Electrokinetic based micro- and nanofluidic technologies provide revolutionary opportunities to separate, identify and analyze biomolecular species. Key to fully harnessing the power of such systems is the development of a robust method for integrated electrodes as well as a thorough understanding of the influence of the electrokinetic surface properties with and without different surface modifications. In this work, we demonstrate a surface micromachined fabrication approach for integrated addressable metal electrodes within centimeter-long nanofluidic channels using a low-temperature, xenon diflouride dry-release method for novel biosensing applications, as well as recent results from a joint theoretical and experimental study of electrokinetic surface properties in nano- and microfluidic channels fabricated with fused silica. The main contribution of this fabrication process involves the addition of addressable electrodes to a novel dry-release channel fabrication method, produced at <300°C, to be used in nanofluidic electronic sensing of biomolecules. We also show a novel method with which to coat our channels with silane based chemistries. Finally, we show a theoretical model that we developed to robustly characterize the surface. Our theoretical model consists of three parts: (1) a chemical equilibrium model of the wall, (2) a chemical equilibrium model of the bulk electrolyte, and (3) a self-consistent Gouy--Chapman--Stern triple-layer model of the electrochemical double layer coupling between (1) and (2). We use experimental data with aqueous potassium chloride and buffer solutions in silica nanochannels to confirm theoretical predictions. In addition to describing techniques used to measure electrokinetic properties within experimental micro- and nanofluidic systems, show three separate experimental validations of this electrokinetic surface model.

Flexible microsystems are a member of the flexible electronic family, yet provide important and unqiue functions. Various devices and microsystems have been developed such as the micro-machined infrared bolometer, piezoelectric actuators, piezoelectric pressure sensors, and micro-fluidics. Surface acoustic wave (SAW) devices are essential for electronics, microsensors, microsystems and lab-on-a-chip; however there is no report on the development of flexible SAW device yet. This paper reports the first development of flexible SAW devices on cheap, bendable and disposable plastic films.
SAW devices were made on ZnO thin films deposited on polyimide substrates by sputtering. By optimizing the process, we were able to obtain ZnO films on polymers with good quality and low stress of <100 MPa, suitable for direct device fabrication. Crystal structure characterization revealed the ZnO films are dominated by (0002) crystal orientation with the grain sizes of ~70 nm. The SAW devices were made by photolithograph and lift-off process. The results are highlighted as follows,
(1). Deposition process has significant effects on performance of SAW devices, high deposition pressure resulted in high quality ZnO films, and the transmission signal become larger as the pressure was increased.
(2). For a fixed wavelength, the ZnO film thickness affects the SAW characteristics strongly, the thicker the ZnO film, the better performance of the SAW devices.
(3). Flexible SAW devices showed two resonant modes. The resonant frequency of the first mode increases from 198 to 447 MHz as the wavelength was decreased from 32 to 10 mu;m, while the zero mode decreases from ~200 to 34 MHz. Large signal amplitudes up to 18dB were obtained from devices with different wavelengths.
(4). Theoretical analysis and modeling revealed that the zero mode resonance is the Rayleigh wave, while the first mode resonance is the Lamb wave. Modeling showed the Rayleigh wave velocity increases with decrease in wavelength, while the Lamb wave velocity decreases, in excellent agreement with the experimental results.
(5). The resonant frequencies of the SAW devices decrease linearly with increase in temperature, with the temperature coefficients of frequency of ~442 and ~245 ppm/K for the two mode waves respectively.
(6). The flexible SAW devices have been used to test the ability for microfluidics and particle sorting, two of many lab-on-a-chip functions. A streaming velocity up to ~3.5 cm/s was obtained. Also it showed the flexible SAW devices are able to perform microparticles concentration and sorting in a few tens of seconds.
The results showed that the flexible SAW devices have high performance for electronics, sensors and microfluidics, compatible to those made on rigid substrates, demonstrated great potential for widespread applications.

A MEMS floating element shear stress sensor has been developed for flow testing applications, targeted primarily and ground and flight testing of aerospace vehicle and components. Skin friction is one of the major components of drag in vehicles, but is difficult to measure using existing techniques such as oil film interferometry, boundary layer profile surveys, or thermal methods. Direct floating element MEMS sensors address these issues by providing real time, momentum transfer based unsteady shear measurements at a surface with the potential for low topology and array sensing. However, concerns remain about the interaction of the flow with the mechanical elements of the structure at the microscale. In particular, there are concerns about the validity of laminar flow cell calibration to measurement in turbulent flows, and the extent to which pressure gradients may introduce errors into the shear stress measurement. In order to address these concerns, a numerical model of the Tufts floating element shear stress sensor has been constructed.
The model describes the behavior of the mechanical components, fluid interaction and electrostatics of a micromachined nickel floating element fluid shear stress sensor. Three 3-D numerical simulations were performed on the geometry of this sensor: A finite element model of the static element; a computational fluid dynamics (CFD) model of the floating element, for flat and textured versions of the floating shuttle; and a finite element model of the capacitive sensing combs. The static, large displacement, finite element model subjected the sensor element to applied loads of 2Pa to 12Pa. This model shows linear deflection in the flow direction on the order of 10^-10m and deflection in the transverse flow direction on the order of 10^-11m at 12Pa. In the CFD model, the shear sensor experienced internal duct flows of 5CFH to 40CFH where the geometry of the floating element was studied to determine which features contributed to what percentage of the total applied force and sensitivity. Pressure gradient was determined to contribute to 25% of the sensitivity of the flat sensor and close to 60% for the textured shuttle showing that a textured shuttle surface adds to the pressure gradient sensitivity and non-linearity. The finite element model of the capacitive combs shows the electrostatic coupling and fringe effects and how they contribute to the sensitivity of the sensor. With the combination of these three numeric models, sensor output as a function of steady state fluid flow parameters can be predicted and directly compared to experimental data. The computational model allows us to quantify the contributions (e.g. pressure gradient vs shear, top surface vs. lateral surfaces) to the sensor output in a manner that is difficult by purely experimental means.

The measurement of precise submicron displacements is essential in several MEMS applications. For instance, the measurement of mechanical parameters of biological cells requires repeatable measurement of displacements in the nanometer regime. This paper presents a method to make precise displacement measurements in an aqueous media. Optical microscopes are limited by the diffraction of light and a maximum resolution of about 200 nm is typically achievable; pattern matching algorithms allow further refinement of the displacement measurements to about 100 nm. The measurement technique developed by Yamahata et al. [Yamahata et al., JMEMS, Vol. 19, No. 4, pp. 1273-1275, 2010] allows displacement measurements in the nanometer regime using only a conventional optical microscope setup. This technique is based on applying a FFT phase-shift comparison of movable spatially periodic comb structures with known pitch of the comb. Implementation of the Yamahata method using MUMPS structures operating in air demonstrated 5 nm repeatability [Warnat et al, 24th Canadian Congress of Applied Mechanics, Saskatoon, June 2-5, 2013]. However, estimation of mechanical parameters of live biological cells requires measurements in aqueous media. Implementation of the Yamahata method in aqueous environment resulted in a 35-50 nm repeatability [Warnat et al, 24th Canadian Congress of Applied Mechanics, Saskatoon, June 2-5, 2013], this is due to poor optical contrast between the periodic comb structures (Poly-Si) and the substrate (SiN). This paper shows an approach to overcome the material contrast problem in water by analyzing static color microphotographs. While the substrate and the moving comb intensities are similar and of low contrast, the color profiles are different and of higher contrast. The image analysis was complicated by the variable nitride layer thickness 600 +/- 70 nm and thus varying color. The contrast is improved by subtracting the substrate average RGB color from the image, resulting in a higher contrast image that is then processed by the Yamahata technique. This paper shows displacement measurements in an aqueous media with a repeatability of 10 nm. This approach is therefore a good alternative to electrostatic displacement measurement techniques, which need additional and often expensive electronic measurement devices.

Nanopores and nanochannels have been widely used as a useful platform for electrical detections of single-particles to study the fast translocation dynamics [D. Branton et al., Nat. Biotechnol., 26, 1146 (2008), B. M. Venkatesan et al., Nat. Nanotechnol., 6, 615 (2011)]. The sensing mechanism involves measurements of temporal blockage of the ionic current through a fluidic channel, which is called resistive pulses, by an individual particle passing through it. In recent years, there has been increasing number of research efforts in this field reporting miscellaneous features in the electrical signatures of single-molecule or -particle translocations, e.g. ionic current blockage, enhancement, or even composites of these two aspects. Although systematic and statistical analysis of numerous electrical signals gives comprehensible explanations on the underlying physics in many cases, electrical detections alone cannot ensure whether a resistive pulse is associated with envisaged translocation of individual analyte particles [S. Alon et al., Rev. Sci. Instrum., 81, 014301 (2010)].
In the presentation, I present a single-particle tracking in a fluidic channel by combined electrical and optical sensing. This approach offers the missing function of the electrical method for identifying the presence and absence of analyte in a nanopore or nanochannel at the moment when a resistive pulse is detected.
In experiments, a fluidic microsystem lithographed on a glass substrate was used to measure the ionic current through the microchannel. Meanwhile, fluorescent imaging of dyed polystyrene particles was conducted simultaneously from the back of the substrate. I will show that this technique enables unambiguous sorting of resistive pulses to translocation-derived and non-translocation-derived, particle-particle collisions and temporal blocking, and even discriminates single-particle and two-particle-cluster translocations. It is also found that the combined optical/electrical sensing reveals an important role of particle history in the pre-translocation regime in determining its electrophoretic dynamics in a microchannel by allowing detections of single-particle motions at extensive spatial and dynamic ranges. The capability of this combined approach to guarantee the presence of an analyte in a specific volume may be useful in developing electrode-embedded nanopore sensors [M. Zwolak et al., Rev. Mod. Phys., 80, 141 (2008), M. Tsutsui et al., Sci. Rep., 1, 46 (2011)], which requires elucidating whether the measured transverse current originates from a particle or a molecule of concern.

Capacitive energy storage is distinguished from other types of electrochemical energy storage by short charging times, the ability to deliver significantly more power than batteries and long cycle life. A key limitation to this technology, which is based on electric double-layer capacitance, is its low energy density. For this reason, there is considerable interest in exploring materials which exhibit pseudocapacitive charge storage where the fast and reversible redox reactions that occur with transition metal oxides lead to energy densities which are several times larger than traditional double layer capacitance.

Our results with mesoporous TiO₂, MoO₃ and Nb₂O₅ films establish that this porous architecture is very beneficial for capacitive energy storage. Through the co-assembly of inorganic sol-gel precursors with diblock copolymers, an interconnected mesoporous network is formed that enables the electrolyte to access the redox-active pore walls, thus ensuring that the entire film is electrochemically active. Because of the mesoporous morphology, pseudocapacitive contributions dominate the charge storage process. Moreover, these mesoporous materials exhibit significantly higher levels of charge storage with far better kinetics than the corresponding oxides without a well-defined pore-solid architecture. Mesoporous films of MoO₃ and Nb₂O₅ exhibit enhanced pseudocapacitive charge storage because of an additional contribution associated with lithium ions being inserted into preferentially oriented crystalline layers. The pseudocapacitive behavior exhibited by the mesoporous transition metal oxide films represents a very promising direction for designing electrochemical capacitors that can achieve increased energy density while still maintaining high power density.

This talk is about the electromechanical performance of highly grain-oriented (K0.50Na0.50)NbO3 ceramics fabricated using KNN single crystals as templates. Smart materials, that include piezoelectric and ferroelectrics, play a crucial role in the development of micro-electromechanical systems (MEMS), for applications as optical displays, acceleration sensing, radio-frequency switching, drug delivery, chemical detection, and power generation and storage. The increasing importance of MEMs in microelectronics industry has also been addressed by the International Road Map for Semiconductors (ITRS) within the concept of functional diversification called “More than Moore” [1]. Pb(Zrx,Ti1-x)O3 (PZT) is still the most widely used piezoelectric for such applications, but environment protection and related legislations - are pushing research of lead free piezoelectrics and ferroelectrics. Among them K0.5Na0.5NbO3 (KNN) is one of the leading lead free piezoelectric materials in the perovskite group being considered as a candidate material for MEMS. Undoped K0.5Na0.5NbO3 (KNN) has inferior electromechanical properties as compared to PZT and thus various efforts are on going to tailor and improve the electrical and electromechanical properties of KNN. Some of the strategies include doping, chemical substitutions, or microstructure texturing [2]. In this work, we report the fabrication and electrical properties of highly grain-oriented (K0.50Na0.50)NbO3 ceramics via the use of KNN single crystals as templates. This is an alternative strategy to the current use of hetero-templates (such as plate-like NaNbO3 particles) and not reported before. KNN single crystals were grown by a modified flux method [2]. The grown KNN crystals present a high dielectric permittivity of 29,100 at the tetragonal to cubic phase transition temperature, remnant polarization of 19.4 µC/cm2, piezoelectric coefficient of 160 pC/N and revealed a long range ordered domain pattern of parallel 180° domains with zig-zag 90° domains. The role of homo-texturing on the orientation dependence of the electromechanical properties is presented and the relationships between the properties, processing, and performance of KNN are established. The obtained results are benchmarked against the current available information published in the literature.
References:
[1] International Technology Roadmap For Semiconductors 2012.
[2] Y. Saito, H. Takao, T. Tani, T. Nonoyama, K. Takatori, T. Homma, T. Nagaya, M. Nakamura, Lead-free piezoceramics. Nature, 2004, 432, 84-87.
[3] Muhammad Asif Rafiq, M. Elisabete V. Costa, Paula M. Vilarinho, Establishing the domain structure of (K0.5Na0.5)NbO3 (KNN) single crystals by piezoforce-response microscopy, submitted.

Aiming at harvesting ambient mechanical energy for self-powered systems, e.g. self-powered microelectromechanical systems (MEMS), triboelectric nanogenerators (TENGs) have been recently developed as a highly efficient, cost-effective and robust approach to generate electricity from mechanical movements and vibrations, on the basis of the coupling between triboelectrification and electrostatic induction. However, all of the previously-demonstrated TENGs are based on vertical separation of triboelectric-charged planes [1, 2], which requires sophisticated device structures to ensure enough resilience for the charge separation, otherwise there is no output current. In this paper, we demonstrated a newly designed TENG based on an in-plane charge separation process using the relative sliding between two contacting surfaces. Using Polyamide 6,6 (Nylon) and polytetrafluoroethylene (PTFE) films with surface etched nanowires, the two polymers at the opposite ends of the triboelectric series, the newly invented TENG produces an open-circuit voltage up to ~1300 V, and a short-circuit current density of 4.1 mA/m2, with a peak power density of 5.3 W/m2, which can be used as a direct power source for instantaneously driving hundreds of serially-connected light-emitting diodes (LEDs). The working principle and the relationships between electrical outputs and the sliding motion are fully elaborated and systematically studied, providing a new mode of TENGs with diverse applications. Compared to the existing vertical-touching based TENGs, this planar-sliding TENG has a high efficiency, easy fabrication and suitability for many types of mechanical triggering. Furthermore, with the relationship between the electrical output and the sliding motion being calibrated, the sliding-based TENG could potentially be used as a self-powered displacement/speed/acceleration sensor in MEMS. [3]
[1] Fan, F. R.; Tian, Z. Q.; Wang, Z. L. Nano Energy 2012, 1, 328-334.
[2] Wang, S. H.; Lin, L.; Wang, Z. L. Nano Lett. 2012, 12, 6339-6346.
[3] Wang, S. H.; Lin, L.; Xie, Y. N., Jing, Q. S., Niu, S. M., Wang, Z. L. Nano Lett. 2013, 13, 2226-2233.

