The challenge the researchers set out to address is fundamental to how direct current cables fail. In an HVDC transmission system, insulation must withstand hundreds of kilovolts across just a few centimeters of polymer. Under such intense fields, charge carriers are injected from the electrodes into the material, where they become trapped, migrate, and accumulate as space charge. This trapped charge distorts the local electric field, creating hot spots that can be far more intense than the applied field itself. Over years of service, these distortions accelerate aging, trigger partial discharges, and ultimately drive the insulation toward breakdown. Any material that hopes to serve as the next generation of cable insulation must therefore suppress charge injection, slow carrier mobility, and keep the internal field as uniform as possible.

Polypropylene has long been attractive for this role. It offers excellent electrical properties, high melting temperature, low cost, and, crucially, it is thermoplastic and fully recyclable, unlike cross-linked polyethylene, which is permanently cured and difficult to reprocess. Impact polypropylene copolymer, which blends a polypropylene matrix with an ethylene-propylene rubber phase for toughness, is a particularly promising variant. Yet until now, its optimization has been hampered by a lack of rigorous, quantitative correlations between structure and dielectric performance. Modifications were often evaluated only at the macroscopic level, leaving researchers to guess at the molecular and microscopic mechanisms responsible. The Tianjin team, led by You Wu, Zhonglei Li, and Boxue Du, set out to close that gap.

Their approach centered on a deliberately designed monomer. AOB packs three chemically distinct features into a single molecule: benzene rings, carbonyl groups, and ether bonds. Each of these contributes polarizable electrons and permanent dipoles, and together they create energy states within the polymer that can capture injected charge carriers. Using an aqueous suspension grafting method, the researchers covalently attached AOB onto the polypropylene chains, ensuring that the modifier would not simply migrate or bleed out of the insulation over decades of thermal and electrical stress, a known weakness of physically blended additives.

To understand what the grafted molecules were doing at the smallest scale, the team turned to density functional theory calculations and Kelvin probe force microscopy. The DFT simulations showed that the specific molecular architecture of AOB introduces deep traps for both electrons and holes, corresponding to the highest occupied and lowest unoccupied molecular orbitals of the modified system. These deep energy wells capture charge carriers that would otherwise roam freely through the polymer. The experimental signature came from KPFM measurements of surface potential decay: when the modified material was charged and then allowed to discharge, the surface potential lingered far longer than in the unmodified polymer. The decay time stretched from 3.9 seconds in the pristine material to 9.2 seconds after grafting, a clear indication that carriers were being held in place rather than moving and recombining rapidly.

Deep traps might sound like a liability, since trapped charge is precisely what forms space charge accumulation. The subtlety lies in where and how the charge is held. By capturing carriers close to where they are injected and immobilizing them in a dense distribution of shallow-to-deep trap sites, the grafted molecules prevent large, coherent packets of charge from building up deep inside the insulation. The net effect, confirmed at the macroscopic scale, was the dramatic 85 percent reduction in space charge accumulation. Slower carrier mobility also translated directly into lower measured conductivity, since conduction in these polymers is largely a matter of charge transport through amorphous regions and along interfaces.

The second half of the story unfolds at the mesoscale, in the polymer’s crystalline morphology. When AOB was grafted onto the material, it acted as a nucleating agent, prompting the formation of more numerous, smaller spherulites and increasing the overall crystalline density. This refinement matters because crystalline and amorphous regions of a semicrystalline polymer have different dielectric constants. Wherever these phases meet, the mismatch in permittivity causes local distortion of the electric field, concentrating stress at phase boundaries. Using finite element simulations built on a Voronoi algorithm to model the realistic geometry of the spherulitic structure, the researchers quantified this effect: the morphological changes induced by AOB reduced the dielectric mismatch between the crystalline and amorphous phases, cutting interfacial electric field distortion from 43 percent to 21 percent.

What distinguishes this study from much of the prior literature is how these scales were connected. Rather than treating molecular, microscopic, and macroscopic observations as separate narratives, the team first established direct quantitative correlations within each specific scale and only then integrated them into a comprehensive model. The deep traps characterized at the molecular scale explain the suppressed injection and mobility seen in surface potential measurements; the refined spherulitic morphology measured at the mesoscale explains the homogenized internal field; and the combination of these two mechanisms accounts for the macroscopic outcomes. The synergy is the point. Trap capture alone would slow charge but leave field distortions intact, while morphology refinement alone would homogenize the field but do nothing about the sheer quantity of injected charge. Together, they attack the failure cascade from both ends.

The macroscopic results follow logically from that dual mechanism. With fewer carriers injected and those carriers immobilized, space charge accumulation collapsed by 85 percent. With carrier transport suppressed throughout the bulk, conductivity dropped by an order of magnitude, which in turn reduces Joule heating and thermal aging of the insulation. And with the internal field distribution smoothed by the refined crystalline structure, the material withstood a direct current breakdown strength 30 percent higher than the unmodified polymer before failure. Each of these metrics is directly relevant to cable design, where higher breakdown strength allows thinner insulation walls, and lower conductivity and space charge translate into longer service life and higher permissible operating temperatures and voltages.

Beyond the specific numbers, the study offers a template for rational materials design in electrical insulation. Instead of screening additives empirically, the framework demonstrated here, from molecular orbital engineering through trap spectroscopy and morphological modeling to macroscopic dielectric testing, provides a quantitative theoretical basis for predicting how a given chemical modification will perform. As grids worldwide expand high-voltage direct current transmission to move renewable electricity over long distances, the demand for recyclable, high-performance insulation will only grow. A modified polypropylene that can be grafted, characterized, and optimized with this level of mechanistic clarity brings that future measurably closer, and suggests that the next generation of power cables may be built on plastics engineered one molecular feature at a time.

Subject of Research: Multiscale structural modulation of grafted polypropylene for high-voltage direct current cable insulation

Article Title: Enhanced dielectric performance of polypropylene for hvdc cable insulation via 4‑allyloxybenzil grafting for multiscale structural modulation

Article References: Wu, Y., Li, Z., Wang, H., Cao, G., Zheng, Z., Du, H., & Du, B. (2026). Enhanced dielectric performance of polypropylene for hvdc cable insulation via 4‑allyloxybenzil grafting for multiscale structural modulation. Advanced Composites and Hybrid Materials. https://doi.org/10.1007/s42114-026-02115-9

Image Credits: AI Generated

DOI: 10.1007/s42114-026-02115-9

Keywords: polypropylene, HVDC cable insulation, dielectric properties, 4-allyloxybenzil, chemical grafting, space charge, deep traps, spherulites, breakdown strength, density functional theory, Kelvin probe force microscopy, multiscale modeling

Rubrene has long been a benchmark material in the organic electronics community. In single-crystal form it can carry charges at mobilities of up to 20 square centimeters per volt per second, rivalling some amorphous inorganic semiconductors. The catch is that those spectacular numbers come from painstakingly grown individual crystals, not from the thin films that real devices need. When rubrene is evaporated directly onto silicon dioxide, the standard insulator of transistor technology, thermal annealing rarely produces uniform, large-area crystalline domains. The result is a patchwork of grains with wildly varying quality, and devices whose performance cannot be trusted from batch to batch. Reproducibility, not raw mobility, has become the bottleneck for commercialization.

The Korean team’s strategy was to insert an ultrathin polystyrene interlayer between the silicon dioxide and the rubrene, and then to engineer that interlayer’s chemistry with remarkable precision. They prepared three versions of the polymer surface. The first was plain, as-cast polystyrene. The second, which they call PS-x, was irradiated for twenty minutes with 365-nanometer ultraviolet light in ambient air. The third, dubbed PS-net, received the same UV dose and was then rinsed briefly in isopropyl alcohol. Twenty nanometers of rubrene were subsequently evaporated onto each surface under high vacuum, and the films were annealed in a nitrogen glovebox.

The ultraviolet treatment sounds crude, but it triggers a rich photochemistry. UV photons cleave carbon-hydrogen bonds in the polystyrene backbone, generating carbon radicals that react with oxygen in the air to form peroxy species. Some of these radicals link neighboring polymer chains together into a crosslinked network, while others drive the opposite process, beta-scission and ring-opening reactions that chop the polymer into small, polar, low-molecular-weight fragments. Fourier-transform infrared spectroscopy confirmed the chemical transformation: the characteristic benzene-ring and carbon-hydrogen absorption bands weakened after irradiation, while new carbonyl signals appeared between 1700 and 1850 per centimeter, marking the increased polarity of the modified surface.

The decisive evidence for crosslinking came from a clever solubility test. The researchers patterned UV light onto selected regions of a polystyrene film and then rinsed the whole sample in isopropyl alcohol, a solvent that normally cannot touch pristine polystyrene. Scanning electron microscopy showed that only the irradiated regions survived the rinse. Crosslinking had rendered the exposed polymer insoluble, while the untouched areas dissolved away. The subsequent alcohol wash on the PS-net samples serves a second, subtler purpose: it strips away the soluble chain-scission fragments, leaving behind a cleaner, purely crosslinked network at the surface.

When rubrene was deposited and annealed on these different surfaces, the contrast was dramatic. On bare silicon dioxide and on untreated polystyrene, no crystalline domains appeared at all, even after annealing. On PS-x, spherulitic crystals formed only partially and unevenly. But on PS-net, the film blossomed into highly uniform spherulitic domains that swept across the entire surface. Polarized optical microscopy, which exploits the birefringence of rubrene crystals, revealed a classic nucleation-and-growth mechanism: new nuclei appeared during annealing, expanded radially, and after roughly sixty seconds the whole film had crystallized. Prolonged annealing beyond that point caused no further ripening, indicating the crystals had settled into a deeply stable energy state, with the final grain size fixed by the early nucleation density.

The speed is the headline. Previous polymer-assisted approaches to rubrene crystallization were far slower. One study by Bronshtein, Rybtchinski and colleagues needed more than 300 seconds of rapid thermal processing under high vacuum with argon flow to grow single crystals from a rubrene-polystyrene blend. Another group required roughly ten minutes of solvent vapor annealing in chloroform. The new interlayer method achieves uniform, large-area crystal growth in just sixty to ninety seconds, a tenfold or better acceleration that matters enormously for manufacturing throughput and, crucially, avoids the solvent-blending route whose crystal growth is notoriously sensitive to solvent choice and ambient conditions.

Temperature proved to be a razor-thin dial. Below 130 degrees Celsius, nothing crystallized; the thermal energy was simply insufficient to mobilize the rubrene molecules. At 140 degrees, small domains emerged that previous work assigns to the triclinic crystal phase. Between 150 and 160 degrees, triclinic regions gradually gave way to orthorhombic domains, the crystal structure favored for charge transport. Only in the narrow window of 170 to 180 degrees did large orthorhombic spherulites blanket the entire film, with triclinic domains vanishing entirely. That window, consistent with earlier findings by Holmes and Fielitz, is punishingly tight, and it underscores how little slack process engineers have when optimizing organic transistor fabrication.

What makes the crosslinked interlayer so special? Surface measurements offered clues but no complete answer. Atomic force microscopy showed all three polymer films were remarkably smooth, with roughness below one nanometer, though PS-net was slightly rougher at 0.804 nanometers than its counterparts at roughly 0.49 nanometers, perhaps aiding nucleation. Contact-angle measurements traced a rising surface energy, from 27.5 millijoules per square meter for pristine polystyrene to 38.5 for PS-x and 44.1 for PS-net, the latter closely matching rubrene’s own surface energy. Yet the energy-matching story collapsed as a full explanation: silicon dioxide and PMMA, with even higher surface energies, supported no polycrystalline growth, while HMDS-treated silicon dioxide, whose surface energy resembles PS-net, also failed. Something beyond wetting physics was at work.

The missing ingredient, the authors argue, is the thermal dynamics of the interlayer itself. Differential scanning calorimetry showed a clear glass transition for pristine polystyrene but none for the UV-modified films, a signature of restricted segmental motion caused by crosslinking. In the 170-to-180-degree window, the crosslinked network apparently provides just enough surface mobility for rubrene molecules to diffuse and reorder into orthorhombic crystals. Too rigid a surface, like HMDS-treated silicon dioxide, blocks diffusion; too mobile a surface, like linear polystyrene, lets the polymer chains churn and disrupt ordering. The residual chain-scission fragments in unrinsed PS-x likely interfere with long-range molecular ordering, echoing earlier pentacene studies in which shortened polystyrene chains suppressed surface diffusion. The researchers caution that this mechanistic picture remains partly speculative, but the practical recipe is clear: remove the unstable fragments, tune the crosslinking, and the polymer underlayer becomes a template that guides small molecules into the crystal phase devices need. For organic field-effect transistors, where structural uniformity and batch-to-batch reproducibility are prerequisites for any commercial future, a sixty-second annealing step on a UV-crosslinked plastic film may prove to be exactly the kind of unglamorous enabling technology the field has been waiting for.

Subject of Research: Interfacial engineering of polymer interlayers to control rubrene thin-film crystallization for organic electronics

Article Title: Interfacial engineering of rubrene thin films with crosslinked polymer layers

Article References: Im, J., & Choi, H. H. (2025). Interfacial engineering of rubrene thin films with crosslinked polymer layers. Advances in Industrial and Engineering Chemistry, 1(1), Article 20. https://doi.org/10.1007/s44405-025-00023-y

Image Credits: AI Generated

DOI: 10.1007/s44405-025-00023-y

Keywords: rubrene, organic semiconductors, polystyrene, crosslinking, spherulites, thin films, crystallization, surface energy, UV irradiation, organic field-effect transistors, polymer interlayers, thermal annealing