Elan 6000 ICP-MS

I am just wanting to throw this out there into the forum and see if there is anyone else out there who has experienced this issue with this particular ICP-MS model or even one of the newer models such as the Elan 9000, or Elan DRCe.

After countless months of running along smoothly with the plasma igniting with a nice, clean and crisp pop on the first try, I walked into my laboratory this past Tuesday to start the ignition sequence like usual and there was no pop, not even an attempt to ignite.  (Note that the instrument had performed without any issues the previous day, and absolutely nothing was touched on the instrument from the end of Monday's run to the morning of the following day). 

I ran through all of the following troubleshooting protocols:

1-checked the faults history log to see if any unusual error messages were being displayed within the application section, and other that they expected, "ICP failed to ignite" message, there were no other faults listed.
2-Checked the Argon supply for issues with gas flow regulation and did not note any problems, but due to receiving some bad tanks in the past, I did go ahead and switch to a new tank just to be on the safe side.  This switch did not make any difference.
3-Checked the regulator on the back of the instrument to make sure that the gas pressure was equal to or above 50 psi, and it was within normal and acceptable parameters (which I expected to be the case, as the faults log did not display the, "argon pressure has fallen below 48 psi" error message.
4-Checked all gas line components to ensure that no gas leaks were occurring at any juncture or line, and also started the ignition sequence and then separated the interface from the torch box to manually check for argon gas flow, and this proved to be acceptable and properly functioning.
5-Removed and replaced entire torch assembly, as well as replaced neb tips (gas and sample) along with the o rings, ferrules and attachment screws and this made no difference.
6-Checked the alignment of the torch box and all was acceptable.
7-Removed ignition rod and cleaned with fine grain sand paper and then re-inserted securely and this had no impact on the issue.
8-Removed ignition rod and replaced with a spare rod from another instrument, and this did not eliminate the problem either.
9-Checked the load coil for wear and pitting and did notice some slight pitting, so again I initiated the ignition sequence and directly after the sequence ended, I opened the chamber once again to check for heat on the coil, to see if it was in fact getting a current, and it was warm.
10-Did go ahead and replace the load coil with a spare that was in somewhat better position and attempted to light the plasma again and still no change in the situation.
11-Replaced the skimmer and sampler cones since the mass spec was down anyway
12- Opened up the right hand side of the instrument and checked into the electronics section and power source as well as the induction coil and tested all for current, and nothing was out of the ordinary.
13-Entered service mode to check that all grid and other voltages did rise at the appropriate times during plasma ignition sequence and still no changes.
14-Tried hooking up a another argon tank and still no change.\
15-also changed one of the gas lines feeding the load coil to a new one with all new ferrules and fittings and still no change.
16-Both myself, and my service engineer are baffled.  He is ordering a new ignitor/transformer to see if it is possibly a result of that component going bad.

Basically, I do have a PE Senior service engineer assisting, but since it will be next week before the part comes in,  I am just asking if anyone else has experienced such an issue with their Elan ICP-MS and if so, is there anything that you did in your troubleshooting methodologies that I have overlooked?

Amy advice would be appreciated.

Thanks

Hi ToxLabRat,

All the things you checked did need checking, but there are 2 things missing :

(1) is there enough RF power available from the generator to ignite and sustain
the plasma ?  If the RF power tube has failed, the presence of correct grid and
other Voltages won't be able to assure the supply of RF power.

(2) is there a a sufficient ignition Voltage to trigger the plasma ?   The service
engineer should be able to answer this when the new ignitor/transformer has
been installed.

Cheers,

Rube

Research Scientist

Thank you for your reply.  I was able to solve this issue about a week after this post, so all is well now.  As to your points:

(1) is there enough RF power available from the generator to ignite and sustain
the plasma ?  If the RF power tube has failed, the presence of correct grid and
other Voltages won't be able to assure the supply of RF power.

and

(2) is there a a sufficient ignition Voltage to trigger the plasma ?   The service
engineer should be able to answer this when the new ignitor/transformer has
been installed.

-yes I did check this in steps 12 and 13 in my list (see below):

12- Opened up the right hand side of the instrument and checked into the electronics section and power source as well as the induction coil and tested all for current, and nothing was out of the ordinary.

