Search

Overlay metrology is crucial to process control in manufacturing semiconductor devices. Diffraction-based overlay (DBO) is an effective overlay measurement approach because it exhibits multiple advantages. This study analyzed measurement errors caused by sidewalls in the bottom gratings of DBO targets. Accordingly, improvement was proposed using a neural network. First, rigorous coupled wave analysis was employed to calculate the pupil images generated by an overlay target. These images were then used as a data set. Next, two-directional two-dimensional principal component analysis was used to reduce the dimension of features in these images. The features were then used to train a neural network and determine weighting coefficients in each network layer to create a DBO model. This study used virtual metrology to analyze 30 types of overlay targets and generated 18900 pupil images to create a data set. Each overlay target model was measured 10 times, and shot noise, dark noise, and quantization noise in the pupil images were accounted for. The simulation results revealed that when the dose level was 1000 mJ/ $\text{s}\cdot $ cm2, the overlay mean square error of the testing data was 0.40nm2, indicating notable improvement in the measurement results of overlay targets with bottom grating sidewalls. Therefore, the proposed neural network-based DBO model can be applied to overlay targets with sidewalls and effectively improve the overlay accuracy.

Titanium nitride (TiN) thin films were prepared by RF magnetron sputtering on a bare Si surface and a silicone-oil-covered Si surface as surface-enhanced Raman scattering (SERS) substrates. The TiN film deposited on the bare Si surface had a flat surface with densely packed grains measuring 18 nm in diameter, whereas the TiN film deposited on the oil surface had a wrinkly pattern made of grains measuring 19 nm in diameter. To analyze the SERS effect, Rhodamine 6G (R6G) was used as a Raman probe molecule at various concentrations ranging from 10−4 to 10−8 M. The analytical enhancement factor (AEF) values at 10−6 M R6G were respectively 6.2 × 103 and 1.3 × 104 for the flat TiN film and wrinkly TiN film. The SERS effect could be attributed to the enhancement of the local electromagnetic field at the top of the nano-sized grains. The wrinkly TiN film had a higher AEF than the flat TiN film, possibly due to the increased area of the Raman active spots. The wrinkly TiN film exhibited high level of SERS activity and achieved an R6G detection limit of 10−8 M.

This study used a digital micromirror device (DMD) to produce point-array patterns and employed a self-developed optical system to define line-and-space patterns on nonplanar substrates. First, field tracing was employed to analyze the aerial images of the lithographic system, which comprised an optical system and the DMD. Multiobjective particle swarm optimization was then applied to determine the spot overlapping rate used. The objective functions were set to minimize linewidth and maximize image log slope, through which the dose of the exposure agent could be effectively controlled and the quality of the nonplanar lithography could be enhanced. Laser beams with 405-nm wavelength were employed as the light source. Silicon substrates coated with photoresist were placed on a nonplanar translation stage. The DMD was used to produce lithographic patterns, during which the parameters were analyzed and optimized. The optimal delay time-sequence combinations were used to scan images of the patterns. Finally, an exposure linewidth of less than 10 µm was successfully achieved using the nonplanar lithographic process.

As the complexity of mask patterns increases, the standard off-axis illumination in 193-nm immersion systems is increasingly becoming inadequate for high-quality lithography. In this study, a source optimization (SO) technique was employed for improving lithographic resolution. A pixel-based optimization process was developed for constructing freeform sources on a PC platform to enhance contrast and improve process window (PW). In the proposed SO algorithm, the metaheuristic binary ant colony optimization algorithm was applied for constructing the freeform sources. Moreover, the proposed algorithm converged to the optimal solution between the 300th and 400th iterations. The performance of this algorithm was assessed using the half pitch (hp) 45-nm line/space (L/S) and 55-nm SRAM contact hole (CH) patterns. The fitness function was quality of aerial image, exposure latitude (EL), and depth of focus (DOF); the constructed freeform sources enhanced the contrast and extended the PW for the patterns. The DOF increased from 0.91 to 1.1 μm and 0.042 to 0.061 μm for the L/S and SRAM CH patterns, respectively, at an EL of 5% when using the freeform source. Moreover, the freeform source improved the mask error enhancement factor by 22% for the SRAM pattern.

Digital-mask lithography systems, with a digital micromirror device (DMD) as their central piece, have been widely used for defining patterns on printed circuit board (PCB). This study designed optical module parameters for point-array projection lithography based on field tracing technique to improve the quality of the aerial image on the exposure plane. In the realized optical module for the point-array projection lithography, a DMD was used as the dynamic digital-mask, and a 405-nm-wavelength laser was used to illuminate the DMD. The laser was then focused through the micro-lens array in the optical module to form a point array and was projected onto a dynamic scanning stage. By calculating the beam-overlapping rate, stage velocity, spot diameter, and DMD frame rate and programming them into the stage- and DMD-synchronized controller, the point array formed line patterns on the photoresist. Furthermore, using pulse width modulation (PWM) technique to operate the activation periods of the DMD mirrors effectively controlled the exposure and achieved a feature linewidth of less than 10 μm.

With the dramatically increase of design complexity and the advance of semiconductor technology nodes, huge difficulties appear during design for manufacturability with existing lithography solutions. Sub-resolution assist feature (SRAF) insertion and optical proximity correction (OPC) are both inevitable resolution enhancement techniques (RET) to maximize process window and ensure feature printability. Conventional model-based SRAF insertion and OPC methods are widely applied in industrial application but suffer from the extremely long runtime due to iterative optimization process. In this paper, we propose the first work developing a deep learning framework to simultaneously perform SRAF insertion and edge-based OPC. In addition, to make the optimized masks more reliable and convincing for industrial application, we employ a commercial lithography simulation tool to consider the quality of wafer image with various lithographic metrics. The effectiveness and efficiency of the proposed framework are demonstrated in experimental results, which also show the success of machine learning-based lithography optimization techniques for the current complex and large-scale circuit layouts.

Diffraction-based overlay (DBO) accuracy is critical to the intelligent nanolithography process control for producing advanced semiconductor fabrication nodes. Optical gratings located on various layers are commonly used as the targets for the detection of the overlay displacement offset in DBO measurement. The asymmetry in intensity between the 1st and -1st beams diffracted by the targets is used for the prediction of grating displacement offset. This paper describes the effect of grating targets with sidewall angles (SWAs) on asymmetry in intensity and proposes an artificial neural network (ANN) method for enhancing the accuracy of grating displacement offset prediction. Grating targets with a 1:3 line-to-pitch ratio and SWA profiles varying from 86° to 90° were employed in this paper. The asymmetry in the intensity of the designed targets was computed for incident beams with transverse-electric and transverse-magnetic polarization at visible wavelength. An ANN feed-forward model was developed for the displacement offset prediction. The ANN, the conventional linear model, and the regression models were evaluated using diffraction data calculated by a numerical electromagnetic solver. The mean square error and the mean of the residual indicated that using the ANN model and incident beams at wavelengths of 600, 650, and 750 nm is substantially more effective for prediction than the conventional linear model is.

This paper investigates the manufacturability-aware process of p-n junction formation for photovoltaic cells involving with Si nanoparticle layer. The furnace-based dopant diffusion process of forming a p-n junction consumes a substantial amount of energy. In addition, repetitive production steps prevent the possibility of Si ink-based cells integrating onto flexible substrates. This research examined the local heating dopant diffusion process by using a fiber laser at a wavelength of 1064 nm. The infrared beam is delivered onto the wafer stack with a nanoparticle carbon layer and n-type Si ink layer on p-type Si substrates. The nanoparticle carbon film absorbs infrared beam energy and converts photon energy as a thermal source to diffuse the n-type dopant in Si ink into the p-type Si wafer. The Si ink in this paper contains a mixture of Si nanoparticles and an n-type spin-on dopant solution. The TEM results show that Si nanoparticles are uniformly dispersed on the Si wafer surface. This research investigated sheet resistance as a function of laser parameters, including laser power, scanning speed, and pulse frequency for the samples coated with Si ink. Secondary ion mass spectroscopy measurements indicate the presence of an n-type dopant in p-type substrates, with an approximate diffusion depth of 100 nm. The results indicate that the proposed infrared laser treatment technique is promising for the formation of p-n junctions with Si ink-based photovoltaic cells.

This paper describes the use of ArF immersion lithography to verify the feasibility of a self-developed freeform illumination source that exposes features on masks and forms resist patterns. After development inspection (ADI) was used to measure the results and inspect whether the critical dimension (CD), process window, line-edge roughness (LER), and line-width roughness (LWR) of the resist pattern attained the standards of existing manufacturing processes. The testing features include nine horizontal and eight vertical features. Ant colony optimization (ACO) can provide optimal freeform illumination sources for these horizontal and vertical features. This paper considers each feature’s particular weighting in order to test the complexity and adaptability of the ACO algorithm applied to semiconductor manufacturing processes. The results of source calculations show that all 17 testing features can be satisfied with the X symmetry freeform source. When using this freeform source in an ArF lithography system, the ADI resist CD of all testing features is located within ±10% target CD tolerance. The ADI target CD set by dense patterns is 40 nm. Resist pattern ADI CD measurements show that, under a through-focus condition, the resist pattern CD between 40 and 42 nm is within tolerance, the depth of focus can reach up to $0.1~\mu \text{m}$ , LER is approximately 5 nm, and LWR is approximately 7 nm. These ADI CD measurements show that this ACO-based freeform source can be integrated into existing advanced semiconductor lithography processes.