The battery-less sensor system has been investigated for the health monitoring system of infrastructure because of the maintenance-free and the low cost. These devices operate by the energy from external environmental sources. Vibration energy harvesting by the piezoelectric films is one of the method to generate the electric power. In this method, the piezoelectric films are required the high flexibility as well as the good electromechanical properties because the large displacement of piezoelectric films leads the large energy generation. Therefore, it is important to get the piezoelectric films on flexible substrates. In addition, environmental friendly lead-free piezoelectric materials must be used and have superior piezoelectric and electromechanical properties.
In the present study, 3 mu;m-thick (K,Na)NbO3 films were deposited on metal foil substrate (25 mu;m in thickness) at low temperature of 240omicron;C by hydrothermal method. Hydrothermal method is the wet process and deposit films at low temperature below 300omicron;C for large area.
We succeeded to make the flexible (K,Na)NbO3 films because these films did not break by the large displacement. Clear hysteresis loop was observed in polarization- electric filed and strain-electric field relationships. d33 value and remanent polarization were 28 pm/V and 7 mu;C/cm, respectively. Moreover, the resonance frequency value was 20 Hz. This value of (K,Na)NbO3 films on flexible substrate was smaller than that of (K,Na)NbO3 films on rigid substrate. These low resonance frequency is useful because the environmental frequency is generally less than about 1 kHz.

Here we present a primary microbattery with high capacity and high power density. The microbattery is based on interdigitated three-dimensional nanoporous bicontinuous electrodes, which enables short ion diffusion lengths and high discharge power. The microbatteries utilize new primary battery chemistry to achieve a capacity of 8.67 µAh/cm2µm, which is 3X the average capacity of our previously reported high power rechargeable microbatteries [1]. The microbattery capacity is increased by using lithium as an anode and two phases of manganese oxide as the cathode to achieve a large discharge voltage range of 3.3 - 0.6 volts with high capacity. Increasing the discharge voltage range takes advantage of the low voltage requirements of state of the art microelectronics for portable devices [2]. The interdigitated three-dimensional nanoporous bicontinuous electrodes allow for micro-integration of the anode and cathode and shorten the ion diffusion and electron conduction lengths to achieve high power density. The 3D electrodes were formed by electrodepositing nickel through a colloid assembly formed from 500 nm diameter colloidal particles, resulting in a Ni inverse of the colloid structure, on a gold patterned glass substrate. Electrodeposited lithium was used for the anode and electrodeposited manganese oxide for the cathode active materials. Lithium is an excellent primary anode because of its high energy density and low, constant voltage relative to the cathode. During galvanostatic discharge the manganese oxide has two voltage plateaus with mean voltages of 2.6 volts and 0.8 volts, corresponding to two difference Mn valence states. The footprint area of the tested cells was ~4 mm2. The microbattery cells can deliver a high power density of 3,300 µW/cm2µm, which is close to the highest power reported for microbatteries [1]. Additionally the microbatteries can alternate between a high power pulsed discharge (740 µW/cm2µm pulsed 36 times for 0.1 seconds) and a low power sleep discharge (0.3 µW/cm2µm), which is critical for applications requiring wireless communication and device actuation. High capacity and high power density primary microbatteries can enable smaller power sources, faster computation, stronger actuation, and longer data transmission distances for microsystems with short to medium lifetimes.
[1] J. H. Pikul, H. Gang Zhang, J. Cho, P. V. Braun, and W. P. King, "High-power lithium ion microbatteries from interdigitated three-dimensional bicontinuous nanoporous electrodes," Nature communications, vol. 4, p. 1732, 2013 2013.
[2] G. Chen, M. Fojtik, K. Daeyeon, D. Fick, P. Junsun, S. Mingoo, C. Mao-Ter, F. Zhiyoong, D. Sylvester, and D. Blaauw, "Millimeter-scale nearly perpetual sensor system with stacked battery and solar cells," in Solid-State Circuits Conference Digest of Technical Papers (ISSCC), 2010 IEEE International, 2010, pp. 288-289.

Measuring the motion of microelectromechanical systems is key to understanding system performance and reliability and is rapidly becoming an area of great commercial importance. These measurements can be difficult due not only to the size of the measured devices, but also to the size of the characteristic motions of the devices, which are often many orders of magnitude smaller. Such is the case for many stepwise microactuators and closed-loop acccelerometers, in which the characteristic motions may be on the order of nanometers. Although such small dimensions lie beneath the Rayleigh limit for optical resolution, the motions may still be measured optically using super-resolution analysis techniques, provided that there is sufficient contrast in the image. One way to provide sufficient contrast for these techniques is through fluorescence microscopy. By labeling a MEMS device with multiple fluorescent nanoparticles and measuring the motions of these particles, translations and rotations of the MEMS devices can be determined. Measurement uncertainties of 1.85 nm and 100 microradians have been achieved, with localization precision of 130 pm.

Phonons — the quanta of lattice vibrations — are the major heat carriers in semiconductors and insulators, and a better understanding of their nanoscale transport will inform the engineering of materials for thermoelectric and microelectronic cooling applications. To this end, we have developed a microfabricated phonon spectrometer utilizing superconducting tunnel junctions (STJ) as phonon transducers.[1,2] Al-AlxOy-Al STJs are utilized for emission and detection of tunable and non-thermal acoustic phonons with frequency ~100 to 870 GHz in silicon microstructures. [1,2] We show that phonon spectroscopy with STJs offers a spectral resolution of ~15-20 GHz, which is ~10 times better than thermal conductance measurements, for probing nanoscale phonon transport. We discuss the design, fabrication, and low-noise instrumentation needed to implement a phonon spectrometer. We show that the phonons propagate ballistically through silicon microstructures with submicron spatial resolution. With a spectrally resolved measurement of phonon transport such as ours, the understanding of phonon propagation and scattering in nanoscale structures will be greatly improved; hence, the exploitation and engineering of phonons for thermoelectric devices, microelectronic coolers, and phononic devices should become more feasible. In addition, with the ability to distinguish acoustic phonon modes, we can further understand the acoustic loss mechanisms in micromechanical resonators.
References:
[1] J.B. Hertzberg, O. O. Otelaja, N. J. Yoshida, and R. D. Robinson. Rev. Sci. Instrum. 82, 104905 (2011).
[2] O.O. Otelaja et al, New J. Phys. 15 (2013), 043018

This work is devoted to capillary nucleation rate with time. It is well known that the environment in which microelectromechanical systems operate significantly affects their performance. It is therefore important to characterize micromachine behavior in environments that are well controlled. To this end, we have modified and characterized an advanced test system that allows for environmental control (pressure, relative humidity, and gas composition) while retaining full micromachine characterization capability (long working distance interferometry, electrical probe connectivity, actuation scripting). The system also includes a load lock with in-situ plasma cleaning that renders surfaces uniformly hydrophilic. A microcantilever crack healing experiment is conducted and surface adhesion energy measurements are tracked over time after a step change in humidity is applied. The experimental sample includes twenty silicon cantilevers, of lengths 1,050 µmle;Lle;2,000 µm in 50 µm increments. The height of the step-up support post is h=1.8 µm, the cantilever width is w=20 µm and the cantilever thickness is t=2.5 µm. After increasing relative humidity from 30% to 53%, the crack healing rate begins at 4 µm/minute and gradually decreases to 0 after 410 minutes. An FEM model has been generated to simulate the near-crack tip capillary adhesion forces. From this model, it is found that the capillary nucleation rate decreases from 0.00525 to 0.0000797 bridges/µm2/min as the crack heals. In turn, the capillary nucleation rate can be understood via a thermal activation process in which the activation energy depends on the local asperity gap. We will test this model by measuring crack healing rates at different temperatures and humidity levels. Capillary formation is important in nature (granular materials, insect locomotion) and in technology (disk drives, adhesion), and the methodology described here enables us to measure and study capillary bridge nucleation rates.

The NIST microscale “theta” specimen, which is shaped like the Greek letter Θ, acts as a mechanical test specimen when loaded in compression; generating tensile stress in the central web of the specimen or bending stress in the outer frame. MEMS-scale lithographic and etching processes can be used to microfabricate many specimens such that statistically-relevant numbers of strength tests can be performed for MEMS device reliability predictions. Here, the tensile strength distributions of Si specimens resulting from different etch recipes are used to predict reliability in two different ways. In the first, the ability of a component to support a required load over a specified lifetime is considered: The loading spectrum applied to a component is varied, and the proportion of intact components remaining as a function of time is determined. Both explicit and stochastic variations of the maximum applied stress with time are demonstrated. In the second, the effect of a change in component loading geometry is considered: The specimen geometry is altered and the proportion of intact components remaining as a function of applied load is determined. In both approaches, the theta specimen is used to great effect to gain quantitative insight into the role of etching-induced surface features on the manufacturing yield and operational reliability of MEMS components.

The characterization and understanding of the mechanical response of freestanding nano-objects are of prime interest for fundamental research and for supporting reliability analysis of various Micro- and NanoElectroMechanical Systems (MEMS/NEMS). A versatile on-chip mechanical testing platform has been developed using MEMS-based structures in order to explore the mechanical properties of freestanding nano-objects. These structures take advantage of the mismatch stress present in a long silicon nitride beam to apply a deformation to a specimen beam attached to it owing to the release of the underneath sacrificial layer. The connection between the both beams is ensured by an overlap. At the specimen beam ends, dogbone shapes are located to concentrate the loading where the width is uniform. The in-plane dimensions of both, the actuator and the specimen beams, are varied to induce different strains and stress levels. Then, a large range of deformation can be applied to the specimen allowing the extraction of the strain-stress curve as well as the creep/relaxation behavior. Thousands of tests can be performed using a single processed wafer making statistical data analysis possible. The specimen stress and strain are extracted from the measured total displacement of the structure after its release using an analytical model based on beam theory.
Recently, in order to improve the accuracy of the stress and strain estimation, additional geometrical and material features contributing to the total displacement of the structure have been implemented into the analytical model. These geometrical features are the specimen dogbones, the overlap between the actuator and specimen, and the under etching of the specimen and actuator fixed parts after the structures release. The presence of a stress gradient in the actuator constitutes an additional feature that must be taken into account. 3D finite elements simulations have been performed to evaluate the improved model precision and to determine what extent these additional features have to be considered or neglected.
The technique has been used to study the deformation and fracture of phosphorus-doped polysilicon thin films. Two different film thicknesses have been tested: 240 and 40 nm. The Young&’s modulus and the fracture strain have been extracted. For the two film thicknesses, the fracture strain increases with decrease specimen surface area. The thinnest film exhibits the lowest fracture strains. More surprising, post mortem analyses of broken samples reveal a change of fracture mode between the two film thicknesses, from transgranular to intergranular. The fracture strain difference and the fracture mode change observed between the two films thicknesses will be discussed in the light of surface roughness and microstructural characterization using the ACOM-TEM method.

Microsystems have the potential to impact biology by providing new ways to manipulate cells and the microenvironment around them. Simply physically manipulating cells or their environment—using microfluidics, electric fields, or optical forces—provides new ways to separate cells and organize cell-cell interactions. One example illustrating the power of microscale manipulation of cells is to sort cells based on their intrinsic electrical properties. Electrical properties have previously been correlated with important biological phenotypes (apoptosis, cancer, etc.), but a sensitive and specific method approach has been lacking. We have developed a method called iso-dielectric separation that uses electric fields to drive cells to the point in a conductivity gradient where they become electrically transparent, resulting in a continuous separation method specific to electrical properties. With this method, we have screened the entire genome of an organism to understand the biological basis of electrical properties, finding that the relationship between genetics and intrinsic properties has both intuitive and non-intuitive features. We have also applied those results to develop a point-of-care assay for immune cell activation for monitoring sepsis. Microfluidics can also be used to manipulate the environment around cells. For example, we have developed arrays of microfluidic perfusion culture chambers that use fluid flow to create a convection-dominated transport environment, allowing control over local cell-cell diffusible signaling. This in turn provides a more controlled soluble microenvironment in which to study diffusible signaling in cell systems. In particular, we have examined the impact of diffusible signaling on self-renewal and neural specification of embryonic stem cells. Using these microsystems, we have identified the existence of previously unknown autocrine loops involved in fate specification, and have delineated the effects of shear itself on self-renewal. Together, these new microscale tools provide ways to exploit cells&’ potential for both basic science and applied biotechnology.

We present a new method for prototyping polymer microsystems, termed as Solvent Immersion Lithography (SIL) [1] that enables ultra-fast polymer imprinting, and bonding; for example, a complete assembly of a microfluidic system can be typically achieved in less than one minute, and under minimal instrumentation requirements. SIL is based on directed polymer disolution, induced by immersing a polymer into a favourable solvent. The same immersion step enables also the penetration of molecules mixed with the solvent into the polymer walls. This leads to channel surface functionalization, and adds novel functions to the imprinted features such as catalytic or sensing capabilities. Microfluidics [2], optofluidics [3], cell growth microreactors, as well as microstructured oxygen sensors prototyped by SIL will be discussed.
In directed polymer dissolution, the solvent penetrates into a polymer film at depths determined by the underlying polymer-solvent interactions [4]. This type of interaction generates a surface gel that can be imprinted and subsequently irreversibly bonded with non-treated surfaces. The imprinting depth can be controlled through the compatibility between the solvent and polymer solubility parameters, as well as using solvent mixtures. In this talk we will focus on polystyrene and various organic solvents, as well as water-solvent mixtures. The latter was found to inhibit imprinting even at minute amounts of water, hence giving rise to another, and tightly regulated imprinting mechanism. These effects were analyzed in detail using surface chemical imaging, such as micro-Raman and confocal microscopy, as well as computational methods [1].
The details and performance of SIL will be presented, along with microfluidic and optofluidic applications. Regarding optofluidics, oxygen sensing moieties were integrated in the imprinted surfaces of a microfluidic channel, enabling the monitoring of oxygen concentrations within. Furthermore, as an example in microbiology, the growth of pure and mixed microbial populations was investigated in polystyrene micro-pore scale models. In these, a double mean generation lifetime was revealed for the thermophile anaerobe cellulose fermenter Clostridium Thermocellum [5] than under ideal culture conditions.
[1] A. E. Vasdekis, et al. under consideration (2013).
[2] G. M. Whitesides, Nature 442, 368 (2006).
[3] D. Psaltis, S. R. Quake, C. Yang, Nature 442, 381 (2006).
[4] E. Kim,, Y. N. Xia, X. M. Zhao, G. M. Whitesides Advanced Materials 9, 651 (1997).
[5] A. L. Demain, M. Newcomb, J. H. D. Wu, Microbiology and Molecular Biology Reviews 69 124 (2005).

