The growing use of nanoscale zero-valent iron (NZVI) in the remediation of contaminated groundwater raises concerns regarding its transport in aquifers. Laboratory-scale sand-packed column experiments were conducted with bare and sucrose-modified NZVI (SM-NZVI) to improve our understanding of the transport of the nanoparticles in saturated porous media, as well as the role of media size, suspension injection rate and concentration on the nanoparticle behavior. As the main indicative parameters, the normalized effluent concentration was measured and the deposition rate coefficient (k) was calculated for different simulated conditions. Overall, compared to the high retention of bare NZVI in the saturated silica column, SM-NZVI suspension could travel through the coarse sand column easily. However, the transport of SM-NZVI particles was not very satisfactory in a smaller size granular matrix especially in fine silica sand. Furthermore, the value of k regularly decreased with the increasing injection rate of suspension but increased with suspension concentration, which could reflect the role of these factors in the SM-NZVI travel process. The calculation of k-value at the tests condition adequately described the experimental results from the point of deposition dynamics, which meant the assumption of first-order deposition kinetics for the transport of NZVI particles was reasonable and feasible.

INTRODUCTION

Injection of nanoscale zero-valent iron (NZVI) is certified as one of the effective technologies for the remediation of contaminated groundwater aquifers. Compared to the traditional in situ remediation technologies, iron nanoparticles possess some distinct advantages of mild reaction condition, rapid reaction rate and especially the widespread applicability that is not limited by the depth of the pollutants (Jiao et al. 2009). However, bare nano-iron particles (B-NZVI) tend to reunite due to their higher surface energy (Cheng et al. 2007); consequently, the reactive ability of the nano-iron will weaken significantly. Attempts have been made to modify the nano-iron using various dispersants/stabilizer by many researchers (Saleh et al. 2005; Phenrat et al. 2007). Sucrose is another green material that has been proved effective for the dispersion of nanoparticles.

In addition to improving the dispersion of nanoparticles, enhancing their transport in the subsurface environment is another important objective of surface modification for nanoparticles. Forming aggregates may block and inhibit the nanoparticle transport through saturated porous media (Zhang et al. 2002), so that the particles hardly contact with the pollution source/plume, thus leading to an unsatisfactory remediation effect. In this regard, many researchers have made efforts to enhance the transport of nanoparticles in saturated porous media. Yang et al. (2007) concluded that PAA (polyacrylic acid) modified nano-iron slurry could move far more effectively through a sand column, and compared to unmodified NZVI, its sticking coefficient decreased to 0.1078 from 0.5628 implying a longer transport distance for PAA-NZVI. Alberto & Rajandrea (2009) selected guar gum as stabilizer for nano-iron to finish their migration test. They found guar gum was able to ensure nanoparticle transport in complete contrast to basically immobile bare nano-iron in sandy porous media. Very recently, Bahngmi et al. (2014) reported that the stability of NZVI was increased through adsorption of polymers due to the enhanced steric stabilization among the nanoparticles; thus their transport ability in porous media was significantly strengthened.

Several studies have demonstrated that both nanoparticle concentration and grain size of media can strongly affect the deposition and transport of colloidal particles in porous media under various conditions. Wang et al. (2012) investigated the effect of input concentration on the transport of silica nanoparticles in porous media; the results indicated that lower particle input concentration would bring higher deposition on the surface of media. Similarly, the study from Liang et al. (2013) showed that silver nanoparticles in porous media increased at higher input concentrations. These effects of input concentration of colloid particle were explained by the adsorption and blocking of particles reaching the maximum deposition sites of the matrix (Bradford & Bettahar 2006). In regard to the influence of media, Polyxeni et al. (2013) investigated the combined effect of ionic strength and sand grain size on colloid fate ad transport in unsaturated porous media, and the results suggested that relatively smaller attachment was observed onto fine than onto medium quartz sand under the same ionic strength. Unlike the condition in unsaturated environments, Vasiliki & Constantinos (2011) found that no obvious relationships between mass recoveries of three biocolloids and grain size could be established in water-saturated porous media, from the transport experiments. The study from Tian et al. (2011) also obtained the similar conclusion that grain size had little effect on the transport behavior of modified carbon nanotubes in saturated porous media. However, as well as general characteristic of nanoscale particle size, the magnetic property and larger specific gravity of the nano-iron particles determine easier aggregation and deposition in porous media of underground aquifers. Therefore, the granule media and input concentration of the slurry may have different effect, and more investigation is needed directed toward understanding the transport and deposition of bare or modified NZVI particles. Furthermore, the injection rate of the slurry may take an important role in the transport process of nanoparticles, which should be given more attention.

The overall objective of this work was to confirm the enhanced transport ability of modified NZVI with sucrose in saturated porous media, and determine the effect of media size, input concentration and injection rate of the suspension. Laboratory columns packed with quartz sand of three grain sizes (0.1–0.25, 0.25–0.5 and 0.5–1.0 mm) were used as experimental porous media. Sucrose-modified NZVI (SM-NZVI) suspension at the concentration range of 1.82–9.72 g/L was introduced to the columns at different rate, and effluent concentration data were obtained for breakthrough curves and deposition kinetics analysis.

MATERIALS AND METHODS

Nanoparticle slurry preparation

The bare nano-iron (70–80 nm by transmission electron microscopy, TEM) and sucrose-modified nano-iron (100–150 nm) particles were synthesized by chemical reduction method in aqueous solutions. The chemical and physical properties of bare and sucrose-modified nano-iron, as well as their specific synthesis procedure, have been described in our previous study (Li et al. 2013). A portion of prepared slurry was diluted in deionized (DI) water purged with N2 gas to obtain a 2 g/L stock suspension. It was subsequently agitated under sonication for 20 min to break any aggregates that might exist, and was then stored in an anaerobic environment. All chemicals used in the preparation process, including iron sulfate heptahydrate (FeSO4·7H2O), potassium borohydride (KBH4), sucrose and ethanol, were purchased from Technology Development Corporation (Tianjin, China), and they were analytical pure grade (99+%) and used without further purification.

Matrix of groundwater aquifer

The quartz sand supplied by Jinheng Commercial Corporation was used as the porous matrix of the groundwater aquifer. The sand was uniform with diameter range of 18–35 mesh (0.5–1 mm), 35–60 mesh (0.25–0.5 mm) and 60–140 mesh (0.1–0.25 mm), respectively. Prior to packing the column, the sand was acid washed as previously described (Pelley & Tufenkji 2008) to remove organic and inorganic impurities. The columns were uniformly packed wet in 1 cm increments and tamped to remove air bubbles, giving the column porosity as shown in Table 1.

Table 1

Physical parameters of silica columns in migration experiment

Packed media Media size, diameter (mm) Packing mass (g) Packing density (g/cm3Column porosity 
Coarse sand 0.5–1 94.5 1.50 0.48 
Medium sand 0.25–0.5 95.2 1.51 0.48 
Fine sand 0.1–0.25 95.5 1.52 0.45 
Packed media Media size, diameter (mm) Packing mass (g) Packing density (g/cm3Column porosity 
Coarse sand 0.5–1 94.5 1.50 0.48 
Medium sand 0.25–0.5 95.2 1.51 0.48 
Fine sand 0.1–0.25 95.5 1.52 0.45 

Column test

Column experiments were conducted to investigate the transport ability of SM-NZVI under different conditions. Polymethyl methacrylate columns, with an inner diameter of 5 cm were wet-packed to a height of 20 cm. Wall effects were assumed to be insignificant given a column-to-grain diameter ratio above 20. DI water was drawn into the column using a peristaltic pump to saturate the packed sand and then 4–5 pore-volume (PV) of nanoparticle suspension was introduced from the bottom of the column. The solution of each test was applied at a certain constant flow rate, and the effluent solution was collected continuously using a fraction collector and monitored for iron concentration by phenanthroline spectrophotometric method. When introduction of the nanoparticle solution ended, the sand column was flushed with DI water until there no iron concentration could be detected in the sample solution. After completing the transport process, the column was divided into sections for exploring the deposition of B-NZVI and SM-NZVI in the sand. The bottom end-piece was removed without disturbing the packed bed and the sand was removed at 2 cm intervals. The iron in each section was digested with 4 mL HCl (1:1 volume ratio of HCl:DI water) and total iron concentration was measured in the spectrophotometer.

