03 SC 377

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

P1: GAD

Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

Structural Chemistry, Vol. 14, No. 4, August 2003 (°


C 2003)

GIAO Calculations of Chemical Shifts


in Heterocyclic Compounds1

Ibon Alkorta2 and José Elguero2,3

Received April 17, 2002; revised June 7, 2002; accepted June 12, 2002

In this review, the GIAO calculations of absolute shieldings and their relationship with experimental
chemical shifts for aromatic heterocycles will be summarized. Automatic assignment, conformational
analysis, E/Z isomerism, and, in particular, tautomerism, will be discussed in detail. Solid-state and
solvent effects will be examined, as well as the problem of heteroaromaticity. The review ends with the
discussion of some methodological problems with special emphasis on the calculation of references,
such as TMS and nitromethane.

KEY WORDS: GIAO; DFT; Heteroaromatic compounds; tautomerism; NMR.

INTRODUCTION (gauge independent or invariant or including atomic or-


bital). We have searched in the Chemical Abstract (be-
If one excludes solid-state chemists, molecular and tween 1987 and 2001) for these three acronyms and we
supramolecular chemists alike rely mainly on NMR spec- have represented the result in Fig. 1 (we used Collective
troscopy for their research. This is not to diminish the Indexes 12CI, 1987–1991, and 13CI, 1992–1996).
importance of mass spectrometry, crystallography, mi- It appears that LORG has always been a minority
croscopy, IR and UV spectroscopies, etc., but NMR dom- option, while IGLO is less quoted. Although GIAO is
inates the panoply of analytical tools used today and will also used for other applications (circular dichroism, for in-
continue so for the next years. This is related both to the stance), most references deal with calculation of absolute
simplicity of its use and to the wealth of direct information shieldings or with Schleyer’s NICS (nuclear independent
it provides. chemical shifts) [1]. Part of its enormous success (474
At the beginning of NMR use, it was 1 H NMR, a spec- references in the explored period of time) is related to its
troscopy that provided at the same time chemical shifts and inclusion in the Gaussian package [2], but also to its effi-
1
H-1 H coupling constants. This is no longer the case with ciency, compatibility with the DFT approach [3–5] (often
many other nuclei, among them 13 C and 15 N, that in many in its B3LYP variety [6–9]), and excellent results. GIAO
cases are recorded in conditions that yield only chemical is based on London’s pioneering work [10] and developed
shifts (because they are less abundant). Most people now to its present state by Ditchfield [11].
rely more on chemical shifts than on coupling constants These methods afford the NMR shielding tensors,
if one excludes the famous Karplus relationship. From a σ11 , σ22 , and σ33 , from which the isotropic shieldings
theoretical point of view, this is fortunate because the cal- can be calculated [σiso = 1/3(σ11 + σ22 + σ33 )], as can the
culation of coupling constants is much more demanding anisotropic shieldings [σaniso = σ33 –1/2(σ22 + σ11 ), where
than the calculation of chemical shifts. σ33 > σ22 > σ11 ]. For most organic applications, σiso is the
Three are the most used methods to calculate chem- only valid information, since it is related to chemical shifts
ical shifts: IGLO (individual gauge localized orbital), by adding or substracting the reference.
LORG (localized or localorbital/local origin), and GIAO
1 Dedicated in memoriam, to Professor Lech Stefaniak.
2 Instituto
DISCUSSION
de Quı́mica Médica (C.S.I.C.), Juan de la Cierva, 3, E-28006
Madrid, Spain.
3 To whom all correspondence should be addressed; email: iqmbe17@ We will report the literature results concerning
iqm.csic.es GIAO calculations of heteroaromatic compounds in two

377
1040-0400/03/0800-0377/0 °
C 2003 Plenum Publishing Corporation
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

378 Alkorta and Elguero

Fig. 1. Plot of the number of references in function of time.

sections. The first one will deal with methodological and Because many different types of computational
systematic studies and the second one with specific ap- methods have been used in the references cited in
plications. The experimental data will be chemical shifts this review, we thought it useful to briefly summarize
(δ), i.e., relative to a given reference, in solution or in the them (the corresponding references can be found in
solid state. The calculated values correspond to different the original articles and many of them in the Gaussian
ab initio or DFT approximations, thus preventing them manual [2]).
from constituting a consistent set of data. In some cases, Calculation level:
the reference was also calculated; in other cases, only HF: Hartree–Fock
relative values are provided. In many cases, the signal of RHF: restricted Hartree–Fock
the reference is far away from the signals of the studied CHF: coupled Hartree–Fock
compounds, for instance, in 13 C NMR between TMS and SCF: self-consistent field
aromatic carbons, or in 15 N NMR between MeNO2 and DFT: density-functional theory
heterocyclic nitrogen atoms. B3LYP: Becke 3 Lee, Yang, and Parr
We should note that this review, concerned with B3PW91: Becke 3 Perdew–Wang
15
N NMR, is influenced mainly by the group from the MP2: Møller–Pleset 2
Polish Academy of Sciences (Michal Witanoswki, Jerzy CCSD(T)
W. Wiench, the late Lech Stefaniak, and their co-workers) Basis sets:
as well as by their usual collaborator Graham A. Webb STO-3G**
(University of Surrey). 3-21G
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

GIAO Calculations of Chemical Shifts in Heterocyclic Compounds 379

3-21G(d,p) [3-21G**]
6-31G
6-31G(d) [6-31G*]
6-31G(d,p) [6-31G**]
6-31+G(d) [6-31+G*]
6-31+G(d,p) [6-31+G**]
6-31++G(d,p) [6-31++G**]
6-311G
6-311G(d,p) [6-311G**]
6-311+G(d,p) [6-311+G**]
6-311+G(2d,p)
6-311+G(2df,2p)
6-311++G
6-311++G(d, p) [6-311++G**]
6-311++G(2df, p)
6-311++G(3df, 2p) Conformational Analysis and E/Z Isomerism
cc-aug-pVTZ
QZP The comparison of calculated values for two forms
We will discuss the performance of the different lev- (conformers, isomers, tautomers, different geometries in
els in Appendix I. the crystal, etc.) with the experimental results has been
used in a number of cases that will be discussed here and
in the two next following sections. Both isomers, the Z
(9a) and the E (9b), were observed in mesoionic thiatri-
Methodological and Systematic Studies azole 9 and 1 H, 13 C, and 15 N calculated absolute shield-
ings [21]. The 1 H and 13 C chemical shifts of the rotamers
The problem of transforming calculated absolute of furan-2-carboxaldehyde (furfural) (10) have been de-
shieldings σ into experimental chemical shifts δ and the termined at low temperature and satisfactorily compared
problem of the References [12] will be discussed in detail with SCF/6-31G** calculations [22]. By comparing the
at the end of this review. calculated 1 H chemical shifts of two conformations of
1-benzyloxypyrazole (11) with the experimental results,
Automatic Assignment the most stable one was ascertained [23]. Kleinpeter [24]
has calculated the chemical shifts of 2-benzo[b]furyl-
It is an old dream of spectroscopists to automatically quinoxalines (12) as a function of the dihedral angle link-
assign the bands or signals. The use of GIAO calculations ing both heterocycles to determine the conformation of
for this purpose is possible, although not very interest- these compounds. Ferraro has carried out an exemplar
ing, due to the efficiency of the experimental techniques. conformational study of dipyridylureas 13 by calculat-
Moreover, the most interesting problems are those related ing their isotropic 1 H and 13 C nuclear magnetic shielding
to close signals where the crossing of calculated values constants [25].
is not unexpected. For instance, CHF/6-311G** calcula-
tions have been said to be sufficiently accurate for use
in assigning the 13 C NMR spectra of solid amino acids
(including histidine 1) [13]. Other authors, using other
basis sets and different nuclei, have reached the same con-
clusion for tetrafluoroindazole 2 [14], 2-mercaptopyridine
(3) and related compounds [15], and the quaternary salts
derived from tetrazolo[1, 5-a]pyridine (4) [16]. Mikenda
et al.[17] have discussed the utility of MP2-based cal-
culations to assign 1 H, 13 C, and 15 N chemical shifts in
nitrogen-containing compounds, like purine (5); Wawer
et al. [18] have done the same for quinoacridinium salts 6,
Stefaniak et al. [19] for benzotriazoles 7, and others [20]
for 1,2,4-diazaphospholes 8.
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