We have developed a MEMS filtration device fabricated with patterned vertically aligned carbon nanotubes (VACNT) and polydimethylsiloxane (PDMS). The microfluidic channel wall of the droplet-shaped device is made of a VACNT forest in the dimension of 50 µm in height, ~10^7 counts/cm^2 in density and 15 nm in single VACNT diameter. The VACNT is synthesized by aerosol based chemical vapor deposition on a patterned Fe catalyst thin film. The Fe catalyst is deposited on a silicon substrate and patterned by lift-off fabrication in droplet geometry. The top PDMS cover is fabricated by SU-8 molding technique with two fluidic access through holes, one inlet and one outlet. Oxygen plasma surface treatment is applied on both PDMS and VACNT to achieve permanent bonding and sealing. The dead-end filtration is performed in the assembled device. The inlet port is established as a sample reservoir. The outlet port is connected to vacuum source with a sample collection trap. The sample at the inlet reservoir is driven by vacuum source connected outlet and transported through the filtration membrane. The membrane is characterized to have permeability of ~10^14 m^2, porosity of ~99.6% and cut-off size of ~124 nm. The measured permeability and porosity match with the theoretical estimation by Darcy's model at the same order of magnitude. The cut-off size is estimated by combining flow resistance experimental data to a VACNT forest geometrical model, and later confirmed to be within the range of 110-140 nm with filtration experiments using fluorescently labeled nanoparticles of different sizes. The membrane is able to achieve high throughput nano-scale liquid filtration due to its high porosity and nanometer range cut-off size. Furthermore, we demonstrate the blood albumin measurement after plasma separation from human whole blood. The results show that plasma from the blood can be extracted and collected at outlet vacuum trap while the blood cells are blocked and confined at the enclosed droplet chamber. The measured blood albumin concentration matches with that measured after conventional plasma extraction by centrifugation.

Rotary nanomotors, a critical but less developed type of NEMS devices, is extremely important for further advancing NEMS technology. In this work, we report innovative design, assembly, and rotation of ordered arrays of nanomotors. The nanomotors were bottom-up assembled from nanoscale building blocks with nanowires as rotors, patterned nanomagnets as bearings, and quadrupole microelectrodes as stators. Arrays of nanomotors can be synchronously rotated with controlled angle, speed (to at least 18,000 rpm), and chirality by electric fields. The fundamental nanoscale electrical, mechanical, and magnetic interactions in the nanomotor system were investigated, which adds new knowledge and understanding for design of various metallic NEMS devices. The innovations from concept, design, to actuation are on multi-levels and may bring transformative impact to NEMS, microfluidics, and lab-on-chip architectures.

MEMS-based biological and chemical sensors have gained increasing attention due to their ability to provide a technique for high-precision (zepto-gram), label-free sensing. However, to date, the widespread application of this class of sensors has been impeded in part by the lack of a technology capable of performing high resolution measurements of large arrays of MEMS devices. We present an experimental demonstration of a technology capable of filling that need.
MEMS-based sensors consist of either free-standing microcantilevers or microbridges, coated in a functionalization substance which preferentially bonds to the chemicals being searched for. MEMS sensors have two modes of operation: static (where height position is measured) and dynamic (where resonance frequency is measured). To measure the state of the cantilevers, our design integrates a silicon photonics waveguide and diffraction grating beneath the sensing cantilever. By making the underside of the cantilever reflective, a resonant optical micro-cavity is formed beneath the cantilever. The amount of light coupled back into the waveguide from the grating is an extremely sensitive measure of the position of the cantilever.
Silicon photonics optical components were fabricated using deep ultra-violet (DUV) lithography. Microcantilevers were then fabricated above these components using surface micromachining. The optical output from the waveguide was measured as a function of the height of the suspended microcantilever.
The measurements show that the position of the suspended MEMS is measured with accuracy of single picometers. We have demonstrated the measurement of the first two harmonics of cantilever resonant mechanical motion that is thermally stimulated at room-temperature.

MIT Lincoln Laboratory has a long history of MEMS development and application from RF MEMS and optical switches, to microfluidics for chemical and biological applications. This talk will cover some development of fully packaged electrostatic actuated RF MEMS switches and varactors, application of commercial varactors to tunable RF circuits, novel liquid metal switches, microfluidic technologies for chemical and biological applications, and liquid crystal imagers for IR applications. In all cases, the development and technology transfer path of these technologies will be discussed. Additionally, the integration of these technologies with control and microelectronics will be addressed.

A fast, low power, transistor-type switching device has been proposed as a potential CMOS replacement for computer logic. In this technology, called piezotronicsdagger;, piezoelectric and piezoresistive materials are employed in a stacked sandwich structure of nanometer dimension. Actuation of the piezoelectric material results in a pressure induced insulator-to-metallic transition in the electrical conductivity of the piezoresistive material, turning the switch on from the normally off state. Of particular interest to this program is the functionality of the high aspect ratio piezoelectric 70Pb(Mg1/3Nb2/3)O3-30PbTiO3 (PMN-PT) component. Dense PMN-PT films of approximately 350 nm in thickness were made by chemical solution deposition using a 2MOE solvent. These films were phase pure and strongly {100} oriented by XRD with dielectric constants exceeding 1400 and loss tangents of approximately 0.01. The films showed slim hysteresis loops with remanent polarizations of about 10 mu;C/cm2 and breakdown field of over 1500 kV/cm. Fully clamped films exhibited large signal strain of approximately 1%, with a d33,f coefficient of approximately 90 pm/V. By laterally subdividing the blanket PMN-PT film into smaller driving pixels, the piezoelectric response is declamped from the substrate while reducing footprint of an individual piezotronic pixel. Reactive ion etching (RIE) has been employed to pattern features in the PMN-PT film down to one micron in spatial scale with vertical sidewalls. Upon lateral scaling, an increase in both small and large signal dielectric properties has been observed, including an approximate 50% increase in permittivity in PMN-PT structures with 1 mu;m wide features. A modified Mirau single beam interferometer is currently being built at Penn State University to accurately measure the evolution of the piezoelectric properties as the films are laterally subdivided.
dagger; Reference: D.M. Newns, B.G. Elmegreen , X.-H. Liu, and G.J. Martyna, “The piezoelectronic transistor: A nanoactuator-based post-CMOS digital switch with high speed and low power”, MRS Bulletin, 37, 1071 [2012 ].

Ohmic micro- or nanorelays have been recently regarded to complement transistors in applications where electrical current leakage is becoming a problem. Although the solid state metal oxide silicon field effect transistor (MOSFET) has fueled a global technology revolution, it is now reaching its performance limits because of device leakage. To avoid electric field-induced damage in MOSFETs, operating voltage and hence threshold voltage must be reduced as linewidth is reduced. However, below a limit, the current cannot be turned off. The ohmic relay approach solves this problem because an air gap that separates the electrical contacts provides excellent electrical isolation when the relay is open. Some applications require these relays to perform billions of cycles, yet typical devices that are exposed to ambient environment degrade electrically after just a few thousand cycles. The main challenge here is that trace amounts of volatile hydrocarbons in air adsorb on the electrical contact surfaces for a large variety of coating materials, causing an insulating deposit to form that prevents signal transmission during switch closure. We investigate how to mitigate this challenge by exploring the effect of contact materials and operating environment on device lifetime. We select Pt, a common coating material in switch applications due to its resistance to wear, and RuO2, which is believed to be somewhat resistant to hydrocarbon adsorption. We test our devices in N2 and O2 background environments with controlled hydrocarbon contaminant concentrations. We firstly demonstrate that RuO2, a conducting oxide, does not exhibit contaminant-induced degradation even at very high hydrocarbon presence as long as O2 is also present. We also demonstrate that electrical contacts that have been pre-contaminated with hydrocarbons can be restored to their original performance level by cycling the devices in clean N2-O2 environment. Finally, we illustrate that the insulating hydrocarbon deposit can be electrically broken-down and its resistance lowered. Nanorelays, which can tolerate higher resistance values than microrelays, can potentially take advantage of this phenomenon.

Direct integration of electronics on piezoelectrics is attractive for large array transducers, as it allows for the possibility of reconfigurability and active surfaces. In particular, row-column address schemes for large area arrays would be useful in enabling individual element control with a greatly reduced number of connections. In this work, ZnO thin film transistors (TFTs) were integrated with lead zirconate titanate (PZT) piezoelectric films to assess whether interposer electronics on PZT can serve as a bridge between CMOS electronics and MEMS devices. Prior to electronics deposition, Nb-doped PZT films were deposited using solution based processing on platinum coated silicon substrates. The PZT was found to have a dielectric constant of 1100 with a loss of less than 3% and an average remanent polarization of greater than 27mu;C/cm2 before transistor deposition. Using plasma enhanced atomic layer deposition, ZnO TFTs were fabricated alongside reactive ion etched PZT capacitors. It was found that there were no significant changes in either the ZnO TFT characteristics (mobilities ~ 20 - 25 cm2/Vs) or the baseline PZT properties following integration. Furthermore, it was shown that these TFTs can be used to power piezoelectric actuators.

Using the piezoelectric polarization charges created at metal-GaN nanobelt (NB) interface under strain to modulate transport of local charge carriers across the Schottky barrier, piezotronic effect(2) is employed to convert mechanical stimuli applied on wurtzite-structured GaN NB into electronic controlling signals, based on which GaN NB strain-gated transistors (SGTs) have been fabricated. By further assembling and integrating GaN NB SGTs, universal logic devices such as NOT, AND, OR, NAND, NOR and XOR gates have been demonstrated for performing mechanical-electrical coupled piezotronic logic operations. Moreover, basic piezotronic computation such as one-bit binary addition over the input mechanical strains with corresponding computation results in electrical domain by half-adder has been implemented for the first time. The strain-gated piezotronic logic devices may find novel applications in human-machine interfacing, active flexible/stretchable electronics, MEMS, biomedical diagnosis/therapy and prosthetics.
References:
1. Yu, R. M.; Wu, W. Z.; Ding, Y.; Wang, Z. L. GaN Nanobelts Based Strain-Gated Piezotronic Logic Devices and Computation. ACS Nano (Submitted) 2013.
2. Wang, Z. L. Piezotronic and Piezophototronic Effects. J Phys Chem Lett 2010, 1, (9), 1388-1393.

Most commercial micro-electro-mechanical systems (MEMS) are fabricated out of silicon based on extensive industrial processing experience. However, this reliance on silicon as a structural material often places limitations on device functionality, operating conditions or reliability. Highly engineered metallic alloys that can be precisely sculpted to submicron resolution would offer improved strength and thermal stability and may lead to a wide array of new MEMS applications. Many of these new opportunities exist in elevated temperature environments. Ni-based alloys are excellent candidates for these extreme environments. Free-standing sputtered thin films of several low thermal expansion alloys have been fabricated and characterized. The unique microstructures resulting from the sputtering process influences both the mechanical behavior as well as the thermal stability. The link between microstructural evolution and performance will be discussed with emphasis on MEMS integration and systematic identification of other promising alloy systems.

Fast subsurface Faradic redox reactions and charge transport rate are vital factors for achieving high energy and power pseudocapacitors. However, the typical thin film configuration hardly fulfills the high energy requirement while thick electroactive material deposition restricts ion diffusion path. In addition, high mass loading could further lower fault tolerance. Here, we demonstrate the fabrication of open-ended Ni(OH)2 nantube arrays growing on nickle foam via a sacrificial ZnO nanowire template. The increasing accessible electrochemical sites benefiting from nanoarchitectured inner/outer walls of nanotubes lead to swift and reversible redox reactions. The improved electrolyte percolation rate as well as the high fault tolerance endow the Ni(OH)2 nanotube arrays superior electrode materials. The protocol allows a highest Ni(OH)2 nanotube loading of 3.18 mg cm-2 which yields a high areal capacitance of 5.88 F cm-2 (1850.6 F g-1) with remaining high rate capacity. The assembled Ni(OH)2 nanotubes/KOH/activated carbon supercapacitor possesses an operational potential window of 1.4 V for actual device applications.

Development of frequency-adaptive micro and nanoresonators is a fundamental task for engineering new applications. Dynamic control of mechanical resonances in nanoelectromechanical systems (NEMS) has been obtained using electrostatic fields [P. Quirin et al. Nature 458, 07932 (2009)] or exploiting thermal expansion given by electro-thermal heating [S. C. Jun et al. Nanotechnology 17, 1506 (2006)]. However, these solutions lack of memory effects. A different approach, based on correlated electron materials showing phase transition at nanoscale such as VO2 [E. Merced et al., Smart Mater. Struct. 21 (2012) 035007], can thus introduce new paradigms for the development of novel integrated nanoresonators and mechanical devices. VO2 attracted lot of attentions due to its still debated Metal-Insulator Transition (MIT) above room temperature. It shows more than three-orders-of-magnitude hysteretic change of electrical resistance nearby 68°C associated with a change of its crystal structure from monoclinic (M) to rutile (R) -type.
In this work, we report the possibility of programming multiple eigenfrequency states of a VO2-based microresonator with a high degree of control and reproducibility using confined Joule effect. In our previous work [L. Pellegrino et al Adv. Mater. 24, 2929 (2012)], we introduced VO2-based Joule-assisted two-terminal multiple resistive memories, where volatile and non-volatile resistive states were programmed by electrical current pulses. Here, our prototype device is a mechanical thin film microresonator made of a VO2(200nm)/TiO2(110) (20 nm) heterostructure (approximately 20um x 30um x 200 nm with 5um wide arms). The key issue of this work is controlling in a selected region the electronic phase of nanometric metallic clusters, driving the metallic/insulating ratio by Joule heating effect, with the consequent change of the mechanical stiffness of the device. The total amount of current determines the number and the size of metallic clusters and thus the overall rigidity of the microresonator itself. The hysteretic nature of VO2 together with the efficient Joule heating on free-standing regions allow us controlling resonator eigenfrequency and implementing a mechanical memory having volatile/non-volatile multistate programmable behaviour. The results we present do not show finite size effects because of the intrinsic nanoscale size of VO2 domains on TiO2(110). Downscaling to NEMS devices is thus feasible. Our studies can be also extended to more complex geometries, where the opportunity of controlling the reciprocal ratio of phases with two mechanical stiffness values at nanoscale opens promising perspectives for developing devices with reconfigurable mechanical parameters.

