Orientations of the 17O electric-field-gradient tensor and chemical shift tensor in benzamide: NMR of dipolar coupled spins

Article abstract of DOI:10.1016/S0009-2614(00)01463-9

We report a solid-state ¹⁷O NMR study of [α-¹³C,¹⁷O] benzamide. The orientations of the ¹⁷O electric-field-gradient and chemical shift tensors were determined from analysis of magic-angle spinning and stationary ¹⁷O NMR spectra. The largest electric-field-gradient component lies in the amide plane and perpendicular to the ¹³C-¹⁷O dipolar vector, whereas the intermediate electric-field-gradient component is along the C=O bond. It is also found that the principal component of the ¹⁷O chemical shift tensor with the least shielding, δ₁₁, is approximately 18° off the C=O bond and that the component with the most shielding, δ₃₃, is perpendicular to the amide plane. The present study confirms our earlier results of quantum chemical calculations.

Full text of DOI:10.1016/S0009-2614(00)01463-9

9
February 2001
Chemical Physics Letters 334 (2001) 265±270
www.elsevier.nl/locate/cplett
1
7
Orientations of the O electric-®eld-gradient tensor
and chemical shift tensor in benzamide: NMR of dipolar
coupled spins
*
Gang Wu , Shuan Dong
Department of Chemistry, Queen's University, Kingston, Ontario, Canada K7L 3N6
Received 7 September 2000; in ®nal form 20 November 2000
Abstract
We report a solid-state O NMR study of [a- C; O] benzamide. The orientations of the O electric-®eld-gradient
and chemical shift tensors were determined from analysis of magic-angle spinning and stationary O NMR spectra.
1
7
13
17
17
1
7
1
3
17
The largest electric-®eld-gradient component lies in the amide plane and perpendicular to the C± O dipolar vector,
whereas the intermediate electric-®eld-gradient component is along the C@O bond. It is also found that the principal
1
7
component of the O chemical shift tensor with the least shielding, d11, is approximately 18° o€ the C@O bond and that
the component with the most shielding, d33, is perpendicular to the amide plane. The present study con®rms our earlier
results of quantum chemical calculations. Ó 2001 Elsevier Science B.V. All rights reserved.
1
. Introduction
quantum chemical calculations have become an
increasingly important tool even to experimental-
ists. However, it is always desirable to obtain
experimental con®rmation for the results of
quantum chemical calculations. The ideal experi-
mental NMR technique for determining the ab-
solute orientation of an NMR tensor in the
molecular frame is single-crystal NMR. Unfortu-
nately, single-crystal NMR experiments require
large single-crystal samples that very often are
dicult to obtain. Growing large single-crystals
Recently we attempted to determine the orien-
17
tations of the O electric-®eld-gradient (EFG)
tensor and chemical shift (CS) tensor for the car-
bonyl oxygen in amides [1±3]. Our approach was
17
to obtain the relative orientation between the
O
EFG and chemical shift tensors by analyzing
NMR spectra of a stationary powder sample at
more than one magnetic ®eld strength, and then
deduce the absolute tensor orientations in the
molecular frame from quantum chemical calcula-
tions. Recent advances in computational chemistry
have made it possible to accurately evaluate NMR
tensors at least for light elements [4±6]. As a result,
17
containing O labels is even more challenging
since the quantity of the O-labeled sample is
usually small.
17
In this contribution, we chose to use an alter-
native approach to determine the absolute orien-
17
tation of O NMR tensors in the molecular
frame. In particular, we synthesized singly and
*
Corresponding author. Fax: +1-613-545-6669.
E-mail address: gangwu@chem.queensu.ca (G. Wu).
