DF-750是半导体行业中超高纯气体测量的优选

Gen-7 Podcast transcript

DB: My name is Douglas Barth. I’m the USTC product manager for the DF-700, and I’m here today with…

PR: I’m Phil Rogers, and I am the senior applications engineer here at the USTC for Servomex.

DB: We are going to introduce you to the Generation Seven NanoTrace DF-700 analyzer. It’s a modern analyzer. It’s been recently re-engineered, redesigned…

PR: Re-engineered, updated.

DB: …For the modern LCD and LED manufacturing processes that require ultra-trace quality measurement for moisture, contaminants, and ultra-high-purity electronic grade gases. In such a demanding application, users need analysis capable of delivering high accuracy and low, ultra-low detection limits in multiple background gases. No matter how demanding the application requirement, you will want a device that reduces preventative maintenance costs, maximizes uptime and has a long life in the marketplace.

We don’t believe you should have to compromise, and that’s one of the reasons why we are so bullish on the DF-700 product. So, given that the previous generation is going to end its production run, Phil, can you tell some of the listeners why this new project has been undertaken?

PR: Oh, sure. The DF-700 has been around in its current iteration for about 20 years. That’s quite a long run with the same basic architecture. The circuit board designs are all old, very complicated wiring. It’s essentially an analog analyzer because, you know, the signal gets digitized at the end and… running out of suppliers for a lot of these old components, so we had to update the electronics and the processing power of the analyzer to something that’s current, that will be serviceable for years to come, and get us ready for the next 20 years.

DB: So this is a completely digital analyzer?

PR: Yes, it is, in fact, completely digital.

DB: Wow. Excellent. Yeah, I heard a lot about digitization in the industrial gas side of the business. Great to see that digitization is coming to the semiconductor side of those Servomex products. This new Generation Seven DF-700, how’s it better positioned for our customers in the future?

PR: I wouldn’t say so much that is better positioned, but certainly positioned well in a modern platform. The signal processing means that what we have at our disposal now with the digital instrument is much greater. This allows us to get a quieter signal out of the analyzer which will result in improved detection limits, response time, and elimination of events that are not moisture or oxygen-related. And so that in itself should make a huge difference to our end-users, especially those in the CQC world, the semiconductor manufacturers that are keeping these things running 24-7, 365.

DB: So you’re telling me that all-new electronics, new PCB, hard drive, operating system… so everything is contemporary within the analyzer for this new digital platform? Have you updated the laser cell?

PR: The laser cell itself is essentially unchanged. Herriott cell design doesn’t, you know… you can’t really improve on that. It’s a simple design that’s robust and durable. And so, we are using the exact same cell, exact same lasers, the exact same mirrors. But what we’re doing is we’re getting the signal out of there in a different fashion.

DB: That’s awesome. So I can understand now how this digital platform with all these new pieces are coming together to improve the operation of the instrument. What will the actual customers see on their side of the analyzer from all these new additions that are made in this redesign?

PR: You know, from the outside, the analyzer looks essentially the same. The screen is much bigger, and much brighter, so it’s easier to read and more information can be displayed on the screen in a clear and understandable way. You can see what the analyzer is telling you from across the room, as opposed to having to put your readers on and get in front of the analyzer.

If the unit requires service, the experience will be much better. The sensors, and all components, can be swapped in situ really by a competent field service engineer or technically savvy end-user. Everything is right there, easy to get to, calibration will be following the moisture cell instead of being on the hard drive, the calibrations are on the moisture cell, or if you get a 760, the calibrations are on the oxygen cell, so all of these things make a much smoother service experience. If or when the inevitable service call comes in, it should be easy to do and make the customer happy.

DB: I know that some of the complaints for the previous generation have been about field serviceability. With all these new components, have you designed in some features that will allow the instrument to be serviced in the field?

PR: You know, everything is accessible, from the power supplies that provide power for the CPU and the moisture sensor, to the boards that contain the relays, analog outputs and serial communications. The solid-state hard drive is right there, easy to access, as is the CPU. Again, getting down to the sensors themselves, if something happens and the unit requires service, a new sensor or sensor swap, for troubleshooting purposes even, which does happen, it’s just a matter of pulling it and putting the sensor into the analyzer. The analyzer will pull the circuit board for that sensor, get all the calibration off of it and you’re up and running. Right? Just like that.

DB: So it’s the hard drive, the CPU, the PCBs, the display, the gas panel, all of those items now can be serviced in the field?

PR: Yes.

DB: Wow. That’s a big step forward. That’s the meat of the components within the analyzer that service engineers usually touch. That’s fantastic. You mentioned that the Herriott cell has been around for quite a long time. How long has it been around and could you tell us a little bit about the Herriott cell?

PR: Yeah, the Herriott cell is old technology that was invented in 1965 by the aptly named Donald Herriott. And it consists of two spherical mirrors, with an aperture in one of them, that allows the light from the laser to enter and exit the Herriott cell. It gives you a very long path length, up to 93 passes, which is about a 50-meter path length. So that’s essentially what a Herriott cell is, spherical mirrors with the light from the laser bouncing back and forth between them.

DB: So I know that the Herriott cell is where the actual sample or measurement is taken. Specifically, Servomex uses Tunable Diode Laser Absorption Spectroscopy inside that Herriott cell. Tell the listeners a little bit about the benefits and features of Tunable Diode Laser.

PR: The Tunable Diode Laser is a neat technology, but we’re looking at it as an absorption spectrometer. So we have to know the wavelength at which moisture absorbs light. And in this case we use 1854 nanometers where you get the laser output tuned to that output frequency by adjusting the temperature on the laser. Each laser has its own unique characteristics and so the laser temperature is unique for each laser.

