- Email: [email protected]
Selective iron sorption for on-line reclaim of chromate electroplating solution at highly acidic condition
Chemical Engineering Journal 281 (2015) 434–443
Contents lists available at ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Selective iron sorption for on-line reclaim of chromate electroplating solution at highly acidic condition Lan-Ping Wei a, Po-Yi Liu a, Kai-Chen Lin a, Je-Ming Yi a, Cyuan-Wun Liou a, Guo-Chuan Yang a, Hsi-Yen Hsu b, Tseng-Chang Tsai a,⇑ a b
Department of Applied Chemistry, National University of Kaohsiung, Kaohsiung 816, Taiwan Div. of Chemical Engineering, Material and Chemical Res. Lab., Industrial Technology Research Institute, HsinChu 300, Taiwan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Maximum Fe contaminant level in
Stable, Regenerable SIR for Selective Fe Removal from Cr Bath
a r t i c l e
i n f o
Article history: Received 21 March 2015 Received in revised form 17 June 2015 Accepted 18 June 2015 Available online 3 July 2015 Keywords: Fe selective sorption DEHPA SIR resin Sorption isotherm Cr electroplating Fe(II)–Fe(III) analysis
Regeneration > 22 cycles
Sorption Capacity (mg/g)
chromate electroplating limited below 50 mg/L. DEHPA impregnated S957 for selective Fe sorption and on-line reclaim of contaminated chromate bath. Complementary sorption selectivity of Fe(III) by DEHPA and S957. Improving electroplating performance with enhancing current efficiency and reducing waste water.
Desorption Efficiency Fe 100% Cr 50%
140 120 100 80
D2-E-S957 Fe de-Fe Cr de-Cr
60 40 20 0 0
5
10
15
http://dx.doi.org/10.1016/j.cej.2015.06.080 1385-8947/Ó 2015 Elsevier B.V. All rights reserved.
30
On-line reclaim process by selective ion sorption for the purification of highly acidic Fe contaminated chromate electroplating bath has been developed. Among many industrial functional resins, Lewatit VPOC1026 implanted with di-(2-ethylhexyl) phosphoric acid (DEHPA) extractant in the polymeric backbone exhibited the highest selectivity of Fe versus Cr but low sorption capacity of Fe. Impregnation of DEHPA on functionalized S957 resin (a sulfonated mono-phosphonic acid resin) exhibited a much enhanced Fe sorption capacity with high Fe selectivity. The enhancement in the adsorption performance of S957 is due to the selective adsorption of DEHPA for Fe(III) complex formation. For the best electroplating performance, the maximum Fe contaminant level in the chromate electroplating bath was determined as 50 mg/L. The developed process could not only maximize electroplating performance but also minimize the usage of original water and the production of waste water. Ó 2015 Elsevier B.V. All rights reserved.
Metal corrosion always leads to massive economic loss in gross natural product. It was accounted for around 4–5%. Traditionally electroplating of secondary metal as a protection layer on the surface of metal substrate could improve its corrosion resistance, mechanical hardness and surface smoothness. Chromium (Cr)
E-mail address: [email protected] (T.-C. Tsai).
25
a b s t r a c t
1. Introduction
⇑ Corresponding author. Fax: +886 7 5919348.
20
Sorption Cycle Number
coating, traditionally the most effective coating, is mostly fabricated by electroplating process or chromate conversion coating (CCC) process using very strong acidic chromate solution (pH < 2). During the electro-deposition of carbon steel or alloy, many metal components of the substrate such as iron (Fe), copper, etc. are apt to leach out into the highly acidic CCC and electroplating bath solutions and accumulate to form contaminants after long service time. Minor contaminants in bath solution particularly Fe could reduce the current efficiency during electro-deposition and downgrade electroplating product quality. To keep the contaminant level as
435
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
low as possible, electroplating industry usually shortens the service life of CCC bath solution with excessive changeover rate. The typical practice not only requires additional fresh water for preparing CCC bath solution but also generates large amount of Cr waste water solution which are harmful to the Earth. In principle, it is technology feasible to selectively remove Fe ion impurity from spent electroplating solutions such as Cr pickling solution, electroplating baths, and passivation baths, etc. Some novel adsorbents for selective adsorption of heavy metals have been reported, such as titanate nanoflower for selective sorption of Cd(II) relative to Zn(II) and Ni(II) [1], magnetic MCM41 for selective adsorption of Cr(VI) in magnetic field [2]. In strong acidic electroplating solution, on one hand, sorption equilibrium could be more favorable for desorption to adsorption; on the other hand, hard metal ion has greater sorption affinity than soft metal ion. As such, selective sorption of Fe(II) and Fe(III) ions (intermediate metal ions) from Cr(VI) and Cr(III) (hard metal ions) electroplating solution is a great technology challenge. Many ion-exchange resins are available nowadays for metal sorption. While many chelating resins have been successfully used for removing heavy metals from wastewaters and spent baths, e.g., Lewatit TP-207 [3,4], Dowex M4195, Amberlite IRC748 [5], and Purolite S-950 [6], only few resins, such as Purolite S957 (sulfonated mono-phosphonic resin), have been reported for selective removal of iron ions from Cr(III) passivation bath [7,8]. Metal extraction using organic extractant has been proposed for metal recovery. For example, for the recovery of zinc from Fe(II)– Fe(III) containing spent pickling solution, organophosphorus extractant tributylphosphate (TBP) [9], octylphenyl acid phosphate (OPAP) and di-(2-ethylhexyl)phosphoric acid (DEHPA) were used [10]. The sorption process could be complementary to extraction process in terms of small equipment size, low energy consumption, sustainable operation in acidic solution, etc. In a recent review paper, Kabay et al. discussed comprehensively the preparation and application of various extractant-containing resins [11]. Two different approaches have been developed for the preparation of extractant-containing solid sorbents. The Levextrel resins were patented using polymerization process, by mixing the extractant with a mixture of monomers (styrene and divinylbenzene), for implanting extractant into polymer backbones. The Levextrel resins available today include extractant TBP, DEHPA and di(2,4,4-trimethylpentyl)phosphinic acid (DTMPPA). The Lewatit VPOC1026 implanted with DEHPA functional group is useful for the extractive sorption of divalent metal ions such as Zn(II), Cu(II) and Cd(II). Other preparation method so-called solvent-impregnated resin (SIR) by impregnation of extractant on various solid sorbents is versatile and flexible in tailoring the types and contents of
extractant, solid sorbent and preparation procedures, etc., providing a wide range of metal separation characteristics. Because of high mobility of ligands, the SIR resin in most of time could perform superior separation property to some widely used ion-exchange resins and chelating resins [12]. Numerous extractants have been used for the preparation of SIR resins, including organophosphorous extractant such as DTMPPA [13], DEHPA [14]; amine group such as tri-n-dodecyl-ammonium chloride [15]; quinolinol/quinone group such as 7-(4-ethyl-methyloc tyl)-8-quinolinol (Kelex 100) [16]; and macrocyclic functional group [17]. Some macroporous resins, dodecyl acrylate – divinylbenzene crosslinked polymeric organogels [18] and silica-based composites [19] have been used for solid sorbent. Various SIR solid sorbents have been reported such as DEHPA modified Amberlite 200 for enhanced sorption of Ni, V and Mo [20], DEHPA modified XAD-2 for extraction of Cu ion [12], DEHPA modified HZ-803 for extraction of In(III), Ga(III) and Zn(II) [21], etc. The present study develops an on-line reclaim process for highly acidic electroplating solution by selective removal of Fe impurity using DEHPA impregnated resins. The SIR modified S957 by DEHPA impregnation exhibited a much enhanced Fe sorption capacity and selectivity, exceeding the DEHPA implanted VPOC1026 resin. The reclaim process could significantly extend the service life of bath solution, reduction of usage of raw materials, fresh water and production of Cr waste water; and also improve the current efficiency during electroplating and electro-deposition product quality. 2. Experimental 2.1. Materials Some sorbents (Table 1) including industrial resins and SIR resins were tested in this study. Sodium hydroxide, nitric and hydrochloric acid, and hexane (Merck) were used for the preparation of the different solutions. DEHPA was provided by Tokyo Chemical Industry. Three industrial electroplating bath solutions provided by local galvanic industry (deployed as B1, B2 and B3 solution) were analyzed by Perkin Elmer model Optical 2100DV Inductively-Coupled Plasma Atomic Emission Spectroscopy (ICP-OES) for the determination of the concentrations of metals (Cr, Fe, and Cu) and ion chromatography for anions (chloride and sulfate). The Cr(III)–Cr(VI) and Fe(II)–Fe(III) partitions in the electroplating solution were determined with UV spectroscopy The methodology for the determination of Fe(II)–Fe(III) will be discussed in Section 3.2. For determining Cr partition, the concentrations of
Table 1 Description of commercial and modified resins. ID
Commercial name
Functional group
Group A–SAC resin A200C PK228
Amberlite 200C Diaion PK228
Sulfonic acid Sulfonic acid
Group B–CKL resin C100 TP207 VPOC1026 S957
Amberlite Chelex 100 Lewatit Mono Plus TP207 Lewatit VPOC1026 Purolite S957
Iminodiacetic acid Iminodiacetate groups DEHPA (0.84 mmol/g) Sulfonated monophosphonic acid
Group C–AKL resin P4V DM4195
2% Poly(4-vinylpyridine hydrochloride) Dowex M4195
4-Vinylpyridine hydrochloride Bis-picolylamine
Solvent impregnated resin D2-E-S957 D2-H-S957 D2-E-XAD
S957 impregnated with ethanol solution of DEHPA (0.82 mmol/g) S957 impregnated with hexane solution of DEHPA (0.62 mmol/g) XAD-2 impregnated with ethanol solution of DEHPA (0.58 mmol/g)
436
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
Cr(VI) and total Cr ion were measured from the absorbances at 540 nm of UV spectra of the pristine electroplating bath and the one pre-treated with ammonium persulfate solution by which Cr(III) was oxidized into Cr(VI) form. The ratio of UV absorbances at 540 nm of the two spectra would give the percentage of Cr(VI) ion in the electroplating bath. For a typical electroplating bath B1, the partitions of metal ions were around 47% Cr(III) – 53% Cr(VI) and 35% Fe(II) – 65% Fe(III). The formation of Cr(III) could be attributed to the reduction of Cr(VI) in highly acidic solution [22].The stock solutions of Fe(III) and Cr(VI) (deployed as C1, PF and PC solution) for simulating the bath solution B1 were prepared by dissolving quantified amounts of chromium oxide (CrO3, Riedel-deHaën) and FeCl3 (Riedel-deHaën) in deionized water and were standardized by ICP-OES. All the electroplating solutions were used without pH adjustment. 2.2. Solvent impregnation process S957 and XAD-2 were used as sorbents subjected for incorporating DEHPA. S957 was first pretreated in 4 M HCl solution of methanol–water mixed solvent for 12 h to remove some possible impurities [23], then immersed in distilled-deionized water for 12 h and at last dried. The dried S957 was stirred in 25 ml of DEHPA containing n-hexane (or ethanol) solution at 25 °C for 6–48 h, and then filtered and dried at 70 °C to form the DEHPA impregnated S957 (denoted as D2-H-S957 or D2-E-S957). Since DEHPA is a weak acid with pKa of 1.4, the DEHPA content in the filtrate of SIR preparation and the ethanol extract solution of D2-E-S957 resin were determined with NaOH titration. The amount of DEHPA impregnated in VPOC1026 was evaluated by washing a proper amount of resin with ethanol. Ethanol could completely elute the ligand. Subsequently the ethanol effluent was titrated by NaOH. The total extractant capacity of the resin was 0.84 mol kg1 dry resin. 2.3. Sorption isotherm of metal ion Single or multicomponent sorption capacities of pure or mixed metal ion solutions were determined in batch operation. During sorption experiment, 1 g of ion-exchange resin was in contact with a stock solution of Fe or Cr concentration in the range of 0.2–2.2 g/L at various solid/liquid ratios for 1 h under constant magnetic stirring at 55 °C, then the solution was centrifuged and the solid was removed and washed into the solution. The sorption test reached equilibrium composition after sorption time of 1 h (Referring to Supplementary Information (SI) SF1). Therefore, most sorption tests were conducted for 1 h. The concentration of metal ion in the filtrate was analyzed by Perkin Elmer model Optical 2100DV ICP-OES. The equilibrium capacity of the resin was calculated after measuring the final (equilibrium) iron concentrations of the solution. 2.4. Breakthrough curves of different resins The breakthrough curves of different resins were measured in a continuous flow sorption system. One ion-exchange resin of 150 ml was loaded into the sorption chamber. Prior to the sorption measurement, the ion-exchange resin was activated initially with HCl solution. An electroplating solution was injected by a mini peristaltic pump (DG Instrument, model C9ES DG-100 N) at a preset flow rate going through the sorption chamber kept at 55 °C and atmospheric pressure for sorption measurement. The effluent at the sorption exit was collected periodically and analyzed with Perkin Elmer model Optical 2100DV ICP-OES. For regeneration treatment, 14 wt% HCl solution of 4 bed volume (BV) was passed through the ion-exchange resin chamber.
2.5. Electroplating The quality of electroplating solution particularly the effect of impurity was evaluated in a miniature electroplating unit so-called Hull Cell (R.O. Hull and Co., Inc.) with scribed solution level line of 267 ml. The Hull Cell 1 in Lucite was a trapezoidal structure in which the cathode using brass plate was placed at an oblique angle with respect to the anode. Electro-deposition was carried out at 55 °C temperature and 5–7 v with variable amperage.
