When molecular contamination technologist David Soules and his colleagues at the Jet Propulsion Laboratory, Pasadena, Calif., were asked to develop a complicated ultrahigh vacuum (UHV) chamber for measuring the effects of molecular contamination on sensitive optical surfaces, they enlisted two consultants to help with its design and then submitted a complete set of CAD drawings to Huntington Laboratories, Mountain View, Calif. (800227-8059), for construction of the chamber and most of its subassemblies.
Can't afford to hire consultants to help you design your UHV chamber? Couldn't produce an AutoCAD drawing to save your life? Not to worry--you're actually the more typical custom chamber customer. Chamber manufacturers are well-versed in walking people like you through the specification process.
"Customers come to us with anything from fully engineered CAD drawings to sketches on the back of the proverbial cocktail napkin," says Tom Bogdan, technical sales manager at MDC Vacuum Products Corp., Hayward, Calif. (800-443-8817). "The back-of-the-napkin chambers can be more fun because you get to design them with the customers. Often we're working with scientists or graduate students who know what they want to do with the chamber, but can't decide how it should be constructed. We try to give them design ideas that help them achieve their goals."
Although you needn't be an expert, you should arrive at the chamber manufacturer's door with some firm design ideas in mind. Here's what experts say about how your design process should proceed and the pitfalls you should avoid along the way.
"Due to its complexity and importance to the overall functionality of the complete system, the first component to be considered should be the sample handling system," advises Michael Ricks, VP of sales and marketing at Thermionics Vacuum Products, Port Townsend, Wash. (800-962-2310).
Do think about "how the sample will be manipulated within the system from load lock introduction to processing, surface analysis, and end-use," Ricks says. Include in your deliberations "the size and number of samples to be processed at any one time, how the samples need to be manipulated, whether the samples will be able to reach all of the required stations at the required angles, and whether the samples will need to be heated or cooled to a specific temperature."
An example of the problems you can get into if you don't give careful thought to sample handing was provided by a company who approached MKS Vacuum Products Group, Boulder, Colo. (800-345-1967), to design a chamber that would operate more cost-effectively than the one then in use. The company is in the business of supplying specialty coatings for mirrors.
"They were placing a small tray of mirrors in a chamber as large as an office," says MKS special products engineering manager John Hanzelka. "It would take half a day to pump the chamber down. If the products needed to be rotated during the process, the chamber was vented, the products were moved by hand, and the chamber was pumped down again."
The engineers at MKS put together a system that included an antechamber, a process chamber, and an automated feed track. The antechamber takes only 15 min to pump down while the process chamber remains at [10.sup.-8] torr. The automated feed track transfers the products to the process chamber and manipulates them while there. As a result, coatings are now applied more consistently and with dramatically less downtime.
Do concern yourself with the requirements of the processes that will take place inside your chamber. These requirements include such considerations as base pressure, operating pressure, sample throughput, and film thickness and quality if it is to be a deposition system. In this ease, you also need to decide the type of deposition source and the most appropriate means of film thickness monitoring.
These issues, taken together, define what will go inside the chamber, making it a fairly straightforward task to design the envelope that will enclose everything. Port placements are next for such things as ion gauges, sample and source shutters, vacuum pumps, feedthroughs, and analytical instruments--to meet both present and foreseeable needs. All that remains is a final check of the complete package. The check should ensure compatibility of all of the components, including those you intend to add after the chamber is delivered to your door. You should also check to make sure that your system is easy to maintain.
"Customers often don't think of maintenance, but maintenance problems become very obvious when they get the system in their lab," says Mike Dwyer, director of engineering at MeiVac, San Jose, Calif. (408-362-1000).
Now comes the hard part. You need to critique your overall design from a manufacturing standpoint, because in all probability the design in its present form is either unmanufacturable, horrendously expensive, or both.
Don't overspecify. Since research chambers are inherently expensive, it makes perfect sense to design them to be as versatile as possible--up to a point, that is. As Dwyer puts it, "More is not always better; a lot of times it is impossible." He was thinking of a researcher who needed a chamber for a process that takes place at 700 [degrees] C, but who wanted to specify the chamber for 1400 [degrees] C, a temperature that probably would have melted many of the components inside even if the chamber had double-wall construction for cooling.
Chamber geometry is another area that leads people into financial difficulties. If you really need a rectangular chamber, fine, but expect to pay a premium for it. That's because spherical or cylindrical walls handle pressure much better than flat plates. So the walls of a rectangular chamber need to be relatively thick.
Don't worry about permeability as far as wall thickness is concerned. Concern yourself only with chamber stability, says Steve Greuel, UHV product manager at Nor-Cal Products, Yreka, Calif. (800-824-4166). "You want the chamber to remain stable so that port positions do not change." Unless you have extraordinary stability requirements or need a rectangular chamber, the standard wall thicknesses manufacturers offer will do just fine.
The wall thickness you require is, of course, a function of the wall material you choose. Normally that would be stainless steel, but many people nowadays are turning to aluminum. "The demand for aluminum has increased by leaps and bounds," says Jeff Vaudreuil, marketing manager at A&N Corp., Williston, Fla. (800-FLANGE1). "One of the reasons is that aluminum is lighter and somewhat cheaper," Vaudreuil continues. "But a more important reason is that aluminum dissipates heat faster and dissipates radioactivity much more quickly, too. That is a real plus."
Do specify aluminum rather than stainless steel when working with radioactivity. Aluminum dissipates radioactivity quickly in the sense that the aluminum radioisotopes that are formed have a much shorter lifetime than steel radioisotopes would.
Something to keep in mind if you do specify aluminum is that welding aluminum for UHV is tricky business, so you should choose a vendor who has experience doing it. Another consideration is that an aluminum chamber means aluminum Conflat flanges. You should specify flanges whose knife-edges have been coated to make them hard enough to resist bending over when biting into a copper or aluminum seal.
As far as stainless steel is concerned, you have two basic choices--304L or 316L. Deciding between them depends on the final base pressure that you will be operating at. "The majority of people I work with still use 304L grade, and that will get them down to [10.sup.-10] torr, assuming that they're using metal seals," says Nor-Cal's Greuel. "If they're going lower than that, we advise people to specify 316L grade along with vacuum firing."
Vacuum firing involves placing the chamber in a vacuum furnace and holding it at a temperature of about 950 [degrees] C for several hours. The idea is to drive out hydrogen and water vapor and anything else whose outgasing characteristics would inhibit the chamber from achieving a very low base pressure in a reasonable time.
Don't apply tight tolerances across the board, experts say. The premium you pay is not related to machining to such tolerances, but to the fact that welding changes part dimensions slightly. Tight tolerances generally mean remachining parts after the welding step, which can contribute significantly to a high price tag. Of course, if a tight tolerance is critical or if welding makes a sealing surface no longer fiat, the extra cost is unavoidable.
Avoiding tight tolerances is particularly tough if you are designing a chamber as an OEM. But even so, you should scrutinize your design for opportunities to do so. "This may entail slight design changes, but it may just mean changing some tolerances without moving anything," says Nor-Cars Greuel. "If a tolerance requires additional machining after welding, you want to find out whether that tight a tolerance is absolutely needed. If it isn't, it is costing you money to leave that spec there."
Now that you've finished your design, you need to decide where you will get the flanges, fittings, feedthroughs, manipulators, valves, traps, heater jackets, etc., that will turn your chamber into a complete system. Your answer to this question will not only determine who you choose to manufacture the chamber, but will greatly affect the overall price you pay as well. All chamber manufacturers want to sell you more than just a chamber, and they want to do so badly enough to offer you components at cut-rate prices. What they will offer you, however, will be restricted to their current product line. So you will find that some vendors will be able to offer deposition sources and complete downstream contamination solutions, while others will not. Generally speaking, a components vendor will not provide you with a turnkey system if doing so would mean going to other vendors for some parts.
There are two reasons for this. The first is that they do not want to infringe on customers of theirs who are in the business of making turnkey systems. Secondly, as A&N's Jeff Vaudreuil says, "We try to stay focused on our areas of expertise because we do not want to get into a situation where we could not support the customer 100 percent."
To obtain a turnkey system, you need to go either to a system manufacturer who buys components from many vendors or is associated with a general vacuum equipment company. The Modified Standard Products Group at Varian Vacuum Technologies, Lexington, Mass. (781-860-5407), for example, is in the business of making custom turnkey systems.
Turnkey-system vendors also would prefer to sell you a complete system, rather than merely a chamber, so they are generally willing to give you a deal on the system's overall price. "Customers have found that it is pretty cost-effective to allow us to put a system together for them," says Varian product manager Dave McCarthy. "Because of our expertise, we can usually put a system together more cheaply than they could."
Do make one last decision: Whether you want the system shipped to you under vacuum. If you will be using the system to work at extreme ultrahigh vacuum levels, you might want to specify that the vendor pump down the system to [10.sup.-8] or [10.sup.-9] torr and ship it to you in this condition.
RELATED ARTICLE: Special Enclosures Enable Special Research Capabilities
Special research facilities have special requirements--so Tom Bogdan, technical sales manager at MDC Vacuum Products Corp., discovered when he sat down with an engineer from the Basic Energy Sciences Synchrotron Radiation Center Collaborative Access Team (BESSRC-CAT). BESSRC-CAT performs experiments using a wide variety of techniques at two sectors of Argonne (Ill.) National Laboratory's Advanced Photon Source (APS). "One sector consists of an undulator and a bending magnet, with four experimental stations," explains Pedro Montano ([email protected]), who is director of the CAT. "The other sector has an elliptical multipole wiggler and is split into three beamlines with three experimental stations." The undulator intercepts the APS's high-brilliance x-ray beam and converts it into a series of highly focused, discontinuous high-energy pulses, while the wiggler creates a continuous fan of circularly polarized radiation covering a spectrum of energies.
Between the insertion devices and the experimental stations are versatile double-crystal, fixed-exit monochromators, which BESSRC-CAT designed with help from MDC. Each monochromator contains two crystals affixed to a rotatable stage mounted on a backplane. To dissipate the high energies produced by the APS, the crystals need to be cooled with liquid nitrogen, which created special requirements for the backplane.
"One of the constraints was that we needed a very stable backplane for the monochromators," says Mark Beno, deputy director of BESSRC-CAT. "Most chambers are made out of metal that is very thin, but these needed stainless steel backplanes that are a couple of inches thick."
An additional feature was added by members of other CATs, who wanted similar monochromators that provided easier access by means of a hinged door. "MDC did the engineering, so now some monochromators here have a hinged door rather than the door we put on," says Beno.
Partly because of the versatile monochromators BESSRC-CAT pioneered, BESSRC-CAT users can now avail themselves of numerous experimental techniques, including x-ray absorption fine structure, x-ray scattering (bulk and surface), small-angle x-ray scattering, single-crystal diffraction, powder diffraction, high-energy diffraction, Compton scattering, and magnetic scattering. These capabilities serve researchers interested in materials science, chemical science, atomic physics, solid-state physics, and geoscience.