Devices

Sustainable Semiconductor Manufacturing through Hyperspectral Imaging

© DIVE imaging systems Gmbh

In the joint research project NEST (New Screening Tool for Efficient Semiconductor Manufacturing), Fraunhofer IPMS, Fraunhofer IZM, and DIVE Imaging Systems GmbH are working within the GreenICT@FMD competence center to make semiconductor production more efficient, cost-effective, and sustainable. The focus is on introducing hyperspectral imaging as an advanced screening tool for wafer inspection and process optimization.

Semiconductor production involves up to 1,500 complex process steps, including etching, deposition, and lithography, which require extremely rigorous quality control. Up to 50% of these steps are dedicated to metrology, and thousands of control wafers are used monthly. This leads to high resource consumption, increased CO₂ emissions, and added cost.

The NEST project revealed that targeted inspection technologies can significantly reduce the environmental footprint of chip manufacturing. In a 28 nm node with 25,000 wafer starts per month, the deployment of hyperspectral imaging could reduce control wafer usage by at least 25%, cutting CO₂ emissions by over 118,000 kg/month. Additional benefits include reduced water and chemical usage, minimized tool load, and increased wafer yield due to early defect detection.

Cleanroom-Ready Hyperspectral Inspection by DIVE and Fraunhofer IPMS

The DIVE VEpioneer® inspection system is the first of its kind to operate reliably in industrial cleanroom environments. It combines optical spectroscopy and AI-based image analysis to detect surface deviations and contamination within just 20 seconds, non-destructively and without halting production.

Fraunhofer IPMS provides critical expertise in cleanroom qualification and optical design, enabling the system’s integration into semiconductor fabs. Following the project's conclusion, the VEpioneer® remains in operation at the Fraunhofer IPMS Center Nanoelectronic Technologies (CNT) for continued evaluation and joint development efforts, including automation of wafer handling and real-time data integration.

Energy-Efficient Semiconductor Technologies for Digital, Analog, and RF Applications

CEA-Leti and Fraunhofer extend their collaboration in the FAMES pilot line, a pioneering project to promote semiconductor technologies in Europe. This initiative supports the EU Chip Act, which aims to strengthen the EU's technological sovereignty.

The FAMES pilot line is dedicated to advancing next-generation energy-efficient semiconductor technologies for digital, analog, and high-frequency (RF) applications. The project focuses on the development and scaling of five key technologies:

  • FD-SOI platforms at 10 nm and 7 nm nodes to improve power efficiency and performance in advanced digital circuits
  • A range of embedded non-volatile memory (eNVM) technologies, including OxRAM, FeRAM, MRAM, and FeFETs, for data retention in ultra-low-power systems
  • High-frequency components such as RF switches, filters, and capacitors, supporting 5G, IoT, and edge computing applications
  • Two innovative 3D integration approaches:
    • Heterogeneous integration for combining different chiplets and materials
    • Sequential integration to stack transistors layer-by-layer for higher density
  • Integration of miniaturized inductors to enable compact and efficient DC-DC converters in power management ICs (PMICs)

These developments support a sustainable and competitive European semiconductor industry, enabling compact, high-performance, and energy-efficient electronics for future technologies.

Sustainable Sensor Technology: Resource-Efficient ISFET Development for pH Sensing

Close-up of a circuit board with ion-sensitive field-effect transistors (ISFETs)
© Fraunhofer IPMS
Close-up of a circuit board with ion-sensitive field-effect transistors (ISFETs)
Control box for Nb2O5 ISFETs and an AG/AgCl reference electrode with and without housing including the connections for analog readout.
© Fraunhofer IPMS
Control box for Nb2O5 ISFETs and an AG/AgCl reference electrode with and without housing including the connections for analog readout.

To promote sustainability in sensor technology, current research focuses on developing resource-saving manufacturing processes for ISFET-based pH sensors. Traditional materials like tantalum, though effective, are costly and critical in supply. The goal is to replace them with low-cost, energy-efficient alternatives offering comparable performance.

Key innovations 

  • Bonding-free ISFET integration to eliminate rare metals like palladium and silver alloys
  • Wafer-level sensor functionalization via micro-precision dispensing to reduce material waste
  • Avoidance of flat coating processes, which cause over 99% material loss
  • Reduced use of reactive and toxic substances, improving occupational safety and environmental compatibility

These advancements significantly lower production costs, minimize environmental impact, and support the development of next-generation eco-friendly semiconductor sensors.

Compact Evaluation Kit for Nb₂O₅-Based ISFETs

Fraunhofer IPMS offers a plug-and-play evaluation kit for Nb₂O₅-based ISFETs for pH sensing. The USB-C powered device enables easy testing with integrated reference electrode support and temperature compensation. A low-power analog variant is also available for battery-operated systems. Both options are compact, user-friendly, and optimized for accurate and efficient pH measurement.

Neuromorphic Computing: Energy-Efficient and Sustainable AI Hardware

Neuromorphic computing offers a breakthrough for energy-efficient and sustainable electronics. By mimicking the adaptive processes of the human brain, this technology enables ultra-low-power data processing, which is ideal for smart sensors, edge AI, and decentralized systems. It is particularly well-suited for future-oriented applications in Industry 4.0, smart cities, and autonomous mobility, where performance and efficiency must go hand in hand.

Fraunhofer IPMS: Experts in Neuromorphic Computing

Fraunhofer IPMS is a leading expert in neuromorphic hardware and materials. We are developing specialized materials, technologies and complete hardware solutions that are particularly optimized for use in the edge area - for example in intelligent sensors and decentralized systems that can process data and make decisions independently.


👉 Learn more by following the link in the blue box.