Do we all need to use WB/CAS protocol?
Posted: Thu Jun 20, 2019 5:56 pm
We’ve been chasing the source of the Helios cell signal artifacts that have been “solved” by the use of wide-bore and CAS. Clearly not all machines suffer from these artifacts, and not all machines that do manifest them suffer from them all the time.
The main characteristic of the artefact is that it is frequently time-dependent within a single sample. Unfortunately this is not the way that most data are presented. We also see that the artefact may be masked by the routine 50sec acquisition delay that most users employ.
We read with great interest the recent report from Lee et al in BioRXiv and were intrigued to see the differences in the severity of the artefact in different machines. We had the impression that these artifacts were relatively stable, but analysis of our own data over a period of several years has shown us that they are not in our hands. Importantly they are NOT user or sample specific. They are clearly related to the state of the Helios – we can see the quality of our data declining over the course of months.
So what controls the difference? We first tried adjusting the Z axis, which is not changed during the standard tuning that adjusts X and Y. We can move the Z of the sample injector and the entire torch body. We saw no real differences in the frequency of the artifacts (see attached file).
So what has changed whether the artefact appears or not?
We have only 2 real pieces of data that relate to what caused the artifacts to disappear for periods of up to months. The first is when we had the vacuum hose replaced. The second is when our interface vacuum pump was purged. Both these data points indicate that it may be the vacuum level that is crucial, and that is what we have been chasing recently.
When our machine was installed, the interface vacuum (plasma off) was 4 x 10^-4 (and the sheet said specs were < 7 x 10^-4). At the PMs over the last 2 years, it crept up to 8 x 10^-4 (on a new report sheet that no longer gave a spec value), then 9.9 x 10^-4 in October 2018 and currently about 1.3 x 10^-3. We just had the pump replaced, and the value is now 1.5 x 10^-3 – so there are clearly other elements in the pathway that need to be fixed as well.
What I would like to know is what the vacuum levels are on other machines, with, if possible, correlation with the occurrence of the artefact at any time on a particular machine. I have the impression that most users have now switched to a WB plus CAS and may no longer be seeing the artefact – but if you have not replaced any vacuum components since changing over to WB/CAS, your vacuum values will still be relevant. Probably the most important are the values BEFORE the plasma is turned on in the morning – they are on the LEDs behind the door on the LHS of the machine.
BTW when the artefact was a major problem before our recent vacuum pump purge, we saw the same artefact after a full day’s running with WB plus CAS. This indicates to me that our machine is still performing suboptimally.
The main characteristic of the artefact is that it is frequently time-dependent within a single sample. Unfortunately this is not the way that most data are presented. We also see that the artefact may be masked by the routine 50sec acquisition delay that most users employ.
We read with great interest the recent report from Lee et al in BioRXiv and were intrigued to see the differences in the severity of the artefact in different machines. We had the impression that these artifacts were relatively stable, but analysis of our own data over a period of several years has shown us that they are not in our hands. Importantly they are NOT user or sample specific. They are clearly related to the state of the Helios – we can see the quality of our data declining over the course of months.
So what controls the difference? We first tried adjusting the Z axis, which is not changed during the standard tuning that adjusts X and Y. We can move the Z of the sample injector and the entire torch body. We saw no real differences in the frequency of the artifacts (see attached file).
So what has changed whether the artefact appears or not?
We have only 2 real pieces of data that relate to what caused the artifacts to disappear for periods of up to months. The first is when we had the vacuum hose replaced. The second is when our interface vacuum pump was purged. Both these data points indicate that it may be the vacuum level that is crucial, and that is what we have been chasing recently.
When our machine was installed, the interface vacuum (plasma off) was 4 x 10^-4 (and the sheet said specs were < 7 x 10^-4). At the PMs over the last 2 years, it crept up to 8 x 10^-4 (on a new report sheet that no longer gave a spec value), then 9.9 x 10^-4 in October 2018 and currently about 1.3 x 10^-3. We just had the pump replaced, and the value is now 1.5 x 10^-3 – so there are clearly other elements in the pathway that need to be fixed as well.
What I would like to know is what the vacuum levels are on other machines, with, if possible, correlation with the occurrence of the artefact at any time on a particular machine. I have the impression that most users have now switched to a WB plus CAS and may no longer be seeing the artefact – but if you have not replaced any vacuum components since changing over to WB/CAS, your vacuum values will still be relevant. Probably the most important are the values BEFORE the plasma is turned on in the morning – they are on the LEDs behind the door on the LHS of the machine.
BTW when the artefact was a major problem before our recent vacuum pump purge, we saw the same artefact after a full day’s running with WB plus CAS. This indicates to me that our machine is still performing suboptimally.