Abstract
Advances in polymer technology have provided humankind with a plethora of polymers for use in various sectors, such as tissue engineering, packaging, and personal care and hygiene products. However, the disposal of spent plastics imposes a serious environmental burden and has resulted in biodegradability being considered an essential criterion for single-use polymeric films. Herein, a non-solvent-induced phase separation (NIPS)-based electrospinning method was adopted for the synthesis of biodegradable, porous polycaprolactone–zinc oxide (ZnO) composite nonwoven films. Structural analysis confirmed the uniform incorporation of ZnO nanoparticles without altering PCL chemistry. The selection of solvent mixture ratios was performed using Hansen solubility parameters and by calculating the resulting centroidal distance. This novel centroidal framework enables the selection of appropriate solvent mixtures with balanced affinity toward the polymer and the nanofiller. Chloroform, dimethyl sulfoxide (DMSO), and acetone were used as unary and binary solvent systems. The effects of the Hansen parameters and centroidal distances on the biodegradation properties of the electrospun films were established. Among the three solvents, DMSO exhibits the lowest volatility and the highest conductivity and dielectric constant, whereas chloroform has high volatility, a low dielectric constant, and low conductivity. When combined as a solvent mixture for PCL–ZnO and electrospun, this resulted in fibers with numerous inherent pores, indicating significant phase separation. The film exhibited appreciable degradability when exposed to environmental conditions. A low centroidal distance of the solvent mixture correlated with enhanced porous structures, increased accessibility to hydrolytic and microbial degradation, and a reduced chemical footprint. This predictive methodology thus minimizes experimental iterations and solvent waste, leading to a sustainable route for designing electrospun scaffolds with enhanced environmental degradability.
(i) Easy approach for solvent selection (reduces trial, chemical waste and toxicity). (ii) Ecofriendly route for biodegradable scaffolds. (iii) Balanced affinity towards polymer and nanoparticles.
1. Introduction
Growing environmental concerns associated with the persistence of conventional plastics have intensified research into environmentally degradable polymers. Widely used polymers such as polyethylene and polypropylene are ubiquitous and pose a serious challenge owing to their non-degradable nature arising from long chain hydrocarbon backbones, leading to their accumulation in natural environments.1 Beyond their long-term environmental persistence, petroleum-based conventional plastics contribute significantly to greenhouse gas emissions across their entire life cycle, encompassing fossil resource extraction, polymer synthesis, transportation, and end-of-life disposal. Consequently, the growing carbon footprint associated with plastic production and waste management has emerged as a critical global environmental challenge. Replacement of conventional environmentally hazardous polymers with biodegradable counterparts is thus imminent. Biodegradable polymers eventually degrade into carbon dioxide and water through hydrolytic and enzymatic reactions mediated by microorganisms present in the soil.2
Polycaprolactone (PCL), a linear aliphatic polyester containing ester linkages (–COO–), is considered one of the most promising biodegradable polymers. The susceptibility of PCL to physiological, hydrolytic and soil-enzymatic biodegradation, along with its biocompatibility, has rendered it extensively useful in the domain of biomedical research, including tissue engineering scaffolds, drug delivery systems, and biodegradable implants.3 The functional performance of PCL-based materials can be further enhanced through the incorporation of nanoparticles such as TiO2, ZnO, and Ag, which impart improved mechanical strength, antibacterial activity, and controlled degradation behaviour.4 Among these, PCL/ZnO nanocomposites have attracted considerable attention due to their favourable physicochemical and antimicrobial properties for biomedical applications.5 ZnO nanoparticles are known to exhibit antibacterial activity through mechanisms such as generation of reactive oxygen species and release of Zn2+ ions which lead to the disruption of bacterial cell membranes.
Many applications of polymer and polymer-nanocomposite films demand high porosity.6 Several strategies such as the use of bath collectors,7 additives,8 sacrificial polymer blends,9 breath figure formation,10 and phase separation11 have been explored to enhance porosity in electrospun films. Among these approaches, phase separation offers a relatively simple and effective route for generating porous architectures during electrospinning. In this method, thermodynamic instability within the polymer solution results in separation into polymer-rich and polymer-deficient phases, which subsequently form the structural matrix and porous domains, respectively.12
In the present work, a non-solvent-induced phase separation (NIPS) strategy is employed using binary solvent systems within a single polymer-nanocomposite formulation.13 The influence of binary solvent systems on the biodegradation behaviour of electrospun PCL/ZnO films has not been previously reported. Chloroform (good solvent), acetone (moderately poor solvent), and dimethyl sulfoxide (DMSO, poor solvent) were used to initiate and control the phase separation process. Multiple PCL/ZnO electrospun films were fabricated using different combinations of unary and binary solvent mixtures. Hansen and Hildebrand solubility parameters of the solvent systems14 were analysed using Teas graphs and solubility sphere representations to elucidate solvent–polymer–nanoparticle interactions. The influence of solvent composition on the morphological and structural characteristics of the electrospun films was subsequently correlated with their biodegradation behaviour.
Previous studies employing NIPS-based electrospinning are largely based on empirical approaches to optimize solvent compositions for achieving porous structures.11,13,15 While Hansen solubility parameters (HSPs) have been reported to predict solvent–polymer compatibility and guide solvent selection for electrospinning, their application has majorly been restricted to binary (polymer–solvent) systems. Recently, a few research groups have employed HSPs to guide NIPS-modified electrospinning of pure polymers for diverse applications, including seawater desalination.16,17 Prior studies primarily utilize HSP analysis to ensure polymer dissolution and optimize electrospinning conditions through experimental screening. However, for multicomponent systems such as polymer–nanofiller composites, solvent selection must simultaneously satisfy interactions with both the polymer matrix as well as the dispersed nanofiller. To the best of our knowledge, a predictive framework for the optimal selection of solvents in nanofiller-infused functional polymer composites has not been reported to date. The centroid-based framework observed in this work addresses this limitation by considering the solvent, polymer and nanofiller as a unified interaction system in the Hansen solubility space. The centroid distance provides a quantitative measure of interaction balance among all three constituents, thereby offering a predictive tool for solvent selection. Using Hansen and Hildebrand solubility parameters, the centroid (centre of mass) of the solvent–polymer–nanoparticle interaction triangle was determined, and the distance between this centroid and the midpoint of the segment joining the two solutes was calculated. It is proposed that solvent mixtures producing minimal centroidal distances provide balanced interactions with both polymer and nanoparticle components, thereby enabling uniform nanoparticle dispersion, controlled phase separation, and formation of highly porous electrospun films.
