Adsorption of nickel(II) from aqueous solution onto activated carbon prepared from coirpith

Adsorption of nickel(II) from aqueous solution onto activated carbon prepared from coirpith

Separation and Purification Technology 24 (2001) 497– 505 Adsorption of nickel(II) from aqueous solution onto activate...

133KB Sizes 7 Downloads 54 Views

Separation and Purification Technology 24 (2001) 497– 505

Adsorption of nickel(II) from aqueous solution onto activated carbon prepared from coirpith K. Kadirvelu *, K. Thamaraiselvi, C. Namasivayam Department of En6ironmental Sciences, Bharathiar Uni6ersity, Coimbatore-64 1 046, Tamil Nadu, India Received 31 May 2000; received in revised form 2 March 2001; accepted 9 April 2001

Abstract Activated carbon has been prepared from coirpith by chemical activation and characterized. Carbonised coirpith is able to adsorb Ni(II) from aqueous solution. It was noted that a decreasing in the carbon concentration with constant Ni concentration, or an increase in the Ni concentration with constant carbon concentration resulted in a higher nickel uptake per unit weight of carbon. The Langmuir and Freundlich models for dynamics of metal ion uptake proposed in this work fit the experimental data reasonably well. The adsorption capacity (Q0) calculated from Langmuir isotherm was 62.5 mg Ni(II) g − 1 at initial pH of 5.0 at 30°C for the particle size 250– 500 mm. The adsorption of Ni increased with pH from 2 to 7 and remained constant upto 10. The recovery of Ni(II) after adsorption can be carried out by treatment of the Ni loaded carbon with HCl. Desorption studies confirms adsorption is ion exchange. As coirpith is discarded as waste material from coir processing industries, the carbon is expected to be economical product for metal ion remediation from water and wastewater. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Activated carbon; Carbonization; Adsorption; Adsorption isotherms

1. Introduction Recovery of heavy metals from wastewaters and industrial wastes has become a very important environmental issue. Ni has many useful applications in our life and is harmful if discharged into natural water resources [1]. Ni(II) is present in the effluents of silver refineries, electroplating, zinc base casting and storage battery in* Corresponding author. Tel.: + 91-422-880716; fax: +91422-422387. E-mail address: [email protected] (K. Kadirvelu).

dustries [2]. In India, the acceptable limit of Ni in drinking water is 0.01 mg l − 1 and for discharge of industrial wastewater is 2.0 mg l − 1 [3]. At higher concentrations, Ni(II) causes cancer of lungs, nose and bone. Dermatitis (Ni itch) is the most frequent effect of exposure to Ni, such as coins and costume jewelry. Ni carbonyl [Ni(CO)4] has been estimated as lethal in humans at atmospheric exposures of 30 ppm for 30 min [4]. Acute poisoning of Ni(II) causes headache, dizziness, nausea and vomiting, chest pain, tightness of the chest, dry cough and shortness of breath, rapid respiration, cyanosis and extreme weakness [5]. Hence, it

1383-5866/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 8 6 6 ( 0 1 ) 0 0 1 4 9 - 6


K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505

is essential to remove Ni(II) from industrial wastewaters before mixing with natural water sources. Conventional methods for the removal of Ni(II) from wastewaters include chemical precipitation, ion exchange, filteration, chemical reduction, electrodeposition and adsorption on activated carbon [6]. But due to operational demerits and high cost of the treatment, some new technologies have been tried for a long time. Among them less expensive non conventional adsorbents are being investigated. Rice hull [7], sphagnum peat [8], peat moss [9], soya been and cotton seed hulls [10], blast furnace slag [11], apple waste [12], peanut hull carbon [13] and straw [14] have been investigated to remove Ni(II) from wastewater. India, the third largest producer of coconut in the world, produces about 8160 million coconuts every year [15]. Coirpith constitutes as much as 70% of the coconut husk, and is a light, fluffy material generated in the process of the separation of fibers from the coconut husk in coir-fiber industries [16]. It is estimated that the production of coirpith in India is about 0.5 million ton/year, while world production is about 3.6 million ton/year [17]. Coirpith is disposed as waste and its accumulation around coir-proces sing industries is creating large problems [18]. Different methods of preparation of activated carbons from coirpith, the treatment of dyeing industry wastewater and adsorption kinetics of Cu(II), Hg(II) and Pb(II) from aqueous solution have been reported recently [19– 24]. Many reports have appeared on preparation of activated carbon from cheaper and readily available materials [13,19,25–27]. The objective main of this study was to investigate the feasibility of using carbonized coirpith for the removal of Ni from aqueous solution by varying parameters of agitation time, Ni(II) concentration, carbon concentration and pH.

