I agree with the usefulness of the LSC, but I would like to share a
little sample prep experience:

1.  For aqueous samples, be careful to use enough cocktail in proportion
to the sample.  We use Scintiverse BD ("biodegradable"), and 0.5mL
aqueous/buffer to 8mL BD.  For 1mL, you need AT LEAST 15mL BD.  It is
best to add BD to all vials, THEN to add samples as required.

2.  It is important to vortex each vial right after adding two phases
together, sample by sample, to optimize the sensitivity of the
cocktail.  Incomplete mixing results in inaccurate counting (visible as
separation of phases in the vial).

3.  Match the matrix of sample and standard.  If you have sample
dissolved in organic phase, add the same amount of this to the reference
vials (if you are measuring addition or recovery of tracer) to account
for the quenching that organic solvents cause in the scintillation
fluid.  If the sample phase is water, quenching seems to be much less of
a problem.

Finally, be aware that the effect of too much tracer is that the energy
distribution of the beta particles shifts to lower energy, so your
counting window should be set wide enough to accommodate the activity
that you expect to be measuring.
"Suckow.A" wrote:
>
> What a nice question, Steve, since this is exactly one of the debates in our
> lab at the moment. Not really for 14C but for tritium, but anyway: I am
> really looking forward to the debate in this forum.
>
> Here are my thoughts: (PC = proportional counter, LSC = liquid scintillation
> counter)
>
> It clearly depends on what you want.
>
> I do not know anybody selling proportional counters for radiocarbon dating
> on a comercial basis. So if you do not have a good electronic engineer and
> precision mechanic at hand, you have no choice. If you have them at hand
> anyway, you might take into consideration, that the material for a PC system
> is much cheaper than a commercial LSC. But the problem is, that you need a
> lot of experience (several years) to run a state of the art proportional
> counting system. Taking this into acount for somebody without experience,
> LSC wins.
>
> I hope the community will agree, that LSC is the technique of choice, if you
> have samples of elevated activity, as for example in biology where you use
> 14C-labelled substances. LSC have automated sample-changers which you will
> hardly find for PC. The PC inlet needs a vacuum system and it is more
> difficult to handle it. So for high activities and a sample throughput of
> tens of samples per day, again LSC clearly wins.
>
> There is an old paper coming to the result, that PC is the better system
> concerning background and stability of parameters for radiocarbon dating:
>
> Geyh, M.A. (1972): A Comparison: Proportional Counter and Liquid
> Scintillation Spectrometer for Radiocarbon Dating.  Proceedings of the
> VIIIth International Conference on Radiocarbon Dating, B, pp. 81-93;
> Wellington, New Zealand.
>
> But as you already stated, things changed in the last quarter of the
> century.
>
> And there is the question of sample preparation: For LSC you need at least
> three steps: conversion to CO2, conversion to acetylene, conversion to
> benzene. For elevated activites there are commercial systems on the market
> doing that, but I do not know anybody using these for radiocarbon dating. A
> PC system can run with pure CO2, which brings you to two steps (conversion
> to and purification of CO2). Slight advantage for the PC systems.
>
> Take into consideration that you might not want to measure only 14C, but
> perhaps also tritium. There you need again the conversion into a gas for the
> PC (there are systems running with a Hydrogen-Argon-CH4 mixture, or with
> CH4, or with C2H6), but you just mix the destilled water with the coctail
> for LSC. Clear advantage for LSC on this side. On the other hand, for
> environmental samples you will have to electrolytically enrich the water,
> since the rain nowadays has something like 10-20TU on the northern
> hemisphere, and the detection limit of a good LSC is around 5-10TU (our PC
> for tritium without enrichment have a detection limit of 1TU). If you do not
> want to enrich, PC has an advantage, if you have to enrich anyway, LSC is
> the clear choice.
>
> To bring the things to the point: if you have money and no experience, take
> the LSC. If you are looking for the scientific adventure and have no money
> at all, you might try PC.
>
> Axel
>
> Dr. Axel Suckow
> Institut für Geowissenschaftliche
> Gemeinschaftsaufgaben (GGA);
> Joint Geoscientific Research Institute (GGA);
> Geochronology and Isotope Hydrology (S3);
> Stilleweg 2;
> 30655 Hannover;
> Phone: +49 511 6432527;
> Fax:   +49 511 6433665;
> e-mail: [log in to unmask] <mailto:[log in to unmask]>
>
> -----Ursprüngliche Nachricht-----
> Von: Steve Nelson [mailto:[log in to unmask]]
> Gesendet am: Freitag, 8. September 2000 18:11
> An: [log in to unmask]
> Betreff: 14C counting
>
> I'd welcome the thoughts of the community on the relative merits of
> gas proportional versus liquid scintillation counting for 14C
> measurements--especially in the light of modern equipment.
> --
>
> ********************************
> Steve Nelson
> Dept. of Geology
> S389 ESC
> Brigham Young University
> Provo. Utah  84602
>
> voice:  1-801-378-8688
> lab:    1-801-378-5124
> FAX:    1-801-378-8143
>
> CONTROL, ALT, DELETE:  the 3 most
> frequently used keys on a PC.
> *********************************

--
Irene Ellis
Mass Spectrometry Lab
U.S.D.A. Human Nutrition Research Center at Tufts University
711 Washington St.
Boston, MA  02111

(617)556-3182 (LAB)
(617)556-3164 (NEW NUMBER)