# How are integrated circuits fabricated?

How are integrated circuits (e.g. a microprocessor) fabricated from start to finish? For example, there must be some wiring with resistors, capacitors to store energy (bits) in a field, transistors, etc....

How is this done? What machinery and chemical processes are required for building an integrated circuit?

• Related: electronics.stackexchange.com/q/7042/8159. You can get your custom ICs manufactured, but for small runs this is not cheap. – Renan Apr 29 '13 at 21:28
• Take a look at these slides. – Kaz Apr 29 '13 at 21:31
• Intel recently constructed a new fabrication facility near where I live, to make chips with 14-nm process technology. It will go on-line this summer. Cost: $5 billion. – tcrosley Apr 29 '13 at 22:22 • RepRap has been trying to figure out how to print circuits. Of course, these would be orders of magnitude bigger than any IC, but probably the closest realistic thing out there. – Phil Frost Apr 29 '13 at 22:36 • Probably the easiest way would be to enroll in a university to become the type of engineer who does this for a living, which has the advantage of also giving you an education and hopefully a career. – Jeanne Pindar May 2 '13 at 0:03 ## 7 Answers No big deal really. First you get a pile of silicon. A bucket of ordinary beach sand contains a lifetime supply if you're going to make your own chips. There is lots of silicon on this planet, but it's mostly all so annoyingly bound up with oxygen. You have to break those bonds, discard the non-silicon stuff, then refine what's left over. You need very very pure silicon to make useful chips. Just smelting the silicon oxide into elemental silicon isn't anywhere near enough. The bucket of sand was mostly silicon dioxide, but there will be little bits of other minerals, bits of snail shells (calcium carbonate), dog poop, and whatever else. Some of elements from this stuff will end up in the molten silicon mix. To get rid of this, there are various ways, most having to do with very carefully allowing the silicon to crystalize at just the right temperature and rate. That ends up pushing most of the impurities in front of the crystallization boundary. If you do this enough times, enough of the impurities get pushed to one end of the ingot, and the other end might be pure enough. To be sure, you wave a dead fish over it during a full moon while thinking only pure thoughts. If it turns out later that your chips are no good, then one possibility is you botched this step by using the wrong species for fish or that your thoughts weren't pure enough. If so, repeat back from step one. Once you have pure crystalline silicon, then you're almost done, just another 100 steps or so that all have to be just right. Now cut your pure silicon into wafers. Maybe that can be done with a table saw or something. Check with Sears to see if they sell silicon-ingot-cutting blades. Next polish the wafers so that they are very very smooth. All the rough stuff from the table saw blade needs to be gone. Preferably get it down to a wavelength or so of light. Oh, and don't let oxygen at the open surface. You'll have to flood your basement with some inert gas and hold your breath for a long time while you finish the polishing. Next you design the chip. That's just wiring a bunch of gates together on a screen and running some software. Either spend a few 100s of k$ or make your own if you've got a few 10s of man-years free. You can probably do a basic layout system, but you'll have to steal some trade secrets to be able to do the really good stuff. The people that figured out the really clever algorithms spent many M$doing it, so don't want to give out all the cool bits for free. Once you have the layout, you'll have to print it on masks. That's just like regular printing, except for a few orders of magnitude finer detail. After you have the masks for the various layers and photolithography steps, you need to expose them onto the wafer. First you slather on the photoresist, making sure it has a uniform thickness to within a fraction of the wavelength of the light you will use. Then you expose and develop the resist. That leaves resist over some areas of your wafer and not over others, just like the mask specified. For each layer you want to build up or etch or diffuse into the chip, you apply special chemicals, usually gasses, at very precisely controlled temperatures and times. Oh, and don't forget to line up the masks for each layer in the same location on the wafer to a few 100 nm or better. You need really steady hands for that. No coffee that day. Oh, and remember, no oxygen. After a dozen or so mask steps, your chips are almost ready. Now you should probably test each one to find out which ones hit impurities or got otherwise messed up. No point putting those into packages. You'll need some really really tiny scope probes for that. Try not to breath as you're holding a dozen probes in their targets to within a few µm on the special pads you designed into the chips for that purpose. If you've done the passivation step already, you can do this in a oxygen atmosphere and take a breath now. Almost done. Now you cut up the wafer into chips, being careful to toss out the ones you found earlier were no good. Maybe you can snap them apart, or saw them, but of course you can't touch the top of the wafer. You have the chips now, but you still need to connect to them somehow. Soldering on silicon would make too much of a mess, and soldering irons don't have fine enough tips anyway. Usually you use very thin gold bond wires that are spot welded between the pads on the chip and the inside of the pins of whatever package you decide to use. Slap on the top and glob on enough epoxy to make sure it stays shut. There, that wan't so bad, was it? • This was very entertaining. I found myself holding my breath every time I read the words silicon/wafer. – apalopohapa Apr 30 '13 at 1:40 • Awesome answer, this is exactly why I read "hot questions" section. Taught me a lot with a laugh. – Paystey Apr 30 '13 at 8:17 • Minor nit: Usually the last step in the masking/etching process is a thick glass (silicon dioxide) "passivation" layer. After this point, you can stop holding your breath. – Dave