Contact Us

Dynamic 3D meta-holography in visible range with large frame number and high frame rate | Science Advances

tagsHdmi Splitter 1 In 3 Out

These authors contributed equally to this work.

Holograms are an ideal method for displaying three-dimensional images that are visible to the naked eye. The metasurface composed of sub-wavelength structures shows great potential in light field manipulation, which is very useful for overcoming the shortcomings of ordinary computer-generated holography. However, there are long-standing challenges in realizing dynamic meta-holography in the visible range, such as low frame rate and low frame number. In this work, we demonstrate a meta-holographic design that can achieve 2.

Different holographic frames and extremely high frame rate in the visible range (9523 frames per second). The design is based on the space channel metasurface and high-speed dynamic structured laser beam modulation module. The spatial channel is composed of silicon nitride nanopillars with high modulation efficiency. This method can meet the needs of holographic displays and is useful in other applications, such as laser manufacturing, optical storage, optical communication, and information processing.

Holographic technology is a technology that records and reconstructs light wave fronts, and is an ideal method for naked-eye three-dimensional (3D) display (

), optical data storage (

), and optical information processing (

). However, traditional holograms cannot create virtual objects or dynamically displayed holograms. To overcome these limitations, in 1966, Brown and Lohman (

) Invented the computer-generated hologram (CGH), which uses the theory of physical optics to calculate the phase map on the interference pattern. In addition, by using digital devices (such as spatial light modulators (SLM) or digital micromirror devices (DMD)), CGH can also perform dynamic holographic displays (


). However, for applications with large pixel sizes (such as small field of view (FOV), dual imaging and multi-order diffraction), CGH using SLM/DMD faces long-standing challenges


In recent years, with the rapid development of nanofabrication technology, metamaterials and metasurfaces have opened up a new era for hologram research and other research fields in Engineering Optics 2.0 (

). Metamaterials are composed of sub-wavelength man-made structures with innovative functions that exceed the limits of bulk materials. The manufacture of 3D metamaterials is very difficult. Therefore, the metasurface plays a significant role as an optical device in the visible light range. As a 2D metamaterial composed of subwavelength nanostructures, metasurfaces provide powerful tools for achieving amplitude, phase, and polarization modulation. The results show that the conventional optical ruler should be recast into a more general form (

). The research on metasurfaces can be divided into static and dynamic metasurfaces. The design of dynamic or active metasurfaces is based on the use of different materials and mechanisms, such as phase change materials (

),liquid crystal(


), mechanical strain (

), charge injection (

), thermo-optical effect (

), chemical and structural methods (

), and many more. Today, metasurfaces have been used to manufacture many different types of functional devices, such as metal materials (

), beam splitter (

), catenary optics (

) And Orbital Angular Momentum (OAM) equipment (

In terms of the second wavelength unit structure, metasurface holography has several major advantages, including larger FOV, high resolution, and elimination of high-order diffraction (

). Meta-holograms can be divided into three categories representing different physical mechanisms, namely phase-only meta-holograms (

), plain hologram (

) And complex amplitude hologram (

). Most of the ultra-holographic research in the visible light range is designed as a static device, which can only display a frame with a single meta-surface. However, dynamic design is essential for an ideal sub-holographic smooth display. In order to achieve this goal, there are two important considerations. The first is the frame number, which refers to the number of different frames that a single meta-hologram element can display. The second is the frame rate of the sub-holographic display (the reciprocal of the switching time between two frames), which can be quantified by "frames per second (fps)". The ultimate goal is to perceive and interpret discrete reconstructed holographic frames as smooth videos due to the permanence of the eyes. It is generally believed that video display with a frame rate higher than 24 fps is continuous to the human eye (

). A higher frame rate corresponds to a finer and smoother video display.

The latest development of dynamic meta-holographic technology in the visible light range is summarized. Dynamic meta-holography can be divided into two groups with different operating principles. The first group uses active metasurfaces, which will undergo physical and chemical changes through external control. Many different methods have been developed to achieve this goal, such as the use of phase change materials [e.g. Ge



(GST)] (

), the application of stretchable substrates (

), changing optical properties through chemical reactions (

), and rewrite the graphene oxide supersurface with a femtosecond laser (

). The other group is using static multiplexing metasurfaces. A variety of multiplexing methods have been adopted in recent studies, such as wavelength (

), incident angle (

) And polarization multiplexing (


). The complex modulation of incident light has been used in equipment such as OAM (

). In addition to the above-mentioned holographic holography performed by a single-type multiplexing method, a multiple multiplexing method is also proposed, which displays the recording of 63 holographic images multiplexed by wavelength and polarization in the same hologram (

). All these representations of dynamic meta-holography are progressive and inspiring. However, the realization of dynamic meta-holographic technology in the visible light range still faces severe challenges, as shown in Figure 1.

