In this editorial, Editor-in-Chief Chris Mack declares this the Year of Stochastics.

This paper explains how an electric field and a reticle interact and describes the different kinds of damage that can be caused to a reticle through its exposure to electric field. It is shown why electrostatic reticle damage has changed from ESD damage (which causes yield to suddenly drop precipitously) into a gradual and cumulative form of degradation that is very difficult to diagnose. It is explained why some of the approaches that have been taken to reduce ESD damage in the semiconductor factory, such as equipotential bonding and the use of static dissipative plastics for making reticle pods, actually increase the risk of this cumulative type of electrostatic degradation in reticles. When assessing the risk to reticles and designing an effective protective strategy for reticle handling, it is shown why one must take into account the temporal characteristics of a reticle’s interaction with electric field—including the effect that the reticle’s immediate surroundings will have on that interaction—as well as considering the strength of any electric field in the reticle handling environment. Solutions are presented that would allow the electrostatic risk to reticles to be reduced significantly, without requiring major changes to operating procedures in semiconductor manufacturing facilities.

In optical lithography tools, thermal aberration of a projection lens, which is caused by lens heating effect, leads to degradation of imaging quality. In addition to in-line feedforward compensation technology, thermal aberration can be reduced by optimizing optical design of a projector. Thermal aberration analysis of a projection lens benefits the optimization of optical design. A model of lens heating effect for a lithographic projector is introduced, which is capable of evaluating the synthetical thermal aberration of a projector as well as analyzing the contribution of an individual optical element. Simulation results by the introduced model show that not only the deformation of lens surface, the variance of refractive index but also the change of optical path, which depends on optical design, should be considered in thermal aberration analysis. The contributions of optical elements at different locations of the projector are also analyzed. Based on the model and the simulation results, an optimization method is proposed. A projector for i-line lithography is optimized by the proposed method. Main aberrations Z5, Z9, and Z17 are reduced by about 40%. The image quality of the lithographic projector in steady state is also improved.

With the continuous shrinking of semiconductor manufacturing technology node, the pattern size has become smaller and the pattern density has become higher. This can cause the lithography defects to be more difficult to remove, because the developing fragments have a larger relative surface area, and thus more sticky to the patterned substrate surface. Although the defect can be reduced or removed by the rinse process, it may require more time and effort. As we know, the process optimization is related to the co-optimization of the rinse dispense volume, the nitrogen gas dispense recipe, and the wafer rotation speed. In the limit that we push the three parameters to the highest level, the defect can be removed. However, the adverse effect is that the resist patterns may be damaged, such as pattern collapse. It may also cause an increase in the cost for the process and more trouble shooting time in finding an optimizing recipe. It will be very helpful that we can develop a simulation program that can do the optimization. We present a model based on viscous fluid dynamics and calculate the removing force distribution across the 300-mm-diameter wafer for the defect residual. We assume that the defect, mostly the partially deprotected and developed photoresist polymer residual. We assume that once the removing force made by a sum of the flowing rinse water, the nitrogen gas, and the centrifugal force by wafer rotation reached certain threshold level, the defect can be removed. We have performed some simulation study and compared our results with previous studies. We found that we can reproduce the defect distribution patterns, such as the two rings, from the previous studies. From the simulation, we have learned the interrelations between the three parameters and find that we can get the minimally required strength from the three parameters for defect removal. We have also studied the situation of the 450-mm diameter wafer, and we found that we can get the defect clean result with reduced wafer rotation speed. In summary, we have built a defect model for photolithography development that can model the defect removal mechanism during the rinse process.

