1. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. Improvements in Mask House processing due to thinner chrome have driven the binary blank migration path toward OMOG. The purpose of this study is to validate its lithographic performance. 10nm of chrome is used for OMOG mask processing and is removed before use in the wafer fab. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

2. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. (a) An example of an electromagnetic field (EMF) amplitude effect is the EMF Bias, defined as the required bias for a thick mask to print identical to an ideal thin mask approximation (TMA). EMFB is shown for polarized dipoles at 1.35NA where the anchor is where TMA prints a 45-nm mask space to a 50-nm wafer trench. (b) For binary imaging, small spaces tend to produce better contrast but worse MEEF than large spaces. This effect is amplified by EMF effects that make spaces behave effectively smaller than they are. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

3. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. 100-nm pitch (1.35NA, polarized c-quad illuminator): (a) simulation and (b) AIMS experiments show that smaller spaces improve contrast and degrade MEEF, both of which are amplified by EMF effects. The mask duty cycle is defined as the linewidth-to-pitch ratio. The predictive power of thin mask approximation (TMA) is important. TMA-P metric is defined to compare mask blanks as the difference in prediction between (c) EMF and TMA simulation predictions or (d) AIMS data and TMA simulation prediction. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

4. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. (a) Verification of EMF simulation by measuring the zeroth-order intensity of subresolution gratings on an AIMS tool. (b) Comparable 1-D process windows are seen on wafer data during preliminary analysis of 32-nm-node mask blanks. Duty-cycle-dependent trade- off between EL and MEEF was generally observed. (c) 32-nm contact data shows how, for identical mask dimensions, a thicker mask produces slightly worse MEEF for comparable mask sizes. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

5. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. EMF phase effect can be seen as a drift in best focus position through pitch and is about 2 to 3× stronger for BIM than OMOG seen from both (a) wafer and (b) AIMS data. AttnPSM is observed to have about 2× stronger effects than OMOG. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

6. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. The need for more computationally intense EMF-aware OPC is minimal with OMOG, as seen by the relative improvement in EPE with EMF-OPC compared to TMA-OPC on various 32-nm and 45-nm test masks. The benefit of EMF-OPC is significant with BIM, minimal with OMOG, and mild for attnPSM. The average EPE reduction with EMF-OPC (multiple levels, constant threshold model) was 1.20nm for BIM, 0.15nm for OMOG, and 0.50nm for attnPSM. BIM and OMOG analysis was done for 32-nm-node levels, whereas attnPSM analysis was done for 45-nm levels. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

7. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. Bright-field MEEF through pitch (wafer data) shows comparable performance for OMOG and attnPSM. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

8. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. 80-nm pitch imaging: (a) TMA simulations show comparable best case imaging with a slight advantage for attenuated PSM. EMF effects begin to minimize or reverse any advantage, seen both by (b) EMF simulation and (c) AIMS data. Using the TMA-P metric defined in Sec., attnPSM is seen to deviate more from TMA via (d) simulation and (e) AIMS data. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

9. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. Simulations show that for (a) bright-field imaging, attnPSM fares better through pitch but is subject to more TMA predictability issues. (b) For dark-field imaging, OMOG performs better through pitch, and again TMA predictability is stronger. Polarized dipole illumination is used. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

10. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. Two-dimensional effects: Phase shifting makes proximity effects more nonintuitive, as seen by (a) simulation of an isolated contact with on-axis coherent illumination. As the contact size decreases, significant sidelobe electric fields emerge until complete image reversal occurs. For (b) multiple contacts, comparable performance is possible for dense structures, but binary is generally more robust through pitch or random design. Images are shown for λ=193nm and NA=1.35. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712

11. Date of download: 6/2/2016 Copyright © 2016 SPIE. All rights reserved. Example 22-nm test mask where TMA OPC was done separately for each mask type. With a mix of clear- and dark-field patterns, the better blank is case dependent, but in general, binary fares better. Figure Legend: From: Lithographic qualification of new opaque MoSi binary mask blank for the 32-nm node and beyond J. Micro/Nanolith. MEMS MOEMS. 2010;9(1):013010-013010-10. doi:10.1117/1.3295712