Home A metasurface color router facilitating RGB-NIR sensing for an image sensor application
Article Open Access

A metasurface color router facilitating RGB-NIR sensing for an image sensor application

  • Yoon Jin Hong ORCID logo , Byeong Je Jeon ORCID logo , Yu Geun Ki ORCID logo and Soo Jin Kim ORCID logo EMAIL logo
Published/Copyright: January 22, 2024
Become an author with De Gruyter Brill

Abstract

CMOS image sensor (CIS) plays a crucial role in diverse optical applications by facilitating the capture of images in the visible and near-infrared spectra. The enhancement of image resolution in CIS by an increase in pixel density is becoming more significant and realizable with the recent progress of nanofabrication. However, as pixel size decreases towards the diffraction limit, there is an inevitable trade-off between the scale-down of pixel size and the enhancement of optical sensitivity. Recently, to overcome this, an entirely new concept of spectral sensing using a nanophotonic-based color router has been proposed. In this work, we present a metasurface-based spectral router to effectively split the spectrum from visible to near-infrared and redirect through the four optical channels to the targeted pixel surfaces. We optimize the metasurface that simultaneously controls the phases of the transmitted light of targeted spectra, i.e. red (R), green (G), blue (B), and near-infrared (NIR), which is the largest number of channels reported based on a single layered metasurface and has an optical efficiency that surpasses the efficiency of conventional color filter systems.

1 Introduction

The development of a CMOS image sensor (CIS) built on a silicon photodetector has been gaining great attention in various fields of IoT applications due to its ability to effectively capture images of visible and near-infrared light. To acquire higher-quality images, it’s important to increase the number of pixels in a given surface area, which enables the increased resolution of images. However, as the pixel size has been reduced to be extremely small, the light receiving area per unit pixel decreases significantly, and this results in decreased optical sensitivity. In particular, as the pixel size approaches the level of the diffraction limit, optical signals can be far below the electrical noise of an integrated CIS system, and such an optical condition ultimately renders a captured image blurry. As such, there remains an inevitable trade-off between the scale-down of pixel sizes and the enhancement of optical sensitivity in effectively detecting optical images.

Meanwhile, the recent progress of metasurfaces has drawn attention from researchers for decades due to the potential ability to control the flow of light for designer purposes. In particular, dielectric metasurfaces constructed from the subwavelength array of lossless nanomaterials have been used to develop practical optical applications, including optical elements and imaging devices. Research on developing lenses [1]–[9], color printing [10], [11], [12], and color filters [13], [14], [15], [16] using nanoscaled structures has been extensively conducted, and metasurface-based devices such as polarimeters [17], [18], [19], [20] and photodetectors [21], [22], [23], [24] and ultrathin imaging elements [25], [26] have been demonstrated. Furthermore, to overcome the aforementioned limit on the loss of light in the CIS system, the concept of a nanophotonic color router has been proposed and intensively studied recently [27]–[35]. Unlike conventional color filters which only transmit targeted spectral bands and reflect or absorb out-of-band power, nanophotonic color routers are designed to re-direct light with a specific targeted spectral band to the designated pixels without seriously deteriorating optical sensitivity. Moreover, nanophotonic routers made from dielectric materials have advantages over conventional filters in that they are less susceptible to environmental conditions such as humidity, temperature, and degradation from light exposure. There have been demonstrations of color routers using metasurfaces that effectively sort the targeted spectrum. Examples of these include the two-channel router, i.e., visible and near-infrared (NIR) [33], [34], or the three-channel router that splits the red (R), green (G), and blue (B) colors to the targeted pixel areas [27]–[31]. A router that affords the separation of four channels, including three visible colors and near-infrared, attracts researchers recently due to its advantages of detecting images at hazed environment and potential applications to IoT sensors for image fusion, face recognition, bio imaging, and many other purposes [36], [37]. However, the development of such router is considered to be more challenging and still elusive due to the increased complexity of phase design and the creation of the library for various meta-structures. Nevertheless, a recent demonstration of an inversely designed nanophotonic router demonstrates efficient splitting of visible and near-infrared light [32], [35].

In this work, we present a metasurface router with four distinct channels constructed from a single nanostructured layer that facilitates the selective separation of the visible and NIR spectra. In contrast to the previously reported three dimensionally structured nanophotonic routers which pose comparatively greater fabrication challenges [32], [35], the proposed design can be fabricated using conventional techniques and effectively separate the spectra into the targeted four bands without significant loss of light occurring in conventional filters. The proposed metasurface-based spectral router shows 32 % averaged optical efficiency over operating RGB and NIR wavelengths, which is the maximum reported number of channels of single layered metasurfaces and exceeds the efficiency of the conventional color filter system.

2 Results and discussion

A comparison between the conventional filter system and the single-layered spectral router is illustrated in Figure 1(a) and (b). In the case of the conventional system (Figure 1(a)), each color filter only transmits the targeted spectrum and reflects or absorbs the remaining parts of spectra to perform selective detection of RGB light. For instance, the “R” cell, which is designed to capture red light, only utilizes a spectrum near the 630 nm wavelength and discards all other wavelengths of light. Based on the working principle, the conventional system with the inclusion of a near infrared channel is restricted to utilizing only 25 % on average of the incoming light.

Figure 1: 
Conceptual schematics of designed spectral router. (a), (b) Conceptual cross-sectional images of conventional color filter system and spectral router system, respectively. (c) Schematic of designed spectral router with different types of meta-atom. (d)–(g) 3D images of the light receiving mechanism for the spectral router. The black boundary line at the top indicates the area where a targeted pixel can collect and absorb the light (A
absorption). The yellow boundary line indicates the physical surface area of a photodetector (A
PD), which implies that A
absorption is larger than the A
PD.
Figure 1:

Conceptual schematics of designed spectral router. (a), (b) Conceptual cross-sectional images of conventional color filter system and spectral router system, respectively. (c) Schematic of designed spectral router with different types of meta-atom. (d)–(g) 3D images of the light receiving mechanism for the spectral router. The black boundary line at the top indicates the area where a targeted pixel can collect and absorb the light (A absorption). The yellow boundary line indicates the physical surface area of a photodetector (A PD), which implies that A absorption is larger than the A PD.

