Your Spectrum

We should perhaps make a few important points before going any further - the title of this chapter is highly ambitious We certainly cannot promise to turn you, the reader into expert interpreters in the time it takes you to read this section. Experience is essential and to become really proficient in this area, you need to critically examine literally thousands of spectra. However, be that as it may, by establishing some sound principles and cultivating a critical approach to the spectra you encounter, this book should prove useful in helping you along the way. 

Our initial observations are aimed at improving your understanding of 1-D proton spectra, though many of the principles we will try to establish will be equally applicable to other nuclei too. We will discuss issues specific to 13C interpretation in Chapter 9. 

As we mentioned in the Introduction, it is ironic that one of the major problems encountered when dealing with NMR spectra, is the sheer quantity of information that you are presented with. Unless you are practiced in the art of interpretation, you may find yourself swamped by it. Clearly, a methodical and universally applicable approach would be advantageous. There is not necessarily a right or a wrong way to approach a spectrum, but some ways are probably better than others These are our top ten recommendations, for what they are worth. 

Take a moment to survey the spectrum and ask yourself if it is fit for purpose Of course, if you have run it yourself, then it should be fine but this may not always be so with walk-up systems. Is the line shape and resolution up to standard Has the spectrum been phased correctly Is the vertical scale well adjusted so that you can see the tops of all the peaks (except perhaps, obvious 

Essential Practical NMR for Organic Chemistry S. A. Richards and J. C. Hollerton 

---

Of course, you don t have to use either of the above standards at all. In the case of samples run in deutero chloroform/methanol and dimethyl sulfoxide, it is perfectly acceptable, and arguably preferable, to reference your spectra to the residual solvent signal (e.g., CD2HOH) which is unavoidable and always present in your spectrum (see Table 2.2). These signals are perfectly solid in terms of their shifts (in pure solvent systems) though the same cannot be said for the residual HOD signal in D2O and for this reason, we would advise adhering to TSP for all samples run in D20. 

Probably the most basic parameter that you will be able to set is the number of spectra that will be co-added. This is normally called the number of transients or number of scans . As mentioned elsewhere in the book, the more transients, the better the signal to noise in your spectrum. Unfortunately, this is not a linear improvement and the signal to noise increase is proportional to the square root of the number of transients. As a result, in order to double your signal to noise, you need four times the number of scans. This can be shown graphically in Figure 3.1. 

As you are no doubt aware, integrals are one of the key parameters in the interpretation of proton spectra and are pivotal in quantification. They measure the area under a peak and this is directly proportional to the number of protons (in the case of proton NMR) in that environment. Most software will automatically try to identify the peaks in your spectrum and integrate them for you. If you need to do it yourself, then it is a fairly trivial matter of defining the start and end point of the integrals of interest. The only complication is that you may need to tweak the slope and bias of the integral. This should be unnecessary if you have got the phase and baseline of your spectrum correct. If you find that you need to adjust slope and bias, we suggest that you go back and try to sort out baseline and phase a bit better. 

Does your proposed structure exhibit any special features likely to have a significant effect on your spectrum (e.g., chiral centres, sites of potential restricted rotation, abnormal stereochemistry, etc.)... 

If you have any reasonable cause for doubt (e.g., because some key signals in your spectrum are obscured, etc.), would the acquisition of more NMR data be helpful If so, consider exactly what you wish to achieve and select the appropriate technique and gather the data. 

Re-evaluate all data again and again until you are as happy as possible with all aspects of your spectrum. Guard against complacency Is it watertight Check on this by asking yourself if you would be happy to stand up in a court of law and defend your efforts. 

This is demonstrated once more with our familiar morpholine compound in Spectrum 9.2. The DEPT sequences are of course, still relatively insensitive, though they are probably a little more sensitive than the standard 1-D, fully decoupled 13C spectrum. We find it convenient, particularly with complex molecules, to combine the 1-D 13C spectrum with the DEPT-135 spectrum, which is plotted above it at the same expansion, of course This enables you to differentiate the different types of carbon in your spectrum at a glance. 

Once you know, or can guess, the field limits of your spectrum, setting the center field and sweep width values is not very difficult. The center field corresponds to the middle of the spectrum and a sufficiently large sweep width chosen so that all of the spectrum is recorded. 

If you do not know the field range occupied by your spectrum in advance, the center field must be chosen by educated guess set the sweep width 2-4 x greater than the expected width. Hopefully, you will see at least a piece of your spectrum and can make appropriate adjustments to zero in on the correct settings. 

In most cases, you will use the first harmonic and the normal first-derivative of the absorption spectrum will be presented. If your spectrum has very good S/N and has some regions where you would like better resolution, a second-derivative presentation may help. However, second derivatives from second harmonic detection are very costly in terms of S/N ratio and so you really do have to have a strong signal ... 

IR interpretation can be as simple or as complicated as you d like to make it. You ve already seen how to distinguish alcohols from ketones by correlation of the positions and intensities of various peaks in your spectrum with positions listed in IR tables or correlation tables. This is a fairly standard procedure and is probably covered very well in your textbook. The things that are not in your text are... 

Carries out a phase correction on your spectrum using the 0 and the T order parameters last defined during interactive phase correction (see chapter 5). 

The Linewidtli option in the Analysis pull-down menu allows you to measure the linewidths at half height. The ability to recognize different linewidths in your spectrum is important because it may indicate additional molecular processes going on in solution. Broadening of some of the resonances may be indicative of additional non-resolved couplings, dynamic processes or different types of relaxation mechanisms selectively affecting a particular observed nucleus. This option will also be used to estimate linewidths for use as input data for WIN-DAISY as described in detail in Modem Spectral Analysis, volume 3 of this series. 

In the box Layout Elements to be plotted choose Parameters and Title. Click on the Edit Title... button in the right side button panel to enter a title. Clicking on the appropriate buttons, define the acquisition, processing and plotting parameters that you require to be printed with your spectrum. 

Following the instructions given above, select one single row of your spectrum which includes signals as close as possible to both ends of your spectral window. K this is not possible select up to three different rows for this purpose. 

Use the same series of data and follow the same procedure as before to try out the four Traficante window types. Use the Interactive option, as a starting value set LB = 0, which yields a horizontal line and corresponds to no weighting. Increase and decrease the LB value in small steps using the scroll bar and try to predict the effect on your spectrum. Use different values to enhance the resolution, store the results and compare the resulting spectra using the multiple display. 

---