Mobility of zero valent iron nanoparticles and liposomes in porous media

Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

Contents lists available at ScienceDirect

Colloids and Surfaces A: Physicochemical and Engineering Aspects journal homepage: www.elsevier.com/locate/colsurfa

Mobility of zero valent iron nanoparticles and liposomes in porous media K. Terzi a,b , A. Sikinioti-Lock a , A. Gkelios a , D. Tzavara a,c , A. Skouras a,c , C. Aggelopoulos a , P. Klepetsanis a,c , S. Antimisiaris a,c , C.D. Tsakiroglou a,∗ a

Foundation for Research and Technology Hellas - Institute of Chemical Engineering Sciences, Stadiou street, Platani, 26504 Patras, Greece Department of Chemical Engineering, University of Patras, 26504 Patras, Greece c Laboratory of Pharmaceutical Technology, Department of Pharmacy, University of Patras, 26504 Patras, Greece b

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

• The presence of oleic phase triggers the lipid membrane disruption.

• CMC-coated nZVIs are highly mobile in any form.

• Empty liposomes co-flowing with nZVI coalesce and are trapped in pore network. • Liposomes encapsulating CMCcoated nZVI retain, in part, their mobility. • Trapped liposomes are easily remobilized by injecting water.

a r t i c l e

i n f o

Article history: Received 2 April 2016 Received in revised form 14 July 2016 Accepted 19 July 2016 Available online 22 July 2016 Keywords: Porous media Zero valent iron Nanoparticles Flow Liposomes Stability Mobility Particle size distribution

a b s t r a c t Suspensions of zero valent iron nanoparticles (nZVI) are commonly used for the in situ remediation of groundwater contaminated with chlorinated solvents. Stable aqueous suspensions of zero-valent nano-particles (nZVI) are prepared by wet chemistry techniques and stabilized with a carboxyl-methylcellulose (CMC) coating. To enhance their penetration length along with their capacity to attach on oil/water interfaces, nanocomposites are prepared where the CMC-coated nZVI suspension is encapsulated in liposomes. The liposomes might be regarded as vehicles for the safe delivery of nZVI to hydrophobic pollutant targets. The integrity of synthesized liposomes membranes is evaluated with batch tests and flow-through tests in a pre-saturated with oil glass-etched pore network. For assessing the mobility and longevity of nZVI suspensions under flow-through conditions, visualization flow tests are performed on the glass-etched pore network for three types of suspensions: (i) CMC-coated nZVI; (ii) CMC-coated nZVI encapsulated in liposomes; (iii) mixture of CMC-coated nZVI and empty liposomes. The measured iron and lipid concentration breakthrough curves for all cases are interpreted by accounting for

∗ Corresponding author. E-mail address: [email protected] (C.D. Tsakiroglou). http://dx.doi.org/10.1016/j.colsurfa.2016.07.054 0927-7757/© 2016 Elsevier B.V. All rights reserved.

712

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

the transient changes caused on the particle size distribution in the suspension collected from the outlet. Albeit the CMC-coated nZVI are always very mobile, stable, and detectable in the effluent, a fraction of the liposomes or no liposomes are detected in the effluent, when injecting CMC-coated nZVI encapsulated in liposomes or mixed with empty liposomes, respectively. The flushing of the pore network with water, remobilizes the liposomes and withdraws them completely from the pore system. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Most lab-scale studies and field-scale applications of nanoscale materials for soil/groundwater remediation has focused on nanoscale zero-valent iron (nZVI) and related products. The advantages of nZVI may be summarized as follows [1]: • Fast reaction: (i) short treatment time; (ii) less cost; (iii) less exposure for workers, fauna and flora. • Complete reduction pathway to non-toxic byproducts: (i) less exposure for workers, fauna and flora. • In situ treatment: (i) less equipment and above-ground structures required; (ii) less costs. The zero valent iron nanoparticles (nZVI) are synthesized by a variety of bottom-up and top-down methods, reported in detail elsewhere [2–7]. However, a disadvantage of nZVI suspended in aqueous media is the agglomeration of particles to each other and the fast attachment of agglomerates to the soil surface. Agglomeration may be caused by groundwater conditions (pH, ionic strength), surface properties of the particles, the age of materials, or shipping conditions [8,9]. Modifications to enhance the mobility, reactivity, or stability of nanoscale iron particles have been made by using polymers or surfactants. Surface modifiers increase the surface charge of the nanoparticles thereby providing electrostatic stabilization. They can also create a surface brush layer that engenders long-range strong steric repulsion forces, usually insensitive to high ionic strengths for which double layer repulsions would be greatly shielded. Examples of iron nanoparticles stabilization methodologies include: (1) Coatings such as polyelectrolyte or triblock polymers which are added in the suspension [9–14] to stabilize the iron nanoparticles and improve their mobility. (2) Nanoparticle encasement in emulsified vegetable oil droplets (EZVI) [15] to improve their stability and reactivity (by facilitating their contact with the contaminant). The emulsions have the

potential to partition into non-aqueous phase liquids (NAPLs) due to hydrophobic oil continuous phase that may be miscible with NAPL. (3) Development of multi-functional nano-composites (MFNC) by incorporating nZVI into porous sub-micron particles (nanocomposites) of silica [16,17] or carbon [18] to prevent agglomeration and couple iron reactivity with high silica/carbon adsorption capacity. (4) The use of guar gum as a stabilizing agent of aqueous suspension to reduce the attachment efficiency of NP in soil grains [19]. Liposomes are artificially prepared vesicles composed of a lipid bilayer which encapsulate a region of aqueous solution inside a hydrophobic membrane. They can be used as vehicles for administration of nutrients, drugs and imaging agents Nanoliposomes have been recently used to encapsulate Ultra Small Paramagnetic Iron Oxide (USPIO) in order to target large quantities of NPs, and direct them to target sites by using small amounts of targeting ligands (for in vivo imaging purposes) [20]. The physicochemical properties of liposomes (size, surface charge and hydrophilicity, membrane fluidity, integrity) can be easily manipulated by using different preparation techniques and modulating the composition of the lipid bilayer or by adding specific polymeric coatings (e.g. PEG, chitosan etc.) on their surface [21]. Depending on the specific application, liposomes can be constructed to have optimal properties and release (of encapsulated materials) kinetics. In fact, by using a specific methodology, known as dried-rehydrated vesicles (DRV) technique [22], very high entrapment of USPIOs (<20 nm mean diameter) in nanoliposomes (mean diameter ∼100 nm) has been achieved [20]. Perhaps, the distribution, dispersion, and penetration of active NPs in polluted soils can be modulated in a beneficial way, by encapsulating them in nanoliposomes (Fig. 1a). In this manner, active particles (nZVI) will be released at high concentration in the areas where most pollutants reside, by enhancing the pollutant remediation efficiency. The general concept of nanoparticles delivery close to trapped NAPL ganglia, after the liposomes (Fig. 1a) membrane disruption, is shown schematically in Fig. 1b.

