Is there a site the sums up all the IC processing techniques?

Qacer

Platinum Member
Apr 5, 2001
2,721
1
91
Hello all,

I'm curious if there is a site that has information about general process steps on how certain processing techniques are done. I've been scouring the IEEE Xplore site and Google for various explanation on how etch-back, back sputtering, and other techniques are done. One with pictures would be helpful.

If not a site, is there a book? My main goal is to get a better overview of the processes used especially for multi-level interconnects. I've read about sputtering, CVD, etc.. But I would like to get a better overview on the others that I've encountered / read and don't know clearly.

Thanks!

 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
One book? Try 100 books,,,,
It is a HUGE field, at least if you want to really understand every step in the process (which is usually not neccesary, there are several techniques that I know how to USE, but I do not really understand in detail, I just know which buttons to push;)).

If you "only" want to understand how these processes are done commercially it gets a little bit easier since most techniques are only used in research, only some can be scaled up to commerical production.

If you give us the names of the technoques you would want to know more about someone might be able to point you in the right direction.




 

orion7144

Diamond Member
Oct 8, 2002
4,425
0
0
There is a video out there that walks you through the whole process from sand to packaging. How the Ingot is made and everything. Can't remember the name off the top of my head but it was pretty good. I had to watch it when I first started in the industry 8yrs ago.
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
I forgot to ask: Are you only interested in Si technology or are other materials also interesting (III-V, ferromagnetic- and ferroelectric materials etc)?
 

buleyb

Golden Member
Aug 12, 2002
1,301
0
0
Originally posted by: orion7144
There is a video out there that walks you through the whole process from sand to packaging. How the Ingot is made and everything. Can't remember the name off the top of my head but it was pretty good. I had to watch it when I first started in the industry 8yrs ago.

I'd be curious to see this...think you can find out the name?
 

Qacer

Platinum Member
Apr 5, 2001
2,721
1
91
Originally posted by: orion7144
There is a video out there that walks you through the whole process from sand to packaging. How the Ingot is made and everything. Can't remember the name off the top of my head but it was pretty good. I had to watch it when I first started in the industry 8yrs ago.


The video is called Silicon Run. You can actually view it online from here:
http://www.multimedia.vt.edu/ee5545/

As for any process, I'm interested in all materials not just Silicon. The ferromagnetics one also sounds interesting because I've been reading up on spintronics stuff.

Right now, I'm curious about etch back and back sputtering. I'm just curious because I really want to get an internship while in graduate school. I just don't seem to have enough industry experience in this area, so I'm limited to learning it through books and other resources.


Thanks!


 

Eskimo

Member
Jun 18, 2000
134
0
0
I'm on vacation at the moment so I'll have to keep my response fairly brief (not really supposed to be on the internet ;)).

The two techniques you mentioned are just extensions of more general process areas. Etch and physical vapor deposition (PVD). I'm not sure how familiar you are with process technology in general. When I get back home this weekend I can give you some better references that helped me in college. Now to your 2 initial queries:

Etchback is simply the reduction of a film thickness by a controlled etch. An etchback is done after you have deposited or grown a film on your wafer. I won't go into all the ways you can deposit films as that would take me too much time at the moment. The purpose of an etchback is two-fold, it can serve to planarize and it can allow you to better control the resulting thickness of a deposited film. Many deposited films are conformal in nature, meaning they deposit on all surfaces at an equal rate. Thus unless you deposit a very thick film on a wafer with a lot of topography (think gate stacks sticking up from your substrate) that topography will simply be transferred to your deposited film. Peaks in a film have more surface area exposed to an etchant and therefore the realized etch rate of the peaks will be greater than the depressions in the film. This is how etchback results in a more planar film.

We can't always control deposition as closely as we want. Most diffusion processes are still done in a batch manner where 100-150 wafers are processed in the same chamber, there is bound to be a thickness uniformity range within your batch. Rather than deal with unmanufacturable control limits at the diffusion step we simply deposit what is deemed to be a sufficient thickness with limits that ensure that the thinnest film thickness will still exceed what we want our end thicknes to bed. Then using metrology tools we can measure the the thickness of each wafer or a sample of wafers after deposition and feed those measurements ahead to a dry etch chamber tool where the processing time is altered to etch to the desired target.

If you want to know more about how etch in general works (wet and/or dry) you'll have to ask someone else or wait for me to get home.

Back sputter, not entirely sure what you are referring to here but i'll take a shot at a couple possibilities.

Backside sputtering is often used in academic or older technology. In it you use sputter to deposit a conductive film onto the backside of your wafer. This provides you with an easy substrate contact (see your device physics class/book for reasons why might use Vsub) to contact with either a grounded/biased chuck during testing.

In case you are referring to sputter in general it is technically referred to as PVD. The system pumped down under vaccum. There is a target of the material you desire to deposit onto the wafers. An inert gas is inserted (usually argon) and using a DC or RF power source a large electric field is applied across the argon gas ionizing it. The Ar ions are accelerated towards the target due to a potential bias and impact the surface of the target. If the ions impact with sufficient energy they can dislodge small amounts of the target material which is 'sprayed' onto the wafers, the longer you expose the wafers to the sputter the thicker your film.

Since your previous question was about an etch process you may also be referring to a by product of dry (plasma) etch. Similar to how sputter ionizes an inert gas to transfer material from the target to the wafer a Dry etch system uses a combination of inert and reactive gasses excited to a plasma state and bombarded at the wafer in order to remove material. Some of the material that is physically removed by the bombardment of ions into the wafer can 'back sputter' onto other parts of the wafer and your chamber walls. Sometimes the chemistry of your gases can use this to your advantage and deposit polymer byproducts onto the walls of your via to help promote anisotropic etch.
 

MadAd

Senior member
Oct 1, 2000
429
1
81
Originally posted by: Qacer
Originally posted by: orion7144
There is a video out there that walks you through the whole process from sand to packaging. How the Ingot is made and everything. Can't remember the name off the top of my head but it was pretty good. I had to watch it when I first started in the industry 8yrs ago.


The video is called Silicon Run. You can actually view it online from here:
http://www.multimedia.vt.edu/ee5545/

As for any process, I'm interested in all materials not just Silicon. The ferromagnetics one also sounds interesting because I've been reading up on spintronics stuff.

Right now, I'm curious about etch back and back sputtering. I'm just curious because I really want to get an internship while in graduate school. I just don't seem to have enough industry experience in this area, so I'm limited to learning it through books and other resources.


Thanks!



For anyone thats interested, if you want more than just the silicon_run_i video, go to here:

http://www.multimedia.vt.edu:16080/ee-mse/

There you can watch silicon_run_1 and 2 as well as a few others

(dont get annoyed with video 4 silicon magic and 5, dicing, the links dont appear to work)


hth
 

Qacer

Platinum Member
Apr 5, 2001
2,721
1
91
Originally posted by: Eskimo

If you want to know more about how etch in general works (wet and/or dry) you'll have to ask someone else or wait for me to get home.


Thanks! I can wait. I'm definitely interested in knowing the texts that you've use in your college heydays. And yes, I would like to know more about etching in general.

 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
Have you tried any of the "standard texts" on thin-film processing? Like Ohring "Materials Science of thin films"
 

Eskimo

Member
Jun 18, 2000
134
0
0
Originally posted by: f95toli
Have you tried any of the "standard texts" on thin-film processing? Like Ohring "Materials Science of thin films"

Speaking from personal experience the Ohring book sucks. We tried using it as our text for my thin films course and it was just an overall bad experience. Especially bad were the 'questions' at the end of each chapter. They often drew on knowledge outside of the scope of the text that you either had to make up or know already and sometimes weren't even related to the content in the chapter at all. There are hardly any examples and the diagrams often don't even match up with what the author is describing. I think I used the first edition of the text, maybe it's gotten better. Sorry for my mini-rant :)

Is there a site the sums up all the IC processing techniques?

A quick google search turned up
http://selland.boisestate.edu/jjozwiak/ST181/ST181.htm
There is a ppt of a relatively modern CMOS process at
http://selland.boisestate.edu/jjozwiak/ST181/10_cmosflow.ppt


Thanks! I can wait. I'm definitely interested in knowing the texts that you've use in your college heydays. And yes, I would like to know more about etching in general.

