The review, led by Hajar Naemah Sohif and colleagues, takes aim at a persistent gap in the literature. While previous surveys have catalogued TPU printing parameters, modification strategies and end-use applications, the authors contend that the connections between what goes into a filament formulation, how the material’s microstructure develops during printing, and what properties ultimately emerge from the nozzle have remained insufficiently integrated. Their proposed solution is a processing-microstructure-property framework, a conceptual map that links material formulation and processing conditions to phase morphology, crystallinity, interlayer adhesion and filler dispersion, and ultimately to the mechanical and functional performance of the finished part.

Understanding why this framework matters requires a brief tour of TPU’s inner architecture. Thermoplastic polyurethanes are block copolymers composed of alternating soft segments, which confer elasticity, and hard segments, which act as physical crosslinks and provide strength. The relative proportions and degree of ordering of these segments, a phenomenon known as microphase separation, governs much of the material’s mechanical behaviour. Studies cited in the review have visualised and quantified how varying hard segment content changes microphase separation and, with it, tensile performance. This internal structure, however, is only part of the story; in fused deposition modeling and related extrusion-based processes, the way molten layers bond to one another and how fillers distribute themselves within the melt can make or break the final part.

The most heavily explored strategy for upgrading TPU is polymer blending, and the review devotes considerable attention to three partner polymers in particular: polylactic acid, polycaprolactone and acrylonitrile butadiene styrene. Blending TPU with PLA, a rigid and bio-based polymer, can dramatically increase stiffness and has been shown to improve fracture toughness in ways that depend strongly on the blend ratio. PLA/TPU blends have also demonstrated shape memory behaviour, opening the door to eco-friendly filaments that can be programmed into temporary shapes and recover their original form on heating. Researchers have even prepared PLA-TPU copolymers via in situ ring-opening polymerization to improve compatibility between the two otherwise immiscible phases, a reminder that in polymer blending, chemistry at the interface is often as important as the bulk composition.

Polycaprolactone, by contrast, brings a different set of talents to the table. With its low melting temperature and biocompatibility, PCL has been blended with TPU to create shape memory materials aimed at biomedical use, including self-knotting sutures that tighten when warmed to body temperature. Composites of TPU/PCL blends with hydroxyapatite have been explored for bone-related applications, combining the elastomer’s compliance with the ceramic’s bioactivity. ABS, the workhorse of consumer 3D printing, offers yet another route: ABS/TPU blends balance the stiffness and dimensional stability of ABS with the toughness and flexibility of TPU, and studies have shown that blend composition directly affects printability, weld line quality and mechanical performance. Carbon nanotube-modified ABS/TPU nanocomposites have further demonstrated improvements in fluidity and fracture behaviour.

Blending, however, is not a free lunch. The review is candid about the pitfalls. Most polymer pairs, including TPU with PLA or ABS, are thermodynamically immiscible, meaning they tend to separate into distinct phases rather than mixing at the molecular level. The resulting phase morphology, whether droplets, fibers or co-continuous networks, depends on composition, viscosity ratio and processing history, and it exerts decisive control over properties. Poor interfacial adhesion between phases can create weak points that crack propagation exploits. This is where compatibilizers enter the picture: molecules designed to sit at the interface, reduce interfacial tension and anchor the two phases together. The review highlights compatibilization as a critical lever, noting that well-chosen compatibilizers can transform a brittle, phase-separated blend into a material with genuinely useful combinations of stiffness and toughness.

Additives beyond compatibilizers round out the formulation toolbox. Fillers of various dimensions, from nanoparticles such as silica-coated titanium dioxide and zinc oxide to carbon nanotubes, graphene, boron nitride and short carbon fibers, can impart electrical conductivity, thermal conductivity, flame retardancy or reinforcement. Conductive TPU composites loaded with multiwalled carbon nanotubes have been 3D printed into stretchable strain sensors, while graphene-filled TPU filaments produced by melt compounding show enhanced thermal conductivity for thermal management applications. The catch, the authors stress, is dispersion: filler agglomeration is a recurring failure mode that undermines both mechanical performance and functionality, and filler size itself matters, as demonstrated by studies of boron nitride particles in flexible thermally conductive TPU composites. Plasticizers, meanwhile, soften the material and modify the microphase-separated morphology, with documented effects on the electromechanical properties of TPU films.

The practical payoff of all this formulation science is already visible across four application sectors the review identifies. In medicine, TPU-based filaments are used for tissue engineering scaffolds, tracheal constructs, surgical planning models and shape memory devices, where biocompatibility and tunable mechanics are paramount. In electronics and sensors, stretchable conductive TPU composites enable wearable and flexible devices that survive repeated deformation. Non-pneumatic tires, which replace air with engineered lattice structures, exploit TPU’s energy absorption and rebound behaviour, with 3D-printed lattice architectures offering designers unprecedented control over cushioning and durability. Textiles and sports equipment round out the list, with flexible printed TPU finding roles in functional garments, protective gear and auxetic structures designed for energy absorption.

