Recently, the 16th International Nanotechnology Industry Expo (CHInano) was held at the Suzhou International Expo Center. During this expo, Xinshiye launched the "Q&A @ CHInano" series of interviews, exploring product forms, yield rates, costs, orders, equipment, and ecosystem constraints behind nanotechnology roadmaps. This article features an exclusive interview with Luo Junsheng, Commercial Director of Zhongcai New Materials Research Institute.
Testing and analysis are the "eyes" of chip R&D, yet 90% of this third-party testing institution's clients are universities and research institutes. Serving 4,000 to 5,000 research groups annually, Luo Junsheng describes it as a business with "low average transaction value." Regarding domestic self-supply, he offers a sober assessment: The gap in equipment mainly lies in basic components, while there is virtually no difference between domestic and international software. What kind of business is micro/nano characterization for semiconductor materials and devices? With these questions in mind, Xinshiye interviewed Luo Junsheng.
The following interview content has been edited based on on-site recordings without altering the original meaning.
Q1: What exhibits did you bring to this year's CHInano? How is your company's business layout structured?
Luo Junsheng: We are a third-party testing service organization specializing in microscopic structural characterization of materials. Targeting the semiconductor and chip sectors, we primarily conduct Failure Analysis (FA), measurement of nano-level coatings and layers during manufacturing processes, compositional analysis, and reverse engineering. Our business model is similar to leading institutions in this field, focusing on FA and Material Analysis (MA). Currently, 90% of our clients come from universities and research institutes, while 10% are semiconductor companies.
Q2: What technical reserves have you accumulated in sample preparation? How do you ensure data reliability for sensitive materials?
Luo Junsheng: Sample preparation defines the boundaries, while equipment reveals the scales. We have extensive technical reserves in FIB processing, ion beam cross-section polishing (CP), and failure point localization. Over 80% of our business involves metal substrate and composite material samples; we have experience handling issues such as stress differences between different metals, phase transformations during processing, and dislocations. For sensitive samples like semiconductor thin films and quantum dots that are not resistant to irradiation and prone to thermal damage, we use low-dose imaging techniques such as iDPC (integrated Differential Phase Contrast) to ensure the samples retain their original morphological state, so that the images reflect the true appearance of the sample.
Q3: In the field of testing and analysis, how far has domestic supply progressed? What is the gap between hardware and software?
Luo Junsheng: On the equipment side, aberration-corrected electron microscopes, transmission electron microscopes, and scanning electron microscopes are still dominated by overseas brands like Thermo Fisher and Zeiss. We are also collaborating with domestic equipment manufacturers for trial use, but overall gaps remain—mainly stuck in basic components such as electron microscopes, vacuum systems, and electron beam systems. Foreign players have decades or even centuries of accumulation, whereas we are still in the reserve stage. The software side is different. Our company has 8 software engineers and has independently developed the MicroEagle data analysis toolset, which is open to scientific research users for free. It has been used over 100,000 times cumulatively, analyzing nearly 800,000 data points. At the software level, there is no significant difference between domestic and international standards.
Q4: How do the needs of scientific research clients differ from those of corporate clients?
Luo Junsheng: The differences are very obvious. Scientific research users seek breakthroughs; they need to characterize points that others haven't characterized before to publish in high-impact journals, requiring higher precision. Corporate clients need to solve practical problems in production and R&D, with high timeliness requirements, often demanding results within 24 or 48 hours.
The service methods also differ: For corporate clients, we help design comprehensive testing and analysis plans and streamline the ordering process as much as possible. For scientific research clients, we mostly listen to their characterization objectives and execute according to their proposed plans. Additionally, advanced characterization methods typically appear in scientific research applications 3 to 5 years earlier than in industrial applications. For example, techniques like Atom Probe Tomography (APT), 4D-STEM, and Precession Electron Diffraction (PED) only started being adopted by the industry in the last two years.
Q5: What considerations led to co-building the Chip Semiconductor Testing R&D Center with Zhejiang University South China Institute?
Luo Junsheng: It is mainly about equipment resources. High-end equipment is too expensive; a single aberration-corrected electron microscope costs tens of millions of CNY. If enterprises purchase them themselves, the utilization rate is low, and the cost-benefit ratio is unbalanced. Initially, we collaborate with universities and research institutes, using their equipment while our personnel operate and analyze. Only when the volume of high-end characterization business increases will we consider purchasing large-scale equipment ourselves. Our company was established 8 years ago, and most of our test engineers have over 10 years of experience. They focus on characterization analysis daily, and their capabilities are no worse than most university testing centers. The relationship between us and schools is complementary.
Q6: What new requirements do third-generation semiconductors and advanced packaging impose on micro/nano characterization?
Luo Junsheng: Third-generation semiconductors and advanced packaging involve a large number of functional and protective nano/micro coatings, along with various deposition processes and combinations, imposing higher demands on refined characterization. Many samples are not resistant to irradiation and are prone to thermal damage, requiring customized sample preparation and characterization schemes, making them harder to prepare and image than conventional samples. Analyzing and testing new materials like 2D materials, MOFs, and COFs is also highly difficult. However, most of these demands are currently still in the reverse engineering and domestic substitution stages, with small volumes. Detection demand will explode only after these products enter pilot production, mass production, and tape-out phases, where initial yields are low and there are many failure and defect samples.
Q7: Can you provide lightweight end-to-end services for startups and research teams?
Luo Junsheng: We provide one-stop solutions ranging from plan formulation, sample preparation, testing, to data analysis and interpretation. For corporate clients, we streamline the ordering and communication process; clients just need to state the problem they want to solve or what they want to prove, and we design the entire testing scheme and provide data support. For scientific research clients, we follow their instructions for execution. Clients do not need to worry about the characterization workflow and can focus their time on experimental design and sample preparation.
Q8: How do you guarantee data credibility? How is the laboratory quality management system arranged?
Luo Junsheng: We hold CMA and CNAS qualifications and have passed ISO9001 system certification. We implement full-process digital management: Samples are tagged with QR codes upon receipt, and every step and handler must scan the code, ensuring traceability of sample preparation methods, testing methods, and corresponding personnel. If the final delivery does not meet client expectations, we can trace back step by step to identify where the issue occurred and optimize accordingly.
Q9: What is the competitive landscape of the industry? Which links require the most industrial collaboration?
Luo Junsheng: Competition is fierce. Universities are acquiring more equipment, reducing spillover demand; capital and talent are also flowing into this industry. Overall, it is a deflationary market—if suppliers increase by10 companies, clients may only increase by 5%, leading to intense internal competition (involution). At expos, we exchange ideas with equipment manufacturers about latest product iterations and application strategies, and learn from the tactics of leading institutions. We do not engage in malicious competition or easily enter segments where others excel; instead, we deepen our expertise in areas we are good at to improve the client experience.
Q10: What are your future plans for testing categories and regional layout? What gains have you made from CHInano?
Luo Junsheng: Regionally, we target nationwide coverage with two bases in Guangzhou and Ningbo, serving a balanced distribution of clients. Domestic clients cover regions including South China, East China, North China, and Northeast China, primarily consisting of key universities and research institutes. Overseas clients include renowned universities and research institutes in Singapore, Europe, and South Korea. In terms of categories, we stick to doing what we are best at.
This is our fourth time participating in CHInano. The main purpose is to increase exposure and acquire customer leads. CHInano attracts a precise audience of R&D professionals, which aligns well with our target market.