Broadband molecular dynamics simulation of confined Brownian motion
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.
Kai: We're looking at the paper: "Broadband molecular dynamics simulation of confined Brownian motion".
Mira: Here's the gist of the paper. Broadband molecular dynamics simulation of fluid inertial effects in confined Brownian motion shows that confinement modifies the full hydrodynamic response across a broad spectrum of timescales by reshaping the velocity autocorrelation function.
Kai: Hydrodynamic Memory and Observables Hydrodynamic memory governs Brownian motion over a broad range of timescales, from acoustic wave propagation at short times to diffusive relaxation at long times.
Mira: The velocity autocorrelation function (VACF) is identified as a central quantity because it contains dissipation, inertia, and memory in a single time-dependent quantity.
Kai: In the bulk, the simulations recover compressibility, added mass, the hydrodynamic long-time tail, and Stokes-Einstein diffusion without adjustable parameters.
Title and authors: Kai: So let's talk about who wrote this, the paper "Broadband molecular dynamics simulation of confined Brownian motion." The authors are Thomas Salez and Clara Mefo Sop with Maxime Lavaud and Yacine Amarouchene, plus Pascal Damman.
Mira: I see they’ve got a good mix here. Lavaud is involved in the modeling side, which is key for getting these molecular dynamics setups right, and the others are bringing in different expertise to look at the physics.
Lev: I'm interested in seeing how they set up the molecular model itself because that dictates what we can actually measure on real hardware down the line.
Kai: Yeah, that's exactly what I wonder about. Because this paper uses a neutral buoyant colloidal particle in an explicit solvent, which sounds like a solid experimental setup you could try to build to test these ideas.
Mira: And they are fixing things like the solvent density, viscosity, and sound velocity independently so they can measure the bulk added mass without having to guess any of those constants.
Lev: That’s good because it means the starting point for their calculations is really grounded in measured fluid properties, not just some theoretical assumptions about the liquid.
The paper's summary: Kai: So, what's the main thrust of this "Broadband molecular dynamics simulation of confined Brownian motion" paper? It’s about seeing how confinement modifies the full hydrodynamic response across all timescales.
Mira: They use the velocity autocorrelation function as their central tool because it captures everything: dissipation, inertia, and memory in one single time-dependent quantity.
Lev: By looking at that VACF, they can see how the fluid's memory is reorganized when you introduce a boundary condition like a wall.
Kai: In bulk, they get the standard results—compressibility and the long-time tail of diffusion—but near the wall, things change dramatically in how fast those correlations decay.
Mira: They show that near a rigid wall, you get different algebraic tails for parallel motion versus perpendicular motion; specifically, they see C parallel decaying like t to the power of minus five over two and perpendicular decaying like t to the power of minus seven over two.
Lev: Those specific exponents are what really tell us how the wall reshapes that hydrodynamic kernel, as they call it.
The paper's improvements: Kai: Now, let's talk about what they suggest as improvements or new insights from this work. They really push the idea of using the VACF to do a lot more than just check bulk stuff.
Mira: They are suggesting that because you have that VACF, you can extract both the low-frequency mobility and the high-frequency inertial response using different parts of it.
Lev: That’s a big deal for hardware simulation because it means we can potentially measure both things without needing two separate experimental setups or two different ways to analyze the data.
Kai: They also suggest that this approach gives us a practical way to check the validity range of continuum hydrodynamics, by comparing their MD results against theoretical predictions for added mass, like Yang’s compact expression for the wall-induced mass amplification.
Mira: And they point out that confinement allows us to discriminate between different physical mechanisms dominating those effects—whether it’s steady drag or elastic storage or inertial coupling—just by looking at the time scales involved.
Conclusion: Kai: So to wrap up, this paper on "Broadband molecular dynamics simulation of confined Brownian motion" shows that confinement modifies the full hydrodynamic memory from acoustic to diffusive regimes in a way that is direction-dependent.
Mira: The main physical result they highlight is the coexistence of reduced mobility and increased inertia near a wall, which happens because the low-frequency response is sensitive to no-slip boundary conditions while the high-frequency response is governed by an effectively inviscid flow constrained by that same boundary.
Lev: From an engineering standpoint, this confirms that MD simulations are well-suited for probing free thermal motion across all those regimes—acoustic, inertial, hydrodynamic memory and diffusive ones.
Kai: It’s a great tool because it connects the two extremes of the response spectrum in a way that traditional methods can’t easily do.
Mira: They suggest that using the VACF-based analysis can help us distinguish whether the confinement effects are dominated by steady drag, elastic storage, inertial coupling, or acoustic propagation.
Lev: I just think this work is a solid numerical validation of known wall corrections and shows how we can use these simulations to test continuum predictions over a wide dynamical range.
Kai: Yeah, it’s important for understanding how these systems behave when they are constrained in real devices. That’s the big picture here for the paper "Broadband molecular dynamics simulation of confined Brownian motion."
Laboratoire Interfaces & Fluides Complexes, University of Mons · University of Bordeaux, CNRS, LOMA · Faculty of Physics, University of Vienna · Mechanics Department, Ecole Polytechnique Institut Polytechnique de Paris
cond-mat.soft, cond-mat.stat-mech, physics.flu-dyn
Submitted: 2026-06-23
Updated: 2026-11-06
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
Importance score: 92/100
The gist: Broadband molecular dynamics simulation of fluid inertial effects in confined Brownian motion shows that confinement modifies the full hydrodynamic response across a broad spectrum of timescales by…
Key concepts
- Velocity Autocorrelation Function (VACF)
- The VACF is a single time-dependent quantity that captures all aspects of fluid motion, including how energy dissipates, how inertia behaves, and the memory of past interactions. It is the central tool used to analyze Brownian motion across different timescales.