2. Experimental

2.1. Adsorbent Waste coirpith was collected from Senthil coir processing industry, Pollachi, Coimbatore district,

Tamil Nadu, India and dried in sunlight. Preparation and characteristics of the carbon have been recently reported and are presented in Table 1 [22]. Carbon as prepared by mixing one part of coirpith, 1.8 parts of concentrated sulphuric acid and 0.1s part of ammonium persulphate [NH4S208] and keeping it in hot air oven at 80°C for 12 h. The carbonized material was washed with distilled water to remove free acid and soaked in 1% sodium bicarbonate solution overnight to remove any residual acid. The material was then washed with distilled water and dried at 80°C. The carbon was stored in plastic containers and used for further treatment studies.

2.2. Adsorbate A stock solution of 1000 mg l − 1 of Ni(II) was prepared by dissolving 4.4790 g of ultra pure Ni sulphate [NiSO4.6H20] in double distilled water, acidified with nitric acid to prevent hydrolysis. All the solutions were made with double distilled water.

Table 1 Characteristics of activated carbon Parameter


pH Conductivity (mS cm−1) Specific gravity Bulk density(g ml−1) Porosity (%) Moisture (%) Ash (%) Volatile matter (%) Fixed carbon (%) Surface area (m2g−1) Decolorizing power (mg g−1) Ion exchange capacity (mequiv. g−1) Solubility in water (%) Solubility in 0.25 M HCl (%) Sodium (w/w,%) Potassium (w/w,%) Calcium (w/w,%) Iron (w/w,%) Silica (w/w,%) Iodine number (mg g−1)

8.63 47.80 0.86 0.22 75.00 11.43 11.05 53.17 34.06 592 296 0.5 1.60 2.60 1.45 0.02 0.13 0.13 3.61 572

K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505


Fig. 1. Effect of agitation time and initial concentration of Ni(II) on the adsorption of Ni(II). Conditions — initial pH 5.0; carbon concentration, 20 mg/50 ml.

2.3. Batch mode adsorption studies

2.4. Batch mode desorption studies

Batch adsorption studies were conducted to determine the equilibrium time needed to reach saturation. Adsorption kinetics were carried out using 50 ml of metal ion solution containing the desired concentration (10– 40 mg l − 1) at initial pH 5.0 and 20 mg of adsorbent in 100 ml conical flasks (agitation speed 160 rpm). At predetermined time intervals, samples were separated by centrifugation and analyzed by spectrophotometrically using glyoxime [28]. All experiments were carried out at initial pH 5.0 (except when the pH effect was studied) where the adsorption is significant but below the pH where metal hydroxide precipitation occurs. Adsorption isotherms were carried out with different initial Ni(II) concentration and fixed adsorbent concentration of 30 mg for 50 ml of solution of 20 and 40 mg l − 1 with 20 mg of carbon. 0.1 M l − 1 of HCl and NaOH was used to adjust the pH.

After adsorption, the metal ion loaded carbons were separated and slightly washed with 50 ml of HCl of various strengths ranging from 0.025 to 0.15 M l − 1 and were analyzed as before. All the chemicals are used of analytical reagent grade and were obtained from SD fine chemicals, Mumbai, India. All the experiments were carried out in duplicate and mean values are presented. Maximum deviation was 5.0%. 3. Results and discussion

3.1. Adsorption kinetics (effect of agitation time and initial metal ion concentration) Fig. 1 shows the effect of agitation time and initial metal ion concentration on adsorption. The equilibrium was reached within a short period (40 min) for all concentrations of metal ion. Time required to reach equilibrium and equilibrium adsorption capacities are given in Table 2. Ac-