380 Alkorta and Elguero

Tautomerism

This is one of the most studied cases, probably be-


cause tautomers differ much more than isomers or con-
formers in their NMR properties, thus making inter-
polation more reliable. No less than 27 references de-
scribe tautomeric studies (either prototropic or otherwise)
[15,17,19,23, 26–48]. We will classify the different classes
of tautomerism according to the same principles contained Before the advent of accurate calculations of chemi-
in our book on the tautomerism of heterocycles [49]. cal shifts, the interpolation used the “blocked” N -methyl
i. Annular Tautomerism. This prototropic tau- derivatives, ignoring the effect of the N -methylation [49].
tomerism involves only ring nitrogen (aromatic) and In the case of purine, the authors estimated an N(7)-
carbon atoms (nonaromatic). The most representative ex- H/N(9)-H ratio in DMSO of 33:66 (from 15 N chemical
ample is the azoles (pyrazoles, imidazoles, triazoles, tetra- shifts) and 37:63 (from 13 C chemical shifts) [17]. Our
zoles, and their benzo derivatives). group has applied the same approach to the whole family
a. Azoles. Two cases should be considered: first, of azoles (including 14 and 15) [44], to fluoropyrazoles
when the tautomers are identical (degenerate tau- 16[32], to 3,5-bistrifluoromethyl-pyrazole (17) [38], and
tomerism) and second, when they are different, for in- to chloro- and bromo-1,2,4-triazoles 18 [48]. In this last
stance, benzotriazole itself (7, R = H) has three tautomers, work, the usefulness of the mixed method was qualified
two identical 1H (a) and 3H (c) and another different, as disappointing.
2H (b). The equilibrium between 7a and 7c is impor-
tant, particularly in NMR spectroscopy, since only aver-
age signals are observed corresponding to a 50:50 mixture.
When there is a substituent on the carbocycle (7, R 6= H),
tautomers a and c become different. Stefaniak et al. [19]
have studied this case, combining experimental observa-
tions with GIAO calculations.

On the other hand, for compounds like 17, that


present prototropic tautomerism in the solid state (solid-
state proton transfer, SSPT), the calculations were es-
sential to ascertain that there is a certain amount (about
40 ± 10%) of dynamic disorder [38].
The same approach has been used for histamine b. Other compounds. Only two examples belong to
and its 4-iodo derivative [although this is a classical this section [50]. One is free-base porphyrin (also called
case of annular tautomerism of imidazole; for histamine porphine) (19) and the other is azepines 20 and diazepines
the tautomers are called N(1)-H and N(3)-H] [29,45]. A 21.
systematic study of the effect of different levels of theory
on the GIAO results (HF, MP2, BLYP, and B3LYP with
two basis sets 6-31G** and 6-31++G**) was applied to
pyrazole (14) (average signals in CDCl3 and individual
signals in DMSO), to imidazole (15) (always average sig-
nals), and to purine (5) [17]. When average experimental
signals are compared to calculated ones, if the equilibrium
is degenerate (50:50, cases of pyrazole and imidazole), the
usual procedure is to average the calculated ones. If the
equilibrium involves tautomers of a different energy [case
of purine, here also the numbering is not conventional,
tautomers are called N(9)-H and N(7)-H], the position of
the equilibrium can then be estimated by interpolating the The double-proton transfer in porphyrins in the solid
experimental value between the calculated ones. state is still a matter of controversy. Pulay et al. [35]
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

GIAO Calculations of Chemical Shifts in Heterocyclic Compounds 381

demonstrated that the observed, individual as well as aver- tautomer—a conclusion already known from experimen-
age, signals are well reproduced by the calculations—15 N tal work [49, p. 348]. The simplified tautomerism (only
being the most sensitive. Koch, Wiedel, and Wentrup [27] three tautomers instead of four) of 1-phenylpyrazolones
have carried out a very complete study of the different the- 24 has been studied very competently by Kleinpeter and
oretical approximations in the case of compounds 20 and Koch [46]. Using 13 C NMR experimental and calculated
21 and their nonaromatic CH tautomers. Their main con- results, they determine the major tautomers under differ-
clusions were: (1) HF and B3LYP NMR calculations are ent conditions. This work is also remarkable for the cal-
preferred over BLYP methods; (2) B3LYP calculations are culation of both nonspecific and specific solvent effects
better at predicting the correct order of chemical shifts; on the shieldings (see below). The same authors have
(3) medium-sized basis sets (e.g., 6-31+G**) give used an approach based on 13 C and 15 N NMR to dis-
better results than extensive ones [6-311+G*, cuss the tautomerism of compound 12 involving both a
6-311+G(3df,2p)], with the latter generally leading five-membered (benzo[b]furan) and a six-membered ring
to downfield shifts of relative values; (4) HF single-point (quinoxaline) [42].
calculations both on HF and MP2 geometries give the b. Six-membered rings.
lowest average errors, but B3LYP single points employing
an MP2 geometry seem to be consistently more reliable;
(5) IGLO and LORG do not improve the quality of the
results; (6) judging from these calculations, it seems
important to include polarization functions in the basis
set.
ii. Functional Tautomerism. This tautomerism in-
volves an exocyclic group, like OH, SH, NHR, CHRR’, Some of us have calculated the absolute shield-
and, in some cases, can exist concomitantly with annular ings (1 H, 13 C, 15 N and 17 O) of the two tautomers
tautomerism, for instance, in N -unsubstituted pyrazolones of 2-hydroxypyridine (25) and the three tautomers of
23. 4-hydroxypyrimidine (26) and used them to discuss
a. Five-membered rings. the structure of these compounds in the solid state
(both oxo tautomers) [43]. In the case of the 2,8-
bis(trifluoromethyl)quinoline, the 4-hydroxy tautomer
(27) predominates in acetonitrile solution [31]. Blaze-
jowski et al. [30] have calculated the 1 H and 13 C isotropic
magnetic shieldings of the amino (28) and imino tau-
tomers of 9-acridinamine and concluded that they corre-
late only qualitatively with experimental data and do not
exclude the existence of tautomerism. Finally, the case of
Begtrup et al. [23] reported that the calculated 2-phenacylpyridines (experimentally observed keto and
shieldings (1 H, 13 C, 15 N) of compound 22 are closer to enol tautomers) was studied at the B3LYP/6-311G//HF/3-
the N -hydroxy tautomer 22a than to the N -oxide one 21G (second row) and 6-31G** (third row) [48].
22b. This publication also contains a principal com- c. Fused systems. Only two publications have
ponent analysis (PCA) of the calculated 13 C and 15 N dealt with [5,6] fused heterocycles and both concern
chemical shifts of 13 monosubstituted benzenes and triazolo[1,5-a] pyrimidines.
21 1-substituted pyrazoles, as well as some method-
ological conclusions: 1) GIAO calculations give bet-
ter results than either IGLO or LORG; 2) of the
three methods used [B3LYP/6-311+G(2d,p)//B3LYP/6-
31+G*, B3LYP/6-311+G(2d/p)//B3LYP/6-31G**, and
HF/6-311+G(2d,p)//B3LYP/6-31G**], the former gave
the best results; (3) the chemical shifts of nonmeasured
compounds could be predicted.
The tautomerism of 3-methyl-5-pyrazolone (23a)
was studied by Alderete et al. [39] who concluded that in
DMSO the calculated values (13 C, 15 N) are in better agree- Koch and Kleinpeter [28] studied the tautomerism
ment with the 3-hydroxy tautomer 23b than with any other of derivatives 29 (X = O, S, NH) at different levels
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

