Infundibulum, thank you very much for thinking along...

Being a science-leaning person , one of the first tests I did was what I think you would call a basic dilution curve: I added 2 Fiji print-screens where I analysed 1/16, 2/16, 3/16 and 4/16th of a capillary with a reference standard of M. Same TLC-plate, UVC-exposed as well as ninhydrin developed (for 10 min at room temperature). I chose the option '32 bit' in Fiji which maks them black & white and subjectively more rich in contrast. In the end it didn't seem to influence the results much though.

What you can clearly see is that both methods demonstrate a linear relationship of concentration and spot intensity. The UV-plate however gives a lot of noise while the curves for ninhydrin are very smooth.

I follow your suggestion that UV would make the method more accessible because peaople wouldn't have to acquire ninhydrin. However, as I stated in my notes, in contrast to pure M-solutions, UV-quantification cannot be used when using plant samples because of the aspecific 'smear' that covers the spots. Ninhydrin selectively colors the alkaloids and circumvents this problem.

'Rolling ball background subtraction' sounds totally cool by the way. Since blue.magic has got expertise in image processing, he might have suggestions as to the practical application of the technique. Fiji does have a 'background subtraction' option. I tried it but it doesn't make much of a difference. It can be tweaked however, so I might just have applied it in an ineffective way...

Hi An1cca,

It is really useful to know that the relationship is linear - and I hear you about the drawbacks of just UV.

Do you think that if the plant material gets pre-extracted somehow the smear will disappear and will make the TLC bands clearer? What I have tried in the past is to prepare crude extracts from small amounts of plant material (10-50 mg) using aqueous solution of high ionic strength (10% NaOH and 10% NaCl), then use acetone as the non-polar solvent. Acetone separates very well from aqueous solutions of high ionic strenth and carries along freebases leaving behind a fair amount of interfering substances.

Indeed Infundibulum, I thought the same thing. Apart from avoiding smear, I also tried to micro-extract because it would eliminate the 'snot' problem when using fresh samples. I tried a few things, the most succesful was a micro-extraction in an insulin-needle where a dried (heat-gun) and pulverised sample was sandwiched between 2 cotton balls. This setup was packed with the plunger and then the syringe was filled with a solvent of choice. This was allowed to gravity-extract the sample and the extract was dried immediately on the plate by putting it on a steam-heated surface. The whole setup (and thus solvent) was kept hot by putting a glass bowl over it with spacers to allow for venting of the solvent. After some experiments it became clear that the repeatability of the process and the following reproducibility of the results was hard to attain. Doing quantitative (!!) micro-scale extractions is not easy at all. In fact, I abandoned this research-thread in favor of the Direct Chromatographic method as explained in the document. Direct as in 'no extraction' and thus no messing around with milligram quantities and unacceptable error-margins. Getting your hands on ninhydrin (in any form) is far easier than getting microscale-extractions right. I searched for a picture of the above process but could only find one of a preliminary setup. But I think you'll get the idea.

On the use of acetone: since we are really interested in quantifying M-content, I would be hesitant to use anything that is even slightly miscible with water. It would take quite some experimentation and extrapolation to check just how much M is left in the aqueous layer. I wouldn't make it too complex: take a punch, denaturate slime by heat if necessary, homogenize, and spot!