Hấp phụ và giải hấp Cd trong môi trường đất
Adsorption and desorption of cadmium by goethite pretreated with phosphate Kaijun Wang, Baoshan Xing * Department of Plant and Soil Sciences, Stockbridge Hall, University of Massachusetts, P.O. Box 37245, Amherst, MA 01003-7245, USA Received 11 October 2001; received in revised form 21 March 2002; accepted 4 April 2002 Abstract The adsorption of Cd by oxides or soils have been extensively studied, however, the desorption has received rela- tively limited attention, especially in the presence of phosphate. In this study, a batch equilibration method was used to investigate Cd sorption and desorption by goethite pretreated with phosphate. Phosphate not only enhanced Cd ad- sorption, but also accelerated the adsorption process. Compared with Cd adsorption by goethite alone, phosphate substantially moved the adsorption curves (edges) to lower pH range, indicative of enhancement of Cd sorption. The Cd adsorption by the pretreated goethite reached apparent equilibrium within 24 h at 20 °C, while such equilibrium was not observed after 4 weeks in the absence of phosphate. Cadmium was more readily released from phosphate-treated goethite. It is believed that phosphate blocked the pores on goethite surface, which lead to the fast adsorption kinetics and high extraction percentage. These results provided strong support for the diffusion of Cd into goethite particles. Ó 2002 Elsevier Science Ltd. All rights reserved. Keywords: Cadmium; Adsorption; Desorption; Phosphate; Goethite; pH 1. Introduction Cadmium (Cd) is one of the toxic trace metals, which can be introduced into and accumulate in soils through agricultural application of sewage sludge and fertilizers, and/or through land disposal of metal-contaminated municipal and industrial wastes. Chemical processes strongly affect the fate and availability of Cd in soils. It is accepted that concentrations of heavy metals includ- ing Cd in soil solution are most likely controlled by sorption–desorption reactions on the surface of soil col- loidal materials (Brummer et al., 1988; Ainsworth et al., 1994; Backes et al., 1995; McLaren et al., 1998). As major components of soil colloidal materials, iron and manganese oxides play very important roles in the sorption of heavy metals. Goethite is the most wide- spread iron oxide in natural environments (Schwert- mann and Cornell, 2000), and has been well studied and used for sorption experiments (Fischer et al., 1996; Strauss et al., 1997; Manceau et al., 2000). The sorption of heavy metals by soils or oxides has been extensively studied (Barrow, 1998; Gray et al., 1998; Eick et al., 1999; Christophi and Axe, 2000; O’Reilly et al., 2001). Results suggest that sorption ap- pears to be a multi-step process involving an initial fast adsorption followed by a slow adsorption and diffusion into solid particles. Such diffusion might explain the low reversibility of heavy metal sorption to oxides (Ains- worth et al., 1994; Brummer et al., 1988; Barrow et al., 1989). However, this suggestion was not directly derived from experiments, but from the fitting of sorption data to various diffusion equations (Barrow, 1986; Krish- namurti et al., 1999). Previous studies (Christensen, 1984; Young et al., 1987; Swift and McLaren, 1991) indicate that desorption Chemosphere 48 (2002) 665–670 www.elsevier.com/locate/chemosphere * Corresponding author. Tel.: +1-413-545-5212; fax: +1-413- 545-3958. E-mail address: bx@pssci.umass.edu (B. Xing). 0045-6535/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 045-6535( 0 2)001 6 7 - 4 is as important as sorption, because it governs the rate and extent of metal ions released from sorbents. Im- proved understanding of desorption characteristics may allow us to better evaluate the bioavailability and po- tential toxicity of trace metals in soils. However, com- pared with adsorption research, desorption of heavy metals received relatively limited attention, especially in the presence of phosphate (Davis and Upadhyaya, 1996; McLaren et al., 1998). Phosphorus, one of the major nutrients for plants, is widely distributed in soils and