Survey of chemically amplified resist models and simulator algorithms

Characterization for the photomask fabrication based on a high-resolution technique with a non-chemically amplified resist and a post-exposure bake

Progress towards 65 nm next-generation lithography requires unprecedented global CD uniformity, with the actual ITRS 2002 roadmap proposing 4.2 nm 3/spl sigma/ (dense lines) for 65 nm binary masks. Since resolution requirements are satisfied only by using chemically amplified resists (CARs), exposure and post-exposure bake (PEB) are key processes to successful mask making, both introducing global CD errors. Develop and etch processes potentially contribute further global CD errors. The global CD uniformity can be improved significantly by adaptive PEB, especially for CARs showing moderate to strong PEB sensitivity, like NEB22. With the 25-zone hotplate of the APB5500 bake system, facilitated through a novel calibration mask with 25 equidistant temperature sensors within the resist plane, an appropriate temperature profile can be applied during PEB. This temperature profile is automatically calculated by an adaptive optimization algorithm, based on 2-dimensional spline fitting of a CD measurement. A CD-uniformity improvement (dense lines) from 4.8 nm 3/spl sigma/ to 3.9 nm 3/spl sigma/ (/spl cong/20%) on a state-of-the-art production mask is achieved for the chrome layer (ASI, after strip inspection).

The second level exposure of Alternating Phase Shift Mask has to be accomplished by e-beam lithography. However, e-beam writing on structured open chrome layers may induce severe charging and yields to deteriorate overlay accuracy. In case of charging problems, applying of conductive top coats on resists help to reduce the distortion of the second layer. ESPACER makes the conductive layer on the e-beam resist and prevent a positioning error during e-beam writing by its shield effect. However, ESPACER required removal of its topcoat before post-exposure bake (PEB) because of the undesirable effect to the resist, due to the acid diffusion from the ESPACER film to high sensitive Chemical Amplified Resists (CARs). Our investigations were focused on the combination of the representative positive tone CAR: FUJIFILM ELECTRONIC MATERIALS CO., LTD and the new-type ESPACER called ESPACER 300F. ESPACER 300F was possible to use the PEB process before its removal, improved by exchanging a lower mobility surfactant than before. This material also had a good wettability, a low resist-thickness losses and kept good pattern shapes of the resist. Additionally, ESPACER 300F have little influence in both overlay accuracy at the variety of pattern density and sensitivity of the resist, therefore it would be the good material for making reticles for 65nm node and beyond.

Improvements in the final uniformity of masks can be shrouded by error contributions from many sources. The final Global CD Uniformity (GCDU) of a mask is degraded by individual contributions of the writing tool, the Post Applied Bake (PAB), the Post Exposure Bake (PEB), the Develop sequence and the Etch step. Final global uniformity will improve by isolating and minimizing the variability of the PEB and Develop. We achieved this decoupling of the PEB and Develop process from the whole process stream by using “dark loss” which is the loss of unexposed resist during the develop process. We confirmed a correspondence between Angstroms of dark loss and nanometer sized deviations in the chrome CD. A plate with a distinctive dark loss pattern was related to a nearly identical pattern in the chrome CD. This pattern was verified to have originated during the PEB process and displayed a (Final CD)/ (Dark Loss) ratio of 6 for TOK REAP200 resist. Previous papers have reported a sensitive linkage between Angstroms of dark loss and nanometers in the final uniformity of the written plate. These initial studies reported using this method to improve the PAB of resists for greater uniformity of sensitivity and contrast. Similarly, this paper demonstrates an outstanding optimization of PEB and Develop processes.

Several types of chemically amplified resists (CARs) have been evaluated in comparison with other types of resists. The evaluated CARs were found to have better performance than conventional resists and to satisfy the requirements for next generation reticle fabrication. Especially CARs have high sensitivity and high contrast enough to make photomasks with e-beam writers and have good dry-etch durability. We evaluated the allowance of baking conditions. It was found that it was important to minimize the dependence on prebake temperature as well as on post exposure bake temperature. The influence of airborne contaminants on post exposure delay (PED) stability of CARs was investigated. PED stability depends on resist materials and the ammonia concentration in the process environment. The use of a chemical filter is effective in improving PED stability. In conclusion, CARs can meet the requirements for resists used for next generation reticle fabrication.

Lithography simulators have become a standard tool in industrial and governmental research and development departments. In contrast to the modeling approaches for the optical system and for the lithographic performance of i-line resists, there is still no consensus on the modeling of chemically amplified resists (CAR). Existing models differ in the description of the kinetics and the diffusion phenomena during post exposure bake (PEB) and in the specification of the development rate. A modeling approach was established, that combines the light induced generation of photoacid, in- and outdiffusion of acid or base components, a generalized deprotection kinetics, Fickian and non-Fickian diffusion of resist components and an arbitrary development rate model. Existing models such as the effective acid model and a standard deprotection model for CAR can be considered as special cases of the implemented model. To evaluate the importance of certain options of the model and of the model parameters we have evaluated the performance of the model by comparing simulated CD data and resist profiles with experimental data.

