spectroscopy, open science
@Alyx we got:
https://publiclab.org/wiki/papercraft-spectrometer
and:
https://spectralworkbench.org/
Shared with us.
I think that a dedicated scanner probably gives "better" results. And the question is whether the results you are looking for justify it.
spectroscopy, open science
@JigmeDatse holy shit! well...you might be right, but you could _probably_ get some quantitative results with this simple method, although it'd take a bit of cleverness. you'd need to make use an "internal standard", probably
but it's tempting to see if such a simple means could be used to get simple metals analyses, looking at emission lines
thanks for the tip! 👍
spectroscopy, open science
@Alyx I think that for, "good results" you would need to work with calibrating on known samples, with a known spectrum. And I think that you probably could run your tests including the known samples, and work to get any corrections you might need.
spectroscopy, open science
@JigmeDatse that's where an "internal standard" comes in. you spike the sample with a known amount of some element that isn't one that you are trying to analyze, but whose emission lines show up in the spectrum and can be used as reference lines
spectroscopy, open science
@Alyx Oh, wait, you could do that? Like run a sample to produce a spectrum you can align your lines with?
I think probably a lot of the ideas that go into this, probably are pretty old. It's just that the ease of creating the image, and processing the image has changed a lot.
In the 1980s, and 1990s (maybe a bit early 2000s) we kind of were connected, but never actually went through the processes.