STORY

Image Sensor

Establishing the technology to reduce noise of image sensors: curious mind leads to delivering new imaging experience

January 11, 2023

Taking photos with a smartphone has become a regular feature of our lives, and it drives a trend for smaller pixels in smartphone cameras. This, in turn, facilitates higher definition pictures with higher resolution, boosting the value of smartphone photography. Not widely known is the fact that this photography experience is made possible by the technology developed to a high degree of perfection by researchers at Sony Semiconductor Solutions Corporation (SSS). To make smaller pixels for smartphones, there was always the same problem, that they became more prone through the manufacturing process to producing noises. The person who resolved this problem is Hideki Minari, a winner of 2021 Sony Outstanding Engineer Award.*1 He identified the mechanism that caused a type of noise known as a “white pixel” and established a foundational technology to significantly reduce the noise. He already saw the problem before it became a known challenge and tackled it to establish a solution. We interviewed him to find out what difficulties he had to face during the research and his work ethics as a researcher.

*1) The Sony Group’s most prestigious award program that recognizes individual engineers for their remarkable contributions in order to inspire engineers to take on new challenges.

Hideki Minari

Sony Semiconductor Solutions Corporation
Imaging & Sensing Device Development Division

Profile:Minari joined Sony Corporation in 2011 and worked on the development of simulation technology for image sensors. In 2012, he was seconded to Belgium to take part in a joint development consortium for advanced semiconductor devices. Returned to Japan in 2014, he joined the team for the development of the SWIR (short-wavelength infrared) image sensor, which adopted compound semiconductors to its photoelectric conversion layer. He has also been pursuing the research to analyze the white pixel phenomenon and develop technology to mitigate it for smartphone image sensors.

A new challenge created by advancing smartphone image quality

拡大 拡大アイコン

There are several factors that are detrimental to the picture quality of an image sensor. The one Minari addressed is a type of noise called a white pixel. You may have seen a photo of a night scene or dark room which has white dots over the dark areas. They are the white pixels, caused often by defects or metal contaminations in pixels. Minari analyzed this white pixel phenomenon using semiconductor simulation technology, identified the mechanism through which the defects developed to cause the white pixels, and established a foundational technology to substantially reduce the noises.

“In an image sensor, there is a part that takes in light, called a pixel. It is made of silicon, and if the silicon’s atoms are perfectly aligned, there is no noise. In a manufacturing process to make an image sensor, however, the silicon undergoes chipping, heating, coating, and so on. If the silicon crystal is damaged during these processes, the silicon produces extra electrons during it converts the light into electrons,” says Minari. This additional electron causes the white dot as the sensor detects it and interprets it as light where there is actually no light.

It is relatively recently that these ‘white pixels’ became an issue in relation to the development of image sensors for smartphones. “When the pixels were still relatively large, they rarely got damaged in the manufacturing process. Today, they have become smaller and are more prone to it.” The white pixel problem has a considerable impact on development schedules. Developing an image sensor is a very time-consuming project. You test the design and proceed with the development, and when you have a prototype made, you find out some white pixels. Then, you will have to solve the problem and start over to make another prototype. In order to deliver the latest image sensor to clients without delay, it is crucial to understand what physically goes on inside the sensor that causes white pixels and other problems.

Project that started from building a team
Probing painstakingly both design and manufacturing process to identify the cause

Minari had experience in researching the simulation technology for advanced logic devices as a student and, after joining Sony, during his overseas posting. “In advanced logic devices, the atomic alignment is crucial in the component materials of a transistor. I had an inkling that we would get to the atomic level physics if we traced the origin of the noise like a white pixel in an image sensor,” he recalled. This tracing is no easy task. The problem may be in the image sensor design or in the manufacturing process. Both possibilities must be scrutinized. He needed many engineers to help with the investigation. However, it was not possible to propose a development of an image sensor for his research, as it was no more than his personal hypothesis. Although there were early signs of gradual increase of white pixels seen in the commercial development of image sensors, he did not have definite proof that justified a major research operation.

His solution was to approach an ongoing development project and ask for cooperation. He proposed to give insights as to the cause of the noise and how to address it, and in return, he would have access to the sensor design, manufacturing process and prototype. With the help of engineers from Sony Semiconductor Manufacturing’s Kumamoto Technology Center, he conducted the investigation and developed a model to show how the white pixel was caused. While the investigation was in progress to identify the causal mechanism to an extent, he had no place to test his ideas for mitigating the white pixels.

