A polyurethane (PU) foam nanocomposite impregnated with iron oxide nanoparticles (IONPs) was developed to remove arsenic (As) from drinking water at ppb concentrations. The effect of synthesis and application parameters such as the size of IONPs, pH levels, weight of adsorbents, and arsenic concentrations on the performance of PU-IONP adsorbents in removing arsenic were studied. The prepared adsorbents were characterized by scanning electron microscopy and energy dispersive X-ray microscopy to evaluate the microstructure of PU-IONPs and the surface adsorption of arsenic species, respectively. Atomic absorption spectrometry was conducted to measure the concentration of arsenic in the treated solutions in order to calculate the removal capacity of PU-IONPs. The experimental results revealed that decreasing the size of IONPs from 50–100 nm to 15–20 nm yields a higher removal capacity. Increasing the weight of the used adsorbents and the contact time led to an increase in the removal capacity as well. As the arsenic species (III and V) concentration increased in the solution, the removal capacity of PU-IONPs decreased. In a column study, a long-term cyclic operation mode was found to be very effective in removing arsenic; 100% removal capacity was achieved when 500 ml of As solution (100 ppb) was treated.

## INTRODUCTION

Water pollution by heavy metal pollutants such as arsenic (As) has serious toxic effects on humans and living organisms. Arsenic is identified as a carcinogenic element by the International Agency for Research on Cancer (IARC) and the US Environmental Protection Agency (Levine et al. 1988; IARC 2012). Bladder, skin, and lung cancers were confirmed due to chronic arsenic exposure, and the potential target organs for cancer from arsenic exposure are liver, kidney, and prostate (NRC 1999). Significant research work, aimed at finding and developing various separation and treatment techniques, has been conducted over the past few decades. Arsenic contamination of ground water poses a serious concern in many countries throughout the world, including the United States (Smedley & Kinniburgh 2002).

The removal of arsenic, essentially, involves a selective separation of As(V) and As(III). The conventional treatment methods of arsenic involve a coagulation with ferric chloride or aluminum sulfate coagulants, followed by the separation of the produced insoluble by settling, or by direct filtration through sand beds. Other treatment techniques for arsenic removal are reverse osmosis (Ning 2002), lime softening (McNeill & Edwards 1997), flotation and adsorption on hydrated iron oxide or activated carbon (DeMarco et al. 2003; Gu et al. 2005; Gupta et al. 2005; Asadullah et al. 2014). These methods have been reported to be effective, mainly, for the removal of pentavalent arsenic As(V). Therefore, a pre-oxidation step is usually required in order to achieve efficient removal of trivalent arsenic As(III). Solid-phase nanoadsorbents are becoming the core of most recent work in removing heavy metals due to their high capacity and affinity to heavy metal ions. Iron oxide nanoparticles (IONPs) have been used in many studies (Nguyen et al. 2006; Vunain et al. 2013; Devi et al. 2014). Nano-adsorbents such as hydrous ferric oxide (Zhang et al. 2008) and MgO (Choi et al. 2014) have been deposited on the surface of porous materials. However, the preparation of these adsorbents often involves complex and costly methods.

In this study, an extended performance analysis for a new bulk modified nanocomposite material (adsorbent) is investigated. PU-IONPs were developed by using IONPs impregnated in open cell polyurethane (PU) foams in order to exploit the inherent advantages of porous PU foam structures. Besides, the incorporation of the adsorbent particles in a foam media allows for easier post-treatment steps. The capability of iron compounds to react with arsenic species by adsorption and ion exchange mechanisms allows higher removal capacities. This system of PU-IONPs offers a potential for arsenic removal at lower costs compared with other conventional arsenic removal systems.

## MATERIALS AND METHODS

### Materials

Particular raw materials were used in the synthesis of PU-IONPs. Polypropylene glycol 1200 (PPG; Sigma Aldrich Co. LLC) dehumidified in a vacuum oven at 70 °C, toluene di-isocyanate (TDI; 2,4 – 80%, 2,6 – 20%, Alfa Aesar), polysiloxane surfactant (Sigma Aldrich), IONPs (Fe3O4, high purity 99.5%, US Research Nanomaterials Inc.) with two size ranges of 50–100 nm and 15–20 nm, nitrogen gas (Airgas, O2 free UHP), and 18.2 MOhm-cm deionized water. For the batch experiments, 1 ppm standard arsenic solution, from Inorganic Ventures Co., was diluted to 100 ppb. The solution contained both As(III) and As(V).

### Preparation and characterization of PU-IONPs

The experimental setup applied in this study is described in a previous publication (Gunashekar & Abu-Zahra 2014). A three-neck round-bottom reaction flask was placed in an oil bath and fitted with a mechanical stirrer and a condenser at the center neck. A nitrogen gas inlet was fitted at the right neck and a dropping funnel was fitted at the left neck. The reaction between PPG and TDI was conducted at 75 ̊C in an inert atmosphere. Figure 1 shows the experimental setup for PU synthesis.
Figure 1

Experimental setup for PU synthesis.

Figure 1

Experimental setup for PU synthesis.

Initially, the three-neck flask was charged with TDI and allowed to stabilize at 75 °C in a saturated nitrogen atmosphere. A dropping funnel was filled with a pre-weighted amount of PPG, which was added dropwise and allowed to react with TDI for 4–5 hours until an initial isocyanate content of 11%–12% was reached, in accordance with ASTM D5155. IONPs were manually added to the mixture in different weight percentages (4%, 8%, 12%, and 16%). Based on the amount of PPG used, a pre-weighted amount of deionized water was added as a blowing agent along with the polysiloxane surfactant. The compound was mixed using a mechanical stirrer at 2,500–3,000 rpm for 10–15 seconds. The reaction of water with the remaining isocyanate groups released CO2 gas to form the final foam structure (Szycher 1999).

The microstructure of PU-IONP adsorbents and the elemental analysis and distribution of IONPs in the nanocomposite foam matrix were characterized using a JEOL JSM-6460 LV scanning electron microscope with energy dispersive X-ray microscope (SEM/EDX). The open cell content and the skeletal density were measured using an AccuPyc II 1340 FoamPyc V2.00 instrument. The adsorption capacity of the PU-IONPs was calculated from the measurements of a Thermo Electron Corporation S4 atomic absorption spectrometer (AAS).

Batch experiments were performed to investigate the effect of many variables on the arsenic removal capacity. Several samples with the same PPG:TDI molar ratio, granular shape and different IONP sizes, i.e., 15–20 nm and 50–100 nm, were soaked in 50 ml of 100 ppb As solution for 6 and 24 hours and shaken in neutral solution (pH = 6.5) at 200 rpm and at room temperature (22 °C) to study the effect of the nanoparticle size and contact time on the adsorption capacity. Further stages were carried out to examine the effect of other variables. The weight of adsorbents (0.5 g, 1 g, 1.5 g and 2 g), pH levels (3.5, 6.5, 8.5 and 10.5) and the concentration of arsenic solutions (100 ppb, 200 ppb, 400 ppb and 600 ppb) were chosen to extend the performance analysis of PU-IONP adsorbents. After each batch test, 25 ml of each treated solution was filtered and preserved with 2% HNO3, for AAS analysis.

### Column study

A column study was performed using a glass column with 19 mm inner diameter and 300 mm height. The column was packed with 8 g (57 ml, fixed bed height 20 cm) of granular PU-IONP adsorbents. A 120 ppb arsenic solution was circulated through the packed column in the upflow direction, using a Fisher Scientific™ Variable-Flow peristaltic pump, at a flow rate of 1.5 ml/min. The column was operated in a closed circulation mode for 9 days and the samples from the column test were collected at a regular number of circulations. The removal capacity of the arsenic species was analyzed by AAS. A schematic illustration of the column study setup is shown in Figure 2.
Figure 2

Schematic illustration of column test.

Figure 2

Schematic illustration of column test.

## RESULTS AND DISCUSSION

### Characterizations analysis

#### SEM and EDX

High-resolution SEM imaging for non-conductive foam samples were obtained using low vacuum mode and a back-scatter detector. Figure 3 shows the porous structure of the PU-IONPs with a combination of open and closed cells. The detected structure offers a higher surface area with IONPs compared with IONPs deposited on the foam surface only (Nguyen et al. 2006). The adsorption capacity of arsenic is improved with the more accessible surface area of the IONPs in the foam.
Figure 3

SEM image of PU-IONPs at 500× magnification.

Figure 3

SEM image of PU-IONPs at 500× magnification.

To identify the IONP distribution inside the foam, an EDX mapping technique was used. A quantitative analysis (EDX spectrum) was performed on individual points, as illustrated in Figures 4 and 5. The elemental analysis of the circled area resulted in 46.05%, 34.69%, 14.93%, and 4.03% as weight percentages for the detected elements Fe, O, C, and Si, respectively.
Figure 4

EDX mapping scan of PU-IONPs.

Figure 4

EDX mapping scan of PU-IONPs.

Figure 5

EDX spectrum for the circled area in Figure 4.

Figure 5

EDX spectrum for the circled area in Figure 4.

### Open cell content

The open cell content of the PU-IONP samples was measured using an AccuPyc II 1340 FoamPyc V2.00 instrument, in accordance with ASTM D6266. The open cell volume in the foam was calculated by the difference between the nitrogen gas volume and the sample volume. The density and open cell content of the PU-IONPs are listed in Table 1.

Table 1

Density and open cell content of PU-IONPs

CharacteristicsDescription
Open cell (%) 81.59%
Closed cell (%) 18.41%
Skeletal density 1.23 g/cm3
CharacteristicsDescription
Open cell (%) 81.59%
Closed cell (%) 18.41%
Skeletal density 1.23 g/cm3

The above physical properties of the PU-IONPs reveal a high percentage of open cell content compared with the closed cell content. This can be associated with the lower amount of isocyanate groups in the foam formulation, which leads to regular foaming with less probability of foam collapse during molding, thus increasing the number of open cells in the foam (Gunashekar 2015).

### Batch sorption analysis

#### Effect of IONP size

Batch adsorption analysis was conducted to study the effect of IONP size on the As removal capacity of PU-IONPs. PU-IONP samples were prepared with the optimum percentage of loaded IONPs (12%) using a PPG:TDI composition ratio of 1:1.75 (Hussein & Abu-Zahra 2016), and two size ranges of the IONPs: 15–20 nm and 50–100 nm. The batch experiments were carried out under two exposure time intervals: 6 and 24 hr. Figure 6 illustrates the As removal capacity for the various IONP sizes and exposure times.
Figure 6

Arsenic removal capacity of PU-IONPs with two size ranges of IONPs: 15–20 nm and 50–100 nm.

Figure 6

Arsenic removal capacity of PU-IONPs with two size ranges of IONPs: 15–20 nm and 50–100 nm.

The results shown in Figure 6 point toward an enhancement in the performance of the adsorbents when the size of the IONPs is decreased from 50–100 nm to 15–20 nm under the same exposure conditions This behavior can be attributed to the higher surface area provided with smaller-size particles and the weaker effect of self-aggregation of the nanoparticles (Yavuz et al. 2006; Mayo et al. 2007). Furthermore, higher removal capacity can be attained by allowing more contact time between the arsenic species and the adsorbent due to the filling of the available binding sites on the surface.

### Effect of solution pH

The effect of the solution pH on the adsorption of As was examined. PU-IONP samples made with a PPG:TDI composition ratio of 1:1.75 in a granular shape were used in this study. The PU-IONP samples were soaked in 100 ppb As solutions with four different levels of pH (3.5, 6.5, 8.5 and 10.5) for 24 hr at room temperature (22 °C). Figure 7 shows the removal amount of As at various pH levels.
Figure 7

Effect of pH on the removed amount of As by PU-IONPs.

Figure 7

Effect of pH on the removed amount of As by PU-IONPs.

The adsorption process of As is affected by the pH level of the contaminated solution. The removed amount of As is higher as the solution becomes more acidic (i.e., pH < 7), compared with the removed amount of As when the solution is more basic (i.e., pH > 7). At a pH level below the pHPZC of an oxide, it produces substantially more positive charges than negative charges on the surface, whereas at pH levels above the pHPZC it produces more negative charges on the surface than positive charges (Mondal et al. 2013). The pHPZC of Fe3O4 is approximately 8, as reported in the literature (Cornell & Schwertmann 1996). Therefore, the surface of the PU-IONPs is predominantly negatively charged above pH = 8. Hence, the chemisorption of arsenic species is less, which lowers the removed amount of arsenic species. On the other hand, increasing the pH from 8.5 to 10.5 provides a slight increase in the removal capacity. At pH 8.5, the surface of the oxide will be close to neutrality while above that the surface is partially negative. This does not eliminate the effect of the partially positive charge, which could attract As species as well.

### Effect of weight

To study the effect of using different sample weights on the removal capacity of arsenic, four samples of PPG:TDI with a molar ratio of 1:1.75 and a granular shape were prepared and soaked in 100 ppb arsenic solution for 24 hr. Figure 8 shows the removed amount of As at different PU-IONP sample weights.
Figure 8

Removal capacity of PU-IONPs with respect to the foam weight.

Figure 8

Removal capacity of PU-IONPs with respect to the foam weight.

The removed amount of As increases from 40 ppb to 60 ppb when the weight of the PU-IONPs increases from 0.5 g to 2 g, respectively. This correlates with the increase of IONP amounts as the foam weight is increased; therefore, more adsorption sites will be available to uptake arsenic species. A similar finding for the effect of adsorbent weight on the removal capacity was reported in the literature (Kanel et al. 2006).

### Effect of As concentration

To investigate the effect of using different concentrations of arsenic solutions (100 ppb, 200 ppb, 400 ppb, and 600 ppb) on the removal capacity of arsenic, one gram granular samples of PPG:TDI with a molar ratio of 1:1.75 were kept in contact with arsenic solutions for 24 hr. The outcomes of this batch experiment are illustrated in Figure 9.
Figure 9

Effect of initial arsenic species concentration on the removal capacity.

Figure 9

Effect of initial arsenic species concentration on the removal capacity.

The removal capacity of arsenic decreases as the As concentration increases in the solution. The lower concentration As solution exhibits a higher removal capacity since the available binding sites for the adsorption process are almost the same for all foam samples; consequently, certain amounts of arsenic will be adsorbed from each concentration. A similar finding for the effect of As concentration on the removal capacity has been reported in the literature (Mandal et al. 2013).

### Column study

A long-term column study was carried out to study the removal capacity of As species. An amount of 500 ml of arsenic solution at 120 ppb was circulated through a packed column in the upflow direction at a filtration flow rate of 1.5 ml/min. The column was packed with 8 g (57 ml, fixed bed height 20 cm) of granular PU-IONP adsorbents with PPG:TDI ratio of 1:1.75. The time needed to complete one flow cycle was measured, using a stop watch, to be 9 hours and 40 minutes. The column was operated for 10 consecutive days. Figure 10 shows the removed amount of As during the whole operating time of the column.
Figure 10

Column study for arsenic species.

Figure 10

Column study for arsenic species.

The adsorption of arsenic species was very rapid in the first few cycles with an approximately 50% arsenic removal within two cycles. After that, a constant increase in the removal rate occurred. All arsenic species were removed in 22 cycles (approximately 9 days) of the operating period. This behavior of removing arsenic can be explained by the abundance of adsorption sites at the beginning of the process, which decreases gradually with more adsorbed arsenic species.

## CONCLUSIONS

This study presents an extended performance analysis of bulk modified nanocomposite material. IONPs impregnated in PU foam matrix were used for arsenic removal from water. The effect of IONP size, pH of the solution, adsorbent weight, and As concentration on the removal capacity were investigated. Foam samples with a smaller IONP size range (15–20 nm) achieved higher removal capacity compared to a size range of 50–100 nm. Increasing the weight of adsorbents led to an increase in the removal capacity under the same conditions. The removal capacity of PU-IONPs decreased as As species concentration increased in the solution. In the column study, a long-term cyclic operation mode was found to be very effective in removing arsenic. A 100% removal capacity was achieved when 500 ml of As solution (100 ppb) was treated. The applications of this study can be extended to surface water sources (e.g., lakes and rivers), where lower levels of arsenic exist, and to industrial waste water at ppm concentrations with proper modifications to the foam composition. Finally, the estimated cost of PU-IONPs ($182/lb) is lower than other filtration materials such as activated alumina ($225/lb) and hydrated iron oxide (\$586/lb).

## REFERENCES

REFERENCES
M.
Jahan
I.
Ahmed
M.
P.
Malek
N.
Rahman
M.
2014
.
J. Ind. Eng. Chem.
20
,
887
896
.
Choi
H.
Woo
N.
Jang
M.
Cannon
F.
Snyder
S.
2014
.
Sep. Purif. Technol.
136
,
184
189
.
Cornell
M.
Schwertmann
U.
1996
The Iron Oxides
.
VCH
,
Weinheim, Germany
.
DeMarco
M. J.
SenGupta
A. K.
Greenleaf
J. E.
2003
.
Water Res.
37
,
164
176
.
Devi
R. R.
Umlong
I. M.
Das
B.
Borah
K.
Thakur
A. J.
Raul
P. K.
Banerjee
S.
Singh
L.
2014
.
Appl. Water Sci.
4
,
175
182
.
Gu
Z. M.
Fang
J.
Deng
B. L.
2005
.
Environ. Sci. Technol.
39
,
3833
3843
.
Gunashekar
S.
2015
A Study on the Synthesis - Structure - Property - Performance Relationship of Bulk Functionalized Polyurethane Foams for Water Filtration Applications
.
PhD Dissertation
,
University of Wisconsin
,
Milwaukee, WI, USA
.
Gunashekar
S.
Abu-Zahra
N.
2014
.
Int. J. Polym. Sci.
2014
,
article ID 570309
.
Gupta
V. K.
Saini
V. K.
Jain
N.
2005
.
J. Colloid Interface Sci.
288
,
55
60
.
Hussein
F.
Abu-Zahra
N.
2016
.
J. Water Process Eng.
13
,
1
5
.
IARC
2012
A Review of Human Carcinogens. C. Metals, Arsenic, Fibres and Dusts
.
International Agency for Research on Cancer
,
Lyon
,
France
.
Kanel
S.
Greneche
J.
Choi
H.
2006
.
Environ. Sci. Technol.
40
,
2045
2050
.
Levine
T.
Rispin
A.
Chen
C.
Gibb
H.
1988
Special Report on Ingested Inorganic Arsenic: Skin Cancer; Nutritional Essentiality
.
USEPA
,
Washington, DC, USA
, .
Mandal
S.
Sahu
M.
Patel
R.
2013
.
Water Res. Ind.
4
,
51
67
.
Mayo
J. T.
Yavuz
C.
Yean
S.
Cong
L.
Shipley
H.
Yu
W.
Falkner
J.
Kan
A.
Tomson
M.
Colvin
V. L.
2007
.
8
(
12
),
71
75
.
McNeill
L.
Edwards
M.
1997
.
J. Environ. Eng.
123
(
5
),
453
460
.
Mondal
P.
Mohanty
B.
Majumder
B.
2013
.
Chem. Eng. Sci. J.
1
(
2
),
27
31
.
Nguyen
T.
Vigneswaran
S.
Ngo
H.
Pokhrel
D.
Viraraghavan
T.
2006
Iron-coated sponge as effective media to remove arsenic from drinking water
.
41
(
2
),
164
170
.
Ning
R.
2002
.
Desalination
143
(
3
),
273
241
.
NRC
1999
Arsenic in Drinking Water
.
,
Washington, DC, USA
.
Smedley
L.
Kinniburgh
G.
2002
.
Appl. Geochem.
17
,
517
568
.
Szycher
M.
1999
Szycher's Handbook of Polyurethanes
.
CRC Press
,
Boca Raton, FL, USA
.
Vunain
E.
Mishra
A.
Krause
R.
2013
.
J. Inorg. Organomet. Polym.
23
,
293
305
.
Yavuz
C. T.
Mayo
J. T.
Yu
W. W.
Prakash
A.
Falkner
J. C.
Yean
S.
Cong
L.
Shipley
H. J.
Kan
A.
Tomson
M.
Natelson
D.
Colvin
V. L.
2006
.
Science
314
(
5801
),
964
967
.
Zhang
Q. J.
Pan
B. C.
Zhang
W.
Bingjun Pan
B. J.
Zhang
Q. X.
Ren
H.
2008
.
Ind. Eng. Chem. Res.
47
,
3957
3962
.