3. Results and discussions 3.1. Resin evaluation for selective Fe removal As shown in Table 2, all the tested electroplating solutions were very acidic with pH value around 0. Their Cr and Fe contents were in the same range of 50–160 gL1 and 0.2 mg L1–22 g L1, respectively. The bath solution B1 had a much higher Fe content than other bath solutions B2 and B3. The equilibrium sorption capacities of ion-exchange resins were evaluated using different electroplating solutions. Four groups of resins were used (Table 1) classified as Group A for strong acidic ion-exchange (SAC) resin, Group B and C for cationic and anionic chelating resin (CKLR and AKLR), and solvent-impregnated resins. As shown in Table 3, most ion exchange resins except for C100 and P4V possessed higher sorption capacity of Fe for pure component solution PF than those in other binary component solutions. Furthermore, all the tested ion-exchange resins exhibited higher Fe sorption capacity for the sorption of B1 solution than the capacities in the sorption of other electroplating solutions. On the other hand, Cr sorption capacities of all the tested resins except for VPOC1026 were about the same in different electroplating solutions In reference to the other term denoted as Fe sorption selectivity to be defined below, a term denoted as Cr/Fe capacity ratio was defined as ratio of the sorption capacity of Cr versus that of Fe. For example, Fe and Cr sorption capacity of S957 was 26.5 and 105.3 mg/g, respectively for B1 solution, and 0.3 and 105.7 mg/g, respectively for B3 solution. Therefore the Cr/Fe capacity ratio over S957 was 4.0 and 352.3 for B1 and B3 solutions, respectively. A strong effect of the concentration of adsorbate solution on the sorption capacity of adsorbent was reported earlier [24]. The exceptional behavior of VPOC1026 could be associated with its superior Fe sorption selectivity. Therefore, with increasing Fe concentration in B1 electroplating solution, Fe sorption capacity of VPOC1026 increased in compensation of Cr sorption capacity. Apparently, in variation with metal concentration of adsorbate and sorption selectivity of adsorbent, Cr was adsorbed competitively with Fe. As a result, Cr/Fe capacity ratio changed with the ratio of Cr/Fe content of adsorbent solution. The ratio could be used for rough evaluation of sorption selectivity of ion exchange resin. Rigorous definition for Fe sorption selectivity based on the model parameter of sorption isotherm by taking account of the composition effect of the solution will be discussed later (Section 3.3).
Table 2 Compositions of electroplating bath solutions and stock solutions. Solution
H
Cr (gL1)
Fe (g L1)
Cu (mg L1)
Cl (g L1)
1 SO2 ) 4 (g L
B1 B2 B3 C1 PF PC
0 0 0 0 1 0
159.6 50.4 146.7 139.5 0 139.5
21.7 6.5 0.2 21.8 21.8 0
1151 10 31 0 0 0
8.5 1.1 3.0 65.0 48.8 5.0
3.3 0.8 1.5 0 0 0
437
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443 Table 3 Sorption capacity of different ion-exchange resins for different electroplating solutions (resin/solution ratio: 1 g/L mL; temperature: 55 °C; sorption time: 1 h). Solution
Bath solution
Resin
PF
PC
Fe
B1 Cr
Fe
Cr
Fe
Cr
Fe
Cr
Group A PK228 A200C
14.2 3.3
30.6 8.7
0.3 1.5
29.2 9.0
13.0 3.8
40.6 11.8
25.9 6.9
27.9 9.3
Group B C100 VPOC1026 S957
2.1 4.4 26.5
42.5 8.6 105.3
0.5 0.3 0.3
43.0 24.9 105.7
2.3 4.2 31.8
45.0 5.0 122.6
0.2 11.0 30.7
45.0 20.7 124.8
Group C P4V DM4195
0.8 7.5
59.5 119.4
0.2 0.1
60.5 128.7
0.7 8.0
60.0 117.0
0 14.2
61.0 133.6
125.5 115.6
0.3 0.3
130.5 115.5
40.6 36.5
130.0 120.3
41.0 35.8
128.5 127.0
Solvent Impregnated Resin D2-E-S957 38.5 D2-H-S957 32.3
B3
On the other hand, for all the tested resins, the sorption capacities of Fe and Cr for the binary chemical solution C1 were about the same with those for bath solution B1, suggesting that the effect of anion impurity was marginal. Similar observation was reported that the breakthrough sorption behavior of S957 was insensitive to the presence of anion entity in passivation bath but it changed with pH value [5]. The Cr/Fe capacity ratios over SAC resins (Group A) were about 3, 100 and 3 in concentrated and diluted Fe containing Cr bath solution (B1 and B3), and C1 solution, respectively. Among the chelating resins, with respect to SAC resins, the iminodiacetic acid functionalized C100 (Group B), 4-vinylpyridine hydrochloride functionalized P4V and bis-2-(pyridylmethyl)amine functional po lystyrene–divinylbenzene (PS-DVB) polymer DM-4195 (Group C) all exhibited higher Cr sorption capacities and Cr/Fe capacity ratios at low Fe sorption selectivity. On the other hand, in comparing to SAC, sulfonated mono-phosphonic acid functionalized Purolite S957 and DEHPA implanted Lewatit VPOC1026 exhibited enhanced Fe capacities. More interestingly, the SIR modified resins exhibited even more Fe sorption capacities. C100 is functionalized with iminodiacetic acid group. It belongs to weak acidic ion-exchange resin. In literature, iminodiacetic acid functionalized chelating resins have been known having stronger chelating power to Cr3+ and In3+ than Fe3+, Cu2+ and Fe2+ [25]. Indeed, C100 had very high Cr/Fe capacity ratio (Table 3). On the other hand, Group C includes pyridine chelating anionic resins. According to Kratz et al. [26], bis-pycolylamine functionalized Lewatit MonoPlus TP220, and possibly also DM4195 with the same functional group, exhibited stronger metal ion affinity of Fe3+ than Cr3+ and much stronger than Co2+, Cu2+ and Ni2+ in tetrahydrofuran solutions. Unfortunately application of this resin is severely limited by lacking of swelling ability in water. Presumably the chelating power of Cr with pyridine in aqueous solution was stronger than Fe, P4V and DM4195 both exhibited strong sorption capacity of Cr [27]. The trend of increasing Cr/Fe ratios of those resins suggest that the anionic chelating resin and iminodiacetic acid functionalized resin all are not applicable for selective sorption of Fe. On the other hand, Levextrel VPOC1026 contains DEHPA functional group highly dispersed in crosslinked polystyrene framework. The phosphoric-acid-ester DEHPA has been widely used as a liquid/liquid-extraction solvent for the extraction of metals from aqueous solutions. Immobilization of DEHPA in VPOC1026 could significantly reduce the separation cost of DEHPA in liquid extraction process. Due to DEHPA’s tendency for hydrolysis and desorption under these conditions, VPOC1026 resin is particularly applicable for low pH fluids with high electrolyte content but not
C1
applicable in neutral and basic media. As shown in Table 3, Fe impurity could be selectively removed from Cr bath solution with low Cr sorption capacity. Nevertheless, its sorption capacity of Fe was very low. The Fe sorption capacity of S957 for high Fe containing Cr bath solution B1 was 26.5 mg/g. For bath solution B3 containing low Fe concentration, Fe impurity could be completely removed using S957 with 0.3 and 105.7 mg/g sorption capacity of Fe and Cr, respectively. While VPOC1026 possessed excellent Fe sorption selectivity but low capacity, S957 possessed high Fe capacity with unsatisfied selectivity. Interestingly, the Fe sorption selectivity of S957 could be improved by impregnation of DEHPA.
3.2. Fe(II) and Fe(III) sorption selectivity A colorimetric experiment shown in Fig. 1 was conducted for studying the selective adsorption of Fe ion using S957 resin. Cu(II), Fe(II) and Fe(III) ion solution exhibited absorbance at 320 nm, 350 nm, 395–450 nm, respectively, showing the color of cyan, weak light yellow and light yellow, respectively. The mixture of Cu(II) solution with Fe(II) and Fe(III) solution was in light cyan and light green, respectively, corresponding to the peak intensity at 330 nm and 330–400 nm. Subjected to S957 adsorption, the absorption peak of the Fe(III)–Cu(II) mixture changed to 330 nm with color change into light blue color, which was about the same color of Cu(II), indicating the selective adsorption of Fe(III) from the
(A)
(B) a Fe(III) b Cu(II) c Fe(III)-Cu(II) d Fe(III)-Cu(II)/S957
a Fe(II) b Cu(II) c Fe(II)-Cu(II) d Fe(II)-Cu(II)/S957
Absorbance (arbitray uuit)
Capacity (mg/g)
Pure component solution
a
c
a
d
b
200 300
c
d b
400 500
600
Wavelength (nm)
200
300
400
500
600
Wavelength (nm)
Fig. 1. Colorimetric test and UV spectra during the sorption of S957 for (A) Fe(II)– Cu(II); and (B) Fe(III)–Cu(II) mixtures (Solution: Cu(NO3)22H2O 5000 ppm, Fe(II)Cl24H2O 8000 ppm, FeCl32H2O 5000 ppm; pH = 0; Solution/S957 = 3/7 v/v).
438
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
1.8
(A)
65% Fe(III) + 35% Fe(II) mixture
1.5 1.2
20,000ppm
78ppm
0.9
Absorbance
0.6
10ppm
0.3 0.0 1.8
(B)
Spectra Gussian model simulation Peak 260 nm Peak 350 nm Peak 450 nm
1.5 1.2 0.9
HFeCl42TBP and HFeCl4C921, respectively [7]. On the other hand, Biswas and Begum reported that the stoichiometric ratio of Fe(III) to DEHPA was in the range of 1.5 and 3.0, depending on the concentrations of Fe(III), H+ and HCl and DEHPA [28]. They found that the stoichiometric ratio increased with decreasing HCl/Cl ratio. Recently Ma et al. proposed a serial structures in which one Fe(III) cation formed complex with five DEHPA ligands in chloride solution but formed with three DEHPA ligands in sulfate solution [29]. A 35% Fe(II)Cl2 – 65% Fe(III)Cl3 mixture with total Fe concentration of 20 g/L was used as the adsorbate solution in a continuous flowing sorption unit loaded with S957 resin. The UV spectra of the effluent collected at different BV (bed volume, equivalent to time-on-stream (TOS) by a factor of resin volume/flow rate ratio) are shown in Fig. 3.
0.6
BV ¼
0.3 78ppm
200
300
400
500
600
Wavelength (nm) Fig. 2. (A) UV spectra of 10–20,000 ppm Fe(II)–Fe(III) mixture solution, Fe(II):Fe(III) = 0.35: 0.65; (B) Experimental and simulation spectrum of 78 ppm Fe(II)-Fe(III) mixture.
mixture solution by S957 (Fig. 1(B)). On the other hand, the color of Fe(II)–Cu(II) mixture after S957 adsorption did not change (Fig. 1(A)), suggesting a minor adsorption of Fe(II) by S957. Similar experiments were conducted for the adsorption using VPOC1027, indicting low sorption capacities for both of Fe(III) an Fe(II) ions. An UV spectra of 35% Fe(II) – 65% Fe(III) mixture solution in the range of 10–20,000 ppm concentration is shown in Fig. 2(A). Fig. 2(B) presents the spectrum of 78 ppm mixture as one example. The spectrum could be de-convoluted into three curves centered at 260, 350 and 450 nm, among which the last two curves belong to Fe(III) spectra and the former two curves are assigned to Fe(II). The appearance of characteristic tailing curve in the right shoulder of the measured spectra at around 450 nm is an indication of the presence of Fe(III). With the proposed UV methodology, a series of calibration curves were constructed for the determination of characteristic parameters for those Fe species (Referring to SI SF2–S4). A calibration curve for Fe(III) in the range of 2000– 20,000 ppm (0.2–2%) was developed in SI SF4. A linear relationship between the peak area of 450 nm peak of UV spectra and Fe(III) concentration was observed. The linearity extended to a range at lowest one down to 2000 ppm (0.2%). Accordingly, the relative composition of Fe(II) and Fe(III) in mixture solution could be semi-quantitatively determined. By applying the UV methodology, bath solutions B1 and B2 were analyzed exhibiting a distribution of Fe(II): Fe(III) at around 35%: 65%. The UV analysis methodology was used to study the chelation stoichiometry of Fe – DEHPA complex. Various HNO3 solutions of FeCl3 (pH = 0.7) at different concentration were subjected to the extraction with hexane solution of 0.15 M DEHPA at room temperature. The number of mole of Fe(III) and DEHPA in the extractant phase was analyzed with UV spectroscopy and NaOH titration, respectively. The stoichiometric molar ratio of DEHPA to Fe(III) was determined as 3.0. On the other hand, from similar experiment using FeCl2 solution, DEHPA showed no extraction power for Fe(II). According to literature, the stoichiometric ratio changes with the type of ligand and the composition of solution. Dessouky et al. reported that tributylphosphate (TBP) and commercial trioctyl phosphine oxide (C921) could extract Fe(III) but not Fe(II) in strong acidic chloride medium in the stoichiometric ratio as
ð1Þ
The UV spectra indicates that at BV shorter than 1.2 BV (TOS of 13 min.), the effluent was merely Fe(II) solution. With extending BV, Fe(II) and Fe(III) were co-eluted through S957; until BV of 7.2 (TOS of 80 min), the effluent showed identical spectra of the adsorbate solution as an indication of saturation of S957 sorption. Accordingly, S957 exhibited sorption selectivity of Fe(III) higher than the sorption of Fe(II). 3.3. Sorption isotherm The sorption isotherms of Fe and Cr on different resins were measured in batch mode. Langmuir isothermal sorption model (Eq. (2)) and Freundlich isothermal sorption model (Eq. (3)) were used for curve fitting the experimental data
qe ¼
K L qm C e ðLangmuir isotherm modelÞ 1 þ K LCe
ð2Þ
where Ce is the liquid phase equilibrium concentration (mg/L); qe is the equilibrium capacity in solid phase (mg/g); qm is the maximum sorption capacity (mg/g); KL is the Langmuir constant (L/mg).
qe ¼ K F C 1=n ðFreundlich isotherm modelÞ e
ð3Þ
where Ce is the liquid phase equilibrium concentration (mg/L); qe is the equilibrium capacity in solid phase (mg/g); n is the Freundlich empirical constant representing sorption strength; and KF is the Freundlich empirical constant (mg11/n L1/n g1) representing relative sorption capacity. Fig. 4 depicts the sorption isotherm of binary chemical solution C1 over S957. The sorption isotherms of Fe and Cr were
1.8
1.2 BV 2.9 BV 5.2 BV 7.2 BV
1.5
Absorbance
0.0
Flow rate ðml=hÞ TOS Resin volume
1.2 0.9 0.6 0.3 0.0 200
300
400
500 600 700
800
Wavelength (nm) Fig. 3. UV Spectra of the effluent from the sorption of Fe(II)Cl2–Fe(III)Cl3 mixture solution through S957 (Fe mixture: total Fe concentration: 20 g/L, 35% Fe(II) + 65% Fe(III); resin: 150 mL S957; flow rate: 13.5 mL/min.).
439
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
22
5
(A)
(B)
18
ln qe (g/L)
Ce/qe (g/L)
20
16 14 12
4 3 2
10 0
400
800
1200
1 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
1600
Ce (mg/L) 100
ln Ce (mg/L) 4
(C)
(D)
3
80
ln qe (g/L)
Ce/qe (g/L)
90
70 60 50
2 1
40 0
400
800 1200 1600 2000 2400
0 4.5 5.0
Ce (mg/L)
5.5
6.0 6.5
7.0
7.5
8.0
ln Ce (mg/L)
Fig. 4. Sorption isotherm of binary chemical solution C1 solution over S957 (A) Fe sorption isotherm fitted with Langmuir model; (B) Fe sorption isotherm fitted with Freundlich model; (C) Cr isother fitted with Langmuir model; and (D) Cr sorption isotherm fitted with Freundlich model (Sorption experimental conditions: 55 °C, 2 h, 0.2 g resin, 10 ml solution, C1 solution concentration: 5–60 mM).
Table 4 Sorption isotherm fitting parameters (Experimental conditions: 55 °C, 2 h, 0.2 g resin, 10 ml solution, C1 solution concentration: 5 60 mM). (A) Langmuir isotherm model Metal LSFe
Fe
Parameter
KL(L/mg)
qm(mg/g)
R
KL(L/mg)
qm(mg/g)
R
6.9 104 2.0 103 2.6 103 1.3 104 5.7 103
145.8 57.8 71.8 8.6 8.0
0.983 0.992 0.992 0.411 0.965
6.7 104 4.3 103 3.8 103 3.3 104 1.0 105
36.6 6.8 9.9 2.8 9.9
0.999 0.983 0.982 0.714 0.415
S957 D2-H-S957 D2-E-S957 D2-E-XAD VPOC1026
4.0 8.5 7.3 – –
Cr
(B) Freundlich isotherm model Metal FSFe
Fe
Parameter
KF (mg11/nL1/n g1)
n
R
KF (mg11/nL1/n g1)
n
R
0.24 0.87 0.79 0.44 1.45
1.25 1.85 1.87 0.81 4.65
0.995 0.991 0.999 0.957 0.889
0.10 0.49 0.51 0.06 0.07
1.40 2.93 2.59 0.66 0.88
0.996 0.914 0.954 0.985 0.989
S957 D2-H-S957 D2-E-957 D2-E-XAD VPOC1026
2.4 1.8 1.5 7.3 20.7
Cr
curve fitted with Langmuir and Freundlich models in Fig. 4(A) and (B), 3(C) and 3(D), respectively. The parameters of the isotherm models for different resins were determined with the best fitting technique giving R value mostly greater than 0.99 (Table 4(A) and (B)). Whereas the sorption isotherms of S957, D2-H-S957 and D2-E-S957 could be fitted well with either Langmuir or Freundlich model, the isotherm of Levextrel VPOC1026 and D2-E-XAD were fitted better with Freundlich model than Langmuir model. Demirbas et al. determined a selectivity factor as the ratio of sorption capacity of different metal ions over Amberlite IR-120 in reference arbitrarily to a particular metal ion, such as Cu(II) [30]. The methodology was widely used. Nevertheless, the selectivity factor the same with Cr/Fe capacity ratio discussed above, changes with metal concentration of the adsorbate solution. Juan and Shao defined another selectivity factor as the capacity ratio of two metals divided by the concentration ratio of a binary adsorbate solution [31]. To deal the concentration effect in a more
representative way, Fe sorption selectivity, namely LSFe and FSFe, in using a particular resin could be defined as the ratio of Langmuir model parameter qm or Freundlich model parameter KF, respectively.
LSFe ¼
qm ðFeÞ ðLangmuir modelÞ qm ðCrÞ
ð4aÞ
FSFe ¼
K F ðFeÞ ðFreundlich modelÞ K F ðCrÞ
ð4bÞ
As shown in Table 4(A), the pristine S957 and the DEHPA modified ones all exhibited LSFe numbers greater than unity as a clear indication of Fe sorption selectivity. Moreover, the LSFe number of DEHPA modified S957 increased. Although its maximum sorption capacity (qm) decreased, the Fe sorption capacity of DEHPA modified S957 increased (Table 3). Presumably, due to the
440
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
complementary sorption selectivity of Fe(III) and Fe(II) respectively from DEHPA and S957, impregnation of DEHPA could induce synergy effect on S957. The R values for curve fitting the sorption isotherms of VPOC1026 and D2-E-XAD using Langmuir model were relatively low (Table 4(A)) the LSFe determined from Langmuir model parameter qm could be erratic. Instead, with improved R numbers greater than 0.9, FSFe values of VPOC1026 and D2-E-XAD were determined from the Freundlich model parameter KF. As shown in Table 4(B), VPOC1026 and D2-E-XAD both had high FSFe values and thus superior Fe sorption selectivity. Since there are fundamental differences in the basic assumptions between Langmuir and Freundlich models, LSFe and FSFe are in different scale. They could not be back-to-back compared. In the scoping type work (Section 3.1), Cr/Fe capacity ratio was evaluated in an empirical approach by which selectivity would be affected dramatically by the experimental conditions during sorption experiment, such as metal concentration, solution/resin ratio, etc. In comparison, the Fe sorption selectivity defined in Eq. (4) is relatively systematic and rigorous by using the parameters determined from isotherm model for the analysis of sorption isotherm. The selectivity either LSFe or FSFe is applicable for a range of metal concentration. It is more useful for the choice and design of resin. Referring to the supplementary information S5, the effect of pH of the metal ion solution on the uptake of metal ion during the sorption test of D2-E-S957 was examined. With increasing pH, the uptakes of most of metal ions such as Cu, Co and Zn increased, the uptakes of Fe and Cr ions decreased accordingly. This phenomenon could be associated with DEHPA’s tendency for hydrolysis and desorption in neutral and basic media. Therefore the DEHPA modified S957 is particularly applicable for low pH fluids.
at 55 °C; the effluent was periodically collected and analyzed for its concentration of Ce. The breakthrough sorption capacity (qb, mg/g) was determined from the breakthrough curves following Eq. (5),
qb ðmg=gÞ ¼
The total breakthrough capacity was determined in a continuous flow sorption unit. An electroplating solution with concentration of Cf flew continuously through a resin sorption chamber kept
(A)
VPOC-B3 Cu Cr Fe
0.4 0.2 0
(D)
1.2
5 10 BV Number
0.6 S957-B3 Cu Cr Fe
0.2 0
5
D2-H-S957-B3 Cu Cr Fe
0.4 0.2 0
20
40 60 80 BV Number
100
D2-E-S957-B3 Cu Cr Fe
0
20
(F)
1.2
40 60 80 100 BV Number
1.0
0.8 0.6 A200C-B3 Fe Cr Cu
0.4 0.2 0.0
0.6
0.0
10 15 BV Number
Ce/Co
Ce/Co
0.6
0.8
0.2
1.0
0.8
(C)
0.4
(E)
1.2
1.0
Ce/Co
0.8
0.0
ð5Þ
1.0
0.4
15
Ce dt 103 1 Cf
1.2
Ce / Co
Ce / Co
Ce / Co
0.6
0.0
0
1.0
0.8
0.0
ts
(B)
1.2
1.0
Z
where Q is the inlet flow rate (mL/min); W is the sorbent weight (g); d is solution density; Cf and Ce were the metal ion concentrations at the inlet and at the exit of the sorption unit, respectively; and ts is the sorption saturation time (min) at which the effluent concentration equals to the inlet concentration. As shown in Fig. 5(A), when sorption unit was loaded with VPOC1026, it exhibited a well described breakthrough curve of Fe in using bath solution B3, reaching breakthrough and saturation at 4.8 and 9.8 BV number, respectively. Breakthrough sorption capacity was then determined according to Eq. (5). Furthermore, the pristine S957 and D2EPHA modified S957 all exhibited similar breakthrough curves among which the later exhibited a higher BV for breakthrough and saturation (Fig. 5(B –D)). Group A (SAC type) resins including A200C and PK228 and other Group B (chelating) resins C100 and TP-207 were also tested for references. As shown in Table 5, all the SAC type resins (A200C, Fig. 5(E), and PK228) and the iminodiacetic acid functionalized chelating resin (C100 and TP207) exhibited fast breakthrough at low BV numbers and low Fe sorption capacities. The SAC resin PK228 and A200C both showed selective Cu and Fe adsorption versus Cr adsorption with low sorption capacity. Their breakthrough capacities were determined from the breakthrough curves according to Eq. (5). On the other hand, TP207 (Fig. 5(F)) showed comparable sorption rate for Cr, Fe and Cu, i.e., those three metal ions conducted competitive and non-selective sorption on TP207. Similar experiment was conducted for the measurement of breakthrough capacities in concentrate bath solution B1. Interestingly, VPOC1026, D2-E-S957 and D2-H-S957 all exhibited enhanced Fe sorption capacity with reduced Cr sorption capacity (Table 5). As a result, Fe sorption selectivities of those three resins were much more enhanced in using bath solution B1 than B3. The phenomena could be realized with competitive sorption mechanism
3.4. Breakthrough capacity
1.2
Qd W
0
2
4 6 8 BV Number
10
0.8 0.6
TP207-B3 Fe Cr Cu
0.4 0.2 0.0
0
1 2 BV Number
3
Fig. 5. Breakthrough curves of (A) VPOC1026, (B) S957, (C) D2-E-S957, (D) D2-H-S957, (E) A200C and (F) TP207 for low Fe containing Cr bath solution B3 (Bath solution containing Fe: 340 mg/L, Cr: 100.7 g/L, Cu: 31 mg/L; resin: 150 mL; flow rate: 10.0 mL/min., 4.0 BV/h).
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443 Table 5 Breakthrough capacity of different resins in bath solution B1 and B3 (resin: 150 mL; flow rate: 10.0 mL/min., 4.0 BV/h). Solution
Bath solution
Capacity (mg/g)
Resin
B1
B3
Fe
Cr
Fe
Cr
Group A A200C PK228
– –
– –
0.5 0.3
35.7 24.7
Group B C100 TP-207 VPOC1026 S957
– – 7.6 50.2
– – 17.9 141.7
0.2 0.2 2.8 1.2
38.0 63.8 46.8 49.1
138.5 102.5
3.2 4.0
229.0 231.1
Solvent impregnated resin D2-H-S957 65.9 D2-E-S957 50.2
441
other hand, S957 exhibited slight decay in sorption capacity of Fe and stable Cr sorption capacity (Fig. 6(B)). As shown in Fig. 6(C), during the period of first 13 cycles, D2-E-S957 exhibited stable Fe sorption capacity (shown as legend Fe) and quantitative Fe desorption load during regeneration treatment (shown as legend de-Fe). The sorption capacity for electroplating bath B2 and desorption load from regeneration were about the same as 25 mg/g. However, Cr desorption load (shown as legend de-Cr) was always only around 30–50% of the sorption capacity (legend Cr). For example, the Cr sorption capacity during 7th cycle was 107 mg/g and desorption weight was only 33 mg/g. Because of the incomplete desorption, Cr sorption capacity declined slightly. After 13 cycles, the Cr and Fe sorption capacities both decreased; Fe sorption capacity dropped to around 80% of initial capacity. The reduction in Fe sorption capacity could be arisen from DEHPA leaching or incomplete desorption.
