Removal of metal ions from an acidic leachate solution by nanofiltration membranes

Desalination 227 (2008) 204–216

Removal of metal ions from an acidic leachate solution bynanofiltration membranes

Lina M. Ortegaa, Rémi Lebrunb*, Jean-François Blaisc, Robert Hauslerd

aInstitut des sciences de l’environnement, Université du Québec à Montréal,C.P. 8888, Succ. Centre-Ville, Montréal, Qc, Canada, H3C 3P8

bDépartement de Génie Chimique, École d’ingénierie, Université du Québec à Trois-Rivières,C.P. 500, Trois-Rivières, Qc, Canada, G9A 5H7

Tel./Fax: +1 (819) 373-6011; email: [email protected] national de la recherche scientifique (INRS-Eau Terre et Environnement), Université du Québec,

490 rue de la Couronne, Québec, Qc, Canada, G1K 9A9dStation expérimentale des procédés pilotes en environnement, École de technologie supérieure,

1100, rue Notre-Dame Ouest, Montréal, Qc, Canada, H3C 1K3

Received 13 July 2006; Accepted 20 June 2007

Abstract

This paper presents the feasibility of the application of two commercial nanofiltration (NF) membranes(Desal5 DK and NF-270) in the removal of metal ions from an acidic leachate solution generated from acontaminated soil using H2SO4 as a soil washing agent. The experimental results of soil washing indicated that H2SO4is highly effective in removing metal ions from contaminated soil. Following this process, the treatment of this acidicsolution by nanofiltration membranes showed good metal ion rejection (between 62% to 100%) where divalent ionswere better rejected than monovalent ions. For characterization purposes, the membrane experiments were conductedusing K2SO4 solutions at different pHs. Membrane performance criteria were evaluated according to membranepermeability and ionic retention in the tank and permeate, taking into account different operating conditions suchas pressures, flow rate and pH. These results demonstrated the effectiveness and feasibility of the application ofnanofiltration treatments in the cleaning-up of contaminated water residues generated during soil washing processes.

Keywords: Nanofiltration; Soil washing; Acidic leachate; Ionic separation factor; Dynamic permeability

1. Introduction

Every year, millions of tons of metals aregenerated and spilled into the environment,

*Corresponding author.

making an enormous impact on it. In 2002,industries worldwide released 22,000 tons of cad-mium, 93.900 tons of copper, 783.000 tons oflead, and 135.000 tons of zinc [1].

Although metals are naturally present in soils,

0011-9164/08/$– See front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.desal.2007.06.026

L.M. Ortega et al. / Desalination 227 (2008) 204–216 205

this contamination is attributed to human acti-vities such as vehicle emissions, use of fertilizersand pesticides, mining activities, and, principally,industrial processes [2]. Once released into thesoil, they are mobilized either by leaching or byuptake into plants, persisting for many yearsbecause metals are not biodegradable and aregenerally not mobile [3]. In high concentrations,metals can affect all groups of organisms anddifferent ecosystems, including microbial acti-vities [4]. These wastes are dumped arbitrarily,without waste management techniques. Forexample, in the United States, approximately1200 sites on the National Priority List contain amix of organic-metal contamination, of which63% are polluted with metals [5]. In Quebec(Canada), according to the Ministère de l’Envi-ronnement et de la Faune du Québec [6], 5125sites are polluted, and 11% of these sites arecontaminated with metals.

During the last decades, several technologieshave emerged that try to give innovative solutionsto the soil contamination problem. Some exam-ples are soil washing, soil flushing, vitrification,bioremediation, phytoremediation, or a combina-tion of them [5]. However, despite numerouspromising decontamination options, many tech-nologies have not been successfully implemented.This is because some of them are inadequate toreduce the metal concentration to acceptableregulatory standards, or because of the extremecost of the removal process, excavation andlandfill disposal, or simply due to disappointingresults [7]. For that reason, much effort has beenmade to find ways to remove metals from soil.One of the most common and simple treatmentsfor soil metal remediation is soil washing [8].This technique extracts metals from contaminatedsoils by solubilizing them in a washing solution[9,10]. This technique usually uses washingagents such as inorganic acids, organic acids,bases, chelating agents, or a combination of them[7]. However, the problem with the application ofthese reagents is that they generate secondary

waste products that may require additional waste-water treatments.

For that reason, in the search for waste watertechniques, there has been a growing interestduring the last decade, for “cleaner” treatmentssuch as membrane filtration (reverse osmosis(RO), nanofiltration (NF), ultrafiltration (UF),and microfiltration (MF)). Principally theapplication of NF membranes after soil treatmenthas been seen as a favorable option due to the factthat soil treatment operations (e.g., soil washing)generates large volumes of contaminated water[11].

Hence, the present work evaluates the perfor-mance of two commercial nanofiltration mem-branes (Desal5 DK and NF-270) for the removalof metal ions from an acidic leachate solutiongenerated from contaminated soil.

2. Nanofiltration

Nanofiltration with a molecular weight cut-off(MWCO) ranging between 200 and 1000 g/mol,displays separation characteristics betweenreverse osmosis (RO) and ultrafiltration mem-branes (UF) [12]. The major separation mechan-ism in salt separation using NF process can beexplained in terms of steric (organic solutes)and/or charge effects (inorganic solutes) [13].Principally charge effect is responsible of theremoval of ions from waste waters. Other factorsthat could influence membrane performance arethe electrolyte concentration and the acidiccharacteristics existing in the feed solution [14].

3. Experimental

This section is divided into two parts: gene-ration of the acidic leachate solution coming fromcontaminated soil, followed by membrane charac-terization and nanofiltration experiments of theacidic leachate solution, to evaluate membraneperformance.

L.M. Ortega et al. / Desalination 227 (2008) 204–216206

3.1. Site information

Contaminated soil from the Pointe-aux-Lièvres (PAL) site located in Quebec, Canadawas collected and used for this work. The fractionof soil used was superior to 20 µm and less than2 mm. The initial metal concentrations in the soilwere as follows: 1.4 mg Cd/kg, 120 mg Cr/kg,175 mg Cu/kg, 457 mg Mn/kg, 80 mg Ni/kg, 817mg Pb/kg and 565 mg Zn/kg.

Cu, Pb and Zn concentrations were above thecriteria B given by the Ministry of Environmentand Fauna of Quebec [15] which are, respec-tively, 100, 500 and 500 mg/kg. For that reason,this soil cannot be used for agricultural, resi-dential or recreational purposes.

3.2. Soil experimental procedure

500 g of soil was added to 1.5 L of wateragitated by a variable speed mixer at 800 rpmwith a stainless steel propeller (SS-316L, LabcorTechnical Sales, Montreal, QC., Canada) fixed toa caframo RZR50 rotor (Labcor Technical Sales).To this suspension was added H2SO4 (36 N) untila pH of 3 was obtained. This leachate was agi-tated for 25 min. Following this procedure, 5 mLof the reactant Percol E-10, an anionic polymericsolution (Ciba Specialty Chemicals Canada Inc.,Mississauga, Ont.), was added to facilitate thedecantation.

This acidic solution was passed through aWhatman no. 4 membrane filter with a porositybetween 20 and 25 µm (Whatman BioscienceInc., Newton, MA, USA) under a vacuum, toeliminate any trace of soil particles from theleachate solution. The leachate was kept at roomtemperature until membrane filtration experi-ments were realized.

H2SO4 was used for this study as a soil wash-ing agent not only for its inexpensive cost butalso because in practice, acid washing is afrequent and effective method for metal-contamination removal [8,16].

3.3. Membrane procedure

The steps used in membrane experiments areas follows:

3.3.1. Membrane compactionThe membranes were compacted with distilled

water for 3 h at room temperature, a constantrecirculation flow rate (1.01×10!4 m3/s), and apressure of 1.0×106 Pa.

3.3.2. Membrane pure water permeability(Aipw)

After membrane compaction, the pure waterpermeability of each membrane was measuredbefore and after the test of each solution. Theoperating conditions for all the experiments werethe following: three different pressures of 0.5×106, 0.8×106, and 1.0×106 Pa and a constant feedflow rate 1.01×10!4 m3/s at room temperature.Pure water permeability was measured before andafter the measurement of the dynamic permea-bility of each inorganic solution to evaluatemembrane integrity in terms of permeability.

Between each NF run with a solution (K2SO4(at different pHs) and acidic solution), themembranes were washed by circulating distilledwater with a electric conductivity of #2 µS/cm,for approximately half an hour without pressurein order to minimize experimental errors.

The permeate flux values for pure water andthe inorganic solutions, including the leachate, atdifferent operating pressures were measured andplotted against applied pressure. The slope of thecurve (straight line) was the value to determinethe membrane permeabilities.

Aipw was calculated using the followingequation:

(1)pipw

JA

Pμ

=Δ

where μ is water viscosity, Jp is water permeateflux and ΔP is the transmembrane pressure.


Jp was defined by:

(2)p p mJ Q S=

where Qp is the permeate flow rate and Sm is thesurface of the membrane.

3.3.3. Membrane dynamic permeability (Aid)Following the determination of the water

permeability, the dynamic permeability andselectivity of each membrane was measured inthe following order K2SO4 + H2SO4 at pH 6.7, 3.7and 2.2, and the principal solution of this study,the acidic leachate solution. The concentration ofK2SO4 was kept constant (11 mol/m3).

