This PDF file contains the front matter associated with SPIE Proceedings Volume 12807, including the Title Page, Member Companies, Table of Contents, and Welcome Note.

This was a panel discussion that took place at Bay Area Chrome Users Society Symposium 1983.

What is High Technology? We hear a lot about the Silicon Valley Micro Electronics, Mos, gingabis, laser induced micro electronics, but what exactly is it? Webster’s dictionary* indicates that technology is a technical method of achieving a practical purpose. Now that’s a plausible explanation, but it’s still not quite what we’re looking for. Let’s look at the history of American technology.

The title of our topic this morning will be Inter-die vs. Intra-die Feature Size Control. The mask industry is traditionally focused on die to die control, whereas today we’re going to take a closer look at the intra-die feature size control and what role it has to play as a source of error for feature size.

The table of contents of this speech: First, photo mask image and hard surface mask. Second is photo mask blank chrome see through tantalum photomask blank. Third is hot mask making process. Fourth is photomask inspection. Fifth is future hot mask specifications.

Anyhow, basically starting out with the first slide, what I’m trying to show here is that essentially the KLA system, the Klara systems can be easier subdivided into three main systems. One is the data base conversion, that converts other formats into the KLA format. The second, being what they call the RIA (Reticle Inspection Adapter) that converts the KLA format into a bit map representation and sends that to the 201 and the third being the 201 or inspection station that scans the plate under the optics and projects that onto a photo diode array and from there digitizes that and compares that to the data from the RIA.

I’m grateful for this opportunity to discuss Ultratech Stepper lx reticles and the mask making requirements. Two years ago the Ultratech lx reticle was seen as a nearly impossible mask making feat. Today, thanks to talented and determined people, Ultratech lx reticles are being made in volume production to support leading edge devices

Greetings to Bacus members and guests. Thank you for the invitation to be here this morning. When I look at the title I hope it’s not misleading. GCA’s future in mask making is and will be, we are sure. We have a few comments this morning on the future which is the year coming. I want to comment a little bit on some R & D activities that are to some degree underway already. Initially here we are talking about the 3696 photo repeater and looking at some features which introduce elements, which are in fact elements of automation. The Arc II Automatic Reticle Changer and the RMS 1, which is also an automatic reticle changer, but an ackronym for Reticle Management System. The difference between the two, the ARC has a magazine of 10 reticles. It has a surface dust detector, which the reticle is scanned top and bottom before introduction to the platen where alignment is automatic and where there are apertures which close fully from the center of the field outward. These are individually controlled. The RMS1 is a reticle management system that is introduced with the idea of being able to manually handle reticles which are pellicalized. This particular system will handle reticles that have pellicles or it will handle reticles without pellicles.

Two systems are now available that compares VLSI mask design images to their source tapes This will allow for the first time verification of design integrity at the mask or reticle and potentially to the wafer itself before the product is fabricated in silicon. The systems are the Klaris, built by KLA Instruments, Santa Clara, CA, and the Chipcheck, produced by Cambridge Instruments Inc., Cambridge, U.K. In a comparison of the two instruments, the operating methods, each uses the range of application demonstrated, each needed to be highlighted. Also, of interest are the applications these machines will serve that cannot be met today or attempted manually. And as a background to the machine’s applications, I’d like to deal with the area of mask design and creation that can be audited by these systems.

At present, the production of the 64K devices has been accelerated. The R/D manufacturing of the 256K is almost complete and production has begun in the US and Japan. In the making of these types of VLSI, especially the 256K, quartz masks have been predominately used for wafer fabrication, utilizing the wafer stepper and project aligner. Also, the characteristics of the quartz glass with extremely low thermal expansion and high transmittance to DUV light, allows it to be used for contact printing.

I would like to take a few moments this morning to talk about a subject that I have heard alluded to several times during this conference, and over the past several months. And that is this. It would be nice to have a water soluble, water developable material compatible with water soluble etches for electron beam applications and indeed this has been looked at over the past several years. I would like to share some thoughts with you on the chemistry of those systems and why the AZ materials perform the way they do in the electron beam machines. And some possible solutions to those problems that we, as well as several other companies, are actively pursuing.

Good afternoon. The paper that I’m going to present this afternoon, entitled “An Overview of Magnetron Sputtering”, will examine briefly some of the sputtering techniques other than Magnetron sputtering, lead into a discussion of post and planar Magnetron technologies. Some comparisons will be made of the above mentioned techniques, their advantages and disadvantages and a discussion of their production capability with respect to process repeatability and product through-put.

Recently, I’ve been involved in our program in AMD to come up with design rules for the next generation of bipolar products. For those of you who have never been involved with coming up with new design rules, what it consists of is everyone who is responsible for a process and technology, getting together and blaming the other guy for not being able to do as well as we’d like. And what I’d like to do today is talk about how we can, through improvement in the quality of reticle, take some of the burden off the lithographer.

Thin Films were originally produced by the evaporation of material in vacuum. The heating and evaporation of material was and still is achieved e.g. by putting the material into a boat of high melting point material. By passage of electric current the boat is then heated to a temperature high enough to evaporate the material in the boat. A more modern method is electron beam evaporation. The material to be deposited is heated by an intense electron beam. In sputtering the ejection of material from a surface (e.g. chrome target) is achieved by momentum transfer. This can e.g. be done by accelerating argon ions towards said surface. There are many methods to produce energetic argon ions. Hence many different types of sputtering systems were developed e.g. diode sputtering, triode sputtering, post magnetron sputtering. The planar magnetron as sputtering source brought the break-through for sputter deposition as production method for thin films.

The first Murphy’s Law is that all photo science graduates from RIT typically think they know more than they’re speaking about, because no one understands what they are talking about anyway. So it leads them into a falsehood of understanding. But in either case, bear with me for a few moments as I discuss and go through the history of sensitometry as it deals with silver photographies films. And that will lead into the discussion of photoresist.

The purpose of my paper today is to increase the survival rate of the technician and the engineer in the photo mask shop.

You recall that earlier this morning Reed Welsh mentioned that, with the tolerances that we’re experiencing today in our product, any kind of a difference between measurement standards is important. To go ahead and document our study to this group, we used the Nikon NPA’s that were in our mask lab. As you know you can use any kind of a standard to calibrate the machines. We used the Roger Sherman standard originally, and we went to use the NBS standard for this one customer. As you also might be aware, a line width measure to Roger Sherman three remeasured at NBS will give you two different values. We had two NBS standards in our lab and that they both tracked each other very well. So I’m just going to group those two together for this presentation.

This is the closing address of the Bay Area Chrome Users Society Symposium 1983.