Uptake of trivalent chromium from aqueous solutions by xanthate pine bark: Characterization, batch and column studies

Uptake of trivalent chromium from aqueous solutions by xanthate pine bark: Characterization, batch and column studies

Accepted Manuscript Title: Uptake of trivalent chromium from aqueous solutions by xanthate pine bark: characterization, batch and column studies Autho...

1MB Sizes 0 Downloads 6 Views

Accepted Manuscript Title: Uptake of trivalent chromium from aqueous solutions by xanthate pine bark: characterization, batch and column studies Authors: Aline L. Arim, Margarida J. Quina, Lic´ınio M. Gando-Ferreira PII: DOI: Reference:

S0957-5820(18)30909-1 https://doi.org/10.1016/j.psep.2018.11.001 PSEP 1559

To appear in:

Process Safety and Environment Protection

Received date: Revised date: Accepted date:

20 September 2018 31 October 2018 1 November 2018

Please cite this article as: Arim AL, Quina MJ, Gando-Ferreira LM, Uptake of trivalent chromium from aqueous solutions by xanthate pine bark: characterization, batch and column studies, Process Safety and Environmental Protection (2018), https://doi.org/10.1016/j.psep.2018.11.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Uptake of trivalent chromium from aqueous solutions by xanthate pine bark: characterization, batch and column studies Aline L. Arim1,2, Margarida J. Quina1, Licínio M. Gando-Ferreira1*

1

*

IP T

CIEPQPF, Department of Chemical Engineering, Faculty of Sciences and Technology, University of Coimbra, Pólo II, Rua Sílvio Lima, 3030-790 Coimbra, Portugal 2 UNIPAMPA - Federal University of Pampa, Campus Bagé, RS, Brazil Corresponding author: E-mail address: [email protected]

PT

ED

M

A

N

U

SC R

Graphical abstract

HIGHLIGHTS

New biosorbent was developed by Xanthation (XPB) and characterized from pine

CC E



bark



Feasibility of XPB as biosorbent was verified and maximum Cr(III) uptake was

A

56.50 mg/g



The functional groups C=C, C=O and C-S are the main responsible for sorption



Ion-exchange involving Cr ions and Na released by was confirmed



Breakthrough curves of Cr(III) using XPB in fixed-bed was investigated.

Abstract 1

In this study, the performance of chemically modified pine bark for removal of Cr(III) from solutions was investigated. Initially, several chemicals were tested (NaOH, C5H9NO4, H3PO2 and CS2). The xanthate pine bark (XPB) obtained with CS2, was screened as the best adsorbent and thus, it was characterized in respect to morphological and textural properties. Sulfur groups after xanthation reaction were identified on XPB by FTIR and EDS spectroscopies. Equilibrium isotherm and kinetics were determined.

IP T

The equilibrium data were well described by the Langmuir isotherm allowing to

calculate the maximum adsorption capacity (56.5 mg/g). The kinetics is fast and follows

SC R

a pseudo-second order model. Furthermore, an ion-exchange sorption mechanism

between Cr(III) and Na+ was proposed. Among the desorbing agents tested, the best results were achieved with 2.0 M H2SO4. Moreover, in the column tests, a reduction on the breakthrough time and the stoichiometric time from 36.5 and 22.8%, respectively,

U

was observed, when the feed concentration increased from 230 to 500 mg/L. The

N

breakthrough curves were well modeled by Bohard-Adams, Thomas or Yoon-Nelson equations. Globally, XPB revealed to be a promising adsorbent for uptake Cr(III). The

M

treatment of real effluents.

A

scale-up design and the economic assessment indicated potential applicability for the

ED

Keywords: Cr(III); biosorption; xanthation process; equilibrium; ion-exchange; fixed-

PT

bed column.

1. Introduction

CC E

Pollutants such as heavy metals are often found in natural ecosystems in high

concentrations because of uncontrolled discharge of effluents generated by industrial activities. These contaminants are mainly characterized by resistance to chemical and/or biological degradation, high environmental mobility and strong tendency to

A

bioaccumulate in the food chain [1,2]. Chromium is an example commonly found in effluents due to its wide use in industrial applications, including electroplating, leather tanning, metal finishing, textile manufacturing, etc. At the industrial level, there is a high interest in recovering Cr(III), not only for environmental protection, but also for economic reasons. Unlike chromium hexavalent, which is considered carcinogenic and mutagenic to living organisms, Cr(III) has not been much discussed in the literature [3]. 2

Among the physical-chemical processes for removal of metals from liquid effluents, the adsorption process is considered as a reliable technology, due to the flexibility of the design, easy operation, and in certain cases with the possibility of regeneration of the adsorbent to reduce the total cost of the process [1–5]. Consequently, there has been a significant increase in research on the development of new adsorbent materials, inexpensive and with great potential for adsorption of metal ions. Thus, several attempts have been made by researchers to improve the biosorption

IP T

potential of lignocellulosic materials [6,7]. Indeed, the superficial properties of these materials are suitable to promote modifications by chemical or physical treatments

SC R

resulting in increased adsorption capacities for the removal of specific ions [8]. It is well known that the adsorbents containing sulphur-bearing groups (such as dithiophosphates, sulfides, thiols, and xanthates) have a high affinity for heavy metals, while low affinity for lighter metals is observed [9,10]. The xanthates may be formed by the reaction of an

U

organic substrate containing hydroxyl with carbon disulphide (CS2) under alkaline R-OH + CS2 + NaOH → R-OCS2Na + H2O

N

conditions (Eq. 1)[11].

(1)

A

For this purpose, some lignocellulosic materials were functionalized by the

M

xanthate process and investigated for the removal of heavy metals from aqueous solutions. Some studies have shown that prepared xanthate materials may exhibit great

ED

potential to form stable xanthate-metal complexes with heavy metallic ions [12–14]. Liang et al. (2009) chemically modified orange peel, by introducing sulphur groups

PT

with CS2 treatment under alkaline condition, for sorption of Pb(II) ions [12]. This treatment increased the maximum sorption capacity, qmax, of the modified material to 204.55 mg/g compared to an initial capacity of 89.77 mg/g. Also, Bediako et al. (2015)

CC E

functionalized waste Lyocell with xanthate groups for the removal of Cd(II), Pb(II) and Cu(II), and high qmax values of 505.64, 531.29 and 123.08 mg/g, respectively, were found [14]. Other lignocellulosic materials functionalized by the xanthate process for

A

uptake of metals include: camphor leaves (qmax = 34.4 mg/g for Pb(II)) [15]; sugarcane bagasse (qmax values of 219.2, 327.4, 147.9, 156.9 and 184.9 mg/g for Cd(II), Pb(II), Ni(II), Zn(II) and Cu(II), respectively) [16]; rice husk (qmax = 138.8 mg/g for Cd(II)) [17] and nano banana cellulose (qmax = 154.26 mg/g for Cd(II)) [9]. Although some studies have been devoted to the xanthation modification of biosorbents for removal of metal ions, to the best our knowledge the application of xanthate lignocellulosic materials for removal of Cr(III) was not reported in the literature. 3

In 2010, Pinus pinaster commonly known as Pine Bravo occupied 23% of the forest area in Portugal, corresponding to about 714x103 ha [18]. One of the by-products is the bark (PB), which can represent 20 to 40% of the volume of the logs [19]. The PB mainly consists of total lignin (50.04%), cellulose (29.50%) and hemicellulose (14.86%) [20]. The presence of these three major polymers makes PB a material rich in hydroxyl and phenolic groups and these can be chemically modified to prepare adsorbent materials with novel properties [21]. Besides, these components have binding

IP T

sites for ion-exchange with metals [22].

In this context, the main objective of this work was to investigate the biosorption

SC R

potential of xanthate pine bark (XPB) for the removal of Cr(III) ions from aqueous solutions. Thus, an extensive characterization was performed before and after the

material functionalization. The effect of relevant operational parameters (the contact time, initial pH, and adsorbent dosage) on the Cr(III) uptake by XPB was examined. In

U

addition, studies were performed to gain insight into the sorption mechanism based on

N

the ion-exchange between chromium ions in solution and the Na+ ion present in the modified sorbent by xanthate. Some works have proposed the ion-exchange and

A

complexation mechanisms for sorption of heavy metals using biosorbents functionalized

M

with xanthates[12,14,15]. However, the analysis of the ion-exchange process in these biosorbents was little explored and Cr(III) was not addressed yet. Furthermore, it was

ED

investigated the influence of the feed metal concentration on the efficiency of removal of Cr(III) in a fixed-bed column. The scale-up design of the column as well as the

PT

integration of a nanofiltration process for the recovery of Cr (III) was proposed. In addition, an economic evaluation of the cost of preparation of the biosorbent developed

CC E

was carried out.

2. Materials and methods

A

2.1.Materials and reagents Pine bark (PB) was obtained from the North region of Portugal, washed with

water to remove mud, sand and the other contaminants. Then, PB was oven-dried at 90 ºC for 24 h. This material was milled and sieved to a particle size of 0.088 - 0.149 mm and kept in a desiccator prior to its use. Solutions of Cr(III) were prepared by dissolving predetermined amount of Cr(NO3)3 9H2O in ultrapure water.

