Date

2014

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Mechanical Engineering

First Adviser

Coulter, John P.

Other advisers/committee members

Jedlicka, Sabrina S.; Webb, Edmund B.; Zhang, Xiaohui; Angelov, Aleksandar K.

Abstract

The rapidly advancing field of micro and nano-manufacturing is continuously offering novel advantages to existing technologies. Micro-injection molding provides a unique opportunity to create substrates capable of controlling the mechanical environment in stem cell culture in a high throughput industrially relevant manner. The modification of such polymer surfaces to match the target surface stiffness of relatively more compliant biological tissues necessitates the movement towards higher aspect ratio smaller dimension features. The requirements provide a significant manufacturing challenge which is approaching a solution. The development of high aspect ratio large feature density polymer microarrays requires the synergistic optimization of design, material, mold tooling, and processing. A conventional mold base with steel inserts and controllable resistance heating was assembled to incorporate interchangeable inserts with microfeatured silicon inlays. Ultraviolet (UV) lithography with dry etching was used to impart microfeatures into silicon wafers with a variety of different geometries containing aspect ratios ranging from 0.92 to 6. Multiple polymer resins, including polystyrene (HIPS, PS), low density polyethylene (LDPE), cyclic olefin copolymer (COC), and thermoplastic polyurethane (TPU), were used to test replication and cellular response to materials with different bulk stiffness and topography-modified surface stiffness. The maximum achieved microfeature aspect ratio was 9.3 (high impact polystyrene), owed to tensile stretching during part ejection. For non-stretched substrates, the maximum molded aspect ratio was 4.5 (LDPE) and highest replication quotient (RQ = feature height / tooling feature depth) was 0.97 (COC). The maximum aspect ratio molded with consistent features across the entire surface was 2.1 (TPU).Parameters shown to enhance replication were mold temperature (Tmold = Tg was a critical replication transition point), injection velocity at higher mold temperatures, holding time, holding pressure, and nozzle temperature. The importance of certain parameters was material dependent, but mold temperature consistently had a relatively large impact.A concern that was addressed for a high density array of microfeatures was the consistency of replication, which is vital for the intended application and seldom address in published literature. Increased consistency was attained through strategic placement of temperature control, modification of the main cavity design, and optimized silicon tooling with reduced microcavity nanoroughness.Silicon tooling was fabricated with the initial objective being to achieve high aspect ratio negative features. However, with the realization of molding and demolding limitations, the tooling microfeature profiles were altered to include a taper and reduction of sidewall scalloping. Sophisticated methods of dry etching were used, in which a novel etching technique known as "passivation compensation," was utilized to manufacture microchannels containing low levels of roughness, a well-controlled tapered profile, and the prospect of high aspect ratios. With the new tooling, topography consistency was dramatically enhanced for both COC and TPU, with Taguchi orthogonal array optimization leading to RQs of 0.82 (aspect ratio = 2) and 0.85 (aspect ratio = 2.1), respectively.Water contact angle (WCA) measurements for both COC and TPU generally increased with an increase in surface roughness (dictated by microfeature dimensions), reaching WCA measurements of 139.8o and 141.1o, respectively. WCA hysteresis appeared to increase with roughness up to a critical value for COC while continuing to increase for TPU with a transition, which is thought to be the result of material properties. Moreover, hydrophobic surfaces containing high levels of hysteresis were attributed to the "petal" effect associated with hierarchical surface structures. Hydrophobicity has been shown to be related to biological cell behavior, and thus is an effective characterization technique to measure interfacial properties.Simulation of the injection molding process using conventional methods was used to describe general conditions present at microchannel inlets. The sprue gate and an increase in plate thickness gave the microfeatured region additional time to fill microfeatures prior to generation of a frozen layer. The delayed solidification is attributed to the low thermal conductivity associated with the polymer melt.A cell sensing model was developed based on the mechanical interaction between cell and substrate. The model provides a useful design map by which nanofeatured polymer geometry and material choice can be made to achieve a particular apparent surface stiffness. Bending mechanics were simulated for a few specific examples, providing an indication of the limitations associated with using higher aspect ratio nanostructures. A bending example was applied to a manufactured tapered pillar to note the stiffness reduction achieved through use of the substrates molded during the current study. Cell culture studies showed that the presence of topography had a dramatic effect on cellular morphology and on stress fiber thickness, causing an increase in thickness compared to flat controls. The cytoskeletal re-arrangements occurring may be indicative of a differentiation event, and future results will indicate whether that is the case. Unconventional morphology was observed in the presence of low aspect ratio COC microtopography, ranging from alignment with the micropattern to a circular conformation where adhesion is taking place exclusively in the middle of the cell. Micromolding of tensile bars was conducted to better understand the processing effects on mechanical and thermal properties of microscale molded components. Such results could provide useful general trends for the consideration of mechanical properties of molded microfeatures being exposed to cellular mechanical traction forces, especially considering the extreme processing conditions necessary to fill increasingly small and high aspect ratio features.Results revealed that the mechanical properties of COC is largely unaffected by a wide range processing conditions, but is reduced to approximately 41% of the value obtained from traditional tensile test results. TPU showed a dramatic dependence on molding properties, with higher injection velocities and lower mold temperatures resulting in reduced elastic modulus. Simulation was used to further elucidate the cause for varying properties. Microscale elastic modulus average values approximately 31% higher compared to traditional tensile test results. Trends in thermal properties were not apparent, and were difficult to detect from relatively weak melting peaks. The use of two different polymer lots elicited drastically different results, prompting the further investigation of the differences. Crystallinity, viscosity, and chemical bond structure was found to be very different from one lot to the other.The successful fabrication of uniform tapered microfeatures with middle range aspect ratios were manufactured, and the robust mold design and the tooling fabrication method provides a blueprint for achieving higher aspect ratios with a significant level of fidelity in the future. The enhanced macro and microscale mold design, combined with a deeper understanding of processing induced mechanical thermal microscale properties, can be used to tailor the substrate bio-interface properties to the desired mechanical structure for controllable hMSC behavior.

Share

COinS