Statistical optimization of activated carbon from thapsia transtagana stems and dyes removal efficiency using central composite design

The effect of activation temperature, impregnation ratio and activation time on iodine number IN, methylene blue index MB index and removal efficiencies of methyl violet MV, methyl orang

Journal Pre-proof

Statistical optimization of activated carbon from Thapsia transtagana stems and dyes

removal efficiency using central composite design

A Machrouhi, H Alilou, M Farnane, S El Hamidi, M Sadiq, M Abdennouri, H.

Tounsadi, N Barka

DOI: https://doi.org/10.1016/j.jsamd.2019.09.002

Reference: JSAMD 251

To appear in: Journal of Science: Advanced Materials and Devices

Received Date: 29 March 2019

Revised Date: 21 August 2019

Accepted Date: 7 September 2019

Please cite this article as: A Machrouhi, H Alilou, M Farnane, S El Hamidi, M Sadiq, M Abdennouri,

H Tounsadi, N Barka, Statistical optimization of activated carbon from Thapsia transtagana stems and dyes removal efficiency using central composite design, Journal of Science: Advanced Materials and Devices, https://doi.org/10.1016/j.jsamd.2019.09.002

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition

of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record This version will undergo additional copyediting, typesetting and review before it is published

in its final form, but we are providing this version to give early visibility of the article 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.

© 2019 Publishing services by Elsevier B.V on behalf of Vietnam National University, Hanoi.

Statistical optimization of activated carbon from Thapsia transtagana stems and dyes

removal efficiency using central composite design

A Machrouhi 1, H Alilou 1,2, M Farnane 1, S El Hamidi 1, M Sadiq 1, M Abdennouri 1, H

* Corresponding author: Tel.: +212 645 20 85 64; E-mail: hananetounsadi@gmail.com

** Corresponding author: Tel.: +212 661 66 66 22; fax: +212 523 49 03 54; E-mail:

barkanoureddine@yahoo.fr

Aicha Machrouhi ; E-mail: machrouhi.aicha90@gmail.com

Meryem Farnane; E-mail: farnane.meryem@gmail.com

Sanaa El Hamidi ; E-mail : sanaa.elhamidi@gmail.com

M’hamed Sadiq ; E-mail : sadiqmhamed@hotmail.com

Statistical optimization of activated carbon from Thapsia transtagana stems and dyes

removal efficiency using central composite design

A Machrouhi 1, H Alilou 1,2, M Farnane 1, S El Hamidi 1, M Sadiq 1, M Abdennouri 1, H

* Corresponding author: Tel.: +212 645 20 85 64; E-mail: hananetounsadi@gmail.com

** Corresponding author: Tel.: +212 661 66 66 22; fax: +212 523 49 03 54; E-mail:

barkanoureddine@yahoo.fr

Abstract

This study focused on the preparation of activated carbons from Thapsia transtagana

stems by boric acid activation and their evaluation for dyes removal Central composite design and response surface methodology were used to optimize the preparation conditions The effect of activation temperature, impregnation ratio and activation time on iodine number (IN), methylene blue index (MB index) and removal efficiencies of methyl violet (MV), methyl orange (MO) and indigo carmine (IC) were fully evaluated Activated carbon samples prepared in optimal conditions were characterized by FTIR, XRD, SEM-EDX, Boehm’s titration, and point of zero charge (pHPZC) The equilibrium data for dyes sorption onto optimum activated carbons were best fitted with Langmuir isotherm

Keywords: Thapsia transtagana stems; Dyes removal; Chemical activation; Central

composite design

1 Introduction

Nowadays, the extensive uses of textile dyes are considered the main sources of water pollution [1] It has been estimated that 10 % to 15 % of the dye used during the manufacturing of textile products are released into the environment worldwide annually [2] Moreover, many of these organic compounds can cause allergies, skin irritation or even cancer and human mutations [3] It is therefore essential to remove dyes from wastewater and water reuse to avoid contamination and destruction of natural resources

Currently, there are numerous methods employed to remove dye molecules from aqueous solutions including adsorption [4], precipitation [5], ion-exchange [6], coagulation [7], membrane filtration [8] and photocatalytic degradation [9] etc Among these processes, the adsorption is more applicable because it is an efficient, simplest and economic method for the removal of dyes from aqueous solutions [10-11] For that, various types of low-cost, easily available and highly effective adsorbents are reported such as activated carbon, zeolite, clay, polymer, and nanomaterials [12-16]

From economic point of view, the process of adsorption onto activated carbon is advantageous due to the plentiful accessibility of low cost raw material Also, activated carbon is basically referred as carbonaceous materials, with a high physicochemical stability, high porosity, high sorption capacity and with immense surface area

Recently, many studies have been carried out to investigate the use of inexpensive biomasses to produce low-cost activated carbons using agricultural solid wastes including

coffee ground [17], Carob shell [18], Diplotaxis harra [19], Glebionis coronaria L [20],

maize corncob [21], beetroot seeds [22], apricot stones [23], hazelnut shells [24] and loofah sponge [25]

Activated carbon can be produced in two-step process: carbonization and activation Carbonization is usually conducted via pyrolysis at the temperature of 400-850°C in the absence of oxygen [26] The activation process converts carbonized materials to activated

carbon via heating Carbon dioxide, air or steam is used as a physical activation method to develop the porosity of the carbonaceous materials The increase of surface area and the pore volume is achieved through the elimination of internal carbon mass and the removal of volatiles [27] However, in chemical activation, the carbonization temperature is done only between 400 and 600 °C This method highlights an impregnation of the precursor or raw material with dehydrating agents such as alkali metal hydroxide or acid This method produces an activated carbon with higher yield and well developed microporosities

The objective of this research was to investigate the feasibility of activated carbon

produced from Thapsia transtagana stems biomass, by H3BO3 activation and their ability for cationic and anionic dyes removal from aqueous solution Central composite design (CCD) combined with response surface methodology (RSM) was used to optimize the process The factors chosen are impregnation ratio, activation temperature and activation time Five responses including iodine number (IN), methylene blue index (MB index) and removal efficiencies for methyl violet (MV), methyl orange (MO) and indigo carmine (IC) are investigated

2 Material and methods

2.1 Material

All the chemicals/reagents used in this study were of analytical grade H3BO3 (100%), HCl (37%), I2 (99.8-100.5%), Na2S2O35H2O, Na2CO3, NaHCO3 (99.5-100.5%), commercial activated carbon (powder form) (100%), methyl violet, methyl orange and indigo carmine (100%) were purchased from Sigma-Aldrich (Germany) (100%) Methylene blue was purchased from Panreac (Spain) (100%) HNO3 (65%) was provided from Sharlau (Spain) NaOH (≥ 99%) from Merck (Germany), potassium iodide (KI) (100%) was obtained from Pharmac (Morocco)

2.2 Preparation of activated carbons

The Thapsia transtagana plant was collected from the region of Oued zem, Morocco

Steams were cut into small pieces and were powdered to a particles of size < 125 µm using a domestic mixer 15 g of the biomass were impregnated with H3BO3 as the activating agent at desired mass ratio Later, the sample was loaded in a stainless steel vertical tubular reactor placed into a furnace under purified nitrogen atmosphere The obtained activated carbons were washed with distilled water and dried at 105 °C for 24 h The powder was sieved in particles of size lower than 125 µm using a normalized sieve and kept in a hermetic bottle for

a further use

The impregnation ratio of the activating agent with the precursor was computed using Eq (1): Impregnation ratio = (dried weight of H3BO3 / precursor of TTS) (1)

2.3 Design of experiments using central composite design

Central composite design (CCD) was used to study the individual and synergetic effect

of the three factors towards defined responses This method can reduce the number of experimental trials required to evaluate main effect of each parameter and their interactions [28] It is characterized by three operations namely: 2n factorial runs, 2n axial runs and six center runs [29] For this case, it’s translated into eight factorial points, six axial points and six replicates at the center which gives a total of 20 experiments as calculated from Eq (2):

Total number of experiments (N) = 2n +2n+nc (2)

where n is the number of factors, nc is the number of center points (six replicates)

The independent variables were coded as +1 and -1 which represent the eight factorial points at their low and high levels respectively The six axial points were located at (±α, 0, 0), (0, ±α, 0), (0, 0, ±α), and the six replicates were at the center (0, 0, 0) were run to examine the experimental error and the reproducibility of the data Where α is the distance of axial point from center which makes the design rotatable and had value fixed at 1.682 This value of

rotatability α, which depends on the number of parameters in the experiment, was obtained from the following equation [30]:

α = Np1/4 (3)

In this study, the independent variables studied were activation temperature (A), impregnation ratio (B) and activation time (C) These three variables with their respective ranges were selected based on the literature and preliminary studies as given in Table 1 Table 1: Process factors and their levels

Variables Code Unit Coded variable levels

Activation temperature A °C 366 400 450 500 534

The responses were determined using the optimal quadratic model predictor Eq (4)

given as:

where Y is the predicted response, bo is the constant coefficients, bii the quadratic coefficients,

bij the interaction coefficients and xi, xj are the coded values of the activated carbon preparation variables considered

The quality of the fit of polynomial model was expressed by the correlation coefficient (R2) The significance and adequacy of the used model was further explained using F-value (Fisher variation ratio), probability value (Prob > F), and adequate precision (AP) [31]

2.4 Iodine number (IN)

Iodine number is a measure of micropore content (0–2 nm) by adsorption of iodine from solution The iodine number is defined as the milligrams of iodine adsorbed by 1.0 g of carbon when the iodine concentration of the filtrate is 0.02 N It was determined according to the ASTM D4607-94 method [32]

2.5 Methylene blue index (MB index)

The methylene blue index is a measure of mesoporosity (2–50 nm) present in activated carbon Sorption equilibrium was established for different methylene blue initial concentration between 20 and 500 mg/L for 12 h at room temperature Residual concentrations were determined by spectrophotometric method at the wavelength of maximum absorbance of 665 nm [33]

2.6 Dyes removal

Stock solutions of methyl orange, methyl violet and indigo carmine at 500 mg/L were prepared by dissolving 0.5 g of each dye in 1 L of distilled water Sorption experiments were investigated in a series of beakers containing 50 mL of dyes solutions at 500 mg/L and 50 mg

of each activated carbon The mixtures were stirred for 2 h without any pH adjustment After each sorption experiment, samples were centrifuged at 3400 rpm for 10 min and the dyes concentration was determined using UV–vis spectrophotometer

The adsorption capacities of the dyes at equilibrium were defined as the amount of adsorbate per gram of adsorbent (in mg/g) and were calculated using following equation:

q = (5)

where q is the adsorbed quantity (mg/g), Co is the initial dye concentration (mg/L), C is the residual dye concentration (mg/L), and R is the mass of activated carbon per liter of aqueous solution (g/L)

2.7 Surface and chemical characterization

Textural properties of optimized activated carbon were observed by scanning electron microscopy (SEM) using TESCAN VEGA3-EDAX equipped with an Energy-Dispersive X-Ray detector (EDX) The functional groups present on the surface of the starting material and

AC was determined by Fourier Transform Infrared (FTIR) spectroscope (FTIR-2000, Perkin Elmer) in a range of 4000-400 cm-1 Crystallographic characterization was examined by XRD

measurements in 2θ range from 10 to 70° using a Bruker-axs D2-phaser advance diffractometer operating at 30 kV and 10 mA with CuKα The acidic and basic functional groups on the surface of ACs were determined quantitatively by the Boehm’s titration method [34] The pH of point of zero charge (pHpzc) was determined according to the method described by Noh and Schwarz [35]

3 Results and discussion

3.1 Experimental results

The experimental results obtained at designed experimental conditions according to the central composite design are presented in Table S1 From this table, it could be seen that activated carbon sample activated at 500 °C for 145 min with an impregnation ratio of 2 g/g gives the optimum of MB index (188.75 mg/g), MO adsorption (116.84 mg/g) and MV adsorption (140.76 mg/g The greater iodine number of 794.58 mg/g is obtained for the activated carbon prepared at 450 °C for 130 min with an impregnation ratio of 2.34 g/g Under these same conditions, the optimum for IC adsorption (44.87 mg/g) is also acquired

On the other hand, the regression analysis was performed to fit the response functions with the experimental data Table 2 shows the values of regression coefficients obtained According to this table, the three studied factors present a positive effect on the five responses The table also indicates that the targeted responses are more influenced by activation temperature and impregnation ratio than by activation time

Table 2: Values of model coefficients of the five responses

b23 5.24 2.93 -0.16 -2.27 0.30

3.2 Analysis of variance (ANOVA)

The analysis of variance (ANOVA) was used to determine the significance of the curvature in the responses at a confidence level of 95% After discarding the insignificant terms, the ANOVA data of the coded quadratic models, for the five responses are presented in supplements (Tables S2-S6) The effect of a factor is defined as the change in response produced by a change in the level of the factor This is frequently called a main effect because

it refers to the primary factors of interest in the experiment The ANOVA results showed that the equations adequately represent the actual relationship between each response and the significant variables The F value implies that the models are significant and the values of

“Prob > F” less than 0.05 indicate that models terms are significant Especially larger F-value with the associated P value (smaller than 0.05, confidence interval) means that the experimental systems can be modeled effectively with less error Therefore, interaction effects

as adequate model terms can be used for modeling the experimental system

3.2.1 Iodine number

According to the ANOVA analysis for the iodine number, the significant terms are the activation temperature (A), impregnation ratio (B), activation time (C), the interaction between activation temperature and impregnation ratio (AB), the interaction between impregnation ratio and activation time (BC), the quadratic term of activation temperature (A2) and the quadratic term of activation time (C2) Eq (6)

Y1 = 703.26 + 44.10 A + 56.46 B - 0.74 C - 1.75 AB + 1.75 BC - 28.35 A2 -20.94 C2 (6) The activation temperature, the impregnation ratio and the interaction between impregnation ratio and activation time showed a positive effect on the iodine number

Although, the activation time, the interaction between activation temperature and impregnation ratio, the quadratic term of activation temperature and the quadratic term of activation time showed a negative effect on the iodine number Besides, the impregnation ratio has the largest significant effect on the iodine number due to the high F-value (99.05) followed by the activation temperature, the quadratic term of activation temperature and the quadratic term of activation with an F-value of 60.43, 26.61, and 14.52 respectively (Table S2) Hence, it could be seen that the number of micropores are higher with the impregnation ratio of 2.34 g/g in the studied domain In fact, at high level of the significant model terms, the activation reaction may take place rapidly producing a development of the porosity of the obtained activated carbons and an increase in the microporosity

3.2.2 Methylene blue index

The most significant effects for the methylene blue index are activation temperature (A), impregnation ratio (B), activation time (C), interaction between activation temperature and impregnation ratio (AB) and the quadratic term of impregnation ratio (B2) Eq (7)

Y2 = 133.98 + 13.17 A + 29.11 B + 4.97 C + 10.04 AB - 9.90 B2 (7) The activation temperature, impregnation ratio, activation time and interaction between activation temperature and impregnation ratio showed a positive effect on the methylene blue index response Although, the quadratic term of impregnation ratio presented a negative effect

on the development of mesoporores According to Table S3, the impregnation ratio has the most significant effect on methylene blue index due to the higher F-value (43.55) After that, the activation temperature, the quadratic term of impregnation ratio and the interaction between activation temperature and impregnation ratio with F-values of 8.92, 5.41 and 3.03 respectively

3.2.3 Methyl orange and methyl violet removal

Based on the ANOVA data for methyl orange and methyl violet removal responses, the most significant factors are activation temperature (A), impregnation ratio (B) activation time

(C), interaction between activation temperature and impregnation ratio (AB), interaction between impregnation ratio and activation time (BC) and quadratic term of impregnation ratio (B2) Eq (8) and Eq (9)

Y3 = 93.97 + 9.69 A + 16.57 B + 2.18 C + 0.63 AB - 0.16 BC – 5.38 B2 (8)

Y4 = 113.21 + 9.89 A + 14.66 B + 4.46 C + 4.71 AB - 2.27 BC – 5.09 B2 (9)

The activation temperature, impregnation ratio, activation time and interaction between activation temperature and impregnation ratio showed a positive effect on the methyl orange and methyl violet removal response Although, the interaction between impregnation ratio and activation time and the quadratic term of impregnation ratio presented a negative effect From Table S4 and Table S5, it could be seen that the impregnation ratio have a greatest effect on the methyl orange and methyl violet removal with an F-value of 53.15 and 26.10 respectively, followed by activation temperature and quadratic term of impregnation ratio for MO and MV removal

3.2.4 Indigo carmine removal

The significant model terms for indigo carmine removal are the activation temperature (A), impregnation ratio (B), activation time (C), the interaction between activation temperature and impregnation ratio (AB), the interaction between activation temperature and activation time (AC) and the quadratic term of activation temperature (A2) Eq (10)

Y5 = 27.67 + 3.76 A + 10.63 B + 1.83 C + 2.95 AB - 0.39 AC – 3.22 A2 (10)

The activation temperature, impregnation ratio, activation time and interaction between activation temperature and impregnation ratio showed a positive effect on the indigo carmine removal response However, the interaction between activation temperature and activation time and the quadratic term of activation temperature presented a negative effect on indigo carmine removal According to Table S6, it could be seen that the impregnation ratio has the most effect on indigo carmine removal based on the highest F-value of 60.03 Whereas activation temperature, quadratic term of activation temperature and interaction between

activation temperature and impregnation ratio with an F-value of 7.52, 5.94, and 2.72 respectively

3.3.Response surface analysis

The mathematical models for the iodine number, MB index and dyes removal were used

to build response surfaces as well as to determine the optimal conditions of the process Fig.1 present the 3D response surfaces plots for the significant interactions

For the iodine number, the most significant interactions were the impregnation ratio/activation temperature and the activation time/impregnation ratio The Fig.1(a) indicates that the iodine number increased with the increase of activation temperature and impregnation ratio Fig.1(b) shows that the iodine number increased with increasing of the impregnation ratio and the decreasing of activation time when the activation temperature is fixed at 500 °C For MB index, the most significant interaction was the impregnation ratio/activation temperature From Fig.1(c), it can be observed that the MB index increased with the increase

of the activation temperature and the impregnation ratio The maximal MB index response was obtained at an activation time of 145 min

In the removal of MV and MO dyes, the same significant interactions are accursed, including the impregnation ratio/activation temperature and the activation time/impregnation ratio From Fig.1(d)-1(f), it can be observed that the MV and MO removal increased with increasing of the activation temperature and impregnation ratio The maximal MV and MO removal responses were obtained at an activation time of 145 min Fig.1 (e)-1(g), shows that the MO and MV removal increased with increasing impregnation ratio and decreasing activation time when the activation temperature is fixed at 500 °C

For the removal of IC, the most significant interactions were the impregnation ratio/activation temperature and the activation time/activation temperature From the 3D response surface plot as shown in Fig.1(h), it was observed that the indigo carmine removal increases with increasing of the impregnation ratio and activation temperature, when the

activation time is fixed at 130 min Fig.1(i) shows that when the activation time decreases and the activation temperature increases, the IC removal response increased

In general, the impregnation of the precursor allows the development of the internal structure of the activated carbon by the creation of new pores and the enlargement of existing pores In this context, several parameters including the activation time, the activation temperature and the impregnation ratio play an important role in the development of the porosity of activated carbons and consequently the evolution of the adsorption performance

In fact, during activation, the boric acid catalyzes the dehydration and promoting the formation of aromatic structures during pyrolysis [36] In addition, the formation of an impenetrable glassy coating on the solid surface from boric acid decomposition products inhibits the release of volatile substances, which also promotes the formation of carbon [37-38] Then, this vitreous coating prevents the diffusion of oxygen and prevents the propagation

of exothermic combustion reactions [39]

Fig.1 Surface response plot for the iodine number (a-b), MB index (c), MV removal (d-e),

MO removal (f-g) and IC removal (h-i)

3.4.Diagnostic model

Table S7 summarizes the information of the proposed models of statistic actual and predicts values for testing the significant effects of the regression coefficients Predicted values obtained were compared with experimental values These values for the models are close, which indicates a correspondence between the mathematical model and the experimental data The correlations between the theoretical and experimental responses, calculated by the model, are satisfactory Therefore, the R2 are in reasonable agreement with the Radj2 In addition, the model F-value of the iodine number, methylene blue index, methyl

orange, methyl violet and indigo carmine removal including 28.21, 12.44, 13.06, 6.60 and 13.00 respectively These values implicate that models are significant

3.5.Normal probability plot of residuals

The normal probability plot of the residuals was presented in Fig.2 The normality of the data can be checked by plotting a normal probability plot of the residuals If the data points on the plot fall fairly close to the straight line, the data are normally distributed [40] It’s appears that in the iodine number, methylene blue index and dyes responses, the data points were fairly close to the straight line and it indicates that the experiments come from a normally distributed population