has a high affinity for iron hydroxide surfaces. Several in- vestigators have shown that the adsorption of phosphate on iron oxides may enhance cation adsorption (Kuo and McNeal, 1984; Diaz-Barrientos et al., 1990; Venema et al., 1997). However, the reduction of Cd adsorption in the presence of phosphate in soils was reported by Krishnamurti et al. (1999); they attributed this reduction to the formation of Cd–phosphate complexes in solu- tion. We want to further examine Cd release in the presence of phosphate. Therefore, the objectives of this research were to study the Cd sorption by goethite pretreated with phosphate, and to determine the effect of phosphate treatment on Cd desorption kinetics from goethite. 2. Materials and methods 2.1. Goethite preparation Goethite was prepared using the method of Schw- ertmann et al. (1985). Briefly, ferrihydrite was precipi- tated by rapidly adding 400 ml of a 1 M FeNO 3 solution to 3600 ml of KOH solution to yield a final OH À con- centration of 0.7 M. The suspension was then stored in polyethylene bottles at 10 °C for 2 months. After syn- thesis, the goethite was washed thoroughly with de- ionized water to remove NO À 3 and K þ , and finally dried at 60 °C and ground in an agate mortar. The prepared sample showed a typical X-ray diffraction pattern of goethite, and only 0.8% of total iron remained soluble in oxalate solution, indicative of nearly complete crystal- lization from ferrihydrite. Our goethite was the same sample as used in a surface topographic study by Fischer et al. (1996). The goethite had a BET surface area of 75 m 2 /g measured by the method described by Kruse and Lagaly (1988). 2.2. Adsorption measurement Goethite suspensions were prepared in 0.01 M Ca(NO 3 ) 2 with 20 mg oxide per milliliter. After ultra- sonic dispersion, 1 ml aliquots were dispensed into 15 ml polycarbonate tubes (20 mg/tube). Small amounts of HNO 3 and NaOH were gradually applied to the sus- pensions to achieve a pH range between 3 and 8. The tubes were then made up to 8 ml with 0.01 M Ca(NO 3 ) 2 and shaken on an rotary shaker for 14 days at 20 °Cto reach constant pH. Because it was extremely difficult to achieve the same pH for replicates, a large number of samples were used at various pH values rather than employing duplicates or triplicates at a same pH. Phosphate treatment was carried out as follows. One milliliter 0.01 M phosphate [Na 2 HPO 4 in 0.01 M Ca(NO 3 ) 2 ] was added to the goethite suspension. After the samples were thoroughly mixed on a rotary shaker at 30 rpm for 7 days at 40 °C, 1 ml 10 À4 M Cd [Cd(NO 3 ) 2 in 0.01 M Ca(NO 3 ) 2 ] was added to the tubes (i.e., the goethite suspensions) to make up to the 10 ml final volume. Then, the suspensions were ultrasonically dis- persed and shaken for various reaction times of 15 min, 24 h, 7 days, and 4 weeks at 20 °C. At the end of each reaction time, the suspensions were centrifuged and the supernatants were collected for measurements of pH and Cd concentration. The Cd adsorption in the absence of phosphate was also prepared following the same procedure at the same time. The Varian atomic absorption spectrometer (AAS) with graphite furnace (SpectrAA 55) was used to determine the Cd concen- tration. 2.3. Desorption measurement Cadmium extractions were carried out immediately after Cd adsorption. The 5 M HCl solution was used as an extractant. Briefly, samples were centrifuged at 7840 g at the end of adsorption. After the supernatants were removed carefully, fresh 5 M HCl (10 ml) was added and mixed with solid particles by ultrasonic dispersion. The new suspensions were again mixed on the rotary shaker for a predetermined extraction time. Due to the disso- lution of goethite in 5 M HCl (Cornell et al., 1976), extraction times were designed to be 5 min, 30 min, 2 h, and 4 h for each tube, that is, each tube was sequentially extracted four times. At the end of each extraction, the samples were centrifuged at 7840 g for 10 min. The su- pernatants were decanted from the tubes and the con- centration of Cd was measured by AAS. 3. Results and discussion Goethite and phosphate are commonly