Symposium S.CT07—Micro-Assembly Technologies and Heterogeneous Integration—Fundamentals to Applications

The bottom-up approaches recently become quite attractive for nano-scale fabrication. For example, block copolymers, nano-colloidal particles, and anodized aluminum are known to form nano-scale patterns. In addition, these bottom-up self-assembly can be combined with conventional top-down lithography resulting large scale periodic pattern can be fabricated. Several applications to nano/micro-electronics have been reported. Such as patterned magnetic media for a hard disk drive (HDD), light extraction structures of light-emitting diodes (LEDs), organic light-emitting diodes (OLEDs), anti-reflection (AR) structures for displays, metal wire-grid polarizers (WGP), and transparent conductive film (TCF) using nano-meshed metal thin film were demonstrated by these manners. The guided, aligned, or directed self-assembly (DSA) is a key technique, that the larger guides are made by conventional top-down methods then bottom-up self-assembled structures are formed along the guides. The challenges are continuing to fabricate the semiconductor memories. We would explain these potential applications fabricated by self-assembly.

In this work, a new metallization to form metal nano wires are studied with nanocomposite consisting of block co-polymer and metal compounds or metal nano particles by advanced directed self-assembly (DSA) technology. To evaluate the impact of the metals, we simply characterize the standard DSA behavior with block co-polymer alone. Si trenches with a length of 10 μm and width of 3 μm are fabricated by dry Si etch processes used for 3D integration with TSV (Though-Silicon Via) applications. This trench size is quite larger than that used for typical DSA works. Surprisingly, we have successfully formed lamella (multiple annular) and nano cylindrical structures in the large Si trenches by using polystyrene-block-polymethacrylate (PS-b-PMMA) with the PS:PMMA composition of 1:1 and 2:1, respectively. When we used PS-b-PMMA with a PS:PMMA molecular weight ratio of 15K:15K and 57K:25K, the resulting lamella and cylinder pitches are 20 nm and 40 nm. The pitches depend on the molecular weight of the nanostructure. We evaluate the impact of DSA parameters such as heating time and temperature on morphological behavior and the resulting nanostructure of the block copolymers. The nanostructures were formed even at below 200degreeC. In addition, it is found that the nanostructure can be controlled by adding the homo polymer of PS or PMMA into the block copolymers. From mean field approximation simulation based on Self-Consistent Field (SCF) theory, if a metal compound well interacted with the PMMA or PS block is added into PS-b-PMMA, the metal in the nanocomposite will be kept to be distributed into nanocylindrical PMMA or PS to fabricate lateral or vertical ultrafine-pitch interconnection. Here, we employed a nanocomposite with an Fe compound and PS-b-PMMA for nano wire formation by DSA. According to Moore’s law limitation, metal nano wires are promising candidate to breakthrough the interconnect scaling. It was interesting to note that SEM observation revealed Fe nano wires were induced by the phase separation of the PS-b-PMMA in the Si microstructure. The EDX analyses confirmed the added metal was either attached or aligned near the PMMA of the block copolymer. In addition, another nanocomposite is prepared with PS-b-PMMA and PMMA-modified Au nano particles. By DSA processing with simple heating, the nanocomposite shows nano wires with a width of 30 m and the length of 500 nm or more. The electrical characterization is done with a nano probing system, indicating the liner relationship between current and voltage although the resistance is relatively high. By using DSA with the block co-polymer/metal nanocomposite, this is the first time to show the ohmic behavior applicable to interconnection. In this presentation, another DSA approach without nanocomposites will be proposed and their feasibility works are introduced.

Directed self-assembly of block copolymers (BCPs) provides a scalable route for rationally fabricating dense nanoscale patterns beyond the resolution of conventional lithography at highly precise locations on a surface. Historically, there have been two strategies—graphoepitaxy and chemoepitaxy—for directing the self-assembly of BCPs into useful patterns. We have recently discovered a third paradigm—termed “boundary-directed epitaxy”—in which templates consisting only of planar, low-resolution features are used to drive the formation of more complex BCP patterns with enhanced feature resolution. The templates are comprised of spatial boundaries separating regions on a surface with different composition, formed at the edges of isolated stripes on a background substrate.

Vertical BCP lamellae are pinned by chemical contrast at each stripe/substrate boundary, align parallel to the boundaries, selectively form on the stripes (whereas horizontal lamellae form on the background substrate), and register to wide and incommensurate stripes to multiply the feature density. Isolated BCP line arrays with half-pitch of 6.4 nm are demonstrated on stripes wider than 80 nm. Directed assembly of isolated line arrays, superstructures of isolated line arrays, and complex features such as T-junctions, jogs, and bends is demonstrated, which may be particularly useful for fabricating critical features in integrated circuits. Boundary-directed epitaxy circumvents the need for topographic structures used in graphoepitaxy or ultra-narrow guiding features used in chemoepitaxy to direct assembly of sub-10 nm BCP features, and provides an attractive path towards nanofabrication beyond the resolution of conventional lithography.

Direct emission displays made with inorganic µLEDs can potentially challenge LCD and OLED displays with better performance, efficiency and lifetime. Markets where microLED displays are likely to first appear are large information displays requiring high brightness, home theater applications benefitting from increased dynamic range and automotive displays in which high contrast and reliability are necessary. These displays are made possible by the integration of two mainstream technologies, namely GaN LED developed for general lighting and the large area active matrix backplanes developed for LCD displays.
The unique manufacturing challenge for µLED displays is development of a technology to position and connect millions of small devices with zero defects and low cost. Conventional pick and place tools are capable of positioning µLEDs but the throughput of serial assembly is inadequate while mass-transfer methods have difficulty changing the pixel pitch. To address these challenges eLux has developed a massively parallel fluidic assembly technology which positions each µLED by capture of the device in a well structure that also contains the connecting electrodes.
Fluidic assembly technology uses simple low-cost equipment to achieve assembly rates up to 50 million µLEDs per hour and it is anticipated that production equipment could assemble an 8K display with 100 million LEDs in a few minutes. The technology is suitable for µLEDs from 10 to 200 µm diameter offering flexibility to make a wide variety of displays with resolutions from 400 to 10 ppi. Similarly, the microLED emitter area can be changed independently of the assembly technique to increase brightness and it is possible to produce displays over 5,000 cd/m². Fluidic assembly applies relatively low force on the device so brittle materials such as red µLEDs fabricated from AlGaInP can be assembled in the same way as blue and green emitting GaN µLEDs.
A fully integrated manufacturing system for µLED fluidic assembly will ensure that the assembled µLEDs have acceptable performance with special emphasis on eliminating shorted or high resistance devices. Testing and binning used in the general lighting industry are not practical at the micron scale, so wafers are mapped by cathodoluminescence to identify defective µLEDs. Based on the defect map, a selective harvest technique diverts bad µLEDs from the suspension so only known good die are used for assembly. Fluidic assembly relies on an excess of µLEDs to ensure sufficient capture attempts for complete assembly, so it is also necessary to capture, filter and recycle unassembled µLEDs.
With a suitable choice of complementary characteristic for the device and the trap site, it is possible to use fluidic assembly techniques to position nanowires or nanoparticles as well as µLEDs. The presentation will give an overview of the integration issues with some suggestions for other possible applications and approaches.

Conventional manufacturing is able to assemble products composed of many parts of the size approximately 1mm and above. At the nanometer scale, techniques such as deposition, lithography, and self-assembly are used to make a myriad of heterogeneous products. However, at the mesoscale (1-1000μm) these techniques don’t overlap to produce structures at scale. There is a need for robust and scalable technologies which are able to join structures which can be developed on the micro-scale into large area, heterogeneous assemblies. To address this, we are developing a printing platform which is able to use micro-objects as ink—microLEDs, sensor chips, integrated circuits, micro- and nanoparticles. These objects are dispersed in a dielectric fluid and can be printed similar to a xerographic process, but maintain single object precision and orientation. To do this, an optically-addressed electrostatic actuator array manipulates the particles to move to target locations. From there, an integrated transfer system removes the particles from the assembly platform to be placed on a final substrate while maintaining their relative positioning. This transfer system can either be batch with adhesive transfer, or continuous with a non-contact electrostatic belt. This talk will describe the process in which we are able to rapidly sort, manipulate, orient, assemble, and transfer particles in parallel over a 1-inch wide area. We have demonstrated 1-µm registration, assembly of hundreds of particles in seconds, and roll-to-roll like transfer. With the use of computer vision to aid the assembly process, the technology has the potential to produce highly ordered, custom-patterned, heterogeneous assemblies accurately over a large area much faster than traditional manufacturing techniques while still maintaining fidelity.

Radar technology is a field in which the measures (and hence the counter-measures) are continuously emerging. Constant vigilance on the air borders require stringent performance of the Radar stations and hence the counter measures are required to be developed using either the improved engineered structures of aircrafts or improved materials (which could be coated on the body of the aircrafts). A combination of both these approaches would yield much smarter and effective approach.
Micro-assembled, engineered thin film structures of silver arrays have been used along with different carbon nanocomposite thin films on flexible substrates for shielding the X-Band (8.2 GHz to 12.4 GHz) Radar range of the electromagnetic spectrum. In this presentation, we discuss on synthesis of flexible thin films of silver arrays of split ring resonators, with varied unit cell dimensions, which are deposited using standard micro-printing technology. These layers are stacked with carbon nanocomposite films (either of multiwalled carbon nanotubes (MWCNTs), graphene sheets and carbon conducting ink; in Polyvinyl dimethyl fluoride / polyacrylonitrile matrix) to obtain wide band electromagnetic shielding. Nanocomposites were synthesised using standard chemical methods and thin films were synthesised using dip-coating method. The sandwiched (2 layers of nanocomposites and one layer of resonator array) was packed using hydraulic press and standard lamination technique to yield a compact flexible film, which can be inserted in the X-Band wave guide for further testing. Vector Network Analyser was used for further studies of absorption. Shielding in entire X-band with the power loss of 27 dB (99.98 % loss) and a minimum of 21 dB (99.92 % loss) is demonstrated here, for MWCNT. Similar results for other carbon composites are discussed in this presentation. A strong interplay of carbon nanocomposites with the array of L-C-R equivalent resonators of varied resonant frequencies shows concomitant suppression in the incident electromagnetic power; the results supported by simulation (CST Microwave Studio) studies as well. This work opens up a strategy to obtain wide band shielding using a combination of optimised nanomaterial composites, heterostructured with engineered metallic resonators.

