Professional Documents
Culture Documents
UF Membrane Process
UF Membrane Process
Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere
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
a r t i c l e i n f o a b s t r a c t
Article history: In this study, a pilot-scale adsorption-UF process equipped with an aerated system is established for
Received 2 January 2018 arsenic removal from As-spiked Songhua river water. A newly synthesized amino-functionalized coffee
Received in revised form cellulose adsorbent (PEI-coffee) which is derived from spent coffee powder is fully characterized and
22 March 2018
used for arsenic removal from water. The batch experiments revealed that the adsorption process could
Accepted 23 March 2018
Available online 24 March 2018
be well described by Langmuir model with a maximum adsorption capacity of 13.2 and 46.1 mg/g for
As(III) and As(V), respectively. The negative value of △H and △G indicated the exothermic and spon-
Handling Editor: Tsair-Fuh taneous nature of As adsorption on PEI-coffee. The effects of operating parameters such as pH, initial
concentration and adsorbent dosage, were optimized by response surface methodology (RSM) based on a
Keywords: central composite design (CCD). The combined adsorption - UF process was employed for arsenic
Arsenic removal from As-spiked Songhua river water. It was demonstrated that aeration not only increased the
Adsorption removal efficiency by oxidizing As(III) to As(V), but mitigated the membrane fouling process. Besides of
Cellulose the adsorption process, UF membrane could also reject arsenic through the electrostatic repulsion be-
Membrane
tween arsenic species and membrane surface. After UF filtration, the dissolved As, suspended solids (SS),
River water
and TOC can be effectively eliminated. The saturated adsorbent was regenerated by using an eluting
agent of 10 wt% NaCl and 10 wt% NaOH, the regenerated adsorbent still sustained a very high adsorption
capacity after 6 cycles of adsorption-regeneration process.
© 2018 Elsevier Ltd. All rights reserved.
1. Introduction
* Corresponding author. College of Marine and Environmental Sciences, Tianjin
University of Science & Technology, Tianjin 300457, PR China. Intensive research efforts have been pursued to remove arsenic
E-mail addresses: xilan.2009@tust.edu.cn (L. Hao), lgj69@tust.edu.cn (G. Li).
https://doi.org/10.1016/j.chemosphere.2018.03.159
0045-6535/© 2018 Elsevier Ltd. All rights reserved.
L. Hao et al. / Chemosphere 202 (2018) 768e776 769
(As) contamination from wastewater or groundwater, but limited As(III) and As(V) of 10 mg/L are prepared by dissolving NaAsO2 and
attention is given to arsenic contamination in the natural water Na3AsO4$12H2O, respectively, in Songhua river water, and all the
system such as rivers or lakes, which are the important sources to required concentrations are diluted by the real Songhua river water.
provide potable water supply to millions of people living in A hollow fiber UF membrane is used, which is offered from
different countries. Unlike of excessive arsenic in groundwater, the Tianjin Motian Membrane Eng. & Tech. Co., Ltd. (China). The char-
elevated arsenic concentration in surface waters (such as river or acteristic of the membrane is listed in Table S1. The membrane
lakes) was mainly caused by anthropogenic activities, wherein surface has been modified from hydrophobic to hydrophilic and
As(III) and As(V) are always co-existing (Hao et al., 2016). Arsenic carries a great deal of negative charge, which results in low arsenic
pollution in the natural water sources at elevated concentrations affinity on the membrane surface.
were well documented in many countries such as America, France,
Bangladesh, China, India and Mexico (Guan et al., 2012; Lee et al.,
2017). The baseline concentrations of arsenic in rivers or lakes 2.2. Preparation of the modified coffee cellulose adsorbent
have been reported in the range of 0.1e2.0 mg/L (Rahimsouri and
Modabberi, 2013). Gomati river (Ganga Plain, northern India) had The spent coffee powders were collected from local coffee shops
arsenic concentrations in the range of 1.29e9.62 mg/L due to the and washed several times to remove the soluble compounds. 30 g
geothermal influence and anthropogenic causes (Singh et al., 2010). of spent coffee powder was mixed with 500 mL of NaClO solution
Alpine/Mediterranean Var River water (France) showed arsenic (5%, w/v) in a conical flask overnight, which is called the bleaching
concentrations in the range of 0.1e263 mg/L (Barats et al., 2014). In process. The bleaching of the spent coffee was conducted repeat-
the Stampede and Slate Creek watersheds (Alaska, USA), arsenic edly for several times until the color of the spent coffee powder
concentrations in stream waters were as high as 239 mg/L in the changed from black to white. The floating cellulose fibers were
year of 2010 (Ritchie et al., 2013). Manchar Lake (Pakistan) was separated by Buchner funnel, and then stirred with 200 mL of 10%
reported to have arsenic concentrations of around 60.45 mg/L (Arain polyethyleneimine (PEI) solution for 2 h. 0.5 mL of 25% glutaralde-
et al., 2009). Moreover, water pollution accidents through anthro- hyde was added dropwise into the flask and kept for 1 h. The final
pogenic activities occurred frequently owing to the development of product was washed with deionized water to remove excess re-
global economy and urbanization processes (Hu et al., 2016). agents, dried in a vacuum oven and designated as PEI-coffee. The
Arsenic removal from water can be achieved by conventional schematic diagram of the synthetic process is shown in Fig. S1.
treatment processes such as chemical precipitation, ion exchange
and electro-coagulation (Hering et al., 2017; Nidheesh and Singh,
2017; Smith et al., 2017). Among these possible treatment pro- 2.3. Experimental procedure
cesses, adsorption techniques are extensively studied due to its
high removal efficiency, flexibility, low cost and ease of operation Adsorption equilibrium: Isothermal adsorption experiments
(Yu et al., 2016). Recently, Membrane separation processes such as: were conducted in a shaking water bath (150 rpm) with a tem-
ultrafiltration (UF), nonofiltration (NF) and reverse osmosis (RO), perature controller. Batch experiments were performed in Erlen-
have been increasingly used for the treatment of arsenic to meyer flasks (250 mL) containing 1 g/L adsorbent equilibrated with
decrease the amount of produced wastewater (S anchez et al., 2016). 1, 5, 10, 15, 20, 25, 35, 50 mg/L As(III) or As(V) for 6 h at 20, 30 and
A combination of adsorption e UF technique proves to be a 40 C, respectively. Finally, the samples were filtered through a
promising technology for separation of arsenic from effluents 0.22 mm membrane, and residual arsenic concentration was deter-
(Chen et al., 2016). The use of readily available low-cost adsorbent mined by using an inductively coupled plasma-optical emission
in combination with UF is a hybrid approach to concentrate spectroscopy (ICP-OES).
selectively and to recover valuable metal resources. The advantages Kinetic experiment: The experimental procedures for three
of adsorption -filtration process are the high separation selectivity consecutive runs were conducted. For the first run, the feeding
due to the use of a selective adsorbent and low energy re- arsenic solution of 1 mg/L (0.5 mg/L As(III) þ 0.5 mg/L As(V)), which
quirements involved in these processes (Kai et al., 2016). was prepared by using real Songhua river water, was added into the
This work focused on investigation of the adsorption - UF pro- main reactor until the total volume reached 10 L, the adsorbent
cess for removal of arsenic from As-spiked Songhua river water dosage was 0.5 g/L. The solution was kept to be stirred by an
using modified coffee cellulose as an adsorbent. To highlight the aeration pump with a gas flow rate of 0.2 L/min. The UF membrane
arsenic removal performance, the main operating parameters such module was started to obtain the permeate samples, which are
as pH, adsorbent dosage, and initial concentrations, have been collected at 1, 2, 3, 5, 7, 10, 15, 20, 30, 40, 60, 90, 120, 180, 240, 360,
investigated. The thermodynamic and kinetic studies were also 480, 600, 720, 1440 min to determine the arsenic concentration.
conducted to better understand the properties of the modified After that, the residual solution in the reactor was pumped out
coffee cellulose adsorbent. The aerated stirring method was per- through UF membrane, but the adsorbent was intercepted in the
formed to assess the effect of aeration on the removal efficiency of reactor. For the second run, the same feeding solution was added
total arsenic and membrane fouling. into the reactor, and the procedure was repeated. After the second
run, the residual solution was pumped out through UF membrane,
2. Materials and methods the third batch of feeding solution was added into reactor, and the
procedure was repeated.
2.1. Materials To compare the effects of mechanic and aerated stirring styles
on arsenic removal efficiency, three different initial concentrations
All chemicals are analytical grade and supplied by Shanghai of 0.1 (0.05 mg/L As(III) þ 0.05 mg/L As(V)), 0.5 (0.25 mg/L
Chemical Co., Ltd. (China). All solutions are prepared with ultrapure As(III) þ 0.25 mg/L As(V)) and 1 (0.5 mg/L As(III) þ 0.5 mg/L As(V))
water of conductivity less than 1 ms/cm. The solution pH is adjusted mg/L were prepared and the adsorbent dosage is 0.5 g/L. The me-
by adding 0.1 M H2SO4 and 0.1 M NaOH. The experimental water chanic stirring rate is 150 rpm, the gas flow rate of aeration is 0.2 L/
samples are collected from Songhua river (Heilongjiang province, min. After reaction of 6 h, the samples were filtered through a
China), and the baseline concentration of arsenic in Songhua river 0.22 mm membrane, and residual arsenic concentration was
water is detected to be less than 0.01 mg/L. The stock solutions of determined by using ICP-OES.
770 L. Hao et al. / Chemosphere 202 (2018) 768e776
Fig. 2. (A) XRD patterns of coffee cellulose; (B) TGA curves of raw spent coffee and
modified coffee cellulose; (C) N2 adsorption/desorption isotherms of raw spent coffee
at 77 K; (D) N2 adsorption/desorption isotherms of modified coffee cellulose at 77 K.
772 L. Hao et al. / Chemosphere 202 (2018) 768e776
Table 1
Experimental ranges and levels of the independent variables.
a 1 0 1 a
Solution pH A 2.5 4.5 6.5 8.5 10.5
Initial arsenic conc., (mg/L) B 0.05 0.1 0.5 1 5
Adsorbent dosage, (g/L) C 0.2 2.1 5.1 8.1 10
1 1 1 1 55 63 2.67
2 0 1.68 0 100 100 1.57
3 0 0 0 100 100 0.00
4 0 0 1.68 100 100 1.43
5 1.68 0 0 82 81 0.17
6 0 1.68 0 91 85 1.58
7 0 0 0 100 100 0.00
8 0 0 1.68 70 65 1.45
9 0 0 0 100 100 0.00
10 1 1 1 99 96 1.05
11 1 1 1 93 97 1.19
12 1 1 1 99 91 2.65
13 1 1 1 100 99 0.10
14 0 0 0 100 100 0.00
15 1 1 1 99 95 1.18
16 0 0 0 100 100 0.00
17 1.68 0 0 64 65 0.15
18 1 1 1 53 54 0.11
19 0 0 0 100 100 0.00
20 1 1 1 70 73 1.07
removal of 99.9% was reached according to the CCD model. The The DH value of 16.56 kJ/mol for As(III) is higher than that of
optimized parameters were validated with the experimental result 11.91 kJ/mol for As(V), which means As(V) adsorption is easier than
of 99.9%, which demonstrated that the result of modeling is reliable As(III). The negative values of DG (45.61 ~ 26.02 kJ/mol) in-
to predict arsenic removal under different operational conditions. dicates the arsenic adsorption process is spontaneous with high
affinity towards PEI-coffee. The decrease of DG with a rise in tem-
perature suggests the stronger affinity at higher temperatures. The
3.3. Thermodynamics of arsenic adsorption on PEI-coffee adsorbent values of DG for As(V) are obviously lower than that of As(III),
indicating As(V) adsorption process is much easier than As(III),
Batch experiments for As(III) and As(V) adsorption were carried these results are consistent with the changes of DH values. The
out at 293, 303 and 313 K to determine the thermodynamic pa- positive value of DS implies an increase of randomness at the solid/
rameters. As can be seen from Fig. 3, adsorption capacities solution interface. For many adsorption in aqueous solutions, the
increased with an increase in temperature within 20e40 C. positive values of DS were frequently calculated (Liu et al., 2010; Hu
Langmuir and Freundlich isotherms were used to determine the et al., 2014). In this study, the positive DS was explained as follows:
parameters of adsorption affinity and the maximum adsorption Before arsenic adsorption, some co-existing ions such as Cl, NO 3
capacities at different temperatures. The calculated isotherm con- may be pre-adsorbed on the adsorbent surface. When arsenic
stants were given in Table 2. The maximum adsorption capacity of species was added, the pre-adsorbed ions were dissociated from
13.2 and 46.1 mg/g for As(III) and As(V), respectively, were obtained the adsorbent, so as to increase the random of solid/solution
at temperature of 313 K according to the Langmuir isotherm. The interface. The possible reactions were shown as follows:
parameters of adsorption affinity (kL) towards As(III) and As(V) also
increased with an increasing temperature, indicating the compar- S NH3þ $Cl þ H2 AsO þ
4 4S NH3 $H2 AsO4 þ Cl
(5)
atively higher temperature could facilitate As(III)/As(V) adsorption
on PEI-coffee within 293e313 K. The maximum adsorption affinity
S NH3þ $NO þ
3 þ H2 AsO4 4S NH3 $H2 AsO4 þ NO3 (6)
constants kL for As(III) and As(V) were 0.27 and 0.55 L/mg,
respectively. The higher kL value demonstrated the stronger
adsorption affinity to As(V) than As(III).
The thermodynamic parameters were obtained according to
Equations (3) and (4). 3.4. Effect of aeration on removal efficiency
InðK0 Þ ¼ DS=R DH=RT (3) As(III) and As(V) are always co-existing in natural water sources
(Ben et al., 2011). The species of arsenic is greatly influenced by the
DG ¼ RT$lnðKL Þ (4) redox potential in water (Guo et al., 2017), therefore it is necessary
to investigate the effect of dissolved oxygen (DO) on RE%. The
where R is the universal gas constant 8.314 J/mol/K, T is tempera- comparison between the mechanic and aerated stirring is shown in
ture (K) and KL is the thermodynamic equilibrium constant (L/mol), Fig. 4 (A). It is revealed that RE% is significantly enhanced by using
which is derived from Langmuir equation. aerated stirring at different initial As concentrations of 0.1, 0.5 and
As shown in Table 2, the positive values of DH confirm the 1 mg/L. Compared with the mechanical stirring, the average RE% is
endothermic nature of As(III) and As(V) adsorption on PEI-coffee. increased from 74% to 88% by using the aerated stirring method
L. Hao et al. / Chemosphere 202 (2018) 768e776 773
Fig. 4. (A) Comparison of the mechanic and aerated stirring for total arsenic removal
efficiency; (B) Effect of UF membrane on the removal efficiency of arsenic.
Fig. 3. Adsorption isotherms for (A) As(III) and (B) As(V) adsorption on PEI-coffee at
20, 30, 40 C (initial As concentration of 1, 5, 10, 15, 20, 25, 35, 50 mg/L, respectively, the oxidation time (Wang et al., 2016; Yokohira et al., 2010), how-
adsorbent dosage 1 g/L, pH 5.1). ever, the excessive addition of these reagents is also posing a po-
tential threat to human health. Therefore the aerated stirring is
used as an environmental-friendly method to expect the oxidation
when the initial concentration is 1 mg/L and the adsorbent dosage
of As(III) to As(V). It needs to be mentioned that some co-existing
is 0.5 g/L (The mechanic stirring rate is 150 rpm, the gas flow rate of
ions, such as phosphate in Songhua river water, may have a nega-
aeration is 0.2 L/min). Because PEI-coffee exhibited much better
tive effect on arsenic removal. The competition between arsenate
adsorption performance for As(V) than As(III), the idea that As(III) is
and phosphate is well documented (Kanematsu et al., 2011). The
pre-oxidized to As(V) is considered. It was reported that many
dissolved Fe3þ can facilitate arsenic (III, V) adsorption by forming
oxidizing reagents, such as active chlorine, sodium hypochlorite,
FeAsO4 precipitate (Onanguema et al., 2010). High valance of
ozone and potassium permanganate, could significantly shorten
Table 2
Isothermal and thermodynamic parameters of linear fitting for arsenic adsorption on PEI-coffee.
As(III) 293 0.25 11.1 0.987 1.82 2.09 0.954 26.02 16.56 0.18
303 0.26 11.4 0.969 2.19 2.13 0.967 27.17
313 0.27 13.2 0.958 2.70 2.32 0.968 28.52
As(V) 293 0.49 44.5 0.996 7.68 1.42 0.914 45.56 11.91 0.13
303 0.46 44.9 0.986 9.32 1.52 0.905 45.69
313 0.55 46.1 0.972 11.4 1.72 0.892 46.61
774 L. Hao et al. / Chemosphere 202 (2018) 768e776
and mercury in surface waters and bed sediments from Aq-Darreh river, Takab, concentrations by the combined pre-oxidation and nanofiltration membrane
northwest Iran. J. Environ. Sc. Water Resour 2, 75e81. process. Desalin. Water Treat 57, 1e10.
Reddy, K.O., Maheswari, C.U., Dhlamini, M.S., et al., 2017. Preparation and charac- Watson, M.A., Tubi c, A., Agbaba, J., et al., 2016. Response surface methodology
terization of regenerated cellulose films using borassus fruit fibers and an ionic investigation into the interactions between arsenic and humic acid in water
liquid. Carbohydr. Polym. 160, 203e211. during the coagulation process. J. Hazard Mater. 312, 150e158.
Ritchie, V.J., Ilgen, A.G., Mueller, S.H., et al., 2013. Mobility and chemical fate of Wu, Y.Y., Zhou, S.Q., Qin, F.H., et al., 2010. Modeling physical and oxidative removal
antimony and arsenic in historic mining environments of the Kantishna Hills properties of Fenton process for treatment of landfill leachate using response
district, Denali National Park and Preserve, Alaska. Chem. Geol. 335, 172e188. surface methodology (RSM). J. Hazard Mater. 180, 456e465.
S
anchez, J., Butter, B., Chavez, S., et al., 2016. Quaternized hydroxyethyl cellulose Yokohira, M., Arnold, L.L., Pennington, K.L., et al., 2010. Severe systemic toxicity and
ethoxylate and membrane separation techniques for arsenic removal. Desalin. urinary bladder cytotoxicity and regenerative hyperplasia induced by arsenite
Water Treat 57, 25161e25169. in arsenic (þ3 oxidation state) methyltransferase knockout mice. A preliminary
Sarker, M., Ji, Y.S., Jhung, S.H., 2017. Adsorption of organic arsenic acids from water report. Toxicol. Appl. Pharmacol. 246, 1e7.
over functionalized metal-organic frameworks. J. Hazard Mater. 335, 162e169. Yu, C., Peltola, P., Nystrand, M.I., et al., 2016. Arsenic removal from contaminated
Singh, M., Singh, A.K., Swati, et al., 2010. Arsenic mobility in fluvial environment of brackish sea water by sorption onto Al hydroxides and Fe phases mobilized by
the Ganga Plain, northern India. Environ. Earth Sci. 59, 1703e1715. land-use. Sci. Total Environ. 542, 923e934.
Smith, K., Li, Z., Chen, B., et al., 2017. Comparison of sand-based water filters for Yu, X., Tong, S., Ge, M., et al., 2013. Synthesis and characterization of multi-amino-
point-of-use arsenic removal in China. Chemosphere 168, 155e162. functionalized cellulose for arsenic adsorption. Carbohydr. Polym. 92, 380e387.
Wang, Q., Zhang, G., Chen, H., et al., 2016. Enhanced As(III) removal at low