Photolithography: SU-8 2035 Soft Lithography Process (UC Berkeley BNC)

Oxygen plasma converts hydrophobic PDMS surfaces (contact angle ~108°) to hydroxylated states through UV excitation and radical attack. The five-step chemical mechanism generates Si-OH groups that enable irreversible siloxane bonding between surfaces. Optimal bonding occurs when contact angle falls below 10°, correlating with maximum hydroxyl group density. High-pressure low-power conditions minimize substrate damage while achieving effective surface activation.

MEMS assembly employs multiple bonding techniques: anodic bonding uses high voltage and heat to create charge depletion regions enabling lattice fusion; plasma bonding modifies surface energy to create hydrophilic surfaces forming siloxane linkages. Oxygen plasma treatment converts hydrophobic polymers to hydrophilic states, enabling clean bonding without adhesive layers. This is essential for polymer MEMS where thick adhesives would compromise micro features.

Plasma surface treatment improves adhesion by simultaneously cleaning surfaces through reactive species that break down contaminants and activating surfaces through oxygen plasma to increase surface energy, enabling better bonding with adhesives, inks, and coatings; the treatment works effectively on various materials including plastics, glass, ceramics, and metals, though activation duration varies significantly between rigid materials (lasting days) and flexible materials (lasting only minutes).

In top-gate oxide TFTs, oxygen plasma time during PD SiO2 deposition critically affects device performance. Longer plasma times (2-4 seconds) create more positive source voltages and reduce mobility, while shorter times (0.3 seconds) produce very negative source voltages. Hall measurements confirmed significant carrier concentration variations due to plasma treatment. Oxygen plasma fills oxygen vacancies in the semiconductor, increasing metal-oxygen bonding while reducing vacancy concentration. However, shorter plasma times leave more hydrogen impurities in the film because reduced chemical reactivity allows unreacted precursors to remain. This hydrogen-oxygen balance determines whether hydrogen acts as a beneficial passivator or causes unwanted conductivity, highlighting the complex trade-offs in plasma process optimization.

Metrology drives innovation in precision engineering for aerospace components with tight specifications. When new ideas are generated, R&D creates computer-aided designs, and manufacturing produces components requiring new measurement methods and instruments. Innovative metrology involves inventing new instruments and fixtures for complex profiles. Metrology objectives include determining appropriate instruments for plant activities, ensuring maintenance through periodical calibration, suggesting instruments for quality control (in-process vs off-process gauging, range and accuracy requirements), and providing training for operators. Statistical quality control uses control charts with upper and lower limits to monitor process stability, taking corrective action when measurements exceed control limits.

This course covers the fundamentals of metrology and quality control in industrial settings, teaching students how to properly measure and verify products using common measurement equipment including micrometers (exterior, interior, thread), precision balances, thickness gauges, depth gauges, mechanical comparators, goniometers, dynamometers, torque wrenches, nanometers, roughness meters, and thickness gauges. Students learn to develop quality control plans, select appropriate measurement equipment for specific situations, and avoid measurement errors to ensure product quality.

The 'metrology bubble' refers to the isolation of metrology professionals within their own laboratories, where they may believe they are doing everything correctly without external perspective. Passing an audit does not mean a laboratory is doing everything right—it only means they met minimum requirements at that moment. PFA (Probability of False Accept) and PFR (Probability of False Reject) are critical quality control concepts. PFA occurs when defective instruments are incorrectly accepted, while PFR occurs when good instruments are incorrectly rejected. Both are equally problematic for customers. Metrology training benefits from multiple learning modalities, with kinesesthetic (hands-on) learning being particularly important for most technicians. Resolution is the primary factor that differentiates trained technicians in their measurements. The Unit Under Test (UUT) resolution should be included in the uncertainty budget reported to customers but not in the scope of accreditation uncertainty.

This segment explains ABNT standards requiring engine assemblers to perform all metrology and verification procedures. The instructor highlights that machining shops often fail to properly measure bushings before shipping, creating quality risks. The video demonstrates how to identify problematic bushing lots by comparing measurements from multiple samples, showing how variations of 0.25mm between bushings from the same lot indicate quality issues. Assemblers must verify all measurements and return non-compliant parts for adjustment.

Measurement is the process of determining the quantity of any physical quantity. Metrology is the science of measurement, encompassing establishing standards, their applications, verification, and precision. The main purposes include establishing universal measurement standards for consistency, developing measurement methods, determining uncertainty, and eliminating errors. Physical quantities are measurable quantities requiring units for representation. Fundamental units (independent) include Length (meters), Mass (kilograms), Time (seconds), Temperature (Kelvin), Electric Current (Amperes), Amount of Substance (Moles), and Luminous Intensity (Candela). Derived units are obtained through multiplication or division. Measurement standards are categorized into Primary (highest accuracy), Secondary, and Working Standards. Repeatability ensures consistent results under identical conditions. The measurement process involves determining objectives, selecting instruments, checking calibration, verifying environmental conditions, performing measurements, analyzing results, and documenting findings. Measurement plays a crucial role in quality control by ensuring products meet specifications, maintaining accuracy and reliability, ensuring compliance with standards, controlling manufacturing processes, identifying defects, maintaining customer satisfaction, and supporting audits. Five main factors affect measurement accuracy (SWIPE): Standards, Workmanship, Instrument quality, Personnel, and Environment. Linear measurement instruments include Vernier Calipers (0.1mm least count), Micrometers (0.01mm least count), Depth gauges, Height gauges, and Slip gauges. Gauges for direct measurement include Plug gauges, Wire gauges, and Fillet gauges. Measurement uncertainty refers to doubt in results arising from instrument limitations, environmental effects, human errors, and calibration issues. Errors are classified as Systematic (consistent direction, correctable through calibration), Random (unpredictable, reduced by multiple measurements), and Human (reading/recording mistakes).

