All-Glass, Large Metalens at Visible Wavelength Using Deep-Ultraviolet Projection Lithography

Abstract

Metalenses, planar lenses realized by placing subwavelength nanostructures that locally impart lenslike phase shifts to the incident light, are promising as a replacement for refractive optics for their ultrathin, lightweight, and tailorable characteristics, especially for applications where payload is of significant importance. However, the requirement of fabricating up to billions of subwavelength structures for centimeter-scale metalenses can constrain size-scalability and mass-production for large lenses. In this Letter, we demonstrate a centimeter-scale, all-glass metalens capable of focusing and imaging at visible wavelength, using deep-ultraviolet (DUV) projection stepper lithography. Here, we show size-scalability and potential for mass-production by fabricating 45 metalenses of 1 cm diameter on a 4 in. fused-silica wafer. The lenses show diffraction-limited focusing behavior for any homogeneously polarized incidence at visible wavelengths. The metalens’ performance is quantified by the Strehl ratio and the modulation transfer function (MTF), which are then compared with commercial refractive spherical and aspherical singlet lenses of similar size and focal length. We further explore the imaging capabilities of our metalens using a color-pixel sCMOS camera and scanning-imaging techniques, demonstrating potential applications for virtual reality (VR) devices or biological imaging techniques.

Metasurfaces for biomedical applications: imaging and sensing from a nanophotonics perspective

“…For example, in a study by Zhang et al [9,80,81], metasurface devices in the mid-infrared wavelength range were fabricated using photolithography, and in a study by Chen et al [32] and Hu et al [82], metasurface devices were designed in the terahertz and microwave frequencies. Recently, there has been a trend in using deep-UV photolithography with an excimer laser source to fabricate metasurfaces at the visible [14] and the near-infrared wavelength [10,83] in an attempt for scalability and mass production.…”

“…In fact, this is a hot research topic in the entire metasurface community. Recently, demonstrations of largearea metalenses with single wavelength operation in the visible and near-infrared wavelength ranges [10,14,83,189] and achromatic focusing of small area metalenses [190,191] have been reported. Hence, it is promising that this challenge can be solved soon with the advancement in both metasurface design methodologies and fabrication techniques [44].…”

“…Such metasurface-based flat devices represent a new class of optical components that are compact, flat, and lightweight. Numerous devices based on metasurfaces have been developed, including metasurface lenses (metalenses) [9][10][11][12][13][14][15][16][17][18][19], waveplates [20][21][22][23], polarimeters [24][25][26], and holograms [27][28][29]. For metalenses, in particular, numerical aperture (NA) as high as 0.97 was demonstrated [15].…”

Visible to long-wave infrared chip-scale spectrometers based on photodetectors with tailored responsivities and multispectral filters

“…Recent history has seen the rapid miniaturization of optical components through the development of nanophotonic optical elements with functions akin to conventional optics such as lenses [1,2], color filters [3,4], polarizing beam splitters [5], wave-plates [6], spatial light modulators [7], photodetectors [8][9][10], and light sources [11,12]. Now this trend has been extended to include more sophisticated, chip-scale optical systems such as interferometers [13,14], systems to exploit quantum mechanical properties [15][16][17], as well as multispectral imaging systems [18,19].…”

“…For FADA microspectrometers R n (λ) T n (λ)R 0 (λ), where T n (λ) is the transmission of the n th filter and R 0 (λ) is the intrinsic responsivity of the detector array. To estimate the incident spectrum, S(λ) using equation (1) and real experimental data there exist several linear regression methods. Researchers have used techniques such simulated annealing [44,48,49], recursive least squares [38,40,47], Tikhonov (L 2 ) regularization [39,45,48,50], Lasso (L 1 ) regularization [51], as well as non-negative least squares [47,52], since S(λ) is always non-negative.…”