To overcome such limit and utilize NIR signal effectively, we have designed a single-layered spectral router capable of sorting R, G, B, and NIR light to the four distinct channels (Figure 1(c)). Our design can simultaneously perform selective sorting and focusing without the need for an NIR cut-off filter or microlens elements. For a more comprehensive description, Figure 1(d)–(g) illustrates the light-receiving regions of CIS including the NIR channel. In the case of the red cell of the router design depicted in Figure 1(d), light from adjacent areas (solid black line, A absorption) is focused onto the targeted photodetector, and a significant portion of received light can be utilized compared to the typical CIS system that only detects the light incident on each designated color filter area. With the working principles, we have developed a single-layer metasurface router for RGB and NIR channels, which could replace the functionality of constituent microlens, color filter, and NIR cut-off filter of conventional system simultaneously.

To optimally perform four channel (R, G, B, and NIR) sorting within a single-layer metasurface, it is important to design the appropriate meta-atoms. The optimization of phase delay, valid and individually controllable for the four channels, is performed simultaneously by adjusting each single meta-atom. Among the previously pioneered available methods [38]–[45], we adopt the property of propagation phase delay of resonant nanostructures which typically support polarization-independent responses. The phase modulation generated by the nanostructure is obtained by Equation (1).

(1) Δ φ = 2 π λ Δ n eff h

Based on the equation, the phase delay (Δφ) depends on the height (h) and effective refractive index (Δn eff) of the nanostructure. Meta-atoms fabricated using high refractive index materials, such as silicon or germanium, exhibit more effective control of phase delay compared to low-index materials. However, such an advantage comes at the cost of increased optical loss within the visible and near-infrared spectrums. Taking these into consideration, we designed the nanostructure using silicon nitride (Si3N4), and two different prototype geometries are analyzed as illustrated in Figure 2(a) and (b), which show lossless properties and a relatively larger phase library for design flexibility.

Figure 2: 
Optimization and analysis of a constituent meta-atom. (a) and (b) Side and top view of a meta-atom. Period (P) is fixed to 400 nm, and Si3N4 based nanostructure is designed on a SiO2 substrate. Light is incident to the back side of the substrate and transmitted into free space (n = 1). (c) and (d) Normalized electric fields (|E

x
|norm) of a square and cross-shaped nanostructure illustrated in (a) and (b) under the two representative wavelengths of blue (450 nm) and NIR (800 nm), respectively. Width (w

sq
, w

x
) and height (H) are fixed to be 230 nm, 112 nm (with l/w

x
 = 1.5), and 1250 nm, respectively. The cross-sectional electric fields at the insets are monitored at the height of 630 nm, which are indicated as black dashed lines. (e)–(f) Simulated phase delay of transmitted light by sweeping height and width of the nanostructure under x-polarized incident light. Noted that for convenience, l/w

x
 is fixed to 1.5 in (f). White dashed lines indicate the position of H = 1250 nm, at which the close to full 2π phase modulation is attained.
Figure 2:

Optimization and analysis of a constituent meta-atom. (a) and (b) Side and top view of a meta-atom. Period (P) is fixed to 400 nm, and Si3N4 based nanostructure is designed on a SiO2 substrate. Light is incident to the back side of the substrate and transmitted into free space (n = 1). (c) and (d) Normalized electric fields (|E x |norm) of a square and cross-shaped nanostructure illustrated in (a) and (b) under the two representative wavelengths of blue (450 nm) and NIR (800 nm), respectively. Width (w sq , w x ) and height (H) are fixed to be 230 nm, 112 nm (with l/w x = 1.5), and 1250 nm, respectively. The cross-sectional electric fields at the insets are monitored at the height of 630 nm, which are indicated as black dashed lines. (e)–(f) Simulated phase delay of transmitted light by sweeping height and width of the nanostructure under x-polarized incident light. Noted that for convenience, l/w x is fixed to 1.5 in (f). White dashed lines indicate the position of H = 1250 nm, at which the close to full 2π phase modulation is attained.

Figure 2(c) and (d) show the electric field images of a meta-atoms with square and cross-sectional shapes, which resonate with different modes at the two representative wavelengths of blue (450 nm) and NIR (800 nm) channels. Incident light is polarized in the x-direction and illuminated on the backside of the SiO2 substrate.

It is observed in the analysis that the 5th-order Fabry–Pérot (FP) resonance mode is supported at a wavelength of 800 nm (Figure 2(c), right), and the higher-order FP mode with seven anti-nodes at the vertical directions is supported at a shorter wavelength of 450 nm (Figure 2(c), left). The same analysis can be applied to a cross-shaped meta-atom in Figure 2(d), which shows 7th and 4th-order modes at blue and NIR channels. The additional field diagrams at the wavelengths of 540 nm (G) and 630 nm (R) (Supplementary note S1) show different order of modes, which implies that a meta-unit with a height of 1250 nm simultaneously covers low-order and high-order FP modes at the targeted spectral channel.

