Assuming your layers are thin, and once you start etching the Ta layer, then try to do the composition analysis, if Ta still exists do your etching again, verifying by compositional analysis. continue to do so till you hit Co Fe B layer, and thereby stop the etching treatment. Probably you  can do this with AES Auger electron spectroscopy equipment. 

Thank you for your response. Some clarifications, 

1. numbers in the parentheses are thicknesses in nm.

2. sample is [Si/SiO2]substrate/ Ta(3)/ CoFeB(1)/MgO(2)/SiO2(5) I want to etch away CoFeB(1)/MgO(2)/SiO2(5).

3. There is a photolithography step prior to etching which would define my magnetic structure.    

Hi Harsha,

Modern ion beam etching systems like the IBE's from e.g. Veeco very often have optical emission spectroscopy (OES) or mass spectrometry (SIMS) endpoint detection. With these you can very easily see the change in layers by monitoring the correct emission line (in this case for Ta, Co and Fe) or checking the change in mass of the sputtered and ionised species in the vacuum chamber. Both work very well, are inline tools and give you real-time measurement. With the signals on the screen you can determine how much over etch you want to apply. If you do not have a system with integrated endpoint detection you can fit an OES monitor or an RGA. These system are sold nowadays for less then 10k and can be fitted on any vacuum port.

Thank you peter, An OES is available in a different department, though I cannot do this insitu, I will use several samples and etch different times and then check with OES.