Sunday 15 November 2015

Geography of Silicon Chip

The making of silicon chips is a Light Industry. This means the products a small light and easily transported.
The inputs processes and outputs of the creation of the silicon chip is:

Silica Sand
Hydrofluoric acid
Nitric acid

Heating in Furnaces 
Fractional Distillation
Wire Saw 
Chemical Mechanical Polishing

Silicon Chip
Broken Silicon Chips

That is only to name a few inputs processes and outputs of make the silicon chips.
FABs is the name given to the clean room facility that make the silicon due to tiny dirty particles destroying silicon chips. An example of these FABs is in Leixlip in County Kildare Ireland. Sit for Fabs are specially pick for flat stable ground and good transport to receive and transport goods and resource. These sit can cost Billions to build and run.
Silicon Chip are use all over the world in many devices from computer to even toys we use them every day but you now know how they are made.

Earth Matters Folens by Eddie Guilmartin and Edwina Hynes

Saturday 14 November 2015

Finishing the Chip

Now a gate is add to the silicon. A gate is a thin layer of silicon oxide which acts as an insulator. The first part is put between the source and the drain. This is accomplished by chemical vapour deposition (CVD). CVD is done by filling a furnace with gases so a chemical reaction takes place on the surface of the silicon. Next a layer of conductive silicon is put over the silicon oxide layer. CVD is use for this as well.
Billions of MOSFETs put on the wafer but now they need to be connected. Fist we must insulate the silicon with a layer of silicon dioxide by CVD. Now that the surface is insulated can't make connection to the Source, Drain or Gate. To remake the connection they use a method called double damascene. Need to make tungsten connecting pin and copper interconnect. Hydrofluoric acid etches holes into the silicon dioxide through photoresist. Next the interconnection tracks are etched in. A top layer of copper is electroplated on. It fills the holes to make connections in the MOSFETs. The pins through the insulating layer are called vias. Now the top layer of copper is removed using chemical mechanical polishing. The pin pins in the tracks are left. A final insulating layer of silicon dioxide is connected using the vias.
The wafer has hundreds of chip on it and they have to be checked if they are working or not.

This is done with a wafer probe. Uses pins that touch the contact of the chip and pass electrical signals. Sometimes if part of the chip does not work it is sold as a lower end chip.
Finally the chips are cut out and pack and sent off.
I referenced

Intel Museum website

Tech Radar

Sunday 1 November 2015

Making the Circuit

Now that we have a silicon wafer we have to put the circuit on the wafer. The circuit is tiny and use very precise methods to put the advance circuits on the chip.

First the wafer is heated the silicon reacts with the oxygen to make silicon dioxide layer. Then a layer photoresist and is spun in vacuum and then dried the spinning is done so the layer is even. The wafer is then exposed to ultraviolet light through a photographic mask. The mask is used to block the light to make the circuit on the silicon this step is repeated for each chip on the wafer. Different masks are use to build up different layers. The wafer is the developed in a alkaline solution then area exposed the ultraviolet light on the photoresist is washed away. Hydrofluoric acid is used to dissolve the silicon oxide where the photoresist was washed away.

Now that the circuit is put on it is time to add the transistor to the processor it is called MOSFET. 
There are two types of MOSFET P-channel made of p type materials and n-channel made n type material. 
First they must create a P type and n type areas on the circuit.
The wafer has a beam of boron ions they implant themselves into the silicon in the gaps in the photoresist called p wells which will be use in the n channel MOSFET. A different photoresist pattern is now applied to make n well for p channel. Yet anther photoresist beam is use to create the n type region in the p well this will act as the source and drain of the n channel. Next a layer of silicon germanium doped with boron is applied. 

Saturday 31 October 2015

Making A Silicon Wafer

The silicon ingot is still not suitable semiconductor due to its polycrystalline structure. In order for the silicon to become a semiconductor it must be made into one regular structured crystal.
It is done by melting the silicon in a quartz crucible which rotates and is heated to 1414 °C.
A seed crystal of silicon is lowered into the crucible rotating in the opposite direction of the crucible. The seed cause the molten silicon to crystallize around it the crystal is slowly removed from the crucible. The rod which has formed is called boule. This process is called Czochralski.
Next the boule is cut into thin discs which are called wafers. The boule is cut with a wire saw. A 300mm diameter wafer is cut 0.775mm thick. Next the wafers are polished by lapping to be flat with in 2μm. Then is etched in nitric hydrofluoric and acetic acids. The nitric acid gives the wafer a layer of silicon dioxide which the hydrofluoric acid dissolve to leave a clean surface the acetic controls the reaction.

I referenced
Intel Museum website
Tech Radar

Wikipedia for pictures

Youtube video on Czochralski 

Monday 12 October 2015

Making Silicon

The first topic I will cover in my blog is silicon chips. I will first be covering how silicon chips are made for computers.
Silicon is the 2nd most abundant material on earth but only found in bonded states and I'm going to explain how it is purified. 


Sand has silica (silicon dioxide). That is why it is the starting material for making silicon chips. But no ordinary sand from the beach will work. They use a purer form of sand called silica sand. This sand is gotten from quarries.

The sand needs further purification by removing the oxygen from the silica in order to made silicon. They do this by mixing silica and carbon and put it in a special furnace called the electric arc furnace and heat it till over 2000°C. The oxygen reacts with the the carbon to form carbon dioxide. The silicon then settles to the bottom of the furnace. Next the silicon is treated with oxygen to remove any calcium and aluminium from the silicon. Now the silicon is of a metallurgical grade and is up to 99% pure. This is still not pure another for the silicon chips.
The silicon is ground to a fine dust which is reacted with gaseous hydrogen chloride in a fluidised bed reactor which is heated to 300°C which make a liquid form of silicon called trichlorosilane. The impure material reacts into their chloride which is then removed through a process of fractional distillation. The trichlorosilane is vaporized and reacted with hydrogen gas at 1100°C so they can change it back to silicon. The silicon is poured over an ultra pure silicon rod to make a silicon ingot of electronic grade which is 99.999999% pure.

Making Silicon Chips

I referenced
Intel Museum website

Tech Radar

Thursday 1 October 2015