Figure 2(e) and (f) show the analysis of phase delay as a function of height and width (w sq , w x ) at a fixed period of 400 nm. At each targeted wavelength of RGB and NIR, the phases of square-shaped (Figure 2(e)) and cross-shaped (Figure 2(f)) meta-atoms are displayed, respectively. As the height rises, the transmission phase is delayed by more than 2π, which confirms that the designed nanostructures effectively support vertically oriented FP resonances. At the blue channel, a more abrupt change in phase delay occurs due to the higher-order modes of FP resonance compared to the example of the NIR channel where a gradual change in phase modulation occurs by the relatively lower-order of FP resonance. At the height of 1250 nm which is indicated by white dashed lines in the panels, close to full 2π phase modulations are achieved for the four designated channels by altering the width of nanostructures. It is noteworthy that the electric field is more densely concentrated in the squared geometric nanopost (Figure 2(c)) than the field concentration in the cross geometric nanopost (Figure 2(d)), which implies that the square-shaped meta-atom has higher mode index than the cross-shaped one. Such a higher mode index facilitates relatively larger phase variation of squared nanopost (Figure 2(e)) than the phase variation of cross-shaped nanopost (Figure 2(f)) as the width w sq and w x equally varies. Thus, the square-shaped meta-atoms primarily address most of phase delay, while the cross-shaped meta-atoms (Figure 2(f)) serve to compensate for the phase delay that the square-shaped meta-atoms cannot sufficiently provide.

The phase library constructed by the nanostructures with a fixed height of 1250 nm is illustrated in Figure 3(a) and (b). They show the amplitudes and phases of transmitted light from square- and cross-shaped meta-atoms, respectively, with respect to the changes in the cross-sectional size of each nanostructure. As confirmed in Figure 3(a), close to full 2π phase coverage can be obtained in all four channels by adjusting the width of square-shaped meta-atoms (w sq ). Despite the effective modulation, at the NIR regime, the square-shaped nanostructure by itself does not provide complete 2π phase modulation due to the effect of lower-order FP mode, which results in relatively smaller phase variations over the control of the geometric structure. To achieve complete control of the 2π phase at the fixed height, the cross-shaped nanostructure is additionally designed and analyzed, which has more degrees of freedom in geometry than the square-shaped nanostructure. By effectively utilizing and allocating the two geometric types of meta-atom on the pixel surface, we obtained high transmission and complete 2π phase coverage for all four spectral channels. Figure 3(b) shows the simulated result with the ratio between l and w x being fixed to 1.5. More detailed phase library is shown in Supplementary note 2.

Figure 3: 
Analysis of phase profiles for the design of metasurface router. (a), (b) Amplitude and phase of transmitted light at square- and cross-shaped meta-atoms. The inset image shows the schematics of meta-atoms. For square-shape, width (w

sq
) is swept from 180 nm to 280 nm (a), and for cross-shape, width (w

x
) is swept from 92 nm to 132 nm with the ratio between l and w

x
 fixed to be 1.5. (c) Designed phase profile for the simultaneous routing of four channels at each center wavelengths of 450 nm, 540 nm, 630 nm, and 800 nm, respectively.
Figure 3:

Analysis of phase profiles for the design of metasurface router. (a), (b) Amplitude and phase of transmitted light at square- and cross-shaped meta-atoms. The inset image shows the schematics of meta-atoms. For square-shape, width (w sq ) is swept from 180 nm to 280 nm (a), and for cross-shape, width (w x ) is swept from 92 nm to 132 nm with the ratio between l and w x fixed to be 1.5. (c) Designed phase profile for the simultaneous routing of four channels at each center wavelengths of 450 nm, 540 nm, 630 nm, and 800 nm, respectively.

Since the designed router should operate as a multi-functional lens that simultaneously focuses on four different locations of pixels at each targeted wavelength, it is important to optimize the design which concurrently satisfies the phase distributions at the four spectral channels of R, G, B, and NIR. To satisfy this, we analyze the required phase profile for the positions x and y at each center wavelength using the following equations.

(2) ϕ B x , y = 2 π λ B x x B 2 + y y B 2 + f 2 x B 2 + y B 2 + f 2 + C B

(3) ϕ G x , y = 2 π λ G x x G 2 + y y G 2 + f 2 x G 2 + y G 2 + f 2 + C G

(4) ϕ R x , y = 2 π λ R x x R 2 + y y R 2 + f 2 x R 2 + y R 2 + f 2 + C R

(5) ϕ NIR x , y = 2 π λ NIR x x NIR 2 + y y NIR 2 + f 2 x NIR 2 + y NIR 2 + f 2 + C NIR

In each equation, the variables x R , G , B , NIR and y R , G , B , NIR represent the positions of the focal point for the red, green, blue, and near-infrared channels, respectively, and f denotes the focal length of the lens. In order to facilitate the effective routing of light to the four channels, each focal point is positioned at the center of four pixel regions on the focal plane. To achieve concurrent modulation of phase throughout all spectral channels and compensate for the errors, the fitting constant C is employed. The optimal value of the constant C is achieved by iteratively updating it through cascaded feedback between the temporarily produced phase profile and the accessible phase library of meta-atoms. As illustrated in Figure 3(c), such an iterative process creates a phase distribution profile that closely matches the values of experimentally implementable meta-atoms while minimizing the inaccuracy in the overall control of focusing light at each desired spectrum.

Based on the steps, the values of C are determined as being C R = 0, C G = 0.3307, C B = 2.9762, and C NIR = 5.6218, and total 100 points are sampled from the continuous phase profile set to construct the desired metasurface.

The schematic of the constructed metasurface is illustrated in Figure 4(a). Si3N4-based meta-atoms are arranged with a combination of nanostructures with different cross-sectional shapes and sizes on a 4 μm × 4 μm metasurface, which is assumed to be periodic. Detailed design information of the metasurface is summarized in Supplementary note 3. To demonstrate the ability of NIR detection, we adopt the conventional RGB Bayer pixel arrangement and replace one green pixel with an NIR pixel. The total transmittance of the metasurface is 64.7 % for blue (λ = 450 nm), 76.2 % for green (λ = 540 nm), 88.9 % for red (λ = 630 nm), and 91.9 % for near-infrared (λ = 800 nm), which implies that the designed metasurface is highly transmissive for broadband spectral range. As depicted in Figure 4(b), the electric field intensities at the targeted four channels, i.e. λ = 450 nm, 540 nm, 630 nm, and 800 nm, are effectively detected on the focal plane. Although there is a non-negligible spectral crosstalk between the nearby spectrum, the overall power of incident light is predominantly focused near the center of the targeted pixel area, which confirms the effective routing and focusing abilities of the designed metasurface. Such routing phenomena are observed both under the x- and y-directed polarizations of incident light in a similar fashion (Supplementary note 4). This is due to the geometric properties of the meta-atoms that satisfy C4 symmetry and negligible optical coupling with neighboring meta-atoms that mitigate the effect of broken rotational symmetry at the level of the super-cell scale. It is also noteworthy that light is illuminated from SiO2 substrate to the metasurface. Such configuration of backside illumination alleviates the efficiency degradation under oblique incident light than the configuration of the direct coupling of light from air. Further optimization can be achieved by placing a metasurface sandwiched between SiO2 layers for monolithic integration to CIS.

Figure 4: 
A designed metasurface and its optical performance. (a) Schematic of the overall metasurface-based group pixel which accommodates RGB and NIR. Yellow bold text indicates the location of each pixel (left). The scale bar is 1 μm. Vertical schematic of designed spectral router. The focal length is 5.15 μm (right). (b) Normalized electric field intensities at each center wavelength analyzed under x-polarized incident light (








E


x






2




${\left\vert {E}_{x}\right\vert }^{2}$




norm). Focal spots for each wavelength can be observed at each targeted region, which shows the abilities of effective sorting of RGB and NIR. (c) The optical efficiency of the metasurface over a broadband spectral range from visible to NIR. The black dashed line indicates the maximum efficiency that can be achieved in a conventional four channel filter system.
Figure 4:

A designed metasurface and its optical performance. (a) Schematic of the overall metasurface-based group pixel which accommodates RGB and NIR. Yellow bold text indicates the location of each pixel (left). The scale bar is 1 μm. Vertical schematic of designed spectral router. The focal length is 5.15 μm (right). (b) Normalized electric field intensities at each center wavelength analyzed under x-polarized incident light ( E x 2 norm). Focal spots for each wavelength can be observed at each targeted region, which shows the abilities of effective sorting of RGB and NIR. (c) The optical efficiency of the metasurface over a broadband spectral range from visible to NIR. The black dashed line indicates the maximum efficiency that can be achieved in a conventional four channel filter system.

To evaluate the optical efficiency, we calculate the integrated power density on a targeted 2 μm × 2 μm sub-pixel divided by the total optical power incident on a unit of the metasurface, i.e., 4 μm × 4 μm super-pixel. As exhibited in Figure 4(c), the optical efficiencies over a broad wavelength range of blue (400–500 nm), green (500–600 nm), red (600–700 nm) and NIR (700–1100 nm) are calculated, and the spectral peaks at each pixel exhibit the highest value at the designated spectral range. Specifically, the averaged optical efficiencies for each pixel under the incident light of two orthogonally linear polarizations are 26.5 %, 28.1 %, 37.1 %, and 39.8 % at the aforementioned four wavelengths of blue, green, red, and near-infrared, respectively. It is noteworthy that the averaged efficiencies of the metasurfaces surpass the maximum efficiency of 25 % attainable using conventional color filters of four spectral channels which is outlined as a horizontal dashed line in Figure 4(c). Further reduction in pixel size can be achieved through the compensation of spectral crosstalks and focal lengths (Supplementary note 7). Advanced nanofabrication techniques such as deep reactive ion etching (DRIE) and metal aided chemical etching (MACE) can be used to create the proposed nanostructures with high aspect ratio [46], [47]. We further investigate the effect of fabrication imperfections such as slanted profiles on the performance of the metasurface routers (Supplementary note 5).

The optical noise caused by the spectral crosstalk can be further minimized by additionally applying the conventional spectral filter above the surface of the photodetector array, which blocks the side band transmission and transmits the routed target spectrum. To realize such a hybrid-type metasurface router, conventional spectral filter layer is integrated with metasurface router as illustrated in Figure 5(a). The spectral shapes for filters are modelled based on the sigmoid function with the variable k in the function σ(kx) being ±0.2 to define the spectral slope of transmission band edge. The profile is configured to exhibit an ideal transmittance of 100 % within the target wavelength range, with a center wavelength denoted by λ i .

Figure 5: 
Analysis of hybrid-type metasurface router. (a) Schematical illustration of hybrid-type spectral router. Conventional spectral filter is applied near the focal plane of detecting surfaces. Inset image shows the profile of idealized spectral filter with a peak transmittance 100 % for a single filter. λ

i
 is 450 nm, 540 nm, 630 nm, and 800 nm for each targeted filter. (b) Optical efficiency of conventional filter system (dashed line) and hybrid spectral router (solid line).
Figure 5:

Analysis of hybrid-type metasurface router. (a) Schematical illustration of hybrid-type spectral router. Conventional spectral filter is applied near the focal plane of detecting surfaces. Inset image shows the profile of idealized spectral filter with a peak transmittance 100 % for a single filter. λ i is 450 nm, 540 nm, 630 nm, and 800 nm for each targeted filter. (b) Optical efficiency of conventional filter system (dashed line) and hybrid spectral router (solid line).

The comparison between conventional filter and hybrid-type metasurface router is performed numerically, as depicted in Figure 5(b). Solid line denotes the spectral efficiency of hybrid metasurface router, which exceeds the efficiency of conventional filter array in an overall spectral region as denoted by dashed line. In particular, the efficiency of red and NIR spectra increased approximately by 50 % and 60 %, respectively, compared to the conventional system.

Table 1 shows the summary and comparison of the optical performance with previously pioneered design of nanophotonics or metasurface-based routers. Compared to the previous metasurface, which routes two (visible and NIR) or three (RGB) channels, the main progress of our proposed metasurface is the ability to route four channels, including the near-infrared spectrum. And it achieves an average optical efficiency of over 30 % for the broadband wavelength regime, which still exceeds the efficiency limit of conventional color filter systems.

Table 1:

Comparison with previously pioneered routers.

Ref Material Configuration Operating wavelength (nm) Efficiency (%)
[48] TiO2 Single cubic 400 nm, 500 nm Not reported
[49] Ag Single nanorod aNot specifically

mentioned
Not reported
[27] SiN Single-layered metasurface 400–700 nm 25 % (B), 50 % (G), 65 % (R)
[29] GaN Single-layered metasurface 430 nm, 532 nm, 633 nm b15.9 % (B), 37.9 % (G), 27.6 % (G), 38.3 % (R)
[31] Si3N4 Single-layered metasurface 400–700 nm 49 % (B), 59 % (G), 58 % (R)
[32] Si3N4, TiO2 3D metamaterial (inverse design) 400–800 nm c>99 % (B, G, R, NIR)
[33] SiN Single-layered metasurface 400–1100 nm d59 % (Vis), 60 % (NIR)
[34] Si3N4 Single-layered metasurface 400–1000 nm e71.2 % (Vis), 80.36 % (NIR)
[35] Si3N4 3D metamaterial (inverse design) 400–800 nm f29 % (B), 38 % (G), 39 % (R), 36 % (NIR)
This work Si3N4 Single-layered metasurface 400–1100 nm 26.5 % (B), 28.6 % (G), 37.1 % (R), 39.1 % (NIR)
  1. All described efficiencies (except [33]) are calculated (or measured) on a center wavelength. For [27], [29], [31], two pixels for green channel are used. aOperating channels are defined as red-blue-red or green-red-green for a single Ag nanorod. bEfficiencies are measured by the single wavelength laser illuminated on the spectral router. The second and third efficiencies are for the green pixels. cOver 99 % of optical efficiency has been achieved in this report. However, the proposed design is based on the inverse design multi-layered metasurface. dTwo channels (visible and NIR) were adopted. Described efficiencies are averaged over the operating wavelength range. eMeasured efficiencies at the center wavelengths are described above. fSimilar with [32], the inverse design multi-layered metasurface is provided in the research.

3 Conclusions

To summarize, we present a single layered metasurface capable of sorting a broad range of light to four spectral channels that the silicon photodetector can respond to. The basic optical properties of a meta-atom are additionally analyzed, which support FP resonance and afford full coverage of phase delays. We believe that the proposed design offers the potential solutions to realize selective sorting of light over a broad spectral range for the applications of miniaturized pixel-based image sensors. As a prospective research direction, further scaling down of the metasurface router for reduced pixels should be investigated, as should the development of advanced fabrication techniques and the exploration of various candidates of emerging materials in order to realize advanced nanophotonic color routers for miniaturization, simplification, and enhanced quality of imaging for future optical devices.

4 Methods

4.1 Optical simulation

To optimize and evaluate the optical performance of the device, commercial finite-difference time-domain (FDTD) software from Lumerical Solutions is used. Arrays of Si3N4 nanostructure with a 1250 nm height (refractive index obtained from the data measured by Phillip in 1973) are arranged on the SiO2 substrate (refractive index from the Handbook of Optical Constants of Solids by Edward D. Palik). The period of the unit cell is designed to be 400 nm, and the metasurface array is designed as 4 μm by 4 μm for x- and y-directions. Periodic boundary conditions are used in the x- and y-directions, and perfectly matched layer (PML) boundary conditions are applied along the z-direction for the simulation of both a unit cell and a metasurface. Transmittance and optical efficiencies are calculated by monitoring transmitted power at the focal plane.


Corresponding author: Soo Jin Kim, School of Electrical Engineering, Korea University, Seoul, Korea, E-mail:

Award Identifier / Grant number: RS-2023-00256247

Funding source: National Research Foundation (NRF)

Award Identifier / Grant number: 2022R1A2C4001246

Award Identifier / Grant number: 2022R1A4A1034315

Award Identifier / Grant number: RS-2023-00254920

  1. Research funding: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2022R1A2C4001246, No. NRF-2022R1A4A1034315, No. RS-2023-00254920) and in part by the Technology Innovation Program (Korea-led of K-Sensor technology for market leader-ship) (RS-2023-00256247) funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea).

  2. Author contributions: Y.J.H. and S.J.K. conceived the idea for the design and simulation. Y.J.H. and B.J.J. performed the design and simulation of the metasurface. Y.J.H., B.J.J. and Y.G.K. conducted the theoretical analysis of the results. S.J.K. supervised the project. All authors contributed to the interpretation of results and participated in the preparation of the manuscript.

  3. Conflict of interest: Authors state no conflicts of interest.

  4. Data availability: The datasets generated and analysed during the current study are available from the corresponding author upon reasonable request.