Fig. 1. (a) The general concept of a liposome encapsulating nanoparticles. (b) The liposomes are used as vehicles for the release of nanoparticles at the vicinity of NAPL/water interfacial regions.

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

713

The objective of the present study is to assess the mobility of ZVI-nanocomposite suspensions in porous media. Three types of nZVI-based aqueous suspensions are prepared: (i) carboxylmethyl-cellulose (CMC)-coated nZVI suspension; (ii) CMC-coated nZVI suspension encapsulated in liposomes; (iii) mixture of CMCcoated nZVI and empty liposome suspensions. The visualization of the transport of the various types of suspensions in a glass-etched pore network is combined with measurements of (i) the iron and lipid concentration breakthrough curves and (ii) the temporal evolution of the particle-size distribution (PSD) in effluents to assess the mobility of nZVI in porous media. 2. Materials and methods 2.1. Synthesis of nZVI suspension Aqueous suspensions of nZVI were prepared by adding NaBH4 solution under anoxic conditions in an aqueous solution of FeSO4 . 7H2 O pre-grafted with carboxy-methyl-cellulose (CMC). The protocol of He and Zhao [12,23,24] was followed by keeping the w/w ratios, Polymer/Fe = 0.003, and NaBH4 /Fe = 2. A 250 mL threeneck flask was placed inside a cold bath at ∼4◦ C, and two sensors measuring the pH, and redox potential, were inserted in the two necks. First, the flask was filled with 100 mL of the polymer solution of concentration 10 g/L, and 40 mL of deionized and degased water. In order to keep anoxic conditions, a steady flow of nitrogen was injected in the solution during the entire procedure. The solution was stirred steadily at ∼800 rpm, and 10 mL of deionized and degased FeSO4 . 7H2 O solution of concentration 100 g/L were added in it. After 15 min, a yellowish or green solution evolved. Afterwards, 50 mL of a fresh NaBH4 solution of concentration 5.4 g/L was added in the flask at rate of 5 mL/min Gradually, the solution became black indicating the generation of nZVI suspension. The stirring was continued for 10 min to allow all bubbles of evolved hydrogen to release. Then, the suspension was undergone vacuum filtration, degassed and finally stored in amber glass vials which were filled with N2 on top and closed tightly. A series of tests confirmed the results of [25,26] that the average size of nanoparticles decreases with the temperature decreasing, and this was attributed to the very slow reaction rate at the lowest temperature tested. 2.2. Synthesis of liposomes and membrane integrity tests Phosphatidyl choline [egg lecithin] (PC) and 1,2-distearoylsn-glycerol-3-phosphatidylcholine (DSPC) were purchased from LIPOID AG (Germany), and Cholesterol (Chol) was purchased from Sigma-Aldrich. Typically, lipid solutions were prepared by using the thin film hydration method [20]. For this, the lipids were initially solvated in a chloroform/methanol (2:1 v/v) mixture, at concentrations between 10 and 20 mg lipid/ml of organic solvent, and once the lipids were thoroughly mixed in the organic solvent, the solvent was removed to yield a lipid film. For small volumes of organic solvent (<1 mL), the solvent was evaporated using a dry nitrogen or argon stream under a fume hood. For larger volumes, the organic solvent was removed by rotary evaporation deriving a thin lipid film on the sides of a round bottom flask. The lipid film was thoroughly dried to remove residual organic solvent by placing the vial or flask on a vacuum pump overnight. Hydration of the dry lipid film was accomplished simply by adding an aqueous medium to the container of dry lipid followed by agitation. The temperature of the hydrating medium was above the gel-liquid crystal transition temperature, Tc , of the lipid. After the addition of the hydrating medium, the lipid suspension was maintained above the Tc during the hydration period. Disruption of lipid suspensions using sonic energy (sonication) typically produces small, unilamellar vesicles

Fig. 2. TEM image of liposome encapsulating a suspension of CMC-coated nZVI.

(SUV) with diameters in the range of 15–150 nm. The mean size and size distribution of liposomes are influenced by the composition & concentration, temperature, sonication time & power, volume, and sonicator tuning. The liposomes (average size = 150 nm) were prepared in two buffers, (i) 10 mM PBS (NaH2 PO4 1.44 g/L, KH2 PO4 0.24 g/L, NaCl 8 g/L, KCl 0.2 g/L, pH = 7.4) and (ii) 1 mM NaHCO3 (pH = 7.1), at a concentration of 1 mg/mL. The following types of liposomes were tested with respect to their membrane integrity: 1) 1,2-distearoyl-sn-glycero-3-phosphocholin/Cholesterol, DSPC/CHOL (2:1, mole/mol) 2) Phosphaditidyl choline/Cholesterol, PC/CHOL (2:1, mole/mol) 3) Phosphaditidyl choline, PC A small amount of a hydrophilic fluorescent dye (calcein) was encapsulated in liposomes during their formation, and fluorescence spectroscopy was used to detect the fluorescence intensity, FI, (␭ex = 470 nm, ␭em = 520 nm, 25 ◦ C) which was proportional to the quantity of calcein that has escaped from liposomes due to their membrane disruption. The latency percentage (L) was calculated by using the FI measurements, according to L=

c f FIAT − FIBT cf FIAT

X100

(1)

where FIBT and FIAT are the measured FI values before and after the addition of Triton X-100 which is a nonionic surfactant used to disorganize liposomal dispersions, and cf = 1.1 is a correction coefficient due to dilution (after Triton addition). Therefore, the latency, L, can be regarded as a measure of the percentage of undisrupted liposomes. To examine the integrity of liposome membranes under conditions simulating those prevailing in soils, two types of experimental tests were carried out: (1) batch tests; (2) continuous flow tests in a glass etched pore network as it is explained in extent below. The liposome stability batch studies were done for two lipid (liposome) concentrations, 1 g/L and 3 g/L, at 25 ◦ C for 48 h. To assess the effect of oleic phase on the liposomal membrane integrity, the batch studies were iterated with the presence of some drops of n-dodecane. 2.3. Encapsulation of CMC-coated nZVI in liposomes To encapsulate CMC-coated nZVI in liposomes (PC/Chol, molar ratio = 4/1), the following procedure was adopted [20]. Multilamellar vesicle (MLV) liposomes were prepared by the thin film

714

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

Table 1 Properties of the glass-etched pore network. Width x Length (cm)

10.3 × 14.5

Topology Pore orientation with respect to the main flow direction Porosity Total pore cross-sectional area (m2 ) Mean pore width (␮m) Standard deviation of pore width distribution (␮m) Mean pore depth (␮m) Standard deviation of pore depth distribution (␮m) Pore length, Lp (␮m) Absolute permeability, k (Da)

Square lattice ␪ = 45◦ 0.65 5.9 × 10−6 774 274 125 20 1840 19.6

hydration method and hydrated with 1 mL of distilled water. The liposome dispersion was sonicated for 15 min using a probe sonicator (Sonics and Materials) to create small unilamellar liposome vesicles (SUV). Then, the SUV were freeze dried and rehydrated in presence of the CMC-coated nZVI suspension. The rehydration was performed in three steps, and under anoxic conditions to prevent the oxidation of nZVI. First, 100 ␮L of CMC-coated nZVI suspension was added and the sample was incubated for 30 min in 40 ◦ C; in a second phase, 100 ␮L of distilled water was added and the sample was incubated for 30 min in 40 ◦ C; finally, 800 ␮L of distilled water was added in the sample which was further incubated for 30 min in 40 ◦ C. The size of dried rehydrated vesicles (DRVs) was decreased by extrusion (20 times) through 2 stacked polycarbonate membranes (pore size 100 nm) in a extruder system (Avestin) and the vesicles produced by this method are referred to as extruded-DRVs (Fig. 2). As evidenced by TEM morphological studies, performed immediately after extruded-DRV preparation, the CMC-coated nZVI nanoparticles are entrapped in the liposomes, the liposome-membrane having the role of a physical barrier (Fig. 2). The concentration of lipids was measured by the Stewart assay [27] with UV-spectroscopy (Shimatzu UV mini 1240). The volume-based particle size distribution for all types of suspensions was measured with ␨-Nanosizer (MALVERN). 2.4. Visualization studies of suspension flow in glass-etched pore networks The mobility of the various types of nZVI suspensions was tested with visualization flow experiments performed on a glass-etched pore network (Fig. 3a, Table 1) used in earlier multiphase transport studies [26,28,29]. The capacity of CMC-coated nZVI to remediate a pore network from trapped ganglia of per-chloro-ethylene (PCE) has been demonstrated in earlier visualization studies [26]. A schematic diagram of the experimental apparatus is shown in Fig. 3b. The entire system was placed in a thermostatted chamber to keep constant temperature and avoid any undesired changes in fluid properties. The suspension was injected in the pore network through four inlet ports and expelled from it through four outlet ports (Fig. 3a). A syringe pump (HARVARD PHD 2000) was

Fig. 3. (a) Glass-etched pore network and pore space morphology. (b) Schematic diagram of experimental apparatus.

used for the injection of nZVI suspension, and a CCD camera (PANASONIC AW-E300), connected through an image USB stick grabber to the host computer, was used to capture images and store them directly on the hard disk of the PC (Fig. 3b). The micro-model was illuminated by a transmitted light source placed beneath it. The effluent was collected automatically in the tubes of a fractional collector (ELDEX) (Fig. 3b). Initially, the pore network was fully saturated with distilled and degased water, and the nZVI suspension was injected for 3 h. Afterwards, the feed solution was replaced by distilled and degased water and the injection was continued for another 2–3 h. The total flow rate was set equal to 0.025 and 0.05 mL/min. The collection of each effluent sample lasted for 1 h. Each sample was digested by adding 0.1 N HNO3 solution and the total iron concentration was measured with atomic absorption spectroscopy (AAS). In addition, flow experiments were performed on the glassetched pore network to investigate the effect of the oleic phase (n-C12 ) on nano-liposome membrane integrity under continuous flow conditions. Initially the n-C12 was displaced by the buffer solution (PBS) under a flow rate 0.02 mL/min, until attaining a residual oil saturation. Then nano-liposomes dispersed in PBS were injected under a flow rate 0.0025 mL/min. Samples were collected from the outlet and the percentage of calcein latency was measured and compared with corresponding measurements from batch (control) tests.

Table 2 Summary of the results of batch tests. Lipids

Solution

1 mg/mL

3 mg/mL

PC CHOL PC

PBS NaHCO3 PBS NaHCO3

It does not disrupt It disrupts after 18 h It does not disrupt Moderate disruption after 24 h

It does not disrupt It disrupts after 24 h It does not disrupt It does not disrupt

PBS-C12 NaHCO3 -C12 PBS-C12 NaHCO3 -C12

It disrupts after 24 h It disrupts after 18 h It disrupts after 10 h It disrupts after 18 h

Moderate disruption after 24h It disrupts after 24 h It disrupts after 10 h It disrupts after 18 h

With n-dodecane (n-C12 ) PC CHOL PC

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

715

(a)

100

Calcein latency, L (%)

80

60

40

PC PC PC PC PC PC PC PC

20

1mg/ml PBS 1mg/ml PBS-C12 3mg/ml PBS 3mg/ml PBS-C12 1mg/ml NaHCO3 1mg/ml NaHCO3 -C12 3mg/ml NaHCO3 3mg/ml NaHCO3 -C12

0 0

10

20

30

40

50

40

50

Time, t (hr)

Caclein latency, L (%)

100

(b)

80

60

40

PC:CHOL PC:CHOL PC:CHOL PC:CHOL PC:CHOL PC:CHOL PC:CHOL PC:CHOL

20

0 0

10

1mg/ml PBS 1mg/ml PBS-C12 3mg/ml PBS 3mg/ml PBS-C12 1mg/ml NaHCO3 1mg/ml NaHCO3 -C12 3mg/ml NaHCO3 3mg/ml NaHCO3 -C12

20

30

Time, t (hr)

100

(c)

Calcein latency, L (%)

80

60

DSPC:CHOL (2:1) Δείγμα ελέγχου (DSPC:CHOL (2:1)) PC:CHOL (2:1) Δείγμα ελέγχου (PC:CHOL (2:1)) PC Δείγμα ελέγχου (PC)

40

20

0 0

5

10

15

20

Time, t (hr) Fig. 4. Transient response of the percentage of calcein retention in liposomes: (a) batch tests with PC liposomes; (b) batch tests with PC: CHOL (2:1) liposomes; (c) flow-through tests in glass-etcehd pore network for various types of liposomes (Fig. 5).