3.5. Effect of Fe impurity on electro-deposition of Cr arisen from concentration effect. For the low Fe containing solution, Cr concentration was much greater than Fe by a factor of 730, leading to a lower Fe sorption strength than Cr. Among various resins, Fe selectivity decreased in the order of (Group B) VPOC1026 > D2-E-S957 D2-H-S957 > S957 > (Group A) A200C > PK228 > (Group B) C100 > TP-207. The Fe selectivity ranking was consistent with the results of isotherm measurement (Table 5). Nevertheless VPOC1026 had low sorption capacities of Fe and Cr. Both DEHPA modified S957 resins exhibited higher breakthrough capacities with enhanced Fe sorption selectivity than those of S957. The enhanced sorption capacity of S957 with DEHPA modification could be due to Fe(III)–DEHPA complex formation for complementary sorption of Fe on S957. Ortiz et al. reported that Fe(III) bonds strongly to the functional groups of resin and is difficult to be eluted from resin [8]. In the present study, the recyclability of resin was tested by repeatedly measuring the breakthrough sorption capacity cycle-by-cycle. Resin was regenerated at room temperature in between cycles by using 14 vol% HCl solution at 7 BV/h flow rate for 40 min. The VPOC1026 exhibited stable sorption capacity of Fe and Cr (Fig. 6(A)). On the
A continuous flow sorption unit loaded with D2-E-S957 resin of 60 L was constructed. The sorption unit was installed as a side-stream of an industrial electroplating unit for continuously on-line purification of a bath solution B1 with operating time for over two months. The purified bath solution sampled at different operating time was labeled as S1–S5 and subjected for analysis and electro-deposition test. As shown in Table 6, with increasing operating time, Fe content of the purified baths greatly reduced with slightly reduction in Cr content. The quality of electroplating solution particularly the effect of impurity was evaluated in a miniature electroplating unit so-called Hull Cell. While electric current of 7.7 A was generated in bath solution B1 having Fe content of 17.64 g/L, it increased sharply to 8.9 A by reducing Fe content in the purified bath S1 down to 11.75 g/L. Further reduction in Fe content all the way to sample S5, the electric current maintained about the same at 8.9 A with reduced electric potential. The high electric current with low electric potential in the purified bath solution indicated that with decreasing Fe content, electric resistance of the bath solution decreased and current efficiency increased accordingly.
Fig. 6. Recyclability test for Fe and Cr sorption using resin (A) VPOC1026; (B) S957; (C) DEHPA impregnated D2-E-S957 (Bath solution B2 containing Fe: 6.5 g/L, Cr: 50.4 g/L; resin: 90 mL; flow rate: 7.0 BV/h).
442
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
Table 6 Effect of Fe contaminant in electroplating bath on the electroplating performance tested in Hull Cell. Solution ID
B1
S1
S2
S3
S4
S5
Solution compositions Cr (g/L) Fe (g/L)
129.6 17.64
122.7 11.75
119.5 8.82
116.0 5.87
111.2 1.60
109.4 0.05
55 °C 7.7 6.8 4
55 °C 8.9 6.5 4
55 °C 8.9 6.6 4
55 °C 8.9 6.4 4
55 °C 9.0 6.0 4
55 °C 9.0 5.8 4
Operating conditions in Hull Cell test Temperature Current (A) Voltage (V) Time (min) Electroplating products
On the other hand, when Fe concentration in the bath solution was higher than 5.8 g/L (solution S3), clear rainbow strips started appearing on the edges of Cr deposit. As Fe concentration was less than 50 mg/L (solution S5), uniform Cr deposition on Cu platelet was obtained. According to the experimental result, optimum electroplating operation would require a maximum Fe concentration in the electrolyte less than 50 mg/L. Plant test showed that removal of Fe metal ion from the electroplating solution led to electricity saving by 15% and improvement in the product quality of Cr electroplating carbon steel.
of SIR decayed after 13 sorption–desorption cycles, which could be due to partial DEHPA leaching and incomplete desorption of Cr.
4. Conclusion
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2015.06.080.
During the widely applied chromate electro-deposition process, Fe dissolution from steel substrate in the highly acidic electroplating bath is inevitable. In the present study, the maximum Fe contaminant level in bath solution for optimum Cr electroplating operation was determined as 50 mg/L. High Fe contaminant could reduce electric current efficiency with increasing energy consumption, and downgrade the product quality of Cr electro-deposition. Solvent impregnated resin (SIR) with di(2-ethylhexyl) phosphoric acid (DEHPA) for selective sorption of Fe with respect to Cr has been developed for on-line reclaim of Cr electroplating bath solution. Plant test showed that application of SIR could selectively remove Fe metal ion from the electroplating solution, leading to electricity saving by 15%, improvement in the product quality of Cr electroplating carbon steel, extension of service life of Cr electroplating bath solution, and reduction of fresh water usage as well as Cr electroplating solution waste. Semi-quantitative analytical procedure using de-convoluted UV spectra for determination of Fe(II) and Fe(III) distribution in electroplating solution has been proposed. A rigorous definition of Fe sorption selectivity from the qm or KL parameter derived from Langmuir or Freundlich isotherm model was proposed for better representation for a range of metal concentration of real electroplating solution. In reference to strong acidic resins, while the anionic chelating resins such as P4V and DM-4195 exhibited higher Cr sorption capacities, the sulfonated mono-phosphonic acid functionalized Purolite S957 and DEHPA implanted Lewatit VPOC1026 exhibited enhanced capacities of Fe. S957 was found to adsorb Fe(III) more selective than Fe(II). On the other hand, DEHPA extractant is very selective for the sorption of Fe(III) with stoichiometric ratio of 3. By the complementary Fe sorption selectivity of DEHPA to S957, DEHPA impregnated S957 resins exhibited further improved Fe sorption selectivity. Regeneration of SIR modified S957 using HCl solution could quantitatively desorb Fe but only 50% desorb Cr. The Fe sorption capacity
Acknowledgements This study was financially supported by the Ministry of Science and Technology of R.O.C. with Grand Nos. NSC-102-2113-M-390-003-MY3 and NSC-101-2120-S-006-008. Appendix A. Supplementary data
References [1] J. Huang, Y. Cao, Z. Liu, Z. Deng, F. Tang, W. Wang, Efficient removal of heavy metal ions from water system by titanate nanoflowers, Chem. Eng. J. 180 (2012) 75–80. [2] X. Chen, K.F. Lam, K.L. Yeung, Selective removal of chromium from different aqueous systems using magnetic MCM-41 nanosorbents, Chem. Eng. J. 172 (2011) 728–734. [3] E. Maranˇón, F. Suárez, F. Alonso, Y. Fernández, H. Sastre, Preliminary study of iron removal from hydrochloric pickling liquor by ion exchange, Ind. Eng. Chem. Res. 38 (1999) 2782–2786. [4] C. Lasanta, I. Caro, L. Pérez, Theoretical model for ion exchange of iron(III) in chelating resins: application to metal ion removal from wine, Chem. Eng. Sci. 60 (13) (2005) 3477–3486. [5] F.D. Mendes, A.H. Martins, Selective sorption of nickel and cobalt from sulphate solutions using chelating resins, Int. J. Mineral. Process. 74 (1–4) (2004) 359–371. [6] R. Kiefer, A.I. Kalinitchev, W.H. Höll, Column performance of ion exchange resins with aminophosphonate functional groups for elimination of heavy metals, React. Funct. Polym. 67 (12) (2007) 1421–1432. [7] I. Fernández-Olmo, A. Ortiz, A. Urtiaga, I. Ortiz, Selective iron removal from spent passivation baths by ion exchange, J. Chem. Technol. Biotechnol. 12 (2008) 1616–1622. [8] A. Ortiz, I. Fernández-Olmo, A. Urtiaga, I. Ortiz, Modeling of iron removal from spent passivation baths by ion exchange in fixed-bed operation, Ind. Eng. Chem. Res. 48 (2009) 7448–7452. [9] S.I. El Dessouky, Y.A. El-Nadi, I.M. Ahmed, E.A. Saad, J.A. Daoud, Solvent extraction separation of Zn(II), Fe(II), Fe(III) and Cd(II) using tributylphosphate and CYANEX 921 in kerosene from chloride medium, Chem. Eng. Process. 47 (2008) 177–183. [10] F. Principe, G.P. Demopoulos, Comparative study of iron(III) separation from zinc sulphate–sulphuric acid solutions using organophosphorus extractants, OPAP and D2EHPA, Part II. Stripping, Hydrometallurgy 79 (2005) 97–109. [11] N. Kabay, J.L. Cortina, A. Trochimczuk, M. Streat, Solvent-impregnated resins (SIRs) – methods of preparation and their applications, React. Funct. Polym. 70 (8) (2010) 484–496. [12] J.L. Cortina, R. Arad-Yellin, N. Miralles, A.M. Sastre, A. Warshawsky, Kinetics studies on heavy metal ions extraction by Amberlite XAD2 impregnated resins containing a bifunctional organophosphorous extractant, React. Funct. Polym. 38 (1998) 269–278. [13] J.L. Cortina, N. Miralles, A.M. Saslre, M. Asuilar, A. Profumo, M. Pesavento, Solvent-impregnated resins containing di-(2,4,4-trimethylpentyl)phosphinic
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
acid. II: study of the distribution equilibria of Zn(II), Cu(II) and Cd(II), React. Funct. Polym. 21 (1993) 103–116. J.L. Cortina, N. Miralles, M. Aguilar, A.M. Sastre, Extraction studies of Zn(II), Cu(II) and Cd(II) with impregnated and Levextrel resins containing di(2ethylhexyl) phosphoric acid (Lewatit 1026 Oc), Hydrometallurgy 36 (1994) 131–142. I. Villaescusa, V. Salvadó, J. De Pablo, Solid–liquid extraction of Au(III) from aqueous chloride solutions by tri-n-dodecylammonium chloride impregnated in Amberlite XAD-2 resin, React. Funct. Polym. 32 (1997) 125–130. G. Cote, F. Laupretre, C. Chassagnard, Magic-angle carbon-13 NMR investigation of solvent-impregnated resins: application to 7-(4-ethyl-1methyloctyl)-8-quinolinol (Kelex 100) adsorbed on Amberlite XAD-7, React. Polym. Ion Exch. Sorbents 5 (2) (1987) 141–150. B.A. Moyer, G.N. Case, S.D. Alexandratos, A.A. Kriger, Extraction of copper(II) from sulfuric acid by macrocycle-synergized cation exchange: comparing a novel impregnated resin with its solvent-extraction analog, Anal. Chem. 65 (23) (1993) 3389–3395. S. Nii, S. Okumura, T. Kinoshita, Y. Ishigaki, K. Nakano, K. Yamaguchi, S. Akita, Extractant-impregnated organogel for capturing heavy metals from aqueous solutions, Sep. Purif. Technol. 73 (2010) 250–255. H. Hoshi, Y.Z. Wei, M. Kumagai, T. Asakura, Y. Morita, Separation of trivalent actinides from lanthanides by using R-BTP resins and stability of R-BTP resin, J. Alloys Compd. 408–412 (2006) 1274–1277. J.H. Chen, W.R. Chen, Y.Y. Gau, C.H. Lin, The preparation of di(2ethylhexyl)phosphoric acid modified Amberlite 200 and its application in the separation of metal ions from sulfuric acid solution, React. Funct. Polym. 56 (2003) 175–188. J.S. Li, H. Chen, Z.L. Guo, Y.C. Hu, Selective separation of In(III), Ga(III), and Zn(II) from dilute solution using solvent-impregnated resin containing Di(2ethylhexyl) phosphoric acid, J. Appl. Polym. Sci. 100 (2006) 253–259.
443
[22] Z. Redded, C. Gerende, Y. Andres, P. Le Cloriec, Mechanisms of Cr (VI) and Cr (Ill) from aqueous solutions by sugar beet pulp, Environ. Technol. 24 (2003) 257–264. [23] R.S. Juang, S.H. Lee, Column separation of divalent metals from sulfate solutions using impregnated resins containing di(2-ehtylhexyl)phosphoric acid, React. Funct. Polym. 29 (1996) 175–183. [24] C. Namasivayam, K. Ranganathan, Waste Fe(III)/Cr(III) hydroxide as adsorbent for the removal of Cr(VI) from aqueous solution and chromium plating industry wastewater, Environ. Pollut. 82 (1993) 255–261. [25] S.D. Alexandratos, D.W. Crick, Polymer-supported reagents: application to separation science, Ind. Eng. Chem. Res. 35 (3) (1996) 635–644. [26] M.R. Kratz, D.G. Hendricker, Preparation of polymer-bound 2,20 dipyridylamine and some of its transition metal complexes, Polymer 27 (10) (1986) 1641–1643. [27] C.V. Diniz, F.M. Doyle, V.S.T. Ciminelli, Effect of pH on the sorption of selected heavy metal Ions from concentrated chloride solutions by the chelating resin Dowex M-4195, Sep. Sci. Technol. 37 (2002) 3169–3185. [28] R.K. Biswas, D.A. Begum, Kinetics of stripping of Fe3+–D2EHP complexes from D2EHPA–kerosene phase by aqueous HCl–Cl phase using the single drop technique, Hydrometallurgy 60 (2001) 81–97. [29] H.G. Ma, L. Zhang, D.G. Li, Y.H. Sun, Structure of Fe3+ complexes formed with D2EHPA extractant, Environ. Chem. 30 (2) (2011) 538–543. [30] A. Demirbas, E. Pehlivan, F. Gode, T. Altun, G. Arslan, Adsorption of Cu(II), Zn(II), Ni(II), Pb(II), and Cd(II) from aqueous solution on Amberlite IR-120 synthetic resin, J. Colloid Interface Sci. 282 (2005) 20–25. [31] R.S. Juang, H.J. Shao, Effect of pH on competitive adsorption of Cu(II), Ni(II), and Zn(II) from water onto chitosan beads, Adsorption 8 (2002) 71–78.