K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505


cordingly, the equilibrium was selected as the agitation time for rest of the batch experiments. The contact time required for Ni(II) removal with carbon is very minimal. This is one of the parameters for economical wastewater treatment plant applications. Kinetic curves are modeled using Lagergren equation [29], which allows to compute the adsorption rate kad (1 min − 1) according to: log10(qe − q)=log10 qe −kad

t 2.303


where, q is the amount adsorbed (mg g − 1), qe is the amount adsorbed at equilibrium time (mg g − 1), kad is the adsorption rate constant (1 min − 1). Linear plots of log10(qe −q) versus t show that applicability of Lagergren equation for activated carbon. The kad values at different initial metal ion concentrations were calculated from slope of the plots and presented in Table 2. As the initial metal ion concentration increases, the kinetic adsorption rate constant decreases. The initial adsorption kinetic coefficient g (l mg − 1 min − 1) is also computed [30] and were presented in Table 2. k=

  dC dt

V t “ 0 mC0


where t is the time (min), C is the metal ion concentration at time (t), V is the solution volume (l), m is the activated carbon weight (mg), C0 is the initial concentration (mg l − 1).

3.2. Effect of carbon concentration Fig. 2 shows that adsorption of Ni(II) as function of carbon concentration. Increasing carbon concentration increased the percent adsorption of Ni(II). For complete removal of Ni(II) from 50 ml of 20 and 40 mg l − 1, a maximum carbon concentration of 40 and 60 mg respectively, was required. Increasing carbon concentration increasing removal was due to availability of more surface area and functional groups.

3.3. Adsorption isotherms Langmuir isotherm was applied for adsorption equilibrium [31]. Ce/qe = l/(Q0b)+Ce/Q0)


where, Ce is the equilibrium concentration (mg l − 1), qe is the amount absorbed at equilibrium (mg g − 1), Q0 and b is the Langmuir constants related to adsorption capacity and energy of adsorption. The linear plot of Ce/qe versus Ce shows that the adsorption obeys the Langmuir model. Q0 and b were determined from the slope and intercept of the plot and are presented in Table 3. Values of adsorption capacity (Q0) of the other adsorbents are given in Table 4 for comparison. The essential characteristics of Langmuir isotherm model can be explained in term of a dimensionless constant separation factor or equilibrium parameter RL [41], which is defined by RL = 1/(1 + bC0)


Table 2 Adsorption kinetic results of Ni(II) onto coirpith carbon Ni(II) concentration (mg l−1)

Equilibrium time Equilibrium capacity Initial adsorption coefficient (g) (l (min) (qe) (mg g−1) mg−1 min−1×105)

Adsorption rate constant (kad) (min−1×10−2)

10 20 30 40

40 40 40 40

10.59 8.93 9.09 6.37

20.63 39.5 45.0 64.38

7.81 8.28 6.14 5.55

K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505


Fig. 2. Effect of carbon concentration on adsorption of Ni(II). Conditions: initial pH 5.0; agitation time 1 h.

where b is the Langmuir constant, C0 is the initial concentration of metal ion. According to Hall et al. [41] it has been shown using mathematical calculations that the parameter RL indicates the shape of isotherm as follows: RL value RL\1 RL =1 0BRLB1 RL =0

Type of isotherm Unfavorable Linear Favorable Irreversible

RL values between 0 and 1 (Table 3) indicate favorable adsorption of Ni onto coirith carbon. The Freundlich isotherm is represented by Eq. [42]. qe = kf Cel/n


where Ce is the equilibrium concentration (mg l − 1), qe is the amount adsorbed at equilibrium (mg g − 1), kf and n are constants incorporating all factors affecting the adsorption process such as adsorption capacity and intensity. Linear plots of in qe versus ln Ce show that the adsorption followed Freundlich models well. kf and n were calculated from the intercept and slope of the plots. The constants were presented in Table 3. According to Treyball [43] it has been shown using mathematical calculations that n values between 1 and 10 represents beneficial adsorption.