In my intro to microelectronics course I used Microchip Fabrication, by Peter Van Zant
Several other courses relied on a combination of Volume I (Robert F. Pierret, Gerold W. Neudeck) and Volume V (Richard C. Jaeger) of Modular Series on Solid State Technology.

A wealth of information is available for a price in Silicon Processing for the VLSI Era, Vol. 1: Process Technology by Stanley Wolf and Richard Tauber. More up to date information is available in Silicon Processing for the VLSI Era, Vol. 4: Deep-Submicron Process Technology, also by Wolf. Fantastic books and Mr. Wolf is a great speaker if you ever get a chance to meet him.

For lithography I recommend my former professor's book Microlithography Science and Technology by Sheats and Smith.

Most of the know how in the industry isn't captured in textbooks however. You really have to either have some fantastic professors with personal experience that can relate it to you through their coursework or you simply have to get on the job training and learn from those that have come before you.
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
I don't use Ohrings book that much either, but I asked some friends that work a lot with film growth/processing and they said it is the best book they have (I do work in the cleanroom occasionaly, but I have learned most of what I know from the people I work with, not from texts).

In the thin-film/microfabrication courses at my university we use texts from various sources, not a single book. The problem is that you then need to know which texts to look for. But I still think Ohrings book might be a good starting point. You can also try to find review papers on very specific topics.

But Eskimo is right, I don't think you can truly learn this stuff without actually working with it. Nothing beats beeing able to see the result of a process by looking at a sample in a microscope/SEM/AFM.


 

Eskimo

Member
Jun 18, 2000
134
0
0
As for etch... there are two main types, wet and dry. The general purpose of an etch step is the controlled selective removal of material from the wafer. Etch can be used for planarization, thickness control, complete material removal, and feature definition.

Let's start with some basic terminology. In almost any etch step there is what is referred to as a masking layer. This is a layer of material that has a different chemical/physical composition than the material you wish to etch and therefore will not react to the chemistry being used in the same manner. This can be any material, the most common of which are: photoresist, silicon nitride, silicon dioxide, and various metals (W, Ti, TiN, Al, Cu). An important term you'll hear about in regards to etch is isotropy. This describes how directional the etch is. An isotropic etch will etch in all directions at an equal rate, an anisotropic etch will etch much faster in one direction than the others. Selectivity, another term, is a ratio of the etch rates (distance/time) of two different materials. Often we speak of the selectivity of the film we wish to etch and the underlying or masking layer. General rule of thumb is you need a minimum of 10:1 selectivity for a process to work, 100:1 is desired.

Wet etch is the oldest and simpler of the techniques. Traditionally the wafers are processed in a batch process of 2 lots at a time (50 8" wafers). In the 300mm generation some wet processes are moving to single wafer processing for increased control and flexibility. In the batch process the wafers are placed by robot (or human) into a bath of chemical. The processes are almost exclusively timed, which means an etch rate is determined and the wafers are left in the chemical for the time necessary to remove the desired thickness. The chemical used is chosen for its selectivity to remove the desired film. Cleaning chemistry often include an oxidizing agent such as hydrogen peroxide to oxidize the surface and then a reduction agent to remove the newly oxidized surface. This allows the cleans to work on a variety of materials.

Common wet etchants include:
Phosphoric Acid (H3PO4) ? Used for removal of Silicon Nitride (Si3N4)
Hydrofluoric Acid (HF) ? Used for removal of Silicon Dioxide (SiO2)
Potassium Hydroxide (KOH) ? Used for removal of Silicon
Sulfuric Acid (H2SO4) ? Used for the removal of organic materials such as photoresist and organic anti-reflective coatings
SC1/APM (Ammonium-Peroxide-Water mixture) ? Used to remove organic contaminants in clean steps
SC2/HPM (Hydrochloric acid-Peroxide-Water mixture) ? Used to remove metallic contaminants in clean steps
Al Etch (phosphoric, nitric, and acetic acids) ? Used to etch aluminum, acetic and nitric are also used to etch chrome for photomasks

Dry etch is the newer and more advanced technique. Referred to by a variety of names you might see it called plasma etch or RIE (reactive ion etch). Most dry etch processes are single wafer performed in a chamber platform system. If you?ve never seen a chamber system before the basic description is a common load port where a cassette of wafers is placed. Directly behind the load port is a common robotic transfer stage/platform which takes wafers off the cassette and places them inside of individual process chambers located around the transfer stage. These are called cluster tools. In each chamber the wafer is placed on a chuck. The entire chamber is evacuated to a vacuum state so the chuck used to hold the wafer is often electrostatic since obviously you can?t use a vacuum to hold a wafer in place within a vacuum. A large potential difference using an RF power source is used in the presence of reactive and inert gases. At sufficient potential and under vacuum these gases will be excited into the plasma state of matter. This plasma ionizes the gases by stripping off electrons.