found in soils. Under natural conditions, the interaction between phosphate and goethite is usually in equilibrium before any addition of exogenous heavy metals into soils. In this respect, the sorption characteristics of Cd by goethite pretreated with phosphate may resemble the sorption– desorption reactions occurring in the natural environ- ment. 666 K. Wang, B. Xing / Chemosphere 48 (2002) 665–670 3.1. Cadmium adsorption Cadmium adsorption by goethite was highly pH- dependent. Almost all Cd was adsorbed at high pHs for the initial concentration of 10 À5 M. After 15 min reaction, nearly all Cd were adsorbed at pH P 7:2 in the absence of phosphate, and at pH P 5:5 in the pres- ence of phosphate. Fig. 1 shows that Cd adsorption was greatly enhanced in the presence of phosphate. Al- though phosphate did not change the shape and slope of adsorption curves, the adsorption edge shifted distinctly to lower pH, indicating the substantial enhancement of Cd adsorption. The pH 50 (the pH value at which 50% of heavy metal ions were adsorbed) was reduced by up to 1.7 pH units (Table 1). At the reaction time of 15 min, for example, about 75% at pH 5 and 100% at pH 6 of Cd ions were adsorbed by the goethite pretreated with phosphate, while only 1.5% and 23% were adsorbed at the same pHs in the absence of phosphate. The en- hanced Cd adsorption could be attributed to the re- duction of the pH zpc (zero point of charge) of goethite and surface potential after the phosphate treatment (Kuo and McNeal, 1984). Phosphate sorption increases surface negative charge and decreases the electrostatic potential near the solid surface, which might cause an increase in the Cd surface loading. This result is con- sistent with that reported by Venema et al. (1997). Phosphate also altered the adsorption kinetics of Cd by goethite. The continuous decrease of pH 50 as a function of reaction time implies that the Cd adsorption on the untreated goethite did not reach equilibrium after 4 weeks reaction (Table 1). At pH 6, for example, 23% of Cd was adsorbed at 15 min, while 32%, 37%, and 55% of Cd was adsorbed at reaction times of 24 h, 7 days, and 4 Fig. 1. Percent Cd adsorption by the phosphate-treated goethite () and untreated goethite () as a function of pH at different reaction times and a constant temperature of 20 °C. Table 1 pH 50 a values of Cd at initial concentration 10 À5 M and 20 °C Reaction time Phosphate-trea- ted goethite Untreated goethite Difference 15 min 4.69 6.39 1.70 24 h 4.55 6.24 1.69 7 days 4.54 6.17 1.63 4 weeks 4.52 5.94 1.42 a pH 50 is defined as such a pH value at which 50% of initial metal ions are adsorbed. K. Wang, B. Xing / Chemosphere 48 (2002) 665–670 667 weeks, respectively. However, except for 15 min contact time, little change in the amount of Cd adsorbed was observed between times in the presence of phosphate. These results indicate that the Cd adsorption on goethite in the presence of phosphate needed 6 24 h to reach apparent equilibrium at 20 ° C. The acceleration of Cd adsorption may be caused by blocking and/or occupy- ing of meso- and micro-pores on goethite surface or between goethite domains by phosphate (Madrid and de Arambarri, 1985; Torrent et al., 1990; Strauss et al., 1997). As a result, Cd ions might not be able to diffuse further into the pores and the overall adsorption process was faster. These meso- and micro-pores have been proposed as pathways for diffusion of metal ions into oxide matrix following initial fast adsorption (Brummer et al., 1988; Bibak et al., 1995). Fischer et al. (1996) reported that the pores on goethite surface were 20–30 nm wide and be- came narrower towards the interior of the crystals to 2 nm or less. The diffusion of ions may be not limited from the pore entrances (20–30 nm), but with increasing pore depth and decreasing pore size, only small ions can penetrate into the narrow pores. Because phosphate ions are bigger than Cd ions (0.22 nm in radius for phos- phate, 0.097 nm for Cd 2þ ), it is possible for phosphate ions to diffuse into and block such pores during pre- treatment. For some micro-pores, phosphate may only diffuse partway into the pores or block the entrance, resulting in rapid sorption kinetics. 