Micro and nanofabrication

Silicon carbide micromachining

Wide band gap materials such as diamond like carbon, III–nitride, and silicon carbide have been widely applied in many electronic devices, where conditions include high power and high temperature. Among numerous materials, silicon carbide (SiC) is a preferable choice owing to its excellent physical properties along with the availability of wafers. The main obstacles hindering the wider applications of SiC include the high wafer costs and low etching rate of SiC compared to Si. Cubic silicon carbide thin films provide a valid alternative as they can be grown on a silicon substrate and have been used as an excellent platform for MEMS. Epitaxial SiC films grow on large scale Si wafers, taking advantage of low cost Si wafers, and simplifying the fabrication process of SiC MEMS. To increase the application feasibility of SiC in harsh environments, electrical insulators such as SiO2 can prevent the leakage current from the functional layer (SiC) to the substrate (Si). In addition, the SiC on the SiO2 platform is also of significant interest for optical applications such as SiC waveguides or photonic crystals since SiC offers a higher refractive index than SiO2. To date, the most common method to form SiC on an insulator is based on wafer bonding. However, this method typically requires smooth surface treatment and the removal of Si. The NTNLab has devised a novel technique to form SiC on an insulator by thermally oxidizing the Si substrate. This technique is based on the chemical inertness of SiC, which allows the material to remain unchanged during the thermal oxidation process. NTNLab’s experimental data show that thermal oxidation of silicon makes the development of SiC nano structures (eg nanowires) on an insulator possible without the requirements of wafer-bonding and/or silicon removal. The proposed method could result in SiC based MEMS devices including high temperature sensors, and nanowire waveguides.1

Other work in NTNLab examines strain engineering, a topic which has attracted much attention, particularly for epitaxial films grown on a different substrate. Residual strains of SiC have been employed to form ultra-high frequency and high Q factor resonators. However, the highest residual strain of SiC was reported to be limited to approximately 0.6%. Large strains induced into SiC could lead to several interesting physical phenomena, as well as significant improvement of resonant frequencies. NTNLab research examines an unprecedented nanostrain-amplifier structure with an ultra-high residual strain up to 8% utilizing the natural residual stress between epitaxial 3C-SiC and Si. In addition, the applied strain can be tuned by changing the dimensions of the amplifier structure. The possibility of introducing such a controllable and ultra-high strain will lead to investigating the physics of SiC in large strain regimes and the development of ultra-sensitive mechanical sensors.2

Work from NTNLab has explored the thermoresistive property of p-type single crystalline 3C-SiC (p-3C-SiC), which was epitaxially grown on a silicon (Si) wafer, and then transferred to a glass substrate using a Focused Ion Beam (FIB) technique. A negative and relatively large temperature coefficient of resistance (TCR) up to -5500 ppm/K was observed. This TCR is attributed to two activation energy thresholds of 45 meV and 52 meV, corresponding to temperatures below and above 450 K, respectively, and a small reduction of hole mobility with increasing temperature. The large TCR indicates the suitability of p-3C-SiC for thermal-based sensors working in high-temperature environments.3 In other work, the Lab has identified the gauge factor of the p-type nanocrystalline SiC was found to be 14.5, which is one order of magnitude larger than that in most metals. This result indicates that mechanical strain has a significant influence on the electrical conductance of p-type nanocrystalline SiC, which is promising for mechanical sensing applications in harsh environments.4

Polymeric micromachining

Poly(methyl methacrylate) (PMMA) has been used as a basic substrate material for fabrication of microfluidic devices due to its high mechanical stability, good chemical properties and excellent optical clarity. The manufacturing process of PMMA microchips includes several fabrication methods such as hot embossing, injection moulding, LIGA (German acronym for x-ray lithography electro deposition and moulding), and excimer laser (ultra-violet laser) ablation. For replication methods such as hot embossing and injection moulding, the costly and time-consuming mould fabrication process hampers the rapid turnaround of new design although the technology is intrinsically suitable for mass production. The implementations of x-ray lithography and excimer laser ablation are expensive because of the high cost of the x-ray and excimer laser system and the corresponding masks. More recently, the cheaper CO2 laser has been used to create microchips in PMMA. This technique has been proven to be rapid and effective for fabricating microfluidic devices, especially for scientific trials and small-scale production.1