References

[1] A. Arbabi, Y. Horie, M. Bagheri, and A. Faraon, “Dielectric metasurfaces for complete control of phase and polarization with subwavelength spatial resolution and high transmission,” Nat. Nanotechnol., vol. 10, no. 11, pp. 937–943, 2015. https://doi.org/10.1038/nnano.2015.186.Search in Google Scholar PubMed

[2] Y. Wang, et al.., “High-efficiency broadband achromatic metalens for near-IR biological imaging window,” Nat. Commun., vol. 12, no. 1, p. 5560, 2021. https://doi.org/10.1038/s41467-021-25797-9.Search in Google Scholar PubMed PubMed Central

[3] M. Khorasaninejad, et al.., “Polarization-insensitive metalenses at visible wavelengths,” Nano Lett., vol. 16, no. 11, pp. 7229–7234, 2016. https://doi.org/10.1021/acs.nanolett.6b03626.Search in Google Scholar PubMed

[4] M. Khorasaninejad, W. T. Chen, R. C. Devlin, J. Oh, A. Y. Zhu, and F. Capasso, “Metalenses at visible wavelengths: diffraction-limited focusing and subwavelength resolution imaging,” Science, vol. 352, no. 6290, pp. 1190–1194, 2016. https://doi.org/10.1126/science.aaf6644.Search in Google Scholar PubMed

[5] W. T. Chen, et al.., “A broadband achromatic metalens for focusing and imaging in the visible,” Nat. Nanotechnol., vol. 13, no. 3, pp. 220–226, 2018. https://doi.org/10.1038/s41565-017-0034-6.Search in Google Scholar PubMed

[6] Z. Li, et al.., “Meta-optics achieves RGB-achromatic focusing for virtual reality,” Sci. Adv., vol. 7, no. 5, p. eabe4458, 2021. https://doi.org/10.1126/sciadv.abe4458.Search in Google Scholar PubMed PubMed Central

[7] W. T. Chen, A. Y. Zhu, and F. Capasso, “Flat optics with dispersion-engineered metasurfaces,” Nat. Rev. Mater., vol. 5, no. 8, pp. 604–620, 2020. https://doi.org/10.1038/s41578-020-0203-3.Search in Google Scholar

[8] F. Aieta, M. A. Kats, P. Genevet, and F. Capasso, “Multiwavelength achromatic metasurfaces by dispersive phase compensation,” Science, vol. 347, no. 6228, pp. 1342–1345, 2015. https://doi.org/10.1126/science.aaa2494.Search in Google Scholar PubMed

[9] S. Shrestha, A. C. Overvig, M. Lu, A. Stein, and N. Yu, “Broadband achromatic dielectric metalenses,” Light: Sci. Appl., vol. 7, no. 1, p. 85, 2018. https://doi.org/10.1038/s41377-018-0078-x.Search in Google Scholar PubMed PubMed Central

[10] J. Proust, F. Bedu, B. Gallas, I. Ozerov, and N. Bonod, “All-dielectric colored metasurfaces with silicon Mie resonators,” ACS Nano, vol. 10, no. 8, pp. 7761–7767, 2016. https://doi.org/10.1021/acsnano.6b03207.Search in Google Scholar PubMed

[11] V. Flauraud, M. Reyes, R. Paniagua-Domínguez, A. I. Kuznetsov, and J. Brugger, “Silicon nanostructures for bright field full color prints,” ACS Photonics, vol. 4, no. 8, pp. 1913–1919, 2017. https://doi.org/10.1021/acsphotonics.6b01021.Search in Google Scholar

[12] N. Pinton, J. Grant, S. Collins, and D. R. Cumming, “Exploitation of magnetic dipole resonances in metal–insulator–metal plasmonic nanostructures to selectively filter visible light,” ACS Photonics, vol. 5, no. 4, pp. 1250–1261, 2018. https://doi.org/10.1021/acsphotonics.7b00959.Search in Google Scholar

[13] A. De Proft, et al.., “Highly selective color filters based on hybrid plasmonic–dielectric nanostructures,” ACS Photonics, vol. 9, no. 4, pp. 1349–1357, 2022. https://doi.org/10.1021/acsphotonics.1c01983.Search in Google Scholar

[14] S. Yokogawa, S. P. Burgos, and H. A. Atwater, “Plasmonic color filters for CMOS image sensor applications,” Nano Lett., vol. 12, no. 8, pp. 4349–4354, 2012. https://doi.org/10.1021/nl302110z.Search in Google Scholar PubMed

[15] J. Berzinš, S. Fasold, T. Pertsch, S. M. Bäumer, and F. Setzpfandt, “Submicrometer nanostructure-based RGB filters for CMOS image sensors,” ACS Photonics, vol. 6, no. 4, pp. 1018–1025, 2019. https://doi.org/10.1021/acsphotonics.9b00021.Search in Google Scholar

[16] Z. Li, S. Butun, and K. Aydin, “Large-area, lithography-free super absorbers and color filters at visible frequencies using ultrathin metallic films,” ACS Photonics, vol. 2, no. 2, pp. 183–188, 2015. https://doi.org/10.1021/ph500410u.Search in Google Scholar

[17] N. A. Rubin, G. D’Aversa, P. Chevalier, Z. Shi, W. T. Chen, and F. Capasso, “Matrix Fourier optics enables a compact full-Stokes polarization camera,” Science, vol. 365, no. 6448, p. eaax1839, 2019. https://doi.org/10.1126/science.aax1839.Search in Google Scholar PubMed

[18] L. Li, et al.., “Monolithic full-Stokes near-infrared polarimetry with chiral plasmonic metasurface integrated graphene–silicon photodetector,” ACS Nano, vol. 14, no. 12, pp. 16634–16642, 2020. https://doi.org/10.1021/acsnano.0c00724.Search in Google Scholar PubMed

[19] J. Park, J.-H. Kang, S. J. Kim, X. Liu, and M. L. Brongersma, “Dynamic reflection phase and polarization control in metasurfaces,” Nano Lett., vol. 17, no. 1, pp. 407–413, 2017. https://doi.org/10.1021/acs.nanolett.6b04378.Search in Google Scholar PubMed

[20] Z. Yang, et al.., “Generalized Hartmann-Shack array of dielectric metalens sub-arrays for polarimetric beam profiling,” Nat. Commun., vol. 9, no. 1, p. 4607, 2018. https://doi.org/10.1038/s41467-018-07056-6.Search in Google Scholar PubMed PubMed Central