3.2. Cd desorption Several authors have demonstrated that the sorption process of trace metals is not completely reversible, and explanations have been proposed for such observations, including diffusion of trace metals within oxide particles or into micro-pores (Brummer et al., 1988; Backes et al., 1995; Gray et al., 1998), precipitation (Farrah and Pickering, 1978), incorporation of metals into oxides (McKenzie, 1970; Ainsworth et al., 1994), and re- adsorption (Davis and Upadhyaya, 1996). A wide range of solutions were used as extractants in these studies, including Ca(NO 3 ) 2 , HNO 3 , NaCl, EDTA, and NTA. These solutions exhibit differences in their extraction ability. In our study, 5 M HCl was used as an extractant because such a strong acid would effectively release Cd from the goethite particle surface and prevent re- adsorption of Cd (Farrah and Pickering, 1978). In the present work, we focused on the samples with nearly 100% Cd adsorption; in this case, the pHs of initial sorption reaction ranged from 5 to 8. Our results showed that most Cd was released from goethite after a 400 min extraction (cumulative time). Fig. 2 shows the cumulative average percentage of Cd released as a function of extraction time for the phos- phate-treated goethite. Because of a slight difference in Cd recovery between pHs from 5 to 8, the average values were used to demonstrate the overall effect of contact time. As one may see, there was no substantial difference in Cd adsorption for reaction times beyond 24 h (Fig. 1), but the Cd recovery decreased with increasing contact time (Fig. 2). For 15 min adsorption, about 95% and nearly all of Cd adsorbed were released into solution after 5 and 400 min extraction, respectively. But for 4 weeks adsorption, percentages were 82% at 5 min and 93% at 400 min. One possible reason is that not all the pores were blocked by phosphate, which might be still available for Cd diffusion. Another possibility is that the sorption of phosphate by goethite might not reach equilibrium during pretreatment. As a consequence, phosphate could diffuse further into the pores during Cd sorption and Cd diffusion would be extended with in- creasing contact time. Also, the dissolution rate of go- ethite–phosphate complex may become slower with time, which could cause lower Cd extraction. In an attempt to examine the effect of phosphate pretreatment on Cd desorption, the Cd extraction from goethite in the absence of phosphate was carried out for the 4 weeks sorption experiments (Fig. 3). It is clear that the phosphate pretreatment greatly enhanced Cd de- sorption. Moreover, we might have underestimated the difference in Cd desorption between the two systems because goethite treated with phosphate dissolves more slowly in 5 M HCl than the phosphate-free goethite (Strauss et al., 1997); more Cd would be released from phosphate-treated goethite assuming its dissolution rate is the same as that of phosphate-free goethite and, thus, the difference on the percentage of Cd extraction would be even greater than that we reported here. Nevertheless, Fig. 2. Effect of contact time on the average percent Cd re- leased (on a cumulative basis for both time and Cd recovery) from the phosphate-treated goethite. Some error bars are in- visible because the symbols are greater than the error bar size (error bar represents one standard deviation). 668 K. Wang, B. Xing / Chemosphere 48 (2002) 665–670 results support our hypothesis that phosphate can block pores of goethite, which hinders Cd diffusion into the inner surface binding sites as Cd ions did in the absence of phosphate. As a result, more Cd accumulated on the particle surface or near surface in the presence of phosphate. Such Cd distribution would be easier for extraction. In this study, the samples were thoroughly mixed by ultrasonic dispersion. It is reasonable to assume that the goethite particles were evenly dissolved from surface in HCl solution. Under this assumption, our results sup- port the diffusion of Cd into solid phase or the meso- and