Aid was calculated by the next equation [17]:

(3)pid

e

JA

Pμ

=Δ

where

ΔPe = ΔP!ΔΠ (4)

(5)( ) ( )2 3A AX XΔΠ = Π −Π

where ΔPe is the effective pressure gradient andΔΠ is the difference in osmotic pressure betweenthe molar fraction of the concentrated boundarylayer (XA2) and the permeate solutions (XA3).

When there is no concentration polarization onthe surface of the membrane, XA2 tends towardsXA1 (molar fraction of the feed and in the reser-voir), and, therefore, ΔPe = ΔPa. ΔPa is theapparent gradient pressure.Thus,

(6)( ) ( )1 3a A AP P X XΔ = Δ −Π +Π

Consequently,

(7)( ) ( )2 1a e A AP P X XΔ = Δ +Π −Π

The osmotic pressure for dilute solutions wascalculated using the Van’t Hoff equation:

(8)iRTCΠ =∑where Σi is the number of ions per molecule ofsolute, R is the universal gas constant, T is theabsolute temperature, and C is the concentrationof the inorganic solution (mol/m3).

Rejection rate (f) was calculated using thefollowing equation:

(9)1 3 3

1 1

1A A A

A A

X X XfX X−

= = −

3.4. Membrane materials

Two types of thin-film commercial NFmembranes — Desal5 DK and NF-270 — werestudied for their permeation and ionic selectivity.Desal5 DK, manufactured by GE-Osmonics(Minnetonka, MN, USA), is a polymeric mem-brane in which a polyamide selective layer issupported on a polysulfone layer [14]. NF-270,supplied by FilmTec Corporation (FilmTecCorporation, Dow Chemical Co., Midland, MI),is a semi-aromatic piperazine-based polyamidelayer on top of a polysulphone micro-poroussupport reinforced with a polyester non-vowenbacking layer. Supplementary details about themembranes characteristics are found in Table 1.

Table 1Characteristics of the membranes used in this study

Membrane Desal 5DK NF-270

ManufacturerIsoelectric pointpH resistance (20EC)Temp. resistanceCharge

GE Osmonics4 [18,19]2–11 [18,19]90 [19]Positive [18,19]

Dow (FilmTec)3.3 [18]3–10 [18]45 [19]Negative [18]


3.5. Experimental set-up

The filtration experiments have been carriedout at room temperature, in batch mode withrecirculation of substances. The schematicrepresentation of the equipment is illustrated inFig. 1.

In this investigation, the feed solution from thetank was pumped through four different mem-brane cells (made of polyvinyl chloride (PVC)),obtaining itself a retentate that went back into thetank and a permeate that was collected into abeaker for permeate flux and rejection ratecalculation. This permeate was not recirculatedinto the system. The effective membrane area was1.26×10!3 m2. The same NF membranes wereused in all experiments.

This apparatus was designed to test differentmembranes using the same operating conditionsand also to test solutions using different pHs.Additional details of the experimental set-up werepresented and described in the investigation madeby Nöel et al. [20].

3.6. Reference solutes

During membrane characterization, certifiedanalytical grade K2SO4 and H2SO4 were used,purchased from the Aldrich Chemical Company,Milwaukee, WI. H2SO4 was used in the experi-ments to decrease the pH.

3.7. Analytical methods

All the solutions coming from feed andpermeate used during this study were determinedby a conductivity meter (model CDM 81, Radio-meter, Copenhagen, Denmark), a pH meter(Fisher Acumet model 915, Pittsburgh, PA), andplasma emission spectroscopy with a simul-taneous (ICP-AES), Varian model (VarianCanada, Inc., Mississauga, Ont.). Quality controlswere performed with certified liquid samples(multi-elements standard, catalogue number 900-

Fig. 1. Schematic diagram of NF set-up. 1 feed tank,2 pump, 3 manometer, 4 membrane cell, 5 permeateoutlet, 6 pressure valve, 7 flowmeter.

Q30-002, lot number SC0019251, SCP Science,Lasalle, Quebec) to ensure the conformity of themeasurement apparatus.

4. Results and discussion

4.1. Characterization of the membranes

4.1.1. Water permeability (Aipw)Initial and final water permeability for the

membranes is reported in Table 2. This permea-bility was measured at different transmembranepressures. As can be seen from Table 2, themembrane permeability is significantly differentfor the two membranes. The NF-270 membranetype exhibits higher permeate flux valuescompared to the Desal5 DK membrane.

As presented in this table, NF-270 (6) and (7)showed a decrease in permeability of 28.2% and

Table 2Values of initial and final water permeability (Aipw) of themembranes

Membrane Aipw (m), initial Aipw, (m), final

NF-270 (6)Desal5 DK (6)Desal5 DK (7)NF-270 (7)

2.55×10!14

0.72×10!14

0.67×10!14

2.47×10!14

1.83×10!14

0.880.791.56


Table 3Values of dynamic permeability (Aid, m)

Membrane NF-270 (6) Desal5 DK (6) Desal5 DK (7) NF-270 (7)

K2SO4 (pH 6.76)K2SO4 + H2SO4 (pH 2.9)K2SO4 + H2SO4 (pH 2.32)

2.26×10!14

2.01×10!14

1.96×10!14

0.83×10!14

0.93×10!14

1.15×10!14

0.77×10!14

0.85×10!14

1.01×10!14

2.10×10!14

1.98×10!14

1.61×10!14

36.8%, respectively. This behaviour suggests thatthe membranes underwent preferential sorption(irreversible phenomenon).

On the other hand, it was observed that theDesal5 DK membranes presented an increase inpermeability over the initial values (Table 2).These data suggest that the electrolyte solutionscontributed to the increase in membrane permea-bility (swollen effect) [21].

The NF-270 membrane was more permeableduring all the experiments than the Desal5 DK, asseen from whole permeability results (Tables 2and 3).

4.1.2. Evolution of the dynamic permeability(Aid)

Generally it is observed for a membrane thatin the presence of electrolyte solutions under thesame operating conditions, the dynamic permea-bility (Aid) is lower than pure water permeability(Aipw). This behaviour was presented for the NF-270 membrane coupons (Tables 2 and 3).

On the other hand, the Desal5 DK membranespresented the opposite behaviour. It was observedthat under the same operating conditions: thedynamic permeability was higher than pure waterpermeability (Tables 2 and 3). In this case thedynamic permeability increased as the electrolyteconcentration increased (in function of pH). Itseems that the membranes were dependent on theelectrolyte solution. As presented in the inves-tigation made by Lebrun and Xu [17], they pro-posed that the pores can suffer changes ordynamic changes due to the presence of ions inthe solution.

4.1.3. Ionic rejectionFigs. 2–5 illustrate the rejection of K+ and

SO42! ions at different pHs. It can be seen from

these figures that both membranes exhibited anexcellent retention of sulphate and potassium ionsat pH=6.76 (>95%) and good to mediumretention at pH=2.9 and 2.32.

In general it was observed that the rejection ofSO4

2! and K+ decreased with the increasing ofelectrolyte concentration (in function of pH)(Figs. 2–5). Similar data were obtained by theDesal5 DK (6) and NF-270 membrane coupons(6).

As observed in these figures, the membraneretention was dependent on the electrolyteconcentration of the feed solution. In this casehigher rejection at lower feed concentration andlower rejection at higher feed concentration wasobserved, characteristic of charged membranes[22].

For the Desal5 DK membrane, generally simi-lar retentions for SO4

2! and K+ ions at the same pHwere observed (Figs. 2 and 3). This behaviour canbe attributed to the Donnan exclusion pheno-mena where these close retentions keep theelectroneutrality of the solution on both sides ofthe membrane. Similar behaviour has beenobserved by some authors using different solu-tions [12,23].

At pH=6.76 for NF-270, similar retentions forSO4

2! and K+ ions are also presented. As men-tioned before, in order to keep electroneutrality,SO4

2! and K+ are repelled by the membrane.However, at pH < Ip (pH =2.9 and 2.32), as themembrane is charged positively, it is observed


Fig. 2. SO42! separation factor vs. apparent transmembrane

pressure for the Desal5 DK membrane (7) at differentpHs. Salt concentration = 11 mol/m3; Qc = 0.32 m3/s; 24 <T (°C) < 25.

Fig. 4. SO42! separation factor vs. apparent transmembrane

pressure for the NF-270 membrane (7) at different pHs.Salt concentration = 11 mol/m3; Qc = 0.32 m3/s; 24 <T (°C) < 25.

that the membrane repelled the co-ion (K+), andthe counter-ions are rejected (SO4

2!) (Figs. 4and 5).

4.1.4. Acidic retentionFigs. 6 and 7 represent the evolution retention

of H+ as a function of the applied pressure. Fromthese figures it can be seen that the membraneDesal5 DK (7) presented the highest retention atpH=2.9 and lower retention of H+ at pH 6.7. Asthis membrane is charged negatively at pH< Ip,the high ion retention presented by it at low pHsuggests that due to the acidic solution, the

Fig. 3. K+ separation factor vs. apparent transmembranepressure for the Desal5 DK (7) membrane at differentpHs. Salt concentration = 11 mol/m3; Qc = 0.32 m3/s; 24 <T (°C) < 25.

