#engineering #energy
If you were to look across semiconductor manufacturing technology a few common steps would stand out.
Growth of semiconductor boule (crystal).
Post processing steps that cut the boule into wafers, polish those wafers etc.
Doping of wafers as well as deposition of active components on wafer,which will differ from application to application.
Silicon, the most popular semiconductor material, is grown into a boule from the liquid phase. The dopants are mixed in the melt and the big crystal that is formed after the process is finished is chopped into multiple thin small wafers.
Crystallization from liquid phase is quite popular in semiconductor metallurgy but crystal growth does not have to take place from the liquid phase. Yes it is the most popular way to grow silicon crystals but it is not the only way.
More and more fabrication technologies are preferring growth from the vapor phase.
Why? Vapors/gas can be controlled more easily. It can lead to higher purity, faster deposition times as well as fewer defects.
The greatest advantage of Vapor phase crystallization is deposition of thin films on substrates.
Across a wide range of semiconductor products substrate and active regions are grown separately. The substrate only exists to support the growth of the active material serving little useful purpose on its own.
Vapor phase crystallization offers deposition of very thin films of the active region,enough to do what they are supposed to do and nothing more. This saves time and materials.
Such precise control over deposition is extremely hard to achieve using any other method, certainly not from liquid phase growth.
Vapor phase crystallization makes heavy use of organometallic compounds. Due to their lower boiling points as well as the ability of organic molecules to supply some of the energy needed through formation of combustible reactants.
Semiconductors like gallium nitride rely on Metal organic vapor phase epitaxy (MOVPE) technique for their deposition on the substrate. Silicon carbide crystals have also been successfully grown from gasses by sublimation of silicon carbide powder.
The furnace is the device that makes all this happen. It provides the energy needed to make the phase change. From solid to liquid or gas before crystallization takes place. Furnaces are absolutely critical to building semiconductors.
There are many types of furnaces in the literature. This lecture provides a good overview on different types of furnaces.
https://www.youtube.com/watch?v=GUctNyh1FR0
You can easily modify any of these to support crystal growth.
Vegetable oil is a good (cheap and renewable) source of fuel providing temperatures in excess of 1500 degree celcius, some of them even hitting 2000 degree celcius. Wood /husk is another.
Depending upon the type of semiconductor you are growing the fumes as a result of burning may be used as a carbon source ( for example in synthesis of silicon carbide or carbon nanoparticles) or it may be discarded at the exhaust.
One thing to keep in mind is that in the subject of metallurgy, furnaces are usually discussed in the context of alloying or purification of metals.
You will need to adapt the theory to suit your goals of semiconductor crystallization. But it's not too far away from the main topic.
Crystallization is not all that different from casting/sintering. And doping is just an extremely specific case of more general alloying principles.
Purity in semiconductors are important in some applications ,like computing, and are overrated in others like solar cells and led.
Despite all the investment,years of research, and generous Nobel prize distributions, the industry has not delivered on the efficiency front. The number is hitting 20% for solar and just around the same for LED.
This would be a good opportunity to experiment and allow impurities to seep in and introduce energy states into the crystal. Chlorophyll, an organometallic compound, the most efficient, the most widely deployed semiconductor material on the planet is not made in a clean room.
Crystallization in an open atmosphere should be possible using commonly available tools and techniques. The bar set by industry leaders is pretty low and there are many ways to the top of the mountain.
See also
A playlist containing some incredible lectures on epitaxy
https://www.youtube.com/playlist?list=PLKMc2qT6vGRuwTkKP6P1TGlSN9lfTvLmC
A cool technique developed by NTT for Gan semiconductor release
https://www.youtube.com/watch?v=6DGkH_GyY0Y
One of the most popular methods for fabricating semiconductor devices, Vapour phase epitaxy offers a high degree of control over the growth and mixing of dopants in the gas phase before their crystallization on solid substrate. Deposition of materials can be controlled by adjusting composition and temperature of the gas.
Watch this incredible lecture by Dr. Nandita Das Gupta,IIT chennai, that explains the process in detail
https://www.youtube.com/watch?v=tWq2OF3abNI&t=626
A flame is the result of an exothermic chemical reaction. A fuel is burnt in the presence of oxygen to release energy as heat and light. If it's a hydrocarbon fuel the products are usually carbon dioxide and water. Most of the methods we use for creating fire like burning of lpg gas, a matchstick, a candle or a diya are all examples of flames arising from an exothermic chemical reaction. In fact fire and its manipulation goes back to ancient times and it is one of the most important discoveries in human civilization. Fire is used today everywhere from cooking to generation of electricity.
It is common knowledge that burning a fuel gives off energy as heat and light. How much energy is released? How much fuel needs to burn? What products are formed as a result of combustion?