- Hydrodynamic Memory
- This refers to the fluid's ability to retain information about past movements over time. In this simulation, confinement modifies this memory kernel; correlations that usually last a long time decay faster and become dependent on whether the movement is parallel or perpendicular to the wall.
- Mobility vs. Inertia Contrast
- A key finding is that confinement leads to two opposite effects: mobility (how easily it moves) is reduced, while inertia (resistance to acceleration) is increased. This contrast arises because the low-frequency response reflects viscous drag, while the high-frequency response reflects an effectively inviscid flow constrained by the wall.
Terminology
Summary
Broadband molecular dynamics simulation of fluid inertial effects in confined Brownian motion shows that confinement modifies the full hydrodynamic response across a broad spectrum of timescales by reshaping the velocity autocorrelation function. p. 0 While confinement-induced
corrections to Brownian diffusion are well established, how confinement modifies the full hydrodynamic response remains less explored. In this Letter, we use molecular-dynamics simulations of
a neutrally buoyant colloidal particle in an explicit solvent to resolve the velocity autocorrelation
function across a broad hydrodynamic spectrum.Where the paper says it
Hydrodynamic Memory and Observables
Hydrodynamic memory governs Brownian motion over a broad range of timescales, from acoustic wave propagation at short times to diffusive relaxation at long times. The velocity autocorrelation function (VACF) is identified as a central quantity because it contains dissipation, inertia, and memory in a single time-dependent quantity. p. 1 (1)
The three terms are from left to right : the Stokes drag,
the added-mass contribution, and the Basset history
force associated with vorticity diffusion. Equation (1)
makes it explicit why the VACF is a natural observable:
it contains dissipation, inertia, and memory in a single
time-dependent quantity. In this Letter, we use molecular-dynamics simulations of
a neutrally buoyant colloidal particle in an explicit solvent to resolve the velocity autocorrelation
function across a broad hydrodynamic spectrum. In the bulk, the simulations recover compressibility, added mass, the hydrodynamic long-time tail, and Stokes-Einstein diffusion without adjustable
parameters. The molecular model, wall preparation, sampling protocol, and uncertainty analysis are
detailed in the Supplementary Material [24]. The starting point is the unsteady force exerted on
a sphere in an incompressible fluid.Where the paper says it
Where the paper says it
Where the paper says it
Confinement Effects on Dynamics
Boundaries enrich this physics by reorganizing the flow responsible for memory. Near a rigid wall, the velocity correlations become anisotropic, their algebraic tails are modified, and the diffusion coefficients are reduced. p. 0 In the bulk, the simulations recover compressibility, added mass, the hydrodynamic long-time tail, and Stokes-Einstein diffusion without adjustable
parameters. Near a rigid wall, the velocity correlations become anisotropic, their algebraic tails are
modified, and the diffusion coefficients are reduced. Near a rigid wall, the velocity correlations become anisotropic, their algebraic tails are
modified, and the diffusion coefficients are reduced. Most importantly, the short-time dynamics reveals a pronounced enhancement of the effective added mass as the wall is approached. This protocol separates
the two directional memory responses while avoiding an
external trapping potential in the measured direction. Confinement leaves the early acoustic response almost
unchanged, but it strongly modifies the subsequent hydrodynamic relaxation.Where the paper says it
Where the paper says it
Where the paper says it
Frequency Domain Analysis
The VACF yields the diffusion coefficient by integration, and its early-time structure gives the effective inertia of the particle-fluid system. The low-frequency mobility is obtained from the diffusion coefficients, while the high-frequency inertial response is obtained from the early post-compressible normalization. p. 1 Having established that the wall reshapes the memory
kernel itself, we now exploit the fact that a single VACF
encodes two opposite limits of the same response. The
low-frequency mobility is obtained from the diffusion coefficients, as:
Di =
Z ∞
0
Ci(t) dt, i =∥, ⊥, (5)
and the high-frequency inertial response is obtained from
the early post-compressible normalization Ci ≃ kBT /M∗
i
. 4
a result of different limiting manifestations of the same
wall-modified hydrodynamic kernel. The low-frequency
response is sensitive to no-slip boundary conditions and
lubrication.Where the paper says it
Where the paper says it
Validation and Limitations
The agreement of the predictions with simulation data demonstrates that thermal MD captures boundary effects over a wide dynamical range. The approach provides a practical way to assess the range of validity of continuum hydrodynamics, restricting simulations to gaps large compared with solvent molecular size. p. 3 The approach also provides a practical way to assess
the range of validity of continuum hydrodynamics. When
the gap is large compared with the solvent diameter,
the measured VACFs can be interpreted using continuum predictions for the algebraic tails, steady drag, and
added mass. Agreement with both steady mobility corrections and potential-flow added-mass theories
therefore indicates that the MD simulation with explicit
solvent captures the boundary effects over a wide dynamical range. Deviations are expected only when the gap
approaches the molecular diameter, where continuum hydrodynamics, the definition of the particle surface, and
possible slip become ambiguous. Once
regularized, the resulting kernel would allow to identify
the crossover frequencies between compressibility, added
mass, wall-modified vorticity diffusion, and steady mobility. This perspective shows why the VACF is a more
complete observable than the mean-square displacement:
it retains the temporal ordering of hydrodynamic processes that are integrated out in the diffusion coefficient. The same boundary condition therefore produces two opposite signatures: a reduced mobility and
an increased inertia. This contrast is the central physical result of the present work, and the use of an explicit
molecular solvent is essential for connecting the two extreme limits without changing models. The same boundary condition therefore produces two opposite signatures: a reduced mobility and
an increased inertia. This contrast is the central physical result of the present work, and the use of an explicit
molecular solvent is essential for connecting the two extreme limits without changing models.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Summary of Key Findings
-
Confinement modifies the full hydrodynamic memory from acoustic to diffusive regimes. p. 0
Where the paper says it
Bordeaux, CNRS, LOMA, UMR 5798, 33405 Talence, France 3Faculty of Physics, University of Vienna, Boltzmanngasse 5, 1090 Vienna, Austria 4Mechanics Department, Ecole Polytechnique, Institut Polytechnique de Paris, 91128 Palaiseau, France (Dated: June 24, 2026) Hydrodynamic memory governs Brownian motion over a broad range of timescales, from acoustic wave propagation at short times to diffusive relaxation at long times. While confinement-induced corrections to Brownian diffusion are well established, how confinement modifies the full hydrodynamic response remains less explored.
