Efficient Approach for Cd2+ and Ni ... - ACS Publications

Mar 23, 2007 - Cadmium pollution from the manufacture and disposal of. NiCd batteries remains an important problem in developing countries. Among the ...
0 downloads 0 Views 211KB Size
Environ. Sci. Technol. 2007, 41, 3329-3334

Efficient Approach for Cd2+ and Ni2+ Removal and Recovery Using Mesoporous Adsorbent with Tunable Selectivity K O O N F U N G L A M , †,‡ K I N G L U N Y E U N G , * ,† A N D G O R D O N M C K A Y † Department of Chemical Engineering and Environmental Engineering Program, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, P.R. China

Cadmium pollution from the manufacture and disposal of NiCd batteries remains an important problem in developing countries. Among the available remedial technologies, adsorption is popular because of its low cost and simplicity. The mesoporous NH2-MCM-41 displays good adsorption capacities for cadmium and nickel ions. This work demonstrated for the first time the use of the competition between complexant and sorbent to create high adsorption selectivity. The selectivity was manipulated by the judicious use of a chelating agent, thus achieving on-demand 100% selectivity for either Cd2+ or Ni2+ adsorption. Single and binary components adsorption studies, carried out with different metals and EDTA concentrations, solution compositions and pH, showed that NH2-MCM-41 adsorbs only cadmium with a capacity of 0.56 mmol g-1 from binary mixtures at [EDTA]/[M2+] ) 0.5 and pH 5. The NH2-MCM-41 displays 100% selectivity for nickel adsorption at [EDTA]/[M2+] ) 0.5 and pH 2 with a measured adsorption capacity of 0.50 mmol g-1. Pure Cd2+ and Ni2+ solutions were recovered by a simple acid wash, and the regenerated adsorbent could be reused without lost of performance (i.e., adsorption capacity and selectivity).

1. Introduction Cadmium and nickel are common pollutants in the manufacture of batteries, pigments, coatings, alloys, and electronics (1). The manufacture and use of NiCd batteries are still growing in developing countries (e.g., India and China) (2) despite the restrictions and bans imposed by the U.S.A., European Union, and other developed countries. This is mainly due to the higher cost of the alternative lithium-ion and nickel metal hydride batteries (3). The NiCd battery production in China alone showed a growth from less than 10 M cells in 1994 to under 200 M cells in 1999. The cadmium production grew by 1000 metric tons during the same period, and consumption reached 6000 metric tons in 2002. This dramatic growth was fueled mainly by NiCd battery production, which accounted for better than 75% of the worldwide cadmium consumption (4). Indeed, a large quantity of * Corresponding author phone: (852)2358-7123; fax: (852)23580054; e-mail: [email protected]. † Department of Chemical Engineering, Hong Kong University of Science and Technology. ‡ Environmental Engineering Program, Hong Kong University of Science and Technology. 10.1021/es062370e CCC: $37.00 Published on Web 03/23/2007

