Multicolor holograms encoded into color print images are structures that generate holographic images when illuminated with lasers, while showing completely different images when viewed with the eye or with microscopes under white light incoherent illumination. Despite their promising applications in optical document security, they have been the subject of only a handful of research efforts, underscoring the need for further exploration in this area. Here, we propose a hybrid metasurface that achieves this functionality and thoroughly characterize its performance using simulations. In our device, nanohole arrays in an aluminum film function as plasmonic color filters for blue, green, and red channels with low crosstalk. Amorphous titanium dioxide (aTiO2) nanopillars comprise the hologram metasurface, which modulates the outgoing light's phase to produce a holographic image. Due to the subwavelength dimensions of the unit cell of the color filter (e.g., 415 nm for red, 315 nm for green, and 255 nm for blue wavelengths) and metasurface hologram (e.g., 430 nm for red, 360 nm for green, and 305 nm for blue wavelengths), the color print and holographic images can have very high resolution. Simulations reveal that the metasurface can be perceived as a tricolor image under incoherent white light, whereas under illumination from red, green, and blue lasers, three distinct holographic images can be observed.
Metasurfaces have favorable qualities for holography1 and color printed images,2,3 including wide viewing angles (for holograms) and ultra-high resolution (for color printed images). Metasurface holograms are composed of unit cells that contain either metallic nanostructures3–5 or dielectric nano atoms.6–8 For the latter, structural color originates from selective light reflection or transmission through optical phenomena, such as scattering, diffraction, refraction, and interference.9 Structural colors have favorable attributes, such as high color saturation, lack of photo-bleaching,10 and their ability to be produced by a wide range of materials, i.e., both dielectric6 and metallic.11,12
Applications of data encryption and document authentication have been revolutionized by the advent of optical security devices (OSDs). Recent advancements in nanofabrication have created exciting and interesting possibilities for OSDs, leading to efforts to incorporate nanomaterials for enhanced security and authentication.13,14 Additionally, secure holography utilizing wavelength multiplexing has been proposed in Ref. 15, with potential extension to watermarking.16 OSDs produced via nanofabrication often simultaneously use several attributes of light, such as amplitude, phase, polarization, and wavelength.17,18 By contrast, traditional OSDs generally use one specific property of light. For instance, holograms typically modulate the phase, while color prints modulate the amplitude. Holograms can be readily authenticated by using coherent light (e.g., from a laser) to project the holographic image onto a screen located in the Fraunhofer regime (far field), while color prints can be viewed with the eye or a microscope under white light illumination. Early works in the field of OSDs focused on single-function devices (i.e., color print image or hologram, but not both).19–22 In recent years, there has been a push to strengthen the security of OSDs by encoding multicolor holograms to color print images. Here, we achieve this functionality using a structure that has low crosstalk between color channels while being relatively easy to fabricate.
Despite the growing interest in integrated metasurface holograms, the number of published works on optical security documents that combine color printing and holography remains surprisingly low.23–28 This scarcity highlights the need for further research and exploration in this area, as the development of such dual functional documents holds significant potential for enhanced security and visual appeal. In Ref. 23, Lim et al. integrated polymer nanopillars (as a color filter) with a phase plate for simultaneous control of both the spectrum and phase of the outgoing light, thereby achieving dual mode functionality. In Ref. 24, Yoon et al. achieved dual mode functionality using amorphous Si nanoblock dimers. Two sets of dimers were used. The two sets had distinct reflectance spectra while maintaining identical cross-polarization conversion efficiency at the hologram's operating wavelength (635 nm). In Ref. 25, Zhang et al. employed an aluminum (Al) cross as a unit cell and used a femtosecond laser to precisely shape each arm of the Al cross to control resonance. They utilized this technique to create a metasurface capable of displaying both color and holographic images, switchable through polarization control, but with limitations in structural colors due to variations in nanostructure sizes. Their metasurface is capable of showing a color image in reflection channels and a holographic image in the diffraction channel. In Ref. 26, Wei et al. introduced a single layer dielectric metasurface that effectively merged two holographic images into a full-color printed image using amorphous silicon nanofins and nanodimers as their unit cells. When illuminated with white light, their single layer metasurface displayed a small-scale color image. It also concealed two separate holographic images, which could be revealed when subjected to red and green laser beams, projecting these holograms in the far-field. In Ref. 27, Bao et al. introduced a method that offers full control over the brightness of nanoprints and holograms through the use of silicon nanoblocks. Their unit cell consisted of four sub cells: one for the red channel, one for the green channel, and two for the blue channel. Each sub cell contained two Si nanoblocks positioned at an angle with respect to each other. This angle influenced the intensity of the transmitted light but did not affect the resonance properties. As a result, by adjusting the angle, one could control the brightness levels in the resulting nanoprints and holographic images. In Ref. 28, Wen et al. produced a dielectric metasurface comprising TiO2 cones on glass. They controlled the reflection spectrum by varying the period and radii of the cones while controlling the phase of reflection using detour phase.
Despite the benefits of dual functionality, integrated metasurface holograms that combine color printing and holography are hindered by design complexity, which limits the number of wavelengths for holography, and scalability issues. Furthermore, they require balancing the demands of both color printing and holography, resulting in potential tradeoffs. To address these challenges, we propose an OSD approach, which utilizes two distinct layers for dual mode functionalities. By encoding three-level holographic images and one tricolor color print onto the metasurface, we achieve a crosstalk between filter channels of less than 17%, outperforming previous reports presented in Ref. 28 (of 25%) and Ref. 27 (of 31%). Crosstalk is the most significant factor affecting the performance of the holographic image in dual functional metasurface holograms. Crosstalk refers to the unintended transmission of light at undesired wavelengths through a color filter meant for a specific wavelength. This can lead to color mixing in different channels of a holographic image, resulting in less pure colors. When a metasurface hologram is illuminated with red, green, or blue lasers, crosstalk can introduce unwanted noise into the holographic image, compromising its overall quality. For example, areas that should appear green in holographic images may end up having a noisy background from blue and red light. Notably, our approach leverages materials commonly found in the microelectronics industry, including glass, aluminum, and titanium dioxide, to ensure image quality and feasibility.
Figure 1 shows a 3D conceptual image of a small part of the metasurface hologram, with the color print image encoded on to it. It consists of two parts. The first part is the color filter and consists of three layers: an SiO2 layer (200 nm) at the bottom, an aluminum layer (150 nm) and a thick glass layer (500 nm) at the top. The second part, which controls the phase, consists of amorphous TiO2 (aTiO2) nanopillars that sit on the glass layer, all of fixed height (350 nm), but different in diameter and period. In the top left of Fig. 1, the intended color print (that would be seen with transmitted light) is depicted. Illumination of the metasurface by red, green, and blue laser beams would result in a color holographic image being produced in the image plane. The 3D schematic image in this figure depicts a small part of the metasurface. The metasurface contains red, green, and blue sections, each comprising a nanohole array that serves as a color filter. Each section is divided into regions that we refer to as pixels. Each pixel contains nanopillars (height: 350 nm) in square arrays, with the nanopillar diameter determining the phase imparted upon the transmitted light. Due to the square lattice and symmetric nature of the aTiO2 nanopillar, the pixels are polarization-insensitive. We quantize the phase into eight levels, with steps of between them. There are thus 24-pixel designs, i.e., for the three operating wavelengths and the eight phase levels. The geometric parameters of the nanoholes and nanopillars that constitute the metasurface are given in Table I. Due to the subwavelength pixel size of holographic images, this metasurface has a wide field of view and high resolution.
Three-dimensional schematic of a portion of dual mode device. Nanoholes in aluminum layer comprise the plasmonic color filter, while amorphous titanium dioxide (aTiO2) pillars comprise the hologram. Boundaries of red, green, and blue holograms (and color filters) are highlighted by red, green, and blue borders, respectively. The left graphic is a schematic illustration of device.
Three-dimensional schematic of a portion of dual mode device. Nanoholes in aluminum layer comprise the plasmonic color filter, while amorphous titanium dioxide (aTiO2) pillars comprise the hologram. Boundaries of red, green, and blue holograms (and color filters) are highlighted by red, green, and blue borders, respectively. The left graphic is a schematic illustration of device.
Geometric parameters of nanoholes and aTiO2 nanopillars.
