Critical Dimension Variation Research Articles

Extreme ultraviolet pellicle wrinkles influence on mask 3D effects: experimental demonstration.

Extreme ultraviolet (EUV) lithography uses reflective optics and a thick mask absorber, leading to mask 3D (M3D) effects. These M3D effects cause disparities in the amplitudes and phases of EUV mask diffractions, impacting mask imaging performance and reducing process yields. Our findings demonstrate that wrinkles in the EUV pellicle can exacerbate M3D effects. This imbalance results in critical dimension variation, image contrast loss, and pattern shift in mask images. Therefore, the use of a pellicle material with thermodynamic characteristics that minimize wrinkles when exposed to EUV rays is imperative.

CMOS-compatible manufacturability of sub-15 nm Si/SiO2/Si nanopillars containing single Si nanodots for single electron transistor applications

This study addresses the complementary metal-oxide-semiconductor-compatible fabrication of vertically stacked Si/SiO2/Si nanopillars (NPs) with embedded Si nanodots (NDs) as key functional elements of a quantum-dot-based, gate-all-around single-electron transistor (SET) operating at room temperature. The main geometrical parameters of the NPs and NDs were deduced from SET device simulations using the nextnano++ program package. The basic concept for single silicon ND formation within a confined oxide volume was deduced from Monte-Carlo simulations of ion-beam mixing and SiO x phase separation. A process flow was developed and experimentally implemented by combining bottom-up (Si ND self-assembly) and top-down (ion-beam mixing, electron-beam lithography, reactive ion etching) technologies, fully satisfying process requirements of future 3D device architectures. The theoretically predicted self-assembly of a single Si ND via phase separation within a confined SiO x disc of <500 nm3 volume was experimentally validated. This work describes in detail the optimization of conditions required for NP/ND formation, such as the oxide thickness, energy and fluence of ion-beam mixing, thermal budget for phase separation and parameters of reactive ion beam etching. Low-temperature plasma oxidation was used to further reduce NP diameter and for gate oxide fabrication whilst preserving the pre-existing NDs. The influence of critical dimension variability on the SET functionality and options to reduce such deviations are discussed. We finally demonstrate the reliable formation of Si quantum dots with diameters of less than 3 nm in the oxide layer of a stacked Si/SiO2/Si NP of 10 nm diameter, with tunnelling distances of about 1 nm between the Si ND and the neighboured Si regions forming drain and source of the SET.

Overlay and CD uniformity variation due to wafer thermal deformation caused by EUV exposure

The extreme ultraviolet lithography (EUVL) process allows for the manufacturing of better-performance chips. However, the EUVL exposure process might contribute to significant variations in the patterns of circuits, which can cause serious image placement and flatness change. Especially during the exposure, the EUV absorption might cause local wafer thermal deformation that might cause overlay and local critical dimension (CD) variation. In this study, we investigated wafer thermal deformation with different CDs and pitches with EUV exposure and thermo-mechanical behavior of the photoresist using the finite element method simulation. It is found that the thermal deformation caused during EUV exposure is different inside a die. In the direction along the scan, thermal deformation increases as it is near the center of the slit. The deformation in the perpendicular direction increases as it approaches the edges and deforms in the opposite direction relative to the center of the slit. In both directions, the thermal deformation increases gradually as the scan progresses, and it could be as large as ∼1.5 nm in a die. There was no difference in thermal deformation by CD change, but it changed with pitch.

A Study on the Resolution and Depth of Focus of ArF Immersion Photolithography.

In this study, the resolution and depth of focus (DOF) of the ArF immersion scanner are measured experimentally according to numerical aperture (NA). Based on the experiment, the theoretical trade-off relationship between the resolution and depth of focus can be confirmed and k1 and k2 are extracted to be about 0.288 and 0.745, respectively. Another observation for a problem in small critical dimension realization is the increase in line width roughness (LWR) according to mask open area ratio. To mitigate the trade-off problem and critical dimension variation, the photoresist thickness effect on depth of focus is analyzed. Generally, the photoresist thickness is chosen considering depth of focus, which is decided by NA. In practice, the depth of focus is found to be influenced by the photoresist thickness, which can be caused by the intensity change of the reflected ArF light. This means that photoresist thickness can be optimized under a fixed NA in ArF immersion photolithography technology according to the critical dimension and pattern density of the target layer.

