The influence of multi-solvent systems on the electrospinning process

Abstract

The paper analyses the influence of binary solvent systems on the electrospinning process and gelation effect. The characteristics of the electrospinning process are presented, considering the physicochemical properties of polymer solutions and the role of different solvents: chloroform, dimethylformamide (DMF), dimethyl sulfoxide (DMSO) and d-limonene. The shift in transition temperatures between the liquid and the gel phase were analyzed, and the obtained fiber structures were assessed using scanning electron microscopy (SEM). The optimization of solvent systems allows for the control of electrospinning process parameters, opening new possibilities for design advanced polymeric materials.

Introduction

Nanomaterials are currently one of the most promising areas of research in materials science, offering unique properties that go beyond the capacities of their conventional counterparts.

Due to their nanoscale size, they are characterized by exceptional specific surface area, conductivity, optical and mechanical properties^1–5. These properties enable widespread applications of nanomaterials, ranging from modern electronics and chemical engineering to the pharmaceutical and environmental industries^6–12. Among many methods of producing nanomaterials, the electrospinning technique draws special attention, allowing the production of polymer structures with nanometric diameters and well-controlled morphology^13,14. The electrospinning process is currently one of the most commonly used methods for producing nonwoven fabrics in micro- and nanometric samples with a wide application potential^14–17. It belongs, together with electrospray technique, to the group of electrohydrodynamic processes relying on interaction between charged particles (ions) and an electric field and fluid flow^18,19. Electrospun structures are widely used in various applications, starting from medicines as modern drug carriers and tissue scaffolds^20–25, environmental engineering in the form of air filters and water purification membranes^26–29, and ending with advanced electronics and energy storage systems, leading to significant improvements in their efficiency^30,31.

The widespread use of electrospinning in so many fields and applications is due to the ease with which the morphology of the structures produced can be designed (from spheroidal through spindle-shaped to fibers), diameter size and porosity^32–35. This is due to the large number of parameters describing the electrospinning process itself and their mutual correlation.

In the literature, these parameters are most often divided into 3 groups related to apparatus (technical setup), polymer solution (chemistry) and environmental (ambient)^34,36,37, which scope is shown on Fig. 1.

Fig. 1.

Parameters describing the electrospinning process.

Electrospinning process is highly sensitive to the precise adjustment of all constitutional, contributing parameters. A smooth modification of only one of them, will mostly affect the remaining ones, which in turn results in a modification of the obtained morphology of the manufactured structures^36,38. This is very important in the case of applications where it is necessary to use fibers of specific dimensions or with the appropriate pore size. An additional advantage of this method is the number of materials that can be processed, which includes polymers of natural and synthetic origin, as well as metals and ceramic materials^35,39,40. These materials can be easily modified at various stages of their production stage, starting with the preparation and design of the polymer solution and continuing through the modification of the conditions for openwork structures or finished product (e.g. plasma modification)^34,41.

In order to obtain a product with specific morphology and properties, it is necessary to select the optimal process parameters and their correlation, including solvent type. This is due to solubility ratio (of the material or additive), dynamic viscosity, surface tension and electrical conductivity—all of them crucial from processing ease and smoothness^34,36.

Solvent type directly affects the value of the solution viscosity through polymer–polymer and polymer–solvent intermolecular interactions, which alternates cohesive forces of the solution. Indirectly, the viscosity of the polymer solution will also be determined by its concentration and the length/branching of the polymer chains^42,43. However, in the case of combinations of certain solvent systems and materials used in the electrospinning process, interactions between them can cause complications that prevent or significantly limit their application in the electrospinning process.

One of the polymer materials intensively investigated in binary solvent systems is polybutylene succinate (PBS)^44–46. The implementation of this polymer in a wide range is due to its versatility. It is often found as a substitute for such materials as polyethylene (PE) or polypropylene (PP) in the packaging industry for contact with food due to similar properties and in agriculture as various types of nonwoven fabrics and mulching mats^47,48. Additionally, this polymer is bioresorbable, which is why it is often found in medical applications as a carrier of active substances (medicines, biostatics), dressings or tissue scaffolds^49–51. Due to its versatility and ease of use as a carrier, this material is used in the electrospinning process, but there are limitations in this respect.

Although the solubility of most biopolymers in common organic solvents is satisfactory, the situation dramatically changes when it comes to binary liquids. Many authors^44,45,52 report the occurrence of a gelation process as a result of a partial miscibility in a specific volume ratios range. It should be noted, however, that none of them, despite noticing the ongoing process, tried to describe it using boundary conditions that would minimize its occurrence, which left an open research gap for the analysis of this issue. However, it should be remembered that from the process point of view, selection of the appropriate solvent system (in terms of constituents type and mixing ratio) and adjustment of the electrospinning process parameters for the polybutylene succinate solution, enables precise control, which is crucial condition for obtaining nanofibers with high homogeneity and desired properties. Therefore, the main goal of this study was to fill the research gap concerning the understanding of the basic mechanisms responsible for the gelation process of two-solvent solutions and the procedures leading to its elimination, but also to what extent it influences the electrospinning process itself.

Materials and methods

Materials

Commercially available biodegradable plastic granulates were used in the study. The polymer used to prepare polymer solutions was polybutylene succinate (PBS) obtained from two manufacturers: BioPBS FD 92 (Mitsubishi, Thailand) and PBE 003 (NaturePlast, France). Following the manufacturer’s note, the physical properties of these materials are listed in Table 1 below.

Table 1.

Properties of polymer granulates.

Properties	BioPBS FD 92	PBE 003
Density [g/cm3]	1.24	1.26
MFR [190 °C; 2.16 kg]	4.0	4.0–6.0
Melting temperature [°C]	84	90
Impact strength [kJ/m2]	47	–

Four analytically pure solvents were used to dissolve both granulates: chloroform (CHF) (Chempur, Poland), dimethyl sulfoxide (DMSO) (EUROCHEM BGD Sp. zo.o.), dimethylformamide (DMF) (Chempur, Poland), d-limonene (Sigma Aldrich, USA).

Preparation of polymer solution

The preparation of polymer solutions followed standard laboratory procedures, including weighing the substrate and dissolving it in the desired volume of primary organic solvent using magnetic stirrer. Chloroform (CHF) was used as a primer due to its acceptable affinity to all granulates, resulting in high polymer dissolution kinetics. Once dissolving was complete, calculated amount of high boiling point (refer to Table 2) solvent was added to the solution.

Table 2.

Boiling temperatures of used solvents.

Solvent	Boiling temperature [°C]
Chloroform (CHF)	61
Dimethylformamide (DMF)	153
Dimethyl sulfoxide (DMSO)	189
d-limonene	175–178

Similar way of solute preparation was described by other authors and results from differences from solubility parameters of CHF, polymer and additional organic species^53–55.

The amount of second solvent varied from 10 to 30% by volume relating to chloroform. So prepared polymer solution systems were placed on a magnetic stirrer to provide sufficient homogenization.

Table 3 below shows the compositions of the polymer solutions used for the electrospinning process. For better clarity, shorter names are used in the paper.

Table 3.

Compositions of the polymer solutions used in the research and their designations.

Polymer used	Solution system	Polymer concentration (%)	Amount of CHF (%)	Amount of the second solvent (%)	Designation of each solution
BioPBS FD 92	CHF + DMF	10	90	10	A1
	CHF + DMSO	10	90	10	A2
PBE 003	CHF + DMF	10	90	10	B1
	CHF + DMSO	10	90	10	B2

Molecular mass

The average molecular mass of the polymers was determined by gel permeation chromatography (GPC), using Agilent series 1200 device (Agilent Technologies, USA) connected to a MiniDAWN TREOS detector (Wyatt Technology Corporation, USA).

Dynamic viscosity measurements

Dynamic viscosity tests of the polymer solutions were carried out using an IKA rotavisco viscometer. The measurements were carried out in triplicate at ambient temperature (23 °C).

Electrospinning process

The electrospinning process was performed using the Bioinicia Fluidnatek LE-10 device (Bioinicia, Spain). This device is equipped with two infusion pumps, a high-voltage power supply and replaceable collectors. The process was carried out along the horizontal axis.

In this model, it is possible to use a flat or rotary collector and electrospinning with the “sweep” function. The device is not equipped with a thermostatic system or a system maintaining humidity at a given, constant level.

Determination of the transition temperature

In order to determine gelling point, a water circulator Akura G21 (G21, Czech Republic) was used, equipped with a heater and two thermocouples, as well as two laboratory thermometers. All elements were placed in a glass vessel together with a bottle containing the prepared polymer solution and then on a magnetic stirrer. One thermometer was placed in the water column and the other in a suitably prepared bottle to verify the transition temperatures. Additionally, in order to verify the transition conditions, a camera was used for continuous measurement observation. The diagram of the research system is shown in Fig. 2 below.

Fig. 2.

Schematic diagram of the test system for determining the transition temperature with a water circulator.

SEM analysis

The morphology of the fabricated structures was determined by microscopic analysis using a Vega Tescan 3 scanning electron microscope (SEM) (Tescan, Czech Republic). Before analysis, the samples were coated with a layer of gold using Cressington 108 sputter coater (Cressington Scientific Instruments, UK). The sputtering process time was 60 s at a current of 40 mA. SEM photos were analyzed statistically to determine the average fiber diameter, which was done using the ImageJ program. For each sample, areas chosen to be analyzed were distributed among the whole sample. The fiber diameter distribution calculation was based on at least 200 separate fibers.

Results

Gelation process

Observed gelation process occurred in all of the tested solvent systems but differed in kinetics and turbidity development. The process most often manifests itself upon the addition of a second solvent with a large electronegativity difference, which results in its turbidity and a change in viscosity until it becomes a solid or grease-like state. The greatest dynamics of this phenomenon was noticeably pronounced in chloroform systems containing DMF and DMSO. During solution preparation, certain dependencies were observed: solutions based on BioPBS FD 92 granulate revealed complete dissolution (without any opaque traces), while 30% solutions transparency was intensely affected.

Contrary to description above, solutions based on PBE 003 could only be prepared in elevated temperatures, regardless of the solvent ratio in binary system. During observation and analyzing the timeline of the experiment on the camera footage solutions became subsequently opaque after cooling down to ambient conditions (23 °C)—firstly, creation of individual agglomerates was noticed and then solutions became nontransparent, as shown in Fig. 3.

Fig. 3.

Gelation process of polymer solutions: (a) occurrence of agglomerates, (b) complete solidification of solutions.

In the case of d-limonene/chloroform systems, the gelation process occurred only when its content was 30% by volume in relation to chloroform, but even though high d-limonene concentration no solidification was observed.

In order to explain this phenomenon, the molecular structures of all solvents and polymers used in the studies were analyzed in terms of their probable interactions.

Table 4 presents dipole moment values for all solvents used. DMF and DMSO molecules reveal considerably higher dipole moment comparing to chloroform—around 400%. Contrary to that, the moment dipole moment is higher only for 50% relating to chlorinated derivative of methane. That observation directly correlates with the gelation tendency, which occurs more readily at lower concentrations of DMF or DMSO in binary CHF solution, suggesting much probable creation of associates.

Table 4.

Dipole moment of solvents used in the research^44,56.

Solvent	Dipole moment [D]
CHF	1.01
DMSO	3.96
DMF	3.90
d-Limonene	1.57

Moreover, an analysis of the solvent molecule’s geometry (Fig. 4) gives rational support to this thesis. All four molecules (CHF, DMSO, DMF and d-limonene) are polarized due to pronounced bond charge separation, resulting from mainly electronegativity and, giving additional input, presence of non-bonding electron pairs on electronegative atoms. This uneven distribution of electron cloud density within single bonds, is pronounced by molecule specific spatial geometry. Thus, within a whole molecule, one can distinguish the electrophilic part—with an uncompensated, partial positive charge, and the nucleophilic part—in which an increased population of electron density is observed, generating an excess—partial negative charge. The uncompensated charge ratio will depend on:

• the polarization of each single bond in a molecule, i.e. the differences in electronegativity of the atoms that make them up,

• the number of polarized bonds,

• the symmetry of the molecule, which determines the mutual arrangement of polarized bonds and thus influencing net dipole moment.

Fig. 4.

Structure of the molecule: (a) dimethyl sulfoxide (DMSO), b) dimethylformamide (DMF), (c) chloroform (CHF), (d) d-limonene, (e) polybutylene succinate (PBS) with designation of nucleophilic and electrophilic groups.

The chloroform molecule consists of three strongly polarized C–Cl bonds, where due to the high electronegativity of the chlorine atom, a shift in electron density is observed towards its direction. Additionally, their nucleophilic character will be enhanced by the presence of three free electron pairs on the halogen atom. Fourth, C–H bond is also polarized, but towards carbon atom and with much lower bond dipole moment comparing to C–Cl ones.

