A model of thermophoresis of colloidal proteins in water using non-Fickian diffusion currents

This paper demonstrates that Chapman's non-Fickian diffusion current is an indispensable factor for accurately modeling the thermophoretic motion of colloidal proteins in water, as it successfully captures the temperature-dependent variations of the Soret coefficient and aligns strongly with experimental data for Lysozyme, BLGA, and Poly-L-Lysine.

The Gist

Imagine you have a crowded dance floor (a glass of water) filled with dancers (colloidal proteins). Now, imagine you turn on a heater at one end of the room, creating a temperature gradient: one side is hot, the other is cool.

In the old way of thinking, scientists believed the dancers would simply drift toward the cool side or the hot side based on how they liked the temperature, like moths to a flame or people seeking shade. This movement is called thermophoresis.

However, this paper argues that the old map is missing a crucial piece of the puzzle. The authors, Mayank Sharma, Angad Singh, and A. Bhattacharyay, are saying: "Wait a minute! The floor itself changes as the temperature changes, and the dancers' ability to move isn't just about where they are, but how the floor feels under their feet at that specific spot."

Here is the breakdown of their discovery using simple analogies:

1. The Two Types of "Crowd Movement"

The paper introduces two different ways the dancers move, which the authors call currents:

• The "Fickian" Current (The Obvious Drift): This is the standard rule. If there are too many dancers in one spot, they naturally spread out to empty spots. It's like water flowing downhill. This is the "normal" diffusion everyone knows.
• The "Non-Fickian" Current (The Hidden Push): This is the paper's main character. Imagine the dance floor is made of a special material that gets slippery or sticky depending on the temperature.
  • If the floor gets more slippery as it gets hotter, a dancer standing on the edge of the hot zone feels a sudden "push" to move, even if they aren't crowded.
  • The authors call this the Chapman-Itô current. It's a subtle, invisible force that arises because the ability to move (diffusivity) changes from place to place. It's like a river that speeds up as it gets shallower; the water doesn't just flow because of gravity, it flows faster because the channel changed shape.

2. The Three Competing Forces

The authors describe thermophoresis as a tug-of-war between three teams:

1. The Spreading Team (Fickian): "Let's spread out evenly!"
2. The Slippery Floor Team (Non-Fickian): "The floor is changing under us! We are being pushed by the change in the floor's texture!"
3. The Solvation Team (Drift): "The water molecules around us are changing their grip on us due to heat, pulling us one way or another."

The paper argues that for a long time, scientists focused mostly on Team 1 and Team 3. They completely ignored Team 2 (the Non-Fickian push). The authors show that Team 2 is actually a major player, and without it, you can't explain why proteins behave the way they do.

3. The "Soret Coefficient" (The Scoreboard)

Scientists use a number called the Soret coefficient to measure the final result: Do the proteins gather in the hot spot or the cold spot?

• If the number is positive, they like the cold.
• If negative, they like the heat.

The authors built a mathematical model using their "Slippery Floor" theory (the Non-Fickian current) and tested it against real experiments with three different proteins: Lysozyme (found in egg whites), BLGA (a milk protein), and Poly-L-Lysine.

The Result? Their model matched the real-world data almost perfectly.

• They showed that the "Slippery Floor" effect explains why the proteins switch from liking the cold to liking the heat as the temperature changes.
• They even predicted how these proteins behave when they are just sitting in a jar (equilibrium), not just when they are being pushed by a temperature gradient.

4. Why Does This Matter?

Think of it like fixing a broken GPS.

• Old GPS: "The protein moves because it wants to be cool/hot." (This often gave wrong directions).
• New GPS (This Paper): "The protein moves because the road conditions (viscosity, water structure) change with temperature, creating a hidden current that pushes it."

The authors emphasize that you cannot fully understand how proteins move in heat (which is crucial for things like drug delivery, protein purification, and even understanding how life works in extreme environments) without accounting for this "hidden push" caused by the changing nature of the water itself.

The Takeaway

In simple terms: Heat doesn't just make things move; it changes the "terrain" they are moving on. This change in terrain creates a hidden current that pushes particles in ways we didn't fully appreciate until now. By adding this "hidden current" back into our equations, we can finally predict exactly how proteins will behave in a temperature gradient, matching reality with stunning accuracy.

Technical Summary

1. Problem Statement

Thermophoresis (the movement of particles in a temperature gradient) and thermodiffusion (the Ludwig-Soret effect) are complex non-equilibrium phenomena driven by the interplay of diffusion and solvation forces. While traditional models often rely on Fickian diffusion (where the diffusion coefficient $D$ is constant) and solvation forces (drift currents), they frequently fail to capture the full temperature dependence of the Soret coefficient ($S_T$) for colloidal proteins in aqueous solutions.

The authors identify a critical gap: the neglect of non-Fickian diffusion currents. In inhomogeneous media where diffusivity depends on space (due to temperature or concentration gradients), a non-Fickian current arises. This concept was historically established by Chapman (1928) and Itô (1941) but has often been overlooked in modern thermophoresis models. The paper aims to demonstrate that including this non-Fickian current is indispensable for accurately modeling the thermophoretic behavior of proteins like Lysozyme, BLGA, and Poly-L-Lysine.

2. Methodology

The authors develop a theoretical framework based on kinetic theory and stochastic calculus to model the particle current density ($j$) in a temperature gradient.

A. Theoretical Framework
The total particle current is modeled as the sum of three competing components:

1. Fickian Diffusion Current ($j_F$): Driven by concentration gradients ($-\nabla C$).
2. Non-Fickian Diffusion Current ($j_{NF}$): Arises from spatial gradients in the diffusion coefficient ($\nabla D$). According to Chapman and Itô, $j_{NF} = -C \nabla D$.
3. Solvation Drift Current ($j_{solv}$): Driven by temperature-dependent solvation forces ($F_{sol}$).

The governing equation for the current density is:
$$j(T) = -\frac{\partial}{\partial x}[C(T)D(T, C)] + D(T, C)\frac{F_{sol}(T)}{k_B T}C(T)$$

B. Key Physical Assumptions

• Coordinate-Dependent Diffusivity: The diffusion coefficient $D(T, C)$ is not constant. It is modeled using a Stokes-Einstein relation where viscosity $\eta$ and temperature $T$ vary spatially.
  • $D(T, C) = D_0(T) \Lambda(C)$, where $\Lambda(C) = (1 + bC^\beta)^{-1}$ accounts for hydrodynamic crowding effects at high concentrations.
• Viscosity Correction: The local dynamic viscosity $\eta(T)$ is modified by the thermal expansion of water ($\alpha_w$) due to particle-solvent interactions: $\eta(T) = \eta_w(T) - c \alpha_w(T)$. The parameter $c$ represents the coupling between solvation mechanisms and diffusion.
• Solvation Force: Modeled as proportional to the temperature gradient, $F_{sol} \propto T \frac{dT}{dx}$, derived from a local equation of state.

C. Derivation of the Soret Coefficient
By setting the steady-state current to zero ($j=0$), the authors derive an analytical expression for the Soret coefficient ($S_T \equiv -\frac{1}{C}\frac{dC}{dT}$):
$$S_T = \frac{-\frac{a}{T}f(T) + \frac{\partial \ln D}{\partial T}|_C}{1 + \frac{C}{D}\frac{\partial D}{\partial C}|_T}$$
In the strong concentration regime ($bC^\beta \gg 1$), this simplifies to:
$$S_T \approx \frac{1}{1-\beta} \left( \frac{d}{dT}\ln D_0(T) \right) - \frac{a}{1-\beta}$$
This equation highlights that the sign and magnitude of $S_T$ are critically dependent on the exponent $\beta$ (concentration dependence) and the solvation parameter $a$.

3. Key Contributions

1. Reintroduction of Non-Fickian Currents: The paper rigorously applies the Chapman-Itô non-Fickian diffusion current to thermophoresis, demonstrating it is a major contributor to particle transport, not just a minor correction.
2. Unified Model for $S_T$ Temperature Dependence: The model successfully captures the complex temperature dependence of the Soret coefficient (including sign changes from thermophilic to thermophobic behavior) using only three free parameters ($a, \beta, c$).
3. Coupling Mechanism: The authors propose a specific physical mechanism where the solvation force couples to the diffusion process via the temperature-dependent dynamic viscosity (mediated by water's thermal expansion coefficient).
4. Equilibrium Validation: The model is validated not only against non-equilibrium thermophoresis data but also against equilibrium saturation concentration curves ($C_{sat}$), showing that the same non-Fickian framework describes both equilibrium and non-equilibrium states.

4. Results

The theoretical model was compared against experimental data for three polypeptides: Lysozyme, BLGA, and Poly-L-Lysine.

• Quantitative Agreement: The model's predicted $S_T(T)$ curves show a strong match with experimental data and empirical fitting functions across a temperature range of 5°C to 40°C.
• Parameter Sensitivity:
  • $\beta$ (Concentration Exponent): Values were found to be slightly greater than 1 ($\approx 1.18 - 1.26$). This value is crucial as it determines the sign of $S_T$ and scales its magnitude.
  • $c$ (Viscosity Coupling): Positive values of $c$ indicate that solvation mechanisms effectively reduce local viscosity in the solvation layer, likely due to hydrogen bonding changes.
  • $a$ (Solvation Strength): Determines the vertical shift of the $S_T$ curve.
• Current Competition Analysis:
  • The study reveals that the Non-Fickian current and the Solvation drift are the dominant forces.
  • The Fickian current is smaller in magnitude but reverses direction when the concentration gradient changes sign.
  • The transition temperature (where $S_T = 0$) corresponds to the point where the concentration profile reaches a local minimum/maximum.
• Equilibrium Match: The model accurately reproduces the equilibrium saturation concentration ($C_{sat}$) of Lysozyme crystals, confirming that the non-Fickian current is essential for describing local equilibrium distributions in inhomogeneous fields.

5. Significance

• Theoretical Completeness: The paper argues that existing models of thermophoresis are incomplete because they neglect non-Fickian currents. It posits that these currents are fundamental to understanding transport in inhomogeneous fluids, not just a mathematical artifact.
• Predictive Power: By using a minimal set of parameters derived from first principles (kinetic theory and hydrodynamics), the model explains the universal temperature dependence of the Soret coefficient in aqueous protein solutions without relying on ad-hoc empirical fitting for every new system.
• Implications for Engineering: Understanding the precise balance between Fickian, non-Fickian, and solvation currents is vital for applications in microfluidics, protein purification, and drug delivery, where precise control of particle transport via thermal gradients is required.
• Historical Context: The work bridges a gap between early 20th-century theoretical physics (Chapman, Itô) and modern soft matter physics, showing that these "forgotten" currents are essential for explaining modern experimental observations.

In conclusion, the authors demonstrate that Chapman's non-Fickian diffusion current is an indispensable element for a complete physical understanding of thermophoresis, successfully unifying the description of equilibrium and non-equilibrium transport in colloidal protein solutions.