Study of the Extraction of Copper from Aqueous Solutions as Copper Oxyquinolinate with Benzene Using Experimental Design
Abstract Generated abstract
This study examines the extraction of copper from aqueous solutions into benzene as copper oxinate, using experimental design to evaluate the effects of extraction time, solution pH, and oxine concentration. A first factorial series identified the oxine concentration as the dominant factor and located a near-stationary region with extraction above 90 percent. A second, rotatable second-order design produced an adequate regression model for the response surface and was used to estimate optimal conditions. Within the investigated range, the optimum was found at about 77 seconds extraction time, pH 7.4, and an oxine concentration of 5.3 x 10^-3 mol/l, giving a calculated extraction factor of 98.6 percent and experimental values near 99 percent.
Full Text
STUDY OF THE EXTRACTION OF COPPER FROM AQUEOUS SOLUTIONS AS COPPER OXINATE WITH BENZENE BY THE METHOD OF EXPERIMENTAL DESIGN
Yu. A. Karbainov, N. N. Chernysheva
(Presented by the scientific seminar of the Department of Analytical Chemistry)
The following independent variables were considered: \(z_1\)—extraction time, min; \(z_2\)—pH of the solution; \(z_3\)—concentration of oxine in benzene, mol/l. The optimization parameter \(y\) was the copper extraction coefficient \(R\), related to the distribution coefficient \(E\) by the relation:
\[ R := \frac{E}{E+1}\cdot 100\%, \text{ where} \tag{1} \]
\[ E = \frac{[C_{\mathrm{Cu}}]_o}{[C_{\mathrm{Cu}}]_{\mathrm{aq}}}. \tag{2} \]
To prepare the initial solutions with a specified pH value, a universal buffer mixture [1] was used. After extraction, the organic phase was analyzed by the method of extraction AAS. Composition of the background solution: 3 ml of 0.2 M \(NH_4NO_3\) and 1 ml of 0.2 M \(NH_4SCN\) in 96% \(C_2H_5OH\).
In the first series of experiments, the zero point and the variation levels were chosen using the data obtained in [2]. The conditions, design matrix, and results obtained are presented in Tables 1 and 2. Each result is the average of 2 parallel determinations.
Since \(b_{ij} \gg S\{b\}\) and \(b_{ij}\) is of the same order as \(b_i\), except for \(b_3\), the value of which is significantly greater than all the other coefficients, indicating the strong influence of the oxine concentration on the copper extraction process, it may be asserted that the “nearly stationary region” has been reached. This is confirmed by the large values of the extraction factor in the last 4 experiments (\(y > 90\)) (Table 2).
In the second series of experiments, the design was carried out in the stationary region. The variation interval for the variable was narrowed, since it was assumed that the optimal values of the regression coefficients lie near the new center. The response surface is described by second-degree polynomials.
The new design conditions are given in Table 3. Table 4 presents the matrix of rotatable second-order design [3] and the experimental results.
The regression equation obtained was:
\[
\hat{y} = 97 + 5.12x_1 - 1.1x_2 + 0.86x_3 + 1.29x_1x_2 - 1.54x_1x_3 +
\]
\[
+ 1.84x_2x_3 - 4.29x_1^2 - 0.76x_2^2 + 1.63x_3^2,
\tag{3}
\]
Table 1
Conditions of the first series of experiments
| Factor | \(z_1\), sec | \(z_2\) | \(z_3\), mol/l |
|---|---|---|---|
| Basic level | 65 | 7.35 | \(5.005 \cdot 10^{-3}\) |
| Variation interval | 35 | 2.35 | \(4.995 \cdot 10^{-3}\) |
| Upper level | 100 | 9.7 | \(10^{-2}\) |
| Lower level | 30 | 5 | \(10^{-5}\) |
Table 2
Planning matrix and results of the first series of experiments
| No. | \(x_0\) | \(x_1\) | \(x_2\) | \(x_3\) | \(x_1x_2\) | \(x_1x_3\) | \(x_2x_3\) | \(y\) |
|---|---|---|---|---|---|---|---|---|
| 1 | \(+\) | \(-\) | \(-\) | \(-\) | \(+\) | \(+\) | \(+\) | 43 |
| 2 | \(+\) | \(+\) | \(-\) | \(-\) | \(-\) | \(-\) | \(+\) | 56 |
| 3 | \(+\) | \(-\) | \(+\) | \(-\) | \(-\) | \(+\) | \(-\) | 60 |
