[nathanielngliwen]

I'm currently on a project which involves analysis the compounds present in pigments. Currently we've isolated pigments from oil paints and we have totally no idea what possible tests we could run on these pigments. The internet offers many many suggestions for qualitative analysis. I'd like to enquire which tests would be the most efficient in telling us what compounds are present. Currently we're considering these two forms of analysis:
1) MNR (magnetic nuclear resoance)
2) Raman effect

Finally, does anyone know a website which can provide me with compounds present in various common pigments so I can run simple tests, such as those found in QA of A levels.

Thanks!

[billnotgatez]

http://www.chemicalforums.com/index.php?page=searchmsds

Have you looked at the MSDS or the products you have selected?

[Yggdrasil]

For identification, good analyses to start with would be NMR (nuclear magnetic resonance is the more common term) and GC-MS (gas chromatography-mass spectrometry).

[Borek]

Do you have idea whether the pigment is organic or inorganic?

[hmx9123]

MS can be handy; X-ray powder diffraction is good for confirming your material.  Use simple chemical tests to discern solubility, metals present, etc., assuming that you have enough sample.  UV-Vis may also be a good way to detect the materials (since they're colored), assuming that you can find references for the pigment materials.  I'd also advise taking a look at IR if your sample is organic--it can help with some simple functional groups.  Basically, you're going to have to run the gamut to make sure of what you have.

[Borek]

Use simple chemical tests to discern solubility, metals present, etc.

Bingo

My $.02 - start in library. How old is the paint? Different pigments have been used in different times, but these things are described in books about history of art. If the paint is new, you will be able to find books on paint technology they will contain info on contemporary pigments. That will greatly reduce number of possibilities, making your analysis much easier. Sure, it may happen you will find something new, but that's very unlikely.

[nathanielngliwen]

Thanks for the help thus far. Our guess is that pigments are inorganic. I mean what we've learnt is that transition metals usually emit wavelengths of different colours within the visible light spectrum. Organic pigments? Honestly that's pretty new for me.

[UnintentionalChaos]

They mean organic pigments like anthraquinones (although aside from alizarin lake, I don't know what else would be in oil paints) and some of the more recent (cheaper) oil paints use large organic molecules, most likely as metal complexes such as in phthalocyanine blue. What pigments are in a paint depends on what colors, what companies, and what age. I have some (30 or so year old) red oil paint from my great uncle that I know is real vermillion red (HgS...I intend to eventually experiment with one of the tubes), but no company that I can find will sell it anymore for obvious reasons. Most modern "mineral" reds are Cadmium Selenide blended with Cadmium Sulfide. Most also have an inert batrium salt thrown in. Old greens may be Paris green- a blend of copper acetate and arsenic trioxide. Again, you wont see it anywhere today.

[Yggdrasil]

ICP-MS could be useful to identify metal ions present in your pigment.

[hmx9123]

Paris Green is used in the pyrotechnic industry still, though it is copper(II) acetoarsenate, not a mixture, as you stated, so it may be two different materials with the same name.  You can occasionally find it from pyrotechnists.  I have a pound or two at home.

[nathanielngliwen]

Alright, now i'd like to ask what if i'd like to find out exactly the amount of each compound present in each pigment? Like the percentage wise?

[nathanielngliwen]

http://www.chemicalforums.com/index.php?page=searchmsds

Have you looked at the MSDS or the products you have selected?

I've checked it unfortunately they do not show the specific compounds in the paints, specifically ones which show color. they'd usually provide me with the name of the solvents

[hmx9123]

In order to find the exact % of each component, you first have to identify them.  Then you can do GC-MS if the compounds contain compounds compatible with that spectroscopy, or go back to an old-fashionied elemental analysis and back calculate your %ages.  That is tricky, though.  Otherwise, you're stuck with laborious wet chemistry of separating the components and weighing them.