382 Alkorta and Elguero

and concluded that HF/6-31G* give 13 C and 15 N results were not included in the paper, as they were considered too
showing sufficient agreement with the experimental val- speculative.
ues corresponding to the 7-oxo tautomer. Molina and
Dobado [40] reported calculations (1 H, 13 C, and 15 N at
the B3LYP/6-311+G** level) on systems 30 with X = O
and NH and R4 H and OH.
iii. Valence Tautomerism. This nonprototropic tau-
tomerism is also known as ring-chain isomerism. It must
be noted that the barrier separating the isomers is large
enough to observe them by NMR at room temperature.
Therefore, the best way to determine the equilibrium con-
stant is by routine 1 H NMR [51].
a. Azido/tetrazole. Cmoch, Stefaniak, Web, and
co-workers [34,36] devoted two papers to the case
of tetrazolo[1,5-a]pyridines 31a in equilibrium with
2-azidopyridines 31b and another to the related case of Two Polish groups reported GIAO calculated shield-
tetrazolo[1,5-b]pyridazines 32a/3-azidopyridazines 32b ings using each experimental geometry to approach this
[37]. In this last paper, there is an example of functional problem (Z 0 = 2 differing in conformation). It is true that
tautomerism coexisting with ring-chain isomerism. The this procedure is not a proof, but, at least, a very difficult
results based on t z + p//dz + p calculations are in good problem can be handled. In this way the case of the bus-
agreement with the experimental results. pirone analogue 34 [53] and that of the synthon in the syn-
thesis of 5-HT1a agonists 35 [54] were solved. Note that
this is related to the problem of what are the best geome-
tries for reproducing CPMAS results: the crystallographic
or the optimized ones? (See later on the condensed-phase
effects “Solid state.”)

Condensed-Phase Effects
b. Other examples. The valence tautomerism of syd-
nones, isosydnones, and isothiosydnones was studied by Unspecific solvent effects. The main contribution on
the same group [26], while Wentrup reported his care- this topic is one by Jaszunski, Mikkelsen, Rizzo, and
ful study of mesoionic compounds and their open-chain Witanowski [55]. These authors considered the nitro-
ketene isomers [41]. gen atoms of all N -methylazoles (fig. 1, if one excluded
the nonexistent 1-methylpentazole). They calculated the
isolated molecule (gas phase) as well as the molecule
Solid State with Several Independent Molecules (Z 0 6= 1) placed in a spherical cavity immersed in a homoge-
neous, isotropic, linear dielectric medium. The method is
Several years ago, we described the X-ray structure the (multiconfiguration) self-consistent reaction field ap-
and the 13 C CPMAS spectrum of campho[2,3-c]pyrazole proach of Mikkelsen and Ruud. The geometries were op-
(33) [52]. This compound crystallizes with six indepen- timized at the MP2 level with a 6-311G** basis set. The
dent molecules (Z 0 = 6) and for most positions; up to NMR calculations were carried out with a Huzinaga’s ba-
six signals were observed (this very good resolution is sis set and a RAS SCF wave function (the CI expansion
related to the spheroidal shape of 33). To assign each includes approximately 1.25 million determinants). With
line of a given carbon atom to an individual molecule this approach they have calculated the effects reported in
of the unit cell requires the growth of a large single- Table I for the two simple cases of 1-methylpyrazole (37)
crystal and to measure the tensors using an NMR instru- and 1-methylimidazole (38).
ment equipped with a goniometer. To determine the sig- The ideal character of these calculations is that all
nals of a given molecule using 2D-spin diffusion needs the data of Table I (eight points) are linearly correlated
an enrichment in 13 C by a factor of 10 is required to (r 2 = 1.000) and, as well, with a dielectric function
make this a feasible experiment. We attempted hierarchi- p(ε) = (ε − 1)/(ε + 2/3) similar to those known by the
cal clustering and trees similitude but these approaches names of Kirkwood, Onsager, etc. [56]. The effects are
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

GIAO Calculations of Chemical Shifts in Heterocyclic Compounds 383

Table I. Calculated Shielding Constant Increments with Respect to Vacuum (in ppm)
[σ(solvent) − σ(vacuum) ]

Solvent N-1 (pyrazole) N-1 (imidazole) N-2 (pyrazole) N-3 (imidazole)

Vacuum 0.00 0.00 0.00 0.00


Cyclohexane −1.28 −2.44 0.98 3.49
Benzene −1.44 −2.80 1.10 4.03
Diethyl ether −2.35 −4.46 1.83 6.49
Hexanol −3.10 −6.05 2.47 8.73
Acetone −3.32 −6.35 2.61 9.19
Methanol −3.43 −6.60 2.64 9.49
Water −3.52 −6.83 2.77 9.81

opposite on both kinds of nitrogen atoms (the pyrrole-like Bednarek et al. [60] have calculated 5-halouracils
N-1 and the pyridine-like N-2/N-3), with imidazole (46, X = F, Cl, Br, I) surrounded by six water molecules
signals being about 2.5 times more sensitive than the forming hydrogen bonds at the HF/3-21G* level; the water
pyrazole ones (a fact probably related to differences in molecules are necessary to reproduce the 1 H chemical
their dipole moments, 3.77 and 2.25 D, respectively, in shifts.
benzene [57]).

At the same time, de Dios et al. [58] have examined


the case of tetrazoles 43 and 44, as well as that of 1,2,4,5-
tetrazine (45). Using the polarizable continuum model of Facelli [61] has explored the effect on the 15 N chem-
Tomasi and the continuous set-gauge transformation of ical shift in the pyridine–methanol complex (47) the func-
Bader, they have represented the evolution of 15 N NMR tioning of the three parameters A, R, and D. This is similar
shieldings against the solvent dielectric constant: the re- to our work on the complex pyridine-water but only on the
sults are quite satisfactory. Catalán [59] has discussed the energetic aspects [62]. This author carried out a PCA anal-
same experimental data, but used an empirical model to ysis of the nitrogen chemical shift tensors (δ11 , δ22 , and
describe the solvents (SB and SA, solvent basicity and sol- δ33 ) to devise a very complex model of dependence of δ
vent acidity); his approach allows the estimation of chem- in the function of A, R, and D.
ical shifts in the gas phase, which are then well correlated Polish authors [63] have calculated the effect of the
with GIAO calculations. proton migration in an imidazolium/imidazole complex
Specific Solvent Effects (Hydrogen Bonds). Klein- (imidazole · · ·H+ · · ·imidazole 48). They conclude that
peter and Koch [46] calculated the effect of DMSO on there exists a correlation between the N· · ·H+ · · ·N ge-
the ring 13 C chemical shifts of pyrazolones 24 using ometry and 13 C shielding parameters, in particular δ22
the “supermolecule” approach. Three tautomers, a, b (from TMS through the relationship δii = 192.7 − σii ) and
and c, were optimized with one molecule of DMSO and 1δ(C4C5) .
the GIAO calculations carried out. This considerably
improved the results.