Wafer level response to mask deficiencies in 0.55-numerical aperture extreme ultraviolet photolithography

BackgroundThis study investigates the impact of 0.55-numerical aperture (NA) imaging on wafer defectivity when multilayer ripple is present in the extreme ultraviolet (EUV) mask.AimWe investigate the impact of 0.55-NA anamorphic imaging on one-dimensional (1D) and two-dimensional (2D) horizontal and vertical mask feature deficiencies on wafer defectivity. The information in the study is intended to guide experimentation to aid in setting standards for EUV mask blanks in 0.55-NA lithography.ApproachThis study stochastically simulated 1D and 2D horizontal and vertical features using an extreme ultraviolet mask with varied random multilayer ripple configurations. The photoresist critical dimension (CD) was measured from the simulated wafer and used to generate statistical analysis of the simulation.ResultsHorizontal 1D features show an ∼2.5 × improvement in failure ratio versus vertical 1D features. Vertical 2D features appear to have a lower failure ratio compared with 2D horizontal features, but the difference is not as clear as found in the 1D case. The light scattered from the leaf shaped illuminator into the pupil background region due to multilayer ripple from 2D features appears smaller than in the 1D case. Potentially mask deficient structures may benefit from orientating the CD vertically on an anamorphic system to reduce wafer level defectivity.ConclusionsThere appears to be a multilayer ripple saturation level for 1D and 2D features at which defectivity and CD variation become constant. Horizontal imaging appears to be preferable for all CDs. The roughness threshold computational shortcut to mask deficiency characterization may not be sufficient for 2D features. The 2D analysis in this study does not change previous mask standard suggestions.

Effect of wrinkles on extreme ultraviolet pellicle reflectivity and local critical dimension.

Extreme ultraviolet (EUV) pellicles must have an EUV reflectance (EUVR) below 0.04% to prevent the reduction of critical dimension (CD). However, pellicle wrinkles cause localized CD variation by locally amplifying the EUVR. This study demonstrates that wrinkles can increase the pellicle's EUVR by approximately four times, and the CD drop depends on the relative position of the reflected light from the wrinkle to the 0th- or 1st-order diffracted light. The CD decreases by 6 nm. Therefore, even if the pellicle satisfies the requirement for the EUVR, we need to tightly control the generation of wrinkles to suppress CD variation during the entire exposure process.

A self-aligned lithography method based on Fabry-Perot resonator effect

Fabry-Perot (F-P) resonator is widely applied in the field of optics such as micro-ring resonator and photonic crystal, but there are not yet relevant reports about optical imaging. We propose a self-aligned lithography method based on F-P resonator effect and realize self-aligned effect successfully by controlling film stack parameters. The self-aligned structure is composed of Ag/photoresist/Ag film stack and wafer layer which contains SiO2 layer and SiO2/Si gratings. A rapid simulation model based on Rigorous Coupled Wave Analysis algorithm is established to optimize the best parameters of the self-aligned structure. Then rigorous Finite Difference Time Domain model and Finite Element Method model are used to verify the self-aligned effect and the results show that such process has extremely large mask critical dimension (CD) and overlay variation. The mask CD could be tuned more than 50% of structure period while the wafer CD maintains in the range less than 10%, and the maximum mask overlay offset could be ± 25% of half pitch while the wafer pattern center maintains unchanged. All the results show that the proposed method has lots of process robustness and could increase the application area of imaging lithography technology.

Fabrication tolerant coupling between silicon strip and subdiffraction V-groove waveguides

We consider the problem of efficiently coupling light into a recently proposed all-dielectric subdiffraction waveguide exhibiting an ultra-small mode area (An ∼ 10−3 λ 0 2 ), more than one order of magnitude lower than a diffraction limited strip waveguide (An ∼ 10−2 λ 0 2 ). Two prospective coupling solutions are compared with respect to coupling efficiency, fabrication tolerance, and optical bandwidth. The strategy based on adiabatic mode evolution is shown to be superior with respect to fabrication tolerance as it preserves ≥99% efficiency under +/- 10 nm critical dimension (CD) variations, whereas the directional coupling approach achieves only ≥60% efficiency for the same CD errors. Similar results are obtained with respect to optical bandwidth, with the nominal adiabatic mode evolution and directional coupling based designs achieving &gt;95% efficiency over wavelength ranges of &gt;200 nm and ∼50 nm respectively. The superior performance of the adiabatic design requires a coupler length in the range L ≈ 100–250 µm. The results yield a high performance, compact, and straightforward design solution for efficiently interfacing between conventional diffraction limited waveguides and all-dielectric subdiffraction waveguides with an ultra-small mode area.

