March 16, 2018 - Health Ranger - Mike Adams
10:29
| 
| Time | Text |
|---|---|
| - All right, welcome back to ICP-MS Troubleshooting. | |
| My name is Mike Adams, the Health Ranger. | |
| I'm your resident quantum mechanic here at CWCLabs.com. | |
| I've been running the Agilent ICPMS instrument for about two and a half years now, and I've become very good at troubleshooting problems, so I'm sharing some information with you online to help you troubleshoot issues. | |
| Now, here's a very, very difficult issue to track down, so I'll share it with you. | |
| It might save you a couple of days. | |
| If you are having a calibration curve that is slightly out of range at only low concentration ranges, for example, 0 parts per billion through, let's say, 200 or 300 parts per billion, or maybe 500 at the high end, depending on the element, but that calibration curve cleans up at higher ranges, such as 500 plus parts per billion, You know, getting into parts per million and so on. | |
| If that's the pattern that you are observing, then you need to look at your pulse analog tuning. | |
| So remember that this instrument has a detector here with nine orders of magnitude of range. | |
| It accomplishes that by combining two different types of detection technologies. | |
| One is a digital detector, which is called pulse or P, and the other is analog. | |
| Which is for your high-end ranges, and that's, of course, called A. So PA tuning is your pulse analog tuning, and what that does when you properly tune with the PA tuning method, then the software is able to create coefficients at all the different masses that you use in the tune in order to adjust the concentrations that it's seeing during a scan. | |
| And this is crucial. | |
| This is absolutely crucial. | |
| If you change a lot of things, if you change... | |
| The masses that you're looking for, the analytes, if you change your internal standards and do something totally different, even if you move the instrument or if you radically change the environment, you need to redo the PA tuning. | |
| But don't forget to do an EM tune before you do a PA tune. | |
| If you don't do the EM tune first, then your PA tuning isn't going to help you that much. | |
| And remember, the EM tune affects the voltage of, I believe, the RF coil. | |
| The voltage that's affecting the plasma. | |
| So first do the EM tune and then do the PA tune. | |
| Make sure on your PA tuning screen that you have entered the correct masses of your PA tuning solution. | |
| It does you no good whatsoever to run a PA tuning solution that is different from the masses that you have entered into the system. | |
| It's not going to see those. | |
| Also, make sure you check the results of the PA tune and make sure that there's a number for every mass in that tuning range. | |
| If you get a message that says too low sensitivity or something below detection limits, that means you've diluted your PA tuning solution too much. | |
| You need to remake it with a much lower dilution factor I often use just 10 to 1 right out of the bottle of a PA tuning solution that I buy from inorganic ventures. | |
| I just dilute it 10 to 1 and run it, and that gives me good numbers. | |
| Now, here's the really tricky part in this. | |
| You may not realize that the concentrations you're seeing in your final samples Let's say you're testing some piece of food, and you're seeing a concentration in the final test of, let's say, 100 parts per billion. | |
| Let's say you're seeing 1,000 parts per billion of lead at 206 or 208, for example. | |
| You might think that a thousand parts per billion is in the analog range of the detector. | |
| So you think, hey, as long as my analog is good, then that one thousand parts per billion result must be calibrated and aligned. | |
| But you've forgotten something. | |
| You've forgotten the dilution factor. | |
| If you're using 0.5 grams of that food sample in a 50 milliliter vial, then you have a 100 to 1 dilution factor, which means the detector is not seeing 1,000 parts per billion of lead at 208. | |
| Instead, it's seeing 10. | |
| 10 parts per billion at 208. | |
| And 10 parts per billion is in the pulse range. | |
| In other words, the digital range of the detector. | |
| Do not confuse the resulting concentrations of your samples with the direct concentrations of your external standards. | |
| This is a big mistake that people make all the time. | |
| You have to realize that a lot of your resulting calculated concentrations are based on pulse mode detection because of your dilution factor in the food or whatever sample that you're looking at. | |
| If your dilution factor is large, then that means that you're shoving all the counts per second down into the pulse range of the detector, which means your pulse tuning has to be really, really fine-tuned. | |
