Removal of Chromium (VI) from Water Using Orange peel as the Biosorbent: Experimental, Modeling, and Kinetic Studies on Adsorption Isotherms and Chemical Structure

Abstract

The present work aims to assess the effectiveness and efficiency of orange peels as a low-cost biosorbent for removing Cr(VI) from an aqueous solution by the biosorbent process. The orange peels as adsorbent was characterized using different methods, such as FTIR, pH_pzc, equilibrium pH, TGA, XRD, SEM, and (BET). The tests were conducted in the batch mode, and the effects of different parameters, such as the pH, dosage of the bioadsorbent, influent Cr(VI), and time, on the biosorption of Cr(VI) were investigated. The adsorption kinetics proved that a contact time of 90 min resulted in the highest (approximately 97.8%) Cr(VI) removal, with an adsorption capacity of 4.96 mg/g. Moreover, the increase in the biosorbent dosage (from 1 to 10 g/L) resulted in the enhancement in the Cr(VI) removal effectiveness. Moreover, the pH of the solution also affected significantly the effectiveness of the removal. The tests were conducted under acidic pH solution conditions, and the prediction of the pH value at a zero charge (pH _pzc) was confirmed experimentally. Furthermore, the results from the batch-mode assays were successfully tested by an experimental design (full factorial design). The biosorption of Cr(VI) on orange peels occurred mostly according to the pseudo-second-order kinetic model and the uptake of Cr(VI) was satisfactorily described by the Langmuir model.

1. Introduction

Chromium (Cr) is one of the most widely used chemicals in the industry thanks to its valuable effects in processing skin for tannery [1], textiles [2], wood and manufacturing dye [3] as well as in paint and paper [4] and in oil refining [5]. Cr(III) and Cr(VI) are two stable oxidation states of Cr that persist in the environment [6]. Cr (III) is essential for human nutrition, especially in glucose metabolism, while Cr(VI) is toxic for humans, animals, and bacteria. Cr(VI) compounds are known to be carcinogenic [7,8]. According to the World Health Organization (WHO) guidelines, the concentration thresholds of Cr(VI) are 0.1 mg/L and 0.05 mg/L in wastewater discharged into inland surface waters and drinking water, respectively [9].

Various methods are used for removing Cr(VI) substances from wastewater, industrial waters, surface water, and drinking water were reported in the literature [10,11,12,13]; for instance, conventional treatment methods, such as chemical precipitation, re commonly used for the removal of chromium from industrial effluents, such as tannery wastewater [14,15,16,17,18,19,20,21,22]. Using this method, other various substances can be separated in addition to chromium. Ion exchange [23,24,25], electro-coagulation [26], membrane separation [27], coupled coagulation/precipitation [28], and integrated reduced/oxidation [29] have been used. However, they are costly to use to safely remove residual sludges and the chemicals necessary for removing Cr(VI) are expensive. They also suffer from unsatisfactory Cr(VI) removal, whereas adsorption is considered to be a promising method, as it requires low operating costs and, when it is combined with a suitable desorption phase, the concern for the residual sludge disposal is negligible [30].

A number of low-cost adsorbents have been already tested successfully for the removal of toxic pollutants from wastewater [31,32,33,34,35,36]. In recent years, several natural adsorbents, like wool, olive cake, sawdust, pine needles, almond shells, cactus leaves, soot, hazelnut shell, coconut shell, banana peel, orange peel, seaweed, dead fungal biomass, cyano-bacteria, and green algae, have been used to remove Cr from water [37,38,39,40]. However, many of these adsorbents present a low Cr adsorption yield and slow process kinetics. Thus, it is of interest to find other low-cost materials with enhanced performance. In such a context, the aim of this study is to test the efficiency of orange peels in removing Cr(IV) from water. We validate the results using the Minitab software 21.1.0 (Minitab, LLC, State College, PA, USA) and determine the optimal conditions corresponding to the best chromium removal efficiency and effectiveness. Compared to other publications, the novelties of this study include: (1) Creating a mathematical model to promote the adsorption process of water with similar chromium properties, potentially omitting the laboratory optimization step, i.e., only mathematical models are used; (2) Kinetic studies and adsorption isotherms; (3) Chemical structure considerations of orange peel and their reaction with chromium.

2. Materials and Methods

2.1. Biosorbent Preparation

With the aim of developing an inexpensive biomaterial for the removal of Cr(VI) from wastewater, orange (Citrus sinensis) peels were chosen as a natural source for the preparation of the biosorbent. It was obtained according to the following sequence of operation:

• The orange peels were washed with tap water and then with distilled water to remove residual impurities and soluble parts. The washing process was repeated several times until clear washing water was obtained.
• The orange peels were exposed to the sun for several days until completely dry.
• Once dried, the peels were crushed using an electric household grinder.
• The ground peels were sifted using a sieve with a mesh size of 0.315 mm.
• After sieving, the biosorbent was stored for the experimental period in dry bottles.
• Before each use, the orange peels in powder form were dried in an oven at a temperature of 105 °C to constant weight.