A two-stage approach is introduced to improve the accuracy of compact patterning models used in large-scale computational lithography. Each of the two stages uses a separate empirically calibrated regression model whose accuracy at predicting printed feature dimensions has been proven in the usual standalone (single stage) mode. For the first model stage, we choose an established regularized regression model of the kind that accounts for resist non-idealities by suitably modifying the pre-thresholded exposing dose pattern, with the model basis functions taking the form of modified convolutions of adjustable kernels with the optical image. A different class of regression model is used in the second stage, namely a model that accounts for resist non-idealities by making a pattern-dependent local adjustment in the develop threshold, with the model basis functions being characteristic traits of the image trace along feature cutlines. However, rather than applying this second model in the usual mode where it adjusts the develop threshold applied to the exposing optical image, we use it to adjust a threshold that is applied to the improved effective dose distribution provided by the first-stage model. The effectiveness of the proposed method is verified by modeling pattern transfer of critical layers in 14- and 22-nm complementary metal–oxide–semiconductor (CMOS) technology. In our experience, little accuracy improvement is gained by expanding the complexity of standard single-stage models beyond the level of empirically proven model forms. However, even in a basic implementation, inclusion of a second stage of modeling will in itself reduce RMS error by ∼45  %   in our 14-nm example. Moreover, accuracy improvement is further boosted to ∼55  %   by adopting a minimax strategy in which the model is conservatively regularized according to the worst-case outcome in cross-validation tests but is calibrated according to the best-case outcome.

Atomic layer deposition is an efficient method for coating a few nanometer-thick alumina over a wafer scale. This method combined with the standard photolithography process was presented to fabricate metallic nanometer gaps that optically act in terahertz regimes. However, the cross-sectional view of the gap shape of the metal–insulator–metal nanogap structure varies depending on the conditions from the stepwise procedure. In specific, selecting photoresist materials, adding ion milling and chemical etching processes, and varying metal thicknesses and substrates result in various optical gap widths and shapes. Since the cross-sectional gap shape affects the field enhancement of the funneled electromagnetic waves via the nanogap, the control of tailoring the gap shape is necessary. Thus, we present five different versions of fabricating quadrangle-ring-shaped nanometer gap arrays with varying different kinds of outcomes. We foresee the usage of the suggested category for specific applications.

The amount of absorbed light in thin photoresist films is a key parameter in photolithographic processing, but its experimental measurement is not straightforward. The optical absorption of metal oxide-based thin photoresist films for extreme ultraviolet (EUV) lithography was measured using an established methodology based on synchrotron light. Three types of materials were investigated: tin cage molecules, zirconium oxoclusters, and hafnium oxoclusters. The tin-containing compound was demonstrated to have optical absorption up to three times higher than conventional organic-based photoresists have. The absorptivity of the zirconium oxocluster was comparable to that of organic polymer-based photoresists, owing to the low absorption cross section of zirconium at EUV. The hafnium-containing resist shows about twice as high absorptivity as an organic photoresist, owing to the significantly higher absorbance of hafnium. From the chemical composition and crystal structure, the density of the spin-coated films was determined. Using the density of the films and the tabulated data for atomic cross section at EUV, the expected absorptivity of these resists was calculated and discussed in comparison to the experimental results. The agreement between measured and expected absorption was fairly good with some substantial discrepancies due to differences in the actual film density or to thickness inhomogeneity due to the spin coating. The developed method here enables the accurate measurement of the EUV absorption of the photoresists and can contribute to the further development of EUV resists and more accurate lithographic modeling.

This paper reports on our studies of the dynamic process of positive tone photoresist development in real time. Using high-speed atomic force microscopy (HS-AFM) in dilute alkaline developer solution, changes in morphology and nanomechanical properties of patterned resist were monitored. The Bruker Dimension FastScan AFM was used to analyze 193-nm acrylic-based immersion resists in tetramethylammonium hydroxide developer solution. HS-AFM operated in PeakForce Tapping^® (Registered Trademark of Bruker, Inc.) mode can allow for concurrent measurements of resist topography, stiffness, adhesion, and deformation during development. These studies focused on HS-AFM topography data as it readily revealed detailed information about initial resist morphology, followed by a resist swelling process, and eventual dissolution of the exposed resist areas. HS-AFM showed potential for tracking and understanding development of patterned resist films and can be useful in evaluating the dissolution properties of different resist designs. Also discussed are the roles of AFM tip shape and resist feature geometry on the measured line edge roughness using a simulation procedure.

