PBL and filaments

25 Sep 2014

Hi all,

This is a follow up on the scale compression discussion with respect to capillaries, so I wonder what else may be different among instruments. The tought came yesterday after checking the third filament in a row in the last couple of weeks. In all of them we get sensitivity that is 20-25% lower than we are used to and in all of them the mass 49 baseline is way more negative than I ever saw (mass 47 baseline varies among filaments and tunings). The reason I suspect the filament is that it happened to us 2 years ago when we changed filament and that was solved by installing one from an old batch that we had still available. Then I learned that Thermo replaced their filament supplier.

I wonder, can it be that all the labs that has very negative baseline and need do do PBL are using filaments that were purchased from Thermo in the last 2, maybe 3, years and older filament (presumably from the previous supplier) perform better for D47?


Cedric's picture
Joined: Feb 6 2011
Hi Hagit, This is a very good

Hi Hagit,

This is a very good point. We changed our filament about 5 months ago on one of the two instrument, and since then the 47 has been less stable. Also, the 48 behaved completely differently at first, i.e. we even had a strong negative slope in 48 first, which gradually over a month or so stablised to normal. I wonder if anyone else had that issue? But filaments probably play a big role.

I will look into this in more detail on my machine once I have a chance.


Paul Dennis's picture
Paul Dennis
Joined: Sep 16 2011
It seems to me that there is

It seems to me that there is a poor understanding of the details of electron-molecule interactions in the source and the effect that the filament position, alignment, operating conditions may have on the measured isotope ratios. I think the big issue is scale compression, at least it is for us as MIRA is linear. From my early experiments on a test rig (reported at the clumped workshop in Seattle in 2010 - is it really that long ago) the key issue is electron induced molecule fragmentation and recombination. I think the results Huntington et al. reported on the effect of opening the sulfur window support this. Basically once you've generated ions these need to be extracted and accelerated with maximum efficiency before they can recombine within the source. The problem is the e-beam in many sources is poorly constrained. Yes, we have magnetic collimation but I know that our source design on MIRA doesn't produce a really tightly collimated beam. The electron oscillation radius is at least 1.5mm and this mostly as a result of transverse energy due to space charge expansion. As a result we are generating a lot of ions that lie outside the acceptance window of the ion extraction and transfer optics. These ions have a long residence time in the source and recombination leads to scale compression. 

Small changes to filament position, alignment etc., ageing of the filament can lead to small but important changes in the position of the e-beam and it's oscillation radius vis-a-vis the acceptance window of the ion beam forming optics. Hence changes in sensitivity, changes in actual measured ratios due to mass discrimination effects which are a function of the e-beam position and changes in scale compression are observed.

We're currently looking at an improved optics system for the e-beam so as to better constrain the electron path to match closely the acceptance phase space of the ion extraction optics. I hope this will give us much better control on scale compression, reducing it to a minimum.


Paul Dennis, Stable Isotope Laboratory, Environmental Sciences, UEA

Joined: Sep 18 2011
Hi everybody, my experiences

Hi everybody,

my experiences with the PBL are as follow:

About ayear ago we replaced the entire source with a new one, and the PBL (the shape and the magnitude of the ngative backgrounds)  were the same as the old source. In our experience we find that when we vent theinstrument the backgrounds became smaller and they increase with time afterwards.   As we explained in the Meckler et al. paper the shape of the PBL changed a lot when we changed the rotary pump backing the analyzer turbos with a slightly stronger one.  This suggests that the vacuum conditions in the flight tube may have a significant influence on the PBL (more or less scattering of ions?),  at least on the shape of the backgrounds. I wonder if placing a second turbo pump where the H/D collector is would improve the problem with the backgrounds. 

The magnitude of the negative backgrounds will depend on the sensitivity of the mass spec/source, because the negative backgrounds are proportional to the partial pressure of CO2 inthe source.   If you put a filament that gives 50% signal compared to the previous one or your purposedly reduce the sensitivity  of the instrument to 50% by changing the tuning you will end up with a 2 times more negative background for the same beam size because you need a higher source pressure to achieve the same beam.  This will lead to a much steeper heated gas line. 

we have had lots of problems with the filaments on the Delta V instruments, but luckily, so far we have not seen the same problems on the 253 even if the filaments are younger than 3 years.