Lithography Machine Research Articles

Lightweight Hotspot Detection Model Fusing SE and ECA Mechanisms.

In this paper, we propose a lightweight lithography machine learning-based hotspot detection model that integrates the Squeeze-and-Excitation (SE) attention mechanism and the Efficient Channel Attention (ECA) mechanism. These mechanisms can adaptively adjust channel weights, significantly enhancing the model's ability to extract relevant features of hotspots and non-hotspots through cross-channel interaction without dimensionality reduction. Our model extracts feature vectors through seven convolutional layers and four pooling layers, followed by three fully connected layers that map to the output, thereby simplifying the CNN network structure. Experimental results on our collected layout dataset and the ICCAD 2012 layout dataset demonstrate that our model is more lightweight. By evaluating overall accuracy, recall, and runtime, the comprehensive performance of our model is shown to exceed that of ConvNeXt, Swin transformer, and ResNet 50.

Source and mask optimization for stability of reticle and wafer stages

The relative motion of the reticle stage and wafer stage caused by vibration during the scanning exposure process of the lithography machine is an important factor that affects the imaging quality under limited lithography ability. In this paper, the influence of the vibration of the lithography stage system on the lithography imaging quality is studied. The lithography model including the vibration error of the stage is taken into account in the framework of source and mask optimization (SMO). For the 193-nm immersion lithography machine, combined with different design patterns with a line width of 40 nm and different vibration states of the stage when the lithography machine is exposed, the SMO joint optimization research is carried out, and the robustness of the stage under the limit process after SMO optimization is explored. The research results show that the contrast and relative light intensity change of the limit design figure under the influence of moving standard deviation (MSD) will greatly affect the linewidth at the light intensity threshold and then affect the process window (PW). The SMO including MSD allows the MSD to change in the range of 0-6 nm without changing the PW.

Design and fabrication of a Fresnel zone plate with an enhanced depth of focus

A Fresnel zone plate (EFZP) with an extended depth of focus can maintain focused monochromatic light at different distances compared to a general Fresnel zone plate (FZP). The focal distances are determined by dividing the zone plate into multiple areas based on the desired order. The EFZP has potential applications in various research fields such as microscopy, direct laser lithography, and optical coherence tomography. However, manufacturing an EFZP is challenging due to the high precision requirements and difficulties associated with the calculation and simulation processes. In this research, a complete process is presented to design, simulate, and fabricate an EFZP using a Fourier optics design, simulations, and a direct laser lithographic machine. The resulting EFZP has an increased depth of focus of about nine times compared to a general Fresnel zone plate with similar parameters, while maintaining the focal spot diameter. The performance of this EFZP is evaluated through optical verification and mathematical simulation methods.

Novel low-temperature and high-flux hydrogen plasma source for extreme-ultraviolet lithography applications

Inside extreme-ultraviolet (EUV) lithography machines, a hydrogen plasma is generated by ionization of the background gas by EUV photons. This plasma is essential for preventing carbon build-up on the optics, but it might affect functional performance and the lifetime of other elements inside the machine. The interaction of scanner materials and components with hydrogen plasma is investigated in controlled experiments using laboratory (off-line) setups, where the properties of EUV-generated plasmas are mimicked. Here, we present a novel experimental setup at TNO, where a low-temperature hydrogen plasma is generated by means of electron-impact ionization using a high-current, high-pressure electron beam (e-beam) gun. We show that the produced ion flux, peak ion energies, and radical-to-ion ratio are similar to that of the EUV-generated plasma. Since the e-gun has the option of operating the e-gun in the pulsed mode, it is possible to reproduce the time-dependent behavior of the scanner plasma as well. Moreover, as shown by Luo et al. [RSC Adv. 10, 8385 (2020)], electrons that impinge on surfaces mimic EUV photons in the generation of secondary electrons, which often drive radiation-induced processes (e.g., surface oxidation, reduction, and growth of carbon). We conclude that e-beam generated hydrogen plasma is a very promising technology for cost-effective lifetime testing of materials and optics, as compared to setups with EUV sources.