Stretchable electronics have generated great interest in the recent past [Science 327, 1603 (2010)]. In this work, we report structural transformation of traditional electronics to obtain flexible and stretchable silicon fabric. Two structures - serpentine and spiral - are investigated with mechanical modeling and simulation. Next, we experimentally study the mechanical characteristics of both these architectures. The generic process flow involves the use of bulk silicon (100) substrate (or any form of silicon) and consists of standard microfabrication processes.
The design of the fabric consisted of small squares interconnected with either serpentine or spiral structures, to provide the requisite stretchability. These interconnect structures elongate when a longitudinal stress is applied, thus making the fabric itself stretchable [Proc. IEEE IEDM, 217 (2007)]. These structures where simulated using finite element analysis technique in COMSOL to obtain the optimum values for the parameters such as edge width, inner diameter, number of turns etc. It was seen that the spiral structures gave a stretchability of 2 orders-of-magnitude more than the serpentine structures for the same edge width, number of turns and applied force. Among the spiral structures, an edge width of 20 µm, inner circle diameter of 100 µm and edge separation of 3 µm was the most efficient design - giving a strain-to-load ratio of 15 N^-1.
The fabrication process started with the fabrication of CMOS circuitry integrated bulk silicon (100) substrate. A layer of inter-layer dielectric (ILD) was deposited to protect the devices and circuits. A pattern of holes (5 µm diameter and 15 µm center-to-center spacing) was made in the ILD layer using Reactive Ion Etching (RIE). The holes were then etched to a depth of 50 µm into the silicon substrate&’s unused areas using deep RIE. Then we formed an Al₂O₃ based vertical spacer on the sidewalls of the holes. The wafer was then subjected to XeF₂ based isotropic etching of silicon. This isotropic etch undercut the silicon and released the silicon from the bottom bulk substrate. To make the resultant partially released silicon fabric stretchable, we patterned a sequence of squares interconnected with spiral and serpentine wires (20 µm wide), while the silicon squares had side length of 400 µm. The partially released fabric was then subjected to DRIE. This released the silicon fabric completely and patterned the fabric in the interconnected-squares pattern. The released silicon fabric with devices is easily handled, flexed and stretched.
With 5 µm diameter release holes, the total usable real-estate in each square was reduced by only 8.72 %. Thus, an area of 1.46 × 10⁵ µm² was available on each square for device fabrication. The spirals can be used to interconnect the modules on adjacent squares to complete the microprocessor design for stretchable electronics.

Optical control is a real time, reversible and convenient technology in microfluidic, biomechanical, and electro-mechanical devices. These advantages attracted many attentions of related photoresponsive materials recently. In this paper, we created a new photoresponsive and electrostrictive material by mixing the dielectric polymer P(VDFTrFE-CFE), and the organic photoconductive material TiOPc. The newly developed thin film was composed of P(VDF-TrFE-CFE) with 5%~20% TiOPc, which was started by dissolving P(VDF-TrFE-CFE) powder in a dimethylacetamide (DMAc) solvent before adding TiOPc powder. The film was then formed at 60 degree C on the hot plate before being annealed in a 100 degree C vacuum oven for six hours. The photo-responsibility of this material was validated both by spectrum analysis and corresponding actuator performance.
The photo-responsive band of P(VDF-TrFE-CFE)/ TiOPc was measured by a UV-Vis spectroscope (Lambda 900 spectrometer, PerkinElmer). We coated the composite onto a PET plate with ITO electrode to make the cantilever beams. We then evaporated Au on the cantilever beams to serve as the top electrode. The tip displacement of the cantilever beam was measured by a displacement meter (Keyence, LK-H080). It was found that the tip displacement increases under white light illumination, which could be attributed to the decrease of the TiOPc impedance. As experimental results showed that the tip displacement changes were controlled by the impedance and the leakage current mechanism in the composite, it is clear that the changes are dependent on the TiOPc content. Through parametric studies, we identified that the optimal TiOPc concentration was 10% in P(VDF-TrFE-CFE)/ TiOPc for application as actuators. Moreover in this paper, the photoresponsive actuator is validated by controlling the fluid velocity within the micro-channel though light illumination. More specifically, the composite film actuator was applied to exert force onto a capillary tube to drive the fluid within. It was shown that the water level could be controlled by light illumination intensity and time duration.
In summary, we demonstrated that this new photoresponsive composite P(VDF-TrFE-CFE)/ TiOPc could induce large displacement changes with light illumination in this paper. It has clearly shown that the mechanism and the photoresponsive material can potentially be used to pursue light controlling microfluidic, and related electro-mechanical devices.

As a type of contact electrification in which a material becomes electrically charged after contacting with a different material, triboelectric effect has been known for centuries. Recently, this effect was successfully used as an effective means for harvesting mechanical energy, leading to the invention of the triboelectric nanogenerators (TENGs).
In this work, a coaxial cylindrical structured rotating-TENG is developed to harvest mechanical energy from rotation in analogous to an electromagnetic induction based generator. Based on the coupling of contact electrification and electrostatic induction, relative displacement between two contact surfaces of different triboelectric polarities creates uncompensated triboelectric charges that can drive induced free electrons to flow through the external circuit. A rotating-TENG with a total contact area of 12 cm2 contains multiple strip units that are connected in parallel. The instantaneous short-circuit current (Isc) and the open-circuit voltage (Voc) could reach 90 mu;A and 410 V, respectively, for a TENG with 8 strip units at a linear rotational velocity of 1.33 m/s, corresponding to an instantaneous maximum power density of 36.9 W/m2 and an equivalent average direct current of 45.6 mu;A. Higher output power can be achieved by using a higher density of strip units and/or applying larger linear rotational velocity. At a linear rotational velocity of 1.33 m/s, the TENG can be used as a direct power source for simultaneously powering 90 commercial light-emitting diode (LED) bulbs in real time. Since rotation is one of the most common forms of motion in ambient environment, a fully packaged cylindrical rotating TENG has potential applications in harsh environment, outdoors or even under water to harvest energy from water flow.

With the rapid growth of worldwide energy demand, energy harvesting is of great importance in modern life. Triboelectric nanogenerator (TENG) technology is developing quickly to solve this problem owing to its high output power, inexpensive materials, and ease of fabrication. The TENG has two basic working modes: contact-mode and sliding-mode. Although the basic design and operation of the sliding-mode TENG has been established, and several different device structures for different applications have been demonstrated, there still lacks a systematic theoretical model for this working mode. In this paper, the theoretical model for the sliding-mode TENG is first presented. The kernel part of the theory for TENG is the relationship of three parameters: its output voltage V, its charge accumulated at each electrode Q, and the separation distance x (V-Q-x relationship). The finite element method (FEM) is utilized to characterize the distributions of electric potential, electric field, and charges on metal electrodes. Utilizing interpolation method, the semi-analytical and the approximate analytical models are built from these numerical calculation results. The approximate analytical model is compared with semi-analytical model to verify its validity. Then dynamic output property of the sliding-mode TENG with a load resistance is calculated by combining the basic equation of sliding-mode TENG and Ohm&’s law. Through the dynamic output property calculation result, the formation of 3 working regions of sliding-mode TENG is explained, which include short circuit limit, general working region and open circuit limit. And the optimum resistance for sliding-mode TENG is theoretically explained. Finally, the calculation results are further compared with experimental data and match with experimental data quite well. This proves again that the theoretical model presented by this paper is valid. The theory presented here establishes a solid physical background and clarifies the working mechanism of the sliding-mode TENG and could guide experimental design of high output sliding-mode TENG either as a power source or as a self-powered active sensor. [1]
Reference:
1. Theory of Sliding-mode Triboelectric Nanogenerators
By Simiao Niu, Ying Liu, Sihong Wang, Long Lin, Youfan Hu, Yusheng Zhou, and Zhong Lin Wang
Submitted to Nano Letters

Flexible electronics has been intensively studied as it is a very promising technology for widespread applications owing to its excellent flexibility and light-weight. Various flexible electronic devices such as field effect transistors, integrated circuits and generator have been developed. Attempts have also been made to fabricate microsystems on flexible substrates that are able to provide unique functions normal electronic devices are unable to do.
Surface acoustic wave (SAW) resonators are one of the building blocks for electronic devices, microsensors and microsystems. Owing to their high operating frequency, high sensitivity, simple fabrication process and low cost, SAW devices have been intensively explored for applications as microsensors to measure physical parameters as well as biochemical substances, and as microactuators for microfluidics and lab-on-a-chip. Recently we have successfully fabricated flexible SAW devices on polyimide thin films with excellent performance; here we report the temperature and humidity sensors with high sensitivities of the flexible SAW devices.
The SAW devices were fabricated on a Kapton® polyimide film 100H with 60 mu;m thickness. ZnO piezoelectric films were deposited on the polyimide substrates using a direct-current (DC) magnetron sputtering system. The flexible SAW devices with different wavelengths were fabricated. The one with 12µm wavelength have two resonant modes with the center frequencies at fR~132MHz (Rayleigh mode) and fL ~427MHz (Lamb mode), respectively.
The two resonant frequencies of the SAW devices decreased with increase in temperature linearly regardless of the wavelength of the devices. The Rayleigh mode wave has a temperature coefficient of frequency (TCF) of -440ppm/K, while the Lamb mode has a small TCF of -250ppm/K for the device with a wavelength of 20µm, about an order of magnitude larger than those on rigid substrates. The large TCF of the flexible SAW devices is attributed to the much larger thermal expansion coefficient of the polyimide substrate (20 ppm/K for Kapton, ~3.0 ppm/K for Si), demonstrated the flexible SAW devices are high sensitivity temperature sensors.
The humidity responding characteristics of the Rayleigh mode wave were investigated. The fR shifts significantly to low frequency when the relative humidity (RH) was increased. The flexible SAW sensor exhibited a very good humidity sensing repeatability and the sensitivity increases with decrease in wavelength owing to the improved characteristics of the sensors with shorter wavelengths. At low RHs, the flexible SAW devices have a sensitivity of ~16kHz/10%RH, and increase to ~34kHz/10%RH at ~80%RH without any surface treatment and a polymer absorbent, almost an order of magnitude higher than other SAW sensors made on rigid substrates. Both the sensing results show that flexible SAW devices have great potential for applications not only in flexible electronics, but also in flexible sensors and microsystems.

A silicon-based microreactor with a structure that allows in situ characterization by Attenuated Total Internal Reflection Infrared spectroscopy (ATR-IR) [1] of processes driven by an external electrical field (E-field) is presented. The system is used to study E-field driven surface polarization, electron excitation, and carrier drift in the silicon microreactor. The experimental results provide insight into the electronic processes that can occur in the device, and into potential application in the initiation of novel chemistry [2-4]. Pillar microstructures integrated into the flow channel of the microreactor locally densify the electrical field and allow control regimes ranging from polarization to electron conduction. The interference of these structures with the optical behaviour in the IR wavelength region will be presented, and methods to extract valuable chemical information from the microdevices by overcoming these undesired optical side-effects while flowing NH₄⁺ and CO ₂ will be discussed. Another application, e.g. the study of the dynamics of phonons in silicon, generated while helium gas or water is present under different voltages will be addressed. We believe that this type of microreactor can generate new insights in both fundamental chemistry and physics.
[1] B. L. Mojet, S. D. Ebbesen, L. Lefferts, Chem. Soc. Rev. 39, 4655 (2010)
[2] T. I. Kovalevskaya, Journal of Applied Spectroscopy 65, 462 (1998)
[3] Z.-J. Zhou, X.-P. Li, Z.-B. Liu, Z.-R. Li, X.-R. Huang, C.-C. Sun, J. Phys. Chem. A 115, 1422 (2011)
[4] A. Agiral, H. B. Eral, D. van den Ende, J. G. E. Gardeniers, IEEE Transactions on Electron Devices 58, 3518 (2011)

Since Lade zirconate titanate oxide (PZT) has excellent piezoelectricity, it has high potential for piezoelectric device. Recently, micro electro mechanical system (MEMS)-based micro cantilevers with the PZT films (PZT micro cantilever) have been applied for the actuator and the sensor devices. The tips of the PZT micro cantilevers move up and down with an applied voltage. Driving force of the PZT micro cantilevers is the piezoelectricity of the PZT films. Operation and failure analysis of a PZT micro cantilever is required for performance increasing and reliability. However, there are only some experimental reports on actual device using focused X-ray diffraction. In this study, we have applied the micro-Raman spectroscopy for operation and failure analysis of PZT micro cantilever. Raman spectroscopy has some advantages that it was non-destructive, non-contact, high sensitive measurement and relative high spatial resolution measurement method.
The PZT micro cantilevers have Pt / LaNiO3 / PZT / LaNiO3 / Pt / Ti / SiO2 / silicon-on-insulator (SOI) multilayer structure. PZT films were deposited on LaNiO3 / Pt / Ti / SiO2 / Si substrates by chemical solution method. The piezoelectric micro cantilever was fabricated by the MEMS techniques. The micro-Raman spectroscopy was measured using a 514.5 nm-Ar+ laser for an excitation source. The laser beam was focused at about 1 mu;m in diameter on a sample surface. The Raman spectra from PZT micro cantilever showed typical tetragonal PZT phase with no second phase. The intensity of A1(TO)-modes was changed with increasing applied voltage to micro cantilever. It indicated that the a-domains was switched to c-domains with increasing applied voltage. Some top electrodes were peeled from the PZT film under applied voltage and the PZT micro cantilever did not operate. The Raman spectrum of PbO was observed at defect point. This result indicated that the existent of PbO in PZT was decreased the reliance on PZT micro cantilever device. These results were important information about performance increasing and reliability for the PZT micro cantilever device.

A solid-state nanopore device that consists of a nanoscale hole sculpted in a thin membrane has been used as a powerful tool of electrical detection of single-molecules and -particles. In recent year, graphene nanopores attract a great deal of attention in the field of nanotechnology [Nature, 467, 190 (2010)]: the atomic-scale thickness of graphene provides ultra-high spatial resolution and makes itself an excellent substrate for single-molecule sensing. Despite the potential capability, fundamental understanding of the sensing mechanism in such low thickness-to-diameter aspect ratio pores is yet to be established.
In this respect, our group have studied single-particle translocation through a low aspect ratio nanopore mimicking the graphene nanopore architecture [ACS Nano, 6, 3499 (2012)]. In the previous work, we found that the ionic transport outside a low-aspect-ratio pore changes little during single-particle translocation, enabling single-particles by their volume through a resistive pulse technique. In this study, we applied it for discriminations of single-nanoparticles by their surface charge states. In principle, discrimination of equi-sized particles is difficult for nanopore sensors as they exclude essentially the same number of ions in nanopores during translocation. In contrast, I will present in my presentation that low-aspect-ratio nanopores. can discriminate nanoparticles not by their volume but by their surface charge density.
In experiments, we fabricated a nanopore in a thin Si3N4 membrane using an electron beam lithography process. The pore diameter was 0.7mu;m. We performed ionic current measurements by applying an electrophoretic voltage to polystyrene beads (diameter = 0.5 mu;m) with different modifications dispersed in a TE buffer (pH 8.0). Current spikes were observed when particles translocate through the pore because of temporal ionic current blockade by the polystyrene beads. By comparing the height and the width of the spike signals obtained for the nanoparticles modified with COOH, NH2, S, and OH, we showed that it is possible to identify equi-sized particles by the difference in the surface charge density. This method may open the prospect of application to cell sorter and virus screening.

We discuss our recent findings on the use of MEMS, such as microcantilevers (MCs), entirely based on transition metal oxides (TMOs) for the development of sensors and devices. TMOs show a wide range of functionalities (superconductivity, ferroelectricity, piezoelectricity, magnetic, dielectric, semiconductivity) that come from the strong interplay between charge, spin and lattice degrees of freedom [1]. MCs and TMOs are already separately used in different fields of sensors. However, the wide possibilities offered by freestanding structures made of TMOs are not fully explored. MCs are fabricated starting from epitaxial oxide multilayers deposited by PLD and then machined by conventional microlithography. We selectively remove different layers using proper acids, thus leaving free-standing elements. For example we can the fabricate MCs made of crystalline SrTiO3 (STO) using a sacrificial of (La,Sr)MnO3 (LSMO) [2,3] and crystalline free-standing structures made of LSMO or VO2/TiO2 [4] films can be also realized. The suspended elements can also be used as flexible substrates for high-temperature deposition of additional TMO layers on top. After a brief discussion of the main fabrication protocols, we show three different applications of TMO-based free-standing structures.
Strain-tuned devices: TMO-based MCs can be bent mechanically or by applied voltages obtaining an uniaxial strain at surfaces. Strain experiments performed on (La,Sr)MnO3 [2] and (La,Sr)CoO3 [3] films show reversible modulation of their electrical resistance. We analyze opportunities for tuning of the physical properties of TMOs by strain and the actuating mechanisms of the MCs by localized electrostatic gate fields.
Joule-heated devices: Joule-heating in free-standing structure can be efficiently used to increase temperature to very high values in short time with small amount of power, especially when shrinking device dimensions. Hence, we can tune the temperature of TMOs by electrical current and control their phase transitions (magnetic, electrical, structural). We discuss recent results on LSMO films, where we can easily achieve temperatures above 600°C with few mW of power or trigger its para/ferromagnetic transition. Transient analysis and device operation have been simulated by Finite Element Analysis and bolometric applications are also discussed.
Resonant devices: TMO-based MCs can be set into oscillation by mechanical excitations or electric fields. We discuss opportunities for realizing novel mechanical detectors by monitoring changes of TMO physical properties under external stimuli or during phase transitions through detecting shifts of their mechanical resonances. Preliminary measurements on LSMO free-standing cantilevers are shown.
1 D.G. Schlom, et al. J. Am. Ceram. Soc. 91, 8, 2429, 2008
2 L. Pellegrino, et al. Adv. Mater. 21, 2377, 2009
3 M. Biasotti, et al. Appl. Phys. Lett. 97, 223503, 2010
4 L. Pellegrino,et al. Adv. Mater. 24, 2929-2934, 2012

The huge mismatch in mechanical properties of conventional electronics and the human body causes difficulties in the development of new wearable biomedical instruments. Flexible and stretchable electronics have emerged in a pallet of new technologies for realizing smart sensors and actuators for biomedical applications.In contrast to conventional electronics ,which are rigid and planar, flexible electronics can be bent, stretched and twisted.Many researchers have exploited organicelectronic materials or extremely thin inorganic materials or combinationsin order to address the mechanical mismatch problems.One of the biggestissues to be solved withnew flexible electronic structures is simultaneously achieving the performance and reliability of foundry electronics processes. Organic materials are flexible but they have uncertain reliability. In contrast,inorganic materials (like silicon or gallium arsenide)have been used in electronic devices for decades.Therefore, we focus on stretchable electronics through a combination of inorganic material (like Si) and silicon-based organicpolymer (Polydimethylsiloxane(PDMS)). We examine the response of wiring when embedded in these flexible polymers to determine that strain that can be accommodated. We accomplish this through finite element analysis coupled with experimental polymer-wiring composites to determine when cracking and/or delamination occurs. We believe that this work has implications in a diversity of fields including stretchable electronics, biomaterials, and composite materials.

A technique developed for studying the internal friction and energy loss of nano-scale thin metal films on substrate is presented. The test microstructure was designed on the triangular cantilever beam and fabricated by the standard C-MOS processes, which can improve stress distribution non-uniform problem of conventional cantilever beam. The thickness of deposited film on its surface could reduce to several nanometers. Nanocrystalline thin metal films are widely used in the electronic interconnections or MEMS structures. Nanocrystalline Au, Ag, Al and Cu thin film with thickness from 50 nm to 200 nm were performed to observe its internal friction and energy loss response under high vacuum conditions. We measured energy loss through decay of oscillation amplitude of a vibrating structure following resonant excitation. We closely examine those film microstructures and grain sizes with respect to the dynamic properties of films. The results show the measurement system used here can accurately measures the energy loss of thin film. The internal friction measurement results provided evidence for the grain boundary motion and dislocation motion in the nanoscale thin films. Moreover, the length scale dependence on loss mechanism of tested films was observed.

Molecular imprinting is a technique for synthesis of polymers with highly specific recognition sites for target molecule (1). The functional groups of template molecule (for which selectivity is expected) are involved in the formation of molecular interactions with the monomer(s) prior to a polymerization process. After removing of template molecules, we obtain three-dimensional cavities with spatially oriented functionalities; this enables their use in sensing technologies. An important future challenge consists in performing a direct detection of analytes by coupling the recognition and the detection steps into one complete sensor (2). Thus, this kind of sensors can be used to real time monitor analytes in environmental or health applications.
More than a linkage of MIP to an optical transducer, recent studies associated MIP hydrogels with opals (3). Our work aims at combining MIP with opals in a sensor by using a new deposition technique to build opals. The used online-coating device is a novel online-process enabling to deposit large areas of organized monolayers of nanoparticles on a substrate (4). The device was adapted to deposit successive layers, making opals with a controlled and uniform number of beads layers. Five layers opals were built with beads from 300 to 1000nm on polymer (PMMA for example) substrate. Then, MIP pre-polymerisation mixture was infiltrated using dip-coating to give a thin and regular layer of polymer. Dissolution of silica beads and template extraction made it possible to obtain polymeric matrix as an inverse opal, which presents photonic crystal properties. Further template recognition by the matrix leads to a polymeric swelling which rhymes with a shift in the Bragg diffraction peak.
In this work, we focus on the novel synthesis of MIP inverse opal structure for detection of pesticide (2.4D: 2,4-Dichlorophenoxyacetic acid ), its characterizations and its sensor properties. We will underline the potential industrialization of such built sensors.
(1) S. Piletsky, A. Turner, Molecular Imprinting of polymers, Landes-Bioscience, 2006
(2) L. I. Andersson, et al., Tetrahedron Lett. 29, 1988, 5437-5440
(3) N. Griffete, et al., Langmuir, 2012, 28, 1005-1012
(4) Z.Tebby et al. Multi-Material Micro Manufacture, 2010 November, Oyonnax. Research Publishing, 2010

For a few decades, the piezoelectric materials have been considered as important functional materials for various applications, from the microphones to the high technology scanning electron microscopes, actuators, sonar sensors, cell phones, MEMs and etc. Lead-based phases have dominated almost all these applications. However, lead based materials have been found as toxic and hazardous materials and will be prohibited to use within a short period of time. Among various lead-free piezoelectric materials, BZT-BCT based phases are good candidates instead of Pb-based materials due to having a tricritical point compared to other lead-free alternatives. In this study, lead-free Ba(Ti_0.8Zr_0.2)O₃-(Ba_0.7Ca_0.3)TiO₃ (BZT-BCT) thin films were successfully prepared on Pt/TiO₂/SiO₂/Si substrates using CSD (Chemical Solution Deposition) method. The effect of sintering temperatures on microstructure, dielectric and ferroelectric properties were determined using a systematic study. Among the various high-quality BZT-BCT thin films with uniform thickness, the optimum room temperature dielectric and ferroelectric responses were observed for films annealed at 800 °C for 1 h. The thickness was kept constant for all measurements as 500 nm (triple layered films). The average grain sizes were found around 60 nm for samples sintered at 700, 750 and 800 °C. Both morphological and structural analyses showed that BaTiO₃ based Ba(Ti_0.8Zr_0.2)O₃-(Ba_0.7Ca_0.3)TiO₃ composition had homogeneously nucleated growth mechanism throughout the film. This type of growth mechanism yields polycrystalline films as observed in structural analyses and dense-small grained morphology as observed in morphological analysis. BZT-BCT thin films sintered at 800 °C yielded effective remnant polarization and coercive field values of 2.9 mu;C/cm² and 47.5 kV/cm, together with a high dielectric constant and low loss tangent of 365.6 and 3.52 %, respectively, at a frequency of 600 kHz due to pure perovskite phase showing full crystallization and minimum surface porosity obtained at this temperature.

The Casimir Force arises due to quantum mechanical vacuum fluctuations in the electromagnetic field. A 1/r⁴ attractive force sets the Casimir Force apart from more conventional effects. It has been shown to cause nonlinear motion in metallic and semiconducting microresonators when surfaces are brought together to within tens of nanometers. We report progress on a study based on these nonlinearities using a 500mu;m×500mu;m polysilicon plate suspended 2mu;m above two polysilicon electrodes by torsion rods. The structure can then be excited by an AC bias at the plate&’s torsional resonance frequency. If a gold plated polystyrene sphere (sim;200mu;m radius) is biased and suspended close to the plate (sim;100nm) the frequency response of the resonator shifts and becomes nonlinear exhibiting clear hysteresis. By driving the sphere at twice the resonance frequency of the plate we plan to show parametric amplification induced by quantum mechanical effects. Bistability produced by the anharmonic oscillator is useful in applications such as switching when hysteresis can be controlled by varying the initial distance between the sphere and plate as well as the displacement amplitude of the sphere. We believe that by utilizing quantum fluctuations, force detection sensitivity and background noise could be greatly improved upon.

PbTiO3-base piezoelectric films have been attractive for many applications such as sensors and actuators in microelectrical mechanical systems (MEMS) because of their high piezoelectric response. Pb(Mg1/3Nb2/3)O3- PbTiO3 showed larger piezoelectric coefficient than widely investigated Pb(Zr,Ti)O3. However, the reports on film form are still small number compared with those of Pb(Zr,Ti)O3 due to the difficulty of the film formation. Especially, thickness dependency of the electrical and electromechanical properties, especially thick film region beyond 500 nm is still unknown in spite of the importance for various applications.
In the present study, we compared the electrical and electrometrical properties of high quality epitaxial Pb(Mg1/3Nb2/3)O3 [PMN] and 0.6Pb(Mg1/3Nb2/3)O3-0.4 PbTiO3 [PMN-PT] films grown on SrRuO3//SrTiO3 substrates by metal organic chemical vapor deposition. 0.6Pb(Mg1/3Nb2/3)O3-0.4 PbTiO3 composition shows the maximum relative constant at room temperature as film form of Pb(Mg1/3Nb2/3)O3- PbTiO3 system.
Epitaxial films were successfully grown for wide range of film thickness from 500 to 3000 nm. Lattice parameters did not strongly depended on the film thickness for both films. Maximum relative dielectric constant (εmax.) and the temperature showing maximum relative dielectric constant (Tmax.) were about 3500 and 400 K, respectively for PMN-PT films and were almost independent of the film thickness. This is contrast with the clear thickness dependencies of εmax and Tmax. for both of (100) and (111) - oriented PMN films. These results show that PMN-PT films did not show noticeable thickness dependency in thick film region. This is very good for the piezo MEMS application of PMN-PT films.

As a novel candidate for smart materials, La modified Pb(Zr,Sn,Ti)O3 perovskite antiferroelectrics, has attracted much attention in past decays because of its high electric-field-induced longitudinal and volume strains(above 1.7%) accompanying a phase transition between tetragonal antiferroelectric [AFE(T)] to rhombohedral ferroelectric[FE(R)][1]. Bulk single crystal of La modified antiferroelectric Pb(Zr,Sn,Ti)O3 (PLZST), close to the morphotropic phase boundary (MPB), has been grown successfully from a PbO-PbF2-B2O3 flux in our lab [2,3]. PbF2-evaporation-dependent was considered as main result of PLZST single crystal growth in PbO-PbF2-B2O3 flux system. Microstructure and domain of PLZST single crystal were studied by Raman and in situ polarized light microscope (PLM). Polarization studies indicate that the dielectric and ferroelectric properties of PLZST single crystals are strongly orientation-dependent. Electric-field-induced strain and Ps along <111> direction are much higher than that of other else.
Reference
1. W.Y. Pan, W.Y. Gua, L.E. Cross, “Transition speed on switching from a field-induced ferroelectric to an antiferroelectric upon the release of the applied electric field in (Pb,La) (Zr,Ti,Sn)O3 antiferroelectric ceramics”, Ferroelectrics, 1989, 99(1), 185-194
2. Y.Y. Li, Q. Li , L. Wang, Z. Yang, X.C. Chu, “Growth and characterization of (Pb,La)(Zr,Sn,Ti)O3 single crystals” Journal of Crystal Growth, , 318(2011), 806-864
3. Y.Y. Li, Q. Li, Q.F. Yan, Y.L. Zhang, X.Q. Xi, X.C. Chu, W.W. Cao, “Phase transitions and domain evolution in (Pb, La)(Zr, Sn, Ti)O3 single crystal,” Applied Physics Letters, 2012, 110(13), 132904

Recent experimental results have shown that substitution of Bi atom with Sm atom in BiFeO3 film can lead to an R3c-Pnma phase transition, and the Sm-doped BiFeO3 film shows ultra-high electromechanical response near the morphotropic phase boundaries. We constructed a thermodynamic theory for this system, including order parameters of polar mode and out of phase and in phase oxygen octahedron tilting. The composition-temperature phase diagram is plotted and compared with experiments. It is shown that near the R3c-Pnma phase boundary, an electric field can induce a Pnma-R3c phase transition. Calculated out of plane lattice parameter change from Pnma-R3c phase transition is shown to be much larger than that from intrinsic piezoelectric response of R3c phase, which explains the enhanced piezoelectric coefficients observed in experiments. We also employ the phase field method to investigate how ferroelectric R3c domains grow in Pnma matrix.

We are investigating the characteristics of microfabricated PZT/FeGa multiferroic cantilevers. The cantilevers can be driven by AC or DC magnetic and electric fields, and the device response can be read off as a piezo-induced voltage. We can use the multiple input parameters to operate the devices in a variety of manners for different applications. They include electromagnetic energy harvesting, pulsed triggered nonlinear memory devices, and parametric amplification. Electromagnetic energy harvesting was demonstrated using a variable load resistor attached to a cantilever, and the peak power of 0.7 mW/cm3 was captured with a load of 12.5 kOmega; when the input AC magnetic field is 1 Oe. Due to the competition of anisotropy and Zeeman energies, the mechanical resonant frequency of the cantilevers was found to follow a hysteresis behavior with DC bias magnetic field applied in the cantilever easy axis. Another important property of our devices is manifested in our ability to control and tune the occurrence of nonlinear bifurcation in the frequency spectrum. The resulting hysteresis in the frequency spectrum can be used to make switching devices, where the input can be DC electric and magnetic fields, as well as pulses of AC fields. We have also demonstrated parametric pumping of the response from an AC magnetic field using frequency-doubled AC electric field. The detected response to the AC magnetic field can be enhanced with a gain as high as 950, when the pumping voltage is very close to a threshold voltage. The latter application can be used to help increase the sensitivity of the cantilevers used as a magnetometer.