This study demonstrates a prognostic methodology of solvent mixture selection for a polymer–nanofiller conjugate to achieve a porous morphology and consequently enhanced environmental degradation kinetics without extensive experimental effort and chemical waste, supporting the development of more sustainable polymeric materials. The approach is expected to be applicable to other biodegradable polymers such as polylactic acid (PLA), poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), polybutylene succinate (PBS), and their blends, as well as to diverse nanofillers including metal oxides, nanoclays, carbon nanomaterials, and bioactive particles. Beyond electrospinning, the concept may also be relevant to membrane fabrication, coatings, packaging materials, and controlled-release systems where solvent-induced morphology plays a critical role.
2. Method and materials
2.1. Materials
Poly(ε-caprolactone) (PCL) with an average number-average molecular weight of 80 000 Da and zinc oxide (ZnO) nanoparticles with an average particle size of <100 nm were purchased from Sigma-Aldrich, USA. Chloroform (99.8%), acetone (99.9%), and dimethyl sulfoxide (DMSO, 99.9%) were obtained from Merck, India. All chemicals were used as received without further purification.
Selected physicochemical properties of the solvents used in this study, including boiling point, dielectric constant, and their solubility behaviour with respect to PCL, are summarized in Table 1.
Table 1. Physicochemical properties of solvents used in this study.
| Solvent | Chloroform | Acetone | DMSO |
|---|---|---|---|
| Chemical formula | CHCl3 | (CH3)2CO | (CH3)2SO |
| Boiling point (°C) | 61.70 | 56 | 189 |
| Dielectric constant (ε) at 20 °C | 4.81 | 20.7 | 46.68 |
| Solubility of PCL | Good | Moderate | Poor |
| Dissolution temperature (°C) | Room temperature | 60 | 80 |
| Dissolution time (min) | 2 | 10 | Insoluble |
2.2. Methods
2.2.1. Optimization of solvent systems
Different solvent combinations were investigated to study their influence on fibre morphology, fibre diameter, porosity, crystallinity and biodegradation of electrospun PCL/ZnO nanocomposite films. Polymeric solutions containing 15% (w/v) PCL were prepared using three individual solvents (chloroform, acetone, and DMSO) as well as their binary mixtures. The prepared solutions were subsequently electrospun using the Nabond Nanofibre Electrospinning Unit, TL-01, comprising a high-voltage power supply, a syringe pump, and a 10 mL syringe fitted with a 17 mm diameter needle. A direct current (DC) voltage of 20 kV was applied between the needle tip and the collector, maintaining a distance of 15 cm. Each polymer-nanocomposite solution was dispensed at a controlled flow rate of 5 mL h−1 using the syringe pump. A 15 cm × 15 cm aluminium plate served as the collector. Upon complete solvent evaporation, the resulting membranes were carefully peeled off from the aluminium surface for subsequent characterizations. For chloroform–DMSO mixtures, the chloroform-to-DMSO ratio was maintained at 90 : 10 (v/v) to facilitate phase separation while also ensuring sufficient polymer solubility for electrospinning. Similarly, acetone–DMSO mixtures were prepared with an acetone-to-DMSO ratio of 90 : 10 (v/v). For mixtures of chloroform and acetone, two compositions were investigated: C90A10 and A90C10, in which chloroform and acetone were used as the dominant solvents, respectively. All solvent mixtures were therefore designed with a fixed 90 : 10 (v/v) composition to systematically evaluate the effect of solvent quality on electrospinning behaviour and fibre morphology.
Compatibility between the solvent systems and the polymer was further evaluated using Hansen and Hildebrand solubility parameter calculations.18 In addition to solubility considerations, the electrical conductivity of the solvent systems also plays a critical role in determining the electrospinning ability of polymer solutions.19 A total of ten solvent compositions were initially considered. However, only seven solvent systems could be successfully prepared and electrospun. Specifically, electrospinning could not be achieved for the solvent systems D100, D90A10, and D90C10 due to extremely poor polymer solubility arising from the high DMSO content.
3. Results and discussion
3.1. Influence of solvent properties on electrospinning behaviour
Numerous researchers have fabricated electrospun PCL films using chloroform11,20 owing to its excellent solubility and formation of uniform, bead-free fibres. However, due to its high toxicity,21 solvents like DMSO and acetone have been introduced as a partial replacement of chloroform in fabricating electrospun PCL films.22–24 The 90 : 10 good : poor solvent ratio was selected in this study based on extensive literature demonstrating that binary solvent systems containing a predominant good solvent and a limited amount of poor solvent provide an optimal balance between polymer solubility, solution stability, evaporation kinetics, and phase-separation behaviour during electrospinning. Previous studies have shown that the introduction of a small fraction of poor solvent promotes controlled phase separation and pore formation, whereas excessive poor-solvent content can adversely affect solution stability and fibre morphology.25,26 Studies by Katsogiannis et al.11 have revealed that porous, bead-free PCL fibres can be obtained with good to poor solvent ratios between 75–90% (v/v). For a polymer concentration of 15% (w/v), chloroform : DMSO ratio at 90 : 10 (v/v) was selected to maintain good solubility of PCL (due to high chloroform content) as well as to initiate PCL phase-separated electrospinning (due to DMSO content). This solvent system exhibited Taylor cone stability during electrospinning, resulting in uniform fibre formation with the highest porosity. A higher DMSO proportion in the solvent mixture would lead to the formation of merged or beaded fibres, or ribbon-like structures due to reduction in polymer chain entanglement arising from low volatility of the solvent. Solvent ratios used in the present work were designed accordingly. ZnO nanoparticles (3 wt%) were incorporated into the PCL matrix and electrospun using chloroform (CF), acetone (Ac), and dimethyl sulfoxide (DMSO) as unary and binary solvent systems in varying compositions. Resulting film designations are shown in Table 2. The physicochemical properties of the solvents, particularly conductivity, dielectric constant, and solution viscosity, significantly influence jet stability and fibre formation during electrospinning.19
Table 2. Solvents and corresponding films designation.