Thermal interface material (TIM) is a key component to dissipate the accumulated heat in the majority of power electronic systems. In this work, a facile and solid-state ball-milling method is adopted for the solvent-free reduction of exfoliated graphite nanoplatelets (EGNs) into high-quality ball-milled exfoliated graphite nanoplatelet (BMEGN) fillers. In addition, BMEGN fillers are embedded and uniformly dispersed with polydimethylsiloxane (PDMS) matrix to make a highly stretchable BMEGN-embedded PDMS-TIMs (BMEGN/PDMS) with strongly enhanced thermal conductivity. Furthermore, material characterizations were thoroughly investigated using scanning electron microscopy, transmission electron microscope, Raman spectroscopy, thermogravimetric analysis, and x-ray diffraction. Improvements in the thermal conductivity of TIMs by adding BMEGN were compared, the thermal conductivity was observed for BMEGN fillers with 0- to 48-h ball-milling time, and an enhanced in-plane thermal conductivity of 15.04 to 16.91 W/mK and through-plane thermal conductivity of 1.03 to 1.19 W/mK can be experimentally measured. A strong anisotropy was observed in the range of 14.60 (BMEGN12h/PDMS) to 14.21 (BMEGN48h/PDMS). The results reveal that the ball-milled graphene filler network with branched morphology can effectively provide the synergetic effect of a thermally conductive pathway via diffusion of phonon vibration in flexible composites. The combination of thermal conductivity and thermal stability may facilitate the applications in thermal management.

A method for the inline measurement of the tunnel oxide–nitride-blocking oxide (ONO) film thickness in 3D-NAND devices was studied. The ONO film, whose thickness is critical to the device properties, cannot be measured with conventional methods because it is deposited on the sidewall of a memory hole. Thus, a method to measure the thickness of this vertical film is required. We propose a critical dimension-scanning electron microscope (CD-SEM) measurement. The film thickness can be obtained by measuring the hole diameter before and after the film deposition. Namely, the decrease in the hole diameter should be twice of the thickness in principle. However, its applicability to the actual 10-nm-thick ONO film has not been verified. In this study, the measurement precision and the validity of the method were examined with actual ONO film in the 3D-NAND test wafers. The results showed excellent precision (0.08 nm) and good consistency with planar transmission electron microscope (TEM) and ellipsometry results. In addition, the method revealed the subnanometer thickness difference depending on the nominal hole diameter and the hole density. It suggests the impact of inhomogeneity in the source gas supply during the film deposition. These results ensure that the method is sufficiently precise for the inline local thickness measurement of the ONO film. With this method, yield and reliability managements in the 3D-NAND device manufacturing would be improved.

A concept using a glass-based microchip with finger electrodes, under a two-step voltage-controlled manner, to accomplish the formation and collection of liposomes is presented. By the actions of electro-osmotic and electrolytic bubbles in low and high applied voltages, respectively, the liposomes were first formed in an electrolyte reservoir; then, they were successfully moved to a collected reservoir, which is helpful for subsequent on-chip processing. This study evaluated the distribution of an electric field for different couple-finger distances using finite-element analysis software. The simulation result shows that the electric field density increases rapidly with the decrease of finger distance, with a quadratic relationship. Experiment results show that, when the applied AC voltage is 1 Vp-p, at a driving frequency of 10 Hz, the liposomes are first formed near the boundary of the electrolyte reservoir, and when the voltage is higher, the mass production of liposomes is achieved. When the voltage is higher than 4 Vp-p, the electrolytic bubbles grow cyclically to separate the liposomes from the electrolyte reservoir and therefore collect in a specific reservoir. Liposomes encapsulate doxorubicin, and their drug release to the lung cancer cell line H1299 was also implemented to verify their biointeraction characteristics. These experimental results were all identified using an inverted fluorescence microscope.

The micromixer is a very common component in the state-of-the-art lab-on-chip devices and occupies large chip areas to fulfill the rather challenging process of mixing in microscales. Two various design micromixers are introduced, which show a step over efficiency in the microlevel mixing. Finite-element method (FEM) tools were utilized to assess the mixing efficiency of the presented micromixers versus common T- and zigzag-shaped mixers. Using the availability of three-dimensional printing features, the chaotic advection is maximized as a mainstream factor affecting microscale behavior of the mixer. Both FEM and experimental results prove a 95% improvement in the performance of micromixers for low Reynolds numbers at 1 versus 8 cm for conventional devices.

Paper-based microfluidic devices offer great potential as a low-cost platform to perform chemical and biochemical tests. Commercially available formats such as dipsticks and lateral-flow test devices are widely popular as they are easy to handle and produce fast and unambiguous results. Although these simple devices lack precise control over the flow to enable integration of complex functionality for multistep processes or the ability to multiplex several tests, intense research in this area is rapidly expanding the possibilities. Modeling and simulation is increasingly more instrumental in gaining insight into the underlying physics driving the processes inside the channels; however, simulation of flow in paper-based microfluidic devices has barely been explored to aid in the optimum design and prototyping of these devices for precise control of the flow. We implement a multiphase fluid flow model through porous media for the simulation of paper imbibition of an incompressible, Newtonian fluid such as when water, urine, or serum is employed. The formulation incorporates mass and momentum conservation equations under Stokes flow conditions and results in two coupled Darcy’s law equations for the pressures and saturations of the wetting and nonwetting phases, further simplified to the Richard’s equation for the saturation of the wetting fluid, which is then solved using a finite element solver. The model tracks the wetting fluid front as it displaces the nonwetting fluid by computing the time-dependent saturation of the wetting fluid. We apply this to the study of liquid transport in two-dimensional paper networks and validate against experimental data concerning the wetting dynamics of paper layouts of varying geometries.

Methods exist for the creation of antireflective thin film layers; however, many of these methods depend on the use of high temperatures, harsh chemical etches, or are made with difficult pattern materials, rendering them unusable for many applications. In addition, most methods of light blocking are specifically designed to increase light coupling and absorption in the substrate, making them incompatible with some applications that also require blocking transmission of light. A method of forming a simple, patternable light-blocking layer that drastically reduces both transmission and reflection of light without dependence on processes that could damage underlying structures using a light scattering matte coating over a partially antireflective thin film light-blocking layer is presented.

With MEMS technology developing rapidly, the microfluidic droplet technology has been employed into many biological or chemical experiments. To satisfy the increasingly high demands in various applications of the microfluidic chips, the size of the droplet needs accurate control. We have designed and performed a three-dimensional numerical simulation of a droplet generation in a double T-junction microchannel by a level-set method in COMSOL5.0 for discussing the process of the droplet formation and the influences on the size of the droplet. We have studied out that the flow rate ratio, the continuous phase viscosity, the interfacial tension, and the contact angle influence the size of the droplet. The effective droplet diameter increases when the flow rate ratio and the interfacial tension increase. It decreases when the continuous phase viscosity and the contact angle increase.

To advance neuroscience in vivo experiments, it is necessary to probe a high density of neurons in neural networks with single-cell resolution and be able to simultaneously use different techniques, such as electrophysiological recordings and optogenetic intervention, while minimizing brain tissue damage. We first fabricate electrical neural probes with a high density of electrodes and small tip profile (cross section of shank: 47-μm width  ×  16-μm thickness). Then, with similar substrate and fabrication techniques, we separately fabricate optical neural probes. We finally indicate a fabrication method that may allow integrating the two functionalities into the same device. High-density electrical probes have been fabricated with 64 pads. Interconnections to deliver the signal are 120-nm wide, and the pads are 5  ×  25  μm. Separate optical probes with similar shank dimensions with silicon dioxide and silicon nitride ridge single-mode waveguides have also been fabricated. The waveguide core cross section is 250  nm  ×  160  nm. Light is focused above the waveguide plane in 2.35-μm diameter spots. The actual probes present two output focusing gratings on the shank.