Computational Fluid Dynamics-Based Optimisation of High-Speed and High-Performance Bearingless Cross-Flow Fan Designs

To enhance the fluid dynamic performance of bearingless cross-flow fans (CFFs), this paper presents a CFD-based optimisation of both rotor and static casing wall modifications. High-performance CFFs are essential in industrial applications such as highly specialised laser modules in the semiconductor industry. The goal for the investigated rotor modifications is to enhance the CFF’s mechanical stiffness by integrating reinforcing shafts, which is expected to increase the limiting bending resonance frequency, thereby permitting higher rotational speeds. Additionally, the effects of these rotor modifications on the fluid dynamic performance are evaluated. For the casing wall modifications, the goal is to optimise design parameters to reduce losses. Optimised bearingless CFFs benefit semiconductor manufacturing by improving the gas circulation system within the laser module. Higher CFF performance is a key enabler for enhancing laser performance, increasing the scanning speed of lithography machines, and ultimately improving chip throughput. Several numerical simulations are conducted and validated using various commissioned prototypes, each measuring 600mm in length and 60mm in outer diameter. The results reveal that integrating a central shaft increases the rotational speed by up to 42%, from 5000rpm to 7100rpm, due to enhanced CFF stiffness. However, the loss in fluid flow amounts to 61% and outweighs the gain in rotational speed. Optimising the casing walls results in a 22% increase in maximum fluid flow reaching 1800m3/h at 5000rpm. It is demonstrated that the performance of bearingless CFFs can be enhanced by modifying the geometry of the casing walls, without requiring changes to the CFF rotor or bearingless motors.

Design and Fabrication of Continuous Surface Optical Field Modulator for Angular Spectrum Discreteness Compensation.

The light homogenizing element is a crucial component of the illumination system of the lithography machine. Its primary purpose is to realize the uniform distribution of energy. However, it suffers from a common issue, which is angular spectrum discreteness, which significantly impacts light uniformity. To address this, we design and fabricate random micro-cylindrical lens arrays to obtain a small-angle Gaussian optical field, which can compensate for the angular spectrum discreteness. By adjusting the pitches and curvature radii of the micro-cylindrical lenses separately, we are able to manipulate the divergence angle of the emitted sub-beams, enabling precise angular spectrum modulation. By using mask-moving technology, the angular spectrum modulator is fabricated to generate a Gaussian illumination field. The surface profile is measured and determined with a structural roughness below 10 nm. Furthermore, optical test experiments on the modulator have been conducted, achieving an angle error of less than 0.02° and a balance better than 0.5%.

A global plasma and surface model of hydrogen/methane inductively coupled discharge to analyze hydrocarbon plasma–surface interactions in extreme-ultraviolet lithography machines

Hydrocarbon contamination is associated with light transmission losses in modern lithography machines, which contain extreme-ultraviolet-induced plasma. A volume-averaged global and deposition/etch surface model of a reference hydrogen/methane inductive discharge is developed to investigate the plasma–surface interactions. The simulation results are validated against a wide variety of experiments and verified with respect to multiple sets of computational data. The deposition rate is calculated for a variation in methane impurity (10–10 000 ppm), power, pressure, and net mass flow. The simulations conclude that the hydrocarbon plasma deposition can be minimized by reducing methane impurity and excluding solid organic structures.

Interdependence and specialization in the global semiconductor industry

Interdependence between the United States (U.S.), European Union (EU) and Asia in the semiconductor industry, driven by specialization, can serve as a preventive measure against disruptions in the global semiconductor supply chain. Moreover, with rising geopolitical tensions, the cost-intensive nature of the semiconductor industry and a slowdown in demand, interdependence and partnership provide countries with opportunities and benefits. Specifically, by analyzing global trade patterns, developing the Interdependence Index within the semiconductor market, and applying the Grubel-Lloyd Index to the U.S., the EU, and Asian countries from 2011 to 2022, our findings reveal that interdependence enhances regional semiconductor supply chains, such as the establishment of semiconductor foundries in the U.S., Japan, and the EU; reduces dependence on a single supplier, such as the U.S. distancing from China; and increases market share in different semiconductor segments, as demonstrated by Taiwan in automobile chips. The evidence indicates that China heavily depends on foreign sources to meet its semiconductor demand, while Taiwan and South Korea specialize as foundry service providers with lower Interdependence Index values. The U.S. maintains a moderate level of dependence on semiconductor imports due to its strong presence in manufacturing and research, while the EU exhibits a relatively higher level of interdependence, emphasizing its reliance on semiconductor imports. The stage-specific analyses indicate that the U.S. and the EU rely on Asia for semiconductor devices, while China and Taiwan have a higher dependence on American intermediate inputs and European lithography machines.