4

The concentration of chromium was determined by energy dispersive X-Ray fluorescence (EDXRF) using a Nex CG Rigaku spectrometer. 2.2.Preparation of biosorbents Three types of chemical modification were applied to the PB: mercerization (NaOH), acid treatment (glutamic and hypophosphorous) and xanthation (CS2). In the case of the mercerization, about 2 g of PB was stirred with 60 mL of 2.5 M NaOH

IP T

solution for 16 h at 25 ºC using the method described elsewhere [20] and hereafter referred as MPB.

Acid treatment was performed according to the procedure reported by Liu et al.

SC R

(2011) [23], where 2.5 g of PB was suspended in 10 mL of dimethyl sulfoxide:

epichlorohydrin mixture (1:1, v/v), and then 5 mL of 1.0 M NaOH solution was added as a catalyst [23]. After the mixture was stirred for 2.5 h at 40 ºC, activated PB with

U

epoxy groups was obtained. Then, this material was contacted with 5 mL of 2 mol/L

N

Na2CO3 solution containing 1.5 g of glutamic acid (C5H9NO4)for addition of the carboxylate group or 2.5 mL hypophosphorous acid (H3PO2) for addition of the

A

phosphonium group. These mixtures were stirred for 12 h at 60 ºC, Finally, each

M

biosorbent was filtered, washed with distilled water until constant pH and then dried for 24 h at 70 ºC. The materials will be referred to hereafter as PB-Gluta and PB-Hypo,

ED

respectively.

Xanthation process was performed according to methodology indicated in various published works [9, 11–13] with some modifications. Firstly, about 30 g of PB

PT

was suspended in 300 mL 1.0 M NaOH solution and stirred for 24 h at room temperature. This initial step was conducted to remove lignin, hemicellulose, and low

CC E

molecular weight matter. As result, an increase in the reactivity and accessibility of hydroxyl groups in cellulose is obtained [15]. After filtration, the material was washed with distilled water until the solution reaches constant pH, and finally dried for 24 h at

A

70 °C (PB-wash). About 2 g of PB-wash, PB and MPB were treated with 30 mL of 4.0 M NaOH solution and stirred for 3 h. Then, 2 mL of CS2 were added to each mixture and shaken for 3 h. Finally, these materials were washed with distilled water and acetone until constant pH and then dried at 70 °C for 24 h. The prepared biosorbent will be mentioned next as XPB, PB-Xant and MPB-Xant, respectively.

5

The differences between the sorption capacities of the developed biosorbent were assessed with the one-way ANOVA test, with p value (significance level) of 5%. All statistical analyses were performed using Microsoft Excel software, version 2016. 2.3. Biosorbents characterization The specific surface area, average pore diameter and pore volume were obtained by N2 adsorption at 77 K using ASAP 2000 equipment, Micromeritics. The particle size

IP T

distribution was measured by Dynamic Light Scattering (DLS) from Malvern. The zeta potential was measured at different pH (from 2.0 to 10.4) using an analyser Malvern Zetasizer nano ZS. FTIR (Fourier Transform Infrared spectroscopy) was used to

SC R

determine FTIR spectra, with a spectrometer PerkinElmer, Frontier mode. Surface

morphology was analysed before and after of treatment by means of SEM (Scanning

U

Electron Microscopy) and EDS (Energy Dispersive X-ray Spectroscopy). 2.4.Equilibrium experiments

N

The effect of initial solution pH was tested in the pH range of 2.03 to 5.40

A

adjusted by either using 1.0 mol/L NaOH or HCl as required. The batch tests were

M

conducted by putting 0.2 g of adsorbent in contact with flasks containing 20 mL of 500 mg/L of Cr(III) solution. The flasks were agitated in a thermostatic water bath shaker at

ED

25 ºC for 3h, to achieve equilibrium conditions. Aliquots were taken and filtrated with a FILTER-LAB filter with 0.45 μm of pore size. Sorption equilibrium isotherms were obtained with a solution of 500 mg/L of

PT

Cr(III), in which the pH was adjusted to 5.0 using 1.0 mol/L NaOH. In each batch, 20 mL of Cr(III) solution contacted with a dosage of biosorbent ranged from 2 to 15 g/L.

CC E

The suspensions were continuously agitated for 3 h at 25 ºC and then filtrated. The amount of metal adsorbed, 𝑞𝑒 , (mg/g) and the removal efficiency, η, (%) were calculated according to Eqs. (2) and (3), respectively. (𝐶0 − 𝐶𝑒 )𝑉 𝑚 𝐶0 − 𝐶𝑒 = 100 𝐶0

A

𝑞𝑒 =

(2) (3)

where 𝑚 is the mass of biosorbent used (g), 𝐶0 is the initial concentration, 𝐶𝑒 is the equilibrium concentration (mg/L) and 𝑉 is the volume of the solution (L).

6

The adsorption isotherms were evaluated using the Langmuir and Freundlich models, Eqs. (4) and (5), respectively, 𝑞𝑒 =

𝑞𝑚𝑎𝑥 𝐾𝐿 𝐶𝑒 1 + 𝐾𝐿 𝐶𝑒

(4) (5)

1/𝑛

𝑞𝑒 = 𝐾𝐹 𝐶𝑒

where 𝑞𝑚𝑎𝑥 is the maximum adsorption capacity (mg/g) and 𝐾𝐿 is the Langmuir

IP T

constant (L/mg); 𝐾𝐹 is the Freundlich constant (mg1-(1/n) L1/n /g) and 1/n is the equilibrium constant indicative of adsorption intensity and associated to the heterogeneity of the adsorbent surface.

SC R

To evaluate the sorption mechanism of Cr(III) by ion-exchange, it was

determined its relationship with sodium ions released from XPB adsorbent through the definition of an apparent equilibrium constant for the exchange reaction. The speciation

U

of Cr(III) as a function of pH should be considered since it is important to know which

N

predominant hydroxo-Cr(III) species can exchange with the sodium ions. For pH 5.0, acidic conditions used in the determination of the equilibrium data, Cr(III) is mostly in

A

the form of Cr(OH)2+ (70%), while it coexists with other minor species. Thus, the

M

following equilibrium reactions may be considered [24]: pK1(25 ºC) = 3.85

(6)

+ 𝐶𝑟 +3 + 2𝐻2 𝑂 ↔ 𝐶𝑟(𝑂𝐻)+ 2 + 2𝐻

pK2(25 ºC) = 10.06

(7)

ED

𝐶𝑟 +3 + 𝐻2 𝑂 ↔ 𝐶𝑟(𝑂𝐻)+2 + 𝐻 +

From the definition of the equilibrium constants for Eqs. 6 and 7, the Cr(OH)2+ concentration was calculated as:

𝐶𝐶𝑟𝑡 [𝐻 + ] 𝐾2 [1 + + ] + 𝐾1 𝐾1 [𝐻 ]

PT

𝐶𝐶𝑟(𝑂𝐻)2+ =

(8)

CC E

where 𝐶𝐶𝑟𝑡 corresponds to the total molar concentration of all species of Cr(III) in solution at pH 5.0.

The desorption studies were performed by soaking a certain amount (0.08 g) of

A

pre-saturated biosorbent in 40 mL of ethylenediaminetetraacetic acid disodium salt dihydrate (EDTA) (0.05 M or 0.1 M), N,N-bis (carboxymethyl)-DL-alanine trisodium salt (N,Nbis) (0.05, 0.1 M), hypophosphorous acid 50% (H3PO2), HCl (0.5 M or 1 M), HNO3 (0.5 or 1.0 M) and H2SO4 (0.5 or 1.0 M) solutions. The flasks were sealed and agitated in a thermostatic water bath shaker for 24 h. The solid phase was separated by

7

filtration and aliquots of the liquid were analysed. The desorption efficiency, 𝜂𝑑𝑒𝑠. , (%) was determined as follows: 𝜂𝑑𝑒𝑠 =

𝐶𝑑𝑒𝑠 𝑉 100 𝑞𝑒 𝑚

(9)

where 𝐶𝑑𝑒𝑠 is the final ion concentration (mg/L) in the desorption step. Kinetic experiments were carried out by adding 0.2 g of XPB into 20 mL solution with a certain initial Cr(III) concentration at pH 5. The suspensions were

IP T

continuously agitated at 25 °C. In this study, the effect of the initial concentration (250 and 500 mg/L of Cr(III)) was evaluated. Aliquots were taken between 2 and 180 min and the concentration of chromium was analysed. Two kinetic models were tested to

SC R

describe experimental data: the pseudo-first-order (Eq. 10) and the pseudo-second-order (Eq. 11) [21,25]. 𝑞𝑡 = 𝑞𝑒 (1 − 𝑒 −𝑘𝑓 𝑡 ) 𝑞𝑒 2 (𝑘𝑠 𝑡) (1 + 𝑞𝑒 𝑘𝑠 𝑡)

(11)

U

𝑞𝑡 =

(10)

N

where 𝑞𝑡 is the quantity adsorbed (mg/g), kf, is the pseudo-first-order rate constant

A

(min-1), 𝑡 is the time (min), ks is the equilibrium rate of pseudo-second order (g/mg

M

min).

The initial sorption rate may be estimated by deriving Eq. (11) leading to Eq. (12), ℎ = 𝑘𝑠 𝑞𝑒 2

ED

(12)

where ℎ represents the amount of adsorbate per mass of adsorbent and time (mg/g min). The value of ℎ can be used to compare the initial sorption rate of metals on various

CC E

[25].

