6-CMR-8 Assignment of Carbon-13 NMR Signals
The complete assignment of all of the 13C signals of a complex molecule can be a very difficult and time-consuming process, but one that is necessary for the detailed use of 13C spectra for stereochemical determinations and conformational analysis, or isotopic labeling studies in biosynthesis or physical organic chemistry. The techniques that are used are the following:
1. Chemical Shifts. The principal method of roughly grouping 13C signals is by their chemical shift. Hybridization (sp3, sp2, sp) and electronegative groups (α effect of O, N, F, Cl, Br) cause large 13C chemical shift effects which can be used to classify groups of resonances. Within groups there are smaller effects which are useful: resonance interactions within π systems cause predictable upfield and downfield chemical shift effects. Heavy atoms (e.g., iodine, tellurium) cause upfield shifts, as does the accumulation of adjacent sterically crowded carbon atoms (branching effects). For detailed assignments involving stereochemical considerations the γ-effect is important.
Even applying chemical shift arguments in a sophisticated way does not provide a way of assigning closely spaced resonances. In most cases a complete assignment requires a group of compounds with very similar structures. Within a series of model compounds substituent effects and stereochemical effects can provide very powerful tools for assignments of even very closely spaced resonances. This is because the direction and magnitude of a change in chemical shift (Δδ) resulting from a small structural change is much easier to predict than the absolute magnitude of the shift (δ).
2. Attached Proton Tests (APT). There are a variety of techniques for distinguishing carbons signals on the basis of the number of attached protons (CH3, CH2, CH, C).
The simplest methods is to measure the 13C NMR spectrum without decoupling (coupled spectrum). This has the disadvantage that overlapping multiplets are observed for all but the very simplest molecules. Since a given carbon may be coupled to a number of protons two or three bonds removed, in addition to the one-bond coupling of interest, these multiplets can be very complicated.
More efficient are various APT tests, in particular the DEPT pulse sequences, which can be run so that only C-H carbons are visible (DEPT-90), or that all carbons with an even number of hydrogens attached (CH2) give negative signals, and all with odd (CH, CH3) give positive signals (DEPT 135). Quaternary carbons give no signal. Such techniques depend on the size of 1JCH, and can give ambiguous results when these are unusually large or small (such as for acetylenic carbons).
3. Proton-Carbon Correlation. For most molecules, the proton signals are easier to assign than the carbon signals. H-H coupling information is usually easily obtained, and supplements chemical shift arguments. Some techniques such as homonuclear decoupling and 2D-COSY experiments, which are not available for carbon because of its low natural abundance, can be easily used. Thus techniques for making correlations between proton signals and carbon signals, using JC-H, are valuable for assigning 13C NMR spectra. These experiments include various 2D heteronuclear correlation experiments (HETCOR, HMBC, HMQC, see Section 8) and can use either 1-bond or longer range (2-bond, 3-bond) CH couplings for the correlation.
4. Isotopic Labeling. Replacement of an atom attached to carbon by a lighter or heavier isotope results in splitting of the signal from C-D coupling, as well as significant chemical shift changes of the attached carbon, as well as small shift changes of other nearby carbons. If the isotopic labeling is done with high specificity, then assignment of several nearby carbons may be possible.
Replacement of H by D in some positions (deuteration) can be easily carried out, for example by base-catalyzed H/D exchange of protons α to carbonyl groups. There are two effects: the C-D signal will be split into a 1:1:1 triplet by coupling with the spin 1 deuterium nucleus (JCD ≈ JCH/6, typical coupling constant is 20 Hz), and the carbon signal will be isotopically shifted (the center peak of multiplet will be upfield from that of the C-H signal, typically by 0.5 ppm, see Section 7). If there are no protons attached to the C-D carbon, then the signal will also be much weaker because of longer relaxation times (loss of C-H dipole-dipole relaxation), some line broadening (due to relatively short deuterium T1), and loss of NOE intensity enhancement. Carbons two and even three bonds removed from the C-D carbon may also show small chemical shifts and intensity losses (due to broadening by 2JCCD and 3JCCCD, typical couplings of 1-2 Hz), thus distinguishing them from more remote carbons, which will be unaffected.
5. Shift Reagents. Lanthanide shift reagents (Eu or Pr β-diketone complexes, see Section 8.7) will cause diagnostic chemical shift changes of carbons near polar functional groups complexed to the Lewis acidic lanthanide atom.
6. T1 Effects. There are predictable effects on T1 relaxation times which can occasionally be used to make assignments of carbon signals. Carbon signals may be distinguishable either by their relative distance from nearby protons, by the variable effects of anisotropic motion on carbons on or off a principal axis of rotation of the molecule, or by differences in the degrees of segmental motion. See Section 8.1, Dipole-Dipole relaxation.
Previous Section: 4JC-H ·Home