Looking ahead, the review identifies sustainability as the defining frontier. Bio-based TPUs derived from renewable feedstocks, including polylactic acid-derived polyols, are advancing rapidly, and fully bio-based composites with enhanced flame retardancy have been demonstrated using additives such as phytic acid. Recyclability is another theme, with researchers developing polyurethane elastomers that can be reprocessed without catastrophic loss of mechanical properties, and studies of polymer blend recycling offering guidance for circular material flows. Perhaps most intriguingly, the authors point toward multi-material and 4D printing, in which printed objects change shape over time in response to environmental stimuli. Shape memory TPU formulations, hybrid composites combining shape memory polymers with thermoplastic elastomers, and programmable multi-material structures all suggest a future in which printed TPU components are not merely flexible, but actively responsive.

What emerges from the Malaysian team’s synthesis is a discipline coming of age. The era of treating TPU as a single, fixed filament grade is giving way to a design-space approach in which chemists, processors and printers jointly tune soft and hard segment chemistry, blend ratios, compatibilizer architecture, filler type and loading, and printing parameters as one integrated system. The review’s framework, connecting formulation and processing to microstructure and final properties, offers researchers a common language for that effort. If the authors are right, the humble spool of flexible filament on a workbench today is merely the starting point for a generation of smart, high-performance and sustainable elastomeric parts, printed layer by layer and engineered from the molecule up.

Subject of Research: Tailoring the mechanical and functional properties of thermoplastic polyurethane filaments for additive manufacturing through blends and additives.

Article Title: Tailoring mechanical and functional properties of Thermoplastic Polyurethane (TPU) filaments for additive manufacturing: A review of blends and additives

Article References: Sohif, H. N., Manshor, H., Hatta, M. A. M., & Sajab, M. S. (2026). Tailoring mechanical and functional properties of Thermoplastic Polyurethane (TPU) filaments for additive manufacturing: A review of blends and additives. Polymer Bulletin, 83(12), Article 681. https://doi.org/10.1007/s00289-026-06739-1

Image Credits: AI Generated

DOI: 10.1007/s00289-026-06739-1

Keywords: thermoplastic polyurethane, TPU, additive manufacturing, 3D printing, polymer blends, PLA, PCL, ABS, fillers, compatibilizers, shape memory polymers, sustainability

The research, led by Chloé Melin of the École de Technologie Supérieure in Montreal and INSA-Lyon, together with Jean-Fabien Capsal, Ricardo Zednik, Angelo Pommella, Nicole Demarquette, and Jean-Marc Chenal, set out to answer a question that has lingered in the literature for years: does PLA actually help PVDF become a better piezoelectric material, or does it merely look like it does on a spectroscopy plot? The team attacked the problem from two processing directions at once. They melt-blended PVDF with PLA at weight ratios of 95/05 and 60/40 using a twin-screw extruder at 210 degrees Celsius, then compression-molded films and stretched them uniaxially at draw ratio 3 across a range of temperatures. In parallel, they prepared the same blend compositions as electrospun fibrous membranes using a semi-industrial twenty-needle unit, with both polymers dissolved together in a dimethyl sulfoxide and acetone mixture.

The central character in this story is the crystal structure of PVDF. The polymer can solidify into several distinct lattice forms, and they are not equal partners. The alpha phase, which forms readily when PVDF cools from the melt, is non-polar and piezoelectrically useless. The beta phase, in which polymer chains adopt an all-trans conformation, is highly polar and delivers the strongest piezoelectric response when the molecular dipoles are aligned by a high-voltage poling treatment. Converting alpha to beta is therefore the name of the game, and the two classic routes are uniaxial stretching of solid films and electrospinning, in which a jet of polymer solution is whipped by intense electric fields into nanometer- and micrometer-scale fibers, combining extreme elongational deformation with rapid solvent evaporation.

Morphology turned out to be the first fork in the road. In the melt-processed blends, scanning electron microscopy revealed that the 95/05 formulation produced a beaded, droplet-in-matrix structure, with discrete PLA domains dispersed in the PVDF, while the 60/40 blend formed a co-continuous architecture in which both polymers formed interpenetrating networks. Differential scanning calorimetry showed that adding PLA left the overall crystallinity of PVDF essentially unchanged at roughly 42 percent, before and after stretching, a consequence of the micrometric PLA domain size and the fact that PVDF crystallizes at a higher temperature than PLA. But the phase composition told a very different story. In the stretched 95/05 blend, the beta-phase fraction matched that of pure PVDF, whereas in the co-continuous 60/40 blend the beta fraction collapsed to 36 percent from the 86 percent achievable in well-processed samples, leaving the material dominated by the inert alpha phase.

