Optimizing the Shape of the Spinning Electrode for Needleless Coaxial Electrospinning

Abstract

This work deals with designing the optimal shape of the spinning electrode to optimize the distribution of the electric field and suppress the formation of corona discharges on the surface of the electrode during electrospinning using direct current (DC). Some of the solutions used for electrospinning are solved in flammable solvents, such as PVB; therefore, corona discharges are hazardous, as they cause sparks that can cause fires and explosions. The shape optimization was carried out on a plate weir electrode, which uses the principle of free surface spinning. Using the electric field simulation, an analysis of the plate weir spinner was carried out, and its optimization was aimed at minimizing the occurrence of corona discharges, which negatively affect the spinning process. Based on the simulations’ results, the spinning electrode design parameters were adjusted so that an even distribution of the electric field over the entire active surface of the electrode was ensured, and the incidence of corona discharges was prevented. A laboratory experiment was used to validate the function of the design changes in the spinning electrode.

1. Introduction

Thanks to their properties, nanofibrous structures are among the most important materials of this millennium, and their application potential is constantly increasing. For this reason, basic and applied research is devoted to these materials. The areas in which nanofibers appear are diverse and constantly expanding. Currently, nanofibrous structures can be found in tissue engineering [1,2], filtration [3,4], chemical and medical [5] applications [6] and the electrical [7] and electronic industries [8]. Recently, very progressive applications have appeared that expand the fields of application of nanofibrous structures. Flexible sensors and wearable devices can be considered among the most important of these applications [9,10,11]. Nanofibrous structures can be incorporated with various additives for specific applications. Various types of additives can be used, such as activated carbon, drugs, or sorbents [12]. A variety of methods can create pure nanofibers. Among the most used are AC and DC electrospinning [13], melt-blowing [14] and force-spinning [15,16], which are used not only for producing nanofibers for laboratory and research purposes but also for products that can be produced on a mass scale. On the other hand, there are other methods whose products can produce nanofibers, but their technological level ranks them among methods used for laboratory purposes. This is due to their low productivity, and drowning is a typical example [17].

DC electrospinning is currently the most commonly used method for producing nanofibers for research and commercial production, especially for flat nanofiber textiles in many fields. The essence of electrospinning is generally the use of a high electric voltage between the spinning electrode and the collector [18]. Using an electric current, nanofibers can be produced via the DC electrospinning (DC) and AC electrospinning (AC) methods [19]. Although AC electrospinning is a technology with several times higher productivity [20], it has a significantly different fiber morphology than DC [21]. The most significant difference is in the diameters of the fibers produced. However, with DC electrospinning, fibers that are up to several hundred nanometers finer can be achieved than with AC. It is also worth noting that the DC method has a more homogeneous structure [22]. For parametric adjustments of the production technology from the point of view of electrical quantities, the values of the voltages on the positive and negative electrodes and the distance between the electrodes are changed. While AC electrospinning is more variable, it is possible to modify the shapes of the excitation signals, especially their amplitude and frequency. Based on the mentioned differences, the fiber structures obtained using different technologies suit different applications. DC-made structures are more suitable for medical applications or tissue engineering. On the other hand, structures produced using AC are mainly used, for example, in protective textiles, hygiene aids or filter applications.

Needleless electrospinning is spinning from the surface of the free layer of the polymer solution when the action of an electric field destabilizes this layer. This phenomenon is known as Tonks–Frenkel instability [23] This phenomenon uses the so-called self-organization of nozzles on the spinning solution’s free surface using the fastest-growing wave mechanism. The electrospinning process occurs after reaching a specific electric field value (critical electric field value). The so-called Taylor cones are created on the surface of the polymer solution, from which nozzles and nanofibers are formed. The solution is drawn from the spinning electrode in nozzles, passes through the bending instability zone and is transformed into nanofibers [24]. During the passage of the bending instability nozzle, most of the solvent evaporates, and the finished dried nanofibers are deposited on the collector with the opposite charge to the electrode, as shown in Figure 1.

