The research team, led by Prayitno with Sri Rulianah, Wahyuni Ningsih, and Wahyu Widyananda, all affiliated with Indonesian institutions collaborating through the State Polytechnic of Malang, set out to build a quantitative model of thermal demulsification, the process by which emulsified oil droplets are coaxed into merging and separating from water. Their wastewater came directly from a turpentine plant in Trenggalek, where the production of turpentine oil and gum rosin leaves behind a cocktail of cellulose fibers, resin acids, carboxylic acids, and dissolved turpentine oil. These natural surfactants form rigid interfacial films around oil droplets, creating emulsions so stable that conventional treatment methods like flotation, neutralization, and sedimentation struggle to meet regulatory thresholds, particularly for oil and grease.

The experimental setup was elegantly simple. Two liters of filtered wastewater, held at its naturally acidic pH of 1 to 2, was placed in a beaker inside a temperature-controlled water bath. A commercial polymeric demulsifier, Nalco 14750, was dosed at 50, 100, or 150 milligrams per liter, while heating temperatures ranged from 30 to 50 degrees Celsius. Half the experiments ran with vigorous agitation at 5,000 revolutions per minute; the other half proceeded in complete stillness. Each five-minute treatment was then assessed for four regulated parameters: turbidity, total suspended solids, biochemical oxygen demand, and oil and grease. The removal data were fitted to second-order polynomial regression models, producing response surfaces that map exactly how temperature and dose interact to drive pollutant removal.

The mechanism underlying the process reads like a microscopic tug-of-war at the oil-water interface. The non-ionic polymeric demulsifier carries both hydrophilic and hydrophobic segments, allowing it to migrate to the droplet surface, penetrate the rigid film formed by gum rosin and carboxylic acids, and displace the natural emulsifiers stabilizing the emulsion. Meanwhile, moderate heat thins the interfacial film, lowers the zeta potential of the droplets, compresses the electrical double layer, and reduces the viscosity of the surrounding water, all effects that encourage droplets to collide, coalesce, and rise. The wastewater’s extreme acidity adds another lever: at pH 1 to 2, weakly acidic groups on the natural emulsifiers become protonated, reducing the negative surface charge on droplets and weakening electrostatic repulsion before the demulsifier even arrives.

Under agitation, the best single result came at 100 milligrams per liter of demulsifier, where turbidity fell by 92.21 percent, total suspended solids by 96.36 percent, biochemical oxygen demand by 99.84 percent, and oil and grease by an impressive 99.29 percent. But the response surfaces revealed a clear ceiling: temperatures above roughly 40 degrees Celsius combined with high-speed stirring actively sabotaged the process. The researchers traced this deterioration to well-established fluid dynamics. At 5,000 rpm, raising the temperature lowers the water’s viscosity, pushing the impeller Reynolds number higher and intensifying turbulence. According to Kolmogorov-Hinze theory, greater turbulent energy dissipation shrinks the maximum stable droplet diameter, meaning the stirrer begins slicing coalesced oil droplets back into tiny, newly stabilized fragments faster than the demulsifier can disrupt their regenerated interfacial films. In effect, the mixer re-emulsifies what chemistry has just separated.

The quiescent experiments told a strikingly different story. Without agitation, oil and grease removal ranged from 98.05 to 99.49 percent, consistently outperforming the stirred condition at comparable temperature and dose combinations, with a strong model fit of R-squared equal to 0.9312. Once the demulsifier and moderate heat had destabilized the emulsion, undisturbed conditions allowed the coalesced droplets to rise continuously to the surface, exactly as Stokes’ law predicts for creaming in a low-viscosity, quiescent fluid. Total suspended solids removal still reached up to 96.11 percent, aided by a slightly higher optimum temperature of around 43 degrees Celsius that compensated for the absence of mechanical collision energy. Turbidity removal peaked near 32 degrees Celsius at doses of 100 to 113 milligrams per liter.

The models also exposed a subtle danger in overdosing. Beyond roughly 100 milligrams per liter, removal efficiencies for oil and grease and biochemical oxygen demand began to falter. The explanation lies in surfactant physics: when the demulsifier concentration approaches the critical micelle concentration, excess molecules spontaneously assemble into micelles, spherical structures with oily interiors that encapsulate hydrocarbons and keep them dispersed in the water phase. Simultaneously, surplus surfactant adsorbing onto droplet surfaces enhances steric stabilization, blocking the very collisions the treatment is meant to promote. The result is a secondary, self-inflicted emulsion that resists settling, flotation, and even analytical detection, since standard methods count micelle-trapped oil as part of the oil and grease load.