Microsphere-assisted, nanospot, non-destructive metrology for semiconductor devices

As smaller structures are being increasingly adopted in the semiconductor industry, the performance of memory and logic devices is being continuously improved with innovative 3D integration schemes as well as shrinking and stacking strategies. Owing to the increasing complexity of the design architectures, optical metrology techniques including spectroscopic ellipsometry (SE) and reflectometry have been widely used for efficient process development and yield ramp-up due to the capability of 3D structure measurements. However, there has been an increasing demand for a significant reduction in the physical spot diameter used in the SE technique; the spot diameter should be at least 10 times smaller than the cell dimension (~30 × 40 μm2) of typical dynamic random-access memory to be able to measure in-cell critical dimension (CD) variations. To this end, this study demonstrates a novel spectrum measurement system that utilizes the microsphere-assisted super-resolution effect, achieving extremely small spot spectral metrology by reducing the spot diameter to ~210 nm, while maintaining a sufficiently high signal-to-noise ratio. In addition, a geometric model is introduced for the microsphere-based spectral metrology system that can calculate the virtual image plane magnification and depth of focus, providing the optimal distance between the objective lens, microsphere, and sample to achieve the best possible imaging quality. The proof of concept was fully verified through both simulations and experiments for various samples. Thus, owing to its ultra-small spot metrology capability, this technique has great potential for solving the current metrology challenge of monitoring in-cell CD variations in advanced logic and memory devices.

Multiscale simulations for exploring photo-chemical processes to mitigate the critical dimension variability of contact holes in EUV lithography

In extreme ultraviolet lithography (EUVL), the critical dimension (CD) variability of contact hole patterns results in pattern failures such as bridging or missing holes, which affects production yield.

Optimizing etching process recipe based on Kernel Ridge Regression

Abstract Exploring optimal recipes to reduce dimensional variations is critical in etching processes. Variations in critical dimensions that were acceptable previously can become problematic because of smaller node sizes and more complex structures. Dry etch can be a major source of variations and will be the focus of this research. Advanced Process Control (APC) has been widely studied in semiconductor manufacturing. Even though different APC methods have been developed to adjust recipes, it is challenging to explore an optimal recipe to achieve multiple critical dimensions. In this paper, a learning method based on Kernel Ridge Regression (KRR) is proposed to generate optimal recipes for multi-input multi-output (MIMO) systems. A KRR parameter optimization method is developed. To improve the recipe optimization process, a feedback fine tuning method is proposed. Experimental data in a dry etch process were collected and processed for model construction and recipe optimization. The results demonstrate the effectiveness of the proposed method in exploring optimal recipes for MIMO systems.

Real-time dose control for electron-beam lithography

Shot-to-shot, or pixel-to-pixel, dose variation during electron-beam lithography is a significant practical and fundamental problem. Dose variations associated with charging, electron source instability, optical system drift, and ultimately shot noise in the e-beam itself conspire to critical dimension variability, line width/edge roughness, and limited throughput. It would be an important improvement to e-beam based patterning technology if real-time feedback control of electron-dose were provided so that pattern quality and throughput would be improved beyond the shot noise limit. In this paper, we demonstrate control of e-beam dose based on the measurement of electron arrival at the sample where patterns are written, rather than from the source or another point in the electron optical column. Our results serve as the first steps towards real-time dose control and eventually overcoming the shot noise.