3.4. Adsorption mechanism approach Fig. 3 shows the effect of pH on the removal of Ni(II) by adsorption onto activated carbon and also by hydroxide precipitation [Ni (OH)2]. Adsorption increases with increase of pH from 2 to 10. Adsorption of metal ion is efficient for the concentrations studied at below precipitation pH. One of the conventional method of removing metals from aqueous solution is the precipitation of metal hydroxides using an alkali. This method has some demerits in that the complete removal of metals is not possible due to the solubility product of metal hydroxide. The above precipitation pH Ni(II) is removed by both adsorption as well as precipitation. Removal of metals was strongly dependent on the pH of the solution and removal increased 0–100% for 20 mg l − 1 and 80% for 40 mg l − 1 in a narrow pH range 2–7. The increase in metal ion removal as pH increases can be explained on the basis of a decrease in competition between proton (H+) and positively charged metal ion at the surface sites, and by decrease in positive charge which results in a lower repulsion of the adsorbing metal ion. The removal efficiencies of metal ion is affected by the initial metal ion concentration with the removal decreasing as the concentration increases at constant pH. This can be explained as follows. At low metal/carbon

K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505


Table 3 Langmuir and Freundlich constants Ni(II) concentration (mg l−1)


10 20 30 40

0.297 0.174 0.123 0.095

Q0 (mg g−1)

b (l mg−1)

kf (1−l/n ll/ng−1)








ratio, the metal ion adsorption involves the high energy sites, as the metal/carbon ratio increases, the higher energy sites are saturated and adsorption begins on the lower energy sites, resulting in a decrease of adsorption efficiencies. pH curves are shifted to alkaline regions, as it has been previously reported by several authors [13,44– 46]. In pH range 4–7 maximum removal was observed which might be due to partial hydrolysis which might be due to formation of Ni(OH) and Ni(OH)2. Low solubilities of hydrolysis metal ion species may be another reason for maximum adsorption. The above statement agrees with earlier reports [13,44– 46]. The mechanism of metal ion adsorption may also be explained based on ion exchange model. In activated carbon, carbon– oxygen complexes are present. The surface oxygen complexes hydrolyzed water molecules as shown below: 2COH + M+ “(CO)2M+ +2H+


COH22 + M+ “ COM2 + M+ +2H+.


The above mechanism has been confirmed by an increasing initial metal ion concentration in aqueous solution as the final pH of solution decreases. This clearly indicates that uptake of more metal ions release of H+ increases. Since carbon is prepared upon treatment with H2SO4, (NH4)2S2O8 and NaHCO3, groups such as CONa+, CONa22 + , CSO3Na and CONH4 are present. Na+ in the above functional groups also exchange with H+ in the aqueous medium as follows. CONa+ +H+ “ COH +Na+ CONa +2H “COH +


2+ 2


(8) +

CONa22 + +2H+ “ COH22 + +2Na+

(9) (10)

COSO3Na +H+ “ CSO3H+ + Na+.


Fig. 4 shows the effect of initial pH an final pH of Ni(II) concentrations of 20 and 40 mg l − 1 for coirpith carbon. The curve obtained under conditions such that [Ni(II)= 0]. Eqs. (8)–(11) contribute to an increase in pH in blank curve. At the same time reactions lead to the release of Na+. Excess of Na+ was introduced into carbon when it was soaked with NaHCO3 to neutralize free sulphuric acid during the preparation of carbon. When Ni(II) is present in solution, its adsorption Table 4 Comparison of adsorption capacity of Ni(II) onto various adsorbents Adsorbent

Q0 (mg g−1)


Rice hull Dye stuff treated rice hull Red mud Clay treated with NaCl Clay treated with HCl Natural clay Fly ash Polymerized onion skin Peanut hull carbon Granular activated carbon Fe(III)/Cr(III) hydroxide Sphagnum peat Bituminas coal Chemically treated coal H2O2 treated coal Melon seed husk Peat moss Aspergillus niger Rhizopus nigricans Blast furnace slag Coirpith carbon Soyabean hull Cotton seed hulls NaOH treated rice hull

5.58 6.16 15.0 14.54 10.93 12.5 0.683 7.55 53.65 1.49 22.94 14.69 6.47 7.29 8.12 5.9 9.18 1.1 1.0 55.75 62.5 89.52 46.57 12.31

[7] [7] [32] [33] [33] [34] [35] [36] [13] [13] [37] [8] [38] [28] [38] [39] 9 [40] [40] [11] This work [10] [10] [10]

