Theory / Background

Nuclear magnetic resonance (NMR) was first described independently by Nobel Prize winners Felix Bloch and Edward Mills Purcell in 1946. Purcell had worked on the development and application of RADAR during World War II at MIT, where he focused on the production and detection of electromagnetic radiation radio frequency range (RF ~ 60 MHz). He was specifically interested in the absorption of RF energy by matter.

They noticed that magnetic nuclei, like the hydrogen proton, could absorb RF energy when placed in a magnetic field of a strength specific to the identity of the nuclei. When this absorption occurs, the nucleus is described as being in resonance. Interestingly, for analytical scientists, different atoms within a molecule resonate at different frequencies at a given field strength. The observation of the resonance frequencies of a molecule allows a user to discover structural information about the molecule.

Thus, like electrons, the protons of certain atoms are considered to have spin. The spinning of these (+ ) charged particles generates a magnetic moment along the axis of spin, so that these nuclei act like tiny bar magnets. Thus, according to quantum mechanics, if a proton is placed in an external magnetic field, its magnetic moment can be aligned in either of two ways: with or against the field. Alignment with the field is more stable, and energy must be absorbed in order to "flip" the tiny proton magnet over to the less stable alignment (against the field).

The energy necessary to flip the proton depends on the strength of the external field. The stronger the field, the greater the tendency for the proton to remain aligned with the field, and the greater the energy (or higher the frequency) of the radiation needed to do the job. In practice, the the frequency of the RF radiation is held constant, and the strength of the magnetic field is varied. At some specific value of the applied field strength, the energy required to flip the proton is equal to the energy of the RF radiation.  The resulting absorption occurs, and a signal is observed. Such a spectrum of absorption vs. field strength is called a nuclear magnetic resonance or NMR spectrum.

                        NMR Spectrum

The frequency at which a proton absorbs depends on the magnetic field experienced by the proton. This effective field strength varies, depending on the environment of the proton. The environment is determined primarily by such factors as:

1) Local electron density

2) Number of nearby protons.

Each set of equivalent protons will have a unique environment, and hence will require a different applied field strength in order to produce the same effective filed strength: the particular field strength at which absorption occurs. At a given RF frequency, all protons absorb at the same effective filed strength, but they absorb at different applied filed strengths. It is this applied filed strength that is measured, and against which the absorption is plotted. The resulting absorption peaks give detailed information about the structure of a molecule, including:

a)  Number of signals: Number of different equivalent protons.

b)  Positions of the signals: Electronic environment of the proton. 

c)  Intensites of the signals: How many protons of each equivalent set.

d)  Splitting of the signal: Environment of the proton with respect to other protons.

Chemical Shifts (Peak Position)

Just as the number of signals in the NMR spectrum tells us how many kinds of protons a molecule contains, so the positions of the signals help to tell us what kinds of protons they are: aromatic, aliphatic, primary, secondary, tertiary, benzylic, vinylic, acetylenic, adjacent to halogen or to other atoms or substituent groups. These different kinds of protons have different electronic environments.

Thus, induced magnetic fields are generated when electrons are caused to circulate in a magnetic field. Circulation of electrons about the proton itself generates a filed aligned in such a way that - at the proton - it opposes the applied filed. The field felt by the proton itself is thus diminished, and the proton is said to be shielded.

Circulation of non-bonding pi electrons about nearby nuclei generates a field that can either oppose or reinforce the applied field at the proton, depending on the proton's location. It the induced field opposes the applied filed, the proton is shielded. If the induced field reinforces the applied filed, then the field felt by the proton is enhanced, and the proton is said to be deshielded.

Compared with a naked proton, a shielded proton requires a higher applied field strength - and a deshielded proton requires a lower applied filed strength - in order to provide the particular effective filed strength at which absorption occurs. Shielding thus shifts the absorption upfield, and deshielding shifts the absorption downfield.

Thus, the total magnetic field experienced by a nucleus includes local magnetic fields induced by currents of electrons in the molecular orbitals. The electron distribution of the same type of nucleus (e.g. ¹H, ¹³C, ¹⁵N) usually varies according to the local geometry (e.g. bonded atoms, bond lengths, bond angles), and with it the local magnetic field at each nucleus. This is reflected in the spin energy levels and resonance frequencies.

