روش های جدید برای اندازه گیری دقیق کنترل کیفیت تصویر

New metrology methods have been developed to characterize exposure tool performance: SPIN for monitoring lens aberration, Z-SPIN for focus measurement, and EFSCP for checking effective light source profile. These methods, with design intention directed to measurement accuracy, measurement speed, and process robustness, provide increased accuracy and reduced time by a factor of 15 and 1/6, respectively, compared to traditional SEM metrology. SPIN allows direct measurement of lens aberration, achieving reproducibility of 0.8 mλ RMS. Z-SPIN translates focus (i.e. displacement in Z-axis direction) into an X–Y shift. Reproducibility of Z-SPIN measurement mostly depends on repeatability of the metrology tool, i.e. overlay metrology tool, which is ∼1 nm, and the measurement time of Z-SPIN is 1/8 of the conventional SEM measurement. EFSCP is a function to measure effective light source profile with high precision, achieving the repeatability of 0.00028 NAill for light source shift. While the traditional exposure tool performance inspections using SEM provide measurement reproducibility of 20 nm for IFD, 15 nm for astigmatism, and 2 nm for LR-CD, the new metrology methods provide 7 nm, 1 nm, and 0.3 nm, respectively, when estimated on the same scale. Also, the measurement result using the new methods shows very good agreement with the average of five SEM measurements for all the three inspection items. From this fact, it can be said that the traditional SEM-based inspection can be replaced by the new methods. To insure exposure tool quality in an appropriate manner, we at Canon recommend employing the new metrology methods in characterizing the exposure tool performance.

The endless drive toward smaller geometries, as represented by ITRS [1], requires continuous efforts to enhance CD control capabilities of exposure tools. Key contributors to CD control are lens aberration, focus, and effective light source profile of the illuminator. Traditionally, inspections of these CD control factors are performed based on SEM measurements; lens aberration is examined by measuring astigmatism, image field deviation (IFD), and left–right CD difference (LR-CD). SEM-based inspections, however, provide measurement repeatability of 3.6 nm for CD and 20.7 nm for focus, which accuracy is unsatisfactory for the inspection of the CD quality currently required of the exposure tool. It is therefore necessary to establish alternative inspection methods with higher speed, higher accuracy, and greater process robustness which can replace SEM-based inspections. Considering this, Canon have developed the following new inspection methods: (1) SPIN (Slant projection through a PINhole): Lens Aberration Monitor (2) Z-SPIN (Zex type of SPIN): Focus Monitor (3) EFSCP (Effective source check plus command): Effective Light Source Monitor Astigmatism is characterized by Zernike coefficients [2] using SPIN [3], and IFD is checked by obtaining focus values at each field position using Z-SPIN [4]. Also, LR-CD is characterized by Zernike coefficients using SPIN and by effective light source profile using EFSCP [5]. In this paper, we will describe the principle and performance of the new methods and present some experimental results in comparison with conventional SEM tests.

Table 1 summarizes measurement repeatability/accuracy and measurement time for astigmatism, IFD, and LR-CD, discussed in this study, to compare the conventional SEM inspection and the new method. It has been made obvious that the new method can be a good substitute for SEM when comparing the result of the new method with the average of the SEM measurements as discussed in Section 3. The new method provides improved measurement repeatability/accuracy, 3–15 times as high as the conventional method, which is sufficient for inspecting the exposure tool performance required in the future. Also, the new method significantly reduces measurement time to 1/3–1/6 of the conventional method.From the above consideration, Canon proposes replacing the conventional SEM-based exposure tool performance check with the inspection using the SPIN and Z-SPIN methods and the EFSCP function. In doing so, it is necessary to reconsider inspection items to match the new methods because the current inspection items have been defined assuming that SEM is used. For instance, the exposure tool performance is evaluated in terms of LR-CD when using SEM, but with the new method, it should be inspected focusing on coma and 3θ components of lens aberration and effective light source profile characteristics. As CD control is becoming tighter with shrinking linewidth, it will be more important to guarantee the exposure tool performance on a module basis. In this regard, the SPIN and EFSCP methods can be used to insure the quality of the projection lens and the illuminator, respectively, and will work as a key metrology means to guarantee exposure tool performance in the future.