[durgandallas]

I am working on a personal project to retrieve various substances from dirt.
Essentially, I want to separate the mixture that is dirt into a number pure substances and then discard the organic matter. This is no trivial task, but I believe it should be made considerably easier if I know which specific compounds make up the dirt. As such, I am currently searching for a way to identify the various components. So far, only spectroscopy seems promising, but the equipment required makes me feel that I am overcomplicating the matter.
So then my question is this: what other ways exist for me to identify what compounds are present in the dirt?

I have also found that there are many university labs that will take a soil sample and, for a small price, identify the types and concentrations of various compounds in the soil - I believe this might be what I would like to accomplish. How do they do it? I'm thinking they introduce reactants that only produce a chemical reaction in the presence of a specific compound; this could be beneficial to my project, however, it might not be viable as there are too many possible compounds to be efficient.

If any clarification is required, please say so.

[Ψ×Ψ]

If you want to remove all the organic matter, just extract with dichloromethane and water.  (At least that's what they did where I worked over the summer...)

[durgandallas]

Well, its not the organic matter that I want to remove/use- quite the opposite; I want to be able to use the purified inorganic compounds for other reactions, etc, so even with the organics removed, I still have the task of separating/purifying the rest of the components of the mixture.

[enahs]

I think you will find that if you send the dirt sample into a lab like that for test, they will only report back to you the "typical" finding. Such as the nitrogen, sulfur, phosphate content, pesticides, known pollutants, etc. It will most likely not be a complete workup, that you seem to be looking for.

If there are some more specific compounds you are looking for, it might be possible to determine with simpler and cheaper wet chemistry methods.


The problem with spectroscopy is (other then the cost and complication) that you will need to disolve the compound in solution; and because "dirt" is such a complicated mixture, it could take dozens of solvents, and this means dozens of separations, as well.


You mentioned removing all organic compounds, so if you are after inorganic compounds, possibly mainly metals/metal compounds?

[durgandallas]

I think you will find that if you send the dirt sample into a lab like that for test, they will only report back to you the "typical" finding. Such as the nitrogen, sulfur, phosphate content, pesticides, known pollutants, etc. It will most likely not be a complete workup, that you seem to be looking for.

Ah, this is precisely what I was thinking- thank you for confirming it.

If there are some more specific compounds you are looking for, it might be possible to determine with simpler and cheaper wet chemistry methods.

What exactly do you mean? Like using a different method for each compound type or something similar? Do you know of any examples?

The problem with spectroscopy is (other then the cost and complication) that you will need to disolve the compound in solution; and because "dirt" is such a complicated mixture, it could take dozens of solvents, and this means dozens of separations, as well.

Yes, this was part of my hesitation to go out and buy a spectrometer in the morning (not that it wasn't tempting).

You mentioned removing all organic compounds, so if you are after inorganic compounds, possibly mainly metals/metal compounds?

Yes, mainly metals and metal compounds, such as those that contain Zn, Mg, Na, K, Fe, Cu, etc etc (I'm basically after as many different elements as I can get my hands on, and the majority are metals and the nonmetals that can be found bonded to them). Does the question have a specific point? That is, is there some method that will allow me to identify/separate metals/metal compounds more easily than others?

Edit: To clarify, the reason I indicated discarding the organic material is that I figured it would be much harder to break down into usable form, and so I would ignore it for the first few experiments and then work on it after I had the rest of the processes figured out.

[enahs]

Yes, I asked specifically about the metals, because there are many wet chemistry based methods for determining the concentration of metals in the soil.

However, unless you live in a place in the world with an unusually high concentration, you are talking about parts per million to parts per billion levels of concentrations of these metals. That is, if you could and wanted to say extract zinc from the soil, assuming 100% efficiency, to get 1 gram you would have to work with 1 million to 1 billion grams of soil. This would cost thousands in labor and fuel.


There are key places in the world where certain compounds are mined directly from the soil, not in actual "mines". But these are very rare. These companies pay chemists and geologist to go around looking for these types of places.

[durgandallas]

What exactly do you mean by "wet chemistry"? I understand that most elements will be in extremely low quantities, but I am not as interested in these trace elements; I'm more interested in the elements/compounds that are at least in concentrations of 1:10,000, for example (in fact, some might be acceptable at 1-10ppm).

Edit: Perhaps Zinc was a bad example. I'm not so concerned with which elements I get, as long as I can at least recover around 80-90%+ of the dirt in pure compound forms.

[enahs]

What exactly do you mean by "wet chemistry"?

http://en.wikipedia.org/wiki/Wet_chemistry

The spectroscopic, and other more advanced techniques would not fall under wet chemistry, even if you have to do some to get them to the point of analysis.

It is the kind you can do without spending thousands and millions. As the equipment you where talking about and going out and buying the next morning is really expensive.

[enahs]

For instance, you mentioned Copper.
http://cat.inist.fr/?aModele=afficheN&cpsidt=3669631

But each metal will require different reagents and techniques, and in this case they where removing a possibly toxic amount; it is still however nowhere near enough to produce a usable amount.

[durgandallas]

I've been looking around for ways to separate various substances from mixtures and one method that seems to be useful for a number of things is to dissolve the mixture in a liquid and then filter the liquid, so that the undissolved content remains.
What kind of filter might be used? Will just any filter paper work?

[Borek]

I have read whole thread - and to be honest I am still not sure what your goal is. I have a feeling that you lack some basic chemical knowledge yet you want to do things that are beyond the reach of well equipped labs.

Try to give some more specific examples of what you want to achieve.

Note: every dirt has different composition. Usually it will be mostly SiO₂. Organic matter can be removed either by the method ?*? described, or by roasting.