 2007 American Chemical Society

cadmium and nickel released into the environment is from the metal refining, manufacturing, and disposal of NiCd rechargeable batteries (5). Cadmium is very toxic and a human carcinogen (6). Prolonged exposure causes kidney failure, in addition to anemia, cardiovascular diseases, growth impairment, and loss of taste and smell (7). Recent reports showed a strong correlation between cadmium exposure and renal damage in the workers at Chinese NiCd battery factories (8), and cases of outright cadmium poisonings had been reported by Chinese newspapers (9). The situation could become aggravated as many of the factories are situated in rural areas near some of the important farming regions in China, and unregulated waste discharge could lead to the repetition of the “Itai-itai” incidence. The painful “Itai-itai” disease was first reported in Japan back in 1967 and was caused by the long-term consumption of rice grown on cadmium contaminated soil (10). Established remediation methods including chemical precipitation and adsorption are effective for removing the toxic metals from the wastewater effluent. Recent reports showed that it is possible to separate and recover cadmium from multicomponents systems using new solvent extraction (11) and electrodeposition (12). Adsorption remains popular because of easy operation, low-energy consumption, simple maintenance, and large capacity (13). However, common adsorbents such as activated carbons are nonselective and adsorb all metal species making their application in “recovery for reuse” a difficult task. Recent years have seen the emergence of new selective adsorbents based on mesoporous materials including MCM-41 and SBA-15. The early work of Feng et al. (14) demonstrated that grafting thiolpropyls on MCM-41 produces adsorbents with high affinity for mercury adsorption. More recently researchers have developed MCM41 adsorbents that are selective to precious metals including silver (15), gold (16), palladium, and platinum (17). Other mesoporous adsorbents including mesoporous silica containing surface aminopropyl-, aminoethyl-, and propionamidephosphonate groups were shown to perform well for actinide adsorption (18), while selective cesium sorption was achieved using copper ferrocyanide immobilized on mesoporous silica (19). Chromate and arsenate were adsorbed using mono-, di-, and triamino functionalized adsorbents through the ionic interactions between the amine heads and the negatively charged oxyanion (20, 21). Despite these progressive developments, the design and production of selective adsorbents remained difficult. This work reports the selective adsorption, removal and recovery of Ni2+ and Cd2+ from binary components solutions using mesoporous NH2-MCM-41 adsorbent. The adsorption selectivity was directly manipulated by judicious use of chelating agent (i.e., EDTA), thus achieving on demand 100% selectivity for either Cd2+ or Ni2+ adsorption. The adsorbed metals were recovered as high purity metal salt solution by a simple acid wash and the adsorbent was regenerated for reuse.

2. Experimental Section 2.1. Preparation and Characterization of Mesoporous Adsorbents. The MCM-41 mesoporous silica powders were prepared from alkaline synthesis solutions in a 1-L batch (16). Four grams of cetyltrimethylammonium bromide (CTABr, 99.3%, Aldrich) was dissolved in a liter of 1.1 M ammonium hydroxide (NH4OH, 28-30 wt %, Fisher Scientific) solution. Sixteen milliliters of tetraethyl orthosilicate (TEOS, 98%, Aldrich) was added under vigorous mixing to VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3329

FIGURE 1. (a) Electron microscope pictures, (b) X-ray diffraction plots, and (c) FTIR spectra of mesoporous adsorbents. produce a solution containing 6.6 SiO2:1 CTABr:292 NH4OH:2773 H2O. White powders were formed within 10 min of adding the silica precursor, but the mixture was allowed to age for 24 h at room temperature (295 ( 2 K) to crystallize the mesopore structures. The MCM-41 powder was filtered, washed, and dried before calcining in air at 823 K for 24 h to remove the organic template molecules (i.e., CTA+). The mesoporous adsorbent, NH2-MCM-41, was prepared by grafting aminopropyl (RNH2) groups on the pores of the MCM-41 according to the procedure described by Ho et al. (22). Calcined MCM-41 (2.5 g) was refluxed in 250 mL of dry toluene (>99.5%, Mallinckrodt) containing 0.1 mol of 3-aminopropyltrimethoxysilane (97%, Aldrich) for 18 h. The NH2-MCM-41 powder was filtered, washed with toluene, and dried in an oven at 383 K overnight. The dried adsorbent was ground and sieved to obtain a free flowing powder. The structure and morphology of the mesoporous adsorbent were examined by X-ray diffraction (XRD, Philips PW 1830), transmission electron microscopy (TEM, JEOL JEM 2010), and scanning electron microscopy (SEM, JEOL JSM 6300F). The textural properties of the adsorbent, including surface area, pore volume, and pore size distribution, were determined by nitrogen physisorption (Coulter SA 3100). The grafted organic moieties on the NH2-MCM-41 were characterized by Fourier transformed infrared spectroscopy (FTIR, Perkin-Elmer GX 2000) and X-ray photoelectron spectroscopy (XPS, Physical Electronics PHI 5000) and quantified by elemental analysis (EA, Elementar Vario EL III) and thermogravimetric and differential thermal analyses (TGA/DTA, Setaram 31/1190). 2.2. Single and Binary Components Adsorptions. The equilibrium adsorption of Cd2+ and Ni2+ on NH2-MCM-41 was measured for both single and binary components systems. The binary components adsorptions were conducted using equimolar concentrations of Cd2+ and Ni2+. The adsorptions were carried out using 0.1 g of NH2-MCM41 adsorbent for 100 mL of aqueous solutions containing the metal ions. Cadmium(II) nitrate tetrahydrate (99%, RDH) and nickel(II) nitrate hexahydrate (99%, Aldrich) were used 3330