Fig. 5. K+ separation factor vs. apparent transmembranepressure for the NF-270 membrane (7) at different pHs.Salt concentration = 11 mol/m3; Qc = 0.32 m3/s; 24 <T (°C) < 25.

protons are able to neutralize the negative sites onthe membrane surface, thus reducing the anionrepulsion effect caused by the membrane surfacecharge [24]. On the other hand, the NF-270 (7)membrane presented an average H+ retention.Fig. 7 shows that an increase in pH results in anincrease of membrane retention. It is assumed thatthe membrane surface charge becomes lessnegative as the pH increases.

4.2. Soil washing treatment

Soil washing has been widely used for many


Fig. 6. H+ separation factor vs. apparent transmembranepressure for the Desal5 DK membrane (7) at differentpHs. Salt concentration = 11 mol/m3; Qc = 0.32 m3/s; 24 <T (°C) < 25.

Fig. 7. H+ separation factor vs. apparent transmembranepressure for the NF-270 membrane (7) at different pHs.Salt concentration = 11 mol/m3; Qc = 0.32 m3/s; 24 <T (°C) < 25.

years as a technique for decontaminating soil. Asmentioned before, this method is a water-basedsoil treatment that consists of the extraction oforganic and inorganic contaminants using a wash-ing solution [9–11]. Soil washing agents such asorganic acids (citric acid, acetic acid, etc.), che-lating agents (EDTA, ADA, NTA, DTPA), oxi-dative agents (KMnO4, H2O2, etc) and principallyinorganic acids (H2SO4, HCl, HNO3, etc.), havebeen widely applied to extract metal ions fromsoil [8,10,25]. The result after this method is amulti-component system mixture, characterizedby richness in organic and principally inorganiccompounds.

Commonly the application of inorganic acidssuch as HCL, HNO3, etc., or, as in this case, the

use of H2SO4, is done to remove metal ions fromsoils; however, the particular washing solutionapplied to the soil can depend on the metalinvolved, the specific metal compound and thespecies involved in the removal [10].

4.2.1. Leaching of ions with H2SO4

Values of the acidic leachate solution char-acterization before NF treatment are depicted inTable 4. The results reported in this table indicatethat the acidic leachate contains a variety of ions,indicative of the effectiveness of the washingsolution (H2SO4). The electric conductivity pre-sented by this solution principally came from thehigh concentration of H2SO4. The leachate pro-duced is a yellowish solution due to the presenceof humic acids. This acidic solution is not sub-jected to filtration pre-treatment for the removalof any type of contaminants.

4.2.2. Membrane dynamic permeability of theacidic leachate solution

As presented in Fig. 8, the flux increased withincreasing pressure with no linearity. This evolu-tion clearly indicates that the membranes aredependent on the solution concentration. Thistrend is possibly due to the presence of theaccumulation of inorganic and organic moleculesand suspended macromolecules on the membranesurface. This behaviour was expected due to thehigh concentration of the solution measured byelectric conductivity (Table 4).

As can be seen from Table 5, the membranesperform differently. Membrane pore clogging byorganic or inorganic compounds probably occursparticularly for the NF-270 membrane becausethe permeability was not recuperated after the runwith deionized water (Table 5).

On the other hand, for the Desal5 DK mem-brane, it was observed that the dynamic per-meability increased comparing to pure waterpermeability (Table 5). As mentioned before, thepores can suffer changes or “dynamic changes”


Table 4Characterization of the acidic leachate solution before NFtreatment

Parameter Value

pHORPConductivity, ms/cmTOC, mg/LH+, mol/LAs5+, mol/LB3+, mol/LCr3+, mol/LCu2+, mol/LFe2+, mol/LK+, mol/LMg2+, mol/LMn2+, mol/LNa+, mol/LNi2+, mol/LPO4

3!, mol/LPb2+, mo./LSO4

2!, mol/LZn2+, mol/L

3.15450.73.3619.407.08×10!1

1.82×10!3

5.30×10!2

3.60×10!3

9.76×10!2

2.09×10!1

6.73×10!1

5.925.20×10!1

9.641.09×10!2

772×10!1

2.47×10!2

3.26×101

7.87×10!1

Fig. 8. Permeate flux vs. effective pressure for the acidicleachate solution. pH = 3.15, Qc = 0.32 m3/s; 24 <T (°C)< 25.

due to the presence of ions in the solution, thusproducing an increase in permeability [17]. Thisbehaviour indicates that the dynamic permeabilityof this membrane is dependent on the electrolytesolution. Similar behaviour has been observed bysome authors using different solutions [17,20].

