I got many answers from ammrl@ ammrl.org .Here is the summary without contributors' name
Xe gas sealed in an ampoule with a little bit of O2 to speed up relaxation works best. Use Cynthia Jameson's formula to calculate the shift (it's 1 ppm per amagat, I believe) based on the Xe density of your sample (this means - use an accurate, calibrated pressure gauge and a known volume to transfer the Xe to your tube). Simply dip it in liquid nitrogen to freeze the Xe and flame-seal it. If you don't break it, it will last forever.
Neat XeOF4 is the IUPAC standard. Could presumably be put into a small sealed capillary tube.
As it said in my Ph.D. Thesis, you should measure gas-phase 129Xe at several pressures, and extrapolate to zero pressure!!! This only took me something like a month to accomplish back in graduate school!! Hah!! Seriously, you should use the IUPAC standard method of absolute referencing, back to the 1H shift of TMS in the 1H NMR spectrum. The IUPAC lists XeOF4 (neat) as the zero reference, using a gamma ration of .27810186 vs. the absolute frequency of the 1H resonance of TMS. We now do all of our referencing of nuclei other than 1H using the "Absolute Referencing" to the 1H frequency of TMS (or DSS for aqueous samples). Of course to use this method, you need to know the exact larmor frquency for TMS for the same sample. I attached the most recent reference to the IUPAC method of shift referencing that we use on all of our instruments for all nuclei.
I tried to seal liquidized Xe, but the glass wall is not enough to seal it. I got the tube broken after raising to room temperature. It is better to seal into small tube with thick wall, but must check the pressure. Xe-129 was purchased in cylinder with 500 PSI presure.
I did more 129Xe than anyone... I started it all, and I taught Cathy how to do it (and got her started). The largest chapter of my Ph.D. dissertaition is 129Xe NMR. Cathy's reference was based upon mine, which originall came from Cecil Dybowski's work... where we measured gaseous 129Xe vs. pressure, and extrapolated to zero pressure. The point of absolute referencing is that you never need to acquire a reference spectrum of the 129Xe nucleus as long as you know where zero is in the 1H spectrum. There is a nice set of macros that do this automatically that we use (usually the "xref" macro). That way you never need to get XeOF4 to refrence to XeOF4.
As the 129Xe chemical shift depends on pressure. A mixture of 500 tor of Xe gas and 50 tor of O2 gives a 0.0 ppm 129Xe chemical shift. I expect it would help you.
The usual standard for 129Xe is the gas line in the limit of 0 Pressure. Of course, at 0 pressure you have no signal. By 10 atmospheres you usually have plenty of signal. I forget the shift in ppm per bar. Remember T1 is long. Look for papers by Jameson. you might also be able to use Xe dissolved in a solvent. See for example Yoong Ho Lim, and A. D. King Jr., J. Phys. Chem., 1993, 97 (47),
XeOF4 is the accepted standard, but is a labile compound that can release HF in contact with water.
When I have done 129Xe work, I have used Xe at 0 pressure as determined from a extrapolated plot of a series of pressure experiments as I varied the pressure of xenon in a high pressure NMR tube. Once the spectrometer frequency corresponding to that 0 pressure value is determined, then I set that number as my reference.
Use a high pressure tube (J Young, perhaps) and put in a couple of atmospheres of the gas.
I believe that it is a liquid. Not sure where of if you can buy it now, however. It can be made by the controlled reaction of XeF6 with H2O. If you can't get these, you can try Xe gas at high pressure.The chi value is 27.810186 MHz, so you can always reference to this. Simply take the 1H frequency of zero ppm on your machine, divide by 100.0 and multiply the chi value by this number to get the absolute frequency of zero ppm for 129Xe on your machine. Not sure what type of instrument that you are using, but referencing this way is much easier on Bruker machines than Varian instruments.We use this method all the time for things like 199Hg and 77Se.
The O2 will help for relaxation, but perhaps it may affect the shift? I imagine that there also might be a pressure correction term for Xe-129. Probably 1 atm is defined as 0 ppm, or it may be the shift limit approached as the pressure goes to zero. I expect the shifts from O2 or deviation from 1 atm to be small. There are journal articles on the temp and pressure corrections.
I noticed you are looking for a reliable Xe-129 NMR standard. I’m not sure what type of NMR experiments you are conducting, but solid XeF 2 is a reliable reference and is commercially available, but must be handled with caution under inert conditions. Please see:
Michelle A. M. Forgeron, Roderick E. Wasylishen, and Glenn H. Penner
J. Phys. Chem. A , 2004 , 108 (21), pp 4751–4758
An alternative method for reference Xe-129 NMR spectra is using the absolute frequency scale. Please see:
Solid-State 129 Xe and 131 Xe NMR Study of the Perxenate Anion XeO 6 4-
<<perxenate.pdf>> Michelle A. M. Forgeron, Roderick E. Wasylishen, Michael Gerken, and Gary J. Schrobilgen
Inorg. Chem. , 2007 , 46 (9), pp 3585–3592
Excerpt from experimental section of reference 2:
Solid-State 129/131 Xe NMR Spectroscopy. The primary NMR reference for 129/131 Xe NMR spectra, neat XeOF 4 (l) at 24 ° C, 3,27 is not readily available. An alternative 129/131 Xe NMR reference is xenon gas; however, the 129/131 Xe chemical shifts are both temperature- and pressure-dependent, making xenon gas a less than ideal reference. The frequency ratio scale has been employed in the present work and is highly recommended for referencing 129 Xe NMR spectra, as previously described. 71 This method enables 129 Xe NMR spectra to be referenced with respect to the primary Xe NMR reference, XeOF 4 [neat liquid; 24 ° C, ¥ ( 129 Xe) ) 27 810 <<XeF2.pdf>> 184 Hz] without direct use of the primary reference. The 131 Xe NMR resonance of XeOF 4 (l) cannot be observed as a consequence of rapid quadrupolar relaxation; thus, 131 Xe NMR spectra of the perxenate salts were indirectly referenced to XeOF 4 (l) using the isotropic 129 Xe chemical shifts of XeO 6…