Hydrogen, H2, has vital applications in the fuel cell, aeronautic, metallurgy, semiconductor, petroleum and pharmaceutical industry. There are, however, risks associated with its use as it is can be an explosive gas, and a colorless, odorless asphyxiant. Therefore, a fast, reusable and reliable sensor is needed to limit accidental deaths and injuries caused by H2 leaks. Much attention has been given to Pd for H2 sensing due to its ability to absorb approximately 900 times its own volume of H2 by forming a reversible PdHx phase. Current technologies utilizing this PdHx phase typically employ thin films deposited via sputter coating, pulsed laser deposition (PLD) or chemical vapor deposition (CVD) from metal targets. Such films while capable of detecting H2, suffer from stress induced deformations upon repeated cycling. We present a method which avoids the use of thin films, expensive deposition techniques and time intensive fabrication routes. Nanoparticle arrays provide a higher surface-to-volume ratio which leads to more adsorption sites, an enhanced sensitivity and the ability to withstand multiple H2 cycling without significant degradation. These properties can lead to sensors which are reusable and have longer shelf lives. Preliminary research suggests that coupled Au and Pd nanoparticles can track changes in the partial pressure of H2 by correlating the plasmonic response to the uptake of H2. Metallic Au nanoparticles possess unique optical properties. For instance, the plasmon band of Au nanoparticles is a sensitive probe to changes in the local environment, e.g., the dielectric properties of the medium. We expect that arrays of nanometer sized dewetted Au bridged by Pd nanoparticles will provide an ideal platform for future nano-H2 sensor technologies. Discrete dipole approximation simulations are performed in parallel to validate the experimental findings. The influence of morphology (i.e., hollow vs solid), shape and spacing of the dewetted Au on the H2 induced plasmonic shift will be reported.

Multiwall carbon nanotubes (MWCNTs) are elongated anisotropic molecular size cylinders that can form a liquid crystal (LC) phase in lyotropic solutions. When dispersed in LCs their nematic directors couple. The liquid crystal properties of the carbon nanotubes are one of the few ways of controlling their orientation and order parameter. This approach can be used to create microelectromechanical devices, for example electric or optical nano-switches. We studied the effects of multiwalled carbon nanotubes at low concentrations (0.01 wt %) on the Freedericksz transition of a 4-Cyano-4&’-pentylbipenyl (5CB) liquid crystal using transmission ellipsometry. We calibrated the altitudinal angle of CNTs as a function of the electric field and directed the azimuthal angle which gave us complete control of the 3D orientation of the CNTs. Our results show that in the presence of CNTs the voltage and width for the Freedericksz transition are reduced by close to a factor of two as compared to the control electro-optic cell without CNTs. Our results of molecular dynamics simulations to understand the reasons for the downshift of the transition voltage show that polarization effects in the nanotubes in external electric field are partially responsible for the downshift.

Piezoelectric microelectromechanical-systems (MEMS) are widely used for actuators, sensors and other devices. Pb-contained piezoelectric materials such as Pb(Zr,Ti)O₃ [PZT] and Pb(Mg,Nb)O₃-PbTiO₃ [PMN-PT] with superior piezoresponse and electromechanical properties are commonly used in piezo-MEMS. However, Pb-based materials are toxic, and there is an urgent need to replace those materials. In this work, we propose Pb-free (Bi,Sm)FeO₃ [BSFO] as a new piezoelectric material which has a perovskite structure. Epitaxial BSFO films prepared on (100)cSrRuO₃//(100)SrTiO₃ substrates by pulsed laser deposition display robust piezoresponse at the morphotropic phase boundary composition. The origin of piezoelectricity is investigated by in-situ synchrotron-HRXRD under pulsed applied electric field. We find that the large piezoresponse in BSFO at this composition originates from an extrinsic effect of field-induced phase transition from an antiferroelectric phase to a ferroelectric phase. Maximum longitudinal piezoresponse (d₃₃) of approximately 240 pm/V was detected based on displacement of diffraction peaks measured under electric field.

Solar power provides unique promise for the next energy revolution; however, current commercially available photovoltaic (PV) solar power systems have significant limitations in fulfilling this vision. First of all, even with recent dramatic price reductions, solar power is still not cost competitive in most energy markets without government subsidies. Furthermore solar power is limited to locations that can utilize heavy, rigid solar panels. If these limitations are to be overcome, there needs to be a significant change in the cost, performance, and functionality of solar PV. MEMS technologies offer an avenue to utilizing scale effects that exist within solar PV cells, modules, and systems where new functionality, improved performance, and reduced costs are possible as solar cells are reduced in size to the micro-scale domain. By taking advantage of these scaling effects we have demonstrated prototype solar power cells and systems with unprecedented functionality and performance. We call this microsystem-based approach to PV “Microsystem Enabled Photovoltaics (MEPV).”
Within our program, we have created crystalline silicon solar cells with cell conversion efficiencies up to 14.9% with a thickness of 14 to 20 microns; a 10X reduction in Si material usage with comparable performance to commercially available c-Si solar cells. These cells have micro-scale lateral dimensions, typically from 250 to 750 microns across. We have also demonstrated 4 micron thick GaAs cells with similar lateral dimensions with efficiencies above 20% and are working to develop other III-V micro-scale cells.
We have created interconnected modules of over 500 c-Si microscale cells through a parallel assembly technique that have demonstrated extremely high flexibility (~ 1 mm bend radius), a module conversion efficiency of 14%, and a specific power of >440 W/kg; providing a highly flexible, high performance, lightweight PV solution that is not available with any other PV technology. The unique functionality of this flexible PV module is ideal for mobile solar power applications.
We have also exploited scale benefits that manifest themselves through concentrated photovoltiac (CPV) systems where lenses focus light onto cells to reduce the required cell area and thus reduce cell costs. By using very small cells, we have demonstrated prototype CPV systems that have a 1 cm focal length instead of the 10 to 50 cm focal lengths of current commercial systems. A short focal length allows the creation of a CPV module that is essentially the same thickness as flat-plate one-sun modules but with the performance and cost advantages of CPV. This approach provides significant module and balance of system cost reductions relative to wafer silicon, thin film, or traditional CPV systems thus providing the potential to make solar PV power cost competitive with retail grid prices in most of the United States.

Fabrication of widely used complex 3D microstructures like microlenses, microtrapezoids, microtips etc with controlled dimensions in minimized process steps and least complexity has always been a target for many researchers over the decades. In order to fabricate such kind of patterns, intricate and expensive processes like RIE, stereo lithography, UV micro stamping etc have been utilized, which are not only costly but also have limitations in terms of controlling the pattern dimensions of these microstructures.
In this study, we have proposed a simple, convenient, cost effective and commercially applicable method, double diffuser lithography(DDL), for the fabrication of such complex 3D microstructures by adding a pair of diffusers in conventional photolithography process and examining the effects of change in exposure energy on this proposed technique. The pair of diffusers rendering the Gaussian and Lambertian scattering phenomena, when inserted in the photolithography tool above the photomask in the passage of incident UV light diffracts the incident rays of light at wide angles before approaching the photoresist. Keeping the photoresist thickness and mask pattern dimensions constant, PR patterns with µm to sub-µm dimensions were conveniently fabricated by just changing the exposure energy of the incident light and the combination of the diffusers used. The increase in exposure energy of the UV light causes a decrease in the dimensions of the microstructures and vice versa, moreover controlling the height and shape of patterns by changing the type of the diffusers. The morphology of the fabricated patterns was analyzed using the FE-SEM images and 3D-profiler data.
With the help of this facile, congruous and coherent DDL technique, the above mentioned complex 3D microstructures with controlled dimensions have been successfully fabricated eliminating the need of large number of complex and expensive processes.

We report a microelectromechanical systems (MEMS)-based heat engine based on pyroelectric energy conversion from a 200 nm thick BaTiO3 film. The device consists of a microfabricated platinum heater and SrRuO3 microelectrodes, which enable fast cycling of temperature and electric fields. The device produces a maximum power density of 30 W/cm3 from the pyroelectric Ericsson cycle when operated for a temperature range of 20 - 120 °C, an electric field range of 100 - 125 kV/cm, and at cycle frequency 3 kHz. This power density is much higher than has been previously reported for pyroelectric conversion, owing to the fast cycling times enabled by the MEMS-based platform. For example, previous studies have examined pyroelectric energy conversion in bulk samples and thick polymer films with cycle frequency lower than 0.1 Hz, reaching power density as high as 0.1 W/cm3 [1]. When operated at 3 kHz, our device produces an energy density of 0.01 J/cm3, which is much lower than has been reported for bulk samples [2]. The low energy density is due to the relatively low pyroelectric coefficient of the BaTiO3 film (~70 µC/m2K) and large dielectric losses.
[1] G. Cha and Y.S. Ju, Sensors and Actuators A, 189, 100-107 (2013).
[2] F.Y. Lee, H.R. Jo, C.S. Lynch, and L. Pilon, Smart Mater. Struct., 22, 025038 (2013).

This talk is about nanoporous ferroic BiFeO3 and the role of nanoporosity on the ferroic response towards optimised performance for applications as smart materials. Micro- and nano- electromechanical systems (MEMS/NEMS) encompass currently much more than silicon based microelectronics. Promising MEMS and NEMS with added functionalities include nanomaterials, smart materials (piezoelectric and ferroelectric materials) and biomaterials, among others. Within the family of ferroelectric materials, BiFeO3 with both ferroelectric and antiferromagnetic properties at room temperature stands out as a multifunctional material. Magnetoelectric effect (coupling interaction between ferroelectric and magnetic order parameters) in multiferroic materials arouses potential applications in information storage, spintronics and multiple-state memories. The best properties for BiFeO3 were obtained in epitaxial thin films with a Pr value of 90 micro C/cm2 and a magnetoelectric coefficient of 3 V cm Oeminus;1.[1] The large Pr value makes BiFeO3 one of the most promising candidate for capacitors in high-density memories. However, some drawbacks, including the weak magnetic properties at room temperature and high dielectric loss restrict the application of BiFeO3.
So far, the strategy to improve BiFeO3 properties has been focused on Bi and Fe atoms substitution.[2] However the properties of materials are also related to structure and microstructure. In common sense, porosity is adverse to the electromecahnical properties being responsible for the decrease of their ferroelectric and dielectric properties. However, our recent results indicate that BiFeO3 thin films with ordered nanoporous structure have stronger magnetic response than dense films.
In this work, nanoporous BiFeO3 crystallites were fabricated by low- temperature hydrothermal process using a block co-polymer (PS-b-PEO) or Cetyl Trimethyl Ammonium Bromide (CTAB) as templates. The polymer builds micelles in the solution, which will be orderly distributed in BiFeO3 matrix, and removed by thermal treatment leaving an ordered array of nanopores.[3] The morphology of nanoporous BiFeO3 crystallites was observed with transmission electron microscopy (TEM) and high-resolution TEM. Nitrogen adsorption measurements and mercury intrusion porosimetry characterization were carried out. The effects of nanoporous morphology on the crystal structure, magnetic and dielectric properties of BiFeO3 were systematically investigated. The relationships between structure, material properties and processing were established for nanoporous BiFeO3.
References
[1]J. Wang, J.B. Neaton, H. Zheng, V. Nagarajan, S.B. Ogale, B. Liu, D. Viehland, V. Vaithyanathan, D.G. Schlom, U.V. Waghmare, N.A. Spaldin, K.M. Rabe,M. Wuttig, R. Ramesh, Science 299 (2003) 1719.
[2] Hiroshi Ishiwara, Current Applied Physics 12 (2012) 603-611
[3]Clément Sanchez, Cédric Boissière, David Grosso, Christel Laberty, and Lionel Nicole, Chem. Mater. 20(2008)682-737.

Microtransfer printing is an emerging fabrication process that allows thin (0.1 to 10 mu;m) semiconductor and metal layers to be aligned, stacked, and bonded. The process utilizes a stamp to retrieve and handle the released patterned layers and bonding between the layers is achieved through direct adhesion and subsequent annealing. While microtransfer printing has been widely used in the fabrication of flexible electronics, there are only a handful of demonstrations for its use in the construction of micromechanical structures. Microtransfer printing provides design flexibility in MEMS by allowing for the integration of layers of various thicknesses, permitting the integration of a broad range of materials at different levels, and enabling the control over residual stresses in thick and multi-layer devices. The use of microtransfer printing for MEMS manufacturing requires retrieval and release of patterned thin structures using a stamp, micron-level alignment during stacking, and robust bonding between layers. In this talk, we will present our work in each of these areas. This includes the development of unique stamp geometries for transfer of patterned structures and experiments demonstrating controlled optical alignment between layers using microtransfer printing. In addition to these results, the fabrication and characterization of a transfer-printed MEMS 2D scanner will be described.

Optomechanical systems offer one of the most sensitive methods for detecting mechanical motion using shifts in the optical resonance frequency of the optomechanical resonator. Presently, these systems are used for measuring mechanical thermal noise displacement or mechanical motion actuated by optical forces. Meanwhile, electrostatic capacitive actuation and detection is the main transduction scheme used in RF MEMS resonators. The use of electrostatics is convenient as it allows direct integration with electronics used for processing the RF signals.
In this talk, I will introduce a method for actuating an optomechanical resonator using electrostatic forces and sensing of mechanical motion by using the optical intensity modulation at the output of an optomechanical resonator, integrated into a monolithic system fabricated on a silicon-on-insulator (SOI) platform. I will discuss new applications enabled by this hybrid system including Opto-Acoustic Oscillators (OAO) and Opto-Mechanical Gyroscopes (OMG).

Fixed-fixed beams are ubiquitous MEMS structures that are integral components for sensors and actuation mechanisms. However, residual stress inherent in surface micromachining can affect the mechanical behavior of fixed-fixed structures, and even can cause buckling. A self-tensioning support post design that utilizes the compressive residual stress of trapped sacrificial oxide to control the stress state passively and locally in a fixed-fixed beam is proposed and detailed. The thickness and length of the trapped oxide affects the amount of stress in the beam. With this design, compression can be reduced or even converted into tension. An analytical model and a 3D finite element model are presented. The analytical model shows relatively good agreement with a 3D finite element model, indicating that it can be used for design purposes. A series of fixed-fixed beams were fabricated to demonstrate that the tensioning support post causes a reduction in buckling amplitude, even pulling the beam into tension. Phase shifting interferometry deflection measurements were used to confirm the trends observed from the models. Controlling residual stress allows longer fixed-fixed beams to be fabricated without buckling, which can improve the performance range of sensors. This technique can also enable local stress control, which is important for sensors.

Microelectromechanical systems (MEMS) made of thin film silicon open up the possibility of using MEMS in large area applications. Low temperature processing allows the use of flexible, transparent or low-cost alternative substrates. In addition, thin-film silicon MEMS are potentially compatible with CMOS backend processing allowing monolithic integration of MEMS with ICs.
Although the literature has extensive studies of the growth, structure and optoelectronic properties of amorphous and nanocrystalline silicon thin films, their mechanical properties are relatively little studied. A clear understanding of these mechanical properties and how they impact the electromechanical properties of MEMS using these films as structural layers is required to properly design and characterize thin film silicon MEMS resonators.
The total mechanical stress in a thin film results from contributions of the intrinsic stress, related to its as-deposited internal structure, and of thermal stress, related to the different thermal expansion coefficients between the film and substrate.
In this work, bridge resonators were fabricated by surface micromachining on glass substrates in a four layer lithographic process, at temperatures of 100°C. The n-type a-Si:H structural layer is deposited by rf-PECVD, using a mixture of gases of silane, hydrogen and phosphine. The air gap between the resonator and the bottom electrode is defined by a sacrificial layer released by wet etching. The resonators are electrostatically actuated in vacuum and its displacement is optically and electrically detected.
The thermal stress component is isolated by fabricating resonators with similar structural layer, deposited at 100°C, and using three different sacrificial layers (aluminum, silica, and photoresist) with different thermal expansion coefficients. Different thermal strains between the structural and sacrificial layers during microfabrication generate stresses that, after sacrificial layer removal, act on the beam affecting its performance. The effect of the sacrificial layer, geometrical dimensions, and design of the anchors on the resonance frequency and quality factor of the microresonators will be described. This work shows that consideration of the mechanical properties of the sacrificial and structural layers can be used to engineer the dynamic response of thin film MEMS resonators.
It was also found that post processing thermal annealing sharply increase the quality factor of the resonators made with silicon films with tensile type of stress (from Q~2000 to Q~8000, after a 1h annealing at 150°C). The tensile stress of the bulk silicon film was shown to increase after the thermal annealings. These variations were related to changes on the hydrogen content within the film measured by FTIR. The implementation of a process to prime the microstructure after fabrication is essential to ensure the optimization and reproducibility of the MEMS performance and will be described.