Contents lists available at ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Selective iron sorption for on-line reclaim of chromate electroplating solution at highly acidic condition Lan-Ping Wei a, Po-Yi Liu a, Kai-Chen Lin a, Je-Ming Yi a, Cyuan-Wun Liou a, Guo-Chuan Yang a, Hsi-Yen Hsu b, Tseng-Chang Tsai a,⇑ a b
Department of Applied Chemistry, National University of Kaohsiung, Kaohsiung 816, Taiwan Div. of Chemical Engineering, Material and Chemical Res. Lab., Industrial Technology Research Institute, HsinChu 300, Taiwan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Maximum Fe contaminant level in
Stable, Regenerable SIR for Selective Fe Removal from Cr Bath
a r t i c l e
i n f o
Article history: Received 21 March 2015 Received in revised form 17 June 2015 Accepted 18 June 2015 Available online 3 July 2015 Keywords: Fe selective sorption DEHPA SIR resin Sorption isotherm Cr electroplating Fe(II)–Fe(III) analysis
Regeneration > 22 cycles
Sorption Capacity (mg/g)
chromate electroplating limited below 50 mg/L. DEHPA impregnated S957 for selective Fe sorption and on-line reclaim of contaminated chromate bath. Complementary sorption selectivity of Fe(III) by DEHPA and S957. Improving electroplating performance with enhancing current efficiency and reducing waste water.
Desorption Efficiency Fe 100% Cr 50%
140 120 100 80
D2-E-S957 Fe de-Fe Cr de-Cr
60 40 20 0 0
5
10
15
http://dx.doi.org/10.1016/j.cej.2015.06.080 1385-8947/Ó 2015 Elsevier B.V. All rights reserved.
30
On-line reclaim process by selective ion sorption for the purification of highly acidic Fe contaminated chromate electroplating bath has been developed. Among many industrial functional resins, Lewatit VPOC1026 implanted with di-(2-ethylhexyl) phosphoric acid (DEHPA) extractant in the polymeric backbone exhibited the highest selectivity of Fe versus Cr but low sorption capacity of Fe. Impregnation of DEHPA on functionalized S957 resin (a sulfonated mono-phosphonic acid resin) exhibited a much enhanced Fe sorption capacity with high Fe selectivity. The enhancement in the adsorption performance of S957 is due to the selective adsorption of DEHPA for Fe(III) complex formation. For the best electroplating performance, the maximum Fe contaminant level in the chromate electroplating bath was determined as 50 mg/L. The developed process could not only maximize electroplating performance but also minimize the usage of original water and the production of waste water. Ó 2015 Elsevier B.V. All rights reserved.
Metal corrosion always leads to massive economic loss in gross natural product. It was accounted for around 4–5%. Traditionally electroplating of secondary metal as a protection layer on the surface of metal substrate could improve its corrosion resistance, mechanical hardness and surface smoothness. Chromium (Cr)
E-mail address: [email protected] (T.-C. Tsai).
25
a b s t r a c t
1. Introduction
⇑ Corresponding author. Fax: +886 7 5919348.
20
Sorption Cycle Number
coating, traditionally the most effective coating, is mostly fabricated by electroplating process or chromate conversion coating (CCC) process using very strong acidic chromate solution (pH < 2). During the electro-deposition of carbon steel or alloy, many metal components of the substrate such as iron (Fe), copper, etc. are apt to leach out into the highly acidic CCC and electroplating bath solutions and accumulate to form contaminants after long service time. Minor contaminants in bath solution particularly Fe could reduce the current efficiency during electro-deposition and downgrade electroplating product quality. To keep the contaminant level as
435
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
low as possible, electroplating industry usually shortens the service life of CCC bath solution with excessive changeover rate. The typical practice not only requires additional fresh water for preparing CCC bath solution but also generates large amount of Cr waste water solution which are harmful to the Earth. In principle, it is technology feasible to selectively remove Fe ion impurity from spent electroplating solutions such as Cr pickling solution, electroplating baths, and passivation baths, etc. Some novel adsorbents for selective adsorption of heavy metals have been reported, such as titanate nanoflower for selective sorption of Cd(II) relative to Zn(II) and Ni(II) [1], magnetic MCM41 for selective adsorption of Cr(VI) in magnetic field [2]. In strong acidic electroplating solution, on one hand, sorption equilibrium could be more favorable for desorption to adsorption; on the other hand, hard metal ion has greater sorption affinity than soft metal ion. As such, selective sorption of Fe(II) and Fe(III) ions (intermediate metal ions) from Cr(VI) and Cr(III) (hard metal ions) electroplating solution is a great technology challenge. Many ion-exchange resins are available nowadays for metal sorption. While many chelating resins have been successfully used for removing heavy metals from wastewaters and spent baths, e.g., Lewatit TP-207 [3,4], Dowex M4195, Amberlite IRC748 [5], and Purolite S-950 [6], only few resins, such as Purolite S957 (sulfonated mono-phosphonic resin), have been reported for selective removal of iron ions from Cr(III) passivation bath [7,8]. Metal extraction using organic extractant has been proposed for metal recovery. For example, for the recovery of zinc from Fe(II)– Fe(III) containing spent pickling solution, organophosphorus extractant tributylphosphate (TBP) [9], octylphenyl acid phosphate (OPAP) and di-(2-ethylhexyl)phosphoric acid (DEHPA) were used [10]. The sorption process could be complementary to extraction process in terms of small equipment size, low energy consumption, sustainable operation in acidic solution, etc. In a recent review paper, Kabay et al. discussed comprehensively the preparation and application of various extractant-containing resins [11]. Two different approaches have been developed for the preparation of extractant-containing solid sorbents. The Levextrel resins were patented using polymerization process, by mixing the extractant with a mixture of monomers (styrene and divinylbenzene), for implanting extractant into polymer backbones. The Levextrel resins available today include extractant TBP, DEHPA and di(2,4,4-trimethylpentyl)phosphinic acid (DTMPPA). The Lewatit VPOC1026 implanted with DEHPA functional group is useful for the extractive sorption of divalent metal ions such as Zn(II), Cu(II) and Cd(II). Other preparation method so-called solvent-impregnated resin (SIR) by impregnation of extractant on various solid sorbents is versatile and flexible in tailoring the types and contents of
extractant, solid sorbent and preparation procedures, etc., providing a wide range of metal separation characteristics. Because of high mobility of ligands, the SIR resin in most of time could perform superior separation property to some widely used ion-exchange resins and chelating resins [12]. Numerous extractants have been used for the preparation of SIR resins, including organophosphorous extractant such as DTMPPA [13], DEHPA [14]; amine group such as tri-n-dodecyl-ammonium chloride [15]; quinolinol/quinone group such as 7-(4-ethyl-methyloc tyl)-8-quinolinol (Kelex 100) [16]; and macrocyclic functional group [17]. Some macroporous resins, dodecyl acrylate – divinylbenzene crosslinked polymeric organogels [18] and silica-based composites [19] have been used for solid sorbent. Various SIR solid sorbents have been reported such as DEHPA modified Amberlite 200 for enhanced sorption of Ni, V and Mo [20], DEHPA modified XAD-2 for extraction of Cu ion [12], DEHPA modified HZ-803 for extraction of In(III), Ga(III) and Zn(II) [21], etc. The present study develops an on-line reclaim process for highly acidic electroplating solution by selective removal of Fe impurity using DEHPA impregnated resins. The SIR modified S957 by DEHPA impregnation exhibited a much enhanced Fe sorption capacity and selectivity, exceeding the DEHPA implanted VPOC1026 resin. The reclaim process could significantly extend the service life of bath solution, reduction of usage of raw materials, fresh water and production of Cr waste water; and also improve the current efficiency during electroplating and electro-deposition product quality. 2. Experimental 2.1. Materials Some sorbents (Table 1) including industrial resins and SIR resins were tested in this study. Sodium hydroxide, nitric and hydrochloric acid, and hexane (Merck) were used for the preparation of the different solutions. DEHPA was provided by Tokyo Chemical Industry. Three industrial electroplating bath solutions provided by local galvanic industry (deployed as B1, B2 and B3 solution) were analyzed by Perkin Elmer model Optical 2100DV Inductively-Coupled Plasma Atomic Emission Spectroscopy (ICP-OES) for the determination of the concentrations of metals (Cr, Fe, and Cu) and ion chromatography for anions (chloride and sulfate). The Cr(III)–Cr(VI) and Fe(II)–Fe(III) partitions in the electroplating solution were determined with UV spectroscopy The methodology for the determination of Fe(II)–Fe(III) will be discussed in Section 3.2. For determining Cr partition, the concentrations of
Table 1 Description of commercial and modified resins. ID
Commercial name
Functional group
Group A–SAC resin A200C PK228
Amberlite 200C Diaion PK228
Sulfonic acid Sulfonic acid
Group B–CKL resin C100 TP207 VPOC1026 S957
Amberlite Chelex 100 Lewatit Mono Plus TP207 Lewatit VPOC1026 Purolite S957
Iminodiacetic acid Iminodiacetate groups DEHPA (0.84 mmol/g) Sulfonated monophosphonic acid
Group C–AKL resin P4V DM4195
2% Poly(4-vinylpyridine hydrochloride) Dowex M4195
4-Vinylpyridine hydrochloride Bis-picolylamine
Solvent impregnated resin D2-E-S957 D2-H-S957 D2-E-XAD
S957 impregnated with ethanol solution of DEHPA (0.82 mmol/g) S957 impregnated with hexane solution of DEHPA (0.62 mmol/g) XAD-2 impregnated with ethanol solution of DEHPA (0.58 mmol/g)
436
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
Cr(VI) and total Cr ion were measured from the absorbances at 540 nm of UV spectra of the pristine electroplating bath and the one pre-treated with ammonium persulfate solution by which Cr(III) was oxidized into Cr(VI) form. The ratio of UV absorbances at 540 nm of the two spectra would give the percentage of Cr(VI) ion in the electroplating bath. For a typical electroplating bath B1, the partitions of metal ions were around 47% Cr(III) – 53% Cr(VI) and 35% Fe(II) – 65% Fe(III). The formation of Cr(III) could be attributed to the reduction of Cr(VI) in highly acidic solution [22].The stock solutions of Fe(III) and Cr(VI) (deployed as C1, PF and PC solution) for simulating the bath solution B1 were prepared by dissolving quantified amounts of chromium oxide (CrO3, Riedel-deHaën) and FeCl3 (Riedel-deHaën) in deionized water and were standardized by ICP-OES. All the electroplating solutions were used without pH adjustment. 2.2. Solvent impregnation process S957 and XAD-2 were used as sorbents subjected for incorporating DEHPA. S957 was first pretreated in 4 M HCl solution of methanol–water mixed solvent for 12 h to remove some possible impurities [23], then immersed in distilled-deionized water for 12 h and at last dried. The dried S957 was stirred in 25 ml of DEHPA containing n-hexane (or ethanol) solution at 25 °C for 6–48 h, and then filtered and dried at 70 °C to form the DEHPA impregnated S957 (denoted as D2-H-S957 or D2-E-S957). Since DEHPA is a weak acid with pKa of 1.4, the DEHPA content in the filtrate of SIR preparation and the ethanol extract solution of D2-E-S957 resin were determined with NaOH titration. The amount of DEHPA impregnated in VPOC1026 was evaluated by washing a proper amount of resin with ethanol. Ethanol could completely elute the ligand. Subsequently the ethanol effluent was titrated by NaOH. The total extractant capacity of the resin was 0.84 mol kg1 dry resin. 2.3. Sorption isotherm of metal ion Single or multicomponent sorption capacities of pure or mixed metal ion solutions were determined in batch operation. During sorption experiment, 1 g of ion-exchange resin was in contact with a stock solution of Fe or Cr concentration in the range of 0.2–2.2 g/L at various solid/liquid ratios for 1 h under constant magnetic stirring at 55 °C, then the solution was centrifuged and the solid was removed and washed into the solution. The sorption test reached equilibrium composition after sorption time of 1 h (Referring to Supplementary Information (SI) SF1). Therefore, most sorption tests were conducted for 1 h. The concentration of metal ion in the filtrate was analyzed by Perkin Elmer model Optical 2100DV ICP-OES. The equilibrium capacity of the resin was calculated after measuring the final (equilibrium) iron concentrations of the solution. 2.4. Breakthrough curves of different resins The breakthrough curves of different resins were measured in a continuous flow sorption system. One ion-exchange resin of 150 ml was loaded into the sorption chamber. Prior to the sorption measurement, the ion-exchange resin was activated initially with HCl solution. An electroplating solution was injected by a mini peristaltic pump (DG Instrument, model C9ES DG-100 N) at a preset flow rate going through the sorption chamber kept at 55 °C and atmospheric pressure for sorption measurement. The effluent at the sorption exit was collected periodically and analyzed with Perkin Elmer model Optical 2100DV ICP-OES. For regeneration treatment, 14 wt% HCl solution of 4 bed volume (BV) was passed through the ion-exchange resin chamber.
2.5. Electroplating The quality of electroplating solution particularly the effect of impurity was evaluated in a miniature electroplating unit so-called Hull Cell (R.O. Hull and Co., Inc.) with scribed solution level line of 267 ml. The Hull Cell 1 in Lucite was a trapezoidal structure in which the cathode using brass plate was placed at an oblique angle with respect to the anode. Electro-deposition was carried out at 55 °C temperature and 5–7 v with variable amperage.