The NTNLab presents a laser-micromachined polymeric membraneless fuel cell. The membraneless fuel cell, constructed with three PMMA layers, takes advantage of two laminar flows in a single micro channel to keep the fuel and oxidant streams separated yet in diffusional contact. The packed fuel cell has been electrochemically characterized by an electrochemical analyser. The Lab demonstrated for the first time the use of hydrogen peroxide in sulfuric acid as the oxidant. The new oxidant composition allows a simple recycling process and better fuel utilization.2

PMMA due to its cost, transparency and mechanical and chemical properties, is of interest to the microfluidic research community. Meanwhile, the more flexible polydimethylsiloxane (PDMS) is well suited for pneumatic actuation. However, PDMS is permeable to gases and absorbs molecules from the sample liquids. The NTNLab has developed a simple and reliable technique for bonding PMMA to PDMS. A 25 µm thick adhesive was laminated onto a clean PMMA surface. The PDMS/adhesive membrane acts as the pneumatic actuator for the micropump. Pressurized air was switched to the three pneumatic actuators by solenoid valves and control electronics. The micropumps can achieve a flow rate as high as 96 µ/min, allowing the integration of microfluidic components made of both PMMA and PDMS in a single device.3

Low-cost alternative devices

Flexible and multifunctional electronic devices show potential for various applications including human-motion detection and wearable thermal therapy. The key advantages of these systems are (1) highly stable, sensitive and fast-response devices, (2) fabrication of macroscale devices on flexible substrates, and (3) integrated (lab-on-chip) and multifunctional devices. However, their fabrication commonly requires toxic solvents, as well as time-consuming and complex processes. The NTNLab is successfully researching low-cost, rapid-prototyping and user-friendly fabrication of flexible transducers using recyclable, water-resistant poly(vinyl chloride) films as a substrate, and ubiquitously available pencil graphite as a functional layer without using any toxic solvents or additional catalysts. The flexible heaters showed good characteristics such as fast thermal response, good thermostability (low temperature coefficient of resistance) and low power consumption. The heaters with their capability of perceiving human motion were shown to be effective. Results from the Lab indicate that a wide range of electronic devices fabricated from the environmental-friendly material by this simple and user-friendly approach could be utilized for cost-effective, flexible and low power consuming thermal therapy, health monitoring systems and other real-time monitoring devices without using any toxic chemicals or advanced processes.1

Paper-based microfluidics and sensors have attracted great attention as low-cost alternative devices. Work in the NTNLab examines the development of a self-sensing paper-based magnetic-actuator. The actuating component of the device was formed by employing nanoferromagnetic particles absorbed into porous paper. A graphite layer was used to create the mechanical sensing element for the paper-based actuator. For the proof of concept, a cantilever-type device was fabricated using a clean-room free process, demonstrating the actuating and self-sensing functions integrated in a single paper-based device.2

NTNLab is also investigating graphite on paper (GOP), in which graphite layers are deposited on paper by either graphite-ink printing techniques or manual pencil drawing techniques. The excellent electrical conductivity of graphite and the low thermal conductivity of porous paper make GOP an attractive material for thermoresistive sensors. Consequently, GOP, as a transducing element for thermal-based sensors, has remarkable advantages such as low cost, cleanroom-free fabrication and high sensitivity.3. The Lab is also investigating solvent-free fabrication of a biodegradable hot-film flow sensor for noninvasive respiratory monitoring. Our research shows that the solvent-free and cleanroom free fabrication of a degradable flexible sensor could be affixed to the upper lip to noninvasively monitor human respiration in real-time with a high sensitivity. The sensors offer high sensitivity, low power consumption, high signal-to-noise ratio (SNR) and good long-term stability. Results indicate that sensors fabricated by this simple but effective approach can be utilized to produce low-cost, environment-friendly, sustainable, flexible and wearable sensors for noninvasively monitoring human respiration in realtime.4

NTNLab is also investigating flexible and stretchable electronics that have a wide variety of wearable applications in portable sensors, flexible electrodes/heaters, flexible circuits and stretchable displays. Spinnable carbon nanotubes (CNTs) constructed on flexible substrates are a potential material for wearable sensing applications owing to their high thermal and electrical conductivity, low mass density and excellent mechanical properties. The Lab is exploring wearable thermal flow sensor for healthcare using lightweight, high strength, flexible CNT yarns as hotwires, pencil graphite as electrodes, and lightweight, recyclable and biodegradable paper as flexible substrates, without using any toxic chemicals.5