While dip coating is a widely used solution-based assembly method to achieve large scale assembly of nanomaterials, its integration with the flexible substrate is hindered by the difficulty to choose a good wetting system of solvent, nanomaterials and polymer substrate. The good wetting solvent may swell or even dissolve the polymer substrate. Here, to the contrary, we utilize the hydrophobic nature of a Polydimethylsiloxane (PDMS) substrate and nanomaterials (e.g., graphene) and create a “dry” dip coating method by choosing water as the solvent. Unlike the traditional dip coating method, there are two important innovations. First, the dip coating speed can reach 5.4 m/min (in comparison with 0.001-0. 2 m/min for commercial dip coater) because solvent evaporation is required. A dry substrate with assembly nanomaterial is removed from the solution because water does not wet PDMS and nanomaterials. Second, a weak sono field is applied to the solution during the dip coating process to significantly speed up the assembly process and maintain the disperse status of nanomaterials. The principle can be generalized to a wide range of nanomaterials such as carbon black, carbon nanotube, and two-dimensional materials (graphene, MoS₂, and h-BN). The film thickness can be controlled through dip coating times and solution concentration. This strategy can be further integrated with microfabrication processes such as lithography to manufacture arrays of hybrid multilayer devices with different nanomaterials. As a proof-of-concept, strain sensor, electromagnetic interference (EMI) shielding coating, and multilayer capacitor are demonstrated. This ultra-fast dry dip coating strategy paves the way towards affordable flexible devices.

We report direct fabrication of nanoparticle-polymer composite microstructures via a novel light-induced self-writing approach. Light-induced self-writing is an optical phenomenon wherein a light beam permanently inscribes ‘channels’ or ‘fibers’ across the depth of a photo-reactive medium. Self-writing is a scalable, safe and rapid technique that offers a direct single-step route for precise fabrication of polymer composite structures. Herein, we investigated the organization of nanoparticles during light-induced self-writing to assess feasibility of control over composite morphology, as core-shell and nanoparticle embedded structures have found use in various applications such as anodes for secondary batteries, as chemical sensors, and MEMS. By varying the light intensity and beam size employed in self-writing, the photopolymerization kinetics and path length of nanoparticle diffusion can be controlled, which enables directed assembly of nanoparticles and tunable morphology. In this work, a mixture of Silicon nanoparticles in an acrylate type monomer was used to self-write vertically aligned composite pillar-like structures that were then pyrolysed to obtain Carbon/Silicon structures. Use of these structures as an anode for Li-storage is examined. Scanning electron microscopy, X-ray diffraction and Raman spectroscopy were used to characterize morphology and composition of the structures, and other monomer-nanoparticle mixtures are under investigation. This work provides a novel technique to fabricate microscale architectures for energy storage applications and explores ways to control structure morphology by varying photopolymerization process parameters.

Robotic instruments are miniature robots and mechatronic systems that can manipulate small objects and precisely interact with different surfaces. In this talk, I will discuss our recent work in robotic instruments for manipulation and assembly using two techniques: 1) acoustic manipulation devices that can automatically manipulate/assembly a large amount of particles on a surface using the field gradient generated by a single acoustic transducer, both in ambient air and submerged; 2) robotic electromagnetic needle that can precisely extract and independently manipulate/assemble magnetic microparticles in the presence of neighboring magnetic particles by contact or contactless.
References:
1. Latifi, K., Wijaya, H, and Zhou, Q., “Motion of heavy particles on a submerged Chladni plate”, Physical Review Letters, Volume 122, Issue 18. 5. DOI: 10.1103/PhysRevLett.122.184301, 2019.
2. Seon, J.A., Cenev, Z., and Zhou, Q., “Automatic Non-Contact Extraction and Independent Manipulation of Magnetic Particles Using Electromagnetic Needle”, under review.
3. Cenev, Z., Zhang, H., Sariola, V., Rahikkala, A., Liu, D., Santos, H.A., and Zhou, Q., “Manipulating Superparamagnetic Microparticles with an Electromagnetic Needle”, Advanced Materials Technologies, 3 (1), 1700177, 2018.
4. Zhou, Q., Sariola, V., Latifi, K., Liimatainen, V., “Controlling the motion of multiple objects on a Chladni plate”, Nature Communications, 7, 12764, 2016.

Materials scientists have been fabricating nanomaterials with unique functionalities for decades. However, there remains a need to develop tools to assemble these materials into devices. To date, there are a variety of assembly techniques that utilize electric fields, magnetic fields, inertial forces, capillary forces, as well as robotic pick and place methods to obtain the desired structure. Among these techniques, electrophoretic deposition (EPD) has the advantage of being cost-effective, fast, and easy to perform for most material systems and substrate geometries. EPD uses electrophoresis to assemble charged nanomaterials from a suspension onto a substrate. However, EPD is currently limited by the fact that the deposition rate decreases with time, limiting the overall yield and thickness of the resulting material. Previously, this limitation has been attributed to the decrease in both the particle concentration as well as the electric field in the suspension during deposition. But, more recent studies have shown that the electrophoretic mobility of the electrostatically stabilized particles is also decreasing with time. Here, we show that decreasing particle mobility is the most critical factor in determining the EPD yield. To demonstrate this, we used alumina nanoparticles suspended in ethanol as a model system. By monitoring the changes in the pHe (pH in non-aqueous solvents) of the suspension during EPD, we observed the particle mobility of the alumina nanoparticles decreased with time. We also showed that this decrease in particle mobility significantly decreased the deposition rate. To overcome this, we developed a new suspension replenish EPD approach that allows us to maintain near constant particle mobility and particle concentration with time. This suspension replenish method results in a linear increase in the mass of the deposited film with time, overcoming the limitation of decreasing deposition rate in conventional EPD. We also saw a 6x improvement in the thickness of the film compared to the conventional EPD. This approach is applicable to a wide range of materials that can be electrostatically stabilized/charged by the adjustment of the pH of the suspension. Thereby, enabling a path to significantly improve the ability to assemble thick films of nano- and micro- materials.

Controlled macroscopic assembly of carbon nanotubes (CNTs) while preserving their excellent optoelectronic properties still remain a key challenge. Here, we investigated the molecular dynamics of single-wall carbon nanotube (SWCNT) alignment inside various viscous media under electric fields. An analytical model based on dielectrophoresis induced torque considering the viscosity and conductivity of the medium was used to obtain molecular dynamics of the SWCNTs. An alternating current electric field was applied to a liquid solution of SWCNT containing several solvents/surfactants such as DIW, DMF, DPSF, SDS, and DOC. The time required for the SWCNTs to get aligned to the applied AC electric field was calculated for different initial conditions for all the solutions. The effects of CNT length, CNT diameter, and frequency of the electric field on the SWCNT network formation were theoretically studied. Longer and thinner SWCNTs prompted to faster alignment in the SWCNT network. Furthermore, the effect of the concentration of surfactant on arrangement time was examined. Slower SWCNT alignment was observed in medium with higher viscosity. The findings of this report will be helpful for establishing an effective technique of producing large-scale aligned CNT films for diverse applications.

Displays that use combinations of microscale LEDs and ICs within each pixel to control and direct light emission have the potential to be colorful, bright, fast and power efficient. High-throughput mass transfer technologies that accurately and cost-effectively integrate large arrays of wafer-fabricated microscale LEDs onto non-native display substrates with high yield are key enablers for microLED displays. Transfer-printing with elastomer stamps is a candidate mass transfer technology for making next generation displays. A variety of displays, including displays driven by transfer-printed microICs, have been designed and fabricated using transfer-printed microscale microLEDs.

Indium Phosphide based materials are essential for realisation of lasers, modulators and detectors operating in the 1300 nm to 1600 nm range enabling widespread applications in telecommunications, data communications and sensing. The material structure for different devices are different expensive resulting in expensive components. With the emergence of silicon photonics (SiPh) as a platform for integrated photonics there is an urgent practical need to identify a scalable integration technology that can combine the unique functionality provided by InP with the routing and electronic capability enabled by SiPh in a low cost manner. We have developed a heterogeneous integration approach by using micro-transfer printing of InP devices to Si using either direct bonding or by using a thin intermediate adhesive polymer layer. The performance of transferred lasers is equal or superior to those on the native substrate after transfer due to enhanced thermal dissipation through Si. The lasers are coupled to the silicon waveguides either by butt coupling or by evanescent coupling methods.
Here, we perform an in-depth study on the adhesion properties of directly bonded InP coupons to Si. Coupons of 1 micron thick InP layer with lateral dimensions of 600 mm x 60 mm are formed by dry etching. The coupons are released from the InP substrate by selective undercut etching of an AlInAs release layer in a FeCl₃ solution while being registered and held in place on the substrate by resist tethers. The coupons are picked and transferred to silicon wafers using a PDMS stamp. The silicon has been prepared with different surface treatments to make the surface hydrophobic (HF) or hydrophilic (H₂SO₄, O₂ plasma). Additionally, coupons are transferred to gold coated silicon for potential electrical injection. Following transfer, the resist tethers are removed by an O₂ plasma. A high pick and transfer yield is obtained for coupons to all surfaces. Post-print treatments include annealing in different atmospheric environments and under vacuum. The adhesion is assessed by scratch test, acoustic microscopy and Raman spectroscopy. Experiments to perform epitaxy on the transferred coupons will be described.