2.2. Biosorbent Characterization

For the physical and chemical characterization of the bioadsorbent (orange peels in powder size), several techniques were used. The Fourier-transfer infrared (FTIR) spectrum was used to examine the nature of the chemical bonds with a JASCO FT/IR -4600 type device (JASCO, Tokyo, Japan). A thermogravimetric analysis (TGA) was performed to study the thermal stability with the TGAQ50 module from the Thermal Analyzes device. The crystalline structure of the orange peels was evaluated by X-ray diffractometer using Cu with wavelength K-Alpha (1.54) at 40 mA and 40 kV with a scan analysis. The Scanning electron microscopic (SEM) analysis of the bioadsorbent was recorded with a scanning electron microscope Quanta 650 (FEI Company, Hillsboro, OR, USA). The Brunauer–Emmett–Teller (BET) method was employed to measure the surface area of the orange peel powder. Moreover, the identification of the points of zero net charge, iodine number, and contact-free pH was determined.

2.2.1. pH Value at the Point of Zero Charge

The pH at the point of zero charge (pH_pzc) of materials is a good indicator of the chemical and electronic properties of functional groups on their surface. The pH_pzc of the biosorbent was calculated as follows: Solutions of 50 mL of NaCl (0.01 M) were placed in flasks. The pH in each flask was adjusted by adding drops of solutions of NaOH or HCl (0.1 M) to achieve initial pH (pH_i) values ranging between 2 and 12. In each flask, 0.15 g of the biosorbent (Orange peel) was added. The flasks were kept closed at room temperature for 48 h, and then the final pH (pH_f) was determined. On the graph of pH_f = f(pHi), the intersection of the curve with the quadrant bisector represents the iso-electric point (pH_pzc) [41,42].

2.2.2. Contact-Free pH

The contact-free pH was determined by introducing 1 g of the biosorbent in 100 mL of distilled water, and the sample was stirred at 300 rpm for 24 h. The sample was centrifuged and, finally, the pH of the filtrate was measured using a pH meter (Jenway model 3540, Camlab, Cambridge, UK).

2.2.3. Iodine Number

The iodine number (Id) was calculated according to the following procedure involving two tests. The first test was the blank test conducted with 10 mL of a 0.1 N iodine solution placed in a beaker and dosed with a 0.1 N solution of sodium thiosulfate (Na₂SO₃) until the color disappeared with the adding of a few drops of a starch solution (indicator). The second test consisted of adding 0.2 g of the biosorbent into a beaker that contained 15 mL of a 0.1 N iodine solution and mixing it for 4 min with a stirring equipment. After that time, a centrifugation process was carried out and 10 mL of the filtrate (which contained iodine) was dosed with a 0.1 N solution of sodium thiosulfate (Na₂SO₃) until the color disappeared with the adding of a few drops of a starch solution (indicator) [43].

Id is calculated according to the following Equation (1).

$$Id=\frac{\left(\left(V_B-V_S\right)\times N\times 126.9\times \raisebox{1ex}{$15$}\!\left/ \!\raisebox{-1ex}{$10$}\right.\right)}{m}$$

(1)

where

• (V_B − V_S): difference between the titration volumes calculated using the blank test and the biosorbent with V_B = 12.9 mL and V_S = 9 mL of 0.1 N solution of thiosulfate.
• N: normality of the sodium thiosulfate solution (0.1 N).
• 126.9: the atomic mass of iodine (g/mol).
• m: mass of the biosorbent (0.2 g).

2.3. Adsorption Tests

2.3.1. Chromium Solutions

A Cr(VI) stock solution of 1 g/L was prepared by dissolving 2.828 g of potassium dichromate (K₂Cr₂O₇) salt in 1000 mL of distilled water. Seven solutions at different Cr(VI) concentrations (i.e., 10, 20, 30, 50, 60, 80, and 100 mg/L) were obtained from the stock solution by the appropriate dilution rates. The pH was adjusted by adding 0.1 N hydrochloric acid (HCl) and 0.1 N caustic soda (NaOH) solutions, and measured by using a pH meter (type 3505, JENWAY).

2.3.2. Operating Procedure

The adsorption tests were conducted according to the following procedure:

(1)

Preparation of a suspension (bio-sorbent/contaminated solution) for a ratio r (S/L) = 1.5 and 10 g/L as the biosorbent dosage;

(2)

The initial Cr(VI) was set at the desired concentration, as reported in Section 2.3.1;

(3)

The biosorbent suspension was stirred at 300 rpm for 24 h at room temperature (22 ± 1 °C);

(4)

Volumes of 5 mL were collected with a syringe from a suspension at different reaction times during the first period (350 min) of experimentation in order to determine the equilibrium time;

(5)

The pH was adjusted by adding a HCl or NaOH solution in the range of 2–10;

(6)

The samples collected from the suspension were filtrated with a Millipore filter (0.45 µm);

(7)

The filtrated fraction was analyzed by a UV–visible spectrometer SHIMADZU UV-160A model.

The biosorbent efficiency expressed as the Cr_removal (%) was assessed by calculating the amount of Cr(VI) removed according to the following Equation (2):

$${Cr }_{removal}\left(\%\right)=\frac{\left(C_0-C_e\right)}{C_0}\times 100$$

(2)

whereas the quantity of the adsorbed pollutant (q) expressed in mg of pollutant per g of adsorbent (mg/g) was calculated according to the following Equation (3):

$$q=\frac{(C_0-C_t)V}{m}$$

(3)

where

• C₀ = the initial concentration of Cr(VI) in mg/L.
• C_e and C_t = the residual concentrations of Cr(VI) at the equilibrium and “t” times, respectively, in mg/L.
• m = mass of the biosorbent (g).
• V = volume of the Cr solution (mL).

2.4. Effect of the Parameters

The effects of the following parameters were investigated: solid–liquid ratio (r), operating pH, and the initial concentration of Cr(VI).

The examination of the influence of the solid–liquid ratio (r) on the adsorption capacity of Cr(VI) on the orange peels was conducted by varying the biosorbent dosage among 0.1, 0.5, and 1 g in 100 mL working solution.

To optimize the pH, a study of the Cr(VI) adsorption as a function of the pH varying withing the range from 2 to 10 was carried out with an initial Cr(VI) concentration of 50 mg/L and a dosage of the adsorbent of 10 g/L.

Finally, in order to study the effect of the initial concentration of Cr(VI), the following values were considered: 10, 20, 30, 50, 60, 80, and 100 mg/L.

2.5. Study of the Effects and Interactions between the Parameters Affecting the Cr(IV) Adsorption Process

To understand in depth the effect of the operating parameters and their interactions on the Cr(VI) removal efficiency, a full factorial design study was conducted and is described in the following subsections.

Full Factorial Design

In order to obtain the optimal conditions for adsorption, a full n^k factorial design was performed, where n = the number of levels and k = the number of factors being evaluated. For the present study, n = 2 and k = 4, as detailed in

Table 1. Minimum and maximum levels of the studied parameters.

Variables	Symbols	Domain and Levels
		−1	+1
Mass (mg)	X1	500	2000
pH	X2	2	10
Contact time (min)	X3	5	90
Initial concentration of Cr(VI) (mg/L)	X4	10	200

. Thus, the total number of required experiments was 2⁴. If Y is the response variable, the four parameters’ regression equation and their interactions are evaluated through the following Equation (4):

Y = b₀ + b₁X₁ + b₂X₂ + b₃X₃ + b₄X₄ + b₁₂X₁X₂ +b₁₃X₁X₃ + b₁₄X₁X₄ + b₂₃X₂X₃ + b₂₄X₂X₄ + b₃₄X₃X₄

(4)

where b₀, b₁, b₂, b₃, and b₄ are the linear coefficients, and b₁₂, b₁₃, b₁₄, b₂₃, b₂₄, and b₃₄ are the second-order interaction terms. X₁, X₂, X₃, and X₄ are the coded dimension factors of the following studied parameters: the biosorbent dosage (m), pH, the initial concentration (C) of Cr(VI), and contact time (t), respectively. Table 1 reports the low and high levels for the studied parameters. The minimum and maximum values of the variables in Table 1 were selected based on several conditions for the preliminary experimentations, for example, the contact time was chosen when the equilibrium was reached. For the initial concentration, we chose to ascribe to the synthetic solutions a character like that of wastewater effluents, which nearly reach these concentrations. For the pH, it was better to use a large interval in order to optimize the available values.

2.6. Kinetic Study

The kinetic studies of the adsorption process on the biosorbent were carried out at 22 °C (room temperature) and different initial Cr(VI) concentrations (from 10 to 100 mg/L). The efficiency of the adsorption process was assessed at regular time intervals.

The mathematical expressions of the pseudo-first- [44] and pseudo-second-order [45] models as well as the intra-particle kinetic model are reported in the following Equations (5)–(7):

$$\ln \left(q_e-q_t\right)=\ln q_e-\frac{K_1}{2.303}t$$

(5)

$$\frac{t}{q}=\frac{1}{K_2.{q_e}^2}+\frac{1}{q_e}t$$

(6)

$$q_t=K_{in}. t^{0.5}$$

(7)

where

• q_e (mg/g): adsorbed amount at equilibrium.
• q_t, (mg/g): adsorbed amount at time t (min).
• K₁ (min⁻¹): equilibrium constant for the adsorption rate of the pseudo-first-order equation.
• K₂ (g/mg·min): equilibrium constant for the adsorption rate of the pseudo-second-order equation.
• K_in (mg/g·min^1/2): equilibrium constant for the adsorption rate of the intra-particle model equation.

