26

New Materials Selection Has Been an Trend

Materials selection has become an increasingly serious challenge for the entire semiconductor supply chain.

At present,new material supply is a kind of market demand, and the new material content in the chip is increasing along with higher-required density and functionality. But the trend to solve the material problem begins sweeping the market even though not all products are in their newest process nodes. What we need in the market is a kind of long-lasting device. But the trend to solve the material problem begins sweeping the market even though not all products are in their newest process nodes.

Dominic Miranda, the Business Development Manager at Brewer Science, said, ‘One of the misconceptions about AI is that it's all about training and reasoning, but the data that input into these systems also plays an important role. There are many sensors used in networks, factories, and cities, So the relevant data involved and the transfer speed between them will become a problem worth considering, making the sensor's response speed to various data also become a top priority. What we need to know is that the material has a big impact on how quickly the sensor reacts to the stimulus.’

Miranda also said: 'The way in which materials functionalize these devices will affect the characteristics of the sensor and the sensing mode. The more complex the device is , the more you need to deal with noise which  may be ambient noise, such as noise from machine operation. We find that the market is moving toward two directions. You can remove the sensor from the shelf and apply it to the system, or you can use a customized design to get a clearer signal.'

Noise is a growing problem, especially on advanced nodes which have much smaller tolerances than the old ones. Nowadays thinner (<10/7nm) gate oxide layers and higher densities has made the noise caused by power supply, electromagnetic interference and heat an increasingly intractable problem, even in digital circuits.

Marching into 2D market

One of the facts we face in shrinking devices is that, like most materials, silicon itself is three-dimensional. Even though the silicon layer has only the thickness of one atom, it still contains hangling bonds that extend from the surface. These bonds need to be passivated to avoid undesirable interactions and add surface roughness that leads to carrier scattering and degrading mobility. In contrast, there are no out-of-plane bonds in two-dimensional semiconductors. The monoatomic layer is structurally 'complete', self-passivating, thereby reducing or eliminating Short-channel effects.

Nonetheless, developing these promising structural features in manufacturable devices remains a challenge. Graphene is the first two-dimensional semiconductor which is without a band gap. At a typical operating temperature, black phosphorus is unstable, therefore, current Researches have focused on transitioning metal disulfides such as molybdenum disulfide, WS2 and WSE2. At the Spring Meeting of the Materials Research Association in April and the IEEE Electronics Conference in December last year, several papers examined the physics and materials science of these compounds.

The first challenge in business applications is how to make two-dimensional materials simply. A large number of research samples can be obtained by stripping - separating the graphene by using tape to pull the layer out of the bulk graphite, but manufacturing precision and quality requirements desire for a more controllable method.

While CVD is an obvious choice for thin layer deposition, two-dimensional CVD appears to be a more complex material than it seems at first glance. For example, a 2D material can be placed on a substrate but not bonded to it. Therefore, growing a 2D semiconductor typically involves etching or ablating a nucleation layer beneath the semiconductor monolayer to separate it. Xiangfeng Duan, a professor of chemistry and biochemistry at the University of California, Los Angeles, explained in a presentation that the multilayer stack required for the device needs higher chemical compatibility with the substrate, the stacked components, and the process gases. Process conditions suitable for a single layer may lead to chemical or thermal degradation of the next layer.

However, when the two-dimensional semiconductor heterostructure was successfully deposited, the results were obvious. Duan's team produced horizontal WSE2 / WS2 p-n diodes with atomic thickness, which interacted along a quantum wire. For vertical stacking, they are investigating the insertion of electrically passive materials (electronic passive materials) into existing stacks. This approach may separate the monolayers from each other without increasing the complexity after the substrate is removed .

But it is not enough to separate a single layer. In two-dimensional materials, defects can completely prevent the movement of the carrier: they cannot leave the plane to find another path. Ali Javey, a professor of electrical engineering and computer science at the University of California at Berkeley, pointed out in his research report relating to optoelectronic devices that the defects existed Is the non-radiative recombination center. Therefore, the quantum yield Is the reasonable measure of the defect level.

Once high quality semiconductor materials are obtained, low-resistance contacts become the next challenge. A contact with good electron mobility may block holes and vice versa. Javey's team showed photo-emissive devices that used alternating current to provide the first hole, then the electrons, which were recombined in the shed MoS2 layer to emit light. 

In a study published on IEDM, PhD student Xuejun Xie and colleagues at the University of California, Santa Barbara described the use of photosensitive MoS2 FETs in artificial retinal devices. This device has potential application value for neuromorphological image recognition. Although memristive crossbar arrays are often proposed for use as artificial synapses, they cannot directly 'see' the image.

Capturing image information and writing it to a crossbar array is a potentially significant bottleneck that may be mitigated by combining image sensing and analysis in a single device. To this end, the Santa Barbara team created a metal MoS2 quantum dot array using electron beam mode on the semiconductor MoS2 channel. Quantum dots attract a large amount of electrons from the conduction band of the semiconductor, moving the Fermi level to the valence band. Besides, the resistance is increased as the holes are limited. However, as the current flows, the moving electrons are recombined with the holes, causing the resistance to Decline with time. There are more carriers when the light source is on, so the device 'detects' and 'remembers' the bright portion of the image.

Flexible materials

One of the challenges of materials engineering today is to innovate standard chip formats. There is a new set of flexible hybrid electronics, including film temperature sensors, electronic inks, etc., each with its own unique features and defects. And they are making these materials run on schedule under a new and sometimes unexpected set of operating conditions, which will be much more difficult.

Norman Chang, chief technologist at the semiconductor division of ANSYS, said, ‘There are a range of sensors that can detect glucose, pH, humidity and temperature. The problem is that we are using components with different thermal gradient, which can have an impact on performance. We are actually studying 3D geometry input, which requires co-simulation of flexible bases and packages because they affect the electrical performance of these devices. All of this must be simulated together. If you study printed RF, you must know that the millimeter wave behaves differently in different area.'

One of the new methods being developed is called geometry wrapping, by which the circuit can be wrapped around any device, even can stretch across buildings. For example, the US Air Force Research Laboratory announced earlier this year that it was working with NextFlex to develop A flexible sensor systems for the Internet of Things (IoT) in military and commercial fields. The goal was to implement stretchable electronics that can withstand high G loads and temperatures.

Flexible sensors are also used in some applications such as water and environmental testing. Brewer Science's Miranda said, 'In water testing, the challenge is to design a sensor to ignore everything, and just consider what you want to test. Solid-state materials can't do it, but flexible sensors can. You may have seen a video on YouTube saying that some people heard Yanni, but some heard Laurel, all depending on the wavelength they heard. Actually, the difference comes from the wavelength they heard. But the materials can be used to detect what you want to detect and make sure that what you hear is what you should hear .'

Conclusion

The electronics industry is paying more attention to new materials. It can be applied to autonomous driving, artificial intelligence, 5G, and industrial and medical fields which have higher requirements for the flexibility, noise sensitivity, and signal throughput of chips used in electronic devices.

What is Materials Engineering?


You May Also Like: