Last summer, the semiconductor industry reached a significant milestone: one of the world’s top-tier fabs had begun production of the first 3D NAND chip with more than 200 layers. The announcement was significant but not a shock. Several other fabs had been progressing toward breaking the 200-layer barrier, so reaching the milestone was not a matter of if but when.
As significant as this advance is, the high-volume manufacturing challenges of producing high-aspect ratio (HAR) 3D NAND chips are considerable. One challenge is the ability to measure the tungsten (W) recess to the bottom of a 3D NAND device following the replacement gate process. Presently, there is no in-line process control solution that can accomplish this. The reason for this is known: beyond just a few layers in the stack, the W recess becomes opaque in the ultraviolet/visible/ near-infrared region, the realm of many OCD systems, after just a few layers in the HAR stack. Additionally, increased wordline slit pitch scaling further reduces the already minimal optical signal from the top of the 3D NAND structure to the bottom.
For decades, Moore’s Law has been a way to measure performance gains in the semiconductor industry, but the ability to double the density of transistors on a chip every twoyears is becoming increasingly challenging. With scaling reaching its limit, manufacturers are looking to advanced packaging innovations to extend the performance gains that the industry, and the world at large, have grown to depend on. Cu-to-Cu hybrid bonding is one way the industry is looking to extend ever-increasing I/O density and faster connections, all while using less energy.
Abstract
Ultrafast pump-probe measurements are used to characterize various samples, such as biological cells, bulk, and thin-film structures. However, typical implementations of the pump-probe apparatus are either slow or complex and costly hindering wide deployment. Here we combine a single-cavity dual-comb laser with a simple experimental setup to obtain pump-probe measurements with ultra-high sensitivity, fast acquisition, and high timing precision over long optical delay scan ranges of 12.5 ns that would correspond to a mechanical delay of about 3.75 m. We employ digital signal balancing to obtain shot-noise-limited detection compatible with pump-probe microscopy deployment. Here we demonstrate ultrafast photoacoustics for thin-film sample characterization. We measured a tungsten layer thickness of (700 ± 4) Å with shot-noise-limited detection. Such single-cavity dual-comb lasers can be used for any pump-probe measurements and are especially well-suited for ultrafast photoacoustic studies such as involving ultrasonic echoes, Brillouin oscillations, surface acoustic waves and thermal dynamics.
A mother steps on the brakes, bringing her car to a stop as she drops her kids off for dance lessons. At the time, she doesn’t notice anything wrong, but when she takes her car in for its regular service appointment, the mechanic conducts a diagnostic check and discovers that the primary brake system on the car had failed because of a faulty braking controller without anyone realizing it. Fortunately, the car was able to stop successfully due to the vehicle’s system redundancies, and the dealer’s diagnostic test confirms that since that first chip failure, another one has not occurred. The braking systems are behaving normally.
Following that, the dealership sends the information about the braking failure to the manufacturer, where an analyst notes that over the last 60 days, and around the country, six other brake failures traced back to the same controller system have been reported for the same make and model. In each of these situations, the backup system successfully brought each car to a complete stop. And, as in the case with the mother who dropped her kids off at dance class, the analyst looks at the reporting samples for these six other failures and determines that each is isolated and non-recurring.
As logic and memory semiconductor devices approach the limits of Moore’s Law, the requirements for accuracy in layer transfer become increasingly stringent. One leading silicon wafer manufacturer estimates that 50% of epitaxial wafer supply for logic will be on nodes equal to or less than 7nm. This is up approximately 30% from earlier in the decade.
To meet the demands of extreme ultraviolet (EUV) lithography, these leading-edge epi-deposited substrates have tighter specifications than previous substrates. Consider 3-5nm logic nodes: the image placement requirement can be as low as 3nm [1].
With the more stringent requirements of EUV lithography in mind, wafer makers are searching for new solutions, such as those addressing the primary reason for inaccuracies in image transfer: macro defects.
Over the past ten years, primarily driven by a tremendous expansion in the availability of data and computing power, artificial intelligence (AI) and machine-learning (ML) technologies have found their way into many different areas and have changed our way of life and our ability to solve problems. Today, artificial intelligence and machine learning are being used to refine online search results, facilitate online shopping, customize advertising, tailor online news feeds and guide self-driving cars. The future that so many have dreamed of is just over the horizon, if not happening right now.
The term artificial intelligence was first introduced in the 1950s and used famously by Alan Turing. The noted mathematician and the creator of the so-called Turing Test believed that one day machines would be able to imitate human beings by doing intelligent things, whether those intelligent things meant playing chess or having a conversation. Machine learning is a subset of AI. Machine learning allows for the automation of learning based on an evaluation of past results against specified criteria. Deep learning (DL) is a subset of machine learning (FIGURE 1). With deep learning, a multi-layered learning hierarchy in which the output of each layer serves as the input for the next layer is employed.
Currently, the semiconductor manufacturing industry uses artificial intelligence and machine learning to take data and autonomously learn from that data. With the additional data, AI and ML can be used to quickly discover patterns and determine correlations in various applications, most notably those applications involving metrology and inspection, whether in the front-end of the manufacturing process or in the back-end. These applications may include AI-based spatial pattern recognition (SPR) systems for inline wafer monitoring [2], automatic defect classification (ADC) systems with machine-learning models and machine learning-based optical critical dimension (OCD) metrology systems [1][7].