Parameter . | Value (nm) . | Description . | Parameter . | Value (nm) . | Description . |
---|---|---|---|---|---|
415 | Nanohole square array period, λ = 635 nm design | 200 | Thickness of coating layer | ||
320 | Nanohole square array period, λ = 532 nm design | 500 | Thickness of glass film | ||
255 | Nanohole square array period, λ = 450 nm design | 210 | Nanohole diameter, λ = 635 nm design | ||
150 | Thickness of aluminum layer | 175 | Nanohole diameter, λ = 532 nm design | ||
145 | Nanohole diameter, λ = 450 nm design | 430 | Nanopillar ( ) square array period, λ = 635 nm design | ||
360 | Nanopillar ( ) square array period, λ = 532 nm design | 305 | Nanopillar ( ) square array period, λ = 450 nm design. | ||
350 | Nanopillar ( height |
Parameter . | Value (nm) . | Description . | Parameter . | Value (nm) . | Description . |
---|---|---|---|---|---|
415 | Nanohole square array period, λ = 635 nm design | 200 | Thickness of coating layer | ||
320 | Nanohole square array period, λ = 532 nm design | 500 | Thickness of glass film | ||
255 | Nanohole square array period, λ = 450 nm design | 210 | Nanohole diameter, λ = 635 nm design | ||
150 | Thickness of aluminum layer | 175 | Nanohole diameter, λ = 532 nm design | ||
145 | Nanohole diameter, λ = 450 nm design | 430 | Nanopillar ( ) square array period, λ = 635 nm design | ||
360 | Nanopillar ( ) square array period, λ = 532 nm design | 305 | Nanopillar ( ) square array period, λ = 450 nm design. | ||
350 | Nanopillar ( height |
(a) Schematic of unit cell and (b) transmission spectra of blue, green, and red color filters. The full width at half maximum (FWHM) values of the blue, green, and red color filters are ∼90, 100, and 125 nm, respectively. Transmission at intended wavelengths of lasers (450, 532, and 635 nm) are denoted by cross systems. (c) Phase imparted on transmitted light by nanoholes, (d) positions of optimized (black stars) and other (open circles) nanohole filters on CIE color diagram, assuming North Sky Daylight (D65) illumination, and (e) simulated color print image.
(a) Schematic of unit cell and (b) transmission spectra of blue, green, and red color filters. The full width at half maximum (FWHM) values of the blue, green, and red color filters are ∼90, 100, and 125 nm, respectively. Transmission at intended wavelengths of lasers (450, 532, and 635 nm) are denoted by cross systems. (c) Phase imparted on transmitted light by nanoholes, (d) positions of optimized (black stars) and other (open circles) nanohole filters on CIE color diagram, assuming North Sky Daylight (D65) illumination, and (e) simulated color print image.
Figure 2(b) shows transmission spectra of the optimized red, green, and blue color filters. These produce the color for the color print image and have low crosstalk (below 17%). Figure 2(c) plots the phase vs wavelength of the color filters. The phases imparted by the blue, green, and red color filters at 450, 532, and 635 nm are , , and , respectively. These are the wavelengths of the lasers that would be used to illuminate the hologram. Figure 2(d) shows the colors of the designed color filters on the CIE diagram, with each blue circle representing a color filter with distinct geometric parameters. The black stars denote the optimized color filters. Figure 2(e) is the simulated image of the color print that would be seen with transmitted light when the input illumination is North Sky Daylight (D65). The colors of some color filters are shown in supplementary material Fig. S1.
We next design our aTiO2 nanopillars, with which we control phase and therefore achieve hologram functionality. We use aTiO2 because it shows no loss at visible wavelengths and relativity high refractive index, making it a more suitable material for dielectric metasurface design than silicon nitride or titanium nitride. The nanopillar layer sits on top of the stack of layers containing the Al nanoholes (i.e., glass layer/Al/SiO2 layer), i.e., it imparts the necessary phase on the color-filtered light. As discussed, the target wavelengths of our device are = 450, 532, and 635 nm. One way to approach this would be to produce a design for one wavelength (e.g., 635 nm), then scale all dimensions to translate this design to the other two wavelengths. This would be in principle feasible because the refractive index of aTiO2 is approximately the same at the three wavelengths. However, it would of course mean that the pillars would have different heights, which would complicate fabrication. To ensure consistency and simplify the fabrication process,30 we decide against having different nanopillar heights. Instead, we opt for a common nanopillar height of 350 nm. We optimize by varying the nanopillar period and radius. We choose the nanopillar height to be 350 nm because it produces the necessary 2π phase for all desired wavelengths.
