Contents 

	macro for plotting of arrays
	champagne NMR

Thanks go to Mark Miskolzie for this contribution.

Champagne NMR
As part of the recent Grand Opening of the m400 undergraduate spectrometer, a live demonstration was given in E1-60 on what can be done with it. The sample was chosen to be champagne, indicating the happy occasion. As crazy as it might seem, but  0.35 mL of champagne were mixed with 0.35 mL of D2O and the bubbles were removed. The spectra show the expected predominance of water, followed by ethanol (11.5% according to the label) and molecules that resemble carbohydrates a lot in the 0.5% range relative to ethanol (the C13 satellites of ethanol, marked by *, serve as an internal standard for this assessment!).

The signal at 5.2 ppm is almost certainly an anomeric proton in the alpha configuration (J = 3.7 Hz). At 4.6 ppm a beta-H1 is visible too (J=8 Hz) although this one partly overlaps with a fairly intense unidentified peak. This is a very dry champagne (also fruity and well balanced according to the label!). It could be expected that a sweeter variety would produce more carbohydrate signals.

The C13 APT spectrum was recorded in only 2.5 minutes (72 scans). If run for a longer time, the smaller components also appear in the C13 spectrum, confirming for the most part their carbohydrate origin.

Needless to say that the price of champagne is not due to the water or the ethanol (nor the bubbles!) but rather due to the smaller components that provide the taste and flavor.

The removal of the bubbles might at first sight just look like a gross insult to the art of champagne making. However, something can be learned from this that goes way beyond champagne: NMR samples need to be as homogeneous as possible. That includes:

	no concentration gradients
	no temperature gradients
	no bubbles
	no solids ("fish")

Temperature gradients are probably the least obvious problem. They result from samples measured at a different temperature than they were stored before insertion into the magnet. The warm (or cold) air flows from the bottom to the top in the probe and hence the lower part of the sample warms/cools faster than the rest. Not much can be done other than waiting a few minutes. This is particularly a problem with aqueous samples. Even going from room temperature to 27C takes a few minutes of equilibration. Not waiting long enough can result in a split field, i.e. the z1-shim gradient is off and all peaks in the spectrum appear split by typically a small separation (less than 1 Hz is not uncommon). Only time and re-shimming after reaching temperature equilibrium can solve this problem.

Any field inhomogenieties, be it from the sample itself or from poor shimming, will degrade the quality of the spectrum. Bubbles in champagne are therefore nothing else but impurities in terms of NMR (sorry, bubbles!).