Droplet based microfluidics

Convective heat transfer in droplet-based microchannel heat sinks can be enhanced by the recirculating vortices due to the presence of interfaces. In rectangular microchannels, the three dimensional structures of the vortices and the ‘gutters’ (ie the space between the curved droplet interface and the corner of the microchannel) can significantly affect the heat transfer process. Numerical simulations of the heat transfer process are performed to study the three dimensional features in droplet-based microchannel heat sinks, with results showing that the ‘gutters’ can hinder the heat transfer process because of its parallel flow. On contrast, the recirculating flow in droplets and in slug regions between successive droplets can enhance the heat transfer by advecting hot fluid towards the center of the droplets/slugs and advecting fresh fluid towards the wall of the channel.1 Research from the NTNLab proposes to continuously generate droplets with programmable concentration using the ion concentration polarization (ICP) phenomenon. Via this method, our fabrication method has the advantages of simplicity, reliability, repeatability and low-cost. Sample droplets with up to 100-fold concentration were generated continuously with our device.2

Bioaffinity mass spectrometry screening is a novel approach using non-denaturing electrospray ionization (ESI) mass spectrometry (MS) in identifying drug leads. This technique can detect and preserve noncovalent protein-active drug ligand complexes under different physiological conditions. The screening challenge being the reduction of sample volume needed. The NTNLab demonstrates that analysis of samples can be performed using droplet-based microfluidics, with results showing that a MS instrument with a conventional ESI source can detect the samples and distinguish it with the separating oil phase.3

The NTNLab is committed to developing new micron-sized droplets. Delicate control of the droplet generation process is needed to address complex applications. At the liquid/liquid interface, an energy imbalance leads to instability and droplet breakup. Controlling the droplet generation process is a central topic of microfluidics.4 Work in the NTNLab presents state-of-the-art technologies to efficiently sort droplets. We classify the concepts according to the type of energy implemented into the system. Combining digital microfluidic devices allows for functional, flexible and powerful systems, covering specific technological areas and applications where electronics alone cannot compete or cannot offer an integrated chemical or biochemical functionality. Most powerful is the compatibility with digital electronics that allows for programmability and storage of droplets and droplet trains.5

Our work introduces an effective method to actively induce droplet generation using negative pressure. Droplets can be generated on demand using a series of periodic negative pressure pulses. Fluidic network models were developed using the analogy to electric networks to relate the pressure conditions for different flow regimes. Experimental results show that the droplet volume is correlated to the pressure ratio with a power law of 1.3.6

Automated droplet measurement (ADM) software has been developed based on OpenCV image processing library that has much higher throughput than currently available software. The process speed was found to be higher than the transfer speed even when visual feedback is enabled. Added to this, to shorten the total time taken on droplet measurement further, our Lab integrated newly developed ADM software and camera software development kit together in a new process flow. This was done by performing video transfer/streaming simultaneously with video processing. The total time for the droplet measurement using the integrated software was found significantly shorter than using the old process flow with DMV software.7

Micro optofluidics

For decades, microelectromechanical (MEMS) systems sensors have been intensively developed using various physical concepts, including piezoresistive, piezoelectric, and thermoresistive effects. The thermoresistive effect, which refers to the electrical resistance change with temperature variation, has many advantages in terms of simplicity in design and implementation. Based on this effect, various micro-sized thermal sensors, which can monitor temperature, flow and acceleration, have been successfully fabricated, thanks to the advancement in MEMS technologies.1

Research from the Nam-Trung Nguyen lab (NTNLab) examines the design and demonstration of an optofluidic in-plane bi-concave lens to perform both light focusing and diverging. With the concave lens hydrodynamically formed in a rectangular chamber with a liquid core cladding (L2) configuration. Our research has found that the focusing mode of the opto- fluidic bi-concave lens can be simply switched to diverging mode by stopping the pumping the liquid from core inlet. The previous cladding liquid (cinnamaldehyde) would join together in the chamber and become the core stream for the diverging mode. The auxiliary cladding liquid would change its function from avoiding light-scattering to becoming the cladding stream for the diverging mode. Combining the performance of focusing and diverging, the tunability of the focal length of the optofluidic lens and the width of the light beam can be greatly enhanced. This improvement of tunability will make this optofluidic bi-concave lens more adaptive to the lab-on-chip applications.2

NTNLab is also examining modelling and experimental results of a liquid-core liquid-cladding optofluidic lens under the combined effect of hydrodynamics and electro-osmosis. To allow the lens to be tuned by a voltage, the cladding fluids are electrically conducting, while the core fluid is non-conducting. Our results show that the interfaces between the cladding fluids and the core fluid have optically smooth arc shape. Under fixed cladding flow rates, the same voltage forms symmetric biconvex lens only. Different voltages can form biconvex lens, planoconvex lens, and meniscus lens.3