micro-pores of goethite particles with increasing contact time (Brummer et al., 1988; Backes et al., 1995). One may try to use re-adsorption or precipitation to explain the results reported here. However, we disagree for the following reasons. First of all, 5 M HCl is such a strong extractant that the possibility of re-adsorption would be ruled out. Secondly, HCl would be able to dissolve any precipitated hydrous oxide species of heavy metals (Farrah and Pickering, 1978). If precipitation was involved in the Cd sorption on goethite, more Cd would be released from goethite at higher pHs of initial sorp- tion reaction due to the higher Cd sorption. In fact, a slightly decreasing Cd recovery was observed with in- creasing pH from 5 to 8 for a given adsorption period (Table 2), which can be explained well by diffusion. In the pH range from 5 to 8, the adsorption of phosphate on goethite decreased from about 190 to about 165 lmol/g after 7 day sorption (Fig. 4). Therefore, more pores would be blocked by phosphate at lower pHs, which would result in the increase of Cd extraction. The opposite would be true for high pHs. 4. Conclusion Phosphate treatment greatly influenced Cd sorption. Not only did phosphate enhance Cd adsorption, but also accelerated the adsorption kinetics. The adsorption edges shifted to lower pH substantially in the presence of phosphate, indicative of enhanced Cd adsorption. The adsorption of Cd did not reach equilibrium after 4 weeks of reaction in the absence of phosphate; however, for the Fig. 3. Average percent Cd released (on a cumulative basis for both time and Cd recovery) from the phosphate-treated go- ethite ( ) and untreated goethite () after 4 weeks adsorption at 20 °C (error bar represents one standard deviation). Table 2 Percent Cd released from the phosphate-treated goethite after 5 min extraction by 5 M HCl Adsorption period pH Percentage released (%) 15 min 5.38 95.4 6.18 95.1 6.51 94.9 6.92 94.0 7.21 92.4 24 h 5.43 89.4 6.07 88.3 6.51 88.4 6.96 87.1 7.23 85.6 7 days 5.36 87.5 6.07 86.4 6.50 86.2 6.93 84.0 7.22 83.9 4 weeks 5.20 83.1 5.92 82.8 6.52 79.4 6.97 79.0 Fig. 4. Amount of phosphate adsorbed by goethite after 7 days at 10 À3 M initial concentration and 40 °C, showing a gradual decrease of sorption with increasing pH. K. Wang, B. Xing / Chemosphere 48 (2002) 665–670 669 treated goethite, Cd adsorption reached apparent equi- librium within 24 h, as indicated by a relatively constant pH 50 . Cadmium was more readily released from the phosphate-treated goethite. Both adsorption and extrac- tion results support the hypothesis that phosphate can block/occupy meso- and micro-pores of goethite parti- cles during the pretreatment, causing fast sorption and desorption processes. Therefore, these results provided strong support for the diffusion of Cd into goethite particles during sorption. Acknowledgements Authors thank Dr. G.W. Brummer for his helpful discussion and for providing the goethite material. References Ainsworth, C.C., Pilon, J.L., Gassman, P.L., Van Der Sluys, W.G., 1994. Cobalt, cadmium, and lead sorption to hydrous iron oxide: residence time effect. Soil Sci. Soc. Am. J. 58, 1615–1623. Backes, C.A., McLaren, R.G., Rate, A.W., Swift, R.S., 1995. Kinetics of cadmium and cobalt desorption from iron and manganese oxides. Soil Sci. Soc. Am. J. 59, 778–785. Barrow, N.J., 1986. Testing a mechanistic model. II. The effect of time and temperature on the reaction of zinc with a soil. J. Soil Sci. 37, 277–286. Barrow, N.J., 1998. Effects of time and temperature on the sorption of cadmium, zinc, cobalt, and nickel by a soil. Aust. J. Soil Res. 36, 941–950. Barrow, N.J., Gerth, J., Brummer, G.W., 1989. Reaction kinetics of the adsorption and desorption of nickel, zinc, and cadmium by goethite. II. Modeling the extent and rate of reaction. J. Soil Sci. 40, 437–456. Bibak, A., Gerth, J., Borggaard, O.K., 1995. Retention of cobalt by pure and foreign-element associated goethites. Clays Clay Miner. 