chapter 19 chlorine, bromine, and iodine solid-state … 19 chlorine, bromine, and iodine...
TRANSCRIPT
Chapter 19Chlorine, Bromine, and Iodine Solid-StateNMR
David L. Bryce, Cory M. Widdifield, Rebecca P. Chapman, andRobert J. AttrellDepartment of Chemistry and Centre for Catalysis Research and Innovation, University of Ottawa,10 Marie Curie Private, Ottawa, ON K1N 6N5, Canada
19.1 Introduction and NMR Properties of theQuadrupolar Halogens 321
19.2 Experimental Aspects 32219.3 Representative Quadrupolar and Chemical
Shift Data and Discussion ofApplications 328
19.4 Conclusions and Future Prospects 343References 345
19.1 INTRODUCTION AND NMRPROPERTIES OF THEQUADRUPOLAR HALOGENS
Chlorine, bromine, and iodine all have isotopesthat are amenable to study by solid-state nuclearmagnetic resonance (SSNMR) spectroscopy.1 – 3 Inthis chapter, we present an overview of the NMRproperties of the spin-active isotopes of Cl, Br,and I, details of the experimental aspects of dataacquisition and interpretation, and a survey of NMRdata available, along with some highlights of im-portant practical applications of chlorine, bromine,
NMR of Quadrupolar Nuclei in Solid MaterialsEdited by Roderick E. Wasylishen, Sharon E. Ashbrook andStephen Wimperis© 2012 John Wiley & Sons, Ltd. ISBN: 978-0-470-97398-1
and iodine SSNMR. The focus is on powdered sam-ples. Shown in Table 19.1 are the NMR properties ofthe spin-active isotopes of chlorine (35/37Cl), bromine(79/81Br), and iodine (127I).4 – 6
In contrast to the more widely studied 19F (I =1/2)7, all of these nuclides are quadrupolar, withspin quantum numbers of 3/2 (35/37Cl, 79/81Br) or5/2 (127I). The natural abundances of these isotopesare moderate (24.22% for 37Cl) to excellent (100%for 127I) and the resonance frequencies range frommoderately low (37Cl and 35Cl are below 15N) tocomparable to 13C (79Br and 81Br). As such, SSNMRstudies of these quadrupolar halogen nuclides canbe relatively straightforward under favorable condi-tions. In addition to the magnetic shielding interac-tion, the NMR spectra of these quadrupolar nucleiare influenced by the quadrupolar interaction (QI) be-tween the nuclear electric quadrupole moment andthe electric field gradient (EFG) at the nucleus (seeChapter 2). This interaction is often described usingthe quadrupolar coupling constant (CQ = eV33Q/h)and asymmetry parameter (ηQ = (V11 − V22)/V33).The main limiting factors in recording SSNMRspectra of 35/37Cl, 79/81Br, and 127I, as with manyquadrupolar nuclei, are the broad line widths, whichmay be observed for powdered samples when thenucleus of interest sits at a site with a large EFG.Halide anions that are either isolated or located at
322 Applications
high-symmetry crystallographic sites have low (orzero) EFGs, and for this reason, the majority ofSSNMR studies of chlorine, bromine, and iodinehave been of compounds wherein these elements arepresent in their ionic halide form (Cl−, Br−, andI−).1,2 Some data have also been recorded for per-halogenate ions (XO−
4 ).8 – 13 Nuclear quadrupole reso-nance (NQR) methods have been applied extensivelyto measure CQ (and ηQ in the case of 127I) for halo-gens in covalent bonding environments.14 It is alsointeresting to note that CQ(35/37Cl) has been mea-sured indirectly by observing the NMR spectrum ofspin-1/2 nuclei which are spin coupled (dipolar andJ ) to chlorine.15
The effects of quadrupolar line broadening tend todominate the SSNMR spectra of Cl, Br, and I, exceptwhen the nucleus is in a perfectly symmetric environ-ment. As a result, the spectra (and the correspondingvalues of CQ and ηQ) are very sensitive to the localenvironment as well as to the longer range crystalpacking. The diagnostic nature of the spectral param-eters has been demonstrated to be useful in crystalstructure refinement,16 – 18 in the characterization ofpolymorphs,19 and in the discrimination of differenthydrates of various compounds17,20,21 (vide infra).
In most cases of interest, the QI is sufficientlylarge to broaden the satellite transitions (STs) to suchan extent that they cannot be easily observed, andexperimentally only the central transition (CT) spec-trum is typically recorded. The CT is not broadenedby first-order quadrupolar effects, but is affected bysecond-order effects. The breadth of the second-orderCT lineshape of a stationary powdered sample (�νCT)may be described with the following equation5:
�νCT =[
3CQ
2I (2I − 1)
]2
× (η2Q + 22ηQ + 25)(I (I + 1) − 3/4)
144ν0
The inverse relationship between �νCT and ν0demonstrates why higher applied magnetic fields arebeneficial when acquiring SSNMR spectra of thequadrupolar halogens: increasing B0 increases thevalue of ν0, and therefore results in a decrease inthe second-order quadrupolar broadening of the CT.High applied magnetic fields also result in an increasein Boltzmann sensitivity. High fields are also veryvaluable for the measurement of chlorine, bromine,and iodine chemical shift (CS) tensors.
When interpreting trends in quadrupolar couplingconstants for Cl, Br, and I, it is important to consider
Table 19.1. NMR properties of the quadrupolarhalogens4–6
N.A.a I γ /107 rad Q/mb 1-γ∞(%) T−1s−1
35Cl 75.78 3/2 2.624198 −81.65(80) 43.037Cl 24.22 3/2 2.184368 −64.35(64) 43.079Br 50.69 3/2 6.725616 313(3) 8181Br 49.31 3/2 7.249776 262(3) 81127I 100.0 5/2 5.389573 −696(12) 163
aNatural abundance.
the Sternheimer antishielding factor, conventionallyexpressed as 1 − γ∞.5,14,17 This factor accounts forthe EFG at the nucleus owing to the polarization ofinner-shell electrons.22,23 That is, the EFG due solelyto ionic charges in the lattice, eqlattice, is modified asshown below24:
eqobs = (1 − γ∞)eqlattice
The quantity eqobs must be considered for the accu-rate discussion of quadrupolar coupling constants andrelative line widths. The antishielding factors for Cl,Br, and I are presented in Table 19.1. Considerationof the large value for iodine explains why it can bedifficult to observe 127I SSNMR spectra in a series ofisostructural halide compounds for which observationof the 35/37Cl and 79/81Br SSNMR spectra is feasible(Figure 19.1).
Given that 127I quadrupolar coupling constants inparticular can be quite large, thereby leading to partic-ularly broad powder patterns, some discussion is pro-vided below regarding the application of suitable dataacquisition methods (e.g., variable radiofrequency(RF) transmitter offsets and signal enhancement). Theutility of complementary 127I NQR measurements isnoted,14,17 as well as guidelines pertaining to thevalidity of second-order perturbation theory in theanalysis of spectral lineshapes.17,25 – 28
19.2 EXPERIMENTAL ASPECTS
19.2.1 Chemical Shift Referencing
19.2.1.1 Chlorine-35/37
The IUPAC CS reference for chlorine is 0.1 moldm−3 NaCl in D2O.4 With this standard, the exactconditions of preparation are very important as they
Chlorine, Bromine, and Iodine NMR 323
127I
127I
79Br
81Br
81Br
79Br
35Cl
35Cl
37Cl
37Cl
0
20
40
60
80
100
120
Q2
I(I+
1)–
3/4
(2I(
2I−1
))2
Cs1
33M
o95
O17
Ca4
3V
51R
u99
B11Li
7
K39
CI3
7
Sc4
5S
33
AI2
7Z
n67
Nb9
3T
i49
Sr8
7C
I35
Zr9
1M
g25
Na2
3G
a71
Bi2
09R
b87
Co5
9T
i47
Cr5
3M
n55
Sb1
21In
115
Cu6
5C
u63
Ba1
37B
r81
As7
5I1
27
Br7
9
La13
9
(a)
(b) (c)
Figure 19.1. Computed central transition line widths for the quadrupolar halogens with axially symmetric EFGs in powdersamples. (a) Comparison of the intrinsic CT line width for various quadrupolar nuclei on the basis of their quadrupolemoments and nuclear spin quantum numbers only (assuming a constant EFG at the nucleus and a constant Larmor frequency).(b) Comparison of the CT powder patterns for the isotopes of chlorine, bromine, and iodine, under the same conditionsas (a) except also including the effect of the relative Larmor frequencies. (c) Same as (b) except also including a scalingof the EFG by the Sternheimer antishielding factor to account for its effect on the effective EFG at the nucleus. Data areplotted in (b) and (c) on ppm scales. Scales are different in (b) and (c), i.e., the breadth of the 127I powder pattern has beenarbitrarily held constant in the two sets of spectra for ease of viewing.
324 Applications
may significantly affect the observed isotropic CS.There is a concentration dependence of the chlorineCS as well as strong solvent effects if H2O is usedinstead of D2O.3,29,30 Solid sodium chloride has beensuggested as an alternative reference, as it does nothave concentration or solvent issues and has a nar-row line width at relatively low magic-angle spinning(MAS) frequencies.3 An alternative secondary refer-ence is solid potassium chloride, which was found tobe ideal as a result of the small dipole–dipole cou-pling between potassium and chlorine along with thevanishingly small EFG at the chlorine nucleus.31 Thechlorine CSs of all the alkali metal chlorides with re-spect to the IUPAC reference are listed in Table 19.2.2
Shown in Figure 19.2 are the known CS ranges forchlorine, bromine, and iodine.
In addition, chlorine is the only quadrupolarhalogen for which a reliable and precise absoluteshielding scale has been determined. Using acombination of experiment and theory, Gee et al.established a value of 974(4) ppm for the chlorine
Table 19.2. Chlorine NMR data for cubic alkalimetal chlorides
Compounds δisoa (ppm)
LiCl 9.93NaCl −41.11KCl 8.54RbCl 49.66CsCl 114.68
.Source: From Ref. 2.aFrom 35Cl SSNMR data; with respect to 0.1 moldm−3 NaCl in D2O.
isotropic magnetic shielding constant (σiso) of aninfinitely dilute aqueous NaCl solution.32
19.2.1.2 Bromine-79/81
The IUPAC CS reference for bromine is 0.01 moldm−3 NaBr in D2O.4 Once again, care must be takenin preparation for the same reasons as noted above
4000 3600
20002400
1000 800 600 400 200 0 −200
1600 1200 800 400 0 −400
Per
ioda
tes,
MIO
4
Per
brom
ates
, MB
rO4
Per
chlo
rate
s, M
CIO
4
Gro
up IV
orga
nom
etal
lic c
hlor
ides
Gro
up 1
3 ch
lorid
es
Alk
alin
e ea
rth
chlo
rides
Cop
per
chlo
ride
Sod
alite
s
Cad
miu
m io
dide
Alk
alin
e ea
rth
brom
ides
Am
mon
ium
brom
ides
Silv
er b
rom
ide
Alk
alin
e ea
rth
iodi
des
Cop
per
iodi
de
Cop
per
brom
ide
Sod
alite
s
Silv
er io
dide
Alk
ali m
etal
iodi
des
Alk
ali m
etal
brom
ides
3200 2800 2400 2000
d (127I) /ppm
d (79/81Br) /ppm
d (35/37Cl) /ppm
0.01 M KI in D2O
0.01 M NaBr in D2O
0.1 M NaCl in D2O
1600 1200 800
Ionic liquids
Ionic liquids
Ionic liquids
400 0 −400 −800
Chloride ion receptorI Hyd
roch
lorid
es
Alk
ali m
etal
chlo
rides
Figure 19.2. Known chlorine, bromine, and iodine chemical shift ranges for solids.
Chlorine, Bromine, and Iodine NMR 325
for chlorine and thus a secondary reference of eithersolid sodium or potassium bromide is convenient. Thebromine CSs of all the alkali metal bromides withrespect to 0.03 mol dm−3 NaBr in D2O are presentedin Table 19.8.1
19.2.1.3 Iodine-127
The IUPAC CS reference for iodine is 0.01 mol dm−3
KI in D2O.4 This suggested standard presents furtherpractical problems in addition to concentration andsolvent CS dependences. As a result of the dilutenature of iodine in the IUPAC suggested solution,the time required to obtain an adequate signal makesit undesirable as a setup sample in our experience.Therefore, in addition to the reasons mentioned abovefor bromine and chlorine, sensitivity issues make itdesirable to use solid potassium iodide or sodiumiodide as secondary standards. The iodine CSs of allthe alkali metal iodides with respect to the IUPACreference are reported in Table 19.11.2
19.2.2 Data Acquisition Techniques
For cases where the lineshape is sufficiently nar-row (i.e., maximally on the order of tens of kilo-hertz), MAS NMR experiments are of great utilityfor observing the SSNMR signals associated with thequadrupolar halogen nuclei. In fact, the 79Br SSNMRsignal of KBr is often used to precisely calibratethe magic angle.33 When MAS is used, SSNMRlineshape broadening due to the dipole–dipole andchemical shift anisotropy (CSA) mechanisms canbe removed. In addition, the residual second-orderquadrupolar broadening will be reduced by roughly afactor of 2.43–3.43.5 Owing to the removal of the ad-ditional line broadening mechanisms, lineshape mod-eling of the MAS SSNMR spectra of the quadrupolarhalogen nuclei is greatly simplified, and the EFG ten-sor information (CQ, ηQ), as well as the isotropichalogen CS, can often be readily determined. Moregenerally, available MAS rates will not be rapidenough and spectra can be obtained under station-ary conditions. Furthermore, in many cases, depend-ing on the magnetic field strength and the chemicalenvironment of the halogen under investigation, spec-tra of the CT of powdered samples cannot be acquiredin a single piece. Potential remedies for data acqui-sition of very broad lineshapes are discussed in theSection 19.2.2.3.
19.2.2.1 Echoes
Although a variety of advanced pulse sequences nowexist to acquire SSNMR data, it is often the casethat more traditional echo experiments (especiallywhen the experiments are performed at very highmagnetic fields) are acceptable to produce high S/Nratio spectra in a reasonable amount of time. Thisis especially true for the 79/81Br and 127I nuclides,as their magnetogyric ratios and natural abundancesare rather high. This is also generally true even inthe case of exceedingly broad (>1 MHz) lineshapes.For the acquisition of 35/37Cl SSNMR spectra, how-ever, other pulse sequences, such as quadrupolarCarr–Purcell–Meiboom–Gill (QCPMG), may be ofuse (see Section 19.2.2.2) because of the lower inher-ent sensitivity of these nuclides. The 90/90 or “solidecho” experiment34 is suggested for situations wherea somewhat strong signal is being observed, as itcan produce more uniform excitation at a given RFnutation frequency relative to the 90/180 or “Hahnecho”35 experiment. A theoretical and experimentaldiscussion of the optimal conditions for echoes per-formed on half-integer quadrupolar nuclei has beenprovided by Bodart et al.36 The relative increase inthe uniform signal excitation afforded using the solidecho experiment comes at the expense of reducedsignal intensity (by roughly a factor of 2) relativeto the Hahn echo experiment; it is important to notethat the exact factor will depend on the phase cy-cling and coherence pathway selection. At a typicalmagnetic field of 11.75 T, and assuming an RF fieldof 100 kHz (for 79/81Br and 127I) and 40 kHz (for35/37Cl), solid echo experiments using one transmittersetting are expected to be ideal until CQ values ex-ceed ∼12–14 MHz for bromine, 25 MHz for iodine,and ∼5 MHz for chlorine.
19.2.2.2 Signal Enhancement Methods
While echo experiments may be useful forthe acquisition of SSNMR signals associatedwith the quadrupolar halogen nuclei, in manysituations sensitivity-enhancing pulse sequenceswill be required. The QCPMG37 pulse sequencetakes advantage of the potentially significantratio between the natural transverse (spin–spin)relaxation time constant of the sample (i.e., T2),and the same relaxation time constant due to B0inhomogeneities (i.e., T2
∗). Through the use of a“train” of refocusing π pulses, the time-domain
326 Applications
60 40 20 0
n (35/37Cl) /kHz
−20 −40 −60 −80
(a)
(b)
(c)
(d)
(e)
(f)
37Cl
35Cl
Figure 19.3. (a) 35Cl QCPMG and (b) π/2 − τ − π/2 − τ
echo and (e) 37Cl π/2 − τ − π/2 − τ echo NMR spectra ofpowdered CaCl2·2H2O acquired at 11.75 T under stationaryconditions. Shown in (c) and (f) are the correspondingbest-fit simulated spectra. For comparison, shown in (d)is the simulated 35Cl NMR spectrum obtained when thechemical shift tensor span is assumed to be zero. Spectralparameters are given in Table 19.4. (Reproduced from Ref.20. © Wiley-VCH, 2007.)
system response resembles a series of evenly spacedspikes (spacing = τ ). Once subjected to Fouriertransformation, the frequency-domain response willalso be a series of spikelets, which carry a separationof 1/τ (Figure 19.3). While experimental resolutionis reduced using this method, the sensitivity isincreased, as multiple echoes are collected perscan.
In addition to the QCPMG experiment, therotor-assisted population transfer (RAPT),45
double-frequency sweeps (DFS),46 and hyperbolicsecant (HS)47 pulse sequences may also prove usefulfor 35/37Cl SSNMR experiments. However, as thesepulse sequences rely upon the manipulation of theST populations, one must be able to reasonablyexcite the full ST manifold. Hence, the RAPT, DFS,and HS pulse sequences are not expected to be ofgeneral use for 79/81Br and 127I unless very highsite symmetry is present. The RAPT, DFS, and HSsequences have also been coupled with QCPMG andafford further sensitivity enhancement.48,49
19.2.2.3 Wideline Methods
Wideband, Uniform Rate, and Smooth Truncation(WURST) PulsesThe use of wideband, uniform rate, and smooth trun-cation (WURST)-type pulses in NMR experimentswas developed a number of years ago as a solutionto the problem of wideband inversion and broad-band decoupling in liquid state experiments.50 Re-cently, a number of studies have shown that WURSTpulses can also be used to uniformly excite broadspectral regions using a single transmitter setting forhalf-integer quadrupolar nuclei.51 In addition, theyhave been coupled with QCPMG.52 These pulses areso efficient at uniformly exciting broad spectral re-gions that the probe bandwidths are the limiting factorfor uniform signal detection, as was demonstratedduring the collection of the 127I SSNMR signal ofone of the sites in SrI2.17
Variable-Offset Cumulative Spectrum (VOCS) DataAcquisitionTo acquire meaningful SSNMR spectra of thequadrupolar halogen nuclei, data acquisition willoften be carried out using multiple RF transmittersettings. Failure to uniformly excite the entire CTpowder pattern will result in incorrect spectralanalysis and erroneous NMR tensor parameters.Unless high site symmetry (i.e., tetrahedral,octahedral) at the halogen nucleus is known a priori,it is advisable to obtain halogen SSNMR spectrausing at least two different transmitter frequenciesto confirm that the entire CT spectral pattern hasbeen acquired. A suggested offset for this endeavor(using a high-power amplifier with an echo pulsesequence) would be ca. 100–200 kHz. To acquirethe CT powder patterns of the quadrupolar halogennuclei that are clearly broadened beyond what canbe uniformly excited using a single transmitterfrequency, the variable-offset cumulative spectrum(VOCS) method is useful.53 This protocol simply in-volves collecting a set of SSNMR spectra, with eachspectrum being acquired at a unique, but uniformlyoffset, transmitter frequency (Figure 19.4). Eachspectrum is to be collected using the same numberof scans and should be processed individually.Once this is accomplished, the resulting spectra arecoadded in the frequency domain, producing thefinal VOCS spectrum.
Chlorine, Bromine, and Iodine NMR 327
6000 4000 2000 0
d (81Br) /ppm
−2000 −4000 −6000 −8000
Figure 19.4. Bromine-81 SSNMR spectrum of3-chloroanilinium bromide acquired at 21.1 T usingVOCS data acquisition. 3072 scans were acquired perpiece, and high-power proton decoupling was appliedduring acquisition. Each of the nine slices are shown, aswell as their sum (top).
19.2.3 Data Analysis
19.2.3.1 Importance of Acquiring Data at TwoMagnetic Fields and Utility of TwoSpin-Active Isotopes of Chlorine andBromine
One of the benefits when extracting information us-ing chlorine and bromine SSNMR spectroscopy isthe availability of two NMR-active nuclides for each.In order to accurately extract EFG and CS tensorinformation using SSNMR experiments on quadrupo-lar nuclei, data acquisition should be carried out atmultiple fields. Owing to the multiple NMR-activeisotopes for chlorine and bromine, multiple field dataacquisition is not a stringent requirement. This is be-cause each nucleus has a unique magnetogyric ratioand nuclear electric quadrupole moment, Q (as sum-marized in “Introduction and NMR Properties of theQuadrupolar Halogens”).6 In effect, if one conductsSSNMR experiments on both NMR-active isotopes,then the equivalent of multiple field data acquisi-tion has been performed using the internal nuclearproperties rather than by adjusting the external ap-plied field. This assumes that isotope effects on themagnetic shielding tensor are negligible, which isa valid assumption when dealing with SSNMR ofpowder samples. An ideal method for SSNMR dataacquisition using chlorine and bromine would be toacquire spectra of each nuclide within a very highB0. While the combination of multiple field and mul-tiple nuclide data acquisition is not a requirement
to extract the relevant NMR tensor information forchlorine and bromine, it becomes worthwhile whenconsidering samples that possess multiple sites, asdemonstrated for SrBr2.21
19.2.3.2 Breakdown of Second-OrderPerturbation Theory
When fitting the SSNMR spectrum of a half-integerquadrupolar nucleus, it is well known that the QIwill at least split the resonance into a multiplet struc-ture, with 2I − 1 STs flanking the CT. In the caseof a very weak QI, first-order perturbation theory hasbeen shown to be adequate. Other than for the cu-bic halides, however, first-order perturbation theoryis rarely sufficient to model the observed SSNMRspectrum associated with a quadrupolar halogen nu-cleus. In these systems, one generally measures theCT only, which will carry additional second-orderbroadening that can be used to extract the EFG andCS tensor information. While second-order perturba-tion theory is often sufficient to model the observedSSNMR lineshapes for quadrupolar nuclei, there ex-ist cases where even second-order perturbation theorydoes not lead to the correct NMR tensor parameters.Of the quadrupolar halogens, this is most likely tooccur when carrying out experiments upon the 127Inuclide. Indeed, an example regarding the interpreta-tion of the 127I SSNMR lineshape of one of the sitesin powdered SrI2 has been reported.17 It was foundthat even though the data were acquired at the high-est possible applied field (21.1 T), the internal QI wasstill strong enough (i.e., CQ = 214 MHz) to producea nonuniform shift in the 127I SSNMR spectrum thatcould not be accounted for using second-order per-turbation theory (Figure 19.5). A recently developedmodel, which combines the Zeeman and QI effectsexactly,28 was applied to arrive at a significantly dif-ferent isotropic CS value than was determined usingsecond-order perturbation theory. Since third-orderquadrupolar effects are zero for the CT, fourth-ordereffects on the CT are likely mainly responsible forthese observations. Third-order effects are non-zerofor the STs, however, and an exact simulation willof course take these and higher order effects into ac-count. This is likely to be important since the STpowder patterns will generally overlap partially withthe CT powder pattern for very large values of CQ.
328 Applications
i
iii
iv
v
ii
4
20000 −20000 −35000
14000 13000 12000 11000 10000
−18000
3000 2000 1000 0 −1000
−19000 −20000 −21000
−36000 −37000 −38000 −39000 d /ppm
d /ppm
d /ppm
d /ppm
22000 21000 20000 19000 18000 d /ppm
d /ppm0
2 0 −2 −4 −6 Δn0/MHz
i
ii
iii
iv
v
(a) (b)
Figure 19.5. Breakdown of second-order perturbation theory. (a) Comparison of 127I SSNMR powder patterns generatedusing second-order perturbation theory (solid red and black traces) with one calculated using exact theory (dashed bluetrace). CQ = 214 MHz; ηQ = 0.316 (values for one site in SrI2); B0 = 21.1 T. Iodine CSA (Ω = 460 ppm) is included inthe black trace simulation. Relative to the second-order perturbation theory simulations, the exact simulation is nonuniformlyshifted to lower frequency. Low-frequency shoulders (‡) are from STs. (b) Horizontal expansions of the regions in (a).These illustrate that the incorporation of a reasonable value for the iodine CSA generally influences the positions of thediscontinuities to a lesser extent than do higher order quadrupolar-induced effects in this case. (Reproduced from Ref. 17.© American Chemical Society, 2010.)
19.3 REPRESENTATIVE QUADRUPOLARAND CHEMICAL SHIFT DATA ANDDISCUSSION OF APPLICATIONS
19.3.1 Chlorine
Among the quadrupolar halogens, 35Cl is the nucleusthat has been studied most frequently with SSNMR,despite its moderately low resonance frequency.1 – 3
This is a result of its smaller quadrupole momentcompared to 79/81Br and 127I which, as mentionedabove, results in significantly less line broadening.While both NMR-active nuclei of chlorine have mod-erately large Q values and low Larmor frequencies,the higher natural abundance of 35Cl makes it themore popular of the two nuclei for study.
A wide variety of materials has been analyzed withchlorine SSNMR, ranging from biologically impor-tant model compounds to catalysts to geological sam-ples. The chlorine NMR properties of these different
classes of materials have been found to vary sig-nificantly, with isotropic chlorine CSs ranging from−100 to over 1000 ppm (with respect to the IUPACstandard) and 35Cl quadrupolar coupling constantsfor chloride ions varying from 0 to 40.4 MHz. Inaddition, the increasing availability of high field in-struments has allowed for the determination of CStensor parameters for many materials, with spans(Ω = δ11 − δ33) up to 800 ppm having been quan-tified.
Many early studies using chlorine SSNMR focusedon cubic salts, as the absence of a QI results innarrow lines. Therefore, the chlorine CSs for thealkali metal chlorides,2,54,55 along with other cubicchlorides (i.e., AgCl,56 CuCl56 – 58) are well knownand cover a significant range. For example, NaClappears at −41.11 ppm, whereas CsCl appears at114.68 ppm (with respect to 0.1 M NaCl in D2O). Forfurther information, the reader is referred to Refs 1and 2. See also Table 19.2.
Chlorine, Bromine, and Iodine NMR 329
Tabl
e19
.3.
Sele
cted
chlo
rine
NM
Rda
tafo
rso
lidor
gani
chy
droc
hlor
ides
Com
poun
ds|C
Q(35
Cl)|/M
Hz
ηQ
δ iso
a(p
pm)
Ω(p
pm)
κE
uler
angl
esb/d
egre
esR
efer
ence
s
Coc
aine
HC
l5.
027(
0.02
)0.
2(0.
05)
−41
——
—Y
esin
owsk
iet
al.38
l-Ty
rosi
neH
Cl
2.23
(0.0
2)0.
72(0
.03)
53.6
(0.5
)<
150
——
Bry
ceet
al.39
2.3(
0.1)
0.7(
0.1)
54(1
)—
——
Ger
vais
etal
.40
Gly
cine
HC
l6.
42(0
.05)
0.61
(0.0
3)60
(5)
100(
20)
0.30
(0.3
0)95
(20)
,0(
4),
0(20
)B
ryce
etal
.41
l-V
alin
eH
Cl
5.89
(0.0
5)0.
51(0
.05)
49(1
0)12
5(40
)0.
35(0
.50)
65(2
0),
0(20
),0(
20)
Bry
ceet
al.41
l-G
luta
mic
acid
HC
l3.
61(0
.01)
0.65
(0.0
2)61
(1)
66(1
5)0.
0(0.
3)9(
20),
77(2
0),
6(20
)B
ryce
etal
.41
Qui
nucl
idin
eH
Cl
5.25
(0.0
2)0.
05(0
.01)
14.0
(10.
0)50
(20)
——
Bry
ceet
al.39
l-C
yste
ine
ethy
les
ter
HC
l3.
78(0
.02)
0.03
(0.0
3)57
.5(0
.5)
47(4
)−0
.8(0
.2)
—B
ryce
etal
.39
l-C
yste
ine
met
hyl
este
rH
Cl
2.37
(0.0
1)0.
81(0
.03)
52.5
(0.7
)45
(15)
——
Bry
ceet
al.39
Cys
tein
eH
Cl
mon
ohyd
rate
3.92
(0.0
1)0.
47(0
.02)
63.1
(0.5
)66
(10)
0.12
(0.1
2)15
5(20
),0(
10),
0(20
)C
hapm
anet
al.42
l-Ly
sine
HC
l2.
49(0
.01)
0.42
(0.0
2)64
(2)
26(1
0)−0
.4(0
.4)
0(20
),52
(20)
,0(
20)
Bry
ceet
al.41
l-Se
rine
HC
l3.
0(0.
3)0.
8(0.
2)79
(30)
<15
0—
—B
ryce
etal
.41
l-Pr
olin
eH
Cl
4.50
(0.0
5)0.
63(0
.05)
−4(5
)63
(5)
−0.5
4(0.
08)
48(2
0),
69(3
),9(
20)
Bry
ceet
al.41
l-Is
oleu
cine
HC
l4.
39(0
.05)
0.25
(0.0
3)55
(20)
75(3
0)>
0.85
20(2
0),
12(2
0),
0(20
)B
ryce
etal
.41
l-Ph
enyl
alan
ine
HC
l6.
08(0
.05)
0.52
(0.0
3)55
(5)
129(
20)
0.26
(0.2
5)91
(20)
,13
(20)
,10
(20)
Bry
ceet
al.41
l-T
rypt
opha
nH
Cl
5.05
(0.0
4)0.
86(0
.03)
63.9
(1.0
)72
(5)
0.1(
0.1)
90(1
5),
20(1
5),
2(20
)B
ryce
etal
.18
d,l-
Arg
inin
eH
Cl
mon
ohyd
rate
2.03
5(0.
020)
0.98
(0.0
2)50
.4(1
.0)
57.5
(3.0
)0.
27(0
.10)
85(1
5),
77.5
(12.
0),
30(3
0)B
ryce
etal
.18
(con
tinu
edov
erle
af)
330 Applications
Tabl
e19
.3.
Con
tinu
ed
Com
poun
ds|C
Q(35
Cl)|/M
Hz
ηQ
δ iso
a(p
pm)
Ω(p
pm)
κE
uler
angl
esb/d
egre
esR
efer
ence
s
l-A
lani
neH
Cl
6.4(
0.1)
0.75
(0.0
6)65
(5)
60(3
0)−0
.3(0
.5)
90(1
5),
0(15
),0(
15)
Cha
pman
etal
.42
l-A
spar
ticac
idH
Cl
7.1(
0.1)
0.42
(0.0
5)61
(5)
75(3
0)−0
.9(0
.1)
0(20
),30
(20)
,93
(20)
Cha
pman
etal
.42
l-H
istid
ine
HC
lm
onoh
ydra
te4.
59(0
.03)
0.46
(0.0
2)52
(1)
<15
0—
—C
hapm
anet
al.42
l-M
ethi
onin
eH
Cl
4.41
(0.0
2)0.
35(0
.03)
58(1
)10
0(20
)0.
3(0.
3)93
(20)
,16
3(15
),7(
20)
Cha
pman
etal
.42
l-T
hreo
nine
HC
l5.
4(0.
1)0.
94(0
.02)
58(1
0)95
(40)
−0.2
(0.5
)95
(15)
,0(
10),
0(15
)C
hapm
anet
al.42
Proc
aine
HC
l4.
87(0
.07)
0.28
(0.0
4)55
(6)
125(
25)
−0.4
(0.3
)95
(15)
,3(
2),
32(8
)H
amae
det
al.19
Tetr
acai
neH
Cl
6.00
(0.1
0)0.
27(0
.04)
30(6
)80
(15)
0.4(
0.3)
60(8
),8(
5),
10(1
0)H
amae
det
al.19
Lid
ocai
neH
Cl
mon
ohyd
rate
poly
mor
ph1
4.67
(0.0
7)0.
77(0
.03)
59(4
)11
0(15
)−0
.85(
0.03
)12
(3),
40(1
0),
80(3
)H
amae
det
al.19
Lid
ocai
neH
Cl
mon
ohyd
rate
Site
1:2.
52(0
.12)
0.95
(0.0
5)44
(10)
20(1
0)−0
.8(0
.2)
90(4
0),
50(5
0),
60(4
0)H
amae
det
al.19
poly
mor
ph2
Site
2:5.
32(0
.10)
0.32
(0.1
0)69
(10)
45(1
0)0.
8(0.
2)5(
5),
50(1
5)40
(40)
Bup
ivac
aine
HC
lm
onoh
ydra
te3.
66(0
.10)
0.72
(0.0
8)55
(10)
100(
25)
0.2(
4)10
5(20
),90
(5),
5(5)
Ham
aed
etal
.19
1-B
utyl
-3-m
ethy
lim
idaz
oliu
mch
lori
deco
mpl
exof
mes
o-oc
tam
ethy
lca
lix[4
]pyr
role
1.0(
0.1)
0.7(
0.1)
79(1
0)50
(5)
0.4(
0.4)
15(3
0),
27(2
0),
60(1
5)C
hapm
anet
al.43
aW
ithre
spec
tto
0.1
mol
dm−3
NaC
lin
D2O
.bE
uler
angl
esde
term
ined
usin
gth
eA
rfke
nco
nven
tion.
(See
Ref
.44
)
Chlorine, Bromine, and Iodine NMR 331
19.3.1.1 Organic Hydrochlorides
Organic chloride and hydrochloride salts have beenextensively studied by chlorine SSNMR, with bothEFG and CS tensor parameters being extracted fora variety of materials (see Table 19.3 for selecteddata).1,2 The first 35Cl SSNMR study of a hydrochlo-ride salt was carried out by Pines and coworkerson powdered cocaine hydrochloride at 7.0 T understatic conditions.38 At this field, the VOCS methodwas required to collect the full spectrum, which wasmodeled to provide a 35Cl quadrupolar coupling con-stant of 5.027(0.02) MHz. Later, Bryce et al. re-ported a 35/37Cl NMR study of multiple hydrochlo-ride salts: l-tyrosine hydrochloride, l-cysteine methylester hydrochloride, l-cysteine ethyl ester hydrochlo-ride, and quinuclidine hydrochloride.39 MAS, Hahnecho, and QCPMG techniques were used to col-lect spectra at 9.4 and 18.8 T, allowing for the de-termination of EFG tensor parameters and chlorineCSs for all the salts. The authors noted an in-verse relationship between the number of hydrogenbonds and the magnitude of the quadrupolar couplingconstant.39
The study of amino acid hydrochlorides wascontinued by Gervais et al., who published amultinuclear SSNMR study of four amino acidhydrochlorides.40 Bryce and coworkers extendedthe use 35/37Cl SSNMR to study several otheramino acid hydrochlorides, noting that they arebiologically relevant models for chloride-bindingproteins (Figure 19.6).18,41,42 A total of 17 aminoacid hydrochlorides have been analyzed with35/37Cl SSNMR. The spectra were found to varysignificantly, depending on the precise nature ofthe chloride environment, and demonstrated thesensitivity of the technique to small changes instructure. The magnitude of CQ(35Cl) was found torange from 2.035 to 7.1 MHz, whereas the chlorineCSs varied from −4 to 79 ppm (with respect to theIUPAC standard).18,41,42 In addition, the collectionof spectra at 21.1 T allowed for the reliabledetermination of the CS tensor parameters and therelative orientation of the CS and EFG tensors forthe majority of these amino acid hydrochlorides,with the CS tensor spans ranging from 26 to129 ppm.18,41,42 Interestingly, a large quadrupolarcoupling constant does not always correlate witha large CS span, demonstrating that the two inter-actions are governed by different factors and thusoffer complementary information. Computationalstudies using the B3LYP (Becke’s three-parameter
1000 5000 0
d /ppm d /ppm
−1000 −500
(a) (e)
(f)
(g)
(b)
(c)
(d)
Figure 19.6. Solid-state 35/37Cl NMR spectroscopy ofl-cysteine hydrochloride monohydrate. Experimental spec-tra of stationary powdered samples: (a) 35Cl at 11.75 T, (c)37Cl at 11.75 T, (e) 35Cl at 21.1 T. Simulations incorporatingEFG and CS tensor parameters listed in Table 19.3 appear in(b), (d), and (f). (g) Depicts a simulation assuming no CSA.(Reproduced from Ref. 42. © Royal Society of Chemistry,2007.)
