Lately in studies nanoparticles whose surfaces were modified and in the way the adsorption capacity was increased are used. For these studies, SPFe3O4-NP were coated with iron phosphate (Zhang et al. 2018), Fe3O4 nanoparticles were encapsulated into calcium alginate coated chitosan hydrochloride hydrogel beads (Yi et al. 2018), graphene [email protected]3O4 magnetic beads were compounded with carboxymethyl chitosan and sodium alginate (Wu et al. 2019), Fe3O4/layered double hydroxides were trapped into calcium alginate (Sun et al. 2018), Fe3O4 nanoparticles were coated with sodium alginate (Serunting et al. 2018), the mixture of titanium oxide and maghemite nanoparticles immersed in polyvinyl alcohol-alginate beads (Majidnia and Idris 2015) can be given.
El-Shamy et al. (2019) used magnetite-alginate nanoparticles for Co ions adsorption from wastewaters. Co ions concentrations were determined by AAS. The adsorption efficiency and absorption capacity were calculated as 95.01% and 33.557 mg g− 1, respectively.
In a study bionanocomposite were produced and used for copper and nickel removal. The bionanocomposite adsorption capacity was evaluated 88.49 mg g− 1 for copper and 86.95 mg g− 1 for nickel (Foroutan et al. 2018).
Hosseinzadeh et al. (2016) obtained nanocrystalline magnetite from hematite and used for Cd2+ removal in aqueous solutions. In order for cadmium not to precipitate as hydroxides, it has not been studied greater than 7 pHs’. It was stated that pH is the most effective parameter among the experimental parameters for metal removal, such as initial concentration, contact time and temperature.
In a batch system, biochar produced from greenhouse crop residue was used as adsorbent for Co removal. For adsorbent activation different materials were applied. The adsoption capacity was calculated as 30.98 mg g− 1 (Iáñez-Rodríguez et al. 2020).
Kataria and Garg (2018) synthesized two different iron oxide nanoparticles and used these nanoparticles metal removal from ground water by batch technique. They studies effects of pH, adsorbent dose, temperature, contact time on adsorption and reusability of adsorbents. Adsorption capacities of nanoparticles were found 51 mg g− 1 and 63.3 mg g− 1.
In a study magnetite − manganese oxide nanoparticles immobilized by alginate. Optimum conditions for Cd2+ removal were selected 6 for pH, 30°C for temperature, 120 min for contact time and 10 mg for the adsorbent amount (Kumar et al. 2021).
In this study nanoparticles coated with alginate and the adsorbent adsorption capacity found high for Cd2+ (31.93 mg g− 1) and Co2+ (58.8 mg g− 1) ions. Because alginate molecules are occurred from carboxyl, carbonyl, and hydroxyl other groups so the obtained nanoparticles can easily adsorb the Cd2+ and Co2+ (Cheng et al. 2019).
Characterization of SPFe3O4/alg-NCM
Spectra of the pretreated alginate, SPFe3O4/alg-NCM and Cd2+/Co2+ ions loaded SPFe3O4/alg-NCM were compared by using FTIR-ATR. FTIR-ATR analyses were used to characterize the changes of functional groups before and after Cd2+/Co2+ biosorption. The same characteristic bands were seen in the FTIR spectra (Fig. 1) for the alginate coated beads, SPFe3O4/alg-NCM and Cd2+/Co2+ ions loaded SPFe3O4/alg-NCM: a wide band between 3400 and 3000 cm− 1 corresponds to the OH stretch vibration. Weak peaks at 2921, 2936, 2936 and 2923 cm− 1 in the spectrum of alginate, SPFe3O4/alg-NCM and SPFe3O4/alg-NCM loaded with Cd2+/Co2+ ions, respectively, and C–H belongs to asymmetric stretching vibration, alginate macromolecule and “egg-box” structures formed by calcium ion limit CH stretching (He et al. 2012). Metal complexes occur in different types; it can coordinate in an ionic or uncoordinated form, via non-identical coordination, via bidentate chelation coordination, and through bicognate bridging coordination. In studies of acetate carboxyl groups, there is a clear relationship between the νasym(COO−) ve νsym(COO−) bands of the FTIR spectrum and the metal coordination type, so the separation of bands (i.e. Δν = COasym-COsym) is also indicative of a specific carboxylate structure. In this study, COO-stretching (asymmetric) spectrum peaks were observed at 1593, 1592, 1601 and 1618 cm− 1, respectively. In addition, COO-stretching (symmetrical) spectrum peaks are also seen at 1404, 1416, 1346 and 1345 cm− 1. Therefore, Δν(COO−)complex values were found to be 189, 176, 255 and 273, respectively, and SPFe3O4/alg-NCM is associated with bidentate bridging coordination and SPFe3O4/alg-NCM loaded with Cd2+/Co2+ ions is associated with undefined coordination. Metal–carboxylate interactions in metal–alginate complexes studied with FTIR spectroscopy. Differently, another important band was detected at 541 cm− 1 which corresponds to the stress vibration of Fe-O bonds in SPFe3O4-NP of SPFe3O4/alg-NCM (Huber 2005). In Cd2+/Co2+ ions loaded SPFe3O4/alg-NCM, this peak shift is seen in 561 and 557cm− 1 and there is no peak in pure alginate.