The advantage of needleless spinning is primarily an increase in the productivity of nanofiber production. Another advantage is a more even distribution of the electric field around the spinner, compared to the inhomogeneous electric field in the case of a system of needles. The article deals with optimizing the shape of a new spinning electrode designed for coaxial electrospinning technology [25]. Coaxial fibers are mainly produced by spinners equipped with needle coaxial electrodes. In general, the disadvantage of needle electrodes is low productivity and the small width of the produced layer of nanofibers. Therefore, the effort is to develop an electrode for ensuring the high productivity of nanofibers from the free surface. Some types of these electrodes are known from previous research, such as the cylinder coaxial or weir coaxial electrodes [26]. In this article, the focus is on optimizing the weir coaxial electrode. Figure 2a shows the new design of a coaxial electrode that allows production from the free surface of the polymer solution with its 80 mm width, while the total length of the electrode is 120 mm. Figure 2b shows a schematic cross-section of the electrode to clarify its functional principle. The spinning electrode consists of two chambers, one for the supply of internal spun material and the other for the supply of external spun material. The chambers feed the spinning material and are divided by a plastic partition. A metal wire with a 1 mm diameter is placed at the outlet of the electrode chambers, to which high electric voltage is applied. Both polymer solutions flood the wire and create two separate layers, with the upper solution forming fibers and the lower solution being drawn into the fibers, forming a core in the resulting coaxial nanofiber. The non-fibrous solution is drained into a collection container, as shown in Figure 2a and reused in the spinning process.

The main disadvantage of this electrode is the high probability of spark discharges in places with a high concentration of electric field intensity. The occurrence of roe discharges adversely affects the spinning process, and, therefore, efforts were made to suppress this phenomenon. An example of the occurrence of spark discharges is shown in Figure 3. From that figure, spark discharges at both ends of the overflow spinner can be observed, which cause the leakage of electrical energy and a reduction in spinning productivity. The development of the electrode further dealt with the design of shape changes to suppress unwanted corona discharges. Eliminating corona discharges is very important, especially from the following points of view, which contribute to the stability of the production process. The most important thing is safety, as polymer solutions based on alcohol and other explosive solvents can ignite and explode. Another point of view is the parasitic consumption of electric power by corona discharges. Therefore, it is essential to minimize energy losses arising due to the non-optimal shape of the electrode.

2. Design and Simulations

2.1. Mathematical Model of the Original Design

The mathematical model of the spinner shown in Figure 4a consists of four essential parts: the electrode, the collector, the close vicinity of the electrode and the collector formed by air and the other space bounded in reality by the frame of the device, also formed by air. The air space was, thus, divided into two parts due to the different sizes of the network elements. The elements were too small for the inner part to obtain results with higher accuracy than for the outer part. The mesh for the entire model was made up of tetrahedrons. The electrode itself was simplified to contain essential parts only. Structural elements such as screws, inlet channels, cavities and collection containers were removed, and the model consisted only of the electrode body and the wire. A relative permittivity of 1 was set for air, with 2.3 for the HDPE electrode body. For the metal parts of the model (collector, wire, metal), the relative permittivity was set to the value 10⁹. To obtain more accurate results, the network of the wire model was thickened in the monitored locations to achieve a relative accuracy of 2 percent, as shown in Figure 4b,c. The boundary conditions were chosen as follows: A voltage of +25,000 V was applied to the surface of the electrode wire, and a voltage of −25,000 V was applied to the surface of the collector. A voltage of 0 V was applied to the outer surfaces of the model since the machine frame was grounded. The electrical stiffness value was entered for all boundary conditions in the same way, namely 10⁹ A/V.

The metal wire mesh on the top of the electrode was densified, and, thus, the subtraction of electric field values was carried out in the three most important areas. Important points where the value of the electric field was monitored were located in the center of the spinning electrode, and this value was marked as the middle value (middle value of the electrical field magnitude—E_mid), and 40 mm from the center of the electrode to each side, marked as the edge value (edge value of the electrical field magnitude—E_edge). This distance resulted from the fact that spinning took place over 80 mm, as mentioned in the previous chapter.

Figure 5 shows the simulation result for the electrode of the original concept. The result shows that the value in the middle of the wire, or the middle of the area where spinning appears (E_mid), is lower than in the extreme positions (E_edge), and the distribution of the electric field along the entire length of the spinning area is not homogeneous, which is undesirable and very problematic for the spinning process.