| Solvent composition (%) | Solvent system designation | Designation of electrospun films | ||
|---|---|---|---|---|
| CF | Ac | DMSO | ||
| 100 | 0 | 0 | C100 | a |
| 0 | 100 | 0 | A100 | b |
| 90 | 0 | 10 | C90D10 | c |
| 90 | 10 | 0 | C90A10 | d |
| 0 | 90 | 10 | A90D10 | e |
| 10 | 90 | 0 | A90C10 | f |
Conductivity and dielectric constant govern the bending instability of the charged jet during electrospinning, thereby affecting fibre stretching and final fibre diameter.19 As shown in Fig. 1(i), the conductivity of solvent mixtures increases with the addition of the poor solvent component. Dielectric constants of the solvent mixtures (εm) were calculated using eqn (1).27
| εm = φ1ε1 + φ2ε2 | 1 |
where φ1 and φ2 represent the volume fractions of the solvents, and ε1 and ε2 denote their dielectric constants.
Fig. 1. (i) Conductivity and dielectric constant of solvents and (ii) viscosity of PCL/ZnO nanocomposite solution in different solvents.
Viscosity also plays a crucial role in fibre formation. Solutions with high viscosity exhibit enhanced entanglement rate of the polymer chains during electrospinning, leading to fibres with larger diameters. On the contrary, solutions with very low viscosity leads to the formation of beads along with fibres.28 The measured viscosities of the PCL/ZnO solutions in different solvent systems are shown in Fig. 1(ii).
3.2. Structural characterization of PCL/ZnO nanocomposite films
The FTIR spectra of neat ZnO, neat PCL and PCL/ZnO electrospun scaffolds fabricated using various solvent systems are shown in Fig. 2(i). The broad absorption peak in the range 442–550 cm−1 is attributed to the Zn–O bond stretching vibration, confirming the presence of ZnO nanoparticles.29 Infrared vibrations of electrospun PCL fibres in the range 2900 to 3000 cm−1 and 2800–2900 cm−1 correspond to the asymmetric and symmetric stretching of C–H2 groups, respectively. A strong peak at approximately 1720 cm−1 is attributed to the carbonyl (C O) bond vibration of the ester linkage in PCL.30 Bands in the ranges 1350–1480 cm−1, 1150–1200 cm−1, 1100–1150 cm−1, 1000–1100 cm−1, 900–1000 cm−1, and 500–900 cm−1 correspond to C–H scissoring along with symmetric deformation, O–C–O stretching, C–O stretching, C–C stretching, C–O–C symmetric stretching and CH2 rocking, respectively.31 A peak at 452 cm−1 indicates the successful incorporation of ZnO nanoparticles into the PCL matrix.32 Notably, the FTIR spectra of the nanocomposite films prepared using different unary and binary solvents showed changes in intensities but not in peak positions. This indicates that variations in the solvent systems did not alter the chemical structure of the polymer–nanoparticle composite.33
Fig. 2. (i) FTIR, (ii) XRD and (iii) TGA plots of pure ZnO, pure PCL and PCL/ZnO nanocomposite films prepared in different solvents.
X-ray diffraction (XRD) patterns of the nanocomposite films are shown in Fig. 2(ii). Diffraction peaks at 2θ values of 21.74°, 22.07° and 23.08° correspond to the orthorhombic structure of PCL,34 while those at 31.62°, 34.59°, and 36.37° correspond to the (100), (002), and (101) planes of ZnO, confirming the hexagonal wurtzite structure (JCPDS file no. 00-0361551). The addition of ZnO resulted in a semi-crystalline structure of the composite films with slight shifts in the characteristic PCL peaks.32 Films produced from different solvent systems exhibit variations in peak intensities, but not in peak positions. This confirms changes in the crystallinity of the nanocomposites while maintaining phase purity.32 Crystallinity index (CI) and crystallite size (Dhkl) were calculated using eqn (2) and (3).35
![]() |
2 |
where Ac and Aa denote the areas of crystalline and amorphous peaks, respectively.36
![]() |
3 |
where K is the Scherrer constant (0.9), λ is the X-ray wavelength (1.54 Å), β is the full width at half maximum of the diffraction peak, and θ is the diffraction angle.
Thermogravimetric analysis (TGA) thermograms representing the thermal stability of the electrospun films are shown in Fig. 2(iii). All the films undergo one-step degradation with distinct onset temperatures and mass losses over the temperature range of 25–535 °C. It is observed that films with higher crystallinity indices have higher onset temperatures as well as lower mass losses, indicating greater thermal stability. Table 3 enumerates the salient structural, thermal, and morphological properties of the composite films prepared using various solutions.
Table 3. Table of morphological, structural and thermal properties of PCL/ZnO films prepared in different solvents.
| Sample ID | a | b | c | d | e | f |
|---|---|---|---|---|---|---|
| Average fibre diameter (µm) | 5.80 ± 3.61 | 3.57 ± 2.04 | 2.09 ± 0.90 | 2.57 ± 1.59 | 2.04 ± 1.41 | 1.36 ± 0.77 |
| Average pore area (µm)2 | 36.76 ± 40.45 | 30.27 ± 9.86 | 61.55 ± 10.39 | 42.80 ± 17.29 | 27.75 ± 13.34 | 50.19 ± 26.76 |
| Porosity (%) | 15.67 ± 1.12 | 7.43 ± 0.49 | 30.94 ± 1.71 | 29.73 ± 2.89 | 17.52 ± 1.06 | 20.65 ± 1.43 |
| Crystallinity index (%) | 36.37 ± 2.19 | 65.74 ± 3.84 | 50.42 ± 2.99 | 53.90 ± 3.46 | 59.85 ± 3.10 | 53.95 ± 2.79 |
| Crystallite size (nm) | 19.24 | 21.27 | 22.46 | 16.17 | 23.09 | 21.34 |
| Onset temperature (°C) | 224.80 | 255.50 | 236.40 | 238.90 | 254.30 | 244.50 |
| Mass loss (%) | 98.49 | 97.83 | 98.47 | 98.27 | 98.02 | 98.26 |
3.3. Effect of solvent composition on fibre morphology
Morphology and porosity of electrospun fibres can be tailored by the solvent evaporation rate during jet stretching.37 In binary solvent systems, a difference in the rates of evaporation of the two solvents leads to phase separation during electrospinning and subsequent control over pore formation within the fibres.27
SEM images of the electrospun PCL/ZnO films prepared using different solvent compositions are presented in Fig. 3. The low dielectric constant and high volatility of chloroform resulted in the formation of thick fibres with low porosity. Contrarily, solvents with high dielectric constants and high conductivities allow further stretching of the fibres during this electro-hydrodynamic process, reducing the diameter of the resulting polymer nanocomposite. Hence, the incorporation of acetone or DMSO into chloroform resulted in enhanced stretching of the electrospinning jet and formation of fibres with smaller diameters.38 The average fibre diameter followed the trend: a > b > d > c > e > f. This trend is consistent with the corresponding decrease in solution viscosity. Solutions with lower viscosities undergo greater elongational stretching under the electric field, producing finer fibres.29 The PCL/ZnO film prepared in pure acetone showed successful formation of fibres, but of non-uniform sizes (σ = ±2.04 µm). With the addition of 10% chloroform, fibres with uniform dimensions (σ = ±0.77 µm) could be obtained, while inversion in the solvent ratio resulted in a beaded-fibre morphology. Beads were found to be eliminated by the addition of a solvent with high conductivity, such as DMSO. Porous fibre structures were particularly evident in samples containing small amounts of DMSO (c and e). The presence of the poor solvent induces non-solvent-induced phase separation during electrospinning, leading to heterogeneous regions within the jet and the formation of surface pores.39 These porous morphologies lead to an increase in the surface area of the fibres and an enhancement of the biodegradation kinetics of the films.40
Fig. 3. SEM images and frequency distribution plots of fibres of PCL/ZnO nanocomposite films prepared using various solvents (i) a, (ii) b, (iii) c, (iv) d, (v) e and (vi) f.
3.4. Influence of solvent system on crystallinity
Crystallinity is a key factor governing the mechanical properties, thermal stability, and biodegradation behaviour of polymeric materials.41 Long-chain macromolecules exhibit significantly different crystallization behaviour compared with small molecules. Polymeric chains with higher molecular weight, such as PCL with a molecular weight (Mn) of 80 000 Da, show limited crystallization due to extensive chain entanglement.42 This intertwining of polymeric chains of PCL restricts chain mobility.43 For crystallization to occur, polymeric chains must be aligned and folded to form ordered lamellae, and this process is hindered when the chains are highly knotted. During crystallization, these entangled segments are trapped within amorphous domains between crystalline lamellae.44 When a solvent is added to the polymer system, the solvent molecules can aid in the disentanglement of the polymer chains, thereby enabling greater interactions among the chains.45 Interactions between the solvent and the polymer lead to a stable entropy with reduced entanglements, in turn reducing the amorphous region of the sample and increasing its crystallinity.46 Furthermore, the incorporation of nanofillers such as ZnO nanoparticles into the polymer solution can influence the crystallization behaviour of the PCL matrix. At lower doping levels (up to 1 wt%), ZnO nanoparticles promote nucleation and enhance the crystallinity of the PCL matrix. However, at higher doping levels (e.g., 3 wt%), strong interactions between ZnO nanoparticles and PCL chains restrict chain mobility, resulting in a reduction in crystallinity.47 The dispersion of ZnO nanoparticles follows the same order as the solubility of PCL in the selected solvents, namely CF > Ac > DMSO.48
Solvent mixtures containing highly polar DMSO showed reduced crystallinity due to the large difference in Hansen solubility parameters between the solvent components. It can be seen from Table 3 that the difference in overall solubility parameter (δt) between chloroform and acetone is only 1.14, whereas those between chloroform and DMSO and between acetone and DMSO are 7.9 and 6.76 MPa1/2, respectively. In the case of a good match of the HSP values of the two solvents, they further diffuse to reach the polymer, producing a homogenous mixture.44 The diffusion of DMSO as the minor solvent during intermixing is hindered due to a mismatch in HSP values and results in inhomogeneity and delayed solvent evaporation, which disrupts the alignment of polymer chains during fibre formation. Consequently, films prepared using the C90D10 solvent system exhibited lower crystallinity compared to those prepared using other solvent mixtures.
3.5. Biodegradation behaviour of electrospun films
Biodegradation studies were conducted on porous films synthesized using binary solvents. Samples (2 cm × 2 cm) were immersed in soil–water slurry for up to 90 days. The degradation behaviour was monitored by measuring weight loss at different time intervals using eqn (4).49
![]() |
4 |
where, W1 and W2 represent the initial and final weights of the samples, respectively. Biodegradation of the composite films in soil-water slurry was recorded in terms of weight loss (%) over time in days (Fig. 4). The degradation mechanism of the PCL/ZnO films proceeds through hydrolytic cleavage of the ester bonds present in PCL and is accompanied by enzymatic oxidation by the microorganisms present in soil. These reactions generate low-molecular-weight byproducts such as carboxylic acids, alcohols, and oligomers, which are eventually converted into CO2 and H2O.50 Among the prepared composite films, ‘c’ exhibited the highest biodegradation rate of 63% weight loss within 27 days. This enhanced degradability can be attributed to higher porosity, larger pore area, and lower crystallinity of the scaffold ‘c’.51 These structural characteristics facilitate water diffusion and microbial penetration into the polymer matrix.52,53 Beyond 40 days, ‘e’ and ‘f’ continued to exhibit consistent slow degradation rates and recorded an increase in degradation from the 65th day. The acceleration in the degradation rate further in time could be due to loss of crystallinity of the polymer.54 The substantially higher degradation observed for films c and d as compared with e and f highlights the critical role of solvent composition in determining the scaffold microstructure. In the chloroform-dominated systems (C90D10 and C90A10, resulting in films c and d), effective polymer dissolution combined with controlled phase separation promote the formation of highly porous fibres with reduced crystallinity. Such structures provide increased amorphous surface area and facilitate penetration of water and microorganisms into the polymer matrix, thereby accelerating hydrolytic and enzymatic degradation. Conversely, the acetone-dominated systems (A90D10 and A90C10), forming films e and f, exhibited less pronounced phase separation and comparatively denser fibre architectures. The resulting reduction in pore connectivity and higher structural order restricted water transport and microbial accessibility, leading to slower degradation kinetics. These observations indicate that biodegradation is governed not only by solvent composition but also by the solvent-induced changes in porosity and crystallinity that control the availability of the polymer matrix to the degradation environment.
Fig. 4. (i) Weight loss due to biodegradation, (ii) maximum weight loss of PCL/ZnO films in soil-water slurry (“ll” denotes fragmentation of films into irrecoverable parts).
In general, films exhibiting higher crystallinity and lower porosity showed slower degradation rates due to the dense packing of polymer chains within crystalline domains, which resist enzymatic and hydrolytic attack.47
3.6. Centroid-based solvent selection model
Effective polymer–solvent interaction is the key to prepare a polymeric solution that is capable of efficient electrospinning. The ideal interaction of the polymer with a compatible solvent system causes swelling and expansion, which in turn provides homogeneity to the solution.55 While Hansen solubility parameters provide useful insights into solvent–polymer interactions, they are insufficient for predicting optimal solvent systems for polymer–nanofiller composites.18
Therefore, a centroid-based approach was introduced to evaluate the combined interactions among solvent, polymer, and nanofiller components. Table 4 shows detailed information about the Hansen solubility parameter (HSP) values which help to select the compatible solvents for PCL and ZnO. For a solvent mixture, three solubility parameters (δd, δp and δh) and the total solubility parameter (δt) follow additive rule. Solubility parameters for each solvent mixture have been calculated following eqn (5a) and (5b).18
| δ = δIϕI + δIIϕII | 5a |
where ϕI and ϕII and δI & δII are the volume fractions and solubility parameters of components I & II, respectively, and δ is the solubility parameter of the solvent mixture.18
![]() |
5b |
‘D(S−s)’ is the distance between the solute and the solvent in the solubility space where Hansen solubility parameters written in subscript ‘S’ and ‘s’ denote that of the solute and the solvent, respectively (eqn (6)).18
| DS−s = [4(δdS − δds)2 + (δpS − δps)2 + (δhS − δhs)2]1/2 | 6 |
Table 4. Solubility parameters, D(S−s) and effective interaction radius of solvents w.r.t. PCL and ZnO R0 = 5.5 and 6.7 for PCL and ZnO, respectively. RED (=D(S−s)/R0).
| Sample ID | Solvent composition (%) | δ t (MPa1/2) | Hansen solubility parameters | D (S−s) PCL | D (S−s) ZnO | RED PCL | RED ZnO | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CF | Ac | DMSO | δ d (MPa1/2) | δ p (MPa1/2) | δ h (MPa1/2) | f d (%) | f p (%) | f h (%) | ||||||
| PCL | — | — | — | 20.31 | 17.70 | 6.20 | 7.80 | 55.85 | 19.56 | 24.61 | — | — | — | — |
| ZnO | — | — | — | 21.46 | 17.25 | 6.75 | 10.83 | 49.53 | 19.38 | 31.09 | — | — | — | — |
| C100 | 100 | 0 | 0 | 18.80 | 17.80 | 3.10 | 5.70 | 66.92 | 11.65 | 21.42 | 3.75 | 6.39 | 0.68 | 0.95 |
| A100 | 0 | 100 | 0 | 19.94 | 15.50 | 10.40 | 7.00 | 47.11 | 31.60 | 21.30 | 6.14 | 3.21 | 1.12 | 0.48 |
| D100 | 0 | 0 | 100 | 26.70 | 18.40 | 16.40 | 10.20 | 40.67 | 36.78 | 22.55 | 10.60 | 9.94 | 1.92 | 1.48 |
| C90A10 | 90 | 10 | 0 | 18.93 | 17.60 | 3.83 | 5.83 | 54.84 | 19.56 | 24.61 | 3.08 | 5.83 | 0.56 | 0.87 |
| A90C10 | 10 | 90 | 0 | 19.70 | 15.73 | 9.67 | 6.87 | 62.8 | 15.57 | 21.63 | 5.33 | 5.74 | 0.97 | 0.86 |
| C90D10 | 90 | 0 | 10 | 19.40 | 17.86 | 4.43 | 6.15 | 48.75 | 29.97 | 21.28 | 2.44 | 8.12 | 0.44 | 1.21 |
| A90D10 | 0 | 90 | 10 | 20.59 | 15.80 | 11.00 | 7.30 | 46.33 | 32.26 | 21.41 | 6.14 | 6.25 | 1.12 | 0.93 |
| D90C10 | 10 | 0 | 90 | 25.66 | 18.34 | 15.10 | 9.75 | 42.50 | 34.92 | 22.59 | 9.13 | 8.67 | 1.66 | 1.29 |
| D90A10 | 0 | 10 | 90 | 25.98 | 18.11 | 15.80 | 9.88 | 41.36 | 36.08 | 22.56 | 9.86 | 9.26 | 1 0.79 | 1.38 |
Relative Energy Dispersion (RED), is the ratio of DS−s to the interaction radius of the solute (R0), RED = DS−s/R0 and represents the solute to solvent interactions. RED < 1 and RED > 1 refer to higher and lower affinities between solvent and solute.18
Teas graph, a ternary diagram representing all the solubility parameters in a 2-D plot and a solubility sphere in a 3-D model representing the solubility space of a polymer and nanofiller are plotted in Fig. 5. Dispersive, polar and hydrogen bond related forces for each solvent are plotted and classified in groups.18 A ternary mapping of the solvents can be obtained by plotting the fractional solubility parameters, fd,fp and fh as in (7) and (8).18
![]() |
7a |
![]() |
7b |
![]() |
7c |
| fd + fp + fh = 100 | 8 |
Fig. 5. Hansen solubility parameters (i) Teas graph of fractional solubility parameters representing polymer and solvent systems, (ii) 3D representation of Hansen solubility parameters as solubility spheres and interactions of different solvents with PCL and ZnO.
