Launch of the automated TEM, to be inevitable tool for inspecting semiconductors
With the advancing miniaturization of semiconductors, inspection by a SEM (scanning electron microscope) as conducted in the past, began failing to satisfy the customers’ requirements. Thus, the TEMs (transmission electron microscope) turn came. With a conventional instrument, speedy acquisition of data was difficult since the operation was mostly manual. JEOL Ltd. launched a new model enabling an automated sample image and data acquisitions. It is expected to become a common tool for inspecting and analyzing semiconductors.
Product incorporating the best technologies of JEOL
Accompanied by the greatly advancing miniaturization of semiconductors, the resolution of SEMs, which used to be a major tool, is not satisfactory in testing semiconductors anymore. Instead, TEMs began to be used for this purpose. However, in order to fully make use of TEMs, an experienced engineer is needed for operation and analysis and it is time-consuming because of the manual operation. Thus, responding to the needs of the semiconductor industry, JEOL Ltd. developed a TEM for inspecting and analyzing semiconductors. The high throughput analytical electron microscope JEM-ACE200F (ACE200F) was launched at the end of 2018. The ACE200F provided automation of field adjustment and serial acquisition of data.
Shuji Kawai, leader of development project, briefs about the start of development.
“Along with further miniaturization of the semiconductor device structures, the process development has become much complicated and required for the atomic level analyses and data acquisitions on the volume samples. Being strongly approached by several major semiconductor makers, we built up the system while absorbing the requests from these makers. “
Noriaki Endo is in charge of application support and marketing support for this new ACE200F. Endo says, “We can call it a product with all of the best technologies we have built up for many years. In addition to the latest technologies both mechanically and electrically, this product employs the most advanced system.” says Endo.
Developed by a cross-sectional project
Kawai also says, “It does not merely employ the outcome from TEM development team”. “We had discussions with the manufacturing people many times and built the system so that stable manufacturing is possible. It was a cross-sectional project from technology development, design, application, manufacturing, services, and shipping department. And as a leader, I wrapped up the team. It was difficult to properly absorb various opinions into the product since various opinions came up. However, behind the opinions there are feelings of everyone who wants to make a better product and reduce troubles. These opinions will help brush up the product if well absorbed. “. The application department which Endo belongs to, has the role to connect the customer with JEOL.
“Since we know the customers’ needs best, we evaluate the product from the customers’ point of view, and convey the opinions resulting from the evaluation to the development team. Ultimately, we evaluate the product, make a demo, and obtain a PO with the collaboration with our technical team and sales team. Therefore, we engaged in this product from the initial stage of development up to discussion on software interface. There is no goal with completion of ACE200F. When semiconductor advances, the customers’ needs will change. It is our mission to let the development team move to respond to such a change. The new model became a product which we can offer to our customers with full confidence.”
Contributes to cost reduction in inspecting semiconductors
The semiconductor manufacturers used to use the TEM by labor intensive method for inspecting semiconductors. So, the automation will make great reduction of costs.
“How well our product can satisfy the customers’ requirements, it was very tough. It was required to obtain highly precise data, by automatic focusing with the observation area of tens of atoms level. In order to adjust the field of view, the sample stage needs to be moved frequently. But the sample stage needs to look motionless in order to take the image data. Therefore, we devised it to reduce movement. High speed image acquisition is another device. Thus, the instrument is made by combining several technologies. In order to achieve serial unattended and stable shooting, we needed to ask for wisdoms from various departments. “ recalls Kawai.
Endo tells about the advantage of automation. “It was also first time for JEOL to automate image acquisition of such a great amount of samples by using TEM for material observation. The data accuracy of TEM depends on the skill of the operator. So, automation provides improvement of speed as well as standardization of the outcome, which are other advantages. The operating engineer, who becomes free due to automation of TEM, can dedicate himself/herself to the analysis of data. Thus, the time required for a series of flow from image acquisition to analysis can be reduced.”
According to Kawai, “That we achieved serial observation of a 20 nanometer observation area allowing observation of the atom was the greatest joy for me”. Kawai says.
Endo claims that the supporting system is important for any of our TEM customers, including the ACE200F, since they will continue using the instrument at least for 10 years.
“In the past, some customers switched from our competitor’s to JEOL’s. Of course, the performance of instrument matters. But what I emphasize to customers is our decent support. In particular, for non-governmental companies it could often be the key of selection. With ACE200F, a closer relationship will be required, I guess.” says Endo.
AI to be utilized in the future
Regarding the development direction next to automation, Kawai says,” automation will lead to AI (artificial intelligence). In the future, I would like to take in the AI to be utilized for analysis. That means, the AI will judge whether the automatically taken data can be used for analysis or not.”
TEMs have long been used at academic scenes such as academic researches. With the automation achieved, use at manufacturing scenes is being expected.
“Automation of TEM has raised new needs. TEMs used to be a special instrument for basic researches. The new automation capability has brought the TEM into an inspection tool. In the future, TEM will become inevitable not only for material development, but also quality control by non-governmental companies. Further usability improvement needs to be considered.” says Endo on the future perspective.
SEMs and TEMs have made great contributions to the development of semiconductors. In the future, the JEM-ACE200F will promote further progress of semiconductors.