The reason for that collapse is a lesson in mechanics. In a co-continuous blend, the macroscopic deformation applied during stretching is shared between the two continuous polymer networks, so the PVDF phase simply never sees enough local stress to drive the alpha-to-beta conformational transformation. Meanwhile, in the beaded 95/05 blend, the team observed cavitation, tiny voids opening at the PVDF-PLA interfaces during stretching, which relaxed local stresses and prevented the hoped-for enhancement of stress transfer to PVDF crystals. Earlier hypotheses in the literature had suggested that PLA could act as a stress concentrator, boosting beta formation through cavitation and heterogeneous deformation. This study’s data call that idea sharply into question: the beta fraction in the stretched blend was no better than in neat PVDF, and in some conditions worse.

Then came the electrical measurements, and the results were even more striking. After poling at 100 volts per micrometer for one hour in dielectric oil, the stretched 95/05 blend showed a piezoelectric coefficient d33 that was 34 to 45 percent lower than pure PVDF stretched under equivalent conditions, despite having a nearly identical beta-phase fraction. The culprit, the researchers showed, is dielectric mismatch. PVDF has a relative permittivity of about 10.8, while PLA’s is only 2.7. During poling, electric field lines preferentially route through the low-permittivity PLA inclusions and through the air-filled cavities left by stretching-induced voids, starving the surrounding PVDF matrix of the field it needs to rotate its dipoles. Finite element simulations in COMSOL Multiphysics, built on a representative volume element containing a PLA inclusion surrounded by an air ellipse, reproduced exactly this field concentration. When the team measured the piezoelectric response of pure PVDF poled at the reduced effective field of roughly 65 volts per micrometer, they obtained a d33 of about 6 piconewtons per coulomb, in close agreement with the blend’s measured value. The physics, in other words, checks out.

The co-continuous 60/40 blend, being overwhelmingly alpha phase, registered no measurable piezoelectric response at all, a d33 of zero. But the electrospinning results added their own twist. Pure electrospun PVDF membranes outperformed their stretched-film counterparts, delivering a d33 roughly 30 percent higher, thanks to about 29 percent greater overall beta-phase content and the superior chain alignment imparted by the enormous elongational forces inside the spinning jet. Interestingly, electrospinning also introduced a significant gamma-phase fraction, an intermediate polar conformation rarely seen in the stretched films, which the authors attribute to the rapid solvent evaporation and extreme deformation rates preserving the initially beta-nucleated chains only partially, letting them relax toward the gamma conformation rather than all the way to alpha.

Yet even electrospinning could not rescue the blend. The electrospun PVDF/PLA 95/05 membranes, despite achieving an electroactive phase fraction of about 86 percent, essentially identical to neat electrospun PVDF, showed a d33 of exactly zero after poling. The team points to a convergence of factors: the finely dispersed, highly elongated PLA domains formed during fiber formation disrupt the local electric field and hinder chain and lamellar orientation; the vastly increased interfacial area places PLA, which is stiffer than amorphous PVDF at the poling temperature, immediately adjacent to the crystal-amorphous interfaces where dipole switching initiates; and interfacial interactions locally stiffen the material, impeding the conformational kinks that must propagate through PVDF chains during poling. Raising the poling temperature triggered electrical breakdown, and lowering the field produced no effective dipole alignment, leaving the membranes piezoelectrically mute.

The study also challenges another widespread assumption. None of the electrospun membranes in this work showed measurable piezoelectric activity before an external poling step, even though the poling field used was nearly twice that of most previous studies reporting self-poled electrospun PVDF. The through-thickness dipole orientation required for a d33 response, the authors note, evidently does not emerge from the electrospinning process alone under these conditions, contradicting a body of literature that has treated self-poling as an intrinsic benefit of the technique.

The broader takeaway is a warning against a seductive shortcut. In the drive toward sustainable, flexible sensors and energy harvesters, blending PVDF with biobased PLA seems like an obvious win: PLA is renewable, biodegradable, stiff, and easy to process. But this work demonstrates that a high beta-phase fraction alone does not guarantee functional piezoelectricity. Electric field distribution, interfacial cavitation, domain size, chain mobility, and processing route all conspire to determine whether a material that looks piezoelectric under an infrared spectrometer actually produces a measurable voltage under stress. For engineers designing the next generation of self-powered wearables and structural health monitors, the message is clear: composition, morphology, and processing must be engineered together, and the role of PLA in enhancing the beta phase of PVDF, once considered promising, is now very much in doubt.

Subject of Research: Piezoelectric performance of PVDF/PLA polymer blends prepared by melt processing and electrospinning

Article Title: Piezoelectric properties of PVDF/PLA blends prepared by melt processing and electrospinning

Article References: Melin, C., Capsal, J.-F., Zednik, R., Pommella, A., Demarquette, N., & Chenal, J.-M. (2026). Piezoelectric properties of PVDF/PLA blends prepared by melt processing and electrospinning. Journal of Materials Science: Polymers, 1(1), Article 13. https://doi.org/10.1007/s44493-026-00010-9

Image Credits: AI Generated

DOI: 10.1007/s44493-026-00010-9

Keywords: PVDF, PLA, piezoelectricity, polymer blends, electrospinning, melt processing, beta phase, energy harvesting, flexible sensors, poling, crystallinity, dielectric permittivity