Polymer-Regulated Freezing of Water Droplets Revealed by Synchrotron X-ray Imaging and Raman Spectroscopy

By combining synchrotron X-ray imaging and Raman spectroscopy, this study reveals that polyvinyl alcohol regulates the freezing of water droplets by inducing heterogeneous polymer segregation, which slows the freezing front, suppresses bubble entrapment, and blunts the characteristic tip singularity.

The Gist

Imagine you have a tiny drop of water sitting on a cold surface. If you freeze a drop of pure water, it doesn't just turn into a smooth ice ball. Instead, as the ice grows from the bottom up, it squeezes the remaining liquid into a tiny, sharp point at the top, like a needle. At the same time, the air dissolved in the water gets pushed out by the growing ice and gets trapped inside, forming little bubbles that look like tiny pearls in a necklace.

Now, imagine you add a special ingredient to that water: a polymer called Polyvinyl Alcohol (PVA). Think of PVA like a long, sticky string of spaghetti dissolved in the water. When you freeze this "spaghetti water," something magical happens. The sharp needle at the top disappears, replaced by a smooth, rounded dome. Also, those tiny trapped bubbles vanish.

This paper is like a high-tech detective story that figures out why this happens. The researchers couldn't just look at the ice with their eyes because the ice is cloudy and the inside is hidden. So, they used two super-powered tools:

1. Super X-Ray Vision: They used a very strong X-ray beam (from a giant machine called a synchrotron) to see right through the cloudy ice. This let them watch the freezing process in slow motion and see the inside structure in 3D.
2. Chemical Flashlight (Raman Spectroscopy): After freezing, they sliced the ice open and used a laser to take a "chemical fingerprint" of different spots. This told them exactly where the "spaghetti" (PVA) was hiding.

Here is what they discovered:

The "Traffic Jam" at the Ice Front
When pure water freezes, the ice front is like a smooth, marching army. But when PVA is added, the ice front becomes rough and bumpy, like a burr or a jagged edge. As the ice tries to grow, it pushes the "spaghetti" strings away because they don't fit into the ice crystal.

The Hidden Pockets
Instead of the spaghetti spreading out evenly, it gets pushed into the gaps between the ice crystals. The X-rays showed that the inside of the frozen drop isn't just solid ice; it's a sponge-like structure filled with tiny, interconnected channels and pockets that are rich in PVA. The Raman "flashlight" confirmed that these dark pockets seen in the X-rays are exactly where the PVA is concentrated.

Why the Tip Gets Blunt
In pure water, the ice squeezes everything into a sharp point because the ice is much denser than the water. But in the PVA drop, the "spaghetti" gets stuck in those tiny pockets near the top. These pockets act like a cushion. Because the material at the tip is a mix of ice and these PVA-rich pockets (which are less dense), the ice doesn't need to squeeze as hard to fit everything in. The result? The sharp needle never forms; instead, you get a soft, rounded dome.

Why the Bubbles Disappear
In pure water, the air has nowhere to go but to get trapped as bubbles. But in the PVA drop, the air seems to stay dissolved inside those PVA-rich pockets. Because the pockets are "incompletely frozen" and full of the polymer, the air doesn't need to pop out and form a bubble. It just stays hidden inside the sponge-like structure.

The Rough Skin
The researchers also noticed that the outside of the frozen drop looks rougher and scatters light differently. The X-rays and chemical maps showed that the "spaghetti" also piles up on the very surface, creating a rough, bumpy skin rather than a smooth ice shell.

The Big Picture
The main takeaway is that when you freeze water with polymers, it's not a simple, uniform process. The polymer doesn't just change the water's properties everywhere at once. Instead, it gets pushed around and creates a complex, patchwork world inside the ice. The ice is a mix of solid ice crystals and these special, polymer-filled pockets. This "patchwork" nature is what changes the shape of the drop and stops the bubbles from forming.

The authors suggest that understanding this "patchwork" behavior could help improve processes that rely on freezing, such as making special porous materials (freeze-casting) or preserving biological samples (cryopreservation), but they focus primarily on explaining the physics of how the drop freezes.

Technical Summary

Technical Summary: Polymer-Regulated Freezing of Water Droplets

Problem Statement
The freezing of sessile water droplets is a complex phase transition involving supercooling, latent heat dissipation, volumetric expansion, and solute redistribution. In pure water, this process typically results in a self-similar spherical freezing front that focuses the remaining liquid into a sharp conical apex (tip singularity) and leads to the nucleation and entrapment of discrete gas bubbles. While it is known that adding polymers, such as polyvinyl alcohol (PVA), can suppress this tip singularity and alter freezing times, the underlying physical mechanisms remain unclear. Previous explanations relied on macroscopic bulk properties (e.g., effective density ratios and thermophysical changes) or general solute transport models. However, these approaches fail to address the specific spatial organization of the rejected polymer within the droplet, how it accumulates during solidification, or how this distribution dictates the evolving internal structure and external morphology. Crucially, the interior of freezing droplets is obscured by polycrystalline ice, making it difficult to probe the internal dynamics of solute redistribution using visible light.