13-Entered service mode to check that all grid and other voltages did rise at the appropriate times during plasma ignition sequence and still no changes.

In step 12, I did actully remove the RF power supply as well as the induction coil, and used a meter to test for current, and both components gave results which were consistant with properly functioning components.  (I even tested in duplicate just to be sure).  So, I knew that the issue was not in the generation of power.

In step 13, I did address yuour other issue, as If there was not enough power available to supply the RF Generator in the necessary amounts as to allow it to convert inffluent energy into RF energy, then it would have showed up while in the service mode, as in the Perkin Elmer PE-Sciex software there is a tab where you can start the ignition sequence (while in service mode) and then watch the energy peaks as the sequence plays out.  If you do not see the correct voltage jumps, then it will be indicated here, as the numbers will either stay at zero or will not rise high enough.

I did mean to post on here the conclusion that I came to, as to help others in the future as well as just to have it on here for some troubleshooting reference.  So here is the conclusion:

Since it was not a leak in the gas system, no issues with the main hardware, no issues with mechanical roughing and backing pumps, no issues with the turbomolecular pump or the power supply, induction coils, etc. I decided to examine the ICP-Controller board (the brain).  I took one out of another unit that I had in the lab (basically a unit I use for parts), and removed the ICP Controller board, and replaced with the other ICP-Controller board.  Once the switch was made, I made another attempt to light the plasma, and plasma lit on first try with a nice strong pop.  So in short, it was the ICP Contoller which ended up being the root cause to the plasma ignition issues, and luckily, the ICP Controller board is one of the more easy electrical components within the instrument to replace, so it took less time than 30 minutes to make the switch.

Thanks again for your reply.

Ashley
Tox

Thanks for the final chapter of this Elan story Ashley.  Was there something that the
ICP-Controller board was (or wasn't) doing then, that stopped the ICP ignition - and
yet allowed it to pass checks (12) and (13) ?

We readers applaud your perseverance,

Rube

I did not determine that until just last week, when I had a power disrutption in the lab which took down my chiller and a few ovens.  When I went to re-start the chiller I noticed that it would not kick on, but instead everytime I flipped the switch to "on" it would cause the oven to cycle on and off.  It was all very odd, but I remembered it happening once before several months ago, but by the time I got the electrician down there it just seemed to "magically" correct itself, and just like a car when you take it to the shop, it would never reproduce that probelm again whie the electrician was there.  So he left without making any changes.  It was not until last week when this happened again and instead of trying to do anything at all with it, I called the electrician before the thing went back to normal.  He got there and as it turned out we had two HUGE power supply wires (blue and black) in the secondary fuse box which controlled all of the circuits that the ICP-MS, chiller, ovens FAA, and several other instruments and equipment on, were arching horrible, to the point of green and blue sparks.  He had to kill the power to the entire place because just like everywhere, the main power usually feeds into Box A, then that feeds into Box B, etc.  He corrected the issue and now so many other small electrical issues we were noticing have resolved.

To make it short, I cannot say with 100% certainty, but I believe that it is because of what had been going on with the circuit, which caused the intermittant issues with the ICP-Controller.  They are not really like a mother board, so they can handle a little bit more stress, but I think it finally just had all it could take of the power disruptions.  It is also possible that it was old and on its way out as it was and just was working sporadicly until the end.

Hi Mr ToxLabRat, My name is Ahmad, i have Elan DRC-e 6100. In last 3 days, i have encountered a problem with going out the plasma unexpectedly.(after about 2 hrs and a half from turning on the instrument) I thought it was due to the argon gas cyliders (which i always use), you know the effect of using Argon purity<%99.999. Unfortunately, it did not make any difference. During several trials to diagnose this incident, i found that the vacuum had been broken for a while then returned back. the plasma went out accordingly. I cleaned the spray chamber and neb. tips and tightened their components securely, in order to avoid any air leak>>> the problem still remaind. I checked the interface roughing pump's oil, found coffee color>>changed.>>>The problem remained. I checked the interface components: Torch, Torch mount nuts, tightened RF coil connecting nuts.>> No change. I checked the the sampler and skimmer cones and replaced their O-rings>>>>> No change. I re-checked the cones, i found that the sampler cone had a defect at the rim toward the inteface chamber not the torch side. I lit the plasma after a good vacuum pressure had been achieved..... The plasma sustained for 1 hr, but unfortunately the working day was over, so i was forced to turn off the instrument before getting any good indication that my problem is over. I look forward to hear from you, your advices and recommendations would be appreciated. Best Regards, Eng. Ahmad Lab Engineer/ Chemical Engineer

