Potato Peels as an Adsorbent for Heavy Metals from Aqueous Solutions : Eco-Structuring of a Green Adsorbent Operating Plackett – Burman Design

Treatment of wastewater is becoming a concern of an increasing prominence. Trace amounts of toxic metalloids and heavy metals (HMs) would contaminate large volumes of water. Being present as traces, removal of these ultratrace contaminants from wastewater is challenging. Adsorption of HMs onto raw (RPP) and burnt (BPP) potato peels (PP) is presented in the current treatise. Both adsorbents (RPP and BPP) proved to be efficient in removing Cd(II), Co(II), Cu(II), Fe(II), La(III), Ni(II), and Pb(II) from aqueous solutions. BPP was a more efficient adsorbent compared to RPP. Ecodesign of a model, green adsorbent was structured executing a multivariate approach, design of experiments (DoE). *e purpose of using DoE is to maximize the efficiency of BPP (carbonaceous biomass) as a versatile adsorbent. Plackett–Burman design (PBD) was used as a screening phase. Four factors were considered: pH, contact time (CT), heavymetal concentration (HMC), and the adsorbent dose (AD).*e Pareto chart of standardized effects shows that the most influential factor is the HMC.*ese data were confirmed by analysis of variance (ANOVA). Derringer’s function was operated to find the best factorial blend that maximizes the adsorption process. *e percentage (%) removal of Cd(II), for example, was maximized hitting 100%. Adsorbent surface characterization was performed using FTIR, BET, SEM, TGA/dTG, and EDX analyses. Adsorption was found to be physisorption that follows Temkin isotherm with sorption energy 66 kJ/mole. Adsorption kinetics was found to be pseudo-first-order. Adsorption capacity (qm) for BPP was 239.64mg/g. *e diffusion inside the particles was very limited, while the initial rate of the adsorption was extremely high as shown by the Elovich plot.


Introduction
With the progress in human life and the consequent conversion to industrialization, water pollution is becoming a worldwide concern.Many heavy metals (HMs), such as cadmium, lead, copper, and mercury, which originate from anthropogenic activities, were detected in industrial wastewaters.Existing as traces, being hazardous to the ecosystem and human health and being non-biodegradable and difficult to remove even at low concentrations, HMs represent an ecological burden that has attracted great attention.Moreover, the annual toxicity of all metals exceeds the combined total toxicity of the radioactive and organic wastes generated annually [1,2].
According to the US Environmental Protection Agency (EPA), the most toxic HMs are mercury, lead, aluminum, cadmium, arsenic, chromium, and nickel [3].Cadmium (II) is a water contaminant transferred mainly by discharge from metal refineries, galvanized pipes corrosion, and natural deposits erosion as well as smoking.Having a long half-life, a capability to harm the DNA, and consequently a cumulative toxic effect, exposure to Cd(II) as per many regulatory agencies is associated with cancer.Other toxic effects associated with Cd(II) include osteoporosis, height loss in men, kidney damage, elevated blood pressure, and cardiovascular diseases [3][4][5][6].
Removal of HMs from contaminated samples has been performed by a variety of approaches.Examples for these techniques include ion-exchange, reverse osmosis (RO), chemical precipitation, electrolysis, and sorption [7][8][9].Yet, most of these techniques are time, cost, chemical, and energy consuming.Moreover, most of these procedures are not efficient for HM traces and might cause production of secondary pollutants [10][11][12].Adsorption, on contrary, is not only simple but also cost-effective and efficient.Modelling a standard adsorbent that possess all these profitable criteria becomes an exacting task [13,14].
In 2016, around 376 million tons of potatoes were produced globally.Processing industries are one of the fates of potatoes, where nearly 14% of the harvested crop worldwide is predestined to produce chips, starch, and crisps.It is probably understood that the rest of the crop would be domestically used.For either fates, peeling of potatoes is usually the preprocessing approach commonly followed.
e produced peels might be used for a variety of purposes, including production of biogases, feedstock, and fertilizers [16].
Yet, accumulation of these peels would represent an environmental burden.Recycling of these wastes into a lowcost, versatile adsorbent for the removal of HMs, dyes, drugs, etc. would be a solution to get rid of these wastes and simultaneously clean wastewater.Being rich in high-value constituents, specially polysaccharides and lignin [16,20], which can be converted into carbonaceous material by burning, PP would be an elegant green adsorbent.
Few studies have been reported in the literature employing PP as an adsorbent.Moreover, to the best of our knowledge, none of these approaches was chemometricsassisted, and factors usually affecting the adsorption process, e.g., pH, dose of adsorbent (AD), heavy metal concentration (HMC), and contact time (CT) of the adsorbent with the contaminant were rather investigated using the traditional univariate-based procedure [21][22][23][24][25][26][27].ough being the approach commonly used by researchers, univariate analysis (UVA), with exhaustion of capitals, time, and efforts, unenviably affects the method greenness.A comparison between the present approach and the reported studies employing PP as an adsorbent is shown in Table 1.
Modelling of a green adsorbent from PP using a facile, cost-effective, and efficient procedure, optimization of the adsorption process, and maximizing the removal efficiency of PP were the tasks undertaken in the current approach.A multivariate platform exploiting Plackett-Burman design (PBD) was the maneuver of choice [28,29].A comparison between the RPP and BPP was carried out.Surface characterization of adsorbents was performed using SEM, EDX, BET, and TGA/dTGA.FTIR together with TGA/DTG were used to decide on the functional groups on the adsorbent surface that might be responsible for the adsorption process.Prepared adsorbent samples (BPP) were further used for equilibrium and kinetics studies.

