User:Mcannos/sandbox/heteronuclear NMR
For a more general article, see Nuclear magnetic resonance
Heteronuclear NMR spectroscopy is an analytical technique which probes Spin-active Nuclei other than 1H, 2H, or 13C by Nuclear Magnetic Resonance. Nuclei can be studied directly by exciting those nuclei, or indirectly by spin-spin coupling (1D) or interaction (2D) to other nuclei. Isotopes studied must have a non-zero spin quantum number to be studied. While the possible nuclei are diverse, generally nuclei (other than 1H, 19F, and 31P) suffer from poor spin-active isotopic abundance, gyromagnetic ratio, and/or high quadrupole moment which decrease sensitivity and increase broadness. Generally, these techniques are considered less routine as a result. To compensate for low innate sensitivity, techniques such as isotope enrichment or spin polarization transfer (eg. INEPT) are employed. Techniques including main group elements and transition metals are generally useful in inorganic chemistry and organometallic chemistry for structural characterization of compounds, especially with multiple spin-active nuclei.
History
[edit]Due to high sensitivity, initial NMR work focused on 1H, 19F, and 31P.195Pt capabilities existed early on, but were largely limited due to low sensitivity. With the use of Fourier-Transform and development of superconducting/higher magnetic field instruments, sensitivity increased enough to allow detection of many other spin active nuclei.[1]
Signal
[edit]As with all modern NMR, a signal is obtained by pulsing the resonant frequency of a given isotope while under the influence of a magnetic field. Utility of a signal is mainly governed by possible concentration, the signal strength, and the broadness of signal.
In Heteronuclear NMR, concentration is complicated by "isotopic dilution." Many nuclei have more than one abundant isotope which reduces the amount of nuclei that able to be excited. Being that NMR techniques require a relatively high concentration, this can often lead to sensitivity issues. The strength of signal as well as resonant frequency for a given magnetic field is directly proportional to the gyromagnetic ratio (γ) of a nucleus. Many useable nuclei have spin greater than ½ giving it an electric quadrupole moment. Quadrupolar nuclei tend to have shorter relaxation times and broadened lines, often leading to difficulties in obtaining quality spectra directly.
Coupling
[edit]Coupling in heteronuclear NMR is useful to determine structural aspects. Oftentimes spectra are run decoupled, eg. {1H}, to increase signal by Nuclear Overhauser effect NOE, and to reduce signal dilution due to splitting, especially with nuclei with low signal.
![](http://upload.wikimedia.org/wikipedia/commons/thumb/0/06/PT2Dms2Me4.png/220px-PT2Dms2Me4.png)
Coupling to heteronuclei in other spectrum (eg. in a 1H spectrum) is characterized by generally large coupling constants (relative to 1H-1H coupling). In many cases, the spin number I is not ½, giving rise to equivalent eg. triplets and quartets by the equation:
- Number of signals=2nI+1
where n is the number of coupling nuclei and I is the spin quantum number.
Spin Dilute Systems
[edit]As a consequence of naturally abundant isotopes having different spins, a mixture of coupling is observed. With nuclei with only 1 spin active nucleus, this results in “satellite” peaks (eg. Pt, C). With multiple spin active nuclei, the pattern is more complicated, having different coupling pattern with a coupling constant superimposed on one another. Examples are listed below.
P-Block Nuclei
[edit]Note: This is not an all inclusive list, but rather a collection of several most easily studied isotopes.
10B and 11B
[edit]Boron has two NMR active nuclei. 10B has a nuclear spin of 3 (I = 3) and it is 19.9%[2] abundant. 11B has a nuclear spin of 3/2 (I = 3/2) and is 80.1%[2] abundant. Due to an inherent low abundance of 10B, 10B has lower NMR receptivity of 22.1 relative to 13C, while the relative receptivity of 11B is 754.[3] Thus, 11B is more commonly used.
