Novolak resists have been widely used in IC production and are used to this day in the production of flat panel displays (FPDs) and MEMS. However, with the advent of high-definition devices, FPDs must meet growing requirements for finer dimensions. These trends have generated requirements for higher sensitivity, higher resolution, and wider process margins for novolak resists. Using a lithography simulator with the goal of improving the performance of novolak resists, we examined various approaches to improving resist materials. This report discusses efforts to improve resolution and to broaden process margins using a novolak resin that exhibits a higher degree of fractionation than in the previous report (maximum fractionated resin) with the addition of low molecular weight phenol resins.
Novolak resists have been widely used in IC production and are still used in the production of flat panel displays (FPDs) and MEMS. However, with the advent of high-definition products, FPDs increasingly face requirements for finer dimensions. These trends have generated requirements for higher sensitivity, higher resolution, and wider process margin for novolak resists. Using a lithography simulator with the goal of improving the performance of novolak resists, we examined various approaches to improving resist materials. This report discusses efforts to improve resolution and sensitivity using highly fractionated novolak resins and adding low molecular weight phenol resins.
We obtained development parameters for a chemically amplified resist from calculations involving the conversion of the relationship between exposure dose and development rate to the relationship between protection ratio and development rate using the conventional ABC parameter[1] and development rate data (RDA data) [2]. However, calculations by this method require the ABC parameter. Since chemically amplified resists have no bleaching effect, the C parameter must be measured by the FT-IR [3-5] or coumarin addition method [6-8]. Given this constraint, we examined the method of obtaining development parameters based on the film reduction observed in the resist exposed or the film reduction observed after PEB, without using the ABC parameter. This paper presents the results.
This paper describes a study of a cross-linking reaction model for chemically amplified negative-type thick-film resists. Profile simulation is a major technique used to acquire experimental indicators. For this reason, numerous reports address simulation techniques, and many studies have focused in particular on chemically amplified positive-type resists, due to their role as mainstream resist materials used in the production of ICs. However, virtually no research has been performed on the profile simulation of chemically amplified negative-type thick-film resists. We measured the cross-linking reaction of a chemically amplified negative-type thick-film resist and created a new cross-linking reaction model. Our study demonstrates that this new model is more effective for thick-film resists than conventional models.
F2 lithography and 193nm immersion lithography are considered candidates for 65nm node lithography technology. Of these two, 193nm immersion lithography, the latest incarnation of ArF lithography, has attracted more attention. Immersion lithography is different from conventional dry lithography in that the resist is exposed in liquid. Thus, the resist materials leaching from the resist film during exposure and the dissolution of acids generated by the exposure cause problems. Particularly, the resist materials leaching tends to contaminate the surface of the lens. We have been conducting studies on the leaching during exposure using the QCM method. In the present work, we apply this method to the immersion exposure. We report here the results of an in-situ measurement of the resist mass change during immersion exposure and discuss our analysis regarding the resist materials leaching from the resist film during the exposure.
XP SU-8 3000 (hereinafter referred to as “SU-8”) thick-film resist is a chemically amplified negative resist based on epoxy resin. Here, we report on the profile simulation for this resist. Profile simulation is an important technique for planning experiments. Thus, there have been many reports on simulation techniques. In particular, many studies have been conducted on chemically amplified positive resists, as they are major resist materials used in the IC industry. However, there have been few simulation studies concerning chemically amplified negative resists. Under these circumstances, we have considered performing simulations on chemically amplified negative resist. The results of the simulation and the SEM observations are in good agreement. This study demonstrates that simulation is possible for a chemically amplified negative resist (SU-8).
This report describes the results of a study on resist profile simulation in proximity printing, using light intensity distribution and actually measured dissolution rate values, a method that takes the gap effect into consideration (the effect of the distance between mask and wafer on the aerial image and resist profiles). We calculate the light intensity distribution with the gap effect based on the Van Cittert-Zernike theory and on the Hopkins equation as a model of light intensity distribution of proximity printing in resist film. Dissolution rate values are obtained using an apparatus to measure resist film thickness during development. The resist profile simulation is carried out using the combined data thus obtained. To verify the validity of this simulation, we use an SEM to observe resist profiles obtained from a diazonaphthoquinone (DNQ)-novolak resin positive-type resist for thick films, varying the proximity gaps using the mask aligner, which uses light in the broadband wavelengths of 350 mm to 450 mm, and compare the results with the simulation. The results of simulation and those of the SEM observation are in agreement, proving the validity of our method.