Materials and Reagents.
All chemicals used in the current investigation were analytical grade reagents and solvents.Potatoes were purchased from local markets (Doha, Qatar).All solutions were prepared using ultrapure deionized water (18.2MΩ).A stock solution of a mixture of 250 ppm of HMs (Cu, Zn, Ni, Cd, Fe, Pb, Co, and La) was used as the artificially contaminated water.Dilutions of the previous stock were initially tested to find the HM that is most adsorbed by either adsorbent using inductively coupled plasma-optical emission spectrometer (ICP-OES, ermo Scientific-iCAP 6500).Cadmium nitrate tetrahydrate (Cd(NO 3 ) 2 •4H 2 O; MW 308.48 g/mol, Sigma-Aldrich, USA) was used to prepare the single-element solution (a stock solution of 500 ppm) for flame atomic absorption spectrometric (FAAS, Shimadzu 6800) measurements.Nitric acid (5%, BDH, USA) solution was used to wash all the glassware (pipettes, volumetric flasks, beakers, and watch glasses) followed by deionized water (DIW).Hydrochloric acid and sodium hydroxide (2% solution, Sigma-Aldrich) were prepared to adjust pH of the solution.A ermo Scientific centrifuge (SL8) was used to centrifuge the samples after treatment with the adsorbent at 4200 rpm and for 30 min.Sterilized Whatman syringe filters (0.45 μm) were used to filter the desired samples.

Adsorbent Preparation and Characterization.
Collected potatoes were washed using DIW to remove dirt and impurities, and then peels were soaked in DIW for 3 h.Clean peels were placed in ovens ( ermolyne furnace 48000) at 100 °C until dryness.Dry PP (10 g) was grinded using a nonmetallic blade mill operated at the maximum speed for 15 min and divided into two portions.e first portion was labelled raw potato peels (RPP) and was kept in a sealed container in the desiccator.e second portion was burned in furnace at 500 °C for 30 min and labelled burnt potato peels (BPP).
Topographic features of the prepared adsorbent were determined using a field emission scanning electron microscopy (SEM) (FEI, Quanta 200, USA).Energy-dispersive X-ray spectroscopy (EDX) was used for the elemental analysis.Disintegration of the adsorbent with respect to temperature was studied using a thermal gravimetric analyzer (TGA, PerkinElmer-TGA 400).Fourier-transform infrared spectroscopy (FTIR, PerkinElmer Spectrum BXII, USA (transmission mode), PerkinElmer Spectrum 400, USA (ATR unit, reflection mode, Central Lab Unit, Qatar University), and Bruker Alpha, USA (ATR unit, reflection mode)) were used to determine the functional groups in both RPP and BPP.Spectra of samples were reported in the range of 4000-400 cm −1 , with a resolution of 4 cm −1 and implementing KBr pellets.
Brunauer-Emmett-Teller (BET) surface area and pore size distribution for the prepared samples were determined using a Micromeritics ASAP 2020 Accelerated Surface Area and Porosimetry system.Degassing of the sample was first performed at 100 °C followed by the nitrogen 2 Journal of Chemistry Journal of Chemistry adsorption-desorption processing.Surface area was thereafter calculated based on the nitrogen isotherms measured at 77 °K and implementing the BET equation.Pore volume was obtained using the t-plots and the Barrett-Joyner-Halenda (BJH) equation.e total pore volume was evaluated at a relative partial pressure of 0.99 cm 3 /g.

Design of Experiments (DoE).
Experimental design was built up using Minitab ® 17 software.Four factors with their upper and lower boundaries were fed to the software, and Plackett-Burman design (PBD) was executed.Factors, their codes, and their real values at the three levels (low, mid, and high) are displayed in Table 2.

Procedure
(1) ICP-OES Measurements.A preliminary study was performed to test the ability of PP in removing HMs from a multielement aqueous solution.Accordingly, different amounts of RPP (0.1-0.5 g) and a volume of 10 mL of a 1 ppm multielement HM solution were mixed into a 15 mL centrifuge tube.On the other hand, blanks (no HMs or untreated HM solutions) were similarly prepared and centrifuged at 4200 rpm for 30 min.e supernatant was then separated and filtered using a syringe filter and transferred to a new set of 15 mL tubes.A similar procedure was followed employing BPP.All filtered solutions were analyzed using ICP-OES, and the % removal of each metal by the RPP and BPP (Y) was calculated using equation (1), where C i and C f are the initial and final concentrations of the HM.
(2) FAAS Measurements.A stock solution, 500 ppm of Cd (II), was prepared by dissolving the respective amount of Cd(NO 3 ) 2 •4H 2 O in ultrapure DIW.Serial dilutions to obtain solutions with concentrations of 10-200 ppm were prepared in the same solvent.FAAS was calibrated by using Cd(II) solutions of 10, 20, 26, 30, 40, and 50 ppm with R 2 � 0.9969.A Cd(II) hallow cathode lamp was operated at 228.8 nm and 8 mA, while the burner is completely in a cross position.e planned PBD design was executed as shown in Table 3, where 13 base runs, with one added central point (Ct.Pt.), were generated.As shown in this table, a specific amount of Cd(II) and the stated adsorbent dose (AD) of BPP were added to each tube and left to react for the set out contact time (CT).e pH of all samples was varied between 3.00 and 9.00 ± 0.20 using small aliquots of 2% solutions of HCl/NaOH.Samples were centrifuged at 4200 rpm, filtered into a new set of labelled tubes using a Whatman syringe filter, and analyzed using FAAS.
(3) Adsorption Kinetics and Isotherms Using BPP.A series of Cd(II) solutions (5-500 ppm) were prepared in 15 ml tubes using DIW without adjusting the pH at the room temperature.An amount of 0.1 g BPP was added to each tube.Prepared solutions were kept in a shaker for 3 h at 150 rpm at room temperature.Following this treatment, samples were kept aside for 30 min, and then the supernatant was decanted and filtrated using a 0.2 μm syringe filter.Samples were then analyzed by FAAS.For the kinetics study, 300 mL of a 100 ppm solution of Cd(II) was prepared.Solution was then divided into three equal portions, and an exact amount of 0.506 g of BPP was added to each beaker when the solution reaches the assigned temperature (20, 31, and 41 °C).e reaction medium was kept under stirring at 280 rpm, while 2 mL were withdrawn every 5 min, filtrated with 0.2 μm syringe filter, completed to 15 mL with DIW, and finally analyzed by FAAS.

Preliminary Results of ICP-OES for Both RPP and BPP.
e main goal of this study was to test the efficiency of asprepared PP for removal of HMs from artificially contaminated water samples.As shown in Figure 1, BPP seems to be a more efficient adsorbent compared to RPP. e later, in turn, showed a reasonable adsorption efficiency in case of Cd(II) and Pb(II), with a % removal of 64.09 and 77.34%, respectively.
is superior capability of BPP will be explained later when it comes to the surface characterization.

Screening and Structuring of a Regression Model.
e efficiency of PP as an adsorbent was optimized using a multivariate approach.As the purpose was to develop an ecodesign-based adsorbent, it was necessary to acquire the needed data by performing the lowest number of experiments.Multivariate analysis (MVA), in addition to its competency in diminishing the waste in experimentations, chemicals, and resources, serves to produce a top-notch data which can be further exploited with a great degree of inevitability.
In the current investigation, Plackett-Burman design (PBD) was implemented as a screening approach.PBD, in comparison to other screening and response surface methodologies (RSM), is one of the commonly operated approaches in robustness tests used for analytical method validation [28,29].Running this design would guarantee a low number of trials, where only the key effects are measured.Variable-variable interactions, on the other hand, are profoundly confounded with these key effects.2), were utilized to signify the "weighty" factors among all investigated variables.As shown in Figure 2, HMC (B) is the most in uential factor, followed by the AD (C).e following regression formula was obtained following a response transformation process [30,31]: ( As shown in equation ( 2), the value of R 2 (predicted) is relatively high, implying the capability of the proposed model to calculate new observations as appropriately as it ts the current data set.Experimental and predicted values are shown in Table 3. Comparing values predicted after response transformation to the observed data shows an excellent match.
Statistical validation of screening data was performed using ANOVA testing, Table 4. Similar conclusions can be obtained.By and large, the increase in AD corresponds to a consequent elevation in the adsorption surface area and hence an increase in the removal e ciency.However, as shown in equation ( 2), the AD (C) negatively in uences the removal e ciency of BPP. is case was previously reported [32][33][34].Many justi cations were given, e.g., possible hindrance caused by the interaction of the active sites of adsorbent when the dose is increased.Other elucidations count  on the influence of the blend of experimental conditions employed on the adsorbent itself where aggregates of the adsorbent might form at a high concentration, an issue that hinders the metal uptake.Taking into consideration the large size of Cd(II) and hence existence of possible steric hindrance, HM ions would be adsorbed only on the freely reachable adsorbent sites.
(2) Main Effects, Cube, and 2D Plots.To examine the difference between each level means on the response, main effects plot was drawn for all variables (effectual model variables and pH).As shown in Figure 3 there is a main effect when altered levels of the same variable influence the response differently.us, HMC (B), which has the steepest line slope (upper right panel), would have the greatest effect on the response.In contrary, CT (D) has a line that is parallel to the x-axis, indicating no main effect.
Cube plots, on the other hand, (Figure 4) show the relation between variables and the % removal (Y).As shown in the graph, two cubes display all the patterns of a variable setting for both significant and insignificant variables.A blend of, for example, CT 5 min, pH 9, AD 3 g/L, and HMC 200 ppm, translates to a % removal of 95.5589.
A 2D view of the relationship between a fitted response and the interaction of two variables is shown in the contour plot (Figure 5); dashed lines are used to represent the contours.Dark purple regions disclose zones with the highest % removal as well as the corresponding factorial setup.

Response Optimization.
One response was considered in this study, % removal (Y), and the goal was to maximize the response.A common approach is the operation of desirability plots.In this approach, factorial settings that can achieve maximum/minimum/target response can be controlled following the determination of the goal.A value of D "composite" or d "individual" desirability close or equal to 1.0000 infers a better optimality and hence a preferentiality to the target [35].As shown in Figure 6, ideal settings are denoted as "curr.".For example, a factorial blend of HMC of 200 ppm, AD of 0.1 g/L, and CT of 5 min.at pH of 6.00 ± 0.2 would achieve 99.19% of removal.Moving the vertical solid line to the right or the left shows the impact of different factorial setting on the desirability.