17
doubly isotopic labeled benzamide samples, [ O]
0009-2614/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 6 1 4 ( 0 0 ) 0 1 4 6 3 - 9

2
66
G. Wu, S. Dong / Chemical Physics Letters 334 (2001) 265±270
1
3
17
benzamide and [a- C; O] benzamide. Using the
C± O dipolar vector as an internal reference in
the molecular frame, we were able to determine the
then poured into cold water and the white product
was extracted with ethyl ether. The product was
characterized by IR, NMR and melting point. The
theoretical O enrichment level in [a- C; O]
benzamide is 26.4%.
13
17
17
17
13
17
absolute orientations of the O chemical shift and
EFG tensors in benzamide. Although NMR of
dipolar coupled spins has been extensively studied
for spin-1/2 systems [7±14], we are unaware of any
study in which the half-integer quadrupolar nu-
cleus under direct observation is dipolar coupled
to a spin-1/2 nucleus.
2.2. Solid-state NMR
17
Solid-state O NMR spectra were recorded on
a Bruker Avance-500 spectrometer operating at
1
17
5
00.13 and 67.78 MHz for H and O nuclei, re-
spectively. Polycrystalline samples were packed
into zirconium oxide rotors (4 mm o.d.). A Bruker
2. Experimental
4
mm MAS probe was used for both MAS and
2
.1. Sample synthesis
static experiments. The sample spinning frequency
was 14 kHz in the MAS experiments. The RF ®eld
strength for O was approximately 80 kHz. In the
17
O static experiments, the whole-echo acquisition
17
17
Water ( O 52.8% atom) was obtained from
17
ISOTEC (Miamisburg, Ohio). Benzoic acid- O
was prepared by reacting a; a; a-trichlorotoluene
and H
Benzoic acid- O was then converted to [ O]
benzoyl chloride ( O 52.8% atom) by re¯uxing
2
approach [15] was used to avoid acoustic ringing
from the probe and to enhance sensitivity. In
particular, the pulse width for the ®rst and second
pulses was 1 and 2 ls, respectively. The delay be-
tween the ®rst and second pulses was 120 ls,
which is suciently long to allow the detection of
the whole echo signal. To the whole-echo NMR
signal was applied a large ®rst-order phase cor-
rection to obtain the distortion-free line shape. A
recycle delay of 10 s was used in all experiments.
17
2
O in a sealed tube at 120°C for 15 h.
17 17
2
17
1
7
thionyl chloride and benzoic acid- O for 1 h
2
followed by distilling o€ the excessive thionyl
chloride. [ O] benzamide was synthesized by re-
acting [ O] benzoyl chloride with ammonia in
methanol at room temperature for 3 h.
17
17
13
17
2
.1.1. [a- C; O] Benzoyl chloride
Chemical shifts were referenced to a liquid H O
2
1
3
A total of 0.344 g (2.8 mmol) of [a- C] benzoic
sample. Spectral simulations were performed with
WSOLIDS (Drs. Klaus Eichele and Rod Wasyli-
shen, Dalhousie University).
13
acid ( C 99.2 atom% obtained from CDN Iso-
topes, Pointe-Claire, Quebec, Canada) was ref-
luxed with excessive thionyl chloride for 1 h.
13
[a- C] Benzoyl chloride was obtained after re-
moving the remaining thionyl chloride by distilla-
3. Results and discussion
17
tion. 0.05 g (2.8 mmol) of H
[
2
O was added to
a- C] benzoyl chloride. The reaction tube was
then sealed and left at 60°C for 5 h, which yielded
13
Fig. 1 shows the experimental and calculated
13
17
17
O MAS NMR spectra of [a- C; O] benzamide
13
17
solid [a- C; O] benzoic acid when cooled to
obtained with a spinning rate of 14 kHz. The
spectrum exhibits a typical central-transition
NMR line shape arising only from the second-or-
der quadrupole interaction. This is because the
13
17
room temperature. [a- C; O] Benzoic acid was
13
17
readily converted to [a- C; O] benzoyl chloride
by the same procedure as mentioned above.
