what are the appropriate glasswares for volume measurement

[jiangds06]

used for standard solution preparation. what are the appropriate glassware can be used? Volumetric flask, defnitely. but how about small volume measurement, such as 100ul?

[James_Ball]

The Hamilton Microliter syringes I believe are considered as Class A volumetrics. We use them for making most of our low volume dilutions. If you have good technique you can pull solution and diluent into the same syringe to make a dilution.

[KM-USA]

This topic probably should be in the Sample Preparation section, but that's not my problem.

We've used the Hamilton dispensers almost 20 years, and find them to be reliable and accurate. Of course, in todays Uber-regulated environment of GMP, GLP, ISO, we do send them back to Hamilton every two years for an "official" test. Otherwise each 6 months we weigh water a bunch of times and document.

I do have a question, of course semi-unrelated. Does it state anywhere in USP, ICH, or FDA about how to determine if a class A volumetric pipet is dirty or unsuitable for use? We have some such pipets that are invisibly coated from one of our product solutions, and subsequent uses leave bubbles of that new solution entrapped, so we know that we are not delivering the "full" volume stated. Should we do a study and see if weighed water is still within the stated class A volume tolerance and ignore the bubbles of solution entrapped, or do USP, ICH, or FDA somewhere state that visibly dirty glassware cannot be used? If so, reference please.

We have tried various solvents to clean out the coating, but nothing seems to work.

We don't yet have an SOP on electronic digital pipet use, so we cannot instead use those yet and let the disposable tips be the cure. Thanks.

[MaryCarson]

After switching over and doing IC-ICP-MS for a while, I am now a convert to using a balance to prepare serial dilutions. When you have an instrument that can have over 6 orders of magnitude of linearity, using volumetric pipets/glassware that maybe is accurate within just 1% doesn't cut it anymore.

[James_Ball]

This topic probably should be in the Sample Preparation section, but that's not my problem.

We've used the Hamilton dispensers almost 20 years, and find them to be reliable and accurate. Of course, in todays Uber-regulated environment of GMP, GLP, ISO, we do send them back to Hamilton every two years for an "official" test. Otherwise each 6 months we weigh water a bunch of times and document.

I do have a question, of course semi-unrelated. Does it state anywhere in USP, ICH, or FDA about how to determine if a class A volumetric pipet is dirty or unsuitable for use? We have some such pipets that are invisibly coated from one of our product solutions, and subsequent uses leave bubbles of that new solution entrapped, so we know that we are not delivering the "full" volume stated. Should we do a study and see if weighed water is still within the stated class A volume tolerance and ignore the bubbles of solution entrapped, or do USP, ICH, or FDA somewhere state that visibly dirty glassware cannot be used? If so, reference please.

We have tried various solvents to clean out the coating, but nothing seems to work.

We don't yet have an SOP on electronic digital pipet use, so we cannot instead use those yet and let the disposable tips be the cure. Thanks.

I remember back in college we used Potassium Permanganate solutions to clean pipettes, you can also use chromic acid or other concentrated acids for tough to remove contamination. Of course if you are analyzing for chrome or potassium or manganese these may not be a good remedy.

[lmh]

After switching over and doing IC-ICP-MS for a while, I am now a convert to using a balance to prepare serial dilutions. When you have an instrument that can have over 6 orders of magnitude of linearity, using volumetric pipets/glassware that maybe is accurate within just 1% doesn't cut it anymore.

I suppose that's always going to be the problem with serial dilutions: they're a cumulative error, but usually cumulative in a way that doesn't leap out and hit you (which can be a bad thing: it's best if errors are obvious). But is there anything wrong with doing them like trains, or lifts in a very high building: do a couple of "express" dilutions through to several points, and smaller dilutions from these locally, thereby covering the entire concentration range without any individual standard being more than 3 pipettings away from the original stock?

I've never (not intentionally, at least) worked in a 6-order-of-magnitude situation and would, frankly, find it scary...

[MaryCarson]

Once you get used to it, it's actually pretty fast to prepare a set of calibrants at the balance, enter the weights in a spreadsheet, and calculate exact concentrations with 3 or 4 significant figures. In my case, I can also cut and paste from the spreadsheet into the calibration table of the data processing software. If I goof (no, never!) on the volume of dilution, no problem. I still have the actual weight and just end up with a slightly different calibrant concentration.

Carry-over issues can become very significant when you have that much dynamic range to play in.

[Peter Apps]

I'm with Mary on this. The most accurate and precise instrument in any lab is the analytical balance, and as long as you record actual masses you do not need to waste time getting close to a target addition. You can use the same approach for sample prep.

When working with volatile solvents you need to take some precautions: APPS, P.J. and ARCHER, M. 2010. Evaluation of the source of bias caused by losses of solvent vapour during sample preparation. Journal of Accreditation and Quality Control 15: 171–180.

