NMR spectra have signals, not peaks. Typical 13C NMR chemical shifts show up in four regions:
Aliphatic (sp3) carbons give rise to a signal in the 0 - 50 ppm region; however, if an aliphatic (sp3) carbon is next to a high-χ atom the carbon signal will show up above 50 ppm.
Aromatic or alkene (sp2) carbons show signals in the 100 - 160 ppm region.
Carbonyl (C=O) carbons show signals in the region between 165 - 220 ppm.
For historical reasons we call a signal at high ppm values being "downfield", a signal at low ppm values "upfield".
Alkyl carbons: electronegative elements (O, F, N, Cl) shift an sp3 signal downfield, whereas the signal of a carbon next to an electropositive element (Si, Li) will be upfield.
The inductive effect is additive.
Alkyl substituents also cause a carbon resonance to shift downfield. The effect is small, but noticeable.
CAUTION: the inductive effect argument no longer works for aromatic/alkene carbons. Only the chemical shift of an sp2 carbon bound to an oxygen by a single bond is noticeably downfield in the characteristic region of 150 - 160 ppm. The effect of other electron-withdrawing or electron-donating groups is not easily rationalised.
The chemical shift of a substituted benzene can be estimated as 128 + value for C1 to C4 from the table below. If the benzene has more than one substituent, the contributions from each substituent are roughly additive.
CAUTION: The chemical shifts of a double bond carrying a polar group can be rationalised by looking at the resonance structures. A carbon with a δ+ charge will have a 13C NMR signal at higher ppm values, whereas the signal of a carbon with a δ– charge will turn up at lower ppm values.
13C NMR chemical shifts can be predicted quite reliably using, for example, ChemBioDraw Ultra, or estimated by looking up similar compounds in a database.
DEPT spectra allow you to identify CH3, CH2, CH and C signals. The top DEPT spectrum shows CH and CH3 signals up, whereas CH2 signals face upside down and are thus easily recognised. The bottom DEPT spectrum shows normally only CH signals (up), so that you can use it to distinguish between a CH and CH3. Quaternary carbon (C) signals without any hydrogens attached do not show up in a DEPT spectrum and are therefore deduced, indirectly, by comparing the DEPT with the broadband-decoupled 13C NMR spectrum.
Every non-equivalent carbon will give rise to a 13C NMR signal. Always tabulate ALL experimental 13C NMR signals (but NOT solvent signals) and DO NOT quote ranges.
Unless you run a C,H-correlation (2D NMR) of your compound, you won't be able to unambiguously assign every carbon to a particular signal, particularly when shifts get close (<10 ppm). This is not essential. What is important is that the tabulated number of signals equals the number you would expect theoretically, by looking at the structure and counting the number of non-equivalent carbon atoms. Use DEPT spectra to do an additional carbon inventory (i.e. count the numbers of CH3s, CH2s etc.). This is the first thing a trained chemist will do when he or she checks your NMR data. If the numbers do not match, then the sample may be impure, or some signals might overlap, or are missing or, worse still, the proposed structure of the compound is wrong.
It is possible to draw conclusions about the substitution pattern of a benzene from the number and type of aromatic 13C NMR signals. Symmetry will always reduce the number of signals, whereas lack of symmetry can result in up to 6 aromatic 13C NMR signals for a substituted benzene.
13C NMR spectra allow you to distinguish between ketones, aldehydes and carboxylic acids (or carboxylic acid derivatives).
13C NMR chemical shifts and multiplicities of common deuterated NMR solvents
The ortho-proton of an aromatic ring is most affected by the nature of the substituent.
The chemical shift of alkene protons is also influenced by electron-withdrawing/donating substituents. Note that the signal of the proton furthest away from the substituent is shifted most.
The chemical shift of OH and NH signals can vary over a wide range. Signals are usually broad singlets since exchange with e.g. water or acid wipes out all coupling. The OH and NH chemical shift depends on solvent, concentration, temperature, water content. Finally, integration is often too small due to integration limits being set too close to the signal.