Due to the sp³ hybridization of the carbon atom, the four-bonded central atom will impose high symmetry (C_3V) on the molecule. For this reason, the chloroform molecule, although characterized by the presence of three strongly polarized C–Cl bonds, has a relatively small total dipole moment.

Considerations carried out for DMF and DMSO molecules lead to similar observations. In both compounds there are nucleophilic parts built of atoms with high electronegativity, in particular S=O in DMSO and C=O in DMF. An additional factor enhancing the nucleophilic nature of both groups is the presence of free electron pairs on oxygen (terminal atom). The electrophilic centers of DMSO and DMF molecules are methyl groups occurring as substituents of the central atom. Despite these similarities, the DMF molecule has a lower dipole moment value due to its spatial structure. The double bond (S=O) in the four-bonded sulfur implies planar symmetry (C_2V/Cs) for DMSO with one axis of symmetry along the S=O bond, which is a favorable factor in the discussion of the net dipole moment⁵⁷.

This mechanistic description of possible CHF–DMSO interactions found an experimental evidence in FT-IR and Raman spectra. In the work of⁵⁸ remarkable shifts were observed for binary systems with low DMSO concentration indicating close interaction among organic species.

The symmetry of the DMF molecule is lower (C_2h) It has a carbon atom in sp² hybridization, and a three-bonded nitrogen atom in a trigonal pyramidal symmetry (flattened, the fourth vertex of the pyramid is occupied by a non-bonding electron pair), which causes partial “cancellation” of bond dipole moments during vectorial addition. For this reason, the dipole moment of the DMF molecule is smaller than DMSO, as shown in Table 4.

When considering possible interactions of molecules in solution, it should be taken into account that the dipole moments given in Table 4 refer to isolated molecules, and this assumption is far from the situation occurring in a real systems of these solvents. High values of molecular dipole moments lead to the formation of associates, both DMSO (or DMF) dimers and DMSO–CHF or DMF–CHF combinations.

In the case of the DMSO–CHF system, the formation of associates is basically limited to DMSO dimers (tri- and tetramers are of marginal importance) and bimolecular systems (1:1) of DMSO–CHF, which is confirmed by molecular dynamics simulations and H¹NMR studies⁵⁹. The interaction between DMSO molecules is of a hydrogen bond nature with a distance of 2.38 Å and energy of − 19.8 kJ/mol, while the formation of the DMSO–CHF associate is more energetically favorable (− 34.4 kJ/mol), with a similar value of the hydrogen bond length (2.33 Å). In the three-component system (PBS–CHF–DMSO/DMF) under study, possible interactions between solvent molecules and the polymer chain should still be considered. Taking into account the considerations on electron density distribution given above, one can expect the development of a strong interaction between electrophiles—methyl groups of DMSO (or DMF) and two lone electron pairs present on the oxygen atom (Fig. 4). These specific molecular arrangement were also confirmed by additional molecular dynamics calculations revealed in the work of^60,61.

The attachment of spherical molecules (such as DMSO or DMF) through hydrogen interaction will cause a decrease in the mobility of the polymer chain and, as a result, its solubility in a mixture of two solvents. Therefore, this process may additionally promote the formation of the observed gelation process of the system. Due to the values of the dipole moments of DMF and DMSO, it should be assumed that the interaction with the polymer chain will be stronger in the case of the sulfur–derivative, which seems to be confirmed in the conducted experiment. The transition temperature of the solution from the gel to clear phase was 49.2 °C for the CHF–DMF mixture and 50.6 °C for the CHF–DMSO mixture, when their concentration was 10%. With the increase in the content of the second solvent, the temperature required for there-transition to the liquid state also increased. In the case of the 20% CHF–DMF mixture, the measured transition temperature was already 53.4 °C and 55.8 °C for CHF–DMSO system.

The difference in the transition temperature may be due to the greater strength with which the methyl groups of DMSO interact with the oxygen atoms in the polymer chain^57–59.