Chemically amplified resists (CARs) are the current workhorse for photolithography, where higher resolution and smaller feature size represent a continual driving force for the semiconductor industry. As the feature size decreases to sub-30 nm, LWR and gate critical dimension (CD) control become serious concerns. In order to reach the goals in the ITRS, an unprecedented level of control of photoacid diffusion while maintaining the high resist sensitivity and resolution during image formation is required. CARs require a post exposure bake (PEB), typically performed on a hot plate at 90-150°C for 30-120 seconds, to complete the resist deprotection after photoacid generation. This bake step is a primary influence on resist performance as the time/temperature profile controls both the diffusion of photogenerated acids and the deprotection of the resist backbone. Sufficient time must be provided to achieve the level of deprotection required for the solubility switching in a developer, but the seconds timeframe of conventional hot plate PEB leads to an undesirable amount of acid diffusion. As long as the activation energy of diffusion is less than that for deprotection, higher temperatures for optimized time durations will result in reduced diffusion. However, traditional hot plate PEB cannot access times shorter than a few seconds. We utilize a laser (CO₂) based scanned heating system to achieve sub-millisecond to milliseconds in heating durations with temperatures up to the thermal decomposition limit of the resist. This research is aimed at using synthetic techniques to vary the structure of the photoacid generator (PAG) in order to learn about the role of PAG size and structure on acid diffusion during sub-millisecond heating. A variety of PAGs with different anion sizes have been synthesized and tested on the CO₂ laser system, and their lithographic performance and effect on acid diffusion has been studied.

Real-time control is the trend for photoresist processing in the advanced lithography. In this paper we develop an in-situ ellipsometer integrated with a programmable thermal heating system, which could perform real-time monitoring and control for critical dimension (CD) latent image through the entire post-exposure bake (PEB) process. The inline ellipsometry measurements are fitted into an electromagnetic wave model built on the rigorous coupled-wave analysis (RCWA) theory for CD latent image characterization. Thereafter a real-time control scheme is proposed by applying the programmable thermal bake-plate. The experimental outcome demonstrates a significant improvement on the final CD result comparing to the conventional baking method.

Recent studies on dynamic temperature profiling and lithographic performance modeling of the post-exposure bake (PEB) process have demonstrated that the rate of heating and cooling may have an important influence on resist lithographic response. Measuring the transient surface temperature during the heating or cooling process with such accuracy can only be assured if the sensors embedded in or attached to the test wafer do not affect the temperature distribution in the bare wafer. In this paper we report on an experimental and analytical study to compare the transient response of embedded platinum resistance thermometer (PRT) sensors with surface-deposited, thin-film thermocouples (TFTC). The TFTCs on silicon wafers have been developed at NIST to measure wafer temperatures in other semiconductor thermal processes. Experiments are performed on a test bed built from a commercial, fab-qualified module with hot and chill plates using wafers that have been instrumented with calibrated type-E (NiCr/CuNi) TFTCs and commercial PRTs. Time constants were determined from an energy-balance analysis fitting the temperature-time derivative to the wafer temperature during the heating and cooling processes. The time constants for instrumented wafers ranged from 4.6 s to 5.1 s on heating for both the TFTC and PRT sensors, with an average difference less than 0.1 s between the TFTCs and PRTs and slightly greater differences on cooling.

Physically-based photoresist models, such as those in PROLITH, have been very successful in describing photolithography from a continuum standpoint. These models allow engineers to accurately predict the final resist CD on the wafer and to analyze process robustness, such as calculation of focus-exposure process windows. However, as feature sizes continue to shrink, we are beginning to see yield-limiting phenomena that are due to the molecular nature of photoresist materials. One example of this is line-edge roughness (LER). LER is believed to be due to fluctuations during the exposure process (shot noise) and post-exposure bake (thermal diffusion and reaction). We present a model that explicitly takes into account the molecular nature of the photoresist during the exposure and post-exposure bake processes. We do this by writing a master equation that describes the probability that acid molecules are generated during exposure, and then describes the evolution of the acid, quencher, and blocking-group probability distributions during the bake process. We show how all the parameters in this model can be simply derived from the parameters in a calibrated PROLITH continuum model. Finally, we demonstrate prediction of LER from an accurately tuned PROLITH continuum model and compare the LER predictions with experimental results.