In principle, two boron isotopes should have different chemical shifts due to a primary isotope effect.[3] However, the difference is negligible compared to line width of the resonance.[3] The boron NMR chemical shift range spans about 210 ppm, ranging from -120 to 90 ppm.[4] The reference compound used is boron trifluoride etherate (BF3•OEt2).
- Limitation
- A broad singlet in the background of the spectrum is usually observed.[4] Regular NMR tubes are made of borosilicate glass, giving additional boron signals.[4] To overcome this, quartz tubes are preferably used; however, quartz tubes are much more expensive and more fragile than regular tubes.[4]
See also Boron.
14N
[edit]Two natural nitrogen isotopes are NMR active but each isotope presents considerable problems.[3] 14N has a nuclear spin of 1 (I = 1) and is 99.636% abundant.[2] It is medium responsive to NMR measurements. However, because of its spin, the nucleus possesses a quadrapole moment which results in peak broadening.[5]
The 14N chemical shift range is very wide, ranging from -400 to 600 ppm. Peak widths are often disperse between 100 and 1000 Hz (14 to 140 ppm at 2.35 T).[3] Thus, the resolution becomes poor and signals may not be observable on high resolution NMR spectrometers. In addition, nitrogen nuclei in different chemical environments can be as little as 50 ppm apart from other nuclei in such a large range (approximately 1000 ppm), making difficult for structural investigations.[3]
Both 14N and 15N use the reference compound CH3NO2 recommended by IUPAC, although sometimes liquid NH3 is used as an alternative reference.
15N
[edit]15N has a nuclear spin of ½ (I = ½) and it is only 0.364% abundant [2] thus the sensitivity is very small.
15N has a negative magnetogyric ratio γ, leading to a negative Nuclear Overhauser effect (NOE). Unlike peak broadening in 14N, 15N NMR produces sharp signals. 15N is more preferred than 14N NMR. Its low sensitivity can be improved by projection of a heteronuclear correlation. The presence of two isotopes in nitrogen causes a primary isotope effect.[3] 15N chemical shifts become fundamentally different from 14N.[3] However, the difference of the same sample be less than the experimental error (0.2 ppm).[3]
The range of chemical shifts for N NMR spans 1000 ppm: as examples, when reference to nitromethane, nitrosos range from -540 to -480 ppm, nitrites -210 to -180 ppm, nitriles 110 to 140 ppm, ammonium ions 300 to 360 ppm, and cyanomides 330 to 390 ppm. Referencing is difficult in nitrogen NMR, because the chemical shift of the standard can vary several ppm due to molecular interactions with the solvent or other solutes. To alleviate this problem, external references are generally used for routine nitrogen NMR, The reference compound used is CH3NO2 recommended by IUPAC. The chemical shift in nitrogen NMR is very sensitive to electronic perturbations of the nitrogen environment, making it useful for determining the structural changes and the intermolecular interactions of a compound because of this the chemical shift is useful in quantifying tautomeric equilibria.[6]
Coupling of 14N to other nuclei is seldom observed due to its electric quadrupole; however, 15N coupling is a valuable diagnostic tool. The values of 1J(15N, 1H) coupling covers a range of -50 to -140 Hz, with values of -85 to -95 Hz being the most common. The magnitude of the coupling generally increases with increasing s-character of the bond, though caution must be used as there are several notable exceptions. In conjunction with the chemical shift, 1J(15N, 1H) coupling has also been used to quantify tautomeric equilibria, especially in biochemical systems. Overall, values of 1J(15N, 13C) coupling tend to be negative and less than 35 Hz. Like with hydrogen, the greater the s-character of the bond, the larger the magnitude of the coupling constant, but, again, this is not always the case, and this trend cannot be taken as a rule.[7]
