Optimization of calcium alginate beads production by electrospray using response surface methodology

Sodium alginate was obtained from (Sigma-Aldrich, Munich, Germany) to make viscous alginate solutions at different concentrations 1–3 w/v %. Calcium chloride was purchased from (Merck, Darmstadt, Germany) to prepare a 3 w/v % gelling solution. All experiments were carried out at room temperature, about 27 °C.

The alginate droplet surface at the tip of the nozzle was subjected to electrical shear stress along with the gravitational force and the liquid capillary was elongated downward. Consequently, droplets were separated from the nozzle. The sodium alginate droplets dripped into a calcium chloride container; this produced calcium alginate beads. The schematic of the process is shown in figure 1.

Zoom In Zoom Out Reset image size

The size analyses of dried beads were carried out by analyzing the digital images captured by a CCD camera installed on a microscope, using Motic Images Advanced software (Hong Kong, China). The average diameter of these beads was reported as the mean particle size. More details about the experimental setup and methods were given by Moghadam et al [12, 16, 17]. The size of the droplets formed in electrospraying was affected by two series of parameters: solution characteristics and process variables. As the solution properties (The surface tension, electrical conductivity, viscosity, and density) were affected by the concentration change, the alginate concentration, along with the flow rate, nozzle diameter, and voltage were set as the variable parameters.

The surface morphology of the produced beads was studied by scanning electron microscopy (XL30 Philips, the Nether-lands). The samples were mounted on metal stubs, coated with gold using sputter coater (Bal-Tec Company, Swiss).

RSM is a collection of mathematical and statistical techniques which response on interest is influenced by several variables and the objective is to optimize this response [22, 24]. The most extensive application of RSM can be found in the industrial world, in situations where several input variables influence some performance measures, called the response. This method is sequential and includes the construction of the response surface by interpolation of the data and optimization of the objective function on the response surface. The variable levels Z_i were coded as X_i according to the following equation such that Z₀ corresponded to the central value:

$$\begin{eqnarray}&&{X}_{i}=\displaystyle \frac{{Z}_{i}-{Z}_{0}}{{\rm{\Delta }}{Z}_{i}},i=1,\,2,\,3,\,\ldots ,\,k\end{eqnarray} \tag{ 1 }$$

where X_i is the dimensionless value of an independent variable; Z_i is the natural value of an independent variable; Z₀ is the natural value of an independent variable at the center point and ΔZ_i is the step change [25]. Polynomial-based response surfaces are usually required to approximate the response. In most cases, the second order model is adequate. This polynomial response function is defined as follows [22]:

$$\begin{eqnarray}&&y={\beta }_{0}+\displaystyle \sum _{i=1}^{k}{\beta }_{i}{X}_{i}+\displaystyle \sum _{i=1}^{k}{\beta }_{ii}{X}_{i}^{2}+\,\displaystyle \sum \displaystyle \sum _{i\lt j}{\beta }_{ij}{X}_{i}{X}_{j}+{\epsilon }\end{eqnarray} \tag{ 2 }$$

where y is the response, k is the number of design variables, X_i (i = 1, ..., k) and β_i (or β_ij) (i,j = 1, ..., k) show the design variables and regression coefficients, and ${\epsilon }$ represents the noise or error observed in the response. The unknown coefficients of the polynomial are obtained from a standard least-squares regression method [21]. The experimental design was a 2^k full-factorial central composite experimental plan with four variable parameters (k = 4), i. e., c the oncentration of polymer, nozzle diameter, flow rate, and voltage. In this study, the second order polynomial coefficients were calculated and analyzed using the 'Minitab' software (Version 15, Minitab Inc., State college, PA, USA).

The surface morphology of electrosprayed alginate beads was studied by using SEM technique. Figure 2 shows a SEM photograph of one sample from beads obtained under alginate concentration; 20 g l⁻¹, nozzle diameter; 1 mm, flow rate; 20 ml h⁻¹, and voltage; 8000 v. As it is seen, the beads had a relatively smooth surface and an admissible spherical shape. Detailed investigation of beads surface morphology has been reported in our previous works [17, 26].

Zoom In Zoom Out Reset image size

In the present work, experiments were planned to obtain a quadratic model consisting of 2⁴ trials plus a star configuration and six center points. The lower and upper bounds for natural variables are presented in table 1.

The model presented a high determination coefficient (R² = 94.21) explaining 94% of the variability in response (table 2). The value of the adjusted determination coefficient is also very large indicating a high significance of the model. The following formula obtained by RSM is the exemplary functional form of the bead radius as a function of the four coded design variables in the present space.

$$\begin{eqnarray}&&\begin{array}{l}r=1.7231+0.1678{X}_{1}+0.2040{X}_{2}+0.1014{X}_{3}\\ \,-\,0.6422{X}_{4}-0.0373{X}_{1}^{2}-0.0783{X}_{2}^{2}-0.2004{X}_{3}^{2}\\ \,-\,\,0.2670\,{X}_{4}^{2}+0.2101\,{X}_{1}{X}_{4}-0.0141\,{X}_{2}{X}_{4}+0.1299\,{X}_{3}{X}_{4}\end{array}\end{eqnarray} \tag{ 3 }$$

where X₁, X₂, X_3, and X₄ are coded values of the natural variables, alginate concentration, nozzle diameter, flow rate, and voltage, respectively. In equation (3) r is the response that is the alginate beads radius. Estimated regression coefficients for r are shown in table 2.