A temperature control algorithm for lithography machine based on generalized predictive control and BP neural network PI control

Temperature stability is a critical factor affecting the performance of the most subsystems in the lithography system, due to the high precision and sensitivity of system components to temperature variations. The temperature control system of the lithography machine is characterized by its large inertial constant, time delay characteristics, as well as susceptibility to multiple disturbances. The temperature control system of the lithography machine chiefly requires response speed, high accuracy, and stable and constant temperature control. The contribution of this study is not only avoiding complex precision modeling processes based on real-time parameter estimation and neural network self-tuning but also improving the performance of temperature control in real time under external disturbances. A novel adaptive algorithm with a cascade structure based on generalized predictive control (GPC) and backpropagation (BP) neural network proportional-integral (PI) control is successfully proposed for high accuracy temperature control of lithography machine with a large inertial constant, time delay, and multiple disturbances. In this study, firstly, the liquid circulating temperature control system is developed based on heat exchanger and heater. Secondly, an adaptive controller composed of GPC and BP neural network PI control is successfully proposed. A BP neural network is employed to enable the parameters of the PI controller to adjust in real time, and the mathematical model parameters of the control system are identified in real time by the least square method. Also, the performance of the proposed controller is evaluated comparing with conventional PI controller and GPC controller in terms of robustness and quantitative study of error analysis. Finally, the temperature stability and robustness of the temperature control system controlled with the proposed adaptive GPC-PI algorithm has been investigated by the simulation results carried out in different working scenarios. The simulation results show that the steady-state error from the proposed algorithm is less than 0.01°C under the action of disturbance input. It can effectively counteract the influence of environmental interference and time-varying system parameters. The results of the simulation experiment indicate that the proposed adaptive GPC and PI control algorithm exhibits significant advantages in terms of control accuracy, anti-interference ability, and robustness compared to the conventional control method.

Development of a dynamic gas lock inhibited model for EUV-induced carbon deposition.

The optical surface of extreme ultraviolet (EUV) lithography machines is highly vulnerable to contamination by hydrocarbons, resulting in the formation of carbon deposits that significantly degrade the quality and efficiency of lithography. The dynamic gas lock (DGL) has been proven as an effective approach to alleviate carbon deposition. However, the majority of existing studies on carbon deposition neglect the influence of the DGL. This paper is dedicated to investigating the phenomena of hydrocarbon adsorption, desorption, and cleavage with considering the effects of the DGL. A comprehensive mathematical model of the carbon deposition process is established, and the impact of radiation intensity, temperature, and hydrocarbon types on the depositing rate is considered. The results suggest that the primary cause of carbon deposition is the direct cracking of hydrocarbons induced by photons with a wavelength range between 12.5 and 14.5nm. Additionally, it has been observed that the carbon deposition rate decreases exponentially as clean gas flow increases when EUV radiation intensity exceeds 50 mW/mm2. Conversely, at low EUV radiation intensity, clean gas flow has little effect on the carbon deposition rate. An effective approach to mitigate carbon deposition is to elevate the temperature of the optical surface and employ light hydrocarbon materials in the EUV process.

Development frontier and future challenges of lithography machines for integrated circuit manufacturing

Development frontier and future challenges of lithography machines for integrated circuit manufacturing

Production of 13.5 nm light with 5% conversion efficiency from 2 μ m laser-driven tin microdroplet plasma

We demonstrate the efficient generation of extreme ultraviolet (EUV) light from laser-produced plasma (LPP) driven by 2 μm wavelength laser light as an alternative for 10 μm CO2 gas LPP currently employed in EUV lithography machines for high-volume manufacturing of semiconductor devices. High conversion efficiencies of laser light into “in-band” EUV photons up to 5.0% are achieved by homogeneously heating the plasma that is laser-generated from preshaped tin microdroplet targets. Scaling the laser pulse duration, spot size, and intensity yields a high in-band EUV energy output of up to 12.5 mJ. The EUV emission source size is studied under a similar parameter range and is shown to match typical etendues of EUV optic columns. Our findings make 2 μm LPP a particularly promising candidate to power future EUV nanolithography.

