Talk:Stepper
![]() | This article is rated C-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||||||
|
First start
[edit]I wrote this article, and all and all I think it's a good start. Here are some needed improvements:
- Diagram showing the major subassemblies of the stepper and their relative positions with respect to the light path.
- Explanation of the types of light used, such as mercury lamps and excimer laser--Done
- Detailed sub-section under the "Major subassemblies" section for each subassembly
- Free license picture of a stepper
- Free license picture of a wafer showing grid pattern
- Microscope picture of circuit patterns and lines
- Notice that steppers may soon be superseded by other machines as design node goes below 45 nm--Done
- Discription of various techniques for finer resolution such as higher lens NA, manipulating the illumination light through various aperatures, phase-shift reticles, and others.--Started, not complete yet.
This list is just a beginning. Everyguy 18:43, 18 August 2005 (UTC), updated: --Everyguy 19:09, 15 March 2007 (UTC)
Some edits
[edit]I edited the first couple of paragraphs of the article, but it wasn't logged in, so it came out with my ip address rather than my user name. That is okay, because I left some spelling errors I had to go back and correct. This article still needs illustrations and more details.
Everyguy 06:38, 26 April 2006 (UTC)
Article naming
[edit]I propose to move this article to Projection lithography, and rewrite the introduction accordingly. --Smack (talk) 05:03, 14 February 2007 (UTC)
- I'm not sure that's a good idea. I think this article is about the machine, not the technology, although there is naturally overlap between the two. I would propose instead that an article on projection lithography be written that covers the more general topic, with links to this page in the appropriate section. --Everyguy 20:26, 2 March 2007 (UTC)
Inaccurate/missing information
[edit]I read through this article and noted several inaccuracies.
- I have addressed the various points raised below.Everyguy 21:37, 30 November 2007 (UTC)
A. Scanners such as PerkinElmer scanners, although they are used for photolithography, are not really steppers. Scanners are used to produce lower resolution images on the wafer, whereas stepper are always used to produce high resolution images. Scanners are faster (than steppers) however, so most fabs will have both scanners and steppers.
- Modern scanners are not the old PerkinElmer types. The old scanners scanned the entire wafer at once. Modern scanners scan each shot. They are the mainstay of industry that makes devices with the highest design rules. They are steppers, with the added function that they scan at each exposure site on the wafer to increase the area that can be exposed in one shot. An example can be viewed at http://www.nikonprecision.com/products/nsr_s610c.html
-- Inaccurate. Scanners were developed to reduce the size of the lens, to improve focus/leveling over the wafer surface, and to average lens aberrations through the scan. Almost all current model steppers and scanners have a 26x33 exposure field. But on a scanner, the light field is only 9x33, scanning increases the field to 26x33 without an equivalent increase in the reduction lens size. Also, steppers have higher throughput than scanners. It takes less time to expose the entire field, than it does to scan it.
B. The statement "Currently (as of 2006) the most detailed patterns in semiconductor device fabrication are transferred using a type of stepper called a scanner, which moves the wafer and reticle with respect to each other during the exposure, as a way of increasing the size of the exposed area." is incorrect as it appears to describe a Perkin Elmer type of system--which is not really a stepper. Also, these types of system are used to produce relatively low resolution images. The most advanced steppers to my knowledge are ones that use KrF lasers--Cymer is primary manufacturer of these kinds of excimer lasers in the US. Both Nikon and Canon are now making thier own laser systems based on the reverse engineering of the Cymer laser. Most steppers are UV light based steppers, the most common light source for these types of steppers is the mercury arc lamp (i-line and g-line steppers).
- As I mentioned above, most of the current generations of steppers also scan at each shot (and are called scanners). They use KrF and ArF light sources. Nikon (and probably Canon and AMSL) do not manufacture the lasers. They buy them from Cymer, Komatsu, or Gigaphoton, the main makers. I-line steppers are still used for non-critical layers in some processes.
-- Komatsu is actually Gigaphoton. Also as mentioned above, the scanning function is not to increase the exposure field.
C. Stepper comes from the term DSW (Direct Step on the Wafer). This was the term was used to describe early GCA systems. The term was shortened to just "stepper".
- This may be the case. I'm not familiar with where the term came from.
D. GCA Corporation pioneered the first wafer steppers and no mention of them is made in this article.
- Information like this and item C above would be appropriate in a History of Steppers section in the article, if you would care to write one.
E. No mention of whether the resist process for steppers is positive or negative-- or what the difference is.
- The positive or negative quality of the resist really doesn't affect the stepper. This issue is discussed in the Photolithography article.
F. No mention of the yellow light (It is non actinic to the photo-resist)
- This issue is also not really connected with the stepper, but part of the photolithography process.
G. The descripton of a stepper process is inaccurate because it describes a scanner process (PerkinElmer). The basic process for a wafer stepper process is roughly as follows.
1. A reticle or photomask of the die is placed above the lens of the stepper. A reticle is a glass plate with a chrome outline of the die. The reticle represents a particular overlay layer in the chips architecture. A single chip type may have dozens of layers and hence dozens of reticles
- This is useful information, and the article could be clarified on these points.
2. The lens will reduce the image of the reticle 4X. The light source above the reticle is typically in the UV range. This is usually generated by a mercury arc lamp or a KrF laser. This light source will expose the photo resist on the wafer. The light that hits the wafer will be washed away in the development process (positive development).
- The article already mentions this. Modern steppers can be either 5x or 4x.
-- No, all modern steppers and scanners are 4x reduction lenses. The last Nikon to have a x5 reduction lens is the i14 and EX14.
3. Once the photo has been taken on the wafer, the stages on wafer stepper system will move to the next exposure location. The stages are said to have stepped to the next location. this continues in zig zag pattern until the wafer is entirely exposed.
- The article does mention this.
4. The wafer is then developed and sent out for etching and other processes such as doping.
- The article does touch on this, although I think the details rightly belong in the Photolithography article, because this is happening away from the stepper.
5. Eventually the wafer makes its way back to the litho room for subsequent layers. The process of matching a new layer to a previous layer is called overlay or registration. When the wafer comes back to the stepper, it will use various alignment methods to make sure the registration process is performed to very tight specifications.
- Overlay registration is very important, and the article could be improved by someone writing a section that describes this process in detail, although this may be difficult because the systems are proprietary to each stepper maker, and the details are not generally published because of patent concerns.
In short, I would suggest revising this article to remove all reference to "scanners" because as I mentioned, these are not steppers. Since wikipedia does not have an entry for PerkinElmer or photolithography scanners, I would move that information to those sections. I would also suggest that GCA be noted and possibly a new section on GCA be created since this wiki has no entry for this former company. The section on how the how the stepper works needs to be heavily revised. The other sections in this article appear to be accurate. —Preceding unsigned comment added by Technogrep (talk • contribs) 09:12, 27 September 2007 (UTC)
- Several suggestions made in this comment could improve the article as I mentioned above. However, the reference to scanners should stand because the most modern steppers are actually step and scan systems, which are known in the industry as "scanners." A separate article on GCA could be created to describe the company and its history. A section on stepper history could be added to the Stepper article, and could include how GCA created the first stepper. Details about resist are covered in the Photolithography article. Everyguy 21:37, 30 November 2007 (UTC)
-- Nikon produces steppers. Currently the most advanced being the SF155. The most advanced scanner being the S620, which uses immersion technology.
-- The article is also inaccurate in that the pictures it shows are of mask aligners. -- Comments from Nikon service engineer with 12 years field experience.
Spam
[edit]Please see Talk:Stepper_motor#Spam. - Mdsummermsw (talk) 19:51, 28 December 2007 (UTC)
Image for Article
[edit]The image used for this article shows two contact printers. These are not steppers by any stretch of the imagination - the information is false and misleading. —Preceding unsigned comment added by 152.14.125.113 (talk) 20:37, 13 May 2011 (UTC)
- Also just noticed this, agreed. Will replace them shortly. a13ean (talk) 20:37, 13 September 2012 (UTC)
"Twenty years ago, the ultraviolet "g-line" (436 nm) of the mercury spectrum was used..."
[edit]I'm no expert on any of this, but isn't 436 nm just plain violet? Zyxwv99 (talk) 14:08, 11 May 2013 (UTC)
Price of such a machine
[edit]I just read that those machines can cost up to 130 million USD, which sounds unbelievable. Maybe this could be an interesting information for the article. Renek78 (talk) 09:19, 23 February 2019 (UTC)
Clarification in the Illumination section
[edit]"Since practical light sources with wavelengths narrower than these lasers have not been available,..." Should this say "linewidths" rather than "wavelengths"?
Aligner ref
[edit]Dleebrick added
- "Introducing a Mask Aligner Capable of Submicron printing by Projection", Solid State Technology, Vol. 18, #12, pg 18 (December 1975)
to the end of the article. It is not clear what this is supporting so I've moved it here for review. ~Kvng (talk) 16:07, 4 December 2022 (UTC)
- This Wikipedia article is about wafer steppers. This early SST article describes one of the early wafer steppers that had submicron printing capability. Dleebrick (talk) 19:38, 4 December 2022 (UTC)