-
The wall reshapes the memory kernel, causing long-lived hydrodynamic correlations to decay faster and become direction-dependent. p. 1
Where the paper says it
The central result is not merely that diffusion changes near a wall, but that the temporal structure of the memory kernel is reshaped: long-lived hydrodynamic correlations decay faster and become direction-dependent. Having established that the wall reshapes the memory kernel itself, we now exploit the fact that a single VACF encodes two opposite limits of the same response.
-
The low-frequency response is sensitive to no-slip boundary conditions and lubrication. p. 3
Where the paper says it
4 a result of different limiting manifestations of the same wall-modified hydrodynamic kernel. The low-frequency response is sensitive to no-slip boundary conditions and lubrication.
-
The high-frequency response is governed by an effectively inviscid flow constrained by the same boundary. p. 3
Where the paper says it
The low-frequency response is sensitive to no-slip boundary conditions and lubrication. The high-frequency response, by contrast, is governed by an effectively inviscid flow constrained by the same boundary.
-
Confinement increases in fact the amount of fluid that is dynamically accelerated with the particle, despite apparent mobility suppression. p. 1
Where the paper says it
We quantify this by the relative added-mass deviation: ∆M∗ i (ϵ) = M∗ i (ϵ) − M∗ i (∞) M∗ i (∞) , i =∥, ⊥ . (6) This counterintuitive enhancement reveals that confinement increases in fact the amount of fluid that is dynamically accelerated with the particle, despite the apparent
Figure Interpretation
Figure 1 shows a schematic of the system, illustrating the colloidal particle immersed in a dense WCA solvent and confined near a frozen planar wall. Figure 2 displays bulk VACF data, showing short-time compressibility and long-time asymptotic behavior. p. 1 Figure 2 shows the bulk VACF computed from MD
simulations. At very short times, the solvent is compressible and the acoustic time is directly measured as
τc = R/cs (Fig. Figure 2 shows the bulk VACF computed from MD
simulations. At very short times, the solvent is compressible and the acoustic time is directly measured as
τc = R/cs (Fig. Figure 3 shows that the measured
VACFs indeed display this wall-induced anisotropy. The
central result is not merely that diffusion changes near
a wall, but that the temporal structure of the memory
kernel is reshaped: long-lived hydrodynamic correlations
decay faster and become direction-dependent. Figure 3 shows that the measured
VACFs indeed display this wall-induced anisotropy. The
central result is not merely that diffusion changes near
a wall, but that the temporal structure of the memory
kernel is reshaped: long-lived hydrodynamic correlations
decay faster and become direction-dependent. Also shown are theoretical predictions as well as experimental results from the literature, as indicated. (b) Relative
added-mass deviation, [M∗
(ϵ) − M∗
(∞)]/M∗
(∞), as a function of the normalized distance to the wall ϵ/R. Also shown are theoretical predictions as well as experimental results from the literature, as indicated. (b) Relative
added-mass deviation, [M∗
(ϵ) − M∗
(∞)]/M∗
(∞), as a function of the normalized distance to the wall ϵ/R. For normal motion, the smallgap limit recovers the lubrication scaling D⊥/D0 ≃ ϵ/R,
while the parallel component displays the weaker logarithmic reduction expected for shear-dominated flow. This confirms that the explicit molecular solvent reproduces the zero-frequency confined response. For normal motion, the smallgap limit recovers the lubrication scaling D⊥/D0 ≃ ϵ/R,
while the parallel component displays the weaker logarithmic reduction expected for shear-dominated flow. This confirms that the explicit molecular solvent reproduces the zero-frequency confined response. (9) over the full
frequency range. In practice, this requires a controlled
transform of a noisy, finite-time VACF and careful treatment of the acoustic oscillations at the shortest times
and algebraic tails at the longest times. More complex situations, such as
patterned walls, soft interfaces, liquid-liquid boundaries,
or viscoelastic media, should affect both the real and
imaginary parts of Γ˜
i(ω) in a geometry-specific manner. The same VACF-based analysis could then distinguish
whether an observed confinement effect is dominated by
steady drag, elastic storage, inertial coupling, or acoustic
propagation.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Conclusion and Future Directions
The same boundary condition produces two opposite signatures: a reduced mobility and an increased inertia. This contrast is the central physical result of the present work. p. 3 The same boundary condition therefore produces two opposite signatures: a reduced mobility and
an increased inertia. This contrast is the central physical result of the present work, and the use of an explicit
molecular solvent is essential for connecting the two extreme limits without changing models. A two-wall geometry mainly
changes the low-frequency part of the response and is
therefore included as an additional validation in the Supplementary Material. More complex situations, such as