| If you're running, let's say, a serial dilution set of standards here, and in this case I'm running multi-element standards at 0.8 parts per billion here, all the way up to 3,000 parts per billion here. | |
| I have a very wide range of calibration, and I use factors of five from vial to vial. | |
| There's a very important reason for that. | |
| I'll talk about that in an upcoming episode on pipetting. | |
| But take these same solutions, make another set, and place it over in your unknown samples section Run through the standards in your batch, get a good calibration curve, and then direct the probe to sample all those five solutions as unknown samples and see if it gives you the same results that it got in the calibration curve. | |
| If it does not, I mean, this is one good way to check the consistency of the instrument and the integrity of your curve. | |
| If you're seeing very different numbers on the unknown sample side, even though they're the same solutions as your calibration standards, then you have a big issue, and you need to look, you know, depending on the pattern, you have to look lots of different places to find out why that is, what's happening. | |
| The other thing that I like to do is I put a high acid matrix rinse here. | |
| I use 20% nitric and 2% hydrochloric in this vial, and I hit this vial before I do a background subtraction. | |
| If you look at your background counts, I'm going to set this down so I don't dump it on myself. | |
| If you look at background counts, when you first turn on the instrument and you get into your first vials, your background counts can be artificially high. | |
| And you need to hit the entire plumbing system with some very powerful acid in order to kind of clean that system out before you take your background's. | |
| And a good tip on this is if you're looking at your counts per second as you're doing the calibration runs, and if you start to see a lot of negative numbers, especially in the low range of your calibration curve, that means that your background sample is too high and you did not clean out the system properly before taking the background. | |
| And if you don't know what I mean by background, then you need to go back to, like, step one, because you have to have a background subtraction in order to get a good calibration curve. | |
| I guess that's another video, but there are going to be contaminants in your acids, in your nitric acid, in your hydrochloric, maybe in your vials. | |
| There are elements coming in through the air, the atmosphere that's coming into the plasma. | |
| So, you know, hey, you've got to take out background, right? | |
| Anyway, those are some tips. | |
| I hope that's been helpful to help you solve problems with the ICP. This is a... | |
| I'm running the Agilent 7700X. It's an outstanding instrument. | |
| I have never, to this day, I have never found that the instrument was at fault. | |
| I've always found, in every single case, that if I was seeing strange numbers or unexpected results, I've always found that that is due to me doing something wrong or somebody else doing something wrong or sample prep incorrectly or standards being wrong, things like that. | |
| It's never been the instrument. | |
| So keep that in mind. | |
| I know it's... | |
| A lot of people tend to blame the instrument right away, but the instrument is not lying to you. | |
| There's something that you are doing, or something that your calibration standards... | |
| Maybe something's wrong. | |
| Maybe you have pipette carryover from one to the other. | |
| Maybe you have... | |
| Done something wrong in your calculations of concentrations. | |
| Who knows? | |
| All kinds of problems that could come up. | |
| Don't ever think that the instrument is lying to you. | |
| It doesn't lie. | |
| And it doesn't make elements. | |
| That's for sure. | |
| This instrument does not make elements. | |
| Otherwise, I would have tuned it to gold a long time ago, and I'd be cranking out a gold factory here. | |
| Strangely, the only thing I can get out of it is aluminum. | |
| So, no. | |
| In any case, this is an outstanding instrument. | |
| I hope this video has been somewhat helpful in helping you troubleshoot possible challenges that you might have with running it. | |
| My name is Mike Adams. | |
| I'm known as the Health Ranger. | |
| I used to be a clean food activist. | |
| Now I'm a forensic food scientist working for cwclabs.com. | |
| If you want to use our services for testing your foods, if you're a food company, then please contact us at cwclabs.com. | |
| And thank you for watching. | |
| Share the video with all your other ICP friends, like all two of them. | |
| There's not too many of us out there, right? | |
| So we've got to stick together. | |
| Have fun running the instrument. |
Export Selection