🔬 physics

Hydrogen-Bond Donor-Acceptor Imbalance in Low-Frequency Terahertz Water Spectra

This study identifies the excess low-frequency terahertz response in liquid water as a localized, mixed nuclear-electronic dielectric phenomenon driven by transient imbalances in hydrogen-bond donor-acceptor numbers, evidenced by temperature and isotopic variations and supported by advanced molecular dynamics simulations.

Original authors: Lilian Najm Alsayed, Florian Pabst, Giuseppe Cassone, Ali Hassanali, Fabio Novelli

Published 2026-07-09

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Lilian Najm Alsayed, Florian Pabst, Giuseppe Cassone, Ali Hassanali, Fabio Novelli

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

The Big Picture: Listening to Water's "Hum"

Imagine liquid water not just as a wet substance, but as a bustling crowd of people (molecules) holding hands. These "handshakes" are called hydrogen bonds. Usually, everyone in the crowd is perfectly balanced: two people hold your hand on the left, and two hold your hand on the right.

Scientists have long known that if you shine a special kind of light on water (called Terahertz light, which is like a very high-pitched hum), the water reacts in a specific way. They usually describe this reaction with two main "notes":

The Slow Note: A deep, slow rumble caused by groups of people turning around together.
The Fast Note: A quicker, jittery sound that scientists have struggled to explain.

This paper asks a simple question: What is causing that "Fast Note"?

The New Idea: The "Stuck" Car Analogy

The researchers propose a new way to listen to that fast sound. Instead of thinking of it as a second type of turning, they suggest it's like a car that starts to move but immediately hits a wall and bounces back.

The Old View (Drude Model): Imagine a car driving down a highway. It speeds up and slows down smoothly. This works for metals (where electricity flows freely), but water isn't a metal; electricity can't flow through it easily.
The New View (Drude-Smith Model): Imagine a car in a tiny, crowded parking garage. It tries to drive forward, but it hits a pillar, bounces backward, hits another pillar, and gets stuck in a loop. It never actually gets anywhere (zero "DC conductivity"), but it is still vibrating and moving locally.

The authors found that describing water's fast sound using this "stuck car" model (where the car is trapped in a local spot) fits the data perfectly.

The Discovery: Imbalance is Key

Why does the water get "stuck" or vibrate this way? The paper suggests it's because of imbalances in the handshakes.

In a perfect crowd, everyone has two hands on the left and two on the right. But in reality, at any split second, some people are holding three hands on the left and only one on the right, or vice versa.

The Imbalance: When a water molecule has an unequal number of "donated" and "accepted" handshakes, the electrical charge inside that tiny spot becomes lopsided.
The Effect: This lopsided charge creates a local "jitter" that the new "stuck car" model detects.

Testing the Theory: Heat and Heavy Water

To prove this, the scientists tested two scenarios:

Heating the Water (20°C to 50°C):
- Analogy: Imagine heating the crowd up. People start dancing more wildly, breaking their perfect handshakes. More people end up with uneven handshakes (3 on one side, 1 on the other).
- Result: The "stuck car" vibration got stronger. The model showed that as the crowd got more chaotic, the signal from these imbalanced molecules grew.
Using Heavy Water (D2O):
- Analogy: Imagine the people in the crowd are wearing heavy winter coats (deuterium atoms are heavier than normal hydrogen). They move slower and hold hands more tightly. The crowd becomes more orderly, and fewer people have uneven handshakes.
- Result: The "stuck car" vibration got weaker. The signal dropped because the crowd was more balanced.

The "Ghost" in the Machine: Nuclear vs. Electronic

Finally, the researchers used super-computers to look inside the molecules to see what was actually moving. They found that the "Fast Note" isn't just one thing; it's a mix of two things happening at once:

Nuclear Motion: The physical atoms (the nuclei) wiggling around.
Electronic Motion: The invisible "cloud" of electrons shifting and redistributing charge.

The Metaphor: Imagine a dancer (the nucleus) spinning on a stage. Usually, we just watch the dancer. But in water, the dancer's costume (the electron cloud) is also changing shape and color at the exact same time. The "Fast Note" we hear is the sound of both the dancer spinning and the costume shifting.

Summary

This paper suggests that the mysterious high-frequency sound of water isn't just about molecules turning around. It is a specific signal caused by temporary imbalances in how water molecules hold hands. When the handshakes are uneven, the electrical charge gets "stuck" in a local spot, creating a unique vibration. This vibration is a mix of the atoms moving and the electrons shifting, and it gets stronger when the water is hot (chaotic) and weaker when the water is heavy (orderly).

