| SETTING UP YOUR EXPERIMENT | ||||
| Designing an NMR experiment: | ||||
| Nuclear Overhauser effect (NOE). | ||||
|
When 1H's are decoupled from another nuclide X (such as 13C or 29Si), the J (or scalar) splittings disappear, and the peak collapses to a singlet at the average frequency. The nuclear Overhauser effect, or NOE, is another consequence of the 1H decoupling process, and it leads to changes in the intensities of the observed X resonances. When 1H's are decoupled from 13C, all of the protons are saturated, i.e. the low- and high-energy populations are forcibly equalized. However, this places the entire system (including all the NMR-active atoms in the sample) into a higher-than-normal energy state, since there is usually a slight excess in the low-energy condition. The entire system can only rectify this unbalanced situation by shifting more of the other nuclides (13C's, for example) into lower-energy states. More of these nuclides are, therefore, able to be excited by the rf pulse, producing a larger signal than would be observed in the absence of decoupling. For 13C, the maximum observeable NOE is about three. In other words, some peaks may appear three times are large as expected in the presence of 1H decoupling! |
||||
|
At
first glance, this might appear to be an unquestionable benefit. Aren't
larger signals always better? For 13C, the main problem arising
from the NOE is that not all carbons will achieve the same degree of
enhancement. There are many factors which influence the exact value
of the NOE for any nucleus, but in general, 13C's with more
directly attached 1H's experience larger increases (CH3
> CH2 > CH, for example); quaternary carbons see little
or no effect. If one is trying to compare relative peak areas to aid
in identification or for quantitative analysis, this uneven enhancement
will complicate spectral interpretation by skewing the areas. Furthermore,
for certain nuclides with negative magnetogyric
ratios (such as 29Si or 15N), the
NOE can actually reduce the signal intensity, even to the point of making
a peak disappear!
|
||||
|
Fortunately, it is possible to suppress the NOE during an experiment, while still decoupling 1H's. This is accomplished by gating the decoupler, as shown below: |
||||
|
|
||||
|
This sequence exploits the fact that the NOE requires time to develop, whereas decoupling takes places instantaneously. The decoupler is only on during the relatively short acquisition period, not during the much longer pulse delay (these time periods are not shown to scale in the figure above). In this way, there is insufficient time for an appreciable NOE to develop, but the resonances are still completely J-decoupled. The decoupler should always be gated whenever highly accurate quantitation of peak areas is desired. |
||||
|
|
|
|||