D2O test: adding a drop of D2O to a solution of a sample in CDCl3 and shaking it, will convert OH and NH groups into OD or ND groups and, as a result, their signals will vanish when you subsequently record the 1H NMR spectrum again.
To a first approximation, protons just couple to protons on the carbon(s) next door. If there are n protons on one or more neighbouring carbons, coupling splits the signal into n + 1 lines.
When a proton signal shows no recognisable symmetry or multiplicity pattern at all, we refer to it as a multiplet.
In the absence of roofing or higher-order effects, line intensities in a multiplet follow Pascal’s triangle.
However, when coupling spins get close to each other in chemical shift, line intensities will be distorted (due to roofing). An example of this are two doublets which turn from an AX to an AB spin system where the chemical shifts are no longer at the centre of each doublet but at the weighted average between the two lines of each doublet.
Move the slider to see how an AB spin system is affected by the difference, Δν, in the chemical shift:
J = Hz
Intensity ratios of multiplets are often distorted, the more so the closer the chemical shifts of the coupling partners. Line intensities get distorted, chemical shifts are no longer at the centre of the signal, and small EXTRA lines may appear.
You also get higher-order effects in symmetrical molecules:
AA’XX’ and AA’BB’ spin systems are far more common and not limited to substituted benzenes.
The n + 1 multiplicity rule is valid only when the n coupling partners are all equivalent. For example, a doublet of doublets indicates that there are TWO different coupling partners nearby, one with a BIG coupling constant, one with a SMALL(ER) coupling constant. A triplet of doublets indicates that there are TWO different coupling partners nearby, TWO with a BIG coupling constant, ONE with a SMALL(ER) coupling constant.
Doublets of doublets and triplets of doublets are a characteristic feature of ortho-disubstituted benzenes. In a meta-disubstituted benzene the multiplicity can be similarly derived by counting the number of ortho-protons (with a large 8 Hz coupling) and the number meta-protons (with a small 1.5 Hz coupling).
A doublet of doublets has two coupling constants, a small coupling and a large coupling. The difference in Hz between the first and second (or the third and fourth line) equals the small coupling constant, whereas the difference in Hz between alternating lines, such as the first and the third line, equals the large coupling constant.
Coupling constants can provide additional information about a proton’s neighbours.
Geminal couplings are couplings over 2 bonds. The 2J coupling of protons attached to the same carbon is usually about 10 - 14 Hz. Geminal couplings are often negative. While this makes no difference to a first-order spin system, it can drastically alter the appearance of a higher-order ABX, ABC, AA’XX’, AA’BB’ etc. spin system. So, unless you are sure about the sign of a geminal coupling, best quote the absolute value of 2J to be on the safe side.
Geminal couplings can vary a lot in size. Adjacent double bonds (C=C, C=O) increase the coupling, whereas wide angles and heteroatoms decrease it.
Vicinal couplings are couplings over 3 bonds. Here are some representative 3J coupling constants:
Coupling constants also vary with the angle between the two coupling protons. Coupling constants are large if the two protons are at a 0° or 180° angle. At a 90° angle, the coupling constant is virtually zero.
These are couplings over 4 or more bonds and occur only when the protons are held in a fixed position suitable for coupling, usually following a perfect zig-zag or W-shaped path. Long-range couplings are often seen in rigid rings or for protons in an allylic position, as both double bonds and rings prevent bond rotation.
Finally, protons also couple to other nuclei such as 19F and 31P if present. Since every organic compound also contains 13C, a closer look at any 1H NMR spectrum will reveal tiny little satellites on both sides of a CH, CH2 or CH3 signal and at an intensity of 0.5%. The typical 1JC-H coupling constant is about 125 - 160 Hz, with higher values usually reserved for sp2-hybridised carbons; for alkynes and some exotic strained carbocycles the C-H coupling constant can even reach values as high as 200 - 250 Hz.
The two protons of a CH2 group can occasionally become non-equivalent i.e. diastereotopic. Diastereotopic protons have different chemical shifts and, because of this, they also couple to each other. If a molecule has a chiral centre, the protons on a CH2 group anywhere in the molecule will always be diastereotopic.
Z-test: Replace each of the CH2 protons in turn with some other group (Z = D, Cl, R ...). If you end up with enantiomers, the CH2 protons are magnetically equivalent and undistinguishable by NMR. If you get a pair of diastereoisomers, the two CH2 protons are diastereotopic.
Diastereoisomers are easy to recognise: they have 2 or more chiral centres.
If a compound shows a 1H or 13C NMR spectrum where some signals (other than an OH or NH) are unusually broad whereas most other signals remain sharp, this is often an indication of a dynamic process that occurs at the NMR timescale. NB: In case all signals are broadened, this could be due to (i) a very unusual dynamic process affecting all signals, or (ii) a polymer or, most often, (iii) a badly made up NMR sample.
NMR spectroscopy is able to detect dynamic processes with energy barriers around 9 - 20 kcal mol-1, such as a hindered rotation, a slow ring inversion, a dynamic equilibrium between two stable species, and much more. The appearance of the NMR spectrum will then become dependent on temperature. At low temperature, the NMR spectrum will show sharp signals of two species, whereas at elevated temperature the NMR spectrum will have just one set of signals which is an average of the two. We call the dynamic process slow on the NMR timescale when there are two sets of signals, and fast on the NMR timescale when we see only one averaged set of signals. In between the two extremes comes a point when the signals of the two components move closer together and finally merge — always accompanied by a characteristic line broadening of the NMR signals involved.
A textbook example is the hindered rotation around the single bond between the carbonyl carbon and the amide nitrogen in N,N-dimethylformamide (DMF). Note that the methyl groups just change position, but their chemical shifts remain unchanged.
At room temperature, the 1H NMR spectrum of DMF shows two singlets, one for each of the methyl groups. When the temperature is raised above 100 °C, the methyl signals (but not the CHO signal at 8 ppm) start to broaden, then move closer, subsequently coalesce (= merge to one broad singlet with just one maximum), and finally converge into one sharp singlet.
Move the slider to see :
Δν = Hz
The temperature at which two signals merge is called the "coalescence temperature". The rate of exchange, k (in s-1), at the coalescence temperature is approximately two times the difference Δν in Hz between the initially sharp lines.
The rate constant k is linked to the free energy of activation (ΔG‡) by the Eyring equation which you will encounter in your B19PC lectures;
At 60 MHz, the two methyl singlets of DMF are just 10 Hz apart and coalesce at 337 K. The rate constant k at the coalescence temperature is approximately 2.2 × Δν = 2.2 × 10 Hz = 22 s-1. Using these values the energy barrier for the hindered rotation around the partial amide double bond can be calculated as 8.314 J K-1 mol-1 × 337 K × (22.96 + ln 337 - ln 10) = 74 kJ mol-1. Note that, unlike chemical shifts, Δν (in Hz) will vary with NMR spectrometer frequency and, if you were to measure the NMR spectrum of DMF at higher field, the coalescence temperature for the two methyl signals will be substantially higher.
The simplistic equations for analysing the temperature-dependent NMR spectra of a dynamic system are, however, only valid when two singlets of equal intensity exchange with each other. When more than two signals merge together or, worse, the dynamic process involves more complex multiplicities, you need to simulate of the exchange process using a technique called NMR lineshape analysis.
You need to recognise residual NMR solvent signals, as well as common impurities such as water and acetone. The position of the water signal, and OH signals in general, can vary due to exchange.
Minor impurities in the NMR spectrum are often the result of solvents used during a reaction or sample purification. The following table lists a few examples, showing the data you would see from common impurities in an NMR spectrum taken in chloroform.
|Impurity||Number of signals||Chem. shift||Multiplicity||Assignment|
|tert-Butyl methyl ether||2||1.19||s||C(CH3)3|
|Grease or alkanes||2||0.86||m||CH2CH3|
|1.26||m||CH2CH2CH2 / CH2CH2CH3|
An excellent compilation of chemical shifts of a variety of impurities in different NMR solvents, can be found in the following article: