The observed and calculated 1H and 13C chemical shifts of tertiary amines and their N-oxides

Article abstract of DOI:10.1002/mrc.2750

A series of model tertiary amines were oxidized in situ in an NMR tube to amine N-oxides and their ¹H and ¹³C NMR spectra were recorded. Next, the chemical shifts induced by oxidation (Δλ) were calculated using different GIAO methods investigating the influence of the method [Hartree-Fock (HF), Moeller-Plesset perturbation, density functional theory (DFT)], the functional applied in the DFT (B3LYP, BPW, OPBE, OPW91) and the basis set used [6-31G*, 6-311G**, 6-311 + + G** and 6-311 + + G(3df,3pd)]. The best results were obtained with the HF/6-311 + + G** and OPBE/6-311 + + G** methods. The computation/experiment comparison approach was used for the configuration prediction of chiral amine N-oxides-(R) and (S)-agroclavine-6-N-oxide.

Full text of DOI:10.1002/mrc.2750

Research Article
Received: 14 December 2010
Revised: 25 January 2011
Accepted: 04 February 2011
Published online in Wiley Online Library: 31 March 2011
(wileyonlinelibrary.com) DOI 10.1002/mrc.2750
1
13
The observed and calculated H and C
chemical shifts of tertiary amines and their
N-oxides
∗
Radek Pohl, Martin Dra cˇ ínsk y´ , Lenka Slav eˇ tínsk a´ and Milo sˇ Bud eˇ sˇ ínsk y´
A series of model tertiary amines were oxidized in situ in an NMR tube to amine N-oxides and their H and 13C NMR spectra
were recorded. Next, the chemical shifts induced by oxidation (ꢀδ) were calculated using different GIAO methods investigating
the influence of the method [Hartree-Fock (HF), Moeller-Plesset perturbation, density functional theory (DFT)], the functional
1
∗
∗∗
∗∗
applied in the DFT (B3LYP, BPW, OPBE, OPW91) and the basis set used [6-31G , 6-311G , 6-311++G and 6-311++G(3df,3pd)].
∗
∗
∗∗
The best results were obtained with the HF/6-311++G and OPBE/6-311++G methods. The computation/experiment
comparison approach was used for the configuration prediction of chiral amine N-oxides – (R) and (S)-agroclavine-6-N-oxide.
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: NMR; H; 13C; in situ oxidation of tertiary amines; calculated chemical shifts; HF; MP2; DFT
1
Introduction
Therefore, we decided to test the experiment/calculation
comparison approach in the prediction of chiral N-oxide con-
figuration. This article deals mainly with the in situ oxidation of
model achiral tertiary amines 1a–5a to amine N-oxides 1b–5b
N-oxides of tertiary amines bearing different substituents on the
nitrogen atom are chiral compounds with N-central chirality.
Typical examples of such compounds can be found in various
classes of alkaloids isolated from natural sources, e.g. (+)-
(Fig. 1), a description of the methodology and a screening of
appropriate calculation methods. The application of the method
is then demonstrated on diastereoisomeric N-oxides – (R) and
[
1]
bulbocapnine-β-N-oxide fromGlauciumfimbrilligerum, (+)-5,17-
(S)-agroclavine-6-N-oxides.
[
2]
dehydromatrine N-oxide from Euchresta japonica, pericine N-
oxide[ from Kopsia arborea and (4S)-corynoxeine N-oxide[4] from
Uncariarhynchophylla. Chiral N-oxides have also found application
in asymmetric catalysis, e.g. in asymmetric cyanosilylation of
3]
Experimental
ketones,[ in borane-mediated reduction of ketones or in
5]
[6]
All the amines 1a–5a, the deuterated solvents (CDCl3 and
CD2Cl2) and m-chloroperbenzoic acid (MCPBA) used in this
work were obtained commercially from Sigma-Aldrich Co. The
amine N-oxides 1b–5b were prepared by an in situ oxidation of
corresponding amines 1a–5a with MCPBA in an NMR tube with a
CDCl3 solution.
[
5]
asymmetric allylation of aldehydes.