We simulated the process parameters of a pattern formation reaction that included during-the-exposure and post exposure bake (PEB) processes using an originally developed simulator. From the simulation results, the relationship between process parameters of pattern formation reaction and quencher concentration has been clarified. Moreover, we simulated the present target process parameters of extreme ultraviolet (EUV) resist for breaking the RLS trade-off. In this simulation, the process parameters were calculated from lithographic results (sensitivity, LWR, and CD) using real SEM images. This methodology was used to determine the process parameters required to break the RLS trade-off to obtain the required lithographic target of the EUV resist. We simulated the present lithography performance target using the process parameters of pattern formation reactions. These simulation results showed that a large reaction radius is necessary to break the RLS trade-off. Furthermore, we confirmed that increasing the PEB temperature leads to an improvement in the reaction radius. However, there is a discrepancy between the target radius and the controllable range of reaction radius that can be obtained by varying the PEB temperature.

A method for simulating exposure, post exposure bake and development process is described and applied for analyzing the problem of reflective stepped substrate. The method is based on the numerical solver of Helmholtz equation and the rigorous boundary condition which considers the entire system consisted of projection optics, photoresist and substrate. Diffusion equation and the ray tracing model are used for simulating the post exposure bake and the development process, respectively. The developed photoresist pattern on the stepped reflective substrate topography are obtained, and it shows irregular resist profiles due to the reflection of light at the step edge.

Post Exposure Baking (PEB) has been widely utilized in advanced lithography. PEB simulation is critical in the lithography simulation flow, as it bridges the optical simulation result and the final developed profile in the photoresist. The process of PEB can be described by coupled partial differential equations (PDE) and corresponding boundary and initial conditions. Recent years have witnessed growing presence of machine learning algorithms in lithography simulation, while PEB simulation is often ignored or treated with compact models, considering the huge cost of solving PDEs exactly. In this work, based on the observation of the physical essence of PEB, we propose DeePEB: a neural PDE Solver for PEB simulation. This model is capable of predicting the PEB latent image with high accuracy and >100 × acceleration (compared to the commercial rigorous simulation tool), paving the way for efficient and accurate photoresist modeling in lithography simulation and layout optimization.

Nanoimprint lithography (NIL) is the preferred technique for next-generation nanometer-scale patterning because of its cost effectiveness and simplicity when compared to conventional technology. The replicated pattern depends on the mold pattern, hence efficient methods for fabricating the fine mold are critically required and have recently become intensive subjects of research. This study reports a new method for the fabrication of a fine mold: this method uses low-acceleration-voltage (4kV) electron beam lithography with spin on glass (SOG), which is used as a positive-tone inorganic resist. Although our SOG process does not contain chemically amplified material, postexposure bake (PEB) was characteristically used in the authors’ process. The PEB process caused the annealing effect for the SOG and also suppressed the proximity effect. Consequently, a line-and-space pattern nanoimprint mold with a sub-50nm spacing was fabricated at 4kV using their process, and ultraviolet NIL was successfully carried out using the fabricated mold.

Post exposure bake (PEB) Diffusion effect is one of the most difficult issues in modeling chemically amplified resists. These model equations result in a system of nonlinear partial differential equations describing the time rate of change reaction and diffusion. Verifying such models are difficult, so numerical simulations are needed to solve the model equations. In this paper, we propose a high speed 3D resist image simulation algorithm based on a novel method to solve the PEB Diffusion equation. Our major discovery is that the matrix formulation of the diffusion equation under the Crank– Nicolson scheme can be derived into a special form, AX+XB=C, where the X matrix is a 3D resist image after diffusion effect, A and B matrices contain the diffusion coefficients and the space relationship between directions x, y and z. These matrices are sparse, symmetric and diagonal dominant. The C matrix is the last time-step resist image. The Sylvester equation can be reduced to another form as (<i>I</i>&otimes;<i>A</i> + <i>B</i>^<i>T</i>&otimes;<i>I</i>) X =C, in which the operator &otimes; is the Kronecker product notation. Compared with a traditional convolution method, our method is more useful in a way that boundary conditions can be more flexible. From our experimental results, we see that the error of the convolution method can be as high as 30% at borders of the design pattern. Furthermore, since the PEB temperature may not be uniform at multi-zone PEB, the convolution method might not be directly applicable in this scenario. Our method is about 20 times faster than the convolution method for a single time step (2 seconds) as illustrated in the attached figure. To simulate 50 seconds of the flexible PEB diffusion process, our method only takes 210 seconds with a convolution set up for a 1240&times;1240 working area. We use the typical 45nm immersion lithography in our work. The exposure wavelength is set to 193nm; the NA is 1.3775; and the diffusion coefficient is 1.455×10^-17m²/s at PEB temperature 150&deg;C along with PEB time 50 seconds with image resolution setup to be 1240&times;1240.