Simulation results for the nanopillars designed for = 635 nm are shown in Fig. 3. We plot the transmission coefficient (amplitude) and phase as a function of nanopillar period and normalized radius . The phase plotted in Fig. 3 is the difference between the phase imparted to the wave when nanopillars are present and the phase without the nanopillars. The mechanism behind the phase generation can be understood by thinking about the aTiO2 nanopillars as optical waveguides with fixed lengths. Changing the radius of the aTiO2 nanopillar changes the propagation constant of the circular waveguide. This is not the only mechanism that applies phase; from Fig. 3, it can be seen that there are features related to resonance. These originate from Mie magnetic dipole and electric dipole resonances in the dielectric cylinder and enable us to achieve a phase range of 0–2π with the nanopillar height being moderate. For the range of normalized radius and period plotted in Fig. 3, this Mie resonance feature starts at ∼ 350 nm and ∼ 0.9 and extends to ∼500 nm and ∼ 0.38. It can be seen here that for = 430 nm (denoted by horizontal dashed line), the transmission coefficient is above 72% and the phase of the transmitted light varies between 0 and 2π. We can thus achieve full control over the phase of the transmitted light by varying the radius. By repeating this design process for the other two wavelengths, we find that having = 360 and = 305 nm allows us to achieve full control over phase for the green ( = 532 nm) and blue ( = 450 nm) channels, respectively. It does this while maintaining transmission >72% for red, >65% for green, and 80% for blue. See supplementary material S2 for further details at other wavelengths.
Transmission amplitude and phase of amorphous TiO2 pillar as a function of normalized pillar radius and pillar period, for a wavelength of = 635 nm. It can be seen that, for pillar period , transmission exceeds 70% and the phase can be varied from 0 to by changing the normalized pillar radius.
Transmission amplitude and phase of amorphous TiO2 pillar as a function of normalized pillar radius and pillar period, for a wavelength of = 635 nm. It can be seen that, for pillar period , transmission exceeds 70% and the phase can be varied from 0 to by changing the normalized pillar radius.
The design process for the metasurface hologram is as follows. We choose a tricolor image (Fig. 1) for the color print, i.e., we select the specific regions that will transmit laser beams at the wavelengths 450 nm (blue), 532 nm (green), and 635 nm (red). Figure 4(a) plots the simulated transmission of the color print (i.e., of the color filters) at the target wavelengths for illumination at normal incidence. We next construct the aTiO2 nanopillar hologram designed for 635 nm illumination in the red region of color print image, we construct the aTiO2 nanopillar hologram designed for 450 nm in the blue region, and the aTiO2 hologram designed for 532 nm in the green region, using a technique similar to that used in Ref. 26. The holograms are designed using MATLAB (MathWorks Corporation).31 The design process involves the following steps. First the target holographic images for three channels are read and stored in three matrices (each being 512 × 512). To avoid overlap between the twin image and the target image (main signal), we use an off-axis configuration for generating the phase of the target images.24 The next step involves applying the Gerchberg–Saxton algorithm to design a hologram for each color channel. The algorithm is iterated 50 times to calculate the necessary hologram phase, which is then rounded to eight levels. We then use the previously designed 24 pixels comprising the aTiO2 nanopillars. The image field of the designed hologram is determined by conducting a fast Fourier transform (Fraunhofer range). Finally, the three channels are recombined to generate the reconstructed image.
Figures 4(b) and 4(c) show the schematic of a measurement setup that could be used for characterizing the proposed metasurface hologram. Here, two sources are used for characterization: an incoherent white LED and a visible-wavelength supercontinuum laser producing a coherent beam. In Fig. 4(b), a CCD camera is used to capture the image from the color printed image of the metasurface hologram. In Fig. 4(c), we put the CCD camera in the Fourier plane of a lens to see the holographic images of the metasurface hologram.
(a) Simulated intensity transmission (T) of color print image (see also Fig. 2) at three intended wavelengths ( = 635, 532, and 450 nm). Schematic of measurement setup for characterization of the metasurface hologram with incoherent white LED light (b), and with supercontinuum laser filtered for red, green, or blue light (c). Simulated outputs of the CCD cameras are shown next to the CCD camera for both cases.
(a) Simulated intensity transmission (T) of color print image (see also Fig. 2) at three intended wavelengths ( = 635, 532, and 450 nm). Schematic of measurement setup for characterization of the metasurface hologram with incoherent white LED light (b), and with supercontinuum laser filtered for red, green, or blue light (c). Simulated outputs of the CCD cameras are shown next to the CCD camera for both cases.