Enhanced interaction of nanoscale structures with soft and curvilinear biological systems requires to reduce their large mechanical mismatches at the interface. Recently, successful development of vertically-aligned silicon nanoneedle (SiNN) arrays integrated onto flexible substrates provides a mechanically elastic interface between SiNNs and biological systems to enable their effective interactions, resulting high-efficacy delivery of biomolecules. To create more versatile platforms, transfer printing methods to integrate vertically-aligned SiNN arrays onto diverse, soft and flexible substrates necessitates further development. Here, we introduce two different transfer printing techniques to integrate vertically-aligned SiNN arrays onto diverse soft elastomer substrates for enhanced interaction between SiNNs and biological systems. First, we discuss the swelling-driven transfer printing technique which employs the volume expansion phenomena of elastomers soaked in solvents to controllably break the SiNNs for their transferring from rigid Si substrates to soft and flexible substrates. Second, we develop the double transfer printing method to realize the heterogeneous integration of the vertically-aligned SiNN arrays onto soft and flexible substrates vulnerable to dissolution in solvents. Both methods successfully transfer the targeted nanoscale structures to diverse, soft and flexible substrates that can benefit the enhancement of interaction between nanoscale devices and biological systems for wide applications.

Acknowledgement
This research was supported by the MOTIE (Ministry of Trade, Industry, and Energy) in Korea, under the Fostering Global Talents for Innovative Growth Program (P0008748, Global Human Resource Development for Innovative Design in Robot and Engineering) supervised by the Korea Institute for Advancement of Technology (KIAT)

With increasing demand for electronics System on Chip (SoC) and System in Package (SiP) design in many applications, including consumer electronics and automotive electronics, integrated inductor, which is an essential passive component, has been widely used in numerous integrated circuits (ICs) such as voltage regulators and RF circuits, and the pursuit of its heterogeneous integration into various platforms for intended use has attracted intense interest.
There are inductor in package demonstrations or even early product, however, they are either air core without magnetic core materials that render large size yet low inductance density. Other types using close to surface mount type that can incorporate magnetic core materials, that has larger thickness around 1mm, not suitable for low profile, high inductance density in -package applications, where each organic substrate layer can be much thinner in the tens of µm range. There are prior attempts to make low height profile magnetic core materials in in-package inductor applications, yet typically using ferrite that has low relative permeability, and other processing challenges such as temperature restriction.

There have been work of integrating magnetic core inductor on package substrate, without achieving high inductance increase and small foot print, typically with low inductance increase after incorporation of magnetic core materials, typically ferrite that has low permeability. We plan to use high performance magnetic core materials such as CoZrTaB to achieve high inductance density, and low cost integration process.

In this work, soft ferromagnetic core material, for e.g. amorphous Co-Zr-Ta-B, are incorporated into in-package inductors in order to scale down inductors and improve inductors performance in both inductance density and quality factor. Design of the select inductor structure are performed and optimized through electromagnetic simulation; while magnetic film screening and selection for in package application, deposition, as well as characterization of films in both DC and AC modes are carried out.

In order to maintain performance over multiyear lifetimes, many existing and emerging wearable, flexible devices require robust packaging strategies to inhibit water ingress and prevent premature failure. We report on a scalable, open-air plasma process to deposit highly transparent and dense SiO₂ thin film barriers to prevent moisture ingress. Our plasma process is suitable for direct deposition on thermally sensitive devices and is compatible with flexible, polymeric substrates as shown by 100 cycles of bending tests with a 1.25mm bending radius. Additionally, the open-air plasma process is suitable for large area devices. We show that over a field of view of over 10cm², there is a pinhole density of only 2.67 pinholes/cm². Considering that the deposition process occurs in an open-air system, where there is no regulation of particles in the deposition environment, the low pinhole density confirms the high film quality and is a key reason for its excellent barrier functionality.

We then incorporate the SiO₂ film into a multilayer barrier design comprising heterogenous organic and inorganic layers—all of which are deposited using solely open-air processes—and achieve low water vapor transmission rates on the order of 10^-3 g/m²/day at an accelerated aging condition of 38°C and 90% relative humidity (RH). When deposited directly onto devices, they retained over 80% of their initial performance for over 660 hours in a 50°C 50%RH accelerated aging environment. The ability of the multilayer barrier to enable device resistance to humid environments is crucial towards realizing longer operating lifetimes of emerging biomedical devices and flexible/wearable electronics.

Complex 3D structures enable advanced functionalities that are otherwise out of reach. However, processing issues such as mechanical stability, conformity, alignment, as well as cost, are challenging to address. The overarching physical principle of the Self-Rolled-up Membrane (S-RuM) nanotechnology is strain-driven spontaneous deformation of 2D membranes into 3D architectures. It opens up a new paradigm to achieve 3D functional hierarchical architectures without the difficulties of processing in 3D, and promises to break the constraints of size, weight, and performance of many types of devices. In this talk, we demonstrate S-RuM inductors, transformers, capacitors, and filters with > 10X smaller footprint and unique performance characteristics, through global and local strain engineering and material hierarchical integration.

Perimeter recombination in micron-sized III-V photovoltaic (PV) materials has been studied extensively for performance at high concentrations, but minimal work toward understanding and optimizing edge recombination for Si-based photovoltaics has been pursued. Here, through both modelling and microcell fabrication, we explore the performance of Si heterojunction (SHJ) solar cells with varying edge to total surface area ratios by examining performance characteristics at different micron-scale lateral dimensions (between 200 µm × 200 µm and 1000 µm x 1000 µm microns) and device thicknesses (80 - 200 microns). We explore edge passivation strategies by determining the impact of a-Si:H and oxide- and nitride-based passivation coating materials on recombination at the edges and consequently device open-circuit voltage, V_OC. Finally, we compare the impact of deep reactive ion etching (DRIE) and micro-laser cutting for solar cell dicing on edge surface quality and process scalability. Modelling results indicate that microcell device short-circuit current, J_SC, increases with device thickness (up to 35 mA/cm²) and is generally independent of lateral dimensions, while device V_OC decreases with thickness and increases with lateral dimensions, exceeding 700 mV at most geometries with high quality edge passivation. Consequently, device power conversion efficiency (PCE) is largely independent of thickness and increases with lateral dimensions, with a maximum PCE of > 20% at the largest lateral sizes. Notably, we found that edge passivation quality, rather than device geometry, dictates performance via the magnitude of edge recombination and is therefore the most important variable in this study to optimizing performance; high-quality a-Si:H was determined to be the best passivation material by virtue of minimizing edge recombination velocity (as indicated by dark current measurements) and thus maximizing V_OC. Corroboration of modelling results with that of fabricated and tested SHJ microcell devices will be discussed, along with DRIE and micro-laser cutting routes to fabrication of micron-sized SHJ PV devices and the associated high-level cost-benefit analyses. This work will contribute to understanding of micro-scale Si-based PV and provide a framework for optimizing device parameters for integration with large-scale energy-harvesting systems.

Well-controlled ‘pick-and-place’ of nanoparticles has vast potential in technological applications such as sensing, energy harvesting and transport, data storage, light composites, interfacial and functional modifiers, as well as other applications. These nanoparticles are anisotropic, including two-dimensional (2D) (e.g., nanosheets, nanolayers, nanoplatelets, nanoribbons) and one-dimensional (1D) (e.g., nanorods, nanotubes, nanowhiskers, nanofibers) nanomaterials that are chemically composed of carbon, semiconductor, and metal elements. While the organization and alignment of anisotropic nanosomic materials have been explored in many pathways, each method has its strengths and weaknesses. For example, non-directed methods (e.g., casting and drying) are simple and quite elegant but are somewhat limited in the morphologies they can access. As well, non-directed methods can be susceptible to factors such as solution concentrations and drying thermodynamics. Surface modification significantly expands the versatility of anisotropic nanomaterial assembly, but often requires challenging surface chemistry to modulate nanomaterial properties. Traditional use of prefabricated templates can control both the positional and orientational orders but requires the potentially difficult fabrication of a template. Therefore, our research aims to develop improved assembly strategies at microscales, coherently implemented into additive manufacturing (AM) infrastructure, to readily and simultaneously achieve orientational and positional control of anisotropic nanomaterials over arbitrarily large areas and volumes. The AM proposed here is an in-house designed process supported by layer-by-layer deposition of polymers with alternative and selective addition of nanoparticles. The printing and deposition mechanisms from the two-step manufacturing are compatible on the same processing platform that can greatly improve the precision and throughput.

Our proposed 3D printing mechanism will control the processing and material parameters along the working path such that each voxel of printed material itself can be a spatially varying architecture of particles giving rise to a response that differs locally from highly anisotropic to isotropic. Hence, this manufacturing enables the design and fabrication of composites with site-specific optimization of particle orientations, programmable surface morphology, and designable interfacial transitions. One-dimensional (1D) carbons, namely carbon nanofibers (CNF), will be demonstrated as an example here due to their superior anisotropic properties especially the higher conductivity or sensitivity along the fiber axis as compared to that along with the cross-section directions. We used 3D printing for surface morphologies controls and the CNFs were deposited via layer-by-layer self-assembly (LBL) techniques. The samples with different roughness and pattern features were compared for the efficiency of directed self-assembly. The solid-liquid-air interfaces were proven to be critical in determining the selective deposition of CNF for surface conductive paths and sensitivity degrees. The electricity characterizations showed a strong anisotropy along with the fiber alignment directions, showing resistivity of 0.27 kW/sqr along with the CNF band as compared to 1.21 kW/sqr along with the normal directions. This phenomenon was taken advantage of their sensing capabilities. We observed significantly improved sensitivity to strain, temperature, and chemical liquid/gas as compared to randomly distributed CNF microstructures. The high sensitivity with exposure to different volatile chemicals was obtained from our research and suggests the potential in health monitoring, human-robotics interactions, microelectronics design, and environmental sustainability.

