it depends on how many components your crude edxtract contains and on the analytical techiques you can use. In general, TLC, GLC, HPLC and Py/GLC do not require any large amounts of the crude extract /maybe a few tenth of gram would do/. You shall have to consider also certain amount of the extract necessary for a number of preliminary experiments /to find the optimal conditions of analysis, or for possible derivatization of the extract components/.
If you are interested in structure and composition of each component you will need maybe even several grams of extract /elementary analysis of components, UV, FTIR, 1H NMR, 13C NMR, HPLC/MASS, and mybe some other methods/.
Prior to application of the mentioned special analytical methods you shall have to solve the problem of separation and isolation of each component of the crude extract using a preparative column chromatography /maybe HPLC/. I guess several tenth of gram of each isolated component should be a sufficient amount, the total amount of extract may vary from 1 to 2 g when everything goes smoothly.
The best what you should do in the beginning is to look for some articles dealing with a similar problem.
actually try to known the number of molecules in your crude by simple test like TLC.then by using extraction depends on polarity differences you can differentiate between some of them. use FT-IR with ATR to identified functional group in each type of molecule isolated .sometimes you can select specific regent to react selectively with a single molecule in the .crude to isolated it .
All depends on the no of components (here compounds) present and their quantity.
eg if their are 10 compounds in the crude and your target compound is 90% in it then you will need less amount of crude to isolate and characterize your target compound.
but if your target compound is 5% in it then you will need more amount of crude to isolate and characterize your target compound.
In my opinion to solve the above query you can go for HPLC the no of peaks will tell you the no of compounds in the crude and by observing the peak area you may decide the quantity of your target compound.
Thankyou so much for the help. Just one thing more that as I am working on antiparatic compounds, then from crude extract's HPLC, how would I know that which one is my target compound?