Four groups of experiments were conducted to compare the transport ability of bare NZVI and SM-NZVI, and examine the effect of porous media size, the injection rate and concentration of the slurry on the mobility of SM-NZVI, respectively. Table 2 summarizes the experimental conditions for column operation. The environment of the packed bed and iron solution was adjusted to neutral pH.

Table 2

Iron nanoparticle transport experiments conducted in saturated glass columns packed with silica sand

Influence factors   I.T. M.S. I.C. I.R. PV 
Iron type (I.T.) NZVI   0.5–1 4.37 0.042 
SM-NZVI 
Media size (M.S., mm) 0.1–0.25 SM-NZVI   4.42 0.042 
0.25–0.5 
0.5–1 
Injection rate (I.R., cm/s) 0.027 SM-NZVI 0.5–1 4.53   
0.032 
0.042 
0.053 
0.069 
0.08 4.5 
0.096 3.5 
0.106 
Input concentration (I.C., g/L) 1.82 SM-NZVI 0.5–1   0.042 
3.79 
5.91 
7.76 
9.72 
Influence factors   I.T. M.S. I.C. I.R. PV 
Iron type (I.T.) NZVI   0.5–1 4.37 0.042 
SM-NZVI 
Media size (M.S., mm) 0.1–0.25 SM-NZVI   4.42 0.042 
0.25–0.5 
0.5–1 
Injection rate (I.R., cm/s) 0.027 SM-NZVI 0.5–1 4.53   
0.032 
0.042 
0.053 
0.069 
0.08 4.5 
0.096 3.5 
0.106 
Input concentration (I.C., g/L) 1.82 SM-NZVI 0.5–1   0.042 
3.79 
5.91 
7.76 
9.72 
Table 3

Mass balance calculations of B-NZVI and SM-NZVI in transport process

Type of iron Accumulated outflow (g) Deposition on sand (g) Total mass (g) Percentage of outflow 
NZVI 171.51 536.01 707.52 24 
SM-NZVI 457.78 265.99 723.77 63 
Type of iron Accumulated outflow (g) Deposition on sand (g) Total mass (g) Percentage of outflow 
NZVI 171.51 536.01 707.52 24 
SM-NZVI 457.78 265.99 723.77 63 

RESULTS AND DISCUSSION

Transport behavior of B-NZVI and SM-NZVI particles in sand-packed columns

The hypothesis that the effluent of B-NZVI and SM-NZVI through the sand column would reflect their transport capability in porous media was studied in this work. Representative breakthrough curves for bare and sucrose-modified NZVI particles are shown in Figure 1(a). C/C0 (C: iron concentration of effluent through column; C0: initial concentration of inflow) was calculated as the normalized effluent concentration obtained from the breakthrough curves during the initial phase of particle elution. It can be seen that the sampling concentration gradually increased to a maximum value of 0.32 and 0.67 corresponding to B-NZVI and SM-NZVI respectively. Furthermore, the relative clear suspension of B-NZVI was observed to pass through the sand and most of the particles were retained on the bottom of the column. Unlike the non-ideal passing of B-NZVI particles, the transport of SM-NZVI through the sand column was much easier and the particle deposition in sand reduced significantly as seen in Figure 1(b). The mass balance of the iron during the transport process was calculated (Table 3), and the mass ratio of effluent and influent of B-NZVI and SM-NZVI was 0.24 and 0.63, respectively, which meant that the transport ability of SM-NZVI increased by 1.6 times compared with bare nanoparticles due to sucrose modification.