Pollutant inhibition in an extreme ultraviolet lithography machine by dynamic gas lock

Dynamic gas lock (DGL) is an effective method to inhibit carbon deposition on optical surface in extreme ultraviolet (EUV) lithography machines. However, due to the complexity of the internal flow in DGL, relevant theoretical analyses remain incomplete. In this work, a direct simulation Monte Carlo (DSMC) method based on OpenFOAM platform is employed to study the transport process of rarefied gas in DGL and a project optics box (POB). The dsmcFoam + solver is extended and its applicability to DGL problems is verified. The results show that the number density of hydrocarbon in the POB decreases exponentially with the increase of the flow rate of clean gas, whereas the DGL inhibition rate is insensitive to the change of the pressure in the wafer chamber. Reducing the cone angle of the DGL, lowering the inlet position of the clean gas and using the clean gas with higher molar mass are all helpful to improve the inhibition effect of the DGL on the hydrocarbon diffusion. Additionally, the height of the DGL is relevant: with increase of the height, the inhibition rate rises at the same distance from the wafer surface; the number density of hydrocarbons is larger at higher heights. This study is expected to provide guidance for design of DGL technology under different EUV lithography conditions.

Ultrahigh-printing-speed photoresists for additive manufacturing.

Printing technology for precise additive manufacturing at the nanoscale currently relies on two-photon lithography. Although this methodology can overcome the Rayleigh limit to achieve nanoscale structures, it still operates at too slow of a speed for large-scale practical applications. Here we show an extremely sensitive zirconium oxide hybrid-(2,4-bis(trichloromethyl)-6-(4-methoxystyryl)-1,3,5-triazine) (ZrO2-BTMST) photoresist system that can achieve a printing speed of 7.77 m s-1, which is between three and five orders of magnitude faster than conventional polymer-based photoresists. We build a polygon laser scanner-based two-photon lithography machine with a linear stepping speed approaching 10 m s-1. Using the ZrO2-BTMST photoresist, we fabricate a square raster with an area of 1 cm2 in ~33 min. Furthermore, the extremely small chemical components of the ZrO2-BTMST photoresist enable high-precision patterning, leading to a line width as small as 38 nm. Calculations assisted by characterizations reveal that the unusual sensitivity arises from an efficient light-induced polarity change of the ZrO2 hybrid. We envisage that the exceptional sensitivity of our organic-inorganic hybrid photoresist may lead to a viable large-scale additive manufacturing nanofabrication technology.

Research progress of laser lithography

Lithography is one of the key technologies that restrict the development of the semiconductor industry and its important role continues to be highlighted. This paper will review laser projection lithography and laser maskless lithography based on the theoretical knowledge of laser lithography, present the advantages of laser maskless lithography, discuss the latest progress of laser lithography in application fields, explore its development prospects and trends, and provide some ideas and inspiration for the further development of human laser lithography. The result shows that laser lithography can be used in the field of material processing, and lithography is one of the key technologies to fabricate semiconductor devices. The world’s most advanced lithography machine is the EUV lithography machine of the Dutch ASML. This EUV lithography machine can be used for the production of 5nm chips, so 5nm is the most advanced chip manufacturing process that EUV lithography machine can achieve. Since lithography is a high degree of composite technology, each component plays an irreplaceable role. Therefore, if further improvements are to be made to the chip process, researchers should look at all parts of the exposure system, photoresist and process technology to make them work together in harmony.