K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505


Fig. 3. Effect of pH on Ni(II) removal by precipitation and adsorption. Conditions: carbon concentration, 20 mg/50 ml; agitation time 1 h.

will free some H+ and the pH will increase lower than in the blank (Eqs. (12), (13) and (16)). At the same time Na+ will also be released according to Eqs. (14)–(17). 2COH+ +Ni2 + “(CO)2Ni2 + +2H+


2+ “ CONi2 + +2H+ COH+ 2 +Ni


2CONa+ + Ni2 + “(CO)2Ni2 + +2Na+


CONa22 + +Ni2 + “ CONi2 + +2Na+


2CSO3H+ + Ni2 + “ (CSO3)2 Ni +2H+


2CSO3Na + Ni2 + “(CSO3)2 Ni +2Na+.


3.5. Desorption studies Desorption studies help elucidating the mechanism adsorption and recovering precious metals from wastewater and adsorbent. Studies were carried out to desorb Ni(II) from the metal loaded carbons with various molarities of HCl. The quantitative recovery of metal ion (Fig. 5) is possible. This is further more evidence that

ion exchange is involved in the adorption process.

4. Conclusion The adsorption kinetic experimental data may be useful for environmental technologists in designing heavy metal containing wastewaters. In batch mode adsorption studies, removal increased with the increase of contact time, carbon concentration and pH. The adsorption of metal ion in this system depends on the solution pH, initial metal ion concentration and carbon concentration. Langmuir and Freundlich parameters show that the Ni(II) metal ion adsorption process is favorable onto activated carbon. The pH effect on metal ion removal revealed that the removal efficiency, in the presence of carbon was higher by adsorption than by metal hydroxide precipitation at below pH of precipitation. Quantitative recovery of metal ion from carbon is possible. The desorption studies by HCl indicates that ion exchange seems to be an important process in the


K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505

Fig. 4. Effect of initial pH on final pH on Ni(II) removal.

Fig. 5. Effect of HCl concentration on Ni(II) desorption.

adsorption of metal ion by carbon. Since coirpith is disposed as waste material from coir processing industries the resulting carbon is expected to economically feasible and environmentaly safe.

Acknowledgements Authors K.K. and K.T. are grateful to Professor B. Ilango, Vice-Chancellor, and Professor V.

K. Kadir6elu et al. / Separation/Purification Technology 24 (2001) 497–505

Gopal, Department of Environmental Sciences Bharathiar Unversity, Coimbatore, India, for encouragement and help, also grateful to CSIR, Government of India for award of senior research fellowship.

References [1] S. Al-Asheh, Z. Duvnjak, Adv. Environ. Res. 1 (1997) 194 – 212. [2] K. Kadirvelu, PhD Thesis, Bharathiar University, Coimbatore, India, 1998. [3] Y.C. Sharma, G. Prasad, D.C. Rupiainwar, Int. J. Envirion. Stud. 40 (1992) 41 –535. [4] R.P. Beliles, The lesser metals, in: F.W. Oehme (Ed.), Toxicity of Heavy metals in the Environment, Part 2, Marcel Dekker, New York, 1979, p. 383. [5] P. Parker, Encyclopedia of Environmental Sciences, 2nd edn, McGraw Hill, New York, 1980. [6] J.D. Dean, F.L. Bosqui, K.H. Lanouette, Environ. Sci. Technol. 6 (1972) 518 –522. [7] R. Suemitsu, R. Venishi, I. Akashi, M. Nakana, J. Appl. Polym. Sci. 31 (1986) 75 –83. [8] T. Viraraghavan, M.M. Drohamraju, Environ Sci. Hlth 28A (1993) 1261 – 1276. [9] Y.S. Lo, D.A. Johnwase, C.F. Forster, Water Res. 29 (1995) 1327 – 1332. [10] W.E. Marshall, E.T. Champagne, W.J. Evans, J. Environ. Sci. Hlth. 9(A) (1995) 1922 –1977. [11] S.V. Dimitrova, Water Res. 30 (1996) 228 – 232. [12] E. Maranon, H. Sastre, Bioresource Technol. 38 (1991) 39 – 43. [13] K. Periasamy, C. Namasivayam, Waste Mgmt. 15 (1995) 63 – 68. [14] V.J. Larson, H.P. Schiernp, J. Environ. Qual. 10 (1981) 188 – 193. [15] A.C. Lourduraj, T.S Raveendran, J. Ganapathy, K. Rajamanikam, F.J.S. Kennedy, Ind. Coconut J. 22 (1991) 10. [16] P. Kchampan, Handbook on Coconut-Palm, IBM, New Delhi, 1981, p. 272. [17] C.S. Pillai, K. Gopakumar, P.K. Rohatgi, J. Sci. Ind. Res. 40 (1981) 154 – 165. [18] P. Ajithkumar, Ind. Coconut J. 22 (1991) 9. [19] K. Kadirvelu, C. Namasivayam, Environ. Technol. 21 (2000) 1091 – 1097. [20] C. Namasivayam, K. Kadirvelu, Bioresource Technol. 63 (1997) 123 – 127. [21] C. Namasivayam, K. Kadirvelu, Chemosphere 34 (1997) 377 – 399.