Table 5Values of dynamic permeability (Aid, m) and final waterpermeability (Aipw, m)

Membrane Aipw Aid Aipw, final

NF-270 (6)Desal5 DK (6)Desal5 DK (7)NF-270 (7)

1.99×10!14

0.86×10!14

0.76×10!14

1.71×10!14

1.74×10!14

0.99×10!14

0.92×10!14

1.52×10!14

1.83×10!14

0.88×10!14

0.79×10!14

1.56×10!14

4.2.3. Removal of metal ions: permeate qualityThe permeate quality is expressed in terms of

membrane retention during the nanofiltration ofthe acidic leachate solution. In this case, for allmembranes the permeate obtained was a clear andcolourless solution with relatively low amounts ofions measured by ICP-AES. Table 6 exhibits thepermeation results. As presented in this table, ingeneral divalent ions were better rejected thanmonovalent ions, a typical behaviour of chargedmembranes.

As the feed pH can considerably influence theion rejection during filtration processes due to thechange of membrane surface charge, in the caseof Desal5 DK it seems that the good ion retentionindicates that the increased protons (due to thelow pH) in the acidic solution might neutralizethe negative sites on the membrane surface.

The high retention presented by NF-270 canbe explained by the charge of this membrane. Inthis case, the cations are rejected by the mem-brane charge, as this membrane is positivelycharged at pH< Ip.

Those results present the efficacy of NF mem-branes in the treatment of soil washing solutions.

4.2.4. TOC analysisIt is difficult to specify the type of organic

contaminants present in the soil and at the sametime in the acidic leachate solution. However, inthis type of contaminated soil, one of the majorcomponents of the total organic carbon is


Table 6Ionic separation factor of the acidic leachate solution

Membrane Pressure (PSI) Cu2+ Fe2+ K+ Mg2+ Mn2+ Na+ S2+ Zn2+

NF-270(6)

Desal5 DK (6)

Desal5 DK (7)

NF-270 (7)

75115150

75115150

75115150

75115150

9999

1009898

1009999

10098

100100

100100100

84100100

89100100

89100100

85100100

8669

100vj68

72100

80100100

100100100100100100100100100

99100100

10099

1009999

100100100100

98100100

877898806298816498839898

0983

1009389

1009491

10089

100100

100999999

10099999999989999

Fig. 9. TOC separation factor (%) vs. apparent pressurefor the acidic leachate solution. pH = 3.15, Qc =0.32 m3/s; 24 <T (EC) < 25.

composed of non-specific humic-type substancesprincipally humic acids such as tannins, lignins,fulvic acids, etc.

Fig. 9 presents the removal of TOC retentionas a function of the applied pressure. The resultshows that all the membranes presented similarretentions (between 65 and 80%). As mentionedbefore, due to the yellowish color in the feedsolution (acidic leachate) given by humic acidsand soils containing humic-type substances, it isassumed that this retention is due to the presence

Fig. 10. H+ separation factor (%) vs. apparent pressure forthe acidic leachate solution. pH = 3.15, Qc = 0.32 m3/s;24 <T (°C) < 25.

of those organic components and mainly byhumic acids due to their chelating ability to binddifferent metal ions.

4.2.5. H+ separation factorDuring the filtration of the acidic leachate

solution, H+ retention increased with an increasein pressure for both membranes, showingnegative retentions as depicted in Fig. 10. As pre-sented in this figure, the retention of H+ ions wasslightly higher for the Desal5 DK than for the


NF-270 membrane. This behaviour suggests thatin acidic conditions, the positive surface chargerejects cations, leading to enhanced permeation ofacid [26]. This negative proton retention when themembrane is positively charged has beeninvestigated by several authors [14, 18, 19, 27]and this behaviour has been attributed to thehigher mobility of the H+ compared to othercations in solution. This low retention of acid andthe high retention of metal ions, suggest thepossibility of acid recuperation through NFtreatment.

5. Soil washing integrated into NF process

Several studies, at both laboratory and indus-trial scale, have investigated the performance ofthe membrane process (RO, NF, UF) on theseparation of organic and inorganic pollutantsfrom waste waters coming from soil treatments.These results have demonstrated the effectivenessof incorporating membrane processes into soiltreatments [10,28–30].

Soil washing techniques have been charac-terized to be an economical and feasible processdue to the application of effective and low costchemicals (particularly the use of inorganicacids). The application of this method can lead tothe reduction of the volume of hazardous mater-ials and at the same time the transformation ofcontaminants to non-hazardous substances[10,31].