In the early years of NMR when the configuration of chiral
[
7]
[8]
N-oxides was determined by chemical evidence or X-ray, the
effect of the N-O group in amine N-oxides was observed for the
1
first time in H NMR. The reports dealing with the configuration
determination of chiral N-oxides by NMR spectroscopy are mainly
based on nuclear Overhauser effect (NOE) contacts of alkyl
substituents on the nitrogen atom with other parts of the
The NMR spectra were measured on a Bruker Avance 600
1
13
(with H at 600.13 MHz and C at 150.9 MHz frequency) using a
5–mm TXI cryo-probe and about 5–10 mg of sample in 0.6 ml of
molecule.[
1,9–12]
The JC,C coupling constants were also used for
1
1
CDCl . The chemical shifts are given in δ-scale [with the H shifts
3
[
13]
13
the discrimination of N-epimeric amine oxides.
referenced to TMS and the C referenced to CDCl using δ(CDCl )
3
3
In the literature, there are several examples of the combination
ofexperimentalNMRdatawithcalculatedNMRparameters(shield-
ing nuclear constants and scalar couplings) addressing various
at 77.00 ppm].
1
The typical experimental conditions for the H NMR spectra
were 16 scans, a spectral width of 6 kHz and an acquisition time
of 5 s, yielding 60 K data points. The free induction decays (FIDs)
were zero-filled to 128 K data points.
stereochemical problems, e.g. the conformation of sparteine N-
oxides,[
14,15]
theconformationandconfigurationoftertiaryamines
1
3
via GIAO-derived C chemical shifts and a multiple independent
variable regression analysis, the conformation of N-substituted
piperidines and pyrrolidines or the relevance of the calculated
H, C and 15N shieldings in amines in conformational analysis.[18]
[
16]
[
17]
∗
Correspondence to: Milo ˇs Bud eˇ ˇs ínsk y´ , Institute of Organic Chemistry and
1
13
Biochemistry, Czech Academy of Sciences, Prague, Czech Republic.
E-mail: budesinsky@uochb.cas.cz
GIAO NMR calculations combined with probability analysis were
also used for the assignment of stereochemistry when the NMR
Institute of Organic Chemistry and Biochemistry, Czech Academy of Sciences,
Prague, Czech Republic
[
19]
data were available only for one diastereoisomer.
Magn. Reson. Chem. 2011, 49, 320–327
Copyright ꢀc 2011 John Wiley & Sons, Ltd.

Observed and calculated H and 13C chemical shifts
1
netic shielding constants were calculated at different levels of
theory including the Hartree-Fock (HF) approximation, Moeller-
Plesset perturbation (MP2) theory and density functional theory
(
DFT). In the DFT calculations, B3LYP, BPW and the exchange-
correlation functionals OPBE and OPW91 were used. In all the
∗
∗∗
∗∗
methods, four basis sets [6-31G , 6-311G , 6-311++G and
-311++G(3df,3pd)] were used.
6
Results
Experimental NMR data
The oxidation reactions were monitored by the H and 13C NMR
1
Figure 1. The structures of model tertiary amines and their N-oxides.
spectra as shown for the in situ oxidation of N-methylpyrrolidine
3
a in Fig. 2. Since MCPBA could in general also protonize the
ThetypicalexperimentalconditionsfortheAPT13CNMRspectra
were 64 scans, a spectral width of 30 kHz and an acquisition time
of 1 s, yielding 60 K data points. The FIDs were zero-filled to 128 K
data points.
starting amine, we measured the NMR spectra of ammonium
trifluoroacetates obtained by an in situ protonation of the
corresponding amines with trifluoroacetic acid (TFA) to determine
the resonances that could originate from protonized forms.
1
The two-dimensional (2D) homonuclear (H,H-COSY and H,H-
rotating frame NOE spectroscopy) and 2D heteronuclear [H,C-
heteronuclear single-quantum correlation and H,C-HMBC) exper-
iments were performed when needed for the structural assign-
ments of the signals (with standard 2D NMR pulse sequences of
Bruker software being used). For compounds 4a and 5a, the low-
Protonationismanifestedinthe HNMRspectrabytheappearance
of an NH proton and its coupling with protons in vicinal positions.
However, theprotonatedformoftertiaryamineswasnotobserved
after the addition of MCPBA to amines 1a–5a probably due to the
low acidity of MCPBA (pKa = 3.82) in comparison with TFA (pKa
= 0.23). The protonation of 1a–5a does not significantly change
1
13
◦
13
13
temperature H and C NMR spectra (at 0, −25, −50 and −75 C)
the C chemical shifts and the protonation-induced C chemical
in CD2Cl2 were also measured.
shift is −3.0 to 0.6 ppm for C-α, whereas the oxidation-induced
1
3
1
The geometry optimizations and chemical-shift calculations
were performed using the Gaussian 03 or Gaussian 09 software
package.[ The molecular geometries were optimized at the
C chemical shift is 9.7–13.4 ppm for C-α. In H NMR, both the
1
protonation and oxidation of 1a–5a change the H chemical
shifts dramatically and the induced H chemical shifts for H-α are
0.45–1.40 and 0.89–1.69 ppm, respectively.
20]
1
∗∗
B3LYP/6-311++G level of theory in vacuo. The nuclear mag-
Figure 2. The in situ protonation and oxidation of N-methylpyrrolidine 3a monitored by the 1H (left) and APT 13C NMR (right) spectra: (a) N-
methylpyrrolidine 3a in CDCl3; (b) the spectra after an addition of TFA – the formation of protonized 3a; (c) the spectra after an addition of MCPBA – the
formation of 3b.
Magn. Reson. Chem. 2011, 49, 320–327
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

R. Pohl et al.
Table 1. The H and 13C mean absolute errors (in ppm) of selected computational methods for the calculation of the induced chemical shift in
1
N-oxidation of 1 and 2
Basis set
HF
MP2
B3LYP
BPW
OPBE
OPW91
1
MAE for the H induced chemical shift
∗
6
6
6
6
-31G
0.39
0.41
0.44
0.42
0.37
0.41
0.45
0.46a
0.34
0.34
0.39
0.37
0.31
0.31
0.34
0.33
0.33
0.35
0.36
0.35
0.32
0.34
0.36
0.35
∗
∗
-311G
∗
∗
-311++G
-311++G(3df,3pd)
MAE for the 13C induced chemical shift
∗
6
6
6
6
-31G
0.75
0.79
0.60
0.74
2.48
2.47
2.10
3.36a
2.12
2.27
1.70
1.81
2.30
2.43
1.87
1.98
1.77
1.69
1.06
1.12
1.77
1.88
1.06
1.12
∗∗
-311G
∗∗
-311++G
-311++G(3df,3pd)
a
Calculated for 1 only.
1
Selection of calculation method
Two simple amines 1a and 2a and their N-oxides 1b and
b were used for the selection of the appropriate calculation
method. The molecular geometry of 1a and 2a and their oxidized
products 1b and 2b are well defined, and there is no worry
about a possible conformation equilibrium. The nuclear magnetic
shielding constants were calculated (see Supporting information)
for 1a, 1b, 2a and 2b employing the methods listed in Table 1.
Calculation of the H and 13C chemical shifts (δ values)
The values of ꢀδ are not always accessible especially if one
wonders about an unknown N-oxide configuration of one
particularcompound. Inthiscase, thereferencingofthecalculated
shielding constant to a reference compound is necessary. The
most used reference compound is TMS, but methane
the simplest hydrocarbon was also used for the H and 13
2
[
27]
as
C
1
referencing in the theoretical calculations. We referenced the
calculated nuclear shielding constants of 1a, 1b, 2a and 2b to
TMS [δref(H) = 0 ppm, δ (C) = 0 ppm] and methane [δ (H) =
Calculation of H and 13C chemical shifts induced by oxidation
1
ref
ref
(
ꢀδ values)
[28]
0
.22 ppm, δref(C) = −4.63 ppm] calculated at the same level of
theory according to Eqn (2):
The calculated oxidation-induced chemical shifts (ꢀδcalc
=
σamine − σaminoxide) were compared with the experimentally
observed values (ꢀδobs = δaminoxide − δamine), and the mean
absolute errors (MAEs) were calculated according to Eqn (1):
δi = σref − σi + δref
(2)
The results are summarized in Table 2. It is obvious that using
a different reference compound will provide different values of δ.
n
ꢀ
1
MAE = n
|ꢀδi(calc) − ꢀδi(obs)|
(1)
∗
∗∗
Going from 6-31G to 6-311++G , the differences are smaller.
i=1
1
For the H data, the chemical shift differences using two reference
compounds are 0.32, 0.28, 0.03 and 0.12 ppm for HF/6-31G , HF/6-
∗
The results obtained by the different calculation methods are
compared in Table 1.
We experienced very similar H MAE for all the tested methods
∗∗
∗
∗∗
311++G , OPBE/6-31G and OPBE/6-311++G , respectively.
