Solid-State 17O NMR Investigation of the Carbonyl Oxygen Electric-Field-Gradient Tensor and Chemical Shielding Tensor in Amides

Article abstract of DOI:10.1021/ja0008315

We have presented a systematic experimental and theoretical investigation of the carbonyl oxygen electric-field-gradient (EFG) tensor and chemical shielding (CS) tensor in crystalline amides. Three 17O-labeled secondary amides, R1C[17O]-NHR2, have-been synthesized: benzanilide (1), N-methylbenzamide (2), and acetanilide (3). Analysis of 17O magic-angle spinning (MAS) and stationary NMR spectra yields not only the magnitude but also the orientation of the carbonyl 17O EFG and CS tensors. For compounds 1-3, the carbonyl 17O quadrupolar coupling constant (QCC) and the span of the chemical shift tensor are found to be in the range of 8.5-8.97 MHz and 560-630, ppm, respectively. The largest 17O EFG component lies in the amide plane and is perpendicular to the C=O bond, whereas the smallest component is perpendicular to the N-C=O plane. For the carbonyl 17O CS tensor, the principal component with the largest shielding, δ₃₃, is perpendicular to the amide plane, and the tensor component corresponding to the least shielding, δ₁₁, is in the amide plane approximately 20<*> off the direction of the C=O bond. Extensive quantum chemical calculations using density functional theory (DFT) have been performed for both isolated and hydrogen-bonded molecules of compounds 1-3. The calculated carbonyl 17O EFG and CS tensors from the latter molecular models are in reasonably good agreement with the experimental values. In particular, the B3LYP/D95^** EFG calculations overestimate the carbonyl 17O QCC by approximately 0.5 MHz. The B3LYP/D95^**/GIAO shielding calculations yield a linear correlation between the calculated and experimental data (slope=1.125 and R²=0.9952). The quantum chemical calculations indicated that the intermolecular C=O...H-N hydrogen-bonding interactions play an important role in determining the carbonyl oxygen EFG and CS tensors for an amide functional group.

Full text of DOI:10.1021/ja0008315

11602
J. Am. Chem. Soc. 2000, 122, 11602-11609
17
Solid-State O NMR Investigation of the Carbonyl Oxygen
Electric-Field-Gradient Tensor and Chemical Shielding Tensor in
Amides
Kazuhiko Yamada, Shuan Dong, and Gang Wu*
Contribution from the Department of Chemistry, Queen’s UniVersity, Kingston, Ontario, Canada, K7L 3N6
ReceiVed March 7, 2000. ReVised Manuscript ReceiVed July 11, 2000
Abstract: We have presented a systematic experimental and theoretical investigation of the carbonyl oxygen
electric-field-gradient (EFG) tensor and chemical shielding (CS) tensor in crystalline amides. Three 17O-labled
secondary amides, R1C[17O]-NHR2, have been synthesized: benzanilide (1), N-methylbenzamide (2), and
acetanilide (3). Analysis of 17O magic-angle spinning (MAS) and stationary NMR spectra yields not only the
magnitude but also the orientation of the carbonyl 17O EFG and CS tensors. For compounds 1-3, the carbonyl
17O quadrupolar coupling constant (QCC) and the span of the chemical shift tensor are found to be in the
range of 8.5-8.97 MHz and 560-630 ppm, respectively. The largest 17O EFG component lies in the amide
plane and is perpendicular to the CdO bond, whereas the smallest component is perpendicular to the N-Cd
O plane. For the carbonyl 17O CS tensor, the principal component with the largest shielding, δ33, is perpendicular
to the amide plane, and the tensor component corresponding to the least shielding, δ11, is in the amide plane
approximately 20° off the direction of the CdO bond. Extensive quantum chemical calculations using density
functional theory (DFT) have been performed for both isolated and hydrogen-bonded molecules of compounds
1-3. The calculated carbonyl 17O EFG and CS tensors from the latter molecular models are in reasonably
good agreement with the experimental values. In particular, the B3LYP/D95** EFG calculations overestimate
the carbonyl 17O QCC by approximately 0.5 MHz. The B3LYP/D95**/GIAO shielding calculations yield a
linear correlation between the calculated and experimental data (slope ) 1.125 and R2 ) 0.9952). The quantum
chemical calculations indicated that the intermolecular CdO‚‚‚H-N hydrogen-bonding interactions play an
important role in determining the carbonyl oxygen EFG and CS tensors for an amide functional group.