On the other hand, the advantages of the appli-cation of NF membranes among other filtrationprocesses not only are higher fluxes, lowoperating pressures, relatively low operation andmaintenance cost but also higher removalefficiency of organic and inorganic components,reducing the volume of waste products and apossibility for acid reuse [30–32]. For that reason,the combination of soil treatments in the NFprocess can bring about the removal and recoveryof contaminants from a great variety of aqueous

streams and a total water recirculation in thesystem, eliminating the negative problems relatedwith the release of trace contaminants into theenvironment as a result of soil washing processes[30].

6. Conclusions

The removal of pollutants from soils is one ofthe most challenging environmental tasks. Forthat reason, this study was intended to evaluatetwo NF membranes for the recovery of ionscoming from contaminated soil. The experimentalresults have successfully demonstrated that theuse of the washing solution H2SO4 has a highextraction efficiency in the removal of ions fromcontaminated soils. During the application of NFmembranes in the treatment of the acidic leachatesolution, the commercial membrane showed highion selectivity. Additionally it demonstrated agood removal of TOC during the filtration experi-ment (70–89%). During the filtration experi-ments, the Desal5 DK membrane demonstratedan increase in dynamic permeability, showing itsdependence to the electrolyte solutions. On theother hand, the NF-270 membrane presented anopposite behaviour: a decrease in dynamic per-meability. These results indicate that the bestcandidate for the removal of metal ions wasDesal5 DK, which presented good membranepermeability combined with good ion retentionand acidic resistance.

During this investigation it has been demon-strated that the application of NF processes willbring about the elimination of metal ions fromwaste waters, a possible recovery of acid duringthe process and a reduction of the amount ofinorganic and organic waste.

7. Symbols

Aipw — Pure water permeability of mem-brane, m


Aid — Dynamic solution permeability ofmembrane, m

C — Concentration, mol/m3

f, fN — Overall and intrinsic separation fac-tor separation

Ip — Isoelectric pointJp — Solution permeate flux, m/sM — Molar concentrationORP — Oxido-reduction potential, mVP — Pressure, PaPa — Apparent gradient pressure, PaPe — Effective pressure gradient, PaQc — Feed flow rate, m3/sQp — Permeate flow rate, m3/sSm — Membrane area, m2

T — Temperature, ECTOC — Total organic carbon, mg.O2/LX — Molar fractionXA1 — Molar fraction of the feedX2 — Molar fraction of the concentrated

boundary layerX3 — Molar fraction of the permeate

Greek

Δ — Gradientμ — Fluid viscosity, Pa.sΠ — Osmotic pressure, PaΣi — Number of ions per molecule of

solute

Acknowledgements

The authors wish to thank FQRNT, NSERCand the Canada Research Chair program for thefinancial support and Hayka Industry for pro-viding the experimental set-up. Thanks are alsodue to Myriam Chartier (INRS-ETE) for ICP-AES analysis.

References[1] O.V. Singh, S. Labana, G. Pandey and R. Budhi-

rahja, Phytoremediation: an overview of metallic iondecontamination from soil. Appl. Microbiol. Bio-technol., 61 (2003) 405–412.

[2] R. Kiem, H. Knicker, B. Ligouis and I. Kögel-Knabner, Airborne contaminants in the refractoryorganic carbon fraction of arable soils in highlyindustrialized areas. Geoderma, 114 (2003) 109–137.

[3] N. Vassilev, J.P. Schwitzguébel, T. Thewys, D. vander Lelie and J. Vangronsveld, The use of plants forremediation of metal-contaminated soils. Sci. World,4 (2004) 9–34.

[4] E. Baath, Effects of heavy metals in soil microbialprocesses and populations (a review). Water. Air SoilPollut., 47 (1989) 335–379.

[5] C.N. Mulligan, R.N. Yong and B.F. Gibbs, Remedi-ation technologies for metal-contaminated soils andgroundwater: an evaluation. Eng. Geol., 60 (2001)193–207.

[6] MEFQ, Bilan sur les terrains contaminés/Statistiquesgénérales en décembre 2001. http://www.menv.gouv.qc.ca/sol/terrains/bilan-2001/index.htm.

[7] C.N. Mulligan and F. Eftekhari, Remediation withsurfactant foam of PCP-contaminated soil. Eng.Geol., 70 (2003) 269–279.

[8] N. Meunier, J.-F. Blais and R. D. Tyagi, Selection ofa natural sorbent to remove toxic metals from acidicleachate produced during soil decontamination.Hydrometallurgy, 67 (2002) 19–30.

[9] A. Barona, I Aranguiz and A. Elias, Metals asso-ciations in soils before and after EDTA extractivedecontamination: implications for the effectivenessof further clean up procedures. Environ. Pollut., 113(2001) 79–85.

[10] R.W. Peters, Chelant extraction of heavy metals fromcontaminated soils. J. Hazard. Mater., 66 (1999)151–210.

[11] S.K. Sikdar, D. Grosse and I. Rogut, Membranetechnologies for remediating contaminated soils: acritical review. J. Membr. Sci., 151 (1998) 75–85.

[12] Y. Ku, S.W. Chen and W.Y. Wang, Effect of solu-tion composition on the removal of copper ions bynanofiltration. Sep. Purif. Technol, 43 (2005) 135–142.

[13] J. Schaep, B. Van der Bruggen, C. Vandecasteele andD. Wilms, Influence of ion size and charge in nano-filtration. Sep. Purif. Technol., 14 (1998) 155–162.

[14] S. Bandini, J. Drei and D. Vezzani, The role of pH


and concentration on the ion rejection in polyamidenanofiltration membranes. J. Membr. Sci., 264(2005) 65–74.

[15] MEFQ, Politique de protection des sols et deréhabilitation des terrains contaminés. Ministère del’Environnement et de la Faune du Québec, Québec,Qc, Canada, 1999.

[16] C.G. Rampley and K.L. Ogden, Preliminary studiesfor removal of lead from surrogate and real soilsusing a water soluble chelator: adsorption and batchextraction. Environ. Sci. Technol., 32 (1998) 987–993.

[17] R.E. Lebrun and Y. Xu, Dynamic characterization ofnanofiltration and reverse osmosis membranes. Sep.Sci. Technol., 34 (1999) 1629–1641.

[18] G. Hagmeyer and R. Gimbel, Modelling the rejectionof nanofiltration membranes using zeta potentialmeasurements. Sep. Purif. Technol., 15 (1999) 19–30.

[19] J. Tanninen and M. Nyström, Separation of ions inacidic conditions using NF. Desalination, 147 (2002)295–299.

[20] I.M. Noël, R.E. Lebrun and C. Bouchard, Electro-nanofiltration of a textile direct dye solution.Desalination, 129 (2000) 125–136.

[21] A. Hamza, G. Chowdhury, T. Matsuura and S.Sourirajan, Study of reverse osmosis separation andpermeation rate for sulfonated poly(2,6-dimethyl-1,4-phenylene oxide) membranes of different ionexchange capacities. J. Appl. Poly. Sci., 58 (1995)613–620.

[22] Y. Xu and R.E. Lebrun, Investigation of the soluteseparation by charged nanofiltration membrane:effect of pH, ionic strength and solute type. J.Membr. Sci., 158 (1999) 93–104.

[23] K. Mehiguene, Y. Garba, S. Taha, N. Gondrexon and

G. Dorange, Influence of operating conditions on theretention of copper and cadmium in aqueous solu-tions by nanofiltration: experimental results andmodeling. Sep. Purif. Technol., 15 (1999) 181–187.

[24] K. Mehiguene, S. Taha, N. Gondrexon, J. Cabon andG. Dorange, Copper transfer modeling through ananofiltration membrane in the case of ternaryaqueous solution. Desalination, 127 (2000) 135–143.

[25] B.J.W. Tuin and M. Tels, Continuous treatment ofheavy metal contaminated clay soils by extraction instirred tanks and in a countercurrent column.Environ. Technol., 12 (1991) 178–190.

[26] J. Tanninen, S. Platt, A. Weis and M. Nyström,Long-term acid resistance and selectivity of NFmembranes in very acidic conditions. J. Membr. Sci.,240 (2004) 11–18.

[27] E. Childress and M. Elimelech, Relating nano-filtration membrane performance to membranecharge (electrokinetic) characteristics. Environ. Sci.Technol., 34 (2000) 3710–3716.

[28] K. Linde and A.-S. Jönsson, Nanofiltration of saltsolutions and landfill leachate. Desalination, 103(1995) 223–232.

[29] T.A. Peters, Purification of landfill leachate withreverse osmosis and nanofiltration. Desalination, 119(1998) 289–293.

[30] K. Volchek, D. Velicogna, A. Obenauf, A. Somers,B. Wong and A.Y. Tremblay, Novel applications ofmembrane processes in soil cleanup operations.Desalination, 147 (2002) 123–126.

[31] R. Semer and K. Reddy, Evaluation of soil washingprocess to remove mixed contaminants from a sandyloam. J. Haz. Mater., 45 (1996) 45–57.

[32] X. Lu, X. Bian and L. Shi, Preparation and charac-terization of NF membrane. J. Membr. Sci., 5032(2002) 1–9.

Documents

Removal of metal ions from an acidic leachate solution by nanofiltration membranes