🔬 condensed matter

II. Temperature trends in the properties of simple monohydric alcohols. Molecular dynamics simulations of united atom UAMI-EW model

This study utilizes isobaric-isothermal molecular dynamics simulations with the UAMI-EW united atom force field to investigate the temperature-dependent properties of methanol, ethanol, and 1-propanol, validating the model against experimental data and exploring the mixing behavior of methanol-propanol mixtures.

Original authors: M. Aguilar, E. Núñez-Rojas, O. Pizio

Published 2026-03-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: M. Aguilar, E. Núñez-Rojas, O. Pizio

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where molecules are like people at a crowded party. Some are small and energetic (like Methanol), some are medium-sized (Ethanol), and some are a bit taller and carry a longer coat (1-Propanol). These "people" are alcohols, and they love to mix with each other and with water.

This scientific paper is essentially a digital field report from a team of researchers who built a virtual simulation of this party to see how these molecules behave when the room gets hotter or colder. They wanted to know: Does our computer model of these molecules act like real life?

Here is the breakdown of their findings, translated into everyday language:

1. The Virtual Lab: Building the "Digital Twin"

The researchers didn't just guess; they built a specific set of rules (called a force field) to tell their computer how these alcohol molecules should interact. Think of this like creating a video game character. If the character's physics are wrong, the game looks fake.

They tested their "character" (the UAMI-EW model) against real-world data. They checked four main things:

How heavy they are (Density): Does the crowd get tighter when it gets cold?
How they stick together (Dielectric Constant): How well do they hold hands or repel each other electrically?
How they hold their shape at the edge (Surface Tension): If you have a drop of alcohol, how hard is it to stretch that drop?
How fast they dance (Self-Diffusion): How quickly do they move around the room?

The Verdict: The model was a star performer. It predicted the density, surface tension, and movement of Methanol, Ethanol, and Propanol with impressive accuracy across a wide range of temperatures. It was like having a crystal ball that could tell you exactly how a liquid would behave before you even poured it.

2. The Party Mix: Methanol vs. Propanol

Once they were confident in their single-molecule models, they threw a new party: a mix of Methanol (the small one) and Propanol (the tall one).

In chemistry, when you mix two things, the total volume or energy doesn't always equal the sum of the parts. Sometimes, mixing them makes the total volume shrink (like packing a suitcase efficiently), and sometimes it expands. This is called "Excess Mixing."

The Volume Test: When they mixed Methanol and Propanol, the computer predicted the volume changes very well. It correctly showed that the mixture behaves almost like an "ideal" solution (where things mix perfectly without drama), though the model slightly overestimated the "bumpiness" of the mix.
The Energy Test (Heat): Mixing things usually releases or absorbs heat. The model correctly predicted that mixing these two alcohols requires a tiny bit of heat (it's endothermic). However, the computer predicted this heat effect was a bit stronger than what happens in real life. It's like the digital molecules were a little more dramatic about hugging each other than real molecules are.

3. The "Hand-Holding" (Hydrogen Bonds)

Alcohol molecules are famous for "holding hands" via hydrogen bonds. The researchers looked at how many hands each molecule held in the mix.

The Surprise: Even though the temperature changed (making the party hotter and the molecules dance faster), the number of hands held didn't change much.
The Twist: Even though the "hand-holding" count stayed steady, the energy of the mix changed significantly with temperature. It's like a group of friends standing in a circle: if they start running in place (higher temperature), they might feel more tired (higher energy) even if they are still holding the same number of hands. This showed that small changes in movement can lead to big changes in how the mixture feels energetically.

4. The Electric Personality

The researchers also checked the dielectric constant, which is basically a measure of how "electrically friendly" the liquid is.

The Result: The model was excellent at predicting this. It correctly showed how the mixture's electrical personality changes as you add more Methanol or Propanol. It even beat out other, more complex computer models that tried to do the same job.

5. The "Speed Limit" (Diffusion)

Finally, they watched how fast the molecules moved.

The Observation: As expected, the smaller Methanol molecules zipped around faster than the larger Propanol molecules.
The Mix: When mixed, the speed of each molecule changed depending on how much of the other was present. The computer simulation matched the trends seen in real experiments perfectly, proving that the model understands the "dance floor" dynamics well.