Exciting progress has been made recently on magnetoelectric sensors with electromechanical resonance frequencies < 200kHz, which are highly sensitive magnetometers based on magnetic control of electrical polarization in magnetic/piezoelectric magnetoelectric heterostructures [1-7]. An optimum DC bias magnetic field is required for magnetoelectric sensors to reach maximum magnetoelectric coupling coefficient and sensitivity, which results in additional source of noise and makes it hard for integration. Exchange bias has been most recently introduced to achieve magnetoelectric sensors at zero bias magnetic field [7], which made possible self-biased thin film magnetoelectric sensors at the cost of a significantly reduced magnetoelectric coupling coefficient.
In this talk, we demonstrated a self-biased magnetoelectric sensor based on a 215MHz AlN/(FeGaB/Al2O3)*10 magnetoelectric nanoelectromechanical systems (NEMS) nano-plate resonator with a highly piezomagnetic and low loss FeGaB/Al2O3 multilayer. A new mechanism was used in the NEMS magnetoelectric sensor by measuring the DC magnetic field dependence of the admittance of the magnetoelectric NEMS nano-plate resonator, of which the DC magnetic field sensitivity is linearly proportional to the resonance frequency. By increasing the resonance frequency of the NEMS magnetoelectric sensor to 215MHz, a low limit of detection for DC magnetic fields of ~600pT in unshielded environment was achieved at zero bias magnetic fields [8]. The efficient on-chip piezoelectric actuation and sensing of a high frequency bulk acoustic mode of vibration in a nano-plate structure enables extremely sensitive magnetometers. The nano-scale, self-biased and CMOS compatible fabrication process make the NEMS magnetoelectric sensor high-potential in application of real magnetic sensor device.
1. C.W. Nan, et al., J. Appl. Phys. 103, 031101, (2008).
2. N.X. Sun, and G. Srinivasan, Spin 02, 1240004, (2012).
3. M. liu, et al., Adv. Mater., 25, 1435 (2013).
4. T. Nan, et al. Appl. Phys. Lett. 100, 132409, (2012).
5. M. Liu, et al., Adv. Funct. Mater. 21, 2593, (2011).
6. J. Lou, et al., Adv. Mater. 21, 4711, (2009).
7. E. Lage, et al., Nature Materials 11, 523-529, (2012).
8. T. Nan, et al. Scientific Reports, 3, 1985 (2013).

Relaxation oscillations occur in a circuit where a DC source is connected to a load resistor in series with a parallel combination of a capacitor and resistance switching element. The capacitor charges while the switching element is in the high resistance state until the capacitor voltages reaches a threshold value at which the switching element changes to its low resistance state and the capacitor discharges. Capacitor voltage is low after the discharge allowing the switching element to change back to the high-resistance state, thus oscillations are produced. We have observed relaxation oscillations in a circuit in which the switching element is a nanocrystalline Si microwire (length: 1-5 µm) where the high resistance state is the solid, crystalline phase and the low resistance state is the liquid phase¹. Oscillations as large as 2-20 mA have been observed experimentally with frequencies on the order of 1 MHz. Additional experiments have been performed on silicon-on-insulator wires (length < 1 µm) where oscillations with frequency of ~ 7 MHz are achieved with three times smaller supply voltage and ~40 times smaller peak power². Oscillation amplitude and frequency are determined by load resistance, wire resistance in solid state, pulse voltage, capacitance, and the time scale of melting and resolidification. The devices typically fail after ~1000 cycles of solid-liquid oscillations due to significant electromigration of the Si⁺ ions in the liquid state if they are not encapsulated.
In some cases, we observe a pulsed waveform of the current (0-7 mA) during the relaxation oscillations, suggesting the wire is completely disconnecting and reconnecting during the melting and freezing process which may be attributed to the combination of a deformation of the wire during a pervious melting/freezing cycle³ and the 5% volume decrease of Si upon melting⁴. The pulses have sharp rise and fall times (< 400 ps), and also have consistent pulse shape and duration⁵.
We have modeled the phase change oscillation phenomenon using a finite element, 2-D or 3-D physical model for the Si wire (COMSOL) coupled with a SPICE model for load resistance, capacitance and pulse voltage. Temperature dependent electrical resistivity and thermal conductivity for doped crystalline Si are used in the simulations, and the energy requirement for phase change is modeled through a spike in the heat capacity function at the melting temperature, representing the latent heat of fusion. Simulations show good agreement in the waveform shapes between experimental and simulated data, suggesting that the model captures the physical processes that take place during phase-change oscillations, and suggest that it possible to achieve GHz frequency through phase change oscillations for a device scaled down to nanowire size with capacitance on the order of 10 fF⁶.
1 Cywar et al., 2009
2 Cywar et al., 2011
3 Bakan et al., 2009
4 Glazov et al., 1969
5 Cywar et al., 2010
6 Cywar et al., 2012

The modeling, fabrication and characterization of a new serpentine geometry for platinum IR bolometers is realized and demonstrated using plasma-enhanced atomic layer deposition (PEALD). The measured responsivity of 7.5x107V/WA of the proposed structure is an order of magnitude higher than the previously published atomic layer deposition (ALD) bolometers [1,2] while a greatly reduced thermal time constant (~300µs) leads to a 70% faster response time.
The use of ALD platinum for bolometers has led to a greatly simplified fabrication process because the nm-thick layer can simultaneously fulfill the role of infrared absorber, thermistor and structural layer [1]. However, the traditional geometry of bolometers, which consists of a square absorber membrane suspended by beams [3], has a significant drawback; The majority of the thermistor resistance is located in the suspension which exhibits only a moderate temperature increase from infrared (IR) absorption.
We propose a serpentine-like Pt structure that is approximately 10nm in thickness and has an overall lateral pixel dimension of 25 microns (um) per side with a 26-fold increase in electrical resistance in the critical central absorber region.
An analytical model using thickness-dependent values for electrical and thermal conductivity as well as temperature coefficient of resistance was created to evaluate the serpentine design. Responsivity measurements were recorded in vacuum under exposure to a black body source. The measured values for a design with seven 3mu;m wide beams are in good agreement with the model (to be published as a graph in the final paper). By increasing the number of beams to nine and decreasing the beam width to 2.4mu;m, a 40% higher responsivity compared to the seven-beam design was measured, while the footprint area only increases by 13%. The nine-beam serpentine bolometer with 5.5nm of platinum exhibits a responsivity of 7.5x107V/WA, an estimated noise equivalent temperature difference (NETD) of 97mK (assuming negligible 1/f noise) and a thermal time constant 10x smaller than previously reported VOx bolometers [4].
This novel bolometer design has demonstrated significant improvement in IR measurement performance without increasing pixel area while using a simplified Pt-PEALD process flow.
REFERENCES
===========
[1] Yoneoka, S., Interfacial Phenomena In Vacuum-Encapsulated Micro- And Nano-Structures, PhD thesis, Stanford University, 2011.
[2] Yoneoka, S. et al., “ALD-metal uncooled bo-lometer,” IEEE Micro Electromechanical Sys-tems (MEMS), pp. 676 -679, 2011.
[3] Skatrud, D. D. et al., Uncooled Infrared Imaging Arrays and Systems, Academic Press, 1997.
[4] Wang, H. et al., “IR microbolometer with self-supporting structure operating at room temperature“, Infrared Physics & Technology 45, pp. 53-57, 2004.

Flexible electronics such as flexible displays, artificial skin and health monitoring devices have gained great attention recently [Nature 477, 45 2011), PNAS 101, 9966 (2004)]. Looking closely at the components of these devices, although microelectromechanical systems (MEMS) actuators and sensors can play critical role to extend the application areas of flexible electronics, fabricating movable MEMS devices on flexible substrates is highly challenging [Adv. Mater. 22, 1840 (2010), Nat. Mater. 6, 413 (2007)]. Specially limited thermal budget of traditionally used flexible substrates like plastic or polymer substrates hinders its usage for MEMS processing. However, flexible electronics is widely used for large panel displays. At the same time, one of the most commercially successful MEMS based technology is DLP processor for display applications. Hence, it is critical to explore the opportunities merging MEMS devices on flexible platform. In the recent past we have demonstrated a generic batch process to transform low-cost bulk mono-crystalline silicon (100) substrates with pre-fabricated devices (traditional electronics, MEMS, etc.) into flexible (bending radius of 5 mm) and semi-transparent (12% transmittance) silicon fabric (5 - 25 mu;m thick) [MEMS 2012, MRS Fall Meeting 2012, Appl. Phys. Lett. 102, 064102 (2013)]. Additionally after releasing the top portion of the silicon substrate with pre-fabricated devices, we perform chemical mechanical polishing (CMP) of the remaining substrates to fabricate another set of devices and to release the top portion of the remaining substrate. In this way, we can recycle a 0.5 mm thick 4” wafer up to six times. Based on these progresses, we report a process for fabricating free standing and movable MEMS devices on flexible silicon substrates by fabricating MEMS flexure thermal actuator. Flexure thermal actuators consist of two arms: a thin hot arm and a wide cold arm separated by a small air gap; the arms are anchored to the substrate from one end and connected to each other from the other end. The actuator design was modified by adding etch holes in the anchors to suit the process of releasing a thin layer of silicon from the bulk silicon substrate. Selecting materials that are compatible with the release process was challenging. The structural layer of the devices was partially etched during silicon release. Furthermore, the thin arm of the thermal actuator was thinned during the fabrication process but optimizing the patterning and etching steps of the structural layer successfully solved this problem. Simulation was carried out to compare the performance of the original and the modified designs for the thermal actuators and to study stress and temperature distribution across a device. A fabricated thermal actuator with a 250 µm long hot arm and a 225 µm long cold arm separated by a 3 µm gap produced a deflection of 3 µm - taking a step forward towards expanding MEMS applications into flexible electronics.

Superb electromechanical properties of graphene, where large elastic deformation is achievable without significant perturbation of electrical properties, provide a substantial promise for flexible electronics and advanced nanoelectromechanical devices. Here, we report a rapid and scalable method of texturing 2-dimensional (2D) graphene by using soft-matter transformation of shape-memory polymers into 3-dimensional (3D) electromechanical sensors. We demonstrate that the thermally-induced transformation of graphene on a polymeric substrate creates 3D textured graphene. Quantitative analysis shows that both the wavelength and height of textured graphene are a few micrometers at an applied strain up to 300% and that the 3-dimensionality of graphene (i.e., wavelength and height of texturing) can be controlled by the processing parameters. We further characterize the electrical and mechanical properties of 3D graphene, and demonstrate the robust electromechanical properties of 3D textured graphene. Finally, we explore electromechanical device applications by constructing an array of strain sensors. We believe our approach to forming textured graphene by soft-matter transformation offers a unique avenue for creating advanced electromechanical devices in the future.

Ferroelastic domains observed in single-domain, epitaxial ferroelectric thin films are often associated with dislocations near the film/substrate interface in compressively strained films. The morphology of these domains depends on the film thickness and local stress; in relatively thin films the domains extend through the film while in relatively thick films the domains remain localized at the film/substrate interface. In this study, phase field simulations with supporting TEM observations are used to investigate the transition between the observed ferroelastic domain morphologies as a function of film thickness and its effect on switching in an epitaxial Pb(Zr0.2,Ti0.8)O3 (PZT) film. Analyses of the thermodynamic state of the PZT film reveal the existence of discontinuous changes in the first derivatives of the free energy of the film and hysteretic transition behavior, indicating the transition is first order. The impact of the transition on switching around an embedded ferroelastic domain is also considered. Partial ferroelastic domains are easily switched in thick films and upon removal of the applied bias return to their initial states. In contrast, ferroelastic domains in films with a thickness in which the transition shows hysteresis do not return to their initial configuration. Rather, an additional bias must be applied to return them to their original state. The mechanical and electrostatic states of the epitaxial film throughout the transition and switching process are discussed.

A novel micro-resonator-based interfacial fatigue testing technique was used to investigate the fatigue properties of atomic-layer-deposited (ALD) ultrathin coatings along the sidewalls of deep-reactive-ion-etched mono-crystalline silicon thin films. This technique ensures loading conditions relevant to MEMS devices, including kHz testing frequency and fully reversed cyclic stresses. Specifically, the fatigue properties of ALD titania (22 nm thick), a material known to be more stable in humid environments, were compared to that of ALD alumina (4.2 nm, 12.6 nm, 25 nm, and 50 nm in thickness) in two environments (30 °C, 50% relative humidity (RH) and 80 °C, 90% RH). Fatigue damage, in the form of channel cracks and delamination of the coatings, was found to accumulate slowly over more than 10^8 cycles. A few tests under static load were conducted for which no delamination (or crack growth) occurred, demonstrating that the governing cohesive and interfacial fatigue mechanisms are cycle-dependent. For both materials, the average delamination rates increase with increasing energy release rate amplitude for delamination, modeled with a power law relationship. However, the alumina coatings exhibit fatigue rates that are roughly one order of magnitude higher in the harsher environment, whereas no significant environmental effects were observed for titania. The governing cohesive and interfacial fatigue degradation mechanisms for these nanoscale coatings are discussed in light of these results. This study highlights a promising technique to identify robust coating strategies that are required for MEMS sensors exposed to the environment.