Lee–Yang–Parr exchange-correlation functional)hybrid DFT method and the restricted Hartree–Fock(RHF) method also played an important role inthese studies, and aided in the interpretation ofthe SSNMR spectra. The two methods, as wellas different basis sets, were tested against theexperimental data and the neutron diffractionstructures of several amino acid hydrochlorides, andan optimized method for calculating chlorine NMRparameters was determined.18,41 The optimal methodwas applied to calculate the expected chlorine NMRparameters of a chloride ion channel and showedthat chlorine NMR on complex systems should befeasible, in that prohibitively broad CT lineshapesare not expected.41 In addition, the importanceof optimizing hydrogen atom positions in caseswhere only an X-ray structure is available andthe effectiveness of including point charges forimproving the agreement between computationaland experimental values were shown.18,42 The
332 Applications
amino acid hydrochloride series of data was alsoutilized to test the accuracy of the GIPAW–DFTmethod in calculating chlorine NMR parameters inorganic hydrochlorides.43 The results showed againthat the optimization of hydrogen atom positionswas essential and that while both the value of|CQ(35Cl)| and the CS tensor span were overesti-mated by calculations, the experimental trends inboth parameters were reproduced. While both thecluster model calculations using B3LYP/RHF andGIPAW–DFT methods using periodic boundaryconditions are useful, it appears that the lattertechnique is slightly more accurate relative to thepresently available experimental chlorine NMRdata.18,42,43
In 2008, 35Cl SSNMR was shown to be an effectivemethod to distinguish polymorphs in a study of phar-maceuticals by Hamaed et al.19 Four local anestheticpharmaceuticals were examined at 9.4 and 21.1 T:procaine HCl, tetracaine HCl, monohydrated lido-caine HCl (LH), and monohydrated bupivacaine HCl(BH). The EFG and CS tensors, along with their rel-ative orientations, were reported for all four salts, aswell as high temperature polymorphs of LH and BH.The parameters were found to be in a similar rangeto those observed for the amino acid hydrochlorides,with the CSs ranging from 30 to 77 ppm (with respectto the IUPAC standard), CS tensor spans rangingfrom 20 to 160 ppm, and CQ(35Cl) magnitudes rang-ing from 2.52 to 6.00 MHz. Notably, 35Cl SSNMRwas found to be more effective than powder X-raydiffraction and 13C SSNMR at distinguishing poly-morphs in these cases.
Owing to the moderately large Q of the twospin-active chlorine nuclei, the QI typicallydominates the CT lineshape when the chlorine isin a noncubic environment. A 35/37Cl SSNMRstudy43 carried out at 9.4 and 21.1 T of the1-butyl-3-methylimidazolium chloride complex ofmeso-octamethylcalix[4]pyrrole, a known anionreceptor, provides a counter example. Specifically,at 21.1 T, the 35/37Cl NMR spectra collected understationary conditions exhibited lineshapes typicalof spin-1/2 nuclei, i.e., the powder pattern isdominated by CSA. The value of |CQ(35Cl)| wasfound to be low compared to those of the aminoacid hydrochlorides, at only 1.0 MHz, but is withinthe range observed for perchlorates, ionic liquids(vide infra), and a series of several alkylammoniumchlorides studied by Honda.59
19.3.1.2 Metal Chlorides
Compared to the organic hydrochlorides, there aresignificantly fewer inorganic noncubic chloride saltsthat have been analyzed by chlorine SSNMR,60,61
although the number has grown significantly in re-cent years (see Table 19.4 for chlorine NMR data forselected metal chlorides). In 2007, a study of sev-eral alkaline earth chloride hydrates, in which bothEFG and CS tensor information was extracted, wascarried out at magnetic fields of 11.7 and 21.1 T(Figure 19.3).20 The values of |CQ(35Cl)| observedfor the hydrates were in the range 1.41 to 4.26 MHz,whereas the CS tensor spans ranged from 41 to72 ppm, both on the order of the values observedfor the organic hydrochlorides. The chlorine CSs,however, are higher than those for the organic hy-drochlorides. In addition, the study demonstrated thepower of chlorine SSNMR to distinguish pseudopoly-morphs in the case of the anhydrous, dihydrate, andhexahydrate forms of SrCl2, which displayed verydifferent chlorine NMR tensor parameters. It wasnoted in this study that the CSs decreased as hydra-tion increased. In addition, the study demonstratedthat using the GIPAW-DFT method to calculate chlo-rine NMR tensor parameters yielded good agreementwith experimental results for all samples for which aneutron diffraction structure was available.
Calcium chloride and several of its hydrates werefurther investigated in a subsequent report.62 Thestudy presents 35Cl NMR spectra of the anhydrousand hexahydrate forms at 11.75 and 21.1 T as wellas GIPAW–DFT calculations of the chlorine NMRparameters for other pseudopolymorphs. It was notedthat the new data for CaCl2 were consistent with theexpected values for CQ(35Cl) and δiso, on the basisof its crystal structure and prior correlations betweenthese parameters and local structure.20 The experi-mental CS and CQ(35Cl) values for the anhydroussalt appeared at 105 ppm (with respect to the IUPACstandard) and 8.82(8) MHz, both significantly higherthan the values for CaCl2·6H2O.62
Rossini et al. have studied several organo-metallic compounds, including Cp2TiCl2, CpTiCl3,Cp2ZrCl2, Cp2HfCl2, Cp∗
2ZrCl2, CpZrCl3, Cp∗ZrCl3,Cp2ZrMeCl, Cp2ZrHCl, and (Cp2ZrCl)2(μ-O)(Cp = cyclopentadienyl; Cp∗ = pentamethylcyclo-pentadienyl) (Figure 19.7).61 These compoundsexhibit relatively large chlorine QIs and there-fore the VOCS–QCPMG method and a veryhigh magnetic field (21.1 T) were employed inmany cases to collect the full CT spectrum. The
Chlorine, Bromine, and Iodine NMR 333
Tabl
e19
.4.
Sele
cted
chlo
rine
NM
Rda
tafo
rso
lidm
etal
chlo
ride
s
Com
poun
ds|C
Q(35
Cl)|/M
Hz
ηQ
δ iso
a(p
pm)
Ω(p
pm)
κE
uler
angl
esb/d
egre
esR
efer
ence
s
CaC
l 2·2H
2O
4.26
(0.0
3)0.
75(0
.03)
68.9
(2.0
)72
(15)
0.6(
0.2)
90(1
0),
82(5
),0(
20)
Bry
cean
dB
ultz
20
BaC
l 2·2H
2O
Site
1:2.
19(0
.08)
0.00
122.
3(2.
0)50
(25)
−0.8
(0.2
)85
(20)
,32
(10)
,60
(20)
Bry
cean
dB
ultz
20
Site
2:3.
42(0
.08)
0.31
(0.1
0)11
5.5(
2.0)
50(2
5)0.
20(0
.25)
20(1
5),
8(10
),0(
20)
—M
gCl 2
·6H2O
3.02
(0.0
5)0.
0033
.9(1
.0)
<75
——
Bry
cean
dB
ultz
20
SrC
l 2∼0
n/a
147.
1(1.
0)—
——
Bry
cean
dB
ultz
20
SrC
l 2·2H
2O
1.41
(0.0
2)0.
80(0
.10)
101.
0(1.
0)41
(10)
0.5(
0.2)
86(1
5),
75(5
),37
(10)
Bry
cean
dB
ultz
20
SrC
l 2·6H
2O
3.91
(0.0
5)0.
0049
.3(1
.0)
45(2
0)−1
.00(
10),
90(1
0),
0(10
)B
ryce
and
Bul
tz20
CaC
l 28.
82(0
.08)
0.38
3(0.
015)
105(
8)13
5(15
)0.
0(0.
3)90
(20)
,90
(5),
0(5)
Wid
difie
ldan
dB
ryce
62
CaC
l 2·6H
2O
4.33
(0.0
3)<
0.01
57(3
)40
(8)
−1n/
a,90
(7),
0(8)
Wid
difie
ldan
dB
ryce
62
Cp 2
TiC
l 2c
22.1
(0.5
)0.
61(0
.03)
500(
500)
——
—R
ossi
niet
al.61
Cp 2
ZrC
l 216
.0(0
.5)
0.72
(0.0
4)30
0(15
0)80
0(50
0)0.
0(0.
5)2(
10),
72(2
0),−7
0(20
)dR
ossi
niet
al.61
Cp 2
HfC
l 217
.1(0
.4)
0.65
(0.0
5)40
0(50
0)—
——
Ros
sini
etal
.61
Cp∗ 2
ZrC
l 2e
16.7
(0.4
)0.
73(0
.03)
400(
400)
——
—R
ossi
niet
al.61
CpT
iCl 3
15.5
(0.4
)0.
54(0
.05)
500(
150)
750(
400)
−0.4
(0.5
)80
(30)
,5(
15),
5(30
)R
ossi
niet
al.61
Cp 2
ZrM
eCl
13.7
(0.4
)0.
75(0
.10)
400(
400)
——
—R
ossi
niet
al.61
(Cp 2
ZrC
l)2μ
-O16
.3(0
.4)
0.43
(0.0
7)30
0(40
0)—
——
Ros
sini
etal
.61
Cp∗
ZrC
l 3Si
te1:
12.8
(0.5
)0.
10(0
.10)
400(
200)
500(
400)
0.4(
0.8)
10(9
0),
15(3
0),
0(90
)dR
ossi
niet
al.61
Site
2:13
.3(0
.5)
0.12
(0.1
0)40
0(20
0)50
0(40
0)0.
4(0.
8)10
(90)
,15
(30)
,0(
90)d
—Si
te3:
14.6
(0.5
)0.
88(0
.10)
200(
200)
200(
200)
——
—Si
te4:
14.0
(0.5
)0.
80(0
.10)
200(
200)
200(
200)
——
—C
pZrC
l 3(m
ultip
lesi
tes)
14.8
–18
.60.
7–
0.8
300
——
—R
ossi
niet
al.61
Cp 2
ZrH
Cl
19.7
(0.3
)0.
20(0
.04)
80(5
0)—
——
Ros
sini
etal
.61
AlC
l 322
.5(1
.0)
0.63
(0.1
0)28
4(10
0)30
0(20
0)−0
.5(0
.5)
90(2
0),
90(2
0),
0(20
)C
hapm
anan
dB
ryce
60
InC
l 324
.5(1
.0)
0.52
(0.1
0)33
4(10
0)50
0(20
0)0.
5(0.
5)20
(20)
,90
(20)
,30
(20)
Cha
pman
and
Bry
ce60
GaC
l 231
.2(0
.7)
0.15
(0.1
5)15
9(10
0)20
0(20
0)−0
.5(0
.5)
90(2
0),
90(2
0),
0(20
)C
hapm
anan
dB
ryce
60
32.0
(0.7
)0.
2(0.
2)15
9(10
0)20
0(20
0)−0
.5(0
.5)
90(2
0),
90(2
0),
20(2
0)—
GaC
l 340
.4(2
.0)
0.03
(0.0
3)15
9(10
0)—
——
Cha
pman
and
Bry
ce60
38.1
(2.0
)0.
09(0
.05)
109(
100)
——
——
28.3
(2.0
)0.
48(0
.05)
209(
100)
——
——
aW
ithre
spec
tto
0.1
mol
dm−3
NaC
lin
D2O
.bE
uler
angl
esde
term
ined
usin
gth
eA
rfke
nco
nven
tion
unle
ssot
herw
ise
note
d.cC
p=
cycl
open
tadi
enyl
.dR
ose
conv
entio
nus
edfo
rE
uler
angl
es.
eC
p∗=
pent
amet
hylc
yclo
pent
adie
nyl.
334 Applications
Simulationexperiment
∗
30000
1000 500 0 −500 −1000 −1500
20000 10000 0 −10000 −20000 −30000 −40000 ppm
kHz
Cp2TiCl2
Cp2ZrCl2
Cp2HfCl2
Cp*2ZrCl2
CpTiCl3
Figure 19.7. 35Cl QCPMG SSNMR spectra and analyti-cal simulations of the spectra (solid traces) for Cp2TiCl2,Cp2ZrCl2, Cp2HfCl2, Cp∗
2ZrCl2, and CpTiCl3 (B0 = 9.4T). See Table 19.4 for parameters. Satellite transitionsare visible in the spectra of Cp2ZrCl2, Cp2
∗ZrCl2, andCp2HfCl2. The asterisk in the spectrum of Cp∗
2ZrCl2 de-notes a discontinuity of a satellite transition. (Reproducedfrom Ref. 61. © American Chemical Society, 2009.)
CQ(35Cl) magnitudes observed are notably higherthan those observed for the alkaline earth metalchlorides, ranging from 12.8(5) MHz for one of thefour sites in Cp∗ZrCl3 to 22.1(5) MHz in Cp2TiCl2.The authors were also able to extract CS tensor datafor three of the complexes. The CS tensor spansobserved were very large compared to others inthe literature with the largest reported value being800(500) ppm for Cp2ZrCl2.
The range of CQ(35Cl) magnitudes, and thereforechloride ion environments, which may be practicallyobserved with chlorine SSNMR, was expanded in astudy of four group 13 metal chlorides.60 Again, theVOCS–QCPMG technique and a very high magneticfield (21.1 T) were required to collect the full CTspectra of both 35Cl and 37Cl. The compounds, whichincluded the catalysts AlCl3 and GaCl3, were foundto have strong chlorine QIs, with CQ(35Cl) magni-tudes ranging from 22.5 MHz in AlCl3 to 40.4 MHz
for one site in GaCl3. The chlorine CSs range from109 to 334 ppm (with respect to the IUPAC stan-dard), which is consistent with those observed forother chloride-containing organometallics,61 and sig-nificantly higher than the values observed for organichydrochlorides.2,42 It was noted that the chlorine CSincreased as the M–Cl bond length increased in allcases.60 CS tensor data were also extracted for threeof the chlorides, with CS tensor spans ranging from200 to 500 ppm.
