Talk:Wafer (electronics)

Grammar question
In the context of electronics, is it permissible to use "wafer" as an uncountable noun? For instance, "we examined the effects of various slurry components on silicon wafer." Just curious... -- Visviva 04:52, 28 April 2006 (UTC)


 * I've never heard it used in that context. If I saw that written, I would assume it was written by a non-native English speaker. --Phil Holmes 11:58, 30 April 2006 (UTC)

A web search shows plenty of usage in both forms. I think it is ok as quoted above - they're examining effects on a material (silicon wafer), not on some specific wafers or on silicon in general. Lisamh 22:14, 17 September 2006 (UTC)


 * Wafer, in the initial sentence, is a countable noun. You can have one or many silicon wafers. The sentence dropped a preposition and it should say "we examined the effects of various slurry components on a silicon wafer". The argument made by Phil Holmes is ridiculous. There is nothing wrong with the grammar and no correlation whatsoever with "non-native English speaker(s)" and their ability or inability to write properly. You can be a native Russian or Spanish and write better than an average native English speaker (we all know "it's/its" and "should of/should have" are a huge problems, particularly in the U.S.A.).


 * ICE77 (talk) 17:51, 7 July 2018 (UTC)

Proposed merges
I think the Wafer prober article could best be merged into Wafer testing.Lisamh 22:16, 17 September 2006 (UTC)

I don't think so. —Preceding unsigned comment added by 84.83.197.159 (talk • contribs) 02:32, 30 October 2006


 * Vote for merge. DFH 22:09, 2 January 2007 (UTC)

Hi! Voting is evil. :) (It's better to reach concensus through discussion.)  I think Wafer prober and Wafer testing should both be merged into the Wafer (electronics).  They're all about the same thing (wafers), so it makes sense to have them in one article.  Since they're so closely related, they share a lot of the same terminology, and the articles will benefit from integrated context.  They're all currently stubs or near-stubs, so putting them together yields one article of a better length, rather than three too-short articles.  If, over time, it gets fleshed out to the point where it gets unwieldy, then we can do a split into sub-articles or whatever, as needed.  If people think the articles shoukd not be merged, please explain why.  Thanks! —DragonHawk (talk) 05:44, 17 January 2007 (UTC)


 * Oppose. A prober is a distinct piece of equipment, like an electron microscope, and fully deserves its own article. -- Atlant 20:50, 25 January 2007 (UTC)


 * Oppose. Wafer probing and wafer testing would be better suited under microfabrication or Semiconductor_manufacturing.  The wafers article is about the wafers themselves, not about all the processes used to make circuits.   Wafer probing/testing is a bit of a misnomer because its not the wafers being probed/tested but the circuits that were built from the wafers. Dspark76 (talk) 01:32, 24 February 2008 (UTC)

Possibly dated information
The article mentions the following: "This wafermap is then sent by a network or floppy disk to the die attachment process" Is that current? because I can't imagine floppies being used. Perhaps someone that has more recent familiarity to production could correct this if needed. Wwhat (talk) 17:33, 10 December 2007 (UTC)

I guess it's gone now - but I'm working in wafer testing, and (very occasionally) companies will send you wafer sort (testing) data in text files on a floppy. User:ItaniuMatrix10:38, 25 June 2008 —Preceding unsigned comment added by 199.106.103.254 (talk)

Standard Wafer Sizes
Many modern fabs have moved to 300mm wafers; this should be added to the list of wafer sizes. Many of the smaller sizes shown may not be considered standard anymore! -- Gretchenpatti (talk) 18:26, 4 January 2008 (UTC)

Is the size given as a diameter or an area? it is stated both ways in the article. This may be confusing to an outsider. It says 200mm square in the Formation section. I'm assuming this is a diameter. Wmcleod (talk) 20:28, 17 August 2010 (UTC)

450mm tools have already been developed and are being installed; the 450mm size is not a standard yet for chip manufacturers, but it is for tool makers, due to the lead time required, and I've updated the page. Scutigera (talk) 18:39, 15 June 2013 (UTC)

"In the future"
The "in the future" citation is ambiguous and unknowingly dated. Why add a comment like this? The reference offers no further information, except a 2008 copyright. A better citation for this standard is needed. —Preceding unsigned comment added by 98.244.56.189 (talk) 06:39, 21 April 2009 (UTC)

I think a date column needs to be added to the charts to show when the newer sizes and technologies came into existence and possibly when the older sizes were deprecated. I would think all of the "inch" sizes are obsolete and by now would be considered legacy. — Preceding unsigned comment added by 2600:8806:8501:26A0:886:95CB:B014:36F7 (talk) 13:10, 27 May 2018 (UTC)

Proposing merger of Wafer dicing into this article
The article Wafer dicing has been unreferenced since December 2007. It has received almost no non-automated changes in the past year. My recommendation is that we move over some of that information here and redirect that article. This article is not so long and it already has a brief section on wafer dicing. More importantly, my hope is that since this page is a little more active, people here will be more motivated to get it referenced. Should a merge not be approved, I will have to nominate Wafer dicing for deletion, as it has been unsourced for too long, and I personally have no idea how to find relevant sources (as I'm guessing the relevant sources will be in print, not on the internet). Qwyrxian (talk) 23:54, 6 October 2010 (UTC)
 * I agree. -- ANDROBETA 15:15, 27 December 2010 (UTC)
 * I believe it is more appropriate to merge Wafer dicing into Semiconductor device fabrication --GianniG46 (talk) 15:40, 27 December 2010 (UTC)
 * Wow...I totally forgot about my merger proposal. If either of you would like to perform the merge, feel free to do so.  I personally don't really have enough knowledge of the subject itself to determine what the better merge target is.  GianniG46, do you think you can merge the relevant info into Semiconductor device fabrication?  Qwyrxian (talk) 22:17, 27 December 2010 (UTC)


 * Unfortunately, I just know that wafer dicing is a necessary step in IC fabrication, but have no idea of its technology, so I cannot help. Perhaps one can simply copy the material into "Semiconductor device fabrication", leave on "Wafer dicing" just a few words to explain what is it and to redirect to the main article, and hope that other editors will be able to review and quote it. --GianniG46 (talk) 08:22, 28 December 2010 (UTC)

Lithography
Reading this as a layperson, this article can use a section on uses. It mentions lithography as using 50% of all wafers... but why? Justin15w (talk) 01:32, 20 May 2014 (UTC)


 * That's not exactly what the article says: it says that 50% of the cost of wafer processing is due to lithography. Converting to larger 450 mm wafers would reduce price per die only for process operations such as etch where cost is related to wafer count, not wafer area. Cost for processes such as lithography is proportional to wafer area, and larger wafers would not reduce the lithography contribution to die cost. Lithographer Chris Mack claimed in 2012 that the overall price per die for 450 mm wafers would be reduced by only 10-20% compared to 300 mm wafers, because presently over 50% of total wafer processing costs are lithography-related.
 * Most steps in wafer processing operate only a whole wafer at once (e.g. wafer etch, wet chemical cleans, chemical-mechanical polishing, etc.). However, photolithography cannot pattern a whole wafer at once (at least not at modern wafer sizes), so instead a stepper exposes an identical pattern multiple times across the same wafer.
 * Because of this scaling, whole-wafer steps tend to get cheaper per unit area as the wafer size increases, but lithography-related costs will not benefit in the same way.
 * Does that clarify the issue?
 * Moxfyre (ǝɹʎℲxoɯ | contrib) 20:34, 20 May 2014 (UTC)


 * Absolutely - just misinterpreted it. Thanks for the detailed answer - much appreciated. Justin15w (talk) 21:26, 20 May 2014 (UTC)

Epi wafers
Here are thousands of sources for epi wafers aka epitaxial wafers, in addition to the source(s) in the Epiwafer article:, , , , , , , , , , ,.


 * I'm pretty familiar with epi wafers, just not "epiwafers". I moved the article.  Dicklyon (talk) 06:28, 3 January 2015 (UTC)

Archived link for "Capability for 300 nm" - Exists in archive.is, not in archive.org
The Semiconductor.net source ("Capability for 300 nm..") isn't valid anymore. It appears to be archived at Archive.is, but doesn't seem to exist in archive.org. But the archive.is page says the page was gathered from archive.org, from http://web.archive.org/web/20070615062640/http://www.semiconductor.net/article/CA47551.html. What is the current consensus on what to do with archive.is archived pages? Jimw338 (talk) 03:35, 27 August 2015 (UTC)
 * There is an RfC at Archive.is RFC 4 with the proposal "Remove archive.is from the Spam blacklist and permit adding new links (Oppose/Support)".

450 mm is dead
450mm is dead, at least for the next 5 to 10 years, if not forever. The G450C has shut down and disbanded. The article should be updated to reflect this. --Westwind273 (talk) 20:05, 18 May 2017 (UTC)


 * good, the next generation needs to be 500 mm. Keep it in 100 mm increments.  — Preceding unsigned comment added by 2600:8806:8501:26A0:886:95CB:B014:36F7 (talk) 13:11, 27 May 2018 (UTC)

The case both of you are making is silly. The cost is not worth the investment at this point in time and this is explained in the article. You should at least provide some references.

ICE77 (talk) 17:54, 7 July 2018 (UTC)

Proposed merge January 2018
Die (integrated circuit) is essentially a one paragraph definition, and could easily and profitably be merged here. This merge proposal came out of a suggestion at Talk:Integrated circuit. --Wtshymanski (talk) 20:58, 4 January 2018 (UTC)
 * Oppose (as before, as at Talk:Integrated circuit)  This is just Wtshymanski back to his old disruptive tricks. Andy Dingley (talk) 21:56, 4 January 2018 (UTC)
 * I remain neutral. A merge could work but it is also legit to leave this to develop and see what happens. Die (integrated circuit) is a notable topic and not at risk of being deleted. Die (integrated circuit) has been expanded since this merge proposal was made but much of the new material is actually about Wafer (electronics). ~Kvng (talk) 15:34, 12 June 2018 (UTC)
 * Oppose This is like plank and table - one is the base material, the other the product. --Zac67 (talk) 16:58, 12 June 2018 (UTC)

Comments on the article
1. "The size of wafers for photovoltaics is 100–200 mm square and the thickness is 200–300 μm. In the future, 160 μm will be the standard.[8]"

This sentence seems to claim that the standard, nowadays, would be 1 and 2 inches wafers when we know they are much bigger than that. Also, reference [8] is a dead link. This sentence should be revised.

2. "However, this was not without significant problems for the industry."

This is a vague claim. If you look at a single die, there is no difference if it's fabricated with a larger or a smaller wafer after the wafer is diced. The sentence should provide clarification.

3. "Nikon planned to deliver 450-mm lithography equipment in 2015, with volume production in 2017.[19][20] In November 2013 ASML paused development of 450-mm lithography equipment, citing uncertain timing of chipmaker demand.[21]".

These sentences need to be updated and/or revised.

4. "On the ramp up to 450 mm are that the crystal ingots will be 3 times heavier (total weight a metric ton) and take 2–4 times longer to cool, and the process time will be double.[25]"

Reference [25] is dead (as noted).

5. "Wafers of 200 mm diameter and above use a single small notch to convey wafer orientation, with no visual indication of doping type."

If that's the case then how can you tell the doping?

6. In terms of compound semiconductors, GaN is not the only option. This article does not list SiC which is equally important and promising. It should be listed.

ICE77 (talk) 17:57, 7 July 2018 (UTC)