| 4 | \(+\) | \(+\) | \(+\) | \(-\) | \(+\) | \(-\) | \(-\) | 52 |
| 5 | \(+\) | \(-\) | \(-\) | \(+\) | \(+\) | \(-\) | \(-\) | 96 |
| 6 | \(+\) | \(+\) | \(-\) | \(+\) | \(-\) | \(+\) | \(-\) | 94 |
| 7 | \(+\) | \(-\) | \(+\) | \(+\) | \(-\) | \(-\) | \(+\) | 98 |
| 8 | \(+\) | \(+\) | \(+\) | \(+\) | \(+\) | \(+\) | \(+\) | 99 |
| 9 | \(+\) | 0 | 0 | 0 | 0 | 0 | 0 | 99 |
| Regression coefficients | 75 | 0.5 | 2.5 | 22 | \(-2.25\) | \(-0.75\) | \(-0.75\) |
\[ S\{y\}=1.0 \qquad f=2 \qquad S\{b\}=0.353 \]
Table 3
Conditions of the second series of experiments
| Factor | \(z_1\), sec | \(z_2\) | \(z_3\), mol/l |
|---|---|---|---|
| Basic level | 60 | 7.8 | \(5.05 \cdot 10^{-3}\) |
| Variation interval | 30 | 2.5 | \(2.95 \cdot 10^{-3}\) |
| Upper level | 90 | 10.3 | \(8 \cdot 10^{-3}\) |
| Lower level | 30 | 5.3 | \(2.1 \cdot 10^{-3}\) |
Table 4
Planning matrix and results of the second series of experiments
| Run no. | $x_0$ | $x_1$ | $x_2$ | $x_3$ | $x_1x_2$ | $x_1x_3$ | $x_2x_3$ | $x_1^2$ | $x_2^2$ | $x_3^2$ | $y$ |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | + | − | − | − | + | + | + | + | + | + | 95.5 |
| 2 | + | + | − | − | − | − | + | + | + | + | 98 |
| 3 | + | − | + | − | − | + | − | + | + | + | 86 |
| 4 | + | + | + | − | + | − | − | + | + | + | 96 |
| 5 | + | − | − | + | + | − | − | + | + | + | 97.3 |
| 6 | + | + | − | + | − | + | − | + | + | + | 96 |
| 7 | + | − | + | + | − | − | + | + | + | + | 97.5 |
| 8 | + | + | + | + | + | + | + | + | + | + | 99.0 |
| 9 | + | 1.682 | 0 | 0 | 0 | 0 | 0 | 2.828 | 0 | 0 | 65 |
| 10 | + | −1.682 | 0 | 0 | 0 | 0 | 0 | 2.828 | 0 | 0 | 99 |
| 11 | + | 0 | −1.682 | 0 | 0 | 0 | 0 | 0 | 2.828 | 0 | 94 |
| 12 | + | 0 | −1.682 | 0 | 0 | 0 | 0 | 0 | 2.828 | 0 | 90 |
| 13 | + | 0 | 0 | −1.682 | 0 | 0 | 0 | 0 | 0 | 2.828 | 99.5 |
| 14 | + | 0 | 0 | +1.682 | 0 | 0 | 0 | 0 | 0 | 2.828 | 98 |
| 15 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 99 |
| 16 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 95.5 |
| 17 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 97 |
| 18 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 96 |
| 19 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 96.5 |
| 20 | + | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 99.5 |
| Regression coefficients | 97 | 5.12 | −1.1 | 0.86 | 1.29 | −1.54 | 1.84 | −4.29 | −0.76 | 1.63 |
$S\{y\}=7.73 \qquad f=11 \qquad S\{b_0\}=0.916 \qquad S\{b_i\}=0.752 \qquad S\{b_{ii}\}=0.909 \qquad S\{b_{ij}\}=0.634$
Here \(x_1, x_2, x_3\) are coded variables related to the original natural variables by the relation:
\[ x_i=\frac{z_i-z_{i0}}{\Delta z_i}, \tag{4} \]
where \(z_i\) is the value of the natural variable;
\(z_{i0}\) is the base level;
\(\Delta z_i\) is the interval of variation.
Equation (3) adequately describes the experimental results,
\[ F_{\text{calc}}=4.9 < F_{0.05}(5,5)=5.1. \]
The response surface in the neighborhood of the singular point (at the center of this surface) was studied. To find the singular point of surface (3), a system of equations of the type
\[ \frac{\partial \hat{y}}{\partial x_i}=0, \tag{5} \]
where \(i=1, 2, 3\), was solved.
The obtained coordinates of the center correspond to:
\[ \begin{aligned} x_1&=0.56 &&(77\ \text{s});\\ x_2&=-0.15 &&(7.4);\\ x_3&=0.08 &&(5.3\cdot10^{-3}\ \text{mol/l}). \end{aligned} \]
This point corresponds to an extraction-factor value of 98.6%. The experimental values of the extraction factor at this point are 99%.
The indicated deviation lies within the experimental error; therefore, the found values of the parameters \(z_1, z_2, z_3\) may be considered optimal in the investigated experimental region.
REFERENCES
- Yu. Yu. Lur’e. Handbook of Analytical Chemistry. Moscow, Goskhimizdat, 1962.
- L. A. Chernova, G. Martynenko. ZhAKh, 32, No. 1, 1972.
- V. V. Nalimov, L. A. Chernova. Statistical Methods for Planning Extreme Experiments. Moscow, “Nauka,” 1965.