3. Results and discussions 3.1. Resin evaluation for selective Fe removal As shown in Table 2, all the tested electroplating solutions were very acidic with pH value around 0. Their Cr and Fe contents were in the same range of 50–160 gL1 and 0.2 mg L1–22 g L1, respectively. The bath solution B1 had a much higher Fe content than other bath solutions B2 and B3. The equilibrium sorption capacities of ion-exchange resins were evaluated using different electroplating solutions. Four groups of resins were used (Table 1) classified as Group A for strong acidic ion-exchange (SAC) resin, Group B and C for cationic and anionic chelating resin (CKLR and AKLR), and solvent-impregnated resins. As shown in Table 3, most ion exchange resins except for C100 and P4V possessed higher sorption capacity of Fe for pure component solution PF than those in other binary component solutions. Furthermore, all the tested ion-exchange resins exhibited higher Fe sorption capacity for the sorption of B1 solution than the capacities in the sorption of other electroplating solutions. On the other hand, Cr sorption capacities of all the tested resins except for VPOC1026 were about the same in different electroplating solutions In reference to the other term denoted as Fe sorption selectivity to be defined below, a term denoted as Cr/Fe capacity ratio was defined as ratio of the sorption capacity of Cr versus that of Fe. For example, Fe and Cr sorption capacity of S957 was 26.5 and 105.3 mg/g, respectively for B1 solution, and 0.3 and 105.7 mg/g, respectively for B3 solution. Therefore the Cr/Fe capacity ratio over S957 was 4.0 and 352.3 for B1 and B3 solutions, respectively. A strong effect of the concentration of adsorbate solution on the sorption capacity of adsorbent was reported earlier [24]. The exceptional behavior of VPOC1026 could be associated with its superior Fe sorption selectivity. Therefore, with increasing Fe concentration in B1 electroplating solution, Fe sorption capacity of VPOC1026 increased in compensation of Cr sorption capacity. Apparently, in variation with metal concentration of adsorbate and sorption selectivity of adsorbent, Cr was adsorbed competitively with Fe. As a result, Cr/Fe capacity ratio changed with the ratio of Cr/Fe content of adsorbent solution. The ratio could be used for rough evaluation of sorption selectivity of ion exchange resin. Rigorous definition for Fe sorption selectivity based on the model parameter of sorption isotherm by taking account of the composition effect of the solution will be discussed later (Section 3.3).
Table 2 Compositions of electroplating bath solutions and stock solutions. Solution
H
Cr (gL1)
Fe (g L1)
Cu (mg L1)
Cl (g L1)
1 SO2 ) 4 (g L
B1 B2 B3 C1 PF PC
0 0 0 0 1 0
159.6 50.4 146.7 139.5 0 139.5
21.7 6.5 0.2 21.8 21.8 0
1151 10 31 0 0 0
8.5 1.1 3.0 65.0 48.8 5.0
3.3 0.8 1.5 0 0 0
437
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443 Table 3 Sorption capacity of different ion-exchange resins for different electroplating solutions (resin/solution ratio: 1 g/L mL; temperature: 55 °C; sorption time: 1 h). Solution
Bath solution
Resin
PF
PC
Fe
B1 Cr
Fe
Cr
Fe
Cr
Fe
Cr
Group A PK228 A200C
14.2 3.3
30.6 8.7
0.3 1.5
29.2 9.0
13.0 3.8
40.6 11.8
25.9 6.9
27.9 9.3
Group B C100 VPOC1026 S957
2.1 4.4 26.5
42.5 8.6 105.3
0.5 0.3 0.3
43.0 24.9 105.7
2.3 4.2 31.8
45.0 5.0 122.6
0.2 11.0 30.7
45.0 20.7 124.8
Group C P4V DM4195
0.8 7.5
59.5 119.4
0.2 0.1
60.5 128.7
0.7 8.0
60.0 117.0
0 14.2
61.0 133.6
125.5 115.6
0.3 0.3
130.5 115.5
40.6 36.5
130.0 120.3
41.0 35.8
128.5 127.0
Solvent Impregnated Resin D2-E-S957 38.5 D2-H-S957 32.3
B3
On the other hand, for all the tested resins, the sorption capacities of Fe and Cr for the binary chemical solution C1 were about the same with those for bath solution B1, suggesting that the effect of anion impurity was marginal. Similar observation was reported that the breakthrough sorption behavior of S957 was insensitive to the presence of anion entity in passivation bath but it changed with pH value [5]. The Cr/Fe capacity ratios over SAC resins (Group A) were about 3, 100 and 3 in concentrated and diluted Fe containing Cr bath solution (B1 and B3), and C1 solution, respectively. Among the chelating resins, with respect to SAC resins, the iminodiacetic acid functionalized C100 (Group B), 4-vinylpyridine hydrochloride functionalized P4V and bis-2-(pyridylmethyl)amine functional po lystyrene–divinylbenzene (PS-DVB) polymer DM-4195 (Group C) all exhibited higher Cr sorption capacities and Cr/Fe capacity ratios at low Fe sorption selectivity. On the other hand, in comparing to SAC, sulfonated mono-phosphonic acid functionalized Purolite S957 and DEHPA implanted Lewatit VPOC1026 exhibited enhanced Fe capacities. More interestingly, the SIR modified resins exhibited even more Fe sorption capacities. C100 is functionalized with iminodiacetic acid group. It belongs to weak acidic ion-exchange resin. In literature, iminodiacetic acid functionalized chelating resins have been known having stronger chelating power to Cr3+ and In3+ than Fe3+, Cu2+ and Fe2+ [25]. Indeed, C100 had very high Cr/Fe capacity ratio (Table 3). On the other hand, Group C includes pyridine chelating anionic resins. According to Kratz et al. [26], bis-pycolylamine functionalized Lewatit MonoPlus TP220, and possibly also DM4195 with the same functional group, exhibited stronger metal ion affinity of Fe3+ than Cr3+ and much stronger than Co2+, Cu2+ and Ni2+ in tetrahydrofuran solutions. Unfortunately application of this resin is severely limited by lacking of swelling ability in water. Presumably the chelating power of Cr with pyridine in aqueous solution was stronger than Fe, P4V and DM4195 both exhibited strong sorption capacity of Cr [27]. The trend of increasing Cr/Fe ratios of those resins suggest that the anionic chelating resin and iminodiacetic acid functionalized resin all are not applicable for selective sorption of Fe. On the other hand, Levextrel VPOC1026 contains DEHPA functional group highly dispersed in crosslinked polystyrene framework. The phosphoric-acid-ester DEHPA has been widely used as a liquid/liquid-extraction solvent for the extraction of metals from aqueous solutions. Immobilization of DEHPA in VPOC1026 could significantly reduce the separation cost of DEHPA in liquid extraction process. Due to DEHPA’s tendency for hydrolysis and desorption under these conditions, VPOC1026 resin is particularly applicable for low pH fluids with high electrolyte content but not
C1
applicable in neutral and basic media. As shown in Table 3, Fe impurity could be selectively removed from Cr bath solution with low Cr sorption capacity. Nevertheless, its sorption capacity of Fe was very low. The Fe sorption capacity of S957 for high Fe containing Cr bath solution B1 was 26.5 mg/g. For bath solution B3 containing low Fe concentration, Fe impurity could be completely removed using S957 with 0.3 and 105.7 mg/g sorption capacity of Fe and Cr, respectively. While VPOC1026 possessed excellent Fe sorption selectivity but low capacity, S957 possessed high Fe capacity with unsatisfied selectivity. Interestingly, the Fe sorption selectivity of S957 could be improved by impregnation of DEHPA.
3.2. Fe(II) and Fe(III) sorption selectivity A colorimetric experiment shown in Fig. 1 was conducted for studying the selective adsorption of Fe ion using S957 resin. Cu(II), Fe(II) and Fe(III) ion solution exhibited absorbance at 320 nm, 350 nm, 395–450 nm, respectively, showing the color of cyan, weak light yellow and light yellow, respectively. The mixture of Cu(II) solution with Fe(II) and Fe(III) solution was in light cyan and light green, respectively, corresponding to the peak intensity at 330 nm and 330–400 nm. Subjected to S957 adsorption, the absorption peak of the Fe(III)–Cu(II) mixture changed to 330 nm with color change into light blue color, which was about the same color of Cu(II), indicating the selective adsorption of Fe(III) from the
(A)
(B) a Fe(III) b Cu(II) c Fe(III)-Cu(II) d Fe(III)-Cu(II)/S957
a Fe(II) b Cu(II) c Fe(II)-Cu(II) d Fe(II)-Cu(II)/S957
Absorbance (arbitray uuit)
Capacity (mg/g)
Pure component solution
a
c
a
d
b
200 300
c
d b
400 500
600
Wavelength (nm)
200
300
400
500
600
Wavelength (nm)
Fig. 1. Colorimetric test and UV spectra during the sorption of S957 for (A) Fe(II)– Cu(II); and (B) Fe(III)–Cu(II) mixtures (Solution: Cu(NO3)22H2O 5000 ppm, Fe(II)Cl24H2O 8000 ppm, FeCl32H2O 5000 ppm; pH = 0; Solution/S957 = 3/7 v/v).
438
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
1.8
(A)
65% Fe(III) + 35% Fe(II) mixture
1.5 1.2
20,000ppm
78ppm
0.9
Absorbance
0.6
10ppm
0.3 0.0 1.8
(B)
Spectra Gussian model simulation Peak 260 nm Peak 350 nm Peak 450 nm
1.5 1.2 0.9
HFeCl42TBP and HFeCl4C921, respectively [7]. On the other hand, Biswas and Begum reported that the stoichiometric ratio of Fe(III) to DEHPA was in the range of 1.5 and 3.0, depending on the concentrations of Fe(III), H+ and HCl and DEHPA [28]. They found that the stoichiometric ratio increased with decreasing HCl/Cl ratio. Recently Ma et al. proposed a serial structures in which one Fe(III) cation formed complex with five DEHPA ligands in chloride solution but formed with three DEHPA ligands in sulfate solution [29]. A 35% Fe(II)Cl2 – 65% Fe(III)Cl3 mixture with total Fe concentration of 20 g/L was used as the adsorbate solution in a continuous flowing sorption unit loaded with S957 resin. The UV spectra of the effluent collected at different BV (bed volume, equivalent to time-on-stream (TOS) by a factor of resin volume/flow rate ratio) are shown in Fig. 3.
0.6
BV ¼
0.3 78ppm
200
300
400
500
600
Wavelength (nm) Fig. 2. (A) UV spectra of 10–20,000 ppm Fe(II)–Fe(III) mixture solution, Fe(II):Fe(III) = 0.35: 0.65; (B) Experimental and simulation spectrum of 78 ppm Fe(II)-Fe(III) mixture.
mixture solution by S957 (Fig. 1(B)). On the other hand, the color of Fe(II)–Cu(II) mixture after S957 adsorption did not change (Fig. 1(A)), suggesting a minor adsorption of Fe(II) by S957. Similar experiments were conducted for the adsorption using VPOC1027, indicting low sorption capacities for both of Fe(III) an Fe(II) ions. An UV spectra of 35% Fe(II) – 65% Fe(III) mixture solution in the range of 10–20,000 ppm concentration is shown in Fig. 2(A). Fig. 2(B) presents the spectrum of 78 ppm mixture as one example. The spectrum could be de-convoluted into three curves centered at 260, 350 and 450 nm, among which the last two curves belong to Fe(III) spectra and the former two curves are assigned to Fe(II). The appearance of characteristic tailing curve in the right shoulder of the measured spectra at around 450 nm is an indication of the presence of Fe(III). With the proposed UV methodology, a series of calibration curves were constructed for the determination of characteristic parameters for those Fe species (Referring to SI SF2–S4). A calibration curve for Fe(III) in the range of 2000– 20,000 ppm (0.2–2%) was developed in SI SF4. A linear relationship between the peak area of 450 nm peak of UV spectra and Fe(III) concentration was observed. The linearity extended to a range at lowest one down to 2000 ppm (0.2%). Accordingly, the relative composition of Fe(II) and Fe(III) in mixture solution could be semi-quantitatively determined. By applying the UV methodology, bath solutions B1 and B2 were analyzed exhibiting a distribution of Fe(II): Fe(III) at around 35%: 65%. The UV analysis methodology was used to study the chelation stoichiometry of Fe – DEHPA complex. Various HNO3 solutions of FeCl3 (pH = 0.7) at different concentration were subjected to the extraction with hexane solution of 0.15 M DEHPA at room temperature. The number of mole of Fe(III) and DEHPA in the extractant phase was analyzed with UV spectroscopy and NaOH titration, respectively. The stoichiometric molar ratio of DEHPA to Fe(III) was determined as 3.0. On the other hand, from similar experiment using FeCl2 solution, DEHPA showed no extraction power for Fe(II). According to literature, the stoichiometric ratio changes with the type of ligand and the composition of solution. Dessouky et al. reported that tributylphosphate (TBP) and commercial trioctyl phosphine oxide (C921) could extract Fe(III) but not Fe(II) in strong acidic chloride medium in the stoichiometric ratio as
ð1Þ
The UV spectra indicates that at BV shorter than 1.2 BV (TOS of 13 min.), the effluent was merely Fe(II) solution. With extending BV, Fe(II) and Fe(III) were co-eluted through S957; until BV of 7.2 (TOS of 80 min), the effluent showed identical spectra of the adsorbate solution as an indication of saturation of S957 sorption. Accordingly, S957 exhibited sorption selectivity of Fe(III) higher than the sorption of Fe(II). 3.3. Sorption isotherm The sorption isotherms of Fe and Cr on different resins were measured in batch mode. Langmuir isothermal sorption model (Eq. (2)) and Freundlich isothermal sorption model (Eq. (3)) were used for curve fitting the experimental data
qe ¼
K L qm C e ðLangmuir isotherm modelÞ 1 þ K LCe
ð2Þ
where Ce is the liquid phase equilibrium concentration (mg/L); qe is the equilibrium capacity in solid phase (mg/g); qm is the maximum sorption capacity (mg/g); KL is the Langmuir constant (L/mg).
qe ¼ K F C 1=n ðFreundlich isotherm modelÞ e
ð3Þ
where Ce is the liquid phase equilibrium concentration (mg/L); qe is the equilibrium capacity in solid phase (mg/g); n is the Freundlich empirical constant representing sorption strength; and KF is the Freundlich empirical constant (mg11/n L1/n g1) representing relative sorption capacity. Fig. 4 depicts the sorption isotherm of binary chemical solution C1 over S957. The sorption isotherms of Fe and Cr were
1.8
1.2 BV 2.9 BV 5.2 BV 7.2 BV
1.5
Absorbance
0.0
Flow rate ðml=hÞ TOS Resin volume
1.2 0.9 0.6 0.3 0.0 200
300
400
500 600 700
800
Wavelength (nm) Fig. 3. UV Spectra of the effluent from the sorption of Fe(II)Cl2–Fe(III)Cl3 mixture solution through S957 (Fe mixture: total Fe concentration: 20 g/L, 35% Fe(II) + 65% Fe(III); resin: 150 mL S957; flow rate: 13.5 mL/min.).
439
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
22
5
(A)
(B)
18
ln qe (g/L)
Ce/qe (g/L)
20
16 14 12
4 3 2
10 0
400
800
1200
1 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
1600
Ce (mg/L) 100
ln Ce (mg/L) 4
(C)
(D)
3
80
ln qe (g/L)
Ce/qe (g/L)
90
70 60 50
2 1
40 0
400
800 1200 1600 2000 2400
0 4.5 5.0
Ce (mg/L)
5.5
6.0 6.5
7.0
7.5
8.0
ln Ce (mg/L)
Fig. 4. Sorption isotherm of binary chemical solution C1 solution over S957 (A) Fe sorption isotherm fitted with Langmuir model; (B) Fe sorption isotherm fitted with Freundlich model; (C) Cr isother fitted with Langmuir model; and (D) Cr sorption isotherm fitted with Freundlich model (Sorption experimental conditions: 55 °C, 2 h, 0.2 g resin, 10 ml solution, C1 solution concentration: 5–60 mM).
Table 4 Sorption isotherm fitting parameters (Experimental conditions: 55 °C, 2 h, 0.2 g resin, 10 ml solution, C1 solution concentration: 5 60 mM). (A) Langmuir isotherm model Metal LSFe
Fe
Parameter
KL(L/mg)
qm(mg/g)
R
KL(L/mg)
qm(mg/g)
R
6.9 104 2.0 103 2.6 103 1.3 104 5.7 103
145.8 57.8 71.8 8.6 8.0
0.983 0.992 0.992 0.411 0.965
6.7 104 4.3 103 3.8 103 3.3 104 1.0 105
36.6 6.8 9.9 2.8 9.9
0.999 0.983 0.982 0.714 0.415
S957 D2-H-S957 D2-E-S957 D2-E-XAD VPOC1026
4.0 8.5 7.3 – –
Cr
(B) Freundlich isotherm model Metal FSFe
Fe
Parameter
KF (mg11/nL1/n g1)
n
R
KF (mg11/nL1/n g1)
n
R
0.24 0.87 0.79 0.44 1.45
1.25 1.85 1.87 0.81 4.65
0.995 0.991 0.999 0.957 0.889
0.10 0.49 0.51 0.06 0.07
1.40 2.93 2.59 0.66 0.88
0.996 0.914 0.954 0.985 0.989
S957 D2-H-S957 D2-E-957 D2-E-XAD VPOC1026
2.4 1.8 1.5 7.3 20.7
Cr
curve fitted with Langmuir and Freundlich models in Fig. 4(A) and (B), 3(C) and 3(D), respectively. The parameters of the isotherm models for different resins were determined with the best fitting technique giving R value mostly greater than 0.99 (Table 4(A) and (B)). Whereas the sorption isotherms of S957, D2-H-S957 and D2-E-S957 could be fitted well with either Langmuir or Freundlich model, the isotherm of Levextrel VPOC1026 and D2-E-XAD were fitted better with Freundlich model than Langmuir model. Demirbas et al. determined a selectivity factor as the ratio of sorption capacity of different metal ions over Amberlite IR-120 in reference arbitrarily to a particular metal ion, such as Cu(II) [30]. The methodology was widely used. Nevertheless, the selectivity factor the same with Cr/Fe capacity ratio discussed above, changes with metal concentration of the adsorbate solution. Juan and Shao defined another selectivity factor as the capacity ratio of two metals divided by the concentration ratio of a binary adsorbate solution [31]. To deal the concentration effect in a more
representative way, Fe sorption selectivity, namely LSFe and FSFe, in using a particular resin could be defined as the ratio of Langmuir model parameter qm or Freundlich model parameter KF, respectively.
LSFe ¼
qm ðFeÞ ðLangmuir modelÞ qm ðCrÞ
ð4aÞ
FSFe ¼
K F ðFeÞ ðFreundlich modelÞ K F ðCrÞ
ð4bÞ
As shown in Table 4(A), the pristine S957 and the DEHPA modified ones all exhibited LSFe numbers greater than unity as a clear indication of Fe sorption selectivity. Moreover, the LSFe number of DEHPA modified S957 increased. Although its maximum sorption capacity (qm) decreased, the Fe sorption capacity of DEHPA modified S957 increased (Table 3). Presumably, due to the
440
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
complementary sorption selectivity of Fe(III) and Fe(II) respectively from DEHPA and S957, impregnation of DEHPA could induce synergy effect on S957. The R values for curve fitting the sorption isotherms of VPOC1026 and D2-E-XAD using Langmuir model were relatively low (Table 4(A)) the LSFe determined from Langmuir model parameter qm could be erratic. Instead, with improved R numbers greater than 0.9, FSFe values of VPOC1026 and D2-E-XAD were determined from the Freundlich model parameter KF. As shown in Table 4(B), VPOC1026 and D2-E-XAD both had high FSFe values and thus superior Fe sorption selectivity. Since there are fundamental differences in the basic assumptions between Langmuir and Freundlich models, LSFe and FSFe are in different scale. They could not be back-to-back compared. In the scoping type work (Section 3.1), Cr/Fe capacity ratio was evaluated in an empirical approach by which selectivity would be affected dramatically by the experimental conditions during sorption experiment, such as metal concentration, solution/resin ratio, etc. In comparison, the Fe sorption selectivity defined in Eq. (4) is relatively systematic and rigorous by using the parameters determined from isotherm model for the analysis of sorption isotherm. The selectivity either LSFe or FSFe is applicable for a range of metal concentration. It is more useful for the choice and design of resin. Referring to the supplementary information S5, the effect of pH of the metal ion solution on the uptake of metal ion during the sorption test of D2-E-S957 was examined. With increasing pH, the uptakes of most of metal ions such as Cu, Co and Zn increased, the uptakes of Fe and Cr ions decreased accordingly. This phenomenon could be associated with DEHPA’s tendency for hydrolysis and desorption in neutral and basic media. Therefore the DEHPA modified S957 is particularly applicable for low pH fluids.
at 55 °C; the effluent was periodically collected and analyzed for its concentration of Ce. The breakthrough sorption capacity (qb, mg/g) was determined from the breakthrough curves following Eq. (5),
qb ðmg=gÞ ¼
The total breakthrough capacity was determined in a continuous flow sorption unit. An electroplating solution with concentration of Cf flew continuously through a resin sorption chamber kept
(A)
VPOC-B3 Cu Cr Fe
0.4 0.2 0
(D)
1.2
5 10 BV Number
0.6 S957-B3 Cu Cr Fe
0.2 0
5
D2-H-S957-B3 Cu Cr Fe
0.4 0.2 0
20
40 60 80 BV Number
100
D2-E-S957-B3 Cu Cr Fe
0
20
(F)
1.2
40 60 80 100 BV Number
1.0
0.8 0.6 A200C-B3 Fe Cr Cu
0.4 0.2 0.0
0.6
0.0
10 15 BV Number
Ce/Co
Ce/Co
0.6
0.8
0.2
1.0
0.8
(C)
0.4
(E)
1.2
1.0
Ce/Co
0.8
0.0
ð5Þ
1.0
0.4
15
Ce dt 103 1 Cf
1.2
Ce / Co
Ce / Co
Ce / Co
0.6
0.0
0
1.0
0.8
0.0
ts
(B)
1.2
1.0
Z
where Q is the inlet flow rate (mL/min); W is the sorbent weight (g); d is solution density; Cf and Ce were the metal ion concentrations at the inlet and at the exit of the sorption unit, respectively; and ts is the sorption saturation time (min) at which the effluent concentration equals to the inlet concentration. As shown in Fig. 5(A), when sorption unit was loaded with VPOC1026, it exhibited a well described breakthrough curve of Fe in using bath solution B3, reaching breakthrough and saturation at 4.8 and 9.8 BV number, respectively. Breakthrough sorption capacity was then determined according to Eq. (5). Furthermore, the pristine S957 and D2EPHA modified S957 all exhibited similar breakthrough curves among which the later exhibited a higher BV for breakthrough and saturation (Fig. 5(B –D)). Group A (SAC type) resins including A200C and PK228 and other Group B (chelating) resins C100 and TP-207 were also tested for references. As shown in Table 5, all the SAC type resins (A200C, Fig. 5(E), and PK228) and the iminodiacetic acid functionalized chelating resin (C100 and TP207) exhibited fast breakthrough at low BV numbers and low Fe sorption capacities. The SAC resin PK228 and A200C both showed selective Cu and Fe adsorption versus Cr adsorption with low sorption capacity. Their breakthrough capacities were determined from the breakthrough curves according to Eq. (5). On the other hand, TP207 (Fig. 5(F)) showed comparable sorption rate for Cr, Fe and Cu, i.e., those three metal ions conducted competitive and non-selective sorption on TP207. Similar experiment was conducted for the measurement of breakthrough capacities in concentrate bath solution B1. Interestingly, VPOC1026, D2-E-S957 and D2-H-S957 all exhibited enhanced Fe sorption capacity with reduced Cr sorption capacity (Table 5). As a result, Fe sorption selectivities of those three resins were much more enhanced in using bath solution B1 than B3. The phenomena could be realized with competitive sorption mechanism
3.4. Breakthrough capacity
1.2
Qd W
0
2
4 6 8 BV Number
10
0.8 0.6
TP207-B3 Fe Cr Cu
0.4 0.2 0.0
0
1 2 BV Number
3
Fig. 5. Breakthrough curves of (A) VPOC1026, (B) S957, (C) D2-E-S957, (D) D2-H-S957, (E) A200C and (F) TP207 for low Fe containing Cr bath solution B3 (Bath solution containing Fe: 340 mg/L, Cr: 100.7 g/L, Cu: 31 mg/L; resin: 150 mL; flow rate: 10.0 mL/min., 4.0 BV/h).
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443 Table 5 Breakthrough capacity of different resins in bath solution B1 and B3 (resin: 150 mL; flow rate: 10.0 mL/min., 4.0 BV/h). Solution
Bath solution
Capacity (mg/g)
Resin
B1
B3
Fe
Cr
Fe
Cr
Group A A200C PK228
– –
– –
0.5 0.3
35.7 24.7
Group B C100 TP-207 VPOC1026 S957
– – 7.6 50.2
– – 17.9 141.7
0.2 0.2 2.8 1.2
38.0 63.8 46.8 49.1
138.5 102.5
3.2 4.0
229.0 231.1
Solvent impregnated resin D2-H-S957 65.9 D2-E-S957 50.2
441
other hand, S957 exhibited slight decay in sorption capacity of Fe and stable Cr sorption capacity (Fig. 6(B)). As shown in Fig. 6(C), during the period of first 13 cycles, D2-E-S957 exhibited stable Fe sorption capacity (shown as legend Fe) and quantitative Fe desorption load during regeneration treatment (shown as legend de-Fe). The sorption capacity for electroplating bath B2 and desorption load from regeneration were about the same as 25 mg/g. However, Cr desorption load (shown as legend de-Cr) was always only around 30–50% of the sorption capacity (legend Cr). For example, the Cr sorption capacity during 7th cycle was 107 mg/g and desorption weight was only 33 mg/g. Because of the incomplete desorption, Cr sorption capacity declined slightly. After 13 cycles, the Cr and Fe sorption capacities both decreased; Fe sorption capacity dropped to around 80% of initial capacity. The reduction in Fe sorption capacity could be arisen from DEHPA leaching or incomplete desorption.
