Controlling viscosity to engineer focal conic domains in photonic cellulose nanocrystal films

TL;DR

This work shows that viscosity, modulated by salt-induced Debye screening and sonication-induced bundle fragmentation, governs evaporation-driven self-assembly of CNCs into cholesteric films. By systematically varying [NaCl] and u_s across 24 samples, the authors link flow dynamics and tactoid coalescence to the photonic structure, identifying a narrow window where focal conic domains (FCDs) form reproducibly with submicron pitch and narrow spectral width. Low-viscosity conditions promote large, uniform cholesteric domains with minimal iridescence, while higher viscosity induces disorder and diffuse scattering; FCDs arise only when kinetic arrest is delayed enough for deformable cholesteric layers to buckle under evaporative stresses. The study provides design rules for tuning structural color in sustainable CNC photonic films through accessible processing parameters, enabling reduced iridescence and controllable defect architectures with potential applications in coatings, sensing, and anti-counterfeiting.

Abstract

Cellulose nanocrystals (CNCs) form cholesteric architectures that can have color specific reflectivity and enable sustainable photonic films. However, achieving uniform color, suppressing iridescence, and accessing ordered defect structures such as focal conic domains remain challenging. Here, we control the photonic properties of CNC films by steering the self assembly process. Across 24 dish-cast films with varying salt concentrations and sonication doses, we combine viscosity measurements, timelapse polarized optical microscopy, and angle-resolved reflectance spectroscopy to correlate evaporation dynamics with photonic structure. We show that viscosity, jointly controlled by NaCl-mediated electrostatic screening and sonication-induced bundle fragmentation, dictates the extent of tactoid coalescence. Low-viscosity suspensions generate large, homogeneous cholesteric domains and narrow spectral responses, while high viscosity leads to arrested, heterogenous domains and increased diffuse light reflection. Critically, within a narrow parameter window of intermediate ionic strength and moderate sonication, we reproducibly engineer photonically active focal conic domains. These results identify viscosity-driven flow as a key, previously underappreciated factor in CNC self-assembly and establish design rules for producing structurally colored films with tunable photonic response, reduced iridescence, and controllable defect architectures.

Figures (16)

• Figure 1: (A) Schematics illustrating how processing parameters influence CNC interactions and the resulting film color. Left: The blue arrow indicates that increasing salt concentration reduces the Debye length in solution, blue-shifting the reflected color of the dried film. Center: The red arrow indicates that increasing sonication dose fragments CNC bundles, red-shifting the reflected color. Right: Schematics of a CNC suspension evaporating in a petri dish with time progressing from top to bottom, leading to cholesteric self-assembly (with a half-pitch rotation of the CNCs depicted in the grey inset) and structural coloration in the final film. (B–D) Arrays of reflection-mode optical micrographs (captured with white light illumination through a left circular polarizer), each paired with a reflection spectrum from the same region. $\lambda_\text{max}$ is reported for each spectrum, denoted by a white dot. Panels are ordered by NaCl concentration: (B) 0 mmol/kg of CNC, (C) 150 mmol/kg, and (D) 300 mmol/kg. Within each panel, spectra and images are arranged from top to bottom by sonication dose $u_s$ (0–1440 J/mL). Scale bar: 200 µm.
• Figure 2: An array of micrographs of CNC films taken in transmission mode, captured between crossed polarizers. Bright regions are birefringent and are thereby have anisotropic, liquid crystalline ordering, while dark regions are isotropic. The CNC films were made from suspensions prepared with varying amounts of NaCl concentration (increasing from top to bottom) and sonication dose (increasing from left to right). Each image is taken at the center of its respective film (within a 2 mm radius).
• Figure 3: (A–E) Frames from the timelapse video of a [120 J/mL, 150 mmol/kg] CNC suspension during evaporation. (A: 0 min) Evaporation begins with the solution in the isotropic regime; (B: 2640 min) the critical cholesteric concentration is reached, tactoids nucleate and organize into lines; (C: 2820 min) tactoids grow and coalesce, a tactoid is highlighted in the solid red inset (scale bar: 10 µm); (D: 2960 min) kinetic arrest is reached; (E: 3361 min) the film is compressed as evaporation is completed. FCDs are highlighted with gray dashed squares. (F) SEM image of the film cross-section and exposed glass–air interface. Scale bars: (A–E) 200 µm, (C) 30 µm.
• Figure 4: SEM image of the cross-section and exposed glass–air interface of the [120 J/mL, 150 mmol/kg] film, showing multiple concavities indicative of FCDs. Scale bar: 100 µm.
• Figure 5: Frames taken from three timelapse videos of evaporating CNC suspensions: [1440 J/mL, 0 mmol/kg], [8 J/mL, 150 mmol/kg], [0 J/mL, 450 mmol/kg]. All timelapse videos begin in the isotropic regime at [CNC] = 3 wt.%. Tactoid formation and coalescence follow, and convection cells are observed in the bottom two samples. The second-to-last frame (K.A., kinetic arrest) marks the point at which tactoid motion ceases; however, this frame does not represent the exact moment of kinetic arrest, which cannot be determined visually as a lack of texture evolution can also arise from a high volume fraction of cholesteric phase. The final frame corresponds to complete evaporation. The scale bar for all frames is shown on the top left.