1
3
17
C± O dipolar interaction is completely averaged
17
13
17
1
13
2
.1.2. [a- C; O] Benzamide
by MAS and Jꢀ C; O† is negligibly small. The
A cold ammonia-methanol solution was added
only indirect spin±spin coupling constants between
13
13
17
17
dropwise to [a- C; O] benzoyl chloride until the
temperature was stabilized. The solution was stir-
red at room temperature for 3 h. The solution was
C and O nuclei reported in the literature are
17
1
13
those for acetone ( Jꢀ C; O† ˆ 22 Hz) [16], CO
( Jꢀ C; O† ˆ 16:4 Hz) [17] and CO ꢀ Jꢀ C; O†
1
13
17
1
13
17
2

G. Wu, S. Dong / Chemical Physics Letters 334 (2001) 265±270
267
tween the dipolar vector and the ith principal
component of the chemical shift tensor ꢀi ˆ 1; 2; 3†.
The general appearance of a dipolar NMR spec-
trum is simply a sum of two overlapping chemical
shift powder patterns [14]. From the experimental
dipolar splittings, it is possible to obtain the ori-
entations of the principal elements of the chemical
shift tensor with respect to the dipolar vector. The
essence of the dipolar NMR approach is that the
dipolar vector serves as an internal reference.
For half-integer quadrupolar nuclei, the situa-
tion is more complex. Here we will focus only on
NMR spectra from the central-transition
Fig. 1. Experimental (upper trace) and calculated (lower trace)
13
17
17
O MAS NMR spectra of [a- C; O] benzamide at 11.75 T.
Spinning rate, 14 kHz. Recycle delay, 10 s. Number of tran-
sients, 1306.
ꢀ
� 1=2 $ ‡1=2†. Under such circumstances, two
types of second-rank tensors are operative, e.g.,
chemical shift tensor and EFG tensor. The resul-
tant NMR spectrum usually exhibits more singu-
larities than does a normal chemical shift powder
pattern. Detailed theoretical analysis of such
NMR powder spectra can be found in the litera-
ture [18±22]. In general, it is possible to obtain
information about the relative orientation between
the chemical shift and EFG tensors from analyzing
stationary NMR spectra obtained at multiple
magnetic ®elds. Of course, the absolute orienta-
tions of these tensors in the molecular frame of
reference are still unavailable from such an anal-
ysis.
ˆ
16:1 Hz† [17]. These provide an upper limit for
13 17 17
1
Jꢀ C; O† in amides. Analysis of the O MAS
spectrum yields the following NMR parameters:
2
d
iso ˆ 300 Æ 1 ppm, e qQ=h ˆ 8:40 Æ 0:01 MHz
and g ˆ 0:37 Æ 0:02. These parameters are identi-
17
cal to those reported in our earlier study of [ O]
benzamide [1].
17
Before we present detailed experimental
O
NMR results for stationary samples, it is worth
discussing brie¯y the basic principles of dipolar
NMR spectroscopy. For magnetically dilute spin-
1
/2 nuclei, an NMR spectrum obtained for a sta-
Similar to the dipolar NMR for spin-1/2 nuclei,
if the quadrupolar nucleus under observation is
dipolar coupled to a neighboring spin-1/2 nucleus,
dipolar splittings will also be observed at spectral
singularities. However, in this case, the singulari-
ties may not correspond to the positions of any
principal component of either chemical shielding
or EFG tensors. Rather these singularities are
known as the critical frequencies or critical points
[19]. In general there are up to six critical points in
a powder NMR spectrum arising from a combi-
nation chemical shift and second-order quadru-
polar interactions. As illustrated in Fig. 2A, a
typical O NMR powder spectrum without any
dipolar coupling exhibits both steps (labeled as S)
and peaks (labeled as P). In order to illustrate the
basic features of dipolar NMR spectra for half-
integer quadrupolar nuclei, three di€erent situa-
tions where a neighboring spin-1/2 nucleus is pre-
sent are considered in Fig. 2B±D. Here the
tionary powder sample often exhibits a line shape
known as the chemical shift powder pattern. From
the three singularities of the chemical shift powder
pattern, it is straightforward to obtain the princi-
pal components of the chemical shift tensor. In the
absence of any local molecular symmetry, it is
generally dicult to know the directions of these
principal components in the molecular frame of
reference. When the spin-1/2 nucleus under ob-
servation is adjacent to another spin-1/2 nucleus of
di€erent kind, spectral features known as the di-
polar splittings [13,14] are often observed at the
singularities of the chemical shift powder pattern.
The magnitude of the dipolar splittings can be
written as
17
2
Dm ˆ R ꢀ1 � 3 cos # †
where RDD is the dipolar coupling constant,
ꢀ1†
ii
DD
ii
3
2
ꢀl0=4p†ꢀ1=r †c c ꢀh=4p †, and #ii is the angle be-
1
2

2
68
G. Wu, S. Dong / Chemical Physics Letters 334 (2001) 265±270
point S2 is approximately two times larger than
those at S1 and P1. The dipolar splittings at P2
and P3 are very small. As seen from Fig. 2C, when
the dipolar vector is at (0°, 90°), which means that
the dipolar vector is along the smallest EFG
component, VXX, the dipolar splittings at S1 and
S2 are much smaller than that at P1. Similarly,
when the dipolar vector is along the intermediate
EFG component, VYY, the largest dipolar splitting
17
in the O NMR spectrum appears at S1, as illus-
trated in Fig. 2D.
1
7
The experimental and calculated stationary
13
O
17
17
NMR spectra of [ O] benzamide and [a- C; O]
benzamide are shown in Fig. 3. The analysis of the
17
17
O NMR spectrum for [ O] benzamide has been
described in detail, in our previous studies [1±3].
From the analysis, we obtained the following pa-
rameters: d11 ˆ 500 Æ 20, d22 ˆ 400 Æ 20, d33 ˆ 0 Æ
2
0 ppm, a ˆ 8 Æ 5°, b ˆ 90 Æ 2°, c ˆ 72 Æ 5°
where the three Euler angles de®ne the relative
orientation between the EFG and chemical
shielding tensors. These results suggest that VXX is
parallel to d33 and that d11 makes an angle of
1
8 Æ 5° with respect to VYY [1±3].
Comparison of the NMR spectra shown in
Fig. 3 clearly reveals the presence of dipolar
splittings at the critical points of the spectrum for
13
17
[
a- C; O] benzamide. The largest dipolar split-
ting was observed at approximately 580 ppm (de-
noted as S1), 3868 Æ 50 Hz. Meanwhile the dipolar
splitting at 200 ppm (P1) is much smaller,
1
7
Fig. 2. Typical O NMR powder spectrum (A) and dipolar
spectra (B, C, D). The positions of the six critical frequencies
are labeled as S1, S2 and S3 for steps and P1, P2, and P3 for
peaks. In (B), (C) and (D), the orientation of the dipolar vector
is given.
1
616 Æ 50 Hz. The smallest dipolar splitting is at
2
0 ppm (P3), 1176 Æ 50 Hz. As mentioned earlier,
17
17
analysis of the O NMR spectrum of [ O] ben-
zamide yielded the relative orientation between the
17
direction of the dipolar vector in the principal-
axis-system (PAS) of the EFG tensor was de®ned
by (/; h) where h is the polar angle and / the az-
imuthal angle.
Fig. 2B shows the theoretical NMR line shape
when the dipolar vector lies along the largest
component of the EFG tensor, VZZ. Clearly dipo-
lar splittings are observed at all critical points. As
expected, the magnitude of each dipolar splitting
depends on the relative orientation of the dipolar
vector with respect to the direction that gives rise
to the critical frequency. For example, as seen
from Fig. 2B, the dipolar splitting at the critical
O EFG and chemical shielding tensors. The ob-
servation of dipolar splittings in the O NMR
17
13
17
spectrum of [a- C; O] benzamide allowed us to
obtain the orientation of the dipolar vector in the
PAS of the EFG tensor. In this case, we found that
the use of a dipolar vector direction of (90°, 90°) in
spectral simulation yielded the best agreement
between the observed and calculated line shapes.