PT

adsorbents at similar conditions (pH, solid/liquid ratio, initial metal concentrations, etc.)

2.5. Fixed-bed column experiments Column tests were performed in a glass column of an internal diameter of 1.5 cm

A

and a length of 2.5 cm packed with 1.8 g of biosorbent particles. A peristaltic pump Minipuls3-Gilson was used to feed the column with the inlet

solution at a predefined flow rate of 10 mL/min. Initially, a low flow rate of ultrapure water was percolated to moisten the bed and prevent a preferential flow during the dynamic tests. Cr(III) solutions with different initial concentrations, CE, (230 and 500 mg/L) were percolated through the column in a down flow mode at room temperature. Aliquots were collected at the exit of the column and analysed with respect to 8

chromium concentration. In the regeneration test, the procedure used was the same as the one employed in the saturation experiments, but the solution fed to the bed (previously saturated) was the regenerate agent. The dynamic behaviour of columns was assessed from the analysis of the breakthrough curves expressed as outlet concentration in function of time [26]. The breakthrough curve may be described based on two important moments: the breakthrough time (tbp) and the stoichiometric time (tst). For practical reasons, tbp is

IP T

defined as the time required for the exit concentration, C, is about 1% of 𝐶𝐸 and tst is defined by Eq. (13) [27], (1 −

0

𝐶 ) 𝑑𝑡 𝐶𝐸

(13)

SC R

𝑡∞

𝑡𝑠𝑡 = ∫

The area under the breakthrough curve represents the total mass of metal adsorbed, 𝑚𝑎𝑑𝑠 (mg), for a given feed concentration and flow rate, the amount of metal adsorbed

U

by mass of adsorbent, 𝑞𝑒𝑎𝑑𝑠 (mg/g), and the desorption efficiency, 𝜂𝑑𝑐 (%), can be

(15)

(16)

ED

𝜂𝑑𝑐

A

𝑚𝑎𝑑𝑠 𝑚 𝑡𝑓𝑖𝑛𝑎𝑙 𝑄 𝐶𝑑 𝑑𝑡 ∫0 1000 = ( ) 100 𝑞𝑎𝑑𝑠

𝑞𝑒𝑎𝑑𝑠 =

(14)

M

𝑚𝑎𝑑𝑠

𝑡𝑓𝑖𝑛𝑎𝑙 𝐶𝐸 𝑄 𝑡𝑓𝑖𝑛𝑎𝑙 𝑄 =( ∫ 𝐶 𝑑𝑡) )− ( 1000 1000 0

N

determined as follows:

PT

where 𝑄 is the volumetric flow rate (mL/min) and 𝐶𝑑 𝑖𝑠 the concentration of metal desorbed (mg/L).

Three different empirical models were fitted to the experimental breakthrough

CC E

curves, which allowed to estimate kinetic parameters related to the Cr(III) biosorption in the fixed-bed column: Bohard-Adams, Eq. (17), Thomas, Eq. (18), and Yoon-Nelson,

A

Eq. (19), models [28,29],

𝐶𝑡 = 𝐶𝐸

𝑒𝑥𝑝 (𝑘𝑏𝑎 𝐶𝐸 (𝑡 −

𝐿 )) 𝑢0

𝑘𝑏𝑎 𝑞𝑏𝑎 𝐿𝜌𝑝 𝐿 𝑒𝑥𝑝 (𝑘𝑏𝑎 𝐶𝐸 (𝑡 − )) + 𝑒𝑥𝑝 ( )−1 𝑢0 𝑢0 𝐶𝐸 𝐶𝑡 = 𝑘 𝑞 𝑚 1 + exp ( 𝑡ℎ 𝑡ℎ − 𝑘𝑡ℎ 𝐶𝐸 𝑡) 𝑄

(17)

(18)

9

𝐶𝐸

𝐶𝑡 =

(19) 1 + exp (𝑘𝑦𝑛 (𝜏ℎ − 𝑡)) where 𝐶𝐸 is the feed concentration (mg/L), 𝑘𝑏𝑎 is the rate constant of Bohard-Adams model (L/mg min), 𝑞𝑏𝑎 is the maximum adsorption capacity of Bohard-Adams model (mg/g), 𝑢0 is the superficial velocity (cm/min), 𝑘𝑡ℎ is the Thomas rate constant (mL/mg min), 𝑞𝑡ℎ is the maximum adsorption capacity of Thomas model (mg/g), 𝑘𝑦𝑛 is the rate constant of Yoon-Nelson model (min-1) and τh is the time required for 50% solute

IP T

breakthrough, 𝑡50% (min).

The fraction of saturated bed (𝐹𝑆𝐵) is necessary to scale-up design and

SC R

represents the ratio between the total mass of Cr(III) adsorbed in the column until the breakthrough time, tbp, and the mass of chromium fed into the column. FSB was calculated as: 𝐸

∞ ∫0 (1

𝐶 − ) 𝑑𝑡 𝐶𝐸

(20)

N

𝐹𝑆𝐵 =

𝐶 (1 − 𝐶 ) 𝑑𝑡

U

𝑡𝑏𝑝

∫0

A

To determine the height of the adsorption column in the scale-up, the length of unused bed (LUB) model was applied. This method uses the LUB at the tbp as a

M

parameter to characterize the breakthrough behavior [30]. Accordingly, the LUB is 𝐿𝑈𝐵 = ℎ − ℎ𝑠𝑡

(21)

ED

given by:

where ℎ and ℎ𝑠𝑡 are the height of the laboratory column and the location of the

PT

stoichiometric front (cm), respectively. Since the ℎ𝑠𝑡 moves with the same velocity as the real front, the travel velocity, 𝑢𝑧 , can be defined by using either the real tbp or the stoichiometric time, 𝑡𝑠𝑡 . ℎ𝑠𝑡 ℎ = 𝑡𝑏𝑝 𝑡𝑠𝑡

CC E 𝑢𝑧 =

(22)

where 𝑢𝑧 is the traveling velocity of the mass transfer zone (MTZ) within the column.

A

By combining Eqs. (21) and (22) gives: 𝐿𝑈𝐵 = 𝑢𝑧 (𝑡𝑠𝑡 − 𝑡𝑏 ) =

𝑡𝑠𝑡 − 𝑡𝑏 ℎ 𝑡𝑠𝑡

(23)

For scale-up purposes, by setting the breakthrough time value for the operation at industrial scale (𝑡𝑏𝑟 ), the height of column (ℎ𝑟 ) can be calculated by Eq. (24) [30]. 𝑡𝑏𝑟 =

1 (ℎ − 𝐿𝑈𝐵) 𝑢𝑧 𝑟

(24)

10

The required mass of adsorbent for the industrial column can be determined by Eq. (25), 𝑚 (25) 𝜌𝑏 = 𝑉 where 𝜌𝑏 is bulk density (kg/m3).

3. Results and discussion

IP T

3.1. Effect of chemical modification of pine bark The effect of the chemical treatment of PB on the sorption capacity (mg/g) and

removal efficiency (%) of Cr(III) ions is shown in Fig. 1 (a) and Fig. 1 (b), respectively.

SC R

Bars identified with the same letter are not significantly different, according to ANOVA test at p < 0.05 (significance level of 5%). The untreated PB exhibits a very low

U

adsorption capacity for chromium ions (8.40 mg/g).

The treatment with glutamic and hypophosphorous acids did not significantly improve

N

the sorption capacity, contrary to some literature studies [31–34]. A superior sorption

A

capacity (qe  22 mg/g) was obtained with the MPB, which corresponds to a removal

M

efficiency of 40.8% of Cr(III). It is worth mentioning that the treatment with NaOH (mercerization) has been widely used to improve the surface properties of

ED

lignocellulosic residues leading to better sorption capacities towards the metal ions compared to the initial raw materials [20,33,35]. However, the xanthate materials (treated with CS2) presented in this study the highest Cr(III) removal efficiencies.

PT

Indeed, as reported in the literature, the xanthate group has a high tendency to form stable complexes with heavy metal ions [9,12,16]. The materials that underwent

CC E

xanthation treatment (MPB-Xant, XPB and PB-Xant) presented removal efficiencies of 99.0 (qe = 50.3 mg/g), 95.2 (qe = 48.4 mg/g) and 83.7 % (qe = 42.5 mg/g), respectively. However, according to ANOVA results, the differences between MPB-Xant and XPB

A

were not statistically significant (marked in Fig. 1 with the same letter: d). Thereby, the MPB-Xant material was not selected for future studies because in this case, an additional treatment of mercerization precedes xanthation, which increases the costs and difficulties of production of the final adsorbent. Thus, XPB material was chosen to proceed with further studies.

11

3.2. Characterization of the selected adsorbent 3.2.1. N2 adsorption analysis To elucidate the textural properties of PB and XPB, adsorption-desorption isotherms of N2 at 77 K were obtained (Fig. 2). Some results provided by these isotherms are summarized in Table 1. Both PB and XPB adsorbents exhibited low BET

IP T

surface areas and very low porosities.

No significant differences with respect to the surface area (SBET) were detected between

SC R

PB and XPB. However, it was found that the xanthation modification of PB decreased

the average pore diameter and the pore volume due to the blockage of internal porosity by incorporation of CS2. The N2 isotherms for PB and XPB showed in Fig.2 are of Type IV, according to IUPAC classification, which corresponds to the presence of a large

U

amount of mesopores and/or macropores. The XPB exhibits a more pronounced

N

hysteresis probably as result of the higher occurrence of mesoporous in its structure

A

when compared with the PB. Regarding the particle size distributions of the materials, it is noticeable that 50% of the particles (d50) of PB and XPB have diameters less than

ED

caused an increase in this property.