Associations (Dimers and Beyond). A study has


been undertaken to determine the structure of bulk pyri-
dine from the experimental observation that δ 14 N shifted
5.2 ppm between pure pyridine and infinite dilution in
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

384 Alkorta and Elguero

heptane [64]. Five different dimers and a tetramer were arises if experimental equilibrium (generally from X-ray
calculated. The double hydrogen-bonded dimer involving measurements) or optimized geometries are most con-
H-α (49a, too small effect) and the tetramer (α, γ , α, β venient. Pulay et al. [35] prefer geometries obtained
connectivity, too large effect) were excluded. It remains at the B3LYP level for their work on porphine (19).
as the two dimers involving one Cβ-H· · ·N (49b) or one Schurko and Wasylishen [66] explain that they would
Cγ -H· · ·N hydrogen bond (49c). For these dimers, effects use MP2/6-311G* geometry-optimized structures, which
of about 5.5 ppm were calculated. are very close to experimentally determined geometries.
These last authors reported the effects on the [15 N] la-
beled pyridine when it is coordinated to the 59 Co of
cobaloximes. The calculations correctly predict a negative
coordination shift (i.e., increased nitrogen shielding upon
coordination).
In the already discussed work by Kleinpeter and
Koch on the tautomerism of 1-phenyl-3-methylpyrazolin- Heteroaromaticity Studies
5-ones 24 [46]. These authors, in an effort to reproduce
the experimental observations, have calculated the ef- Several publications deal with the problem of the aro-
fect of the formation of homo dimers CH/CH (24a/24a), maticity of heterocyclic compounds, either through the
OH/OH (24b/24b), and NH/NH (24c/24c). The latter two calculation of the shielding of significant atoms or using
are linked by O-H· · ·N and N-H· · ·O hydrogen bonds as Schleyer’s NICS (nuclear independent chemical shifts)
well as the corresponding homo trimers (we should note calculated on the center of the ring or slightly above it.
that it is possible that hetero dimers would be more stable,
for instance, a NH/OH dimer was found in the solid state
for compound 24, R H [65]). We have summarized their
results in Table II.
Although the use of a supermolecule, including
DMSO, improves the results, as was commented earlier,
the autoassociation (agglomeration) does not, even for the Chesnut and Quin [67] have compared phosphole
solid-state results. (50) to pyrrole (51), furan (52), and thiophene (53). While
Solid State. In addition to the work reported above, the last three show aromatic character (as deduced from
other authors have calculated absolute shieldings to the absolute shieldings compared with the correspond-
compare them with chemical shifts determined in the ing saturated compounds), phosphole does not. Alagona
solid state by the CPMAS technique. The problem then et al. [68] reached the same conclusion for indole (54),
carrying out calculations in the center of the six- and five-
Table II. Experimental and Theoretical 13 C Chemical Shifts (ppm) membered rings (both negative and around –13 ppm, typi-
cal of aromatic compounds). Schleyer et al. [69] addressed
Method C-3 C-4 C-5
the question of stability and aromaticity of fused hetero-
24a, experimental (DMSO) 156.4 43.0 170.5 cycles, for instance, 55 and 56. The NICS values, around
Monomer 24a 150.7 38.9 166.2 −10 ppm for both rings, correspond to aromatic com-
Monomer 24a/DMSOa 154.6 39.3 170.3
Dimer 24a/24a 154.2 37.6 166.9
pounds. The same author has studied the aza[10]annulene
Trimer 24a/24a/24a 157.1 39.7 170.6 system, which presents several isomers, for instance,
24b, experimental (DMSO) 148.5 89.0 155.2 57. At the point marked by a black dot, the NICS
24b, experimental (CPMAS) [65] 150.1 88.3 155.2 amounts to −13.6 ppm, which allows these compounds
Monomer 24a 149.0 78.5 148.4 to be classified as the next higher aromatic analogs of
Monomer 24a/DMSOa 149.9 78.1 150.4
Dimer 24a/24a 157.1 78.8 153.7
pyridine [70].
Trimer 24a/24a/24a 154.7 78.0 153.7
24c, experimental (DMSO) 148.6 92.3 160.4
24c, experimental (CPMAS) [65] 150.1 94.1 163.6
Monomer 24a 156.8 100.8 162.5
Monomer 24a/DMSOa 160.8 96.4 164.0
Dimer 24a/24a 158.8 97.2 167.2
Trimer 24a/24a/24a 163.9 95.1 166.9
a Vide supra.
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

GIAO Calculations of Chemical Shifts in Heterocyclic Compounds 385

NICS have also been used to determine the aro- (46 without water molecules) [89], 2,2’-bipyridyl-3,
matic character of 1, 2-dithiole-2-thione (58a) and 1,2- 3’-diol (67) [90], carbazoles (68, R H, CH3 , C2 H5 ) [91],
dithiole-3-one (58b) (weak in both cases) [71], of other and carbolines (derivatives of 68 with N atoms in the car-
sulfhur-rich heterocycles [72], of bowl-shaped molecules bocycle) [92].
59 (local aromaticity is present in the six-membered
rings) [73], 1-(2-hydroxy-4-bromophenyl)-4-methyl-4-
imidazolin-2-one (60, R H) (−9.2 ppm in the imidazoli-
none ring) [74], and in perfluoroazepines (nonaromatic)
[75].
Recently, a paper has appeared that seems to recon-
cile the points of view of Katritzky and Schleyer [76].
Katritzky published a memorable paper in 1989 about
the multidimensional nature of aromaticity [77], while
Schleyer always preferred a single parameter criteria, like Olah et al. [93] calculated (GIAO-MP2/tzp) sev-
NICS [1,78,79]. We also contributed to this problem, eral compounds formed exclusively by nitrogen atoms
studying magnetic vs. geometric criteria of aromaticity and, although they reported the geometry of N10 (bipen-
using NICS/GIAO/B3LYP/6-31G* calculations on ben- tazole, D2d ), they do not report its 15 N nuclear shield-
zene and distorted benzenes, the conclusion being that a ings (at the B3LYP/6-311++G**//B3LYP/6-311++G**
relationship exists between these criteria, but only for spe- level, we have calculated σ N-1 −50.15, σ N-2 −140.60,
cific subsets [80]. The controversy between both schools and σ N-3 −177.06 ppm). Mathey et al. [94] have reex-
reported in reference [76] is very much alive, but clearly amined the problem of phosphole (50).
they have not succeeded in reaching an agreement. It Jusélius and Sundholm [95] have calculated the NMR
is very unusual that a joint publication ended with two shieldings of porphins (19), chlorins, and bacteriochlorins
conclusions, one by Cyranski, Krygowski, and Katritzky using the Karlsruhe split-valence basis sets of Ahlrichs.
(chemical criteria) and the other by Schleyer (magnetic The 1 H NMR shieldings of the inner hydrogens correlate
criteria). In sofar as the present review is concerned, Ref- well with the calculated current susceptibilities and thus
erence [76] reports NICS(0) and NICS(1) (1 Å above the can be used as an experimental measure for the aromaticity
ring center) for thirty pentagonal heterocycles, including of free-base porphyrins.
furan, thiophene, pyrrole, phosphole, and their aza and
phospha derivatives (pyrazole and imidazole are called Some Methodological Problems
2-azapyrrole and 3-azapyrrole).
There are some problems that are general to these
Examples of Application types of calculations and, consequently, are beyond the
limits of the present review. Nevertheless, they deserve
The 77 Se absolute shielding of selenophene was re- some comment. The first one concerns the transforma-
ported and the methodological problems associated with tion of absolute shielding (isotropic shifts) σ into relative
this nucleus discussed [81]. Stefaniak, Webb, Wiench et chemical shifts δ (both in ppm). Three approaches are
al. have devoted a series of papers to the experimental found concerning this point: (1) the corresponding refer-
study and GIAO calculations of mesoionic compounds ences are not calculated (TMS for 1 H and 13 C, MeNO2
61–65 [26, 82–85]. Witanowski has studied, in detail, the for 15 N, etc.). Instead the authors use a correlation of δ
case of 3-methylsydnone (65, R CH3 , X O) and related vs. σ without discussing the intercept; 2) the references
compounds [86]. (δ = 0) are calculated and used to establish the regres-
sion equation; (3) the calculated value of the reference
is compared to the experimental absolute shielding when
available.
For a given nucleus x X (13 C, 15 N,. . .) in the ideal
case δ x X = σref − σ x X, because δref = 0 by definition. For
practical purposes most authors calculate (by least-squares
fitting) the regression line δ x X(exp) = a − bσ x X. If b = 1,
then σref = a (the intercept of the regression line), but
Other publications concern 1,2,4,5-tetrazine (45) in actual studies where b 6= 1, σref = a/b. Therefore, the
[87], the tricyclic N -pyrrolylborane (66) [88], uracil frequent use of a as σref is not correct. For instance, in
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

