# # Collective hydrogen-bond rearrangement dynamics in liquid water

## # Simulations

Classical molecular dynamics (MD) simulations; 895 SPC/E water molecules in a cubic box of edge length 3 nm with periodic boundary conditions. Gromacs with a Berendsen weak coupling thermostat and barostat with a relaxation time of 1 ps for a fixed temperature of 300 K and a pressure of 1 bar. The time step is 2 fs; every 20 fs, the coordinates of every water molecules are stored and the total simulation time is 10 ns.

Why we need Berendsen thermostat algorithm?

The Berendsen thermostat is an algorithm to re-scale the velocities of particles in molecular dynamics simulations to **control the simulation temperature**.

Relaxation time

Ralaxation time is equal to the start point to the equilibrium?

In the last part of section 1: These slow processes are followed by a quasi-continuum of faster processes with characteristec times below 2 ps, which is consistent with the experimental finding that water relaxation processes span a wide range of time scales. What is the meaning relaxation processes here?

## # What they focused?

- Why they do this research? Despite the fundamental importance of the molecular dynamics (MD) in liquid water, the understanding of collective water restructuring events, such as the switching of a single H-bond from one accepting water molecule to another, remains challenging. "I just want to know" is their motivation.
- They focus on three water molecules meet some conditions.

## # How the study the collective features?

- How the locate the water molecules the want to focus?
- Markov state models: How did they define the states and transition probabilities?
- 12-dimensional configurational space of three water molecules. How did they choose which three water molecues in a large set of water molucules?

Markov state models

Markove state models (MSMs) are a class of models for modeling the long-timescale dynamics of molecular systems. They model the dynamics of a system as a series of memoryless, probabilitic jumps between a set of states. Practically, the model consists of (1) a set of conformational states, and (2) a matrix of transition probabilities between each pair of states.

## # What is the conclusions?

- In other words, collective processes are not necessarily slower than single-molecule processes.
- Using transition path analysis, we classify all possible pathways describing the H-bond switching from one accepting water molecule to a second accepting water molecule. The dominant transition pathways correspond to a direct transition to the new H-bond acceptor without a broken H-bond as an intermediate state, which make up about 66% of all transitions and have been investigated before.

## # What you can do next?

- Not select water dimmer or trimmer like this article.
- Study the vacancies and thickness relations for the the air-water interfaces.

## # What are the methods to study the dynamic processes?

[1] R. Schulz et al, J. Chem. Phys. 149, 244504 (2018)

# # Slow hydrogen-bond switching dynamics at the water surface revealed by theoretical two-dimensional sum-frequency spectroscopy

## # Simulations

MD simualtion with the E3B water model is performed in the canonical ensemble. The simulation box consists of 500 water molecules. The size of the rectangular box is 2.46nm x 2.46nm x 7.39nm, producing liquid and vapor slabs. Standard 3D periodic boundary conditions are applied. Electrostatic interactions are calculated using particle-mesh Ewald summation, and the Lennard–Jones interactions are truncated at 9 Å. The system is maintained at constant temperature (298 K) by means of a Berendsen thermostat with coupling parameter τ = 0.5 ps. The equations of motion are propagated with a 1-fs time step. After an equilibration run of 1 ns, the production run of 20 ns is performed and sampled every 10 fs. The GROMACS package, home modified for implementing the E3B model, is used to perform the simulation.

## # How they measure

- H-bond time-correlation function C(t)
- 2DSFG spectroscopy

## # Conclusions

- The E3B model has been developed with the expectation that it will be more accurate than two-body water models for sumulating water in heterogeneous environments such as the liquid/vapor interface.
- We find that the dynamics for hydrogen-bond rearrangement at the surface are a factor of about 3 slower than in the bulk due to the lower density of the surface.
- The comparison between theory and this experiment will provide more insight into the nuture of the water surface.