First metasurface. Upper left panel: target image, which consists of a red apple (with green leaf and white reflection), blue-colored blackberry, and text (green font). Lower left panel: reconstructed image. Right panels: reconstructed image of each color channel.
First metasurface. Upper left panel: target image, which consists of a red apple (with green leaf and white reflection), blue-colored blackberry, and text (green font). Lower left panel: reconstructed image. Right panels: reconstructed image of each color channel.
Here, and are the target image and the reconstructed image, respectively, and and are the indices for the rows and columns of those images. In addition, the mean square error of two images is . These calculations consider the image dimension ( rows, columns) and the data type's maximum range ( ), where is 1 for double-precision floating point data and 255 for 8-bit unsigned integers.31 For the first image, i.e., the red apple with green leaves and black berry, the total SNR is ∼3.7 dB and the PSNR is ∼19.24 dB. For the red, green, and blue channels, the SNR values are ∼5.07, 2.09, and 4.25 dB, respectively. The PSNR values for the red, green, and blue channels are 24.7, 23.37, and 24.38 dB, respectively.
The second metasurface shows an individual in three running and jumping poses that are colored red, green, and blue. The reconstructed image is shown in Fig. 6 and show good agreement with the targets. Reconstruction is performed as described above, i.e., using the fast Fourier transform. At the center of reconstructed images, we see a white light spot. This is the zeroth order image of the target images, which is suppressed by using off-axis technique we mentioned before. For the second metasurface, the total SNR is calculated to be ∼3.95 dB and the PSNR is ∼19.4096 dB. For red, green, and blue channels, the SNR values are ∼4.9, 3.61, and 3.93 dB, respectively. The PSNR for the red, green, and blue channels are 25.3, 22.89, and 25.12 dB, respectively. We propose a fabrication procedure for multicolor hologram encoded into color print in the supplementary material Sec. 3.
Second metasurface. Upper left panel: target image, which consists of an individual in three running/jumping poses. Lower left panel: reconstructed image. Right panels: reconstructed image of each color channel.
Second metasurface. Upper left panel: target image, which consists of an individual in three running/jumping poses. Lower left panel: reconstructed image. Right panels: reconstructed image of each color channel.
In conclusion, we propose a hybrid metasurface that performs both color filtering and phase control of the transmitted light. This monolithically integrated metasurface consists of an aluminum film into which nanoholes are formed and aTiO2 nanopillars. The color filtering by the nanoholes enables the generation of a tricolor image under illumination by incoherent white light. In conjunction with the phase from the nanopillars, this also results in a holographic image when the metasurface is illuminated by RGB laser beams. Our work enables the achievement of heightened security levels and improved visual attractiveness in metasurfaces. To validate our approach, we simulate two holographic images. The simulation results show very good agreement with the target images with low crosstalk between color channels.
SUPPLEMENTARY MATERIAL
See the supplementary material for the following: (1) Section 1—showing the simulation of nanoholes color filter for different periods and different diameters, (2) Section 2—illustrating the transmission and phase of the aTiO2 nanopillar at 450 and 532 nm, and (3) Section 3—proposing a fabrication procedure for the device (multicolor hologram encoded into color print).
This work was supported in part by the Australian Research Council Centre of Excellence for Transformative Meta-Optical Systems (Project No. CE200100010). S.S.M.K. acknowledges receiving a scholarship from the University of Melbourne–Shanghai Jiaotong University (SJTU) Joint Ph.D. program.
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
Seyed Saleh Mousavi Khaleghi: Conceptualization (lead); Data curation (lead); Formal analysis (lead); Investigation (lead); Methodology (lead); Project administration (lead); Software (lead); Validation (lead); Visualization (lead); Writing – original draft (lead). Dandan Wen: Conceptualization (equal); Investigation (equal); Supervision (equal); Writing – review & editing (supporting). Jasper Cadusch: Conceptualization (supporting); Investigation (supporting); Supervision (supporting); Writing – review & editing (supporting). Kenneth B. Crozier: Conceptualization (supporting); Funding acquisition (lead); Project administration (lead); Resources (lead); Supervision (lead); Writing – review & editing (lead).
DATA AVAILABILITY
The data that support the findings of this study are available from the corresponding author upon reasonable request.