One-bond coupling has also been studied on 15N, 31P, 19F and 135Pt. The values of 1J(15N, 15N) coupling are positive and extensively appear in the range of 10-20 Hz, with the largest found for nitrosoamines and the smallest for hydrazino moitites and molecular nitrogen. On the other hand, 1J(15N, 31P) coupling covers a broader range and can be positive or negative, ranging from -680 to 170 Hz: the value seems to be dependent on the dihedral angle between the nitrogen and phosphorus lone pairs. This wide range makes interpretation of the spectra very difficult if only the magnitude of the coupling is known. The values of 1J(15N, 19F) all tend to be positive and cover the range of 150-460 Hz, while the values of 1J(15N, 135Pt) are also positive and usually ranging from 100-580 Hz. They vary significantly with the nature of the other ligands on platinum, with the largest contributor to the value of the coupling constant being the ligand that is trans to the nitrogen.[6]
Nitrogen geminal coupling is seen frequently with 1H, 13C, 15N, and 31P. The values of 2J(15N, 1H) across saturated carbons are positive and small in magnitude: the value depends on the orientation of the C-H bond with respect to the nitrogen lone pair, with the largest values arising when the lone pair is cis to the bond. The coupling constant is larger for unsaturated carbons, rising to about 15 Hz. In amino-type moieties, the 2J(15N, 1H) is greatly varied based on the electronic properties of the nitrogen in question. When the nitrogen lone pair is free, the coupling is negative and large in magnitude; if the lone pair is in some way occupied, such as by protonation, then the magnitude of the coupling decreases significantly. Values of 2J(15N, 13C) are also negative, but often much smaller in magnitude than in the case of one-bond coupling. As with one-bond coupling, the magnitude is generally much larger for an unsaturated carbon, which provides a useful means of distinguishing saturated and unsaturated fragments. Again, the value depends on the orientation of the lone pair on nitrogen relative to the C-C bond. The magnitude of 2J(15N, 15N) is often near zero, and is only significant across unsaturated carbons where the coupling constant can be as large as 11 Hz. 2J(15N, 31P) are studied primarily in transition metal complexes, and can be used to assess the geometric orientation of the nitrogen and phosphorus groups, as larger coupling occurs when the groups are trans to each other.[6]
Vicinal coupling between nitrogen and hydrogen is often larger than that in their geminal derivatives. Additionally, the value does not change appreciably across saturated and unsaturated moieties. The values of 3J(15N, 13C) are negative and small, usually less than 5 Hz. The coupling constant magnitudes of 3J(15N, 1H) and 3J(15N, 13C) depend on the dihedral angle and follow the Karplus relation, but attempts to determine the dihedral angle through the use of 3J(15N, 13C) coupling has given rise to substantial errors due to the small magnitude of the value.[6]
- DEPT
- 15N is used for DEPT with increasing sensitivity.
19F
[edit]See Fluorine-19 NMR.
27Al
[edit]27Al is 100% abundant[2] and has a nuclear spin of 5/2 (I = 5/2). It is highly receptive to NMR but the experimental signals are broadened due to a qudrapole moment. 27Al NMR is studied much less than other common NMR active nuclei.[8] 27Al NMR is mainly used to detect the presence of aluminium in target compounds.
27Al has a chemical shift ranging from -200 to 200 ppm.[8] The commonly used reference compound is Al(NO3)3 in aqueous solution.