716

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

Fig. 5. Successive snap-shots of the injection of liposome dispersions in glass-etched pore network (trapped n-C12 is coloured red). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3. Results and discussion

3.2. Flow tests in glass-etched pore network

3.1. Membrane integrity tests

It is evident that the CMC-coated nZVI pass through the pore network, but it is unable to confirm visually the attachment of nano-particles on the pore-walls and their gradual deposition in the porous medium (Fig. 6a). On the other hand, it is apparent that the injection of CMC-coated nZVI encapsulated in liposomes leads to the permanent entrapment of iron (black) particle aggregates in the pore network, particularly with reference to the vicinity of inlet ports (Fig. 6b). Such an entrapment in the porous medium is more evident when a mixture of CMC-coated nZVI and liposomes is injected (Fig. 6c). When flushing the pore network with distilled water, any entrapped iron aggregates are remobilized (Fig. 6a–c). For all aforementioned experiments, the maximum iron concentration in the effluent ranges between 0.75 and 0.9, and hence a percentage 10–25% of injected iron nanoparticles are deposited on the pore-walls (Fig. 7a). Moreover, in all cases, the residual iron mass is completely withdrawn from the pore network when injecting water (Figs. 6a–c, 7a). Concerning the breakthrough curve of CMC-coated nZVI encapsulated in liposomes, only a fraction of the injected lipid is detected in the outlet, while the trapped liposomes are displaced completely when injecting water (Fig. 7b) indicating that it is still in the form of liposomes. Accounting for the corresponding low attachment of nZVI particles on the pore-walls (Fig. 7a), it turns out that either the nZVI escaped from the liposomes or only a fraction of the iron was initially encapsulated in the liposomes. The later possibility is highly likely since it is known that an encapsulation of around 10–12% (calculated as percent of initial iron/lipid concentration ratio), which was achieved for encapsulation of 00904-USPIOs (Guerbet), is considered as a high encapsulation efficiency (EE) for iron nanoparticles in nano-sized liposomes [20]. An EE close to 24%, reported for larger magnetoliposomes (with approx. mean diameter of 360 nm), may not be accurate since cryo-TEM revealed that most magnetic nanoparticles were in aggregated form and outside of MLs [30]. Thereby, it is very possible that the fraction of lipid being eluted from the pore network (Fig. 7b) may correspond to the nZVI-encapsulating liposomes. The fact that nZVI elution from the pore network in the nZVI-liposomes is not modified (compared to when eluted alone)

The liposome integrity is quantified by the transient response of calcein latency measured from batch tests. The percentage of calcein latency is substantially reduced after 20 h of incubation in dispersion medium, especially when the dispersion medium contains n-dodecane (Fig. 4a,b). Therefore, the presence of oleic phase is expected to destabilize the liposomes and speed up the disruption of membranes. Regarding the dispersion medium, we observed that the lipid composition PC:CHOL = 2:1 appear to have the greatest stability when dispersed in PBS solution (Fig. 4a,b) in agreement with an earlier study where the stabilizing effect of cholesterol on lipid membranes was emphasized [20]. In NaHCO3 solution, this behaviour is reversed since the PC lipid composition has now the greatest percentage of calcein latency (Fig. 4a,b). Based on the summary presented in Table 2, it turns out that: (1) the liposomes disrupt more easily at low concentrations (1 mg/mL); (2) the disruption of lipid membranes is favoured by the presence of NaHCO3 ; (3) the disruption of lipid membranes is favoured by the presence of n-C12 . It is evident that the integrity of liposome membranes within the porous medium is comparable to that measured in the batch (control) tests under identical conditions, whereas no liposome entrapment in the pore network was observed (Fig. 4c). It’s worth mentioning that the liposomes stimulate the displacement of nC12 from the pore space (Fig. 5). The snap-shots at t = 0, show the residual n-C12 remaining in the pore network after its displacement by PBS (Fig. 5). At next steps, the liposomes enter the pore network and may gradually displace the oleic phase either in part or totally (Fig. 5). As mentioned in the introduction, liposomes consist of amphiphilic molecules (mainly phospholipids) that have the tendency to create lamellar bilayer membranes. It may thus be thought that during advanced stages of the flow process, the disruption of lipid membranes leads to the generation of (mixed) micelles that can adsorb on the oil/water interfaces, and consequently decrease the interfacial tension, resulting in the mobilization and/or solubilisation of trapped oil ganglia (Fig. 5).

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

717

Fig. 6. Snap-shots of the injection of an aqueous suspension in the pore network and the subsequent injection of distilled water, all at flow rate Q = 0.05 mL/min. (a) CMC-coated nZVI; (b) CMC-coated nZVI encapsulated in liposomes; (c) CMC-coated nZVI mixed with empty liposomes.

(Fig. 7a), indicates that liposomal-nZVI’s retain, at least in part, their mobility. Regarding the mixture of CMC-coated nZVI and empty liposomes, it is evident that all liposomal lipid is trapped in the pore network during the injection phase, and released when injecting

water (Fig. 7c). The different lipid elution profiles between CMCcoated nZVI encapsulating liposomes and their simple mixtures (Fig. 7b,c) indicates that the entrapment of nZVI in the liposomes modifies their flowing characteristics.

