Atomic Force Microscopy of Glass Transition and Adhesion in Thin Polystyrene Films

Abstract

Thin polymer films exhibit relaxation behaviors that are highly sensitive to temperature and film thickness. Conventional bulk instruments cannot resolve these dynamics at the nanometer scale. In this study, we used atomic force microscopy (AFM) force–distance curves to probe the relaxation dynamics and thickness‑dependent glass transition temperature (T_g) of normal thin polystyrene (PS) films supported on silicon. The adhesion force (F_ad) between the AFM tip and the PS surface was measured in situ while varying temperature and film thickness. The abrupt change in F_ad as the temperature decreased allowed us to pinpoint T_g. We found that T_g decreases monotonically as the film thickness is reduced, offering insights into the relaxation dynamics of normal thin polymer films.

Background

Nanoscience has driven the widespread use of polymer films with thicknesses in the nanometer regime. The physical properties of these thin films differ markedly from their bulk counterparts due to confinement effects. Multiple studies have reported a reduction in T_g with decreasing film thickness, causing the films to relax at temperatures far below the bulk value. This phenomenon limits the use of thin polymers as dielectrics in micro- and nano‑devices, where premature dielectric loss can occur before breakdown. Therefore, precise, nanoscale measurements of relaxation behavior are essential for advancing nanotechnological applications.

Atomic force microscopy offers nanometer‑scale resolution and high sensitivity, making it ideal for measuring surface morphology, mechanical, electrical, and magnetic properties of nanostructures. Previous AFM studies have explored surface dewetting, T_g depression via electric force microscopy, and viscosity of short‑chain PS films. Mechanical properties such as friction, adhesion, and viscoelasticity—key determinants of relaxation dynamics—have been examined using friction force microscopy, lateral force microscopy, and force–distance spectroscopy.

Because AFM tips can detect extremely weak forces, they provide a direct, highly sensitive method to probe adhesion and, indirectly, the mechanical state of polymer surfaces. In this work, we quantified the adhesion force between an AFM tip and normal thin PS films as a function of temperature and film thickness, enabling us to extract the thickness‑dependent T_g of the films.

Methods

Materials

Polystyrene (MW = 4000) was obtained from Alfa Aesar; chlorobenzene (Sinopharm) and single‑side polished silicon wafers (Silicon Quest) were used as received. PS films ranging from 18 to 127 nm were spin‑coated from chlorobenzene solutions of varying concentration and spin speed. Films were annealed at 358 K for 2 h and their thickness measured by AFM.

Instruments

Force–distance curves and adhesion forces were recorded on a Dimension Icon AFM (Bruker). A silicon nitride V‑shaped tip with a nominal spring constant of ~0.1 N m⁻¹ was employed in contact mode to capture the tip–sample interaction.

Adhesion Force Measurements

Figure 1 illustrates the AFM approach–retract sequence used to measure F_ad. The tip first approaches the surface without contact, then makes contact, indents the film under load, and finally withdraws, generating the pull‑off force that corresponds to adhesion. Measurements were performed while cooling from above the bulk T_g at a rate of 2 K min⁻¹ under <10 % relative humidity to suppress capillary contributions.

Figure 1. Schematic of adhesion force measurement on thin polymer films supported on silicon. a Initial approach; b contact; c indentation; d–e retraction.

Modulus Measurements

In earlier work, we monitored surface potential to study ultrathin PS and PMMA films. Those films exhibited thickness‑independent T_g values of 328 K and 358 K, respectively. To compare PS/PMMA blends, we spin‑coated a 37 nm thick PS–PMMA blend and imaged morphology, modulus, and adhesion at 298 K and 548 K. At 548 K, PS chains dewetted, reducing the film thickness to 22 nm (Figure 2). The significant contrast in modulus and adhesion maps (Figure 2h–i) confirms that adhesion force changes can be used to quantify T_g of normal thin PS films.

Figure 2. Surface morphology (a), modulus mapping (b), and adhesion force mapping (c) of PS–PMMA blends at 298 K; (d)–(f) at 548 K; (j)–(k) AFM topography at 298 K and 548 K.

Results and Discussion

Friction force microscopy is sensitive to lateral viscoelasticity, whereas adhesion force reflects vertical mechanical properties and is obtained from a single point measurement, reducing substrate interference. The tip indentation depth provides a direct measure of viscoelasticity. In situ AFM heating/cooling was used to monitor temperature‑dependent adhesion. Cooling was performed from above bulk T_g at 2 K min⁻¹, with each temperature held for 5 min to ensure thermal equilibration.

Figure 3 shows force–distance curves for a 93 nm PS film at 393, 373, 353, and 343 K. At 393 K the curve has a pronounced tail, indicating a softer surface with a 208 nm indentation. As temperature decreases, the indentation reduces to 109 nm (373 K) and 89 nm (353 K), and the curve becomes steeper at 343 K, reflecting a stiffer surface.

Figure 3. Force–distance curves for a 93 nm PS film at 393 K (a), 373 K (b), 353 K (c), and 343 K (d). The dashed line marks indentation depth.

We recorded 300 force curves at each temperature to build statistical distributions. The resulting histograms (Figure 4) show adhesion forces of 91 nN at 393 K, 30 nN at 353 K, and 26 nN at 323 K. The abrupt drop in adhesion force indicates the glass transition.

Figure 4. Histogram of adhesion force at 393 K (a), 343 K (b), and 303 K (c).

Figure 5 presents the temperature dependence of F_ad for PS films of 18–127 nm. At high temperatures, the adhesion force is large due to cooperative relaxation of chain segments. As temperature decreases, a sharp transition occurs, followed by a plateau at lower temperatures, corresponding to the glassy state. The intersection of the two linear regimes marks T_g for each thickness.

Figure 5. Temperature dependence of adhesion force for PS films from 18 to 127 nm.

Capillary, van der Waals, and contact forces contribute to F_ad. In our dry, low‑humidity environment, van der Waals and electrostatic contributions are negligible; thus, the dominant terms are contact and capillary forces. The tip indentation depth reflects the film’s viscoelasticity and contact area. AFM topographies of a 20 nm PS film at 403 K, 373 K, and 298 K (Figure 6) show a reduction in surface roughness from 1.13 nm to 0.56 nm with cooling. At high temperatures, the softer, rougher surface yields a larger real contact area and a viscous liquid bridge, enhancing adhesion. In the glassy state, the flat, rigid surface reduces contact area and capillary contributions, leading to a constant, lower adhesion force.

Figure 6. AFM topography of a 20 nm PS film at 403 K (a), 373 K (b), and 298 K (c); (d) temperature dependence of roughness; (e–f) schematic of contact area and liquid bridge contributions to F_ad.

The calculated T_g values (Table 1) show that films thicker than 100 nm retain the bulk T_g of 363 K, while thinner films exhibit a pronounced depression. Figure 7a plots T_g versus thickness, confirming the trend.

Figure 7. a Thickness dependence of T_g for normal thin PS films; b Three‑layer model illustrating the reduction of T_g with increasing top‑layer fraction.

An empirical relation proposed by Keddie et al. captures this behavior:
\[T_g(d)=T_g^{\text{bulk}}\bigl[1-(A/d)^{\delta}\bigr]\]
with A ≈ 3.2 nm and δ ≈ 1.8. This equation predicts that T_g approaches the bulk value for thick films. Two‑ and three‑layer models attribute the depression to a liquid‑like surface layer that enhances chain mobility, while an interfacial “dead” layer near the substrate suppresses motion. As the film becomes thinner, the relative thickness of the mobile surface layer increases, leading to a lower overall T_g.

Conclusions

We have demonstrated that AFM force–distance spectroscopy can sensitively track adhesion changes in normal thin PS films, enabling the extraction of thickness‑dependent T_g. The glass transition temperature decreases systematically with decreasing film thickness, a behavior consistent with the presence of a mobile surface layer. These findings enhance our understanding of relaxation dynamics in thin polymer films and may inform the design of polymeric components in nanodevices. Further investigations are warranted to resolve remaining controversies regarding T_g thickness dependence.

Abbreviations

AFM

Atomic force microscopy

AFMAM

Atomic force microscopic adhesion measurement

EFM

Electric force microscopy

F_ad

Adhesion force

FFM

Friction force microscopy

LFM

Lateral force microscopy

PS

Polystyrene

PtBuA

Poly(tert‑butyl acrylate)

SFM

Scanning force microscopy

T_g

Glass transition temperature