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 9, 2007

to prepare the metal salt solutions. The batch adsorption experiments were conducted in a shaker bath kept at a constant temperature of 295 ( 2 K. The adsorption of the metal cations was rapid, and equilibrium was reached within 30 min; however, a longer adsorption time was used in the study to allow for the possible effects of EDTA on the adsorption. The adsorption was monitored by taking samples of the adsorption solution at fixed time intervals until the end of the fifth day. The initial and final concentrations of the metals in the solution were analyzed by inductively coupled plasma, atomic emission spectrometer (ICP-AES, Perkin-Elmer Optima 3000XL). Three measurements were made for each sample, and the results were averaged. Calibration was made before each set of measurements using two ICP standards solutions: 1000 ppm Cd (99.999%) in 2% HNO3 and 1000 ppm Ni (99.999%) in 2% HNO3 purchased from High-Purity Standards. The effects of pH and EDTA (99%, Aldrich) on the adsorption were also investigated.

3. Results and Discussion NiCd batteries remained popular in developing countries because of their low price, although there is growing awareness of their detrimental impact on health and environment (23). The cadmium content in the portable rechargeable batteries could be as high as 17%, and a C-size battery contains up to 10 g of cadmium (24). Their manufacture and disposal remained the primary source of cadmium pollution in the developing countries. The former from the effluent and runoff, while the latter is mainly due to the lack of proper infrastructure for collecting and recycling spent batteries (25). Adsorption remains the cheapest and simplest among the available remedial technologies. The possibility of recovering high purity metals by selective adsorption provides an added economic incentive. This work introduces a new approach to Ni2+ and Cd2+ removal and recovery using mesoporous adsorbents with tunable selectivity. 3.1. Mesoporous NH2-MCM-41 Adsorbent. Figure 1a shows the mesoporous NH2-MCM-41 powder has a disklike

TABLE 1. Properties of the Mesoporous Adsorbents

MCM-41 NH2-MCM-41

moiety

surface area (m2/g)

pore size, (nm)

specific pore volume (cm3/g)

loading of functional groups (mmol/g)

-OH -RNH2

1140 750

3.1 2.9

0.97 0.56

2.3

shape with an average diameter of 0.80 ( 0.20 µm and a thickness of 0.10 ( 0.03 µm as shown in the SEM picture. This is consistent with light scattering measurement showing the powder has a narrow particles size distribution with a mean diameter of 0.75 ( 0.15 µm. The free flowing powder displays little tendency to agglomerate (Figure 1a). The adsorbent displays an ordered hexagonal pore structure that can be seen from the high-resolution transmission electron micrograph in Figure 1a inset. The average pore diameter was calculated to be 2.92 nm according to the method reported by Kruk et al. (26) based on the d-spacing measured by XRD (Figure 1b) and specific pore volume from the N2 physisorption experiments (Table 1). This is smaller than the original 3.09 nm pore diameter of the MCM-41 and is the result of grafting aminopropyls on the pore wall. The mobility of the aminopropyls also introduced some degree of disorder (27) as indicated by the weak (110) and (200) X-ray diffraction peaks of NH2-MCM-41 (Figure 1b). The NH2-MCM-41 has a smaller BET surface area (i.e., 750 vs 1140 m2 g-1) and a specific pore volume (i.e., 0.56 vs 0.97 cm3 g-1) compared to the original MCM-41. The elemental analysis of NH2-MCM-41 shows the combustible content to be 3.54 wt % N, 9.34 wt % C, and 3.41 wt % H, which is equivalent to a loading of 2.5 mmol of aminopropyls per gram of adsorbent. Thermogravimetric and differential thermal analyses gave a comparable aminopropyl loading of 2.3 mmol g-1. The FTIR spectra in Figure 1c show that the characteristic signal for silanol groups at 3743 cm-1 disappeared after grafting the aminopropyls, while

adsorption capacity mmol/g (mg/g) Cd2+ Ni2+ 0 0.71 (79.8)