patterned walls, soft interfaces, liquid-liquid boundaries,
or viscoelastic media, should affect both the real and
imaginary parts of Γ˜
i(ω) in a geometry-specific manner. More complex situations, such as
patterned walls, soft interfaces, liquid-liquid boundaries,
or viscoelastic media, should affect both the real and
imaginary parts of Γ˜
i(ω) in a geometry-specific manner. The same VACF-based analysis could then distinguish
whether an observed confinement effect is dominated by
steady drag, elastic storage, inertial coupling, or acoustic
propagation. Optical trapping and
atomic force microscopy experiments measure confined
Brownian motion with precision, but the external trap
modifies the low-frequency response and the compressible
regime is often beyond temporal resolution. By contrast,
the MD trajectories provide direct access to free thermal motion over the acoustic, inertial, hydrodynamicmemory, and diffusive regimes.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
References
[1] B. Duplantier, Le mouvement brownien, divers et ondoyant. p. 4 damman@umons. Duplantier, Le mouvement brownien, divers et ondoyant. damman@umons. Duplantier, Le mouvement brownien, divers et ondoyant.Where the paper says it
Where the paper says it
[2] X., Bian, C. Kim, and G. p. 4 sèminaire Poincaré 1, 155 (2005). Kim, and G. sèminaire Poincaré 1, 155 (2005). Kim, and G.Where the paper says it
Where the paper says it
[3] B. J. Alder and T. p. 0 Momentum transferred from a particle to the surrounding liquid is transported through the
fluid and subsequently feeds back onto the particle motion. This hydrodynamic memory gives rise to algebraic
long-time tails in the velocity autocorrelation function
(VACF), as first revealed by Alder and Wainwright [3]
and later described by fluctuating and unsteady hydrodynamics [4–8]. Momentum transferred from a particle to the surrounding liquid is transported through the
fluid and subsequently feeds back onto the particle motion. This hydrodynamic memory gives rise to algebraic
long-time tails in the velocity autocorrelation function
(VACF), as first revealed by Alder and Wainwright [3]
and later described by fluctuating and unsteady hydrodynamics [4–8]. Alder and T. Wainwright, Decay of the velocity
autocorrelation function, Phys. Alder and T. Wainwright, Decay of the velocity
autocorrelation function, Phys.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[4] R. Zwanzig and M. p. 4 A 1, 18 (1970). Zwanzig and M. A 1, 18 (1970). Zwanzig and M. A 1, 18 (1970). Zwanzig and M. A 1, 18 (1970). Zwanzig and M. Zwanzig and M. Bixon, Hydrodynamic theory of
the velocity correlation function, Phys. Zwanzig and M. Bixon, Hydrodynamic theory of
the velocity correlation function, Phys.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[5] R. Zwanzig and M. Bixon, Compressibility effects in the hydrodynamic theory of Brownian motion, J. p. 4 Zwanzig and M. Bixon, Compressibility effects in
the hydrodynamic theory of Brownian motion, J. Zwanzig and M. Bixon, Compressibility effects in
the hydrodynamic theory of Brownian motion, J. Bixon, Compressibility effects in
the hydrodynamic theory of Brownian motion, J. Fluid
Mech. Bixon, Compressibility effects in
the hydrodynamic theory of Brownian motion, J. Fluid
Mech. Bixon, Compressibility effects in
the hydrodynamic theory of Brownian motion, J. Fluid
Mech. Bixon, Compressibility effects in
the hydrodynamic theory of Brownian motion, J. Fluid
Mech.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[6] E. J. Hinch, J. Application of the Langevin equation to fluid suspensions, Fluid Mech. p. 4 69, 21 (1975). Application of the Langevin equation to
fluid suspensions, Fluid Mech. 69, 21 (1975). Application of the Langevin equation to
fluid suspensions, Fluid Mech.Where the paper says it
Where the paper says it
[7] H. J. H. Clercx and P. p. 4 72, 499 (1975). Clercx and P. 72, 499 (1975). Clercx and P. Clercx and P. Schram, Brownian particles in shear flow and harmonic potentials: A study of
long-time tails, Phys.Where the paper says it
Where the paper says it
Where the paper says it
[8] T. Li and M. G. Raizen, Brownian motion at short time scales, Ann. p. 4 A 46, 1942 (1992). Raizen, Brownian motion at short time
scales, Ann.Where the paper says it
[9] S. Jeney, B. Lukić, J. A. Kraus, T. Franosch, and L. p. 4 525, 281 (2013). Franosch, and L. Franosch, and L. Forró, Anisotropic memory effects in confined colloidal
diffusion, Phys. Franosch, and L. Forró, Anisotropic memory effects in confined colloidal
diffusion, Phys.Where the paper says it
Where the paper says it
Where the paper says it
[10] Li, Tongcang, Kheifets, Simon, Medellin, David and Raizen, Mark G Measurement of the instantaneous velocity of a Brownian particle, Science 328 1673 (2010)
[11] T. Franosch et al. p. 4 525, 281 (2013). Franosch, and L. 525, 281 (2013). Franosch, and L.Where the paper says it
Where the paper says it
[12] Kheifets et al., Observation of Brownian motion in liquids at short times: instantaneous velocity and memory loss, science 343 1493 (2014)
[13] H. Brenner, The slow motion of a sphere through a viscous fluid towards a plane surface, Chem. p. 4 [12] Kheifets, Simon, Simha, Akarsh, Melin, Kevin, Li, Tongcang and Raizen, Mark G, Observation of Brownian motion in liquids at short times: instantaneous velocity and