Technical Summary: Hydrogen-Bond Donor-Acceptor Imbalance in Low-Frequency Terahertz Water Spectra

Problem Statement
The low-frequency dielectric response of liquid water (0.14–1.21 THz) is traditionally modeled as a dominant Debye relaxation (~8–10 ps) accompanied by faster contributions. While the main Debye term is widely accepted as representing collective dipolar reorientation, the microscopic origin of the additional faster contributions remains debated. Previous attempts to model this "excess" response have utilized additional Debye terms, log-normal distributions, or Lorentzian functions, often associating them with weakly hydrogen-bonded structures. However, these models have yielded inconsistent trends regarding temperature and isotopic substitution, leaving the physical nature of this spectral feature unresolved.

Methodology
The authors employed a multi-faceted approach combining experimental spectroscopy, empirical modeling, molecular dynamics (MD) simulations, and ab-initio calculations:

Experimental Spectroscopy: Terahertz time-domain spectroscopy (THz-TDS) was used to measure the complex refractive index and extinction coefficient of pure H₂O at 20°C and 50°C, and heavy water (D₂O) at 20°C.
Empirical Modeling: The experimental dielectric functions were fitted using two competing models:
- The standard double-Debye approach (two relaxation time constants).
- A single Debye plus Drude-Smith approach. In this model, the Drude-Smith term is constrained to the zero-dc-conductivity limit (localisation parameter $c = -1$ ), representing a localized, non-transporting dielectric response.
Molecular Dynamics Simulations:
- Neural Network MD: Path-integral deep potential molecular dynamics (PI-DPMD) simulations were conducted for H₂O and D₂O at 20°C to include nuclear quantum effects. These simulations classified water molecules based on their instantaneous hydrogen-bond (HB) coordination (donated vs. accepted).
- MB-pol Simulations: Classical MB-pol simulations were used to assess temperature dependence (comparing 17°C and 57°C) and to calculate the imaginary part of the dielectric function.
Ab-Initio Spectra Calculations: Ab-initio based calculations using the MB-pol model compared two dipole definitions:
- A fixed partial-charge approximation (excluding explicit electronic polarization and charge redistribution).
- A neural network trained on Wannier centers (including nuclear motion, electronic polarization, and intermolecular charge redistribution).

Key Results

Model Equivalence: Both the double-Debye and the single Debye plus Drude-Smith ( $c=-1$ ) models reproduce the experimental dielectric response between 0.14 and 1.21 THz with comparable residual errors.
Drude-Smith Spectral Weight Trends: The effective spectral weight ( $\omega_P^2$ $ω_{P}^{2}$ ) of the Drude-Smith term exhibits distinct, consistent trends:
- It increases by approximately 18% when heating H₂O from 20°C to 50°C.
- It decreases by approximately 10% when substituting H₂O with D₂O at 20°C.
Correlation with HB Imbalance: MD simulations reveal a direct correlation between the Drude-Smith spectral weight and the population of water molecules with an unequal number of donated and accepted hydrogen bonds (donor-acceptor imbalance):
- The population of imbalanced molecules increases by ~22% upon heating (17°C to 57°C), matching the experimental spectral weight increase.
- The population of imbalanced molecules decreases by ~14% in D₂O compared to H₂O, matching the experimental spectral weight decrease.
Microscopic Origin: Ab-initio calculations indicate that the 0.1–1 THz response is not purely nuclear or purely electronic. The difference between the Wannier-center (full) and fixed-charge spectra in this frequency range is non-zero and negative, indicating anticorrelated nuclear-electronic motions. This suggests the response arises from a mixed mechanism involving nuclear-network rearrangements, intermolecular charge redistribution, and polarization effects associated with local HB asymmetries.

Significance and Claims
The paper claims that the low-frequency THz excess response in water is a localized, mixed nuclear-electronic dielectric response correlated with transient donor-acceptor imbalances in the hydrogen-bond network.

Novel Interpretation: Unlike previous models that attribute the fast response to weak HBs or collision processes, this work proposes that the Drude-Smith formalism (with $c=-1$ ) provides a physically intuitive description of localized charge fluctuations arising from structural asymmetry (unequal donor/acceptor counts).
Consistency: The Drude-Smith approach resolves the inconsistencies found in previous multi-Debye fits by providing a parameter (spectral weight) that scales consistently with structural changes induced by temperature and isotopic substitution.
Physical Mechanism: The study establishes that the dielectric response in this regime is driven by the interplay of nuclear motion and electronic charge redistribution (polarization/charge transfer) within asymmetric local environments, rather than long-range transport or simple reorientation.

The authors conclude that this empirical approach complements standard double-Debye procedures, offering a simpler analytical form that yields deeper microscopic insights into the relationship between water's hydrogen-bond network topology and its dielectric properties.