Ultra-thin two-dimensional materials with atomic thickness can not only manufacture microprocessors, but also advance the innovation of traditional microprocessors, making them widely used in flexible electronics. Recently, researchers from the Vienna University of Technology in Vienna and the EU's graphene flagship project have made breakthroughs in this field, and are expected to further promote the development of applications such as the Internet of Things and intelligent hardware.
Technical keyword
2D materials, microprocessors, flexible electronics
Background introduction
Two-dimensional material
A few days ago, John was in "Heavy: Scientists use 2D nanomaterial inkjet printing transistors! In the article, the invention of the use of two-dimensional materials (graphene, tungsten diselide, boron nitride) inkjet printing transistors by Irish scientists is introduced. At the same time, the two-dimensional materials are also introduced. Let us review again:
"Two-dimensional material" refers to a material in which electrons can move freely (planar motion) on a non-nano scale (1-100 nm) of two dimensions, such as graphene, boron nitride, transition metal compounds (molybdenum disulfide, Tungsten disulfide, tungsten disilicide, black phosphorus, and the like.
Two-dimensional materials, usually composed of one or several layers of atoms, our previous focus on graphene is a famous two-dimensional material. In addition, some materials similar to graphene, such as transition metal disulfide compounds, are also two-dimensional materials. They are not only small in size, light in weight, soft, but most importantly have excellent semiconductor characteristics, and are very suitable for flexible electronic devices. .
microprocessor
"Microprocessor" is the core device of the contemporary electronics industry. Whether it is consumer electronics such as smart watches, smart phones, smart home appliances, or high-tech products such as supercomputers, automobile engine controls, CNC machine tools, and missile precision guidance, they are inseparable from the role of microprocessors.
A microprocessor, generally consisting of one or several large-scale integrated circuits, can read and execute instructions, exchange data with external memory and logic components, and is the core operation control part of a microcomputer.
Today, the materials used in microprocessor manufacturing are essentially all silicon. The bottleneck of silicon material mainly has the following two aspects:
Performance bottleneck, as I was in Moore's Law, is it survival or destruction? As stated in the article, the performance of semiconductor chips for silicon materials is slowing down and close to physical limits.
Without flexibility, silicon materials cannot be used in the field of flexible electronics.
Innovation exploration
Dr. Thomas Mueller of the Photonics Institute at the Technical University of Vienna has been working on two-dimensional materials. He believes that two-dimensional materials are ideal candidates for the future manufacture of microprocessors and other integrated circuits. Molybdenum disulfide (MoS2), composed of molybdenum atoms and sulfur atoms, has a thickness of only three atoms, which is such a two-dimensional material.
Therefore, he led the research team of the Technical University of Vienna and the researchers of the EU graphene flagship project to create a transistor consisting of the two-dimensional material "molybdenum disulfide" MoS2. 115 such transistors constitute a new type of microprocessor. At present, this kind of microprocessor can perform one-bit logic operation, and it is expected to expand to multi-bit operation in the future.
Key technology
Microprocessor architecture
Illustration:
(a) The block diagram shows the arithmetic logic unit (ALU), accumulator (AC), control unit (CU), instruction register (IR), output register (OR) and program counter (PC) of the two inputs A and B. The enable signals (EA and EO) and the operation selection code (A/O) provided to the respective subunits by the CU. The clock signal generator and memory are implemented off-chip.
(b) Timing diagram of N instruction cycles. In the sequence of fetching instructions, the contents of the memory are loaded into the IR, and the address stored in the PC is incremented. In the sequence of instruction execution, the commands stored in the IR are executed.
(c) Microprocessor instruction set: NOP is a non-operational instruction; LDA instruction transfers data from memory to AC; AND and OR instructions perform logic operations.
Characterization of MoS2 transistors and inverters
Illustration:
(a) Inverter circuit (top) using a front gate technology and a single MoS2 transistor (bottom) schematic
(b) Loading (W/L=45/2) and pull-down (W/L=7/5) transistor conversion characteristics
(c) Output characteristics of gate voltage between 1V and 5 V
(d) NMOS inverter circuit schematic
(e) The output voltage VOUT of the inverter is illustrated given the input voltage VIN. The blue symbol represents the load curve and the red line is the output characteristic of the pull-down transistor. The intersection of the two curves determines VOUT.
(f) The solid line represents the voltage transfer characteristic measured by the inverter. Through the mirror curve (dashed line), a butterfly-like image is obtained from which the NM can be extracted by the largest square embedded in the shaded area of ​​the gray.
Implementation of two-dimensional semiconductor devices
Illustration:
(a) A microscopic image of the microprocessor. The two layers of metal are presented in different colors and are connected by via holes. All subunits have metal test pads for testing. The marked pads are used to connect the device to external devices (memory, clock generator, power supply, output), and the other parts are wired together to achieve internal connections. The scale is 50 microns. (b) D-Latch and (c) ALU circuit schematics
Device operation
Illustration:
(a) The waveform measured on the chip by running the sample program. The CLK1, CLK2, and CKL3 signals are generated externally to the chip. Each instruction takes three clock cycles and 1/TCLK is the clock frequency. The CLK2 pulse is short enough to trigger a sensitive PC input, avoiding the use of more complex master-slave designs. A0 is the address provided by the PC. OP0, OP1 and D0 represent signals from the memory, the first two are instructions and the latter one is data. A and B are the input signals of the ALU, and OUT is the output signal. TCLK = 500 ms.
(b) The results of some of the calculations. In order to run a longer program, a 4-bit PC is implemented externally. The meaning of the curve is the same as in a except that A0 here represents a 4-bit address signal (A0-A3). In addition, this device can provide the expected values ​​as shown by the numbers at the top.
Innovative value
So far, the MoS2 microprocessor is one of the most advanced circuits made of two-dimensional materials. Of course, John also introduced "a revolutionary breakthrough in the semiconductor field: the birth of the smallest 1 nm transistor in history", which also uses molybdenum disulfide.
Tests that run simple programs show that the microprocessor has excellent signal quality, low power consumption, and gets the correct results.
This ultra-thin MoS2 transistor is small in size and flexible, and is beneficial for manufacturing into flexible electronic devices such as wearable devices and smart hardware, and is widely used in the field of Internet of Things.
In this regard, Dr. Thomas Mueller believes that:
“Overall, becoming a soft material will bring new applications. One possibility is to combine these processor circuits with light emitters made of MoS2 to create flexible display devices and electronic paper, or They are integrated into the logic of the smart sensor."
face the challenge
Modern microprocessors, typically made of silicon, have millions of transistors on a single chip, so the device with only 115 transistors made of MoS2 is quite simple. However, Stefan Wachter, a Ph.D. student in Dr. Mueller's research group, believes that:
“Although this is inferior to current silicon-based industry standards, it has become an important breakthrough in research in this area. Now we have proof of concept, and in principle, there is no reason to see the future. Progress."
This chip just shows the early results of this new technology, we can believe that it will have more results in the future. For the team's next plan, Mueller explained:
"Our goal is to achieve larger circuits that can do more with useful operations. We want to complete a full eight-bit design, or more bits, on a smaller chip size."
But because of design and manufacturing, this goal will be challenged, so Stefan Wachter says:
"Adding more bits will of course make everything more complicated. For example, adding just one bit will double the complexity of the circuit."
Future prospects
Two-dimensional materials such as molybdenum disulfide are expected to be a substitute for silicon. However, for making more complex circuits, these circuits require thousands or even millions of transistors, and it is not possible to rely on this new technology for the time being.
After all, the process of producing two-dimensional materials and further processing two-dimensional materials is still in its infancy. Therefore, it is currently only a complementary technology to silicon semiconductor technology.
In this regard, Dr. Mueller explained:
"Our circuits are more or less handcrafted in the lab, and such complex designs are clearly beyond our capabilities. Each transistor must operate as planned so that the processor can work as a whole. â€
In addition, improving the multi-level design process in the future will be an important step in developing a high-volume production solution for the MoS2 microprocessor. Because of all the factors, the transmission of large areas, double-layer MoS2 to the wafer, will bring a high failure rate. Therefore, Dmitry Polyushkin of the Technical University of Vienna believes that:
“Our solution is to raise the process to a point where we can reliably produce chips through tens of thousands of transistors. For example, growing directly on the chip will avoid the transfer process. This will result in higher yields, We can produce more complex circuits."
Researchers believe that this technology will have a variety of new industrial applications in the next few years. Flexible electronics are an example, such as medical sensors and flexible displays. Because two-dimensional materials have stronger mechanical flexibility than traditional silicon materials.
Shenzhen Esun Herb Co.,Ltd. , https://www.szyoutai-tech.com