Predictive design and fabrication of complex micro and nano patterns via wrinkling for scalable and affordable manufacturing

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2014.Cataloged from PDF version of thesis.Includes bibliographical references (pages 189-193).There is a demonstrated need for scalable and affordable manufacturing of complex micro and nano scale structures for applications such as fluidics-based medical diagnostics and photonicsbased sensing. Although high-rate patterning of these structures is feasible via template/stamp based processes, scalability and affordability are often limited by expensive and slow template fabrication processes. The purpose of this work was to develop a wrinkling-based manufacturing process that eliminates the need for an expensive and slow template fabrication step. Wrinkling of thin films is a low-cost buckling-driven patterning process that provides an alternate route to manufacturing scale-up. However, this process is currently of limited practical import because predictive design is restricted to a small set of simple/elementary patterns. In this work, predictive models, tools, and techniques were developed to enable the design and fabrication of a variety of complex wrinkled patterns. The shape and size of wrinkled patterns is determined by the interaction among geometry, material properties, and loading. Complexity in wrinkle patterns often arises due to uncontrolled/undesirable non-uniformities in these process parameters. Due to the confounding effect of simultaneously acting non-uniformities, it is difficult to link wrinkle shape/size to process parameters for such systems. To solve this problem, experimental and computational tools were developed to (i) individually tune/control the non-uniformities in these parameters and (ii) probe the effect of these parameters on the wrinkled pattern. The data generated from these tools was then used to link process parameters to pattern complexity. Contributions were made in the following specific areas: (1) Tunable hierarchical wrinkled patterns via geometric pre-patterning Although complex hierarchical wrinkled patterns have been fabricated via geometric prepatterning in the past, predictive design of such patterns is not feasible. This is primarily because of lack of appropriate process models that can accurately capture the physical effect of prepatterns on the wrinkle generation process. Herein, an analytical model was developed and verified to predict the hierarchical patterns that arise due to geometric non-uniformity. This model elucidates and captures the fundamental energetic response of the system to pre-patterns that distinguishes a pre-patterned system from a flat non-patterned system. The ability to capture this essential physics provides valuable insight into the process that is not available via empirical models that are based on a limited data set. This insight enables one to (i) explore the full design space and identify optimal operating regions, (ii) design and fabricate tunable wrinkled patterns that can be deterministically switched across hierarchical and non-hierarchical states and (iii) explain and predict patterns that are theoretically feasible yet practically inaccessible. (2) Period doubling via high compressive strains Although period doubling at high compressive strains has been demonstrated in the past, models that accurately predict the onset of period doubling are not available. This is due to the inability of existing models to capture the nonlinear stress versus strain material response as the physical basis for period doubling. For example, existing models erroneously assume that the onset of period doubling is independent of stiffness moduli. Herein, models driven by finite element analysis were developed to accurately link the onset of period doubling to nonlinearities that arise during large deformations. These models enable one to accurately separate the effect of individual process parameters on the onset of period doubling. This enables one to extract valuable process-relevant information from the observation of the period doubling phenomenon that is otherwise not available. (3) Attractors and repellers as localized material defects for curved wrinkles Curved wrinkles have been demonstrated in the past during equibiaxial compression of nonuniform materials. However, predictive design of target patterns via curved wrinkles is not feasible at present due to the lack of simple yet effective design rules. Generation of such design rules is hindered by the confounding effect of biaxial compression and localized material nonuniformities. Herein, an elegant design rule for in-plane bending of wrinkles has been generated for the case of uniaxial compression. Based on this, the concept of attractors and repellers as distinct localized material defects has been proposed and experimentally verified. Attractors are relatively compliant defects that pull wrinkles toward them; whereas repellers are relatively stiffer defects that push wrinkles away. These defects may be used as the building blocks to locally alter the in-plane orientation of otherwise uniformly aligned wrinkles during uniaxial compression. The set of predictive models generated in this work would enable a designer to perform inverse pattern design of complex wrinkled structures, i.e., to select the appropriate set of process parameters that are required to fabricate the desired pattern. By replacing other expensive processes for complex patterning, this would reduce the time and cost of manufacturing a variety of functional micro and nano scale patterns by a factor of ~10 for low-volume applications and by ~10% for high-volume applications.by Sourabh Kumar Saha.Ph. D