19.3.1.3 Perchlorates
Perchlorates (ClO4−) have been well studied by chlo-
rine SSNMR due to the nearly tetrahedral symmetryabout the chlorine site (Table 19.5).8,63 – 71 This ar-rangement results in a relatively small QI, and con-sequently small chlorine CQ values and CT spectralbreadths. In addition, the chlorine chemical environ-ment in these materials is interesting as large para-magnetic contributions to the shielding tensor leadto chlorine CSs that are much larger than those forchloride ions.
Jurga et al. have published several spectra of mul-timethylammonium perchlorates.63 Relatively small35Cl quadrupolar coupling constants ranging between0.238 and 1.12 MHz were extracted from spectra col-lected under MAS and stationary conditions. In ad-dition, one phase of dimethylammonium perchloratewas found to have a chlorine CS highly deshieldedat 1012 ppm (with respect to the IUPAC standard).Tarasov et al. extended the study of perchloratesthrough the collection of 35Cl SSNMR spectra ofthree alkali metal perchlorates, CsClO4, RbClO4, andKClO4, at 7.04 T.66,67 The 35Cl quadrupolar cou-pling constants observed were similar to those ob-served by Jurga, with magnitudes ranging from 0.51to ∼0.63 MHz. In 1999, a thorough study was pub-lished by Skibsted and Jakobsen concerning a largeseries of perchlorates (Figure 19.8).8 The authorslooked at 13 samples of anhydrous and hydratedperchlorates, including those examined by Tarasov.Chlorine-35 MAS, satellite transition spectroscopy(SATRAS), and echo experiments for all of the saltsat 14.1 T allowed for accurate determination of thechlorine NMR parameters. In select cases, 37Cl NMRspectra were also collected. Their results were con-sistent with earlier observations, and chlorine CSswere all greater than 988.5 ppm (with respect to theIUPAC standard), although the relative range of 35Clquadrupolar coupling constants was more substantial,
Chlorine, Bromine, and Iodine NMR 335
Table 19.5. Selected chlorine NMR data for solid perchlorates
Compounds |CQ(35Cl)|/MHz ηQ δisoa (ppm) References
(CH3)3NHClO4 —phase II 0.318 — — Jurga et al.63
(CH3)3NHClO4 —phase III 0.32 to 0.35 0.6 to ∼1 —(CH3)2NH2ClO4 —phase I ∼0 n/a 1012 Jurga et al.63
(CH3)2NH2ClO4 —phase II 0.238 0 —(CH3)2NH2ClO4 —phase III 1.12 0 —CH3NH3ClO4 —phase II 0.258 0 — Jurga et al.63
CH3NH3ClO4 —phase III 0.932 0.75 —NaClO4 0.887(0.014) 0.92(0.02) 1003.2(0.5) Skibsted and Jakobsen8
NaClO4·H2O 0.566(0.009) 0.90(0.02) 998.8(0.3) Skibsted and Jakobsen8
LiClO4 1.282(0.008) 0.34(0.01) 993.1(0.5) Skibsted and Jakobsen8
LiClO4·3H2O 0.695(0.004) 0.00(0.03) 1004.8(0.5) Skibsted and Jakobsen8
KClO4 0.51(0.01) (at 296 K) 0.52(0.10) — Tarasov et al.67
0.440(0.006) 0.88(0.02) 1008.1(0.3) Skibsted and Jakobsen8
RbClO4 Temperaturedependence of CQ
monitored
0.53(0.02) −3(5)b Tarasov et al.67
— 0.537(0.015) 0.87(0.03) 1008.3(0.3) Skibsted and Jakobsen8
CsClO4 Temperaturedependence of CQ
and ηQ determined
0.55 — Tarasov et al.66
— 0.585(0.008) 0.86(0.02) 1006.6(0.3) Skibsted and Jakobsen8
Mg(ClO4)2 2.981(0.007) 0.57(0.01) 995.1(0.5) Skibsted and Jakobsen8
Mg(ClO4)2·6H2O—site 1 0.309(0.006) 0.00(0.08) 1005.5(0.3) Skibsted and Jakobsen8
Mg(ClO4)2·6H2O—site 2 0.475(0.008) 0.00(0.05) 1004.4(0.3)Ba(ClO4)2 2.256(0.008) 0.58(0.01) 988.5(0.5) Skibsted and Jakobsen8
Ba(ClO4)2·3H2O 0.383(0.005) 0.00(0.03) 999.5(0.3) Skibsted and Jakobsen8
Cd(ClO4)2·6H2O 0.328(0.005) 0.00(0.03) 1003.3(0.3) Skibsted and Jakobsen8
(CH3)4NClO4 0.307(0.004) 0.00(0.03) 1008.2(0.3) Skibsted and Jakobsen8
aWith respect to 0.1 mol dm−3 NaCl in D2O, unless otherwise noted.bWith respect to 0.1 mol dm−3 RbClO4(aq).
at 0.307 to 2.981 MHz. Skibsted and Jakobsen im-proved the accuracy of the earlier data because of theinclusion of CSA in their analyses and the analysisof the spinning sideband manifolds for the STs.8
19.3.1.4 Other Systems
In addition to the classes of compounds discussedabove, other materials have been analyzed usingchlorine SSNMR (Table 19.6). For example, severalchlorine-containing glasses have been studied with35Cl MAS SSNMR. In a 2004 report, a series ofsilicate and aluminosilicate glasses were studied anda correlation was noted between the M–Cl bonddistance and the chlorine CS.72 The value of CQ(35Cl)in these glasses ranged from 2.9 to 4.4 MHz.72
A new class of materials was analyzed by Gordonet al. in 2008 when they used chlorine SSNMRto examine four ionic liquids, which are solid atroom temperature.73,74 The magnitudes of CQ(35Cl)extracted were quite small, ranging from 0.805 to1.500 MHz, and demonstrated the small chlorine QIin these salts. The CSs observed ranged from 60.6to 91.7 ppm (with respect to the IUPAC standard).In addition, it was shown that the four crystal-lographic chlorine sites in ethylmethylimidazoliumchloride could be resolved in a 35Cl MAS spectrumcollected at 21.1 T.
Covalently bound chlorine in organic moieties ischaracterized by CQ(35Cl) magnitudes on the orderof 60–80 MHz.14,15 Acquiring the 35/37Cl SSNMRspectrum of covalently bound chlorine in a powdersample is technically feasible in a high magnetic
336 Applications
1040 1038(ppm)
1040 1038(ppm)
200 150 100 50 0
(kHz)
−50 −100 −150 −200
(a)
(b)
(c)
(d)
Figure 19.8. 35Cl MAS NMR spectra (νr = 7.0 kHz) ofthe satellite transitions for (a) Ba(ClO4)2·3H2O and (c)Cd(ClO4)2·6H2O shown with the central transition cutoffat ∼1/10 of its total height. The inset in panel (a) illustratesthe lineshape for the central transition for Ba(ClO4)2·3H2O.Simulated spectra of the spinning sideband manifolds fromthe satellite transitions are shown in panels (b) and (d)for Ba(ClO4)2·3H2O and Cd(ClO4)2·6H2O, respectively,and employ the 35Cl NMR parameters in Table 19.5. Asimulation of the central transition for Ba(ClO4)2·3H2O isshown as the inset in panel (b). (Reproduced from Ref. 8.© American Chemical Society, 1999.)
field (≥18.8 T), but is also quite time consuming andtypically impractical.75
19.3.2 Bromine
Relative to 35Cl and 37Cl SSNMR spectroscopy, sig-nificantly fewer 79Br and 81Br SSNMR data are avail-able, although a number of similar systems have beenstudied.1,2 Typical 81Br CQ values for noncubic ionicbromide sites are on the order of 10–20 MHz, whichrepresents a static line width of about 100–500 kHzat 11.75 T. As such, while 79/81Br SSNMR exper-iments are generally applicable for the study ofbromide-containing materials, the VOCS method willoften be required for the acquisition of complete CT79/81Br SSNMR signals. Although the experimentalobservation of bromine CSA is rare, typical values ofthe CS tensor span for bromide-containing materialsare on the order of 100–200 ppm.1,2,21,74,76
19.3.2.1 Organic Hydrobromides
A few organic hydrobromide systems have been stud-ied using 81Br SSNMR spectroscopy, primarily usingsingle crystals; however, the effects of bromine CSAwere neglected in all cases.77,78 The EFG tensordata pertaining to these systems are summarized inTable 19.7. Although the data are somewhat sparse,it is observed that CQ(81Br) values for these sys-tems range from 11.26 to 49.0 MHz, and hence futurestudies on organic hydrobromide systems should bepossible using standard spectrometer systems. Owingto the low symmetry present at the bromide anionsin the systems studied to date, ηQ values are seento strongly deviate from zero (ranging from 0.59 to0.86). Bromine-81 SSNMR experiments using a sin-gle crystal of deuterated glycyl-l-alanine HBr·H2Oestablished that the V33 component of the bromineEFG tensor oriented approximately along the shortestH–Br hydrogen bond,78 which highlights the sensi-tivity of 79/81Br SSNMR experiments to weak inter-molecular interactions in solids.
19.3.2.2 Inorganic Bromides
The alkali metal bromides, of the general formMBr (M = Li, Na, K, Rb, Cs), were the first com-pounds to be studied using solid-state 79/81Br NMR
Chlorine, Bromine, and Iodine NMR 337
Table 19.6. Selected chlorine NMR data for other chlorine systems
Compounds |CQ(35Cl)|/MHz ηQ δisoa (ppm) References
Cl-containing silicate and aluminosilicate glasses 2.9(0.2) to 4.4(0.4) 0.7 −85(11) to 107(22) Sandland et al.72
Butyldimethylimidazolium chloride 0.978(0.004) 0.10(0.02) 71.60(0.02) Gordon et al.73,74
Butylmethylimidazolium chloride 1.500(0.002) 0.390(0.005) 71.65(0.05) Gordon et al.73,74
Ethylmethylimidazolium chloride Site 1: 0.808(0.004) 0.95(0.01) 91.72(0.03) Gordon et al.73,74
Site 2: 0.805(0.005) 0.20(0.02) 74.98(0.04)Site 3: 0.884(0.004) 0.86(0.01) 71.36(0.04)Site 4: 0.972(0.005) 0.80(0.01) 60.60(0.03)
Butylmethylpyridinium chloride Site 1: 0.857(0.008) 0.525(0.005) 83.15(0.07) Gordon et al.73,74
Site 2: 0.889(0.008) 0.08(0.08) 70.56(0.04)
aWith respect to 0.1 mol dm−3 NaCl in D2O.
Table 19.7. Selected bromine NMR data for solid or-ganic hydrobromides
Compounds |CQ|/MHz ηQ References
Deuteratedglycyl-l-alanineHBr·H2O
19.750a 0.8328 Kehrer et al.78
l-Leucine HBr 49.0b 0.59 Persons andHarbison77
l-Tyrosine HBr 11.26b 0.86 Persons andHarbison77
aFrom 81Br SSNMR data.bFrom 79Br SSNMR data.
spectroscopy79 owing to the essentially cubic envi-ronment at the bromide anions, which greatly reducesthe QI relative to noncubic environments.2,54,55 Thelack of a significant QI makes the alkali metal bro-mides ideal compounds with which to set up furtherexperiments on either NMR-active bromine nucleus.The CS values of this series relative to 0.03 mol dm−3
NaBr in D2O have been precisely determined underMAS conditions, and are reported in Table 19.8.1
Additional simple inorganic bromide salts (NH4Br,AgBr, CuBr, and TlBr) have been studied using79/81Br SSNMR, including three crystalline phasesof NH4Br, where the bromine CS and spin-lattice re-laxation time were both found to be sensitive to thephase of this material.80,81
Many alkaline earth metal bromides and selectedhydrates have also been studied using 79/81BrSSNMR spectroscopy (see Table 19.9 for EFG andCS tensor parameters).16,21 Owing to the extremesensitivity of the 79/81Br nuclei to the EFG, theseexperiments, when combined with the GIPAW–DFTcomputational method and ultrasoft core electron
Table 19.8. Selected bromine NMR data for solid inor-ganic bromides—cubic systems
Compounds δiso (ppm) References
LiBr 119.33(0.15)a Widdifield et al.1
NaBr 1.57(0.09)a Widdifield et al.1
KBr 54.51a Widdifield et al.1
RbBr 126.13(0.11)a Widdifield et al.1
CsBr 282.76(0.25)a Widdifield et al.1
AgBr 223.66(0.07)b Hayashi and Hayamizu56
CuBr −79.83(0.20)b Hayashi and Hayamizu56
aFrom 81Br SSNMR data with respect to 0.03 mol dm−3
NaBr in D2O.bFrom 79Br SSNMR data with respect to 0.03 mol dm−3
NaBr in D2O.
pseudopotentials, were used to propose a modifiedstructure for MgBr2.16 None of the anhydrousMBr2 series compounds (M = Mg, Ca, Sr, Ba)studied to date possess cubic local symmetry, whichleads to the observation of relatively large bromineQIs (CQ(81Br) ranges from 7.32 to 62.8 MHz).Bromine-79/81 SSNMR experiments were shownto be sensitive to the hydration level of thealkaline earth metal bromides through decreasesin the bromine CQ and δiso values upon samplehydration.21 A rare example of 79/81Br MASNMR data for a noncubic sample was presentedfor BaBr2·2H2O, which possessed unusually lowCQ(79/81Br) values. While the QI is expected todominate the line broadening in the alkaline earthmetal bromides, even in high magnetic fields, theeffects of bromine CSA in the observed SSNMRspectra were quantified, with Ω values ranging from50 to 250 ppm (Figure 19.9).21
338 Applications
Tabl
e19
.9.
Sele
cted
brom
ine
NM
Rda
tafo
rso
lidal
kalin
eea
rth
met
albr
omid
es
Com
poun
ds|C
Q(81
Br)
|/MH
zη
Qδ i
so(p
pm)a
Ω(p
pm)
κE
uler
angl
esb/d
egre
esR
efer
ence
s
CaB
r 262
.8(0
.4)
0.44
5(0.
02)
335(
50)
250(
150)
0c27
0c,
90(2
0),
180c
Wid
difie
ldan
dB
ryce
21
tris
-Sar
cosi
neC
aBr 2
21.9
0.64
——
——
Erg
eet
al.82
MgB
r 221
.93(
0.20
)0.
02(0
.02)
340(
10)
——
—W
iddi
field
and
Bry
ce16
MgB
r 2·6H
2O
19.0
(0.2
)0.
23(0
.03)
112(
7)50
(20)
0.7(
0.3)
170(
10),
57(1
0),
180c
Wid
difie
ldan
dB
ryce
21
SrB
r 2—
site
110
.3(0
.3)
0.07
(0.0
4)47
7(5)
50(2
0)−1
d90
c,
90(1
5),
180(
5)W
iddi
field
and
Bry
ce21
SrB
r 2—
site
218
.1(0
.2)
0.03
(0.0
2)46
5(8)
85(2
5)−1
d90
c,
90(1
0),
180(
8)W
iddi
field
and
Bry
ce21
SrB
r 2—
site
325
.6(0
.2)
0.69
5(0.
015)
375(
10)
110(
30)
0.3(
0.4)
42(8
),90
(10)
,23
5(20
)W
iddi
field
and
Bry
ce21
SrB
r 2—
site
453
.7(0
.6)
0.33
(0.0
2)35
5(50
)—
——
Wid
difie
ldan
dB
ryce
21
SrB
r 2·6H
2O
27.7
(0.3
)<
0.01
150(
15)
70(3
0)−1
d21
0c,
90(2
0),
180(
10)
Wid
difie
ldan
dB
ryce
21
BaB
r 2—
site
123
.5(0
.3)
0.17
(0.0
2)33
5(10
)20
0(20
)−0
.6(0
.2)
0c,
47(7
),18
0cW
iddi
field
and
Bry
ce21
BaB
r 2—
site
227
.2(0
.3)
0.07
0(0.
015)
535(
15)
170(
30)
0.1(
0.2)
180c
,18
(7),
180c
Wid
difie
ldan
dB
ryce
21
BaB
r 2·2H
2O
7.32
(0.0
3)0.
76(0
.02)
272.
7(1.
0)86
(5)
−0.2
0(0.
15)
70(5
),95
(8),
253(
5)W
iddi
field
and
Bry
ce21
aFr
om81
Br
SSN
MR
data
,w
ithre
spec
tto
0.03
mol
dm−3
NaB
rin
D2O
.bG
iven
asα
,β
,an
dγ
acco
rdin
gto
the
“ZY
Z”
conv
entio
n.(S
eeR
ef.
44.)
cSi
mul
ated
SSN
MR
lines
hape
isno
tse
nsiti
veto
vari
atio
nin
this
para
met
er.
dA
ssum
edon
the
basi
sof
crys
tallo
grap
hic
site
sym
met
ry.
Chlorine, Bromine, and Iodine NMR 339
100 100
100 50 0 −50
80 60 40 20 0 −20
400
500 400 300 200 100 0 −100
100 80 60 40 20 0
300 200 100 600
1000 500 −5000
400 200 0 −200d /ppm d /ppm
d /ppm d /ppm
80 60 40 20 n (81Br) /kHz n (81Br) /kHz
n (79Br) /kHzn (79Br) /kHz
(a)
(b)(c)
(f)(g)
(d)
(e)
(h)
(i)
Figure 19.9. Analytical simulations incorporating quadrupolar and CSA effects (a, d, f, h) and experimental static solidecho (b, c, e, g, i) 79/81Br{1H} SSNMR spectra of powdered BaBr2·2H2O, acquired at B0 = 21.1 T (b, c, e) and 11.75 T (g,i). See Table 19.9 for parameters. In spectrum (c), 1H decoupling is not applied. (Reproduced from Ref. 21. © AmericanChemical Society, 2010.)
19.3.2.3 Other Systems
Solid-state bromine (primarily 81Br) NMR experi-ments have also been carried out on a variety of alkalimetal perbromates,9,11,12 n-alkyltrimethylammoniumbromides,76 ionic liquids,74 and mesoporousmaterials.83 Selected bromine EFG and CS tensordata for these systems are located in Table 19.10.Owing to the high symmetry present at the brominesites in the alkali metal perbromates, very smallCQ(81Br) values were measured, ranging from1.32 MHz in CsBrO4 to 3.35 MHz in KBrO4.Although these experiments were carried out understatic conditions, bromine CSA was neglected duringthe spectral modeling. These systems could thus beof great use as setup samples, if one wished to havean appreciable second-order quadrupolar lineshape(which would not be present in the alkali metalhalides).
A series of n-alkyltrimethylammonium bromide(n = 1, 12, 14, 16, 18) systems have been studiedusing 81Br SSNMR under both static and MAS con-ditions at high B0.76 As with many of the compounds
studied above, the 81Br nuclei were found to besensitive probes of minute structural variations. Forexample, the bromine-81 δiso, CQ, and ηQ parameterswere found to possess weak dependencies upon thealkyl chain length for the n> 1 series. The variationin the CQ(81Br) value was attributed to an increasein the Br–N distance as the alkyl chain lengthincreased. The 81Br SSNMR data were acquired atmultiple MAS frequencies, followed by a numericaldata fitting procedure, which allowed for theextraction of bromine CS tensor span values around110 ppm (Figure 19.10). The sensitivity of 81BrSSNMR measurements to molecular dynamicshas also been observed using a single crystal oftris-sarcosine CaBr2.82,85
A series of four bromide-containing ionic liquidswere subjected to 79Br SSNMR experiments at bothstandard and very high magnetic fields.74 Owing tothe rather isolated nature of the bromides in thesesystems, CQ(79Br) values were found to lie withina relatively narrow range of 5.12–17.50 MHz, whilethe CSs varied between 122 and 172 ppm relativeto 0.01 mol dm−3 NaBr in D2O. Bromine CSA and
340 Applications
300 200 10081Br chemical shift /ppm
0 −100 −200 −300
(a)
(b)
(c)
(d)
Figure 19.10. Modeling of 81Br spectra of hexade-cyltrimethylammonium bromide: (a) νMAS = 30 kHz, (b)νMAS = 14 kHz, (c) νMAS = 7 kHz, and (d) νMAS = 0 kHz.In each of the four parts, the experimental spectrum is shownat the top, the model in the middle, and the difference spec-trum at the bottom. (Reproduced from Ref. 76. © AmericanChemical Society, 2009.)
CS/EFG tensor interplay were also observed in allsamples (Table 19.10).
An account by Alonso et al. noted the possibilityof carrying out 81Br SSNMR experiments on meso-porous materials;83 however, it appears as thoughMAS experiments were not fruitful for the systemsthat were considered. The authors attribute the diffi-culty in direct observation to a spatially nonuniformcharge distribution (and hence a distribution in theQI parameters) when the probe molecule is withinthe mesoporous material. At the same time, it wasnoted that information pertaining to the 81Br en-vironment could potentially be gained via indirectmethods, such as 1H{81Br} TRAPDOR experiments.Indeed, it was observed using both 81Br and 14NSSNMR data that the bromide anions were distributedclose to the alkyltrimethylammonium head group, then-alkyl chains, and the phenyl groups on the siloxanesurface.
Covalently bound bromine in organic systems istypically characterized by CQ(81Br) magnitudes onthe order of 400–600 MHz.15 Acquiring the 81BrSSNMR spectrum of covalently bound bromine in apowder sample is impractical with currently availablemagnetic field strengths.
19.3.3 Iodine
Owing to the relatively large quadrupole moment forthe iodine-127 nucleus, much of the 127I SSNMRdata reported in the current literature involves eitherionic iodides or periodates.1,2,13,17 Although thesesystems are chosen to minimize the resulting QI, typ-ical CQ(127I) values for ionic iodides and periodatesare still large (up to 214 MHz, which translates into aCT line width of about 10 MHz at 21.1 T). Hence,the VOCS method will often be used for the ac-quisition of 127I SSNMR signals. The experimentalobservation of iodine CSA is very rare: there existsonly a handful of reliable measurements.13,17 Thisis due to the generally small impact of the CSAon the second-order quadrupolar-dominated lineshape(in systems observed to date) as well as the techni-cal difficulties associated with acquiring broad 127ISSNMR spectra.
19.3.3.1 Inorganic Iodides
Owing to the cubic lattice symmetry present for thealkali metal iodides (general form MI, M = Li, Na,
Chlorine, Bromine, and Iodine NMR 341
Tabl
e19
.10.
Sele
cted
brom
ine
NM
Rda
tafo
rot
her
solid
brom
ine-
cont
aini
ngsy
stem
s
Com
poun
ds|C
Q|/M
Hz
ηQ
δ iso
(ppm
)Ω
(ppm
)κ
Eul
eran
gles
/deg
rees
Ref
eren
ces
KB
rO4
3.35
(0.0
3)a
0.71
(0.0
5)—
——
—Ta
raso
vet
al.11
RbB
rO4
2.36
(0.0
4)a
0.37
——
——
Tara
sov
etal
.11
CsB
rO4
1.32
(0.0
4)a
024
00(2
00)b
——
—Ta
raso
vet
al.9
NH
4B
rO4
2.27
(0.0
5)a
0.99
——
——
Tara
sov
etal
.12
(CH
3) 4
NB
r6.
03(0
.002
)a0.
02(0
.01)
100.
2(0.
4)c
33(3
)0.
6(0.
3)9(
9),
5(6)
,9(
2)d
Alo
nso
etal
.76
(C12
H25
)(C
H3) 3
NB
r7.
39(0
.10)
a0.
11(0
.02)
113.
1(0.
9)c
117(
3)−0
.3(0
.3)
39(9
),68
(2),
50(2
8)d
Alo
nso
etal
.76
(C14
H29
)(C
H3) 3
NB
r7.
74(0
.17)
a0.
16(0
.03)
111.
2(0.
8)c
112(
4)−0
.5(0
.1)
26(7
),11
5(2)
,47
(6)d
Alo
nso
etal
.76
(C16
H33
)(C
H3) 3
NB
r8.
03(0
.03)
a0.
18(0
.02)
110.
4(0.
8)c
106(
6)−0
.5(0
.2)
23(2
0),
116(
1),
59(2
4)d
Alo
nso
etal
.76
(C18
H37
)(C
H3) 3
NB
r8.
08(0
.07)
a0.
19(0
.01)
108.
7(0.
5)c
105(
8)−0
.3(0
.1)
29(6
),11
6(2)
,63
(21)
dA
lons
oet
al.76
(C4H
9)N
H3B
r17
.50(
0.02
)e0.
01(0
.01)
137.
0(0.
5)f
75(2
)0.
05(0
.05)
0,0,
0,g
Gor
don
etal
.74
1-E
thyl
-3-m
ethy
limid
azol
ium
brom
ide
12.4
0(0.