13
17
This result means that the C± O dipolar vector
is along the intermediate component, VYY. The
13
17
dipolar coupling constant, RDDꢀ C; O†, used in
the spectral simulation was )2100 Æ 50 Hz,
which corresponds to a C@O bond length of

G. Wu, S. Dong / Chemical Physics Letters 334 (2001) 265±270
269
17
17
Fig. 3. Experimental (A) and calculated (B) stationary O NMR spectra of [ O] benzamide at 11.75 T. Recycle time, 10 s. Number of
13
17
17
transients, 2924. Experimental (C) and calculated (D) stationary O NMR spectra of [a- C; O] benzamide at 11.75 T. Recycle delay,
1
0 s. Number of transients, 4991.
ꢀ
1
with that determined from a low-temperature
:249 Æ 0:009 A. This is in excellent agreement
ꢀ
neutron di€raction study, 1.245 A [23].
Interestingly, the observed dipolar splitting at
13
17
S1, 3868 Æ 50 Hz, is nearly twice of RDDꢀ C; O†,
which means the direction of S1 is nearly parallel
to the C@O bond. Using Eq. (1), we calculated the
angle between the direction of S1 and the C@O
bond to be 13 Æ 2°. Since S1 represents the highest
frequency end of the spectrum, to a very good
approximation, it corresponds to the direction of
the tensor component with the least shielding, d11
.
Therefore, the value of 13 Æ 2° is also consistent
1
7
Fig. 4. Illustration of the orientations of the O NMR tensors
in benzamide.
with our previous result that d11 makes an angle of
1
8 Æ 5° relative to the C@O bond [1±3]. Of course,
the precise direction of S1 must be calculated using
the equations given by Chu and Gerstein [19].
The orientations of the O NMR tensors in the
molecular frame of benzamide are depicted in
Fig. 4. The largest component of the EFG tensor,
con®rm our earlier results of quantum chemical
calculations [1±3].
17
Finally, it should be mentioned that the fact
that the dipole±dipole coupling tensor is axially
symmetric imposes an intrinsic limitation on the
dipolar NMR approach. That is, dipolar NMR
spectra are invariant to any simultaneous rotation
of the NMR tensors about the dipolar vector. In
the present case, since the tensor component as-
sociated with the most shielding, d , is known to
V
ZZ, is in the amide plane and perpendicular to the
C@O bond. The smallest EFG tensor component,
V
XX, is perpendicular to the amide plane. For the
O chemical shielding tensor, the principal com-
1
7
ponent with the largest shielding, d , is perpen-
33
dicular to the amide plane. Most importantly, the
tensor component with the least shielding, d11, is
3
3
be always perpendicular to the amide plane, the
absolute orientations of the O NMR tensors
17
18 Æ 5° o€ the C@O bond. These orientations

2
70
G. Wu, S. Dong / Chemical Physics Letters 334 (2001) 265±270
depicted in Fig. 4 can be unambiguously deter-
mined. Based on results of the present and previ-
ous studies [1±3], we can conclude that the
References
[
[
1] G. Wu, S. Dong, K. Yamada, H. Grondey, J. Am. Chem.
Soc. 122 (2000) 4215.
2] K. Yamada, S. Dong, G. Wu, J. Am. Chem. Soc. 122
17
absolute orientations of the O NMR tensors for
an amide functional group show very little varia-
tions among di€erent amides. However, the mag-
(
2000) 11602.
[3] S. Dong, K. Yamada, G. Wu, Z. Naturforsch. 55a (2000)
21.
17
nitude of the O NMR tensors is very sensitive to
both inter- and intra-molecular interactions. It is
[
[
[
[
4] W. Kutzelnigg, U. Fleischer, M. Schindler, NMR-Basic
Principles and Progress, vol. 23, Springer-Verlag, Berlin,
17
this latter aspect of the O NMR tensors that may
be useful in future solid-state O NMR studies of
chemical and biological systems.