Table 2.  The least squares fit and parameter estimates (significance of regression coefficient).

Model terms	Parameter estimate	Standard error	Computed t-value	Computed P-value
Intercept	1.7231	0.13147	13.116	0.0001
X1	0.1678	0.03567	4.705	0.0001
X2	0.2040	0.03643	5.599	0.0001
X3	0.1014	0.03545	2.861	0.007
X4	−0.6422	0.03866	−16.611	0.0001
${{{\rm{X}}}_{1}}^{2}$	−0.0373	0.06035	−0.619	0.539
${{{\rm{X}}}_{2}}^{2}$	−0.0783	0.06127	−1.278	0.208
${{{\rm{X}}}_{3}}^{2}$	−0.2004	0.14463	−1.385	0.173
${{{\rm{X}}}_{4}}^{2}$	−0.2670	0.05034	−5.303	0.0001
X1. X4	0.2101	0.05506	3.816	0.0001
X2. X4	−0.0141	0.05955	−0.237	0.814
X3. X4	0.1299	0.04683	2.776	0.008

R² = 0.9421; Adj. R² = 0.9273.

The significance of each coefficient was determined by t-values and P-values, which are listed in table 2. The larger t-value and smaller P-value indicate the high significance of the corresponding coefficient.

It can be seen from table 2, the variable with the largest effect was the linear term of voltage, which may due to large t-value (−16.611) and small P-value (0.0001). The great effect of voltage on beads radius in electrospraying method is also emphasized in the literature [12, 16, 17, 26, 27]. The electrical field represses the surface tension of the polymer and therefore, accelerates the fluid motion downward and consequently obtains the smaller beads [28]. This experimental study also revealed that the radius of alginate beads decreased to an almost minimum value with increasing of the applied voltage.

Response surface analysis method reveals that alginate concentration, nozzle diameter, and flow rate have less significant effects on beads radius in comparison to voltage (t-values of 4.704, 5.599, and 2.861, respectively). In previous works from this research team [12, 16, 17, 26], the linear effects of parameters on beads radius were explained in details. These results are demonstrated that the increase of the nozzle diameter generally lead to the production of beads with a larger radius. There was a maximum beads radius for alginate concentration about 20 g l⁻¹. The result also suggested an initial increase and a slight reduction in the bead radius with a flow rate increase. This trend could be due to the transition from dripping to the jet mode of spray. Similar results have been reported about linear effects these parameters [15].

The response surface methodology was used to evaluate the linear and interaction effects of the four numerical variables (alginate concentration, nozzle diameter, flow rate, and voltage) on the alginate beads radius produced by electrospraying. ANOVA results indicate that some of the regression coefficients in this model (equation (3)) are not significant, i.e., the quadratic effect of concentration, nozzle diameter, and flow rate. Furthermore, the combined effect of voltage and nozzle diameter is not important (P-value = 0.814).

The 3D response surface and 2D contour plots are the graphical representations of the regression equation. Plots are shown in figures 3–8. The main goal of response surface methodology is efficiently searching for the optimum values of the variables such a way that the response is minimized. The plots (figures 3 and 4) show the interactions of voltage, alginate concentration and voltage, flow rate, respectively. The 3D plots in these figures along with small P-values (0.0001 and 0.008) represent relatively significant interactions between voltage, alginate concentration and voltage, flow rate. We know that each contour curve represents an infinitive number of combinations of two test variables with the other two maintained at their respective zero levels. Elliptical contours are obtained when there is a perfect interaction between the independent variables [19, 22]. Contour plots in figures 5 and 6 are not perfectly elliptical showing low interactions among the independent variable corresponding to the response surface. Optimum beads radius obtained from figures 5 and 6 are in harmonious with the minimum radius obtained at the intersection of high voltages (> 7000 v), low alginate concentrations (<20 g l⁻¹), and low flow rates (<50 ml h⁻¹).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The large P-value (0.814) and small t-value (−0.237) in table 2 indicate that the interaction between nozzle diameter and voltage is not significant. Regression coefficients for interactions between alginate concentration, nozzle diameter, and flow rate, nozzle diameter cannot be estimated. However, the 3D plots (figures 7 and 8) depicted interactions among these independent variables. The results indicate that there are fewer interactions between nozzle diameter, flow rate and nozzle diameter, alginate concentration. Furthermore, the large P-value showed that there is no significant interaction between the two factors. Therefore, we can conclude that interactions of nozzle diameter with other variables have not important effect on produced beads radius.

One of the most important parts of this study was to determine the optimum operational conditions where minimum beads radius can be obtained. A numerical method presented by Montgomery [22] was used to solve the regression equation (3). As shown in table 3, optimum natural values of model variables are alginate concentration; 10 g l⁻¹, nozzle diameter; 0.001 m, flow rate; 20 ml h⁻¹, and voltage; 9000 v. The minimum beads radius obtained by using these optimized values of variables is 0.3 mm. The minimum beads radius obtained experimentally was 0.2945 mm showing good agreement with the model prediction. Therefore, the developed model could successfully predict the radius of the beads.