Due to its high conductivity, copper is an important interconnect material for high density as well as large area electronic and optoelectronic products, such as ICs, TFT LCDs, and OLEDs. The reliability of the copper line is critical to these applications. In general, the copper interconnect line needs a bottom adhesion layer as well as a passivation layer. The adhesion layer effect on the line reliability has been well studied. The passivation layer is commonly formed by adding a CVD or sputter deposited dielectric or metallic layer. There are some studies on the influence of these passivation layers on the reliability of the copper line (1,2). However, there are few studies on the copper oxide passivation effect on the copper line's reliability.
Since oxide can be easily formed on the copper pattern under a well-controlled condition, it can be an effective self-aligned passivation layer. In this paper, authors report the reliability study of the copper oxide passivated copper line that is etched with the plasma-based etch process (3,4,5). The copper oxide layer was formed from the oxygen plasma method. Copper lines of various thicknesses and widths with and without the oxide passivation layer were prepared. The line breakdown phenomena under the constant current density stress condition are discussed. The composition and structure of the passivation layer before and after the high current stress condition have been examined and discussed. In summary, copper oxide is a potentially important passivation layer for the copper interconnect line in many products.

[1] M. Li, J. Q. Su and Y. Kuo, ECS Trans., 89(3), 87-92 (2019).
[2] J. Q. Su, M. Li, Y. Kuo and S. Hamaguchi, ECS Trans., 92(5), 39-46 (2019).
[3] Y. Kuo and S. Lee, Japan J. of Appl. Phys., 39(3A), L188 (2000).
[4] Y. Kuo and S. Lee, Appl. Phys. Lett., 78(7), 1002-1004 (2001).
[5] Y. Kuo, Proc. 6th Intl. Conf. Reactive Plasmas and 23rd Symp. Plasma Processing, 29-30 (2006).

In the past several decades, the US has experienced a rising concern over toxic pollution from heavy metal sources. For example, the pollution based natural disaster of Flint, Michigan garnered an abundance of media attention for its devastating impact on the lives of those in the local community. Although this negative press was rightly deserved, there are many places in America with much higher recorded heavy metal poisoning rates – three thousand to be exact. A big concern of much of this pollution comes in the form of water toxicity. Often, heavy metals find their way into water supplies for local communities which results in the indirect poisoning of citizens. Therefore, there is great need for a commercially available and portable device which would provide an easy method to test the level of poisonous elements in community water supplies.
Herein, the creation of a novel microfluidic device for the measurement of trace heavy metal ions is described. The work presented within focuses on these ions due to their health hazards and low concentration limits lending to an interesting challenge in detection.
The first effort in this project revolved around determining the form factor of a MEMS device which would act as an environmental sensor for minute concentrations of lead. Much background research into similar devices was performed before starting to conceptualize the preliminary design. Taking the equipment of the San Diego State University Clean Room facility into consideration, a rough model was presented and refined over several weeks. A final design then followed with a basic 3D model and fabrication process presented.
After much time testing and redesigning, results were generated for the measurements of trace heavy metal ions in solutions of water. Additional results were then obtained by altering various parameters within the experiment such as the analyte concentration. They were then compared to other experimental and commercial devices to determine the efficacy and practicality of the device. Finally, a study was performed to determine future work and technological extensions of the tool presented within.

For more than a decade, researchers have studied cell-to-cell signaling peptides using various analytical methods. Signaling peptides have been known to have important role in the physiological mechanisms. Currently available approaches for analyzing peptides have drawbacks such as loss of peptides and complicated steps for sample preparation. We developed an efficient platform for mass analysis of neuropeptides that simultaneously concentrates and desalts the analytes via coffee ring effect. Our approach is to use C18-Au coated patterned substrate. When peptide solution with high salt content is dropped onto the patterned C18-Au surrounded by donut-shaped hydrophilic anchors, salts are removed by outward flow generated during evaporation, while peptides are collected onto the central C18-Au region. Our approach greatly simplifies the sample preparation for matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). We also found that our patterns can be regenerated and used 5 times or more without any performance degradation. We have developed a simple kinetic model that describes the correlation between the evaporation rate and the surface coverage of peptides. Moreover, the performance could be dramatically increased by concentrating absorbed peptides into a crystal using PDMS contact printing. Applying these drop-dry methods to MALDI-MS sample preparation potentially enables high-throughput analyses, including time-lapse mass analysis. Our approach is easily accessible to researchers who do not have special expertise in mass spectrometry, so it is possible to contribute to bringing new ideas to neuropeptide analysis.

Since the nano world can be observed with SEM and SPM, nanomaterials have been actively developed. Among them, nanoparticles are often used for both metals and non-metals, such as cosmetics, home appliances / electrical electronic products, paints / inks, etc. In general, nanoparticles are produced by a vaper phase method or a liquid phase method. However, it is still difficult to produce nanoparticles with a desired shape such as raspberry particles. For example, the prototype raspberry particles are on the order of the micro level at the minimum.
In this study, the surface of two kinds of large and small nano level silica nanoparticles were coated with different reactive chemically adsorbed monomolecular film. Thereafter, the particles were bonded to each other by the interaction of the functional groups at the ends of the chemically adsorbed monomolecular film on the surface of each particle to produce nano sized raspberry particles. As an application example of nano raspberry particles, the nano raspberry particles were coated on the surface of a glass substrate to form a nano fractal surface structure. Furthermore, a light transmitting super water repellent glass substrates could be produced by forming a fluorocarbon based chemical adsorbed monomolecular film on the surface without damaging the surface of the fractal structure.

Self-Assembly of amphiphilic block copolymers in selective solvents can yield a variety of organized structures at the nanoscale of different sizes, shapes, and interfacial characteristics. Such rich structural polymorphism has been tuned and subsequently utilized in this work to controllably prepare a variety of colloidal systems of nanoparticles by templated growth in the self-assembled nanodomains. Nanomaterials spanning compound semiconductors with 0-, 1-, and 2-D morphologies to metallic nanostructures grown in poly(ethylene oxide)-b-poly(propylene oxide)-b-poly(ethylene oxide) assemblies will be presented. The metallic systems in particular were applied as antibacterial agents exhibiting enhanced and sustainable performance due to the block copolymer colloidal configuration.

In dense urban environments, with limited rooftop space, communities must look towards new photovoltaic (PV) technologies that can be integrable into the existing building envelope. Luminescent solar concentrators (LSCs), which sustain performance in both direct and diffuse lighting conditions and boast tunable transparencies, present an attractive option for building integrated (BI) PV. In this talk, we report performance optimization and characterization for varied illumination conditions of Si heterojunction (SHJ) microcells embedded within an LSC for BIPV applications. For this design, we propose a grid-like array of SHJ microcells, beneath a poly-lauryl methacrylate (PLMA) waveguide that has embedded InAs/InP/ZnSe quantum dots as the luminophore. Given the proposed architecture, the microcell can be illuminated from the top surface of the cell as well as each of its device edges; this illumination results from each direct sunlight and from quantum dot photoluminescence (PL) within the LSC waveguide, respectively.

In this talk, we compare SHJ microcell performance when illuminated on each i) the top surface, ii) one device edge, and iii) all four device edges simultaneously. We present both simulated and experimental measurements to corroborate our results around photonic interactions at each surface, and preform the following characterizations and measurements: light current-voltage (LIV), dark current-voltage, photocurrent mapping, external quantum efficiency (EQE), and internal quantum efficiency (IQE). By varying the dimensions (width and length) of the SHJ micro cell (from 400um to 800um), we further report how cell performance for each surface illumination (top vs. edge) varies for different sized devices. Results of these measurements show that Jsc increases when illuminating the microcell on the edge (in comparison to top surface illumination) due to an increased absorption depth. However, Voc drops for edge illumination vs. top illumination, due to an additional darkness factor and increased recombination. The mechanisms and photonics behind the contrasting microcell performance at each of these surfaces will be discussed during this talk. By understanding microcell performance at each interface, we are able to better design a SHJ/LSC integrated device, with optimized luminophore PL to match the edge illumination performance, for BIPV applications.

Fifty years of microelectronics scaling have opened up new opportunities for building complex robotic systems that are microscopic in size. Electronic, magnetic, and optical systems now offer an unprecedented combination of complexity, small size, and low cost, and could readily be appropriated to form the intelligent core robots that are too small to be resolved by unaided human vision. But one major roadblock exists: there is no micron-scale actuator system that seamlessly integrates with semiconductor processing and responds to standard electronic control signals. In this talk, I show how to overcome this barrier, enabling a new generation of microscale robots that utilize on-board, integrated silicon electronics. First, I will present a new class of voltage-controllable electrochemical actuators that operate at low voltages (200 mV), low power (10 nW), and are completely compatible with silicon processing. Next, I'll present a full process flow that uses lithographic fabrication and release techniques to build sub-hundred micrometer robots that walk. Every step in this process is performed massively in parallel, producing over one million robots per 4-inch wafer. I'll show that in spite of their small size, microscopic robots are remarkably robust: they can survive harsh chemistries, large changes in temperature, and can even be injected through hypodermic needles. Finally, I will close by showing our first steps towards transforming sophisticated CMOS circuitry into programmable, autonomous robots too small to be seen by the naked eye.

As Moore’s law for digital computation reaches its inevitable limits, future advances in system level integration will lead to the next generation of electronics. Approaches such as heterogeneous integration will enable small and high-performance micro-systems (optical and electronic integration or System-In-Package) necessary for the growing consumer areas such as wearables, the Internet of Things, and Augmented Reality. Beyond electronics, tools and techniques for heterogenous micro-assembly can lead towards new smart materials (i.e. defect-free constructed materials, digitally tailored electro-mechanical active materials with micro-scale properties and macro-scale performance, etc.) and the micro-assembly of biological and soft components.

While there are many research approaches towards tackling heterogenous micro-assembly (i.e. self-assembly, transfer printing, etc.), in this work, we will discuss our approach of directed robotic assembly using micro/milli robots. Directed micro/milli-robotic assembly has some unique advantages over other methods, such as being relative task agnostic and generalizable to various assembly problems. Furthermore, unlike many batch processes that are open-loop, directed micro-robotic assembly can be closed-loop by integration of metrology and material validation during the assembly process. Thereby swarms of micro/milli-robots can parallelize not only micro-assembly but also micro-metrology, enabling a method of overcoming statistical and thermodynamic limitations of conventional approaches.