The above explanation in terms of HSP values is adequate to explain the solubility of the polymer in various solvent mixture compositions. However, when a nanofiller is also to be used in conjunction with the matrix, HSP alone is insufficient to indicate the choice of suitable solvent compositions for the composite. In this approach, the Hansen solubility parameters of the solvent system, polymer, and nanoparticle are represented as vertices of a triangle in three-dimensional solubility space. The centroid of this triangle represents the combined interaction centre of the system. The distance D, between the centroid and the midpoint of the polymer–nanoparticle interaction line was calculated to evaluate solvent compatibility. Fig. 6 shows the grouping of polymer, nanofiller and various solvent systems in terms of their HSP parameters. Co-ordinates of the centroid for each solvent system and distance between centroid and mid-point of the polymer–filler interaction side have been calculated following (9) and (10) respectively and are presented in Table 5.
![]() |
9 |
where, δd, δp, and δh represent the Hansen solubility parameters of the respective solvent system.
![]() |
10 |
Fig. 6. HSP parameters of solvent system, polymer and nanofiller are plotted as vertices of a triangle in 3D space. Centroid is plotted and its distance from the midpoint of the polymer–nanofiller interaction side is denoted by D.
Table 5. Centroid and its distance from the midpoint of the polymer–nanofiller interaction side for different solvent compositions.
| Solvents used | Co-ordinates of solvents (MPa1/2) | Co-ordinates of centroid (MPa1/2) | D (units) | ||||
|---|---|---|---|---|---|---|---|
| δ d | δ p | δ h | x 1 | y 1 | z 1 | ||
| C100 | 17.80 | 3.10 | 5.70 | 17.58 | 5.35 | 8.11 | 1.65 |
| A100 | 15.50 | 10.40 | 7.00 | 16.82 | 7.79 | 8.54 | 1.66 |
| D100 | 18.40 | 16.40 | 10.20 | 17.78 | 9.79 | 9.61 | 3.33 |
| C90A10 | 17.57 | 3.83 | 5.83 | 17.51 | 5.60 | 8.15 | 1.45 |
| A90C10 | 15.73 | 9.67 | 6.87 | 16.89 | 7.54 | 8.50 | 1.46 |
| C90D10 | 17.86 | 4.43 | 6.15 | 17.60 | 5.80 | 8.26 | 1.26 |
| A90D10 | 15.79 | 11.00 | 7.32 | 16.91 | 7.99 | 8.65 | 1.74 |
| D90C10 | 18.34 | 15.07 | 9.75 | 17.76 | 9.34 | 9.46 | 2.89 |
| D90A10 | 18.11 | 15.80 | 9.88 | 17.69 | 9.59 | 9.50 | 3.13 |
Fig. 7(i) shows the SEM images of the PCL/ZnO film ‘c’ after biodegradation in soil for 27 days. Surface morphology of the film became rougher and less uniform with the appearance of several cracks, scratches and pores, which are preferentially seen in the amorphous regions.47 The irregular texture confirms surface erosion, indicating initiation of degradation from the surface of the fibres towards the core. The solvent system possessing the minimum centroidal distance produced films with the highest porosity, lowest crystallinity, and fastest biodegradation rate (Fig. 7(ii)). This observation demonstrates that solvent mixtures with balanced affinities towards both the polymer and nanoparticle promote favourable phase separation and microstructural development during electrospinning. The significance of the centroidal distance can be interpreted in terms of the interaction balance within the polymer–nanofiller–solvent system. A smaller centroidal distance indicates that the solvent possesses a more balanced avidity toward both PCL and ZnO, thereby creating a more homogeneous interaction environment during electrospinning. Such balanced interactions promote controlled phase separation in the polymer matrix and uniform nanofiller dispersion throughout the matrix, ultimately facilitating the formation of highly porous fibrous architectures with well-dispersed nanofiller distribution. Conversely, larger centroidal distances reflect a greater disparity in solvent affinity toward one or both of the solute components, which may lead to heterogeneous phase separation, non-uniform nanofiller distribution, and less developed pore structures.
Fig. 7. (i) Fibre morphology of film ‘c’ after degradation, (ii) trend of biodegradation and centroid distance of films electrospun using various solvent mixtures.
Incorporation of ZnO nanoparticles into the PCL matrix resulted in a decrease in crystallinity, particularly at 3 wt% ZnO loading. This reduction can be attributed to the restriction of PCL chain mobility caused by strong polymer–nanoparticle interactions, which hinder the rearrangement and folding of polymer chains into ordered crystalline domains during solidification.47 Furthermore, ZnO nanoparticles may interfere with lamellar crystal growth by acting as physical obstacles within the polymer matrix, thereby increasing the amorphous fraction of the material.47,56 Consequently, solvent systems exhibiting smaller centroidal distances generate scaffolds with a combination of higher porosity and lower crystallinity, both of which promote water penetration, microbial attack and hydrolytic cleavage of ester bonds of PCL. This synergistic effect explains the enhanced biodegradation observed for samples possessing lower centroidal distances.
Thus, the centroidal distance provides a predictive basis for identifying optimal solvent systems in polymer–nanofiller electrospinning processes, reducing the need for extensive experimental screening.
4. Conclusions
In this study, we established a sustainable and predictive strategy for producing biodegradable PCL/ZnO electrospun films using a solvent-mediated centroidal approach guided by Hansen solubility parameters. By selecting binary solvent mixtures with minimal centroidal distance, controlled phase separation was achieved, producing porous fibres with enhanced environmental degradability. The incorporation of ZnO nanoparticles aided in increasing the amorphous domains of the polymer composite and the films demonstrated accelerated biodegradation due to improved microbial accessibility.
This work highlights that solvent selection is a critical lever for environmental degradation performance, allowing the design of polymer–nanoparticle composites with tailored porosity and minimal carbon footprint, without relying on experimental screening. In addition, the adoption of biodegradable polymers with ester linkages, coupled with reduced solvent usage and waste generation during processing, substantially decreases the overall environmental impacts associated with material development. While chloroform was employed, its partial substitution with acetone and DMSO has been reported. Moreover, the centroidal framework is transferable to greener solvent systems, enabling broader adoption in environmentally responsible polymer processing.