43, 141–149. Brummer, G.W., Gerth, J., Tiller, K.G., 1988. Reaction kinetics of the adsorption and desorption of nickel, zinc and cad- mium by goethite. I. Adsorption and diffusion of metals. J. Soil Sci. 39, 37–51. Christensen, T.H., 1984. Cadmium soil sorption at low concentrations: II. Reversibility, effect of changes in solute composition, and effect of soil aging. Water Air Soil Pollut. 21, 115–125. Christophi, C.A., Axe, L., 2000. Competition of Cd, Cu, and Pb adsorption on goethite. J. Environ. Eng.––ASCE 126 (1), 66–74. Cornell, R.M., Posner, A.M., Quirk, J.P., 1976. Kinetics and mechanisms of the acid dissolution of goethte (a-FeOOH). J. Inorg. Nucl. Chem. 38, 563–567. Davis, A.P., Upadhyaya, M., 1996. Desorption of cadmium from goethite (a-FeOOH). Water Res. 30, 1894–1904. Diaz-Barrientos, E., Madrid, L., Contreras, M.C., Morillo, E., 1990. Simultaneous adsorption of zinc and phosphate on synthetic lepidocrocite. Aust. J. Soil Res. 28, 549. Eick, M.J., Peak, J.D., Brady, P.V., Pesek, J.D., 1999. Kinetics of lead adsorption/desorption on goethite: residence time effect. Soil Sci. 164 (1), 28–39. Farrah, H., Pickering, W.F., 1978. Extraction of heavy metal ions sorbed on clays. Water Air Soil Pollut. 9, 491– 498. Fischer, L., Zur Muhlen, E., Brummer, G.W., Niehus, H., 1996. Atomic force microscopy (AFM) investigations of the surface topography of a multidomain porous goethite. Eur. J. Soil Sci. 47, 329–334. Gray, C.W., McLaren, R.G., Roberts, A.H.C., Condron, L.M., 1998. Sorption and desorption of cadmium from some New Zealand soils: effect of pH and contact time. Aust. J. Soil Res. 36, 199–216. Krishnamurti, G.S.R., Huang, P.M., Kozak, L.M., 1999. Sorption and desorption kinetics of cadmium from soils: influence of phosphate. Soil Sci. 164, 888–898. Kruse, H.H., Lagaly, G., 1988. Automatische apparatur zur volumetrischen gasadsorptionsmessung. GIT Fachzeitsch- rift fur das labor 32, 1096–1100. Kuo, S., McNeal, B.L., 1984. Effects of pH and cadmium sorption by a hydrous ferric oxide. Soil Sci. Soc. Am. J. 48, 1040–1044. Madrid, L., de Arambarri, P., 1985. Adsorption of phosphate by two iron oxides in relation to their porosity. J. Soil Sci. 36, 523–530. Manceau, A., Nagy, K.L., Spadini, L., Ragnarsdottir, K.V., 2000. Influence of anionic layer structure of Fe-oxyhydrox- ides on the structure of Cd surface complexes. J. Colloid Interf. Sci. 228, 306–316. McKenzie, R.M., 1970. The reaction of cobalt with manganese dioxide minerals. Aust. J. Soil Res. 8, 87–106. McLaren, R.G., Backes, C.A., Rate, A.W., Swift, R.S., 1998. Cadmium and cobalt desorption kinetics from soil clays: effect of sorption period. Soil Sci. Soc. Am. J. 62, 332– 337. O’Reilly, S.E., Strawn, D.G., Sparks, D.L., 2001. Residence time effects on arsenate adsorption/desorption mechanisms on goethite. Soil Sci. Soc. Am. J. 65, 67–77. Schwertmann, U., Cambier, P., Murad, E., 1985. Properties of goethites of varying crystallinity. Clays Clay Miner. 33, 369–378. Schwertmann, U., Cornell, R.M., 2000. Iron Oxides in the Laboratory. Wiley–VCH, Weinheim. Strauss, R., Brummer, G.W., Barrow, N.J., 1997. Effects of crytallinity of goethite: II. Rates of sorption and desorption of phosphate. Eur. J. Soil Sci. 48, 87–99. Swift, R.S., McLaren, R.G., 1991. Micronutrient adsorption by soils and soil colloids. In: Bolt, G.H. et al. (Eds.), Interactions at the Soil Colloid–Soil Solution Interface. Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 257–292. Torrent, J., Barron, V., Schwertmann, U., 1990. Phosphate adsorption and desorption by goethites differing in crystal morphology. Soil Sci. Soc. Am. J. 54, 1007–1012. Venema, P., Hiemstra, T., van Riemsdijk, W.H., 1997. Inter- action of cadmium with phosphate on goethite. J. Colloid Interf. Sci. 192, 94–103. Young, T.C., DePinto, J.V., Kipp, T.W., 1987. Adsorption and desorption of Zn, Cu and Cr by sediments from the Raisin River. J. Great Lakes Res. 13, 353–366. 670 K. Wang, B. Xing / Chemosphere 48 (2002) 665–670 . example, 23% of Cd was adsorbed at 15 min, while 32%, 37%, and 55% of Cd was adsorbed at reaction times of 24 h, 7 days, and 4 Fig. 1. Percent Cd adsorption. 1988. Reaction kinetics of the adsorption and desorption of nickel, zinc and cad- mium by goethite. I. Adsorption and diffusion of metals. J. Soil Sci.