29Si
[edit]There exist three stable isotopes of silicon: 28Si, 29Si, and 30Si. Both 28Si and 30Si have a spin of 0, meaning they are NMR-inactive; the 29Si nucleus, with natural abundance of 4.70%, has a spin of ½ and is therefore useful for NMR. 29Si has a negative gyromagnetic ratio, leading to unfavorable Nuclear Overhauser Effect (NOE) consequences. One example of this is its very low sensitivity in normal decoupled experiments: this can be overcome by performing an inverse gated decoupling experiment, which still provides a decoupled spectrum but without the loss of sensitivity. The resonance frequency of 29Si is 0.19867 relative to hydrogen, slightly lower than of 13C.[9]
The range of chemical shifts for Si NMR spans from about -400 to 600 ppm. Referencing silicon NMR is rather intuitive as tetramethylsilane (TMS), the standard for 1H and 13C, is often used. TMS has the benefit of being inert, of low boiling point, and having a short relaxation time. The extreme chemical shift values are found for divalent silicon, while tetravalent silicon tends to resonate near 0 ppm, with a range of about 200 ppm. One trend found by Ernst et al is that trends in chemical shifts follow a parabolic curve when plotted against the sum of the electronegativities of the silicon substituents: if this sum is greater than 9.5, then an increase in electronegativity leads to an increase in shielding; if the sum is less than 9.5, a decrease in shielding in observed. In silicon-containing ring systems, the silicon is much more shielded than in the equivalent aliphatic system. [9]
1J(29Si, 1H) coupling is visible in 1H NMR spectra, and could therefore be used even before 29Si NMR was feasible. These one-bond couplings are always negative in sign, and generally range from -70 to -400 Hz. The magnitude is heavily dependent on the magnitude of s-character in the Si-H bond, with increasing magnitude of coupling characteristic of increasing s-character. In addition, the coupling constant tends to increase when the other substituents on Si are electronegative. 1J(29Si, 13C) coupling is often difficult to observe, as both 29Si and 13C are spin dilute isotopes. Therefore, the only way to determine the coupling constant is through the use of very low intensity satellite peaks. When available, these coupling constants range between 44 and 107 Hz. [9]
One-bond coupling has also been studied on 19F, 15N, 31P, and 29Si. The values of 1J(29Si, 19F) are positive and range from 120 to 400 Hz. Unfortunately, there are no straightforward trends describing the magnitude of coupling relative to the compound structural features, though variable-temperature 29Si NMR has been used to determine the fluxional character of fluorine bridges in some compounds. 1J(29Si, 15N) couplings are positive values, with few exceptions, and are in usually fairly small (0-10 Hz). It is hypothesised that the few negative coupling constants are due to the influence of the lone pair on nitrogen. Values of 1J(29Si, 31P) coupling are positive, except in cases where the lone pair on phosphorus is engaged with another atom, with magnitudes lying in the 10-70 Hz range. Due to the large number of Si-Si containing compounds, 1J(29Si, 29Si) coupling has been heavily studied. The values of these coupling constants tend to fall in the range of 20-160 Hz, with the only negative values having been reported for silyl anions containing Si-Si bonds. These magnitudes are primarily dependent on the electronegativity of the substituents on silicon, and can be related to the s-character of the Si-Si bond. Like in 1J(13C, 12C), the 1J(29Si, 29Si) coupling increases in going from disilanes to disilenes and disilynes. [9]
Geminal and vicinal coupling of silicon has been observed, but is generally not useful for structural determination and, therefore, is often unreported. Values of 2J(29Si, 1H) are positive and are small, ranging from 3-10Hz. Larger values of 20-70 Hz have been noted for Si-M-H (where M is a transition metal), and is rationalized as a sign of an agostic interaction with the metal. Values of 3J(29Si, 1H) are of roughly the same magnitude as 2J(29Si, 1H), but are of the opposite sign. 2J(29Si, 13C)is rarely seen due to the extremely low intensity and small magnitude (<10 Hz) of the coupling. 2J(29Si, 29Si) coupling is important in siloxane (Si-O-Si) derivatives, with values ranging from 0.5-14 Hz. Also in 2J(29Si,29Si,29Si ) systems, the coupling is positive, ranging from 3-13 Hz in stress-free environments, and from 14-24 Hz in polycyclic systems. However, due to the decreasing magnitude of 1J(29Si, 29Si) coupling with ring strain, it can be difficult to distinguish between one- and two-bond couplings.[9]
31P
[edit]See Phosphorus-31 NMR spectroscopy.
