Invisibility cloak with image projection capability

Investigations of invisibility cloaks have been led by rigorous theories and such cloak structures, in general, require extreme material parameters. Consequently, it is challenging to realize them, particularly in the full visible region. Due to the insensitivity of human eyes to the polarization and phase of light, cloaking a large object in the full visible region has been recently realized by a simplified theory. Here, we experimentally demonstrate a device concept where a large object can be concealed in a cloak structure and at the same time any images can be projected through it by utilizing a distinctively different approach; the cloaking via one polarization and the image projection via the other orthogonal polarization. Our device structure consists of commercially available optical components such as polarizers and mirrors, and therefore, provides a significant further step towards practical application scenarios such as transparent devices and see-through displays.

Scientific RepoRts | 6:38965 | DOI: 10.1038/srep38965 see-through 3D displays and augmented reality technologies 27,28 that could be beneficial to the thrust for transparent electronics in the future.
In this Letter, we experimentally demonstrate an invisibility cloak where a large object can be concealed in the cloak structure and any images can be projected on it simultaneously. This unique functionality can be obtained through the use of one polarization of light for the cloaking and the other orthogonal polarization for the image projection due to the insensitivity of human eyes to the polarization. Such approach has never been captured by any other reports in the past.

Results
Mechanism of our invisibility cloak with the image projection capability. Our cloak structure consists of commercially available optical components; polarizers for oblique incidence (P o, [1][2][3][4] ) and for normal incidence (P n,1 ), and mirrors (M 1-4 ), as shown in Fig. 1(a). We regard our structure as a twelve-port device and each port is labeled by 1 to 12. In the description of the mechanism of our structure below, we consider that the scene behind the cloak structure is built at the observation point in front of the cloak structure, assuming ideal performance of each component. The regions of the scene are labeled by I to IV, and here we assume that regions I and IV (II and III) are observed via the horizontally (vertically) polarized light. Due to the symmetry of the cloak structure, we consider here the half of the structure for the explanation of the cloak mechanism. To illustrate the , and a polarizer for normal incidence (P n,1 ), where port 4 has the polarizer P n,1 , and port 9 does not have a polarizer for comparison. Labels of transmission coefficients of (b) a polarizer for oblique incidence, (c) a polarizer for normal incidence, and (d) a mirror for the use in Table 1. (e) Experimental observation of the cloak structure. The inset shows the experimental setup. A toy car is placed behind the cloak structure and a cylindrical object is partly inserted in the cloak structure. The white number "1" or "2" printed on a black paper reversely in the left and right is attached on each lateral side, port 4 or port 9.
Scientific RepoRts | 6:38965 | DOI: 10.1038/srep38965 image projection capability, one lateral side (port 4) has the polarizer P n,1 and the other lateral side (port 9) does not have a polarizer for comparison.
Consider first the unpolarized light traveling in the − y direction from region II of the scene behind the cloak structure in Fig. 1(a). The light is reflected by mirror M 1 into the + x direction. Here we assume that at ideal obliquely incident polarizers P o1-4 , the vertically polarized light is bent by 90° (T po,bn,V = 1 and T po,st,V = 0 in Fig. 1(b)) and the horizontally polarized light passes straight through (T po,bn,H = 0 and T po,st,H = 1 in Fig. 1(b)). Thus, the vertically polarized light is bent twice at P o,1 and P o,2 and then comes out at port 6 via the reflection at mirror M 2 . Likewise, for the unpolarized light coming from region I, the horizontally polarized light goes through the two polarizers P o,1 and P o,2 and comes out at port 5. Other orthogonal polarized lights coming from areas I and II go out at port 3 (light paths are not presented). In other words, the scene behind the cloak structure is built at the observation point through the use of 50% incidence.
In our structure, the capacity for another 50% light is used for the image projection. The information sources "1" and "2" are placed at ports 4 and 9, respectively, which are reversed in the left and right due to the mirror reflection. Consider the light paths from port 9. The horizontally polarized light goes through polarizer P o,3 and is reflected by mirror M 3 and then goes out at port 7 while the vertically polarized light comes out at port 8 by a 90° bent at the polarizer P o, 3 . As a result, we observe "2" in both regions III and IV. On the other hand, from port 4, we observe "1" in only region I by inserting polarizer P n,1 that allows the transmission of the vertically polarized light. Therefore, the information can be projected in any region(s) as desired. Table 1 shows the summary of light paths with labels of ideal components defined in Fig. 1(b-d). We comprehend the mechanisms of the cloaking and the image projection presented above.
Experimental demonstration of the invisibility cloak with the image projection capability. The experimental setup is shown in the inset of Fig. 1(e). Our cloak was implemented with four wire-grid polarizer cubes (89-604, Edmund Optics) and four right angle mirrors (45-595, Edmund Optics). White numbers "1" and "2" are printed on black papers reversely in the left and right, respectively, and these papers are placed at lateral sides. A wire-grid polarizer film (47-102, Edmund optics) is attached on the black paper having "1", where the vertically polarized light passes through. A cylindrical object is placed within the cloak structure, where the cylindrical object is partly in the concealed region while remaining upper part is exposed outside the cloak structure. There is a toy car behind the cloak structure and we capture screen shots through a camera in front of the cloak structure. Figure 1(e) shows an experimental observation of the cloak structure from the camera. The cylindrical object becomes invisible in the cloaking area and the car behind the cloak structure is observed. We observe "1" in region I and "2" in regions III and IV, as we designed. Therefore, the mechanism of our invisibility cloak with the image projection capability has been experimentally verified for human eyes.
Performance of the image projection capability. We further investigate the image projection ability, particularly, the placement of the image projection. We show that an image can be projected at the middle of neighboring regions as well as each region. As expected from the experimental demonstration of Fig. 1(e), the text "CLOAK" appears at region I ( Fig. 2(b)) by selecting the vertical polarization ( Fig. 2(a)), which is directly reflected by the side polarizer P o2 , and appears at region II (Fig. 2(e)) with the horizontally polarized light going through the polarizer P o2 and being reflected by the mirror M 2 instead (Fig. 2(d)). In this experimental demonstration, the wire-grid polarizer film (47-102, Edmund optics) was rotated by 90° for the polarization change. The text "CLOAK" can be projected at the middle of regions I and II (Fig. 2(h)) via the horizontally polarized light for "CL" and the vertically polarized light for "OAK" (Fig. 2(g)). Therefore, proper use of polarizers for an image enables the projection in any places on the cloak structure while the invisibility cloak phenomenon is maintained (Fig. 2(c,f,i)).
View angle dependence. The dependence of our structure on the view angle is experimentally and numerically investigated and shown in Fig. 3. To illustrate the main reason for the degradation of the rebuilt scene with the increase of the view angle, the cloak structure now consists of polarizers and mirrors that are attached on glass plates (Fig. 3(a)), instead of glass blocks. This allows us to envision what happens in the cloak structure by optical Output port  Table 1. Transmission coefficients of the cloak structure of Fig. 1(a). The cloak structure is regarded as a twelve-port device and transmission coefficients for six ports are presented due to the mirror symmetry of the structure. Each coefficient in the table is defined in Fig. 1(b-d). Parentheses (T j=V ,T j=H ) represent transmittances for vertical polarization and horizontal polarization, assuming ideal performance of each component (T m,j = 1, T po,st,j = 1 or 0, T po,bn,j = 1 or 0, and T po,st,j T po,bn,j = 0). 1/0 represents that the transmittance can be changed by polarized light.
paths clearly without the effect of the refractive index of glass. The half of the structure is used for this investigation due to the symmetry of our cloak structure; the scene including a car has the same size as the summation of regions I and II. Optical paths from the original scene are calculated and the scene rebuilt at the observation plane is obtained (Fig. 3(b)) by using the commercial optical design software Zemax 29 . The view angle has been set at 0°, 2°, and 4°, respectively. In the normal direction, the rebuilt scene is same as the original scene except for the half of the brightness (Fig. 3(d)) due to the use of 50% light (Fig. 3(e)). As the view angle increases, the rebuilt image is degraded manifesting as the lack of the information (black lines in Fig. 3(g,j)) and the wrong position of the information (the middle part of the red car appears at the left edge in Fig. 3(g,j)), respectively. These can be analyzed and well understood from optical paths tracing as shown in Fig. 3(e,h,k) for different incident angles. Taking the 4° incidence for example, light path A-B (yellow color ray) represents one source of image information lacking as the light ray is reflected to the neighboring wire-grid polarizer P o2 directly at the oblique incidence.
On the other hand, the wrong position of the output image information is credited to the rays near F-G-H path (highlighted as the cyan color), where unpolarized light is directly incident onto the wire-grid polarizer at large angles of incidence and then split into two different paths with one path going to the correct position while the other to the left edge. It is worthwhile to mention that there is another source represented by and near C-D-E (red color) for both the two image degradation results. This portion of light is completely misled to the output edge leaving a black line at the center. Note that the lack and the wrong position of the information mentioned above are consistent with the cloak structure of Fig. 1(e).

Discussion
In recent times, modern electronic devices such as cellphones, smart glasses, and computers are desired to be transparent 30 . The need for transparent objects with the display function is not limited to personal electronics. Tremendous applications have also been envisioned from the automotive industry, household appliances to healthcare sector. Typically, two approaches are adopted to realize transparent devices; firstly, a quest for materials with the unusual combination of high electrical conductivity and high optical transparency such as Indium Tin Oxide, and the composites of carbon nanotubes, graphene, and silver nanowires 31 . Despite the recent significant progress, mass application of these materials has remained challenging. On the other hand, driven by the desire to develop augmented reality experience, transparent or see-though displays are aimed to project virtual images of 2D or 3D objects utilizing optical components such as wedge-shaped prisms, spherical mirrors, array of lenses, etc. 32 . Holography based optical elements have also been used 33 . However, as we discussed before, optical cloaking approaches combined with the imaging systems have not been explored in the past. In our experiment, we have used printed papers as information sources for the experimental demonstrations in order to verify the mechanism and architecture of the display aspect in combination with cloaking the central part of the device. In practical applications, one can implement two thin-type displays having electrically switchable polarizers at ports 4 and 9 and project text messages and movies at any locations of the cloak structure. We believe that combining the invisibility cloaking with the image projection capability promises an alternative route to realize next generation transparent devices.
In conclusion, we have explored a cloak structure that provides the image projection capability. Due to the insensitivity of polarization for human eyes, the cloaking and the image projection are simultaneously obtained via the use of orthogonal polarizations in our structure. The mechanism of our structure was experimentally verified for human eyes. Our cloak structure consists of commercially available optical components and provides a significant further step towards practical applications in see-through displays and electronics. Experimental observations of the scene including a red car ((c,f,i)), the simulation results ((d,g,j)), and the corresponding light paths ((e,h,k)) are presented at view angles of 0° ((c-e)), 2° ((f-h)), and 4° ((i-k)), respectively.