First let me ask you if you are using liquid argon?  Typically I would assume this, but do not want to do that here just in case, because your problem may be that simple.  If you are always loosing your plasma after a few hours, I am wondering if the Ar cylinders that you are getting are Ar gas and NOT Refrigerated Liquid Ar.  If you are getting the small gas cylinders then a few hours is all you are going to get out of them before they are empty, as the Elans (all of the modles) use an average of around 20 L/min of Ar total (neb gas flow + plasma gas flow inlet).  So first of all check that out.  If you are using gas, then you need to switch to the refrigerated liquid argon instead.  This will last you MUCH longer.  You can get them in a 180 Liter or 200 Liter cylinder.  For example, I can usually get about 3-4 weeks of run time @ 8 hours per day-5 days a week, off of one 200 Liter Compressed Liquid Ar Tank.

Now when your gas runs out, you should see a loss in vacuum on the faults log and then it will very quickly regain itself when you hook up a new cylinder.  Technically you are not loosing vacuum, because the mechanical vacuum backing pump and the turbo pump do not cut off, it is just when your Ar flow drops below 48 psi the swift extinguishing of the plasma will cause the vacuum pressure to flux enough to give that fault. 

Have you checked your faults log to see what faults you are getting?  If you have not then do this and tell me what the faults say, such as "user pressed plasma on", "gas flow went below 48 psi", "plasma failed to ignite", etc. along with the times in which they occured (so I can see how spaced out from each other that each fault occurs).

It is also possible that if you had a faulty cone that you had one of your o-rings melt slightly.  Did you notice when you took the fault cone out, if the o-ring appeared damged or if it looked odd in any way?  It does not take much if that sampler cone is off to melt the o-ring around it, and sometimes when that happens some of that can get down into the gate.  The gate has to be able to open and close the interface so that the plasma can light as well as so that it can disengague.  When the gate sticks, then you will usually see it in the "Service Mode".  Go under the vacuum tab while in service mode and then manually on the software tell the gate to open and then tell it to close, and do this a few times and see if it gets stuck.  If it is sticking you will see it say "half-way", instead of "opened" or "closed".  Sometimes just doing this a few times is enough to resolve the issue of the gate sticking, but sometimes it is not.  If some of that o ring melted in there, then you may have to try to clean it out and then apply a little bit of Si lubricant to the vavle and see if that helps.

If you would like you can email me at toxlabrat@hotmail.com  and let me know what kind of gas you are using, because like I said, it could be very simple.  If you are using liquid Ar, then give me the faults info as that will be very helpful.

Ashley
(Tox)

It  happen to me last week. I am using ELAN DRC II .  I did all the trouble shooting that you did but it did not work. then I called the service engineer he suggest me to increase the RF power manually by 50 V maximum 150 V  and he told me if it is not working we have to replce the coil. I increased the RF power 5 0V and the plasma ignited with the first attempt.

Ok, it sounds like you might have an RF induction coil that is on its way out the door.  The next time it gives you an issue try this:

Start the plasma ignition sequence and let it complete.  As soon as the sequence ends (and of course ends with the plasma NOT lighting) go ahead and open up the hood and separate the interface from the torch box by lifting the lever up and pulling the interface and MS open.  Then very carefully stick your gloved hand into the torch box where the torch is sitting inside the RF load coil.  Gently touch the RF load coil and see if it is warm.  If it is not warm at all, then more than likely it is not working properly and not pulling enoough power into the coil to help ignite and sustain the plasma.  If it is warm then other things will need to be assessed.  However if it is NOT warm or cold, then go ahead and remove the torch assembly from the torch box.  Then remove the torch box cover so you have some room to work in.  Then get a flashlight or some type of penlight and examine the RF coil and see if you notice any pitting or worn places.  It is not so much the color of the coil that you need to be concerned about as it is areas where you have indentations, pitting, scratches, things that have changed the smooth surface of the coil.  If you notice a great deal of these worn areas, then it could just be a matter of changing out that load coil.  If you do not have an extra coil lying around, then you can take the one which is currently in the instrument and pull it off of the stage and use some very fine grain sand paper and rub those areas which are pitted and worn to try and smooth them back out.  Sometimes just sanding down the rough areas will be enough to help restore the proper induction current capabilities to the coil. 