By negatively biasing the chuck of the wafer these positive ions can be induced to accelerate towards the wafer surface. Large inert ions such as argon serve to physically remove material from the surface. This is primarily a physical process and is very anisotropic. This allows for deep narrow trenches and contact holes (vias) to be formed. The reactive ions such as fluorine will react in a combination of chemical/physical manner. Chemically they will combine with the surface material to form volatile species which can be evacuated through the vacuum. This process is more isotropic in nature but much more selective than the physical bombardment. The lower the vacuum reached the higher the mean free path of the ions which allows more of the ions to reach the surface and increases your etch rate. It will also increase your anisotropy by reducing ion collisions above the wafer surface. Likewise the flow rate of your gases will determine the amount of ions readily available and change your etch rate.

Temperature also plays a role both in the chemical reactions occurring at the surface as well as eliminating flaking from your chamber walls. Remember that all the material you are removing has to go somewhere, when it?s physically sputtered off the wafer it gets deposited either on the chamber walls or falls back onto the wafer itself creating defects. Eventually the material on the walls of the chamber will start to delaminate and flake off onto your wafers leading to large yield problems. It?s for this reason that every chamber is taken down after so many processing hours and wet cleaned by technicians.

A common use in fabs of dry etch is for the removal of photoresist, called plasma strip. In that process the reactive ion used is oxygen. The basics of dry etch in conjunction with the right process conditions and gases can be used to strip a variety of materials and also perform dry cleans to remove residual material left on wafers after process steps. The gases used for each process are much more varied than those of wet etch and particular to each company/process so I can?t tell you particular ones.

The main advantages of dry etch is lower usage of raw materials compared to wet etch and increased control. As the ions in the plasma react with the material on the wafer different spectra of light are given off depending on the material being etched. Using this phenomena we can implement endpoint detection where the intensity of a wavelength of interest can be used as a trigger to end the etch process. Alternatively using knowledge gained by incoming measurements a timed etch can be performed.

That?s my brain dump on the subject, hope some of it is new or helpful. I should point out i've never worked in Etch professionaly, so most of my knowledge is from working with engineers in those departments and my educational research background.
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
Another usefull technique is ion-milling which is a dry-etch technique that works by accelerating ions with high-energy (usually argon) towards one side of the sample using an electric field (it works like a gun, so it is not like RIE plasma which is a more "diffuse" technique where the whole sample is plasma). You can either use hard masks (either mechanical or hard-baked resist) or, like with RIE, use the fact that some materials eatch much faster than others. Selective etching is the best way if you want to etch small structures, you define the pattern in for example a-carbon and then you etch through the carbon mask. There are also reactive forms of ion-milling.
I am not sure if ion-milliing is used commerically, if you need high-precision it is slow technique (I use if for a process where the etch time is close to two hours, etching through a carbon mask).

Often you use a combination of techniques. In the process we are using (for complex oxides, not silicon) there are many steps which includes both wet- and dry-etching, the later including both plasma-RIE and ion-milling (it also includes ,among other things, two runs in an e-beam, it takes about a week to make a single 5x5 chip, state-of-the-art processing in a research lab is not easy)
 

Qacer

Platinum Member
Apr 5, 2001
2,721
1
91
Thanks! Very informative info on etching. It is certainly another perspective of looking at it. Plus, I finally got a good overview about cluster tools. That was one of the questions that I answered wrong along with ashing.

As for RIE, the residuals were one of the defects that were discuss in our class. I think we are going to cover chemical mechanical polishing. As for the residuals, would you use RIE again to planarize the surface or would CMP be more appropriate?