M

0.152 and 0.128 mm, respectively. With respect to density, the xanthation treatment

Zeta potential

PT

The zeta potentials of PB and XPB measured at different pH are shown in Fig. 3. The surface charge distribution curves are related to the type and quantity of surface

CC E

functional groups [36]. The slope change of these curves can indicate the dissociation of functional groups, due to chemical treatment with NaOH and CS2. The zero charge point (isoelectric point) for PB was located at pH 2.3. For XPB the absolute values of zeta potential become more negative and exhibit all negative values within the pH range

A

tested. Thus, it can be concluded that XPB has more functional groups formed that can ionize making them become more negative in solution [37]. This may also indicate that the XPB has greater physical stability and surface activity [12].

12

FTIR analysis The FTIR spectra of PB and XPB adsorbents are shown in Fig. 4. For comparison purposes, the spectrum of XPB after contacting with a Cr(III) solution (referred to XPB-Cr(III)) was also included in this figure. The intense and wide absorption peaks near at 3342 and 2927 cm-1 can be attributed to O-H and C-H groups stretching features of the cellulose, hemicelluloses and lignin

IP T

[20,38]. The peak at 1601 and 1270 cm-1 may be assigned to symmetric stretching of

C=O in ionic carboxylic groups (−COO-)[12,36,39]. The broad peak observed at 1021

SC R

cm-1 involves the C-O stretching vibrations of aliphatic alcohols in cellulose,

hemicellulose, and lignin [9]. A comparison of FTIR spectra of PB and XPB reveals that the peaks at 3342 and 1601cm-1 increased, which may be due to the NaOH solution used for the xanthation treatment. It is well-known that the treatment of the

U

lignocellulosic material with alkali solution may hydrolyse methyl esters of cellulose,

N

hemicellulose, lignin, and pectin. The bonds associated with fatty methyl ester can be

A

saponified to alcoholic (-OH) and carboxylate (-COO-) ligands [40] as shown by Eq. (26).

M

𝑅 − 𝐶𝑂𝑂𝐶𝐻3 + 𝑁𝑎𝑂𝐻 → 𝑅 − 𝐶𝑂𝑂− + 𝐶𝐻3 𝑂𝐻 + 𝑁𝑎 +

(26)

This result supports zeta potential behaviour, where XPB is more negatively charged

ED

than PB (Fig. 3). In the XPB, the increase in the intensity of the peak at 1021 cm-1 is attributed to C=S and C-O stretching vibrations, which are merged into a broad band

PT

[16,41]. The band at 538 cm-1 represent the C-S stretching vibrations [12,15]. The peaks at 1021 and 538 cm-1 are clearly indicative of the presence of the xanthate group bonded

CC E

to XPB [12,15]. Comparing the spectra of XPB and XPB-Cr(III), it is noticeable that the bands assigned to C=C, C=O, C=S, and C-S stretching decreased in intensity, indicating the contribution of carboxylic and xanthate groups in Cr(III) binding. The sorption of Cr(III) ions by XPB occurs through ion-exchange or complexation, or by a combination

A

of both processes [12]. Similar results also were reported in the literature to the binding mechanism between bivalent metal ion and xanthate materials [9,12,15].

SEM-EDS analysis Scanning electron microscopy (SEM) micrographs at 2000x magnifications and energy dispersive spectroscopy (EDS) are presented for PB, XPB, and XPB-Cr(III) in Figs. 5 (a), (b) and (c), respectively. 13

In Fig. 5 (a) is noticeable that the PB surface is heterogeneous and slightly porous, leading to a very low specific surface area (see Table 1). It has been revealed in a previous work that PB pore size distributions have a predominance of mesoporous sizes with diameters in the range of 10 to 500 Å [24]. The morphology of XPB is markedly different from PB. Changes in the surface properties can occur either by masking or

IP T

removing functional groups or by exposing more binding sites [1]. Chemical treatment with NaOH affects the lignin present in PB and the XPB surface becomes less smooth

SC R

with more visible pores. Comparing the surface of XPB-Cr(III) and XPB, surface modifications are not observed since the sorption process should not modify the

morphology of the adsorbent. With respect to the semiquantitative elemental analysis of EDS, it is shown that PB, MPB, and MPB-Cr(III) are composed mainly of carbon and

U

oxygen, as expected for lignocellulosic materials. The XPB surface revealed a high

N

content of sodium and sulfur due to the treatment of PB with NaOH and CS2. Surface EDS analysis of XPB-Cr(III) depicts that chromium ions were in fact adsorbed on the

A

biosorbent surface. In addition, the sodium content was reduced in this case, which may

ED

3.3. Equilibrium experiments

M

be the result of an ion-exchange between Na+ bound to XPB and Cr(III).

Effect of solution initial pH One of the most important variables in the sorption process is the pH of the

PT

solution which affects the surface charge of the biosorbent and the speciation of the metal. The influence of pH on Cr(III) sorption uptake by XPB is shown in Fig. 6. The

CC E

removal efficiency increased from 35.10 (at pH 2.03) to 98.48% (at pH 5.4).

The lower sorption capacity at lower pH was due to the decrease in the negative charge

A

density, in accordance with the one observed in Fig. 3. According to the literature, the dominant species of chromium at pH 2.0 and 5.0 is Cr3+ and Cr(OH)2+, respectively. The higher sorption capacity at pH 5.4 (50.23 mg/g) may be due to the interaction of Cr(OH)2+ with sulfur groups present on the XPB surface. In addition to xanthate groups (C-S2Na), the carboxylic acids (COOH) are also present in XPB and are responsible for Cr(III) sorption [9] as observed in the FTIR spectra (Fig. 4).

14

Effect of contact time The kinetic behaviour of Cr(III) uptake by XPB, at two different initial metal concentrations, is illustrated in Fig. 7.

It can be observed that the sorption rate and the equilibrium concentration depend on the initial metal concentration. The sorption rate is higher at the beginning of the process because of the greater number of active sites on the surface of the biosorbent. The

IP T

experimental data show that most Cr(III) was removed within 20 min and no significant

changes in terms of the metal removal were observed after 40 min. Besides, the sorption

SC R

rates are fast indicating easy interaction and accessibility of available binding sites by

Cr(III) [14]. The kinetic parameter values obtained from the fitting of the pseudo-first order and pseudo-second order models to experimental data are presented in Tab. 2. The pseudo-second order model describes better the kinetic of Cr(III) sorption for the two

U

concentrations tested, as indicated by the highest coefficient of determination, R2, and

N

the lowest root mean square error, RMSE. Moreover, the estimated amounts of metal

A

adsorbed in equilibrium, 𝑞𝑒 , (23.57 and 46.50 mg/g for 𝐶0 of 250 and 500 mg/L, respectively) from the pseudo-second order equation were close to the experimental

M

data. The initial sorption rate (h) displayed in Tab. 2 shows an increase from 86.84 to 136.52 mg/g min when 𝐶0 rise from 250 to 500 mg/L. Higher h values are associated to

PT

ED

a greater driving force for the diffusion of Cr(III) ions within the adsorbent particles.

Effect of biosorbent dosage and sorption isotherm

The effect of the biosorbent dosage on the removal efficiency and the sorption

CC E

equilibrium isotherm for Cr(III) is shown in Fig. 8. The results display that the removal efficiency is strongly dependent on the biosorbent dosage. For higher XPB dosage values, larger removal efficiency (Fig. 8 (a)) is achieved. For example, with a fixed

A

dosage of 10 g/L of XPB, the removal efficiency of Cr(III) was of 93.7%. Indeed, for high biosorbent dosages, a larger number of sorption sites is available for sorbent-solute interaction, thereby resulting in the increased percentage of metal removal from solution. Similar results were reported by other works in the literature [25,39]. The experimental and calculated equilibrium data for sorption of Cr(III) into XPB are depicted in Fig. 8 (b).

15

The equilibrium isotherm parameters (Table 3) were obtained by nonlinear regression through the fitting of the Langmuir and Freundlich models to the experimental data.

The Langmuir isotherm is the best model for describing the experimental data, considering the higher 𝑅2 and lower RMSE. This model assumes that sorption occurs at specific homogeneous sites on the biosorbent and is used successfully in many

IP T

monolayer adsorption processes. The maximum sorption capacity of Cr(III) onto XPB was 56.47 mg/g and the Langmuir constant, 𝐾𝐿 was found to be 0.178 L/mg. The

separation factors (RL = 1/(1 + 𝐾𝐿 𝐶0 )) of 0.01 is consistent with the requirements for

SC R

favourable biosorption (0<𝑅𝐿 <1) [9,42]. Ion-exchange mechanism of Cr(III)

U

To understand the Cr(III) sorption mechanism by ion-exchange, the equilibrium data were analysed by evaluating the relationship between the number of chromium ions

N

exchanged with the sodium ions released by the XPB. In Fig. 9 (a) it can be seen the

A

effect of the dosage of XPB on the equilibrium concentration of Na+ in solution. The increase in dosage resulted in a higher amount of Na+ released by the XPB, as expected.

