This PDF file contains the front matter associated with SPIE Proceedings Volume 11323, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

While EUV systems equipped with a 0.33 Numerical Aperture (NA) lens are entering high volume manufacturing, ASML and ZEISS are in parallel ramping up their activities on an EUV exposure tool with an NA of 0.55. The intent of this high-NA scanner, targeting a resolution of 8nm, is to extend Moore’s law throughout the next decade. The high-NA optical system, together with the developments in mask and resist, provides an increased contrast, key to control stochastic contributions to EPE and error rate of printing defects. A novel lens design, capable of providing the required NA, has been identified; this lens will be paired with new, faster stages and more accurate sensors enabling the tight focus and overlay control needed for future process nodes. Impact on system architecture and proposed solutions are described in this paper. In addition, we give a status update on the developments at ZEISS and ASML.

For each lithography scanner the optics is a key component. While the NXE:3400 with ZEISS Starlith®3400 optics at Numerical Aperture of 0.33 is entering high-volume manufacturing in customer factories, we are developing high NA optics with a Numerical Aperture of 0.55. This optics consists of a highly flexible illumination system and a projection optics enabling single-exposure sub 8nm half-pitch resolution. In this paper, we give an overview of the progress of ZEISS High-NA EUV program where production of first mirrors and frames has already been started.

Pushing the novel anamorphic NA=0.55 EUV projection optics to k1 values below 0.4 and to its ultimate resolution limit will require an alternative mask absorber stack. This paper describes the application of rigorous imaging simulations in combination with multi-objective optimization to access the performance of novel absorber materials for the NA=0.55 system. Simulations of various use cases and material options indicate two main types of solutions: high k materials (k>0.05, especially for vertical lines/spaces) and low n materials (n ~ 0.9) to provide phase shift mask solutions for contact arrays.

Aberrations must be sufficiently controlled to make moving to a higher numerical aperture worthwhile. Traditional isomorphic imaging systems form the same image regardless of their rotation. Likewise, the aberration basis chosen for isomorphic optics is invariant under rotation. Anamorphic optics are not rotationally invariant though—they are only reflection invariant. We have shown in previous reports that a basis composed from a product of Legendre polynomials represents the balanced aberrations of anamorphic optics. Solutions have been presented under the presence of a circular central obscuration. This paper will examine the properties of these aberrations and their effects on image formation through analogies to the well-known Zernike aberrations. It will be shown that ray tracing simulations of the point spread function of an anamorphic optic in Code V matches predictions made by the proposed basis. A system will be described for computing an anamorphic aberration basis in the presence of an arbitrary obscuration. Based on this system we will analyze the effects of using the basis for the wrong type of obscuration.

Vibration levels in MET5 exposures were reduced from 1.5 nm RMS to 0.8 nm RMS by tuning the vibration isolation system and removing non-compliant hardware. Frequency doubling exposures were improved by replacing the Fourier synthesis pupil scanner mirror. Focus-exposure-matrix outliers have been solved by patching a bug in the control software. 9 nm half-pitch lines and 8 nm half-pitch lines were printed in 11 nm thick MOx resist.

High-NA extreme ultraviolet lithography (EUVL) is going to deliver the high-volume manufacturing (HVM) patterning for sub-3nm technology node for the semiconductor industry. One of the critical challenges is to develop suitable EUV photoresists at high resolution with high sensitivity and low line-width roughness (LWR) at reduced film thickness needed for high-NA by its reduced depth of focus (~1/NA²). The resist performance is generally limited by the resolution-LWR sensitivity (RLS) tradeoff, and it is critical to find new materials to support the future lithography nodes. EUV interference lithography (EUV-IL) is a powerful and efficient technique to test new materials at high resolution. In this work, we evaluate the performance of about 120 EUV resists, including molecular resist, inorganic resist, chemically-amplified resist (CAR), and chemically-amplified resist with a metal sensitizer (metal-CAR). Among all tested resists, we selected and compared six resists with the best performance in dose-to-size, line-width roughness, and maximal exposure latitude for 16 nm and 14 nm half-pitch (HP) lines/spaces (LS). A molecular resist showed the lowest dose to resolve HP 16 nm (29 mJ/cm²) and 14 nm (31 mJ/cm²), still featuring low LWR_unbiased (2.7 nm and 3.1 nm, respectively). We observed that there is steady progress in EUV resist development: the LWR of the resists was improved in comparison with the results of the last two years as we noted that the amount of the resists within given dose and LWR threshold increased from 10 % to 33% for HP 14 nm. For contact holes, CAR gave the lowest LCDU of 2.2 nm for HP 20 nm with the lowest doses (21.1 mJ/cm²). While inorganic resists resolved pillars with highest resolution HP down to 18 nm with LCDU of 2.1 nm.

MOx resists have matured into promising alternatives to conventional CAR resists for advanced-node EUV lithography where these materials offer potential improvements to patterning fidelity and high etch resistance based on metallic components. This is a particular boon for processes with limited exposure latitudes such as High-NA EUV lithography. Creating and employing first-principle models of MOx lithographic processes should speed adoption and development of these materials and represents an important aspect of platform maturation. Stochastic photochemical models of metalcontaining resist systems have previously been developed, but without extension to computational lithography. Likewise, stochastic models derived from CAR systems have been fit to MOx lithographic data using computational lithography software, facilitating limited stochastic lithography studies without capturing fundamental MOx imaging processes. Recently, a rigorous stochastic model built from the ground up using MOx-specific resist principles has been developed. In this contribution, the performance of this MOx-specific model was assessed by comparing simulated and experimental lithography data for a series of MOx resists under a range of exposure and process conditions. Chemical and physical properties of the resists derived independently from X-ray diffraction, EUV absorbance, FTIR spectroscopy, and ellipsometry measurements were parameterized in the context of the simulation, and calibration routines were used to fit simulated data to experimental CD-SEM exposure data produced using an NXE-3300B EUV scanner. Insights from these models may be used to guide MOx resist development and EUV lithography process optimization. Ultimately, these studies will help to identify process windows, processing points, and possibly improvements to the MOx resists.

Stochastic defects are becoming major concern in the future EUV lithography as their probability P_d exponentially increases with decreasing feature size and is highly sensitive to variations in process/mask conditions. Photon shot noises and discrete/probabilistic nature of materials have been blamed as their causes. We introduce models for relating P_d to photon and resist statistics under various exposure and material conditions and analyze their impact in future EUV lithography. Three-dimensional reaction distributions are calculated by a fully-coupled Monte Carlo simulation including discrete photon, photoelectron scattering, and resist stochastics. Then, probability models predict _Pd from statistical data extracted from Monte Carlo results. Stochastic defect generation is enhanced by cascade and/or cluster of correlated reactions among nearby polymers/molecules due to secondary electrons (SE)/acid diffusion and SEs generated along scattered photoelectron trajectories. P_d decreases with increasing reaction density, suppressing effective image blur, and introducing quenchers, where reaction density is limited by SE, PAG and reaction site. Defect probability increases with decreasing target size for the same k1-factor, while strongly dependent on image slope and defocus. Our analyses suggest that applying EUV lithography to smaller target requires careful material choice, extremely precise process control, and further EUV power enhancement.