Understanding the Significance of Local Variability in Defect-Aware Process Windows

Process windows based on Mean CD (“Critical Dimension”) have been an analysis workhorse for estimating and comparing the robustness of semiconductor microlithography processes for more than 30 years. While tolerances for variation of CD are decreasing in step with the target CD, the acceptable number of printed defects has remained flat (Hint: Zero) even as the number of features increases quadratically. This disconnect between two key process estimators, CD variability and defect rate, must be addressed. At nodes that require EUV lithography, estimating the printed defects based solely on a Mean CD process window is no longer predictive. The variability / distribution of the printed CDs must be engineered so that there are no failures amongst the billions of instances, rendering the Mean CD, often measured on just hundreds or thousands of instances, a poor predictor for outliers. A “defect-aware” process window, where the count of printed defects is considered in combination with more advanced statistical analysis of measured CD distributions can provide the needed predictability to determine whether a process is capable of sufficient robustness. Determining process robustness where stochastics and defects are taken into account can be simplified by determining the CD process margin. In this work we study dense via/contact hole arrays exposed with 0.33NA single exposure EUV lithography after both the lithography and etch steps. We describe a methodology for expanding the analysis of process windows to include more than the mean and $3{{\sigma }}$ of the data. We consider the skew and kurtosis of the distribution of measured CD results per focus-exposure condition and compare/correlate the measured CD process window results to the CD process margin.

Self-aligned double patterning for active trim contacts with anisotropic pattern pitches in sub-20 nm dynamic random access memories

Background: With continuous decrease in the technology node of dynamic random access memories (DRAMs) down to sub-20 nm, the self-aligned double patterning (SADP) is an effective approach to generate two-dimensional (2-D) patterns, particularly contact arrays. Aim: We demonstrate a patterning scheme using the SADP technique to produce active trim contacts with anisotropic pattern pitches. Approach: The proposed scheme uses two consecutive spacer-formation processes. Results: By making the ellipsoidal core pillars and minimizing the spacer thickness, 2-D critical dimensions (CDs) for self-generated contacts match well with those for core contacts. In addition, an interesting cross-dependence of X -CD and Y -CD variations for the core and self-generated contacts is observed. Conclusion: This patterning approach is useful for forming active trim contacts in sub-20 nm DRAMs using fewer numbers of ArF immersion photolithography steps.

Optimization of photoresist development and DRIE processes to fabricate high aspect ratio Si structure in 5 nm scale

The fabrication of high aspect ratio (AR) Si structure with extremely small feature size is attracting more and more attention recently. In this work, a top-down scheme for fabricating nanostructures with 4.4 nm line width and 40:1 AR has been proposed and demonstrated. This technique utilizes electron beam lithography for nano-scale patterning and deep reactive ion etching (DRIE) for pattern transfer. In this method, a special development strategy is discussed to improve the perpendicularity of the photoresist mask, and a novel nano-scale barrel is utilized as the testing structure for the DRIE process optimization. To study the critical dimension (CD) variation in the DRIE process, the geometric morphology of the obtained vertical barrels is simulated by an ion transport model. It is found that the asymmetrical lateral etching encroachment decreased the CD loss by 50%. Without using a hard mask, this technique is fully compatible with Si-based microelectronic processes. Therefore, after further study of geometric properties of the obtained nano structure, this proposed microelectronic-compatible approach sheds light on manufacturing 3D nano devices towards 5 nm scale and beyond.

Research on critical dimensions variation based on improved optical lithography imaging model

Stray light exists in optical lithography, and causes the CD (Critical Dimension) Variation. Therefore, it is important to obtain an appropriate description model of imaging with stray light. An improved optical lithography imaging model considering medium-range and long-range flare is proposed by using the PSFF (Point Spread Function of Flare) model and the Kirk flare model. The proposed simulation model is confirmed to have much better precision when compared to the experiment data. The effect of flare on CD is studied, in which the total flare is the same but the ratio of medium-range and long-rang flares are varied. The simulation results show that long-range flare has a bigger impact on CD than medium-range flare. The greater the proportion of long-rang flare, the smaller line-width of CD will be. Thus, long-range flare should be taken fully into account in optical lithography.