Prof. K Ramamurthy,IIT madras explains in this video
https://www.youtube.com/watch?v=C87ooSaRpzE&t=0
Generally we don't need to concern ourselves with the underlying chemistry of the reaction leading to a flame. A diya will work just fine without us needing to know what is the composition of the element that burns and what is the result of its burning. Nevertheless with the understanding of the fuel in its various phases characterization of flame and its temperature becomes easier. This understanding allows us to select fuel for our particular use case and manipulate flammable chemical reactions to our advantage.
For instance deposition of carbon nanotubes on a substrate by pyrolysis of an organic compound like camphor.
While CNTs are a bit tricky to figure out, everyone is familiar with soot. It turns out that soot is infact a semiconductor(a mixture of carbon particles known as polycyclic aromatic hydrocarbons,PAH) with band gap varying with its size. In other words a kind of a quantum dot.
Creation of both carbon nanoparticles and nanotubes are an instance of manipulating the temperature and chemistry of carbon vapor to create the materials we want. A complete combustion of the organic reactants produce only carbon dioxide and water in addition to heat. However perfect combustion rarely happens in practice. Complex forms of carbon along with its oxides are produced which can either lower the temperature of reaction or may ignite themselves to release much higher energy. It's a bit counter intuitive but carbon monoxide itself is a fuel with very high heat of combustion. Similarly carbon dioxide is an insulator that stores the heat energy minimizing its loss to the external environment.
This property of oxygen containing carbon compounds can be utilized to produce carbon nanoparticles in systems where heat is provided from an external source for example electric furnaces where maintaining fresh flow of oxygen is not as important as in a chemical flame driven process. Electric furnaces however suffer from disadvantages of higher upfront cost, higher maintenance cost as well as reliance on electrical supply.
How does a substance organic or otherwise give off energy? Where does this energy come from? The energy given off by a substance during combustion is stored in its chemical bonds. At the time of ignition we supply it with enough energy to break those initial bonds and undergo chemical reactions. Initially the heat of formation of the intermediate products is larger than that of reactants. In other words the intermediate products are much less stable than the reactants. They further react with oxygen in the air to produce more stable oxides.And thus the balance energy is released off as heat.
(To understand this consider can you burn water? No.To break it into constituents external energy needs to be supplied. But the keass table hydrogen and oxygen can react more easily to form water)
Here's when things start to get a bit more interesting. If you don't perform any physical work with this energy released, it will raise the temperature of the system by a proportionate amount. We need very high temperatures to perform semiconductor crystal growth in vapor or liquid phase. Understanding the working of fuels,their release of energy thus becomes very important.
Long story short,we need a chemical compound that we can burn,subsequent to which high enough temperatures can be obtained to enable us to grow semiconductor crystals.
Let us look at a few cheap items that can help us achieve our goal. But before we do that here are a few prerequisites
1. The fuel source must be easily obtainable.
2. The fuel source must be renewable. It should not lead to any net emission of greenhouse gasses.
3. It must be cheap. The operating cost of our semiconductor furnace must be negligible.
The points made above are self explanatory but we need to expand a bit on the net emissions of greenhouse gas part. Technically speaking burning any kind of fuel renewable or nonrenewable does not lead to a net emission. Petrol or diesel, a non renewable resource made of plant matter, must have at one time absorbed all the CO2 it is emitting now. The question is of time scales. If a fuel which is formed by a natural process that goes back millions of years in time and is consumed over a period of a few decades things are going to go bad.
On the other hand if there were a fuel that was produced and burnt at shorter time intervals emission and absorption would be balanced well. Sugar is produced everyday by trillions of leaves all around the world absorbing huge quantities of co2. If sugar was burnt the balance would hold because plants would simply absorb it again. In Fact that's what we already do consume food containing sugar which gives us energy to lead our day to day lives releasing co2 as a result. A similar case could be made for vegetable oil or ethanol produced from plant waste.
What you chose would depend upon your particular requirements. But keep in mind that the quantity of heat produced in a reaction is not everything. Sugar is an incredible fuel that burns out quickly,too quick to help in growth of crystal. Mustard oil is known to contain sulfur that can cause unwanted doping of your crystal. Therefore it's important to choose not just the right kind of fuel but also choose flame chemistry that can be tweaked to support our crystal growth. Where the emission products can form a part of our dopant profile or lead to some kind of alloying to introduce new energy states. If not chemically the products can also change the crystal physically eg by introducing pores.
Alternatively you may choose to make crystals that are simply unreactive towards emission products. The possibilities here are too many to list. But hopefully the information is enough to get you to start experimenting.
Comments
Post a Comment