Inspired by massive swarms of ants or termites assembling macro-scale objects from micro-scale parts, we utilize arrays of magnetic actuators in conjunction with diamagnetic graphite to create a magnetic collective of levitated micro/milli sized magnetic robots with nm level motion precision for 3D micro-assembly. By leveraging diamagnetic levitation, we can reduce factors such as hysteresis and stiction to enable high-precision and repeatable motion of micro/milli-robots. These passively levitated, but electromagnetically steerable micro/milli-robotic platforms can serve many roles: (1) compact high-precision motion platform for the manipulation and delivery of micro-components, (2) high precision sensors, and (3) motion platform for other micro-scaled MEMS sensors and actuators. We will discuss multiple independently controlled robots (mm-sized) with dexterous behavior for assembling carbon fiber trusses and performing micro-assembly and micro-manipulation of 3d microspheres and silicon microparts, as well as some of the control issues for coordinating parallel robotic swarms for micro-assembly. In addition, by using robots cooperatively, we also demonstrate how teams of micro/milli robots can achieve new functionality and enable extremely high densities of independent micro-manipulation as compared to conventional standard robotic solutions.

Achieving robotics at the micron and nano scale enables building reconfigurable machines and metamaterials that will create an unprecedented impact on fields such as healthcare, manufacturing, and smart materials. These applications demand scale-free and material-agnostic systems that can morph into arbitrary but programmable target configurations with precision as controlled by external agents such as heat, light, electricity and chemical environment. Such diverse actuation can be accomplished using active colloids—propellers that harness environmental energy source—that can act as motors, triggered and powered by these stimuli. However, unlike macro-scale motors, self-propelled colloids lack memory and ability to compute required by the conventional swarm-based algorithms to program the organization of motors into prescribed configurations. Here we present a design strategy that is both programmable and compatible with the sub-micron scale physics. Our strategy is to connect these motors such that the interactions among the motors are purely mechanical. The sequence of forces on these motors causes the system to morph into a unique configuration. We show how to reverse engineer the system’s design given a target configuration using kilobot (centimeter-scale robot) experiments and Brownian dynamics simulations. We also discuss methods of navigation and reconfiguration of the system. We demonstrate designing complex behavior such as a dynamic gripper, a dynamic Pacmac, and the letter M.

Unique properties of one-dimensional assemblies of particles have attracted great attention during the past decades, particularly with respect to the potential for anisotropic magnetism. We have demonstrated directed assembly of ferromagnetic particles of Gd₅Si₄ into one-dimensional chains in a magnetic field-assisted inkjet printing process. The chains exhibit anisotropic magnetic properties, and are much easier to manufacture than, for example, corresponding single crystals. We have shown that three characteristic times, namely, the critical time for solvent imbibition into the porous substrate (t_im), the time it takes for particles to form chains in the presence of the magnetic field (t_ch), and the time in which the particles reach the substrate in the direction normal to the substrate (t_pz), control the assembly process. The characteristic time ratios of t_pz/t_ch>>1 and t_pz/t_im<<1, favor particle chaining and uniform particle deposition, respectively. Such an ability to control magnetic particle deposition and assembly through inkjet printing is a promising approach for many engineering applications.

Work at Ames Laboratory is supported by the Division of Materials Science and Engineering of the Office of Basic Energy Sciences, Office of Science of U.S. Department of Energy (DOE). Ames Laboratory is operated for the U.S. DOE by Iowa State University of Science and Technology under Contract No. DE-AC02-07CH11358. Work at Virginia Commonwealth University (VCU) is partially supported by National Science Foundation (1357565 and 1726617) and VCU startup funds.

3D chip stacking approaches with TSV (thorough-silicon vias) are mainly classified into three categories: chip-to-chip, wafer-to-wafer, and chip-to-wafer 3D integration. The chip-to-chip technologies are widely used for traditional microelectronic packaging in which KGDs (known good dies) are assembled using one-by-one pick-and-place techniques. Although the assembly yield is high due to the use of KGDs, the assembly throughput is not sufficiently high, compared with wafer-to-wafer approaches. However, wafer-to-wafer technologies have a serious issue in total production yield that is exponentially decreased with the number of stacked layers because the failure dies cannot be removed from the wafers to be stacked. Therefore, chip-to-wafer 3D integration is thought to be a promising candidate to satisfy both the throughput and yield issues. However, chip-to-wafer 3D integration potentially has a trade-off between assembly throughput and alignment accuracies in the conventional robotic pick-and-place techniques. In order to address the problems, we have proposed and developed multichip-to-wafer 3D integration using liquid surface tension. From the technical point of view of chip-to-wafer bonding, we talk about the basic concept and recent studies of the self-assembly based 3D integration. These approaches using surface tension-driven multichip assembly are divided into two methodologies: one is non-transfer stacking without carrier wafers and the other one is transfer stacking with them. These carrier wafers are typically used for temporary wafer bonding in standard 3D integration processes. Here, we also introduce temporary multichip bonding technologies applicable to 3D integration with multichip transfer stacking in which an inorganic temporary adhesive and a debonding layer sensitive to near UV lasers are utilized. In this presentation, we introduce our multichip-to-wafer 3D integration studies with non-transfer and transfer stacking particularly focused on solder microbump bonding, oxide-oxide direct bonding, and hybrid bonding through a new material of Cu nano-pillars. Finally, I will show the massively parallel self-assembly works with tiny dies with the side of less than 100 μm for micro-LED display application.

It is often desirable to assemble conjugated polymer molecules into highly ordered macroscopic structures to fully utilize their potentials in a variety of organo-electronic applications. However, due to their intrinsic flexibility, conjugated polymer chains tend to entangle and form disordered bundles in solution. Here we show that centimeter-long conjugated polymer (poly(3-hexylthiophene) (P3HT)) fibers can be assembled under an external DC electric field. In general, assembly under external stimuli such as magnetic or electric fields provides a way to arrange samples without using template materials, leading to a relatively simple procedure. The electric field is chosen here because of its low cost and tunability of field strengths and orientation during the assembly process. In this work P3HT molecules are pre-aggregated into high aspect ratio nanowires, then these nanowires can be further aligned and bundled into macroscopic fibers in solution under the applications of external electric field. To fully understand the assembly process, its dependences on nanowire microstructure types, the DC electric field strength, and solvent polarity are investigated by characterization methods including UV-Vis, AFM, and GIWAXS. It is proposed that the electric field-induced dipole in the nanowires aligns the nanowires along the electric field direction, then the dipole-dipole interactions will lead to further assembly of the nanowires into larger bundles. The macroscopic fibers will retain its morphology in solution even after the electric field is removed. Furthermore, curved macroscopic fibers can also be prepared under curved electric field lines. This hierarchy method provides a general way to assemble flexible molecules into macroscopic fibers by pre-forming high aspect ratio secondary nanostructures, then align and bundle them under external fields to retain the anisotropic properties. We hope this method can be extended to other flexible molecules that are difficult to be directly fabricated into macroscopic structures under electric fields.

Studies on the manufacture of displays, smartphones, wearable devices, etc. using MicroLED have become active in recent years.
Manufacturing process of MicroLED requires a fast pick and place assembly as there are so many MicroLED chips to be placed onto its substrate and wafer.
Such Placing shall be accurate and reliable and ideally repairable of each individual MicroLED chip as it takes much time to complete sizable panel.
However, it has been very difficult to transfer a large number of microchips with high accuracy.
In order to achieve solution against such transfer requirement, a large amount of chip transfer technology, inspection technology for detecting defects on microchips, and repair technology are necessary.
We have been successful in both the semiconductor and FPD industries as a business area based on the concept of E & M (engineering and manufacturing).
To date, we have made a series of processes necessary to manufacture MicroLED displays (to provide a total solution)，We have developed the following three types of equipment that apply our core technology such as inspection, bonding, laser process for semiconductor and so on.
1. Non-contact inspection 2. Laser trimming only NG chips based on inspection results
3. MicroLED bonder for high productivity
Especially MicroLED bonder of Toray Engineering Co., Ltd. is enable to place such precise MicroLED chips very fast.
In this presentation we will introduce our solution technology for MicroLED manufacturing method.

Understanding the mechanisms of material structures forming processes enables the design and fabrication of structures with improved or tailored properties. The objective of our work is to investigate the mechanisms underlying materials synthesis and processing, including crystal growth, particle self-assembly, and phase transformation, etc., and thus to tune the structures and functions of materials.
For example, particle aggregation, such as oriented attachment, is accepted as a common pathway and leads to well-organized structures and twin interfaces that possess unique or improved properties. However, many fundamental aspects remain unknown, such as the interplay of structures and factors that control them. Here we use in situ transmission electron microscopy and atomic force microscopy to directly observe process of materials synthesis and processing, such as particle assembly and phase transformation. Integrating theoretical calculations and simulations, we revealed the orientation-dependent attachments of particles, interfacial structures, and the factors that affect them, such as temperature, pH, ionic strength, particle shapes, and solvents, etc. In addition, we also demonstrated that defects (edge dislocation), induced via oriented attachments, improving photoactivity of TiO₂. Phase transformation and its role in materials functions are also investigated. The specific systems that we investigate are metal (Ag and Au), oxides (TiO₂), and minerals (mica), etc. The work provides a deep understanding of mechanisms of materials synthesis and processing and thus provides a guide to design and fabricate the morphologies and structures, which are closely related to material properties and functions.