718

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

CFe,in=0.95 g/L CFe,in=0.48 g/L

0.8

CFe,in=0.83 g/L

*

Dimless concentration, CFe =CFe/CFe,in

1.0

0.6 0.4

(a) nZVI suspension nZVI encapsulated in liposomes Mixture of nZVI and empty liposomes

0.2 0.0 0

5

10

15

20

25

1.0

Clip,in=2.8 g/L

0.8

CMC-coated nZVI encapsulated in liposomes

Suspension flow

*

Dimless concentration, Clip =Clip/Clip,in

Number of Pore Volumes

Distilled water flow

0.6 0.4

(b )

0.2 0.0 0

5

10

15

20

25

Pore volumes injected, PV Clip,in=1.5 g/L

3.5

Mixture of CMC-coated nZVI and liposomes

3.0

*

Dimless concentration, Clip =Clip/Clip,in

4.0

2.5 Distilled water flow

Suspension flow

2.0 1.5

(c)

1.0 0.5 0.0 0

5

10

15

20

25

Pore volumes injected, PV Fig. 7. (a) Transient response of the iron concentration in the effluent of the glass-etched pore network during the injection of the three types of suspensions (Figs. 6a–c) and the subsequent injection of distilled water at flow rate 0.05 mL/min (pore volume: Vp = 0.85 mL). (b) Transient response of the lipid concentration in the effluent of the glass-etched pore network during the injection of CMC-coated nZVI encapsulated in liposomes and subsequent injection of distilled water (Fig. 6b). (c) Transient response of the lipid concentration in the effluent of the glass-etched pore network during the injection of a mixture of CMC-coated nZVI and empty liposomes (Fig. 6c) and subsequent injection of distilled water.

3.3. Particle size distribution The transient evolution of the PSD for the suspension of CMCcoated nZVI encapsulated in liposomes is shown in Fig. 8. Initially, the suspension is characterized by a clearly bimodal particle size distribution (Fig. 8a). The component distribution function of small sizes corresponds to CMC-coated nZVI that either were not encapsulated in liposomes or escaped from liposomes (Fig. 8a) and perhaps also to small nZVI-encapsulating liposomes. The component distribution function of large sizes corresponds to liposomes with or without encapsulated nZVI (Fig. 8a). Since the liposomes

were extruded through polycarbonate filters of 100 nm pore size, it is evident from the particle size distributions measured at time t = 0, that vesicle aggregation occurred (Fig. 8a). During the flow of the suspension of CMC-coated nZVI encapsulated in liposomes, in the effluent, the contribution fraction of small particles (CMCcoated nZVI) to the overall PSD increases, indicating that only a fraction of the large particles, attributed to aggregated liposomes, escape from the porous medium (Fig. 8b–d). When injecting distilled water, progressively the trapped liposomes are remobilized and exit from the network (Fig. 7b), and the overall PSD, attributed only to liposomes, becomes narrower (Fig. 8e–f).

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722 0.8

t=0 min

t=70 min

DLS-measured Bimodal log-normal fitted (μ=854.2 nm , σ=279.0 nm)

2.0

Density probability, dF/d(lnDp)

Density probability, dF/d(lnDp)

2.5

μ2=922.9 nm σ2=164.4 nm c2=0.9183

1.5

1.0

μ1 =82.6 nm

(a)

σ1 =11.9 nm 0.5

c1=0.0817

0.0 10

100

1000

(b)

0.6

σ2=260.0 nm

0.4

μ1=282.3 nm

0.3

σ1=245.3 nm

σ1=8.5 nm

1.0

c1 =0.483 0.1 0.0

10000

100

10

c1=0.5755

(c)

0.6 μ1 =306.16 nm σ1=82.8nm

0.4

c1=0.4245 0.2 0.0 10

100

1000

t=190 min

σ1=18.4 nm

c1 =0.6565

(d)

0.2

0.0 10

100

Probability density, dF/d(lnDp)

Probability density, dF/d(lnDp)

1000

10000

Particle diameter, Dp (nm)

3.0

(e)

2.0 DLS-measured Unimodal log-normal fitted (μ=435.1 nm , σ=59.3 nm)

1.0 0.5 0.0 1000

σ1=229.7nm

0.4

t=270 min 2.5

μ1=459.9 nm

c1=0.3435

10000

3.0

100

10000

DLS-measured Bimodal log-normal fitted (μ=324.8 nm , σ=263.9 nm)

μ1 =66.7 nm

0.6

Particle diameter, Dp (nm)

10

1000

Particle diameter, Dp (nm)

DLS-measured Bimodal log-normal fitted (μ=154.6 nm , σ=141.1 nm)

0.8

c2=0.517

0.2

Probability density, dF/d(lnDp)

Probability density, dF/d(lnDp)

μ1 =42.8 nm

μ2=618.7 nm

0.5

0.8 1.2 t=130 min

DLS-measured Bimodal log-normal fitted (μ=456.3 nm , σ=303.7 nm)

0.7

Particle diameter, Dp (nm)

1.5

719

10000

Particle diameter, Dp (nm)

t=320 min

2.5

(f)

DLS-measured Unimodal log-normal fitted (μ=398.8 nm , σ=46.96 nm)

2.0 1.5

(f)

1.0 0.5 0.0 10

100

1000

10000

Particle diameter, Dp (nm)

Fig. 8. Temporal evolution of the measured particle size distribution in the effluent of porous medium for (a–d) the injection of suspension of CMC-coated nZVI encapsulated in liposomes, (e–f) susbsequent injection of distilled water.

When the mixture of CMC-coated nZVI and empty liposomes was injected in the pore network, the above mentioned behaviour became more evident. (Fig. 9a–d). During the mixture injection, the overall PSD measured in effluent is associated exclusively with the small size particle fraction (that in this case is particularly smaller compared to the previous case of the nZVI-encapsulating liposome sample), which is due to CMC-coated nZVI (Fig. 9b–d). Moreover, the liposomes, initially trapped in the porous medium, eventually coalesced by generating very large particles (Fig. 9e) which were released by injecting distilled water (Fig. 9e,f). It is well known that no liposomal-lipid was detected in the effluent during the injection of the mixture and only when water was injected, all trapped liposomes were released (Fig. 7c). The very large sizes of remobilized liposomes (Fig. 9e–f) are indicative of the coalescence, aggregation or perhaps fusion of smaller liposomes that most probably occurs in the porous medium. It is not clear if this increase of liposomes size increase is initiated or even caused by the presence of the CMC-

coated nZVI, a fact that should be investigated in future studies. However, the fact that liposome integrity is not influenced when liposomes are eluted alone (Fig. 4c) clearly indicates a predominant role of the CMC-coated nZVI’s in this phenomenon. The transient changes of the statistical moments of the particle size distribution measured in the outlet of the porous medium when injecting the suspension of the various types of nanocomposites is shown in Fig. 10. When injecting CMC-coated nZVI, the mean size of nanoparticles collected in the outlet decreases weakly during the suspension flow, due to the deposition of the largest particles, and increases during the distilled water flow, due to the detachment of deposited particle aggregates (Fig. 10a,b). However, when injecting CMC-coated nZVI either encapsulated in liposomes (Fig. 10c,d) or mixed with empty liposomes (Fig. 10e,f), the observed dynamic changes of the size distribution become much stronger due to the gradual entrapment and subsequent remobilization of liposomes (Fig. 7b,c).

720

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

Density probability, dF/d(lnDp)

t=0 min

DLS-measured Trimodal log-normal fitted (μ=204.3 nm , σ=295.5 nm)

1.2 1.0

μ1=37.7 nm σ1=8.85 nm

0.8

c1 =0.492

0.6

μ1=110.7 nm

0.4

c1 =0.3

μ2=731.5 nm

σ1=81.8 nm

(a)

σ2=229.2 nm c2 =0.208

0.2

Density probability, dF/d(lnDp)

1.4

0.0 1

10

100

1000

t=70 min

1.4 1.2

μ1 =19.1 nm 1.0

σ1=4.9 nm c1=0.8

0.8 0.6

μ2=101.5 nm 0.4

0.0 1

10

Density probability, dF/d(lnDp)

Density probability, dF/d(lnDp)

σ1=4.7 nm c1=0.83

0.8

DLS-measured Bimodal log-normal fitted (μ=74.4 nm , σ=104.4 nm)

0.6 μ2 =136.5 nm

0.4

σ2 =130.9 nm c2=0.17

(c)

0.2 0.0 1

10

100

1000

10000

1.0

μ1 =33.3 nm

0.8

c1=0.714

0.6 μ2=177.2 nm 0.4

σ2=152.1 nm

0.0

3.0 μ2=1489.7 nm c2=0.808

1.4

DLS-measured Bimodal log-normal fitted (μ=1248.8 nm , σ=553.6 nm)

1.2 1.0 0.8 0.6

μ1=235.5 nm

0.4

σ1=138.1 nm c1=0.192

(e)

0.2 0.0 1

10

100

1000

c2=0.286

(d)

0.2

10

100

1000

10000

Particle diameter, Dp (nm)

σ2=269.7 nm

1.6

DLS-measured Bimodal log-normal fitted (μ=74.4 nm , σ=104.4 nm)

σ1 =9.1 nm

1

Density probability, dF/d(lnDp)

Density probability, dF/d(lnDp)

t=260 min

1.8

10000

t=190 min

1.2

Particle diameter, Dp (nm) 2.0

1000

100

Particle diameter, Dp (nm)

1.2 1.0

c2 =0.2

0.2

10000

t=130 min

μ1=18.9 nm

σ2=87.6 nm

(b)

Particle diameter, Dp (nm) 1.4

DLS-measured Bimodal log-normal fitted (μ=35.6 nm , σ=51.4 nm)

10000

Particle diameter, Dp (nm)

t=280 min

2.5

DLS-measured Unimodal log-normal fitted (μ=297.9 nm , σ=41.9 nm)

2.0 1.5 1.0 0.5

(f)

0.0 1

10

100

1000

10000

Particle diameter, Dp (nm)

Fig. 9. Temporal evolution of the measured particle size distribution in the effluent of porous medium for (a–d) the injection of suspension of CMC-coated nZVI mixed with empty liposomes, (e–f) and the susbsequent injection of distilled water.

Summarizing the results of the particle size distribution measured in the outlet of the porous medium we emphasize to the following. During the flow of various types of suspensions, (a) the nano-iron exits from the pore network without any significant change of its size distribution (Fig. 10a,b), (b) a fraction of liposomes encapsulating nano-iron along with the non-encapsulated nano-iron are filtered out (Fig. 10c,d), (c) only the nano-iron from its mixture with liposomes is extracted (Fig. 10e,f). During the leaching with water, (a) any deposited nano-iron detaches from the porewalls (Fig. 10a,b), (b) no sensible change is observed on the sizes of liposomes encapsulating CMC-coated nZVI (Fig. 10c,d), and (c) the coalescence and aggregation of empty liposomes, mixed with CMC-coated nZVI, leads to very large particles that are trapped in the pore network and remobilized when injecting water (Fig. 10e,f). Concerning the stability of particulate systems, the ␨-potential remains constant at a negative value (between −20 and −40), and this is due mainly to nZVI that are dominant in the effluent when any suspension is injected (Fig. 11). The negative ␨-potential

decreases respectably only at late times when almost pure water is flowing out (Fig. 11). 3.4. Discussion and perspectives To enhance the spatial distribution of nanoparticles in contaminated zone, and overcome problems associated with preferential flow in heterogeneous soils and contaminant mobilization, the potential to use foams and shear-thinning fluids for nanoparticle delivery in unsaturated zone and aquifer, respectively, has been explored [31,32]. However, in both cases the energy demand for fluid injection is expected to be quite high, due mainly to the increased viscosity of injected amendment. Although such strategies might be efficient for the remediation of dissolved pollutants, neither the foams, nor the shear-thinning fluids (e.g. guar gum) ensure that the nanoparticles will deposit on NAPL/water interfaces, facilitating the degradation of trapped NAPL source zone. From this point of view, the liposomes might be regarded as “protective” vehicles able to prevent the interactions of nZVI with solid

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

200

150

Suspension flow

(a )

100

50 Distilled water flow 0 0

60

120

180

240

300

360

420

480

Standard deviation of PSD, σ (nm)

Mean value of PSD, μ (nm)

200

Suspension flow

150

(b )

100

50

Distilled water flow

0 60

0

120

180

Time, t (min)

300

360

420

480

400

(c)

800 Distilled water flow

Suspension flow 600

400

200

0 0

60

120

180

240

300

360

420

480

Standard deviation of PSD, σ (nm)

Mean value of PSD, μ (nm)

240

Time, t (min)

1000

Suspension flow

350

(d)

300 Distilled water flow

250 200 150 100 50 0 0

60

120

Time, t (min)

180

240

300

360

420

480

Time, t (min)

(e)

1200 1000 800 600

Distilled water flow

Suspension flow

400 200 0 0

60

120

180

240

300

360

420

480

Time, t (min)

Standard deviation of PSD, σ (nm)

600

1400

Mean value of PSD, μ (nm)

721

(f)

500 400

Suspension flow

Distilled water flow

300 200 100 0 0

60

120

180

240

300

360

420

480

Time, t (min)

Fig. 10. Temporal evolution of the mean value, and standard deviation of the particle size distribution for the flow of: (a–b) CMC-coated nZVI suspension; (c–d) CMC-coated nZVI encapsulated in liposomes; (e–f) CMC-coated nZVI mixed with liposomes.

and liquid phases until lipid membrane disrupting. In an earlier study [33], the injection of uncoated nano- and sub-micrometer ZVI particles in a glass-etched pore network led to the formation of large aggregates that blocked the flow-paths and only the presence of humic acids at high concentration stimulated the particle detachment from the pore-walls. In contrast, the mobility of CMC-coated nZVI synthesized and tested in the present work was respectable, regardless of the delivery method (Fig. 7a). On the other hand, it seems that the liposomes encapsulating CMC-coated nZVI pass through the pore network (Fig. 7b). In contrast, empty liposomes, mixed with CMC-coated nZVI, are totally immobilized (Fig. 7c). From the lipid breakthrough curve (Fig. 7b) and the measured particle size distribution as a function of time (Fig. 8a–c), it becomes evident that a fraction of the liposomes encapsulating CMC-coated

nZVI (which are the largest particles) exits from the pore network, while the aggregation of entrapped liposomes is limited (Figs. 8e–f, 10c–d). On the other hand, it seems that the presence and flow of CMC-coated nZVI may trigger the coalescence and aggregation of empty liposomes with the formation of large particles which are trapped in the pore space (Figs. 7c, 9b–d, 10e–f), and remobilized when injecting water (Figs. 7c, 9e, 10e–f). In the future, more emphasis should be placed on the elucidation of the physicochemical properties that control the mobility of liposomes when co-existing with nZVI to optimize their capacity to travel long distances in pore space, and facilitate the disruption of membranes at the vicinity of NAPL/water interfaces so that the encapsulated nZVI are released, and the NAPL source zone remediation is enhanced.

722

K. Terzi et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 506 (2016) 711–722

nZVI suspension nZVI encapsulated in liposomes Mixture nZVI and empty liposomes

0

ζ-Potential (mV)

Suspension flow

Distilled water flow

-20

-40

-60 0

60

120

180

240

300

360

420

480

Pore volumes injected, PV Fig. 11. Transient changes of the ␨-potential of effluent.

4. Conclusions Stable suspensions of CMC-coated nZVI are synthesized and encapsulated in liposomes. Flow tests are performed on a transparent glass-etched pore network with large pore sizes (>100 ␮m) to assess the capacity of nZVI and liposomes to pass through the porous medium. The transient changes of the particle size distribution in the effluents are recorded and used as a tool for the interpretation of the results. The iron concentration breakthrough curves show that only a low percentage of nZVI is deposited on the pore-walls of the medium, regardless of the type of injected suspension. In contrast, only a fraction of liposomes is detected in the effluent or all liposomes are trapped in the pore network, when liposomes encapsulating CMC-coated nZVI or a mixture of empty liposomes with CMC-coated nZVI are injected, respectively. In both cases, the trapped liposomes are remobilized when injecting distilled water. Possible explanations for the later observations may be potential interactions between empty-liposomes and CMC-coated nZVI, which result in an increase of the size distribution of liposomes, or it may be that the flow of empty liposomes through the pore network is inhibited by the co-flowing CMC-coated nZVI. Acknowledgements The research is co-financed by the European Union (European Social Fund-ESF) and Greek national funds in the context of the action “EXCELLENCE II” of the Operational Program “Education and Lifelong Learning” – project no 4118 (project title: “Optimizing the properties of nanofluids for the efficient in situ soil remediationSOILREM”). References [1] N.C. Mueller, J. Braun, J. Bruns, M. Cernik, P. Rissing, D. Rickerby, B. Nowack, Application of nanoscale zero valent iron (NZVI) for groundwater remediation in Europe, Environ. Sci. Pollut. Res. 19 (2012) 550–558. [2] M. Valle-Orta, D. Diaz, P. Santiago-Jacinto, A. Vazquez-Olmos, E. Reguera, Instantaneous synthesis of stable zerovalent metal nanoparticles under standard reaction conditions, J. Phys. Chem. B 112 (2008) 14427–14434. [3] C.B. Wang, W.X. Zhang, Synthesizing nanoscale iron particles for rapid and complete dechlorination of TCE and PCBs, Environ. Sci. Technol. 31 (1997) 2154–2156. [4] J.T. Nurmi, P.G. Tratnyek, V. Sarathy, D.R. Baer, J.E. Amonette, K. Pecher, C. Wang, J.C. Linehan, D.W. Matson, R.L. Penn, M.D. Driessen, Characterization and properties of metallic iron nanoparticles: spectroscopy, electrochemistry, and kinetics, Environ. Sci. Technol. 39 (2005) 1221–1230. [5] Y.P. Sun, X.Q. Li, J. Cao, W. Zhang, H.P. Wang, Characterization of zero-valent iron nanoparticles, Adv. Colloid Interface Sci. 120 (2006) 47–56. [6] R.A. Crane, T.B. Scott, Nanoscale zero-valent iron: future prospects for an emerging water treatment technology, J. Hazard. Mat. 12 (2012) 1765–1775.