. First of all, as mentioned above, it is essential to display many different frames with a meta-hologram element to achieve a real holographic video. However, most current designs can only show a few different frames through experiments. Secondly, the modulation time in most attempts is too long to display smooth holographic video.

It proves that in most holographic research, the frame rate of dynamic sub-holography is never mentioned. The latest research on OAM multiplexing holography shows that it is expected that a relatively smooth dynamic hologram can be achieved.

Frames and 60 fps (

), and then the reconstructed image consists of an array of real points. This design is very creative, and is more suitable for all-optical encryption or storage, rather than holographic display. In the end, there is still a lack of efficient dynamic sub-holographic technology and excellent display quality in the visible light range, and it is impossible to display smooth holographic video with a large number of frames and a high frame rate.

"/" means that there is no relevant data in the reference.

In this research, we demonstrated a new visible light holographic design based on the visible light space spatial multiplexing toric surface, which can achieve 2

Different hologram frames and high frame rate (maximum frame rate 9523 fps). In addition, by using silicon nitride (SiN), the modulation efficiency of each spatial channel is very high (greater than 70%)

) Nano-pillar building block for metasurface.

The design of spatial channel sub-hologram (SCMH) is inspired by the comparison between dynamic sub-hologram and common 2D display technology. The ideal way to achieve dynamic hyperholography is to perfectly control each nanostructure of the metasurface. This means that each pixel of the element needs to be independently controlled at high speed, just like the function of a light emitting diode or a liquid crystal display. The recently published work shows a metasurface with individually controlled linear pixels, showing dynamic beam steering and focusing functions (

), which provides a feasible way to realize dynamic holography in the future. In addition to these pixel displays, there are two other methods that can be used to achieve dynamic 2D displays. One is to divide the entire graph into many different subgraphs and combine them at different times, for example, an electronic scoreboard or a digital tube display on an electronic meter. The other is to display different frames from a continuous video at different times, for example, a traditional movie that is recorded and shown as a movie. It can be concluded that both are spatial channel methods.

The physical mechanism of SCMH is explained in note S1 and explained in the example.

. The traditional hyper-hologram design involves the use of mathematical algorithms (such as the Gerchberg-Saxton algorithm) to calculate and reconstruct the corresponding phase map or amplitude map of the target object on the entire manufactured hyper-surface area. In order to create SCMH, the metasurface is divided into

Different spatial channels, which contain thousands of nanopillars (

). There are two different types of SCMH designs. The first is the spatial channel selective sub-hologram. In this design, if different spatial channels are opened at the same time, all reconstructed images of different spatial channels will overlap each other (

). By controlling the structured laser beam to open different spatial channels in a designed order, continuous frames of holographic video can be displayed at a specified time (

). Another design is the spatial channel multiplexing meta-hologram, where the reconstructed target image of different spatial channels is a sub-image of the entire hologram (

). According to the predetermined sequence, different spatial channels are opened at different times; therefore, there are 2

The combination

Bit space channel simultaneously (

). Different combinations of spatial channels can reconstruct different holographic images (

). By changing the structured laser beam to open different spatial channel combinations at high speed at different times, a dynamic full holographic display can be displayed in a smooth and delicate way (


) The structure of spatial channel sub-hologram elements. (


) Selective sub-hologram design of spatial channel. If all spatial channels are turned on at the same time, all reconstructed images will overlap each other (B). The dynamic sub-holographic display can be realized by opening the space channel in the design sequence (C). (


) Design of spatial channel multiplexing element hologram. The reconstructed images of different spatial channels are sub-images of the whole image (D). Open different spatial channels in different time series to form different spatial channel combinations (E), thereby reconstructing different images (F) to achieve dynamic sub-holographic display (G).

The above discussion shows that there are two important aspects to realize our dynamic meta-holographic design. The first is dynamic beam modulation, which is used to encode the spatial distribution of the incident structured laser beam. The module can be realized by a projection system composed of DMD, lens and microscope objective, as shown in the figure.

. The DMD modulates the incident light at a high speed, for example, the maximum is 9523 Hz in our experiment. Lenses and microscope objectives are used as a 4f system to reduce the structured incident beam to open different spatial channels of the metasurface. Another important aspect is the high-efficiency static spatial channel multiplexing metasurface in the visible light range. In this study, the static metasurface is composed of SiN

Nano pillars, as shown

. SiN absorption coefficient

The material is small enough that SiN

It is almost transparent in the visible light range. Refractive index

Close to 2, much larger than 2 of ordinary glass materials. These characteristics make SiN

This material is suitable for the design of high-efficiency metasurfaces with equivalent refractive index in the visible light range. SiN height

The nanopillars are all the same at 700 nm, the period of the rectangular lattice is 500 nm, and the radius varies from 90 to 188 nm. The nano-pillars were simulated by the finite difference time domain method (FDTD), and six suitable radii were selected for manufacturing. Characterization of SiN's Amplitude Transmission Efficiency and Phase Response

Nanopillars are shown in

As a function of nanopillar radius at 633 nm wavelength.