The future of polymeric nanocomposite films lies in the development of multifunctional materials that integrate mechanical robustness, biodegradability, antimicrobial performance, and environmental sustainability within a single platform. Ongoing research is directed toward achieving improved nanoparticle dispersion, strengthening polymer–nanoparticle interfacial interactions, and designing environmentally benign and biodegradable nanocomposite systems. These advancements are anticipated to accelerate the development of next-generation sustainable packaging materials, biomedical devices, tissue engineering scaffolds, and other high-value functional products.
Overall, the methodology offers a generalizable platform for the rational design of sustainable functional polymer composite materials, including biodegradable packaging, biomedical scaffolds, and environmentally responsive devices, aligning with green chemistry principles by reducing chemical waste, hazard, and energy-intensive experimentation.
Author contributions
Arpita Banerjee: writing the original draft, methodology and investigation. Mousumi Majumder: review and editing, project administration, conceptualization.
Conflicts of interest
There are no conflicts to declare.
Acknowledgments
This work is financially supported by the Council of Scientific and Industrial Research, India in Project No. FTT030502. Authors acknowledge the help of M. B Narjinary, P. Khan and S. Samanta, 4M Division, CSIR-CGCRI for FTIR, XRD and TGA results. One of the authors, A. B. acknowledges support from CSIR, India. ChatGPT 5 has been used for language editing and better readability.
Data availability
All data generated or analysed during this study are included in this published article.
Notes and references
- Yao Z. Seong H. J. Jang Y.-S. Ecotoxicol. Environ. Saf. 2022;242:113933. doi: 10.1016/j.ecoenv.2022.113933. [DOI] [PubMed] [Google Scholar]
- Leja K. Lewandowicz G. Pol. J. Environ. Stud. 2010;19:255–266. [Google Scholar]
- Krumov N. Atanasova N. Boyadzhieva I. Petrov K. Petrova P. Int. J. Mol. Sci. 2025;26:5826. doi: 10.3390/ijms26125826. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shabalina A. V. Kozlov V. A. Popov I. A. Gudkov S. V. Nanomaterials. 2024;14:1753. doi: 10.3390/nano14211753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Permyakova E. S. Manakhov A. M. Kiryukhantsev-Korneev P. V. Leybo D. V. Konopatsky A. S. Makarets Y. A. Filippovich S. Y. Ignatov S. G. Shtansky D. V. Polymers. 2022;14:5364. doi: 10.3390/polym14245364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bari S. S. Chatterjee A. Mishra S. Polym. Rev. 2016;56:287–328. doi: 10.1080/15583724.2015.1118123. [DOI] [Google Scholar]
- Ameer J. M. Pr A. K. Kasoju N. J. Funct. Biomater. 2019;10:30. doi: 10.3390/jfb10030030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Song Q. Cao S. Pritchard R. H. Qiblawey H. Terentjev E. M. Cheetham A. K. Sivaniah E. J. Mater. Chem. A. 2016;4:270–279. doi: 10.1039/C5TA09060A. [DOI] [Google Scholar]
- Noh S. Nguyen D. N. Park C. S. Kim Y. Kong H. J. Kim S. Kim S. Hur S.-M. Yoon H. J. Phys. Chem. C. 2017;121:18480–18489. doi: 10.1021/acs.jpcc.7b02558. [DOI] [Google Scholar]
- Dou Y. Jin M. Zhou G. Shui L. Membranes. 2015;5:399–424. doi: 10.3390/membranes5030399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Katsogiannis K. A. G. Vladisavljevic G. T. Georgiadou S. Eur. Polym. J. 2015;69:284–295. doi: 10.1016/j.eurpolymj.2015.01.028. [DOI] [Google Scholar]
- Chen W. Shao Y. Li X. Zhao G. Fu J. Nano Today. 2014;9:759–784. doi: 10.1016/j.nantod.2014.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wittmar A. S. M. Koch D. Prymak O. Ulbricht M. ACS Omega. 2020;5:27314–27322. doi: 10.1021/acsomega.0c03632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zuaznabar-Gardona J. C. Fragoso A. J. Mol. Liq. 2019;294:111646. doi: 10.1016/j.molliq.2019.111646. [DOI] [Google Scholar]
- Zhang S. Campagne C. Salaün F. Appl. Sci. 2019;9:402. doi: 10.3390/app9030402. [DOI] [Google Scholar]
- Melo G. H. F. Sundararaj U. Macromol. Rapid Commun. 2024;45:2400403. doi: 10.1002/marc.202400403. [DOI] [PubMed] [Google Scholar]
- Su Q. Zhang J. Zhang L.-Z. Desalination. 2020;476:114246. doi: 10.1016/j.desal.2019.114246. [DOI] [Google Scholar]
- Hansen C. M., Hansen Solubility Parameters: A User’s Handbook, CRC Press, 1999 [Google Scholar]
- Sun Z. Deitzel J. M. Knopf J. Chen X. Gillespie Jr. J. W. J. Appl. Polym. Sci. 2012;125:2585–2594. doi: 10.1002/app.35454. [DOI] [Google Scholar]
- Mochane M. J. Motsoeneng T. S. Sadiku E. R. Mokhena T. C. Sefadi J. S. Appl. Sci. 2019;9:2205. doi: 10.3390/app9112205. [DOI] [Google Scholar]
- Schueren L. V. De Schoenmaker B. Altan Ö. İ. K. De Clerck K. Eur. Polym. J. 2011;47:1256–1263. doi: 10.1016/j.eurpolymj.2011.02.025. [DOI] [Google Scholar]
- Bosworth L. A. Downes S. J. Polym. Environ. 2012;20:879–886. doi: 10.1007/s10924-012-0436-3. [DOI] [Google Scholar]
