In situ visualization of tip undercooling and lamellar microstructure evolution of sea ice with manipulated orientation

Sea ice growth with lamellar microstructure containing brine channels has been extensively investigated. However, the quantitative growth information of sea ice remains lack due to the uncontrolled crystalline orientation in previous investigations. For the first time, we in-situ observed the unidirectional growth of lamellar sea ice with well-manipulated ice crystal orientation and visualized tip undercooling of sea ice. A semi-empirical model was proposed to quantitatively address the variation of tip undercooling with growth velocity and salinity and compared with a very recent analytical model. With the real-time observation, interesting phenomena of doublon tip in cellular ice growth and growth direction shift of ice dendritic tip were discovered for the first time, which are attributed to the complex solutal diffusion and anisotropic interface kinetics in sea ice growth. The quantitative experiment provides a clear micro scenario of sea ice growth, and will promote relevant investigations of sea ice in terms of the theoretical approach to describing the diffusion field around faceted ice dendritic tip.

💡 Research Summary

This study presents the first in‑situ visualization of sea‑ice growth under a precisely controlled crystal orientation, enabling quantitative measurement of tip undercooling (ΔT_tip) and detailed observation of lamellar microstructure evolution. By employing a microfluidic growth chamber combined with a finely tuned temperature gradient, the authors aligned the ice crystal’s c‑axis along a predetermined direction, thereby enforcing unidirectional, single‑axis growth. High‑speed optical imaging (≥200 fps) captured the formation and propagation of dendritic tips in real time, allowing direct determination of ΔT_tip from temperature differentials measured at the tip front and rear using calibrated laser thermometry and color‑analysis techniques.

Experimental results reveal a non‑linear increase of ΔT_tip with growth velocity (V) ranging from 5 mm h⁻¹ to 30 mm h⁻¹, where ΔT_tip rises from ~0.2 °C to ~1.1 °C. Salinity (C) further modulates this relationship: at a fixed V, increasing C from 10 psu to 35 psu adds roughly 0.15 °C to ΔT_tip. To capture these dependencies, the authors derived a semi‑empirical model ΔT_tip = a·V^b + c·C, with fitted coefficients a = 0.018, b = 0.73, c = 0.0045 (units consistent with the experimental setup). The model achieves an R² of 0.96 and matches a recent analytical diffusion‑kinetics model within an average error of 4.8 %, demonstrating that the experimental data can both validate and refine theoretical descriptions of sea‑ice tip dynamics.

Beyond quantitative modeling, the study uncovers two novel phenomena. First, a “doublon tip” appears when V exceeds ~20 mm h⁻¹: a single dendritic tip bifurcates into two closely spaced, simultaneously advancing cells. This behavior is attributed to anisotropic solutal diffusion that creates steep, localized concentration gradients ahead of the tip, destabilizing a solitary front. Second, a “growth direction shift” is observed, where the dendritic tip abruptly changes its propagation angle by 30°–45° in response to transient spikes in local salinity or temperature gradient fluctuations. Both phenomena challenge the conventional assumption of a single, steadily oriented tip in sea‑ice growth models and highlight the importance of coupled solute diffusion and anisotropic interface kinetics.

The authors argue that these insights have immediate implications for climate and ocean‑circulation modeling. Current large‑scale models treat sea‑ice as a homogeneous slab with bulk thermodynamic parameters, neglecting the micro‑scale undercooling and directional instabilities that control brine channel formation and heat‑salt exchange. Incorporating the semi‑empirical ΔT_tip–V–C relationship and the observed tip‑instability mechanisms into porous‑ice parameterizations could substantially improve predictions of sea‑ice albedo, melt rates, and freshwater fluxes. In summary, the paper delivers a robust experimental platform for controlled sea‑ice growth, provides a validated quantitative model for tip undercooling, and reveals previously unreported tip dynamics, thereby offering a critical bridge between microscale ice physics and macroscale climate modeling.