. 27 (1997) 589–597. [15] T.S. Chen, K.J.C. Yeh, K.L. Huang, Anion effects on the electrochemical regeneration of Ce(IV) in nitric acid used for etching chromium, J. Hazard. Mater. 152 (2008) 922–928. [16] T.S. Chen, K.L. Huang, K.J.C. Yeh, Kinetic and mass transfer parameters for Ce(III) electro-oxidation in nitric acid with/without anion impurities, J. Electrochem. Soc. 156 (3) (2008) E69–E74. [17] C. Dumas, A. Mollica, D. Féron, R. Basseguy, L. Etcheverry, A. Bergel, Checking graphite and stainless anodes with an experimental model of marine microbial fuel cell, Biores. Technol. 99 (2008) 8887–8894. [18] Y. Wei, B. Fang, T. Arai, M. Kumagai, Electrolytic oxidation of Ce(III) in nitric acid and sulfuric acid media using a flow type cell, J. Appl. Electrochem. 35 (2005) 561–566. [19] T.H. Randle, A.K. Kuhn, Kinetics and mechanism of the cerium(IV)/cerium(III) redox reaction on a platinum electrode, J. Chem. Soc. Faraday Trans. I 79 (1983) 1741–1756. [20] J. Been, C.W. Oloman, Electrochemical regeneration of ceric sulphate in an undivided cell, J. Appl. Electrochem. 23 (1993) 1301–1309. [21] D. Pletcher, F.C. Walsh, Industrial Electrochemistry, Blackie Academic & Professional, New York, 1993, pp. 99. [22] C. Eysseric, H. Chifflet, S. Picart, J.M. Adnet, Kinetic aspects of the behavior of a continuous electrolyzer dedicated to actinides and lanthanides oxidation applied to their separation, http://www-atalante2004.cea.fr/home/liblocal/ docs/atalante2000/P3-04.pdf (2004). [23] A. Paulenova, S.E. Creager, J.D. Navratil, Y. Wei, Redox potentials and kinetics of the Ce3+ /Ce4+ redox reaction and solubility of cerium sulfates in sulfuric acid solutions, J. Power Sources 109 (2002) 431–438. [24] B. Fang, S. Iwasa, Y. Wei, T. Arai, M. Kumagai, A study of the Ce(III)/Ce(IV) redox couple for redox flow battery application, Electrochim. Acta 47 (2002) 3971–3976. [25] D.G. Peters, J.J. Lingane, Anodic formation and chemical analysis of oxychloride films on platinum electrodes, Electroanal. Chem. 4 (1962) 193–217. [26] J.E. Harrar, Controlled-potential coulometric determination of hydrogen peroxide, Anal. Chem. 35 (1963) 893–896. [27] M. Hunsom, K. Pruksathorn, S. Damronglerd, H. Vergnes, P. Duverneuil, Electrochemical treatment of heavy metals (Cu2+ , Cr6+ , Ni2+ ) from industrial effluent and modeling of copper reduction, Water Res. 39 (2005) 610–616.