Thermogravimetric analysis of alginate, SPFe3O4/alg-NCM and Cd2+/Co2+ loaded SPFe3O4/alg-NCM were recorded in the temperature range of 30°C − 600°C (Fig. 2a). Alginate, SPFe3O4/alg-NCM, and Cd2+/Co2+ ions loaded SPFe3O4/alg-NCM exhibited similar behavior, showing the formation of magnetic particles in alginate did not change the original structure of alginate and interactions between alginate chains and magnetite or Cd2+/Co2+ ions are not strong enough to change their thermal stability. Mass losses in alginate, SPFe3O4/alg-NCM and Cd2+/ Co2+ ions loaded SPFe3O4/alg-NCM were observed in three stages.
(i) mass losses of 25% from 35°C to 300°C based upon the release of water and CO2 molecules; (ii) losses from 300°C to 450°C were 80%, 40%, 35% and 39% were attributed to alginate decomposition products, and (iii) losses were 88%, 44%, 44% and 45%, respectively, up to 600°C and decomposition was complete. TGA analyses showed that Cd2+/Co2 + ions loaded SPFe3O4/alg-NCM have a high thermal stability. Also based on TGA-DTA analysis it is very evident that no significant change was observed in the decomposition temperature of SPFe3O4/alg-NCM and Cd2+/Co2+ ions loaded SPFe3O4/alg-NCM.
The peaks obtained from the DTA plot of alginate, SPFe3O4/alg NCM and Cd2+/Co2+ loaded SPFe3O4/alg NCM (Fig. 2b) allowed us to obtain more detailed information about thermal degradation. The decomposition of alginate is endothermic, which appears to provide complete degradation at 435°C, and the SPFe3O4/alg-NCM degradation is exothermic and appears to provide complete degradation at 300°C, which is evidence of the presence of Fe3O4 in it. Cd2+ loaded SPFe3O4/alg-NCM degradation is also exothermic and it can be seen from the peaks at 250°C and 315°C, which are determined to be in two stages, in the same way, the degradation of Cd2+ loaded SPFe3O4/alg-NCM is also exothermic, and it is determined to occur in two stages at 250°C and 290°C it can be seen from the peaks at 250°C and 290°C. These results show that Fe3O4 is loaded and Cd2+/Co2+ were absorbed.
From UV-Vis spectrophotometer analysis it was observed that the color will change from yellow to black immediately when the biomass extract was added to the FeCl3 and FeCl2 solutions and adjust the pH. In addition, the UV-Vis spectra of SPFe3O4-NP have shown in Fig. 3. In this figure it can be seen a broad absorption around 550 to 750 nm after reaction, indicating the production of SPFe3O4-NP. Both color change and UV-Vis absorption around 400–580 showed that SPFe3O4-NPs were successfully synthesized (Huber 2005). Formation of iron oxide nanoparticles is proven with peaks at 298–301 nm (Pattanayak and Nayak 2013).
Samples were analyzed at positions between the angle of 0°-80° 2θ. The 2θ characteristic reflection peaks in the XRD spectrum (Fig. 4) were observed at 29°, 31°, 35°, 38°, 47°, 57° and 62°. From XRD patterns it can be seen sharp peaks. These sharp peaks indicate that the Cd2+ and Co2+ loaded SPFe3O4/alg-NCM have high crystallinity structure. Peaks at 35.34°, 35.43° and 35.57° shows us the presence of iron. The most severe peaks were observed to be 29°.
As it can be seen EDS graph in Fig. 5, the amounts of O, Fe, Na are higher than the other elements. The excess of O and Fe is due to Fe3O4, and the excess of Na is due to sodium alginate. It is also seen that Cd2+ and Co2+ are loaded into the SPFe3O4/alg-NCM.
Shape and size distribution of the nanoparticles were investigated using SEM analysis for distinguishing morphological properties and presented in Fig. 6. SEM visual micrograph of SPFe3O4/alg-NCM represents the nanoparticles spherical and uniform distribution of on the alginate surface. As can be observed that after interaction by Cd2+ and Co2+ ions, the nanoparticles still retain their spherical morphology.
RSM statistical analysis and model fitting
The CCD combined with RSM was selected and performed to optimize of crucial variables and explain the response surface nature. Regression coefficients and ANOVA results reveal that quadratic model (p < 0.0001) nature very important both for Cd2+ and Co2+ ions. While Table 2 presents CCD programme, Table 3 presents ANOVA test results that contain suggested second-order equation coefficients and model terms for Cd2+ and Co2+ ions adsorption on SPFe3O4/alg-NCM. Suggested model variables were tested using Fisher's test. Because with this test, it shows that larger “F values” and smaller “p values” are more meaningful than the proposed model terms. In this study; while p value of less than 0.05 indicated significance at the 95% confidence level, F value of 95.55 represented that the model is important. The "Probe > F" values are a measure of whether the model terms are statistically significant. If the "Probe > F" value is less than 0.05, it means that the term is statistically significant for the model, and a value greater than 0.05 indicates that that term is statistically insignificant for the model. In this case, both Cd2+ and Co2+ all model terms including X1, X2, X3, X1X2, X1X3, X2X3, X12, X22 and X32 values are significant model terms (Tables 2 and 3).
Table 2
Analysis of variance (ANOVA) of the Cd2+ quadratic model
|
Sum of
|
|
Mean
|
F
|
p-value
|
|
Source
|
Squares
|
df
|
Square
|
Value
|
Prob > F
|
|
Model
|
997.50
|
9
|
110.83
|
97.03
|
< 0.0001
|
Significant
|
X1-pH
|
230.28
|
1
|
230.28
|
201.60
|
< 0.0001
|
|
X2-Flow rate
|
6.13
|
1
|
6.13
|
5.36
|
< 0.0001
|
|
X3-Cd initial concent.
|
358.16
|
1
|
358.16
|
313.55
|
< 0.0001
|
|
X1 X2
|
21.45
|
1
|
21.45
|
18.78
|
< 0.0001
|
|
X1 X3
|
102.96
|
1
|
102.96
|
90.14
|
< 0.0001
|
|
X2 X3
|
11.28
|
1
|
11.28
|
9.88
|
< 0.0001
|
|
X12
|
124.47
|
1
|
124.47
|
108.97
|
< 0.0001
|
|
X32
|
0.57
|
1
|
0.57
|
0.50
|
< 0.0001
|
|
X32
|
181.60
|
1
|
181.60
|
158.98
|
< 0.0001
|
|
Residual
|
11.42
|
10
|
1.14
|
|
|
|
Lack of Fit
|
6.46
|
5
|
1.29
|
1.30
|
0.3896
|
Not significant
|
Pure Error
|
4.96
|
5
|
0.99
|
|
|
|
Cor Total
|
1008.92
|
19
|
|
|
|
|
R2
|
0.9887
|
|
|
|
|
|
R2Adj
|
0.9785
|
|
|
|
|
|
R2Pred
|
0.9417
|
|
|
|
|
|
Adeq Precision
|
30.037
|
|
|
|
|
|
*p < 0.01 highly significant; 0.01 < p < 0.05 significant; p > 0.05 not significant. |
Table 3
Analysis of variance (ANOVA) of the Co2+ quadratic model
|
Sum of
|
|
Mean
|
F
|
p-value
|
|
Source
|
Squares
|
df
|
Square
|
Value
|
Prob > F
|
|
Model
|
5130.88
|
9
|
570.10
|
1476.85
|
< 0.0001
|
Significant
|
X1-pH
|
1061.13
|
1
|
1061.13
|
2748.88
|
< 0.0001
|
|
X2-Flow rate
|
117.18
|
1
|
117.18
|
303.56
|
< 0.0001
|
|
X3-Co initial concentc
|
1650.39
|
1
|
1650.39
|
4275.37
|
< 0.0001
|
|
X1 X2
|
43.24
|
1
|
43.24
|
112.03
|
< 0.0001
|
|
X1 X3
|
239.80
|
1
|
239.80
|
621.22
|
< 0.0001
|
|
X2 X3
|
33.21
|
1
|
33.21
|
86.03
|
< 0.0001
|
|
X12
|
324.00
|
1
|
324.00
|
839.34
|
< 0.0001
|
|
X32
|
100.17
|
1
|
100.17
|
259.50
|
< 0.0001
|
|
X32
|
1906.79
|
1
|
1906.79
|
4939.59
|
< 0.0001
|
|
Residual
|
3.86
|
10
|
0.39
|
|
|
|
Lack of Fit
|
2.27
|
5
|
0.45
|
1.42
|
0.3548
|
Not significant
|
Pure Error
|
1.59
|
5
|
0.32
|
|
|
|
Cor Total
|
5134.74
|
19
|
|
|
|
|
R2
|
0.9992
|
|
|
|
|
|
R2Adj
|
0.9986
|
|
|
|
|
|
R2Pred
|
0.9963
|
|
|
|
|
|
Adeq Precision
|
129.914
|
|
|
|
|
|
