Statistical Study of Process Parameters Effect on Crystallinity of Electrospun Polyamide 6 , 6 Fibres

Scaffolds play an important role in tissue engineering; therefore a variety of investigations have been conducted using different strategies for obtaining these architectures. Electrospinning of polyamides has been used for obtaining porous nanofibres that can be employed in a variety of applications; among them their use for fabricating tissue scaffolds has attracted a lot of interest. In order to successfully apply these nanofibres in the biomedical field it is important to investigate the diverse parameters involved in this process. Fibre diameter, network morphology and mechanical properties are some of the most important features that can be tailored to mimic specific target tissues. Furthermore, crystallinity of the polymer is related to the mechanical response and OPEN ACCESS


Introduction
Electrospinning is an efficient and versatile process to obtain fine porous fibres with diameters ranging from several microns to tens of nanometers. Electrospun fibres can be adapted in different ways to combine material properties with different functionalities for various applications. In the electrospinning process, a high voltage is applied between a polymeric solution in a syringe and a metallic collector. When the voltage applied reaches a threshold limit the electrostatic forces overcome surface tension of the solution, the hemispherical surface of the drop in the tip of the syringe (Taylor's cone) elongates and an electrically charged solution jet is generated. As the solution droplet passes from the jet to the collector, the solvent evaporates. Solid fibres are formed and collected in the form of non-woven mats when using a static metallic collector [1].
Electrospun nanofibres have shown great potential for many biomedical applications, including the fabrication of scaffolds for tissue engineering, wound dressing, immobilised enzymes and controlleddelivery of drugs [2,3]. Solution, process and environmental parameters can be tuned to achieve desirable properties of the electrospun fibres [4,5]. Among the parameters of the solution are the concentration, viscosity, conductivity, polymer molecular weight and surface tension. Solvent performs two crucial roles, one is to solvate the polymer molecules ready to form the electrified jet and the other is to carry the solvated polymer molecules towards the collector [1]. Within the process parameters are the flow rate, voltage applied, the diameter of the injector tip, the tip and the tipcollector distance. The voltage applied is linked to the distance between the tip of the needle and the collector surface or the metal plate. The higher the electrical potential, the thinner and lighter are the fibres [5]. The injection distance is measured from the tip of the needle to the collecting platform connected to ground. There is a minimum distance required for the fibres to reach the collector properly but also if the distance is too large the fibres cannot be deposited [3,6]. The collector of fibres or metal plate may be fixed or mobile and can be tuned to achieve the fibre thickness [3]. The most relevant environmental parameters are the temperature and humidity [1].
Electrospinning parameters determine the diameter, fibre morphology, the amount of polymer which reaches the collector and the number of defects in the fibres, as well as their structural characteristics [5]. Electrospun fibre mechanical properties such as modulus, strength and elasticity, strongly depend on the initial conditions of the system, thus by proper handling of these parameters, it is possible to produce fibres with the desired characteristics [7].
Furthermore, a variety of works have showed the potential use of polyamide scaffolds due to the ability of the physical properties of these polymers to suit biomedical applications [8][9][10]. Polyamide 6,6, derived from 1,6-hexamethylene diamine and adipic acid, is a semicrystalline polymer which has good thermal stability and mechanical strength, and it is an important engineering thermoplastic [7,[9][10][11][12]. When polyamides are mixed in a protonic solvent, functional groups yield reactions of chemical exchange resulting in a poly-electrolytic behaviour of the solution [13]. This makes polyamide 6,6 a suitable polymer for electrospinning processing.
Because the dependence of the crystalline structure on the mechanical properties of the electrospun fibres [10,11], this research focuses on the crystallinity features provided by the variation of electrospinning parameters including voltage applied, solution concentration, distance from the tip of the needle to the collector surface and flow rate. The parameters accounting for significant changes on the crystallinity of the polyamide 6,6 nanofibres obtained by differential scanning calorimetry (DSC) are presented. Structural changes on the electrospun fibres were also observed by analysing Fourier transform infrared (FTIR) and Raman spectra.

Results and Discussion
Crystallinity is a property directly associated to the structural changes of polymers. The procedure used to obtain the percentage of crystallinity from DSC is explained in the experimental section. Repeatability of the crystallinity results was corroborated by testing 30 samples under the same working conditions (low solution concentration, 20 kV voltage applied, 15 cm tip-collector distance, at 5 ml/h flow rate). This sample space is suitable according to the law of large numbers and the distribution of the crystallinity percentage values are shown in Figure 1. The resulting average crystallinity percentage was 37.7 with a standard deviation of 1.3. The asymmetry coefficient was -0.3709, which is almost a symmetrical distribution with a tendency to the left and a kurtosis of -0.9526, close to normal (see the upper graph in Figure 1). Therefore, the sampling results are close to a normal distribution. This implies that electrospinning process produces independent and reproducible results with less than 3.5% error margin of crystallinity when working under the same conditions. Crystallinity percentage results obtained by DSC of the samples produced at two different voltages, two flow rate velocities, two tip to collector distances and three solution concentrations are found in Tables 1 and 2. The values of these parameters are included in the experimental section.
Considering only the flow rate variations in Tables 1 and 2, the crystallinity percentage shows increases when the electrospun fibres were obtained at high flow rate when using low solution concentration, 15 kV voltage applied, 15 and 20 cm of distance. On the other hand, a decrease of crystallinity percentage was attained when using a high flow rate, low solution concentration, 20 kV voltage applied, 15 and 20 cm of distance. Thus flow rate modification did not produce any consistent effect on the nanofibre crystallinity percentage. Taking into account only changes in the tip-collector separation distance, the results did not show any systematic variation of the crystallinity percentage when modifying the fibre collection distance. As an example, Tables 1 and 2 show that crystallinity percentage can be enhanced when operating at 15 kV, with a low solution concentration and at a flow rate of 2.5 ml/h while when operating at 20 kV, low solution concentration and at a flow rate of 2.5 ml/h this property is reduced.  Electric field voltage variations have a considerable effect on the crystallinity percentage of the polyamide 6,6 fibres. At high voltage applied the crystallinity obtained by DSC resulted in greater values than when operating at low voltage and the rest of the parameters were kept constant. For example, in the fibres obtained using a medium concentration solution, flow rate of 2.5 ml/h and 15 cm of tip-collector separation, the crystallinity percentage changes from 41.40±0.014 to 42.83±0.015 when the electric field voltage varies from 15 kV to 20 kV. This behaviour is consistent in most experiments except when using a high solution concentration, 20 cm distance of collection and the two different flow rates. The close values of the samples obtained at these two different voltages applied indicate the limit conditions that favour crystallisation of polymer.
The variation of solution concentration did not provide a regular effect on the crystallinity of the fibres. However, when keeping the flow rate at 2.5 ml/h, 15 cm of separation and voltage of 20 kV, crystallinity percentage changed from 44.10±0.015 to 36.89±0.013 when the solution concentration was varied from low to high. The dramatic effect on the crystallinity can be explained in the following manner. When using a low concentration solution the polymer chains have more mobility resulting in a higher molecular orientation during electrospinning process as compared to fibres obtained from highly concentrated solutions [13]. This behaviour is not noticeable at larger tip to collector distances because increasing the jet trajectory favoured the orientation of polymer chains and crystallinity.
Crystallinity percentage data of the electrospun samples were processed in Minitab software to obtain the main effects plots, included in Figure 2. The values displayed in this figure represent the mean of each parameter. The design of experiment consists of three factors (voltage, distance, flow) with two levels and one factor (solution concentration) with three levels. From these plots we can determine that the factors which have a considerable influence on the crystallinity of the samples are those with the highest slopes; therefore, both voltage and concentration provided the major effects on the tested response when compared to the rest of the parameters. SEM allowed the determination of the polyamide 6,6 nanofibre diameters produced by the variations of the electrospinning parameters. Figure 3 shows the entanglement and disorientation of the fibres due to the electrospinning set up employed in these experiments. The fibre diameters were measured from these SEM images using the image analysis software ImageJ ® . The fibre diameter of the samples which resulted in the highest, medium and lowest crystallinity percentage are reported in Table 3. The three samples analysed resulted in similar fibre diameters due to the combination of the parameters employed for the fibre electrospinning.  The characteristic functional groups of polyamide 6,6 were analysed by FTIR spectroscopy [14,15] and they are summarised in Table 4. Figure 4 shows three of the normalised FTIR spectra obtained using different solution concentration while the rest of the working conditions remained constant. The FTIR spectra show similar features regarding the polar group vibrations of polyamide 6,6. However, there are intensity modifications in the peaks related to carbon skeleton vibrations as depicted in the highlighted area of Figure 4a. Modifications in the peaks located at 1020 and 1095 cm -1 corresponding to the C-C stretch are clearly seen in Figure 4b. This finding may be correlated with changes in the crystal structure. Vasanthan et al. found that librational motion of methylene segments can be associated to changes in the packing of the crystalline polymer chains [15].
Functional groups of polyamide 6,6 identified in the Raman spectrograms [17,18] are included in Table 4. Figure 5 shows the normalised Raman spectra of the fibres obtained at different solution concentration and the rest of the working conditions were kept constant. This figure shows that the peaks in the 1150-1600 cm -1 region are overlapped and as the solution concentration is increased the peaks are better defined and their intensities are also increased. Furthermore, the peak located at ~1386 cm -1 is redshifted. This change to lower wavenumber was also dependent of solution concentration. Some peak intensities were also increased as the solution is more concentrated (1060, 1130 and 1297 cm -1 ). Modifications of Raman peaks in the range of 800-1600 cm -1 mainly related to the carbon backbone have been previously associated with crystalline structural changes of polyamides [18][19][20]. Methylene stretches in the 2800-2900 cm -1 region corresponding to methylene stretches have been also related in terms of crystal morphology [18].  Furthermore, FTIR peaks related to vibrational modes of functional groups in the crystalline and amorphous domains of polyamide 6,6 are located at ~935 and ~1140 cm -1 , respectively [16]. As seen in Figure 6a and 6b, the crystalline peak broadness and intensity increases as the solution concentration is increased (notice that the colours follow the same fashion as the samples labelled in the figures above). However, the intensity of the amorphous peak has a different behaviour, where the fibres with the highest crystallinity percentage in that set of samples exhibited the weakest peak. In addition, the crystalline band at 935 cm -1 can be associated the with the band located at 950 cm -1 in Raman spectra [17,10]. Figure 6c displays that the intensity of these peaks increase as the solution concentration is higher. Zimba et al. correlated the amorphous FTIR peaks found at 1140, 1232 and 1338 cm -1 with the peaks present in Raman at similar frequencies. Our spectra only show the peak at ~1230 cm -1 (see Figure 5d) [21]. Table 5 shows comparative measurements between the FTIR and Raman peak heights and the ratio of crystalline to amorphous heights. The evolution of these crystallinity ratios is in agreement with the results obtained from DSC, indicating that the solution concentration has an effect on the structural changes of crystalline domains.

Electrospinning of Nylon 6,6
Three polymer solutions were prepared by dissolving polyamide 6,6 pellets (Ultramid® A3K, BASF) in formic acid (88%, Sigma-Aldrich) and kept under stirring for four hours at 60 °C. Then, the solution was cooled to room temperature for being deposited into a 5 ml syringe (21 gauge, 1ʺ needle). The flow rate of the solution was controlled using a syringe pump (KDScientific 101) at two dose velocities. Two different voltages were applied directly to the needle and two tip-collector distances were employed. A copper plate wrapped with aluminium foil was used. Table 6 shows the electrospinning parameters of the experiments.

Polyamide 6,6 characterisation techniques
DSC tests were carried out in a TA Instruments DSC Q2000. The samples were tested from 25 to 280°C at a heating rate of 10°C/min. A 5 minute isotherm was applied for subsequently cooling down the sample to 25°C, using nitrogen as the purge gas. Polyamide 6,6 crystallinity determination in this experiment was performed by integration of the melting curves in the region from 200°C to 274°C using OriginPro ® . The crystallinity percentage is given by: where ΔHm and ΔH 0 m (197 J/g) are the enthalpies of the nanocomposite and purely crystalline polyamide 6,6 respectively [11,12].
The morphology of the fibres was observed by SEM using a JSM-6060LV JEOL microscope at an accelerating voltage of 20 kV. The samples were mounted on cupper stubs and were vacuum-coated with gold at 7 × 10 −2 mbar using Argon in a sputter coater EMS 550. SEM images allowed the determination of the average diameter of 30 polyamide 6,6 fibres obtained using the software ImageJ ® .
The characteristic groups of polyamide 6,6 were analysed using a FTIR Bruker Vector 33 spectrometer in attenuated total reflectance mode, with a spectral resolution of 1 cm −1 . Raman spectroscopy was performed in a Raman Senterra Bruker machine with a 100× microscope objective, a laser emitting at 785 nm and a resolution 1 cm −1 .

Conclusions
Structural properties of electrospun fibres can be modified through variation of the characteristic electrospinning parameters. The electrospun fibres obtained in this work produced repeatable results on the crystallinity percentage with a deviation of 3.5%. This property was significantly affected by the voltage applied. Modifications of the flow rate, the tip-collector distance and solution concentration did not have consistent effects on the electrospun fibres fabricated. However, FTIR and Raman spectroscopies aided the identification of structural crystalline changes on the polyamide 6,6 fibres when varying the solution concentration used for the electrospinning process. In addition, the evolution of the crystalline and amorphous peaks in these spectroscopies are in aggrement. Understanding of the effects provided on the crystalline features of polyamide fibres is useful for tuning scaffold mechanical properties.