386 Alkorta and Elguero

an equation like δ 15 N(exp) = −112.56 − 0.8804 σ 15 N[96], Table IV. GIAO Calculated 15 N Isotropic Shielding of Ammonia
σref is not −112.56, but −112.56/0.8804 = −127.85 ppm. (ppm)
There are references like TMS that are well repro- Method σ 15 N Ref.
duced by the calculations: 1 H σ(exp) = 30.84 ppm, σ(calc) =
HF/6-311++G**//HF/6-311++G** 266.60 [12]
31.97 ppm, 13 C σ(exp) = 188.1, σ(calc) = 184.75 ppm B3LYP/6-311++G**//B3LYP/6-311++G** 259.37 [12]
(B3LYP/6-311++G**//B3LYP/6-311++G**) or water HF/6-311G*//HF/6-311G* 266.2 [12]
17
O σ(exp) = 334, σ(calc) = 322.3 (B3LYP/6-311++G**// XC/TZP//XC/TZP 262.0 [12]
B3LYP/6-311++G**) [12]. Moreover, TMS values are BLYP/TZP//BLYP/TZP 266.24 [12]
rather insensitive to phase effects allowing them to be B3LYP/6-311+G(2d,p)//B3LYP/6-311+G(2d,p) 255.0 [12]
B3LYP//MP2 260.7 [17]
treated together as gas (as most calculated values are) with HF/6-311G*//MP2/6-311G* 271.9 [100]
solution values. For these references, the choice between HF/6-311G**//MP2/6-311G* 274.0 [100]
options (2) and (3) is not very relevant. There are other ref- BPW91/6-311G**//MP2/6-311G* 272.3 [100]
erences where the problem is much more serious. We will BPW91/6-311G*//BPW91/6-311G* 266.2 [100]
discuss, as an example, the (14)15 N nucleus (see [97] for a BPW91/6-311G**//BPW91/6-311G* 267.6 [100]
discussion of the use of 85% H3 PO4 as reference for 31 P).
The standard reference for this nucleus is neat liquid ni-
tromethane δ 15 N = 0.000 ppm [98,99]. This presents three On the other hand, the ammonia experimental value
difficulties: (1) the absolute shielding of nitromethane has [σ 15 N(exp) = 264.5 ppm] is easy to reproduce and it is not
not been determined, (2) the signal of nitromethane is much dependent on the calculation method (see Table IV).
strongly dependent on the phase, and (3) the GIAO cal- In conclusion, following Barfield [100], we think that
culations of σ 15 N of nitromethane are strongly dependent the best solution for 14(15) N NMR is to calculate ammonia
on the basis set used. at the same level as the azaheterocycles and then refer the
To estimate σ 15 N of nitromethane in the gas phase, values to liquid CH3 NO2 using the 399.3 ppm difference
the experimentally measured absolute shielding of ammo- between it and gas-phase ammonia.
nia in the gas phase (264.5 ppm) is transformed into ni-
tromethane (bulk) by using 396.1 ppm [99], 399.3 [100] or CONCLUSION
400.3 [101] to afford −131.6, −134.8, and −135.8 ppm
[102]. From nitromethane (bulk) to nitromethane (gas), Among the unsolved aspects concerning GIAO cal-
an additional 9.05 ppm should be added (from solution in culations is which is the best theoretical approach (HF,
cyclohexane extrapolated to infinite dilution [86]), that is, DFT or MP) and the best basis set (Pople, Huzinaga,
a reasonable estimate is −143 ppm (between −141 and Ahlrichs, etc.). This should remain an open question be-
−145 ppm). cause it will change quickly due to the increasing perfor-
GIAO calculated values for nitromethane are re- mance of the computers and the evolution of the software
ported on Table III. (see, for instance [23, 81, 83, 96, 100, 105]). The obvi-
ous test for any calculation of nuclear shieldings σ is to
try a linear correlation with δ values in the gas phase and
Table III. GIAO Calculated15 N Isotropic Shielding of Nitromethane use, as criteria, a slope as close to 1 as possible and an
(ppm) intercept close to the experimental absolute value of the
Method σ 15 N Ref. reference (1 H TMS = 30.84, 13 C TMS = 188.1 and 15 N
NH3 = 264.5 ppm).
HF/6-311++G**//HF/6-311++G** −184.92 [12]
B3LYP/6-311++G**//B3LYP/6-311++G** −154.43 [12]
In previous times, before chemical shifts were ac-
HF/6-311+G(2d,p)//B3LYP/6-31G** −246.41 [23] cessible by calculation, NMR spectroscopists organized
HF/6-311+G(2d,p)//MP2(full)/6-31G** −269.41 [23] their papers in two sections. The first one was devoted to
B3LYP/6-311+G(2d,p)//B3LYP/6-311+G(2d,p) −158.03 [23] the assignment of the signals and the second one, to the
B3LYP/6-311+G(2d,p)//B3LYP/6-31+G* −159.49 [23] discussion of the chemical shifts thus determined. Con-
B3LYP/6-311+G(2d,p)//MP2(fc)/6-311G** −159.45 [14]
HF/6-31G**//HF/6-31G** −160.53 [103]
jugation, steric effects, lone pairs proximity, aromaticity,
HF/6-31G**//HF/6-31G** −159.9 [104] etc. were invoked to explain the position of the signal.
HF/6-311G**//HF/6-311G** −179.60 [103] With the advent of the methods described in this review
HF/6-311G(2d,p)//HF/6-311G(2d,p) −179.10 [103] and other methods not examined here, such discussions
HF/6-311G(3d,3p)//HF/6-311G(3d,3p) −179.03 [103] are no longer useful. If the calculated chemical shift is
B3LYP/6-31G*//B3LYP/6-311++G** −179.5 [43]
B3PW91//6-311++G** −138.9 [86]
within reasonable boundaries of the experimental one,
no explanation is needed, because the calculation already
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