*p < 0.01 highly significant; 0.01 < p < 0.05 significant; p > 0.05 not significant. |
Furthermore, the data variation around the model that fits the experimental data is expressed as the model's lack of fit (LOF). The LOF p-values measure the fitness of the model and imply the LOF is not significant relating to the pure error. The LOF value, which is conclusive evidence for the adequacy of the model fit, provides information about the model fit without the effects of additional higher order terms. Based on Tables 2 and 3 LOF p-values of Cd2+ and Co2+ are 0.3896 and 0.3548, respectively. These p-values of LOF confirm the suitable applicability for well-fitting of the reaction. Also, the number of experiments performed was found to be sufficient, for determining the independent variables' effects for Cd2+ and Co2+ ions adsorption on SPFe3O4/alg-NCM. Polynomial models validity was assessed by determining the determination coefficients (R2) and determination adjusted coefficients (R2adj). Because R2 and R2adj are defined the percentage of variability in the response. Moreover, it is desirable to have a proposed model high R2 along with for validity of the model, value of R2 should be greater than 0.75 (Ince and Kaplan Ince 2017). Obtained R2 and R2adj values for the previously mentioned Cd2+ and Co2+ models were satisfactory. Because R2 of the models were obtained for Cd2+ as 0.9887 for Co2+ as 0.9992, it can be said that 98.87% and 99.92% of the model-predicted values matched the experimental adsorbed for Cd2+ and Co2+ values on SPFe3O4/alg-NCM. In addition, adequate precision (AP) measures the signal to noise ratio (S/N) and this ratio greater than 4 is desirable. The S/N should be greater than four and it is measure by AP. These values obtained as 30.037 for Cd model and 129.914 for Co model. It can be stated that is indicating an adequate signal both model. In lights of preliminary experiment studies, three critical parameters affecting Cd2+ and Co2+ adsorption were selected as independent variables. On the other hand, Cd2+ and Co2+ ions adsorption on SPFe3O4/alg-NCM (Y) were considered as the dependent variable. In addition, to express the relationship between independent variables and responses, experimental data were fitted to two second-order polynomial mathematical equation (Eqs. 2 and 3) presented as below for Cd2+ and Co2+:
$$Y \left(mg Cd /g {Fe}_{3}{O}_{4}/alg NCM \right)=+45.18-7.57{X}_{1}+3.51{X}_{2}-4.20{X}_{3}-0.41{X}_{1}{X}_{2}+0.36{X}_{1}{X}_{3}-0.12{X}_{2}{X}_{3}-0.56{{X}_{1}^{2}}^{ }+0.04{X}_{2}^{2}+0.11{X}_{3}^{2}$$
2
$$Y \left(mg Co /g {Fe}_{3}{O}_{4}/alg NCM \right)=+105.78-9.63{X}_{1}-4.78{X}_{2}-4.12{X}_{3}-0.58{X}_{1}{X}_{2}+0.18{X}_{1}{X}_{3}+0.07{X}_{2}{X}_{3}+0.90{{X}_{1}^{2}}^{ }+0.50{X}_{2}^{2}+0.04{X}_{3}^{2}$$
3
For determining the optimum point and achieving the highest treatment performance the best method is layout of the surface plot in the Cd2+ and Co2+ ions adsorption on SPFe3O4/alg-NCM process. The affect of each parameter on Cd2+ and Co2+ ions adsorption on SPFe3O4/alg-NCM and their interaction were given in Figs. 7 and 8. Three-dimensional (3D) response surface graphs are useful to determine response maximum, middle and minimum points and they were obtained from quadratic model. Figures 7a and 8a present 3D response surface plots of the influence of pH-flow rate on the adsorption efficiency of Cd2+ and Co2+ ions on SPFe3O4/alg-NCM. Both Cd2+ and Co2+ ions adsorption amount increased when pH increases from 3 to 7 (p < 0.01), after which an increase in metal uptake were observed. Although the decrease in the flow rate has a serious effect on the adsorption of Cd2+ ions, however, it was observed that Co2+ ions adsorption did not show a statistically significant increase or decrease tendency with flow rate change.