Multilayer technology in semiconductor manufacturing involves using a planarizing bottom layer combined with a hard mask and thin photoresist to improve depth of focus and resolution, enabling continued miniaturization when single-layer photoresist approaches reached their physical limits; this multi-layer stack structure provides the necessary resistance to withstand etching processes while optimizing light exposure at the interface between layers.

The Maskless Aligner MLA150 enables direct photolithography without requiring a physical photomask by using CAD files to expose patterns directly onto coated substrates, streamlining the four-step process of system setup, substrate loading with vacuum chuck, automated alignment using camera systems to compensate for rotation and scaling, and exposure, making complex multi-layer patterning more accessible and efficient.

Modern integrated circuits employ a layered architecture with devices (transistors, diodes) at the bottom and multiple metal interconnect layers above, often exceeding 10 layers. This multi-layer approach solves routing complexity by allowing wires to run in different planes, preventing short circuits that would occur if all connections were confined to a single plane. Photolithography enables this complexity by transferring patterns from masks to silicon wafers through a four-step process: applying photoresist, exposing through a mask, developing the exposed regions, and etching the underlying layer. This technology allows thousands of identical chips to be manufactured from a single silicon wafer, achieving the mass production necessary for affordable consumer electronics.

Moore's Law predicted transistor counts doubling every two years, driving 50+ years of technological progress. By 2015, traditional 193nm lithography reached its physical limits. The solution required EUV lithography using 13.5nm light, 100 times shorter than conventional UV. In the 1980s, Japanese scientist Hiroshi Kinoshita proposed X-ray lithography but faced immediate challenges: X-rays at 10nm wavelengths eject electrons from atoms, causing absorption by most materials and requiring vacuum environments. Traditional lenses cannot focus X-rays, necessitating multilayer mirrors. James Underwood and Trey Barbee developed tungsten-carbon mirrors achieving 6% reflectivity at 4.48nm, proving X-ray lithography was theoretically possible. Despite Kinoshita's 1986 demonstration, the scientific community remained skeptical, dismissing the technology as impractical.

Microfluidics miniaturizes bioseparation from 50-micron capillaries to chip-scale devices, reducing reagent consumption, decreasing waste volume, and enabling rapid high-performance separations. The goal is credit-card-sized microchips performing laboratory capabilities, similar to how integrated circuits enabled microelectronics. Microfluidic chips use electrokinetic movement (electroosmotic flow) and can achieve 10-100x speed improvements over traditional systems. Computer-aided design is essential for creating injector blocks, separation blocks, and detection parts, considering heat dissipation challenges. Multi-channel systems (96 or 384 channels) enable easy multiplexing. These devices can perform synthesis and analysis in space, as demonstrated by International Space Station experiments where compounds with solubility enhancement survived a year (compared to 5 years on Earth).

Microfluidics studies fluid flows at micron scales (50-300 micrometers), integrating mechanical, biological, and chemical disciplines. Lab-on-a-chip integrates laboratory functions onto single chips, performing multiple tests simultaneously with minimal samples (drops vs. milliliters). Key advantages include reduced material consumption, lower power requirements, improved safety, automatic operation, and portability for point-of-care testing. This technology enables rapid, accurate diagnostics comparable to conventional labs but in compact, accessible formats.

Microfluidic drops serve as miniaturized reaction vessels (reactors) for lab-on-a-chip applications. Each picoliter-sized drop acts as an independent experimental chamber, enabling thousands of parallel experiments simultaneously. This dramatic reduction in reagent consumption (from microliters to picoliters) makes high-throughput screening economically feasible where traditional robotic approaches would be prohibitively expensive. Microfluidic chips enable sophisticated operations including splitting drops into equal or unequal portions, merging drops to combine reactants, and sorting drops based on properties using electric fields. These capabilities support complex workflows including sequential reactions and titrations. Directed evolution using microfluidic screening enables rapid identification of improved enzyme variants through high-throughput selection. Libraries of mutated genes are expressed as yeast surface-displayed proteins, each in its own drop. Adding substrate reveals functional variants through fluorescence. Functional variants are selected, amplified, and subjected to further rounds of mutation and selection, rapidly converging on improved enzyme variants through iterative optimization.

Microfluidics manipulates fluids at micro and nanoliter scales, enabling precise control over chemical reactions and biological processes. This technology underlies modern diagnostic tests like rapid antigen tests and pregnancy tests. Key applications include drug screening, where cells are exposed to controlled drug concentrations in microcompartments, and single-cell analysis. Microvalves control fluid flow, with passive valves made from dissolvable polymers enabling automated, programmable fluid handling for cell culture applications. The field represents a convergence of physics, chemistry, biology, and engineering, with applications ranging from medical diagnostics to fundamental research.