Concentrated suspensions of colloidal nanorods can exhibit liquid crystalline (LC) behavior: spontaneous and reversible assembly of nanorods into orientationally ordered nematic phase. Considering to Onsager’s theory predicting the rod concentration threshold (Φ) where the isotropic-to-nematic phase transition occurs, the increased effective radius of charge-stabilized rods can significantly reduce Φ compared to the same rods that are stabilized sterically[i] . This has allowed to prepare relatively dilute colloidal LCs (Φ < 5 vol%) with strongly charged LaPO₄ nanorods in polar media[ii]. However, we recently observed a contradictory phenomenon where the ligand stabilized LaPO₄ nanorods in apolar media self-assemble into a nematic phase at even lower Φ near 0.5 vol% while the theoretically estimated Φ is 17%. Moreover, Φ varies with the concentration of additional surfactants in a similar way as in polar media when increasing a salt concentration leads to surface charge screening. This observation suggests that the ligand-functionalized surface of nanorods may still be partially charged and thus contributes to a largely extended Debye length (effective radius of rods) when suspended in apolar media with low dielectric constants. Such colloidal systems, both sterically and electrostatically stabilized, have an intriguing potential for exquisitely controlling the dynamics of charged colloids avoiding the common problem of aggregation. We present the LC behavior of the dilute nanorod suspension, their control under external electric field, and the impact of the amphiphilic surfactants on phenomena associated to electrostatic surface charging in apolar media.

[i] Onsager et al. Ann. N. Y. Acad. Sci, 51, 627-659 (1949)
[ii] Kim, J. et al. Adv. Funct. Mater, 22, 4949-4956 (2012)

Flexible origami structures can mimic the complicated motions of small creatures that are otherwise difficult to be achieved by rigid robots with limited degree of motion freedom. However, actuating origami structures in a compact and self-contained way has been a critical challenge. Here, we demonstrate a versatile approach of actuating origami micro-robots, by printing self-folding creases on compliant polymer membrane substrates. The self-folding creases are made of a type of stimuli-responsive transition metal hydroxides/oxides that can undergo large actuation under electrochemical or light stimulations. Direct printing of the stimuli-responsive material via a microfluidic electrochemical writing method at the locations of the creases enables the creases to self-fold independently “on demand”, with folding curvatures exceeding 1 mm^-1 under low-intensity visible-light stimulation in ambient conditions, or low-potential electrochemical stimulation in electrolytic environments, with response time as fast as in seconds. Based on the high performance of such active creases, complex miniaturized origami designs powered by hinges activated in an independently controllable way are demonstrated, including self-folding Miura and a full micro-robotic hand with independently programmable finger joints. These results prove a new, versatile paradigm for micro-robotics, where a transferrable approach is applicable to design and fabricate a wide variety of customizable micro-robots with compact construction and complex motions using different stimuli-responsive ceramic-based materials.

Grain boundary defects-topological defects around grain boundaries (GB)-play an essential role in the optoelectronic, magnetic, mechanical, and chemical properties of the polycrystalline materials. When the grain size of the materials decreases to the nanometer scale, the grain boundary defects predominantly determine those properties, opening up a variety of new applications. Although researchers have sought ways to study the contribution of specific grain boundary defects to the properties, non-uniformity in grain size and shape and random misorientation between the grains make these tasks challenging. This presentation will first discuss controlling 3D heteroepitaxy of colloidal faceted nanocrystals to grow uniform multiple grains organized into a well-defined geometry, which gives a new opportunity for overcoming such limitations from the non-uniformity and randomness. Within the resulting nanocrystals, we will see that the 3D patterned strain field exists in the form of disclinations and dislocations. We will then discuss how we can apply the 3D hetroepitaxy to investigate the effect of microscopic grain geometry on the macroscopic properties. For example, we can obtain a correlation between the GB defects and the catalytic property, because the patterned strain creates exact GB structure in a periodic fashion that serves as the active sites for catalysis. Lastly, we will highlight that the epitaxy-based design principles can significantly extend the library of grain structures in the nanocrystalline materials.

Cadmium sulfide (CdS) has been highlighted due to the excellent photoelectrocatalytic properties and activities. Hetero-integrated CdS nanoscale structures onto diverse substrates can bring more enhanced performances in wide photocatalytic applications. Herein, we demonstrate a facile strategy to synthesize CdS nanorods on three-dimensional conductive carbon structures with the enhanced photoelectrocatalytic performance by using hydrothermal methods. The hetero-integrated CdS nanorods with high uniformity on highly conductive carbon substrates not only exhibit the enhanced photocatalytic efficiency in reduction of organic pollutant, but also perform the excellent photoelectrochemical activity and photostability, compared to control samples. The superior performance of the hetero-integrated CdS nanorods on carbon substrates is attributed to the improved charge separation and transport of photo-generated electrons in highly crystalline CdS nanorods and the effective photo-generated electron transfer to carbon substrates from CdS nanorods. Finally, we demonstrate the hetero-integrated CdS nanorods on three-dimensional carbon supports with a variety of shapes by using the developed method. Our strategy to integrate CdS nanoscale structures on diverse substrates can be utilized to achieve high performance photoelectrocatalytic systems in wide energy and environmental applications.

Acknowledgement
This research is supported by Basic Science Research Program (NRF-2018R1C1B6007938) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT of Korea. This study was also supported by the Intelligent Synthetic Biology Center of Global Frontier Project funded by the Ministry of Science and ICT of Korea (NRF2012M3A6A8054889)

Silicon nanocrystals (SiNCs) have many applications in electronic devices due to their efficient and tunable optoelectronic properties. In the many methods for creating SiNCs, they are nearly always either embedded in an inorganic matrix or supported by a substrate. While this has not hindered their use in technologies such as light-emitting devices (LEDs) and solar photovoltaics, it also raises questions about SiNC properties in standalone layers. It is well-known that the local environment of nanocrystals can influence their optical properties, and many of the methods for testing the mechanical behavior of thin layers of SiNCs rely on substrate-supported techniques. Here we present our work on fabrication of freestanding layers of SiNCs, and their optical and mechanical behavior.

SiNCs were synthesized in a nonthermal radiofrequency plasma reactor using silane, argon, and hydrogen gases. Transmission electron microscopy (TEM) and x-ray diffraction (XRD) measurements indicate that the SiNCs are crystalline with an average diameter of 4-5 nm. We created freestanding layers by using inertial impaction of SiNCs directly out of the reactor onto sacrificial layers, such as NaCl and sucrose. We used thermal evaporation to deposit thin layers of NaCl onto substrates such as glass, silicon wafer, and polydimethylsiloxane (PDMS). We also used block-like substrates entirely from NaCl and sucrose, as well as other surface coatings of sucrose, as our sacrificial bolsters for the SiNCs. We deposited SiNC layers of varying thicknesses onto these substrates using the inertial impaction technique. We then liberated the SiNC layers via dissolution in water at controlled temperature. The SiNCs, which are hydrophobic, rose to the surface of the water where we collected them. Our ongoing experiments include testing the SiNC layers' photoluminescence and mechanical properties without the support of a substrate beneath, helping to elucidate their properties as a single system rather than as a layer supported by a bolster.

Ni-rich layered LiNi_1-x-yMn_xCo_yO₂ (NMC, x+y<0.5) oxides have been demonstrated to be dominant cathode materials for high-energy lithium ion batteries. However, NMC cathode materials with high Ni contents usually show unsatisfied capacity decay and voltage fading cycled at high voltages, owing to aggravated side reactions and electrochemical irreversibility during prolonged lithiation/delithiation cycles. Here, we report the Mn-rich Li_0.65Mn_0.59Ni_0.12Co_0.13O_δ (marked as LMNCO) nanocomposite that is integrated of layered Li₂MnO₃ and spinel LiMn_1.5Ni_0.5O₄-type phases as a desired shell for improving high-voltage cycling stability of the Ni-rich LiNi_0.8Mn_0.1Co_0.1O₂ (marked as NMC811) cathode material. The core-shell NMC811@x% LMNCO nanostructure has been fabricated in satisfied structural conformality by using an initial sonofragmentation, followed by the solvent evaporation-induced self-assembly (EISA) and post-annealing processes. The optimized NMC811@5% LMNCO cathode material can deliver an initial discharge capacity of 150.0 mAh g^-1 at 5 C (1 C=200 mA g^-1) in a voltage range of 2.7-4.6 V vs. Li⁺/Li with 83.4 % retention up to 500 cycles, significantly superior to 75.6 % of the bare NMC811 material. The Mn-rich shell also enables to effectively stabilize Ni-rich cathode materials for long-term cycles in such a high voltage range at 55 °C. In addition, the in-situ X-ray diffraction (XRD) has been carried out to probe structural evaluations of the Ni-rich NMC811 cathode material during charge/discharge processes. The operando XRD results demonstrate undesired structural irreversibility of the NMC811 cathode material in the first cycle. Enhanced structural stability for continuous phase transformations in next electrochemical cycles can be implemented via a self-assembly of the Li-rich Li₂MnO₃ coating. The optimized NMC811@LMO-3% cathode can retain initial discharge capacities of 92.9 % at 0.1 C after 100 cycles and 80.4 % at 1 C after 500 cycles, significantly superior to 73.8 % and 61.3 % of the bare NMC811 cathode. This work sheds lights on structural evaluations of Ni-rich cathode materials during lithiation/delithiation processes, and further offers a synthetic prototype for the conformal core-shell-structured fabrications, which could be adopted for the surface modification of various functional materials to achieve enhanced performance in device applications.