M

Fig. 9 (b) shows that there is a linear correlation (R2 = 0.94) between Na+ and Cr(OH)2+ equilibrium concentrations in solution at the initial pH of 5.0. The absolute value of the

ED

slope (Δ𝐶𝑒,𝑁𝑎 /Δ𝐶𝑒,𝐶𝑟(𝐼𝐼𝐼) ) is  2.0. This suggests that the amount of Na+ released from XPB and the adsorbed chromium corresponds to a molar ratio of approximately 2:1.

PT

Therefore, the ion-exchange sorption mechanisms between Cr(III) and Na+ ions may be described by Eq. (27). The concentration of the predominant chromium species

CC E

(Cr(OH)2+) at pH 5.0 was calculated by Eq. (8). Hence, the removal mechanism includes the reaction between the species predominant of Cr(III) complex and the sulphonic group to form Cr(OH)2+/sulphonate surface complex with the release of two

A

ions of sodium, as shown in Eq. (27): 2(−𝑂𝐶𝑆2 𝑁𝑎) + Cr(OH)2+ ↔ (−𝑂𝐶𝑆2 )2 Cr(OH) + 2Na+

(27)

where (−𝑂𝐶𝑆2 𝑁𝑎) and (−𝑂𝐶𝑆2 )2 Cr(OH) are the unreacted sulphonic ligand on the XPB surface (mg/g) and the chromium-sulphonate complex (mg/g), respectively. An equilibrium constant, 𝐾, (g/L) can be derived from Eq. (27), using the law of action mass and assuming ideality for the both phases [43]: 16

𝐾=

[(−𝑂𝐶𝑆2 )2 Cr(OH)] [𝑁𝑎+ ]2 (𝑄𝑚𝑎𝑥 − [(−𝑂𝐶𝑆2 )2 Cr(OH)])2 [𝐶𝑟(𝑂𝐻)2+ ]

(28)

where 𝑄𝑚𝑎𝑥 is the total ion-exchange capacity (mmol/g). To calculate the adsorbed amount of Cr(III) complex Eq. (29) can be obtained by manipulating Eq. (28): 𝑞𝑒,𝑐𝑎𝑙𝑐. = [(−𝑂𝐶𝑆2 )2 Cr(OH)] =

([𝑁𝑎+ ]2 − [𝑁𝑎 + ] √[𝑁𝑎+ ]2 + 4𝑄𝑚𝑎𝑥 𝐾[𝐶𝑟(𝑂𝐻)2+ ] + 2𝑄𝑚𝑎𝑥 𝐾[Cr(OH)2+ ]) 2𝐾[𝐶𝑟(𝑂𝐻)2+ ]

(29)

IP T

Fig. 9 (c) shows an appropriate agreement between the experimental and calculated ionexchange equilibrium data. Thus, the proposed model based on Eq. (29) allows a good

description of the ion-exchange mechanism for Cr(III) sorption by XPB. The estimated

SC R

parameter values of Eq. (29) are: 𝐾 = 208.09 g/L and 𝑄𝑚𝑎𝑥 = 1.346 mmol/g (~ 70 mg Cr /g of XPB). These values were obtained through a non-linear regression with 𝑅2 equal to 0.79 and RMSE of 0.17.

U

Desorption studies

To evaluate the possibility of recovering Cr(III) ions from loaded XPB and

N

consequently to reuse the adsorbent, desorption tests were done with different

A

regenerant agents (HCl, HNO3, H3PO2, H2SO4, EDTA and biodegradable complexing

M

agent N,Nbis). The results are shown in Table 4. The chelating agents, such as EDTA and N,Nbis, exhibit very low desorption efficiencies. This means that the chelating

ED

agents have a very low affinity for chromium ions since the metal is hardly displaced from the resin to form chelates. Among the mineral acids tested, H2SO4 was

PT

significantly better than all others for elution of Cr(III) from XPB-Cr(III). The highest efficiencies were obtained with H2SO4, as follows: 53.10, 59.2 and 67.3%

CC E

at concentrations of 0.5, 1.0 and 2.0 M, respectively. It is known that acids solutions are suitable eluents for most metallic ions because causes the protonation of the biosorbent surface with H+ (pH decreases, and oxygen groups protonates). Thus, the positively charged metal ions adsorbed on the functional groups of the adsorbent are released [36].

A

The most probable mechanism for chromium desorption with H2SO4 is described by the following reactions [36]: (−𝑂𝐶𝑆2 )3 𝐶𝑟 + 3 𝐻 + ↔ 3 (−𝑂𝐶𝑆2 ) + 𝐶𝑟 3+

(30)

(−𝑂𝐶𝑆2 )2 𝐶𝑟(𝑂𝐻) + 2 𝐻 + ↔ 2 (−𝑂𝐶𝑆2 ) + 𝐶𝑟(𝑂𝐻)2+

(31)

(−𝑂𝐶𝑆2 )4 𝐶𝑟2 (𝑂𝐻) + 4 𝐻 + ↔ 4 (−𝑂𝐶𝑆2 ) + 𝐶𝑟2 (𝑂𝐻)2 4+

(32)

17

Analysis of Cr(III) sorption in fixed-bed column Fig. 10 (a) shows the experimental and the simulated breakthrough curves (through Eq. 18), and Fig. 10 (b) depicts the experimental elution curves for initial concentrations (CE) of Cr(III) of 230 and 500 mg/L.

Fig. 10 (a) displays the effect of increasing feed concentration from 230 to 500 mg/L on the breakthrough curve, which becomes steeper and the adsorbent (XPB) reaches

IP T

saturation conditions more rapidly. The breakthrough time (tbp) and the stoichiometric time (tst) are lower for the highest feed concentration of Cr(III). The tbp decreased

SC R

36.48% (from 14.8 to 9.4 min) and tst diminished 22.85% (from 25.99 to 20.05 min) when CE increased from 230 to 500 mg/L. This occurs because at higher initial

concentration, the driving force for mass transfer is stronger, which leads to a faster solute transport from liquid to solid phase [44,45]. Therefore, the binding sites become

U

saturated more quickly in the column, leading to an earlier breakthrough and exhaustion

N

time. To describe the dynamic behaviour of the column, the Bohard-Adams, Tomas and

A

Yoon-Nelson models (Eqs. 17 to 19) were fitted to the experimental data. Table 4 summarizes the parameters found for each model and the respective error parameters

M

(RSME and R2) for the two saturation tests (CE = 230 and 500 mg/L). The values of R2 obtained for the three empirical models are quite satisfactory and the differences

ED

between them are not significant. Hence, any of the empirical models reproduce accurately the experimental breakthrough curves. For this reason, the Thomas model

PT

(extensively used and considered as the most general) was plotted in Figs. 10 (a). The parameters of each model were estimated by fitting the models to the experimental data

CC E

using a nonlinear regression technique. The parameters related to the rate constants (𝑘𝑏𝑎 , 𝑘𝑡ℎ and 𝑘𝑦𝑛 ) decrease when CE rise. With respect to τh, its values are quite close to those found experimentally (𝑡𝑠𝑡 , time

A

required for 50% solute breakthrough). These results were expected due to the good agreement between experimental and calculated breakthrough curves. The parameters related to the maximum sorption capacity, 𝑞𝑏𝑎 and 𝑞𝑡ℎ , varied from 35.05 to 59.87 mg/g and from 30.92 to 53.04 mg/g, when CE increase from 230 to 500 mg/L, respectively. Since in the batch desorption studies (section 3.4.5), the solution 2.0 M H2SO4 was the one with the best performance (efficiency of 67.3%), this was used to regenerate the 18

XPB preloaded with Cr(III) in the column experiments. Fig. 10(b) shows the desorption (elution) curves along the time obtained during the regeneration of XPB previously loaded until saturation with Cr(III) for CE of 230 and 500 mg/L. In both cases, a peak in the elution curves is observed, where the concentrations are near 10 and 14 times higher than 230 and 500 mg/L, respectively. This means that H2SO4 promoted a fast desorption of the Cr(III) from XPB. After 5 min of operation, the elution slows down and controlled by dispersive effects, i.e. the curve exhibits a long tail. The desorption

IP T

efficiency calculated with Eq. (9) was 66.64 and 85.84% for CE of 230 and 500 mg/L,

SC R

respectively.

Scale-up design and economic assessment of XPB as an adsorbent The first step of the design was devoted to the calculation of the column

diameter from a required industrial flow rate and an optimal superficial velocity, 𝑢0 .

U

Thus, several simulations applying the Thomas model were performed using laboratory

N

conditions (Table S1 and Figure S1), which allows the selection of a superficial velocity of 0.038 cm/s, corresponding to an adequate bed utilization (FSB = 79%). The required

A

column diameter is 0.61 m, calculated by Eq. (33), considering a flow rate, 𝑄𝑟 = 400

M

L/h, generated by an electroplating industry located in the center region of Portugal. (33)

ED

𝑄𝑟 √ 𝑢0 4 𝑑= 𝜋

The height of column (ℎ𝑟 = 1.06 m)was calculated through Eq. (24) considering

PT

a breakthrough time of 14 h, to maintain the same ratio of length to diameter (L/D) of 1.7 as in the laboratory column. The required adsorbent (XPB) mass to fill the designed

CC E

column is 99.24 kg. All the input and calculated data used in the scale-up are summarized in Table 6. To implement a sustainable treatment of an industrial effluent, ensuring the reuse of a low-cost adsorbent and chemicals, the plant layout depicted in Fig. 11 is proposed. The

A

configuration of the adsorption system includes two columns of the same dimensions. When one of the columns is subjected to the saturation step (e.g. column 1), the other is in the regeneration phase (e.g. column 2), so that the process is continuous. The regeneration of the adsorbent may be performed with a solution of 2.0 M H2SO4 and then the column is washed with water. It should be noted that the effluent coming from the regeneration process does not generate secondary pollution. The 19

H2SO4 solution is recovered to regenerate the adsorbent and Cr(III) in the form of Cr2(SO4)3 can be reused in the industrial process. Gomes et al. (2010) showed the possibility of separate Cr(III) from an acidic solution of H2SO4 by nanofiltration, with an efficiency of around of 86% [46]. Table 7 summarizes the results from the mass balance calculations for one cycle of operation, e.g., one step of saturation (14h)

IP T

followed by one step of regeneration (50 min).