01)e
0.28
(0.0
1)12
2(1)
f73
(3)
0.95
(0.0
5)54
(3),
81(1
),8(
2)g
Gor
don
etal
.74
1-E
thyl
-1-m
ethy
lpyr
rolid
iniu
mbr
omid
e5.
12(0
.05)
e0.
37(0
.08)
123(
1)f
51(1
)0.
96(0
.04)
97(2
),21
(2),
148(
2)g
Gor
don
etal
.74
aFr
om81
Br
SSN
MR
data
.bW
ithre
spec
tto
1m
oldm
−3K
Br.
cW
ithre
spec
tto
0.03
mol
dm−3
NaB
rin
D2O
.dG
iven
asφ
,χ
,an
dψ
acco
rdin
gto
Ref
.84
.eFr
om79
Br
SSN
MR
data
.f
With
resp
ect
to0.
01m
oldm
−3N
aBr
inD
2O
.gG
iven
asα
,β
,an
dγ
acco
rdin
gto
the
“ZY
Z”
conv
entio
n.(S
eeR
ef.
44.)
342 Applications
Table 19.11. Selected 127I NMR data for solid inorganiciodides—cubic systems
Compounds δiso (ppm)a References
LiI 408.66 Chapman et al.2
NaI 226.71 Chapman et al.2
KI 192.62 Chapman et al.2
RbI 269.87 Chapman et al.2
CsI 562.41 Chapman et al.2
AgI −37.6(1.0) Hayashi and Hayamizu56
CuI 200.3(0.2) Hayashi and Hayamizu56
aWith respect to 0.01 mol dm−3 KI in D2O.
K, Rb, Cs), it should not be surprising that thesesystems were the first to be characterized using 127ISSNMR, and that these compounds currently serveas useful reference materials.2,54,55 In addition to thealkali metal iodides, additional cubic iodides, such asCuI and AgI have been studied.56,58,86 Both CuI andAgI possess iodine CSs that are highly sensitive totemperature (on the order of 0.5 ppm/K). As the res-olution of 127I MAS SSNMR experiments for thesesystems typically results in CSs with errors on the or-der of 1 ppm, the iodine-127 SSNMR signals of thesematerials could be used to measure temperatures. Theiodine CS values for these cubic systems, relative to0.01 mol dm−3 KI in D2O, have been precisely de-termined, and are reported in Table 19.11.
The alkaline earth metal iodides, selected hy-drates thereof, and the semiconducting material CdI2have also been characterized using multiple-field 127I
SSNMR data acquisition and GIPAW–DFT compu-tations (experimental data are in Table 19.12; see alsoFigure 19.11).17 The 127I nuclei were found to be use-ful probes of the hydration state (a decreasing iodineCS results upon sample hydration). Reliable measure-ments of iodine CSA were able to establish a rangeof 127I CS tensor span values from 60 to 300 ppm,although owing to the very large line widths (onthe order of MHz), very high field data acquisitionis a requirement to extract these modest Ω values.Interestingly, high-order (i.e., beyond second-order)quadrupole-induced effects were observed in the CTSSNMR signal when analyzing the spectrum of oneof the two iodine sites in SrI2 (cf. Figure 19.5).This was confirmed using both exact lineshape sim-ulation software,28 and 127I NQR experiments. Thehigh-order effects were such that they lead to an un-derestimation in both (i) CQ(127I) and (ii) δiso whenthe spectra were modeled using software that includesQI effects to only second order.
19.3.3.2 Periodates
In order to perform SSNMR experiments using 127Inuclei that are not present as an iodide anion, veryhigh local site symmetry is required. As mentionedin Section 19.1, this is because of the combina-tion of a large quadrupole moment and a signifi-cant Sternheimer antishielding factor. Favorable con-ditions for observing 127I SSNMR spectra exist ina variety of periodates. Compared to the alkaline
Table 19.12. Selected 127I NMR data for solid inorganic iodides—noncubic systems
Compounds |CQ(127I)|/MHz ηQ δiso (ppm)a Ω (ppm) κ Euler angles/degrees
MgI2 79.8(0.5) 0.02(0.02) 920(50) 120(80) −1b 90c, 90(20), 0c
CaI2 43.5(0.3) 0.02(0.02) 755(10) <50 — —SrI2 —site 1 105.2(0.7) 0.467(0.012) 880(70) — — —SrI2 —site 2 214.0(0.1)d 0.316(0.002)d 720(150)e — — —SrI2·6H2O 133.6(0.1)d <0.01 440(25)e — — —BaI2 —site 1 96.2(0.8) 0.175(0.015) 650(70)e 300(100) <–0.5 0c, 45(20), 180c
BaI2 —site 2 120.9(0.2)d 0.015(0.015)d 1000(80)e — — —BaI2·2H2O 53.8(0.3) 0.53(0.01) 630(20) 60(15) >0.5 45(15), 45(15), —CdI2 (4H polytype) ∼96.6(1.0) 0b ∼1450(100) — — —
.Source: From Widdifield and Bryce.17
aWith respect to 0.01 mol dm−3 KI in D2O.bAssumed on the basis of crystallographic site symmetry.cSimulated SSNMR lineshape is not sensitive to variation in this parameter.dEstablished using both 127I SSNMR and 127I NQR experimental data.eEstablished with the aid of exact simulation software.
Chlorine, Bromine, and Iodine NMR 343
1000
4000
1000
10000 5000 −5000 −100000
500 −500 −10000
2000 0 −2000 −4000 d /ppm
d /ppm
500 0 −500 Δn0 /kHz
Δn0 /kHz
(a)
(c)
(d)
(b)
Figure 19.11. Analytical simulations (a and c) and exper-imental static VOCS Solomon echo (b and d) 127I SSNMRspectra of powdered MgI2, acquired at (b) B0 = 21.1 T and(d) B0 = 11.75 T. Partially excited STs are denoted with“‡”. See Table 19.12 for parameters. (Reproduced from Ref.17. © American Chemical Society, 2010.)
earth metal iodide systems outlined above, the io-dine QIs present in the periodates are much smaller.This is analogous to what has been observed for thechlorine and bromine analogs (vide supra). ObservedCQ(127I) values range from only 1.00 MHz in CsIO4to 43.00 MHz in HIO4.13 A great deal of effort hasalso been expended to determine the temperature de-pendence of CQ(127I) for the periodates, with bothpositive and negative correlations being noted.87 – 93
Although these systems are known to possess verysmall relative QIs, the corresponding isotropic CSvalues are the highest known, due to the largely cova-lent O–I local bonding environment. CSs range fromabout 3500 to 4200 ppm with respect to 0.01 moldm−3 KI in D2O, which highlights the high sensi-tivity of the iodine-127 nucleus to small magnetic(in addition to electric) perturbations. The CS ten-sor spans for the alkali metal periodates appear to be
less than 50 ppm in all cases, with only one valueprecisely reported (Ω = 18(2) ppm for CsIO4).13
Additional instances are reported in the literature94;however, further experiments must be carried out tovalidate these observations. Information pertaining toselected periodate systems near room temperature issummarized in Table 19.13.
19.3.3.3 Other Systems
Few additional systems have been probed using 127ISSNMR, and selected relevant parameters can befound in Table 19.14. Two iodide ionic liquids havebeen characterized, and they possess rather similarCQ(127I) and CS values as the iodides noted above.74
To the best of our knowledge, glycyl-l-alanine hy-droiodide monohydrate appears to be the only hy-droiodide system studied using 127I SSNMR.78,95 Therelatively small CQ(127I) in this system (74.04 MHz)suggests that further studies on the amino acid hy-droiodide compound class may be worthwhile. Afew 127I SSNMR accounts of mixed halogen sys-tems, such as IF7 and [IF6]+[AsF6]−, have also beenreported.96,97 Variable-temperature 127I SSNMR ex-periments were found to be very sensitive toward thedetection of phase transitions and molecular motionin IF7.
Covalently bound iodine in organic systems istypically characterized by CQ(127I) magnitudes onthe order of 1700–2000 MHz.15 Acquiring the 127ISSNMR spectrum of covalently bound iodine in apowder sample is impractical with currently availablemagnetic field strengths.
19.4 CONCLUSIONS AND FUTUREPROSPECTS
Of the quadrupolar halogens, applications of 35/37ClSSNMR spectroscopy and the interpretation of theresults in terms of local structure are the most com-mon. Most studies have understandably focused onchloride ions in organic and inorganic environmentsbecause of the relatively small QI of the chlorine nu-cleus. The spectroscopies of 79/81Br and 127I are lessdeveloped owing to the larger quadrupole momentsand Sternheimer antishielding factors for these iso-topes. Nevertheless, recent work has demonstratedthat a combination of very high magnetic fields,modern signal acquisition or enhancement methods,
344 Applications
Table 19.13. Selected 127I NMR data for solid periodates
Compounds |CQ(127I)| ηQ δiso (ppm)a References/MHz
HIO4 43.00(0.01) 0.75 3527(10) Wu and Dong13
NH4IO4 10.00(0.01) 0.0 4187(10) Wu and Dong13
NaIO4 42.24(0.01) 0.0 4177(10) Wu and Dong13
KIO4 20.66(0.01) 0.0 4187(10) Wu and Dong13
RbIO4 15.65(0.01) 0.0 4187(10) Wu and Dong13
CsIO4 1.00(0.01)b 0.0 4199(2) Wu and Dong13
(CH3)4NIO4 15.7 0 — Klobasa and Burkert87
(C2H5)4NIO4 <1.5 — — Klobasa and Burkert88
(n-C4H9)4NIO4 3.66 0.67 — Burkert and Grommelt89
(CH3)4PIO4 <2.0 — — Klobasa and Burkert88
(C2H5)4PIO4 5.87(0.03) 0 — Klobasa et al.90
(C6H5)4PIO4 4.47(0.05) 0 — Burkert and Klobasa93
(CH3)4AsIO4 ≤1.8 — — Grommelt and Burkert91
(C2H5)4AsIO4 5.55(0.03) 0 — Klobasa et al.90
(C6H5)4AsIO4 4.53(0.05) 0 — Burkert and Klobasa93
(C2H5)4SbIO4 5.64(0.03) 0.41(0.02) — Klobasa and Burkert92
(C6H5)4SbIO4 2.63(0.05) 0 — Burkert and Klobasa93
aWith respect to 0.01 mol dm−3 KI in D2O.bFor this compound, Ω = 18(2) ppm and κ = 1.
Table 19.14. Selected 127I NMR data for other iodine-containing systems
Compounds |CQ(127I)|/MHz ηQ δiso (ppm) References
IF7 — — 3040(40)a Weulersse et al.96
[IF6]+[AsF6]− 2.3–2.9 — — Hon and Christe97
Glycyl-l-alanine HI·H2O 74.04 0.776 — Kehrer et al.78,95
1,2,3-Trimethylimidazolium iodide 36.8(0.9) 0.73(0.01) 143(18)b Gordon et al.74
1,1-Dimethylpyrrolidinium iodide 16.9(0.1) 0.27(0.02) 282.0(0.7)b Gordon et al.74
aWith respect to 5 mol dm−3 KI (aq).bWith respect to 0.01 mol dm−3 KI in D2O.
and proper data analysis render feasible the studyof chlorine, bromine, and iodine anions in noncubicinorganic and bioinorganic compounds. Quadrupo-lar and CS tensors are now available for halidesin a range of chemical environments, and the sen-sitivity of the tensor parameters to factors such aslocal site symmetry, hydration state, and polymor-phism has been demonstrated. It is likely that the35/37Cl isotopes will continue to be the most ac-cessible to study by SSNMR, and it is for chlorinethat the most comprehensive understanding of the re-lationship between local structure and the observedNMR spectrum is currently available. Acquisition of
35/37Cl SSNMR spectra of covalently bound chlorineis technically feasible in very high magnetic fields,but such experiments are most often impractical andtime consuming. Acquisition of 79/81Br or 127I SS-NMR spectra of covalent bromines or iodines is notpractical in currently commercially available mag-netic fields. Sustained advances and applications inthe area of quadrupolar halogen SSNMR are an-ticipated; however, such studies will likely con-tinue to be focused on chlorine, bromine, and iodinein environments in which they exist predominantlyin the ionic form. One interesting future directionis the study of halides that participate in halogenbonds.98
Chlorine, Bromine, and Iodine NMR 345
RELATED ARTICLES IN THEENCYCLOPEDIA OF MAGNETICRESONANCE
Chemical Shift TensorsFluorine-19 NMR;
Fluorine-19 NMR of Solids Containing Both Flu-orine and HydrogenQuadrupolar Nuclei in Liquid SamplesTensor Interplay
REFERENCES
1. C. M. Widdifield, R. P. Chapman, and D. L. Bryce,Annu. Rep. Nucl. Magn. Reson. Spectrosc., 2009, 66,195.
2. R. P. Chapman, C. M. Widdifield, and D. L. Bryce,Prog. Nucl. Magn. Reson. Spectrosc., 2009, 55, 215.
3. D. L. Bryce and G. D. Sward, Magn. Reson. Chem.,2006, 44, 409.
4. R. K. Harris, E. D. Becker, S. M. Cabral de Menezes,R. Goodfellow, and P. Granger, Pure Appl. Chem.,2001, 73, 1795.
5. K. J. D. MacKenzie and M. E. Smith, MultinuclearSolid-State NMR of Inorganic Materials , Pergamon:Amsterdam, 2002.
6. P. Pyykko, Mol. Phys., 2008, 106, 1965.
7. R. K. Harris and P. Jackson, Chem. Rev., 1991, 91,1427.
8. J. Skibsted and H. J. Jakobsen, Inorg. Chem., 1999,38, 1806.
9. V. P. Tarasov, S. A. Petrushin, and Y. K. Gusev, Dokl.Akad. Nauk SSSR, 1987, 293, 1423.
10. V. P. Tarasov, V. I. Privalov, Y. A. Buslaev, andU. Eichhoff, Z. Naturforsch. B , 1990, 45, 1005.
11. V. P. Tarasov, V. I. Privalov, K. S. Gavrichev, V. E.Gorbunov, Y. K. Gusev, and Y. A. Buslaev, Russ.J. Coord. Chem., 1990, 16, 854; [Translated from:Koordinats. Khim., 1990, 16, 1603.]