Despite years of research and development, the fundamental processes of photoionization, secondary electron generation, recombination, diffusion, and resist switching are poorly understood at the atomic level for EUVL. Multiscale modeling of these physical and chemical processes can provide answers to questions that are difficult or impossible to answer with experiment alone. A modeling pipeline that includes Monte Carlo modeling of photon- and electron-matter interactions, along with density functional theory calculations of chemical switching will be introduced in this proceeding. The Hf₄O₂(OMc)₁₂ nanocluster resist will be presented as a case study. Photon and secondary electron yields, electron energy and spatial distributions, and a quantum chemical pathway for negative tone switching will be presented. Fundamental learning from studies like this can be used to improve resist design including improving contrast of these materials.

In this paper we will first review the metrology for detecting stochastic printing failures (or defects): which techniques are available and what is their performance in detecting the different failure types in regular L/S or contact arrays, in terms of sensitivity and wafer area they can cover. Although progress is being made, metrology is still a concern: not all types of failures can be detected with the sensitivity and area-inspection capability that will eventually be required for yield sign-off. Next, we look into the progress we have been making in generating empirical correlators (or predictors) that can reproduce observed dependencies of the position of the stochastic cliffs on patterning or structure parameters, such as focus-dose, illumination mode, mask bias and pitch. In this part we focus on the microbridge-cliff on resist wafers, with some initial data on the missing-contact cliff. Especially the pitch-dependency (proximity dependency) of the cliff position is a challenge, as good models for this dependency are crucial for the development of stochastic-hotspot predictors for the OPC Verification flow. In this paper we will present for the first time a reasonable predictor for this pitch dependency; but although these results a promising step forward, they are not yet good enough as a final solution. So, this type of research will need to be continued.

Stochastic defects are a concern in the lithographic processes used for semiconductor manufacturing, particularly for advanced node extreme ultraviolet lithographic processes. Experimentally determining the defect probability for a lithographic process is extremely time-consuming, requiring expensive metrology equipment and generally limited to simple patterns. Defect probability simulations can be beneficial from time and cost perspective and furthermore should be extensible to more complex patterns. As such, being able to accurately predict the defect probability using lithography simulations would be a valuable complementary option. We show the results of a fast simulation-based methodology for predicting defect probabilities based on a continuum lithographic model calibrated to experimental data. The simulation based-results are compared to experimental microbridge defect probability data where we show a good correlation between the two.

The methods to calculate the probability of success/failure of EUV lithography (EUVL) processes are presented. The success of an EUVL process is defined as a complete removal of the resist material within one set of designated volumes and a complete retention of the resist material within another set of designated volumes in the resist film. We demonstrate that, under certain assumptions, the probability calculation reduces to the well-known problem of calculation of probability of excursion of a certain Gaussian random field. The methods to calculate the probability of success/failure of a lithographic process are presented, including the Monte-Carlo methods, methods based on factorization of a covariance matrix, methods based on Mahalanobis distance, and the methods using Rice’s formula and its variations. A particular attention is paid to the methods applicable to full chip OPC and OPC verification. The results from the proposed methods are tested in simulations and by comparison with experimental data.

Extending extreme ultraviolet (EUV) single exposure patterning to its limits is dependent on eliminating its stochastic defectivity. Along with developments in photoresist platforms, the patterning film stack also needs to be considered. The material immediately underneath the photoresist is expected to have significant impact on both lithographic and pattern transfer performance. By designing the resist substrate interface with high EUV absorbance, there is potential to increase the EUV quantum yield of the exposure process. Increasing the selectivity to organic layer offers the opportunity to modulate stochastic defects through etch strategies. This paper will demonstrate the patterning of various chemically amplified resists on a high-Z metal-based hardmask. The potential for dose reduction, higher etch selectivity, and defectivity improvement from a high-Z hardmask will be discussed. Deposition-trim etch techniques will be used for decreasing the transfer of stochastic defects to the underlying substrate. Sub-32nm pitch trench patterning, defectivity, and electrical yield for this patterning stack will be highlighted.

With aggressive scaling of single-expose EUV lithography to the sub-7 nm node, stochastic variations play a prominent role in defining the lithographic process window. Fluctuations in photon shot noise, absorption and subsequent chemical reactions can lead to stochastic failure, directly impacting electrical yield. Fundamental characterization of the mode and magnitude of these variations is required to define the threshold for failure. In this work, a complementary series of techniques is enlisted to probe the nature and modulation of stochastic variation in single exposure EUV patterning. Unbiased line edge roughness (LER), local critical dimension uniformity (LCDU) and defect inspection techniques are employed to monitor the frequency of stochastic variations leading to failures in line/space and via patterning. Using this methodology, we explore the modulation of stochastic variations by different photoresists and illuminations, with emphasis on material and process down-selection for improved yield at the sub-7 nm node.

The effective dose delivered by an EUV lithography cluster is composite function of the dose provided by the scanner EUV radiation source and illuminator, the reflectance of the EUV mask, the transmission of the scanner projection optics and the PEB conditions experienced by the EUV sensitive imaging resist. Open frame test wafer exposures and the sub-E0 analysis technique described at SPIE2018 have been adopted to characterize and monitor the impact of the factors above on the effective dose stability and uniformity. Wafer exposure sequences and layouts, and the details of the analysis methodology were customized to study adverse dose factors in each of the areas described above.

Mask technology is key to enabling the progression of advanced IC technology nodes into the realm beyond 7nm node logic. In particular, extreme ultraviolet lithography (EUVL) relies heavily on the mask in order to achieve adequate process window (PW) and final yields. In the EUVL environment, mask process development and verification becomes increasingly more difficult and costly. The mask manufacture costs are driven by multiple factors including significantly more expensive mask blanks and increased ebeam write times. Wafer verification of mask process improvements is very difficult with the relatively low number of early adopters of EUVL and high cost associated with processing non-product wafers with that technology. It is therefore useful for mask manufacturers and wafer lithographers to collaborate to develop low cost mask process screening techniques as a precursor to committing valuable EUV exposure time for final verification. Previously, we demonstrated that by utilizing a toolkit of mask and wafer analytical techniques known as advanced mask characterization and optimization (AMCO) [2], we were able to predict wafer defectivity on 1D and 2D metal structures and optimize a mask process to enable 30nm pitch interconnect in a single exposure step using EUVL [1]. In this study, we use similar methodology to develop a mask process for beyond 7nm node logic contact and via layers. These structures pose a different set of challenges than the metal layer. Contact hole area loss, corner rounding (CR), and mask process-induced x-y error on asymmetric holes must be optimized to deliver the required capability. Additionally, sub-resolution assist features (SRAFs) become relevant at this node. Resolving these on the mask is critical. Here, we describe the development of a mask process to overcome these challenges. We use advanced modeling techniques including AMCO to characterize the process improvements and predict wafer performance. Mask process improvements that are characterized include both physical mask process components as well as write data optimization techniques, i.e. mask process correction (MPC).

