Panasonic Announces Technology to Improve CMOS Sensor Sensitivity

Feb 5, 2013
Ikutaro Kojima, Tech-On!
Images taken by a conventional method (left) and the new method (right)
Images taken by a conventional method (left) and the new method (right)
[ If it clicks, the expanded picture will open ]
Color separation by the conventional (left) and new (right) methods
Color separation by the conventional (left) and new (right) methods
[ If it clicks, the expanded picture will open ]

Panasonic Corp invented and developed the "micro color splitter," a device that utilizes diffraction of light for color separation.

By using the device for an image sensor, the company enabled to take color images with a high sensitivity. The new method realizes twice as high a sensitivity as traditional color separation methods using color filters.

According to Panasonic's announcement, for image sensors used in digital imaging devices such as digital cameras and camcorders, medical devices and security devices, a color filter is usually attached to each pixel. In the Bayer array, which is the most widely adopted structure, each of the red, green and blue pixels is equipped with a filter that transmits only red, green or blue light, respectively. As a result, 50-70% of the amount of incoming light is lost.

As the resolutions of image sensors used in mobile devices are increasing and their pixel sizes are decreasing, there are increasing demands for higher sensitivities.

Applicable to CCD, CMOS sensors

The newly developed device has three major features. First, with its high light use efficiency, it became possible to take color images that are twice as bright as before.

Second, it can replace the color filters in conventional image sensors regardless of the type of image sensor (CCD sensors, CMOS sensors, etc). Third, the new device can be manufactured by using inorganic materials and processes that are used for the production of existing semiconductor devices.

The application of the micro color splitter to an image sensor was realized by newly developing the following three elemental technologies. They are (1) a wave analysis/optical design technology that calculates the behavior of light wave at a high speed and accuracy, (2) a device technology for realizing the micro color splitter that controls the phase of light passing through a plate-like high-refractivity transparent material and diffracts light in a minute area for color separation and (3) a layout technology and algorithm that reproduce colors with a high sensitivity and resolution based on signals detected by combining color-separated lights in an image sensor.

High speed, accuracy

For the first elemental technology, Panasonic realized a high speed and accuracy at the same time. As a method to analyze light wave, the FDTD (finite-difference time-domain) method is widely adopted, the company said. But with the method, it takes time to analyze light wave, making it difficult to use the method for designing the micro color splitter, which requires a large amount of calculation.

On the other hand, there is a fast analysis method called "Beam Propagation Method (BPM)." But it has a low accuracy and cannot accurately analyze the phenomenon of color separation.

Therefore, Panasonic newly developed a practical optical design technology that realizes a high-speed, high-accuracy analysis. It separates a space into areas that have different optical constants and applies the BPM to each area. As a result, it became possible to accurately express optical phenomena such as reflection, refraction and diffraction.

This technology can be used not only for the micro color splitter but also for optical designs for minute areas.

Manufacturable with common semiconductor processes

(Continue to the next page)

Manufacturable with common semiconductor processes

The second elemental technology (device technology) enables to make the micro color splitter by using commonly-used semiconductor processes. With the micro color splitter, light is dispersed because of the refractive index difference between a plate-like high-refractivity material (which is thinner than light wavelength) and a material around it. In other words, it is possible to control the phase of transmitted light by changing shape parameters, generate a diffraction phenomenon in minute areas and separate light by color.

The micro color splitter can be formed by using common semiconductor processes. By changing its shape, it is possible to separate light into specific colors and their complementary colors without reducing the amount of light as well as to separate light into colors ranging from blue to red like prism, Panasonic said.

Technology used after color separation

The layout technology and algorithm are necessary for the process that comes after color separation with the micro color splitter. The lights separated by the splitter come to the detector plane of the image sensor in a mixed state. Therefore, it is necessary to design a pixel layout so that it becomes possible to cope with lights in such a state.

By combining the layout and an arithmetic processing technology for mixed color signals and optimizing them, Panasonic enabled to reproduce high-sensitivity, high-resolution color images. For example, with a structure that separates light into primary colors and complementary colors, four colors, which are "white + red," "white - red," "white + blue" and "white - blue," can be obtained. And it is possible to convert them into normal color images without deteriorating resolution by arithmetic processing.

Some of the latest results were published on the online version of Nature Photonics Feb 3, 2013. Panasonic has received or filed applications for 21 patents in Japan and 16 patents in other countries for the micro color splitter.