Peter

[lmh]

ooooh, volatiles. My nightmare. At a trivial level, I haaate making ammonium acetate and ammonium formate buffers at the ammonia pKa; means measuring ammonium hydroxide solution and ammonium salt. The hydroxide solution is of pretty-vaguely unknown concentration because it evaporates, it is murderous to pipette because it dribbles as the gas comes out of solution, and the salt is hygroscopic and half full of water no matter how carefully I try to store it, so it really comes down to mixing two more-or-less random solutions and hoping my pH meter isn't too far off (a big hope) so at least some feature of the final solution is defined in some way.

Yes, I'm a big fan of weighing anything that can be weighed (I suppose if I'm being super-finicky I could mutter darkly that the 1ppm mass accuracy I'd get from a good modern MS instrument is comparable to, and probably verrry slightly better than a top-end microbalance, so there are two instruments fighting it out for top-spot on the precision/accuracy podium, albeit in different arenas: one of them can't make solutions.)

[Peter Apps]

Yes, I'm a big fan of weighing anything that can be weighed (I suppose if I'm being super-finicky I could mutter darkly that the 1ppm mass accuracy I'd get from a good modern MS instrument is comparable to, and probably verrry slightly better than a top-end microbalance, so there are two instruments fighting it out for top-spot on the precision/accuracy podium, albeit in different arenas: one of them can't make solutions.)

So that particular soap box is starting to wobble slightly. The comparison might turn out to be quite interesting - unless the MS can detect a single ion (which I don't think it can, can it ??) you are not getting a measurement of mass for a particular ion (in the sense of a single changed particle) but for a population of ions, all of which have a mass between an upper and a lower bound. The balance traces back to the Kilogram, and the MS (presumably ?) to the Mole, both of which have had metrologists jumping up and down (to the extent that jumping is compatible with such a careful procedure) for several decades.

I vaguely recall that someone (Graham Cooke ?) did use an MS to make solutions - he soft landed charged proteins into drops of PEG having separated them by MS.

Peter

[lmh]

oooh, good points. As I was typing that, I was also thinking about the common features of "good" measuring devices, and that many track back to a measurement of frequency (or time, which is the same thing since all good clocks count oscilations), and for the whole of history, we've been pretty good at measuring time/frequency. Which made me wonder how analytical microbalances work. Time for wikipedia and google!

I suppose the mole is slightly less reliable than the kg, because if it's still the number of atoms in 12g of carbon 12, it's entirely dependent on the kg, but also requires someone to be able to count atoms. And have the metrologists sorted out why the standard kg was losing mass yet? Just the admission that the standard was losing mass is actually an admission that it isn't the standard, in that if it were, all the other ways of assessing its mass would be increasing... That's another story that's gone cold.

The nice thing is that because molecular/atomic masses are relative, I suppose the accuracy of measurement depends entirely on the linearity of the mass spec, and by using lock-masses, we're getting as close to an authentic standard as one can (but some of my colleagues here were idly trying to calculate how far mass accuracy would have to improve before you could determine bond energies by the difference they'd cause in mass (thanks, Mr Einstein!)... but for the moment, I think we can assume that bonded carbon weighs the same as un-bonded. Actually, come to think of it, this means the mole isn't actually defined with perfect precision at the moment. It should be the number of carbon atoms in 12g of carbon vapour at zero pressure (make that!) or very slightly more practically, 12g of 100% pure graphite with no edges, or some such standard).

Oops, time to stop.

[James_Ball]

And here I though trying to make a standard by weighing Freon12 into methanol was difficult

Seems easy now.

[Peter Apps]

The saga of redefining the Kilogram in terms of a natural constant (it is the only one that still relies on a physical artifact) is hideously complex. A silicon28 sphere with a known (counted) number of atoms and a very precisely measured diameter has recently been produced that might link the mole to the kilogram. The other way is with a Watt balance that relates force to Coulombs to the Mole. The current problem is that the two approaches do not agree quite closely enough.

Google silicon kilogram and go from there.

Peter

[lmh]

Strewth, I'm never going to get my work done if people send interesting google-hints like that! That used up a good half-hour of genuine work-time, part of my lunch-break, and a follow-up this evening. And now I'm wondering if our microbalance is just about good enough to detect the change in the strength of the earth's gravitational field between its bench and the floor. It should definitely be able to do upstairs/downstairs. Trouble is, it's far too expensive for me to dare try it. There are experiments whose relevance my boss perhaps wouldn't fully understand...

[Peter Apps]

Yep, it gets scary out there on the margins ! The reason that standard weights have specified materials and shapes, rather than being assorted chunks of corrosion resistant materials is that on the best balances a tall thin weight weighs different from a short fat one. Strictly, if you get a six-figure balance certified it is valid only for the spot that it was certified in. If you move it you should get it re-done.

Peter