A magnetic Fe3O4/graphene oxide nanocomposite material was synthzed and used for Pb2+ removal from water samples by Thy et al. (2020). The optimum Pb2+ ions adsorption onto nanocomposite material conditions were performed by RSM combined with Box-Behnken design approach. In mentioned study, the critical variables such as pH and initial concentration on the adsorption the interactive effects of these parameters were examinated. For determining the model suitability and reliability statistical parameters from quadratic model were analysed using ANOVA. By using the comparison and evaluation with the "p-values" the variables and their interactions effect on the responses were investigated. It was stated that pH "p value" is smaller than 0.0001, thus this factor is highly significant factor for Pb2+ removal from water.
On the other hand, it is clear that the amount of adsorbed both Cd2+ and Co2+ ions were increased when both Cd2+ and Co2+ ions initial concentration (p < 0.0001) were increased (Figs. 7b and 8b). Also, when both Cd2+ and Co2+ ions solutions' pH increases from 3 to 7 (p < 0.01) amount of adsorbed metals were increased. According to ANOVA table, it was confirmed that these two variables had a significant effect on both Cd2+ and Co2+ ions adsorption. Moreover, it was observed that these two variables interaction was statistically significant (p < 0.05).
An investigation was carried out by Rasoulzadeh et al. (2020) for the Pb2+ ions adsorption onto Fe3O4 nanoparticles and chitosan-coated Fe3O4 particles using RSM. According to applied approach model, the optimum conditions were found to be as 0.1 g L− 1 for Pb (II) concentration, 10.95 for pH and 5.5 mg L− 1 for adsorbent dosage. Also, maximum removal efficiency of Pb2+ ions was calculated as 93.6%. It was reported that Pb2+ ions initial concentration directly affects the uptake capacity of the Fe3O4 nanoparticles and chitosan-coated Fe3O4 particles.
Figures 7c and 8c represents combined effect of flow rate-initial concentration of Cd2+ and Co2+ ions removal at constant pH. While the adsorption for both Cd2+ and Co2+ ions increased with the increase in the initial concentration, it was observed that the change of flow rate had a partial effect on the metal adsorption on SPFe3O4/alg-NCM.
As shown in Figs. 7 and 8, factors including pH and initial concentration were found to be very critical with respect to their centre points. Also, it was used for a combined effect of all the factors on a process besides to analyze variation of the factors. Mentioned factors indicate that Cd2+ and Co2+ ions removal was highly affected by these variables.
All diagnostic plots for Cd2+ and Co2+ optimization process by using CCD approach were shown in Fig. 9. For comparing all factors effect on optimum conditions for Cd2+ and Co2+ ions adsorption on SPFe3O4/alg-NCM a perturbation plot was carried out (Fig. 10). All parameters especially initial concentrations and pH indicates that Cd2+ and Co2+ ions removal was highly affected by these variables. Additionally, to check the lambda (λ) value a Box-Cox plot was selected for two metals. For predicting any necessary transformation this plot is often used to enhance the significance of model. No transformation was needed, based on λ values obtained from the plots.
The efficiency of the regression model for both elements and the characteristic charts for evaluation are presented in Fig. 9. Although it is clearly observed that there is a high correlation between the estimated and experimental removal values, the residuals show a normal distribution and this distribution supports the model sufficiently in the real system.
The most important statistical metrics such as lowest p-value and LOF value and along with highest F-value and R2 were examinated to found maximum desirability function (DF) and select the optimal model. The models desirability values were obtained to be 0.98 and 0.75 for Cd2+ and Co2+, respectively (Fig. 11).
The optimum values of the variables used for optimization and the highest desirability values obtained depending on these values are presented in Fig. 12 for both Cd2+ and Co2+.
