High-precision lens element polishing process can cause the mid-spatial frequency surface error and thus affect the image quality. In order to investigate the influence of the mid-spatial frequency errors on the performance of the photolithography illumination system, a partially coherent optical model is used. The mid-spatial frequency error generated by the processing is relatively complex and can be simplified to a random function convolved with a structured mid-spatial frequency error. This paper selects the mid-spatial frequency distribution from an actual processed aspheric, and adopts the penumbra width of the relay lens as an evaluation target. The analysis results shown that the surface closest to the pupil plane might be most impact factor.
Top-Gaussian illumination generation technology has considerable application prospects in photolithography. We propose a detailed design and optimization method for the optical component that generates the top-Gaussian illumination field. This method can generate a field with a specific profile in the scan direction and a rectangular distribution in the nonscan direction. The energy density is reduced by jointly designing the first and second microlens arrays (MLAs). Thus, design freedom becomes more extensive. The requirements of the illumination field are met by the designed top-Gaussian illumination field, whose dimensions in the scan and nonscan directions are insensitive to illumination mode. Furthermore, the influences of the rectangular distribution dimension and the Gaussian distribution σ errors on the dimension of the top-Gaussian distribution are studied. The design results show that the maximum energy density in the second MLA can be reduced to 17.43% of that in the general design method. The analysis results indicate that the full-width-at-half-maximum error of the rectangular illumination field in the scan direction should be restricted within ±0.3 mm, and the σ error of the Gaussian distribution should be restricted within ±0.05 deg. The proposed method should increase the service life of the key component of the photolithography machine, and the cost of the manufacture and maintenance may be decreased.
Source mask and polarization optimization (SMPO) is a promising extension of the widely used resolution enhancement technology, source mask optimization (SMO), to further enhance chip manufacturability beyond 28-nm node. Our work is aimed to develop an efficient gradient-based SMPO method by employing the hybrid Hopkins–Abbe imaging model to fulfill the goal. In addition to source and mask variables, the model is adapted to also include polarization variables to realize the optimization. Compact formulas for forward and backward model application are derived. The computation benefits from precomputed transmission cross coefficients and features high efficiency. Validity of the method is confirmed by case studies. For dense array pattern case, the optimal source and polarization can be found analytically. SMPO optimized results match well with the theoretical expectations. In addition, process window, mask error enhancement factor, and normalized image log-slope for the studied cases all get improved over the counterpart SMO results, which employ commonly used polarization. Runtime analysis shows the method is computationally efficient. Our work provides a valid way to optimize polarization together with source and mask.
As the critical dimensions of integrated circuits continue to grow smaller, overlay error is becoming increasingly important. Overlay error is mainly determined by the telecentricity of the exposure system in a photolithography tool. Existing telecentricity-measuring methods—which are photoresist-based—are complex and cannot obtain the telecentricity in real time. Moreover, the obtained result is influenced by the photoresist performance. We propose a light-cone-central-line method that measures telecentricity without the need of photoresist. Compared with photoresist-based methods, the light-cone-central-line method can be used in real time, accelerating the setup and tuning of the exposure system and thus significantly reducing the complexities and expenses of the existing measurement methods. The proposed method was applied in a lithography tool with a 90-nm resolution, and the results for five field positions (the center position and four corner-field positions) are presented. This method is found to provide sufficient measurement repeatability, and it satisfies the telecentricity measurement requirements of the tool.
The wave plate is an indispensable optical element for transforming polarization state in various optical fields. In the retarder’s applications, precise measurement of the retardance is necessary. However, there is no common and effective way to analyze resolution of the retardance measuring instrument. For ensuring the retardance measurement accuracy, an analysis method involving a wedge wave plate with a tiny angle between two planes is proposed to characterize the resolution. The retardance of the wedge wave plate varies linearly and slightly due to its linear and diminutive variation of thickness. The diminutive variation of retardance versus displacement can be measured when the instrument has enough resolution. And the instrument demonstrates higher resolution when it can measure smaller retardance variation. In the experiment, a wedge wave plate is mounted on a two dimensions stage and its one surface is normal to the detecting light of the instrument. And then the retardance is measured along one certain direction and the results are fitted linearly. The method can be used to evaluate the resolution differences of retardance measurement instruments. The gained resolution of both two instruments is superior to 0.1nm. Through several experiments, it can be demonstrated that the proposed analysis method has the advantages of easy operation, high efficiency and simple configuration.
In the 28nm and below nodes lithography machine, the freeform pupil illumination has been widely used with the application of source mask optimization technology. As one of the most important implementations, the freeform illumination module is equipped with micro-mirror array (MMA), which can generate arbitrary pupil by adjusting its angle position. It is necessary to monitor the angle position of MMA real time for its significant function. However, there are several difficulties in the monitoring: 1) The monitoring unit could not disturb the working light path. This determines that the monitoring light should be glancing incident onto the micro-mirror; 2) The size of micro-mirror is relative small. And its rotation angle range is relative large in two dimensions; 3) There are several thousands of micro-mirrors. The crosstalk should be avoided in the monitoring method. In order to find a suitable monitoring method, the imaging and Fourier transform methods are studied. In the imaging method, the reticle is imaged onto the detector with the reflection of a single micro-mirror. The center of the reticle image is calculated to represent the angle position. In the Fourier transform method, the angular distribution of the light reflected by the micro-mirror is detected. And the angular distribution centroid is used to evaluate the angle position. In order to verify the feasibility and compare the performance of the two methods, the influences of alignment error and the scattered light are analyzed. The simulation results show that the Fourier transform method is insensitive to the scattered light and is independent to the relative position of the micro-mirror and Fourier transform lens.
With the application of source mask optimization (SMO) technology in the 28nm and below nodes photolithography machine, the freeform pupil illumination technology has been widely utilized to achieve resolution enhancement for various complex patterns. The freeform illumination module (FIM) equipped with micro-mirror array (MMA) are proposed, which could realize arbitrary pupil by adjusting the angle position distribution of MMA. Therefore, it is necessary to research the freeform pupil illumination technology in immersion photolithography machine. An excellent performance optical system for FIM mainly including homogenization unit, micro-lens array (MLA), MMA and Fourier transform lens is proposed in this paper. The homogenization unit is used to increase the uniformity of the beam incident onto MMA. The beam incident onto MLA is divided and focused on MMA. The focused sub-beams are reflected by micro-mirrors and then incident into Fourier transform lens. And the freeform pupil is generated at its back focal plane. In order to verify the feasibility of the designed optical system, three freeform pupils optimized by SMO are input into the designed FIM and the corresponding simulated pupils are exported. Furthermore, the photolithography performance simulations of the optimized and simulated pupils are implemented in optical model. The results indicate that their critical dimension (CD) differences are less than 0.5nm RMS for thousands of patterns in 40nm-80nm, such as line end, line space, contact hole, end to line, SRAM et. al., which shows that the excellent performance of the designed FIM.
The resolution limit is one of the key performance specifications of photolithography machine. And off axis illumination is one of the important resolution enhanced technologies. The generally used illumination modes include conventional, annular, quadrupole and dipole. And their performance is expressed by the characteristic parameters. To guarantee these parameters, the pupil correction unit should be adopted. Therefore, it is necessary to study the pupil correction technology for photolithography. In order to achieve flexible pupil correction, a method with correction finger is studied, which could change the regional energy by partial blocking effect. It is available to reduce regional energy by adjusting the width and length of correction finger. As a contrast, a method with grayscale filter is also analyzed. The grayscale filter has uniform transmission distribution in every region. The higher energy region corresponds to lower transmission distribution to achieve the energy balance. The comparison of the two pupil correction methods are analyzed firstly. The analysis results show that the two methods could improve pupil performance significantly and achieve the same correction results. Furthermore, the photolithography performance simulation is implemented. The results indicate that the critical dimension (CD) and H-V bias of the corrected pupils are improved consistently compared with the uncorrected pupils. In the application perspective, the method with correction finger is more flexible because its length could be adjusted to change relative blocked energy. However, the grayscale filter has to be replaced to change its correction effect.
The uniformity of the illumination field in the scanning direction is an important factor that affects the lithographic overlay accuracy as well as Critical dimension uniformity (CDU). With the improvement of lithography resolution illumination integrated uniformity is also increasing. To improving illumination integrated uniformity, a highly intelligent uniformity correction is introduced, which is used to correct the illumination integrated uniformity by inserting a plurality of independent movable correction plate arrays into the illumination field. In addition, a correction algorithm based on step by step is proposed. The simulation results show that the corrected illumination integrated uniformity is better than 0.3%, which is meeting the requirements of illumination integrated uniformity for 65nm node lithography.
Scanning slit is an important element to form illumination area and control exposure dose in step-and-scan lithographic system. If the penumbra of blade’s edge generated by scanning slit is too large, the exposure performance can be affected. Moreover, the distortion in the corner of illumination field also has great impact on the integral uniformity of illumination. Firstly, the generation principle of penumbra of blade’s edge at reticle plane was introduced according to the illumination principle of the step-and-scan lithographic system, and then the calculation expression of penumbra of blade’s edge at reticle plane was derived by analyzing the relationship between the intensity distribution of light and the location and the thickness of blades along the optical axis. Secondly, according to different types of blades, the distortion at the junction of adjacent blades of coplanar scanning slit was analyzed. At last, based on the optical model of step-and-scan lithographic system with the numerical aperture of NA0.75, the illumination fields at reticle plane generated by the four blades in scanning slit were simulated. The results indicate that the penumbra of blade’s edge generated by coplanar scanning slit could be significantly reduced and the distortion in the corner of illumination field could also be improved by optimizing the structure design of the blades of coplanar scanning slit properly.
Pupil parameters are important parameters to evaluate the quality of lithography illumination system. In this paper, a cloud based full-featured pupil processing application is implemented. A web browser is used for the UI (User Interface), the websocket protocol and JSON format are used for the communication between the client and the server, and the computing part is implemented in the server side, where the application integrated a variety of high quality professional libraries, such as image processing libraries libvips and ImageMagic, automatic reporting system latex, etc., to support the program. The cloud based framework takes advantage of server’s superior computing power and rich software collections, and the program could run anywhere there is a modern browser due to its web UI design. Compared to the traditional way of software operation model: purchased, licensed, shipped, downloaded, installed, maintained, and upgraded, the new cloud based approach, which is no installation, easy to use and maintenance, opens up a new way. Cloud based application probably is the future of the software development.
With the constant shrinking of printable critical dimensions in photolithography, off-axis illumination (OAI) becomes one of the effective resolution-enhancement methods facing these challenges. This, in turn, is driving much more strict requirements, such as higher diffractive efficiency of the diffractive optical elements (DOEs) used in the OAI system. Since the design algorithms to optimize DOEs’ phase profile are improved, the fabrication process becomes the main limiting factor leading to energy loss. Tolerance analysis is the general method to evaluate the fabrication accuracy requirement, which is especially useful for highly specialized deep UV applications with small structures and tight tolerances. A subpixel DOE simulation model is applied for tolerance analysis of DOEs by converting the abstractive fabrication structure errors into quantifiable subpixel phase matrices. Adopting the proposed model, four kinds of fabrication errors including misetch, misalignment, feature size error, and feature rounding error are able to be investigated. In the simulation experiments, systematic fabrication error studies of five typical DOEs used in 90-nm scanning photolithography illumination system are carried out. These results are valuable in the range of high precision DOE design algorithm and fabrication process optimization.
One method of realizing color holographic imaging using one thin diffractive optical element (DOE) is proposed. This
method can reconstruct a two-dimensional color image with one phase plate at user defined distance from DOE. For
improving the resolution ratio of reproduced color images, the DOE is optimized by combining Gerchberg-Saxton
algorithm and compensation algorithm. To accelerate the computational process, the Graphic Processing Unit (GPU) is used.
In the end, the simulation result was analyzed to verify the validity of this method.
Illumination uniformity is one of the key specifications of lithography illumination system because of its strong influence on the critical dimension (CD) uniformity in optical lithography. Refractive microlens array (MLA) has been extensively adopted in lithography system to achieve highly homogeneous illumination field with less light loss relative to diffractive element. Off-line homogenization inspection of the MLA provides important data for entire system integration. It is still a challenge work to investigate the optical performance for such high-end MLA with large clear aperture and high sensitivity to the incident light parameters. In order to address these issues, subaperture stitching method has been proposed to be applied and studied in this work. The feasibility of this method has been verified by theoretical simulation of a diffracting homogenizer. In the experiment, a corresponding optical setup is constructed, and a crossed-cylindrical single-plate MLA has been tested. The experimental results are consistent with the simulation ones. It could be concluded that subaperture stitching method is a powerful method to evaluate the homogeneous performance of MLA.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.