To distill all four responses into a single operating recipe, the team applied the Derringer-Suich composite desirability method, treating every parameter as a larger-the-better characteristic and weighting them equally because each is independently regulated. Evaluated over a grid of 90,000 points across the experimental domain, the optimization delivered an unambiguous verdict. Without agitation, the overall desirability reached 0.8258 at approximately 35.2 degrees Celsius and 150 milligrams per liter of demulsifier, with the predicted model delivering oil and grease removal above 99.4 percent, total suspended solids removal above 96 percent, turbidity removal above 93 percent, and near-complete biochemical oxygen demand removal above 99.7 percent. The stirred condition managed only 0.7264, with a sharper, more fragile optimum that would make industrial operation far less forgiving of small deviations in temperature or dosing.

The practical implications extend well beyond one factory in Trenggalek. Skipping agitation eliminates the energy cost of continuous mixing, removes the capital expense of mixer installations, and cuts long-term maintenance for full-scale treatment plants. Compared with the electrocoagulation-Fenton process the same group previously tested, which achieved 99 percent oil and grease removal but demanded constant electrical input, consumable electrodes, and complex sludge handling, moderate thermal-chemical demulsification in a still tank is dramatically simpler. The authors note that the optimal dose of 150 milligrams per liter sits at the boundary of their tested range, meaning the true global optimum may lie beyond it, and they call for future work on demulsifier type, heating duration, and agitation speed. For now, the message to engineers battling oily industrial wastewater is refreshingly concise: dose it, warm it to about 35 degrees Celsius, and then let physics do the quiet work of separation.

Subject of Research: Thermal demulsification modeling of turpentine industrial wastewater using a polymeric demulsifier with multi-response optimization

Article Title: Analysis model of the demulsification process by heating in turpentine industrial wastewater

Article References: Prayitno, Rulianah, S., Ningsih, W., & Widyananda, W. (2026). Analysis model of the demulsification process by heating in turpentine industrial wastewater. Case Studies in Chemical and Environmental Engineering, 14, Article 101477. https://doi.org/10.1016/j.cscee.2026.101477

Image Credits: AI Generated

DOI: 10.1016/j.cscee.2026.101477

Keywords: turpentine wastewater, demulsification, oil and grease removal, wastewater treatment, polymeric demulsifier, response surface methodology, composite desirability, thermal separation, industrial effluent, emulsion breaking, polynomial regression, BOD removal

The study, led by Wenjun Zhang of the Frontiers Science Center for Deep Ocean Multispheres and Earth System at Ocean University of China, together with colleagues at the University of Shanghai for Science and Technology and Shandong University, introduces what the authors call a heterogeneous wetting membrane, or heterowettability membrane. Unlike conventional membranes, which present a single uniform wettability to the fluid they contact, this membrane features spatially patterned regions of hydrophilic and hydrophobic surface chemistry arranged side by side. The design is paired with a bidirectional gradient contraction channel module, and it is this combination that allows one device to do the work of two.

The central insight of the work lies in how the two components divide their labor. The gradient contraction channel is not merely a piece of plumbing. As fluid flows through a channel whose cross-section narrows progressively, the velocity field becomes non-uniform, generating a spatially varying shear stress across the flow. That non-uniform shear field acts directly on the emulsion droplets, deforming them and gradually destabilizing the surfactant shells that keep them intact. In effect, the channel performs a gentle, continuous demulsification before the fluid ever reaches the membrane surface, priming the droplets to coalesce and release their contents at the membrane interface.

Once the partially demulsified dispersion arrives at the membrane, the patterned wettability takes over. Droplets of the dispersed phase encounter surface regions whose chemistry matches their own continuous phase affinity, and the resulting surface energy gradients drive them to migrate in opposite directions across the membrane. Water droplets in an oil continuous phase are drawn toward hydrophilic zones, while oil droplets in water are pulled toward hydrophobic zones. Coupled with the shear-induced demulsification and directional phase transport, this wettability-guided routing allows oil-in-water and water-in-oil emulsions to be processed simultaneously and in parallel, each yielding its own purified phase.

A critical challenge in any such dual-function design is preventing the two phases from crossing into each other’s collection pathways, a failure mode known as continuous phase crossover. The researchers addressed this with what they describe as a pressure field–Laplace anchoring mechanism. By carefully controlling the transmembrane pressure and exploiting the Laplace pressure barriers that arise at the boundary between wetting and non-wetting regions of the membrane, the device suppresses unwanted crossover of the continuous phases. In essence, the capillary pressure required to force a non-wetting liquid through a pore region acts as a self-regulating valve, keeping each phase confined to its designated pathway without the need for external actuation or switching.

The team’s experimental characterization, documented across six main figures and an extensive supplementary package of more than sixty figures and sixteen tables, probed the interfacial restructuring and retained fluid states on the membrane during operation. Contact angle measurements captured in real-time video show directed water transport along the heterogeneous wetting interface, driven by surface energy gradients. Dynamic observations of the oil-water interface on both hydrophilic and hydrophobic membrane regions under varying transmembrane pressure reveal how the fluid states are maintained and how the phase-selective droplet transfer is controlled by the interplay of membrane wettability and channel geometry.

Perhaps the most subtle finding of the study concerns the dual role of shear stress. Shear is essential for demulsification, breaking droplets apart so their phases can be separated. But excessive shear at the membrane surface risks re-emulsification, tearing newly coalesced phases back into fine droplets and undoing the separation. The researchers mapped this trade-off in detail, identifying the operating window in which the gradient contraction channel delivers enough shear to destabilize droplets while the membrane interface experiences conditions that favor coalescence and phase transport rather than droplet re-formation. This balance, they show, is what sustains the device’s remarkable efficiency over extended operation.

The practical implications are considerable. Anti-fouling performance and long-term stability, the two metrics that most often doom membrane technologies in real wastewater service, were both notably enhanced in the new system. Because the opposing emulsion streams are demulsified before contact with the membrane and routed along chemically matched pathways, foulants are less likely to accumulate and clog the pores. The authors also report a techno-economic assessment supported by collaborators at Ocean University of China, suggesting the team has begun to evaluate not just laboratory performance but the commercial viability of the approach. For industries ranging from petroleum refining to food processing and textile manufacturing, where both emulsion types arise in the same facility and are currently handled by separate, sequential treatment trains, a single device that purifies both streams continuously could simplify plant design and reduce operating costs.

The work also builds on, and departs from, a rich lineage of bioinspired separation research. Prior advances include superwetting nanofiber membranes, CO2-responsive switchable membranes, Janus membranes with asymmetric affinities, liquid-infused aerogel systems capable of on-demand emulsification and demulsification, and a Janus channel of membranes reported in Science in 2024 that achieved concurrent oil and water recovery from emulsions. What distinguishes the new study is its explicit targeting of opposing emulsions within a single synchronous process, achieved not by switching a membrane’s properties over time but by distributing different wetting behaviors across space and letting the channel geometry do the rest. The result, the authors write, establishes a new paradigm for synchronous separation and parallel processing of opposite emulsions, paving the way for efficient integrated oil-water separation technology.

Challenges remain before the system can be scaled from laboratory modules to industrial service. Membrane fabrication must be reproducible at large areas, the patterned wettability must withstand years of exposure to surfactants, solvents and temperature swings, and the pressure operating window must tolerate the variability of real waste streams. Yet the fundamental demonstration is striking: a membrane that treats wettability not as a fixed property to be chosen but as a design variable to be patterned, working in concert with fluid mechanics rather than against it. If the reported efficiencies and stability hold up at scale, the humble membrane may finally be ready to handle the full, messy spectrum of emulsified oily waste in one continuous pass.

Subject of Research: A patterned heterowettability membrane with a gradient contraction channel for synchronous, parallel separation of opposing oil-in-water and water-in-oil emulsions.

Article Title: Synchronous and parallel separation of opposing emulsions enabled by a heterowettability membrane

Article References: Zhang, W., Zhang, D., Guan, Y., Li, Y., Geng, S., Li, B., Chen, D., Wang, J., Bao, M., & Wang, Z. (2026). Synchronous and parallel separation of opposing emulsions enabled by a heterowettability membrane. Nature Water. https://doi.org/10.1038/s44221-026-00703-z

Image Credits: AI Generated

DOI: 10.1038/s44221-026-00703-z

Keywords: heterowettability membrane, emulsion separation, oil-water separation, demulsification, shear-induced demulsification, water-in-oil emulsion, oil-in-water emulsion, membrane technology, anti-fouling, Laplace pressure, directional liquid transport, water purification