Adsorbent Characterization
3.4.1.ATR-FTIR Analysis and the Proposed Mechanism for the Adsorption Process.FTIR spectra of RPP and BPP are shown in Figure 7.For RPP, a band between 3000 and 3300 cm −1 was observed confirming the presence of free and H-bonded -OH stretching vibration of the hydroxyl group in the adsorbent structure (derived from polymers composing the adsorbent, e.g., cellulose, hemicellulose, and lignin). is broad peak may be attributed to the existence of water molecules on the adsorbent surface [36,37].Peak observed at 2938 cm −1 can be attributed to aliphatic C-H bond stretching of the methyl and methylene groups in the polymers existing in PP [38].Carbonyl (C�O) stretching vibration of aldehydes at 1632 cm −1 reflects the presence of the aromatic compounds in the lignin component of RPP [39,40].Peak at 1572 cm −1 can be assigned to the aromatic C�C stretching mode.Peaks at 1409 and 1154 cm −1 indicate the presence of aryl OH groups.Peaks at 1081 and 1364 cm −1 can be attributed to -OH and C-O stretching vibrations of the carboxylate group in lignin, hemicellulose, and cellulose [39,41].Yet, the FTIR fingerprint region shows that the main component is hemicellulose [42].On the other hand, spectrum for BPP shows a very broad band between 2700 and 3600 cm −1 and a peak at 1631 cm −1 , which can be attributed to aldehydes or carboxylic acids formation.Presence of different carbon-oxygen functional groups is indicated by the peaks at 1631, 1236 and 1364, and 1071 cm −1 for C�O, C-O-C, and C-O, respectively.

BET and TGA Analysis.
Porosity of prepared adsorbent is an important variable that impacts the performance of the adsorption process.Nitrogen adsorption-desorption measurements would portray a picture on the surface area readily available to the HM. e BET surface area and pore volume of both adsorbents were determined and shown in Table 1.e BET surface area of the RPP was around 3 times smaller than that of BPP, whereas the single point total pore volume was about 2 times smaller in case of RPP compared to BPP.Pore size distribution, another important feature, exemplifies the physical heterogeneity of the adsorbent surface [43][44][45].Figure 8 shows that BPP has type V isotherm, indicating that BPP is porous and that as the concentration increases; adsorption of N 2 gas occurs via multilayer adsorption followed by capillary condensation.Referring to IUPAC guidelines (microporous: pore size <2 nm, mesoporous: pore size � 2-50 nm, and macroporous: pore size � 2-50 nm), RPP is mainly micro-mesoporous where peaks are mainly observed at a pore diameter < 2 nm and between 2 and 50 nm, while BPP showed mainly mesopores with fewer macropores and much less micropores.ese data show the thermal activation of PP is essential for enhancing the porosity and hence the adsorption process.TGA graph shown in Figure 9 characterizes the % weight loss of RPP against temperature in the range of 30-800 °C.Decomposition of RPP is shown where the dashed line represents TGA curve, while the solid line represents differential TGA curve (dTGA).According to the obtained data, the rst decomposition peak at ∼60 °C (4.65%) is probably due to water vapor or humidity in the sample.
e second clear sharp peak at ∼300 °C (64.37%) can be attributed to the decomposition of hemicellulose, starch, and cellulose content of the peels, which are decomposed at 263, 296, and 354 °C, respectively.In addition, the TGA curve shows that the peel consists mainly of hemicellulose and starch with little amount of cellulose.ese ndings agree with FTIR data in the nger print region.e remaining undecomposed peels (∼30.9%)can be attributed to the carbon content which is necessary for the adsorption process [5,42,46].10(a   Journal of Chemistry the capability of BPP to uptake the metal ions [21,47].Consolidating the ndings of FTIR and TGA and in agreement with previous studies, the burning process of agricultural wastes results in increasing the carbon content in the prepared samples due to the destruction of cellulose and organic materials.In addition, FTIR analysis of the burnt material shows the formation of carbonyl, ether, ketones and carboxylic groups and the disappearance of hydroxyl group [48,49]. is also would explain the improved capability of BPP to remove HMs compared to RPP, as is shown in Figure 1.

EDX Analysis.
EDX was used for the elemental analysis of RPP and BPP.Attained data, Figure 11, show that the % of carbon in RPP BPP is 55.85 and 74.76, respectively.On the other hand, oxygen content decreased  from 37.30% in RPP to 16.65% in BPP.ese data con rm that most oxygen functional groups were eliminated by burning of PP and converted to mainly carbon which is responsible for the high adsorption e ciency of Cd(II).EDX shows also the presence of small concentrations of other elements, e.g., Mg, P, S, Cl, K, and Ca, and their concentrations before and after burning were less than 2%.

Equilibrium, Isotherms, and Kinetics Studies of Cd(II)
Adsorption on BPP.Two factors should be considered when designing a good sorbent: (i) the adsorbent should be porous and with a reasonable surface area and (ii) adsorption rate should be high with more preference given to chemisorption [50].

Adsorption Isotherms.
Adsorption can be described using di erent isotherms.Yet, a suitable isotherm is the one that can basically replicate the experimental data.Four isotherms have been considered in the current approach: Langmuir, Freundlich, Temkin, and Dubinin-Radushkevich (D-R).It is noteworthy to mention that the capacity of the adsorption isotherm is crucial in determining the adsorption e ciency of the investigated adsorbent [50].
As shown in Figure 12(a) and Table 5, the Langmuir isotherm reveals that adsorption has three regions: (i) unfavorable region where C 0 ranges from 0 to 0.78 mg/L, (ii) the second region 0.78-200 mg/L in which the favorability increases with increasing the concentration, and (iii) the third region >200 mg/L, where the dimensionless constant becomes close to zero and therefore the adsorption becomes irreversible and precipitation occurs.Adsorption capacity (q m ) was calculated, Table 5, and found to be 239.64 mg/g, which is higher than the previously reported values for adsorbents prepared from PP [21-23, 25, 26], or other adsorbents, e.g., sweet potato [5] where loading capacity of Cd(II) was 18 mg/g.Freundlich isotherm, Figure 12(b), shows two regions, where in both the reciprocal of the exponent of nonlinearity (n) is > the unity.erefore, there is a competition between adsorbate (Cd (II)) and water molecules on adsorption sites.
Although most of the conducted investigations on the adsorption of Cd(II) refer to Langmuir [51] and Freundlich [52] models as proposed isotherms, the adsorption curve is well tted to the Temkin isotherm as shown in Figure 12(c).Temkin constant b T has a value of 66 kJ/mole (less than 80 kJ/mole) [53], so the interaction between adsorbate and adsorbent is physical.R-D isotherm con rms that the adsorption is physical adsorption with energy 5 kJ/mole (less than 8 kJ/mole) [54].

Kinetics Studies.
More than 95% of Cd(II) were consumed within 5 min of the contact time.e consumption then slows down until the equilibrium was reached within 3 h e q e RT/b T ln A T C e q e q m exp(−k ad ε 2 ) Linear form (q e /C e ) q m K L K L q e ln q e ln K F + (1/n)ln C e q e (RT/b T )ln A T + (RT/b T )ln C e ln q e ln q m − k ad ε 2 Parameters Region (I): q m 0.    (c)  at room temperature.e order of the adsorption reaction employing the integration method is not clear; however, Arrhenius' plots show that the reaction could be pseudosecond-order with respect to Cd(II) consumption and pseudo-first-order with respect to adsorption of Cd(II) on BPP (Figures 13(a) and 13(b)).ese findings align with many reported investigations and as shown by Ho et al. [56].e activation energies are found to be 34.46 and 46.63 kJ/mole, while the pre-exponential factor is 4100 L•mg −1 •min −1 and 2479853 min −1 , for Cd(II) consumption and adsorption processes, respectively.Moreover, the reaction order of the adsorption confirms that the process is physical adsorption.
Intraparticle diffusion (Weber-Morris plot Figure 13(c)) shows that there are two segments with coefficient constants of 5.2 × 10 −2 and 2.6 × 10 −2 mg•g −1 •min −1 , while the barrier thickness is 19.07 and 19.16 mg•g −1 for first and second steps, respectively.In addition, the adsorption is not controlled only by intraparticles diffusion, where and as shown by Weber-Morris diffusion plot, a thick layer is formed around the particle limiting the diffusion.Figure 13(d) shows Elovich plot, where desorption constant (β) is 0.0561 g/mg and initial sorption rate (α) equals 0.135 × 10 150 mg g −1 •min −1 , which is extremely high.
Findings of equilibrium, BET surface area, porosity, and kinetics studies are in alignment with each other.Type V N 2 adsorption isotherm, confirming the porosity of BPP, together with Langmuir isotherm, where the adsorption of Cd(II) on BPP is unfavorable at low concentration, then becomes irreversible with increasing Cd(II) content, which are indicative of physisorption.Freundlich isotherm shows that there is a competition between adsorbate and water molecules, while Temkin and D-R isotherms show the adsorption sites are heterogeneous and the adsorption is endothermic and physical.Kinetics studies confirm the adsorption is physical due to the pseudo-first-order of the reaction.
From the above, the suggested mechanism is physical adsorption, where Cd(II) is attracted to the particles under the influence of negative charges on the BPP surface that is rich with oxygen-groups.e Cd(II) fill the pores of BPP in extremely high adsorption rate to form multilayers around the particles.

Conclusion
e aim of this study was to optimize the efficiency of potato peels as a natural and green adsorbent for Cd(II) from contaminated aqueous samples.Two achievements were gained: cleaning of the contaminated water samples from toxic HMs and getting-rid of the peels that would accumulate and act as a source of pollution.Burning of potato peels into a char (carbonaceous biomass) was done through a simple process with no use of chemicals.Obtained results show that BPP is more efficient as an adsorbent compared to RPP.To ensure the method greenness, data were acquired through a minimal usage of efforts and resources using the DoE and in specific, Plackett-Burman design PBD as an approach.Four factors were investigated: pH, adsorbent dose, contact time, and heavy metal concentration.
Optimum conditions were decided upon after three stages: screening, optimization, and then verification using ANOVA and quality tools.A regression model was generated, and three variables were proved to be significant.Adsorbent characterization was done using FTIR, where several functional groups were proved to exist on the adsorbent surface.SEM analysis revealed the existence of pores on the surface of BPP.Furthermore, the equilibrium, BET, and kinetic studies indicate that the BPP is a porous and Cd(II) are adsorbed physically on the BPP surface as multilayers, which is well fitted by Temkin isotherm.

Figure 1 :Figure 2 :
Figure 1: A Comparison of BPP and RPP in the adsorption of HMs from aqueous multielement samples.

Figure 3 :
Figure3: Main e ects plots acquired subsequent to response transformation.A gray background represents a term not in the model.

Figure 4 :
Figure 4: Cube plots for the model terms and pH.

Figure 6 :
Figure 6: Optimization plots.Optimal values are represented by the vertical solid lines.

Table 1 :
Comparison between the performance of PP prepared in the current approach and PP as reported in the literature.Adsorbent (PP) given name and abbreviations are as mentioned

Table 2 :
Variables and their domains for a Plackett-Burman design (PBD) for both RPP and BPP.

Table 4 :
Analysis of Variance (ANOVA) for transformed response.
e : adsorbed amount at equilibrium; q m : maximum adsorbed amount; C e and C 0 : equilibrium and initial concentrations, respectively; K L : Langmuir rate constant; R l : dimensionless constant R l (1/1 − K L C 0 ).K F : Freundlich rate constant; n: exponent of nonlinearity; R and T: ideal gas constant and temperature in Kelvin, respectively.b T : Temkin isotherm constant; A T : Temkin isotherm equilibrium binding constant; ε: Polanyi potential; ε RT ln(1 + 1/C e ); E: adsorption energy E (1/ ad ); k ad : sorption energy constant. q