1990, 165.
17
5] J.A. Tossell (Ed.), Nuclear Magnetic Shieldings and
Molecular Structure, Kluwer Academic, Dordrecht, The
Netherlands, 1993.
6] J.C. Facelli, A.C. de Dios (Eds.), Modeling NMR Chem-
ical Shifts, American Chemical Society, Washington, DC,
4. Conclusions
1
999.
7] D.L. VanderHart, H.S. Gutowsky, J. Chem. Phys. 49
1968) 261.
13
17
We have utilized the C± O dipolar interaction
17
(
13
in [a- C; O] benzamide to obtain the orientations
of O NMR tensors in the molecular frame. The
[8] D.L. VanderHart, H.S. Gutowsky, T.C. Farrar, J. Chem.
Phys. 50 (1969) 1058.
17
13
17
[9] S. Kaplan, A. Pines, R.G. Grin, J.S. Waugh, Chem.
Phys. Lett. 25 (1974) 78.
results for [a- C; O] benzamide con®rmed our
earlier theoretical results, which strongly suggests
that the current quantum chemical calculations are
[10] K.W. Zilm, D.M. Grant, J. Am. Chem. Soc. 103 (1981)
2
913.
[11] R.K. Harris, K.J. Packer, A.M. Thayer, J. Magn. Reson.
2 (1985) 284.
17
capable of yielding reliable results for O NMR
tensor orientations. Since both experimental and
6
17
[12] R.M. Dickson, M.S. McKinnon, J.F. Britten, R.E. Wa-
sylishen, Can. J. Chem. 65 (1987) 941.
theoretical studies suggest that the O NMR tensor
orientations in primary amides are similar to those
in secondary amides, the present results for ben-
zamide are also relevant to the situation in a peptide
bond. The information obtained from the present
study provides a foundation for future solid-state
[
13] K. Eichele, R.E. Wasylishen, J. Magn. Reson. 106 (1995)
6.
[14] W.P. Power, R.E. Wasylishen, Ann. Rep. NMR Spectrosc.
3 (1991) 1.
4
2
[15] D. Massiot, I. Farnan, N. Gautier, D. Trumeau, A.
Trokiner, J.P. Coutures, Solid State Nucl. Magn. Reson. 4
17
O NMR studies of peptides and proteins.
(
1995) 241.
[
[
16] M. Broze, Z. Luz, J. Phys. Chem. 73 (1969) 1600.
17] R.E. Wasylishen, J.O. Friedrich, S. Mooibroek, J.B.
Macdonald, J. Chem. Phys. 83 (1985) 548.
Acknowledgements
[18] J.F. Baugher, P.C. Taylor, T. Oja, P.J. Bray, J. Chem.
Phys. 50 (1969) 4914.
We are grateful to the Natural Sciences Engi-
neering Research Council (NSERC) of Canada for
research and equipment grants. This research was
partially supported by a grant from the Advisory
Research Committee of Queen's University. G.W.
also thanks Queen's University for a Chancellor's
Research Award and the Government of Ontario
for a Premier's Research Excellence Award.
[
[
19] P.J. Chu, B.C. Gerstein, J. Chem. Phys. 91 (1989) 2081.
20] J.T. Cheng, J.C. Edwards, P.D. Ellis, J. Phys. Chem. 94
(1990) 553.
[
[
[
21] W.P. Power, R.E. Wasylishen, S. Mooibroek, B.A. Pettitt,
D. Danchura, J. Phys. Chem. 94 (1990) 591.
22] K. Eichele, J.C.C. Chan, R.E. Wasylishen, J.F. Britten, J.
Phys. Chem. A 101 (1997) 5423.
23] Q. Gao, G.A. Je€rey, J.R. Ruble, R.K. McMullan, Acta
Crystallogr. B 47 (1991) 742.