GIAO Calculations of Chemical Shifts in Heterocyclic Compounds 387

incorporates all the factors the experimentalist could methods; (2) B3LYP calculations are better at predicting
think of. the correct order of chemical shifts; (3) medium-sized ba-
There has been a long and well-established practice sis sets (e.g, 6-31+G**) give better results than extensive
among NMR spectroscopists to study families of com- ones [6-311+G*, 6-311+G(3df,2p)], with the latter gen-
pounds that differ only in the position and nature of the erally leading to downfield shifts of relative values; (4) HF
substituents. The chemical shifts were subsequently dis- single-point calculations, both on HF and MP2 geome-
cussed using Hammett’s relationship or its derivatives tries, give the lowest average errors, but B3LYP single
(Swain and Lupton, Smith and Proulx, Taft, Charton, etc.) points employing an MP2 geometry seem to be consis-
known as LFER or ETR [106,107]. This approach is al- tently more reliable; (5) IGLO and LORG do not improve
ready obsolete, because substituent effects, as perturba- the quality of the results; (6) judging from these calcu-
tions, are particularly well suited for GIAO calculations. lations, it seems important to include polarization func-
Related to the previous topic and also part of LFER tions in the basis set. Finally, publication [23] contains
is the analysis of solvent effects on chemical shifts with some methodological conclusions: (1) GIAO calculations
models like those of Taft, Abraham, Drago, Catalán, etc. give better results than either IGLO or LORG; (2) of the
[108,109]. In this case, calculations would need time to three methods used [B3LYP/6-311+G(2d,p)//B3LYP/6-
supersede them due to the difficulties of accurate model- 31+G*, B3LYP/6-311+G(2d/p)// B3LYP/6-31G**, and
ing of mixed general and specific solvent effects in the- HF/6-311+G(2d,p)//B3LYP/6-31G**] the former gave
oretical chemistry [103]. The same applies to solid-state the best results.
effects—at this moment very difficult to compute. On the In conclusion, for the time being, we advocate the use
other hand, when there is a clear discrepancy between the of B3LYP//6-311++G** for the calculation of absolute
calculated and the measured chemical shift, a new subject shieldings of all types of molecules.
of further research, either experimental or theoretical, has
been found. APPENDIX II. THE PROBLEM OF
CALCULATING NITROGEN MAGNETIC
APPENDIX I. COMPARISON OF THE SHIELDING CONSTANTS
PERFORMANCE OF THE DIFFERENT
CALCULATION LEVELS AND BASIS SETS In a paper reporting the nitrogen chemical shifts of
N+5 the difficulties of calculating 14(15) N chemical shifts
This topic is outside the scope of the present review were discussed [110]. The authors carried out their cal-
because these types of methodological studies are carried culations at the CCSD(T)/QZP level of theory, as rec-
out on simple molecules, not heterocycles. Nevertheless, ommended by Gauss, but pointed out that “An empirical
some of the references we have discussed in the main part correction of −20 ppm was applied to all calculated val-
of this paper contain information about comparative stud- ues, based on a comparison between the calculated and
ies. We ourselves, compared the performance of HF/HF, observed shifts of a number of closely related molecules
HF/DFT, DFT/HF and DFT/DFT approaches for the cal- and ions.” Thus, even at this high level, only reasonable
culation of a large number of small molecules, includ- for small molecules, the authors used a regression equa-
ing the compounds used as references (TMS, water, 85% tion between calculated and experimental values, that is,
phosphoric acid, nitromethane, etc.) [12]. The conclusion the best one can expect concerning nitrogen shieldings is
was that, depending on the nuclei, one or other method, relative values.
performs better. Gauss has discussed, in detail, the difficulties that
We have reported that a systematic study of the ef- the calculation of nitrogen shieldings presents [111–115].
fect of different levels of theory on the GIAO results (HF, In these publications he showed that the calculations of
MP2, BLYP and B3LYP with two basis sets 6-31G** nitrogen derivatives need, at least, GIAO-MP2, GIAO-
and 6-31++G**) was applied to thirty-five 15 N chemical CCSD (coupled cluster singles and doubles), CCSD(T)
shifts [17]. The authors concluded that the choice of the (the noniterative perturbation treatment of triple excita-
optimized geometries seemed to be of minor importance tions) or, better, CCSDT (coupled cluster singles, doubles,
and that the performance of GIAO calculations increased and triples). The case of ammonia was examined in detail.
in the order HF < BLYP < B3LYP. Wentrup et al. [27] Gauss reported that the 15 N absolute shielding of ammonia
have carried out a study of different theoretical approx- was measured experimentally at σ0 = 264.5 ± 0.05 ppm,
imations in the case of compounds 20 and 21 and their but he prefers to take into account the rovibrational cor-
nonaromatic tautomers. Their conclusions were: (1) HF rection (−8.8 ppm) and use σe = 273.3 ± 0.1 ppm [at the
and B3LYP NMR calculations are preferred over BLYP CCSD(T) level, his calculated value is 270.7 ppm]. Gauss
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

388 Alkorta and Elguero

never studied nitromethane, but in the case of N2 O, the 20. Claramunt, R. M.; Lopez, C.; Schmidpeter, A.; Willhalm, A.;
difficulty is even greater than for ammonia. Elguero, J.; Alkorta, I. Spectroscopy 2001, 15, 27.
21. Bocian, W.; Stefaniak, L.; Wiench, J. W.; Webb, G. A. Pol. J. Chem.
In conclusion, σ 15 Ns are more difficult to calculate 1996, 70, 65.
than σ 13 Cs and this is particularly true for CH3 NO2 . Un- 22. Ortı́z, P. J.; Evleth, E. M.; Montero, L. A. Theochem 1998, 432,
fortunately, this compound is firmly established as the 121.
23. Begtrup, M.; Balle, T.; Claramunt, R. M.; Sanz, D.; Jiménez, J. A.;
standard reference in 15 N NMR spectroscopy [98, 99], Mó, O.; Yáñez, M.; Elguero, J., Theochem 1998, 453, 255.
so the only solution for the moment is to use NH3 (gas) as 24. Kleinpeter, E.; Hilfert, L.; Koch, A. J. Phys. Org. Chem. 2000, 13,
a secondary reference. 473.
25. Ferraro, M. B. Theochem 2000, 528, 199.
26. Bocian, W.; Jazwinski, J.; Staszewska, O.; Wiench, J. W.; Stefaniak,
L.; Webb, G. A. Khim. Geterotsikl. Soedin. 1996, p. 1581.
ACKNOWLEDGMENTS 27. Koch, R.; Wiedel, B.; Wentrup, C. J. Chem. Soc. Perkin Trans. 2
1997, p. 1851.
Financial support was provided by the Spanish DG- 28. Koch, A.; Kleinpeter, E. J. Mol. Model. 1997, 3, 375.
29. Mazurek, A. P.; Dobrowolski, J. Cz.; Sadlej, J. J. Mol. Struct. 1997,
ICYT (Project nos. BQU-2000-0252 and BQU-2000- 436–437, 435.
0906) 30. Rak, J.; Skurski, P.; Gutowski, M.; Jozwiak, L.; Blazejowski, J.
J. Phys. Chem. A 1997, 101, 283.
31. Wiench, J. W.; Stefaniak, L.; Grech, E.; Webb, G. A. Bull. Polon.
REFERENCES Acad. Sci. Chem. 1999, 47, 67.
32. Claramunt, R. M.; Alkorta, I.; Elguero, J. Heterocycles 1999, 51,
355.
1. Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.; Jiao, H. H.; van 33. Eckert, F.; Rauhut, G.; Katritzky, A. R.; Steel, P. J. J. Amer Chem.
Eikema Hommes, N. J. R.; J. Amer. Chem. Soc. 1996, 118, 6317. Soc. 1999, 121, 6700.
2. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; 34. Cmoch, P.; Wiench, J. W.; Stefaniak, L.; Webb, G. A. J. Mol. Struct.
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, 1999, 510, 165.
J. A.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; 35. Kozlowski, P. M.; Wolinski, K.; Pulay, P.; Ye, B.-H.; Li, X.-Y.
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J. Phys. Chem. A 1999, 103, 420.
J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; 36. Cmoch, P.; Korczak, H.; Stefaniak, L.; Webb, G. A. J. Phys. Org.
Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, Chem. 1999, 12, 470.
P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; 37. Cmoch, P.; Stefaniak, L.; Melzer, E.; Batoniak, S.; Webb, G. A.
Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Magnetic Resonance Chem. 1999, 37, 493.
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; 38. Alkorta, I.; Elguero, J.; Donnadieu, B.; Etienne, M.; Jaffart, J.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. Schagen, D.; Limbach, H.-H. New J. Chem. 1999, 23, 1231.
A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; 39. Alderete, J. B.; Belmar, J.; Parra, M.; Zuniga, C. Boll. Soc. Chil.
Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, Quim. 2000, 45, 85.
J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, 40. Dobado, J. A.; Grigoleit, S.; Molina-Molina, J. J. Chem. Soc. Perkin
Inc.: Pittsburgh, PA, 1998. 2 2000, p. 1675.
3. Seminario, J. M.; Politzer, P., Eds. Modern Density Func- 41. Plug, C.; Wallfisch, B.; Gade Andersen, H; Bernhardt, P. V.; Baker,
tional Theory, A Tool for Chemistry; Elsevier: Amsterdam, L.-J.; Clark, G. R.; Wah Wong, M.; Wentrup, C. J. Chem. Soc.
1995. Perkin Trans. 2 2000, p. 2096.
4. Bartolotti, J.; Flurcuck, K. Rev. Comp. Chem., 1996, 7, Chapter 4, 42. Kleinpeter, E.; Hilfert, L.; Koch, A. J. Phys. Org. Chem. 2000, 13,
p. 187. 473.
5. Kohn, W.; Becke, A. D.; Parr, R. G. J. Phys. Chem. 1996, 100, 43. López, C.; Claramunt, R. M.; Alkorta, I.; Elguero, J., Spectroscopy
12974. 2000, 14, 121.
6. Becke, A. D. J. Chem. Phys. 1993, 98, 5648. 44. Claramunt, R. M.; López, C.; Sanz, D.; Alkorta, I.; Elguero, J.
7. Becke, A. D. Phys. Rev. A 1988, 38, 3098. Heterocycles 2001, 55, 2109.
8. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. 45. Garnuszek, P.; Dobrowolski, J. C.; Sitkowski, J.; Bednarek, E.;
9. Miehlich, B.; Savin, A.; Stoll H.; Preuss, H. Chem. Phys. Lett. 1989, Witowska, J.; Mazurek, A. P. J. Mol. Struct. 2001, 565–566, 361.
157, 200. 46. Kleinpeter, E.; Koch, A. J. Phys. Org. Chem. 2001, 14, 566.
10. London, F. J. Phys. Radium 1937, 8, 397. 47. Osmialowski, B.; Kolehmainen, E.; Gawinecki, R. Magnetic Res-
11. Ditchfield, R. Mol. Phys. 1974, 27, 789. onance Chem. 2001, 39, 334.
12. Alkorta, I.; Elguero, J. Struct. Chem. 1998, 9, 187. 48. Claramunt, R. M.; López, C.; Garcı́a, M. A.; Otero, M. D.; Torres,
13. He, Y.; Wu, D.; Shen, L.; Li, B.; Webb, G. A. Magnetic Resonance M. R.; Pinilla, E.; Alarcón, S. H.; Alkorta, I.; Elguero, J., New J.
Chem. 1995, 33, 701. Chem. 2001, 25, 1061.
14. Hathaway, B. A.; Day, G.; Lewis, M.; Glaser, R. J. Chem. Soc. 49. Elguero, J.; Marzin, C.; Katritzky, A. R.; Linda, P. The Tautomerism
Perkin Trans. 2 1998, p. 2713. of Heterocycles; Academic Press: New York, 1976.
15. Martı́nez-Merino, V.; Gil, M. J. J. Chem. Soc. Perkin Trans. 2 1999, 50. Claramunt, R. M.; Elguero, J.: Katritzky, A. R. Advan. Heterocycl.
p. 33. Chem. 2000, 77, 1.
16. Cmoch, P.; Wiench, J. W.; Stefaniak, L.; Sitkowski, J. J. Mol. Struct. 51. Minkin, V. I.; Garnovskii, A. D.; Elguero, J.; Katritzky, A. R.;
1999, 477, 119. Denisko, O. V. Advan. Heterocycl. Chem. 2000, 76, 157.
17. Dokalik, A.; Kalchhauser, H.; Mikenda, W.; Schweng, G. Magnetic 52. Llamas-Saiz, A. L.; Foces-Foces, C.; Sobrados, I.; Elguero, J.;
Resonance Chem. 1999, 37, 895. Meutermans, W. Acta Crystallogr., C 1993, 49, 724.
18. Jaroszewska-Manaj, J.; Maciejewska, D.; Wawer, I. Magnetic Res- 53. Szelejewska-Wozniakowska, A.; Chilmonczyk, Z.; Les, A.;
onance Chem. 2000, 38, 482. Wawer, I. Solid State Nucl. Magnetic Resonance, 1998, 13, 63.
19. Wiench, J. W.; Stefaniak, L.; Barszczewicz, A.; Webb, G. A. 54. Maciejewska, D.; Herold, F.; Wolska, I. J. Mol. Struct. 2000, 553,
J. Mol. Struct. 1994, 327, 321. 73.
P1: GAD
Structural Chemistry (STUC) PP872-stuc-465428 June 11, 2003 18:8 Style file version Nov. 07, 2000

GIAO Calculations of Chemical Shifts in Heterocyclic Compounds 389

55. Jaszunski, M.; Mikkelsen, K. V.; Rizzo, A.; Witanowski, M. 85. Jazwinski, J.; Staszewska, O.; Wiench, J. W.; Stefaniak, L.;
J. Phys. Chem. A 2000, 104, 1466. Araki, S.; Webb, G. A. Magnetic Resonance Chem. 2000, 38,
56. Koppel, I. A.; Palm, V. A. In Advances in Linear Free Energy 617.
Relationships; Chapman, N. B.; Shorter, J., Eds.; Plenum Press: 86. Witanowski, M.; Biedrzycka, Z.; Grabowski, Z. Magnetic Reso-
London, 1972; p. 215. nance Chem. 2000, 38, 580.
57. Mauret, P.; Fayet, J.-P.; Fabre, M. Bull. Soc. Chim. Fr. 1975, p. 1675. 87. Witanowski, M.; Biedrzycka, Z.; Sicinska, W.; Grabowski, Z.;
58. Manalo, M. N.; de Dios, A. C.; Cammi, R. J. Phys. Chem. A 2000, Webb, G. A. J. Magnetic Resonance 1997, 124, 127.
104, 9600. 88. Wrackmeyer, B.; Schwarze, B.; Milius, W.; Boese, R.; Parchment,
59. Catalán, J. J. Chem. Soc. Perkin Trans. 2 2001, p. 1117. O. G.; Webb, G. A. J. Organomet. Chem. 1998, 552, 247.
60. Bednarek, E.; Dobrowolski, J. C.; Dobrosz-Teperek, K.; Kozerski, 89. Bednarek, E.; Dobrowolski, J. Cz.; Dobrosz-Teperek, K.;
L.; Lewandowski, W.; Mazurek, A. P. J. Mol. Struct. 2000, 554, Sitkowski, J.; Kozerski, L.; Lewandowski, W.; Mazurek, A. P.
233. J. Mol. Struct. 1999, 482–483, 333.
61. Facelli, J. C. Chem. Phys. Lett. 2000, 322, 91. 90. Kaczmarek, L.; Zagrodzki, B.; Kamienski, B.; Pietrzak, M.; Schilf,
62. Llamas-Saiz, A. L.; Foces-Foces, C.; Mó, O.; Yáñez, M.; Elguero, W.; Les, A. J. Mol. Struct. 2000, 553, 61.
J. Acta Crystallogr. Sect. B 1992, 48, 700. 91. Kupka, T.; Pasterna, G.; Jaworska, M.; Karali, A.; Dais, P. Magnetic
63. Potrzebowski, M. J.; Cypryk, M.; Michalska, M.; Koziol, A. E.; Resonance Chem. 2000, 38, 149.
Kazmierski, S.; Ciesielski, W.; Klinowski, J. J. Phys. Chem. B 92. Corbally, R. P.; Mehta, L. K.; Parrick, J.; Short, E. L. Magnetic
1998, 102, 4488. Resonance Chem. 2000, 38, 1034.
64. Megiel, E.; Kasprzycka-Guttman, T.; Jagielska, A.; Wroblewska, 93. Olah, G. A.; Prakash, G. K. Surya; Rasul, G. J. Amer. Chem. Soc.
L. J. Mol. Struct. 2001, 569, 111. 2001, 123, 3308.
65. Foces-Foces, C.; Fontenas, C.; Elguero, J.; Sobrados, I. An. Quim. 94. Mattmann, E.; Mathey, F.; Sevin, A.; Frison, G. J. Org. Chem. 2002,
Intern. Ed. 1997, 93, 219. 67, 1208.
66. Schurko, R. W.; Wasylishen, R. E. J. Phys. Chem. A 2000, 104, 95. Juselius, J.; Sundholm, D. Phys. Chem. Chem. Phys. 2000, 2,
3410. 2145.
67. Chesnut, D. B.; Quin, L. D. J. Amer. Chem. Soc. 1994, 116, 9638. 96. Witanowski, M.; Biedrzycka, Z.; Sicinska, W.; Grabowski, Z.
68. Alagona, G.; Ghio, C.; Monti, S. Theochem 1998, 433, 203. J. Magnetic Resonance 1998, 131, 54.
69. Subramanian, G.; Schleyer, P. v. R.; Jiao, H. Angew. Chem. Intern. 97. Mroz, P.; Pietrusiewicz, K. M.; Wolinski, K. Mol. Phys. Rept. 2000,
Ed. Engl. 1996, 35, 2638. 29, 205.
70. Bettinger, H. F.; Sulzbach, H. M.; Schleyer, P. v. R.; Schaefer, H. F. 98. Witanowski, M.; Stefaniak, L.; Szymanski, S.; Januszewski,
J. Org. Chem. 1999, 64, 3278. J. Magnetic Resonance 1977, 28, 217.
71. Fabian, J.; Herzog, K. Vib. Spectrosc. 1998, 16, 77. 99. Srinivasan, P. R.; Lichter, R. L. J. Magnetic Resonance 1977, 28,
72. Fabian, J.; Gloe, K.; Wust, M.; Kruger-Rambusch, T.; Rademacher, 227.
O.; Graubaum, H. Phosphorus Sulfur Silicon Related Elements 100. Barfield, M.; Fagerness, P. J. Amer. Chem. Soc. 1997, 119,
1998, 140, 35. 8699.
73. Delaere, D.; Nguyen, M. T.; Vanquickenborne, L. G. Chem. Phys. 101. Witanowski, M.; Stefaniak, L.; Webb, G. A. Annu. Rept. NMR
Lett. 2001, 333, 103. Spetrosc. 1993, 25, 88.
74. Cyranski, M. K.; Wawer, I.; Zielinska, A.; Mrozek, A.; Koleva, V.; 102. Jameson, C. J.; Jameson, A. K.; Oppusungu, D.; Wille, S.; Burrell,
Lozanova, C. J. Phys. Org. Chem. 2001, 14, 323. P. M.; Mason. J. J. Chem. Phys. 1981, 74, 81.
75. Karney, W. L.; Kastrup, C. J.; Oldfield, S. P.; Rzepa, H. S. J. Chem. 103. Zhan, C.-G.; Chipman, D. M. J. Chem. Phys. 1999, 110, 1611.
Soc. Perkin Trans. 2 2002, p. 388. 104. Witanowski, M.; Biedrzycka, Z.; Sicinska, W.; Webb, G. A. J. Mol.
76. Cyranski, M. K.; Krygowski, T. M.; Katritzky, A. R.; Schleyer, P. Struct. 2000, 516, 107.
v. R. J. Org. Chem. 2002, 67, 1333. 105. Rablen, P. R.; Pearlman, S. A.; Finkbiner, J. J. Phys. Chem. A 1999,
77. Katritzky, A. R.; Barczynski, P.; Musumarra, G.; Pisano, D.; 103, 7357.
Szafran, M. J. Amer. Chem. Soc. 1989, 111, 7. 106. Chapmann, N. B.; Shorter, J. Eds., Advances in Linear Free Energy
78. Schleyer, P. v. R.; Freeman, P. K.; Jiao, H.; Goldfuss, B. Angew. Relationships; Plenum Press: London, 1972.
Chem. Intern. Ed. Engl. 1995, 34, 337. 107. Faure, R.; Llinares, J.; Elguero, J.; Goya, P. Bull. Soc. Chim. Belg.
79. Schleyer, P. v. R.; Puchta, R.; Boggavarapu, K.; Mauksch, M.; 1987, 96, 603.
Eikema Hommes, N. v.; Alkorta, I.; Elguero, J. manuscript in 108. Chapmann, N. B., Shorter, J. Eds.; Correlation Analysis in Chem-
preparation. istry; Plenum Press: New York, 1978.
80. Alkorta, I.; Elguero, J. New J. Chem. 1999, 23, 951. 109. Abboud, J.-L. M.; Boyer, Elguero, J.; G.; Cabildo, P.; Claramunt,
81. Bühl, M.; Thiel, W.; Fleischer, U.; Kutzelnigg, W. J. Phys. Chem. R. M. Spectrochim. Acta 1991, 47A, 785.
1995, 99, 4000. 110. Christe, K. O.; Wilson, W. W.; Sheehy, J. A.; Boatz, J. A. Angew,
82. Bocian, W.; Wiench, J. W.; Stefaniak, L.; Webb, G. A. Magnetic Chem. Int. Ed. Engl. 1999, 38, 2004.
Resonance Chem. 1996, 34, 453. 111. Gauss, J. J. Chem. Phys. 1993, 99, 3629.
83. Wiench, J. W.; Stefaniak, L.; Tabaszewska, A.; Webb, G. A. Elec- 112. Gauss, J.; Stanton, J. F. J. Chem. Phys. 1995, 103, 3561.
tron. J. Theoret. Chem. 1997, 2, 71. 113. Gauss, J.; Stanton, J. F. J. Chem. Phys. 1996, 105, 2574.
84. Jazwinski, J.; Staszewska, O.; Staszewski, P.; Stefaniak, L.; 114. Gauss, J.; Stanton, J. F. Phys. Chem. Chem. Phys. 2000, 2, 2047.
Wiench, J. W.; Webb, G. A. J. Mol. Struct. 1999, 475, 181. 115. Gauss, J.; Werner, H.-J. Phys. Chem. Chem. Phys. 2000, 2, 2083.

You might also like