12. V. P. Tarasov, S. A. Petrushin, and Y. K. Gusev, Russ.J. Inorg. Chem., 1988, 33, 452. [Translated from: Zh.Neorg. Khim., 1988, 33, 804.]
13. G. Wu and S. Dong, Solid State Nucl. Magn. Reson.,2001, 20, 100.
14. E. A. C. Lucken, Nuclear Quadrupole Coupling Con-stants , Academic Press: London, 1969.
15. K. Eichele, R. E. Wasylishen, J. S. Grossert, andA. C. Olivieri, J. Phys. Chem., 1995, 99, 10110.
16. C. M. Widdifield and D. L. Bryce, Phys. Chem. Chem.Phys., 2009, 11, 7120.
17. C. M. Widdifield and D. L. Bryce, J. Phys. Chem. A,2010, 114, 10810.
18. D. L. Bryce and G. D. Sward, J. Phys. Chem. B , 2006,110, 26461.
19. H. Hamaed, J. M. Pawlowski, B. F. T. Cooper, R. Fu,S. H. Eichhorn, and R. W. Schurko, J. Am. Chem.Soc., 2008, 130, 11056.
20. D. L. Bryce and E. B. Bultz, Chem. Eur. J., 2007, 13,4786.
21. C. M. Widdifield and D. L. Bryce, J. Phys. Chem. A,2010, 114, 2102.
22. R. Sternheimer, Phys. Rev., 1950, 80, 102.
23. R. Sternheimer, Phys. Rev., 1951, 84, 244.
24. G. Wu and V. Terskikh, J. Phys. Chem. A, 2008, 112,10359.
25. W. P. Power, R. E. Wasylishen, S. Mooibroek, B. A.Pettitt, and W. Danchura, J. Phys. Chem., 1990, 94,591.
26. J. T. Cheng, J. C. Edwards, and P. D. Ellis, J. Phys.Chem., 1990, 94, 553.
27. P. J. Chu and B. C. Gerstein, J. Chem. Phys., 1989,91, 2081.
28. A. D. Bain, Mol. Phys., 2003, 101, 3163.
29. J. Blaser, O. Lutz, and W. Steinkilberg, Z. Natur-forsch. A, 1972, 27, 72.
30. J. Blaser, O. Lutz, and W. Steinkilberg, Phys. Lett. A,1970, 32, 403; [Errata: ibid., 1972, 38, 458.]
31. S. Hayashi and K. Hayamizu, Bull. Chem. Soc. Jpn.,1990, 63, 913.
32. M. Gee, R. E. Wasylishen, and A. Laaksonen, J. Phys.Chem. A, 1999, 103, 10805.
33. J. S. Frye and G. E. Maciel, J. Magn. Reson., 1982,48, 125.
34. I. Solomon, Phys. Rev., 1958, 110, 61.
35. E. L. Hahn, Phys. Rev., 1950, 80, 580.
36. P. R. Bodart, J. P. Amoureux, Y. Dumazy, andR. LeFort, Mol. Phys., 2000, 98, 1545.
37. F. H. Larsen, H. J. Jakobsen, P. D. Ellis, and N. C.Nielsen, J. Phys. Chem. A, 1997, 101, 8597.
38. J. P. Yesinowski, M. L. Buess, A. N. Garroway,M. Ziegeweld, and A. Pines, Anal. Chem., 1995, 67,2256.
346 Applications
39. D. L. Bryce, M. Gee, and R. E. Wasylishen, J. Phys.Chem. A, 2001, 105, 10413.
40. C. Gervais, R. Dupree, K. J. Pike, C. Bonhomme,M. Profeta, C. J. Pickard, and F. Mauri, J. Phys. Chem.A, 2005, 109, 6960.
41. D. L. Bryce, G. D. Sward, and S. Adiga, J. Am. Chem.Soc., 2006, 128, 2121.
42. R. P. Chapman and D. L. Bryce, Phys. Chem. Chem.Phys., 2007, 9, 6219.
43. R. P. Chapman, J. R. Hiscock, P. A. Gale, and D. L.Bryce, Can. J. Chem., 2011, 89, 822.
44. G. Arfken, Mathematical Methods for Physicists , 3rdedn., Academic Press, Orlando, Florida, 1985.
45. Z. Yao, H. T. Kwak, D. Sakellariou, L. Emsley, andP. J. Grandinetti, Chem. Phys. Lett., 2000, 327, 85.
46. A. P. M. Kentgens and R. Verhagen, Chem. Phys. Lett.,1999, 300, 435.
47. R. Siegel, T. T. Nakashima, and R. E. Wasylishen,Chem. Phys. Lett., 2004, 388, 441.
48. R. W. Schurko, I. Hung, and C. M. Widdifield, Chem.Phys. Lett., 2003, 379, 1.
49. R. Siegel, T. T. Nakashima, and R. E. Wasylishen,Concepts Magn. Reson. A, 2005, 26A, 47.
50. E. Kupce and R. Freeman, J. Magn. Reson. A, 1995,115, 273.
51. R. Bhattacharyya and L. Frydman, J. Chem. Phys.,2007, 127, 194503.
52. L. A. O’Dell and R. W. Schurko, Chem. Phys. Lett.,2008, 464, 97.
53. D. Massiot, I. Farnan, N. Gautier, D. Trumeau,A. Trokiner, and J. P. Coutures, Solid State Nucl.Magn. Reson., 1995, 4, 241.
54. Y. Yamagata, J. Phys. Soc. Jpn., 1964, 19, 10.
55. T. L. Weeding and W. S. Veeman, J. Chem. Soc. Chem.Commun., 1989, 946.
56. S. Hayashi and K. Hayamizu, J. Phys. Chem. Solids ,1992, 53, 239.
57. B. D. Guenther and R. A. Hultsch, J. Magn. Reson.,1969, 1, 609.
58. K. D. Becker, J. Chem. Phys., 1978, 68, 3785.
59. H. Honda, Z. Naturforsch. A, 2003, 58, 623.
60. R. P. Chapman and D. L. Bryce, Phys. Chem. Chem.Phys., 2009, 11, 6987.
61. A. J. Rossini, R. W. Mills, G. A. Briscoe, E. L.Norton, S. J. Geier, I. Hung, S. Zheng, J. Autschbach,
and R. W. Schurko, J. Am. Chem. Soc., 2009, 131,3317.
62. C. M. Widdifield and D. L. Bryce, Can. J. Chem.,2011, 89, 754.
63. S. Jurga, G. S. Harbison, B. Blumich, H. W. Spiess, F.Fujara, and A. Olinger, Ber. Bunsen-Ges. Phys. Chem.,1986, 90, 1153.
64. S. L. Segel, S. Maxwell, R. D. Heyding, P. Ingman,E. Ylinen, and M. Punkkinen, Solid State Commun.,1988, 66, 1039.
65. T. J. Bastow and S. N. Stuart, J. Phys.: Condens.Matter , 1989, 1, 4649.
66. V. P. Tarasov, M. A. Meladze, G. A. Kirakosyan,A. E. Shvelashvili, and Y. A. Buslaev, Phys. StatusSolidi B , 1991, 167, 271.
67. V. P. Tarasov, M. A. Meladze, and G. A. Kirakosyan,Russ. J. Coord. Chem., 1992, 18, 707.
68. N. Binesh and S. V. Bhat, Solid State Ionics , 1996,92, 261.
69. H. Ono, S. Ishimaru, R. Ikeda, and H. Ishida, Chem.Phys. Lett., 1997, 275, 485.
70. H. Ono, S. Ishimaru, R. Ikeda, and H. Ishida, Bull.Chem. Soc. Jpn., 1999, 72, 2049.
71. O. Czupinski, M. Wojtas, J. Zaleski, R. Jakubas, andW. Medycki, J. Phys.: Condens. Matter , 2006, 18,3307.
72. T. O. Sandland, L. S. Du, J. F. Stebbins, and J. D.Webster, Geochim. Cosmochim. Acta , 2004, 68, 5059.
73. P. G. Gordon, D. H. Brouwer, and J. A. Ripmeester,J. Phys. Chem. A, 2008, 112, 12527.
74. P. G. Gordon, D. H. Brouwer, and J. A. Ripmeester,ChemPhysChem , 2010, 11, 260.
75. V. V. Terskikh, S. J. Lang, P. G. Gordon, G. D.Enright, and J. A. Ripmeester, Magn. Reson. Chem.,2009, 47, 398.
76. B. Alonso, D. Massiot, P. Florian, H. H. Paradies,P. Gaveau, and T. Mineva, J. Phys. Chem. B , 2009,113, 11906.
77. J. Persons and G. S. Harbison, J. Magn. Reson., 2007,186, 347.
78. A. Kehrer, N. Weiden, and A. Weiss, Z. Phys. Chem.,1992, 178, 1.
79. R. V. Pound, Phys. Rev., 1950, 79, 685.
80. J. Itoh and Y. Yamagata, J. Phys. Soc. Jpn., 1962, 17,481.
Chlorine, Bromine, and Iodine NMR 347
81. J. Itoh and Y. Yamagata, J. Chem. Phys., 1956, 24,621.
82. T. Erge, D. Michel, J. Petersson, and F. Engelke, Phys.Status Solidi A, 1991, 123, 325.
83. B. Alonso, T. Mineva, P. Innocenzi, G. Trimmel, K.Stubenrauch, I. Melnyk, Y. Zub, F. Fayon, P. Florian,and D. Massiot, C. R. Chim., 2010, 13, 431.
84. S. Hayashi, Magn. Reson. Chem., 1996, 34, 791.
85. D. Michel, U. Hacker, T. Erge, and J. Petersson, Phys.Status Solidi B , 1994, 185, 257.
86. S. Hayashi and K. Hayamizu, J. Chem. Phys., 1990,92, 2818.
87. D. G. Klobasa and P. K. Burkert, Z. Naturforsch. A,1984, 39, 1222.
88. D. G. Klobasa and P. K. Burkert, Magn. Reson. Chem.,1987, 25, 154.
89. P. K. Burkert and M. Grommelt, Z. Naturforsch. B ,1988, 43, 1381.
90. D. G. Klobasa, P. K. Burkert, and G. Muller, Z.Naturforsch. A, 1986, 41, 330.
91. M. Grommelt and P. K. Burkert, Z. Naturforsch. B ,1989, 44, 1053.
92. D. G. Klobasa and P. K. Burkert, Z. Naturforsch. A,1987, 42, 275.
93. P. K. Burkert and D. G. Klobasa, Z. Naturforsch. A,1985, 40, 274.
94. S. L. Segel and H. M. Vyas, J. Chem. Phys., 1980, 72,1406.
95. A. Kehrer, S. Q. Dou, and A. Weiss, Z. Naturforsch.A, 1989, 44, 659.
96. J. M. Weulersse, J. Virlet, and P. Rigny, Mol. Phys.,1979, 38, 923.
97. J. F. Hon and K. O. Christe, J. Chem. Phys., 1970, 52,1960.
98. P. Metrangolo and G. Resnati, eds, Halogen Bonding:Fundamentals and Applications , Structure and Bond-ing, Springer, 2008, Vol. 126.