1
13
For the C data, the differences using two reference compounds
∗
∗∗
∗
(
0.31–0.46 ppm), reflecting a serious underestimation of ꢀδ (see
are 5.5, 3.3, 3.0 and 1.3 ppm for HF/6-31G , HF/6-311++G ,
∗
∗∗
also Supporting information). This is in agreement with our
previous observation for the oxidation-induced chemical shift of
sulfoxides,[ which might be explained by neglecting the solvent
OPBE/6-31G and OPBE/6-311++G , respectively. HF/6-31G
1 13
underestimated both the H and C chemical shifts, but the
other methods listed in Table 2 have performed relatively well.
21]
[
22]
[23,24]
∗
effect and vibrational averaging.
Previous calculation
Therefore, even the inexpensive OPBE/6-31G method could be
showed that using the polarizable continuum model for the
inclusion of the solvent effect did not improve the results of the
calculation.
used for the prediction of amine N-oxide chemical shift, especially
C.
1
3
A much more diverse situation is seen for 13C, where the
MAE appeared in the 0.60–3.36 ppm interval depending on the
method. Table 1 shows that the HF method provided the best
results (lowest MAEs) with only a small effect of the basis set.
Rather surprisingly, the highest MAEs were in general obtained
by the MP2. For the DFT method, we tested the popular B3LYP
hybrid functional, which yielded slightly better results than the
BPW functional. The OPBE and OPW91 functionals, which were
Conformation and NMR chemical shifts of 3a–5a and 3b–5b
The shielding of a nucleus is very sensitive to the neighboring
environment, and it is therefore not surprising that chemical
shifts reflect the conformation of a molecule. When we consider
the calculation of chemical shifts, we have to be careful with
the conformational equilibria, and weighted NMR parameters
according to the Boltzmann distribution of possible conformers
shouldbeused.Therefore,wedecidedtoexploretheconformation
equilibria for compounds 3a–5a and 3b–5b.
revealedaspromisingDFTfunctionalsforthecalculationofnuclear
shielding constants,[
25,26]
were found to be more advantageous
than the other DFT functionals. The selection of the basis set
appeared to be important and the addition of diffusion functions
to heavy atoms and hydrogens significantly improved the MAEs.
The conformation of 3a was studied by microwave
[
29]
[30]
spectroscopy
or by gas electron diffraction.
We used the
known concept of pseudorotation[ and calculated the energies
31]
wileyonlinelibrary.com/journal/mrc
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2011, 49, 320–327

Observed and calculated H and 13C chemical shifts
1
Table 2. The calculated referenced and observed chemical shifts for 1a, 1b, 2a and 2b
HF
HF
OPBE
OPBE
∗
6-311++G∗∗
∗
6-31G
6-311++G∗∗
6-31G
Compound
Observed
TMSa
CH4b
TMSa
CH4b
TMSa
CH4b
TMSa
CH4b
1H chemical shifts
1
1
2
a
b
a
2.22
3.44
2.85
2.06
2.77
2.49
1.26
1.37
2.95
1.62
1.45
1.74
2.45
2.17
0.94
1.05
2.63
1.30
1.13
2.06
2.69
2.51
1.28
1.23
2.83
1.61
1.37
1.78
2.40
2.22
0.99
0.94
2.54
1.32
1.08
2.06
3.08
2.68
1.32
1.64
3.01
1.95
1.66
2.09
3.12
2.71
1.35
1.67
3.04
1.98
1.69
2.16
3.06
2.85
1.41
1.50
2.96
1.99
1.65
2.28
3.21
2.97
1.53
1.62
3.08
2.11
1.77
1
.53
.74
1
2
b
3.82
2
.09
.15
2
13C chemical shifts
43.78
1
1
2
a
b
a
47.57
60.17
47.63
41.82
55.21
42.02
24.48
18.67
56.94
23.70
18.39
36.32
49.71
36.52
18.78
13.17
51.44
18.20
12.89
40.49
53.68
42.19
24.37
19.02
56.61
23.63
18.88
45.06
60.26
47.90
27.44
23.23
63.94
28.49
22.56
42.06
57.26
44.90
24.44
20.23
60.94
25.49
19.56
46.40
60.32
51.23
30.47
26.07
66.21
30.85
25.56
47.76
61.68
52.59
31.83
27.43
67.57
32.21
26.92
56.97
45.48
2
6.57
0.58
27.76
2
22.31
2
b
61.07
59.90
2
5.94
9.72
26.29
1
22.17
a
Referenced to tetramethylsilane, calculated at the same level of theory.
Referenced to methane, calculated at the same level of theory.
b
Figure 3. The conformation search for N-methylpyrrolidine 3a. (a) The energy map for 201 conformers with different P and φmax and (b) the energy
◦
profile for φmax = 45 .
∗
∗
(
B3LYP/6-31G ) for 201 conformers with a variation of the pucker
calculated ꢀG298 of optimized conformers 4aax, 4aeq,5aax, 5aeq,
4bax, 4beq,5bax and 5beq (Fig. 4) showed that the predominant
conformationistheonewithaCH3 groupintheequatorialposition,
◦
◦
◦
◦
◦
amplitude φmax (0 → 50 , 5 step) and phase angle P (0 → 360 ,
◦
1
8 step; for the definition of φmax and P, see Supporting infor-
◦
[32–34]
mation). The preferred conformations were found for φmax = 45 ,
which is in agreement with the published knowledge.
◦
◦
◦
1
P = 90 and φmax = 40 , P = 270 as shown in Fig. 3. In the
The experimental H NMR spectra of 4a and 5a showed in CDCl3
◦
subsequent step, the preferred conformers were fully optimized
at 25 C a significant line broadening of the α and γ (in the case
∗∗
using B3LYP/6-311++G , and ꢀG298 was obtained by vibrational
of 4a) protons caused by a nitrogen inversion and the chemical
analysis (Fig. 4).
exchange of the axial and equatorial protons. This process can be
◦
In the case of 3b, we considered the similar conformations with
the CH3 group in the axial and equatorial positions and calculated
the corresponding ꢀG298 (Fig. 4). In six-membered rings 4a, 5a, 4b
and 5b, CH3 can again occupy the axial or equatorial position. The
sloweddownbycoolingthesampleto−75 C.Atthistemperature,
sharp signals of one conformer 4aeq and 5aeq are observed.
1
13
Wecalculatedthe Hand Cnuclearshieldingconstantsfortwo
conformers of 3a,b–5a,b using the OPBE/6-311++G∗ method
∗
Magn. Reson. Chem. 2011, 49, 320–327
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

R. Pohl et al.
Figure 4. The conformation preferences for 3a–5a and 3b–5b.
3
bax –5bax, which demonstrates not only the necessity of a
Table 3. A comparison of the experimental and calculateda 13C
chemical shifts of N-oxides 3beq –5beq and 3bax –5bax
properly chosen molecular geometry for the calculation of the
chemical shifts but also the relevance of the calculated chemical
shifts in conformation analysis.
3
b
3beq 3bax
4b 4beq 4bax
5b 5beq 5bax
exp calc calc exp calc calc exp calc calc
C-α
C-β
C-γ
CH3
68.6 70.4 73.2 66.4 67.2 72.2 65.0 66.5 71.2
21.9 26.6 24.9 20.9 24.2 28.4 61.4 64.5 67.1
Determination of N-O configuration in agroclavine 6-N-oxides
–
–
–
21.3 25.5 25.9
54.4 55.7 55.2 57.8 61.1 53.4 59.0 61.2 52.9
2.6 2.8 2.9 5.6 2.3 6.0
–
–
–
The main goal of our project is to find the method for an
unambiguous assignment of the N-O configuration in chiral amine
N-oxides containing at least one additional element of chirality. To
demonstrate the computation/experiment comparison approach,
we calculated the nuclear-shielding constants and referenced the
chemical shifts for agroclavine 6a, 6S-agroclavine N-oxide S-6b
and 6R-agroclavine N-oxide R-6b (Fig. 5) and compared them with
the already published NMR data.[
MAE
a
∗∗
OPBE/6-311++G , referenced to TMS, calculated at the same level
of theory.
11]
(
see Supporting information). To demonstrate the importance of
As arises from the DFT calculations, agroclavine 6a can exist as a
mixture of two conformers differing by the nitrogen configuration.
First, the geometry of the two conformers of 6a was optimized
using B3LYP/6-311++G with a vibrational analysis providing
the conformation equilibrium (Fig. 6) and then the geometry of
N-oxides S-6b and R-6b was optimized by the same method. For
S-6b and R-6b, only one conformer was found and considered for
the calculation of NMR parameters.
the correct molecular geometry on the calculated chemical shifts,
we compared the experimental and calculated 13C chemical shifts
of N-oxides 3beq –5beq and 3bax –5bax (Table 3).
The results in Table 3 show that the calculated 13C chemical
shifts, especially C-α and CH3 in the vicinity of the N-O group,
of major conformers 3beq –5beq are closer to the experimental
values. The MAEs for 3beq –5beq are lower than those for
∗∗
Figure 5. The structure and numbering of agroclavine 6a, 6S-agroclavine N-oxide S-6b and 6R-agroclavine N-oxide R-6b.
wileyonlinelibrary.com/journal/mrc
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2011, 49, 320–327

Observed and calculated H and 13C chemical shifts
1
Figure 6. The conformers of agroclavine 6a.
Table 4. The observed and calculated chemical shifts for 6a, S-6b and R-6b
6
a
S-6b
R-6b
S-6b
R-6b
obsa
calcb
obsa
calcb
obsa
calcb
ind;obsc
ind;calcd
ind;obsc
ind;calcd
Proton
H-2
6.91
6.85
6.95
6.91
6.96
6.98
0.04
0.52
0.07
0.90
0.74
0.83
0.15
0.57
0.11
0.06
0.05
0.04
0.81
0.06
0.86
−0.15
0.91
0.50
1.05
0.06
0.99
0.16
0.00
0.06
0.03
0.57
0.05
0.11
0.56
0.95
0.97
0.66
0.04
0.01
0.02
0.03
0.05
0.05
0.63
0.13
−0.21
1.44
H-4ax
H-4eq
H-5
2.72
3.33
2.47
2.93
3.24
6.22
3.68
6.91
7.06
7.14
1.80
2.47
2.94
3.21
2.69
2.88
3.30
6.59
3.98
7.21
7.43
7.20
1.82
2.44
3.24
3.40
3.37
3.67
4.07
6.37
4.25
7.02
7.12
7.19
1.84
3.28
3.80
3.06
3.60
3.38
4.35
6.65
4.97
7.37
7.43
7.26
1.85
3.01
2.83
3.89
3.42
3.90
3.90
6.26
3.69
6.93
7.09
7.19
1.85
3.10
2.73
4.65
3.32
3.60
3.77
6.55
3.86
7.17
7.39
7.29
1.91
2.87
0.63
H-7a
H-7b
H-9
0.72
0.47
−0.04
−0.12
−0.04
−0.04
0.09
H-10
H-12
H-13
H-14
H-17
NCH3
0.09
0.43
Carbon
C-2
119.7
113.3
120.7
113.8
120.8
115.0
1.0
−1.9
−4.8
8.9
0.5
−1.4
−5.5
11.5
12.8
−5.5
0.5
1.1
−1.5
−4.7
11.7
13.8
−4.8
0.0
1.7
0.1
C-3
112.1
27.7
112.2
30.4
110.2
22.9
110.8
24.9
110.6
23.1
112.3
25.5
C-4
−4.9
13.5
16.1
0.2
C-5
65.7
64.0
74.6
75.5
77.4
77.5
C-7
61.6
61.8
73.4
74.6
75.4
78.0
11.8
−7.4
0.5
C-8
135.6
121.2
42.1
132.2
121.0
44.6
128.2
121.7
38.7
126.6
121.4
41.4
130.8
121.2
41.8
132.3
120.7
45.7
C-9
−0.2
1.2
C-10
C-11
C-12
C-13
C-14
C-15
C-16
C-17
NCH3
−3.5
−1.9
0.9
−3.2
−1.4
0.6
−0.3
−2.3
0.8
133.0
113.3
123.7
110.1
132.8
127.8
21.2
130.1
111.3
120.6
104.6
129.0
123.2
23.7
131.1
114.2
124.2
110.7
135.6
127.1
21.0
128.7
111.9
120.7
105.2
128.5
122.2
24.0
130.7
114.2
124.1
111.1
135.9
127.5
20.8
128.3
111.5
120.3
105.8
129.2
123.2
23.6
−1.9
0.2
0.5
0.1
0.5
−0.3
1.2
0.6
0.5
1.0
2.8
−0.5
−1.1
0.4
3.1
0.2
−0.7
−0.2
17.3
−0.3
−0.4
8.6
0.0
−0.1
8.7
41.3
40.3
58.5
57.4
49.8
48.9
17.1
a
Measured in CD3OD; see reference [11].
b
c
∗∗
OPBE/6-311++G ; referenced to TMS, calculated at the same level of theory.
Observed chemical shift induced by oxidation (ꢀδobs = δaminoxide − δamine).
Calculated chemical shift induced by oxidation (ꢀδcalc = σamine − σaminoxide).
d
The nuclear-shielding constants were calculated afterwards on
the optimized geometries using the OPBE/6-311++G∗∗ method.
The data for 6a were weighted according to the calculated ratio of
the conformers at 298 K. We calculated the oxidation-induced
chemical shifts (ꢀδcalc = σamine − σaminoxide), experimentally
observed values (ꢀδobs = δaminoxide −δamine) and also the chemical
shifts referenced to TMS, calculated by the same method. The
results are presented in Table 4.
The 1H NMR data clearly reflect the orientation of the
N-O group. For example, in S-6b, where the N-O group is axial,
the H-4ax and H-10 protons (for the structure, see Supporting
information) are significantly deshielded by the known 1,3-diaxial
Magn. Reson. Chem. 2011, 49, 320–327
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

R. Pohl et al.
Figure 7. The optimized geometries of S-6b and R-6b with the specified 1,3-diaxial interaction of the N-O group and the corresponding proton.
Figure 8. The correlation between the calculated (referenced to TMS) and observed 13C and H chemical shifts of 6a, S-6b and R-6b.
1
interaction between a particular proton and the N-O group, but
hardly any change in chemical shift for those protons is observed
for R-6b (Fig. 7), whereas, on the other hand, H-4eq is deshielded
because of the parallel orientation of C-H4eq and the N-O bonds.
The experimentally observed effect is nicely reproduced by the
satisfactory results for both oxidation-induced and referenced
chemical shifts of tertiary amines and their N-oxides.
We confirmed that a correct molecular geometry is an essential
requirement for obtaining realistic chemical shifts by calculation.
Inaddition, thecalculatedchemicalshiftscanhelpinconformation
analysis in some cases.
1
3
calculations as follows from Table 4. A dramatic change in the
C
chemical shift is observed for the carbon of CH3-N. Depending on
its arrangement, the chemical shift can differ by about 10 ppm for
the two diastereoisomers S-6b and R-6b owing to the γ -gauche
interaction of N-CH3 in R-6b. Again, the observed chemical shifts
for N-CH3 correlate very well with those calculated.
The computation/experiment comparison approach was suc-
cessfully applied to the N-O group configuration of 6S-
agroclavnine N-oxide S-6b and 6R-agroclavnine N-oxide R-6b,
indicating that the method can be used for determining the
configuration of chiral N-oxides. This preliminary finding will be
subjected to more extensive study with N-oxides of alkaloids and
other chiral tertiary amines in our next publication.
To present the performance of the OPBE/6-311++G∗∗ calcu-
lation method, we correlated the calculated referenced chemical
shifts with those observed for 6a, S-6b and R-6b as shown in
1
3
Fig. 8. As arises from the correlations, the calculated C chemical
shifts can be used for the unambiguous assignment of N-O con-
Acknowledgements
1
figuration in chiral amine N-oxides. The correlation for H is less
This work was supported by the Czech Science Foundation (Grant
No. 203/09/1919).
satisfactory, but some effects, e.g. 1,3-diaxial interaction, could be
useful in determining the configuration.
Supporting information
Supporting information may be found in the online version of this
article.
Conclusions
We have tested several computation methods of 1H and 13
chemical-shift calculation on a series of model tertiary amines
C
∗∗
References
and their achiral N-O counterparts. The HF/6-311++G method
worked well for the induced chemical shifts. OPBE/6-311++G
provided the best results from the tested DFT methods yielding
∗∗
[
1] A. Shafiee, K. Morteza-Semnani, M. Amini, J. Nat. Prod. 1998, 61,
1564.
wileyonlinelibrary.com/journal/mrc
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2011, 49, 320–327

Observed and calculated H and 13C chemical shifts
1
[
[
[
2] S. Ohmiya, H. Otomasu, J. Haginiwa, I. Murakoshi, Phytochemistry
978, 17, 2021.
3] K. Lim, O. Hiraku, K. Komiyama, T. Koyano, M. Hayashi, T. Kam, J.Nat.
Prod. 2007, 70, 1302.