The Big Picture

What did they learn?
The researchers concluded that their "UAMI-EW" model is a highly reliable tool. It can predict how simple alcohols behave when you heat them up, cool them down, or mix them together.

What's next?
Just like a video game that needs new levels, the researchers want to expand their simulation. They plan to:

Test more complex alcohols (like 2-propanol).
Study mixtures of primary and secondary alcohols.
Look at how these liquids behave under extreme pressure (like deep underwater).
Add more details to the model, like how the liquids flow (viscosity) and how they relax after being shocked.

In a nutshell: The scientists built a very accurate digital twin of alcohol molecules. This twin can now help engineers and chemists predict how these liquids will behave in real-world applications (like fuel, solvents, or medicines) without needing to run expensive and messy experiments every single time. It's a powerful shortcut for understanding the invisible world of molecules.

1. Problem Statement

The theoretical modeling of alcohols and their mixtures is critical for chemical engineering but remains challenging. While significant efforts have been made using experiments and simulations, a complete understanding of the thermodynamic trends of these systems across varying pressures, temperatures, and chemical compositions is lacking.

Specific Gap: Existing united-atom models (like TraPPE) are often parametrized for vapor-liquid equilibrium but may fail to accurately predict miscibility with water or dielectric properties. Conversely, the UAMI-EW (United Atom, Non-polarizable, Explicit Water) model was previously parametrized at a single thermodynamic state ( $T=298.15$ K, $P=1$ bar) to match density, dielectric constant, and surface tension.
Objective: The authors aim to validate the applicability and robustness of the UAMI-EW force field beyond its parametrization point. Specifically, they investigate:
1. The temperature dependence of properties for pure monohydric alcohols (Methanol, Ethanol, 1-Propanol) over a wide temperature range.
2. The accuracy of the model in predicting mixing properties (density, excess volume, excess enthalpy, dielectric constant, diffusion) for Methanol–Propanol (MeOH–PrOH) mixtures at different temperatures.

2. Methodology

Force Field: The study utilizes the UAMI-EW united-atom, non-polarizable force field. In this model, hydrogen atoms attached to carbons are merged with the carbon atoms (united atoms), while hydroxyl hydrogens are treated explicitly. The interaction potential combines Lennard-Jones (LJ) and Coulomb terms.
Simulation Setup:
- Software: GROMACS 5.1.2.
- Ensemble: Isothermal-isobaric (NPT).
- System Size: 3000 molecules in a cubic box with periodic boundary conditions.
- Thermostat/Barostat: V-rescale thermostat ( $\tau_T = 0.5$ ps) and Parrinello–Rahman barostat ( $\tau_P = 2.0$ ps).
- Time Step: 0.002 ps.
- Interactions: Non-bonded cutoff at 1.1 nm; long-range electrostatics handled via Particle Mesh Ewald (PME). LINCS algorithm used for bond constraints.
Protocol: After equilibration, production runs of at least 10 ns were performed. Properties were averaged over 7–10 runs.
Systems Studied:
- Pure Components: Methanol (MeOH), Ethanol (EtOH), 1-Propanol (PrOH).
- Mixtures: MeOH–PrOH binary mixtures at various mole fractions ( $X_{MeOH}$ ).
- Conditions: Temperatures ranging from ~220 K to 360 K (depending on the alcohol's boiling point) at ambient pressure (1 bar).

3. Key Contributions

Validation of UAMI-EW across Temperature: The first comprehensive assessment of the UAMI-EW model's ability to predict thermodynamic and transport properties of simple alcohols over a wide temperature range, not just at the parametrization point (298.15 K).
Mixture Modeling: Application of the pure-component force field to binary alcohol mixtures (MeOH–PrOH) without re-parametrization, testing the transferability of the model.
Comparative Analysis: Direct comparison of UAMI-EW results against experimental data and other advanced models (TraPPE, All-Atom OPLS/AA, and ab initio methods).
Microscopic Insight: Analysis of hydrogen bond networks and their correlation with macroscopic mixing properties (excess enthalpy).

4. Results

A. Pure Alcohol Properties (MeOH, EtOH, PrOH)

Density: The model provides a very good description of density vs. temperature for all three alcohols. Accuracy is highest for MeOH and EtOH, with small deviations (<1%) for PrOH.
Static Dielectric Constant ( $\varepsilon$ ):
- The model captures the general trend of decreasing $\varepsilon$ with increasing temperature.
- MeOH: Slight underestimation compared to experiment, with deviations increasing at lower temperatures.
- EtOH & PrOH: Satisfactory agreement across the temperature range.
Surface Tension ( $\gamma$ ):
- The model correctly describes the linear decrease of $\gamma$ with temperature.
- Trend: Overestimates $\gamma$ for MeOH and EtOH at low temperatures; underestimates for PrOH across the entire range. Absolute deviations are around 5%.
Self-Diffusion Coefficient ( $D$ ):
- MeOH: Simulated values are consistently underestimated compared to experimental data across the temperature range.
- EtOH & PrOH: Excellent agreement with experimental data.
- Note: $D$ was not a target property during the original force field parametrization, making this a successful predictive test.

B. MeOH–PrOH Mixtures

Density: The model underestimates mixture density by approximately 1% across all compositions and temperatures but correctly reproduces the non-linear shape of the density vs. composition curve.
Excess Mixing Volume ( $\Delta V^{ex}_{mix}$ ):
- The model predicts positive excess volumes (expansion upon mixing), consistent with experimental trends.
- It captures the temperature trend: the magnitude of $\Delta V^{ex}_{mix}$ increases with temperature.
- The maximum deviation occurs around $X_{MeOH} \approx 0.6$ .
Excess Mixing Enthalpy ( $\Delta H^{ex}_{mix}$ ):
- The model correctly predicts endothermic mixing (positive $\Delta H^{ex}_{mix}$ ) for the MeOH–PrOH system.
- It reproduces the symmetric curve shape with a maximum near equimolar composition ( $X_{MeOH} = 0.5$ ).
- Limitation: The magnitude of the excess enthalpy is overestimated compared to experiments. However, the temperature trend (increase in magnitude at higher $T$ ) is qualitatively correct.
- Comparison: UAMI-EW performs better than the TraPPE model in predicting the maximum value of $\Delta H^{ex}_{mix}$ .
Hydrogen Bonding:
- The average number of hydrogen bonds per molecule is nearly constant (~2) across the composition range, indicating a chain-like structure dominance.
- Key Finding: Small changes in hydrogen bonding due to temperature changes result in pronounced changes in excess mixing enthalpy, highlighting the sensitivity of thermodynamic properties to microscopic structural changes.
Dielectric Constant of Mixtures:
- The model accurately describes the composition dependence of the static dielectric constant ( $\varepsilon$ ) and the excess dielectric constant ( $\Delta \varepsilon$ ).
- UAMI-EW outperforms both All-Atom (OPLS/AA) and ab initio models in this specific metric.
Self-Diffusion in Mixtures:
- Simulations show that diffusion coefficients increase with methanol concentration (smaller molecules move faster).
- Trends agree qualitatively with experimental data available for MeOH–Ethanol mixtures.

5. Significance and Future Outlook

Model Robustness: The UAMI-EW model is validated as a robust tool for simulating simple monohydric alcohols over a wide $(T, P)$ range, particularly for volumetric, dielectric, and mixing properties.
Efficiency vs. Accuracy: As a united-atom model, it offers a favorable balance between computational efficiency and accuracy, often outperforming more complex all-atom models in predicting dielectric properties of mixtures.
Future Directions: The authors propose extending this work to:
- More complex alcohols (secondary alcohols like 2-propanol, butanols).
- Mixtures of primary and secondary alcohols.
- Investigation of missing properties such as shear viscosity and dielectric relaxation.
- Refinement of all-atom and polarizable models based on these findings.

In conclusion, the paper demonstrates that the UAMI-EW force field, despite being parametrized at a single state point, successfully captures the complex temperature and composition dependencies of alcohol properties, making it a valuable tool for chemical engineering applications involving alcohol mixtures.