These data were used to calculate the activation energy of the gelation process. According to Arrhenius Law, combining temperature dependency with reaction rate, activation energy for CHF–DMSO was estimated as 118 kJ/mol, whereas for CHF–DMF was slightly higher (144 kJ/mol). It should be noted, however, that the intensity of gelation varied depending on the PBS granulate used. In the case of BioPBS FD 92, gelation occurred only when the volume fraction of DMF or DMSO was 20% or more in relation to chloroform. Interestingly, for the PBE 003 granulate, the threshold after exceeding which gelation occurred was much lower; in these solutions, the addition of 5% of the second solvent was sufficient for complete turbidity. In order to determine the basis of this phenomenon, an average molecular weight study was conducted to exclude differences in the structure and size of their polymer chains. The results obtained during the gel permeation chromatography study showed that the granulates did not differ much in molecular weight of about 162 kDa (BioPBS FD 92 161.9 kDa and PBE 003 161.5 kDa). Due to the obtained values, it is assumed that the difference in the intensity of the gelation process between the granulates may result from the difference in their density. Analysis of Table 1 allows to notice that BioPBS FD 92 is characterized by a lower density in relation to PBE 003 (1.24 and 1.26 g/cm³ respectively), which may directly reflect the way in which the solvent particles interact with the single mer units in the polymer chain.

Due to the properties of the solution, only solutions with a concentration of 10% of the second solvent were used for the electrospinning process. The remaining concentrations were rejected due to several processing issues. One of them was rapid solution gelation, which limited the electrospinning time to about 30 min, to less for production of considerable amount of polymer matt. Additionally, the solution evaporation time increased considerably making it impossible to acquire dry product on the collector surface. Mixtures with a higher content of high-boiling point solvent caused the drawn stream of polymer solution to be unable to evaporate sufficiently, which resulted in its inability to solidify into a fiber and deposit on the collector in the form of a film.

Electrospinning process results

After conducting tests on the gelation of the solutions and selecting the appropriate polymer concentrations, they were subjected to the electrospinning process.

The parameters used for the electrospinning process were selected based on a previously established process window allowing it to be conducted in a way that allowed obtaining a continuous fiber from the polymer solution, which was deposited in a dry form on the collector. A quantitative summary of fiber diameter distributions obtained from SEM image analysis is presented in Fig. 5 below. The statistical representation highlights differences in median fiber diameter, data dispersion, and the presence of extreme values between solvent systems and polymer grades.

Fig. 5.

quantitative summary of fiber diameter distribution.

Figure 5 shows a Box-and-whisker plots of fiber diameter distributions for electrospun PBS fibers obtained from solutions A1 (BioPBS FD 92, CHF/DMF), A2 (BioPBS FD 92, CHF/DMSO), B1 (PBE 003, CHF/DMF) and B2 (PBE 003, CHF/DMSO). The boxes represent the interquartile range, the horizontal line indicates the median value, whiskers denote the data range excluding outliers, and individual points correspond to extreme values.

For solutions based on BioPBS FD 92 (series A), a high degree of morphological stability was observed irrespective of the applied secondary solvent. Continuous fibrous structures were obtained for both A1 (CHF/DMF) and A2 (CHF/DMSO) solutions across a wide range of processing parameters, as confirmed by SEM observations. A direct comparison of fiber diameter distributions shown in Fig. 5 reveals clear and systematic differences between the investigated polymer grades and solvent systems. Fibers obtained from BioPBS FD 92 solutions (A1 and A2) are characterized by relatively narrow interquartile ranges and a limited number of extreme values, indicating high dimensional uniformity and stable electrospinning process. In contrast, PBE 003-based systems (B1 and B2) exhibit markedly broader interquartile ranges accompanied by pronounced outliers, reflecting increased variability and local jet instabilities. Among the investigated systems, sample A1 shows the highest median fiber diameter, whereas the lowest median values are observed for DMSO-containing solutions, particularly A2 and B2, confirming the enhanced jet stretching induced by this solvent. The presence and distribution of outliers in series B further correlate with the frequent occurrence of spindle-shaped structures observed in SEM images, linking the statistical dispersion directly to morphological defects.

Figure 6 presents fibrous structures obtained from BioPBS FD 92 solutions with 10% DMF (a–c) (A1) and 10% DMSO (d–f) (A2) at increasing magnifications. In both solutions concentration of the polymer was 10%. Fibers produced from DMF-containing solutions exhibit smooth surfaces, whereas those obtained from DMSO-containing solutions show a porous surface morphology.

Fig. 6.

Fibrous structures obtained for BioPBS FD 92 solution with 10% addition of DMF (a–c) and DMSO (d–f).

The diameter distribution for sample A1 was characterized by a relatively high median value and a narrow interquartile range, indicating good dimensional uniformity and high process repeatability. This structure was created with the following parameters: flow rate of 2.5 ml/h, nozzle-collector distance of 21 cm and voltage of 9.0 kV. Fibers obtained from the A2 solution exhibited a distinctly lower median diameter, accompanied by a moderately broader distribution. This behavior is consistent with the higher electrical potential required for stable electrospinning of DMSO-containing solutions, which enhanced jet stretching and resulted in more pronounced thinning of the polymer fibers. In the case of this sample, a higher voltage of 12.0 kV was used where the rest of the instrument parameters, were identical.

SEM images revealed additional differences in surface morphology between the two solvent systems. Fibers electrospun from the DMF-containing solution exhibited smooth surfaces, whereas those obtained from the DMSO-containing solution showed a clearly porous morphology.

The produced materials differed significantly in terms of the shape of the fiber surface. The use of an appropriate solvent with a high boiling point in the composition of the prepared polymer solutions led to obtaining a smooth (DMF–A1) or porous (DMSO–A2) morphology.

Analysis of the morphology of the structure produced from solution A2 may suggest that the formation of pores was the effect of the phase inversion phenomenon occurring on the surface of the forming polymer fiber. As noted in the previous chapters of the work, DMSO is a highly polar solvent (Table 3) and highly hygroscopic, which results in the absorption of moisture from the process atmosphere in which the electrospinning process was carried out. It should be noted that the process was not carried out in a climate-controlled chamber with strictly defined parameters, but only under ambient conditions, which were: 21 °C and about 60% Rh. Additionally, DMSO does not dissolve polybutylene succinate, which is why this type of porosity can be classified as steam-induced phase separation (VIPS).

In this process, the polymer rapidly precipitates on the surface of polymer fibers, which is caused by the absorption of water vapor (nonsolvent) by the highly hygroscopic DMSO, which most often results in phase inversion on a local scale and the formation of a porous morphology before the solvent evaporates. An interesting issue is the fact that both solvents with which the polymer solutions were doped are highly hygroscopic, but the phase separation phenomenon occurred only in the case of the DMSO solution, which has a relatively lower affinity for water molecules contained in the air compared to DMF.

Interestingly, none of the solutions used in this series of studies showed a tendency to gel, therefore the electrospinning process was repeated for solutions containing PBE 003.

Analyses of the surface morphology of fibrous structures resulting from electrospinning of solutions based on this granulate (B1 and B2) did not show significant differences in terms of surface topography as was the case with BioPBS FD 92 (Fig. 7).

Fig. 7.

Fibrous structures obtained from the solution of PBE 003 with 10% addition of DMF: (a–c) and DMSO: (d–f).

The occurrence of these differences is not entirely clear. Both granulates used were characterized by very similar molecular weights of 161.9 kDa (BioPBS FD 92) and 161.5 kDa for PBE 003. The only differences between the two granulates were the density of 1.24 and 1.26 g/cm³ and the measured dynamic viscosity.

Studies have shown that the measurements of this parameter obtained for the solutions used in the work are characterized by significant discrepancies, as shown in Fig. 8. Regardless of whether the solutions were doped with DMF or DMSO, higher values of dynamic viscosity were characterized by those containing BioPBS FD 92. The measured differences are: 14.5% for DMF (BioPBS FD 92–757.4 Pa∙s ± 66.0; PBE 003–650.9 Pa∙s ± 16.0) and 15.5% for DMSO (BioPBS FD 92–772.9 Pa∙s ± 50.0; PBE 003–652.9 ± 26.0 Pa∙s).The differences in the recorded values may be due to the fact that the granulates used are intended for different processing processes. BioPBS FD 92 is described by the manufacturer as a granulate intended for processes such as extrusion and coating, and PBE 003 mainly for the injection molding process.

Fig. 8.

Values of dynamic viscosity measured for solutions of BioPBS FD 92 and PBE 003.

This means that the differences in the measured dynamic viscosity most likely result from the difference in the structure of the polymer chains and the additives used for a given processing method.

The higher viscosity of BioPBS FD 92 solutions contributes to improved viscoelastic stabilization of the electrospinning jet, facilitating uniform elongation of polymer chains and suppressing the formation of spindle-shaped defects. In contrast, the lower viscosity of PBE 003-based solutions reduces resistance to capillary instabilities, increasing the susceptibility of the jet to local thinning and spindle formation. This effect directly explains the broader diameter distributions and increased occurrence of extreme values observed for series B in Fig. 6.

For the B2 (CHF/DMSO) system, a lower median fiber diameter and reduced dispersion were observed compared to B1, indicating improved dimensional control. This suggests that the higher polarity and electrical conductivity of DMSO partially compensate for the reduced viscoelastic forces by enhancing electrostatic stretching of the jet. Nevertheless, the presence of outliers indicates that this stabilizing effect is not sufficient to fully suppress jet instabilities.

Time-dependent effects were particularly pronounced for PBE 003 solutions due to the ongoing gelation process (Fig. 9).

Fig. 9.

Fibrous structures obtained from the solution of PBE 003 with 10% addition of DMSO after 45 min of cooling.

Increased spindle density and significant heterogeneity in fiber diameters are clearly visible in the SEM image. Fibers spun after prolonged time of solution cooling exhibited a marked increase in diameter dispersion and the appearance of large extreme values, as reflected in Fig. 6 (B2’). These changes are attributed to progressive reduction in polymer chain mobility caused by solvent–polymer associations, which limit elongation under the applied electrostatic forces.

As already noted in the previous chapter, the usability of these solutions was limited in time by the ongoing gelation process, the direct consequence of which for the electrospinning process was initially the breaking of the polymer solution stream, until the process was completely stopped, which occurred after about an hour. For this reason, the effective process window for conducting the process in a stable manner was determined empirically at 30 min, after which the problems described above occurred. The changes occurring in the solution directly affected the electrospinning process and affected the obtained morphological structures. In laboratory testing, this time is sufficient to obtain a small sample of materials, but for an industrial-scale process, it is unacceptable. To extend the shelf life of the electrospinning solution, it is necessary to use heating systems that heat the reservoir alone or in combination with a flow heater. The application of this type of system and its parameters will be the next step in laboratory work. This is clearly visible in Fig. 9 It shows a spindle-shaped structure obtained from the same B2 solution as the material described above with the same process parameters. The only difference is the time it was used. The process was carried out after about 45 min from the time of placing it in the apparatus system. As can be seen, the obtained morphology is significantly different from that in Fig. 7d–f. There is a much greater instability in terms of the obtained diameters.

Overall, the combined analysis of fiber diameter distributions, SEM morphology, and dynamic viscosity measurements demonstrates that BioPBS FD 92 solutions provide superior dimensional stability during electrospinning, while PBE 003-based systems are more sensitive to solvent composition and processing time. These differences arise primarily from variations in solution rheology and gelation behavior, which play a decisive role in controlling jet stability, fiber morphology, and process repeatability.

Conclusions

The article presents the influence of the solvent system selection on the properties of polymer solutions used in the electrospinning process. It has been proven that the use of systems based on solvents with high dipole moment differences such as chloroform, dimethylformamide and dimethyl sulfoxide to dissolve polybutylene succinate (PBS) may negatively affect their affinity for electrospinning. This is due to the presence of methyl groups with a strong affinity for oxygen atoms in the polymer chain, which form molecular associations, which results in a local reduction in the polymer solubility and its precipitation, causing the gelation effect described in the paper.

It was noted, however, that in order to counteract this phenomenon and to transform the solution into a clear form, suitable for conducting the process or even mixing it, heat had to be supplied to the prepared system. This procedure allowed the prepared solutions to be used within a process window of approximately 30 min after reaching ambient temperature. It should be noted, however, that not every PBS granulate was subject to turbidity and gelation.

Despite almost identical molecular weights of BioPBS FD 92 and PBE 003 granulates, the former did not show a tendency to gel regardless of whether the solution was doped with DMF or DMSO. It is suspected that the degree of branching of the polymer chains themselves and the process additives used may have a direct impact on this. This is supported by the slightly different density of both materials and the fact that they are recommended for different processing processes. The next stage of research will involve conducting DSC and NMR analyses to verify these assumptions and also to modify the production system in order to extend the process window and conduct the process in a stable and repeatable manner.

Author contributions

**M. B.:** article conception, writing of the original draft, materials preparation, statistical analysis, methodology, visualization; **K. L.** : edition of the original draft, supervision; conceptualization; **K. S.** : SEM microscope operator, SEM image analysis, edition.

Funding

This research received no external funding.

Data availability

All data obtained during this study is included in this article.

Declarations

Competing interests

The authors declare no competing interests.