Mechanism of Helical Nanofiber Formation via Co‑Electrospinning of Cellulose Acetate and Thermoplastic Polyurethane

Background

Helical nanostructures are prized for their high surface‑to‑volume ratio and intrinsic porosity, making them attractive for sensors, filtration media, oil sorbents, and photovoltaic devices. The addition of a helical geometry to micro‑ or nanofibres enhances mechanical resilience and creates a three‑dimensional, porous mat that can be tailored for specific performance requirements. Natural examples – from plant tendrils to wool fibres – illustrate how differential shrinkage or extension during growth can produce tight helices, a principle that has inspired synthetic fabrication strategies.

Co‑electrospinning has emerged as a simple yet powerful route to generate composite fibres with complex morphologies. Unlike techniques such as chemical vapour deposition, sol‑gel processing, or hydrothermal synthesis, co‑electrospinning allows two polymers to be spun simultaneously, yielding core‑shell or side‑by‑side architectures that can evolve into helical shapes. Prior work has demonstrated helical nanofibres from combinations such as PAN/TPU, Nomex/TPU, and PU/Nomex, highlighting the importance of polymer stiffness, miscibility, and hydrogen bonding in driving curvature.

While previous studies have shown that Nomex/TPU can produce fine helices, the non‑hydrophilic nature of Nomex limits its applicability in biomedical or adsorption contexts. Our work therefore focuses on a CA/TPU system, where CA provides rigidity and TPU offers elastomeric flexibility, to explore how solution chemistry and interfacial forces govern helix formation.

Experimental

Materials

Cellulose acetate (CA, Mw ≈ 100 kDa) was purchased from Acros Organics. Thermoplastic polyurethane (TPU, Desmopan DP 2590A) was obtained from Bayer Materials Science. Solvents – N,N‑dimethylacetamide (DMAc), acetone, tetrahydrofuran (THF) – and lithium chloride (LiCl) were supplied by Shanghai Chemical Reagents Co., Ltd. All reagents were used as received. Experiments were conducted at ~25 °C and 40–50 % RH.

Co‑electrospinning

CA solutions of 14 wt % were prepared by dissolving CA and LiCl in DMAc/acetone (1:2) and stirring for 5 h at room temperature. TPU solutions (18 wt %) were dissolved in either DMAc/THF (3:1; TPU1) or DMAc/acetone (3:1; TPU2). A custom off‑centered co‑electrospinning setup (Figure 1) used syringes and pumps to deliver core and shell solutions to a high‑voltage supply (15–20 kV). The collector rotated at 14.24 cm s⁻¹. The core‑shell configuration mimics plant tendril formation, where the core component contracts more rapidly than the shell, driving helix curvature.

a Schematic of the off‑centered co‑electrospinning system. b Formation mechanism of helical nanofibres.

Characterisations

Fiber Morphology

Resultant fibres were examined by SEM (JSM‑5600LV) after gold sputtering.

DSC

Glass transition temperatures were measured on a DSC‑4000 under nitrogen (scan rate 10 °C min⁻¹, −80 → 300 °C). Samples (5–10 mg) were rapidly cooled to −80 °C after the first melt scan.

FTIR

Infrared spectra were recorded on a Bruker Vector 33 FTIR spectrometer (32 scans, 1 cm⁻¹ resolution). Films were thin enough to satisfy the Beer‑Lambert law. Temperature‑dependent spectra were obtained using a temperature‑controlled cell.

Experimental Results

We first optimized single‑spinning conditions for CA and TPU, then explored two co‑electrospinning systems: CA/TPU1 and CA/TPU2. Figures 2 and 3 illustrate the impact of solution concentration and solvent composition on fibre morphology.

SEM images of CA single‑spinning at varying concentrations (x‑axis) and DMAc/acetone ratios (y‑axis). Applied voltage: 15 kV; working distance: 10 cm; flow rate: 0.15 ml h⁻¹.

SEM of single‑spinning TPU1 (a) and TPU2 (b) at 18 wt % (15 kV, 10 cm, 0.15 ml h⁻¹). TPU2 shows reduced fibre adhesion relative to TPU1.

Co‑electrospinning experiments varied LiCl content in the CA shell and employed either TPU1 or TPU2 as the core. Figure 4 shows that without LiCl, no helices form. At 0.5–1 wt % LiCl, CA/TPU2 fibres begin to bend, while CA/TPU1 remains largely straight or forms bundles. Increasing LiCl to 2 wt % produces many helices in CA/TPU2 but compromises fibre uniformity due to excessive conductivity.

SEM of CA/TPU1 and CA/TPU2 at 0–2 wt % LiCl (20 kV, 15 cm, 0.15 ml h⁻¹).

Across all trials, CA/TPU2 consistently produced a higher density of helices than CA/TPU1, underscoring the importance of solvent compatibility and interfacial tension.

Results and Discussion

Our analysis focuses on three interrelated factors: solution chemistry, hydrogen‑bonding, and miscibility. Together, they explain why the CA/TPU2 system favours helical formation.

Mechanism of CA/TPU Helical Fibres

LiCl plays a dual role. It increases solution conductivity, enhancing jet stability, and it interacts with CA chains via chloride‑hydrogen bonding, aligning the rigid CA backbone and increasing its stiffness relative to TPU. This differential stiffness generates longitudinal stress during jet elongation, promoting a coiled morphology. Figure 5 illustrates the proposed LiCl‑mediated dissolution of CA in DMAc.

Proposed mechanism for CA dissolution in DMAc/LiCl: (a) molecular formula; (b) 3‑D structure.

Solution Properties

Co‑electrospinning demands a delicate balance of viscosity, vapor pressure, and interfacial tension. TPU2’s lower surface tension (25.34 N m⁻¹) and compatible solvent mix with CA reduce interfacial tension, preventing fibre collapse and enabling smooth core‑shell integration. In contrast, TPU1’s higher surface tension (34.45 N m⁻¹) and solvent mismatch with CA lead to fibre adhesion and poor helix formation.

Hydrogen Bonding in Blends

FTIR spectra (Figure 6) reveal a pronounced 3400 cm⁻¹ band in the CA/TPU blend, indicative of –NH···O hydrogen bonds between TPU and CA. This interaction not only stabilises the core‑shell interface but also enhances viscous drag during jet stretching, encouraging the two components to intertwine into helices.

FTIR of pure CA, TPU, and CA/TPU blends: (a) TPU; (b) CA; (c) CA/TPU.

Miscibility Behaviour

DSC thermograms (Figure 8) show two distinct glass‑transition temperatures for the CA/TPU blend, positioned between the individual polymers’ Tg values (−31.24 °C for TPU, 193.74 °C for CA). This intermediate behaviour confirms partial miscibility, which amplifies interfacial stresses when the core and shell experience differential shrinkage, further driving helix formation.

DSC of TPU, CA, and CA/TPU blends: (a) TPU; (b) CA; (c) CA/TPU.

Conclusions

The CA/TPU2 co‑electrospinning system, leveraging solvent‑matched interfacial tension and LiCl‑mediated conductivity, reliably produces helical nanofibres. Enhanced hydrogen bonding and partial miscibility between CA and TPU generate strong interfacial stresses, while the large stiffness differential promotes curvature. These findings elucidate the mechanistic underpinnings of CA/TPU helix formation and broaden the range of materials suitable for advanced helical fibre applications.

Abbreviations

CA

Cellulose acetate

DMAc

N,N‑dimethylacetamide

DSC

Differential scanning calorimetry

HSPET

Poly(ethylene glycol terephthalate)

LiCl

Lithium chloride

Nomex

Poly(m‑phenylene isophthalamide)

PAN

Polyacrylonitrile

PS

Polystyrene

PTT

Poly(ethylene propanediol terephthalate)

PU

Polyurethane

THF

Tetrahydrofuran

TPU

Thermoplastic polyurethane

TPU1

TPU dissolved in DMAc/THF (3:1)

TPU2

TPU dissolved in DMAc/acetone (3:1)

FTIR

Fourier transform infrared spectroscopy