2.6.1. Adsorption Isotherms

Different adsorption isotherm models are available in the literature to process the adsorption process experimental data [46,47]. Among them, in this work, the Langmuir, Freundlich, and Elovich isotherms were used. In what follows, we report the linear relationships of the 3 models: Equations (8), (9), and (10), respectively.

$$\frac{1}{{\mathrm{q}}_{\mathrm{e}}}=\frac{1}{{\mathrm{q}}_{\max }}+\frac{1}{{\mathrm{K}}_{\mathrm{L}}\times {\mathrm{C}}_{\mathrm{e}}{\times \mathrm{q}}_{\max }}$$

(8)

$$\ln \left(\mathrm{q}\right)=\ln \left({\mathrm{K}}_{\mathrm{F}}\right)-\frac{1}{\mathrm{n}}\times \ln \left({\mathrm{C}}_{\mathrm{e}}\right)$$

(9)

$$\ln \left(\frac{{\mathrm{q}}_{\mathrm{e}}}{{\mathrm{C}}_{\mathrm{e}}}\right)=\ln \left({\mathrm{K}}_{\mathrm{E}}{\times \mathrm{q}}_{\max }\right) -\frac{{\mathrm{q}}_{\mathrm{e}}}{{\mathrm{q}}_{\max }}$$

(10)

where

• C_e (mg/L): equilibrium concentration.
• q_e (mg/g): adsorbed amount at equilibrium.
• q_max (mg/g): adsorbed maximum amount.
• K_L, K_F, and K_E are, respectively, the Langmuir constant (L/mg), Freundlich constant (mg⁽¹⁻ⁿ⁾·Lⁿ/g), and Elovich constant (L/mg).
• n: constant relating to energy.

3. Results

3.1. Biosorbent Characterization

3.1.1. FTIR Analysis

FTIR measurements were conducted to understand the possible functional groups existing for interactions between metallic ions and the orange peel adsorbent. Figure 1 shows that the spectrum is characterized by the presence of a characteristic peak at 3311.28 cm⁻¹ attributed to the elongation of the N–H [48,49] or most likely representing O–H group stretching vibrations (of carboxyl, phenols, or alcohols) [50,51]. This is quite logical if one considers that the main constituents of orange peels are cellulose, lignin, and hemicelluloses, all constituents that are rich in different kinds of –OH groups, as well as vicinal phenolic –OHs and –OCH₃ groups belonging to flavonoids, such as hesperidin glucoside, known to be abundant in orange peels. The band at 2926.58 cm⁻¹ generally characterizes aliphatic C-H stretching vibrations [52,53,54]. The band appearing in the orange peel powder spectrum at 1611 cm⁻¹ is attributed to the C=O groups of carboxylic acids, mainly both the acetate groups (COO⁻) of hemicelluloses and of the esters of citric acid, present in all citrus peels, as well as the carbonyl groups of ketones, like the one on the C₄ site of hesperidin, as well as traces of aldehydes or lactones [55]. The presence of a peak at 1358 cm⁻¹ attributed to C–O stretching supports the existence of numerous and varied carboxyl/alcohol/ether/ester functional groups in the absorbent. The probability of the presence of aliphatic fluorinated compounds (C-F), which are characterized by a peak at 1011.5 cm⁻¹ in the FTIR spectrum of the orange peel [56,57], exists but is unlikely, as more likely, it is a confirmation of the existence of aromatic nuclei in the mixture of the material constituents.

3.1.2. pH Value at the Point of Zero Charge

Figure 2 reports the graphical calculation of pH_pzc; at pH = 2.63, the surface charge is zero. For pH values higher than pH_pzc, the biosorbent surface charge is negative [58,59].

Therefore, on the basis of the operating pH, three different conditions can occur:

(1)

At pH < pH_pzc, the orange peel surface is positively charged;

(2)

At pH = pH_pzc, the orange peel surface is not charged (neutral);

(3)

At pH > pH_pzc, the orange peel surface is negatively charged.

3.1.3. Contact-Free pH

The value of the contact-free pH obtained was 4.3 (Figure 3). Therefore, the biosorbent is slightly acid.

3.1.4. Iodine Number

The resulting value of Id was 133.11 mg/g. This value proves that the biosorbent material is characterized by meso-pores and shows a moderately high iodine retention capacity.

3.1.5. Thermogravimetric Analysis

A thermogravimetry analysis (TGA) was carried out to evaluate the biosorbent composition in terms of water, cellulose, hemicellulose, and lignin. From Figure 4, it can be noticed that three steps of weight loss occurred: The first step between 25 °C and 180 °C is due to water evaporation (moisture loss). The second step in the range of 180–380 °C is a consequence of hemicellulose and cellulose degradation. The last step, when temperature is over 380 °C, is attributed to the decomposition of residual lignin. Generally, lignin accounts for 15% of the total dry weight [60,61].

3.1.6. XRD Analysis

Figure 5 shows the XRD spectra of the orange peel powder in its natural form. The graph shows that the constituent mixture is amorphous. This can be explained by the rupture of multiple C-C bonds (the aromatic rings) and the formations of groups and functions on the surface of the biomaterial (powder or orange peels). The spectrum shows two peaks, the first at 2θ = 21° and a large one at 2θ = [30°–45°]; this second one reveals that the adsorbent has a largely amorphous structure. A notable property of well-defined adsorbents is the absence of sharp peaks, which indicates that it has a largely amorphous structure [62,63,64,65].

3.1.7. SEM Analysis

In Figure 6, the surface morphology of the raw orange peels in powder form with the different magnifications of (a: ×1500), (b: ×800), and (c: ×400) confirms the heterogeneous nature of this adsorbent with large pore size. This structure facilitates the process of adsorption. It indicates that the porous structure of the material ensures a high surface area, facilitating adsorption. However, the chemical complexation of Cr(VI) by carbohydrate polymers and oligomers is well documented [66] as well as the ortho-diphenol metallic complexes by flavonoids (such as hesperidin) [67].

3.1.8. BET Surface Area Analysis

A BET analysis was used to measure the surface area, the pore volume, and the pore diameter of the orange peels in powder form. This analysis showed that the pore diameter (pd), the pore volume (Vp), and the specific surface area (SBET) were 152.238 A°, 32.270 m³/g, and 1511.278 m²/g, respectively. The obtained results confirm that the orange peel powder has a large surface area, which implies a high number of vacant sites that vaporize the adsorption of the pollutant.

3.2. Equilibrium Study

As shown in Figure 7, the chemical equilibrium for the adsorption process is reached in almost 90 min, with a satisfactory removal efficiency of approximately 99%. Such a result was obtained with pH = 5.1 and a biosorbent particle size (d) ≤ 0.315 mm. It is worthy of note that the removal efficiency value of 99% was obtained in the preliminary study conducted to assess the time needed to reach the chemical equilibrium condition. The results show that 90 min is a sufficient time to reach the chemical equilibrium. This time was, therefore, used to conduct further tests designed to study the effects of the remaining operating parameters, such as the dosage of the biosorbent, contact time, pH, and the initial concentration of Cr(VI).

3.3. Adsorption Process Performance

Adsorption is affected by several parameters, such as the biosorbent dosage, contact time, pH, and initial Cr(VI) concentration. The effects of each previously mentioned parameter are discussed in the following dedicated subsections.

3.3.1. Effect of the Biosorbent Dosage

According to Figure 8, the adsorption capacity (q) defined by Equation (3) is inversely proportional to the solid–liquid ratio (r), whereas the efficiency of the removal process increases by increasing the dosage of the biosorbent, thus increasing the working adsorbent surface and consequently the number of active sites. The obtained result is confirmed by other researchers [68,69,70].

3.3.2. Effect of the Contact Time

Figure 9 shows the trend of Cr(VI) adsorption. In this figure, two different phases of the process can be clearly noticed: a first stage lasting 15 min, where the process occurs very fast, and a second stage with a slower adsorption and even no adsorption. This trend is actually the result of a progressive saturation of the vacant sites on the surface of the biosorbent.

3.3.3. Effect of the pH

Figure 10 clearly shows that the Cr(VI) removal capacity at the chemical equilibrium condition decreases by raising the initial pH from 2 to 10. Comparing the curves in Figure 10, it can be stated that, for a pH ranging between 2 and 4, the adsorption capacity is at its maximum (4.96 and 4.91 mg/g, respectively), whereas, at the opposite extreme, i.e., at pH = 10, this capacity decreases to 4.88 mg/g.

Figure 11 reports the removal efficiency of Cr (IV) as a function of the pH: Increasing the initial pH from 2 to 10, the removal efficiency of Cr(VI) increases up to 99.21% at pH = 2. The minimum removal rate (i.e., 97.68%) was obtained at pH = 10. Anyway, this value is high and still satisfying, but at a pH higher than 6, another process occurs, which is precipitation, as mentioned in another work [71].

The highest removal rate of Cr(VI) at pH = 2 is a consequence of the protonation of the biosorbent as well as Cr(VI) speciation. Indeed, at a low pH, the proton concentration is high, and the negative charges on the pore surface of the biosorbent are, thus, new adsorption sites that are consequently available with positive charges, which is confirmed the pHpzc result [72]. Conversely, Cr(VI) in a solution can coexist as chromate (CrO₄⁻²), dichromate (Cr₂O₇⁻²), hydrogen chromate (HCrO₄⁻), or chromic acid (H₂CrO₄), according to the reaction scheme reported in Figure 12. At a low pH, ions of HCrO₄⁻ are the most abundant in a solution compared to those of Cr₂O₇⁻², and ions of HCrO₄⁻ are smaller in size than those of Cr₂O₇⁻² [73]. Therefore, monovalent ions of HCrO₄⁻ diffuse smoothly and are adsorbed more easily and in greater amounts by the pore surface than ions of Cr₂O₇⁻². At a pH below 1, H₂CrO₄ predominates in the solution as a polycyclic anhydride species that is difficult to be adsorbed by the pores [74].