0 0.69 (40.5)

infrared signals corresponding to -NH2 stretching (i.e., 3360 and 3288 cm-1), -CN (1600 cm-1), and alkyl chain (i.e., 2845 cm-1 and 2920 cm-1) appeared. The XPS analysis of NH2MCM-41 showed a C1s binding energy of 285 eV typical to the carbons in organic alkyl chains and a N1s binding energy of 399.5 eV common to the amino compounds (28). Table 1 summarizes the physical, chemical, and textural properties of the NH2-MCM-41. 3.2. Single and Binary Components Adsorptions. Figure 2a plots the single component Cd2+ and Ni2+ adsorptions on NH2-MCM-41 at different pH. No Cd2+ or Ni2+ was adsorbed below pH 3.2, the isoelectric point of NH2-MCM-41. At low pHs, the protons compete with the metal cations for the surface aminopropyls resulting in the formation of RNH3+ and a positively charged adsorbent surface. Adsorption starts above pH 2.8 and reaches the constant values of 0.70 mmol g-1 for Cd2+ and 0.60 mmol g-1 for Ni2+ above pH 4. The equilibrium adsorption isotherms of Cd2+ and Ni2+ were measured at pH 5 and plotted in Figure 2b. The adsorption capacities for Cd2+ and Ni2+ are 0.71 mmol g-1 (80 mg g-1) and 0.69 mmol g-1 (40 mg g-1), respectively. It is clear from the results that NH2-MCM-41 has similar affinity for Cd2+ and Ni2+ adsorptions, and no selective adsorption is expected. Indeed, the K and n values of the Freundlich adsorption equation are comparable with 0.64 and 0.11 for Cd2+ and 0.66 and 0.10 for Ni2+. The effects of pH on the binary components adsorption of Cd2+ and Ni2+ on NH2-MCM-41 were investigated, and the data are plotted in Figure 2c. The adsorption of Cd2+

FIGURE 2. Single (a and b) and binary (c and d) components adsorptions of Cd2+ (]) and Ni2+ (0) as a function of solution pH (a and c) and equilibrium solution concentration (b and d). Note: symbols represent experimental data and lines represent model calculations. VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3331

FIGURE 3. (a) Equilibrium metal adsorption capacity as a function of EDTA concentration in the solution ([Cd2+] ) [Ni2+] ) 5 mM; 100 mL solution at pH 5.0 with 0.1 g of NH2-MCM-41 (2.2 mmol NH2/g); T ) 295 K; 5 days), (b) Cd(NO3)2 and Ni(NO3)2 speciation as a function of total concentration of EDTA in the solution including all EDTA species as calculated by Visual MINTEQ (c and d) binary components adsorption of Cd2+ (]) and Ni2+ (0) at (c) pH 5.0 and (d) pH 2.0 ([Cd2+] ) [Ni2+] ) [EDTA]; 100 mL solution with 0.1 g of NH2-MCM-41 (2.2 mmol NH2/g); T ) 295 K; 5 days). from the equimolar Cd(NO3)2 and Ni(NO3)2 salt solutions is similar to the single component data shown in Figure 2a. The Cd2+ adsorption is only significant above pH 3.2 and reached a constant value of 0.60 mmol g-1 above pH 3.5. The Ni2+ adsorption also starts above pH 3.2 but reached a considerably lower value of 0.16 mmol g-1 compared to the single component data. Figure 2d plots the equilibrium adsorption capacity of NH2-MCM-41 as a function of equilibrium metal concentrations in the solution. The binary components adsorption isotherms were obtained from solutions with equimolar metal ions concentrations at pH 5 and 295 ( 2 K. Maximum adsorption capacities of 0.56 and 0.15 mmol g-1 (i.e., 63 and 8.8 mg g-1) were obtained for Cd2+ and Ni2+, respectively. These are significantly smaller than the equilibrium adsorption capacity of the individual cations measured from the single component adsorptions (Figure 2b). The binary adsorption data were fitted to a Freundlich adsorption equation giving K and n values of 0.53 and 0.12 for Cd2+ and 0.15 and 0.07 for Ni2+. 3.3. Tunable Adsorption Selectivity. The previous works on precious metal adsorptions using mesoporous adsorbents (15, 16) showed that excellent selectivity could be obtained by selecting the surface functional groups that have good affinity for the target metals. Adsorbents with 100% selectivity for gold and silver were successfully prepared enabling the recovery of high purity precious metals. This present work investigates the possible use of chelating agents as a means of improving the selectivity of NH2-MCM-41 for cadmium and nickel adsorptions. Chelates such as EDTA are commonly used for metal separation and recovery by precipitation but not in the adsorption process where the presence of chelates often leads to a poorer sorption performance. Indeed, Nowack 3332