Introduction
directly available from solution NMR spectra. For 17O NMR,
the relevant NMR parameters are electric-field gradient (EFG)
tensors and chemical shielding (CS) tensors. Recent advances
in experimental solid-state NMR and computational chemistry
have made it possible to establish a correlation between NMR
tensors and molecular structures,5,6 but to date, most successful
solid-state NMR applications have focused only on spin-1/2
nuclei such as 13C and 15N.
High-resolution solid-state 17O NMR was pioneered by
Oldfield and co-workers.7-9 These authors demonstrated that
useful information can be extracted from central-transition
magic-angle spinning (MAS) 17O NMR spectra. However, the
spectral resolution of MAS spectra is severely limited by
incomplete averaging of the second-order quadrupole interac-
tions. Later, the advent of dynamic-angle spinning (DAS) and
double rotation (DOR) techniques made it possible to obtain
solid-state 17O NMR spectra with a site resolution similar to
those in spin-1/2 spectra.10-14 However, both DAS and DOR
Together with hydrogen, carbon, and nitrogen, oxygen is
among the most common elements in chemical substances.
Although 1H, 13C, and 15N nuclear magnetic resonance (NMR)
spectroscopy has made a tremendous impact on the development
of chemical and biological sciences, much less is known about
17O NMR of organic and biological compounds. This is not
surprising, because unfavorable nuclear properties of 17O (spin
) 5/2, natural abundance ) 0.037%, γ ) -3.6279 × 107 rad
T-1 s-1, Q ≈ - 2.558 × 10-30 m2) have rendered 17O NMR
experiments rather difficult. Nevertheless, solution 17O NMR
has been used for many years by organic chemists in the study
of small molecules.1-4 However, it is also clear that the rapid
quadrupole relaxation property of 17O has made solution 17O
NMR of biological macromolecules nearly impossible.
In principle, solid-state NMR spectra contain more informa-
tion than the spectra obtained from liquids. More specifically,
information regarding the orientation dependence of the fun-
damental NMR parameters (i.e., the tensorial property) is not
(5) Nuclear Magnetic Shieldings and Molecular Structure; Tossell, J.
A., Ed.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1993.
(6) Modeling NMR Chemical Shifts: Gaining Insight into Structure and
EnVironment; Facelli, J. C., de Dios, A. C., Eds.; ACS Symposium Series
732; Oxford University Press: Oxford, 1999.
(7) Schramm, S.; Kirkpatrick, R. J.; Oldfield, E. J. Am. Chem. Soc. 1983,
105, 2483.
(8) Schramm, S.; Oldfield, E. J. Am. Chem. Soc. 1984, 106, 2502.
(9) Walter, T. H.; Turner, G. L.; Oldfield, E. J. Magn. Reson. 1988, 76,
106.
(10) Chmelka, B. F.; Mueller, K. T.; Pines, A.; Stebbins, J.; Wu, Y.;
Zwanziger, J. W. Nature 1989, 339, 42.
* Corresponding author. Phone: 613-533-2644. Fax: 613-533-6669.
E-mail: gangwu@chem.queensu.ca.
(1) Klemperer, W. G. Angew. Chem., Int. Ed. Engl. 1978, 17, 246.
(2) Kintzinger, J. P. In NMR Basic Principles and Progress; Diehl, P.,
Fluck, E., Kosfeld, R., Eds.; Springer-Verlag: Berlin, 1981; Vol. 17, pp
1-65.
(3) McFarlane, W.; McFarlane, H. C. E. In Multinuclear NMR; Mason,
J., Ed.; Plenum Press: New York, 1987; Chapter 14, pp 403-416.