73Ge
[edit]There exists a single NMR-active isotope of germanium, 73Ge, which has a natural abundance of 7.76%. Even given its adequate abundance, germanium NMR is plagued by several complicating features. 73Ge has a spin of 9/2 which leads to very complex coupling, a very small negative gyromagnetic ratio, and a moderately large nuclear quadrupole. In addition, the sensitivity of 73Ge compared to 1H is only 1.08 x 10-4. All of this limits the usefulness of 73Ge NMR. The determination of 73Ge chemical shifts is difficult due to the extreme broadness of the peaks, but can span a range of 1200 ppm. When 73Ge NMR is performed, it is usually referenced to tetramethylgermanium (GeMe4) or germanium(IV) chloride (GeCl4). Coupling to 73Ge is rarely reported. In select cases, of simple symmetric compounds with significant coupling with 1H, 73Ge NMR can be obtained using special pulse sequences such as INEPT. This experiment significantly increases signal-to-noise ratio and decreases the time required for data collection.[10]
Sn
[edit]There are three NMR-active isotopes of tin: 115Sn, 117Sn and 119Sn. Each has a nuclear spin of ½, leading to relatively easy spectra interpretation. Natural abundances, however, make 117Sn and 119Sn more favourable: 117Sn and 119Sn have moderate natural abundance (7.51% and 8.58%, respectively), conversely, the natural abundance of 115Sn is very low at 0.35%, thus making this isotope less desirable for NMR spectroscopy. Compared to 1H, 115Sn, 117Sn and 119Sn have a sensitivity of 0.3272, 0.3563, and 0.3729, respectively. All three tin isotopes exhibit negative gyromagnetic ratios, which results in unfavorable Nuclear Overhauser Effect (NOE), including a drastic decrease of sensitivity in normal decoupled experiments. This can be overcome by performing an inverse gated decoupling experiment, which still provides a decoupled spectrum but without the loss of sensitivity.[10]
The range of chemical shifts for Sn NMR is very large, spanning 5000 ppm. Examples of this are illustrated in a variety of tin compounds: SnCp2 (-2199 ppm), SnI4 (-1701 ppm), SnBr4 (-638 ppm), SnCl4 (-150 ppm), Me3SnOMe (120.9 ppm), Me2Sn(OMe)2 (-126.3 ppm). Changes in Sn bond angles when in a ring drastically affect the chemical shift, wherein a decrease in angle generally increases the observed chemical shift. As an example, the δ119Sn for the six-membered ring shown is -42.5 ppm, while that for the five-membered ring is 53.5 ppm. For all three isotopes of tin, and unlike 1H NMR, chemical shifts are not determined from the magnitude of diamagnetic shielding as it is overwhelmed by the paramagnetic shielding. Chemical shifts are often referenced to either tetramethyltin, Sn(CH3)4, or tin(IV) chloride, SnCl4.[10]
There have been relatively few 1J(119Sn, 1H) measured, in part due to the difficulty and instability of suitable compounds. From available data, the 1J(119Sn, 1H) appears very large, approximately 1800 Hz. As the electronegativity of the substituents increases, the magnitude of this coupling decreases. While 1J(119Sn, 1H) is relatively rare, there is an abundance of 1J(119Sn, 13c) data available. The magnitude of the coupling is related to the C-Sn-C bond angle by the following equation, which has been derived empirically from a large set of dimethyltin(IV) compounds. In addition, for R4Sn (where R represents an alkyl chain), the coupling tends to decrease as the chain length increases. The sign is typically negative for all organotin(IV) compounds.[10]
- |1J(113Sn,13C)|=(10.7±0.5)θ-(778±64)
