Discrepancies in side-by-side ICP-OES

[drolmaeye]

Executive summary: No tank can serve two ICP masters? Side-by-side ICP-OES testing using two popular brands gives overall similar results, but different results for the two elements I was focused on: aluminum and tin.

I started a frag tank about 10 months ago. Because I have very little rock, I used a popular brand of ceramic media in the sump as a suitable home for beneficial bacteria. My first ICP test (ATI) on the tank was about 7 months (September) and everything seemed pretty much okay except aluminum (98 ug/l) and tin (12 ug/l) came back a bit high (and an order of magnitude higher than my display tank which does not use the media).

Fast-forward two months (November) and I decided to confirm the results, and moreover decided to do a side-by-side comparison using two different companies (ATI and Triton). The ATI test (same company as September test) came back with aluminum at 120 ug/l and tin at 6 ug/l, in line with the September results. The Triton test came back with zeros for both aluminum and tin. (In fact, results came back zeros/undetected for *all* heavy elements/pollutants).

Note that I also sent an ATI test in for my display tank at the same time (most expensive testing day in my reefing career, but I was curious). Both aluminum and tin were measurable, but an order of magnitude less than my frag tank at 11 ug/l and 0.6 ug/l. I mention this just to point out that two tanks with the same source for water changes but differing media had quite different results.

QUESTION: Anyone have any insight as to why one test comes back with elevated levels and the other undetectable for aluminum? Is it possible one test could be sensitive to aluminum in, for example, Al2O3 whereas the other test would not be? Does it seem odd/unusual to come back with all zeros (reported as 0.00 ug/l although I don't necessarily know if the vendor claims to be able to resolve down to hundredths of a ug/l) for all heavy elements?

 

[rtparty]

Welcome to ICP testing. It isn't nearly as good as people think/claim. During and after my Ultimate Salt Test, I spent A LOT of time digging into, researching, and talking to ICP companies (including techs that run ICP machines for a living) and found they are nowhere near reliable enough at the prices we are paying.

Oceamo is likely the best available right now with ATI in second. Triton is way down the list for me after seeing way too many issues like you described. Cristoph from Oceamo is here on the board and is the only one I know of showing their testing and protocols. The rest of them don't talk about calibration, cleaning, protocols, etc so it really is a crap shoot in what we get.

I would trust ATI long before Triton and your story fits that. Ceramic media almost always come back with elevated aluminum or tin or other elements depending on what the media is made of. The bigger overlying issue is that ICP testing can not tell us the form each element is in. This is key with some elements as some forms are deadly while others are not

 

[drolmaeye]

Welcome to ICP testing. It isn't nearly as good as people think/claim. During and after my Ultimate Salt Test, I spent A LOT of time digging into, researching, and talking to ICP companies (including techs that run ICP machines for a living) and found they are nowhere near reliable enough at the prices we are paying.

Oceamo is likely the best available right now with ATI in second. Triton is way down the list for me after seeing way too many issues like you described. Cristoph from Oceamo is here on the board and is the only one I know of showing their testing and protocols. The rest of them don't talk about calibration, cleaning, protocols, etc so it really is a crap shoot in what we get.

I would trust ATI long before Triton and your story fits that. Ceramic media almost always come back with elevated aluminum or tin or other elements depending on what the media is made of. The bigger overlying issue is that ICP testing can not tell us the form each element is in. This is key with some elements as some forms are deadly while others are not

I think it is worth reiterating that overall the two tests came back with similar results for the majority of elements. For example, 432 vs. 438 for Ca, 1363 vs. 1357 for Mg, just to name a couple. I just though it was odd that a couple of trace elements came back high enough in one test to make me think about intervening, while the same elements came back undetected in the other test. I think I could alleviate some reef anxiety just by switching testing companies, LOL.

For me, one key to ICP is: do I see anything that causes me to consider taking corrective action? If I measure salinity at 35 PPT with my refractometer but my ICP comes back at 31 PSU, I will take action (reevaluate how I am calibrating my refractometer, or consider a different measurement method to cross-check). Similar thing with aluminum, it is not that I am looking for zero or a particular number, I just don't think I want to be on the order of hundreds of ug/l, so I may replace the media with cycled rock and see if a few months worth of water changes shows a noticeable decrease in .

I have also noticed the results of corrective action with ICP. Iodine was always coming back low. I started dosing iodine and now it comes back at a more reasonable value. Note I am not suggesting dosing iodine is necessary -- maybe my tank would be fine without it -- I'm just saying changes in my ICP results have typically reflected corrective actions I have taken.