And once we get the moisture peak tuned in, we modulate the current to that laser, which essentially causes that output to scan across the moisture peak. The moisture peaks at 1854. So we go from, say 1853 and a half to 1854 and a half, and one nanometer, by modulating the current to that laser and slightly affecting the output frequency of that laser.

So, you’re scanning across the moisture peak literally thousands of times a second, and you’re getting a lot of information out of there. It is a direct reading spectrometer. It will compare with the CRDS technology, which is also widely used in the semiconductor industry. CRDS is Cavity Ring-Down Spectroscopy, and what that does, is that emits a pulse of light within the sample cell. It goes between two highly reflective mirror surfaces and they measure the decay time, essentially, for that light to completely decay beyond detectable limits. And that sounds all well and good, but, you know, you’re measuring time, you’re not measuring moisture. And if those mirrors become fogged, or lose any of the reflectivity, it’s going to greatly impact the sensitivity of the device, and the detection limit of the device.

DB: That’s quite a difference. I mean, the only thing that’s basically the same between those two is the laser. They’re very different after that light enters into the chamber.

PR: They use a different frequency than we do as well. They use 1392 nanometers as opposed to ours, which is 1854. But that’s a minor difference there, really.

DB: You mentioned that we used the 1854 wavelength for our measurement. I picked up on something you said about being able to know the centroid of that 1854 wavelength. Could you tell the listeners a little bit more about that?

PR: Okay. Well, what we do is we have what we call a reference peak that the analyzer utilizes to keep the laser tuned properly. Our sensor is divided into two sections. We’ve got the Herriott cell, which is where the sample gas is, and the lasers, and then we have the laser chamber itself, which is isolated from the Herriott cell, and is hermetically sealed and pressurized deep out to keep out contamination.

And we separate the laser chamber from the Herriott cell with a little sapphire window. There’s a small amount of light that reflects off of that window, probably close to, you know, 1% or even less of the light from the laser, is reflected off that window back through. We have a little cuvette out there that contains moisture, so that reflective laser passes through that cuvette to a separate detector. Now there’s always moisture in there, and so the reference detector always shows a moisture peak. The software of the analyzer is designed to recognize that peak and if it sees that peak moving a little bit one way or the other, up or down in wavelength, it will adjust the current being supplied to that laser just a little bit to ensure that the peak stays in the right location and thus the laser is tuned to the proper output frequency.

DB: So if I understand this right, Phil, you’re saying that we use a secondary detector and a cuvette of moisture, and bleed off a little bit of the laser, constantly tell the analyzer where the moisture peak is, and make sure that it doesn’t deviate from that wavelength.

PR: That’s correct.

DB: Another piece that I heard you say during your explanation of Tunable Diode Laser Absorption Spectroscopy was this moisture peak sweep operation. How does that benefit our customers, that the analyzer and our laser technology sweeps across the entire peak?

PR: The sweeping across the peak, there’s a lot going on there. As you sweep across the peak, you’re measuring the brightness of the laser both on-peak and off-peak as you scan across that wavelength. The nature of this measurement is ratiometric. So unlike the cavity ring down, which you discussed earlier, if we lose a little bit of reflectivity due to optical fogging or who knows what real-life sort of things might happen to an instrument in an industrial application, you lose a little bit of reflectivity.

It’s not a big deal at all to analyze to compensate for that automatically, because the light will decrease, the intensity of light will decrease, both on-peak and off-peak, by the same percentage. So the ratio at any given moisture concentration between the on-peak measurement and the off-peak measurement is going to be the same at any given moisture concentration.

So if you lose 10% of your reflectivity – that’s both on-peak and off-peak – and if the ratio between the on-peak and off-peak measurement is going to stay the same, the measurement accuracy is going to stay the same. We had some years ago, we had an analyzer that lost more than 90% of its reflectivity due to contamination. We fired that up in our lab and tested against our known standards and it was still reading accurately with a greater than 90% reflectivity. I have to say I was surprised when that happened, but it gave me faith in the product.

DB: Wow, a 90% loss in an actual intensity of the laser and still measuring accurately, that’s amazing.

PR: It was astonishing to me, quite honestly. There it was. It worked.

DB: So sweeping across the peak, if you lose reflectivity or intensity, you could lose intensity from either the detector and its ability to pick it up, or the laser source. Would not this also calibrate out differences in the laser source and detector?

PR: Anything that has to do with the intensity of the light, be it the output of the laser degrading over the years, the output characteristics of the detector degrading over the years, or the reflectivity of the mirrors being affected by process conditions, all of this, from an algorithmic standpoint, it doesn’t matter to the algorithm.

DB: Say there was an interfering gas in with the background gas and you went off-peak, and it did absorb the laser, that interfering gas would be attenuated and you could calibrate that out?

PR: Yes, it would just be offset out anyways. You’re looking at the center peak, you’re not looking at anything off to the side. And moisture is rather specific at 1854, so in general, interfering elements are pretty much just offset out of there.

DB: So interfering elements, a difference in the intensity of the laser source, a difference in the ability to detect that source, and any kind of contamination on the mirror is instantaneously tenuously zeroed out of the analyzer for you and corrected for?

PR: Yeah.

DB: That’s amazing. No wonder this technology’s been around so long and is used in so many different places like ultra-high purity, semiconductor applications, specialty gas, the electronics market, LED, and display manufacturing.

Thank you so much for taking your time today, Phil, and discussing the amazing capabilities of Tunable Diode Laser and the new Gen-7 DF-700 product. I’d like to thank you for your time.

PR: You’re entirely welcome. This has been a fun thing to do.

DB: And I would like to remind all our listeners to visit servomex.com and find out more about the Gen-7 analyzer online. Thank you.