One of the most important parameters for the application of a real adsorption

SC R

unit is the cost of adsorbent. Therefore, there is a need for the development of locally

available low-cost materials that can be used more economically on a large scale. Pine bark is a very common waste in Portugal and its chemical modification by xanthation that uses NaOH and CS2 can be considered simple and relatively inexpensive. The cost

U

to produce XPB as well as the commercial price (the cost of transport not included) of

N

some cationic ion-exchange resins with affinity for heavy metals are shown in Table 8 [47,48]. The cost of XPB was estimated considering only the expenses with the reagents

A

used in its production. According to the chemical treatment protocol, 2.8 kg of NaOH

M

and 1 L of CS2 are required to produce 1 kg of XPB. The price found to obtain the XPB is reasonably low when compared with other materials. Therefore, the adsorbent

ED

investigated seems to a promising adsorbent from an economic point of view and with the potential to be used by small and medium industries.

PT

4. Conclusions

In case the material PB is washed with 1.0 M NaOH solution followed by a

CC E

xanthation treatment (XPB) the removal efficiency of Cr(III) may be significantly improved, from 16% to 95%. The treatment with organic acids did not significantly enhance the sorption capacity. The good properties of XPB for uptaking Cr(III) were confirmed through the assessment of several physicochemical parameters. Indeed, the

A

zeta potential, the presence of specific functional groups and morphology revealed the effect of the chemical treatment of PB. The adsorbed amount of Cr(III) is strongly dependent on initial pH, and the removal efficiency increased from 35.1 (at pH 2.03) to 98.5% (at pH 5.4). The sorption rate is very fast since most of Cr(III) in solution was removed within 20 min. The pseudo-second order model describes well the kinetic behavior observed for all concentrations tested. The equilibrium data were well modeled 20

by the Langmuir isotherm, with a maximum Cr(III) sorption capacity of 56.5 mg/g. The sorption mechanism based on the ion-exchange between chromium and sodium from XPB presented a molar ratio of approximately 1:2. An equilibrium constant (K) of 208.09 g/L and the total ion-exchange capacity (𝑄𝑚𝑎𝑥 ) of 1.346 mmol/g (~ 70 mg Cr(III)/g of XPB) was determined. The elution of Cr(III) ions from XPB was more efficiently using 2.0 M H2SO4 (67.3% efficiency). The dynamic behavior in fixed-bed column shows that an increase in the feed concentration of Cr(III) decreased both the

IP T

breakthrough and stoichiometric times. The dynamic curves were well modeled by

Bohard-Adams, Thomas or Yoon-Nelson models. The desorption efficiency on the

SC R

column was 66.64 and 85.84% to feed concentrations 230 and 500 mg/L, respectively.

A plant layout was proposed for large scale treatment of industrial effluents containing Cr(III) ions where the adsorbent is regenerated and chemicals (H2SO4 and Cr2(SO4)3)

U

can be reused.

N

Acknowledgement

The authors acknowledge the Conselho Nacional de Desenvolvimento Científico e

M

A

Tecnológico (Cnpq)” the supporting financial under grant nº 201264/2014-5.

[1]

ED

References

A. Abdolali, W.S. Guo, H.H. Ngo, S.S. Chen, N.C. Nguyen, K.L. Tung, Typical

PT

lignocellulosic wastes and by-products for biosorption process in water and wastewater treatment: A critical review, Bioresour. Technol. 160 (2014) 57–66. doi:10.1016/j.biortech.2013.12.037. F. An, B. Gao, X. Dai, M. Wang, X. Wang, Efficient removal of heavy metal

CC E

[2]

ions from aqueous solution using salicylic acid type chelate adsorbent, J. Hazard. Mater. 192 (2011) 956–962. doi:10.1016/j.jhazmat.2011.05.050.

A

[3]

S. Lata, P.K. Singh, S.R. Samadder, Regeneration of adsorbents and recovery of heavy metals: a review, Int. J. Environ. Sci. Technol. 12 (2015) 1461–1478. doi:10.1007/s13762-014-0714-9.

[4]

D.W. O´Connell, C. Birkinshaw, T.F. O´Dwyer, Heavy metal adsorbents prepared from the modification of cellulose: A review, Bioresour. Technol. 99 (2008) 6709–6724. doi:10.1016/j.biortech.2008.01.036.

[5]

W.S. Wan Ngah, M.A.K.M. Hanafiah, Removal of heavy metal ions from 21

wastewater by chemically modified plant wastes as adsorbents: A review, Bioresour. Technol. 99 (2008) 3935–3948. doi:10.1016/j.biortech.2007.06.011. [6]

V.K. Gupta, P.J.M. Carrott, M.M.L. Ribeiro Carrott, Suhas, Low-Cost adsorbents: Growing approach to wastewater treatmenta review, Crit. Rev. Environ. Sci. Technol. 39 (2009) 783–842. doi:10.1080/10643380801977610.

[7]

M. Fomina, G.M. Gadd, Biosorption: Current perspectives on concept, definition and application, Bioresour. Technol. 160 (2014) 3–14.

[8]

IP T

doi:10.1016/j.biortech.2013.12.102.

M. Ahmaruzzaman, Industrial wastes as low-cost potential adsorbents for the

SC R

treatment of wastewater laden with heavy metals, Adv. Colloid Interface Sci. 166 (2011) 36–59. doi:10.1016/j.cis.2011.04.005. [9]

S.S. Pillai, B. Deepa, E. Abraham, N. Girija, P. Geetha, L. Jacob, M. Koshy, Biosorption of Cd(II) from aqueous solution using xanthated nano banana

N

352–360. doi:10.1016/j.ecoenv.2013.09.003.

U

cellulose: Equilibrium and kinetic studies, Ecotoxicol. Environ. Saf. 98 (2013)

[10] S.E. Bailey, T.J. Olin, R.M. Bricka, D.D. Adrian, A review of potentially low-

A

cost sorbents for heavy metals, Water Res. 33 (1999) 2469–2479.

M

doi:10.1016/S0043-1354(98)00475-8.

York, 2010. [12]

ED

[11] O.B. and J.P.M. Jean-Luc Wertz, Cellulose Science and Technology, First, New

S. Liang, X. Guo, N. Feng, Q. Tian, Application of orange peel xanthate for the

PT

adsorption of Pb2+from aqueous solutions, J. Hazard. Mater. 170 (2009) 425– 429. doi:10.1016/j.jhazmat.2009.04.078. [13] P.L. Homagai, K.N. Ghimire, K. Inoue, Preparation and characterization of

CC E

charred xanthated sugarcane bagasse for the separation of heavy metals from aqueous solutions, Sep. Sci. Technol. 46 (2011) 330–339. doi:10.1080/01496395.2010.506903.

A

[14] J.K. Bediako, W. Wei, S. Kim, Y.S. Yun, Removal of heavy metals from aqueous phases using chemically modified waste Lyocell fiber, J. Hazard. Mater. 299 (2015) 550–561. doi:10.1016/j.jhazmat.2015.07.033. [15] C. Wang, H. Wang, G. Gu, Ultrasound-assisted xanthation of cellulose from lignocellulosic biomass optimized by response surface methodology for Pb(II) sorption, Carbohydr. Polym. 182 (2018) 21–28. doi:10.1016/j.carbpol.2017.11.004. 22

[16] P.L. Homagai, K.N. Ghimire, K. Inoue, Adsorption behavior of heavy metals onto chemically modified sugarcane bagasse, Bioresour. Technol. 101 (2010) 2067–2069. doi:10.1016/j.biortech.2009.11.073. [17] J. Qu, X. Meng, H. You, X. Ye, Z. Du, Utilization of rice husks functionalized with xanthates as cost-effective biosorbents for optimal Cd(II) removal from aqueous solution via response surface methodology, Bioresour. Technol. 241 (2017) 1036–1042. doi:10.1016/j.biortech.2017.06.055.

IP T

[18] J. Sousa Uva, Inventário Florestal Nacional - A dinâmica da ocupação florestal do solo desde o séc. XIX a 2050, 2015.

SC R

[19] A. Oliveira, P. Moura, Boas Práticas Florestais para o Pinheiro Bravo. Centro Pinus - Associação para a Valorização da Floresta, Porto, 1999.

[20] A.L. Arim, D.F.M. Cecílio, M.J. Quina, L.M. Gando-Ferreira, Development and characterization of biosorbent with enhanced capacity for uptaking Cr(III) from

U

aqueous solutions, Can. J. Chem. Eng. 96 (2018) 855–864.

