Vacuum tools morph to meet the new demands of research.
Many of today’s turbomolecular pumps, like these partially assembled examples from Oerlikon-Leybold, must be designed for demanding high-throughput tasks and with a minimum of size and maintenance. Photo: Oerlikon-Leybold Vacuum
The high-vacuum market is difficult to comprehend from a brief glance. Segmentation is substantial, as would be expected from a billion-dollar industry. But what really challenges manufacturers of this type of equipment are the differing requirements of the customers.
In industrial applications, for example, customers require vacuum for anything from electron tube construction to heat treatment. In semiconductor development, high-vacuum is necessary for deposition, etching, lithography, and thin-films processes for photovoltaics and LCD displays. Non-semiconductor thin films technologies, such as ceramics and optical coatings, are also widespread and require vacuum-assisted deposition. Finally, the demands of the process vacuum and instrumentation manufacturers must be addressed.
As a result, vendors are moving past componentry and into turnkey solutions. They are also providing new technology and new architectures to meet these needs. For a laboratory not normally accustomed to working with a vacuum system, the logistics of installing a system may be difficult. And for certain market segments, such as the aforementioned analytical instrumentation, the very format of the instruments themselves is changing radically.
All-in-one vacuum clears the benchtop
For many vacuum pump manufacturers, solutions integration used to mean simply giving customers a wide variety of pumping options.
Such an approach no longer suffices. Lab researchers are strapped for time and dollars, and fewer potential customers wish to start from scratch in building their own systems. Granted, many users will have both the expertise and experience to build a cost-effective, low-maintenance vacuum system carefully designed to suit their needs. But not every customer deals daily with high-vacuum systems, and an increasing number require quick solutions for analytical, biotechnology, and emerging materials R&D areas.
For these users, vendors have begun to diversify their offerings.
Oerlikon-Leybold, Export, Pa., is a good representative of this new market philosophy. Long perceived as a component manufacturer for vacuum systems, the company is developing—through its solutions division—several new product types that may increase the usability of high-vacuum without compromising its performance.
First up is VACVISION, a centralized controller platform that unites many of the company’s components under a single programmable OS aegis.
The VACVISION system, says Brett Rock, market support manager, is still in the introductory phase, but it represents a big step for Oerlikon. “We are still in the process of tweaking the software now, but there will be so much more we can do with this once the basic package is complete.”
That basic package is a touch system interface and a custom operating system designed by Oerlikon engineers. The controller is designed to automatically detect up to three active vacuum gauge heads, five valves, one forevacuum, and one high-vacuum pump—and control them simultaneously. Designed with process automation in mind, signals from the forevacuum transmitters can be used to start up the turbomolecular pumps and the high-vacuum transmitters. Users can also connect venting, seal gas, heaters, and cooling at the turbo pump. The unit communicates via RS-232/485, Internet, and USB input for software upgrades and process data transfer.
“The first VACVISION you see is a foundation,” says Mike Ridenour, sales support manager at Oerlikon-Leybold, referring to the prototype example the company displayed at the Pittcon 2010 in Orlando, Fla., in March.
“The software will follow the same design and package,” he continues. “This version does everything. It controls vacuum, controls for the pumps, gauges, valves. All we have to do is cut out certain parts and give it a smaller box. In the future, as things grow, this will be a customizable system.”
The second innovation is actually a second-generation product for Oerlikon-Leybold. The previous iteration was the BMH70, which had a different type of mechanical pump. When Oerlikon-Leybold upgraded its turbopump line to the TURBOVAC SL 80 pumps, they also updated this particular systems solution, creating the TURBOLAB 80. According to Rock, the instrument has only been shown at a couple of smaller tradeshows and Pittcon 2010, and has been on the market for a year.
Nevertheless, the portable, self-contained high-vacuum system has proved popular with a variety of customers, in part because of its useful performance in a small, convenient package. It is supplied in a CF or ISO-K flange, and is equipped with a 0.8 DIVAC diaphragm backing pump and the new SL 80 turbomolecular pump. The wide-range turbo pump has an integrated frequency converter, air cooling, ceramic ball bearing, and splinter guard. High-vacuum connections are made with DN 63 ISO-K or DN 63 interface, and pumping speed for nitrogen is 65 l/s. The backing pump features dual-stage oil-free operation that achieves a pumping speed of 0.7 m³/h and an ultimate pressure of less than 3 mbar. The power supply is a TURBOPOWER 300 that offers a 24 VDC supplement to the frequency converter. Vacuum gauges and a venting valve can also be installed. It’s a completely dry system.
“The nice thing about this is the portability. Lots of laboratories need a small vacuum system and you don’t want it stuck in the corner where you can’t use it,” says Rock.
The nEXT turbomolecular pump line from Edwards has a new bearing design that does not require vibration balancing during replacement, a design step that simplifies maintenance. Photo: Edwards Vacuum Inc.
At 255 mm wide, 355 mm high, 355 mm deep, and weighing less than 15 lbs., the TURBOLAB 80 is lightweight and shaped like a breadbox. Equipped with a handle, it’s easily moved around and comes in two versions. The full-featured model has a PLC-controlled display and can be built up with gauging so that pressure readings can be obtained. There is a definite element of flexibility that designers wanted when packaging the TURBOLAB 80; yet, size and portability seemed to be the focus.
“As we know in every lab, size is a premium. What we wanted was to have a box that splits in half that you can easily work on, easily repair, and that is also small and has a streamlined look,” says Ridenour. He is familiar with the development of both the BMH70 and TURBOLAB 80 systems as they were both engineered at the company’s Export location using vacuum technology from Leybold Vacuum’s catalog. The development was done entirely in-house, highlighting the company’s system solutions capability. The BMH70 was engineered in the U.S.; the TURBOLAB 80 is made in Cologne, Germany, based on the U.S.-generated design.
According to Ridenour, R&D laboratory customers will typically use the TURBOLAB 80 to pump down dewers, form residual gas analyzers, or use them for simple alignment of gauges down to pressures in the 4, 5, or 6 mbar range.
“It’s a relatively universal tool, and relatively cheap for R&D labs that do these sorts of things,” says Ridenour.
Oerlikon saw from the market performance of the BMH70 that sales to the university market were strong, and according to Ritenour, the TURBOLAB 80 looks to appeal to a similar customer base. Government laboratories such as Argonne National Laboratory and Fermilab have shown interest in this pump, Rock adds. Anywhere space is a premium, an integrated, turnkey vacuum solution, he says, will be appealing.
“This really does fit our business model with this solutions group to respond to customers’ needs when they need a package like this,” says Ridenour. “If a customer wants to put another widget in these types of units we have the ability to do that here.”
A new direction for standalone turbopumps
Size for size, turbomolecular pumps are among the most efficient, capable, and flexible pumps available to the laboratory researcher. Able to supply mid-range to ultra-high vacuum performance, these high-speed pumping solutions are also one of the best options for efficiently pumping light gases, a task common to R&D labs. Their compact size allows them to be fitted into a variety of analytical equipment, from spectrometers to electron beam microscopes.
On the flip side, they tend to be quite expensive. The design of the pump requires high rotational speeds that put wear and tear on moving components, namely the bearing. Bearing technology varies, but typically a grease or oil lubricated ceramic ball bearing is used, often in conjunction with a permanent magnet bearing.
Nevertheless, after a fixed number of hours of operations, a turbopump must be serviced. Typically, this means either significant downtime as the user waits for a pump to return from the servicing, or the added expense of a backup system.
Vendors, as a result, are vying for longer service intervals and more robust components.
Pfeiffer Vacuum Inc., Nashua, N.H., for example, introduced its HiPace line in 2008 that featured an optimized construction and integrated control electronics to help extend service lifetime and reliability of its pumping technology. Other manufacturers such as Oerlikon Leybold have also improved flexibility through the addition of adjustable control panels and a maintenance-free ceramic ball bearing.
For analytical applications, the demands for performance, throughput, and reliability are even higher. The pumps must provide high flow of light gases and be robust enough for a minimum of servicing. Fitted into a larger enclosure, the pumps must also have a streamlined profile but capable control electronics. That’s a tall order that vacuum vendors have been responding to in a variety of ways.