It means that for the selected radius, the transmission efficiency almost remains a high constant value higher than 90%, but it drops at the maximum radius of 188 nm, and the phase response varies from 0 to 2π.

A scanning electron microscope (SEM) image showing the manufacturing results. The scale bar in the SEM image is 1 μm.

) Dynamic spatial beam coding module. The DMD can modulate the incident light at a high speed, for example, the maximum is 9523 Hz in our experiment. Lenses and microscope objectives are used as a 4f system to narrow the encoded incident beam to illuminate different areas of the metasurface. (

Geometry of silicon nitride

Characterization of the amplitude transmission efficiency and phase response of nanopillars and SiN

Nanopillars as a function of the radius of the nanopillars at a wavelength of 633 nm. The illustration is the geometry of SiN

Nano pillars. (

) Scanning electron microscope (SEM) image of the result. Scale bar, 1μm.

As described above, one of the designs for dynamic display involves dividing the entire picture into sub-pictures, and illustrating different frames through the combination of different sub-pictures. This method can also be used in the design of spatial channel multiplexing holograms. In this research, a metasurface holographic digital tube display system was designed and demonstrated, as shown in the figure.

. The whole reconstructed target image is a "88:88" digital tube pattern, which is composed of 28 sub-images. Therefore, the metasurface is divided into 28 different spatial channels, which reconstruct the corresponding sub-images marked with numbers (see Figure S1 for detailed design). Take the frame "12:12" as an example. By encoding the spatial distribution of the incident structured laser beam, the metasurface can reconstruct a large number of different frames, thus representing a shared aperture design. This is a 28-bit design, so the total number of frames is 2

= 268,435,456.

) The structured laser beam opens the specific spatial channel combination and reconstructs the target image. (

) The first and third lines: 10 typical examples, ranging from 00:00 to 99:99; the second and fourth lines: the corresponding spatial channel coding mode of DMD. (

) The produced optical image of the metasurface and an enlarged view of a spatial channel. The scale bars are 100 and 30 μm. (

) Experimental results of dynamic spatial channel multiplexing holograms and corresponding patterns of structured laser beams.

Ten typical examples are proposed

Presentation design. In the first and third rows, the illustrations of the reconstructed target image range from 00:00 to 99:99. At the same time, the second and fourth rows show the corresponding spatial channel coding mode of the DMD. It is worth noting that, as shown in the figure, there are dark gaps between the encoding modes of the DMD, but there are no gaps between the sub-regions of the manufactured metasurface. Due to geometrical optical aberration and diffraction, the narrow laser beam irradiating the sub-area hardly diverges. At the same time, the dark gap may limit the crosstalk of adjacent sub-regions. The size of each spatial channel is 150μmx 200μm, and the overall size of the manufactured metasurface area is 1050μmx 800μm (

). The frame rate of the super holographic digital tube display depends on the switching time of the DMD encoding mode. In our experiment, the minimum DMD switching time is 105μs. Therefore, the frame rate can vary from 0 to 9523 fps, which is much higher than the frame rate associated with visual persistence limitations. The results of the experiment are

, The frame rate in this experiment is 1 fps. Facts have proved that our design can achieve a smooth dynamic sub-holographic display (see movie S1).

Another design of this research is the design of dynamic spatial channel selective sub-hologram, which is similar to the traditional film recorded and projected as a motion picture film. The metasurface samples are divided into many spatial channels, which will represent the reconstruction of different frames in the continuous video. In this design, 20 consecutive frames showing the rotation of the four capital letters "HUST" are selected from the short video (see Figure S2 for detailed design) as the reconstructed frames of the dynamic sub-hologram, as shown in Figure 2.

(For a detailed phase diagram, see Figure S3). The incident structured laser beam is modulated by the DMD into a spatial scanning beam, and illuminates different individual spatial channels of the metasurface in the designed order. Then, the reconstructed frame changes over time to display a dynamic sub-holographic movie, and the frame rate of the holographic video depends on the switching time of the DMD. The experimental results of each frame are shown in

. A short sub-holographic video demonstrates the utility of this method (see movie S2).

) The structured laser beam opens specific spatial channels in the designed order, and (

) Display consecutive frames of holographic video. (

) The dynamic 3D holographic display is realized through spatial channel selective sub-holograms.

The dynamic spatial channel selective meta-hologram design can be used to display 2D and 3D holographic videos. As shown in Figure 3, a metasurface is designed for 3D holographic video display.

. The entire annular sub-holographic element is divided into eight spatial channels, and each spatial channel is designed to reconstruct 3D arrows in free space. The geometric parameters are marked in

, The inner radius of the toroidal hypersurface is

= 150μm, outer radius

= 450 microns. The reconstructed 3D arrow is designed in the center circle with a radius of 125μm and a height of

= 2000μm and

= 2020 microns. Eight 3D arrows are connected end to end in free space. The reconstructed light field of each 3D arrow is followed by a self-made microscope

Axis (for detailed phase diagram and experimental results, see Figure S4). This shows that the design can be used for smooth sub-holographic displays (see movie S3).

Based on the DMD's dynamic spatial encoding of the incident beam and the space division multiplexing metasurface design, dynamic hyperholography with a large number of frames and high frame rate in the visible range has been realized. This article shows two different designs, namely dynamic spatial channel multiplexing and selective sub-holograms. Dynamic spatial channel multiplexing hologram can display 2

Different frameworks (over 200 million in this study). The dynamic spatial channel selective meta-hologram can display complex 2D and 3D holographic videos in the visible range. All three designs can achieve high frame rate (0 to 9523 fps), which is far beyond the limit of visual residual, so our method can display ultra-fine and smooth holographic video. The metasurface is composed of SiN

Nano-pillars designed by equivalent refractive index and simulated by FDTD. It is worth noting that each spatial channel of the manufactured metasurface has a high efficiency (greater than 70%) in the visible light range (for detailed calculations, please refer to Note S2 and Figure S5). Each spatial channel can be partially opened by modulating the spatial duty cycle (for signal-to-noise ratio analysis, please refer to Figure S6). Number of spatial channels

It depends on the manufacturing size, the FOV of the objective lens and the minimum size of each spatial channel. Thousands of spatial channels can be created using our design method (see note S3 for detailed analysis). The large number of frames, high frame rate and high efficiency not only enable this metasurface method to meet the demanding requirements of complex and smooth holographic displays in the visible light range, but also provide hope in many applications including laser manufacturing and optical storage. , Optical communications and information processing.

SiN manufacturing

The metasurface starts from a glass wafer substrate with a thickness of 500μm (Figure S7). Silicon nitride layer (

A 700-nm thick λ = 2.023 (at 633 nm = 2.023) was deposited on the substrate by plasma enhanced chemical vapor deposition. Then, a 20 nm chromium layer was deposited on top of the SiN by electron beam evaporation

The layer serves as a hard mask. Next, a 200 nm photoresist layer (CSAR62) was spin-coated on top of the Cr layer. The hologram pattern is written by electron beam lithography (Vistec: EBPG 5000 Plus) and implemented into the photoresist layer after development. Then the pattern is transferred to the Cr hard mask layer by inductively coupled plasma (ICP) etching (Oxford Plasmalab: System 100-ICP-180), and the residual is removed by an oxygen plasma stripper (Diener electronic: PICO plasma stripper) Photoresist. ). Finally, transfer the pattern to SiN

In the next ICP process, the remaining Cr is removed by the Cr etching solution. The Cr layer is used as a hard mask because the etching selectivity between Cr and SiN is extremely high


The optical components and settings of the dynamic space division multiplexing metasurface are shown in Figure 2. S8. The He-Ne laser (Pacific Lasertec, 25-LHP-991-230) with a wavelength of 633 nm propagates through the spatial pinhole filter and collimator lens, and becomes an extended laser beam with suitable beam quality. Then, the expanded laser beam is high-speed modulated by DMD (Texas Instruments, DLP6500FYE). The coded beam propagates through a 4f system consisting of a lens and microscope objective. The reconstructed holographic frame is collected by Fourier lens or objective lens and recorded by CCD.

For supplementary materials for this article, please visit:

This is an open access article distributed under the following terms:

, As long as the original works are properly quoted, they can be used, distributed and copied in any media without restriction.

Volume 6 Number 28

July 8, 2020

Thank you for your interest in promoting the term "scientific progress".

Note: We only ask you to provide an email address so that the person you recommend to the page knows that you want them to see the email and that the email is not spam. We will not capture any email addresses.

This question is used to test whether you are a visitor and prevent automatic spam submission.

Please log in to add an alert to this article.


: Eaba8595

The spatial channel design method can realize dynamic meta-holography with high frame rate and large number of frames.

Volume 371, Issue 6534


. all rights reserved. AAAS is a partner of the following institutions 


ISSN 2375-2548.