Harnessing the interface mechanics of hard films and soft substrates for 3D assembly by controlled buckling

Abstract

Techniques for forming sophisticated, 3D mesostructures in advanced, functional materials are of rapidly growing interest, owing to their potential uses across a broad range of fundamental and applied areas of application. Recently developed approaches to 3D assembly that rely on controlled buckling mechanics serve as versatile routes to 3D mesostructures in a diverse range of high-quality materials and length scales of relevance for 3D microsystems with unusual function and/or enhanced performance. Nonlinear buckling and delamination behaviors in materials that combine both weak and strong interfaces are foundational to the assembly process, but they can be difficult to control, especially for complex geometries. This paper presents theoretical and experimental studies of the fundamental aspects of adhesion and delamination in this context. By quantifying the effects of various essential parameters on these processes, we establish general design diagrams for different material systems, taking into account 4 dominant delamination states (wrinkling, partial delamination of the weak interface, full delamination of the weak interface, and partial delamination of the strong interface). These diagrams provide guidelines for the selection of engineering parameters that avoid interface-related failure, as demonstrated by a series of examples in 3D helical mesostructures and mesostructures that are reconfigurable based on the control of loading-path trajectories. Three-dimensional micromechanical resonators with frequencies that can be selected between 2 distinct values serve as demonstrative examples.

Keywords: 3-dimensional assembly; buckling; interface mechanics; reconfigurable 3D structures.

Fig. 1.

Delamination processes associated with buckling-guided assembly of 3D mesostructures from 2D precursors. (A and B) Computational model associated with finite element analyses of straight ribbon mesostructures that include both strong and weak interfaces. (C) Delamination ratio $\left(L_d^{film}/L_{weak}\right)$ versus prestrain (ε_pre) for a representative case (L_weak = 3 mm; h_f = 10 µm) with a single region (wave) of delamination. Four stages are identified, including wrinkling (regime I), partial delamination of the weak interface (regime II), full delamination of the weak interface (regime III), and partial delamination of the strong interface (regime IV). (D) Similar results for the case of delamination with 2 separated waves, in which, 2 different stages, wrinkling (regime I) and double-arc delamination of the weak interface (regime V) occur. Here, L_weak = 8 mm and h_f = 10 µm. (E) Representative in situ optical (colorized) images and FEA predictions corresponding to the states marked by a–e in C and f–h in D. (Scale bars, 500 μm.)

Fig. 2.

Design diagram for the buckling-guided 3D assembly of straight ribbon mesostructures. (A) Illustration of the location at which delamination initiates. (B) Colorized experimental images and corresponding FEA predictions for 3 cases with different locations where delamination initiates. Here, L_weak = 4 mm and h_f = 10 µm. (C) FEA results on the evolution of the delamination ratio $\left(L_d^{film}/L_{weak}\right)$ for the 3 cases shown in B. (D) Delamination ratio $\left(L_d^{film}/L_{weak}\right)$ versus prestrain (ε_pre) for ribbons with 3 different film thicknesses (10, 15, and 20 µm), with fixed $L_{weak}/h_f^{3/2}$. (E) Similar results for ribbons with 3 different weak interface energies (2.5, 5.0, and 7.5 J/m²), with fixed ${\gamma }_{weak}/{\overline{E}}_f$ and ${\overline{E}}_f/{\overline{E}}_s$. (F–H) Design diagrams for straight ribbon mesostructures in 3 different material systems, including SU-8/d-Skin in F, silicon/d-Skin in G, and PI/d-Skin in H. The colored regimes in the design diagrams represent wrinkling (green, regime I), partial delamination of the weak interface (yellow, regime II), full delamination of the weak interface (blue, regime III), partial delamination of the strong interface (pink, regime IV), and double-arc delamination of the weak interface (gray, regime V), respectively. (I and J) Colorized experimental images and FEA predictions of deformed configurations in the cases of silicon/d-Skin (in I) and PI/d-Skin (in J), corresponding to the design points marked in G and H, respectively. (Scale bars, 500 μm in B and 200 μm in I and J.)

Fig. 3.

Design diagram for the buckling-guided 3D assembly of helical mesostructures. (A and B) Computational model for FEA of serpentine ribbon mesostructures. (C–E) Design diagrams for helical mesostructures formed with 3 representative arc angles (60° in C, 120° in D, and 180° in E). The colored regimes in the diagrams represent wrinkling (green, regime I), partial delamination of the weak interface (yellow, regime II), full delamination of the weak interface (blue, regime III), partial delamination of the strong interface (pink, regime IV), and double-arc delamination of the weak interface (gray, regime V), respectively. (F–H) Colorized experimental images and FEA predictions of the deformed configurations for 3 different arc angles (60° in F, 120° in G, and 180° in H), corresponding to the design points marked in C, D, and E, respectively. (Scale bars, 500 μm.)

Fig. 4.

Harnessing interface mechanics to achieve reconfigurable 3D mesostructures through loading-path controlled assembly. (A) Conceptual illustration of the strategy through a sequence of FEA results and a pair of colorized experimental images of reconfigurable 3D mesostructures (SU-8; 10 μm in thickness) based on an H-shaped precursor ($L_{weak}^{\left(2\right)}/L_{weak}^{\left(1\right)}$=1.0 and ${\overline{L}}_{weak}^{\left(1\right)}$= 2.31). (B) Design diagram for the assembly of reconfigurable 3D mesostructures from an H-shaped precursor. The colored regimes represent unique stable shapes (brown, regime I; gray, regime III) and bistable shapes (blue, regime II). (C) Colorized experimental images and FEA predictions for reconfigurable mesostructures (SU-8; 10 μm in thickness) with 3 representative length ratios (0.5 in i, 1.5 in ii, and 2.0 in iii), corresponding to the design points marked in B. (D) Three variants of H-shaped reconfigurable 3D mesostructures (PI; 15 μm in thickness). (E) A collection of reconfigurable 3D mesostructures (PI; 15 μm in thickness) with different geometries. (F) Resonant frequencies of the same vibrational mode for the 2 distinct shapes of the second design in E. Insets correspond to FEA images of the vibrational modes for the 2 distinct shapes with 40% prestrain, in which the amplitudes are magnified to clearly illustrate the modes. (G and H) Similar results for the first design in E and the design in A. (Scale bars, 1 mm.)