Author: Zen and the Art of Computer Programming
1. Background introduction
1.1 What is chip manufacturing?
Chip manufacturing, that is, from parts to chip manufacturing, including materials, circuit design, packaging, testing, debugging, production and sales. Its main goal is to use materials, processes, methods and tools to research and develop new high-performance electronic devices and make them into products with certain specifications and performance requirements. In order to improve production efficiency, reduce costs, and ensure quality, attempts at chip manufacturing have been ongoing in the fields of computers, communications, and automation. However, at this stage, there are still few similar manufacturing companies at home and abroad, lacking common technical and management specifications, and there are also great uncertainties and defects in the manufacturing process. Therefore, how to establish a scientific, reasonable and standardized chip manufacturing system has become an important strategic task.
1.2 Why study physical vapor deposition?
The physical gas phase (gas phase for short) is a state of matter in vacuum. Because oxygen in the air dissolves in water, the water can form an air mass-like structure and condense into air masses. The human body serves as a container, filled with juice to form an organic air mass. This air mass is what we usually call climate. How can such a huge air mass be regulated by controlling the flow direction, resistance and temperature changes, and how can it be fixed at a specific position? All this is inseparable from physical vapor deposition and the movement of molecular beams. Physical vapor deposition and the motion of molecular beams have become a hot topic explored by my country's international academic community. According to the physical properties of the physical gas phase, researchers believe that the "physical gas phase" is a complex network with potential feedback mechanisms. These networks can adjust flow direction, resistance, temperature changes and material transport, affecting key parameters of atmospheric physical processes, and can be widely used in various fields. "Physical vapor deposition" can form closed ribbon air masses, rotating jet wind fields, vortex gravity fields, arranged turbulent waves, ocean currents, rainstorms, solar spectrum, etc. through physically carefully designed molecular beams. With a basic understanding of physical vapor phase, we can better understand and improve the application of physical vapor phase in today's global industrial chain.
1.3 Definition and role of molecular beam epitaxy
Molecular Beam Epitaxy (MPP) refers to the phenomenon that surface droplets composed of microscopic particles approach the surrounding tissue through the interaction of microbial molecules. From a structural point of view, the molecular diffusion layer is a space formed by particles, which is generally formed by the aqueous solution interaction of molecules adsorbed on the particles. Due to the autonomous replication of different molecules, the molecular diffusion layer forms a large network, causing the morphology of the diffusion layer to change, and even parallel diffusion of molecules on the order of thousands of times can occur. Therefore, the shape of the diffusion layer determines the range of movement of particles, and the energy exchange and thermal conductivity of particles between molecular diffusion layers will promote the movement of particles. Due to the subtle interaction between the molecular diffusion layer and the surface, the particles surrounded by the molecular diffusion layer gradually push to one side of the diffusion layer, and "molecular beam epitaxy" occurs when passing through the diffusion layer. MPP is closely related to the density of surface droplets and surface friction. In some cases, it even directly leads to surface destruction or damage. For example, in clinical applications, MPP can cause the development of cancer cell diseases and even lead to the occurrence of tumors.
2. Core concepts and connections
2.1 Vapor deposition and molecular beam epitaxy model
According to molecular diffusion mechanics, the molecular diffusion layer is composed of particles of different sizes with different spatial coordinate axes. Each particle has the characteristics of uneven spatial distribution and irregular dynamic evolution. Due to the interaction between different particles, a "physical network" is formed. From the perspective of the particles, the interacting particles form a "molecular beam" in space. As time goes by, the molecular beam will expand and become longer. , and eventually evolved to form a "gas phase" as a whole, with the gas phase ranging from billions to millions of microns in size, as shown in Figure 1. Figure 1 Molecular diffusion network of particles (red circles represent particles, blue represents air masses) For each molecular beam in the molecular diffusion layer, its width will gradually increase over time until its length reaches a specific value. It is called the Pores of Diffusion. Inside the molecular diffusion layer, there are a series of "adsorption zones" where particles can only be adsorbed together through interaction to create a chain reaction. The adsorption zone is surrounded by particles, and depending on their position, the particles interact with other particles attached to the area. The adsorbed particles will quickly move to their respective adsorption points, forming an obvious detour, called the "detour center". Similar circuitous behavior can also be seen in air masses. By analyzing the composition and dynamic characteristics of molecular beams, some physical intuitive feelings can be obtained. For example, in an air mass, each molecular beam has a uniform trajectory. Over time, the molecular beams will form suspended air currents, which is why locals will not fall into the water.
2.2 Main parameters in the MPP model
(1) Amorphous area (Bulk-like Area)
The MPP model assumes that the molecular diffusion layer is a typical Fresnel (Fisherian) structure, and its core area is called the "amorphous face area". This structure originates from the superposition of the particle's molecular structure, spin state, and motion state. These three can act on the molecular diffusion layer to a certain extent to form a tiny Fresnel potential field. The creation of this model is based on experimental findings: under microlenses, the microscopic space of particles can jump, rotate and change freely in a short period of time, which gives the molecular diffusion layer dynamic behavior beyond the crystal structure. However, in practice, the amorphous facet region is far from being able to simulate molecular diffusion. Therefore, it is necessary to combine actual experimental data to establish a parameter model of the amorphous face region.
(2) Relaxation coefficient (Hooke's Constant)
The relaxation coefficient, also called the temperature coefficient, is an important parameter that measures the surface tension and friction of particles. Particle surface tension and friction are both the result of material interactions at the microscopic scale. In the molecular diffusion layer, the surface tension of particles usually depends on the electrostatic potential of the atomic nuclei, and the friction force depends on the magnetic field of the molecules. At the same time, there are different adsorption points inside the entire diffusion layer, causing various interactions between particles. Therefore, different combinations of particle surface tension, friction force, and charge potential energy will affect the direction, speed, angular velocity, tension, and tension of particle movement. acceleration etc. The relaxation coefficient is an important parameter used to describe these interactions of particles.
(3) Surface Temperature
Surface temperature is also called surface cooling temperature, which refers to the surface temperature of particle droplets. At normal temperatures, the molecular surface will not melt, but under certain conditions, the molecular surface will cool and form a droplet surface of particles. The surface temperature determines the movement speed, size, friction and collision force of particles in the molecular diffusion layer. Generally, the higher the surface temperature, the easier it is for the surface of the droplet to cool, and the movement speed and size of the molecules will decrease accordingly.
2.3 MPP model calculation technology and experimental verification
(1) Computing technology
The MPP model is a complex theoretical model, and a large amount of calculations are required to solve the parameters. At present, the computing technologies of the MPP model mainly include the following:
- Method based on statistical simulation: This method simulates all interactions on particles, adsorption sites, molecular diffusion layers and surfaces to estimate the parameters of each molecular diffusion layer. Through simulations under different conditions, the MPP model parameters under different conditions can be calculated.
- Method based on numerical simulation: This method uses computer programs to simulate all interactions on particles, adsorption sites, molecular diffusion layers and surfaces, and then solves for the MPP model parameters. Currently, the two most commonly used numerical simulation methods are Time Integration Method (TIM) and Classic Euler Method (CEM).
- Method of experimental measurement: The parameters of the MPP model can be obtained by collecting and processing experimental data. For example, under common experimental conditions where the droplet interface temperature is normal temperature, the surface thermal conductivity of the particle can be obtained by measuring the surface temperature. By experimentally measuring the surface heat conduction coefficient near the material surface, the friction coefficient can be obtained; near the molecular surface, the charge potential energy coefficient can be obtained by experimentally measuring the large and small charge potential energy.
(2) Experimental verification
For the MPP model, an experimental platform that conforms to the actual situation needs to be established to verify it. For example, in many countries, laboratory facilities are still relatively backward, so you may encounter difficulties in experiments. In addition, the MPP model is only limited to the theoretical and physical levels and cannot accurately predict experimental results. Therefore, it is necessary to combine many aspects, such as microbiology, materials science, mechanics, computer science, statistics, etc., to jointly build a complete verification platform for the MPP model.