(4) Oxygen-17 NMR Spectroscopy in Organic Chemistry; Boykin, D.
W., Ed.; CRC Press: Florida, 1991.
10.1021/ja0008315 CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/10/2000

Carbonyl Oxygen Tensors in Amides
J. Am. Chem. Soc., Vol. 122, No. 47, 2000 11603
experiments are technically demanding and have not found
widespread applications in the study of organic and biological
compounds. The recent invention of multiple-quantum magic-
angle spinning (MQMAS) methodology15,16 has provided ex-
perimentalists with a new tool for studying half-integer qua-
drupolar nuclei and, consequently, triggered a new wave of
research interest in the field of quadrupolar NMR. Several
groups have already reported 17O MQMAS results.17-20 Not
surprisingly, all these 17O MQMAS studies focused on inorganic
systems in which the 17O quadrupole interactions are reasonably
small. Solid-state 17O NMR studies of organic and biological
compounds are still rare in the literature. Notable examples of
solid-state 17O NMR studies of biologically relevant systems
are those by Oldfield and co-workers on various heme protein
models.21
Scheme 1
of 17O enrichment and a high-field instrument, we are able to
obtain high quality solid-state 17O NMR spectra and extract
reliable information about 17O NMR tensors in amides.
Among the numerous oxygen-containing functional groups,
the secondary amide functional group, R1C(O)-NHR2, is of
particular importance. This functional group constitutes the
peptide bond of proteins. It is also present in nucleic acid
bases such as thymine (T), uracil (U) and guanine (G). The
CdO‚‚‚H-N hydrogen bond (HB) plays a major role in
determining the conformations of proteins and nucleic acids.
Ando and co-workers22 reported the first solid-state 17O NMR
investigation of peptides. They showed that 17O NMR param-
eters are a sensitive probe to the hydrogen-bonding interactions.
Unfortunately, the relatively low 17O enrichment level in their
polypeptide samples produced only low-quality 17O NMR
spectra, from which the authors reached an erroneous conclusion
concerning the orientation of the 17O CS tensor in the molecular
frame. Recently, we determined unambiguously the absolute
orientation of the carbonyl 17O CS tensor for a primary amide,
benzamide.23 We also demonstrated the importance of including
a complete network of intermolecular CdO‚‚‚H-N interactions
in the quantum chemical calculations of 17O NMR tensors. In
this contribution, we report a systematic solid-state 17O NMR
study of the carbonyl oxygen EFG and CS tensors for three
17O-labled secondary amides: benzanilide (1), N-methylben-
zamide (2) and acetanilide (3). We chose these secondary amides
as models for the peptide bond in proteins because the carbonyl
oxygen atom in each of the three amides is directly involved in
a CdO‚‚‚H-N hydrogen bond, a situation very similar to those
found in polypeptides and proteins. Using relatively high levels
Theoretical background
In this section, we present the theoretical background for the
analysis of 17O NMR spectra arising from stationary powder
samples. For a half-integer quadrupolar nucleus in a strong
magnetic field, both the chemical shielding and second-order
quadrupole interactions must be considered in order to calculate
the central-transition (+1/2, -1/2) NMR spectra. The resultant
NMR frequency can be written as a sum of three contributions:
(2)
ν ) ν0 + νCS + νQ
(1)
where ν0 is the Larmor frequency, νCS and νQ(2) are the chemical
shielding and the second-order quadrupolar interaction contribu-
tions, respectively.
The nuclear quadrupolar interaction arises from the coupling
between the nuclear quadrupole moment, Q, and the electric-
field gradient (EFG) at the nuclear position. The EFG can be
best described by a traceless second-rank tensor, whose principal
components are defined as |qzz| > |qyy| > |qxx|. In general, two
other quantities are often used to describe a quadrupolar coupling
tensor, namely, the quadrupole coupling constant (QCC), ø )
e2Qqzz/h, and the asymmetry parameter, η ) (qxx - qyy)/qzz. For
a half-integer quadrupolar nucleus (I > 1/2) in a strong magnetic
field, only the second-order quadrupolar interaction needs to
be considered for the central-transition (+1/2,-1/2) NMR
spectra. The frequency contribution from the second-order
(11) Wu, Y.; Sun, B. Q.; Pines, A.; Samoson, A.; Lippmaa, E. J. Magn.