3.5. Effect of Fe impurity on electro-deposition of Cr arisen from concentration effect. For the low Fe containing solution, Cr concentration was much greater than Fe by a factor of 730, leading to a lower Fe sorption strength than Cr. Among various resins, Fe selectivity decreased in the order of (Group B) VPOC1026 > D2-E-S957 D2-H-S957 > S957 > (Group A) A200C > PK228 > (Group B) C100 > TP-207. The Fe selectivity ranking was consistent with the results of isotherm measurement (Table 5). Nevertheless VPOC1026 had low sorption capacities of Fe and Cr. Both DEHPA modified S957 resins exhibited higher breakthrough capacities with enhanced Fe sorption selectivity than those of S957. The enhanced sorption capacity of S957 with DEHPA modification could be due to Fe(III)–DEHPA complex formation for complementary sorption of Fe on S957. Ortiz et al. reported that Fe(III) bonds strongly to the functional groups of resin and is difficult to be eluted from resin [8]. In the present study, the recyclability of resin was tested by repeatedly measuring the breakthrough sorption capacity cycle-by-cycle. Resin was regenerated at room temperature in between cycles by using 14 vol% HCl solution at 7 BV/h flow rate for 40 min. The VPOC1026 exhibited stable sorption capacity of Fe and Cr (Fig. 6(A)). On the
A continuous flow sorption unit loaded with D2-E-S957 resin of 60 L was constructed. The sorption unit was installed as a side-stream of an industrial electroplating unit for continuously on-line purification of a bath solution B1 with operating time for over two months. The purified bath solution sampled at different operating time was labeled as S1–S5 and subjected for analysis and electro-deposition test. As shown in Table 6, with increasing operating time, Fe content of the purified baths greatly reduced with slightly reduction in Cr content. The quality of electroplating solution particularly the effect of impurity was evaluated in a miniature electroplating unit so-called Hull Cell. While electric current of 7.7 A was generated in bath solution B1 having Fe content of 17.64 g/L, it increased sharply to 8.9 A by reducing Fe content in the purified bath S1 down to 11.75 g/L. Further reduction in Fe content all the way to sample S5, the electric current maintained about the same at 8.9 A with reduced electric potential. The high electric current with low electric potential in the purified bath solution indicated that with decreasing Fe content, electric resistance of the bath solution decreased and current efficiency increased accordingly.
Fig. 6. Recyclability test for Fe and Cr sorption using resin (A) VPOC1026; (B) S957; (C) DEHPA impregnated D2-E-S957 (Bath solution B2 containing Fe: 6.5 g/L, Cr: 50.4 g/L; resin: 90 mL; flow rate: 7.0 BV/h).
442
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
Table 6 Effect of Fe contaminant in electroplating bath on the electroplating performance tested in Hull Cell. Solution ID
B1
S1
S2
S3
S4
S5
Solution compositions Cr (g/L) Fe (g/L)
129.6 17.64
122.7 11.75
119.5 8.82
116.0 5.87
111.2 1.60
109.4 0.05
55 °C 7.7 6.8 4
55 °C 8.9 6.5 4
55 °C 8.9 6.6 4
55 °C 8.9 6.4 4
55 °C 9.0 6.0 4
55 °C 9.0 5.8 4
Operating conditions in Hull Cell test Temperature Current (A) Voltage (V) Time (min) Electroplating products
On the other hand, when Fe concentration in the bath solution was higher than 5.8 g/L (solution S3), clear rainbow strips started appearing on the edges of Cr deposit. As Fe concentration was less than 50 mg/L (solution S5), uniform Cr deposition on Cu platelet was obtained. According to the experimental result, optimum electroplating operation would require a maximum Fe concentration in the electrolyte less than 50 mg/L. Plant test showed that removal of Fe metal ion from the electroplating solution led to electricity saving by 15% and improvement in the product quality of Cr electroplating carbon steel.
of SIR decayed after 13 sorption–desorption cycles, which could be due to partial DEHPA leaching and incomplete desorption of Cr.
4. Conclusion
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2015.06.080.
During the widely applied chromate electro-deposition process, Fe dissolution from steel substrate in the highly acidic electroplating bath is inevitable. In the present study, the maximum Fe contaminant level in bath solution for optimum Cr electroplating operation was determined as 50 mg/L. High Fe contaminant could reduce electric current efficiency with increasing energy consumption, and downgrade the product quality of Cr electro-deposition. Solvent impregnated resin (SIR) with di(2-ethylhexyl) phosphoric acid (DEHPA) for selective sorption of Fe with respect to Cr has been developed for on-line reclaim of Cr electroplating bath solution. Plant test showed that application of SIR could selectively remove Fe metal ion from the electroplating solution, leading to electricity saving by 15%, improvement in the product quality of Cr electroplating carbon steel, extension of service life of Cr electroplating bath solution, and reduction of fresh water usage as well as Cr electroplating solution waste. Semi-quantitative analytical procedure using de-convoluted UV spectra for determination of Fe(II) and Fe(III) distribution in electroplating solution has been proposed. A rigorous definition of Fe sorption selectivity from the qm or KL parameter derived from Langmuir or Freundlich isotherm model was proposed for better representation for a range of metal concentration of real electroplating solution. In reference to strong acidic resins, while the anionic chelating resins such as P4V and DM-4195 exhibited higher Cr sorption capacities, the sulfonated mono-phosphonic acid functionalized Purolite S957 and DEHPA implanted Lewatit VPOC1026 exhibited enhanced capacities of Fe. S957 was found to adsorb Fe(III) more selective than Fe(II). On the other hand, DEHPA extractant is very selective for the sorption of Fe(III) with stoichiometric ratio of 3. By the complementary Fe sorption selectivity of DEHPA to S957, DEHPA impregnated S957 resins exhibited further improved Fe sorption selectivity. Regeneration of SIR modified S957 using HCl solution could quantitatively desorb Fe but only 50% desorb Cr. The Fe sorption capacity
Acknowledgements This study was financially supported by the Ministry of Science and Technology of R.O.C. with Grand Nos. NSC-102-2113-M-390-003-MY3 and NSC-101-2120-S-006-008. Appendix A. Supplementary data
References [1] J. Huang, Y. Cao, Z. Liu, Z. Deng, F. Tang, W. Wang, Efficient removal of heavy metal ions from water system by titanate nanoflowers, Chem. Eng. J. 180 (2012) 75–80. [2] X. Chen, K.F. Lam, K.L. Yeung, Selective removal of chromium from different aqueous systems using magnetic MCM-41 nanosorbents, Chem. Eng. J. 172 (2011) 728–734. [3] E. Maranˇón, F. Suárez, F. Alonso, Y. Fernández, H. Sastre, Preliminary study of iron removal from hydrochloric pickling liquor by ion exchange, Ind. Eng. Chem. Res. 38 (1999) 2782–2786. [4] C. Lasanta, I. Caro, L. Pérez, Theoretical model for ion exchange of iron(III) in chelating resins: application to metal ion removal from wine, Chem. Eng. Sci. 60 (13) (2005) 3477–3486. [5] F.D. Mendes, A.H. Martins, Selective sorption of nickel and cobalt from sulphate solutions using chelating resins, Int. J. Mineral. Process. 74 (1–4) (2004) 359–371. [6] R. Kiefer, A.I. Kalinitchev, W.H. Höll, Column performance of ion exchange resins with aminophosphonate functional groups for elimination of heavy metals, React. Funct. Polym. 67 (12) (2007) 1421–1432. [7] I. Fernández-Olmo, A. Ortiz, A. Urtiaga, I. Ortiz, Selective iron removal from spent passivation baths by ion exchange, J. Chem. Technol. Biotechnol. 12 (2008) 1616–1622. [8] A. Ortiz, I. Fernández-Olmo, A. Urtiaga, I. Ortiz, Modeling of iron removal from spent passivation baths by ion exchange in fixed-bed operation, Ind. Eng. Chem. Res. 48 (2009) 7448–7452. [9] S.I. El Dessouky, Y.A. El-Nadi, I.M. Ahmed, E.A. Saad, J.A. Daoud, Solvent extraction separation of Zn(II), Fe(II), Fe(III) and Cd(II) using tributylphosphate and CYANEX 921 in kerosene from chloride medium, Chem. Eng. Process. 47 (2008) 177–183. [10] F. Principe, G.P. Demopoulos, Comparative study of iron(III) separation from zinc sulphate–sulphuric acid solutions using organophosphorus extractants, OPAP and D2EHPA, Part II. Stripping, Hydrometallurgy 79 (2005) 97–109. [11] N. Kabay, J.L. Cortina, A. Trochimczuk, M. Streat, Solvent-impregnated resins (SIRs) – methods of preparation and their applications, React. Funct. Polym. 70 (8) (2010) 484–496. [12] J.L. Cortina, R. Arad-Yellin, N. Miralles, A.M. Sastre, A. Warshawsky, Kinetics studies on heavy metal ions extraction by Amberlite XAD2 impregnated resins containing a bifunctional organophosphorous extractant, React. Funct. Polym. 38 (1998) 269–278. [13] J.L. Cortina, N. Miralles, A.M. Saslre, M. Asuilar, A. Profumo, M. Pesavento, Solvent-impregnated resins containing di-(2,4,4-trimethylpentyl)phosphinic
L.-P. Wei et al. / Chemical Engineering Journal 281 (2015) 434–443
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
acid. II: study of the distribution equilibria of Zn(II), Cu(II) and Cd(II), React. Funct. Polym. 21 (1993) 103–116. J.L. Cortina, N. Miralles, M. Aguilar, A.M. Sastre, Extraction studies of Zn(II), Cu(II) and Cd(II) with impregnated and Levextrel resins containing di(2ethylhexyl) phosphoric acid (Lewatit 1026 Oc), Hydrometallurgy 36 (1994) 131–142. I. Villaescusa, V. Salvadó, J. De Pablo, Solid–liquid extraction of Au(III) from aqueous chloride solutions by tri-n-dodecylammonium chloride impregnated in Amberlite XAD-2 resin, React. Funct. Polym. 32 (1997) 125–130. G. Cote, F. Laupretre, C. Chassagnard, Magic-angle carbon-13 NMR investigation of solvent-impregnated resins: application to 7-(4-ethyl-1methyloctyl)-8-quinolinol (Kelex 100) adsorbed on Amberlite XAD-7, React. Polym. Ion Exch. Sorbents 5 (2) (1987) 141–150. B.A. Moyer, G.N. Case, S.D. Alexandratos, A.A. Kriger, Extraction of copper(II) from sulfuric acid by macrocycle-synergized cation exchange: comparing a novel impregnated resin with its solvent-extraction analog, Anal. Chem. 65 (23) (1993) 3389–3395. S. Nii, S. Okumura, T. Kinoshita, Y. Ishigaki, K. Nakano, K. Yamaguchi, S. Akita, Extractant-impregnated organogel for capturing heavy metals from aqueous solutions, Sep. Purif. Technol. 73 (2010) 250–255. H. Hoshi, Y.Z. Wei, M. Kumagai, T. Asakura, Y. Morita, Separation of trivalent actinides from lanthanides by using R-BTP resins and stability of R-BTP resin, J. Alloys Compd. 408–412 (2006) 1274–1277. J.H. Chen, W.R. Chen, Y.Y. Gau, C.H. Lin, The preparation of di(2ethylhexyl)phosphoric acid modified Amberlite 200 and its application in the separation of metal ions from sulfuric acid solution, React. Funct. Polym. 56 (2003) 175–188. J.S. Li, H. Chen, Z.L. Guo, Y.C. Hu, Selective separation of In(III), Ga(III), and Zn(II) from dilute solution using solvent-impregnated resin containing Di(2ethylhexyl) phosphoric acid, J. Appl. Polym. Sci. 100 (2006) 253–259.
443
[22] Z. Redded, C. Gerende, Y. Andres, P. Le Cloriec, Mechanisms of Cr (VI) and Cr (Ill) from aqueous solutions by sugar beet pulp, Environ. Technol. 24 (2003) 257–264. [23] R.S. Juang, S.H. Lee, Column separation of divalent metals from sulfate solutions using impregnated resins containing di(2-ehtylhexyl)phosphoric acid, React. Funct. Polym. 29 (1996) 175–183. [24] C. Namasivayam, K. Ranganathan, Waste Fe(III)/Cr(III) hydroxide as adsorbent for the removal of Cr(VI) from aqueous solution and chromium plating industry wastewater, Environ. Pollut. 82 (1993) 255–261. [25] S.D. Alexandratos, D.W. Crick, Polymer-supported reagents: application to separation science, Ind. Eng. Chem. Res. 35 (3) (1996) 635–644. [26] M.R. Kratz, D.G. Hendricker, Preparation of polymer-bound 2,20 dipyridylamine and some of its transition metal complexes, Polymer 27 (10) (1986) 1641–1643. [27] C.V. Diniz, F.M. Doyle, V.S.T. Ciminelli, Effect of pH on the sorption of selected heavy metal Ions from concentrated chloride solutions by the chelating resin Dowex M-4195, Sep. Sci. Technol. 37 (2002) 3169–3185. [28] R.K. Biswas, D.A. Begum, Kinetics of stripping of Fe3+–D2EHP complexes from D2EHPA–kerosene phase by aqueous HCl–Cl phase using the single drop technique, Hydrometallurgy 60 (2001) 81–97. [29] H.G. Ma, L. Zhang, D.G. Li, Y.H. Sun, Structure of Fe3+ complexes formed with D2EHPA extractant, Environ. Chem. 30 (2) (2011) 538–543. [30] A. Demirbas, E. Pehlivan, F. Gode, T. Altun, G. Arslan, Adsorption of Cu(II), Zn(II), Ni(II), Pb(II), and Cd(II) from aqueous solution on Amberlite IR-120 synthetic resin, J. Colloid Interface Sci. 282 (2005) 20–25. [31] R.S. Juang, H.J. Shao, Effect of pH on competitive adsorption of Cu(II), Ni(II), and Zn(II) from water onto chitosan beads, Adsorption 8 (2002) 71–78.