Methodology
To resolve the spatiotemporal mechanisms of polymer-regulated freezing, the authors employed a multimodal imaging approach combining synchrotron X-ray imaging and Raman spectroscopy on aqueous PVA solutions.

• Sample Preparation: Droplets (1–10 µL) of PVA solutions with varying concentrations and molecular weights (13–23 kDa and 85–124 kDa) were deposited on a temperature-controlled aluminum substrate and cooled from room temperature to −15°C.
• Synchrotron X-ray Imaging: Experiments were conducted at the Pohang Light Source-II (PLS-II) using a monochromatic 14 keV X-ray beam. This allowed for high-resolution (0.65 µm pixel size) radiography and tomography of the freezing process in situ. X-ray tomography was used to reconstruct 3D internal structures of frozen specimens, revealing domains with low X-ray transmittance.
• Raman Spectroscopy: To correlate structural features with chemical composition, frozen droplets were cryo-sectioned and analyzed using confocal Raman microscopy. Chemical maps were generated based on the C–H vibration band near 1450 cm⁻¹, which serves as a fingerprint for PVA. This technique confirmed the presence of PVA-enriched regions within the frozen matrix.

Key Results
The study reveals that the freezing of PVA solutions proceeds via a spatially heterogeneous pathway rather than a uniform bulk solidification:

1. Tip Suppression and Bubble Reduction: As PVA concentration increases, the sharp tip singularity is progressively blunted, eventually becoming a rounded apex (approaching 180°). Concurrently, the entrapment of discrete gas bubbles is significantly suppressed; at concentrations above 1 wt%, no discrete bubbles were observed at the imaging resolution.
2. Internal Heterogeneity: X-ray tomography of frozen PVA droplets revealed that they do not contain bubbles but instead possess interconnected domains of low X-ray transmittance distributed throughout the interior and at the surface. These domains have widths on the order of 5 µm.
3. Polymer Segregation: Raman spectral mapping confirmed that these low-transmittance domains correspond to regions enriched with PVA. This provides direct evidence of freeze-induced polymer segregation, where the polymer is rejected from the growing ice matrix and accumulates in incompletely frozen domains.
4. Front Morphology: Time-resolved X-ray radiography showed that the freezing front in PVA solutions is not smooth but develops a "burr-like," rough interface that propagates more slowly than in pure water. This roughening is attributed to constitutional supercooling caused by the inability of the rejected polymer to diffuse away rapidly enough.
5. Surface Roughness: The outer surface of frozen PVA droplets exhibits a rough layer with lower X-ray transmittance and enhanced PVA signal, suggesting that polymer enrichment occurs not only in the interior channels but also in a distinct surface layer.

Key Contributions

• Direct Visualization of Segregation: The work provides the first direct, spatially resolved evidence that PVA is redistributed heterogeneously during water solidification, forming polymer-enriched, incompletely frozen domains rather than shifting bulk properties homogeneously.
• Mechanistic Explanation of Tip Blunting: The study proposes that the blunting of the tip singularity is a local phenomenon driven by the accumulation of polymer-rich domains near the apex. These domains alter the local solid-to-liquid density ratio ($\nu$), bringing it closer to 1, which mathematically results in a rounded apex.
• Reinterpretation of Freezing Dynamics: The authors argue that the freezing of polymer solutions is better described as a spatially heterogeneous solid embedded with incompletely frozen domains shaped by polymer exclusion, rather than a uniform ice matrix. This framework explains both the suppression of bubble entrapment (gas remaining in incompletely frozen domains) and the extended freezing times.

Significance
The paper claims that understanding polymer-regulated freezing fundamentally requires accounting for heterogeneous polymer redistribution during solidification. By identifying freeze-induced polymer segregation as the pathway regulating both the external shape and internal structure of freezing droplets, the findings offer a spatially resolved framework for interpreting observed phenomena. These insights are presented as relevant to freezing-based processes such as freeze-casting and cryopreservation, where controlling solute redistribution is critical. The authors note that further investigation is needed to determine the phase state of these polymer-rich domains, the fate of rejected gases, and the systematic dependence of these morphologies on polymer identity and freezing rates.