The coil is pretty simple to remove.  You will find a few nuts which lock that coil onto the stage.  The coil is adjoined to the stage at the back side of the coil (opposite the tip of the injector tube and torch).  You should see some gas lines leading up to the back of the coil as well, so you will be removing those from the back of the coil as well.  I know that some people will try to sand one down while it is still on the stage, however being that it is made from copper, I would not recommend it.

There is nothing wrong with bumping up the RF power some, but if your issue is the coil, it will only get worse and you do not want to bump up the RF power to extremes.  If your instrument is under service contract and you have never replaced or performed any maintenance on your RF load coil, then I would recommend going ahead and opening up a service call the next time you have the issue so that you can have your service engineer show you how to do it.  This way you will know that you are removing and installing the part properly and also you can make sure that you know what to look for as far as pitting and wear and tare on the coil is concerned.

If when you check the coil it is warm and yet the plasma still does not ignite, then give me an email, because it could be anything from a bad ignitor all the way to a bad ICP controller and many things in between.
toxlabrat@hotmail.com

Hello Water and TLR,

- my note of caution for the RF coil : the threaded nuts that attach the coil
are Swagelok imperial-size brass fittings, and they can be mis-threaded if
not handled carefully. I recently saw such a case - there was a small argon
leak due to the thread damage and this leakage produced plasma instability
that was hard to diagnose. The PE service people are doubtless aware of the
need to be careful with these fittings, but it could be a trap for others.

Rube

Most of those nuts should have been changed out by now.  I know that I do not have that old style in my instrument anymore, as I changed mine out for the upgraded nuts which are not so easy to strip and are much easier to remove. 

Make sure to use your alignment tool as well if you decide to change out the coil yourself.  Every time you get a new coil it will come with one of these plastic black alignment tools with it.  The coils are actually shipped already screwed onto this tool. 

But, like I said before, I recommend if you have never changed one of these out in the past, that you have your PE Service Engineer show you how to do this so you can be assured that you are putting the part on properly as well as, can you make sure that you are properly assessing the state of the part.  There is nothing worse than spending a lot of money on a part that you do not need in the first place and/or is not even the root cause to your issues.

Good Luck.

Hi,

Thanks a lot for your post. It really helps me a lot.

I am a graduate student and focus on environmental science using ICP-MS. Our Elan 6100 runs very well in the last several years. However, recently there are several problems.

Last week, the plasma turned off automatically during sample analysis. I retried several times and the plasma could only last less than 1 second after igniting. Thanks a lot for the previous posts in this forum. I checked everything you mentioned and changed several parts such as cone, skimmer and pump oil. Now the plasma is OK. :)

However, the sensitivity became another problem. Each time before my experiment, I have to do all the optimizations including X-Y alignment, Nebulizer Gas Flow, Lens Voltage, AutoLens. And on the next day, when I run the daily performance, the intensities are very low again, around 10% of the previous optimized results. Then I have to redo all the optimizations.

This afternoon the same thing happened again. The instrument was optimized this morning and failed to give good intensity in the afternoon (only 15% compared to the optimized result in the morning). I changed nothing but just waited for 30 min after the igniting. After half an hour, the intensity was about two times higher than the initial results (30% compared to the optimized result in the morning). After another 20 min, it was even higher, although still low compared to the optimized result in the morning.

I am wondering whether it is necessary to replace the torch and coil. Do you have other suggestions? Thanks a lot!

Bo

Bo,

The first thing I would ask you (especially since you are using an Elan and Sciex software) is when you do all of your optimizations are you saving them to the default.dac location?  This is a very common issue with some of the Sciex software packages and is also just a common oversite with all of the software versions.  Sometimes the file location can be accidentally changed through human error, software glitch or even when you have a power disruption.  If this happened and your location is not set on the default.dac location then you can optimize all day long and it will not actually save it to the default.dac file location, which is a necessity if you want to see your instrument reflect the changes that you made.  Basically you would see great counts and all of these positive changes as you move through your optimization flow chart and those changes would stick as long as you did not close out of the software or the main page you are working off of, however as soon as you change methods, workspaces or close out of the software and come back in, then all of those optimization changes will be thrown back out and it will revert back to all of the settings that were there initially (before doing any of your optimiztion procedures).  There are a few versions of the Sciex software that are glitchy as well and I have worked with such a glitchy program and in that case the same thing happens but it is not a result of your location being incorrect it is just a result of the glitch.  So in that case what you may have to do is save each optimization file to several different locations before it will actually keep the new file or you may have to save to the default.dac like you normally would and then close out of the software and then open it back up again in order to get the new settings to permanently save.  You can check to see if this is the case by just simply looking at all of your settings and files the next time you bring the instrument up from standby or shutdown mode.  Look at the nebulizer gas flow and see if that is set to the same L/min rate as what you optimized it to be last.  Do the same thing with the lens voltage as well.  Then go to the tab for your autolens calibration and see if the date at which that calibration was obtained matches the most recent date in which you did an autolens calibration.  Also do this for the dual detector calibration as well.  If just one of these files is not saving then it will cause your isotopes of interest to have extremley low cps as will the overall DP report just look terrible.

If none of the above applies to your case then it would point towards it being a hardware issue, or at the very least an issue with the instrument or its components.  I see where you say that you get higher and higher intensities for your daily performance with every DP you run....basically the longer the instrument is on, the better counts you get.  Well this is normal and in fact it is for this reason that it is good practice to let your instrument warm up for at least 30 minutes.  I like to give mine about 45 minutes if I can because the older these instruments get, the longer they tend to need to warm up.  If you try to run a daily performance too soon after first lighting the plasma then you will see extremley low counts, high and/or jumpy RSDs as will you most likely see higher Ce/CeO ratio, Ba++ % and 220 Background cps.  For example today I tried to run a DP a little earlier than I usually do just because I was trying to get something going before heading out of the lab for another engagement and the Rh counts I got were only 170,000 cps with a Ce/CeO ratio of 0.035 and RSDs all over 7.  I gave it an extra 30 minutes to warm up and did not do anything else to it and just by giving it that extra time my Rh counts came up to 220,000 cps with a Ce/CeO ratio of 0.030 and RSDs all right around 1.  So you can see that was a pretty significant difference just from giving the instrument more warm up time.

Now if none of the above capture your specific issue(s) then I am leaning towards something like an intermittant argon leak, faulty cone, faulty o-ring, faulty lens, something wrong in your sample introduction system or maybe even time for a new detector (electron multiplier).  If you would like you can email me if you need further help.  I am very familiar with your instrument and will give you the best troubleshooting plan I can to try to minimize the use of parts and of course minimize the amount of your time it takes.  One thing I will also throw out there is to think back on when you first started having this problem-is there anything (no matter how small it may seem) that you did, changed, cleaned, etc. with the instrument or in the software which corresponds with the time frame this issue started to come up?  This is key for all root cause analyses and can be very helpful when you are troubleshooting.  Even something as small as moving your computer around or just simply pusching it back an inch can cause issues-so note anything you can think of.

Tox

Tox,

Thanks a lot for all the suggestions above. 
 
We are using the version 3.4 software currently. It was updated from v3.3 last fall. I am pretty sure we saved all the settings after optimizations.

I checked all the hardware issue including argon system, cone, o-ring, sample introduction system, torch and coil.

However, even let our ICP-MS warm up for hours, intensities are still low. The most weird thing is we have to do X-Y alignment every time before experiment. And we indeed can make the intensity much higher by optimizing those two knobs. I marked the positions of these two knobs after optimizations and I am sure nobody changed them until the next experiment. But when I came to the lab the next morning, the sensitivity was low again.

I am still working on this problem and I will update the most recent situation.

Bo