Reson. 1990, 89, 296.
(12) Mueller, K. T.; Sun, B. Q.; Chingas, Q. C.; Zwanziger, J. W.; Terao,
T.; Pines, A. J. Magn. Reson. 1990, 86, 470.
(13) Mueller, K. T.; Wu, Y.; Chmelka, B. F.; Stebbins, J.; Pines, A. J.
Am. Chem. Soc. 1991, 113, 32.
(14) Grandinetti, P. J.; Baltisberger, J. H.; Farnan, I.; Stebbins, J. F.;
Werner, U.; Pines, A. J. Phys. Chem. 1995, 99, 12341.
(15) Frydman, L.; Harwood, J. S. J. Am. Chem. Soc. 1995, 117, 5367.
(16) Medek, A.; Harwood, J. S.; Frydman, L. J. Am. Chem. Soc. 1995,
117, 12779.
(2)
quadrupolar interaction, νQ , can be written as24
(2)
2
νQ ) (νq /12 ν0){3/2 sin2 θ[(A + B) cos2 θ - B] -
η cos 2 φ sin2 θ[(A + B) cos2 θ + B] + (η/6)[A - (A +
4B) cos2 θ - (A + B) cos2 2 φ(cos2 θ - 1)2]} (2)
with
(17) Wu, G.; Rovnyak, D.; Sun, B. Q.; Griffin, R. G. Chem. Phys. Lett.
1996, 249, 210.
(18) Dirken, P. J.; Kohn, S. C.; Smith, M. E.; van Eck, E. R. H. Chem.
Phys. Lett. 1997, 266, 568.
νq ) 3ø/2I(2I - 1)
(3)
(19) Wu, G.; Rovnyak, D.; Huang, P. C.; Griffin, R. G. Chem. Phys.
Lett. 1997, 277, 79.
A ) 3 - 4I(I + 1)
(4)
(5)
(20) Stebbins, J. F.; Xu, Z. Nature 1997, 390, 60.
(21) (a) Salzmann, R.; Kaupp, M.; McMahon, M. T.; Oldfield, E. J. Am.
Chem. Soc. 1998, 120, 4771. (b) McMahon, M. T.; de Dios, A. C.; Godbout,
N.; Salzmann, R.; Laws, D. D.; Le, H.; Havlin, R. H.; Oldfield, E. J. Am.
Chem. Soc. 1998, 120, 4784. (c) Godbout, N.; Sanders, L. K.; Salmann,
R.; Havlin, R. H.; Wojdelski, M.; Oldfield, E. J. Am. Chem. Soc. 1999,
121, 3829.
(22) (a) Kuroki, S.; Takahashi, A.; Ando, I.; Shoji, A.; Ozaki, T. J. Mol.
Struct. 1994, 323, 197. (b) Kuroki, S.; Ando, S.; Ando, I. Chem. Phys.
1995, 195, 107.
B ) 1/4[3/2 - 2I(I + 1)]
where θ and φ describe the orientation of the magnetic field in
(24) (a) Stauss, G. H. J. Chem. Phys. 1964, 40, 1988. (b) Amoureux,
J.-P.; Fernandez, C.; Granger, P. In Multinuclear Magnetic Resonance in
Liquids and Solids-Chemical Applications; Granger, P., Harris, R. K., Eds.;
Kluwer Academic Publishes: Dordrecht, The Netherlands, 1990; pp 409-
424. (c) Taulelle, F. In Multinuclear Magnetic Resonance in Liquids and
Solids-Chemical Applications; Granger, P., Harris, R. K., Eds.; Kluwer
Academic Publishes: Dordrecht, The Netherlands, 1990; pp 393-407.
