## Sunday, March 8, 2015

### Experimental chemical shifts of nitrogen (15N) atoms in formamide, acetamide, and N-methyl acetamide

I recently lamented the lack of experimental nitrogen chemical shifts for small prototypical amides such as formamide, acetamide, and N-methylacetamide. I did some digging and here is what I have came up with so far.

Before we get to the amides this very interesting paper show the effects of various solvents on the N chemical shift of ammonia relative to the gas phase value. Lower dielectric solvents do not necessarily result in a more gas-phase like chemical shifts.  So I only discuss chemical shifts measured in water for the amides since that is most relevant for protein NMR.  I use anhydrous liquid ammonia at 25 $^o$C as the reference for the same reason.

Formamide. Martin et al. measured a N chemical (downfield) shift of 260.0 ppm relative to an external sodium nitrate reference solution. This translates to a chemical shift of (376.5 - 260.0 =) 116.5 ppm.  This value is for 0.2 molar fraction formamide in water. Martin et al. show a plot of the chemical shift wrt to mole fraction and the "nitrate" chemical shift appears to drop to about 259 ppm due further dilution, so 117.5 ppm is perhaps a better value.  As I've mentioned in this post other values have been measured.

Also, the proton chemical shifts have been measured in the gas phase (thanks to +Anders Steen Christensen for the tip.

Acetamide. Kricheldorf and Haupt measured a (upfield) N chemical shift of -263.3 ppm relative to nitrate in a 25 wt% sodium nitrate reference solution. This translates to a chemical shift of (376.2 - 263.3 =) 112.9 ppm.  The value is for 0.001 M and is within 0.01 ppm of the value measured for 0.01 M and within 0.4 ppm of the value measured for 4.0 M.

N-methyl acetamide (NMA). Kricheldorf measured a (upfield) N chemical shift of -263.4 ppm relative to nitrate in a 30 wt% external sodium nitrate reference solution. This translates to a chemical shift of (376.2 - 263.4 =) 112.8 ppm.  The value is for 1.5 M at pH 7. Marchal and Carnet have measured a (upfield) N chemical shift of -266.4 ppm relative to neat nitromethane. This translates to a chemical shift of (380.2 - 266.4 =) 113.8 ppm.  The value is for 2.0 M.  Finally, Exner et al. have recently measured a value of 114.2 ppm for NMA dissolved in phosphate buffer (H2O/D2O = 95:5, 50 mM phosphate, 150 mM NaCl, pH 8) to a final concentration of 50 mM.  The same paper also describes a prediction of the N chemical shift, which was followed by this study a year later.

NMA has the additional complication of having two conformations E and Z where the methyl group is opposite and next to the O atom, respectively.  Fritz et al. have shown that 98.5% of NMA is in the biologically relevant Z conformation in DMSO.

Referencing for QM calculations
Can we reproduce these values computationally?  One of the many problems to address is the referencing.  The experimentally measured chemical shielding of liquid ammonia is known (244.4 ppm) so one option is

$\delta_x = \sigma_{\mathrm{NH_3(liq)}} - \sigma_x^{\mathrm{comp}}$

However, to increase error cancellation one could use another molecule for which gas phase N chemical shielding has been measured, e.g. ammonia (263.5 ppm):

$\delta_x = \sigma_{\mathrm{NH_3(gas)}}^{\mathrm{comp}} - \sigma_x^{\mathrm{comp}} + \left( \sigma_{\mathrm{NH_3(liq)}} - \sigma_{\mathrm{NH_3(gas)}} \right)$

One could also use other molecules and take an average.

Finally, I can recommend this interesting paper entitled Nitrogen NMR and Molecular Interactions