Tweed Apr 30 '13 at 11:28 • Oh, and the materials you're using to form the thin film also coat everything else in the chamber, very quickly clogging up the works. To resolve this, just flood your chamber with pure chlorine trifluoride from time to time and it'll scour the silicon right off those chamber walls. Oh, but you don't want to spill any of that stuff; one drop'll set your concrete floor on fire, and if you're not already at a dead sprint when you see the first drops spraying from the regulator, the clouds of hydrofluoric acid will speed you on your way... to an early grave from HF poisoning. – KeithS Apr 30 '13 at 19:01 • The photolithography part becomes fun when you realize the wavelength of light is around 10 times bigger than the width of the traces you're laying down. – gbarry May 3 '13 at 20:39 This question is equivalent to asking: "I want to build a 747 jetliner in my basement, but I need to do it only from drawings and raw materials." The fact that a question like this does get asked really just shows how little-appreciated the complexity of what is involved in modern semiconductor manufacturing and the pure inventiveness that it entails. The thing to know about processing is that you build everything up from raw materials. Except for the wafers; you can easily buy those. But once you have started, you are layering up the device as you go; it's like baking a cake. You can build your own plane by ordering in engines and carbon composite material separately. But here you have to make everything from raw materials. And the manufacturing complexity to even get working devices is staggeringly difficult. I will just list a small number of the things that need to be considered. ## The industry: 1. There has been more effort in terms of money spent, man-power consumed or papers written, PhDs obtained etc. that any other single technical endeavor that leads to a manufactured product ever in the history of mankind. Without regard to feature size and capability, you need to be aware of the following regardless of what you will be trying. ## Cleanliness: 1. Si wafers are some of the most pure substances that have ever existed on this planet. If I use a standard $15 \: \Omega \cdot \text{cm}$ prime wafer (what is usually used in CMOS) - the dopant density is 1×1015 atoms/cm-3. There are 5×1022 atoms/cm-3 in Si. So that means for every 50 million silicon atoms there is one dopant atom. You really need specialized equipment, handling and procedures to be able to maintain that. 2. In processing, deionized (DI) water is used. This is so pure that the electrical resistance is measured in megaohms. There are so few contaminants in the water that it stops conducting. A major contaminant in early days of semi-conductors processing (discovered by Andy Grove of Intel fame) is sodium. CMOS processes are so sensitive to this contaminant that the sodium from the salt in your sweat that is contained in an average thumb print is sufficient to contaminate 10,000 gallons (25,000 L) of DI water. 3. Operating environment: every square meter of floor space has to have an air plenum above and below to move the air through, filter it and bring it back in. In a standard fab they move millions of cubic meters of air each day. Actually each fab comprises three floors with air handling using up the bottom and top floors, and only the middle having people/equipment. Seems kind of important. ## Nasty kill-you-dead-instantly sort of chemicals or nicer burn-your-face-off-slowly types: 1. Hydrofluoric acid: eats through glass just loves all that tasty calcium in your bones. If dropped on your skin it penetrates through the skin (skin is permeable to this) and heads to the calcium channels in nerves and heads to the bones. Very painful. 2. Specialized etch chemicals: Let's see... my favorite is something called "Piranha etch". It's called that because it eats organic materials, needs to run at 80 to 90 °C but also needs to be actively cooled because it has a tendency to run away and erupt in a boiling mess. 3. Silane - a pyrophoric gas - which means that it bursts into flame and explodes in the presence of oxygen. It is toxic, and when it burns, it leaves behind SiO2 vapour - which means the air is filled with tiny microscopic particles of glass that are perhaps ~900 °C. And this is one of the more benign reactive gasses, there are other chemicals present that when the leakage alarm goes off it is generally felt that there is no point in running: it's already too late. 4. Dopants: Let's not forget about the necessary dopants that will allow creation of N-type and P-type semiconductors. Boron, Phosphorus, Arsenic, Gallium (less common). 5. Let's stop here... it will be too morbid otherwise. And no you don't have a choice, unless you think you can do better than trillions of$$of investment. 6. Materials in general all have to be semiconductor grade. So you have to be in a major center and the local suppliers have to have the material on hand. Some of the raw materials have to be manufactured locally because you can't ship them. ## Here are some sample things about the equipment used: 1. Vacuum pumps: most processes run in vacuum conditions. 2. The oven, you need an oven that can sustain 1200 °C with various chemicals injected in like silane and ultra pure oxygen etc. 3. Implanters: most dopants are introduced into the substrate via a modified nuclear accelerator. The good news is that it can't be too powerful because implanters above 3 MeV tend to turn the substrate radioactive so they don't build them to be too high an energy, but you'll still need at least a 1 MeV implanter. You could choose not to use a high energy implanter but then you have to run the oven for many many hours to allow the dopants to diffuse in. The best bet is to buy used equipment. Unfortunately it's been at least 20 years since anyone has designed and built equipment for 100 mm and 150 mm diameter wafers, and there isn't any on the used market. Various universities have stockpiled equipment. I'd recommend buying used 200 mm equipment. The really good news is that can now be had for only about 15% on the dollar. So what would have been a$10 million stepper (used in imaging the wafers) is now only \$1.5 million.