[22] C. Namasivayam, K. Kadirvelu, Carbon 37 (1999) 79 – 84. [23] C. Namasivayam, K. Kadirvelu, Ind. J. Chem. Technol. 5 (1998) 334 – 337. [24] C. Namasivayam, K. Kadirvelu, Bioresource Technol. 48 (1994) 79 – 81. [25] K. Kadirvelu, M. Palanivel, R. Kalpana, S. Rajeswari, Bioresource Technol. 79 (2000) 263 – 265. [26] J.W. Hassler, Purification with Activated Carbon, Chemical Publishing, New York, 1974, p. 169. [27] C.A. Toles, W.E. Marshall, M.M. Johns, Carbon 35 (1997) 1407 – 1414. [28] H.E. Allen, R.A. Minear, Metallic ions, in: M.J. Suess (Ed.), Examination of Water for Pollution Control, Physical Chemical and Radiological Examinations, vol. 2, Pergamon, Oxford, 1989, p. 141. [29] S. Lagergren, Svenska Ventenskaps. Handl. 24 (1898) 1 – 39. [30] M. Baudu, P. Le Cloriec, G. Martin, Water Sci.Technol. 32 (1991) 1659 – 1665. [31] I. Langmuir, J. Am. Chem. Soc. 40 (1918) 1361. [32] A.I. Zouboulis, K.K. Kydros, J. Chem. Technol. Biotechnol. 58 (1993) 95 – 101. [33] S. Hawash, J.Y. Farash, M.S. El-Geundi, Adv. Sci. Technol. 9 (1994) 244 – 257. [34] S. Hawash, H. El-Abd, M.S. El-Geundi, M.M. Nassae, J.Y. Farah, Adv. Sci. Technol. 9 (1994) 231 – 243. [35] P.P. Viswakarma, K.P. Yadava, V.N. Singh, Pertanika 12 (1989) 357 – 366. [36] P. Kumar, S.S. Dara, J. Polym. Sci. 19 (1981) 397 – 402. [37] C. Namasivayam, K. Ranganathan, Waste Mgmt. 14 (1994) 709 – 716. [38] D. Singh, N.S. Rawat, J. Chem. Technol. Biotechnol. 4 (1997) 39 – 41. [39] F.E. Okieman, V.U. Ohyenkpa, Biological Waste 29 (1989) 11 – 16. [40] A. Kapoor, T. Viraraghavan, Bioresource Technol. 63 (1997) 109 – 113. [41] K.R. Hall, L.C. Eagleton, A. Acrivos, T. Vermeulen, Ind. Eng. Chem. Fund 5 (1966) 212 – 219. [42] H. Freundlich, W. Heller, J. Am. Chem. Soc. 61 (1939) 2228. [43] R.E. Trybal, Mass Tranfers Operations, 3rd edn, McGraw, New York, 1980. [44] M.O. Corupcioglu, C.P. Huang, Water Res. 21 (1987) 1031 – 1044. [45] M.M. Benjamin, K.F. Heyes, J.O. Leckie, J. Water Pollut. Control. Fed. 54 (1982) 472. [46] C. Gabaldon, P. Marzal, J. Ferrer, A. Seco, Water Res. 30 (12) (1996) 3050 – 3060.