(23) Wu, G.; Yamada, K.; Dong, S.; Grondey, H. J. Am. Chem. Soc.
2000, 122, 4215.

11604 J. Am. Chem. Soc., Vol. 122, No. 47, 2000
Yamada et al.
A(R, â, γ) )
cos γ cos â cos R - cos γ cos â sin R +
- cos γ sin â
sin γ sin R
-sin γ cos â cos R - - sin γ cos â sin R + sin γ sin â
cos γ sin R cos γ cos R
sin â cos R sin â sin R
sin γ cos R
cos â
(8)
Now the frequency contribution from the CS tensor can be
expressed as
νCS ) - ν0(σ11X′2 + σ22Y′2 + σ33Z′2)
(9)
with
X′ ) cos γ cos â C + sin γ D - cos γ sin â cos θ (10)
Y′ ) sin γ cos â C + cos γ D + sin γ sin â cos θ (11)
Z′ ) sin â C + cos â cos θ
(12)
Figure 1. (A) Initial alignment of the 17O EFG and CS tensors and
direction of the external magnetic field, B0. (B) Euler angles relating
EFG and CS tensors.
and
C ) sin θ cos φ cos R + sin θ sin φ sin R
D ) sin θ sin φ cos R - sin θ cos φ sin R
(13)
(14)
the principal-axis system (PAS) of the EFG tensor as shown in
Figure 1a.
For a chemical shielding (CS) tensor, we often use the
following convention for the three principal components
Combining Equations (1), (2), and (9), we can calculate 17O
NMR powder line shapes for stationary samples. Clearly, when
(R, â, γ) ) (0, 0, 0), the PAS of the CS tensor becomes
coincident with that of the EFG tensor. Under such circum-
stances, all of the above equations are reduced to those reported
by Baugher et al.27
σ11 e σ22 e σ33
(6)
Therefore, the chemical shielding contribution to the NMR
frequency is given by
Experimental and Computational Aspects
νCS ) -ν0 (σ11X2 + σ22Y2 + σ33Z2)
(7)
Sample Preparations. Benzoic acid-17O2 was prepared by reacting
0.3850 g (2.0 mmol) R,R,R-trichlorotoluene with 0.0720 g (4.0 mmol)
H2O* (50.8% 17O atom purchased from ISOTEC, Miamisburg, Ohio)
in a sealed tube at approximately 120 °C for 15 h. The resultant HCl
gas was removed after cooling the sealed tube to room temperature.
The remaining white crystalline products were recrystallized from
acetone/petroether. Yield: 93%. [17O]Benzoyl chloride was obtained
by refluxing thionyl chloride and benzoic acid-17O2 (molar ratio of 1.1:
1) for 1 h, and subsequently removing the excessive SOCl2 by
distillation. To [17O]benzoyl chloride was added dropwise a cold
phenylamine-methanol solution until the temperature was stabilized,
and the solution was stirred at room temperature for 3 h. The solution
was then poured into water, followed by extraction with diethyl ether.
The white product of [17O]benzanilide (1) (50.8% 17O atom) was
obtained upon removal of the solvent. [17O]N-methylbenzamide (2) was
synthesized by the same procedure as above except that a cold
methylamine-methanol solution was used in the last step of the
procedure. Acetic acid-17O2 was prepared by heating a mixture of 0.3
g (16.7 mmol) H217O and 0.35 g (8.5 mmol) acetonitrile in 8 mL
anhydrous dioxane at 80 °C for 12 h. The solution was saturated with
dry HCl gas before heating. Acetic acid-17O2 was converted to [17O]-
acetyl chloride by reaction with excessive POCl3. [17O]Acetyl chloride
was separated by distillation. [17O]Acetanilide (3) was obtained by
reacting [17O]acetyl chloride with excessive aniline in dioxane. Re-
crystallization yielded 140 mg of [17O]acetanilide (total yield ) 53%).