One-bond coupling with tin is less well-studied for other elements, but some data is available for 15N, 31P, 19F, 119Sn, 207Pb, and transition metals. In general, 1J(119Sn, 15N) tends to be large and negative, while 1J(119Sn, 31P) is often large and positive. 1J(119Sn, 19F) tends to be negative but of smaller magnitude than that of 15N. While some negative values have been observed in systems of small s-overlap due to bulky ligands, 1J(119Sn, 119Sn) are most often small and positive, and dependent heavily on the hybridization of tin. For the same reason, 1J(119Sn, 207Pb) are positive but very small due to poor s-s overlap. Finally, one-bond couplings for many transition metal-tin complexes have been measured using transition metal-SnCl3 complexes. In general, these all tend to be large. This can also be used to probe the geometry of the system, as the magnitude of the coupling is dependent on the trans-effect of the substituent, for example coupling is 28 052Hz in trans-Pt(SnCl3)(Cl)(PPh3)2 and 16 321Hz in cis-Pt(SnCl3)(Cl)(PPh3)2.[10]
Geminal coupling constants for tin, 2J(119Sn, Y), are sensitive to the a number of characteristics: the nature of the atom in between, the Sn-X-Y bond angle, and the stereochemistry of the molecule, as well as those factors affecting one-bond coupling. 2J(119Sn, 1H) have primarily been studied where the intervening atom is carbon, and these commonly result in positive values. The magnitude of the coupling increases with increasing Sn coordination number. As with one-bond coupling to metals, 2J(119Sn, 1H) can be used to determine the stereochemistry of a transition metal complex: larger coupling is observed when H is trans to Sn than when it is cis. 2J(119Sn, 13C) couplings tend to be positive and small when observing an aliphatic moiety, while it is negative and large with an olefinic moiety.[10]
Vicinal and long-range couplings are mostly dependent on the dihedral angle, and follow the Karplus relation. The magnitude for 3J(119Sn, 1H) and 3J(119Sn, 13C) are usually larger than the respective geminal coupling. Similar to what is found in 1H NMR, trans 3J(119Sn, 1H) coupling is much larger than cis 3J(119Sn, 1H).[10]
Transition Metal Nuclei
[edit]47Ti and 49Ti
[edit]Titanium is unique in that two of its natural isotopes, 47 (7.41%) and 49(5.41%)[2], are spin active (I= 5/2 and 7/2 respectively) and appear in the same spectrum. In any Titanium spectrum, two signals are obtained 266 ppm apart. Sensitivity is low, however, due to low abundance and low magnetogyric ratio. [11]
59Co
[edit]See Cobalt-59 NMR
107Ag and 109Ag
[edit]Sliver has two NMR active nuclei, 107Ag and 109Ag. Both isotopes have a nuclear spin of ½ (I = 1/2). Although 107Ag is slightly more abundant and produces sharper signals than 109Ag, its sensitivity is lower due to a low magnetogyric ratio γ.[12] In addition, 107Ag has a resonant frequency beyond that of most standard NMR probes.[12]
Silver NMR is used for the study of silver salts and complexes.[12]
Both 107Ag and 109Ag have a chemical shift range from -25 to 725 ppm.[12] AgNO3 is used as a reference compound.
103Rh
[edit]Rhodium (I=1/2) benefits from a 100% natural abundance. It suffers from low gyromagnetic ratio which can be partially overcome by Polarization transfer experiments such as inverse-INEPT(178x intensity) and HMQC(5610x intensity). Use of such techniques require bonded 1H or 31P. As a result, Rhodium NMR is usually confined to organometallic species, where it finds use in studying catalysis and bonding. Chemical shifts have a large breadth of ppm and are measured against Ru(acac)3. In some cases, the chemical shift has been shown to correlate directly with the cone angle of phosphine ligands.[1]
207Pb
[edit]207Pb has a nuclear spin of ½ (I = 1/2) with medium sensitivity to NMR spectroscopy. 207Pb NMR is usually used for studying organolead compounds.