• I would have voted this several times if I could. Kudos for informing, and keeping the snark at bay! – Vaibhav Garg May 7 '13 at 4:42
• The density units seem a little odd -- since there's a division sign between grams and cubic centimeters, the exponent should be positive. i.e. either atoms/cm<sup>3</sup> or atoms &times; cm<sup>-3</sup>. Unfortunately the changes are too small for a valid edit to be done. – Dan Mašek Sep 7 '16 at 1:30

There are people doing this at home, but it's a bit like trying to build a space program in your back garden. It's much harder than e.g. a 3D printer, and involves some nasty chemistry and astonishingly high precision engineering.

https://code.google.com/p/homecmos/ , although they've not actually produced a device yet.

http://hackaday.com/2010/03/10/jeri-makes-integrated-circuits/ : apparently a working device with more than one transistor.

Edit: for practical purposes, and if you're more interested in electronics than chemistry, start learning Verilog and FPGAs.

On this site the process of making a microprocessor is explained. Well detailed even though it's impossible to illustrate each of the 1500 steps required.

• +1 For mentioning steven's website – m.Alin Aug 28 '13 at 15:08

A more appropriate question is "What and how are electronic circuits combined to create microprocessors?" Electronic circuits aren't implanted onto microprocessors. Microprocessors are comprised of electronic circuits.

Resistors, capacitors, and inductors are passive analog circuit elements. The development/invention/discovery of semiconductors gave way to diodes and transistors. Transistors are configured into basic logic gates which implement boolean algebra, and flip-flops which implement basic memory elements. These basic logic gates are configured into more complex circuits which implement addition (an adder), or subtraction (a subtractor), or multiplexing (switching), or de multiplexing, or left-shifts or right shifts and so on. These complex circuits are jammed together with some control logic to make an ALU, or an instruction decoder, or a memory address decoder or some other interface. This ALU is combined with an instruction decoder, a memory address decoder, a memory or 2, and some other elements to form a CPU or Microprocessor.

All of this takes up Millions (or maybe even billions now) of transistor gates. Some current FPGA technologies use 28 nanometer process technology, which, AFAIK, means that a single gate is 28 nanometers long. Designing and Building large scale (LSI) and very-large scale (VLSI) intergrated circuits is a process that requires very specialized knowledge in physics and chemistry and very specialized and expensive equipment.

If you want to functionally design a microprocessor, that is something that you can do. And you could probably implement it on reconfigurable hardware such as an FPGA. If you want to physically design a microprocessor, that is another story. The people who design integrated circuits generally aren't even specifying out the physical layout of gates. They use design tools, not unlike what software engineers use, to say what they want their integrated circuit to do using something called Hardware Description Language (HDL), and then the tools boil the HDL down to a gate level specification.

You definitely won't be able to get this done at home! Manufacturing chips is a complex process involving lots of precise, expensive, complex machinery.

If you are interested in developing your own microprocessor, start by learning VHDL or Verilog and getting it working on an FPGA. Then you might consider learning chip design at a transistor level and getting an IC manufactured. This is not cheap or simple and requires a very specific skill set.

Lets not forget that, aside from MAKING the actual IC (covered in a very humorous and accurate fashion already here) you also need to know how to design circuits that lend themselves to IC implementation. You'll not find very many passive components within an IC - they are not so well behaved and typically take up a disproportionately large area. Instead, you'll find lots of current mirrors, sources, and sinks. P and N type devices aren't created equal, so you'll need to understand the inequities there, also. Actually, because you're "rolling your own" process, you'll need to shoot some test wafers with varying levels of doping concentration ("rainbow wafers") with a variety of test structures and then spend a lot of time and effort (figure at least 10 man years) to characterize what you end up with - to obtain a library of transistor types. Armed with your library, you can start your circuit design - assuming you have some understanding of layout. Don't forget that AFTER fab, then starts the test and debug. That's a whole NEW chapter!

## protected by clabacchio♦Nov 20 '17 at 15:17

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).