3.3.4. Effect of the Initial Concentration of Cr(VI)

Figure 13 shows that the highest removal yield of 98% is achieved after 30 min, with a Cr(VI) concentration of 50 mg/L at pH = 5.1. With higher Cr(VI) concentrations, the removal efficiency decreases. This result occurs because, for a specific dosage of the biosorbent, when the concentration of Cr(VI) is the lowest, the ratio between the active sites of the surface and the molecules of the metals in the solution is the highest, and thus a higher amount of molecules is retained by the biosorbent and removed from the solution. Conversely, at higher Cr(VI) concentrations, the amount of metal ions adsorbed per unit mass of the biosorbent is greater, thus causing a saturation of the biosorbent surface and, therefore, a relevant amount of metal ions remains dissolved in the solution [75].

3.4. Results of the Full Factorial Design

According to the factorial plan for Cr(VI) removal by adsorption on the orange peels, 16 experimental tests (24) were conducted, as reported in

Table 2. Matrix of the experimental plan.

Test	m (mg)	X1	pH	X2	t (min)	X3	C0 (mg/L)	X4	YCr (%)
1	500	−1	2	−1	5	−1	10	−1	99.05
2	2000	+1	2	−1	5	−1	10	−1	99.18
3	500	−1	10	+1	5	−1	10	−1	93.89
4	2000	+1	10	+1	5	−1	10	−1	98.50
5	500	−1	2	−1	90	+1	10	−1	99.42
6	2000	+1	2	−1	90	+1	10	−1	99.29
7	500	−1	10	+1	90	+1	10	−1	94.88
8	2000	+1	10	+1	90	+1	10	−1	98.87
9	500	−1	2	−1	5	−1	200	+1	97.82
10	2000	+1	2	−1	5	−1	200	+1	98.10
11	500	−1	10	+1	5	−1	200	+1	97.80
12	2000	+1	10	+1	5	−1	200	+1	98.03
13	500	−1	2	−1	90	+1	200	+1	98.11
14	2000	+1	2	−1	90	+1	200	+1	98.70
15	500	−1	10	+1	90	+1	200	+1	98.01
16	2000	+1	10	+1	90	+1	200	+1	98.64

. In the last column of Table 2, the Cr(VI) removal efficiency is indicated.

The mathematical expression to calculate the Cr(VI) removal efficiency (Y_Cr) is as follows:

Y_Cr (%) = 98.02 + 0.64 m − 0.69 pH + 0.22 t + 0.13 C₀ + 0,54 m × pH − 0.012 m × t − 0.43 m × C₀ + 0.051 pH × t + 0.66 pH × C₀ − 0.008125 t × C₀ + ζ

(11)

where

• m: mass of the biosorbent (mg);
• C₀: the initial concentration of Cr(VI) (mg/L);
• t: contact time (min);
• ζ: residual error.

The analysis of variance (ANOVA table) resulted in a determination factor R²_(adjusted) = 58% and R² = 85.89%.

The “p”-value (probability of risk) was lower than 0.05 (5%) for only the mass of the biosorbent, the pH, and the (pH/Cr(VI) concentration) interaction; so, these set factors have the largest effect on the adsorption process of Cr(VI) on the orange peels.

Figure 14 shows that both parameters, the pH and biosorbent dosage, had a relevant effect on the removal efficiency of Cr(VI). The pH value was inversely correlated to the removal efficiency. Actually, an increase in the pH resulted in a decrease in the removal efficiency; therefore, Cr(VI) removal was more efficient in acidic media at pH ≥ 2. This finding can be explained by the predominant presence of HCrO₄⁻ ions when the pH is equal to 2 as well as by the surface of the biosorbent being charged positively because the pH_pzc is higher than the pH value of the solution. The biosorbent dosage also has a significant effect on the removal of Cr(VI), because the amount of active sites capable of adsorbing metal ions depends on it.

