Chromatography Forum

Going back to the original thread:
- colleagues who think that if you can't detect any analyte then the concentration in the sample must be "zero" or similar type assumptions.

Ah yes, there's nothing better than an analytical report that says "0." Or, better yet, "0.000." (Yes, I've seen those.)

At the other end was the filter guaranteed to catch 99.999999% of particles. Yes, that's eight nines.

I once heard from an engineer about a gas cylinder that got dropped near the water's edge and ruptured. It sped through the water like a torpedo. I guess it might have been a scuba type tank.

They'll speed like torpedoes through asphalt. Restrain your high-pressure cylinders, folks!

WK wrote:

- colleagues who think that if you can't detect any analyte then the concentration in the sample must be "zero" or similar type assumptions.

I had the opposite experience recently running a LC-MS cleaning method for some plant guys. When I asked them what limit they wanted so I could make standards they said ideally they wanted me to clean until we saw nothing... we patiently explain we were seeing ppb levels in 50L of solvent and that a limit of nothing seen is a meaningless daft limit under most circumstances.

We have a very awkward customer who has very little understanding about analytical methods and possibilities of experimental errors or instrumental errors: She had me checking and rechecking (went straight to top management) a sample that we had recovered for them because the crude had 0.02% (200ppm!) less water than what we reported in the C of A (she was obviously trying to get a discount!). I couldnt believe that I was actually having to recheck something over and over again because of such a slight difference! And as you all know, when you want results to be exactly the same , they never are!!!

Salma

That's always annoying Salma! We had our quality assurance people in the office worrying that we had a 25% error between two preparations of the same sample. They didn't seem to understand that the difference between 0.20ppm and 0.25ppm is minute, just the percentage looks big. In anny case the pass is not more than 5ppm so both were miles in anyway, I guess QA were short of a job for the afternoon.

[quote="WK"]

- colleagues who think that if you can't detect any analyte then the concentration in the sample must be "zero" or similar type assumptions.

quote]

I asked an organic chemist once why he had his GC attenuation set so high.

Answer:" I don't want to see the minor peaks - it makes my synthesis look less pure"

Ralph

Sigh...

My girlfriend is a synthetic chemist, we have this sort of conversation all the time.
They have a generic 8 minute gradient method on a short column for screening really early phase reactions, but it's basically only useful for checking reaction have gone to completion.
They always complain the first time we get samples and develop a specific method that they prefer the 8 minute results as they have less imps. Look under the main peak guys! It's like my idea of tidying my room when I was younger shoving everything under the bed out of sight.

I have chemists who think that the error on an adhoc method is two high when I tell them that 200 and 250ppm are essentially the same result. However when I tell them that the results on two samples are 0.020 and 0.025% they relpy that the results are the same. It is very tempting to report all results in percentage rather than ppm.

I MUST RESIST!!!!

GCguy

Ah... chemists, you can't live with them but they're paying half the rent.

That might just be me, but I doubt it...

'where's that leak coming from?' is perhaps the thing you least want to hear arriving at work in the morning...

I was going to say: The sound of a low pressure limit alarm echoing down the hall when I get in on a Monday morning.

Especially on those pesky Waters Alliance systems, no leak detector to turn the flow off and an column oven full of absorbent foam stuff. They take ages to dry. (Though of course they provide high quality chromatography and I wouldn't want to insult their performance in any way !)

When I was a student on industrial placement my housemate (who was also a student) came out of the lab saying, it is ok to wash that KF machine with water right? Took us 3 weeks of conditioning to get it useable again...

I like that last one PJ8, I regularly use a KF and everytime I think the beaker looks really filthy, I have to remind myself not to wash it with water!!! Experiece is probably a better teacher than common sense heh?

That remembers me that I heard once someone asking after he/she/it finished a KF titration:
Can I report that 40 ppm water?

my first job was in a QC lab at a time when an integrator with a paper roll was the standard device for data recording. The chart speed was set at 1 cm/min all the time. to save paper, they would re-roll the paper and use the back side again. I thought I had a better way to save paper. The first peak was around 11 min, so I changed the chart speed to 0.1 cm/min for the first 10 minutes, still use 1 cm/min during the peak elution window. the next morning the lab manager called me in: "your peaks were coming out too quickly and your run time was too short!" I told him/her what I did and I got a suspicious look. we were in a non-English speaking country, the instrument manuals were in English. The next week I fixed the GC performance issue (changed the septum!). Then all the sudden the manuals, along with the big English dictionary that used to sit on the manager's desk, all disappeared.

I left that place in one year. I was too much a threat to the manager’s sense of job security.

1. To save you some time I took out all the little screws and have the four rods lined up nice and neat in a row on the bench.

2. How do I get the aspirin in the GC injector? The hole seems too small.

3. Half a sec while I light up, then I'll change the hydrogen tank.

jzt: I understand, and I empathize. I think that's the issue between our QA and me: no one else knows enough to push back, even with all our PhDs.

I've also heard a similar comment (years ago) about increasing the chart speed to increase resolution (and that WAS from a PhD).

"Let's run it by QA to get their input"

"Let's run it by Statistics to get their input"

"You solved the problem in a timely manner, so it must've been easy"