KEYWORDS: Picture Archiving and Communication System, Optical lithography, Photoresist processing, Scanning electron microscopy, Temperature metrology, Systems modeling, Photomasks, Lithography, Microelectromechanical systems, Water
The effect of pre-baking conditions on the resolution and aspect ratio of thick-film resists is examined in order to improve resist processing performance. Resist samples are pre-baked at various temperatures and for various baking times, and a range of resist properties are examined. It is found that the pre-baking conditions affording the best resist pattern profile and development contrast are 125 °C for 7 min. The mechanisms responsible for the observed variations in pattern profile are studied by comparing and simulating the development activation energy, the change in the amount of solvent and photo active compound (PAC) during pre-baking, the residual solvent amount in the resist, and the transmission after pre-baking. The results indicate that there are two factors responsible for retarding the pattern formation process and causing degradation of pattern profile and resolution. One mechanism is N2 bubbling during development, which is caused by N2 trapped in residual solvent during exposure. The other mechanism is thermal decomposition of the PAC during baking, which weakens the retardation of development unexposed resist.
In view of the fact that little analysis of the mechanism for the achievement of high resolution or a high aspect ratio in the thick-film resist process has been performed, we study development properties with respect to differences between the development methods employed for pattern formation using thick-film resist. This study identifies the most effective development method for thick-film resist and reports the mechanism of development. For this investigation, we use a development rate measurement system, a mask aligner, and lithography simulator to examine the dipping development method, the step puddle development method, the vibration development method, and the reverse development method. We employ a thick-film positive resist composed of diazo-naphthoquinone (DNQ) and Novolak resin, which is coated on a silicon substrate to a thickness of 24 micrometers . After pre-baking, the coated substrate is placed in a vacuum dessicator to remove water, followed by immersion in deionized water for a fixed period. A mask pattern is transferred to the resist coated substrate with a Mask Aligner Q4000 made by Quintel Corporation, and then the rate of development is measured. A laser microscope analysis of the result indicates that the step puddle development method gives the highest pattern resolution and sharpness, followed by the vibration development method, the dipping development method, and lastly the reverse development method. The mechanisms of the development are studied by comparing the development contrast and the energy of activation involved in each development method, and by conducting resist pattern simulations. The results indicate that the factors responsible for retarding the progression of the development process and causing a degradation of pattern profile and resolution are development inhibition due to N2 released from inside the resist during the development process, and due to products that are dissolved in the development solution.
In the past there have been almost no analyses of mechanisms to achieve higher resolutions and higher aspect ratios in thick film resist processes. In this work, the authors measures the effects of water in a thick film resist on indenecarboxylic acid generation in thick film resist image formation, and studies the mechanism of resolution enhancement through hydration processes. In addition, a development rate measurement system, a system for analysis of resist reactions during exposure, and lithography simulations are employed to simulate profiles in thick film resists. Calculation results are then compared with actual resist profile. Specifically, a diazonaphthoquinone (DNQ)- novalic positive thick film resist is used; after coating to a thickness of 22micrometers and prebaking, the sample is subjected to dehydration in a vacuum desiccator, samples are prepared both immersed in Water and without such immersion, and photochemical reaction rates are measured during exposure, while also measuring development rates. From the measured results for photochemical reaction rates of the resist during exposure, the state of indenecarboxylic acid generation can be determined. The development rate data is inputted to the SOLID-C lithography simulator, thick film resist simulations are executed, and the results were compared with actual SEM observations of resist profiles. From the results of analyses of photochemical reactions during exposure, it is confirmed that differences in water quantity in the resist affect the generation o findenecarboxylic acid. Simulation results also confirms that by immersing the resist in water, the resolution is enhanced. The general trends of simulation results and actual patterning are in agreement, and it is concluded that in thick film resist, resolution is improved when water is present in the resist in sufficient quantities for idenecarboxylic acid generation.
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.