[21] S. J. Kim, et al.., “Anti-Hermitian photodetector facilitating efficient subwavelength photon sorting,” Nat. Commun., vol. 9, no. 1, p. 316, 2018. https://doi.org/10.1038/s41467-017-02496-y.Search in Google Scholar PubMed PubMed Central

[22] L. Khosravi Khorashad and C. Argyropoulos, “Unraveling the temperature dynamics and hot electron generation in tunable gap-plasmon metasurface absorbers,” Nanophotonics, vol. 11, no. 17, pp. 4037–4052, 2022. https://doi.org/10.1515/nanoph-2022-0048.Search in Google Scholar

[23] S. J. Kim, C. Lee, S. Jeon, J. Park, and S. J. Kim, “Subwavelength sorting of full-color based on anti-Hermitian metasurfaces,” Nanophotonics, vol. 10, no. 2, pp. 967–974, 2020. https://doi.org/10.1515/nanoph-2020-0526.Search in Google Scholar

[24] J. S. Smalley, et al.., “Subwavelength pixelated CMOS color sensors based on anti-Hermitian metasurface,” Nat. Commun., vol. 11, no. 1, p. 3916, 2020. https://doi.org/10.1038/s41467-020-17743-y.Search in Google Scholar PubMed PubMed Central

[25] B. Born, S.-H. Lee, J.-H. Song, J. Y. Lee, W. Ko, and M. L. Brongersma, “Off-axis metasurfaces for folded flat optics,” Nat. Commun., vol. 14, no. 1, p. 5602, 2023. https://doi.org/10.1038/s41467-023-41123-x.Search in Google Scholar PubMed PubMed Central

[26] A. Chen and F. Monticone, “Dielectric nonlocal metasurfaces for fully solid-state ultrathin optical systems,” ACS Photonics, vol. 8, no. 5, pp. 1439–1447, 2021. https://doi.org/10.1021/acsphotonics.1c00189.Search in Google Scholar

[27] M. Miyata, N. Nemoto, K. Shikama, F. Kobayashi, and T. Hashimoto, “Full-color-sorting metalenses for high-sensitivity image sensors,” Optica, vol. 8, no. 12, pp. 1596–1604, 2021. https://doi.org/10.1364/OPTICA.444255.Search in Google Scholar

[28] S. Yun, et al.., “Highly efficient color separation and focusing in the sub-micron CMOS image sensor,” in 2021 IEEE International Electron Devices Meeting (IEDM), IEEE, 2021, pp. 30.31. 31–30.31. 34.10.1109/IEDM19574.2021.9720592Search in Google Scholar

[29] B. H. Chen, et al.., “GaN metalens for pixel-level full-color routing at visible light,” Nano Lett., vol. 17, no. 10, pp. 6345–6352, 2017. https://doi.org/10.1021/acs.nanolett.7b03135.Search in Google Scholar PubMed

[30] P. B. Catrysse, N. Zhao, W. Jin, and S. Fan, “Subwavelength Bayer RGB color routers with perfect optical efficiency,” Nanophotonics, vol. 11, no. 10, pp. 2381–2387, 2022. https://doi.org/10.1515/nanoph-2022-0069.Search in Google Scholar

[31] X. Zou, et al.., “Pixel-level Bayer-type colour router based on metasurfaces,” Nat. Commun., vol. 13, no. 1, p. 3288, 2022. https://doi.org/10.1038/s41467-022-31019-7.Search in Google Scholar PubMed PubMed Central

[32] N. Zhao, P. B. Catrysse, and S. Fan, “Perfect RGB‐IR color routers for sub‐wavelength size CMOS image sensor pixels,” Adv. Photonics Res., vol. 2, no. 3, p. 2000048, 2021. https://doi.org/10.1002/adpr.202000048.Search in Google Scholar

[33] R. Zhong, X. Xu, Y. Zhou, H. Liang, and J. Li, “High-efficiency integrated color routers by simple identical nanostructures for visible and near-infrared wavelengths,” Photonics, vol. 10, no. 5, p. 536, 2023. https://doi.org/10.3390/photonics10050536.Search in Google Scholar

[34] Y. Shao, et al.., “Pixelated NIR–VIS spectral routers based on 2D mie‐type metagratings,” Laser Photonics Rev., vol. 17, no. 8, p. 2300027, 2023. https://doi.org/10.1002/lpor.202300027.Search in Google Scholar

[35] Y. Q. Peng, H. P. Lu, D. S. Zhang, L. C. Wang, Z. Y. Ma, and J. J. Xiao, “Inverse design of a light nanorouter for a spatially multiplexed optical filter,” Opt. Lett., vol. 48, no. 23, pp. 6232–6235, 2023. https://doi.org/10.1364/OL.505205.Search in Google Scholar PubMed

[36] C. H. Park and M. G. Kang, “Color restoration of RGBN multispectral filter array sensor images based on spectral decomposition,” Sensors, vol. 16, no. 5, p. 719, 2016. https://doi.org/10.3390/s16050719.Search in Google Scholar PubMed PubMed Central

[37] Y. Monno, H. Teranaka, K. Yoshizaki, M. Tanaka, and M. Okutomi, “Single-sensor RGB-NIR imaging; high-quality system design and prototype implementation,” IEEE Sens. J., vol. 19, no. 2, pp. 497–507, 2019. https://doi.org/10.1109/JSEN.2018.2876774.Search in Google Scholar

[38] A. J. Ollanik, J. A. Smith, M. J. Belue, and M. D. Escarra, “High-efficiency all-dielectric Huygens metasurfaces from the ultraviolet to the infrared,” ACS Photonics, vol. 5, no. 4, pp. 1351–1358, 2018. https://doi.org/10.1021/acsphotonics.7b01368.Search in Google Scholar

[39] M. Decker, et al.., “High‐efficiency dielectric Huygens’ surfaces,” Adv. Opt. Mater., vol. 3, no. 6, pp. 813–820, 2015. https://doi.org/10.1002/adom.201400584.Search in Google Scholar

[40] W. Zhao, et al.., “Dielectric Huygens’ metasurface for high-efficiency hologram operating in transmission mode,” Sci. Rep., vol. 6, no. 1, p. 30613, 2016. https://doi.org/10.1038/srep30613.Search in Google Scholar PubMed PubMed Central

[41] M. V. Berry, “The adiabatic phase and Pancharatnam’s phase for polarized light,” J. Mod. Opt., vol. 34, no. 11, pp. 1401–1407, 1987. https://doi.org/10.1080/09500348714551321.Search in Google Scholar

[42] J. B. Mueller, N. A. Rubin, R. C. Devlin, B. Groever, and F. Capasso, “Metasurface polarization optics: independent phase control of arbitrary orthogonal states of polarization,” Phys. Rev. Lett., vol. 118, no. 11, p. 113901, 2017. https://doi.org/10.1103/PhysRevLett.118.113901.Search in Google Scholar PubMed

[43] J. Li, et al.., “Mechanisms of 2π phase control in dielectric metasurface and transmission enhancement effect,” Opt. Express, vol. 27, no. 16, pp. 23186–23196, 2019. https://doi.org/10.1364/OE.27.023186.Search in Google Scholar PubMed

[44] N. Yu, et al.., “Light propagation with phase discontinuities: generalized laws of reflection and refraction,” Science, vol. 334, no. 6054, pp. 333–337, 2011. https://doi.org/10.1126/science.1210713.Search in Google Scholar PubMed

[45] M. Pan, et al.., “Dielectric metalens for miniaturized imaging systems: progress and challenges,” Light: Sci. Appl., vol. 11, no. 1, p. 195, 2022. https://doi.org/10.1038/s41377-022-00885-7.Search in Google Scholar PubMed PubMed Central

[46] Y. Q. Fu, et al.., “Deep reactive ion etching as a tool for nanostructure fabrication,” J. Vac. Sci. Technol. B, vol. 27, no. 3, pp. 1520–1526, 2009. https://doi.org/10.1116/1.3065991.Search in Google Scholar

[47] C. Chang and A. Sakdinawat, “Ultra-high aspect ratio high-resolution nanofabrication for hard X-ray diffractive optics,” Nat. Commun., vol. 5, no. 1, p. 4243, 2014. https://doi.org/10.1038/ncomms5243.Search in Google Scholar PubMed

[48] K. Saito and T. Tatsuma, “Asymmetric three-way plasmonic color routers,” Adv. Opt. Mater., vol. 3, no. 7, pp. 883–887, 2015. https://doi.org/10.1002/adom.201500111.Search in Google Scholar

[49] Q. H. Xu, “Color routing at the nanoscale,” Light: Sci. Appl., vol. 8, no. 1, p. 58, 2019. https://doi.org/10.1038/s41377-019-0170-x.Search in Google Scholar PubMed PubMed Central


Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/nanoph-2023-0746).


Received: 2023-10-29
Accepted: 2023-12-27
Published Online: 2024-01-22

© 2024 the author(s), published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Frontmatter
  2. Editorial
  3. Editorial on special issue “Metamaterials and Plasmonics in Asia”
  4. Reviews
  5. Programmable optical meta-holograms
  6. Inverse design in photonic crystals
  7. Unleashing the potential: AI empowered advanced metasurface research
  8. Controlling thermal emission with metasurfaces and its applications
  9. Snapshot spectral imaging: from spatial-spectral mapping to metasurface-based imaging
  10. Research Articles
  11. Ultrafast snapshots of terahertz electric potentials across ring-shaped quantum barriers
  12. Ultra-thin, zoom capable, flexible metalenses with high focusing efficiency and large numerical aperture
  13. On-demand Doppler-offset beamforming with intelligent spatiotemporal metasurfaces
  14. Metasurface absorber enhanced thermoelectric conversion
  15. Achieving asymmetry parameter-insensitive resonant modes through relative shift–induced quasi-bound states in the continuum
  16. Collimated flat-top beam shaper metasurface doublet based on the complex-amplitude constraint Gerchberg–Saxton algorithm
  17. Quantifying robustness against sharp bending in an integrated topological interface of valley photonic crystals
  18. Transmissible topological edge states based on Su–Schrieffer–Heeger photonic crystals with defect cavities
  19. A metasurface color router facilitating RGB-NIR sensing for an image sensor application
  20. Deterministic reflection contrast ellipsometry for thick multilayer two-dimensional heterostructures
  21. Highly-efficient full-color holographic movie based on silicon nitride metasurface
  22. Spatial and time-resolved properties of emission enhancement in polar/semi-polar InGaN/GaN by surface plasmon resonance
  23. All dielectric metasurface based diffractive neural networks for 1-bit adder
  24. The perfect waveguide coupler with universal impedance matching and transformation optics
  25. Realization of all two-dimensional Bravais lattices with metasurface-based interference lithography
  26. The impact of hBN layers on guided exciton–polariton modes in WS2 multilayers
  27. Sample-efficient inverse design of freeform nanophotonic devices with physics-informed reinforcement learning
  28. Spintronic terahertz metasurface emission characterized by scanning near-field nanoscopy
  29. Lithium niobate thin film electro-optic modulator
  30. Holographic communication using programmable coding metasurface
Downloaded on 14.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/nanoph-2023-0746/html?lang=en&srsltid=AfmBOoqJujai4wcdFdgwlFmJzmfSmksMWr9216_TULwvlwxFoo5PVfcF
Scroll to top button