memory loss, science, 343, 1493 (2014). Brenner, The slow motion of a sphere through a viscous fluid towards a plane surface, Chem. [12] Kheifets, Simon, Simha, Akarsh, Melin, Kevin, Li, Tongcang and Raizen, Mark G, Observation of Brownian motion in liquids at short times: instantaneous velocity and
memory loss, science, 343, 1493 (2014). Brenner, The slow motion of a sphere through a viscous fluid towards a plane surface, Chem.Where the paper says it
Where the paper says it
[14] L. P. Faucheux and A. p. 3 By contrast,
the MD trajectories provide direct access to free thermal motion over the acoustic, inertial, hydrodynamicmemory, and diffusive regimes. Additional checks, including the dependence on particle radius, the robustness
of mass extraction, and the two-wall Faucheux-Libchaber
geometry, are reported in the Supplementary Material. Faucheux and A. Libchaber, Confined brownian
motion, Phys.Where the paper says it
Where the paper says it
[15] B. Lin, J. Yu, and S. A Rice, Direct measurements of constrained Brownian motion of an isolated sphere between two walls, Phys. p. 4 E 49, 5158 (1994). Rice, Direct measurements of
constrained Brownian motion of an isolated sphere between two walls, Phys. E 49, 5158 (1994). Rice, Direct measurements of
constrained Brownian motion of an isolated sphere between two walls, Phys.Where the paper says it
Where the paper says it
[16] J. Mo et al., Broadband boundary effects on Brownian motion, Phys. p. 4 Simha, and M. Raizen, Broadband boundary effects on Brownian motion, Phys. Simha, and M. Raizen, Broadband boundary effects on Brownian motion, Phys.Where the paper says it
Where the paper says it
[17] M. Matse et al., Test of the diffusing-diffusivity mechanism using near-wall colloidal dynamics, Physical Review E 96 042604 (2017)
[18] M. Lavaud et al. p. 0 Broadband molecular dynamics simulation of fluid inertial effects in confined
Brownian motion
Quentin Thomas,1 Clara Mefo Sop,1 Maxime Lavaud,2, 3 Yacine
Amarouchene,2, ∗ Thomas Salez,2, 4, † and Pascal Damman1, ‡
1Laboratoire Interfaces & Fluides Complexes, Université de Mons, 20 Place du Parc, B-7000 Mons, Belgium
2Univ. Bordeaux, CNRS, LOMA, UMR 5798, 33405 Talence, France
3Faculty of Physics, University of Vienna, Boltzmanngasse 5, 1090 Vienna, Austria
4Mechanics Department, Ecole Polytechnique, Institut Polytechnique de Paris, 91128 Palaiseau, France
(Dated: June 24, 2026)
Hydrodynamic memory governs Brownian motion over a broad range of timescales, from acoustic
wave propagation at short times to diffusive relaxation at long times. Broadband molecular dynamics simulation of fluid inertial effects in confined
Brownian motion
Quentin Thomas,1 Clara Mefo Sop,1 Maxime Lavaud,2, 3 Yacine
Amarouchene,2, ∗ Thomas Salez,2, 4, † and Pascal Damman1, ‡
1Laboratoire Interfaces & Fluides Complexes, Université de Mons, 20 Place du Parc, B-7000 Mons, Belgium
2Univ. Bordeaux, CNRS, LOMA, UMR 5798, 33405 Talence, France
3Faculty of Physics, University of Vienna, Boltzmanngasse 5, 1090 Vienna, Austria
4Mechanics Department, Ecole Polytechnique, Institut Polytechnique de Paris, 91128 Palaiseau, France
(Dated: June 24, 2026)
Hydrodynamic memory governs Brownian motion over a broad range of timescales, from acoustic
wave propagation at short times to diffusive relaxation at long times. Louyer, and Y. Amarouchene, Stochastic inference of surface-induced effects using Brownian motion, Phys. Louyer, and Y. Amarouchene, Stochastic inference of surface-induced effects using Brownian motion, Phys.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[19] Z. Zhang et al., Unsteady drag force on an immersed sphere oscillating near a wall, J. p. 4 Salez, and A. Maali, Unsteady drag
force on an immersed sphere oscillating near a wall, J. Salez, and A. Maali, Unsteady drag
force on an immersed sphere oscillating near a wall, J.Where the paper says it
Where the paper says it
[20] P. Palacios-Alonso et al. p. 4 , 977, A21 (2023). Palacios-Alonso, R. , 977, A21 (2023). Palacios-Alonso, R. Pérez Peláez, and R. DelgadoBuscalioni, Fast spectral solver for viscoelastic structures
under oscillatory flow in free space or wall-bounded domains: Applications to quartz crystal microbalance and
force spectroscopy, J. Pérez Peláez, and R. DelgadoBuscalioni, Fast spectral solver for viscoelastic structures
under oscillatory flow in free space or wall-bounded domains: Applications to quartz crystal microbalance and
force spectroscopy, J.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[21] N. Bigan et al., Long range signature of liquid’s inertia in nanoscale drainage flows, Soft Matter 20 8804 (2024)
[23] A.P. Thompson et al. p. 4 09193 (2026). Thompson, et al. 09193 (2026). Thompson, et al.Where the paper says it
Where the paper says it
[24] See Supplementaary Material for simulation details, fluid characterization, added-mass extraction, radius dependence, and additional confined geometries
[25] B. U. Felderhof, Effect of the wall on the velocity autocorrelation function and long-time tail of Brownian motion, J. p. 4 [24] See Supplementaary Material (SM) for simulation details, fluid characterization, added-mass extraction, radius dependence, and additional confined geometries. Felderhof, Effect of the wall on the velocity autocorrelation function and long-time tail of Brownian motion,