The melting point, solution NMR, and IR results for compounds 1-3
are in agreement with the literature values.
where X, Y, and Z are the directional cosines of the external
magnetic field in the PAS of the CS tensor. Note that the symbol
σ is used for absolute chemical shielding and δ for chemical
shift, δ11 g δ22 g δ33.
In general, the CS and EFG tensors have different orientations
in the molecular frame of reference. Thus, we must consider
the relative orientation between the two tensors using three Euler
angles (R, â, γ). In this procedure, the new directional cosines
(X′, Y′, Z′) of the CS tensor with respect to the EFG system are
used. Because the Euler angles describing the relative orientation
of two tensors depend critically on the choice of initial frame
and rotation sequence, confusion often arises in the literature
as to which definition was actually used in the study. This
sometimes makes it difficult to reproduce the reported tensor
orientations. There are several different definitions in the
literature concerning the Euler angles relating EFG and CS
tensors and the initial alignment of the two tensors.25 We believe
that the approach of Eichele et al.26 is the most straightforward
procedure. Thus, we follow their procedure in the present study.
The initial alignment of the two tensors is illustrated in Figure
1a. We can write the transformation matrix A(R, â, γ) as
(25) (a) Cheng, J. T.; Edwards, J. C.; Ellis, P. D. J. Phys. Chem. 1990,
94, 553. (b) Power, W. P.; Wasylishen, R. E.; Mooibroek, S.; Pettitt, B.
A.; Danchura, W. J. Phys. Chem. 1990, 94, 591. (c) Chu, P. J.; Gerstein,
B. C. J. Chem. Phys. 1989, 91, 2081. (d) Man, P. P. In Encyclopedia of
Nuclear Magnetic Resonance; Grant, D. M., Harris, R. K., Eds.; John Wiley
& Sons: Chichester, U. K., 1996; Vol. 6, pp 3838-3848.
(26) (a) Eichele, K.; Chan, J. C. C.; Wasylishen, R. E.; Britten, J. F. J.
Phys. Chem. A 1997, 101, 5423. (b) Eichele, K.; Wasylishen, R. E.; Nelson,
J. H. J. Phys. Chem. A 1997, 101, 5463.
Solid-State 17O NMR. All 17O NMR spectra were obtained on a
Bruker Avance-500 NMR spectrometer operating at 500.13 and 67.8
MHz for 1H and 17O nuclei, respectively. Polycrystalline samples were
packed into zirconium oxide rotors (4 mm o.d.). Typical sample
(27) Baugher, J. F.; Taylor, P. C.; Oja, T.; Bray, P. J. J. Chem. Phys.
1969, 50, 4914.

Carbonyl Oxygen Tensors in Amides
J. Am. Chem. Soc., Vol. 122, No. 47, 2000 11605
spinning frequencies were 12-15 kHz. For the stationary 17O NMR
experiments, a Hahn-echo sequence was used to eliminate acoustic
ringing from the probe. Typical recycle delays were between 2 and 10
s. A liquid sample of H2O (25% 17O atom) was used for RF power
calibration as well as for chemical shift referencing. Spectral simulations
were performed with the WSOLIDS program package (Klaus Eichele
and Rod Wasylishen, Dalhousie University).
Quantum Chemical Calculations. All quantum chemical calcula-
tions on 17O EFG and CS tensors were carried out using the Gaussian98
program28 on a Pentium II personal computer (400 MHz, 128 MB RAM,
12 GB disk space). The basis set of Dunning/Huzinaga double-ú,29
including polarization functions, D95**, and the B3LYP exchange
functional30 were employed. The Gauge-Included Atomic Orbital
(GIAO) approach31 was used for chemical shielding calculations. The
experimental geometry of acetanilide determined by a neutron diffrac-
tion study32 was used. For benzanilide and N-methylbenzamide, the
crystal structures obtained from X-ray diffraction studies were used.33,34
However, it is well-known that the N-H bond lengths measured by
X-ray diffraction are generally not accurate as compared to those
measured by the neutron diffraction technique. For this reason, the X-ray
N-H bond lengths for benzanilide and N-methylbenzamide were
corrected to a standard value, r(N-H) ) 1.030 Å.35
To make direct comparison between the calculated chemical shield-
ing, σ, and the observed chemical shift, δ, we used the absolute 17O
chemical shielding scale established by Wasylishen and co-workers36
Figure 2. Experimental (upper trace) and calculated (lower trace) 17
O
MAS NMR spectra for (A) [17O]benzanilide (1), (B) [17O]N-methyl-
benzamide (2), and (C) [17O]acetanilide (3). The peaks marked by an
asterisk (*) arise from the MAS rotor material, ZrO2.