We briefly report on a new edge lithographic technique which exploits a polymer buffer layer interspersed between the substrate and metal masking layers to extend the capabilities of controlled undercutting. This nano-patterning technique is simple and versatile, high yield and rapid for creating nanowires, trenches, dots and fins. It does not require any expensive high resolution lithography but relies on conventional silicon fabrication techniques to produce nanometre features (80-400 nm) over large area substrates. Gold nanowire transparent electrodes which exhibit very low resistivity, similar to that of indium tin oxide, were produced by this technique. The nanoscale mapping of thermal properties of graphene, suspended over 180 nm wide trenches created using the technique, was conducted using standard thermal scanning microscopy (SThM). Ballistic heat transport was observed in suspended graphene while diffusive phonons dominated in adjacent areas of supported graphene.
The paper will also discuss the fabrication of carbon nanotube (CNT) based scanning thermal microscopy nanoprobes (CNT-SThM). Such devices exploit CNT&’s extreme thermal conductance which extends the capabilities of SThM to mapping thermal transport properties in materials with a superior sensitivity, thermal and lateral resolutions, including higher thermal conductivity materials. The CNT-SThM probes were produced using a combination of microfabrication and focused ion beam milling to accurately position the CNT before using a platinum layer deposition to weld the CNT onto the heat sensor of a standard SThM tip. The process can be carefully engineered to taper the CNT tip such that the diameter at the apex is controlled to yield values as low as 10 nm. This is critical to determine the lateral and thermal resolutions but also to reach a fast thermal equilibrium, essential to map specimens in the very low nanometre regime.
By channeling efficiently heat between the sensor and the substrate, through the CNT tip, the CNT-SThM nanoprobe overcomes the major drawback of the standard SThM probe, i.e. the high thermal contact resistance at the apex. The outstanding thermal conductivity coupled with the effective contact area determined by the fabrication process are key to fast and high resolution nanoscale thermal mapping. A few layers of graphene flakes were studied and contrasted against the standard SThM analysis. Likewise, mapping thermal transport in ULSI interconnects revealed very fine structures. Unlike the standard SThM which exhibits a lateral resolution just below 100 nm, the CNT-nanoprobe demonstrated significantly higher heat transport sensitivity and achieved spatial resolution of 20-30 nm in topography and ~30 nm in thermal resolution. In brief, a fabrication platform to accurately record thermal energy transport in materials at the nanoscale will be discussed.
We acknowledge support from the UK RAEng & EPSRC and EU FP7: FUNPROB (269169) & NanoEmbrace (316751).

In this work, the mechanical response of freestanding ultrathin aluminum oxide films has been investigated using an on-chip micro-mechanical testing method inspired from MEMS technology. Aluminum oxide layers are notorious for affecting the mechanical response of its underlying metallic Al substrate, by blocking dislocations and by creating stress concentrations when cracking. Alumina films also show interesting features for MEMS applications as such, e.g. as porous templates enhancing the contact area between a sensor layer and the ambient gas [1]. Our uniaxial tension test utilizes the internal stress in an actuator layer to pull on the test material. The design is such that numerous units with varying dimensions can be patterned on a single silicon wafer, allowing to obtain a complete stress-strain response in a single process [2].
The alumina films were deposited either by reactive magnetron sputtering or by complete anodization of an Al layer. While sputtering generally produces non-porous films, anodization can generate both non-porous and porous structures, depending on the electrolyte being used. All our films, sputtered or anodized, were found to be amorphous. For every film, the internal stress evolution during alumina growth was monitored in-situ, based on high resolution curvature measurements.
The mean internal stress for sputter-deposited films varied between 120 MPa and 350 MPa when increasing the chamber pressure from 1.7 to 6.0 mTorr. Such an increase in pressure leads at the same time to a decrease of both the Young modulus, from 179 to 83 GPa as measured by nano-indentation, and the film density, from 3.44 g/cm3 to 2.85 g/cm3 as measured by spectroscopic ellipsometry. The stress-strain response of the same alumina films showed a linear elastic behavior. These micro-mechanical data also confirmed the decrease in Young modulus observed independently by nano-indentation. Furthermore, a lower bound for the strain to failure of 0.4 % was obtained for films deposited at 1.7 mTorr. Complementary statistical data will be presented as well to unravel its pressure-dependence.
Intrinsic microstructural differences, as characterized by SEM, between the non-porous sputtered films and the porous ones produced by anodization could be linked to the observed differences in their stress-strain response. The latter were found to be significant, and allowed to provide direct evidence for the effect of pore density and pore dimensions on the mechanical response of anodized porous alumina. Finally, our micro-mechanical test structures also proved to be effective to study the recently claimed effect of electrical current on alumina plasticity [3]. To this end, we compared micro-electro-mechanically obtained stress-strain curves for non-porous sputtered and anodized films.
[1]A. Mozalev et al., Int. J. Hydrogen Energy 38 (2013) 8011
[2]S. Gravier et al., J. Microelectromech. Syst. 18 (2009) 555
[3]J. Houser et al., Nature Materials 8 (2009) 415

Atomically stabilized lasers form a key component of many families of atomic clocks. The atomically stabilized lasers have been recently miniaturized to 1-cc scale by the advent of MEMS alkali-metal vapor cells in the context of chip-scale atomic clocks. The miniature vapor cells have enabled the realization of MEMS sensors and actuators that can utilize the atomically stabilized laser wavelength as a distance constant-of-nature in a microsystem. We have utilized this concept to realize NORIS - Nano Optical Ruler Imaging System. This system enables uses a micro-grating to diffract stabilized laser beam to create a pattern that is spatially defined with ppm long-term stability, owing to the wavelength stability. This pattern is then correlated with an imaging chip to determine the precise position of the imaging chip location. Long term position placement to within 10-20nm over a 4-inch wafer has been achieved. We have used the imaging chip as a marker for precision metrology for tip-based nanofabrication and correction of inertial sensor bias. In the specific application of tip-based nanofabrication, where the time to fabricate can be hours or even days, the requirement for a tip to come back to the same position and follow a specific grid becomes critical. We have integrated the imaging sensors onto STM and AFM tips, with NORIS to etch and create patterns in carbon structures. We have used the NORIS guided STM tips to pattern HOPG, and more recently graphene, as a pathway to create graphene NEMS. This work has led to the capability to fine tune a resonator to ppm levels by selectively etching graphene. We have also used the long term stability of NORIS to measure the bias and scale-factor of inertial sensors. This is accomplished by measuring the inertial sensor signals as a function of applied motion measured by NORIS for long term stability. In addition to the applications, this paper will outline possible pathways to implement NIST-on-Chip, a way to create many standards of nature on chip for self-calibrating sensors.

Our group has successfully demonstrated the fabrication and characterization of an on-chip electron-impact ion source for use in a novel micro-mass spectrometer. The system uses a cold cathode field emission source and a three-layer polycrystalline silicon (poly-Si) microelectromechanical systems (MEMS) process. The cathodes employ a carbon nanotube (CNT)-based field-emission source which consumes at least an order of magnitude less power than conventional thermionic sources. The CNT emitter arrays are used as impact ionization sources and have demonstrated electron currents (Ie) in excess of 150 mu;A. Incorporating the CNT field emitters into a MEMS-based electron source has advantages in size, power, and cost without significant loss in sensitivity. However, a common failure mechanism during operation stems from poor adhesion between the CNT emitter arrays and poly-Si MEMS cathode platform. It is important to improve the linkage because compromised adhesion can impact device performance. During high potential fields, separation of the CNTs from the MEMS cathode can lead to a loss in field emission or catastrophic failure due to shorting of the device across the electrodes. CNT alignment and length-control are also crucial for low turn-on field emission in these devices. Patterned CNT growth on conductive poly-Si MEMS substrates presents difficulties due to Fe catalyst diffusion into the poly-Si, enhanced by the higher density of grain boundaries and other defects. Limiting the diffusion provides more control of the catalyst and in turn more control of the CNT growth. To both limit diffusion of the Fe catalyst into the poly-Si and improve adhesion of the resulting CNT array, we have employed a multi-layer metal stack catalyst. The diffusion barrier properties sufficiently retard the inter-diffusion and reduce the Fe catalyst reactions with the Poly-Si allowing for improved CNT growth. Enhanced physical interlocking of the CNT base and MEMS platform result in better adhesion between the poly-Si-Fe-CNT under strong electric fields. In this study, we compared the field emission characteristics of CNTs grown directly on poly-Si MEMS flip-up panels using Fe (control), Fe/Mo, and Fe/Ti catalyst layers. With MEMS development steadily increasing in the field of instruments and sensors, some relying on the integration of CNTs, optimizing the material interaction between CNTs and poly-Si substrates will provide an opportunity for better performance, lifetime, and consistency of such devices.

Here we demonstrate multi-modal approach of simultaneous characterization of polymeric materials in nanogram range using a microcantilever sensor. We integrate nanomechanical thermal analysis with photothermal cantilever deflection spectroscopy for monitoring multiple physical and chemical properties of samples in a single platform.
The changes with temperature in the deflection, resonance frequency, and Q factor of a polymeric material-coated silicon microcantilever are measured simultaneously to characterize thermomechanical and thermodynamic properties of the sample. The elastic and loss modulus of polymeric materials are determined as a function of temperature from the frequency response of the cantilever and the glass transition temperature is identified from the analysis of the deflection signal. In addition, nanomechanical infrared (IR) spectra from the same polymeric material-coated silicon microcantilever are acquired to provide chemical information on molecular structure. Degradation or cross-linking processes of polymeric materials can be monitored by analyzing nanomechanical IR spectra.

The roughness of a surface is of great importance to a number of thin film technologies, as it directly controls many physical aspects of materials behavior (e.g., adhesion, friction, and strength) and thus plays a role in the manufacturing yield and operational reliability of thin film devices. For example, it has been reported that critical flaws associated with surface roughness influence the fracture strength of silicon microelectromechanical systems, and consequently, that a distribution of critical flaw sizes from a fabrication sequence brings about a distribution in fracture strengths. In this presentation, we investigate roughness scaling of three different deep reactive ion etched (DRIE) silicon surfaces with atomic force microscopy and correlate the roughness parameters with fracture strengths from newly introduced “theta” test samples with the same etch features. At small lateral length scales, the resulting height-height correlations, H, reveal power-law behavior with equal scaling exponents for all surfaces, suggesting self-affine roughness inherent to the DRIE process. In contrast, at large lateral length scales, H is sensitive to the etch process; the root mean square roughness values change by a factor of five and are inversely related to the characteristic fracture strengths from the theta test samples. Establishing such relationships is only possible after the regular DRIE features are considered, as this precludes the roughness metrics from becoming artificially inflated.

Standard nano-lithography methods rely on a top down approach. This means a device is etched out of a pre-deposited material after patterning. As structure sizes further shrink, more material is removed to leave ever fewer atoms behind. This approach fundamentally breaks down when devices approach the single atom or molecular limit. We propose a new lithographic method for manufacturable fabrication of nano-scale devices and structures.
To this end we are developing a MEMS based “Fab on a Chip”. Micro-electromechanical systems serve as both actuators and sensors with which all major components of a nano-fabrication facility can be built. Furthermore, their size scale allows the entire micro-fab to fit on a single silicon chip. In this approach a macro-fab is used to build a standardized micro-fab. This can be done in a fast, cost effective way. The micro-fab is then capable of patterning custom nano-structures used for fundamental mesoscopic research.
The central component of the micro-fab is the writer. This dynamic stencil is controlled by orthogonal electrostatic comb drives. A central mask and aperture are positioned with nanometer accuracy, allowing for the precise placement of the desired material (lithographic mask). The intensity of the incoming atom flux is measured with a MEMS based mass sensor (film thickness monitor). Where a large atom flux is achieved through a conventional source, a micro heater can supply a small number of atoms to customize a device (source). Between the aperture and source a MEMS shutter controls the precise timing of the incoming atoms. Thermal sensors and heaters provide further information of the state of the fab and give critical feedback control.
Here we present the current status of this fabrication method, with focus on patterned structures made of chrome and gold and the dynamics of the aperture-shutter ensemble, enabling the controlled deposition of only a few atoms.

MEMS devices play an increasingly important role in adaptive optics systems. Adaptive optics systems minimize wavefront distortions by spatially adjusting the phase of the incoming wavefront. Applications of adaptive optics are found in numerous fields, including confocal microscopy, astronomy, and optical coherence tomography. MEMS are particularly useful because they offer high speed and displacement precision, while consuming very little power. These characteristics are ideal for adaptive optics to produces high resolution wavefront corrections. Furthermore, MEMS can be mass produced at low costs using foundries.
Adaptive optics MEMS have been successfully implemented in correcting waves in the visible portion of the electromagnetic spectrum. We are investigating the precision that can be obtained through MEMS devices and how MEMS-based adaptive optics could be utilized in X-ray applications. We have designed an adaptive optics MEMS device to probe the levels of precision obtainable for correcting high frequency wavefronts.
Our design consists of narrow plates, each with four springs attached to the corners and a single actuation electrode. An array of these plates is used to spatially control the phase of a wavefront by adjusting the path length. To accomplish this, the displacements are typically of order lambda;/2 and precision must be better by an order of magnitude or better. An electrostatic force actuates the plates. This method is chosen because it is both fast and draws negligible power. Several geometries were simulated to determine how to obtain the highest level of control with a minimum number of electrodes. This demonstrated how electrode placement can alter the curvature of the plates from concave to flat to convex. The simulations revealed that the device can be translated multiple wavelengths while remaining flat with respect to the wavelengths scale. This flatness would allow for piston-styled movement in an adaptive optics system suited for wavelengths down to approximately one nanometer.
Using the results from the finite element simulations, we designed and fabricated polysilicon MEMS structures and tested them with an optical interferometer to determine their electromechanical response. We demonstrate the feasibility of using MEMS-based adaptive optics for short wavelength electromagnetic radiation. To conclude, the possibility of dynamically tuning X-ray wavefronts for applications and experiments is discussed.

Plasma-etching employing fluorocarbon gas is one of the most popular MEMS processes in the fabrication of high aspect ratio contact hole on the SiO2 surface. However, there are critical problems in the nano-scale plasma-etching processes such as generation of deformed holes and decrease in aspect ratio. These failures are caused by the side-etching of holes and excessive growth of C-F polymer during the etching process. To eliminate these problems, etchant species are investigated in various ions and radicals by experiments. However, experimental methods are difficult to reveal an atomistic etching mechanism. Then, detailed understanding of chemical reaction dynamics depending on etchant species is necessary for an optimal design of the SiO2 etching process. Therefore, we developed a new etching process simulator based on our original tight-binding quantum chemical molecular dynamics (TB-QCMD) method which realized over 5,000 times acceleration compared to the first-principles molecular dynamics method [1, 2], and applied it to the elucidation of the etching process mechanisms. In the present study, we simulated the SiO2 etching processes by continuous irradiation of CF2 and CF3 radicals with kinetic energy of 150 eV, and compared the different etching mechanisms between these radicals. In the simulations, irradiated CF2 and CF3 radicals cause chemical reactions and dissociate Si-O bonds of the SiO2 surface. Moreover, the generation of C-C bonds is observed, and C-C chain grows by continuous radical irradiation. During the etching processes, we also observe the generation of CO, CO2, and SiF4 molecules, which accords with experimental results. Eventually, etching holes are generated on the SiO2 surface. In comparison of numbers of bond dissociations, CF3 radical dissociates more Si-O bonds and promotes the etching process faster than CF2 radical. C and F atoms are likely to be dissociated from irradiated radicals by radical bombardments on the SiO2 surface, and dissociated C and F atoms react with Si and O atoms on the SiO2 surface. Therefore, it is understood that CF3 radical dissociates more Si-O bonds than CF2 radical by a lot of reactive F atoms, and has an advantage in the etching rate. We successfully revealed the different etching mechanisms between CF2 and CF3 radicals. Our newly TB-QCMD etching process simulator is greatly beneficial to the investigation of the etching process.
[1] Hiroshi Ito, Takuya Kuwahara, Yuji Higuchi, Nobuki Ozawa, Seiji Samukawa, and Momoji Kubo, Jpn. J. Appl. Phys., 52 (2013) 026502.
[2] Takuya Kuwahara, Hiroshi Ito, Yuji Higuchi, Nobuki Ozawa, and Momoji Kubo, J. Phys. Chem. C, 116 (2012) 12525.