[Paper Review] Stability and dewetting of thin liquid films

Karin Jacobs, Ralf Seemann+1 more|ArXiv.org|May 28, 2008

Fluid Dynamics and Thin Films4 references16 citations

TL;DR

This paper investigates the stability and dewetting dynamics of thin liquid films, particularly polymer melts like polystyrene on silicon substrates, using experimental and theoretical approaches. It demonstrates that the effective interface potential—governed by van der Waals forces and Hamaker constants—determines dewetting mechanisms (nucleation vs. spinodal), with in situ AFM and Fourier analysis revealing exponential growth of unstable modes and thermal fluctuations accelerating pattern formation.

ABSTRACT

The stability of thin liquid coatings is of fundamental interest in every- day life. Homogeneous and non-volatile liquid coatings may dewet either by heterogeneous nucleation, thermal nucleation, or spinodal dewetting. Wetting and dewetting is explained on a fundamental level, including a discussion of relevant interactions. The article will also address the various dewetting scenarios and explain how the effective interface potential governs the behavior obtained for various stratified substrates and film thicknesses.

Motivation & Objective

• To understand the fundamental mechanisms governing thin liquid film stability and dewetting on solid substrates.
• To clarify the role of intermolecular forces—especially van der Waals interactions—in determining wetting behavior.
• To experimentally validate theoretical models of dewetting by correlating effective interface potential with observed morphologies.
• To distinguish between nucleation-driven and spinodal dewetting scenarios using controlled film thickness and substrate engineering.
• To enable predictive tailoring of wettability through precise control of interface potential in stratified systems.

Proposed method

• Use of polystyrene (PS) melts as model liquids due to low vapor pressure, chemical inertness, and tunable rheology via molecular weight and temperature.
• Preparation of thin films via spin coating on hydrophobized Si wafers with tunable surface energy via self-assembled monolayers (e.g., octadecyltrichlorosilane).
• Employment of in situ atomic force microscopy (AFM) in non-contact mode to image dewetting dynamics in real time.
• Application of Fourier transform analysis to AFM images to extract power spectral density and identify dominant unstable modes.
• Calculation of Hamaker constants from optical properties to determine long-range van der Waals interactions in stratified systems.
• Use of Minkowski functionals for higher-order correlation analysis in complex dewetting patterns where standard statistics fail.

Experimental results

Research questions

• RQ1 How does the effective interface potential govern the transition between nucleation-driven and spinodal dewetting in thin films?
• RQ2 What role do thermal fluctuations play in accelerating the dynamics of spinodal dewetting in polymer films?
• RQ3 How do film thickness and substrate layer structure (e.g., SiOx thickness) influence the stability and morphology of dewetting films?
• RQ4 To what extent can theoretical predictions of interface potential and Hamaker constants be corroborated by experimental observations of dewetting patterns?
• RQ5 Why does the glass transition temperature of thin films deviate from bulk values, and how does this affect dewetting kinetics?

Key findings

• For a 4.1 nm thick PS film on a 2.4 nm SiOx layer, φ''(4.1 nm) ≈ 0, indicating a critical point where spinodal dewetting is expected, and experimental observations confirm this prediction.
• In situ AFM imaging of a 3.9(2) nm PS(2k) film on a 191 nm SiOx layer shows that dewetting proceeds via hole nucleation and Ostwald ripening, with droplet coarsening observed after 5000 s at 53 °C.
• Fourier analysis of AFM scans reveals exponential growth of the dominant unstable mode, with a fit to exponential growth law matching theoretical predictions.
• Thermal fluctuations significantly accelerate dewetting dynamics, as confirmed by comparison between deterministic simulations and experimental data.
• The effective interface potential, derived from Hamaker constants calculated via optical data, successfully predicts dewetting behavior across various film thicknesses and substrate configurations.
• Minkowski functionals prove effective in detecting higher-order correlations in complex dewetting patterns where conventional statistical methods fail.