δ ) 307.9 ppm - σ
(15)
For the quadrupole interaction, because the quantum chemical
calculations yield EFG tensor components, qii, in atomic units (a.u.),
the following equation was used to convert them to the quadrupolar
tensor components, øii, in MHz
we obtained Q(17O) ) -2.33 fm2 at the B3LYP/D95** level. This
value is in excellent agreement with the calibrated Q value at similar
DFT levels reported by De Luca et al.40
øii [MHz] ) e2Qqiih-1 ) -2.3496Q[fm2]qii [a.u.]
(16)
Results and Discussion
where Q is the nuclear quadrupole moment of the 17O nucleus (in fm2)
and the factor of 2.3496 results from unit conversion. In the literature,
the recommended Q value for 17O is -2.558 fm2.37 However, several
recent studies have demonstrated that, at different levels of theory, it
is more advantageous to use “calibrated” Q values, rather than the
standard Q value, to make direct comparison between the calculated
and experimental 17O QCCs.38-40 Following this calibration approach,
17O EFG and CS Tensors. Figure 2 shows the experimental
and simulated 17O magic-angle spinning (MAS) NMR spectra
for compounds 1-3. Each of the MAS spectra exhibits a typical
line shape arising from the second-order quadrupole interaction.
The large line width of approximately 10-12 kHz indicates
the presence of a sizable quadrupole coupling constant (QCC).
From the spectral simulation, we were able to obtain the 17O
isotropic chemical shift, the QCC, and the asymmetry parameter
for each of the three compounds. It should be noted that, because
the spinning sideband intensities are not completely negligible,
the simulated MAS line shapes shown in Figure 2 differ slightly
from the experimental spectra at the low-frequency end of the
line shape. The observed 17O isotropic chemical shifts for
compounds 1-3, ca. 300 ppm, are typical for amide oxygen
functional groups.4 The magnitude of the carbonyl 17O QCCs
found for compounds 1-3, ca. 8-9 MHz, is also consistent
with previous determinations for the amide oxygen.22,23,41 The
detailed 17O NMR results for compounds 1-3 are summarized
in Table 1.
To obtain the information about the 17O CS tensors, we
obtained the stationary 17O NMR spectra for compounds 1-3.
As shown in Figure 3, the stationary 17O NMR spectra exhibit
line shapes covering a frequency range of approximately 50
kHz, which is much larger than those of the MAS spectra. As
already mentioned, analysis of the 17O MAS spectra has yielded
the isotropic 17O chemical shift, QCC, and η. To analyze the
stationary 17O NMR spectra, we need to determine the remaining
five variables: two independent CS tensor components and three
(28) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K.
N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J.
V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.6; Gaussian, Inc.:
Pittsburgh, PA, 1998.
(29) Dunning, T. H.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., Ed.; Plenum: New York, 1976; Vol. 3, p 1.
(30) (a) Becke, A. D. Phys. ReV. 1988, A38, 3098. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. ReV. 1988, B37, 785. (c) Becke, A. D. J. Chem.
Phys. 1993, 98, 5648.
(31) (a) Ditchfield, R. Mol. Phys. 1974, 27, 789. (b) Wolinski, K.; Hilton,
J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112, 8257.
(32) Johnson, S. W.; Eckert, J.; Barthes, M.; McMullan, R. K.; Muller,
M. J. Phys. Chem. 1995, 99, 16253.
(33) Kashio, S.; Ito, K.; Haisa, M. Bull. Chem. Soc. Jpn. 1979, 52, 365.