Pfeiffer, for example, manufactures the SplitFlow 50, a turbopump that features multiple inlets. Supplying up to 53 l/s of flow, the design is unique in that it can differentially evacuate separate chambers, for example, the analysis and transfer chambers for a mass spectrometer. Depending on the gas load, says Pfeiffer Product Manager for High Vacuum Pumps Edward Ho, a small dry diaphragm pump can back a split-flow system for a significant reduction in costs and footprint. Also suitable in leak detectors that work on the analytical principle such as mass spectrometry, the SplitFlow’s modular design is geared for OEMs.
Bearing technology: The biggest change?
Another prominent turbomolecular pump manufacturer, Edwards, Tewksbury, Mass., headquartered in Sussex, UK, recently introduced its own solution to the problem of turbopump serviceability with the nEXT range. Though it resembles its predecessors, the nEXT pump is a major departure for the company in terms of fundamental design philosophy. And, like turbopumps from other vendors, the nEXT has seen plenty of demand already from mass spectrometer manufacturers.
“Historically, when someone needs a new turbo pump, they generally go back to the drawing board whenever they need something slightly different,” says David Steele, product marketing manager, Diffuse, at Edwards. The new pump architecture allows Edwards to be flexible with the design of the actual pumping mechanism.
There are three main sizes of turbopump, each with multiple variants. There are also different combinations of turbomolecular, drag, or viscous mechanisms available depending on the application.
There are several patents pending on the nEXT. These are associated with the motor, three with the new bearing suspension arrangement and pump envelope optimization. Other existing patents such as booster ports are also being utilized. But the core building block in terms of reliability, says Steele, is common to all of the new nEXT pumps: the drive, motor and bearing mechanism.
“From an end user’s perspective, the ability to service the bearing is a big change from anything in the past, taking minutes instead of ordering a service exchange unit or having to send the pump back for a service as was the case previously,” says Steele.
The HiPace Splitflow from Pfeiffer has multiple compression chambers and is capable of replacing several turbomolecular pumps for certain applications. Photo: Pfeiffer Vacuum
All mechanical bearing pumps have lubrication of some sort, and periodically that needs to be replaced. Typical bearing lifetime is in excess of four years. The nEXT pump is designed to tell a user when it needs a service and what kind of intervention is required. After two years the oil cartridge will need replacing and, two years after that, the bearing itself will also need to be replaced. Both of these services can be performed in situ assuming the base of the pump can be easily accessed. In addition to commonly available workshop tools, two inexpensive specialist tools can be used to swap the cartridge in minutes. This allows the user to be back up and running quickly.
“In previous technologies of turbos, the bearing itself has to be suspended in an adapting mechanism to absorb high-frequency vibrations and prevent rotor harmonics causing damage to the pumps,” Steele says.
Edwards has gone away completely from using an elastomer, the most common format for mechanical bearing pumps, and moved to a compact metal spring damper.
It does two things: it holds the bearing in place and fulfills the function of the damping mechanism. It also means the pump is completely self-aligning.
“It’s simply a case of taking the bearing out and putting a bearing in. There’s no need to test, check, and adjust for vibration in the turbo rotor itself,” says Steele. “I’m not inclined to say it’s as easy as changing a light bulb, but it isn’t far off.”
The cost of ownership for turbopumps is often high because maintenance is both expensive and inevitable. Sending a pump back to the vendor is usually the best option a user has because few users are going to have the training, tools, or time available to align a new rotor after a bearing installation.
Worse, there’s the additional cost involved in shipping the pump back to the vendor. And as for the service itself, it can run to a third or a half of the cost of the pump in the first place. Reducing cost of ownership is a strong driver for Edwards in developing this new technology. Moreover, there are a number of installations in which sending a pump back to the vendor is simply not possible, such as facilities that put restrictions on shipping components off-site.
“For the most part, if someone is using one of our 24-volt turbo pump they simply replace it with the new nEXT pump,” says Steele.
The rotor itself is different, too. Essentially, because the bearing is suspended with a solid damper rather than relying on the movement of the elastomer, Edwards has been able to adopt a new drag mechanism.
The pump is available in three architectures: simplex, duplex, and triplex. This refers to the number of mechanisms in the rotor stack. A simplex pump is a straight turbopump designed for UHV light gas compression or any applications where there is a need for low partial-pressure for gases, such as helium or hydrogen.
The duplex range has a Seigbahn drag mechanism after the turbopumps, and will likely be the most popular format. Because it will most likely exhaust into single-digit millibar backing pressures, the nEXT is capable of dealing with much higher backing pressures.
The triplex adds a fluid dynamic drag stage at the bottom of the Seigbahn stage; according to Steele this is mostly designed around the need for mass spectrometer applications for OEMs, which are pushing the throughput limits of their pumps.
“That triplex mechanism can be used to enable what we call a boost port on the side of the pump. The boost port enables the typically mass spec OEM to take advantage of either higher throughput or higher inlet pressures on their mass spectrometer instrument,” says Steele. This improves sensitivity and throughput.
Although the nEXT only recently has been introduced in the end user market, it is already entered volume manufacturing and mass spectrometry OEMs have been fitting them for about six months.
Pumping speeds are also available in three flavors: 240 l/s; 300 l/s; and 400 l/s. The first two come with an DN100 flange, but with different rotor geometry to maximize pumping speed. The 400 l/s pump has an DN160 flange and a geometry that gives a high, light gas compression capability.
Aside from throughput, the pumps benefit also from integrated drive electronics mounted flush into a pump base feature that is removable and non-intrusive for compact setups. There’s no need for a separate control—the pump will natively and on its own run a cooling fan or a vent valve. A traditional separate control panel—TIC—is available that provides DC voltage and monitoring control for backing pumps, valves, and gauges. The TIC panel can manage diagnostics and process control for a wide variety conditions, including power levels for up to four or five big turbopumps at the same time.
“Venting a turbo is always something that people are concerned about because it’s running very fast. The pump offers a variety of ways to decelerate and stop the turbo,” says Steele, without risk to the mechanics of the pump.
Fans can be mounted on either side of the pump, radially or axially.
Making field serviceability possible is really the biggest innovation with the nEXT for most end users, says Steele, but the commonality of bearing mechanisms and the design of the pump body give the company flexibility moving forward for different designs of rotor geometry and multiple inlet-type arrangements for mass spectrometry applications.
“It has also allowed much simpler development programs for developing custom pumping configuration for high-volume analytical instrument applications,” says Steele. For process-type applications where there are high throughputs of gases with significant condensable by-products, Steele recommends the use of magnetically-levitated turbopumps that can handle corrosive gases or vapors. Moving to that bearing technology allows heating of the pump body to keep condensables in vapor phases, for example.
What’s next for vacuum?
In some respects, vacuum component manufacturers must respond to market demands. This is a position that can engender a cautious approach, especially during difficult economic times. But strong competition among vendors means that materials breakthroughs, computing advances, and emerging markets can’t be ignored. In analytical science, the push for smaller but higher-performing instrumentation is relentless.
“We have to keep up with our OEMs and try to get cost-savings into their machines. In the mass spec world, for example, the largest component in a GC-MS is the pumping system. They want great performance, but they want it at very low cost. That sort of drives all the true pump manufacturers,” says Rock.
Pfeiffer, for example, adapted its turbomolecular technology for use in a portable gas chromatography-tandem mass spectrometry instrument developed by Torion Technologies Inc.: the GuardION7. Torion, American Fork, Utah, which won a 2008 R&D 100 Award for its technology, relied heavily on the capacity for Pfeiffer to deliver a high-throughput, reliable pumping solution inside a small footprint. In designing its HiPace line, Pfeiffer anticipated this type of demand by designing a standardized rotor architecture.
But demand alone won’t allow vacuum vendors to move ahead. They need the right tools to learn where to direct their investments in R&D-related projects. Edwards’ nEXT pump is a prime example. Without access to computational engines that profiled new bearing designs, development of the new bearing design might have been hindered. Why build prototypes when software can rule out dead ends?
“It’s really advances in manufacturing technology, materials technology, and software development tools that have allowed us to develop different rotor geometries mathematically rather than repeatedly cutting metals and testing it,” says Steele.
Experience helps, too. Supplying pumps for integrated applications—and in high volume—demands high levels of reliability, longevity, and performance; and, it’s a combination of those attributes that helps vendors decide where to direct their efforts.
Published in R & D magazine: Vol. 52, No. 2, April, 2010, p.8-13.