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery,
J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,
1
K. N. Kudin,
V. N. Staroverov,
R. Kobayashi,
J. Normand,
4] B. Ma, C. Wu, J. Yang, R. Wang, Y. Kano, D. Yuan. Helv. Chim. Acta
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi,
M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
2
009, 92, 1575.
[
[
[
[
[
5] A. Malkov, P. Kocovsky, Eur. J. Org. Chem. 2007, 29.
6] I. O’Neil, C. Turner, S. Kalindjian, Synlett 1997, 777.
7] N. Mandava, G. Fodor, Can. J. Chem. 1968, 46, 2761.
8] C. Huber, G. Fofdor, N. Mandava, Can. J. Chem. 1971, 49, 3258.
9] G. Caldwell, A. Gauthier, J. Mills, Mag. Reson. Chem. 1996, 34, 505.
R. L. Martin,
K. Morokuma,
V. G. Zakrzewski,
G. A. Voth,
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc.,
Wallingford, CT, 2009.
[
[
[
[
[
10] G. Caldwell, A. Gauthier, J. Mills, M. Greco, Mag. Reson. Chem. 1993,
1, 309.
11] V. Kren, J. Nemecek, V. Prikrylova, Collect. Czech. Chem. Commun.
995, 60, 2165.
12] S. Pospisil, P. Sedmera, P. Halada, L. Havlicek, J. Spizek, Tetrahedron
Lett. 2004, 45, 2943.
3
[21] M. Dracinsky, R. Pohl, L. Slavetinska, M. Budesinsky, Mag. Reson.
Chem. 2010, 48, 718.
1
[22] M. Dracinsky, J. Kaminsky, P. Bour, J. Chem. Phys. 2009, 130, 094106.
[23] M. Dracinsky, P. Bour, J. Chem. Theory Comput. 2010, 6, 288.
[24] M. Dracinsky, J. Kaminsky, P. Bour, J.Phys.Chem.B 2009, 113, 14698.
[25] Y. Zhang, A. Wu, X. Xu, Y. Yan, Chem. Phys. Lett. 2006, 421, 383.
[26] A. Wu, P. Zhang, X. Xu, W. Yan, J. Comp. Chem. 2007, 28, 2431.
[27] V. Ananikov, Cent. Eur. J. Chem. 2004, 1, 196.
13] F. Potmischil, H. Herzog, J. Buddrus, Mag. Reson. Chem. 1998, 36,
240.
14] J. Thiel, W. Boczon, P. Fiedorow, B. Jasiewicz, M. Knychala, J. Mol.
Struct. 2002, 642, 15.
15] B. Jasiewicz, Molecules 2008, 13, 3.
16] A. Sebag, D. Forsyth, M. Plante, J. Org. Chem. 2001, 66, 7967.
17] C. Lobato-Garcia, P. Guadarrama, C. Lozada, R. Enriquez, D. Gnecco,
W. Reynolds, J. Mol. Struct. 2006, 786, 53.
[28] G. Fulmer, A. Miller, N. Sherden, H. Gottlieb, A. Nudelman, B. Stoltz,
J. Bercaw, K. Goldberg, Organometallics 2010, 29, 2176.
[29] W. Caminati, F. Scappini, J. Mol. Spect. 1986, 117, 184.
[30] G. Pfafferott, H. Oberhammer, J. Boggs, J. Am. Chem. Soc. 1985, 107,
2309.
[
[
[
[18] I. Alkorta, J. Elguero, Mag. Reson. Chem. 2004, 42, 955.
[19] S. Smith, J. Goodman, J. Am. Chem. Soc. 2010, 132, 12946.
[20] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
[31] C. Altona, M. Sundaralingam, J. Am. Chem. Soc. 1972, 94, 8205.
[32] E. Eliel, D. Kandasamy, C. Yen, K. Hargrave, J. Am. Chem. Soc. 1980,
102, 3698.
J. R. Cheeseman,
G. Scalmani,
V. Barone,
B. Mennucci,
[33] F. dos Santos, C. Tormena, J. Mol. Struct.(Theochem) 2006, 763, 145.
[34] T. Rosenau, A. Hofinger, A. Potthast, P. Kosma, Polymer 2003, 44,
6153.
G. A. Petersson, H. Nakatsuji, X. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada,
Magn. Reson. Chem. 2011, 49, 320–327
Copyright ꢀc 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc