Removal of Pb(II) from water samples using surface modified core/shell CdZnS/ZnS QDs as adsorbents: Characterization, adsorption, kinetic and thermodynamic studies

2.1 Chemicals

Cadmium chloride (CdCl₂·2.5H₂O, ≥98 %), zinc nitrate (ZnNO₃·6H₂O, ≥98 %), sodium sulphide (Na₂S·9H₂O, ≥ 99.9 %), sodium hydroxide (NaOH, ≥ 99 %), zinc chloride (ZnCl₂, ≥98 %), 2-mercaptoacetic acid (2-MAA, ≥ 98 %), ethanol and 3-mercaptopropionic acid (3-MPA, ≥98 %) purchased from Sigma Aldrich and were used without any distillation.

2.2 Synthesis of surface modified CdZnS (Core)/ZnS (Shell) QDs via different capping agents

Surface modified core/shell QDs were synthesized by following the reported aqueous colloidal method with modification (Yakoubi et al., 2016; Masab et al., 2018). Briefly, 0.0271 g of CdCl₂·2.5H₂O was dissolved in 50 mL of de-ionized water to form 2 mM solution. In similar manner the 1 mM solution of ZnCl₂ was prepared in a 50 mL by dissolving 0.0070 g in de-ionized water. 1 mM solution of 2-MAA was prepared in 100 mL de-ionized water by taking 71 µL with micropipette in a 100 mL in three-necked round bottom flask having 50 mL of de-ionized water. One by one 50 mL of each solution was added into the flask and nitrogen was purge for 15 min. Later flask was clasped on hotplate in a heating mantle with condenser and thermometer in two necks. After 10 min, stirred and heated mixture turns turbid which became clear when pH was adjustment to 8 with 2–2.5 mL of 1 M NaOH. After that prepared 10 mM solution of Na₂S·9H₂O by combining 0.1201 g in 50 mL de-ionized water was reckoned up into mixture solution. When solution attained temperature of 100 °C heating was stop and mixture was left on continuous stirring for approximately 3 h for growth of CdZnS QDs.

Meanwhile, in another beaker 1.1 mL of Na₂S·9H₂O (0.1 M) solution (5 mL de-ionize water by weighing of 0.1201 g), 1.1 mL of ZnCl₂ (0.15 M) solution (0.1022 g in 5 mL de-ionized water) and 2-MAA (50 µL from 0.04 M solution in 50 mL of de-ionized water) was taken. After few minutes, mixture turns turbid which became clear when pH was adjustment to 8 with 2–2.5 mL of 1 M NaOH. After 3 h, the mixture was injected into the crude solution of CdZnS QDs after flux. The reaction was again heated for 1 h at 100 °C with constant stirring for development of CdZnS/ZnS core/shell QDs. After the completion of the process, 2.5 mL of ethanol was added into cooled reaction mixture for the removal of organic impurities through precipitation. For the elimination of inorganic impurities solid material was re-disperse into the solution, 15 mL of de-ionized water was poured and solution get centrifuge to form cluster. The centrifuged QDs were left in oven at 101 °C for 24 h. Likewise, QDs capped with 3MPA and CdZnS (core)/ZnS (shell) of same shape and structure were prepared by following aforementioned method using ZnNO₃·6H₂O instead of ZnCl₂ and 3MPA instead of 2-MAA.

2.3 Instrumentation

CdZnS/ZnS QDs dispersed in distilled water were initially characterized by UV–Visible spectra (Schimadzu, 1800 spectrophotometer) in 200–800 nm wavelength range, S1. The X-ray diffraction (XRD) patterns of QDs were recorded on Shimadzu XRD6000 system with Cu-Kα radiation (0.15406 nm). To study the functional groups involvement in the QDs synthesis FT-IR spectra were recorded with FT-IR Shimadzu 8400 s (S2 and ST1). For morphological analyses of QDs, scanning electron microscopy (SEM) (JSM-6380A, JEOL, Japan) and EDS (EX54175jMU, JEOL, Japan) images were taken. While, atomic absorption spectrophotometer (AAS) from PE-AAnalyst700 (Norwalk, CT, USA) was utilized for determination of Pb(II) in the water samples by QDs. All optical measurements were performed at room temperature and under ambient conditions.

2.4 Plackett–Burman design (PBD)

Two level PBD was carried out to distinguish the significant and insignificant variables and assess their mutual interaction as well as their effect on extraction efficacy. Minitab® 16.1.1 (Minitab Inc., State College, PA, USA) and STATISTICA 8 were used in this research model (design of experiments) which composed only on 24 experiments which is much lower than 64 (2⁶) experiments for full factorial design. Pareto chart was plotted and confirmed the significance as well their effect on the efficiency of the extraction technique. Significant variables were identified as their bars in the said chart were passing beyond the vertical line (t_critical value at p ≤ 0.05) (Dilshad et al., 2021).

2.5 Central composite design (CCD)

Once the significant and insignificant variables were identified, the significant variables were considered for further optimization. Feasible values were assigned to the insignificant variables (assessed earlier through PBD) for further experiments. Significant variables were subjected to further optimization. In this multivariate optimization step, a central 2³⁺ star orthogonal composite design based on 26 experiments was performed to evaluate combined effect of these significant variables against % recoveries of analyte.

2.6 Removal of Pb(II) in batch experiments

The removal of Pb(II) from the water samples using QDs was optimized under experimental parameters such as time, pH, temperature, adsorbent mass, and amount of Pb(II) for by the help of AAS, S3. In typical removal experiments, 0.01 g QDs were added into 0.001 g Pb(II) (or other subjected concentration) water sample and mixture was adjusted to optimized condition (see ST2) to get stirred for 24 h at ambient temperature. After centrifuging the suspensions, the Pb content of the supernatant was determined by the AAS and % removal was calculated using the following formula. $$\%Removal=\frac{C_i-C_e}{C_i}\times 100$$ where, C_i and C_e are the initial and remaining concentration (ppm) of Pb(II) in water sample, respectively (Nawaz et al., 2020). Furthermore, AAS techniques was also used to quantify the presence and recovery of Pb(II) by QDs in real water samples by following same procedure with help of given formula. $$\%Recovery=\frac{\mathit{Extracted}}{\mathit{Added}}\times 100$$ $$\%Recovery=\frac{\mathit{SD}}{\mathit{Mean}}\times 100$$

Moreover, averaged values from three runs of each experiment were presented.

3.1 Structural characterization of CdZnS/ZnS QDs

The SEM images of surface modified CdZnS/ZnS QDs capped with 2-MAA and 3-MPA have signifies that core/shell QDs are crystalline in nature and their size in the range of quantum confinement effect. Fig. 1 (a, b) & (c, d) taken at low and high magnification levels showed the formation of variable nano-size QDs with strong compactness. The analysis of EDS Fig. 2(a, b) & (c, d) has also proved that prepared QDs contained Cd, Zn, S, O and C. The powder XRD spectrum of CdZnS/ZnS with 3-MPA and 2-MAA (Fig. 3) displayed eight peaks at 2θ = 27.7°, 29.1°, 32.4°,33.9°, 36.9°, 39.9°, 44.8°, 52.3° which are indexed to the (hkl) planes: (1 0 0), (0 0 2), (1 0 1), (1 0 2), (1 1 0), (1 0 3), (2 0 0), (0 0 4). These results verify that core/shell QDs have microscopic particle size and have nanocrystalline nature. Furthermore, the array of XRD peaks is present at those pure CdS and ZnS merge states, corresponding to the previously reported suppositions (Liu et al., 2013; Kumari and Kar, 2020). Debye-Scherer equation was also practiced for analysis of synthesized CdZnS/ZnS QDs crystallite size in which D = 0.94λ/β cosθ: where θ is the Bragg angle, β is full circumference at half maximum and λ is wavelength of the worn X-ray (1.5406 Å). The average crystallite size (D) from the planes of first peak is 16 nm.

Close

Fig. 1

SEM Image of (a) CdZnS/ZnS@ 2-MAA QDs (using ZnCl₂ as salt), (b) CdZnS/ZnS@ 2-MAA QDs (using ZnNO₃·6H₂O as salt) & (c) CdZnS/ZnS@ 3-MPA QDs (ZnCl₂ salt) (d) CdZnS/ZnS@ 3- MPA QDs (ZnNO₃·6H₂O salt).

Export to PPT

Close

Fig. 2

EDS Image of (a) CdZnS/ZnS@ 2-MAA QDs (using ZnCl₂ as salt), (b) CdZnS/ZnS@ 2-MAA QDs (using ZnNO₃·6H₂O as salt) (c) CdZnS/ZnS@ 3-MPA QDs (ZnCl₂ salt) (d) CdZnS/ZnS@ 3- MPA QDs (ZnNO₃·6H₂O salt).

Export to PPT

Close

Fig. 3

XRD motifs of CdZnS/ZnS QDs processed via divergent capping agents.

Export to PPT

3.2 Optimization of Pb(II) removal by multivariate study

Plackett–Burman design is known to be an efficient design, widely utilized as an optimization tool in method development. In contrast to traditional univariate optimization where only one variable is studied at one time, this design is more effective, prompt and easier design. Different experimental variable that could possibly affect the efficacy of the proposed extraction technique were screened initially through PBD followed by CCD for optimization. These variables were pH, temperature, adsorbent amount, dilution, and adsorption/desorption time. Levels (higher and lower) of these variables along with their optimum values are shown in Table 1. Plackett–Burman matrix design with 24 experiment and the recoveries obtained are presented in Table 2a.

3.2.1

3.2.1 Central composite design (CCD)

Significant variables, i.e., pH, temperature, adsorbent amount, and adsorption/desorption time were further optimized through CCD. Any variation in the value of significant variables (from higher level to lower level and vice versa) influenced the % recoveries. This variation in % recoveries due to varying the levels of significant variable is depicted in Table 2b, and the same trend has been portrayed in the form of Pareto chart of the standardized effect (Fig. 4). Any variable bearing a t_critical value greater than 1.350 (at the 95.0 % confidence interval) is termed as significant variable. The results obtained with CCD for significant variables are present in the Table 2b. The % recoveries of Pb(II) were found to be 1.8–99.1 %. At optimum values of these variables, maximum recovery of Pb(II) was achieved as shown in experiment no. 1 & 26. On the other hand, any deviation from the optimum value of any variable resulted in decrease of Pb(II) recovery (experiments no. 18–25 of Table 2b).

Table 2b Central orthogonal composite design for the set of factors A, B, C and D (n = 6).

S. No.	A	B	C	D	% R
1.	aAo	aBo	aCo	aDo	99.1 ±
2.	−	−	−	−	37.4 ±
3.	+	−	−	−	36.4 ±
4.	−	+	−	−	25.3 ±
5.	+	+	−	−	14.0 ±
6.	−	−	+	−	45.8 ±
7.	+	−	+	−	78.4 ±
8.	−	+	+	−	12.4 ±
9.	+	+	+	−	12.4 ±
10.	−	−	−	+	35.3 ±
11.	+	−	−	+	24.7 ±
12.	−	+	−	+	25.1 ±
13.	+	+	−	+	26.0 ±
14.	−	−	+	+	27.3 ±
15.	+	−	+	+	45.5 ±
16.	−	+	+	+	14.7 ±
17.	+	+	+	+	25.1 ±
18.	−A1	aBo	aCo	aDo	2.3 ±
19.	+A2	aBo	aCo	aDo	13.2 ±
20.	aAo	-B1	aCo	aDo	9.8 ±
21.	aAo	+B2	aCo	aDo	1.8 ±
22.	aAo	aBo	−C1	aDo	11.3 ±
23.	aAo	aBo	+C2	aDo	9.1 ±
24.	aAo	aBo	aCo	−D1	4.6 ±
25.	aAo	aBo	aCo	+D2	7.3 ±
26.	aAo	aBo	aCo	aDo	97.0 ±

-A¹ = -3, +A² = 17, _aA^o = 7, -B¹ = -20, +B² = 140, _aB^o = 60, -C¹ = -0.0025, +C² = 0.0275, _aC^o = 0.0125, -D¹ = -5, +D² = 35, _aD^o = 15.

Close

Fig. 4

Pareto chart of the effects.

Export to PPT

Results obtained with ANOVA are presented as S3, which were utilized in the order to validate the observed quadratic model at 95 % confidence level. To assess the significance of the proposed model for the efficiency of Pb(II) removal; a comparison of F-values was made, where F_calculated (2.65) was found greater than F_critical (2.42). Beside F-values, the same scenario was observed with p-value which was <0.0001 (al Sadat Shafiof and Nezamzadeh-Ejhieh, 2020). Similarly, the lack-of-fit p-value 0.606 > 0.05 and F_calculated = 0.84 < F_critical = 2.93 confirmed that it is insignificant, which means that this model can the fit the data appropriately.

The Fig. 5(a-d) showed the predicted response surface data obtained for various variables A/B, B/C, A/C and C/D against % recoveries of Pb(II). Once the 3D plots of response surface were plotted, quadratic equation for each pair of variables was obtained. The obtained quadratic equations for all these pairs of variables are given (Eqs. (1.1)–(1.4)).

(1.1)

$$\%Recovery\frac{A}{B}=12.27+4.10x+0.55y-{0.29x}^2-1.19\times {10}^{-2}xy-4.60\times {10}^{-3}{y}^2$$

(1.2)

$$\%Recovery\frac{B}{C}=12.27+4.68x+1.47\times {10}^{-3}y-{0.29x}^2+1.40\times {10}^{-2}xy-0.61y^2$$

(1.3)

$$\%Recovery\frac{A}{C}=-2.95+0.60x+1.74\times {10}^{-2}y-{4.62\times {10}^{-3}x}^2-8.75\times {10}^{-3}xy-1.02y^2$$

(1.4)

$$\%Recovery\frac{C}{D}=-2.70+1.07x+2.24y-{31.41x}^2-28xy-8.71\times {10}^{-3}{y}^2$$

Close

Fig. 5

3D Surface plot for the effect of significant variables on % recovery of analyte (n = 6); (a) pH vs temperature, (b) temperature vs adsorbent amount, (c) pH vs adsorbent amount, (d) adsorbent amount vs time.

Export to PPT

Solution pH can greatly influence the adsorption of any analyte as it affects the surface charge of the adsorbent by distressing the protonation/deprotonation of the functional groups. To inspect the effects of pH on the extraction of the Pb(II), several pH solutions with same amount of CdZnS/ZnS@2-MAA (0.01 g) and same concentration of Pb(II) (100 ppb, 100 mL) were prepared. The pH was controlled with the backing of 1 M HCl and NaOH solution.

No removal of Pb(II) from water sample in strong acidic medium (pH < 2) was observed by the CdZnS/ZnS@2-MAA because the surface charge become positive which result in electrostatic repulsion between Pb(II) and adsorbent. Zero charge pH (pH_pzc) is the pH value where the surface charge of the sorbent is zero, at this pH the surface functionalities (both acidic and basic) do not contribute to the overall pH value of the solution (Anari-Anaraki and Nezamzadeh-Ejhieh, 2015). Due to protonation and decline in the stability of metallic complex, lower adsorption of Pb(II) was observed at pH 4.0. The sorbent surface has a positive charge at pH lower than its pH_pzc, the reason is adsorption of H⁺, thus avoiding the positive charged Pb(II) (Nezamzadeh-Ejhieh and Zabihi-Mobarakeh 2014). The Fig. 5a, shows maximum recovery can be obtained at pH 6.9. Equations (1.1) & (1.2) also confirmed the same pH value.

Since, temperature plays notable role in the extraction of an analyte, thus seven standard sample solutions having CdZnS/ZnS@2-MAA (0.01 g) and Pb(II) (100 ppb) were taken in separate beakers having thermometer with the temperature discreteness of 10 °C. The solutions were heated on the hotplate till temperature reached to require level with constant stirring. The adsorption of Pb(II) was secured at temperature 30 °C and further raised in temperature displayed the decreased in adsorption capacity. Three-dimensional surface diagrams (Fig. 5a-b, and Eqs. (1.1) & (1.2)) revealed that 61.1 °C temperature has the ability to yield 100 % recovery, so this amount was selected as optimum value.

Electrostatic adsorption, coordination adsorption and ion-exchange are primarily incorporated as the adsorptive modes for expulsion of metal ions (Yu et al., 2017). The electrostatic attraction in the adsorption process of Pb(II) with CdZnS/ZnS@ 2-MAA was studied by preparing four standard sample solutions in a 200 mL glass beakers marked as beaker-I, beaker-II, beaker-III and beaker-IV having 100 ppb of Pb(II) in 100 mL de-ionize water and different amount of CdZnS/ZnS@ 2-MAA i.e., as 0.005 g in beaker-I, 0.01 g in beaker-II, 0.015 g in beaker-III and 0.02 g in beaker-IV. It was observed that the Pb(II) adsorption increased promptly when the amount of QDs was increased (0.005–0.01 g). However, decrease in Pb(II) adsorption was noticed when more adsorbent was added (0.01–0.015 g). To calculate the maximum adsorption/desorption time of Pb(II) on CdZnS/ZnS@2-MAA, five standard solutions of Pb(II) in the deionized water were made. The time of standard and sample solutions was synchronized to 5.0 min. Furthermore, the Pb(II) extraction boosted when incubation time increased from 5 to 15 min which start to decline as time passes. The calculation from Eq. (1.3 & 1.4) indicated that the recoveries were quantitative at amount of adsorbent 0.013 g with 15.3 min of time for adsorption and desorption (Fig. 5c-d).

3.3 Adsorption isotherms

The removal of Pb(II) from water using CdZnS/ZnS@2-MAA QDs on the AAS was studied under experimental parameters such as pH, time, temperature, adsorbent mass, and amount of Pb(II) for optimization. Langmuir adsorption model was applied with linear form of Langmuir’s equation to predict the adsorption of analyte at definite localized homogenous sites of adsorbent (Lagergren, 1898).

Close

Fig. 6

Langmuir adsorption isotherm.

Export to PPT

Alternatively, Freundlich adsorption isotherm predicted the occurrence of adsorption on a heterogeneous surface with stronger binding sites occupied first (Naghash and Nezamzadeh-Ejhieh 2015). The adsorption model was applied by using linear form of Freundlich equation (Freundlich, 1907).

Close

Fig. 7

Freundlich adsorption isotherm.

Export to PPT

3.4 Thermodynamic parameter

The thermodynamic studies were conducted between 20 and 100 °C. Thermodynamic factors such as entropy (ΔS^°), enthalpy (ΔH^°), and Gibb’s free energy (ΔG^°) are useful to define the nature of the adsorption. Therefore, Van’t Hoff reaction isotherm was used to examine these parameters (al Sadat Shafiof and Nezamzadeh-Ejhieh, 2020).

Close

Fig. 8

Van’t Hoff plot.

Export to PPT

The negative free energy confirms that the complex formation in between CdZnS/ZnS@2-MAA and Pb(II) is temperature dependent spontaneous process. These findings also indicated that the sorption is temperature sensitive since ΔG becomes less negative at elevated temperature (Nezamzadeh-Ejhieh and Kabiri-Samani, 2013). During complex formation the binding reaction was drive by negative enthalpy (−59.26 KJ/molK) which suggested that during adsorption exothermic reaction might have happened. While, negative enthalpy also means breakage of Pb(II)–H₂O bond in aqueous medium is favorable for the formation of the pb(II)-QDs complex. Whereas negative entropy (−0.16 KJ/molK) suggests that complex formation contributes to decrease in randomness at internal structure of the adsorbent during adsorption with low amount of ion replacement. Thus complex formation due to the chemisorptions remains dominant process in Pb(II) removal (Table 4). Similar mechanisms were also reported for the CIZS-QDs and Ag₂S-QDs for the adsorption of Pb(II) and Cu(II) from aqueous mediums (Jiang et al., 2019; Han et al., 2022).

3.5 Kinetics of adsorption

The pseudo-first order model is useful for the systems where driving force model is linear and adsorption rate is associated to alteration in the saturated concentration and number of active adsorbent sites, indicating that the whole adsorption rate is proportional to the preliminary strength of process. The pseudo-second order model, instead suggests that the speed of mechanism is controlled by chemical adsorption, where universal adsorption rate is square of the preliminary strength of process and in it the molecules join the adsorbent surface (Khan et al., 2019). Therefore, adsorption process of Pb(II) on QDs was investigated by using these two kinetic models. The equations used for Lagergren’s Pseudo first order and Ho Pseudo second order kinetics (Ho et al., 1999) are written as eq. (7) and eq.8. For first order modeling, graph was plotted between log (qe-qt) and t. While, for second order modeling graph was plotted between t/q_t and t.

Close

Fig. 9

(a) Pseudo-first order (b) Pseudo-second order.

Export to PPT

3.6 Pb(II) adsorption in water samples

The detection and recovery of Pb(II) in tap and ground water; samples were collected from FUUAST Gulshan-e-Iqbal Campus, Karachi and Malir Cantt (Karachi) and studied under optimized conditions i.e., 0.01 g CdZnS/ZnS@2-MAA dispersed in100 mL of water sample having pH 4.0 stirred for 20 min at 30 °C. Initially unknown concentration of Pb(II) in real water samples were detected in tap water (0.964 ng.mL⁻¹ with 96.4 % recovery) and ground water (0.948 ngmL⁻¹ with 94.8 % recovery) with small %RSD values 1.55 and 2.11, respectively. In second step 10 ng.mL⁻¹ of Pb(II) was added in both water samples. The result demonstrated that CdZnS/ZnS@2-MAA were able to remove 98.9 % Pb(II) from tap water samples with %RSD 1.31. While, 99.3 % Pb(II) was removed from ground water samples with 6.95 % of RSD (Table 6). Moreover, limit of detection and quantification (LOD and LOQ) of the CdZnS/ZnS@2-MAA QDs was also determined. The developed experimental process showed good conformity due to the low LOD (0.1 ng mL⁻¹) and LOQ (0.90 ng mL⁻¹) in comparison of reported QDs utilized for Pb(II) removal from aqueous medium (Table 7) and maximum permissible limit set by WHO for Pb(II) in drinking water (10 µg/L) (Naushad, 2014).

3.7 Regeneration analyses of CdZnS/ZnS-2-MAA QDs

The high adsorption capacity, effortless separation with low outlay and regeneration makes nano adsorbent convenient economically and technologically (Sari et al., 2007). In this research CdZnS/ZnS-2-MAA QDs were recycled with 99 % ethanol solution for further used. Moreover, recovered QDs have validated their functionality equal to the pure QDs when extraction of Pb(II) from solution was repeated.

3.8 Removal mechanism of Pb(II) by CdZnS/ZnS@2-MAA QDs

In CdZnS/ZnS@2-MAA, 2-MAA was utilized as capping agent which functions as stabilizing agent and brings electrostatic attraction between ZnS (shell) and Pb(II) in water samples. Overall Pb(II)adsorption is found to be favorable on the QDs heterogeneous surface as Freundlich models predicted. The findings indicated that sorption is temperature sensitive since ΔG becomes less negative at elevated temperature. The negative ΔG also means that binding energy of QDs is stronger than Pb-solvent and this difference in energetic potentials of the system's components is what drives the redistribution of Pb(II) in the system and spontaneous sorption on QDs. Additionally, negative ΔH and ΔS showed that sorption is an exothermic process and complex formation-induced chemisorption as the primary mechanism for Pb(II) removal. The system follows the pseudo-second order rates in which rate-limiting step is a chemical reaction that was impacted by Pb(II) ion intraparticle/pore diffusion with QDs.