Molecular dynamics study of perchloric acid using the extended Madrid-2019 force field

This study employs molecular dynamics simulations using the extended Madrid-2019 force field combined with TIP4P/2005 water to accurately predict the thermodynamic, structural, and transport properties of perchloric acid solutions across various concentrations and temperatures, demonstrating excellent agreement with experimental data for densities and moderate agreement for viscosities.

The Gist

The Big Picture: Simulating a "Super-Strong" Acid

Imagine you have a very strong, dangerous liquid called Perchloric Acid. It's used in everything from rocket fuel to cleaning industrial equipment. Scientists know a lot about how it behaves in big buckets (like how heavy it is or how thick it is), but they don't fully understand what's happening at the microscopic level—how the tiny atoms and molecules dance around each other.

This paper is like a virtual laboratory. Instead of mixing dangerous chemicals in a real lab, the researchers built a computer simulation to watch how these molecules interact. Their goal was to create a "digital twin" of perchloric acid that is so accurate, it can predict real-world behavior without the risk of an explosion.

The Cast of Characters: The "Madrid-2019" Team

To build this simulation, the researchers didn't invent new rules from scratch. They used a pre-existing "rulebook" for how molecules talk to each other, called the Madrid-2019 Force Field.

Think of this rulebook like a set of instructions for a dance. It tells the molecules:

• How big they are.
• How much they like to stick together or push apart.
• How they move.

Specifically, they used rules for:

1. Water molecules (the stage).
2. Perchlorate ions (the "anions" – negative charges).
3. Oxonium ions (the "cations" – positive charges, which are basically water molecules that grabbed an extra hydrogen).

The Secret Sauce: "Scaled Charges"
In the real world, ions have a full electric charge (like a battery at 100%). However, in a computer simulation, if you use the full charge, the molecules get too attracted to each other, and the simulation breaks.

The researchers used a clever trick called "scaled charges." Imagine turning the volume down on the electric charge. Instead of 100%, they set the volume to 85%. This "dimmed" charge allows the molecules to move more naturally, mimicking how real water screens (hides) the electric forces between ions.

What They Did: The Virtual Experiments

Once the digital acid was mixed with digital water, they ran the simulation to see what happened. They checked four main things:

1. The Weight (Density)

• The Test: They asked, "If we add more acid to the water, does the mixture get heavier, and by how much?"
• The Result: The computer predicted the weight perfectly. Up to very high concentrations (where the liquid is almost half acid, half water), the simulation matched real-world measurements almost exactly. It's like a digital scale that never lies.

2. The "Weird" Temperature (Temperature of Maximum Density)

• The Concept: Pure water is weird. It gets heavier as it cools, but then, just before it freezes, it gets lighter again (which is why ice floats). This happens at a specific temperature called the Temperature of Maximum Density (TMD).
• The Test: They wanted to know: "If we add acid, does this 'weird' temperature shift?"
• The Result: Yes! The acid pushes this special temperature down. The simulation predicted exactly how much the temperature drops based on how much acid is in the mix. This is a property that is very hard to measure in a real lab, so the computer model is a huge help.

3. The Dance Moves (Structure)

• The Test: They looked at how the water molecules arrange themselves around the acid ions. Do they hug them tightly? Do they form a cage?
• The Result: The simulation showed that water molecules form a specific "shell" around the perchlorate ions. Interestingly, the acid ions didn't clump together into big blobs (precipitate); they stayed happily dissolved, just like in a real bottle of acid.

4. The Flow (Viscosity and Speed)

• The Test: How fast do the molecules swim (diffusion)? How thick and sticky is the liquid (viscosity)?
• The Result:
  • Speed: The simulation predicted how fast the molecules move very well at low concentrations.
  • Stickiness: The simulation was good at predicting how thick the liquid gets, but at very high concentrations, it predicted the liquid was slightly thicker than it really is. It's like the digital liquid was a little too "syrupy." The authors suggest that if they turned the "volume" of the electric charge down even a tiny bit more (from 85% to 75%), the prediction would be even better.

Why Does This Matter?

You might ask, "Why bother simulating something we can already measure?"

1. Safety: Perchloric acid is dangerous. Simulating it is safer than testing it.
2. Prediction: The simulation can predict what happens at temperatures or pressures that are too extreme or expensive to test in a lab.
3. Understanding: It gives scientists a "microscope" to see the invisible dance of atoms, helping them design better rocket fuels or industrial processes.

The Heartfelt Ending

The paper ends with a touching dedication to a scientist named Prof. Stefan Sokołowski, who passed away recently. The authors describe him as a brilliant mentor who loved sharing knowledge, much like the "scaled charges" in their model—making complex, heavy concepts accessible and lighter for everyone to understand.

Summary in a Nutshell

The researchers built a virtual model of perchloric acid using a smart set of rules (Madrid-2019) that slightly "dims" the electric charges of the molecules. This model was so good that it could accurately predict how heavy the acid gets, how thick it becomes, and how its molecules dance, matching real-world experiments almost perfectly. It's a powerful tool for understanding one of chemistry's most important (and dangerous) liquids.

Technical Summary

1. Problem Statement

Perchloric acid ($HClO_4$) is a critical chemical used in propellants, industry, and biology, yet its physicochemical properties in aqueous solutions are not fully characterized by molecular simulations. While experimental data exists for density and viscosity, comprehensive molecular-level insights (such as solvation structures and transport mechanisms) are scarce. Existing computational models often struggle to simultaneously reproduce thermodynamic, structural, and dynamic properties of strong electrolytes without extensive parameter re-tuning. The specific challenge addressed in this work is the development and validation of a predictive molecular model for perchloric acid solutions that accurately captures behavior across a wide range of concentrations (up to 10 molal) and temperatures.