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 9, 2007

(29) reported that there is a substantial effect of organic chelating agent on metal adsorption. Figure 3a plots the cadmium and nickel adsorptions as a function of EDTA concentration in the solution. The EDTA was added to selectively bind to the Ni2+ cations and improve the adsorption selectivity toward Cd2+. The experiments were carried out using equimolar solutions containing 5 mM Cd(NO3)2 and 5 mM Ni(NO3)2 salts at pH 5. The amount of nickel adsorbed decreases to zero from an initial value of 0.15 mmol g-1 as the EDTA concentration approaches 4 mM (i.e., [EDTA]/[M2+] g 0.4). The cadmium adsorption remains unchanged at 0.56 mmol g-1 and is insensitive to EDTA up to a concentration of 8 mM (i.e., [EDTA]/[M2+] g 0.8). This means that between EDTA concentrations of 4 and 8 mM (i.e., 0.4 e [EDTA]/[M2+] e 1.0), the NH2-MCM-41 adsorbs only cadmium and displays 100% selectivity for cadmium adsorption without the loss of adsorption capacity (Figure 3a). The results show that the addition of EDTA chelates improved the adsorption selectivity, instead of decreasing the adsorption capacity as reported in previous works (30, 31). The observed selectivity improvement could be explained by examining the speciation of Cd(NO3)2 and Ni(NO3)2 in the presence of EDTA. The speciation was calculated by Visual MINTEQ (32) and plotted in Figure 3b as a function of EDTA concentration. The calculation shows that EDTA binds strongly to Ni2+ forming the NiEDTA2- complex even at low EDTA concentrations. Indeed, none of the Ni2+ cations remained above [EDTA] of 5 mM (i.e., [EDTA]/[M2+] g 0.5). On the other hand, EDTA does not bind as strongly to Cd2+, and the concentration of the CdEDTA2- complex only becomes significant above [EDTA] of 5 mM (i.e., [EDTA]/

[M2+] g 0.5). No NiHEDTA-, CdHEDTA-, NiH2EDTA, and CdH2EDTA complexes are present at pH 5. The aminopropyl groups of NH2-MCM-41 adsorb metal cations including Cd2+ and Ni2+ by forming dative bonds but not the negatively charged NiEDTA2- and CdEDTA2- complexes. Figure 3a shows that there is a good fit between experimental data and model calculation once the effect of metal speciation was taken into account. The adsorptions of cadmium and nickel were investigated for solutions containing equimolar concentrations of Cd(NO3)2, Ni(NO3)2, and EDTA (i.e., [EDTA]/[M2+] ) 0.5). The solutions were kept at constant pH 5. Figure 3c plots the metal adsorption uptakes as a function of equilibrium concentration of cadmium in the solution. It is clear from the figure that only cadmium and not the nickel in the solution was adsorbed (Figure 3c-inset). The 0.61 mmol g-1 cadmium adsorbed by the NH2-MCM-41 is slightly less than the equilibrium adsorption capacity of 0.71 mmol g-1 (Figure 2a). The selectivity for cadmium was maintained even at very low concentrations indicating that the adsorbent is capable of removing even a trace amount of cadmium with 100% selectivity. Figure 3c shows that there is an excellent agreement between model calculation and experimental data after accounting for the formation of CdEDTA2-. The binary components adsorption experiments were carried out at pH 2 and plotted in Figure 3d. It is clear from that data that at this pH, NH2-MCM-41 adsorbs mostly nickel with excellent selectivity and a capacity of 0.44 mmol g-1 (25.8 mg g-1). Below pH 3.5, the aminopropyls are protonated into RH3+ and the negatively charged NiEDTA2- and NiHEDTAcomplexes as well as a trace amount of CdHEDTA- (