Figure 15 reports the effects of each parameter and their mutual interactions. The red and black lines indicate a different level (i.e., +1 and −1) for each parameter. Regarding the biosorbent dosage, the results display that the amount of 2000 mg is more effective than that of 500 mg for all the mutual interactions. Conversely, for the initial concentration of Cr(VI), it can be seen that the lower the concentration, the more efficient the removal of Cr(VI). Finally, concerning the pH, as already mentioned, at a low value (an acidic medium), the removal of Cr(VI) is more pronounced with a contact time of 90 min. Regarding interaction factors, in Figure 14, it can be noticed that any interaction with the contact time does not have a significant effect on the adsorption process. The interactions that significantly affect the removal rate are as follows: the biosorbent dosage with the pH, the pH with the initial Cr(VI)concentration, and the biosorbent dosage with the initial Cr(VI)concentration. Consequently, the parameters and mutual interactions that showed negligible effects on the Cr(VI) removal efficiency were deleted from Equation (11), thus resulting in a simplified expression as follows:

Y_Cr (%) = 98.02 + 0.64 m − 0.69 pH + 0.22 t + 0.13 C₀ + 0.54 m × pH − 0.43 m × C₀ + 0.66 pH × C₀ + ζ

(12)

According to the ANOVA study:

R² = 85.89%; R² (adjusted) = 73.55%.

Also, the “p”-value was lower than 5% for all the parameters investigated and their mutual interactions, except for the contact time and the initial Cr(VI) concentration.

3.5. Kinetic Study

The values of q_e, K₁, K_2, and K_in were determined from the intercept and slope of the line representing the linear relationship for each kinetic model (Figure 16, Figure 17 and Figure 18).

The results are reported in

Table 3. Kinetic constants of the first-order, second-order, and intra-particle scattering models for the different initial Cr(VI) concentrations.

	Pseudo-First Order		Pseudo-Second Order		Intra-Particle Model
C0 (mg/L)	Correlation Factor (R2)	K1 (min−1)	Correlation Factor (R2)	K2 (g/mg‧min)	Correlation Factor (R2)	Kint (mg/g·min1/2)
10	0.065	0.002	1	−55.608	0.009	0.153
20	0.525	0.003	0.9999	13.033	0.712	0.301
30	0.599	0.009	0.999	1.855	0.257	0.447
50	0.549	0.004	0.9998	16.797	0.576	0.751
60	0.071	0.004	1	20.231	0.674	0.903
80	0.201	0.003	0.9999	−26.633	0.183	1.209
100	0.088	0.002	0.9999	−22.271	0.234	1.509

.

From the linear regressions of the three previously mentioned kinetics (pseudo-first-order, pseudo-second-order, and intra-particle models), it can be noticed that the adsorption of Cr(VI) on the biosorbent surface is satisfactorily modeled by a second-order kinetics for all the initial Cr(VI) concentrations examined. The kinetic constants of each model are reported in Table 3.

3.6. Adsorption Isotherm Modeling

The adsorption capacity is often represented by isothermal curves, which can also provide information on the adsorption mechanism.

Figure 19 shows that the shape of the curve is almost similar to a type (II) isothermal curve. This type indicates a single-layer adsorption at a low concentration, followed by multilayer adsorption at a higher concentration. The lower the concentration at which multilayer adsorption begins, the stronger the interaction occurring between the adsorbed substance and the adsorbent.

This type of isotherm is characteristic of an exclusively physical adsorption on non-porous solids or solids presenting mainly macro-pores [76].

The equilibrium isotherms for the adsorption of Cr(VI) on the orange peel powder samples over a wide range of concentrations (10–100 mg/L) were modeled.

To examine the relationship between the sorbed (q_e) and aqueous concentrations (C_e) at equilibrium, sorption isotherm models are widely employed for fitting the data, of which the Langmuir, Freundlich, and Elovich equations are the ones most widely used.

To ensure equilibrium conditions, the linear form of the three models’ equations cited above (Section 2.6.1) was applied to the experimental data.

The equilibrium parameters obtained with the adsorption model of the Langmuir, Freundlich, and Elovich isotherms are reported in

Table 4. Isotherm constants for the adsorption of Cr(VI) on the biosorbent.

Isotherm Type	Linearization of the Equations	Constants	R	R2
Langmuir	$\frac{1}{\mathrm{q}\mathrm{e}}$ = $\frac{1}{\mathrm{q}\mathrm{m}\mathrm{a}\mathrm{x}}+\frac{1}{\mathrm{K}\mathrm{L}\mathrm{q}\mathrm{m}\mathrm{a}\mathrm{x}}$	qmax = 5.46149 KL = 1.17597	0.92	0.85
Freundlich	$\ln \left(\mathrm{q}\right)=\ln \left({\mathrm{K}}_{\mathrm{f}}\right)+\frac{1}{\mathrm{n}}$·ln (Ce)	n = 1.36479 Kf = 2.66038	0.92	0.85
Elovich	ln$\frac{\mathrm{q}\mathrm{e}}{\mathrm{C}\mathrm{e}}=$ ln (keqm)$-\frac{\mathrm{q}\mathrm{e}}{\mathrm{q}\mathrm{m}}$	qm = 130.03901 Ke = 0.01899	−0.057	0.003

. The values of the correlation coefficient (R²) are higher for the Langmuir and Freundlich isotherms than the Elovich isotherm model. Therefore, the Langmuir model assumes that the uptake of adsorbate molecules occurs on a homogenous surface by monolayer adsorption without any interaction between the adsorbed molecules. The Freundlich model is suitable for non-ideal adsorption on heterogeneous surfaces [77]. This result may indicate the formation of single-layer adsorption on the active sites of the surface at a low concentration in the first time, followed by multilayer adsorption at a higher concentration.

The graphical maximum adsorption capacity of the Langmuir isotherm (q_max) was 5.46 mg/g.

The balancing parameter, R_L, of the Langmuir model was calculated using the following equation:

$$R_L=\frac{1}{1+\left(k_L\times C_e\right)}$$

(13)

where R_L (dimensionless) indicates the tendency of the adsorption process to take place when the following conditions occur:

(a)

0 < R_L < 1, the adsorption is favorable;

(b)

R_L > 1, the adsorption is unfavorable;

(c)

R_L = 0, the adsorption is irreversible;

(d)

R_L = 1, the isothermal representation is linear [78].

The R_L values ranging from 0.205 to 0.826 for the different initial Cr(VI) concentrations studied indicate that the process of Cr(VI) adsorption by the biosorbent is favorable.

3.7. Chemical Structure Considerations

It must be considered that, when one talks of the absorption of Cr(VI) on a porous surface, it is not the generic interaction of a metal ion with the unspecified sites of the surface of the porous medium, but that there are clear interactions by secondary forces and perhaps the coordination complexing of some kind with particular structures of the orange peel constituents, i.e., specific adhesion. First of all, carbohydrate oligomers and polymers are well known to easily complex chromium (VI) and Cr (III) with structures of the types that follow: Cr(VI) being reduced to Cr (III) by the carbohydrates and the carbohydrates being simultaneously oxidized as follows:

This is a simplified form of the specific adhesion mechanism by carbohydrates, with the series of mechanisms involved being considerably more complex but well determined [66].

Equally, coordination complexes can occur with hesperidin glucosides, where complexation, and hence the absorption of Cr(VI), can occur with both the ortho –OH and –OCH₃ groups of the B-ring of its flavonoid as well as with the carbohydrate chain linked to it. Thus, the structure of hesperidin is:

Not only coordination and absorption can occur in carbohydrates, but also in the classical coordination of ortho-oxygens of the B-ring for complexes of this kind [67].

Moreover, different types of coordination complexes can form in lignin [79]. These are:

1. Covalently bonded CrO₄⁼ chromate esters, in which some CrO₄⁼ oxygens are covalently bonded to the aromatic nuclei of phenylpropopane units [80,81].

2. CrO₄⁻ coordination complexes with hydroxyl and methoxy groups of the aromatic nuclei of lignin phenylpropane [80,82]; hence, the same type of coordination complexes, as shown above for hesperidin, can occur. Such complexes can also occur with Cr (III), especially the Cr(III) produced by reduction from Cr(VI) by the oxidation of carbohydrates.

3. Ferrocene-type complexes, presenting structures of the type as shown, with the aromatic rings of lignin or any other aromatic rings present, may also occur [79].

In the above, M is either CrO₄⁼, hence Cr(VI), or Cr(III).

All these types of complexes have already been found and characterized for the adsorption of chromium on lignocellulosic material.

One can think that the citric acid present in orange peels can also react to form esters, but it definitely esterifies both carbohydrates and lignin during the drying of the orange peel, as already determined in the case of other applications of citric acid to lignocellulosic materials [83].

4. Conclusions

A preliminary study of the influence of some parameters on the biosorption process was conducted. The results of this study show that orange peel is efficient as a biosorbent and the adsorption process is fast, since the chemical equilibrium is achieved in a relatively short first time, i.e., from 5 to 15 min. The highest removal efficiency was 99.2%, and it was achieved in an acidic solution (pH = 2), thus confirming the validity of the theory of pH_pzc and contact-free pH.

Moreover, an analytic study on the effects and mutual interactions of the operating parameters, such as the pH, the biosorbent dosage, the initial concentration of Cr(VI), and the contact time, was conducted through a full factorial experimental design. This study confirmed that the adsorption process mainly depended on the biosorbent dosage and the pH, whose optimum value was 2. Furthermore, the study of the effect of the biosorbent dosage was extremely useful to set the optimum value of the solid/liquid ratio (r) at 10 g/L as the dosage of the biosorbent. Under such conditions, the Cr(VI) removal efficiency was around 97% when the initial concentration of Cr(VI) was set at 50 mg/L. In addition, a kinetics study was also performed and the adsorption process was satisfactorily modeled by pseudo-second-order kinetics.

Finally, the experimental results were compared to the Freundlich, Langmuir, and Elovich isotherm models. The results demonstrate that the adsorption of Cr(VI) on the biosorbent are well described by the isotherm models of Langmuir and Freundlich. According to Langmuir’s dimensionless factor (R_L), it can be concluded that the Cr(VI) adsorption process on the orange peels in their natural state is favorable (0 < R_L < 1).