2. Methodology

The authors employed classical Molecular Dynamics (MD) simulations to model aqueous perchloric acid solutions.

• Force Field Model:

  • Water: Modeled using the TIP4P/2005 potential.
  • Ions: The system assumes complete dissociation of $HClO_4$ into hydronium ($H_3O^+$) and perchlorate ($ClO_4^-$) ions. The parameters for these ions were taken from the extended Madrid-2019 force field.
  • Scaled Charges: A key feature of the Madrid-2019 model is the use of scaled ionic charges ($q = \pm 0.85e$) rather than formal charges ($\pm 1e$). This scaling accounts for the lack of electronic polarizability in non-polarizable water models and improves the reproduction of dielectric properties and transport coefficients.
  • Geometry: Both $H_3O^+$ and $ClO_4^-$ were treated as rigid bodies. The $ClO_4^-$ anion was constrained to a perfect tetrahedral geometry using the SHAKE algorithm (necessary because LINCS cannot handle the full set of constraints for this topology in older GROMACS versions).
  • Interactions: Cross-interactions between ions and water, and between the cation and anion, were calculated using Lorentz-Berthelot (LB) combining rules without additional fitting.

• Simulation Setup:

  • Software: GROMACS (version 4.6.5).
  • Ensembles: Isothermal-isobaric ($NpT$) for density and Temperature of Maximum Density (TMD) calculations; Canonical ($NVT$) for viscosity and diffusion calculations.
  • System Size: Systems contained 555 water molecules plus the requisite number of ion pairs to achieve target molalities. For high-concentration validation, larger systems (4440 water molecules) were simulated.
  • Conditions: Simulations were run at 1 bar and various temperatures (283.15 K, 298.15 K, 323.15 K) and molalities (up to ~10 $m$).
  • Analysis:
    • Density: Calculated from $NpT$ averages.
    • TMD: Determined by fitting density vs. temperature data to third-order polynomials.
    • Structure: Radial Distribution Functions (RDFs) calculated for ion-water correlations.
    • Transport: Self-diffusion coefficients ($D$) derived from Mean Square Displacement (MSD) with Yeh-Hummer finite-size corrections; Shear viscosity ($\eta$) calculated via the Green-Kubo formalism using pressure tensor autocorrelation functions.

3. Key Contributions

• First Comprehensive Simulation of $HClO_4$: This study provides the first molecular simulation of perchloric acid solutions that reproduces experimental densities over a wide concentration range (up to 10 $m$) and multiple temperatures without re-parameterizing cross-interactions.
• Validation of Transferability: It demonstrates that the Madrid-2019 force field parameters, originally developed for individual ions, are transferable to the combined perchloric acid system using standard combining rules.
• Prediction of Unmeasured Properties: The study predicts the Temperature of Maximum Density (TMD) for perchloric acid solutions, a property not currently available in experimental literature.
• Microscopic Structural Insight: It provides detailed RDF data for ion-water and ion-ion correlations in bulk solution, offering a reference for future experimental structural studies (e.g., neutron diffraction).

4. Key Results

• Thermodynamic Properties (Density):
  • The model predicts experimental densities with excellent agreement (relative deviations < 0.5%) up to 10 $m$ at 283.15 K, 298.15 K, and 323.15 K.
  • The model successfully reproduces the density maximum shift (Despretz law) as a function of molality.
• Temperature of Maximum Density (TMD):
  • The TMD of pure water (277.3 K) decreases linearly with increasing acid concentration.
  • The calculated Despretz constant ($K_m$) is -21.093 K·kg·mol⁻¹, slightly deviating from the sum of individual ionic contributions (-18 K·kg·mol⁻¹) but consistent with the linear trend expected for dilute electrolytes.
  • At 1 $m$, the predicted TMD is 259.3 K.
• Structural Features (RDFs):
  • $O_w-O_p$ (Water O to Perchlorate O): First peak at ~3.20 Å, consistent with IR and neutron diffraction data.
  • $O_w-Cl_p$: Shows two peaks (3.77 Å and 4.40 Å), indicating water penetration into the perchlorate tetrahedron and hydration outside the tetrahedron.
  • Ion Pairs: The absence of a deep first minimum in the $H_{3O^+}-ClO_4^-$ RDF suggests that contact ion pairs (CIPs) are not prevalent in the simulated conditions; ions remain largely solvated.
• Transport Properties:
  • Diffusion: The model reproduces the concentration dependence of water and perchlorate self-diffusion coefficients well, though it slightly underestimates the infinite-dilution diffusion of $ClO_4^-$ compared to experimental Nernst-Einstein values.
  • Viscosity: At 298.15 K, predicted viscosities agree well with experiments for concentrations below 4 $m$. At higher concentrations, the model slightly overestimates viscosity (a known limitation of the $\pm 0.85e$ charge scaling for transport properties), but the trend remains accurate.

5. Significance and Conclusion

This work establishes the extended Madrid-2019 force field as a robust and predictive tool for modeling perchloric acid solutions. The study confirms that using scaled charges ($\pm 0.85e$) combined with TIP4P/2005 water allows for the accurate reproduction of bulk densities and structural features without the need for system-specific fitting of cross-interaction parameters.

While the model slightly overestimates viscosity at high concentrations (suggesting that a lower charge scaling, e.g., $\pm 0.75e$, might be optimal for transport properties), the overall performance is remarkable. The results provide a critical microscopic reference for understanding the solvation structure of perchloric acid and offer reliable thermodynamic data (including TMD predictions) for process design and further theoretical development in electrolyte modeling. The study also serves as a tribute to the scientific legacy of Prof. Stefan Sokołowski, highlighting the importance of rigorous molecular simulation in complex fluid research.