That said, I fully agree that we should be careful about putting unwavering faith in the results (of *any* single test) as there can always be issues associated with samples/apparatus/individuals that lead to inaccuracies.

 

[Randy Holmes-Farley]

there are a number of issues related to using icp values. The most important of those for trace elements (such as iron) is lack of chemical form information. It is just not true that 5 ppb iron is equally bioavailable in all reef aquaria.

That effect, however, does not explain differences in test results.

There are other effects that reefers are largely in the dark about, which often includes what treatments the samples receive. Is it filtered to remove particulates? If so, what sizes are removed?At what efficiency?

Is there growth of microorganisms during the time between sampling and testing? Are they free floating or attached to the sample container?

Samples treated differently may well give different results for trace elements. Even the times and temperatures experienced between tank and test may play a role.

Then, of course, there is the possibility that results can be inaccurate, suffer interference from other things in solution, may be accidentally recorded, confused with other customers, etc.

Users should not assume that all icp results are equally likely to be accurate. It’s not like taking the temperature of a water sample. Every company likely has somewhat different protocols, calibrations, quality controls, etc.

All that said, I do not know why your sample results diverged, or which may be more accurate. In this case, for tin and aluminum, you are left little better off than with two home test kits with widely divergent results.

 

[taricha]

Fast-forward two months (November) and I decided to confirm the results, and moreover decided to do a side-by-side comparison using two different companies (ATI and Triton).

Could you share the reports, please. I'd be interested

 

[drolmaeye]

Could you share the reports, please. I'd be interested

Let me see if I can figure out how to attach them here.

For reference:
Tank ~100 liter actual total volume
Sump w/ filter sock, small skimmer, small carbon/gfo reactor
3 gallon WC weekly
dose ~6 mL/day All-For-Reef
dose 1 drop/day Aquaforest iodine
dose 0.5-1 mL/week ATI manganese

Parameters on test day as measured at home (Hanna)
Nitrate: 11 ppm
Phosphate: 0.08 ppm
Calcium: 422 ppm
Alkalinity: 8.2 dKH
Magnesium: 1430 ppm
Salinity: 35 ppt (Milwaukee digital refractometer)
Temperature: ~77 F
pH: ~8.0-8.2

 

[taricha]

You have multiple repeat tests with ATI (pretty good ICP vendor) showing elevated Al in a tank that has a physical reason why that might be the case. The elevated Al is in the range that ICP should clearly detect (Al at ~10ppb is a maybe, at 100ppb - definitely should catch it). Your repeated ATI tests also give logical low values for Al in another system that lacks that physical reason (media) for elevated Al.
Triton (a decent ICP vendor) has a high bar to report a number for traces - they prefer to just indicate zero for things that could be noise. So when ATI says, Mn = 0.05ppb - that's just noise indistinguishable from zero. In this case on this one test, I think Triton missed it.
I'd lean toward the ATI values for the stuff over a few ppb: Al, Fe, Sn.

 

[drolmaeye]

You have multiple repeat tests with ATI (pretty good ICP vendor) showing elevated Al in a tank that has a physical reason why that might be the case. The elevated Al is in the range that ICP should clearly detect (Al at ~10ppb is a maybe, at 100ppb - definitely should catch it). Your repeated ATI tests also give logical low values for Al in another system that lacks that physical reason (media) for elevated Al.
Triton (a decent ICP vendor) has a high bar to report a number for traces - they prefer to just indicate zero for things that could be noise. So when ATI says, Mn = 0.05ppb - that's just noise indistinguishable from zero. In this case on this one test, I think Triton missed it.
I'd lean toward the ATI values for the stuff over a few ppb: Al, Fe, Sn.

For the reasons you mention, I tend to agree that the ATI values are probably at least indicative of what's going on in the tank.

QUESTION: What is the general consensus? At over 100 ppb Al, is it recommended to remove my ceramic material and see if the level goes down?

I have four small fish (springeri damsel, pink-streaked wrasse, hectors goby, ruby red dragonet) as well as a sand sifting star, several hermits, several astraea. I have four small CaribSea rocks in the tank itself (maybe 4 lbs worth). I am concerned about doing something that will acutely change the existing balance between bioload and bacteria population.

Most of the coral is growing so fast I can almost see it, but I lost two torches and a trachyphyllia that really hurt (injuries involve guilt, pride, wallet), and a couple of smaller delicate frags. I want to eliminate as many unknowns or question marks as possible without upsetting the current system.

 

[Randy Holmes-Farley]

I’d certainly remove any white media with aluminum over 100 ppb.

 

[drolmaeye]

I’d certainly remove any white media with aluminum over 100 ppb.

When the forum speaks, drolmaeye (usually) listens. I will try to get it out in the next few days and will follow up sometime in the next 3-6 months with an ICP (and report back here) once there's been a chance for water changes to reduce the level.

 

[drolmaeye]

UPDATE: I removed the ceramic media back in December and have kept up with weekly 10% water changes. I recently sent out for another ICP for mostly unrelated issues (I started dosing I, Mn, and LaCl).

Aluminum: 15 ug/l (down from 120 ug/l)
Tin: undetectable (down from 5 ug/l)

I clearly see the impact of using iodine and lanthanum chloride (in terms of test results, not in terms of increased/decreased health of the tank), and I'm not sure why manganese is undetectable.

I always do a full battery of Hanna tests when I send in an ICP and the ICP results agree fairly well with my Hanna (well enough that I feel comfortable using Hanna to monitor long-term or overall trends).

Test is attached.

 

[Randy Holmes-Farley]

Thanks for the update! :)

 

[Mike Sydney]

Executive summary: No tank can serve two ICP masters? Side-by-side ICP-OES testing using two popular brands gives overall similar results, but different results for the two elements I was focused on: aluminum and tin.

I started a frag tank about 10 months ago. Because I have very little rock, I used a popular brand of ceramic media in the sump as a suitable home for beneficial bacteria. My first ICP test (ATI) on the tank was about 7 months (September) and everything seemed pretty much okay except aluminum (98 ug/l) and tin (12 ug/l) came back a bit high (and an order of magnitude higher than my display tank which does not use the media).

Fast-forward two months (November) and I decided to confirm the results, and moreover decided to do a side-by-side comparison using two different companies (ATI and Triton). The ATI test (same company as September test) came back with aluminum at 120 ug/l and tin at 6 ug/l, in line with the September results. The Triton test came back with zeros for both aluminum and tin. (In fact, results came back zeros/undetected for *all* heavy elements/pollutants).

Note that I also sent an ATI test in for my display tank at the same time (most expensive testing day in my reefing career, but I was curious). Both aluminum and tin were measurable, but an order of magnitude less than my frag tank at 11 ug/l and 0.6 ug/l. I mention this just to point out that two tanks with the same source for water changes but differing media had quite different results.

QUESTION: Anyone have any insight as to why one test comes back with elevated levels and the other undetectable for aluminum? Is it possible one test could be sensitive to aluminum in, for example, Al2O3 whereas the other test would not be? Does it seem odd/unusual to come back with all zeros (reported as 0.00 ug/l although I don't necessarily know if the vendor claims to be able to resolve down to hundredths of a ug/l) for all heavy elements?

I know TCIP well as a scientist.
The problem with TCIP is that Aluminum Oxide or Bauxite will be reported as Aluminium, in this form it does not present a problem it is abundant in the environment. The reason is people who are not degree chemists are running amok with this equipment, and when computer says , computer must be right!
Marine pure for example has an amount of bauxite in its composition, and this can cause this false reading, along with other common items such as pump wear rotors and some substrates.
Just about any metal oxides can be falsely detected as metal ions in TCIP, because the sample is turned to plasma

 

[Randy Holmes-Farley]

For the fourth time today, I do not agree with the above assertions.

 

[Mike Sydney]

For the fourth time today, I do not agree with the above assertions.

For the second time Randy , what are your qualifications and so put up or you know what

 

[Randy Holmes-Farley]

For the second time Randy , what are your qualifications and so put up or you know what

Already done in your other thread. Enjoy.

 

[Mike Sydney]

Already done in your other thread. Enjoy.

So what happened Randy?

 

[Randy Holmes-Farley]

So what happened Randy?

You’ll need to be a little more specific if you want to ask a question. :)

 

[ReneReef]

I know TCIP well as a scientist.

Could you share with us what exactly you mean with TCIP?

As far as I can tell ICP (inductively coupled plasma) or TCP (transformer coupled plasma) are used.
I can't find any mention of the abbreviation TCIP in relation to ICP technologiy.

Thanks!

 

[Mike Sydney]

Test Inductively Coupled Plasma, it is us Australians we do everything backwards or
it could have resulted from the way it started in Australia the only test available for years was Trition, and so Triton Inductively Coupled Plasma, or TICP
The TCIP was a mistype on my part, I'm not perfect, and I get ticked off when I see misrepresentation