Sustaining optimum EUV litho-cluster performance requires monitoring the dimensional variance of representative patterns at high speed and sampling density. We present NXE3400 monitoring results obtained with an optical technique whose amplified sensitivity to dimensional variance increases with decreasing pitch. The scalable technique enables quantitative tracking of site-to-site pattern profile and fidelity and it is applied here on 5nm-node-relevant EUV dense features (line/space width, line/space ends in elongated contact holes). We benchmark against conventional monitoring methods to demonstrate capability improvements. We finally introduce complementary methods which enable a robust manufacturing insertion of EUV technology; such as 2-dimensional feature monitoring and stochastic defect prediction.

In this paper we show experimental verification of the feasibility of printing pitch 40x70nm hexagonal holes using EUV single patterning. We show that at a local CDU (LCDU) of 2.7nm and an exposure dose of 54 mJ/cm² a defect rate smaller than 7x10^-9 is observed. This result was enabled by optimization of the illumination source and improvements in the resist. Resist selection identified multiple candidates that show a promising LCDU performance and optimization of the processing conditions resulted in improved performance. Experimental validation of the defect performance was done using HMI eP5 on the baseline process. Assessment of the LCDU performance for EUV single expose at pitches beyond 40x70nm, showed promising results.

As future patterning processes reach the limit of lithographic printability, continuous innovation in mandrel trim or shrink strategies are required to reach sub-20 nm line-space patterning. Growing concerns of lithography defectivity, mask selectivity, line edge roughness (LER), line width roughness (LWR), and critical dimension uniformity (CDU) present significant challenges towards this goal. The authors compare various alternative mandrel trim strategies to highlight potential solutions and drawbacks towards enabling successful printing of mandrels used in extreme ultraviolet (EUV) multi-patterning schemes. Through this comparison, the authors demonstrate the challenges of maintaining adequate pattern transferability while keeping aspect ratio-driven line roughness and material selectivity under control. By process partitioning, the limitations of traditional lithography and etch trimming strategies are highlighted, suggesting the need for new methods of CD reduction after the pattern has been transferred. These new trimming methods offer flexibility in CD control without negatively impacting the mandrel profile and demonstrates better tunability across different material sets, allowing for evaluation of different mask and mandrel material combinations for downstream process optimization.

Multiple ASML’s NXE:3400B scanners are installed at the factories and slated to go to a high volume manufacturing (HVM) phase. The latest generation of the scanners NXE:3400C has an improved performance and availability also due to availability improvement of the EUV sources by implementing modularity concept. In this paper, we provide an overview of a tin laser-produced-plasma (LPP) extreme-ultraviolet (EUV) sources at 13.5nm enabling HVM at the N7 node and beyond. The field performance of the source at 250 watts power including the performance of subsystems such as the Collector and the Droplet Generator will be shown. Progress in the development of key technologies for power scaling towards 500W will be described.

Gigaphoton develops CO₂-Sn-LPP EUV light source which is the most promising solution as the 13.5nm high power light source for HVM EUVL. Unique and original technologies including; combination of pulsed CO₂ laser and Sn droplets, dual wavelength laser pulses for shooting and debris mitigation by magnetic field have been applied. We have developed first practical source for HVM; “GL200E” in 2014. Then it is demonstrated which high average power CO₂ laser more than 20kW at output power in cooperation with Mitsubishi Electri. Pilot#1 is up running and it demonstrates HVM capability; EUV power recorded at 111W on average (117W in burst stabilized, 95% duty) with 5% conversion efficiency for 22 hour operation in October 2016. Availability is achievable at 89% (2 weeks average), also superior magnetic mitigation has demonstrated promising mirror degradation rate (= 0.5%/Gp) at 100W or higher power operation with dummy mirror test. We have demonstrated actual collector mirror reflectivity degradation rate is less than 0.4%/Gp by using real collector mirror around 125W (at I/F clean) in burst power during 30 Billion pulses operation. Recently we have redefined target power higher >330W and its development plan. Also we will update latest challenges for >330W average operation with actual collector mirror at the conference.

With the introduction of the NXE:3400B scanner, ASML has brought EUV to High-Volume Manufacturing (HVM). In this context, ASML is pursuing a dual-path approach towards zero reticle defectivity: EUVcompatible pellicle or zero particles towards reticle by advanced particle contamination control. This paper will focus primarily on the approach of advanced particle contamination control and on the understanding of EUV-induced plasma to control both release and transport of particles within the scanner. This paper will present our advancements in understanding and control of particle forces related to the EUV-induced plasma, for EUV sources up to 250W and beyond. This will combine models and simulations with off-line experiments as well as in-situ scanner tests. It will be shown that our understanding of the underlying mechanisms of plasma-induced release and transport of <1um particles enables us to manage defectivity levels down to be compatible with HVM requirements for sub-10nm node lithography.

With the persistent drive to enable EUV lithography (EUVL) for the continuation of pattern scaling and the close collaborations between suppliers and customers, tremendous progress has been made in the last five years in EUV mask infrastructure development. With the advent of actinic pattern mask inspection (APMI) tool, the only remaining EUV mask infrastructure gap until recently has been closed. We will present real-case examples from inspection of 7nm and 5nm logic node EUV masks with APMI in operation at Intel mask shop and demonstrate that actinic inspection provides defect detection capability beyond the traditional DUV optical and e-beam mask inspection (EBMI) tools for defect control and the guaranty of mask quality. In addition to the main focus on APMI and through-pellicle inspection in this paper, we also provide a brief discussion of other key EUV infrastructure modules for mask production in current EUVL at 0.33NA and future technology extension to enable high NA EUVL at 0.55NA.

Toward logic 3nm node and beyond, alternative absorber material is required to reduce mask 3D effect. One of the promising candidates is attenuated phase shirt type absorber which is effective for specific pattern shape such as contact hole. We developed novel high reflective phase shift type absorber. In this paper, we will report the blank and mask properties for this high reflective phase shift type absorber material.

Ni is a promising candidate for future EUV absorber material due to its superior optical properties. However existing reports on gas phase patterning of Ni are limited and not precise enough to support the application. In this work, we present a two-step reaction ion etch (RIE) and atomic layer etching (ALE) process, in which the a cyclic exposure to chlorine and hydrogen plasmas in the RIE stage removes the bulk of the majority of the exposed Ni thin films, and a cyclic exposure to oxygen plasma and gas phase formic acid tailors the surface chemical contrast to form a directional etching profile. In the ALE step, the final sidewall angle is defined by directional ions that simultaneously control the subsequent etching selectivity and anisotropy. Specifically, the oxygen plasma exposure converts the surface to an oxide layer which has a large chemical reactivity contrast to the organic acid, therefore the reaction self-terminates upon the complete removal of the oxides, leaving an exposed metallic surface on the sidewall with great verticality. The self-limiting nature of this process is backed by the removal rate difference of over 100:1 between etching NiO x and Ni using formic acid vapor. The effect of acid exposure pressure is studied, and a higher pressure results in a more vertical sidewall profile. Structures of post-etch patterns were measured using AFM, SEM and TEM while the chemical compositions are confirmed by XPS and EDS. This approach is free of noble ion bombardment – greatly reducing adventitious sputtering – and has a much more forgiving process windows comparing to other reported approaches. A final sidewall angle of 87 has been observed for the first time using this highly selective self-limiting technique, which can be further improved with process optimization.

