Independent Research-Term 1(updated) Topic 8:Silicon in Computer Chips

Fromto

                                                     The secret of computer chips!

    How is silicon made into Microchips? For those who like computer chips, this may be an interesting article for you. To make this as accurate as possible, do put comments in the comments box for improvements that can be done.

     Firstly, Sand is collected as it has a high percentage of silicon in it. To extract silicon, silica sand is commonly used as it is sand in its purer form, or in other words, less impurities. It is then reduced, or to put it simpler, have its oxygen taken away. This is done by heating a mixture of silica and carbon in an electric arc furnace to a temperature of more than 2,000°C. The carbon reacts with the oxygen in the molten silica to produce carbon dioxide (a by-product) and silicon, which settles in the bottom of the furnace. The remaining silicon is then treated with oxygen to reduce any calcium and aluminium impurities. Sound like a complicated process? It's going to be far worst later!

     The next step is making an ingot out of it. What's that? It's a cylindrical crystal made up of many silicon crystals. What is it use for? That can be explained later.Well, how is it done? Using the Czochralski process, a seed crystal is placed in molten silicon with a dopent, like Boron or Phosphorous, just above its melting point. With careful control, a cylindrical single-crystal ingot can be obtained. Now to what its used for.This Ingot is cut into individual silicon discs called Wafers, which are then polished to a flawless, mirror-smooth surface.Next, a light-sensitive,etch-resistant material called Photoresist is put onto the wafer surface.After it has hardened, it is exposed to ultraviolet light to make it soluble.  Ultraviolet light passes through a mask, then through a lens to print circuit patterns. These are used to make many different components, be it transistors(shown below), resistors, capacitors,etc.
-------------------------------------------------------------------------------------------------------------------
 Making of Transistor
     Now to making a transistor. Even though it doesn't sound close to making chips, it is one of the main components of making chips, ,making the necessary electrical connections within a chip, thus making it just as important to be in the process of making chips. To have a better understanding, we first need to know the fabrication process. If you all are hungry for more extremely detailed information, do visit here (Chpt 2) However, I'll give an overview of the process. This process is extremely simple, basing itself on these steps: oxidation, diffusion(old days)/ion implantation(modern-will touch on this), etching, deposition. That's all!
     
      A sample of silicon crystal is first placed in an oxidising environment to oxidise, forming a layer of silicon dioxide at the surface (a).(Oxidation) This layer makes sure that unwanted impurities will not diffuse into the wafer during high heating. However, this gives the chip a very big problem. As shown below(scroll to bottom of post)P-type impurity accepts electrons while N-type donate electrons. With that exact amount of electricity passing through, an electrical current will be formed. However, now, we are isolating P-type silicon crystal from N-type!

      So, photolithography comes into play. Photolithography is the selective removal of oxide(silicon dioxide) in desired areas. During this process, Photoresist, a photosensitive material, is coated over the silicon dioxide (b). A Photomask is then placed on areas of the silicon dioxide that are not meant for removal. Ultraviolet light of the appropriate wavelength is then directed on the wafer, causing areas of the photoresist which are uncovered to be chemically dissolved (C).
      
     That's not all! Our main motive is for the unwanted silicon dioxide to be removed, which is those under the dissolved photoresist. To make it so, the wafer is dipped into an etching solution, like hydrofluoric acid, or by exposing it to an electronically done plasma etcher. The silicon dioxide can now be etched away, leaving the bare silicon surface it used to be at that area.(etching) The remaining photoresist is then removed using a chemical stripping operation, leaving holes, or 'windows', at desired locations(e). It then undergoes a predisposition and diffusion step, forming p-type or n-type regions(depending on wafer type) where the removed silicon dioxide used to be(f).(Deposition) P and N-type silicon crystals are now free to be connected to each other. All these improves performance and reduces leak

                                           Special Thanks:Analysis and Design of Analog Integrated Circuits                                              By Paul R.Gray, Paul J.Hurst, Stephen H.Lewis, Robert G. Meyer

     Lastly, we have to connect the main wafer to the transistor. Three holes are etched into the insulating area above transistor and filled with copper or another material that makes metal connections.Then, the wafer is then put into copper sulfate solution for copper ions to deposit onto the transistor.After the copper ions have settled as a thin layer of copper on the transistor surface,the excess material is polished ,revealing a pattern of copper.These interconnect the transistors in a silicon chip.The chip is done!
 
    Now to something which may or may not be interesting to do. How they package the transistors. After wafer processing, they are tested to see if they work and then cut into pieces called die. The die is then attached to an individual chip carrier or substrate. With the package body containing an internal cavity where the chip is mounted, these cavities provide ample room for many connections to the chip leads (or pins). The leads compose the second interconnect level
and connect the chip to the global interconnect medium, which is normally a PC board.
-----------------------------------------------------------------------------------------------------------------
     This is a whole lot of steps! However, you may be asking a daunting question.Why is pure silicon deliberately made impure again when it is used to make computer chips? Well, if you haven't, it's time to. Remember N-type and P-type impurities? That's what we are going to talk about.

     Firstly,all crystallised silicon atoms form a covalent  bond as shown on the right.This happens as silicon contains four valence electrons,or in other words,four electrons in its outermost shell.In chemistry,an atom is at its strongest with eight valence electrons.Thus,silicon 'shares' their four electrons with surrounding atoms to have eight.




     Now, as we know, silicon is a semiconductor as it exhibits both properties of a conductor or insulator, depending on temperature. In theory, when electricity is passed through, be it insulators or conductors, there would be a conduction band, where electrons can flow, forbidden band, where no electrons are found, and valence band, where the electrons are attracted to the nucleus. The forbidden band is the one that differentiates insulators from semiconductors from conductors.(as shown on the right) So, in this case, the semiconductor, with the right amount of voltage, can become a conductor. To give you a better understanding, do continue reading.



     For an electric current to succeed, free carriers(electrons and holes) need to be created. However, with silicon crystals so closely bonded(with 8 valence electrons, very strong), dopers, or in other words, impurities need to be added.

 
     Why? When P-Type dopers like Boron are added, there is a hole created in the silicon bond.However, when N-Type dopers like Antimony are added, the community would have an extra electron. Thus, in silicon crystals, chip-makers would add both N-Type and P-Type dopers to create holes and extra valence electrons.

      Sounds funny? But there are reasons. When there is a hole in P-Type, naturally, the community would want to fill it up, just like when you find something strange, you want to find the root cause right? This creates a diffusion of holes from P-Type to N-Type and vice-versa for N-Type electrons to P-Type, causing P-Type semiconductor region becoming more negatively charged while N-Type becoming more positively charged. Problem settled. Electricity travels from negative to positive right? But wait...no.

     The two components of the current formed by the hole and electron movement across the junction add together to form the diffusion current, a phenomena that creates a depletion region, impeding electrical movements.This problem is solved when the right amount of voltage, from the power source, is put through to the N-type region, that will finally closen the depletion region. (Fun fact:If voltage is put through the P-type region, depletion region will be widened and more voltage need to be put in to allow the diffusion.) So here we have it. Why impurities need to be added back to silicon to make computer chips.

 Thus is the magic of silicon chips!

Credits:
1.Intel
2.Georgia State University
3.C.J.Savant,Jr,Martin S.Roden,Gordon L.Carpenter,Electronic circuit design-An engineering approach,pg2-9
4.PVEducation
5.techradar.components
6.Analysis and Design of Analog Integrated Circuits, Paul R.Gray, Paul J.Hurst, Stephen H.Lewis, Robert G. Meyer, Chapter 2