This sounds like a job for ion chromatography. I do phosphate in dried plant material by GC/MS after persilylation. Ion chromatography should be much easier for aqueous solutions.

MRM quantification by LC-MS/MS using triple quad mass spectrometer will show accurate quantification. If you have got access to Ion-mobility mass spectrometer is better for your work..

GC-MS Shimadzu 2010 in SIM mode has LOD at 1 ppb level in SIM mode. If analyte have good selective ions LOD could go to ppt level. Derivatisation of phosphoric acids could be done by BSTFA or similar reagents, providing that you could remove the water from your solvent by filtering trough drying agent.

Protein PTMs (post translational modifications) are routinely identified by LC-MS methods. The loss of phosphate ion and also the phosphate ion will be identified quantitatively using MRM transitions at parent ions and one or two stable daughter ions ions with unmatched accuracy and sensitivity.

Thank you all for your input. I have talked with our LCMS guy and will try that first.

Ion chromatography is the most appropriate method in your case. With IC you can separately quantitatively determine the concentrations of H3PO2, H3PO3, and H3PO4.

I agree: Ion chromatography is a probably the most direct approach for distinguishing the three species. If you were only interested in PO4-, you could use a simpler solution like spectrophotometry (phosphomolybdates).

Here's a reference that might be helpful:

I also agree with Sofia and Robert about ion chromatography as a simplest method because you can separate and determine all three ions simultaneously. Remember that PO4- can adsorb on glass and polyethylene surfaces and you can loss some percent of phosphate-ion

Ion chromatography will be a method of choice as it can quantitate the 3 anions in the presence of each other.

It must be ion chromatography. Acetonitrile in a sample should not be a problem.

Hi

Did you try quantitative 31P-nmr spectroscopy?. Phosphorus is very sensitive atom, you can follow the species in your mixture. I think that each 31P in various species (H3PO2, H3PO3 and H3PO4) possesses a typical chemical shifts.

good luck

I feel IC is the most appropriate technique to address this problem. In aqueous medium there should not be any problem. But with acetonitrile one has to try and see.

If the low concentration of H3PO2, H3PO3 and H3PO4 is a problem for NMR determination, you may increase it /for at list one order/ by freeze drying of larger volume of the sample in an apparatus /flask/ filled with an inert gas /N2 or He and like/. To avoid adsorption of the mentioned acids on the glass surface you may make it inert by sililation prior to use. To avoid loss of analyts in the sample during freeze drying, it could be neutralized prior to drying with addition of small amount of water solution of NaOH. The freeze dryied sample may by analysed by IC, 31P NMR spectroscopy.

You can also derivatize the freez dryied sample of sodium salts of the acids to methyl-, ethyl- or benzyl-esters and analyse it by GLS equipped with EC detector, maybe even ordinary FID detector might serve satisfactorily.

More info. about the mentioned derivatization can be found in: O.P. Bethge and K. Lindstrom in Analyst, 99, 137 /1974/ -

/heterogenous alkylation of phenols and O-acids in alkalized water solution by NaOH to which tetral-buthylammónium hydrogen sulphate is added. To this solution dichlor -methane containing a few ml of CH3I or of other iodate is added. The larger is the molecule of alkyliodate used the beter is the respons of detector /FID/. The alkylesters /in your case/ will be present in the CH2Cl2 layer.

Thank you all for the information and especially the references.

Emmanuel: I have tried 31P NMR, but the concentration of H3PO3 is a bit too low to observe. I have seen O atom addition to DMSO yielding DMSO2 in 1H NMR, but that has 6 equivalent protons making it much easier to see at low concentrations. I'm having issues with the LCMS because we do not own a column which can adequately separate the different phosphorous oxy-acids.

Rastislav: I like your idea a lot. I will look into it.

My next thought it to try some phosphines, but I have to find ones that do not have photochemistry at 308 nm.

Hi Jack

I am surprised indeed few mM of a such atom ( 31P) can be detected without problem, because it is very sensitive, it is not necessary to observe proton, you just must record 1D phosphorus. May be you can increased the number of scans. However, if you observed others species thus you have not H3PO4 or less of mM level! What kinds of probe and nmr machine do you have? The best way is to use a 9.4Teslas apparatus, equipped with a 3 mm or 5mm (Broad Band Observed) BBO probehead. best regards

Emmanuel: I'm not sure what probe we have on our NMR. We're using a newer Bruker 600 MHz AVIII, but beyond that I'm not sure at this point. I can see my starting material (~9 mM H3PO2), but I only expect about 0.2-0.3 mM H3PO3. I've been allowing 30 seconds of time between scans to allow for quantitative measurement, but I ran a qualitative scan (5 seconds between scans) for nearly 250 scans and found what could be a broad peak near 7 ppm, but could also just my imagination because I want to see something. I'm going to look at phosphines now. I think seeing oxygen addition as opposed to oxygen insertion may make the chemistry more favorable.

Hi Jack- I'm late on the question here, but when we do 31P NMr we often let these go for 30000 scans.  It is hard to see mM concentrations at 250 scans.  Try doing both H-coupled and H-decoupled to identify your species.

We've also been developing an IC/HPLC-ICP-MS method for accurate P analysis to sub-uM, but that may be overkill.