Structure–Property Relationships and Force-Field Transferability in 1-Propanol/Water Mixtures from 243 to 298 K

Imagine a glass of water and a glass of rubbing alcohol. If you pour them together, they do not simply sit side by side; they mix into a single, uniform liquid. Yet, on a microscopic level, this mixture is far from simple. The water molecules, which love to hold hands with each other through invisible electrical attractions, must suddenly accommodate the alcohol molecules, which have a water-loving head and a water-avoiding tail. This tug-of-war creates a complex, shifting landscape where the molecules arrange themselves in ways that change the liquid's weight, how easily it flows, and how it behaves at the surface. Understanding these hidden arrangements is crucial for everything from designing better fuels to creating effective disinfectants, but it is incredibly difficult to see these tiny structures directly.

To peer into this invisible world, scientists use computer simulations that act as a virtual microscope. They build digital models of individual molecules and watch how they move and interact over time. The challenge lies in the models themselves: if the digital rules governing how molecules attract or repel each other are slightly off, the entire simulation can lead to the wrong conclusions. A major question in this field is whether a model built to describe pure alcohol can still work accurately when that alcohol is mixed with water, without needing to be completely rebuilt from scratch. This is the core puzzle tackled by a team of researchers at Jordan University of Science and Technology, who set out to test if a specific digital model of 1-propanol, a common type of alcohol, could survive the transition from a pure liquid to a mixture with water.

The researchers created a vast array of virtual mixtures, covering every possible ratio of water to 1-propanol, and watched them behave at temperatures ranging from just above freezing to room temperature. They paired their alcohol model with two different, well-known digital models of water to see which combination produced results that matched real-world experiments. One water model was a simpler three-point version, while the other was a more detailed four-point version. By keeping the alcohol model exactly the same and only changing the water, the team could isolate how the choice of water model influenced the outcome. They measured how dense the mixtures became, how the molecules moved, and how the liquid behaved at its surface, comparing their virtual results against known physical data.

The simulations revealed that the digital alcohol model was surprisingly robust. When mixed with water, it successfully reproduced the way real mixtures contract in volume, becoming slightly denser than a simple average of the two liquids would suggest. This contraction peaked when the mixture contained about forty percent alcohol by number, a point where the water and alcohol molecules seemed to pack together most efficiently. Both water models captured this behavior, though the four-point water model predicted the density of the mixtures with slightly higher accuracy. The study also showed that the molecules did not mix randomly; instead, the alcohol molecules tended to gather at the very surface of the liquid, creating a layer that was richer in alcohol than the liquid beneath it. This surface enrichment explains why adding even a small amount of alcohol to water significantly lowers the liquid's surface tension, making it easier to spread or break apart.

However, the story became more complex when the researchers looked at how the molecules moved. In the middle range of the mixture, where neither water nor alcohol dominated, the alcohol molecules slowed down significantly, reaching a point of minimum movement. At the same time, the liquid's resistance to flow, or viscosity, appeared to reach a maximum. This suggests that in this intermediate zone, the molecules form a temporary, tangled network that hinders movement before the mixture becomes fluid again as more alcohol is added. While the simulations captured this slowing and speeding up, the exact numbers for how fast the molecules moved depended heavily on the size of the virtual container and the specific water model used. The researchers found that correcting for the finite size of their virtual boxes shifted the speed values, but the general pattern of slowing down in the middle of the mixture remained consistent.

The team also examined the internal structure of the liquid by looking at how the oxygen atoms in the molecules were arranged relative to one another. They found that as the mixture changed from mostly water to mostly alcohol, the local neighborhood of the oxygen atoms shifted dramatically. In water-rich areas, oxygen atoms were surrounded by many neighbors in a tight, organized shell. As alcohol was added, this shell became less crowded and more disordered. This structural reorganization happened right in the same composition range where the molecules slowed down and the liquid became more viscous, linking the physical movement of the molecules directly to how they were packed together. The simulations also successfully recreated the patterns seen in real X-ray experiments, confirming that the digital models were capturing the correct large-scale arrangement of the molecules.

Despite these successes, the study highlighted important limitations in how these virtual experiments are conducted. The researchers found that the way they handled the long-range forces between molecules had a massive impact on the surface tension results. When they used a more complete method to calculate these forces, the predicted surface tension increased by several units, showing that their initial, faster calculations had missed a significant piece of the physics. Furthermore, the study could not definitively prove that the liquid remained stable at the lowest temperatures tested, as the simulations were not long enough to rule out the possibility of the water freezing into ice. The viscosity results, while showing the correct trend of a peak in the middle of the mixture, were described as preliminary because the computer runs were not long enough to guarantee statistical certainty.

Ultimately, the work demonstrates that a force field designed for pure alcohol can transfer well to mixtures, but the choice of water model and the details of the simulation method matter deeply. Neither water model was perfect across every measurement; one was better at predicting density, while the other offered different dynamic behaviors. The study did not declare a single winner but instead provided a clear, property-by-property map of where each model succeeds and where it falls short. By showing exactly how the digital models behave under strict, controlled conditions, the research offers a reliable guide for future scientists who need to simulate these complex, everyday liquids. The findings confirm that while the microscopic dance of water and alcohol is intricate, it can be understood through careful, comparative simulation, provided the limitations of the virtual tools are kept in mind.