Photoresist and stochastic modeling

Analysis of physical modeling results can provide unique insights into extreme ultraviolet stochastic variation, which augment, and sometimes refute, conclusions based on physical intuition and even wafer experiments. Simulations verify the primacy of “imaging critical” counting statistics (photons, electrons, and net acids) and the image/blur-dependent dose sensitivity in describing the local edge or critical dimension variation. But the failure of simple counting when resist thickness is varied highlights a limitation of this exact analytical approach, so a calibratable empirical model offers useful simplicity and convenience. Results presented here show that a wide range of physical simulation results can be well matched by an empirical two-parameter model based on blurred image log-slope (ILS) for lines/spaces and normalized ILS for holes. These results are largely consistent with a wide range of published experimental results; however, there is some disagreement with the recently published dataset of De Bisschop. The present analysis suggests that the origin of this model failure is an unexpected blurred ILS:dose-sensitivity relationship failure in that resist process. It is shown that a photoresist mechanism based on high photodecomposable quencher loading and high quencher diffusivity can give rise to pitch-dependent blur, which may explain the discrepancy.

Smartphone Display Based Photolithography to Fabricate Microdevices

We demonstrate a smartphone-based low cost photolithography technique that can be easily adapted, especially in university laboratories. In the initial experiments, we used only the display from a smartphone as the source of light and successfully transferred micrometer sized patterns in an optical mask on to a silicon wafer coated with photoresist microposit S1805. The fabricated patterns were identical to the features in the optical mask and had minimum critical dimension variation of 800 nm. However, the exposure time required to polymerize the photoresist film with adequate dosage of light was long (up to 40 minutes). In the next stage, we optimized the setup by making use of irradiation from UV LED arrays attached to the opposite long edges of the smartphone and reduced the required exposure time by 50%. Finally, we extended the lithography experiments to showcase potential of the proposed smartphone-based lithography as a low-cost tool to perform maskless lithography. For the later, an interdigital transducer (IDT) design was displayed and successfully transferred as a pattern on to a photoresist coated wafer sample. Further metallic electrodes of the IDT was fabricated on the sample through lift-off process, where the transferred pattern was utilized as sacrificial layer in the process. Finally, the fabricated IDT was verified as a sensor to detect conductivity of liquid samples, and thus, demonstrating the usability of the technique in designing laboratory exercises.

Key Effects and Process Parameters Extraction on the CD of Reactive Ion Etching (RIE) Based on DOE Modeling

The critical dimension (CD) is highly influenced by the reactive ion etching (RIE) of silicon in the CMOS technology. The CD has to be well-controlled since it is one of the most important features related to process stability and product quality. However, the RIE process involves a lot of parameters which are highly confounded and entangled and thus it is very difficult to analyze the process. To study and extract the influences of these parameters in terms of CD variation, screening experiments are carried out to obtain the necessary factors to develop a CD-related model. An elaborated model, based on the selected parameters, is then built up via applying the design of experiments and response surface optimization. This model is validated to be capable of characterizing the process as well as predicting the CD.

Influence of a wrinkle in terms of critical dimension variation caused by transmission nonuniformity and a particle defect on extreme ultraviolet pellicle

An extreme ultraviolet (EUV) pellicle is needed for the protection of EUV masks from defects, contaminants, and particles during the exposure process. However, the EUV pellicle can be easily deformed during the exposure process because it has an extremely thin thickness for high transmission of EUV lights. Due to the very thin thickness and the weak structure of the pellicle, a pellicle is easily deformed; a wrinkled pellicle causes an image distortion, which leads to critical dimension (CD) variation. In addition, a particle defect on an EUV pellicle can result from scanner building materials. Added materials of the particle defect on an EUV pellicle can also cause image distortion and CD variation. We investigated the impact of wrinkles and particle defects on the transmission and CD variation for the 5- and 3-nm nodes of isomorphic and anamorphic numerical aperture (NA) systems. The variation in transmission and the critical size of the particle defect with a wrinkled EUV pellicle were calculated to obtain the requirement of a CD variation of 0.2 nm for a EUV pellicle. As a result, a change in transmission of 1.9% (after two pass) resulted in a 0.2-nm variation in the CD for the anamorphic NA system (3-nm node), whereas a transmission variation of 1.3% (after two pass) caused a 0.2-nm CD variation in the isomorphic NA system (5-nm node). From these results, an allowable local tilt angle can be calculated; the allowable local tilt angle of an isomorphic NA system is 0.31 rad and that of an anamorphic NA system is 0.41 rad. When the particle defect is added on a wrinkled EUV pellicle, the critical size of the particle defect is 1.2 μm for the 5-nm node and 2.2 μm for the 3-nm node.