It is of great scientific interest to integrate nanoscale materials onto non-native substrates to synthesize structured architectures with novel functionalities that cannot be achieved by traditional means. To achieve this, we explore the crystallization of nanoparticles on periodically nanostructured substrates.
For the structured substrate we utilize a periodic array of nanocups of pitch 750 nm formed on a poly(dimethylsiloxane) (PDMS) substrate through replication from a polymer mold [1]. The novel feature of the nanocup array is the existence of surface tribo-charge strongly localized at the rims of the nanocups in an annular ring pattern, which was confirmed through Kelvin probe force microscopy of the surface electric potential [1,2]. Dielectric nanoparticles may be polarized by the strong electric fields in the vicinity of the tribo-charge. We expect that such dielectric nanoparticles in a colloidal suspension may be preferentially attracted to the charged regions and driven into nucleation, forming unusual structural motifs, including ring-like patterns, that could not otherwise be produced in conventional nucleation experiments. We hypothesize that the tribo-charge provides a new control variable to fabricate novel colloidal structures within the existing nanostructure.
To test this hypothesis, a PDMS nanostructured array with annular tribo-charge was immersed in a colloidal alumina suspension (5% alumina by volume) for ten minutes and then removed from the suspension. The mean particle size of the suspension was ~50 nm. The array was slowly dried, allowing crystallization. The dried PDMS array was characterized by scanning electron microscopy (SEM). We observed a high density of colloidal nanoparticles crystallized within each nanocup. It is very probable that the charged template has attracted the colloidal nanoparticles to the nanocup wall in the initial phase of growth. This is followed by agglomeration of the nanoparticles within each nanocup. Certain regions of the nanostructure reveal agglomerates of alumina nanoparticles with nearly spherical agglomeration shapes, neatly filling the entire nanocups. There is negligible agglomeration on the flat top surface between the nanocups.
It will be necessary to tailor the microscale assembly conditions that nucleate the colloids or loading material in the first phase, but prevent further agglomeration within the nanocups. This may be achieved by i) using a lower density of nanoparticles in the solution and ii) increasing the concentration of dispersants to reduce agglomeration. kinetics of crystallization will be discussed. These experiments will create a pathway for tailored growth of colloidal crystals and other drug coatings within these charged templates.
These can be biomedically utilized for loading of nano-arrays with drugs since it leads to complete filling of the nanocups with negligible coating left on the flat surface. Thus drug elution would require dissolution of nanoparticles from the nanocups, starting with outer nanoparticles in the outer shell of the agglomerated clusters. Microstructures created by bottom-up technologies can serve as 2-dimensional photonic crystals where strong diffractive effects can be present.
[1] Qiang Li, Akshit Peer, I. Cho, Rana Biswas, Jaeyoun Kim, Nature Communications 9, 974 (2018).
[2] Qiang Li, Akshit Peer, I. Cho, Rana Biswas, Jaeyoun Kim, Back Scatter A nano-volcano array, Physics Today, 71 (5), 72 (May 2018).

Directed assembly at the colloidal scale remains challenging despite advances in using optical forces to manipulate nanoparticles. Optical holography allows for precise positioning of components, but does not scale well to thousands or millions of particles. However in dense colloidal systems particle positions are highly correlated, and so not every particle need be controlled in order to enact reconfigurations of the dense assembly. In this work we use simulation to study how a small number of particles embedded within a colloidal monolayer can exert forces that lead to long-range permanent deformation of the monolayer edges. By framing the problem in terms of the creation and migration of mobile dislocation defects, we uncover design principles for the embedded cluster shapes that result in the accumulation of large shear deformations on lattice planes. We show that arbitrary control of shear deformation localized to a single plane is sufficient to carry out a variety of useful reshaping operations on crystalline colloidal monolayers. This work provides a method to increase the length scale of control achieved with optical forces without imposing additional complexity on the optical system required.

The bottom-up approaches recently become quite attractive for nano-scale fabrication. For example, block copolymers, nano-colloidal particles, and anodized aluminum are known to form nano-scale patterns. In addition, these bottom-up self-assembly can be combined with conventional top-down lithography resulting large scale periodic pattern can be fabricated. Several applications to nano/micro-electronics have been reported. Such as patterned magnetic media for a hard disk drive (HDD), light extraction structures of light-emitting diodes (LEDs), organic light-emitting diodes (OLEDs), anti-reflection (AR) structures for displays, metal wire-grid polarizers (WGP), and transparent conductive film (TCF) using nano-meshed metal thin film were demonstrated by these manners. The guided, aligned, or directed self-assembly (DSA) is a key technique, that the larger guides are made by conventional top-down methods then bottom-up self-assembled structures are formed along the guides. The challenges are continuing to fabricate the semiconductor memories. We would explain these potential applications fabricated by self-assembly.

Conventional manufacturing is able to assemble products composed of many parts of the size approximately 1mm and above. At the nanometer scale, techniques such as deposition, lithography, and self-assembly are used to make a myriad of heterogeneous products. However, at the mesoscale (1-1000μm) these techniques don’t overlap to produce structures at scale. There is a need for robust and scalable technologies which are able to join structures which can be developed on the micro-scale into large area, heterogeneous assemblies. To address this, we are developing a printing platform which is able to use micro-objects as ink—microLEDs, sensor chips, integrated circuits, micro- and nanoparticles. These objects are dispersed in a dielectric fluid and can be printed similar to a xerographic process, but maintain single object precision and orientation. To do this, an optically-addressed electrostatic actuator array manipulates the particles to move to target locations. From there, an integrated transfer system removes the particles from the assembly platform to be placed on a final substrate while maintaining their relative positioning. This transfer system can either be batch with adhesive transfer, or continuous with a non-contact electrostatic belt. This talk will describe the process in which we are able to rapidly sort, manipulate, orient, assemble, and transfer particles in parallel over a 1-inch wide area. We have demonstrated 1-µm registration, assembly of hundreds of particles in seconds, and roll-to-roll like transfer. With the use of computer vision to aid the assembly process, the technology has the potential to produce highly ordered, custom-patterned, heterogeneous assemblies accurately over a large area much faster than traditional manufacturing techniques while still maintaining fidelity.

Direct emission displays made with inorganic µLEDs can potentially challenge LCD and OLED displays with better performance, efficiency and lifetime. Markets where microLED displays are likely to first appear are large information displays requiring high brightness, home theater applications benefitting from increased dynamic range and automotive displays in which high contrast and reliability are necessary. These displays are made possible by the integration of two mainstream technologies, namely GaN LED developed for general lighting and the large area active matrix backplanes developed for LCD displays.
The unique manufacturing challenge for µLED displays is development of a technology to position and connect millions of small devices with zero defects and low cost. Conventional pick and place tools are capable of positioning µLEDs but the throughput of serial assembly is inadequate while mass-transfer methods have difficulty changing the pixel pitch. To address these challenges eLux has developed a massively parallel fluidic assembly technology which positions each µLED by capture of the device in a well structure that also contains the connecting electrodes.
Fluidic assembly technology uses simple low-cost equipment to achieve assembly rates up to 50 million µLEDs per hour and it is anticipated that production equipment could assemble an 8K display with 100 million LEDs in a few minutes. The technology is suitable for µLEDs from 10 to 200 µm diameter offering flexibility to make a wide variety of displays with resolutions from 400 to 10 ppi. Similarly, the microLED emitter area can be changed independently of the assembly technique to increase brightness and it is possible to produce displays over 5,000 cd/m². Fluidic assembly applies relatively low force on the device so brittle materials such as red µLEDs fabricated from AlGaInP can be assembled in the same way as blue and green emitting GaN µLEDs.
A fully integrated manufacturing system for µLED fluidic assembly will ensure that the assembled µLEDs have acceptable performance with special emphasis on eliminating shorted or high resistance devices. Testing and binning used in the general lighting industry are not practical at the micron scale, so wafers are mapped by cathodoluminescence to identify defective µLEDs. Based on the defect map, a selective harvest technique diverts bad µLEDs from the suspension so only known good die are used for assembly. Fluidic assembly relies on an excess of µLEDs to ensure sufficient capture attempts for complete assembly, so it is also necessary to capture, filter and recycle unassembled µLEDs.
With a suitable choice of complementary characteristic for the device and the trap site, it is possible to use fluidic assembly techniques to position nanowires or nanoparticles as well as µLEDs. The presentation will give an overview of the integration issues with some suggestions for other possible applications and approaches.

Studies on the manufacture of displays, smartphones, wearable devices, etc. using MicroLED have become active in recent years.
Manufacturing process of MicroLED requires a fast pick and place assembly as there are so many MicroLED chips to be placed onto its substrate and wafer.
Such Placing shall be accurate and reliable and ideally repairable of each individual MicroLED chip as it takes much time to complete sizable panel.
However, it has been very difficult to transfer a large number of microchips with high accuracy.
In order to achieve solution against such transfer requirement, a large amount of chip transfer technology, inspection technology for detecting defects on microchips, and repair technology are necessary.
We have been successful in both the semiconductor and FPD industries as a business area based on the concept of E & M (engineering and manufacturing).
To date, we have made a series of processes necessary to manufacture MicroLED displays (to provide a total solution)，We have developed the following three types of equipment that apply our core technology such as inspection, bonding, laser process for semiconductor and so on.
1. Non-contact inspection 2. Laser trimming only NG chips based on inspection results
3. MicroLED bonder for high productivity
Especially MicroLED bonder of Toray Engineering Co., Ltd. is enable to place such precise MicroLED chips very fast.
In this presentation we will introduce our solution technology for MicroLED manufacturing method.

Enhanced interaction of nanoscale structures with soft and curvilinear biological systems requires to reduce their large mechanical mismatches at the interface. Recently, successful development of vertically-aligned silicon nanoneedle (SiNN) arrays integrated onto flexible substrates provides a mechanically elastic interface between SiNNs and biological systems to enable their effective interactions, resulting high-efficacy delivery of biomolecules. To create more versatile platforms, transfer printing methods to integrate vertically-aligned SiNN arrays onto diverse, soft and flexible substrates necessitates further development. Here, we introduce two different transfer printing techniques to integrate vertically-aligned SiNN arrays onto diverse soft elastomer substrates for enhanced interaction between SiNNs and biological systems. First, we discuss the swelling-driven transfer printing technique which employs the volume expansion phenomena of elastomers soaked in solvents to controllably break the SiNNs for their transferring from rigid Si substrates to soft and flexible substrates. Second, we develop the double transfer printing method to realize the heterogeneous integration of the vertically-aligned SiNN arrays onto soft and flexible substrates vulnerable to dissolution in solvents. Both methods successfully transfer the targeted nanoscale structures to diverse, soft and flexible substrates that can benefit the enhancement of interaction between nanoscale devices and biological systems for wide applications.

