Keeping Warmer in the Winter: MS Temps for Air Analysis

As winter starts up it becomes time for a lot of us to start thinking of how to keep warm. When it comes to air analysis instrumentation though, more thought is put into keeping things cold. Focusing in the preconcentrator and head of the column are key to getting good results for air methods. When it comes to the mass spec though, a bit of extra heat can help you out.

Higher mass spec temperatures have gotten some attention for semivolatile analysis (see a previous blog by Chris Rattray - Effect of Source Temperature on 2,4-DNP Response at Low Concentrations), with higher temps giving better responses. The popular opinion seems to be that volatiles don’t need extra heat, but what does it do for air analysis? Keen eyed readers of the blog may have noticed that in my ethylene oxide work I began to use higher source temperatures, with my first blog (Cryogenic Cooling for Air Analysis Part 2 – Combining TO-15A and Ethylene Oxide) using 150°C for the quadrupole and 230°C for the source, but by the end (Ethylene Oxide in Air Canisters – Method Performance) it was up to 200°C and 350°C. However, it wasn’t for ethylene oxide, at least not directly. I found that over time my responses were dropping when running humid standards, and I was having to retune every couple days to keep them up. Kicking the source and quad temperatures up changed my tuning schedule from almost daily to once every few weeks or more.

How much did it help? To make sure I was giving both temperatures a fair shake, I did a more controlled test. I set the MS back to the original temperatures of 150°C for the quad and 230°C for the source, let the system equilibrate for 3 days, then performed a BFB autotune. After that, 20 blanks at 80% RH were run each day for 2 days (i.e., 40 injections total), injecting 400 mL on to the instrument. I then repeated this with the higher temps of 200°C for the quad and 350°C for the source. 80% RH was chosen because dry blanks didn’t seem to cause the loss of response, so I wanted to  be aggressive with the amount of water vapor. At 80% RH 400 mL of air has 6.4 µL of water. Even with the Kori water removal module for our Markes and a 3:1 desorption split, an extracted ion chromatogram of m/z 18 shows that a significant amount of that water still reaches the GC-MS.

Fig. 1 – Extracted Ion Chromatogram of m/z 18.

When looking at the change over time of the internal standard responses, the higher temps do improve the stability, although the first internal standard (bromochloromethane) still shows a drop over time, going below 80% of the original response by injection 32. This is far better than at lower temperatures, where it drops below 80% by injection 4. While I only looked at internal standard responses here, it’s possible that some compounds could drop at different rates, leading to drift from the calibration relative response factors. The loss of sensitivity could be a concern for low level results as well. Finally, TO-15 requires samples to have an internal standard response of ±40% from the more recent calibration, and while the high temperature sequence passes this, by injection 25 the low temperature is routinely more than 40% off from the original response for at least one internal standard. All in all, the dramatic loss of response with low MS temperatures is potentially bad news for your results on multiple fronts.

Fig. 2 – Internal standard stability at low (150°C quad, 230°C source) and high (200°C quad, 350°C source)

When setting up the runs, I started 20 injections in the morning, then came back the next day to start the next 20. On injection 21 you can also see the responses jump up slightly after the instrument has been able to sit overnight, followed by a quick drop. It suggests that the MS is able to purge out some excess water vapor if allowed to sit, and that a few humid blanks before running samples for the day may be a good idea get things equilibrated.

If you’re concerned about the higher temperatures affecting your BFB tune results, you’ll be happy to hear that I had no trouble with it. Even with the response change after 40 injections the tune still passed the criteria laid out in Table 14-2 of TO-15A.

Fig. 3 – BFB tune results. Top left – low temp, injection one. Bottom left – low temp, injection 40. Top Right – High temp, injection 1. Bottom right – High temp, injection 40.

As a final bonus, the higher temperatures gave me more than 2.5 times the internal standard response for the same multiplier voltage. This means you can potentially extend your expensive multiplier lifetime by getting the same response at lower voltages. If you’re concerned about wear and tear on the MS source and quad, I can’t find any info that indicates higher temperatures lead to shorter MS lifetimes, and semivolatile labs run at higher temps with no issues that I’ve heard of.

While I’ve focused mainly on air analysis, this could also be applicable to other volatile methods that potentially have water vapor issues, such as purge and trap for 8260, 624, 524, etc. A related trick to share is to keep your column temperature up at 100°C if your GC is going to be idle for a while. Even very trace amount of water vapor and other contaminants can condense on the head of the column over time if you leave it at 35°C for a long while, so keeping a higher idle temperature can make your first few runs looking a bit better.

All in all, this seems to show there are a lot of upsides to keeping things warm. If you’re having trouble with your TO-15 or other volatile method internal standard responses, adding a bit of extra heat might help you out.