Industry has made significant progress advancing EUVL technology the past several years leading to its full adoption for N7 node production. To extend EUVL technology well into 3nm node and beyond, a new high-NA EUV platform is needed. These new EUVL scanner platforms plan to use anamorphic optics with 0.55NA, which introduces several additional mask related imaging complications. Current scanners have light incidence of 6°, while the new 0.55NA systems will use < 6° incidence. At mask level, the combination of EUV light at oblique incidence, absorber thickness, and non-uniform mirror reflectance through incidence angle, creates mask-induced imaging aberrations, known as mask 3D effects (M3D). Additionally, these high NA systems will have non-telecentricity, which has shown to cause H-V bias due to shadowing, pattern shift through focus, and image contrast loss due to apodization by the reflective mask coatings. All these factors point to the need of thinner EUV mask absorber, which will dramatically reduce these effects. A possible mitigation for the M3D effects is to use mask absorber materials with high extinction coefficient κ and refractive coefficient n close to unity. Current EUV masks use Ta-based absorber on top of MoSi reflective multilayer. Multiple studies have shown that optical constants for the 60nm Ta-based absorber thickness at EUV wavelength do not provide optimal wafer imaging for high NA platforms. Alternative absorbers with higher absorptivity than Ta, such as Ni, Co among others have been proposed and, from simulation work, have shown improved high NA imaging at < 40nm thickness. The replacement of Ta-based absorber by new alternative material is an arduous task for the mask industry. The new absorber material must not only meet improved imaging criteria, but it must also meet required material properties which makes its compatibility with different aspects of mask blank and mask manufacturing process very critical. In this work, we have embarked on alternative thin absorber material evaluation and characterization studies to find and assess viability of new absorber material, which best meets EUV mask fabrication requirements for impending N3 and below advanced semiconductor device nodes. While number of high-k materials were considered and evaluated, material selection narrowed our focus on one unique composite material, which met our critical requirements of low EUV reflectivity, fast etching rate and high cleaning durability. Material evaluation consisted of studying mask process modules from absorber etch, ebeam patterning capability, post-exposure processing and cleaning, defect Inspection, and finally EUV wafer printability. In these studies, we also evaluated this new material for critical mask patterning performance looking at key metrics such as resolution, CD linearity, proximity, control (targeting and uniformity), and LER/LWR. We also characterized pattern fidelity looking at complex mask designs for 5nm node and beyond using internally developed advanced mask characterization methodology emphasizing 2D pattern characterization including sidewall angle (SWA) assessment of patterned absorber.

As extreme ultraviolet lithography (EUVL) technology progresses towards and below sub-7nm generations, polarization effects will begin to have an impact. Moving to higher NA will introduce polarization effects at all locations: the mask plane, the optics, and the image plane. Contrast losses at a larger angle can occur from the interference at the image plane that will become significant for smaller geometries approaching the EUVL wavelength. Some of these contrast losses can occur at the mask where the polarized reflectance from an EUV mask is affected by the multilayer reflective stack and patterned features. This paper explores the polarization effects that are induced by EUVL masks for sub-7nm lithography. The results show a polarization-dependent induction and attenuation of current in EUVL mask structures, especially as mask pitch decreases below 6λ.

We present a comparison of experimental techniques for measuring the as-built phase shift of EUV photomasks to meet the unique requirements for EUV lithography at the resolution limit. Attenuated phase-shift masks provide superior image quality for certain applications such as low-k1 contact and pillar arrays, offering increased throughput and reduced stochastic printing failures. But whereas the traditional phase-shift is π, rigorous electromagnetic simulations suggest the optimal phase-shift for an EUV photomask must be adjusted to account for Mask 3D effects, which are themselves difficult to measure. In this work, we explore at-wavelength metrology including reflectometry, scatterometry, and phase imaging for measuring multilayer and absorber reflectance, as well as complex scattering amplitudes for a grating with pitch p = 420nm and duty-cycle D = 0.33. Using rigorous electromagnetic simulations, we find that relying only on reflectometery and the Fresnel thin-mask model predicts the complex scattering amplitudes with 22% accuracy due to 3D effects, whereas a combination of scatterometry and through-focus imaging can achieve a promising 0.6% accuracy, and a combination of scatterometry and Zernike Phase-Contrast imaging can achieve a superior 0.1% accuracy. Experimental results based on imaging and scatterometry clearly display 3D effects that differ substantially from idealized rigorous simulations, suggesting the difficulty of accurately predicting 3D effects and hence the need to accurately measure them.

Mitigation of 3D-mask effects is a requirement for pushing high-NA (0.55) EUV lithography to its limits. Both the absorber and the reflective multilayer parts of the EUV mask contribute to the 3D-mask effects. This paper focuses on the investigation and optimization of the multilayer. The impact of different multilayer parameters on the imaging performance is investigated and used to explain the optimization outcome. Multilayer optimization yields better lithographic performance by including imaging metrics in the merit function instead of reflectivity data only. Different geometrical representations of the multilayer are optimized and their performances are compared. The results show a tradeoff among different lithographic metrics with improvements compared to a reference obtained from reflectivity optimization.

With conventional methods, improvements in optical proximity correction (OPC) runtime and accuracy can be challenging. Often improvements in accuracy have limited impact or require longer runtimes. Conversely, improvements in runtime often come at a sacrifice to overall correction quality. OPC industries have been developing and applying machine-learning (ML) methods to address both issues together, such as the Newron® machine learning family of products, which provides for both faster ML-based correction and more accurate resist models. Benchmark testing shows that ML-based correction prediction can yield runtime improvements of 30% or more without sacrificing pattern fidelity. It also shows that a ML resist model can deliver simulation accuracy 15% better than a conventional lithography model. This paper discusses the conversion flow from baseline OPC recipe to ML-accelerated recipe and presents results of a study that applies this technique to a sub-5 nm EUV test case, as well as results of a study that leverages a ML resist model to improve OPC accuracy.

Finding a process window and improving the yield for EUV single exposure nodes requires an understanding of stochastic defects. Stochastic simulations can be used as a tool to understand the influence of the process on defectivity. This presentation introduces stochastic model calibration with the purpose of matching lithographic observables. The validity of this approach will be shown based on comparison of measured defectivity data and matching stochastic simulations.

Advanced characterization techniques for nanoscale photoresist structure Miquel Salmeron, Slavomir Nemsak, and Patrick Naulleau Lawrence Berkeley National Laboratory, University of California Berkeley The spatial chemical structure of photoresists before and after processing is currently limited to a resolution of 10’s of nanometers, far above the desirable level of mono and sub-nm resolution that can enable the next generation of microelectronic circuits and devices. We will discuss in this presentation novel characterization techniques that will open the door for advanced spectroscopic characterization of polymer interfaces both in the latent image and the developed image. Among the new techniques being currently developed we will describe plasmonically enhanced infrared spectroscopy that provides FTIR spectra with nm x-y resolution along the interface between the photoresist and the substrate/underlayer. Additionally, we will present Standing Wave X-ray Photoelectron Spectroscopy in soft and tender X-ray regime, which provides elemental and chemical state composition in the z-direction across surfaces and buried interfaces. Currently these techniques have provided the first nanoscale characterization of the molecular structure of solid-liquid interfaces, revealing the chemical structure of the electrical double layer, of crucial importance in batteries, corrosion phenomena, and electrocatalysis. Applications to the polymer-substrate interface is currently underway.

As feature sizes continue to shrink, low-k₁ lithographic processes are required to advance chip technologies. In this context, both advances in optical systems and imaging capabilities as well as improvements in EUV photoresist have led to practical gains in resolution. This work focusses on our latest attempt to characterize photoresist performance by a metric that can effectively exclude all other contributors (e.g. scanner) and quantify the true photoresist capabilities. By a cautious analysis of the previous k₄ metric for several use cases and resist platforms (chemically amplified vs. metal-oxide resist), we propose an improved formula that incorporates key resist attributes.

Recently, the metal oxide nanoclusters (or nanoparticles) have attracted much attention as a promising candidate for a resist material used for the high-volume manufacturing of semiconductor devices. The radiation-induced reactions of ligands play an important role in the sensitization of metal oxide nanocluster resists. In this study, the radiation-induced reactions of carboxylic acid ligands were investigated using a pulse radiolysis method. The results obtained in this study indicated that the molecular structures of ligands affect their reactivity to cationic and anion species and the stability of their radical cations and anions.

The most pressing issue facing extreme ultraviolet lithography (EUV) as a multi-node technology is the inability to effectively manage stochastic effects. Fundamentally linked to the discrete nature of absorbed light (photons) and photo or electron-active components in the photoresist, stochastic effects lead to non-idealities in printed features, which for lines manifests itself as line edge and width roughness, and in the extreme cases, line breaks and bridges. However, measurement of stochastic propagation in photoresist has been hampered by the lack of metrology capable of probing the resist at intermediate stages of the patterning process. Here we present the use of atomic force microscopy (AFM) to measure the image in resist prior to photoresist development. In particular, the method was used to analyze the role of bake time on important resist metrics such as critical dimension, edge roughness, and roughness correlation length.

EUV photon absorption by a resist film leads to emission of a photoelectron and several low kinetic energy secondary electrons. The “universal curve”, used in X-ray photoelectron spectroscopy, suggests that the low kinetic energy electrons may travel tens to hundreds of nanometers in solids until they inelastically scatter. The fact that electrons travel long distances before they may initiate chemical reactions ultimately result in blur of the aerial image, reducing the contrast and subsequently resolution of the resist. In this work, we will present an experimental approach to determine secondary electron attenuation length (EAL) – the thickness of resist material required to reduce number of emitted secondary electrons to 1/e of initial. The EAL describes how far secondary electrons can travel in a resist film and is directly related to the electron blur. Possibilities to gain additional information on electron penetration depth in resist films will also be discussed.

For the past years, ArF immersion has been employed as the major lithography tool in the foundry manufacturing to fabricate the patterns of minimum pitch and size. However, for semiconductor scaling beyond N7 the application of EUV lithography is considered to be crucially important to overcome the physical limitation of ArF immersion and to realize even smaller patterns. In the case of ArF photo processes, the best mask size for a specific pitch could be selected with the consideration of optical performances such as NILS, MEEF, etc. In contrast, for the EUV processes the optical and resist stochastic effect should also be taken into account as an important factor in deciding the best mask size. In this paper, we are going to discuss the dose and mask size optimization process for an DRAM contact hole layer with EUV lithography utilizing stochastic simulations; this contains also the stochastic response of the resist. In order to calibrate a predictive stochastic resist model, which is required for this application, measurements of the stochastic resist response are necessary. In addition, the systematic and stochastic errors of CD-SEM measurements have to be estimated. We will compare experimentally obtained NILS and MEEF to simulated results, which are in very good agreement. Also, we show a comparison of experimental and computational analysis of LCDU (Local CD Uniformity).

In this preliminary computational chemistry study, we report excitation selectivity in a model tin-oxo molecular resist. Upon impact ionization, organic side chains connected to 6-coordinated tin atoms (located near the charge balancing ligands) are preferentially destabilized. Upon electron addition, conversely, side chains connected to 5-coordinated tin atoms (located on the central belt) are destabilized. Inferring from the binding energies, the ionization induced processes likely have a smaller spatial extent than electron attachment induced process.

EUV lithography wafer production has begun and consequently maximizing yield gains importance. One key component to high-yield lithography in manufacturing is using a pellicle to hold particles out of the focal plane and thereby minimize their impact on imaging. Using a pellicle simplifies manufacturing by eliminating wafer inspections that are used to indirectly monitor the presence of printable defects on the mask. The CNT-based pellicle – a membrane consisting of a network of carbon nanotubes – offers the advantage of very high EUV transmission and has demonstrated good durability at high EUV scanner power. Moreover, the microscopic properties of the network can be tuned by modifying several CNT membrane parameters, such as the individual CNT type and diameter, the degree of bundling, the density and the coating. The challenge is balancing these CNT material parameters for optimal performance in the EUV scanner: high transmission, low impact on imaging through scattered light, and low probability for particles to pass. Each of these areas will be addressed along with simulated and experimental data illustrating the value of a CNT-based EUV pellicle solution today and for the future.

In this paper, we present two methods for directly measuring the effective complex reflectance function of a patterned EUV mask. Obtaining this measurement can provide important insight into a number of key areas in EUV mask development, including obtaining a deeper understanding of mask 3-D effects, and characterizing and quantifying the amplitude and phase generated by attenuated and etched phase shift masks. The first method, Quantitative Zernike Phase Contrast Microscopy (QZPCM), works by modulating the imaging pupil function with several known phase shifts, and obtaining through-focus images of a target area for each pupil setting. The second method, Lateral Shearing Imaging (LSI), works by splitting and interfering two copies of the complex amplitude function in the image plane separated by a distance s. The resulting fringe pattern gives information about the derivative of the complex amplitude in the direction of the shift. We present results from two experiments: the first demonstrates QZPCM at EUV on the Berkeley SHARP microscope, and the second utilizes LSI in an optical prototype using a visible light laser source.

Coherence control and flexible pupil fill play a key role in the imaging of EUV reticles. This is also true for lensless metrology applications based on coherent diffraction imaging. We describe the concept and the key components of a Fourier synthesis illuminator designed to provide the RESCAN microscope with flexible illumination capabilities and to improve its resolution limit. In particular, we discuss the characteristics of the three mirrors of the new illuminator and the requirements for their multilayer coating.

Stochastics defect detection has been a topic of intense study by the extreme ultraviolet (EUV) patterning fraternity [1]. A large part of this initial feasibility work has been performed using electron microscope-based systems [1,2]. A limited sample area is imaged using electron microscopes and images are analyzed using offline analysis techniques [1,2]. However, to accurately quantify the stochastics failure rate, the entire area of interest needs to be inspected. Given such large area inspection requirements, automated and high throughput solutions are the need of the hour to enable stochastics quantification in HVM (high volume manufacturing). This paper demonstrates Broadband Plasma optical wafer inspection capability to capture two key defects on EUV layers a) missing contact in contact hole array patterns b) line breaks in line- space pattern.

As extreme ultraviolet lithography (EUVL) enters high volume manufacturing (HVM), the integrated circuit (IC) industry considers actinic patterned mask inspection (APMI) to be the last major EUV mask infrastructure gap. For over 20 years, there have been calls for an APMI tool for both the final qualification of EUV masks in the mask shop and for the requalification of EUV masks in the wafer fab1. Actinic, in this context, is matching the 13.5 nm scanner wavelength to that of the inspection tool so that all types of EUV mask defects can be detected. In order to enable EUVL HVM, we have developed and introduced the world’s first commercially available APMI tool. Actinic inspection enables HVM EUVL by ensuring that the EUV mask going to the EUV scanner is free from EUVprintable defects that may have been overlooked during EUV blank manufacturing or occurred during EUV mask manufacturing, cleaning and use. In this paper we will review EUV mask defect requirements from the maskshop and fab perspective, as well as capabilities of different inspection methods available for HVM. Further, we will provide an overview of the history of APMI tool development and highlight challenges and successes made when designing major components for the tool. APMI enables reliable detection of all classes of EUV-printable mask defects: small absorber defects, phase and amplitude defects in the multi-layer, In this paper, inspection performance of the APMI tool will be reviewed using representative cases from programmed defect masks with designs resembling real production cases. Finally, we will provide an outlook for the next steps in tool development including Die-to-Database inspection, throughpellicle inspection and platform extendibility to high NA EUVL.

We explore various resolution enhancement techniques and investigate their patterning benefits for via patterns of the 3-nm logic node using computational lithography. Simulations are performed by the method of source mask optimization (SMO) using the TachyonTM software. Key assessed process parameters include edge placement error (EPE), overlap process window, image NILS, local CD uniformity and NILS depth of focus (nDOF). Simulation results show that the current mask technology employing the standard Ta-based metallic absorber does not offer enough patterning performance for vias of pitch 40 nm and below. SMO results indicate that high-absorption absorbers give a clear improvement in best-focus shift and pattern placement error while phase-shift masks result in a significant increase of NILS and nDOF. EPE improvement of multiple technologies are also investigated. Novel EUV masks together with advanced imaging with low pupil-filling ratio and curvilinear OPC, combined with highresolution and low-roughness resist and enhanced etch process are among the key enabling technologies to extend EUV single patterning to 3-nm logic via layers.

In this work, we present a novel resistless patterning technique on hydrogen-terminated Si (100) using EUV-induced surface reaction. We has successfully demonstrated a direct Si pattern transfer of half-pitch 75 nm Si gratings using the EUV resistless patterning technique. The EUV-induced surface reactions could achieve direct patterning on functionalized Si surface and potentially overcome the resolution limitations of conventional photoresists in the semiconductor industry. Moreover, this approach can be used for better understanding of EUV exposure mechanisms by preparing simple molecular systems and studying the EUV-induced reactions.

Since the expected exposure dose of photoresist is an important parameter in the usability of EUV at a reasonable cost, the Lithography IRDS Focus Team (IFT) developed a photospeed projection as part of the 2019 IRDS lithography roadmap. We used two different methods to do this. Both methods started with the assumption that the 7-nm node critical layers exposed using EUV single exposure just meet the CDU requirement. This is equivalent to assuming that EUV users picked the fastest resist possible that still met requirements. Method one was to separate the required CD control for the 7nm node into a portion due to photon shot noise and a portion due to resist stochastics. Although the literature has widely varying numbers for the relative amounts of photon shot noise and resist stochastics, this approach did provide some insight. If one assumes most of the CDU is mostly due to photon stochastics, then substantial resist improvements from node to node, for example 20% less stochastic CD variation, doesn’t make much difference to the total stochastics. But if there isn’t resist improvement, then even doubling the source power from node to node won’t meet CDU requirements. If we assume that resist stochastics make up almost all of the CDU observed at the 7-nm node, then no reasonable amount of dose increase meets future requirements. Only if a combination of significant resist improvement and significant dose increases is used can you meet requirements. The second approach was to extrapolate using k₄.¹ In this case, assuming a typical current value of k₄, we estimate a typical line and space 7-nm node exposure dose to be to be 36 mJ/cm². We used the reported historical improvement in equivalent k₁ for resists as a guide for how fast resist k₄ can improve. This projected that typical resist exposure doses will have to increase over 30% per node on average to meet CDU and LER requirements for the leading-edge logic devices, leading to dose requirements < 100 mJ/cm² by the 1.5 nm node. We will discuss the challenges in meeting this roadmap and how the industry may adapt to the future challenge of lower noise imaging processes.

With the adoption of extreme ultraviolet (EUV) lithography across the semiconductor industry, there comes a need to develop new resist materials that will become the mainstay enabling technology. This development places pressure on the metrology community to develop adequate non-destructive techniques capable of measuring the sub-10 nm patterns that such resists are meant to produce. Recent progress in grazing incidence and resonant x-ray scattering techniques [1-3] have shown particular promise at being able to address part of these characterization challenges. In this paper, we highlight the potential for various techniques which use energy-tunable x-rays to probe buried structures critical to progress in EUVL. We show how resonant contrast can harness spatially distributed chemical heterogeneities within a resist in order to probe buried structures in a variety of photoresists containing key inorganic components. Specifically, by focusing on inorganic/organic hybrid resists, we demonstrate the potential insights from and correlation between contrast acquired at different elemental absorption edges and chemistries from a given resist.

When it comes to the patterning of periodic structures like lines and spaces or contact hole arrays interference lithography can be effectively applied. By the use of ultrashort wavelengths in extreme ultraviolet (EUV) range, structure sizes can be pushed into the sub-100 nm region. Using advanced interference schemes, such as achromatic Talbot imaging, high-quality large-area patterning can be realized, e.g. for resist stochastics tests as well as for advanced fabrication processes in research and also small-batch production. Since the Talbot lithography only requires medium spatial coherence and accepts broadband emission, the approach can be used not only with coherent radiation as provided by synchrotron facilities but also with compact plasma-based EUV sources. In this contribution the theoretical resolution limit for the achromatic Talbot lithography is determined by simulations for the use of optimized phase-shifting transmission mask concepts illuminated by compact discharge-produced plasma EUV source operating at a main wavelength of 13.5 nm. The further effects that reduce a practical resolution limit, such as mask imperfections, positioner instabilities and resist contrast are also considered. With the realized EUV laboratory exposure tool and polymer-based contact hole phase-shifting masks 28 nm resolution has been demonstrated so far. Main limitations are found in the mask fabrication process that cannot be further down-scaled due to increasing aspect ratios and pattern degradation of the mask structures. To extend the developed nanopatterning technology to the sub-30 nm region, optimized phase-shifting transmission masks have to be designed and fabricated, enabling a contrast-rich intensity modulation in wafer plane. As size of the mask openings is approaching the exposure wavelength, mask geometry has to be optimized for every node. In this paper rigorous simulations of new mask designs optimized for the achromatic Talbot lithography are presented. For selected phase-shifting transmission masks, the influence of the mask material and geometry on the resulting aerial image is evaluated along with an analysis of the theoretical resolution limit.

The application / effect of alternative developers for resist-based pattern defect mitigation was investigated. It was found that compared to the standard aqueous 0.26N TMAH developer solution, the same developer at lower concentration effectively improve the defect margin (indicated here in the form of a defect-sensitive exposure latitude (EL_X) and critical dimension margin (CDM_X) metric). Moreover, the existence of a developer concentration low limit, where EL_X and CDM_X significantly decrease, was observed. Alternative developers TEAH, TPAH and TBAH, when adjusted to the most effective developer concentration were found to improve defect margin. These results show the advantages of pursuing optimal developer solutions in mitigating resist-based defects while maintaining lithographic performance.

The EUV BeamLine 2 (EBL2) is being used to expose samples to EUV radiation for optics and mask lifetime testing. Before and after exposure the samples can be analyzed in-situ by X-ray photoelectron spectroscopy (XPS). During exposure the samples can be monitored in real-time by an imaging ellipsometer. We report on the development of two additional real-time diagnostic systems that further extend the capabilities of the EBL2 system, a thermal imaging system and an EUV reflectometer. The thermal imaging system monitors the sample surface radiation during accelerated lifetime tests and the EUV reflectometer is able to monitor the sample reflectivity in real-time. These diagnostics systems will allow for a more efficient use of the EBL2 beam-time and therefore speed up the development of EUV optics suitable for high source power and high NA imaging.

With the introduction of the NXE:3400B scanner, ASML has brought EUV to High-Volume Manufacturing (HVM). The high EUV power of >200W being realized with this system satisfies the throughput requirements of HVM, but also requires reconsideration of the imaging aspects of spectral purity, both from the details of the EUV emission spectrum and from the DUV emission. This paper will present simulation and experimental results for the spectral purity of high-power EUV systems, and the imaging impact of this, both for the case of with and without a pellicle. Also, possible controls for spectral purity will be discussed, and a novel method will be described to measure imaging impact of varying CE and DUV. It will be shown that CE optimization towards higher source power leads to reduction in relative DUV content, that the small deltas in EUV source spectrum for higher power do not influence imaging. It will also be shown that resulting variations in DUV do not affect imaging performance significantly, provided that a suitable reticle black border is used. In short, spectral purity performance is not a bottleneck for increasing power of EUV systems to well above 250W.

For generations lithographers have worked to overcome the difficulties associated with defect mitigation, and since EUV lithography has become mature enough for HVM this concern is warranting ever increasing attention to make such processes profitable. Even though much of the EUV defect effort is focused on stochastic defects, in this work we attempt to assess and understand process defects associated with the interaction between different films in an EUV stack. By understanding the behavior of specific underlayer materials and their chemistry within a given environment we have attempted to tune the surface energies to match the photoresist in the stack. With the correct process changes being applied, we have then worked to correlate the proper matching of surface energies with process defects. The current focus of our work is specifically line collapse, and we believe that developing a fuller understanding of the film interactions will ultimately lead to a more robust EUV process for HVM. We hereby present our work utilizing the SCREEN DUO coat develop track system with an ASML NXE:3300 in the imec Leuven cleanroom facility.

Extreme ultraviolet lithography (EUVL) systems struggle from both low source brightness and low source throughput through the tool. For these reasons, photon shot noise will play a much larger role in image process development for EUVL than in DUV processes. Furthermore, the lower photon count increases the stochastic variation of all the processes which occur after photon absorption. This causes the printed edge to move away from the mean edge with some probability. This paper will present a model form and calibration flow for including stochastic probability bands in compact models suitable for full chip simulation. This model form relies on calibrating to statistical data from a rigorous EUV stochastic lithography model calibrated to wafer experimental data. The data generation, data preparation, and model calibration flows for the compact stochastic probability bands will be presented. We will show that this model form can predict patterns which are prone to stochastic pattern failure in realistic mask designs, as well as how this model form can be used downstream for full chip correction (e.g., SMO, OPC and/or ILT).

Actinic patterned mask inspection (APMI) is the last major part of the infrastructure required for EUV lithography to be developed. A coherent imaging approach is proposed as a plausible solution requiring incremental improvements to available components. Diffractive optics direct a mW 13.5nm coherent beam to the mask and to interfere at the imaging detector. A motion system translates the mask between exposures. A non-iterative algorithm generates a complex reflectivity (mask image) estimate from each interferogram, in real time with GPU hardware. Realistic simulation results and requirements for the system components are presented. Except for the FEL that is likely required, many elements of the system are simpler and likely less costly than an incoherent imager. The system can be configured to image the mask through a pellicle. Of note, focus tolerance is loose as the focal distance is a software parameter.

This paper describes the impact of Extreme Ultraviolet (EUV) resist thickness on <30 nm Via Local Critical Dimension Uniformity (LCDU) measured during After Development Inspection (ADI) and After Etch Inspection (AEI). For the same post-etch CD targets, increasing resist thickness from 40 to 60 nm helped reduced CD variability. This work was performed via virtual fabrication using Coventor’s SEMulator3D® software and was confirmed experimentally. Post litho and post etch CD variations were introduced by considering lithography dose and focus variations in the model. Lithography and etch models were calibrated using 3D resist profile simulations from HyperLith® and post-etch Cross Section Scanning Electron Microscope (XSEM) images, respectively.

Gigaphoton Inc. has been developing a CO2-Sn-LPP (LPP: Laser Produced Plasma) extreme ultraviolet (EUV) light source system for high-volume manufacturing (HVM) semiconductor lithography. Key components of the source include a high-power CO₂ laser with 15 ns pulse duration and 100 kHz repetition frequency, a solid-state pre-pulse laser with 10 ps pulse duration, a high-speed Sn-droplet generator, a high-speed and high accuracy shooting system, and a magnetic field debris mitigation system. To achieve an in-band power of 330 W with long collector mirror lifetime and stable output, we improved the performance of key system components; especially, the laser beam quality at 27 kW CO₂ laser output power by upgrading the CO₂ laser beam transfer system, the dose stability and suppression of small Tin (Sn) debris, upgrading a shooting control system, improvement of the collector mirror degradation rate by the optimization of H₂ flow condition, and changing the EUV chamber structure. We achieved an in-band power of 250W under DC operation and demonstrated a power scalability up to 365W. This paper presents the key technology update of our EUV light source.

High-volume production of semiconductor devices by extreme ultraviolet (EUV) lithography has started since 2019. A high numerical aperture tool is planned to extend the use of EUV lithography. The trade-off relationships between resolution, line edge roughness (LER), and sensitivity are a significant concern for the extendability of EUV lithography. In previous study, the dependences of chemical gradient (an indicator of LER) on the half-pitch of line-and-space patterns, the sensitivity, the sensitizer concentration, and the effective reaction radius for deprotection were investigated using a simulation on the basis of the sensitization and reaction mechanisms of chemically amplified EUV resists. Although the relationships between resolution, LER, and sensitivity were formulated in the half-pitch range lager than 10 nm, they were deviated from the equations in the sub-10 nm half-pitch resolution region. Recently, the application of information science to the material engineering has attracted much attention. In this study, the feasibility of the application of machine learning to the analysis of trade-off relationships was investigated. The sub-10 nm region was fitted well using lasso.

Although being progressively introduced to mass production, extreme ultraviolet (EUV) lithography still faces major challenges for 5nm and smaller nodes due to the impact of stochastic and processing failures, resulting in very narrow defect process windows. ¹ These failures are strongly linked to critical dimension (CD) variations.² Therefore, careful control of CD is now directly linked to defect reduction in addition to more conventional in-film particles/developer residues reduction. Photoresist profiles are also believed to be one possible limiting factor and improvements via collapse control or increased resist mask thickness for etch transfer need to be considered. In this paper, most recent understandings regarding defect process window limitations and optimization of processes to further enable narrow pitch EUV lithography will be presented.

This paper describes the successful combination of masking or beam shaping elements based on thin membranes with the thermally ultra-stable properties of the ULE [1] glass substrate material. These kind of elements can be used in transmission in the optical beam path of EUV systems.

By applying a combination of new technologies, thin membranes can be directly manufactured on ULE substrates. The thermal expansion properties of the membrane based transmissive optical elements are dominated by ULE substrate. These elements enable a position stability of features in the range of 10-8/Kelvin for element sizes with a length of up to 150 mm.

The combination of technologies enables also the structuring of features with critical dimensions in the range of 100 nm on the membrane. The features on the membranes are for example absorbing or phase shifting structures or even holes. Holes in the membrane with diameters down to 200 nm have been demonstrated using e-beam based technologies.

These kind of transmissive optical elements could be used in EUV exposure systems, actinic mask inspection systems or other optical systems which are using light in the EUV wavelength range.

In EUV lithography, the short wavelength and residual mirror surface roughness increase the flare levels across the slit. As a key research point, the flares of different exposure fields are carefully discussed by numerical simulation. To ensure the effectiveness and practicability of our simulations, the test patterns are generated according to the general design rules for 7nm technology node. The NILS, process variation band (PVB) and MEEFs from mask optimizations and source mask optimizations (SMO) results are compared. From the comparisons, the constant flare has a greater influence on NILS and PVB than that on MEEF. In contrast, the flare map caused more reduction on the MEEF values.

Currently advanced DRAM design is beyond ArFi resolution limit, especially for the challenging processes in memory cell and core circuit pattern [1]. When devices keep shrinking, multi-patterning with ArFi becomes more and more difficult to reach the process requirements in terms of pattern decomposition, process window loss with complex process integration, defect, and immersion resolution limits. Besides multi-patterning also suffers design cost, mask learning cycle and layout restriction. Currently 0.33NA EUV can provide 16nm pattern single exposure and cover all design circuit requirement. High resolution enhances 2D pattern process window for friendly layout design and better OVL control so it is a good choice to introduce EUV process for DRAM manufacturing.

We evaluate to apply EUV in memory cell instead of the two possible solutions of SADP with cut layer and LELE trimming with multi-mask to simplify processes. Memory cell is periodic main feature for the most area on a mask and dominates the most EUV OPC run time in full shot correction. In this paper we try to find a best way to handle cell area OPC and evaluate single mask to accomplish memory cell patterning.

An EUV source optimization technique using compressive sensing is introduced in this paper. The pixelated source pattern is sparsely represented in a set of certain basis functions. Blue noise sampling method is used to select sampling points around the margins of the target layout for imaging fidelity evaluation. Based on the compressive sensing theory, the EUV SO is formulated as an l1-norm inverse reconstruction problem and solved by the linearized Bregman algorithm. Different types of sparse bases are also experimented in this paper to investigate their impact on the SO results. These bases include the 2D-DCT basis, spatial basis, Zernike basis, and Haar wavelet basis. Simulations show that ℓthe Haar wavelet basis results in the best imaging fidelity among the four types of bases.

With the introduction of Extreme Ultraviolet (EUV) lithography, the control of contamination has become crucial. Sources for contamination in EUV lithography scanners are not limited to only particle generation and release inside the scanner environment but may be introduced from outside as well, e.g. through translational and/or rotational (robotic) feedthroughs.

In this contribution we highlight our joint (TU/e and VDL ETG) research efforts aimed at the development of plasma-enabled contamination control strategies. The focus in this research is on airborne particles immersed in a low pressure gas flow that interact with both the afterglow of a plasma and an externally applied electric field.

A flexible experimental setup has been developed and will be introduced which is able to study the interaction between contaminants, plasmas and externally applied electric fields. Our results show that the designed configuration allows to carefully control the residual charges of the particles as well as their positions and trajectories with respect to the gas flow in which they are immersed. These results, together with the understanding of the underlying principle processes, open-up various possibilities to achieve improved cleanliness of the mentioned systems.

Quantum computing has been a long-anticipated emerging computational paradigm to complement and compete with conventional CMOS technologies. The last decade has featured reports of the initial development of using CMOS processing techniques for qubits and the atomistic fabrication of single atom transistors. Will the semiconductor industry embrace this new architype, and if so, how? When quantum devices begin to scale as predicted, will our industry be ready to integrate radically different architectures and device structures? Are there known obstacles not yet addressed that would enable the industry to more readily adopt and benefit from today’s achievements in the laboratory? What are the emerging designs and the potential fundamental challenges that are to be overcome? We have assembled a panel of experts that will share their insights on the state-of-the-art in quantum computing as well as intellectual leaders that will share their vision of the eventual merger of these new technologies with our computing capabilities today, even as lithography is approaching the near-atomic domain. Join us as we discuss the impending critical impact of quantum computing on the semiconductor industry.

SPIE will celebrate the entry of Extreme Ultraviolet Lithography (EUVL) into high volume manufacturing (HVM) with a special retrospective session. Highlights will include perspectives by true pioneers of EUVL, the formation of early teams, the gathering of resources and organization of the industry based EUV LLC and its partnering with the three national laboratory ‘Virtual National Laboratory’, the hand-off to SEMATECH and industry partners, the emergence of key suppliers, and IC manufacturers’ efforts in partnership with ASML and other suppliers.

The topic of this year’s symposium-wide panel is “A toast to Lithography's past: what we learned from technologies not used in HVM.” Over the years, a number of lithographic concepts were pursued for use in manufacturing when optical lithography finally reached its long-anticipated limits. While there appeared to be good reasons initially to invest substantially, in terms of people and money, in these concepts, most were not ultimately adopted for use in high-volume manufacturing (HVM). In this year’s symposium-wide panel, people who were involved in developing these concepts will share their perspectives on what lessons can be learned from their experiences.

ASML EUV Pioneers Photo Montage