[1] Yicun Ni et al, PNAS, 110, 1992, (2013)

# # Hydrogen-bond dynamics in the air-water interface

## # Simulations

Because the molecular electronic structure can be greatly affected by the environment, the dynamical fluctuating-charge force field is a more suitable candidate than other fixed-charge force fields, such as TIP4P or SPC, in studying the two-phase equilibrated systems. WE thus use the TIP4)/FQ water model. We thus use the TIP4P/FQ water model. The TIP4P/FQ water model has the same geometry as the TIP4P water model, but the partial charges can fluctuate in response to environmental changes. For comparison, a new five-site water model, POL5, is chosen. In the meantime, the TIP4P43 water model and SPC/E42 are selected as the representatives for fixed-charge models. We use 3D periodic boundary conditions. The mesh-based approximations to the Ewald sum23 can be used to calculate the electrostatic interactions efficiently.24-27 We use the P3ME24 method. We choose a real-space cutoff of 11.25 Å. The corresponding Ewald splitting parameter is 0.376 Å-1. The choice of an isotropic Ewald splitting parameter requires that the density of k-space vectors remains the same in each Cartesian coordinate.28 For the P3ME method, this translates to a constant density of grid points in each coordinate. The grid- point density we use is 0.5 Å. **We use 3D periodic boundary conditions.** The mesh-based approximations to the Ewald sum23 can be used to calculate the electrostatic interactions efficiently. We use the P3ME24 method. We choose a real-space cutoff of 11.25 Å. The corresponding Ewald splitting parameter is 0.376 Å-1. The choice of an isotropic Ewald splitting parameter requires that the density of k-space vectors remains the same in each Cartesian coordinate. For the P3ME method, this translates to a constant density of grid points in each coordinate. The grid-point density we use is 0.5 Å.

## # How the measure?

- Desity profile along the z axis is calculated to illustrate the structural change from bulk water to the interface.
- Average number of H-bonds per water molecule.
- The H-bond autocorrelation funciton C(t), relaxation times.
- The effects of diffusion.

## # Conclusions

- Our results show a larger probability for hydrogen bonding on the surface than that in the bulk. The inference from these studies is that although there are fewer hydrogen bonds in the interface the water molecules have a stronger tendency to be hydrogen bonded with each other.
- It is found that as a water molecule nears the interface the relaxation of hydrogen bonds becomes faster, corresponding to an increase in the diffusion coefficient in the interfacial region.
- For the air-water interface system, the rate ofhydrogen-bond relaxation increases as a water molecule nears the interface, clearly showing that the role played by diffusion becomes more and more prominent when a water molecule nears the interface.

Faster or Slower?

In PNAS: Slower; Here: Faster?

## # What I learned

- I need to use MB-pol

New ideas?

- Can I use the movie at a certain thickness to determine the thickness of the water? What is the meanning of this application?
- Can I use collide to explain the properties of air-water interfaces?

[1] Pu Liu et al, J. Phys. Chem. B, 109, 2949, (2004)

# # Predicting hydration layers on surfaces using deep learning

[1] Adam S. Foster et al., Nanoscale Adv., 3, 3447, (2021)

# # Vapor–liquid equilibrium of water with the MB-pol many-body potential

## # How to simulate?

MBX, LAMMPS

Why they choose the temperature range 400 K - 600 K?

In this range water is water gas not liquid.

What is the difference between equilibrium and real MD process?

- In equilibrium process, the structure need to be optimized.

How did they get the initial positions of water molecules?

Maybe I can use PyIron to generate the initial positions.

However, when it comes to reproducing simultaneously vapor phase and VLE properties, such as vapor pressures and densities, surface tension, and critical properties, many pairwise rigid and/or non-poloarrizable models fall short, often due to their inability to represent polarization and/or many-body effects.

In the standard MB-pol model, intramolecular degrees of freedom are unconstrained, necessitating a small timestep size in MD simulations to accurately capture the motion of the hydrogen atoms. In order to access longer simulation timescales, we also testes a rigid variant of the MB-pol model in which we constrained the H-O bond length and H-O-H angle to the average values obtained from the flexible MB-pol simulations via the LAMMPS "fix_rigid" command. This rigid variant of the model allowed us to increase the timestep by a factor of 10.

To generate the vapor–liquid interfacial system in the direct coexistence geometry, we first equilibrated a liquid configuration in a cubic periodic simulation box, then rapidly expanded the z-dimension of the simulation box to produce a liquid slab in the x–y plane, with the z-dimension normal to the vapor–liquid interface.

TIP

This method is clever.

## # Is it good or not?

Good.

[1] Maria Carolina Muniz et al, J. Chem. Phys. 154, 211103, (2021)