PVA TePla (Germany)

Improved understanding of gas/liquid transport in electrochemical gas-evolving systems is increasingly demanded for optimizing device performance. However, high-resolution measurement techniques for in-situ imaging remain limited. This work demonstrates the use of volumetric scanning acoustic microscopy (SAM) for quantifying hydrogen bubble distribution in porous nickel electrodes in a customized alkaline water electrolysis cell. By using high-frequency focused ultrasound, SAM enables volumetric imaging with high spatial resolution in the range of tens of micrometers. This allows the distribution of gas bubbles within the complex 3D architecture of porous electrodes to be resolved. Digital image processing methods are used to segment and quantify the gas content in the electrode. Thus, non-destructive SAM imaging is demonstrated to be an accessible and scalable analytical tool for the quantitative investigation of bubble distribution in electrochemical environments. Here, a methodological foundation is established for future studies aimed at optimizing bubble dynamics and cell design under practically relevant operating conditions, ultimately contributing to higher electrolysis efficiencies. • Volumetric SAM imaging quantifies H2 bubbles in porous Ni electrodes. • Demonstrated applicability for both mesh and open cell foam electrode structures. • Provides 10's μm spatial resolution in opaque, technically relevant electrodes. • Non-destructive, scalable technique for gas-evolving electrochemical systems.

Abstract Traditional failure analysis (FA) methods face growing challenges due to the increasing miniaturization and complexity of modern semiconductor devices. In particular, the detection of cracks in the extreme low K layers of flip-chips, utilized in advanced packaging architectures at the back end of line (BEOL), is challenging. To improve detectability of these defects, we propose an AI-based image enhancement workflow, specifically tailored to improve scanning acoustic microscopy (SAM) images. By applying a Deep CNN with Skip Connections and Network in Network (DCSCN) model to lowresolution C-scan SAM images, we are able to improve perceptual quality of images. The approach boosts resolution dependent analysis like AI-based object detection and classification as well as decreases significantly SAM scanning time. Our work demonstrates that AI-based image enhancement can significantly improve FA workflows, offering a robust and efficient solution for modern semiconductor analysis.

Abstract 3D integration and miniaturization techniques get more widely used in conventional integrated circuits but also represent crucial ingredients for future quantum computing devices. This consolidates the need for efficiently detecting increasingly small defects on wafer size. Here we present a time-efficient and accurate way of measuring, localizing and statistically classifying defects down to the micrometer regime, utilizing a combination of scanning acoustic microscopy (SAM), You Only Look Once object detection, semantic segmentation and machine learning super-resolution (ML-SR) approaches. In particular, we test the capabilities of different ML-SR approaches to enable self-supervised quality enhancement of the measured image data. We reveal that the developed AI-powered workflow enhances time-efficiency by a factor of around 4x and 6x for the TSV and delamination analysis, respectively. Yet, the mentioned approach is not limited to SAM image data but presents a general way for speeding-up failure analysis in various fields.

Abstract Minor flaws are becoming extremely relevant as the complexity of the semiconductor package evolves. Scanning acoustic microscopy is one analytical tool for detecting flaws in such a complex package. Minor changes in the reflected signal that could indicate a fault can be lost during image reconstruction, despite the high sensitivity. Because of recent AI (Artificial Intelligence) advancements, more emphasis is being placed on developing AI-based algorithms for high precision-automated signal interpretation for failure detection. This paper presents a new deep learning model for classifying ultrasound signals based on the ResNet architecture with 1D convolution layers. The developed model was validated on two test case scenarios. One use case was the detection of voids in the die attach, the other the detection of cracks below bumps in Flip-chip samples. The model was trained to classify signals into different classes. Even with a small dataset, experiment results confirmed that the model predicts with a 98 percent accuracy. This type of signal-based model could be extremely useful in situations where obtaining large amounts of labeled image data is difficult. Through this work we propose an intelligent signal classification methodology to automate high volume failure analysis in semiconductor devices.

The main aim of this paper is to provide the feasibility of non-destructive testing (NDT) method, such as scanning acoustic microscopy (SAM), for damage detection in ultrasound (US) probes for medical imaging during the manufacturing process. In a highly competitive and demanding electronics and biomedical market, reliable non-destructive methods for quality control and failure analysis of electronic components within multi-layered structures are strongly required. Any robust non-destructive method should be capable of dealing with the complexity of miniaturized assemblies, such as the acoustic stack of ultrasonic transducers. In this work, the application of SAM in an industrial scenario was studied for 24 samples of a phased array probe, in order to investigate potential internal integrity, to detect damages, and to assess the compliance of high-demanding quality requirements. Delamination, non-homogeneous layers with micron-thickness, and entrapped air bubbles (blisters) in the bulk of US probe acoustic stacks were detected and studied. Analysis of 2D images and defects visualization by means of ultrasound-based NDT method were compared with electroacoustic characterization (also following as pulse-echo test) of the US probe through an ad-hoc measurement system. SAM becomes very useful for defect detection in multilayered structures with a thickness of some microns by assuring low time-consuming (a limit for other NDT techniques) and quantitative analyses based on measurements. The study provides a tangible contribution and identifies an advantage for manufacturers of ultrasound probes that are oriented toward continuous improvement devoted to the process capability, product quality, and in-process inspection.

3D-integration is one of the most challenging approaches addressed by researchers in the field of microelectronics in the recent years. With the intension on integrating different components in three dimensions in one device performance and functionality will increase dramatically by reducing the devices footprint. A major challenge in 3D-integration concepts is the electrical interconnection of the stacked individual components. These interconnecting technologies employ micro bumping, temporary wafer bonding, wafer thinning and through silicon vias (TSVs). The increasing complexity and the miniaturization result in new requirements on testing, diagnostics, failure analysis and metrology techniques, methods and tools. Scanning acoustic microscopy (SAM) is a powerful tool for non-destructively inspecting internal structures and features. It employs elastic waves that can be focused and used for imaging and quantitative analyses. However, at conventionally used frequencies (5 MHz – 250 MHz) imaging resolution compromises the application on devices and technologies required in 3D-integration approaches. The current paper reports on the use of acoustic GHz-microscopy for the inspection, defect localization and its ability for identification of abnormalities in through silicon vias. Investigated were artificial and real defects in the TSV-fillings (voids) and the condition of the TSV-walls (rim-delaminations). Acoustic frequencies used in the current work ranged from 400 MHz up to 1.2 GHz allowing for imaging resolutions in the 1 µm - regime. However, highly focused acoustical lenses as employed here require large numerical apertures which inevitably result in a very complex wave propagation and acoustic field inside a solid sample. To improve the understanding and interpretation acoustic intensity fields have been simulated numerically. Results obtained by acoustic GHz-microscopy have been evaluated complimentarily by FIB-cross-sectioning and SEM imaging which gave a valuable insight into the abilities for acoustic TSV-inspection by GHz-SAM.

ABSTRACT In the field of failure analysis inspection methods operating non-destructively are essential for detecting and localizing irregularities and defects applied prior to destructive high-resolution microstructural analyses. Considering the ongoing trend of continuously decreasing structure sizes while extending their complexity to the third dimension inspection methods are challenged to meet new requirements. Recent developments in the field of scanning acoustic microscopy were focused on a dramatically increased resolution and improved sensitivity by extending the applicable frequency range into the GHz band. Along with this trend adapted ultrasonic transducers with numerical apertures are required that both operate at those frequencies and allow for lateral resolutions in the one-micrometer range and below. The presented results show that high resolution GHz-SAM is capable to precisely resolve TSV structures with 3 μm diameter. Furthermore, it was shown that GHz-SAM can be used as non-destructive high-resolution analytical technique for void detection within the TSV structures.

In this paper the performance of multi-narrow-band spectral analysis was evaluated concerning defect detection in microelectronic components with flip-chip contacts. Today, flip-chip technology is widely applied for interconnecting silicon dies to a substrate within high-end semiconductor packaging technologies. The integrity of the bump solder interconnection is of major concern for the reliability in this technology. Non-destructive defect localization and analysis of the flip-chip interconnections operating in a semi-automated mode is strongly desired. Scanning acoustic microscopy (SAM) combined with subsequent signal analysis has high potential for non-destructive localization of defective flip-chip interconnects. Analyzing multiple narrow spectral bands of signals acquired by a scanning acoustic microscope enabled the identification and localization of defective flip-chip interconnects. In the current study a 180 MHz transducer with 8 mm focal length was employed for acoustic data acquisition by SAM. Those data were then analyzed off-line by discrete Fourier transformation, chirp z-transform and cosine transform using custom made MATLAB software. Through multi-narrow band spectral analysis, defective flip-chip interconnects that have not been revealed by standard acoustical imaging methods have been detected successfully. Acoustically found defects have been confirmed by subsequent FIB-cross sectioning and SEM imaging. The high resolution SEM imaging revealed complete and partial delamination at the interface between the die and the bump.

Abstract New semiconductor chip technologies and technologies for 3D integration require information’s of packaging and interface defects in 3 dimensions, that means the lateral dimension of the defect and the location inside the device or package must be defined. In this paper, new methodical approaches for non destructive failure analysis on 3D integrated TSV samples are introduced. The concepts combine improved scanning acoustic microscopy (SAM) imaging hardware with unique software solutions for defect identification and quantitative analysis of mechanical properties using scanning acoustic investigations. In case of MEMS 3D integration, e.g. based on direct bonding, related interface defects must be investigated by SAM. With respect to 3D integration applications, the potential of recent SAM improvements applying specifically adapted hardware and custom-made signal processing algorithms will be discussed. Examples of SAM-based failure detection techniques for the application in 3D integration are demonstrated. New technologies are shown to improve the through put of fully wafer scanning using scanning acoustic microscopy. To improve the defect resolution, a new transducer design was developed to increase defect resolution and signal to noise for interface characterisation.

Abstract Die industrielle Züchtung von immer größeren und perfekteren Silizium (Si) ‐ Einkristallen für Anwendungen in der Mikroelektronik und in der Photovoltaik erfordert eine kontinuierliche Verbesserung der Prozessführung und der Züchtungstechnologie.Parallel dazu ist eine stetige Anpassung und Optimierung der Anlagentechnik hinsichtlich der Zuverlässigkeit, Prozessflexibilität und Dimensionierung erforderlich. Im ersten Teil dieser Artikelfolge werden die Grundlagen der industriellen Silizium‐Kristallzüchtung und die daraus folgenden Anforderungen an die Si‐Prozess‐ und Anlagen‐Technologien beschrieben. Die fortlaufend steigenden Anforderungen an die Leistungsfähigkeit und Komplexität der elektronischen Schaltkreise (Chips) gemäß dem Moore'schen Gesetz haben zur Folge, dass auch die Anforderungen an die Perfektion und Dimensionen der einkristallinen Si‐Wafer und Si‐Kristalle stetig steigen. Nach Einführung der 300 mm‐Si‐Wafer‐Generation in den letzten Jahren wird daher bereits die nächste Si‐Wafer‐Generation (450 mm) diskutiert.Die technologischen und ökonomischen Auswirkungen dieser stetig steigenden Anforderungen an die erforderliche Anlagen‐Technologien werden in den Folge‐Artikeln anhand von aktuellen Si‐CZ‐ KristallzüchtungsAnlagen sowie von neuen AnlagenKonzeptionen vorgestellt und diskutiert.