J. [24] See Supplementaary Material (SM) for simulation details, fluid characterization, added-mass extraction, radius dependence, and additional confined geometries. Felderhof, Effect of the wall on the velocity autocorrelation function and long-time tail of Brownian motion,
J.Where the paper says it
Where the paper says it
[26] M. D. Carbajal-Tinoco et al. p. 4 B 109, 21406 (2005). Carbajal-Tinoco, R. Lopez-Fernandez, and J. Arauz-Lara, Asymmetry in colloidal diffusion near a rigid
wall, Phys.Where the paper says it
Where the paper says it
[27] H. Lamb, Hydrodynamics, 6th ed. p. 4 99, 138303 (2007). Lamb, Hydrodynamics, 6th ed.Where the paper says it
[28] F. Yang, A formula for the wall-amplified added mass coefficient for a solid sphere in normal approach to a wall and its application for such motion at low Reynolds number, Phys. p. 4 (Cambridge University
Press, Cambridge, 1932). Yang, A formula for the wall-amplified added mass
coefficient for a solid sphere in normal approach to a wall
and its application for such motion at low Reynolds number, Phys.Where the paper says it
[29] A. A. Kharlamov et al. p. 4 Fluids 22, 073302 (2010). Kharlamov, Z. Chára, and P. Vlasák, Hydraulic
formulae for the added masses of an impermeable sphere
moving near a plane wall, J.Where the paper says it
Where the paper says it
Figure Descriptions
Figure 1 shows a schematic of the system, illustrating the colloidal particle immersed in a dense WCA solvent and confined near a frozen planar wall. Figure 2 displays bulk VACF data, showing short-time compressibility and long-time asymptotic behavior. Figure 3 presents VACFs measured near a rigid wall, displaying wall-induced anisotropy with long-time tails decaying as C∥ ∼ t −5/2 and C⊥ ∼ t −7/2. Figure 4 shows normalized diffusion coefficients and relative added-mass deviation as functions of the normalized distance to the wall. The text confirms that the explicit molecular solvent reproduces the zero-frequency confined response. The contrast between reduced mobility and increased inertia is the central physical result of this work.
Supporting Equations and Theory
The unsteady force exerted on a sphere in an incompressible fluid is described by Equation (1), which explicitly shows why the VACF is a natural observable. The bulk VACF follows the asymptotic expression Cv(t) ≃ kBT /12ρf(πν)3/2 t −3/2, where ν = η/ρf. p. 1 Operationally, M∗
is obtained from the early
post-compressible VACF normalization and is independently checked from the change of slope of the meansquare displacement in the ballistic regime (Fig. At long times, the same bulk data follows the expected
asymptotic expression:
Cv(t) ≃
kBT
12ρf(πν)
3/2
t
−3/2
, ν = η/ρf
. Once momentum has diffused
over distances comparable to the wall separation, the
flow is reflected and reorganized by the boundary. For
a particle at center-wall distance h = R + ϵ, Felderhof’s
asymptotic theory predicts [25]:
C∥(t) ≃
kBT(3h
2 − R2
)
24ρfπ
3/2
(νt)
−5/2
, (3)
for motion parallel to the wall, and:
C⊥(t) ≃
kBT
8ρfπ
3/2
−
R2
3
(νt)
−5/2 +
h
4
4
(νt)
−7/2
, (4)
for normal motion. Instead of decreasing, the effective inertial contribution increases as the particle approaches the wall. We
quantify this by the relative added-mass deviation:
∆M∗
i
(ϵ) = M∗
i
(ϵ) − M∗
i
(∞)
M∗
i
(∞)
, i =∥, ⊥ . Having established that the wall reshapes the memory
kernel itself, we now exploit the fact that a single VACF
encodes two opposite limits of the same response. The
low-frequency mobility is obtained from the diffusion coefficients, as:
Di =
Z ∞
0
Ci(t) dt, i =∥, ⊥, (5)
and the high-frequency inertial response is obtained from
the early post-compressible normalization Ci ≃ kBT /M∗
i
. Having established that the wall reshapes the memory
kernel itself, we now exploit the fact that a single VACF
encodes two opposite limits of the same response. The
low-frequency mobility is obtained from the diffusion coefficients, as:
Di =
Z ∞
0
Ci(t) dt, i =∥, ⊥, (5)
and the high-frequency inertial response is obtained from
the early post-compressible normalization Ci ≃ kBT /M∗
i
.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Hydrodynamic Limits
In the low-frequency limit, the particle explores a quasi-steady Stokes flow distorted by the presence of the wall, with mobility being controlled by the mobility tensor. In contrast, in the high-frequency limit, the particle accelerates before vorticity has diffused over the gap, and kinetic energy of surrounding fluid is altered by wall presence. p. 3 In the high-frequency limit, the
particle accelerates before vorticity has diffused over the
gap. The relevant quantity is then the kinetic energy of
the surrounding fluid, which is altered by the presence
of the wall. The relevant quantity is then the kinetic energy of
the surrounding fluid, which is altered by the presence
of the wall. The same boundary condition therefore produces two opposite signatures: a reduced mobility and
an increased inertia. The high-frequency response, by contrast,
is governed by an effectively inviscid flow constrained by