N

doi:0.1002/cjce.23029.

[21] R.B. Garcia-Reyes, J.R. Rangel-Mendez, Adsorption kinetics of chromium(III)

A

ions on agro-waste materials, Bioresour. Technol. 101 (2010) 8099–8108.

M

doi:10.1016/j.biortech.2010.06.020.

[22] M. Rafatullah, O. Sulaiman, R. Hashim, A. Ahmad, Adsorption of copper (II),

ED

chromium (III), nickel (II) and lead (II) ions from aqueous solutions by meranti sawdust, J. Hazard. Mater. 170 (2009) 969–977.

PT

doi:10.1016/j.jhazmat.2009.05.066. [23] J. Liu, M. Yan, Y.K. Zhang, K.F. Du, Study of glutamate-modified cellulose beads for Cr(III) adsorption by response surface methodology, Ind. Eng. Chem.

CC E

Res. 50 (2011) 10784–10791. doi:10.1021/ie200857n.

[24] A.L. Arim, M.J. Quina, L.M. Gando-Ferreira, Insights into the sorption mechanisms of Cr(III) by chemically modified pine bark, Chem. Eng. Technol.

A

41 (2018) 1378–1389. doi:10.1002/ceat.201800034.

[25] B. Al-Rashdi, C. Tizaoui, N. Hilal, Copper removal from aqueous solutions using nano-scale diboron trioxide/titanium dioxide (B2O3/TiO2) adsorbent, Chem. Eng. J. 183 (2012) 294–302. doi:10.1016/j.cej.2011.12.082. [26] M. Calero, A. Ronda, A. Pérez, A. Yáñez, M. Carmen, M.Á. Martín-lara, M. Calero, A. Ronda, A. Pérez, A. Yáñez, M. Carmen, A. Ronda, A. Pe, The scaleup of Cr3+ biosorption onto olive stone in a fixed bed column, Desalin. Water 23

Treat. 3994 (2016) 1–13. doi:10.1080/19443994.2016.1150209. [27] A.L. Arim, K. Neves, M.J. Quina, L.M. Gando-Ferreira, Experimental and mathematical modelling of Cr(III) sorption in fixed-bed column using modified pine bark, J. Clean. Prod. 183 (2018) 272–281. doi:10.1016/j.jclepro.2018.02.094. [28] Z. Xu, J. Cai, B. Pan, Mathematically modeling fixed-bed adsorption in aqueous systems, J. Zhejiang Univ. Sci. A. 14 (2013) 155–176.

IP T

doi:10.1631/jzus.A1300029.

[29] M. Calero, F. Hernáinz, G. Blázquez, G. Tenorio, M.A. Martín-Lara, Study of Cr

SC R

(III) biosorption in a fixed-bed column, J. Hazard. Mater. 171 (2009) 886–893. doi:10.1016/j.jhazmat.2009.06.082.

[30] E. Worch, Adsorption technology in water treatment: Fundamentals, processes, and modeling, 1st ed., Berlin, 2012.

U

[31] R. Leyva-Ramos, L.E. Landin-Rodriguez, S. Leyva-Ramos, N.A. Medellin-

N

Castillo, Modification of corncob with citric acid to enhance its capacity for

doi:10.1016/j.cej.2011.11.021.

A

adsorbing cadmium(II) from water solution, Chem. Eng. J. 180 (2012) 113–120.

M

[32] S. Hokkanen, E. Repo, M. Sillanpaa, Removal of heavy metals from aqueous solutions by succinic anhydride modified mercerized nanocellulose, Chem. Eng.

ED

J. 223 (2013) 40–47. doi:10.1016/j.cej.2013.02.054. [33] L.H. Velazquez-Jimenez, A. Pavlick, J.R. Rangel-Mendez, Chemical

PT

characterization of raw and treated agave bagasse and its potential as adsorbent of metal cations from water, Ind. Crops Prod. 43 (2013) 200–206. doi:10.1016/j.indcrop.2012.06.049.

CC E

[34] J. Yang, Y. Park, K. Baek, J. Choi, Removal of metal ions from aqueous solutions using sawdust modified with citric acid or tartaric acid, Sep. Sci. Technol. 45 (2010) 1963–1974. doi:10.1080/01496395.2010.493782.

A

[35] A.E. Ofomaja, E.B. Naidoo, S.J. Modise, Removal of copper(II) from aqueous solution by pine and base modified pine cone powder as biosorbent, J. Hazard. Mater. 168 (2009) 909–917. doi:10.1016/j.jhazmat.2009.02.106. [36] G.R. Bernardo, R.J. Rene, A. Torre, M. Catalina, Chromium (III) uptake by agrowaste biosorbents : Chemical characterization, sorption – desorption studies, and mechanism, J. Hazard. Mater. 170 (2009) 845–854. doi:10.1016/j.jhazmat.2009.05.046. 24

[37] N. Feng, X. Guo, S. Liang, Adsorption study of copper (II) by chemically modified orange peel, J. Hazard. Mater. 164 (2009) 1286–1292. doi:10.1016/j.jhazmat.2008.09.096. [38] S.S. Lam, R.K. Liew, X.Y. Lim, F.N. Ani, A. Jusoh, Fruit waste as feedstock for recovery by pyrolysis technique, Int. Biodeterior. Biodegrad. 113 (2016) 325– 333. doi:10.1016/j.ibiod.2016.02.021. [39] G. Tan, H. Yuan, Y. Liu, D. Xiao, Removal of lead from aqueous solution with

IP T

native and chemically modified corncobs, J. Hazard. Mater. 174 (2010) 740–745. doi:10.1016/j.jhazmat.2009.09.114.

SC R

[40] A. Abdolali, H.H. Ngo, W. Guo, J.L. Zhou, B. Du, Q. Wei, X.C. Wang, P.D. Nguyen, Characterization of a multi-metal binding biosorbent: Chemical

modification and desorption studies, Bioresour. Technol. 193 (2015) 477–487. doi:10.1016/j.biortech.2015.06.123.

U

[41] Y. Zhu, J. Hu, J. Wang, Competitive adsorption of Pb(II), Cu(II) and Zn(II) onto

161. doi:10.1016/j.jhazmat.2012.04.026.

N

xanthate-modified magnetic chitosan, J. Hazard. Mater. 221–222 (2012) 155–

A

[42] M. Rafatullah, O. Sulaiman, R. Hashim, A. Ahmad, Adsorption of methylene

M

blue on low-cost adsorbents: A review, J. Hazard. Mater. 177 (2010) 70–80. doi:10.1016/j.jhazmat.2009.12.047.

ED

[43] S.E. Gomez-Gonzalez, G.G. Carbajal-Arizaga, R. Manriquez-Gonzalez, W. De La Cruz-Hernandez, S. Gomez-Salazar, Trivalent chromium removal from

PT

aqueous solutions by a sol-gel synthesized silica adsorbent functionalized with sulphonic acid groups, Mater. Res. Bull. 59 (2014) 394–404. doi:10.1016/j.materresbull.2014.07.035.

CC E

[44] L. Tofan, C. Paduraru, C. Teodosiu, O. Toma, Fixed Bed Column Study on the Removal of Chromium (III) Ions From Aqueous Solutions By Using Hemp Fibers With Improved Sorption Performance, Cellul. Chem. Technol. 49 (2015)

A

219–229.

[45] M. Jain, V.K. Garg, K. Kadirvelu, Cadmium(II) sorption and desorption in a fixed bed column using sunflower waste carbon calcium-alginate beads, Bioresour. Technol. 129 (2013) 242–248. doi:10.1016/j.biortech.2012.11.036. [46] S. Gomes, A. Cavaco, M.J. Quina, L.M. Gando-ferreira, Nanofiltration process for separating Cr( III ) from acid solutions : Experimental and modelling analysis, 254 (2010) 80–89. doi:10.1016/j.desal.2009.12.010. 25

[47] M. Gupta, H. Gupta, D.S. Kharat, Adsorption of Cu(II) by low cost adsorbents and the cost analysis, Environ. Technol. Innov. 10 (2018) 91–101. doi:10.1016/j.eti.2018.02.003. [48] Alibaba.com Global trade starts here. https://www.alibaba.com. (Acessed 03

A

CC E

PT

ED

M

A

N

U

SC R

IP T

September 2018).

26

60

e

Removal efficiency (%)

e 40

qe (mg/g)

d

d

100

d d

50

30 c 20

b a

60 c b

40 a a

20

a

10

80

0

0

w)

ra B(

P

P

B XP

t

P

an B-X

t

B an MP B-X MP

PB

w) (ra

ta

PB

lu -G

PB

o yp -H

B XP

PB

nt -Xa

t B an MP B-X MP

IP T

P

o yp B-H

ta

lu B-G

A

CC E

PT

ED

M

A

N

U

SC R

(a) (b) Figure 1. Effect of treatment of PB in the (a) sorption capacity and (b) removal efficiency of Cr(III). (𝐶0 = 500 mg/L; initial pH = 5.0; biosorbent dose = 10 g/L; t = 3 h)

27

2.5 Adsorption PB Desorption PB Adsorption XPB Desorption XPB

Adsorbed volume (cm3/g)

2.0

1.5

1.0

IP T

0.5

0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Relative pressure (P/Pº)

A

CC E

PT

ED

M

A

N

U

SC R

Figure 2. Adsorption/desorption isotherms of N2 at 77 K.

28

10

PB XPB

-10

-20

-30

IP T

Zeta potential (mV)

0

-40

-50 3

4

5

6

7

8

9

10

11

SC R

2

pH

A

CC E

PT

ED

M

A

N

U

Figure 3. Zeta potentials of PB and XPB.

29

C-S

C=O

C=S C-O

C=O

C-H

O-H

PB Transmittance Units

XPB

3500

3000

2500

2000

1000

IP T

538

1021

1270

1500

500

Wavenumber (cm-1)

A

CC E

PT

ED

M

A

N

U

Figure 4. FTIR spectra of PB, XPB and XPB-Cr(III).

SC R

4000

1601

3342

2927

XPB-Cr(III)

30

(a) Element C O

C

Spectrum PB Wt% Error [wt%] 72.13 22.40 27.86 9.11

(b)

C

Element C O Na S

O S

(c)

N

U

Na

Spectrum XPB Wt% Error [wt%] 64.43 22.71 29.29 11.78 3.51 0.78 2.77 0.39

SC R

S

IP T

O

S O

A

C

M

Cr

S

ED

Na

Spectrum XPB-Cr(III) Element Wt% Error [wt%] C 57.76 20.09 O 32.88 12.43 Na 0.50 0.19 S 2.12 0.31 Cr 6.74 0.64 Cr

A

CC E

PT

Figure 5. SEM-EDS results for (a) PB, (b) XPB and (c) XPB-Cr(III).

31

100

50

80

60 30 40 20 20

10 0

IP T

qe (mg/g)

40

Removal efficiency (%)

60

0 1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

pH initial

A

CC E

PT

ED

M

A

N

U

SC R

Figure 6. Effect of initial pH on Cr(III) uptake by XPB (𝐶0 = 500 mg/L, biosorbent dose = 10 g/L and contact time = 3 h).

32

60 50

30

20

500 mg/L 250 mg/L Pseudo-first order Pseudo-second order

10

IP T

qt (mg/g)

40

0 50

100

150

200

250

SC R

0

t (min)

A

CC E

PT

ED

M

A

N

U

Figure 7. Effect of contact time on Cr(III) uptake by XPB (biosorbent dose = 10 g/L; pH initial = 5.0; T= 25 oC).

33

60 50

80

40

qe (mg/g)

60

30

40 20

20

experimental data Langmuir Freundlich

10 0

0 0

2

4

6

8

10

12

14

0

16

50

100

150

200

250

Ce (mg/L)

Adsorbent dosage (g/L)

300

IP T

Removal efficiency (%)

100

350

400

450

A

CC E

PT

ED

M

A

N

U

SC R

(a) (b) Figure 8. (a) Effect of biosorbent dosage and (b) sorption isotherms of Cr(III) on XPB at 25 ºC, pH initial = 5.0, 𝐶0 = 500 mg/L.

34

6

150 5

Ce Na+ (mmol/L)

90

60

30

4 3

2 1

0

0 0

2

4

6

8

10

12

14

0

16

1

(a)

3

(b)

4

5

6

SC R

1.2 1.0

qe, Cr (mmol/g)

2

Ce Cr(OH)2+ (mmol/L)

Dosage (g/L)

IP T

Ce Na+ (mg/L)

120

0.8 0.6

U

0.4 qe exp. qe calc.

N

0.2 0.0 0

50

100

150

200

250

300

350

400

450

A

Ce Cr total (mg/L)

A

CC E

PT

ED

M

(c) Figure 9. (a) Na+ concentration as a function of the dose of XPB, (b) Relationship between chromium and sodium concentrations, (c) experimental and calculated ion-exchange equilibrium

35

7000

500 450

6000

400

5000

250 200 150

4000 3000 2000

500 mg/L 230 mg/L Thomas model

100 50

Desorption (500 mg/L) Desorption (230 mg/L)

1000

0

0 10

20

30

40

50

60

70

0

5

10

t (min)

15 t (min)

20

IP T

300

C (mg/L)

C (mg/L)

350

25

30

A

CC E

PT

ED

M

A

N

U

SC R

(a) (b) Figure 10. Effect of feed Cr(III) concentration on the experimental (a) breakthrough curves and (b) elution profiles with H2SO4 (𝑄 = 10 mL/min; H = 2.5 cm; initial pH = 5.0)

36

Nanofiltratio Nanofiltration n membrane

Cr2(SO4)3

membrane

make-up H2SO4

1

Adsorption column

H2SO4

2 2M H2SO4 solution tank

IP T

Water tank

SC R

Effluent

Discharge water line

Feed pump

A

CC E

PT

ED

M

A

N

U

Figure 11. Proposed plant layout for removal of Cr(III) using XPB as adsorbent.

37

Table 1. Physical characteristics of PB and XPB. PB 4.95 0.006 49.7 1.10 0.152

XPB 4.47 0.003 29.7 1.60 0.128

A

CC E

PT

ED

M

A

N

U

SC R

IP T

SBET [m2/g] Pore volume [cm3/g] Average pore diameter [Å] Real density [g/cm3] Median particle size [mm]

38

Table 2. Adsorption kinetic parameters. Pseudo-first order 𝒌𝒇 𝒒𝒆

𝑹𝟐

(mg/L)

(mg/g)

(min-1)

250

22.967

1.493

0.797

500

46.92

1.439

0.626

RMSE

Pseudo-second order 𝒒𝒆 𝒌𝒔

𝑹𝟐

RMSE

h (mg/g min)

(mg/g)

(g/mg min)

0.666

23.444

0.158

0.974

0.236

86.84

2.100

48.205

0.066

0.946

0.616

153.37

∑( 𝑞− 𝑞 *Root mean square error: 𝑅𝑀𝑆𝐸 = √

𝑐𝑎𝑙 )2

𝑁

IP T

C0

, 𝑞 and 𝑞𝑐𝑎𝑙 are the measured and calculated values of

A

CC E

PT

ED

M

A

N

U

SC R

metal uptake, respectively (mg/g), and N is the number of experimental samples.

39

1/n 8.928

𝑹𝟐

RMSE

0.901

3.151

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Table 3. Langmuir and Freundlich model parameters Langmuir Freundlich 𝒒𝒎𝒂𝒙 𝑲𝑳 𝑲𝑭 RMSE 𝑹𝟐 (mg/g) (L/mg) (mg1-(1/n) L1/n/g) 56.468 0.178 0.992 0.914 29.939

40

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Table 4. Desorption efficiencies for removal of Cr(III) from XPB with different regenerant agents. Regeneration agent Desorption Regeneration agent Desorption efficiency (%) efficiency (%) EDTA 0.05 M 6.53±0.00 HNO3 1.0 M 14.75±0.94 EDTA 0.1 M 2.78±0.26 H3PO2 0.5 M 18.37±0.72 N,Nbis 0.05 M 1.95±0.04 H3PO2 1.0 M 26.54±0.87 N,Nbis 0.1 M 1.48±0.00 H3PO2 2.0 M 29.90±2.85 HCl 0.5 M 1.32±0.35 H2SO4. 0.5 M 53.10±0.36 HCl 1.0 M 6.87±0.20 H2SO4 1.0 M 59.19±1.14 HNO3 0.5 M 3.79±1.30 H2SO4 2 M 67.30±0.23

41

Table 5. Parameters from fitting the empirical models to the experimental breakthrough curves of Cr(III). Run 2 (CE=500 (mg/L)) 9.15x10-4 59.87 0.99 14.56

Thomas

k th (mL/min mg) qth (mg/g) R2 RMSE

2.13 30.92 0.97 17.27

0.91 53.04 0.99 14.56

Yoon and Nelson

k yn (1/min) τh (min) R2 RMSE

0.49 24.19 0.97 17.27

0.45 19.44 0.99 14.56

IP T

k ba (L/mg min) 𝑞𝑏𝑎 (mg/g) R2 RMSE

Run 1 (CE=230 (mg/L)) 2x10-2 35.05 0.97 17.27

SC R

Parameter

A

CC E

PT

ED

M

A

N

U

Model Bohard-Adams

42

𝒎 (kg) 99.24

𝒕 𝒔𝒕 (h) 20.05

𝒕𝒃𝒓 (h) 14

𝒅 (m) 0.61

𝒉𝒓 (m) 1.06

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Table 6. Data of parameters used for the scale-up. CE 𝑸𝒓 𝝆𝒃 𝒖𝟎 𝑭𝑺𝑩 (mg/L) (cm3/s) (kg/m3) (cm/s) (%) 111.11 500 320 0.038 79

43

Permeate H2SO4 (kg) 285.30

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Table 7. Mass balances for one-cycle operation in Column 1. Column regeneration Chemicals recovery in nanofiltration Column Saturation (Efficiency = 85.8%) (Efficiency = 86%) Input amount Input amount Input Output of Cr(III) of H2SO4 Cr2(SO4)3 H2SO4 Retentate (kg) (kg) (kg) (kg) Cr2(SO4)3 (kg) 2.75 333.50 9.00 331.75 7.74

44

Table 8. Comparing the cost of adsorbent material. Material

US$/kg 7.27

Amberlite 252

2.43

Purolite S 108

11.42

Activated carbon Hongchang HC-PA05

1.50

Activated carbon Lvyuan

3.00

Acid modified sugarcane bagasse [47]

4.33

XPB

0.39

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Lewatit TP 208

45