207Pb has a wide range of chemical shift -5500 to 6000 ppm.[13] The reference compound is Me4Pb.[13]
195Pt
[edit]Platinum was one of the first metals to be implemented in NMR studies to give structural data. NMR capabilities began in the 1960's. Since then, it has become a popular choice in NMR experiments due to its characteristic splitting pattern, its relative sensitivity (.994% sensitive as 1H), and its propensity to be diamagnetic. It is used in studying structure and reactivity of Platinum complexes, often for potential catalysts, and in studying drug binding kinetics.[14]
Platinum has only one abundant spin active nuclei:195Platinum (I=1/2, 33.8% natural abundance[2]). Its reference varies, but is usually PtCl62- or 21.4 MHz (at 2.35 T). Its chemical shift is noticeably temperature dependent, and is quite sensitive to substitution, geometry, and oxidation state[15]. Emperical methods exist to predict chemical shift, however DFT calculations are considered more precise.[16]
Coupling to Platinum gives good structural information, as coupling constants are generally large and the multiplicity pattern is distinct. Platinum is spin dilute (33.8%), so any signal is split to approximately 1:4:1 multiplet, usually called a singlet with Pt-satellites. This arises due to the uncoupled singlet and the coupled doublet, each containing 16.9% of the signal.[1]
References
[edit]- ^ a b c d Gunther, Harald (2013). NMR Spectroscopy (3 ed.). Weinheim, Germany: Wiley-VCH. pp. 430–500. ISBN 978-3-527-33000-3.
- ^ a b c d e f g Atomic Weights and Isotopic Compositions for All Elements. National Institute of Standards and Technology - Physical Measurement Laboratory. http://physics.nist.gov/cgi-bin/Compositions/stand_alone.pl
- ^ a b c d e f g h i J. H. Nelson, "Nuclear Magnetic Resonance Spectroscopy" Prentice Hall, Upper Saddle River, NJ, 2003
- ^ a b c d Boron NMR. ©Roy Hoffman and Yair Ozery, The Hebrew University, Revised 2013-12-31T12:55+03. http://chem.ch.huji.ac.il/nmr/techniques/1d/row2/b.html
- ^ Nitrogen NMR. ©Roy Hoffman and Yair Ozery, The Hebrew University, Revised 2013-04-09T15:06+03. http://chem.ch.huji.ac.il/nmr/techniques/1d/row2/n.html#n15
- ^ a b c d Webb, GA (1999). Nitrogen NMR. Elsevier Ltd. p. 1790-1799.
- ^ Marek, Radek (2010). 15N NMR Applications. Czech Republic: Elsevier Ltd.
- ^ a b 27Aluminium NMR. ©Roy Hoffman and Yair Ozery, The Hebrew University, Revised 2013-12-31T15:34+02. http://chem.ch.huji.ac.il/nmr/techniques/1d/row3/al.html
- ^ a b c d e Marsmann, Heinrich (1999). 29Si NMR. Elsevier Ltd. p. 2539-2549.
- ^ a b c d e f g Pettinari, Claudio (1999). Heteronuclear NMR Applications (Ge, Sn, Pb). Elsevier Ltd. p. 798-811.
- ^ Braun, S. (1998). 150 and More Basic NMR Experiments: A Practical Course. Weinheim, Germany: Wiley-VCH. ISBN 3257295127.
{{cite book}}
: Check|isbn=
value: checksum (help) - ^ a b c d Silver NMR. ©Roy Hoffman and Yair Ozery, The Hebrew University, Revised 2014-02-23T16:27+03. http://chem.ch.huji.ac.il/nmr/techniques/1d/row5/ag.html
- ^ a b Lead NMR. ©Roy Hoffman and Yair Ozery, The Hebrew University, Revised 2014-02-24T12:11+02. http://chem.ch.huji.ac.il/nmr/techniques/1d/row6/pb.html
- ^ Chem. Soc. Rev. (36): 665–686. 2007. doi:10.1039/B606190G.
{{cite journal}}
: Missing or empty|title=
(help) - ^ Benn, R.; et al. (1985). Journal of Organometallic Chemistry. 282: 291–295. doi:10.1016/0022-328X(85)87180-1.
{{cite journal}}
: Explicit use of et al. in:|last1=
(help); Missing or empty|title=
(help) - ^ Chem. Soc. Rev. (36): 665–686. 2007. doi:10.1039/B606190G.
{{cite journal}}
: Missing or empty|title=
(help)