the same boundary. Agreement with both steady mobility corrections and potential-flow added-mass theories
therefore indicates that the MD simulation with explicit
solvent captures the boundary effects over a wide dynamical range. To keep this comparison free of adjustable parameters, we restrict the present simulations
to gaps ϵ large compared to the solvent molecular size. In
this regime, the continuum predictions for the algebraic
tails, the steady drag, and the added mass apply without
fitting the position of the hydrodynamic wall, the surface of the colloid, or the effective slip length.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Summary of Insights
The coexistence of mobility suppression and inertia enhancement highlights a fundamental aspect of confined hydrodynamics. The low-frequency response is sensitive to no-slip boundary conditions and lubrication, while the high-frequency response is governed by an effectively inviscid flow constrained by the same boundary. p. 3 The low-frequency
response is sensitive to no-slip boundary conditions and
lubrication. The high-frequency response, by contrast,
is governed by an effectively inviscid flow constrained by
the same boundary. The same VACF-based analysis could then distinguish
whether an observed confinement effect is dominated by
steady drag, elastic storage, inertial coupling, or acoustic
propagation. In this sense, the present work is not only a
numerical validation of known wall corrections, but also a
demonstration of fluctuation-based hydrodynamic spectroscopy under confinement. This contrast is the central physical result of the present work, and the use of an explicit
molecular solvent is essential for connecting the two extreme limits without changing models. The approach also provides a practical way to assess
the range of validity of continuum hydrodynamics.Where the paper says it
Where the paper says it
Where the paper says it
Final Remarks
The authors thank Harshit Joshi, Vincent Bertin, Quentin Ferreira and Juliette Lacherez for interesting discussions. The work was supported by the F. p. 3 Quentin Thomas is FRIA grant holder from
the Belgian National Fund for Scientific Research (FRSFNRS). This work was supported by the F. This work was supported by the F. -FNRS
under the research Grant (PDR “Active Matter in Harmonic Trap”) No. -FNRS
under the research Grant (PDR “Active Matter in Harmonic Trap”) No. 20, by the European Union
through the European Research Council under EMetBrown (ERC-CoG-101039103) grant, and by the Interdisciplinary and Exploratory Research program under
MISTIC grant at the University of Bordeaux. 11t is also acknowledged. Computer time for this study was provided by
the computing facilities of the Mésocentre de Calcul Intensif Aquitain. Computer time for this study was provided by
the computing facilities of the Mésocentre de Calcul Intensif Aquitain. The authors thank the Soft Matter
Collaborative Research Unit, Frontier Research Center
for Advanced Material and Life Science, Faculty of Advanced Life Science at Hokkaido University, Sapporo,
Japan, and the CNRS International Research Network
between France and India on Hydrodynamics at small
scales: From soft matter to bioengineering. At smaller
gaps, these notions become less sharp, and outside the
continuum description that is the focus of the present
Letter. The statistical convergence and the sensitivity of
the extracted mass are documented in the Supplementary
Material [24]. Finally, the present formulation suggests extensions
beyond a single rigid wall. A two-wall geometry mainly
changes the low-frequency part of the response and is
therefore included as an additional validation in the Supplementary Material. More complex situations, such as
patterned walls, soft interfaces, liquid-liquid boundaries,
or viscoelastic media, should affect both the real and
imaginary parts of Γ˜
i(ω) in a geometry-specific manner. The same VACF-based analysis could then distinguish
whether an observed confinement effect is dominated by
steady drag, elastic storage, inertial coupling, or acoustic
propagation. The same VACF-based analysis could then distinguish
whether an observed confinement effect is dominated by
steady drag, elastic storage, inertial coupling, or acoustic
propagation. In this sense, the present work is not only a
numerical validation of known wall corrections, but also a
demonstration of fluctuation-based hydrodynamic spectroscopy under confinement.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[1] B. Duplantier, Le mouvement brownien, divers et ondoyant. sèminaire Poincaré 1, 155 (2005)
[2] X., Bian, C. Kim, and G.E. Karniadakis, 111 years of Brownian motion, Soft Matter, 12, 6331 (2016)
[3] B. J. Alder and T. E. Wainwright, Decay of the velocity autocorrelation function, Phys. Rev. A 1, 18 (1970)
[4] R. Zwanzig and M. Bixon, Hydrodynamic theory of the velocity correlation function, Phys. Rev. A 2, 2005 (1970)
[5] R. Zwanzig and M. Bixon, Compressibility effects in the hydrodynamic theory of Brownian motion, J. Fluid Mech. 69, 21 (1975)
[6] E. J. Hinch, J. Application of the Langevin equation to fluid suspensions, Fluid Mech. 72, 499 (1975)
