Why sealed Argon optics and not vacuum optics are the design of choice?

Header Banner

What’s better? Vacuum optics or Argon?

In a word – “Argon”. But why? And if indeed Argon optics are superior, why do we then have vacuum optics based OES at all? To understand this, let’s understand first why either is needed at all – and then, how the optical chamber has evolved.

Why do we need either?

At the outset, it’s important to understand why this question even arises. The optics of a spectrometer analyse the emission spectrum. This emission spectrum extends from the deep ultra-violet (DUV region) of around 120 nm (nanometers) to the near infra-red region of around 800 nm (nanometers). The characteristics of the spectrum and the prerequisites for good analysis however, vary widely across the regions. The deeper one goes into the UV region, the more sensitive the emissions are to absorption by stray impurities like moisture, oxygen or hydrocarbons, in the atmosphere. Therefore, for good analyses, it is imperative for the environment of the optical chamber to be absolutely inert, particularly in the region that contains the UV side of the spectrum. This necessitates either that this section be filled completely with an inert gas, such as Argon, or that it be kept under a high vacuum. Both would ensure that light is not absorbed and the analyses are therefore good.

How it all began

The performance of an OES – in terms of its detection limits, accuracy and precision – is directly proportional to its resolution. The better the resolution, the better all of these output metrics. Now, resolution depends on three key parameters:

When an OES is designed, the researchers, applications scientists and designers first decide what resolution they require to achieve targeted outputs. When the first spectrometers were made, grating technology was nowhere close to the levels it is at today. Gratings didn’t even exceed 1,000 grooves / mm for decades. As a result, the designers employed very large focal lengths (1-3 meters long) to achieve the desired resolution. This made the optical chambers absolutely massive. Ensuring that such voluminous chambers were kept fully inert while being filled with Argon was virtually impossible – and even if it were possible, would have meant tremendous expense just in terms of the volumes of Argon consumed. Hence, the chosen option was to subject these chambers to high vacuum.

Large and powerful vacuum pumps were therefore the default option for these spectrometers.

Why did PMT OES then remain on vacuum even as gratings improved?

Gratings technology rapidly improved and allowed the OES to become much more compact – but the PMT OES ran up against the next big constraint – the PMT itself! PMTs are large devices and a single PMT is needed for every wavelength that is to be analysed. Since many elements need more than a single wavelength analysed to cover their range, this meant that many PMTs had to be accommodated. The space requirements then became a constraint! To arrange all the required PMTs inside the optics, the arc along which they are placed had to be large too. This cannot be achieved if the focal length is small – and this limiting factor saw to it that even the most “compact” PMT OES was unable to achieve focal lengths of < 750 mm.

The result was that the optics chambers of these spectrometers remained large. Despite gratings becoming better and better and densities rising all the time, PMT OES were unable to utilize these as the shorter focal lengths would mandate that they had to reduce the number of PMTs they could fit into the optics!

The above constraints and the inertia of tradition have ensured that PMT OES have continued to use vacuum optics to this day!

How do CMOS/CCD OES manage it then?

CMOS / CCD detectors are small in size. Each of them (the ones used in spectrometers) contains from 2,000 to 3,800+ pixels! In effect, each detector therefore spans 2,000 to 3,800+ individual spectral lines. Using these detectors, space constraints are therefore far smaller. You don’t need much space to accommodate a very large number of detectors! As a result, OES designers can use gratings with far higher groove densities (compared to PMT OES) with consequently, shorter focal lengths to deliver the same or even higher resolution than in PMT OES. This makes the volume of the optical chamber much smaller than that of a PMT OES.

This smaller volume now makes Argon-filled optics feasible and just the right option, indeed. The designers therefore, having the option to using either of Argon or vacuum optics, choose the better one based on performance, risk and economy.

So why are Argon optics better?

Argon optics trump vacuum optics on three key parameters:

  1. Economy (Cost): Vacuum optics require a large vacuum pump connected to the optical chamber. Needless to say, this adds two elements of cost:
    1. It increases capital cost by virtue of the vacuum pump being needed at all.
    2. It dramatically increases operating costs through:
      1. Higher energy consumption.
      2. Consumables for the vacuum pump, such as the oil which needs to be replaced every few months.
      3. Losses due to vacuum pump failure.
  2. Time to stabilize: CMOS/CCD OES use sealed and Argon-filled/purged optics. This causes their stabilization time to be very short. Indeed, for most instruments, the stabilization time is under an hour even from a cold start. Vacuum optics on the other hand take far longer. Indeed, for the larger instruments, creating the vacuum may take several hours!
  3. Risk to the instrument: Argon optics pose no risk to the instrument. The same cannot be said for vacuum optics. A frequent issue with vacuum optics is “back flow” of oil from the vacuum pump into the optical chamber. Such back flow causes serious damage to the optical chamber and this can only be set right at the manufacturer’s Works – necessitating a very long down-time at huge expense. Further, most manufacturers who use vacuum optics do not offer warranty or AMC cover for the vacuum pump – further increasing the risk to the buyer.

Is there anything more to know about OES?

Of course! As with any field, this is merely the beginning. For more, do refer to:

Have Questions? Contact Us