A mussel clinging to a wave-hammered rock has about 30 seconds to glue itself in place before the next surge tries to rip it free. Materials scientists have spent decades trying to understand how the animal pulls this off, and an even longer stretch trying to copy it. Now, a study published in May 2026 in Nature Communications offers a surprisingly concrete answer: the speed of the glue depends not just on what the ingredients are, but on how they meet.
The research team, led by Xian Chen and Zhong-Yuan Lu at the Hong Kong University of Science and Technology, used large-scale molecular dynamics simulations to model what happens when oppositely charged polymers are brought together under different mixing conditions. Their central finding is that controlling the geometry of mixing, specifically how fresh polymer material flows into growing liquid droplets, can accelerate the formation of adhesive coacervates far beyond what classical physics would predict.
Why mixing geometry changes everything
Coacervates form through a process called liquid-liquid phase separation, or LLPS, in which dissolved molecules spontaneously condense into dense liquid droplets. The textbook expectation, known as Ostwald ripening, says those droplets grow at a rate proportional to the cube root of time. In practical terms, that is slow: molecules diffuse randomly, and droplets enlarge only by scavenging material from their smaller neighbors.
Chen, Lu, and their colleagues found that when the mixing setup channels a continuous stream of fresh polyelectrolyte into the growing droplets, a mechanism they call the “flux pathway,” the growth rate jumps dramatically. Under these conditions, the effective scaling exponent roughly doubles compared to the classical baseline. The simulations showed that a droplet could reach half a centimeter across in a matter of seconds, a timeline that matches the rapid adhesion biologists have documented in living mussels.
The study reports multiple scaling regimes depending on how the charged polymers are introduced to each other. Some conditions produced an early growth phase scaling as the square root of time, followed by faster or slower regimes. The headline figure of a doubled growth exponent represents one favorable scenario, not a universal outcome, but even the less dramatic regimes were faster than diffusion alone would allow.
Biological evidence lines up
The simulation results gain credibility from independent experimental work on real marine organisms. Time-resolved spectroscopy of mussel byssus formation has captured structural changes in adhesive proteins beginning roughly 10 seconds after secretion starts, pointing to a tightly choreographed sequence that unfolds on exactly the timescale the simulations now explain.
Separately, laboratory studies on mussel-derived adhesive peptides designated GK-16* have shown that LLPS occurs under seawater-like salt concentrations, driven by hydrogen bonds and electrostatic forces. That work, published in Nature Communications in 2022, used circular dichroism, infrared spectroscopy, and surface forces measurements to map the molecular interactions behind coacervation at specific peptide and salt thresholds.
The sandcastle worm provides another parallel. Its adhesive system relies on a polyelectrolyte coacervate that sets in under 30 seconds, forming self-supporting particles in turbulent water without any external energy input. Foundational research on marine bioadhesion established years ago that coacervate-based glues can assemble and harden in highly saline, turbulent conditions that would strip away most synthetic adhesives. What the new study adds is a specific physical mechanism, the flux pathway, that could explain why these biological systems work as fast as they do.
Gaps between simulation and reality
As the Nature Communications study itself acknowledges, its scaling laws have not been validated against biological measurements. No one has yet filmed coacervate domains growing inside a living mussel foot at the resolution needed to confirm the predicted scaling laws in a biological system. The study’s numerical results come entirely from molecular dynamics simulations with explicit electrostatics and hydrodynamics, not from tracking droplet growth in an organism. The biological analogy is persuasive but indirect: mussels adhere in under 30 seconds, and the simulations produce a mechanism fast enough to account for that speed, yet the two have not been directly linked by experiment. More broadly, the existing literature on mussel byssus formation does not appear to include direct, real-time measurements of LLPS growth-law exponents during live secretion, a gap the authors of the new study highlight as motivation for their computational approach.
Charge asymmetry complicates the picture further. Earlier computational studies by Chen and collaborators, cited in the new paper, established that when positive and negative polyelectrolytes carry unequal charge densities, coarsening slows relative to the symmetric case. (Those earlier results were reported in a separate publication in the same journal; readers can find the specific citation in the reference list of the new study.) The current paper builds on those findings, but the quantitative thresholds for how much asymmetry is tolerable before the flux pathway stalls have not been tested with real polymers in a lab.
There is also the question of chemical realism. Mussel foot proteins are loaded with catechol-bearing residues and post-translational modifications that shape both charge distribution and binding chemistry. The simulations capture generic charge-driven phase separation, not the full molecular complexity of natural adhesives. The flux pathway may be one of several mechanisms acting in parallel in living organisms rather than the whole story.
On the engineering side, the simulations assume well-controlled mixing geometries and concentration gradients that are straightforward to program on a computer but harder to reproduce at scale. Microfluidic devices could, in principle, impose the necessary spatial control over how two polyelectrolyte streams meet, but the study does not demonstrate a physical prototype. Without benchtop validation, it remains unclear whether surgical applicators, catheter tips, or underwater repair tools can reliably recreate the conditions the simulations require.
Why the flux pathway reframes adhesive design
For engineers and materials scientists, the practical message is specific and actionable: how you mix polyelectrolytes may matter as much as what you mix. If the flux pathway holds up under experimental testing, it suggests that adhesive formulations for surgery, dentistry, or subsea infrastructure repair could be paired with applicators engineered to deliver components in controlled, converging streams rather than pre-mixed batches.
That would turn mixing geometry into a performance parameter on par with polymer length or charge density, a variable that formulators have largely ignored. It would also open a new design route for bioinspired glues that set on the same swift timescale as a mussel gripping a rock in the surf, not by copying the animal’s proteins, but by copying the physics of how those proteins find each other.
The next milestone is clear: someone needs to build a device that exploits the predicted flux regime and measure whether real coacervate droplets grow as fast as the simulations say they should. Until that experiment is done, the study stands as a compelling computational explanation for one of biology’s most impressive quick-setting adhesives, and a blueprint that synthetic chemists can start testing today.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
