|
"There is no lack of competition" in FEOL cleaning market, says Dataquest's Freeman |
|
|
|
Nov 29, 2006 at 11:15 AM |
There are lots of challenges---and opportunities---in advanced FEOL wafer cleaning, and a whole lot of competition among the equipment suppliers to provide solutions.
With the launch of SEZ's Esanti system, the competitive space just became more interesting. I posed a series of questions about the new SEZ tool and the FEOL cleaning market to Gartner Dataquest's Dean Freeman via email. Here is an edited version of his replies.
How does SEZ's FEOL tool/approach stack up against the competition?
That is a bit difficult to say, since it is just being released into the market. At this time there really isn’t a FEOL market leader for single-wafer clean tools as the market is fairly small. The [SEZ] tool fits in a bit between the Semitool Raider, which can have, I think, up to 10 stations and the DNS tool, which I believe has five chambers; the SEZ tool can be configured in either a four- or eight-chamber configuration. SEZ is probably the most versatile single-wafer-cleaning company on the market since they have products in both the silicon, front-end, and packaging markets. If the [new] tool is as successful as the DaVinci, I would expect it to be an extremely competitive tool.
What other competitors are making a solid FEOL play?
Round up the usual suspects. Semitool, DNS, and Akrion have all made some form of product announcements. AMAT has its tool and TEL has a single-wafer tool that will be able to be used in the FEOL segment. SES has a product, and there is Solid State Equipment Co. So there is no lack of competition.
What do you hear that the chipmakers want in an FEOL cleaning tool?
They want a tool that will replicate the success they have had with the automated wet stations. Low cost of ownership, high throughput and a minimum number of headaches. They need systems that will clean photoresist and metals and remove particles, thus generating high yields while keeping damage to a minimum.
Which FEOL applications are most in demand?
None yet. The applications that have the most problems are those with fine polysilicon structures that need cleaning. The megasonics are creating damage and, according to most sources I speak to, have been turned off at 90 nm. So a solution needs to be found for removing resist, and then removing particulates that will not damage the fine structures.
Which applications are the most problematic, challenging, etc., and why?
The most difficult will be high-dose implanted resists since they form a crust that is difficult to break through, and when it does come off in a wet clean application, it comes off in clumps. The photoresist removal processes will need to be able to dissolve and remove the resist in such a way that resist particulates are not an issue. You also have the issue of metals removal, [but] this should not be too difficult to accomplish in a single-wafer tool, and this is an easy entry for the tools, especially if they can save on chemical costs.
What is the ultimate extendability of the wet cleaning/wet chemical approach? Will we ever see all-wet-chem cleaning, or will dry keep on keeping on, or even grow?
I think dry strip will be with us for a long time. There might be some specific cleans that wet chemistry is able to perform better than the dry strip. The key is, can the wet cleans remove the photoresist in an economical manner and not have particle issues due to the clumping problems above? For the foreseeable future I think you will see the two go hand in hand. If the wet cleans prove to be more economical and meet the specs, you would see a switch, but at this time dry strip is a cheap process and using dry eliminates a source of concern. There are one or two processes where the etching of the silicon is a critical factor. If the dry strip folks don’t solve this, then you could have one to two processes move to just wet if both the technology and the economics match up.
What are the biggest obstacles or challenges to SEZ making a successful foray into the broader FEOL sector? Do they face any especially sticky sticking points?
Automatic wet stations. Convincing the fabs that a single-wafer tool can operate as economically as a wet bench. A good example is that most of the industry use single-wafer tools for dual-damascene cleans. The exception is Intel, which is still using batch wet stations. Single wafer has been proven to be cleaner and more economical but Intel still looks at its COO [cost of ownership] and sticks to the wet benches.
What is the size of the overall cleaning/stripping/etc. market; and what is the size of the FEOL segment that SEZ is going after now?
The total wet-clean market is $2.2 billion, dry strip is $433 million. The FEOL segment is forecast in 2006 to be $1.1 billion, broken down into $440 million as resist strip and $414 million as critical/prefurnace cleans.
|
Blogs 
