EUV flare and proximity modeling and model-based correction

Zuniga, Christian; Habib, Mohamed A.; Word, James; Lorusso, Gian; Hendrickx, Eric; Baylav, Burak; Chalasani, Raghu; Lam, Michael

doi:10.1117/12.879488

Cited by 15 publications

(8 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to the long-range effects of the flare PSF and the high complexity of I 0 , directly convolving the PSF with I 0 could be very computationally expensive. To tackle this problem, previous work has shown that dividing a layout into suitably sized grids (e.g., 1μm×1μm) and calculating the layout density for each grid can achieve a good approximation of I 0 [13,19]. Then, by convolving the generated density map I D (x, y) at the coordinate (x, y) with the discrete PSF(x, y), we can derive the flare map I F (x, y) of a layout as follows:…”

Section: Flare Map Computationmentioning

confidence: 99%

“…However, the scattered flare reduces the contrast between bright regions (vacant regions) and dark regions (layout patterns), and may result in critical dimension (CD) distortion. Since the flare is proportional to the surface roughness of the optical system and inversely proportional to squared wavelength [12,19], EUVL suffers from rather high level of flare compared to traditional lithography technologies. It is reported by the alpha demo tool (ADT) at IMEC that the intrinsic flare is about 16% [12].…”

Section: Introductionmentioning

confidence: 99%

“…One is applying global CD resizing similar to optical proximity correction (OPC) on pattern features according to the flare value received by each feature [13,19]. However, previous work has reported that one percent change in flare level may cause 10 nm change in CD (CD sensitivity on flare is 10 nm/% Δflare level) at the 22 nm technology node and may be considerably larger for more advanced technology nodes [10].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Simultaneous flare level and flare variation minimization with dummification in EUVL

Fang

Chang

2012

Proceedings of the 49th Annual Design Automation Conference

View full text Add to dashboard Cite

Extreme Ultraviolet Lithography (EUVL) is one of the most promising Next Generation Lithography (NGL) technologies. Due to the surface roughness of the optical system used in EUVL, the rather high level of flare (i.e., scattered light) becomes one of the most critical issues in EUVL. In addition, the layout density nonuniformity and the flare periphery effect (the flare distribution at the periphery is much different from that in the center of a chip) also induce a large flare variation within a layout. Both of the high flare level and the large flare variation could worsen the control of critical dimension (CD) uniformity. Dummification (i.e., tiling or dummy fill) is one of the flare compensation strategies to reduce the flare level and the flare variation for the process with a clear-field mask in EUVL. However, existing dummy fill algorithms for Chemical-Mechanical Polishing (CMP) are not adequate for the flare mitigation problem in EUVL due to the flare periphery effect. This paper presents the first work that solves the flare mitigation problem in EUVL with a specific dummification algorithm flow considering global flare distribution. The dummification process is guided by dummy demand maps, which are generated by using a quasi-inverse lithography technique. In addition, an error-controlled fast flare map computation technique is proposed and integrated into our algorithm to further improve the efficiency without loss of computation accuracy. Experimental results show that our flow can effectively and efficiently reduce the flare level and the flare variation, which may contribute to the better control of CD uniformity.

show abstract

Section: Flare Map Computationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Simultaneous flare level and flare variation minimization with dummification in EUVL

Fang

Chang

2012

Proceedings of the 49th Annual Design Automation Conference

View full text Add to dashboard Cite

show abstract

“…Both the long-range effects of the flare PSF and the high complexity of I0 make directly convolving the PSF with I0 very computationally expensive. To tackle this problem, previous work has shown that dividing a layout into suitably sized grid (e.g., 1μm × 1μm) and calculating the layout density for each grid can achieve a good approximation of I0 [17,23]. Then, by convolving the generated density map ID(x, y) at the coordinate (x, y) with the discrete PSF (x, y), we can derive the flare map IF (x, y) of a layout as follows:…”

Section: Flare Map Computationmentioning

confidence: 99%

“…Flare reduces the contrast between bright regions (vacant regions) and dark regions (layout patterns), which may lead to critical dimension (CD) distortion. Since the level of flare is proportional to the surface roughness of the optical system and inversely proportional to the wavelength squared [15,23], flare is one of the most critical issues in EUVL as its applied wavelength is substantially shorter than those used in previous technologies.…”

Section: Introductionmentioning

confidence: 99%

Simultaneous EUV Flare Variation Minimization and CMP Control with Coupling-Aware Dummification

Liu

Chiang

Chang

et al. 2014

Proceedings of the 51st Annual Design Automation Conference

View full text Add to dashboard Cite

EUV flare and CMP metal thickness are two main manufacturability concerns for nanometer process technology. The two dummification objectives, however, are conflicting with each other in nature, but existing works only tackle them separately, leading to problem-prone solutions because optimizing one would deteriorate the other. This paper presents the first work that simultaneously considers both concerns during manufacturability optimization. Given a system's point spread function, our proposed method first finds an initial solution with better-than-state-ofthe-art EUV flare uniformity, then followed by gradient-guided optimization to iteratively refine density uniformity. Experimental results show the effectiveness of our method.

show abstract