High-Performance Flux Tracking Controller for Reluctance Actuator

To meet the ever-increasing demand for next-generation lithography machines, the actuator plays an important role in the achievement of high acceleration of the wafer stage. However, the voice coil motor, which is widely used in high-precision positioning systems, is reaching its physical limits. To tackle this problem, a novel way to design the actuator using the magnetoresistance effect is argued due to the high force densities. However, the strong nonlinearity limits its application in the nan-positioning system. In particular, the hysteresis is coupled with eddy effects and displacement, which lead to a rate-dependent and displacement-dependent hysteresis effect in the reluctance actuator. In this paper, a Hammerstein structure is used to model the rate-dependent reluctance actuator. At the same time, the displacement-dependent of the model is regarded as the interference with the system. Additionally, a control strategy combining inverse model compensation and the disturbance observer-based discrete sliding mode control was proposed, which can effectively suppress the hysteresis effect. It is worthy pointing out that the nonlinear system is transformed into a linear system with inversion bias and disturbance by the inverse model compensation. What is more, the sliding mode controller based on the disturbance observer is designed to deal with the unmodeled dynamics, displacement disturbances, and model identification errors in linear systems. Thus, the tracking performance and robustness to external disturbances of the system are improved. The simulation results show that it is superior to the PI controller combined with an inverse compensator and even to the discrete sliding mode controller connected with inverse compensator, confirming the effectiveness of the novel control method in alleviating hysteresis.

A Model-Based Approach for Measuring Wavefront Aberrations Using Random Ball Residual Compensation

The projection objective lens holds a pivotal role in lithography, directly influencing imaging system quality and, consequently, the lithography machine’s feature dimensions. Optical inspection methods for this lens require advancements in calibrating systematic error and enhancing alignment precision of auxiliary devices, given their impact on calibration accuracy. In the random averaging method, random ball can give rise to additional wavefront aberrations due to misalignment and numerical aperture mismatch. To mitigate these aberrations and enhance the accuracy of systematic error calibration, this paper introduces a random ball residual compensation (RBRC) model. Additionally, when combined with the random averaging technique, it elevates the calibration accuracy of the measured lens’s wavefront aberrations. The experimental results underscore the method’s effectiveness, accurately determining optical component eccentricities and numerical aperture errors. After eliminating these errors, more accurate values of lens wavefront aberrations are achieved. This research significantly contributes to enhancing error calibration of lithography objective lens systems.

Particle charging during pulsed EUV exposures with afterglow effect

Nanoparticle charging processes along with background spatial–temporal plasma profile have been investigated using 3D particle-in-cell simulations in a pulsed extreme ultraviolet (EUV) exposure environment. The particle charge polarity (positive or negative) strongly depends on its size, location, and background transient plasma conditions. The particle (100 nm diameter in size and conducting material) charge reaches a steady state in a single pulse (20 µs) within the EUV beams compared to particles outside the beam requiring multiple pulses. The larger the particle size, the lower the number of pulses required to reach a steady state. The charge of a particle decreases with pressure at a faster rate outside the beam compared to inside. These results are crucial for particle contamination (defectivity) control strategies for EUV lithography machines.

Enhancing projection based iterative learning control: A set-membership approach

Lithographic machine tools require both high motion accuracy and high motion flexibility. Projection based iterative learning control (P-ILC) is appealing for wafer stages to achieve two goals, simultaneously. P-ILC contains a nonparametric feedforward controller based on ILC, and a parametric feedforward controller with a projection step for feedforward tuning. In this paper, a set-membership based frequency-domain ILC algorithm (SM-F-ILC) is employed in the enhancing P-ILC scheme to improve the performance in the nonparametric feedforward control mode. SM-F-ILC can effectively compensate for repetitive errors, attenuate the nonrepetitive error accumulation and achieve fast convergence speed with model uncertainties. These superiorities also facilitate to improve the performance of P-ILC in the parametric feedforward control mode. The validity of the enhancing P-ILC scheme is demonstrated by experimental results.

The Research on the Fabrication of the Novel Basic Unit for 3D Structure Integrated Circuit

The IC industry in the Chinese mainland has encountered the difficulty caused by the embargo on lithography machines in developing state-of-the-art semiconductor IC processes. For the first time, this work fabricates out the HVTFET which we have independent intellectual property rights. A 50 nm channel length N HVTFET device is fabricated successfully using a 0.35 µm process, and its characteristics are close to those of TSMC’s 55 nm planar N MOSFET. The experimental results of this paper have two important significances. First, it shows that the fabrication of the advanced ICs in the Chinese mainland can get rid of the limitation of the lithography machine by using the HVTFET. Second, it shows that it is possible to decrease the chip area by about 3/4 by using the HVTFET structure for the ICs with 28 nm and above planar MOS processes, which are quite common in China.