Abstract
Diffractive optical elements such as Fresnel zoneplate lenses have many uses at extreme ultraviolet (EUV), particularly in short focal length, high-resolution applications. However, the diffraction efficiency of a pure absorption zoneplate is limited to about 10%, and it suffers additional loss through the membrane support material. To this end, the authors explored the possibility of silicon nitride (Si3N4) as a EUV phase shifting material. At an etched depth of 244 nm, they measured a diffraction efficiency of 18% in the first order and 18% in the zero order, which compares favorably to an amplitude grating of 10% and 25%, respectively. The measured efficiency as a function of etch depth matches the scalar theory quite well using a measured EUV index of refraction 0.9790 + 0.0066i at the wavelength of 13.5 nm. To further increase the efficiency, zoneplates were made freestanding, with the support membrane completely removed, and a 15% absolute efficiency was obtained. Vector electromagnetic calculations showed that at normal incidence, these optics produce excellent wavefront and efficiency for outer zones of 50 nm or larger. Zoneplates of narrower zones or those illuminated obliquely can suffer larger wavefront errors and low efficiency and would require careful design optimization. In the work, the authors also demonstrated a technique to package zoneplates and associated apertures for high precision insertion and removal from a EUV instrument. This technique has yielded alignment accuracy from a few microns to few 10s microns, depending on the exact design.
This work was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Materials Sciences and Engineering Division, under Contract No. DE-AC02-05CH11231.
I. INTRODUCTION
II. EXPERIMENTAL Si_{3}N_{4} RESULTS
III. SIMULATION OF Si_{3}N_{4} DIFFRACTIVE STRUCTURES
IV. MECHANICAL PACKAGING WITH BALLS AND GROOVES
V. SUMMARY AND CONCLUSIONS
Key Topics
- Extreme ultraviolet radiation
- 21.0
- Diffraction gratings
- 11.0
- Diffractive optics
- 11.0
- Dielectric nitrides
- 7.0
- Refractive index
- 7.0
Figures
(Color online) Reflectivity as a function of angle of a Si3N4 on Si sample covering several orders of magnitude from glancing incidence to near normal incidence. The best fit parameters show n = 0.979 + 0.0066i and thickness = 102.5 nm at a wavelength of 13.5 nm.
(Color online) Reflectivity as a function of angle of a Si3N4 on Si sample covering several orders of magnitude from glancing incidence to near normal incidence. The best fit parameters show n = 0.979 + 0.0066i and thickness = 102.5 nm at a wavelength of 13.5 nm.
(Color online) Complex amplitude Z of a wave propagated through a thickness, t, of Si3N4 compared to a wave propagated through free space. The scalar diffraction efficiency of the mth order is equal to (S/(mπ))^{2} sin(mπb), where S is the square of the length of vector.
(Color online) Complex amplitude Z of a wave propagated through a thickness, t, of Si3N4 compared to a wave propagated through free space. The scalar diffraction efficiency of the mth order is equal to (S/(mπ))^{2} sin(mπb), where S is the square of the length of vector.
(Color online) Measured zero order and first order diffraction efficiency of a 600 nm period grating with a line to period ratio, b, of 0.56. The solid lines represent the scalar theory efficiency and agree well with the measurements.
(Color online) Measured zero order and first order diffraction efficiency of a 600 nm period grating with a line to period ratio, b, of 0.56. The solid lines represent the scalar theory efficiency and agree well with the measurements.
(a) Top and (b) cross section view of a “freestanding” Si3N4 structure held together with buttresses etched all the way through the Si3N4 membrane. The period is 160 nm and the thickness is 320 nm. The measured absolute diffraction efficiency of this structure is 15%, which is less than an ideal structure due to the finite width of the buttresses.
(a) Top and (b) cross section view of a “freestanding” Si3N4 structure held together with buttresses etched all the way through the Si3N4 membrane. The period is 160 nm and the thickness is 320 nm. The measured absolute diffraction efficiency of this structure is 15%, which is less than an ideal structure due to the finite width of the buttresses.
(Color online) Freestanding zoneplate array in 100 nm thick Si3N4 used for EUV mask inspection.
(Color online) Freestanding zoneplate array in 100 nm thick Si3N4 used for EUV mask inspection.
(Color online) (a) Electromagnetic calculations of the efficiency and (b) phase error normalized to 2π, of the first diffracted order as a function of period at normal incidence. A Si3N4 thickness of 250 nm was assumed. λ/40 represents an acceptable relative phase error. The responses are flat until a period of about 100 nm at which point it rapidly changes. Zoneplates with outer zones of 50 nm or larger should perform as expected. Zoneplates with smaller outer zones may have additional aberrations due to the high-aspect ratio waveguide nature of the zones.
(Color online) (a) Electromagnetic calculations of the efficiency and (b) phase error normalized to 2π, of the first diffracted order as a function of period at normal incidence. A Si3N4 thickness of 250 nm was assumed. λ/40 represents an acceptable relative phase error. The responses are flat until a period of about 100 nm at which point it rapidly changes. Zoneplates with outer zones of 50 nm or larger should perform as expected. Zoneplates with smaller outer zones may have additional aberrations due to the high-aspect ratio waveguide nature of the zones.
(Color online) (a) Electromagnetic calculations of the efficiency and (b) phase errors normalized to 2π, of the different diffracted order as a function of period, with a 6° of incident angle. A Si3N4 thickness of 250 nm was assumed. λ/40 represents an acceptable relative phase error. The efficiency is flat until a period of about 160 nm at which point it rapidly changes. Zoneplates with outer zones of 80 nm or larger should perform as expected. Zoneplates with smaller outer zones may have additional aberrations due to the high-aspect ratio waveguide nature of the zones.
(Color online) (a) Electromagnetic calculations of the efficiency and (b) phase errors normalized to 2π, of the different diffracted order as a function of period, with a 6° of incident angle. A Si3N4 thickness of 250 nm was assumed. λ/40 represents an acceptable relative phase error. The efficiency is flat until a period of about 160 nm at which point it rapidly changes. Zoneplates with outer zones of 80 nm or larger should perform as expected. Zoneplates with smaller outer zones may have additional aberrations due to the high-aspect ratio waveguide nature of the zones.
(Color online) (a) Precise gap is determined by the ruby ball diameter and the lithographically defined grooves. w is the half width of the groove, D the diameter of the ball, and θ equals to 54.7° defined by the silicon crystal planes. (b) This concept is used to attach an aperture and mate to the stage.
(Color online) (a) Precise gap is determined by the ruby ball diameter and the lithographically defined grooves. w is the half width of the groove, D the diameter of the ball, and θ equals to 54.7° defined by the silicon crystal planes. (b) This concept is used to attach an aperture and mate to the stage.
(Color online) (a) Front side and (b) backside of a packaged EUV zoneplate and aperture pair. The package was glued to a metal plate with through holes for ease of handling.
(Color online) (a) Front side and (b) backside of a packaged EUV zoneplate and aperture pair. The package was glued to a metal plate with through holes for ease of handling.
Tables
Measured diffraction efficiency for 600 nm period gratings of various depths in 400 nm thick silicon nitride membrane at 13.5 nm wavelength.
Measured diffraction efficiency for 600 nm period gratings of various depths in 400 nm thick silicon nitride membrane at 13.5 nm wavelength.
Article metrics loading...
Full text loading...
Commenting has been disabled for this content