Desorption procedure and SPFe3O4/alg-NCM reusability studies
Reusability and regeneration, desorption studies were investigated for SPFe3O4/alg-NCM performans. For this reason, through conducting ten consecutive adsorption-desorption cycles were performed and result were given on Fig. 13. During desorption and reuse process, bio-nano materials can be damaged because of their sensitive structure. For clarifying the adsorption process nature, desorption and reuse studies are useful. In the desorption examination choosing proper eluent is very important. Therefore, various desorption reagents including H2SO4, HCl, HNO3, and CH3COOH desorption efficiency (Fig. 14) were investigated, and the best desorption eulent was obtained as 0.5 M HCl. Under optimal condition, several experimental studies for reusability were performed using HCl (0.5 M) to test SPFe3O4/alg-NCM, for understanding whether desorption process is damaging. Then, by using the best efficient eluent (0.5 M HCl) adsorbed elements onto nano-biocomposite separated from the column were desorbed and supernatant’s Cd2+ and Co2+ ions concentration were measured using ETAAS. Regenerated SPFe3O4/alg-NCM were washed at least five times using ultrapure water to reach neutral pH and it was stored for the next adsorption-desorption cycle. Reusability investigations proved that the SPFe3O4/alg-NCM exhibited excellent renewability and reusability up to at least ten cycles. Furthermore, for Cd2+ and Co2+ ions removal from industrial wastewater, it has been understood that it has significant potential in practical application. The obtained data are compatible with the data in the literature (Ciesielski et al. 2012; Davodi et al. 2020; Hu et al. 2020; Cheraghipour and Pakshir 2020).
Application to real samples
The developed method was carried out to industrial wastewater samples (IW) at optimum conditions (Table 4). Again, in these conditions standard addition method was applied to real samples. Cd2+ and Co2+ ion standards of 20 µg L− 1 and 40 µg L− 1 were added wastewater samples and the results were given in Table 4. For wastewater samples, contamination factors (CF) were calculated and CF was found in the range of 0.64–1.38 for Cd and 0.36–0.54 for Co. The level of metal pollution is expressed in 4 levels. For low contamination, CF is lower than 1; for moderate contamination, CF is in the range of 1 and 3; for considerable contamination, CF is in the range of and for very high contamination, CF is higher than 6 (Barakat et al. 2019). According to results generally wastewater samples CF is lower than 1.
Table 4
Cd2+ and Co2+ contents of industrial wastewater
Industrial wastewater (IW)
|
Cd2+ (µg L− 1)
|
Co2+ (µg L− 1)
|
SA
|
Found
|
Removal (%)
|
CF
|
Found
|
Removal (%)
|
CF
|
IW 1
|
0.0
|
6.6 ± 0.3
|
|
1.32
|
25 ± 2
|
|
0.5
|
20
|
27.1 ± 1.1
|
100
|
|
46 ± 3
|
100
|
|
40
|
47.3 ± 3.2
|
100
|
|
64 ± 4
|
100
|
|
IW 2
|
0.0
|
6.9 ± 0.2
|
|
1.38
|
24 ± 2
|
|
0.48
|
20
|
26.5 ± 2.2
|
99
|
|
45 ± 2
|
100
|
|
40
|
47.7 ± 3.1
|
100
|
|
63 ± 3
|
99
|
|
IW 3
|
0.0
|
5.4 ± 0.2
|
|
1.08
|
27 ± 3
|
|
0.54
|
20
|
26.4 ± 1.8
|
100
|
|
48 ± 4
|
100
|
|
40
|
45.5 ± 2.0
|
98
|
|
68 ± 3
|
100
|
|
IW 4
|
0.0
|
3.2 ± 0.2
|
|
0.64
|
18 ± 1
|
|
0.36
|
20
|
24.4 ± 1.5
|
100
|
|
36 ± 3
|
98
|
|
40
|
43.8 ± 3.3
|
100
|
|
56 ± 2
|
99
|
|
IW 5
|
0.0
|
5.6 ± 0.2
|
|
1.12
|
21 ± 1
|
|
0.42
|
20
|
27.0 ± 2.5
|
100
|
|
42 ± 3
|
100
|
|
40
|
46.2 ± 3.6
|
100
|
|
60 ± 5
|
98
|
|
IW 6
|
0.0
|
4.8 ± 0.1
|
|
0.96
|
25 ± 2
|
|
0.5
|
20
|
24.9 ± 2.2
|
100
|
|
46 ± 4
|
100
|
|
40
|
43.3 ± 1.9
|
99
|
|
66 ± 4
|
100
|
|
*SA: Standard Addition |