- Nyamiati R. D. Rahmawati Y. Altway A. Nurkhamidah S. IOP Conf. Ser.:Mater. Sci. Eng. 2021;1143:012063. [Google Scholar]
- Prat D. Wells A. Hayler J. Sneddon H. McElroy C. R. Abou-Shehada S. Dunne P. J. Green Chem. 2016;18:288–296. doi: 10.1039/C5GC01008J. [DOI] [Google Scholar]
- Megelski S. Stephens J. S. Chase D. B. Rabolt J. F. Macromolecules. 2002;35:8456–8466. doi: 10.1021/ma020444a. [DOI] [Google Scholar]
- Casper C. L. Stephens J. S. Tassi N. G. Chase D. B. Rabolt J. F. Macromolecules. 2004;37:573–578. doi: 10.1021/ma0351975. [DOI] [Google Scholar]
- Jouyban A. Soltanpour S. Chan H.-K. Int. J. Pharm. 2004;269:353–360. doi: 10.1016/j.ijpharm.2003.09.010. [DOI] [PubMed] [Google Scholar]
- Nezarati R. M. Eifert M. B. Cosgriff-Hernandez E. Tissue Eng., Part C. 2013;19:810–819. doi: 10.1089/ten.tec.2012.0671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zamiri R. Rebelo A. Zamiri G. Adnani A. Kuashal A. Belsley M. S. Ferreira J. M. F. RSC Adv. 2014;4:20902–20908. doi: 10.1039/C4RA01563K. [DOI] [Google Scholar]
- Kuppan P. Sethuraman S. Krishnan U. M. J. Biomed. Nanotechnol. 2013;9:1540–1555. doi: 10.1166/jbn.2013.1653. [DOI] [PubMed] [Google Scholar]
- Özyurt M. G. Bayır E. Doğan Ş. Öztürk Ş. Şendemir A. Turk. J. Biol. 2020;44:121–132. doi: 10.3906/biy-1908-42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Banerjee A. Roy P. Chakraborty J. Majumder M. Mater. Today Commun. 2024;40:109452. doi: 10.1016/j.mtcomm.2024.109452. [DOI] [Google Scholar]
- Manholi S. Athiyananthil S. J. Appl. Polym. Sci. 2021;139(10):51720. doi: 10.1002/app.51720. [DOI] [Google Scholar]
- Sownthari K. Suthanthiraraj S. A. eXPRESS Polym. Lett. 2013;7:495–504. doi: 10.3144/expresspolymlett.2013.46. [DOI] [Google Scholar]
- Mannan A. Kazmi K. R. Khan M. S. Khan I. H. Pak. J. Sci. Ind. Res. 2006;49:72–76. [Google Scholar]
- Ibrahim M. Mohammed A. S. Ramizy A. J. Ovonic Res. 2018;14:17–25. [Google Scholar]
- Lasseuguette E. Malpass-Evans R. Tobin J. M. McKeown N. B. Ferrari M. C. Membranes. 2021;11:422. doi: 10.3390/membranes11060422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Van der Schueren L. De Schoenmaker B. Kalaoglu-Altan O. I. De Clerck K. J. Ind. Text. 2017;47:57–70. doi: 10.1177/1528083716634032. [DOI] [Google Scholar]
- Ji S.-H. Yun J.-S. Nanomaterials. 2022;12:404. doi: 10.3390/nano12030404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen V. J. Ma P. X. Biomaterials. 2006;27:3708–3715. doi: 10.1016/j.biomaterials.2006.02.020. [DOI] [PubMed] [Google Scholar]
- Otitoju T. A. Kim C.-H. Ryu M. Park J. Kim T.-K. Yoo Y. Park H. Lee J.-H. Cho Y. H. J. Cleaner Prod. 2024;467:142905. doi: 10.1016/j.jclepro.2024.142905. [DOI] [Google Scholar]
- Fernández-Tena A. Pérez-Camargo R. A. Coulembier O. Sangroniz L. Aranburu N. Guerrica-Echevarria G. Liu G. Wang D. Cavallo D. Müller A. J. Macromolecules. 2023;56:4602–4620. doi: 10.1021/acs.macromol.3c00234. [DOI] [Google Scholar]
- Liu X. Cui W.-Z. Yu W. Chin. J. Polym. Sci. 2022;40:1451–1465. doi: 10.1007/s10118-022-2814-1. [DOI] [Google Scholar]
- Kurz R. Schulz M. Scheliga F. Men Y. Seidlitz A. Thurn-Albrecht T. Saalwächter K. Macromolecules. 2018;51:5831–5841. doi: 10.1021/acs.macromol.8b00809. [DOI] [Google Scholar]
- Pawlak A. Krajenta J. J. Compos. Sci. 2023;7:521. doi: 10.3390/jcs7120521. [DOI] [Google Scholar]
- Saalwächter K. Thurn-Albrecht T. Paul W. Macromol. Chem. Phys. 2023;224:2200424. doi: 10.1002/macp.202200424. [DOI] [Google Scholar]
- Augustine R. Kalarikkal N. Thomas S. Int. J. Polym. Mater. Polym. Biomater. 2016;65:28–37. doi: 10.1080/00914037.2015.1055628. [DOI] [Google Scholar]
- https://www.rsc.org/suppdata/ra/c3/c3ra46895j/c3ra46895j.pdf. https://www.rsc.org/suppdata/ra/c3/c3ra46895j/c3ra46895j.pdf
- Chathalingath N. Kingsly J. S. Gunasekar A. Microbiol. Res. 2023;267:127270. doi: 10.1016/j.micres.2022.127270. [DOI] [PubMed] [Google Scholar]
- Suresh V. Shams R. Dash K. K. Shaikh A. M. Kovács B. J. Agric. Food Res. 2025;20:101788. [Google Scholar]
- Hassanzadeh-Tabrizi S. A. J. Alloys Compd. 2023;968:171914. doi: 10.1016/j.jallcom.2023.171914. [DOI] [Google Scholar]
- Woodard L. N. Grunlan M. A. ACS Macro Lett. 2018;7:976–982. doi: 10.1021/acsmacrolett.8b00424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Göpferich A. Biomaterials. 1996;17:103–114. doi: 10.1016/0142-9612(96)85755-3. [DOI] [PubMed] [Google Scholar]
- Khatiwala V. K. Shekhar N. Aggarwal S. Mital U. K. J. Polym. Environ. 2008;16:61–67. doi: 10.1007/s10924-008-0104-9. [DOI] [Google Scholar]
- Luo C. J. Nangrejo M. Edirisinghe M. Polymer. 2010;51:1654–1662. doi: 10.1016/j.polymer.2010.01.031. [DOI] [Google Scholar]
- Alp B. Cesur S. J. Appl. Polym. Sci. 2013;130:1259–1275. doi: 10.1002/app.39217. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data generated or analysed during this study are included in this published article.
