The variations of nuclear magnetic resonance frequencies of the same kind of nucleus, due to variations in the electron distribution, is called the chemical shift. The size of the chemical shift is given with respect to a reference frequency or reference sample. This is usually a molecule with a minimum of distortion in electron distribution, such as tetramethylsilane (CH3)4Si. Because of the low electronegativity of silicon, the shielding of the protons in the silane is greater than in most other organic molecules. As a result, most NMR signals appear in the same direction from the tetramethylsilane signal: downfield.

The units of chemical shifting are expressed in parts per million (ppm) of the total applied magnetic field. Most chemical shifts have delta values between 0 and 10. 

The chemical shift for a proton is thus determined by its electronic environment. Non-equivalent protons have different chemical shifts. Equivalent protons have the same chemical shift. Enantiomeric protons (with mirror image environments) have the same chemical shift.

Furthermore, it has been found that a proton with a particular environment shows much the same chemical shift, no matter what molecule it happens to be a substituent of. Take, for example, our familiar classes of hydrogens: primary, secondary and tertiary. In the absence of other nearby substituents, absorption occurs at about these values:

All these protons, in turn, differ widely form aromatic protons which, because of the powerful deshielding due to circulation of the non-bonding pi electrons, absorb far downfield.

Attachment of chlorine to the carbon bearing the proton also causes a downfield shift. If the chlorine is attached to the carbon once removed from the carbon bearing the proton, there is again a downfield shift - but this time much weaker.

Two chlorines cause a greater downfield shift. Other halogens show similar effects.

The downfield shift caused by the halogens is what we might expect form its inductive effect. Electron withdrawal lowers the electron density in the vicinity of the proton and thus causes deshielding. The electronegative oxygen of alcohols and ethers also causes deshielding. the effect of a substituent on the chemical shift is the net result of many factors. But we have sufficient evidence to indicate that the inductive effect indeed a  critical factors. 

The NMR spectra of the alkylbenzenes: toluene, p-xylene, and mesitylene (all aromatic-aliphatics or "arenes") illustrate the points we have just made. In each spectrum there are two signals: one for the side-chain protons, and one for the ring protons. (Here, as in some - but not most - aromatic compounds, the ortho, meta, and para protons have nearly identical chemical shifts).

In each spectrum, the ring protons show the low-field absorption we have indicated is characteristic of aromatic protons. Absorption is not only at low field, but at nearly the same field strength for the three compounds (delta values: 7.17, 7.05, 6.78 respectively). The slight difference in these delta values is due to the difference in electronic environments of these three compounds.  

In each of these three compounds, side chain protons (benzylic protons) are close enough to the ring to feel a little of the deshielding effect of the non-bonding pi electrons. Hence, they will absorb somewhat downfield from ordinary alkyl protons (2.32, 2.30, 2.25 respectively). In all three compounds, the environment of the side-chain protons is nearly identical, and so are the corresponding chemical shifts.

The similarity in structure among these three alkylbenzenes is thus reflected in the similarity of their NMR spectra. There is, however, a major difference in their structures. This is a difference in numbers of aromatic and side-chain protons. As we shall see in the next section, this is reflected in a major difference in their NMR spectra.

Spin Coupling (Peak Splitting)

An NMR spectrum shows a signal for each type of proton on a molecule. Complications arise, however, as evidenced by the spectra for the three following compounds - each of which contains only two kinds of protons.

Instead of two peaks, these spectra show five, six, and seven peaks respectively. Let us examine more closely this phenomenon of peak multiplicity.

What we are observing here is the splitting of NMR signals caused by spin-spin coupling. The signal we expect form each set of equivalent protons is appearing as a group of peaks (vs. a single peak). Splitting reflects the environment of the absorbing protons: not with respect to electrons, but with respect to other, nearby protons. It is as though we were permitted to sit on a proton and look around in all directions. We can see and count the protons attached to the C atoms next to our own C atom, and sometimes protons even farther away.

Let us consider the case of adjacent C atoms carrying a pair of both secondary and tertiary protons. First, consider the secondary protons.

The magnetic field that a secondary proton feels at a particular instant is slightly increased (tertiary proton aligned with the field) or slightly decreased (tertiary proton aligned against the field) by the spin of the tertiary proton.

For half the molecules, absorption by a secondary proton is shifted slightly downfield, and for the other half of the molecules the absorption is shifted slightly upfield. The signal is split into tow peaks: a doublet, with equal peak intensities.

Now what can we say about the absorption of the tertiary proton ?

It is affected by the spin of the neighboring secondary protons. But now there are two protons whose alignments in the applied field must be considered. There are four equally probable combinations of spin alignments for these two protons, of which two are equivalent. Thus the tertiary proton feels any one of three fields at any given time, and its signal is split into three equally spaced peaks: a triplet with relative peak intensities 1:2:1, reflecting the combined (double) probability of the two equivalent combinations.

The following figure is an idealized NMR spectrum due to the grouping of -CH-CH2-. We see a 1:1 doublet (from the -CH2-) and a 1:2:1 triplet (from the -CH-). The total area (both peaks) under the doublet is twice the total area (all three peaks) of the triplet, since the doublet is due to absorption by twice as many protons as the triplet.

The distance between peaks in a doublet or triplet is a measure of the effectiveness of spin-spin coupling, and is called the coupling constant, J. Unlike the chemical shift, coupling is not a consequence of induced magnetic fields. The value of the coupling constant (in Hz) remains the same, whatever the applied magnetic field (i.e. whatever the RF used).

The magnitude of the coupling constant depends markedly on the structural relationship between the coupled protons. For example, in any substituted ethylene - or in any pair of geometric isomers - J is always larger between trans protons than between cis protons. Furthermore, the size of J varies in a regular way with the electronegativity of the substituents, so that one can often assign configuration without having both isomers in mind.

In the case of NMR spectrum due to the grouping of -CH-CH2-, we note that identical peak separation distances in both the doublet and the triplet indicate identical values of the coupling constant J. Even if they were to appear in a complicated spectrum of many absorption peaks, the identical J values would indicate that this doublet and triplet were closely related.. I. E. They would indicate that the two protons giving rise to the doublet and the single proton giving rise to the triplet are coupled, and hence are attached to adjacent carbon atoms.  

What splitting can we expect from more than two protons ? The following figure suggests that three (equivalent) protons splits a signal into four peaks - a quartet - with relative intensities of 1:3:3:1.

It can be shown that, in general:

A set of n equivalent protons will spit an NMR signal into n+1 peaks.

If we now return to the original spectra with five, six and seven peaks, what we see now is a doublet / triplet, or a doublet / quartet, or a triplet / quartet. Each spectrum now shows more than simply absorption by two kinds of protons.

Bearing in mind that the peak area reflects the number of absorbing protons, and the multiplicity of splittings reflects the number of neighboring protons, we find that:

1) In the spectrum of CHBr2-CHBr2, we see:  

2) In the spectrum of CH3-CHBr2, we see:  

1) In the spectrum of CH3-CH2Br we see:  

We see chemical shifts that are consistent with the deshielding effect of halogens. In each spectrum, the protons on the carbon carrying the greater number of halogens absorb further downfield (larger delta).

In each spectrum, we see that the spacing of the peaks within one multiplet is the same as within the other, so that even in a spectrum with many other peaks, we could pick out these two multiples as being coupled.    

So which protons in a molecule can be coupled ? We may expect to observe spin-spin splitting only between:

Non-equivalent, neighboring protons, i.e. protons with

Different chemical shifts on adjacent C atoms.

Sometimes protons farther removed from each other may also be coupled, particularly if non-bonding pi bonds intervene. (Also, if protons on the same carbon atom are non-equivalent as they sometimes are - they may show coupling).

We do not observe splitting due to coupling between the protons making up the same -CH3 group, since they are equivalent. We do not observe splitting due to coupling between the protons on C1 and C2 of 1,2-dichoroethane since, although on different carbons, they, too, are equivalent. 

In the spectrum of 1,2-dibromo-2-methylpropane,

we do not observe splitting between the six methyl protons, on the one hand, and the two -CH2 protons on the other hand. They are non-equivalent, and give rise to different NMR signals. But they are not on adjacent carbon atoms, and their spins do not (noticeably) affect each other. The NMR spectrum contains two singlets, with a peak area ratio of 3:1 (or 6:2).

For similar reasons, we do not observe splitting between ring protons and side-chain protons in arenes (such as alkylbenzenes). Even thought they are non-equivalent, they are not on adjacent carbon atoms.

We do not observe splitting between the two vinyl protons of isobutylene since they are equivalent.

On the other hand, we may observe splitting between the two vinyl protons on the same carbon if, as in 2-bormoproene, they are non-equivalent.

The fluorine F-19 nucleus has magnetic properties similar to the proton. It gives rise to NMR spectra, although at a very different frequency / field strength combination than the proton. Fluorine nuclei can be coupled not only with each other, but also with protons. Absorption by fluorine does not appear in the proton NMR spectrum. It is far off the scale. But the splitting by fluorine of proton signals can be seen. For example, the signal for two protons of 1,2-dichloro-1,2-difluoroethane appears as a 1:2:1 triplet with peak spacings of 11 Hz.

The following figures illustrate some of the kinds of splitting we are likely to encounter in NMR spectra.  

1) Isopropyl bromide.

Absorption by the six methyl protons H(a) appears upfield, split into a doublet by the single adjacent proton H(b). Absorption by the lone proton H(b) appears downfield (the inductive effect of bromine) split into a septet by the six adjacent protons (outside peaks barely visible).

2) n-propylbenzene

Moving downfield, we see the expected sequence of signals: a, primary (3H); b, secondary (2H), c, benzylic (2H), d, aromatic (5H). Signals a and c are each split into a triplet by the two secondary protons H(b). The five protons adjacent to the secondary protons - three o none side and tow on the other - are, of course, not equivalent. But the coupling constants J(ab) and J(bc) are nearly identical, and signal b appears as a sextet (5 + 1 peaks). The coupling constants are not exactly the same, however, as shown by the broadening of the six peaks.

3) 1,2-dibromo-1-phenylethane

The diastereotopic protons H(a) and H(b) give different signals, each split into a doublet by H(c). The downfield peaks of the doublets happen to coincide. [The spectrum shows no splitting due to coupling between H(a) and H(b). With the trial run at high gain, however, the spectrum shows this coupling: each doublet is split into a quartet.]

The four-line pattern of c is due to successive splittings by H(a) and H(b). [ If J(ac) and J(bc) were equal- as they would have to be if, for example, H(a) and H(b) were equivalent - then the middle peaks of c would merge to give the familiar 1:2:1 triplet.]

Carbon-13: CMR

Fortunately for the organic chemist, among the atoms that, like the proton, give rise to NMR spectra is one of the isotopes of carbon: C-13. The CMR spectrum is generated in the same fundamental way as the proton NMR spectrum, and the same basic principles apply. However, obtaining a useable spectrum is more difficult for CMR, and requires more sophisticated instrumentation.

Thus, CMR is much less sensitive than proton NMR. This is because the major isotope of carbon, the C-12 isotope, has is not magnetically active. Only the less common C-13 isotope present naturally at 1.1% abundance is magnetically active. Therefore, only the few carbon-13 nuclei present resonate in the magnetic field. The result is large reduction reduced sensitivity.

This also helps significantly in another standard problem area: peak splitting. Since the C-13 isotope is so rare, only occasionally is a C-13 near enough to another c-13 for spin-spin coupling to occur. Therefore, CMR spectra do not ordinarily show carbon-carbon peak splitting, and are thus enormously simplified. (In addition, proton spectra do not show splitting by CMR). There remains, however, the splitting of C-13 signals by protons (see below).

Another complication results from the complexity of spectra due to the coupling constants J between carbon and hydrogen (~ 100 - 250 Hz). This indirect dipole-dipole coupling occurs between two nuclear spins due to the influence of bonding electrons on the magnetic field running between the two nuclei.

In addition, because carbon-13 resonates at 75.47 MHz in a 7 T magnetic field (compared to 300 MHz for a proton), the splitting patterns often overlap and become too complicated to interpret easily. In order to remove this complexity, CMR spectra are proton decoupled to remove the signal splitting. In contrast to a typical proton NMR spectrum with multiplets for each proton position, carbon NMR spectra show a single peak for each chemically nonequivalent carbon atom.

Thus, the CMR spectrum gives much the same kinds of information as proton NMR. But here the information is directly about the carbon skeleton - not simply the protons attached to it. 

1)  The number of signals tells how many different carbons - or different sets of equivalent carbons - there are in a molecule.

2)  The splitting of a signal tells how many hydrogens are attached to each carbon atom.

3)  The chemical shift tells us the hybridization (sp3, sp2, sp) of each carbon atom.

4)  The chemical shift tells us about the electronic environment of each carbon with respect to other, nearby carbon atoms and/or functional groups.
 

CMR: Peak Splitting

As we said earlier, CMR spectra do not ordinarily show carbon-carbon peak splitting, and are thus enormously simplified. (In addition, proton spectra do not show splitting by CMR). There remains, however, the splitting of C-13 signals by protons.

In a CMR spectrum, we cannot see the absorption by protons because these signals are far off the scale. But we can see the splitting of a carbon signal by protons on the carbon atom itself as well as protons on more distant carbons. Such unwanted splitting is removed by decoupling the C-13 spin from that of the proton. This decoupling can be done in either of two principal ways, depending on the frequency of the radiation used. 

One method of decoupling gives a completely proton-decoupled spectrum, showing no splitting at all. It consists of single peaks, one for each carbon - or each set of equivalent carbons- in the molecule. This is the kind of spectrum used most commonly for structural analysis, and is the kind shown here.

See, for example, the proton-decoupled CMR spectrum of sec-butyl bromide. There are four non-equivalent carbon atoms in this molecule. In the spectrum we see four peaks: one for each of these four non-equivalent carbon atoms.

A second method of decoupling (called off-resonance) gives a spectrum which shows splitting of the carbon signal by protons attached to that carbon itself. That is, we see only C(13)-H coupling and not C(13)-C-H and not C(13)-C-C-H coupling. We shall refer to this as a proton-coupled spectrum. 

For each carbon, then, the multiplicity of the signal depends upon how many protons are attached to it. 

Thus, we see a peak splitting present in the proton-coupled CMR spectrum of sec-butyl bromide. Now each peak is a multiplet (doublet, triplet, and two quartets).

Thus, the proton-decoupled spectrum tells us how many different carbons there are, and the proton coupled spectrum tells us how many protons are attached to each of these carbon atoms. Together, these spectra give us a remarkably detailed picture of the molecule.

CMR: Chemical Shifting

Chemical shifts in the CMR spectrum arise in basically the same way as in the proton NMR spectrum. Each carbon nucleus has its own electronic environment, different from the environment of the other, non-equivalent nuclei. It feels a different magnetic filed, and absorbs at a different applied filed strength. But the shifts in CMR differ in several ways from those in proton NMR.

For starters, chemical shifts are much larger in CMR than in proton NMR. As we see in the following figure, the scale extends form delta values of 0 to 200 and beyond - more than 30 times as wide as in NMR.

Of these shifts, the biggest and most important are determined by the hybridization of the carbon. This is something that, of course, is not a factor for the proton. Look, for example, at the spectrum of 1-octene. We see the peaks for the sp3-hybridized carbons upfield, between 14.1 and 34.0, and for the sp2-hybridized carbons over100 ppm downfield from them, at 113 and 140.

Aromatic carbons are also sp3-hybridized, and also absorb downfield, in much the same region as alkene carbons do. In the spectrum of ethylbenzene, we again see two widely separated sets of peaks: upfield, a set from the side-chain (sp3-hybridized) carbons, and downfield, after a 100 ppm gap, a set from the ring (sp2-hybridized) carbons.

CMR spectra for:

1) 1-Octene ; 2) Ethylbenzene ; 3) 1-Hexyne

Absorption by triply bonded (sp-hybridized) carbon falls between the regions for sp3-hybridized carbon, as shown for 1-hexyne.

In relating structure to chemical shift, we begin with the hybridization of carbon.  

Next, we consider the effects of substituents, which are superimposed on the hybridization effects. As in proton NMR, most substituents in most positions deshield the nucleus, and shift the signal downfield. But with carbon these effects are bigger, are felt form farther away, and fall into different patterns. In order to get some idea of what these patterns are like, let us examine the effects of several substituents on absorption by sp3-hybridized carbons. 

Let us look first a the effects of chlorine on the absorption by various carbons of a saturated chain. The spectra of n-pentane and 1-chloropentane, for example, give data that we can summarize like this:

Let us compare, carbon by carbon, the delta values for the two compounds. 

In the signal for C1, chlorine causes a very large downfield shift, form 13.7 to 44.3, a difference of 30.6 ppm. Such a shift, for the carbon bearing the substituent, is called an alpha-effect.

In the signal for C2, chlorine again exerts a downfield shift, form 22.6 to 32.7, a difference of 10.1 ppm. Such a shift, for the carbon once removed from the carbon bearing the substituent, is called a beta-effect.

At C3, we see a reversal of the downfield shift. Absorption here is shifted upfield, form 34.5 to 29.2, a difference of – 5.3 ppm. Such a shift, for the carbon twice removed from the carbon bearing the substituent, is called a gamma-effect.

Beyond the gamma-carbon, we see a negligible effect of chlorine.

Each substituent chlorine exerts effects of about these same sizes on absorption by saturated carbon in a wide variety of compounds. 

Nearly all substituent effects on absorption by sp3-hybridized carbon follow the same pattern as those for chlorine. This includes alpha and beta effects downfield, (alpha greater than beta) and smaller gamma effects upfield. Size effects are typically quite large. For example, consider the alpha effects exerted by these substituents attached to C1 of pentane: F (70.1), Br (19.3), NH2 (29.7), OH (48.3), NO2 (64.5).

Alkyl groups exert smaller effects than other substituents, and follow a somewhat different pattern. Consider the effects of a methyl group using data form n-pentane and n-hexane.

We can calculate the following substituent effects of methyl:

These effects are typical of alkyl groups acting on the absorption by saturated carbon: alpha and beta effects downfield (and of about the same size) and gamma effects smaller and upfield.

 Consider next absorption by sp2-hybridized carbons – in alkenes and aromatic rings. Here many substituents exert alternating effects which will be evident in the spectra.

 The presence of a carbon-carbon double bond can introduce a new factor, geometric isomerism, which has important effects on the absorption by sp3-hybridized carbons. To see how this stereochemical factor works, let us compare the absorption data for propylene with the data for cis-and trans-2-butene.

Let us focus our attention on the methyl carbon of propylene, and see how its absorption is affected by the substitution of a methyl group on one or the other of the vinylic hydrogens. We are, of course, looking at gamma effects.

These effects are both upfield: - 7.3 ppm for the cis isomer and – 1.9 ppm for the trans. But the gamma effect for the cis isomer is stronger, by 5.4 ppm.

The stereochemical influence on gamma effects in alkenes is a general one, and is extremely useful in assigning configuration to geometric isomers

NMR / CMR Spectra of Hydrocarbons

For most molecular compounds, we shall find that, while IR absorption and spectroscopy helps to identify what kind of compound we are dealing with, NMR helps us to understand which specific compound it is.

Alkanes and alkane-like (saturated) groups will give upfield peaks in both CMR and NMR: absorption by sp3-hybridized carbons and the protons attached to them.

Aromatics will show downfield absorption in both CMR and NMR.

Alkenes will show similar absorption to aromatics in CMR, but not in NMR. Vinylic protons absorb well upfield form aromatic protons, and this will often allow us to distinguish between the two (clearly evidenced by IR spectrum.)

Alkynes, with triply bonded sp-hybridized carbons, will give peaks between those of sp3 and sp2-hybridized carbons.

Electronegative atoms - halogen, oxygen, nitrogen - will shift peaks downfield in both CMR and NMR, but not usually outside the region where we expect to see them.

Precise molecular structures can only be elucidated using the number of signals and their splitting.

Alcohols and Ethers. NMR absorption by a hydroxylic proton (O-H) is shifted downfield by hydrogen bonding. The chemical shift that is observed depends, therefore, on the degree of hydrogen bonding, which in turn depends on temperature, concentration and the nature of the solvent. Thus, a signal can appear anywhere in the range 1 - 5 ppm. It may be hidden among the peaks due to alkyl protons, although its presence there is often revealed through proton counting.

A hydroxyl proton ordinarily gives rise to a singlet in the NMR spectrum. its signal is not split by nearby protons, nor does it split their signals. Proton exchange between two identical molecules of alcohol is so fast that the proton - now in one molecule and the next instant in another - cannot see nearby protons in their various combinations of spin alignments, but in a single average alignment.

Presumably through its inductive effect, the oxygen of an alcohol causes a downfield shift for nearby protons. This shift is of about the same size as other electronegative atoms.

Applications

The primary use of nuclear magnetic resonance is in magnetic resonance imaging (MRI) for medical diagnosis. However, it is also widely used in chemical studies, notably in NMR spectroscopy such as proton NMR and carbon-13 NMR. By studying the peaks of nuclear magnetic resonance spectra, skilled chemists can determine the structure of many compounds. It can be a very selective technique, distinguishing among many atoms within a molecule or collection of molecules of the same type but which differ only in terms of their local chemical environment.

By studying previously documented information, a chemist can determine the identity of a compound by comparing the observed absorption frequencies to known frequencies. Further structural data can be elucidated by observing spin-spin coupling, a process by which the precession frequency of a nucleus can be influenced by the magnetization transfer from nearby nuclei.

A relatively recent example of nuclear magnetic resonance being used in the determination of a structure is that of buckminsterfullerene, or "bucky balls". This now famous form of carbon has 60 carbon atoms forming a sphere. The carbon atoms are all in identical environments and so should see the same internal H field. Unfortunately, buckminsterfullerene contains no hydrogen and so Carbon NMR has to be used. However in 1985 the spectrum was obtained @ Rice University and sure enough it did contain just the one single spike, confirming the unusual structure of C60.

NMR extremely useful for non-destructive analysis of samples. Radio waves and static magnetic fields easily penetrate many types of matter and anything that is not inherently ferromagnetic. For example, various expensive biological samples, such as nucleic acids, including RNA and DNA, or proteins, can be studied using NMR for weeks or months before using destructive biochemical experiments. This also makes NMR a good choice for analyzing dangerous samples.

Another use for NMR is data acquisition in the petroleum industry for petroleum and natural gas exploration and recovery. A borehole is drilled into rock and sedimentary strata into which nuclear magnetic resonance logging equipment is lowered. Nuclear magnetic resonance analysis of these boreholes is used to measure rock porosity, estimate permeability from pore size distribution and identify pore fluids (water, oil and gas).

NMR has now entered the arena of real-time process control and process optimization in oil refineries and petrochemical plants. Two different types of NMR analysis are utilized to provide real time analysis of feeds and products in order to control and optimize unit operations. Time-domain NMR (TD-NMR) spectrometers operating at low field (2-20 MHz for 1H) yield free induction decay data that can be used to determine absolute hydrogen content values, rheological information, and component composition.

These spectrometers are used in mining, polymer production, cosmetics and food manufacturing as well as coal analysis. High resolution FT-NMR spectrometers operating in the 60 MHz range with shielded permanent magnet systems yield high resolution 1H NMR spectra of refinery and petrochemical streams. The variation observed in these spectra with changing physical and chemical properties is modeled to yield predictions on unknown samples. The prediction results are provided to control systems via analogue or digital outputs from the spectrometer.