2.2. Mathematical Model of the Optimized Design

The optimized electrode design, as shown in Figure 6, was created. The metal wire was extended to its ends beyond the edges of the electrode body, and discharge eliminators (spheres) were placed on them. It was calculated that reducing the curvature due to the large radius of the spheres would eliminate high electric field values at the edges of the electrode. Through FEM, an analysis of the electric field intensity distribution near the electrode was performed for the original design, both with the wire alone and for the new concept for two types of eliminator material (HDPE and metal).

The spheres’ size at the electrode’s edges was optimized to achieve electric field homogeneity along the spinning area’s entire length. The values mentioned above were observed, and the diameter of the spheres was searched for, where the E_edge value was equal to the E_mid value. The maximum value on the spheres’ surface was also monitored and marked as E_sphere. Simulations were performed for plastic spheres with diameters of 10 mm, 20 mm, 30 mm, 40 mm, 50 mm and 60 mm. The electric field intensity values from the focused locations are in

Table 1. Electric field values for electrodes with plastic spheres.

Ball Diameter [mm]	Electric Field [V/mm]			Emid/Eedge
	Emid	Eedge	Eball
no ball	7809	8986	x	0.869
10	6765	7285	1652	0.929
20	6694	7020	789	0.954
30	6643	6793	523	0.978
40	6601	6576	385	1.004
50	6554	6389	311	1.026
60	6508	6168	267	1.055

. As this is a symmetrical model, the results for both sides are identical and presented in Table 1, as well as in the text for only one side. The E_edge and E_mid values are shown in the graph; see Figure 7. The calculated values show that the optimal diameter of the plastic spheres is 40 mm. In this case, the E_mid and E_edge values ratio is closest to 1.

The optimal value of the sphere diameter can also be reached from the graph in Figure 7. The optimal value corresponds to the intersection of the curve for E_mid and the curve for E_edge. Figure 8 shows the electric field distribution near the electrode and plastic spheres with a diameter of 40 mm.

In the second case, the electric field distribution on the electrode’s surface was analyzed in an arrangement with metal spheres at the ends of the electrode. The magnitude of the electric field was obtained again from the three main points (sphere, amid and edge). The diameters of the metal spheres, including the values of the electric field intensities, are shown in

Table 2. Electric field values for electrodes with metal spheres.

Ball Diameter [mm]	Electric Field [V/mm]			Emid/Eedge
	Emid	Eedge	Eball
1	7809	8986	x	0.869
2	7734	8704	15,618	0.889
3	7647	8305	12,278	0.921
4	7559	7975	10,315	0.948
5	7495	7681	8558	0.976
7	7345	7282	6296	1.009
9	7205	6911	4897	1.043
10	7124	6714	4600	1.061
20	6395	5023	2596	1.273

. The graphical representation of the results is shown in Figure 9. In the graph, we can find the intersection of the E_mid and E_edge dependence curves, showing the optimal diameter of the spheres to be 7 mm and the graphic demonstration of electric field distribution for metal spheres; see Figure 10.

3. Experiments

3.1. Materials and Equipment

Experiments were performed using a Nanospider Lab nanofiber fabrication facility supplied by Elmarco (Liberec, Czech Republic). The fiberized material was a water-soluble polymer solution of polyvinyl alcohol (PVA, Sloviol, Fortischem a. s., Nováky, Slovakia). PVA was chosen because of its biocompatibility, non-toxicity and good spinnability. An aqueous solution of PVA with a concentration of 12 wt.% was prepared as the shell material. The core material was an aqueous solution of PVA at a concentration of 4 wt.% with added red food coloring. The dye was added to make it easy to observe the spinning behavior of the solutions. Cetoni NEMESYS S (Cetoni, Korbussen, Germany) was used to dose polymeric solution to the electrode and syringe pump. Electric field simulations were performed in the FEM software Autodesk Simulation Mechanical 2017. The CoroCAM 1 camera (UVIRCO, Pretoria, South Africa) produced by CSIR was used to display corona discharges during the experiments.

3.2. Methods

Laboratory experiments were conducted to verify the optimized design and analysis of the electric field near the spinning electrodes. The experiments were performed at an ambient temperature of 21–23 °C and a relative humidity of 37–38%. The optimized spinning electrode, tipped with plastic spheres, as shown in Figure 11a, was operated in an air-conditioned spinning chamber. The electrode (wire position) was clamped in the spinner’s stand and placed 110 mm below the spinner collector. After forming a uniform double layer of the polymer solution, a high voltage (HV) was applied to the electrode. The initial HV value was set at +10, −8 kV and was gradually increased. The following value of the applied voltage was recorded: the critical voltage value (the voltage at which the spinning process started). The optimal spinning value and the maximum voltage value were reached; at this point, the spinning process was still smooth and uninterrupted by discharges. Testing began at a +12, −10 kV supply voltage. The voltage was increased up to a value of +25, −20 kV when corona discharges were observed on the plastic spheres, and after a further increase, the first Taylor cones were visible. Optimal spinning was found at an HV of +33, −30 kV. However, the process was accompanied by the creeping discharge observed on plastic spheres. This discharge did not significantly affect the spinning process.

Spinning was recorded using the CoroCAM. Via this equipment, it was possible to monitor the development of the spinning process and the influence of attached spheres. The image from the camera capturing continuous surface spinning is shown in Figure 11b. The image shows many polymer nozzles without corona discharges, which would hurt the spinning process.

Experiments have shown that spheres with optimized diameters positively affect the course of the spinning process. The uniform distribution of the electric field intensity results in uniform electrospinning from the entire surface of the polymer solution. The spinning process was negligibly affected by creeping discharges that arose at the transition point between the metal part of the electrode and the plastic spheres. The spinning process was not negatively affected by corona discharges caused by a place with a high-intensity electric field (at sharp edges) at the electrode, as was the case with the previous variant of the electrode.

The functionality of the optimized electrode with metal spheres, as shown in Figure 12a, was also verified experimentally. The experiment was carried out in the same way and under the same conditions as in the previous case. Figure 12b shows a recording made using a CoroCAM. The same behavior can be observed here as in the case of the electrode with plastic spheres. Using metal spheres placed at the ends of the electrode significantly contributed to eliminating corona discharges. The use of metal spheres further contributed to the uniform distribution of the electric field and, thus, to the more uniform formation of Taylor cones.

4. Discussion

The experiments showed the functionality of the weir electrode and its ability to influence the homogeneity of the electric field along the length of the functional part of the electrode, especially the formation of corona discharges. These findings are essential for increasing the homogeneity of the produced nanofibrous layer and the safety of the spinning device’s operation. However, it should be noted that the research was focused on a narrow profile of the issue, as the electrode was optimized for only one length dimension. In case nanofibrous layers need to be produced with larger widths, it will be necessary to focus on optimizing the electrode with larger sizes in the research to follow. It will be required to check whether the effect of the used spheres on the edges of the electrode will have as significant an impact as that of the electrode studied in this research.

5. Conclusions

The functionality of the optimized electrodes was experimentally proven, and, in particular, there was an elimination of spark corona discharges, which occurred with the previous version of the electrode and negatively affected the electrospinning process. By optimizing the electrode’s shape, the process’ safety was significantly strengthened in eliminating the occurrence of fires and possible explosions in the case of flammable substances used to prepare spinning solutions. The last step of this study was to determine whether the optimized electrode negatively affected the product’s structure. The morphology of the produced nanofibrous layers was observed using the SEM. Images of nanofibers produced via needleless coaxial electrospinning using an optimized electrode with metal spheres are shown in Figure 13. The average values of the nanofiber diameters are 270 ± 60 nm. It can be seen from the pictures that the nanofibers were produced without defects. This work’s results are the basis for further development stages regarding the optimal de-sign of spinning overflow electrodes. Thanks to the suitable design of the electrodes, higher safety and efficiency of the spinning process can be achieved. This work can help not only laboratory researchers in the field of DC electrospinning but also designers of industrial spinning machines. The results have significant overlap, from the laboratory and theoretical level to the field of application.