[7] H. J. H. Clercx and P. P. J. M Schram, Brownian particles in shear flow and harmonic potentials: A study of long-time tails, Phys. p. 4 Clercx and P. Schram, Brownian particles in shear flow and harmonic potentials: A study of
long-time tails, Phys.Where the paper says it
[8] T. Li and M G Raizen, Brownian motion at short time scales, Ann. p. 4 A 46, 1942 (1992). Raizen, Brownian motion at short time
scales, Ann.Where the paper says it
[9] S. Jeney et al. p. 2 The agreement of the predictions with simulation data (see Fig. 4(b)) demonstrates that thermal
D⊥/D0 MD
D∥/D0 MD
Brenner [12]
Perkin [2] (ε ≪ R)
Carbajal [26]
ε/R
Jeney [9]
Mo [16]
Lavaud [18]
Lin [15]
Di/
D
0
0.Where the paper says it
[10] Li et al., Measurement of the instantaneous velocity of a Brownian particle, Science 328 1673 (2010)
[11] T. Franosch et al., Resonances arising from hydrodynamic memory in Brownian motion, Nature 478 85 (2011)
[12] Kheifets et al., Observation of Brownian motion in liquids at short times: instantaneous velocity and memory loss, science 343 1493 (2014)
[13] H. Brenner, The slow motion of a sphere through a viscous fluid towards a plane surface, Chem. Eng. Sci. 16 242 (1961)
[14] L. P. Faucheux and A J Libchaber, Confined brownian motion, Phys. p. 4 Faucheux and A. Libchaber, Confined brownian
motion, Phys.Where the paper says it
[15] B. Lin et al., Direct measurements of constrained Brownian motion of an isolated sphere between two walls, Phys. Rev. E 62 3909 (2000)
[16] J. Mo et al., Broadband boundary effects on Brownian motion, Phys. Rev. E 92 062106 (2015)
[17] M. Matse et al., Test of the diffusing-diffusivity mechanism using near-wall colloidal dynamics, Physical Review E 96 042604 (2017)
[18] M. Lavaud et al., Stochastic inference of surface-induced effects using Brownian motion, Phys. Rev. Research 3 L032011 (2021)
[19] Z. Zhang et al., Unsteady drag force on an immersed sphere oscillating near a wall, J. Fluid Mech. 977 A21 (2023)
[20] P. Palacios-Alonso et al., Fast spectral solver for viscoelastic structures under oscillatory flow in free space or wall-bounded domains: Applications to quartz crystal microbalance and force spectroscopy, J. Chem. Phys. 163 19 (2025)
[23] A.P. Thompson et al., Computer physics communications 271 108171 (2022)
[24] See Supplementaary Material for simulation details, fluid characterization, added-mass extraction, radius dependence, and additional confined geometries
[25] B. U. Felderhof, Effect of the wall on the velocity autocorrelation function and long-time tail of Brownian motion, J.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved understanding of fluid dynamics in confined systems: The simulation resolves
the full time window from the same observable
(VACF) whichcontains dissipation, inertia, and memory in a single time-dependent quantity,
allowing for a unified analysis across acoustic to diffusive regimes. -
Enhanced prediction of confinement effects on mobility and inertia: The system can now simultaneously extract both zero-frequency mobility and high-frequency inertial response by using the same VACF to find
the low-frequency mobility is obtained from the diffusion coefficients
andthe high-frequency inertial response is obtained from the early post-compressible normalization.
-
Simulation of novel hydrodynamic kernels: AI systems can now model how boundaries reshape memory, as the study shows that confinement causes
long-lived hydrodynamic correlations decay faster and become direction-dependent,
reshapingthe temporal structure of the memory kernel itself.
-
Validation and extension of continuum models: The improved system can assess the validity range of continuum hydrodynamics by comparing simulation data against theoretical predictions for added mass, such as
Yang’s compact expression for the wall-induced mass amplification,
and identifying wherecontinuum predictions for the algebraic tails, steady drag, and added mass apply without fitting.
-
Discrimination of physical mechanisms: The system can distinguish between different physical effects by analyzing whether confinement effects are
dominated by steady drag, elastic storage, inertial coupling, or acoustic propagation
using the VACF-based analysis.
Abstract
Molecular dynamics with an explicit solvent resolves the hydrodynamic response of a Brownian colloid in bulk and near a rigid wall over acoustic, inertial, and diffusive time scales. Bulk spectra establish a continuum reference and reveal departures associated with discrete-solvent dynamics. Under confinement, thermal spectra provide a measurement of the space-dependent added mass, while velocity autocorrelations determine space-dependent, anisotropic hydrodynamic memory and diffusion. All together, these measurements connect the enhanced added mass and reduced mobility near a wall, and probe the discrete-continuum crossover.
Sources
Related papers
- Coarse-graining dense, deformable active particles
- Three-body corrections to the Brownian viscosity of hard spheres without hydrodynamic interactions
- Thin active nematohydrodynamic layers: asymptotic theories and instabilities
- Origins of suppressed self-diffusion of nanoscale constituents of a complex liquid
- Soft material mechanical property degradation at high strain rates
- Avalanche criticality emerges by thermal fluctuation in a quiescent glass