Integrating a flow cytometer into microfluidic networks helps to miniaturize the system and make it portable for field use with applications in environmental monitoring, industrial testing and biochemical studies. Integration of optical components, such as lenses, further improves the compactness. However, many current designs suffer from severe light scattering due to roughness of the solid-based lens interface. Our research examines a flow cytometer using an optofluidic lens to focus the light beam. Benefiting from the smooth liquid–liquid lens interface and the refractive-index matching liquid as cladding streams, a light beam can be well focused without scattering. The variations of the signal peak values are reduced, owing to the small beam width at the beam waist.4 We are also undertaking work on an optofluidic aperture stop formed by the liquid-core/liquid-cladding flow;5 and numerically and experimentally investigating and the design of an optofluidic in-plane bi-concave lens to perform both light focusing and diverging using the combined effect of pressure driven flow and electro-osmosis.6

Further work by the NTNLab illustrates the concept of liquid crystal based tunable optofluidic polarizer. Our research to date demonstrates the working principle, fabrication and characterization of a novel optofluidic tunable polarizer suitable for lab-on-a-chip applications. The concept and developed probe can find potential optics and lab-on-a chip applications.7

Lab-on-a-chip and point-of-care diagnostics

Rare cells are low-abundance cells in a much larger population of background cells. It is difficult to isolate and analyse rare cells because of their generally low selectivity and significant sample loss. Microfluidics provides robust solutions including apparent performance enhancements resulting in higher efficiencies and sensitivity levels. It also provides simpler handling of small sample volumes and multiplexing capabilities for high-throughput processing. The NTNLab continue to examine the design considerations of representative microfluidic devices for rare cell isolation and analysis.1

A system of multiple organs integrated on a single chip or human on a chip (HUC) has potential in drug discovery, and could overcome the limitations of animal models including cost and incompatibility with human physiology. The matching length scale of biological structures and micromachined components makes a microfluidic chip the ideal platform to investigate physiological events. The NTNLab continues to study the development of HUC, including the integration of cell culture on a chip to create an ethical human model and to provide insights into the sensitivity of different cell constructs on drugs.2

Understanding the process of fusion of olfactory ensheathing cell (OEC) spheroids will lead to improvement of cell transplantation therapies to repair spinal cord injuries. OECs are neuroglia cells that provide support and protection for olfactory sensory neurons in the peripheral nervous system (PNS). The NTNLab has developed a microfluidic device that enables injection and capture of spheroids of OECs.3

Piezoelectric thin films including zinc oxide and aluminium nitride are used in lab-on-chip applications such as biosensing, particle/cell concentrating, sorting/patterning, pumping, mixing, nebulisation and jetting. Integrated acoustic wave sensing/microfluidic devices have been fabricated by depositing these piezoelectric films onto substrates for making flexible devices, which have the potential to integrate disposable, or bendable/flexible lab-on-a-chip devices into various sensing and actuating applications. The NTNLab continues to engineer high performance piezoelectric thin films, with our research examining critical issues such as film deposition, MEMS processing techniques, control of deposition/processing parametres, film texture, doping, dispersion effects, film stress, multilayer design, electrode materials/designs and substrate selections.4

Current in vitro gut models lack physiological relevance, and various approaches have been taken to improve current cell culture models. For example, mimicking the 3D tissue structure or fluidic environment has shown improved physiological function of gut cells. The NTNLab’s research has incorporated a collagen scaffold that mimics the human intestinal villi into a microfluidic device, providing cells with both 3D tissue structure and fluidic shear. Our research suggests the combination of fluidic stimulus and 3D structure induces improvement in gut functions.5

Significant progress has been made in the fabrication and characterisation of MEMS thermal sensors that use the thermoresistive effect in metals such as platinum and semiconductors including silicon and polysilicon. Drawbacks include cost and the low capability to work in high temperature environments. Research in alternative thermal sensing materials with high thermoresistive sensitivity is being investigated by the NTNLab, as is proper packaging for thermal sensors as thermal expansion at high temperatures can cause cracking in devices. An alternative trend is low-cost materials for thermal sensors, which can be processed using cleanroom-free facilities and user-friendly techniques. For example, the thermoresistive effect in pencil graphite has been intensively investigated, and a proof of concept of thermal flow sensors has been successfully demonstrated.6