Figure 1

Breakthrough curves and deposition conditions of B-NZVI and SM-NZVI.

Figure 1

Breakthrough curves and deposition conditions of B-NZVI and SM-NZVI.

Colloid particles tend to deposit on the surface of porous media due to gravitation and filtration resulting from particle aggregation, which means that particle aggregation can significantly affect the transport behavior of nanoparticles. Many previous studies about NZVI transport have found that the mobility of polymer-coated NZVI was obviously better than that of bare NZVI (Kim et al. 2012), and also confirmed that the space resistance and the electrostatic repulsion between the nanoparticles could be strengthened by modifiers (Franchi & O'Melia 2003). As well as decreased nanoparticle aggregate formation due to existing modifiers, electrostatic stabilization results in decreased deposition onto the sand surface, thus also reducing the likelihood of particle retention in the porous media. The test results mentioned above adequately show that sucrose modification can increase the transport ability of NZVI; meanwhile, sucrose can be more easily utilized by microbes in underground water as a carbon source, due to its relatively small molecular weight.

Effect of media size on SM-NZVI transport

In order to investigate the effect of the media size on SM-NZVI transport behavior, silica sand with diameter range of 0.5–1, 0.25–0.5 and 0.1–0.25 mm were chosen as porous media of the simulated aquifer. The breakthrough curves shown in Figure 2 indicated that the sand grain size had a strong influence on the amount of SM-NZVI transport. In comparison with the coarse and medium sand, nearly no breakthrough of SM-NZVI in the fine sand column was observed and thus the outflow was defined as zero. The mobility of SM-NZVI in coarse and medium sand columns was higher with C/C0 value of 0.67 and 0.45, respectively. The results indicate that the retention of SM-NZVI at a given C0 tended to increase with decreasing sand size, which is consistent with many previous studies about the transport behavior of other nanoparticles such as silver (Liang et al. 2013) and carbon nanotubes (Kasel et al. 2013). The possible reason for this situation may be that a larger pore space exists in the porous media with larger diameter, and thus the nanoparticles can pass through the sand more easily instead of being deposited. Moreover, the specific surface area of the porous media is improved with its smaller size, which therefore results in the stronger adsorption for SM-NZVI. Classical colloid filtration theory from Yao et al. (1971) has also predicted that the increased rate of mass transfer to the collector surface occurs as the grain size decreases, and the prediction is likewise applicable to SM-NZVI particles obviously. The findings about media size effect imply that corresponding measures should be taken for the increased migration distance of SM-NZVI in the underground environment with close-grained media, thereby assuring ideal remediation effect for SM-NZVI applied.

Figure 2

Breakthrough curves of SM-NZVI in different size media.

Figure 2

Breakthrough curves of SM-NZVI in different size media.

Effect of slurry injection rate on SM-NZVI transport

In a practical application, NZVI suspension is generally injected to underground aquifers through injection wells; thus the migration and distribution of the particles in porous media will be affected by suspension injection rate. Figure 3 gives the breakthrough curves of SM-NZVI in the coarse sand column at different injection rate. It indicates that the effluent SM-NZVI tended to reach maximum concentration in less and less time with the increase of suspension injection rate, resulting in the increase of the corresponding value of C/C0 from 0.25 to 0.85, while, as shown in Figure 3(b), the recovery concentration achieved almost the same maximum value when suspension flow velocity increased to 0.106 cm/s. In the light of previous studies, it has been found that greater hydrodynamic shear force could reduce the deposition of colloidal particles on the surface of porous media (Chowdhury et al. 2011), and in consequence the nanoparticles were anticipated to migrate with longer distance. However, the results mentioned above show that the recovery concentration of SM-NZVI did not increase continuously. Therefore, we can draw the conclusion that there exists a maximum critical value of suspension injection rate affecting the transport of SM-NZVI particles; increasing the injection rate to lower than the critical value facilitates the mobility of SM-NZVI dramatically, while little impact could be observed once the inflow rate increases to above this value. The colloid deposition on the surface of media in the transport process includes two aspects: the migration and adsorption of colloid particles onto the matrix surface. The above results imply that large enough shear force can be brought by increasing the flow rate to overcome the adsorption force between colloid and consequently promote the mobility of colloid particles.

Figure 3

Breakthrough curves of SM-NZVI at different injection rate.

Figure 3

Breakthrough curves of SM-NZVI at different injection rate.

The fitted curves and equations were obtained for expressing the relationship between injection rate and migration capability (M) of SM-NZVI, as well as the time required for maximum recovery concentration (Figure 4). The curve in Figure 4(a) shows that M increased with the increasing injection rate (v) until v reached 0.06 cm/s, after which the amplification of M tended to drop significantly. It means that the injection rate may be no longer the dominant factor for SM-NZVI migration; that is, the shear force derived from the rate has made the maximum impact available to conquer the surface adsorption of porous media for nanoparticles. Furthermore, as seen in Figure 4(b), less time was required for maximum value of C/C0, which also indirectly represents the enhancement of SM-NZVI transport due to increasing injection rate.

Figure 4

Fitted curves for the effect of the injection rate on migration capability and the time for maximum recovery concentration.

Figure 4

Fitted curves for the effect of the injection rate on migration capability and the time for maximum recovery concentration.

Effect of input concentration on SM-NZVI transport

Additional column experiments were performed to investigate the influence of SM-NZVI input concentrations. Figure 5 presents the recovery for SM-NZVI in coarse sand at the injection rate of 0.042 cm/s. It shows that the breakthrough curves for different concentration reached a plateau at almost the same time; however, the maximum value of C/C0 decreased from 0.73 to 0.58 with the input concentration change from 1.82 to 9.72 g/L, implying that the transport ability of SM-NZVI particles decreased with the increase of input concentration. The main reasons for the results may be attributed to the pore clogging of media. Furthermore, higher C0 may lead to the collision between nanoparticles and granule media becoming significantly aggravated, which accelerates SM-NZVI precipitation on the surface of the silica sand, and the contact of SM-NZVI particles is enhanced simultaneously, both of which reduce the stability of modified nanoparticles, resulting in more agglomeration and deposition. Kasel et al. (2013) researched the C0 influence on the retention of multi-walled carbon nanotubes (MWCNs) in saturated porous media, and found out that the effluent increased with increasing C0. This trend was explained by the filling up of retention locations over time so that higher input concentration produced increasing recovery mass (Bradford & Bettahar 2006). The effect of C0 on SM-NZVI migration behavior was not the same as on MWCNs due to the magnetic force and surface activity between SM-NZVI particles resulting in intense particle aggregation and pore blockage. Further calculation showed that the fraction of the injected mass of SM-NZVI decreased by merely 20% when initial concentration increased 4.3 times from 1.82 to 9.72 g/L, which more or less reflects the important role of sucrose in improving the dispersion and stability of SM-NZVI particles.

Figure 5

Breakthrough curves of SM-NZVI at different input concentration.

Figure 5

Breakthrough curves of SM-NZVI at different input concentration.

In conclusion, enhanced mobility of SM-NZVI is expected for lower input concentrations. Therefore, C0 effect needs to be considered for more accurately predicting the remediation area, as well as the fate and risks of injected nanoparticles.

Deposition kinetics studies

Under steady-state conditions, the transport of colloidal particles in porous media can be described by convective-dispersive equation (CDE) of first-order deposition kinetics (Kretzschmar et al. 1997), whose general form is given by 
formula
1
where C, D, x, t are the particle concentration and diffusion coefficient, transport distance and time, respectively. Note that the rate of physicochemical filtration is often represented by the particle deposition rate coefficient (k), and larger k value indicates stronger tendency to be deposited on the surface of porous medium, resulting in poor transport ability. Based on the assumption of the CDE equation and filtration theory, k is related to the ratio of effluent and initial particle concentration as follows (Tufenkji & Elmelech 2004): 
formula
2
Here, L is the filter medium packed length and the ratio of approach velocity to the porosity (U/f) is the interstitial fluid velocity commonly used in colloidal particle transport modeling.

The k value of bare and modified NZVI particle of 5.07 × 10−3 s−1 and 1.75 × 10−3 s−1, respectively, were obtained by Equation (2), showing that the deposition rate of SM-NZVI decreased by 65.5% compared with B-NZVI. Moreover, the k value was 3.50 × 10−3 s−1 in medium sand, double that of coarse sand. Figure 6 gives the effect of injection rate (Figure 6(a)) and concentration (Figure 6(b)) on deposition rate coefficient. Obviously, the value of k was found to systematically change with v and C0. In particular, it decreased with increasing injection rate below the critical influence value, and inversely increased with suspension concentration. These trends indicate more deposition occuring in the pore space, resulting from media adsorption and nanoparticle aggregation (Chen et al. 2011). The calculated results of k value are consistent with the description of breakthrough curves and therefore first-order deposition kinetics can be used to reasonably predict the transport behavior of SM-NZVI in saturated porous media.

Figure 6

The effect of injection rate (a) and concentration (b) on deposition rate coefficient.

Figure 6

The effect of injection rate (a) and concentration (b) on deposition rate coefficient.

CONCLUSIONS

The transport behavior of iron nanoparticles in a subsurface environment is important to in situ degradation efficiency of contaminants. To this end, several laboratory column transport experiments were undertaken for bare and sucrose-modified nanoparticles in saturated porous media, and the effects of media size, injection rate and input concentration of nanoparticle suspension were also investigated in depth. The research findings are summarized as follows.

  1. An addition of sucrose as surface modifier would greatly enhance the transport ability of nano-iron suspension for enhanced dispersion. The migration efficiency of SM-NZVI increased by 1.6 times and the deposition rate coefficient (k) decreased by 65.5% compared with bare nano-iron.

  2. The transport behaviors of SM-NZVI could be very different in various porous media mainly depending on their granule size. The highest effluent concentration was obtained in coarse sand because of the relatively large pore space and as such there was hardly any breakthrough of nanoparticles in the fine sand column.

  3. The injection rate of suspension was demonstrated to have a remarkable influence on the transport of SM-NZVI particles and the critical value was approximately 0.06 cm/s. The increasing injection rate below the critical value would enhance the mobility of SM-NZVI and simultaneously reduce the breakthrough time.

  4. Increasing input concentration of suspension was likely to be adverse to SM-NZVI migration and the recovery concentration decreased by 20.5% as C0 increased from 1.82 to 9.72 g/L, it implies that higher concentration of suspension for reaction rate should be prudently chosen in practical application.

  5. As an important kinetic parameter, the value of k can reasonably describe the deposition behavior of SM-NZVI in saturated porous media, and therefore the transport distance of nanoparticles in subsurface can be more accurately predicted.

Such studies are a necessary first step toward understanding the behavior of SM-NZVI in saturated porous media. However, these researches should be extended to a wider range including the exploration of migration mechanism, the quantitative calculation for maximum migration distance and the seeking of a feasible method to enhance the transport ability of iron nanoparticles. These works will be meaningful for promoting the application of iron nanoparticle in remediation of contaminated soil and groundwater.

ACKNOWLEDGEMENTS

This study was supported by the Natural Science Foundation of China (No. 41272256) and Basal Research Fund of Chinese Academy of Geological Sciences (No. YWF201405).

REFERENCES

REFERENCES
Alberto
T.
Rajandrea
S.
2009
Enhanced transport of zerovalent iron nanoparticles in saturated porous media by guar gum
.
Journal of Nanoparticle Research
11
(
3
),
635
645
.
Bradford
S. A.
Bettahar
M.
2006
Concentration dependent transport of colloids in saturated porous media
.
Journal of Contaminant Hydrology
82
(
1–2
),
99
117
.
Cheng
R.
Wang
J. L.
Zhang
W. X.
2007
Comparison of reductive dechlorination of p-chlorophenol using Fe0 and nanosized Fe0
.
Journal of Hazardous Materials
144
(
1–2
),
334
339
.
Chowdhury
I.
Hong
Y.
Honda
R. J.
Walker
S.
2011
Mechanisms of TiO2 nanoparticle transport in porous media: role of solution chemistry, nanoparticle concentration, and flowrate
.
Journal of Colloid and Interface Science
360
(
2
),
548
555
.
Jiao
Y. L.
Qiu
C. C.
Huang
L. H.
Wu
K. X.
Ma
H. Y.
Chen
S. H.
Ma
L. M.
Wu
D. L.
2009
Reductive dechlorination of carbon tetrachloride by zero-valent iron and related iron corrosion
.
Applied Catalysis B: Environmental
91
(
1–2
),
434
440
.
Kretzschmar
R.
Barmettler
K.
Grolimund
D.
Yan
Y.
Barmettler
M.
Sticher
H.
1997
Experimental determination of colloid deposition rates and collision efficiencies in natural porous media
.
Water Resource Research
33
(
5
),
1129
1137
.
Li
H.
Zhao
Y. S.
Zhao
R.
Ma
B. W.
Chen
Z. F.
Su
Y.
Zhou
R.
2013
Characteristics and kinetics of nitrobenzene reduction by sucrose-modified nanoiron
.
Chemical Research in Chinese Universities
29
(
4
),
765
770
.
Liang
Y.
Bradford
S. A.
Simunek
J.
Vereecken
H.
Klumpp
E.
2013
Sensitivity of the transport and retention of stabilized silver nanoparticles to physicochemical factors
.
Water Research
47
(
7
),
2572
2582
.
Pelley
A. J.
Tufenkji
N.
2008
Effect of particle size and natural organic matter on the migration of nano- and microscale latex particles in saturated porous media
.
Journal of Colloid and Interface Science
321
,
74
83
.
Phenrat
T.
Saleh
N.
Sirk
K.
Tilton
R.
Lowry
G. V.
2007
Aggregation and sedimentation of aqueous nanoscale zerovalent iron dispersions
.
Environmental Science & Technology
41
(
1
),
284
290
.
Polyxeni
N. M.
Vasiliki
I. S.
Constantinos
V. C.
2013
Transport of colloids in unsaturated packed columns: role of ionic strength and sand grain size
.
Chemical Engineering Journal
232
,
237
248
.
Saleh
N.
Phenrat
T.
Sirk
K.
Dufour
B.
OK
J.
Sarbu
T.
Matyjaszewski
K.
2005
Adsorbed triblock copolymers deliver reactive iron nanoparticles to the oil/water interface
.
Nano Letters
5
(
12
),
2489
2494
.
Vasiliki
I. S.
Constantinos
V. C.
2012
Transport of biocolloids in water saturated columns packed with sand: effect of grain size and pore water velocity
.
Journal of Contaminant Hydrology
126
(
3–4
),
301
314
.
Wang
C.
Bobba
A. D.
Attinti
R.
Shen
C. Y.
Lazouskaya
V.
Wang
L. P.
2012
Retention and transport of silica nanoparticles in saturated porous media: effect of concentration and particle size
.
Environmental Science & Technology
46
(
13
),
7151
7158
.
Yang
G. C. C.
Tu
H. C.
Hung
C. H.
2007
Stability of nanoiron slurries and their transport in the subsurface environment
.
Separation and Purification Technology
58
(
1
),
166
172
.
Yao
K. M.
Habibian
M. T.
O'Melia
C. R.
1971
Water and waste filtration, concepts and applications
.
Environmental Science & Technology
5
(
11
),
1105
1112
.
Zhang
P. F.
Tao
X.
Li
Z. H.
Bowman
R. S.
2002
Enhanced perchloroethylene reduction in column systems surfactant-modified zeolite/zero-valent iron pellets
.
Environmental Science Technology
36
(
16
),
3597
3603
.