[7] L.B. Hoch, E.J. Mack, B.W. Hydutsky, J.M. Hershman, J.M. Skluzacek, T.E. Mallouk, Carbothermal synthesis of carbon-supported nanoscale zero-valent iron particles for the remediation of hexavalent chromium, Environ. Sci. Technol. 42 (2008) 2600–2605. [8] N. Saleh, H.-J. Kim, T. Phenrat, K. Matyjaszewski, R.D. Tilton, G.V. Lowry, Ionic strength and composition affect the mobility of surface-modified Fe◦ nanoparticles in water-saturated sand columns, Environ. Sci. Technol. 42 (9) (2008) 3349–3355. [9] W.X. Zhang, Nanoscale iron particles for environmental remediation: an overview, J. Nanopart. Res. 5 (2003) 323–332. [10] B.W. Hydutsky, E.J. Mack, B.B. Beckerman, J.M. Skluzacek, T.E. Mallouk, Optimization of nano-and microiron transport through sand columns using polyelectrolyte mixtures, Environ. Sci. Technol. 41 (2007) 6418–6424. [11] N. Saleh, K. Sirk, Y. Liu, T. Phenrat, B. Dufour, K. Matyjaszewski, R.D. Tilton, G.V. Lowry, Surface modifications enhance nanoiron transport and NAPL targeting in saturated porous media, Environ. Eng. Sci. 24 (2007) 45–57. [12] F. He, D. Zhao, Manipulating the size and dispersibility of zerovalent iron nano particles by use of carboxymethyl cellulose stabilizers, Environ. Sci. Technol. 41 (2007) 6216–6221. [13] Y.-P. Sun, X.-Q. Li, W.- Zhang, H.P. Wang, A method for the preparation of stable dispersion of zero-valent iron nanoparticles, Physicochem. Eng. A 308 (2007) 60–66. [14] T. Phenrat, F. Fagerlund, t. Illangasekare, G.V. Lowry, R.D. Tilton, Polymer-modified Fe0 nanoparticles target entrapped NAPL in two dimensional porous media: effect of particle concentration, NAPL saturation, and injection strategy, Environ. Sci. Technol. 45 (2011) 6102–6109. [15] J. Quinn, C. Geiger, C. Clausen, K. Brooks, C. Coon, S. O’Hara, T. Krug, D. Major, W.S. Yoon, A. Gavaskar, T. Holdsworth, Field demonstration of DNAPL dehalogenation using emulsified zero-valent iron, Environ. Sci. Technol. 39 (2005) 1309–1318. [16] J. Zhan, T. Zheng, G. Piringer, C. Day, G.L. Mcpherson, Y. Lu, K. Papadopoulos, V.T. John, Transport characteristics of nanoscale functional zerovalent iron/silica composites for in situ remediation of trichloroethylene, Environ. Sci. Technol. 42 (2008) 8871–8876. [17] T. Zheng, J. Zhan, J. He, C. Day, Y. Lu, G.L. McPherson, G. Piringer, V.T. John, Reactivity characteristics of nanoscale zerovalent iron-Silica composites for trichloroethylene remediation, Environ. Sci. Technol. 42 (12) (2008) 4494–4499. [18] J. Zhan, B. Sunkara, J. Tang, Y. Wang, J. He, G.L. McPherson, V.T. John, Carbothermal synthesis of aerosol-based adsorptive-reactive iron-carbon particles for the remediation of chlorinated hydrocarbons, Ind. Eng. Chem. Res. 50 (23) (2011) 13021–13029. [19] A. Tiraferri, R. Sethi, Enhanced transport of zerovalent iron nanoparticles in saturated porous media by guar gum, J. Nanopart. Res. 11 (2009) 635–645. [20] A. Skouras, S. Mourtas, E. Markoutsa, M.-C. de Golstein, C. Wallon, S. Catoen, S.G. Antimisiaris, Magnetoliposomes with high USPIO entrapping efficiency, stability and magnetic properties, Nanomed.: Nanotechnol. Biol. Med. 7 (2011) 572–579. [21] S.G. Antimisiaris, Encycl. Nanosci. Nanotechnol., H.S. Nalwa (Ed.), Am. Sci. Publ. 18, 455–487 (2011). [22] S.G. Antimisiaris, Preparation of DRV liposomes Methods in Molecular Biology, 605, Clifton, N.J, 2016, pp. 51–75, ISBN: 978-1-60327-359-6. [23] F. He, M. Zhang, T. Qian, D. Zhao, Transport of carboxymethyl cellulose stabilized iron nanoparticles in porous media: column experiments and modeling, J. Colloid Interface Sci. 334 (2009) 96–102. [24] F. He, D. Zhao, C. Paul, Field assessment of carboxymethyl cellulose stabilized iron nanoparticles for in situ destruction of chlorinated solvents in source zones, Water Res. 44 (2010) 2360–2370. [25] J. Zhang, B. Sunkara, L. Le, V.T. John, J. He, G.L. Mcpherson, G. Piringer, Y. Lu, Multifunctional colloidal particles for in situ remediation of chlorinated hydrocarbons, Environ. Sci. Technol. 43 (2009) 8616–8621. [26] C. Tsakiroglou, K. Terzi, A. Sikinioti-Lock, K. Hajdu, C. Aggelopoulos, Assessing the capacity of zero valent iron nanofluids to remediate NAPL-polluted porous media, Sci. Total Environ. (2016), http://dx.doi.org/10.1016/j.scitotenv. 2016.02.010 (in press). [27] J.C.M. Stewart, Colorimetric determination of phospholipids with ammonium ferrothiocyanate, Anal. Biochem. 104 (1980) 10–14. [28] M.A. Theodoropoulou, V. Karoutsos, C. Kaspiris, C.D. Tsakiroglou, A new visualization technique for the study of solute dispersion in model porous media, J. Hydrol. 274 (2003) 176–197. [29] M.A. Theodoropoulou, V. Sygouni, V. Karoutsos, C.D. Tsakiroglou, Relative permeability and capillary pressure functions of porous media as related to the displacement growth pattern, Int. J. Multiph. Flow 31 (2005) 1155–1180. [30] M.-S. Martina, J.-P. Fortin, C. Ménager, O. Clément, G. Barratt, C. Grabielle-Madelmont, et al., Generation of superparamagnetic liposomes revealed as highly efficient MRI contrast agents for in vivo imaging, JACS 127 (2005) 10676–10685. [31] L. Zhong, J. Szecsody, M. Oostrom, M. Truex, X. Shen, X. Li, Enhanced remedial amendment delivery to subsurface using shear thinning fluid and aqueous foam, J. Hazard. Mat. 191 (2011) 249–257. [32] X. Shen, L. Zhao, Y. Ding, B. Liu, H. Zeng, L. Zhong, X. Li, Foam, a promising vehicle to deliver nanoparticles for vadose zone remediation, J. Hazard. Mat. 186 (2011) 1773–1780. [33] Q. Wang, J.-H. Lee, S.-W. Jeong, A. Jang, S. Lee, H. Choi, Mobilization and deposition of iron nano- and sub-micrometer particles in porous media: a glass micromodel study, J. Hazard. Mat. 192 (2011) 1460–1475.