For this problem, he had a plan—to form like-minded team. He started by making a presentation to the engineers in the division to explain the cause of white pixels and ideas for fixing the problem. He says that he prepared this presentation very carefully. “At SSS, we have various engineers coming from different backgrounds. Everybody is passionate about technology, but not all of them studied theories. If the presentation was too academic, they might find it difficult to understand, and it would possibly fail to engage them.” He thus arranged his presentation with illustrations and analogies so that people could easily visualize the physical phenomena that were taking place in an image sensor. “The food wrap is strong against the sideways tension, but it tears easily if you make a tiny incision. It is the same in small pixels. If they get scratched or contaminated during the manufacturing process, they can easily develop faults.” He spoke to many engineers in this manner.

Soon, he gained volunteers to take part in his ideas. It did not take long before they started taking their own initiative and conducted experiments or made observations. Eventually, these efforts culminated in the development of technology that considerably reduced the noise. “I am really impressed that SSS is blessed with so many engineers with a curious mind. I find that the secret of improving the quality and speed of technological development is to stimulate their curiosity.” (Minari)

The decision to join SSS
came from the change of heart: from wanting to know to using knowledge with purposes

Minari’s inquisitive and curious personality developed from early on. He grew up in the countryside where children, even into their teens, had mountains for their playgrounds. When he was a high school student, he saw a mobile phone for the first time in his life. He was not only impressed, but inspired to understand how it worked. The more he understood, the more curious he became to know. It was perhaps a natural outcome that his infinite curiosity led him to the very device that moved electronic devices across the world—semiconductors.

At university, he used to play a tennis game with an assistant professor on a projector in the research lab. Having this laid-back personality on the one hand, Minari was also technology-savvy to be able to connect several desktop computers to form a computer cluster. “The research lab did not have a massive budget to be able to afford to a supercomputer. Outsourcing simulations was also costly. You can spend that money on some desktop computers and connect them yourself. This way is far more cost-effective. For the same budget, you get three times better performance.” He says that he gained knowledge in information technology because his professor was very knowledgeable in the subject. It is evident, also as in the episode of playing games with the assistant professor, that his sociable personality was an enabler to form his team for his R&D project.

Minari’s experimental computer cluster played a significant role in making his research a success. Thanks to the strong computing capacity, he could run large-scale simulations, and his research paper was presented at many international conferences and published in many journals. It was this research that led him to joining SSS. He was considering a teaching career when he was a Ph.D. student. In fact, he had offers of a teaching post from several universities. He changed his mind when he participated in the largest conference on semiconductor devices in Japan. Many companies showed interest in his research and its results, which made him think of a possible chance to put his expertise to good use in the industrial domain. He began contemplating on becoming an engineer. His inquisitive mind made room for the desire to use his knowledge with purposes.

His choice of SSS among many other possibilities was quite opportunistic: he was a postdoctoral researcher when an SSS employee who was a graduate of the same lab came to see him. He introduced him to the professional personnel recruitment opportunity at Sony. From the interview, he had a good impression of the company, that it had many lively and smiley people. It was a major part of his decision, according to Minari. “They seemed very eager to listen to me when I was talking about technology. I thought I would enjoy doing research with people like them.”

Always keeping a positive attitude to gaining knowledge
no matter what the subject may be

Soon after joining SSS, Minari was seconded to Belgium. His mission was to work in a semiconductor joint development consortium of which SSS was a member and, as a member of an advanced logic device development project, acquire the latest expertise on the device. However, the consortium was represented by many companies, and there was already a set of research themes provided. Not all of them were of interest to SSS. Minari formulated a research theme so that it was aligned with the consortium’s research orientation and relevant to SSS's interest. He persuaded the members that his research was useful to gain the knowledge the consortium was seeking.

While successfully negotiating his research theme and working in the consortium, Minari was not exactly taking everything in his stride. He was undertaking the development of transistor (logic device) using compound semiconductor materials, but the crystal alignment did not come out as neatly as it would with silicon. Finally, the time came for him to go back to Japan without having achieved expected device performance. He had to return with some successful experience but also with disappointment. However, the knowledge of compound semiconductor materials he gained from this experience came in useful later when he developed an image sensor using compound semiconductors.

His approach to work, according to Minari, is “to follow my curiosity and try to understand something I don’t know about or understand positively.” For his rich knowledge on the white pixel problem of image sensors, his schedule is filled with requests for discussions by SSS’s colleagues. “I come across many things I don’t know in my work. Sometimes, I feel I’m out of my depth. But I always remind myself that, as a researcher, I should stay truthful to my curiosity and try to gain new knowledge, then I will be able to understand more and better.” His words eloquently show that he always has the desire to know as he did as a child, and it is always important for him that his knowledge is “useful” to the world.

He maintains friendships with people he worked with in Belgium as well as from his university. He says, “There are many excellent talents out there. I believe that it is possible to make society a better place if we can collaborate more closely.” He is all hopeful to thrive in future research activities and make a good environment for conducting them. We also look forward to seeing his future successes.