Acknowledgement
This research was supported by the MOTIE (Ministry of Trade, Industry, and Energy) in Korea, under the Fostering Global Talents for Innovative Growth Program (P0008748, Global Human Resource Development for Innovative Design in Robot and Engineering) supervised by the Korea Institute for Advancement of Technology (KIAT)

Fifty years of microelectronics scaling have opened up new opportunities for building complex robotic systems that are microscopic in size. Electronic, magnetic, and optical systems now offer an unprecedented combination of complexity, small size, and low cost, and could readily be appropriated to form the intelligent core robots that are too small to be resolved by unaided human vision. But one major roadblock exists: there is no micron-scale actuator system that seamlessly integrates with semiconductor processing and responds to standard electronic control signals. In this talk, I show how to overcome this barrier, enabling a new generation of microscale robots that utilize on-board, integrated silicon electronics. First, I will present a new class of voltage-controllable electrochemical actuators that operate at low voltages (200 mV), low power (10 nW), and are completely compatible with silicon processing. Next, I'll present a full process flow that uses lithographic fabrication and release techniques to build sub-hundred micrometer robots that walk. Every step in this process is performed massively in parallel, producing over one million robots per 4-inch wafer. I'll show that in spite of their small size, microscopic robots are remarkably robust: they can survive harsh chemistries, large changes in temperature, and can even be injected through hypodermic needles. Finally, I will close by showing our first steps towards transforming sophisticated CMOS circuitry into programmable, autonomous robots too small to be seen by the naked eye.

Displays that use combinations of microscale LEDs and ICs within each pixel to control and direct light emission have the potential to be colorful, bright, fast and power efficient. High-throughput mass transfer technologies that accurately and cost-effectively integrate large arrays of wafer-fabricated microscale LEDs onto non-native display substrates with high yield are key enablers for microLED displays. Transfer-printing with elastomer stamps is a candidate mass transfer technology for making next generation displays. A variety of displays, including displays driven by transfer-printed microICs, have been designed and fabricated using transfer-printed microscale microLEDs.

Robotic instruments are miniature robots and mechatronic systems that can manipulate small objects and precisely interact with different surfaces. In this talk, I will discuss our recent work in robotic instruments for manipulation and assembly using two techniques: 1) acoustic manipulation devices that can automatically manipulate/assembly a large amount of particles on a surface using the field gradient generated by a single acoustic transducer, both in ambient air and submerged; 2) robotic electromagnetic needle that can precisely extract and independently manipulate/assemble magnetic microparticles in the presence of neighboring magnetic particles by contact or contactless.
References:
1. Latifi, K., Wijaya, H, and Zhou, Q., “Motion of heavy particles on a submerged Chladni plate”, Physical Review Letters, Volume 122, Issue 18. 5. DOI: 10.1103/PhysRevLett.122.184301, 2019.
2. Seon, J.A., Cenev, Z., and Zhou, Q., “Automatic Non-Contact Extraction and Independent Manipulation of Magnetic Particles Using Electromagnetic Needle”, under review.
3. Cenev, Z., Zhang, H., Sariola, V., Rahikkala, A., Liu, D., Santos, H.A., and Zhou, Q., “Manipulating Superparamagnetic Microparticles with an Electromagnetic Needle”, Advanced Materials Technologies, 3 (1), 1700177, 2018.
4. Zhou, Q., Sariola, V., Latifi, K., Liimatainen, V., “Controlling the motion of multiple objects on a Chladni plate”, Nature Communications, 7, 12764, 2016.

As Moore’s law for digital computation reaches its inevitable limits, future advances in system level integration will lead to the next generation of electronics. Approaches such as heterogeneous integration will enable small and high-performance micro-systems (optical and electronic integration or System-In-Package) necessary for the growing consumer areas such as wearables, the Internet of Things, and Augmented Reality. Beyond electronics, tools and techniques for heterogenous micro-assembly can lead towards new smart materials (i.e. defect-free constructed materials, digitally tailored electro-mechanical active materials with micro-scale properties and macro-scale performance, etc.) and the micro-assembly of biological and soft components.

While there are many research approaches towards tackling heterogenous micro-assembly (i.e. self-assembly, transfer printing, etc.), in this work, we will discuss our approach of directed robotic assembly using micro/milli robots. Directed micro/milli-robotic assembly has some unique advantages over other methods, such as being relative task agnostic and generalizable to various assembly problems. Furthermore, unlike many batch processes that are open-loop, directed micro-robotic assembly can be closed-loop by integration of metrology and material validation during the assembly process. Thereby swarms of micro/milli-robots can parallelize not only micro-assembly but also micro-metrology, enabling a method of overcoming statistical and thermodynamic limitations of conventional approaches.

Inspired by massive swarms of ants or termites assembling macro-scale objects from micro-scale parts, we utilize arrays of magnetic actuators in conjunction with diamagnetic graphite to create a magnetic collective of levitated micro/milli sized magnetic robots with nm level motion precision for 3D micro-assembly. By leveraging diamagnetic levitation, we can reduce factors such as hysteresis and stiction to enable high-precision and repeatable motion of micro/milli-robots. These passively levitated, but electromagnetically steerable micro/milli-robotic platforms can serve many roles: (1) compact high-precision motion platform for the manipulation and delivery of micro-components, (2) high precision sensors, and (3) motion platform for other micro-scaled MEMS sensors and actuators. We will discuss multiple independently controlled robots (mm-sized) with dexterous behavior for assembling carbon fiber trusses and performing micro-assembly and micro-manipulation of 3d microspheres and silicon microparts, as well as some of the control issues for coordinating parallel robotic swarms for micro-assembly. In addition, by using robots cooperatively, we also demonstrate how teams of micro/milli robots can achieve new functionality and enable extremely high densities of independent micro-manipulation as compared to conventional standard robotic solutions.

Flexible origami structures can mimic the complicated motions of small creatures that are otherwise difficult to be achieved by rigid robots with limited degree of motion freedom. However, actuating origami structures in a compact and self-contained way has been a critical challenge. Here, we demonstrate a versatile approach of actuating origami micro-robots, by printing self-folding creases on compliant polymer membrane substrates. The self-folding creases are made of a type of stimuli-responsive transition metal hydroxides/oxides that can undergo large actuation under electrochemical or light stimulations. Direct printing of the stimuli-responsive material via a microfluidic electrochemical writing method at the locations of the creases enables the creases to self-fold independently “on demand”, with folding curvatures exceeding 1 mm^-1 under low-intensity visible-light stimulation in ambient conditions, or low-potential electrochemical stimulation in electrolytic environments, with response time as fast as in seconds. Based on the high performance of such active creases, complex miniaturized origami designs powered by hinges activated in an independently controllable way are demonstrated, including self-folding Miura and a full micro-robotic hand with independently programmable finger joints. These results prove a new, versatile paradigm for micro-robotics, where a transferrable approach is applicable to design and fabricate a wide variety of customizable micro-robots with compact construction and complex motions using different stimuli-responsive ceramic-based materials.

Complex 3D structures enable advanced functionalities that are otherwise out of reach. However, processing issues such as mechanical stability, conformity, alignment, as well as cost, are challenging to address. The overarching physical principle of the Self-Rolled-up Membrane (S-RuM) nanotechnology is strain-driven spontaneous deformation of 2D membranes into 3D architectures. It opens up a new paradigm to achieve 3D functional hierarchical architectures without the difficulties of processing in 3D, and promises to break the constraints of size, weight, and performance of many types of devices. In this talk, we demonstrate S-RuM inductors, transformers, capacitors, and filters with > 10X smaller footprint and unique performance characteristics, through global and local strain engineering and material hierarchical integration.

Understanding the mechanisms of material structures forming processes enables the design and fabrication of structures with improved or tailored properties. The objective of our work is to investigate the mechanisms underlying materials synthesis and processing, including crystal growth, particle self-assembly, and phase transformation, etc., and thus to tune the structures and functions of materials.
For example, particle aggregation, such as oriented attachment, is accepted as a common pathway and leads to well-organized structures and twin interfaces that possess unique or improved properties. However, many fundamental aspects remain unknown, such as the interplay of structures and factors that control them. Here we use in situ transmission electron microscopy and atomic force microscopy to directly observe process of materials synthesis and processing, such as particle assembly and phase transformation. Integrating theoretical calculations and simulations, we revealed the orientation-dependent attachments of particles, interfacial structures, and the factors that affect them, such as temperature, pH, ionic strength, particle shapes, and solvents, etc. In addition, we also demonstrated that defects (edge dislocation), induced via oriented attachments, improving photoactivity of TiO₂. Phase transformation and its role in materials functions are also investigated. The specific systems that we investigate are metal (Ag and Au), oxides (TiO₂), and minerals (mica), etc. The work provides a deep understanding of mechanisms of materials synthesis and processing and thus provides a guide to design and fabricate the morphologies and structures, which are closely related to material properties and functions.

While dip coating is a widely used solution-based assembly method to achieve large scale assembly of nanomaterials, its integration with the flexible substrate is hindered by the difficulty to choose a good wetting system of solvent, nanomaterials and polymer substrate. The good wetting solvent may swell or even dissolve the polymer substrate. Here, to the contrary, we utilize the hydrophobic nature of a Polydimethylsiloxane (PDMS) substrate and nanomaterials (e.g., graphene) and create a “dry” dip coating method by choosing water as the solvent. Unlike the traditional dip coating method, there are two important innovations. First, the dip coating speed can reach 5.4 m/min (in comparison with 0.001-0. 2 m/min for commercial dip coater) because solvent evaporation is required. A dry substrate with assembly nanomaterial is removed from the solution because water does not wet PDMS and nanomaterials. Second, a weak sono field is applied to the solution during the dip coating process to significantly speed up the assembly process and maintain the disperse status of nanomaterials. The principle can be generalized to a wide range of nanomaterials such as carbon black, carbon nanotube, and two-dimensional materials (graphene, MoS₂, and h-BN). The film thickness can be controlled through dip coating times and solution concentration. This strategy can be further integrated with microfabrication processes such as lithography to manufacture arrays of hybrid multilayer devices with different nanomaterials. As a proof-of-concept, strain sensor, electromagnetic interference (EMI) shielding coating, and multilayer capacitor are demonstrated. This ultra-fast dry dip coating strategy paves the way towards affordable flexible devices.