(34) Leiserowitz, L.; Tuval, M. Acta Crystallogr. 1978, B34, 1230.
(35) (a) Taylor, R.; Kennard, O. Acta Crystallgr. 1983, B39, 133. (b)
Taylor, R.; Kennard, O.; Versichel, W. Acta Crystallogr. 1984, B40, 280.
(36) Wasylishen, R. E.; Mooibroek, S.; Macdonald, J. B. J. Chem. Phys.
1984, 81, 1057.
(37) Pyykko¨, P. Z. Naturforsch. 1992, 47a, 189.
(38) Eggenberger, R.; Gerber, S.; Huber, H.; Searles, D.; Welker, M. J.
Mol. Spectrosc. 1992, 151, 474.
(39) (a) Ludwig, R.; Weinhold, F.; Farrar, T. C. J. Chem. Phys. 1995,
103, 6941. (b) Ludwig, R.; Weinhold, F.; Farrar, T. C. J. Chem. Phys. 1996,
105, 8223.
(40) De Luca, G.; Russo, N.; Ko¨ster, A. M.; Calaminici, P.; Jug, K. Mol.
Phys. 1999, 97, 347.
(41) (a) Takahashi, A.; Kuroki, S.; Ando, I.; Ozaki, T.; Shoji, A. J. Mol.
Struct. 1998, 442, 195. (b) Yamauchi, K.; Kuroki, S.; Ando, I.; Ozaki, T.;
Shoji, A. Chem. Phys. Lett. 1999, 302, 331.

11606 J. Am. Chem. Soc., Vol. 122, No. 47, 2000
Yamada et al.
Table 1. Experimental 17O CS and Quadrupole Coupling Tensors of the Carbonyl Oxygen in Compounds 1-3a
deg
system
δiso
δ11
δ22
δ33
ø/MHz
η
R
â
γ
[17O]benzanilide (1)
320
287
330
580
520
570
450
380
440
-50
-40
-20
8.97
8.50
8.81
0.15
0.30
0.20
5
5
0
90
90
89
70
78
69
[17O]N-methylbenzamide (2)
[17O]acetanilide (3)
a All chemical shifts are in ppm and relative to liquid water. Errors in the experimental CS principal components and ø are ( 5 ppm and ( 0.02
MHz, respectively. The uncertainty in the Euler angles is estimated to be (4°.
Figure 4. Orientations of 17O EFG and CS tensors in the molecular
frame. Both δ33 and EFGX components are perpendicular to the amide
plane.
As depicted in Figure 4, the principal CS tensor component
with the strongest shielding, δ33, is approximately perpendicular
to the N-CdO plane. The tensor component corresponding to
the least shielding, δ11, lies in the N-CdO plane approximately
20° off the CdO bond direction. This observation is consistent
with the previous determination for benzamide.23 The qxx of the
17O EFG tensor lies perpendicular to the amide plane. The qzz
component lies in the N-CdO plane and is perpendicular to
the CdO bond. The experimental orientations of the 17O EFG
and CS tensors for compounds 1-3 are also in excellent
agreement with the results of several previous quantum chemical
calculations.42-44 As will be discussed in the next section, the
17O CS and EFG tensor orientations shown in Figure 4 are
confirmed by quantum chemical calculations.
Quantum Chemical Calculations. Another objective of the
present work is to evaluate the quality of quantum chemical
calculations for 17O EFG and CS tensors in hydrogen-bonded
systems. Although there has been a considerable amount of
theoretical work on 17O chemical shielding constants in small
Figure 3. Experimental (upper trace) and calculated (lower trace) 17
O
stationary NMR spectra for (A) [17O]benzanilide (1), (B) [17O]N-
methylbenzamide (2), and (C) [17O]acetanilide (3).
Euler angles that define the relative orientation between the EFG
and CS tensors. In general, it is quite difficult or impossible to
determine five unknowns from analyzing stationary 17O NMR
spectra obtained at a single field strength. The ideal technique
for obtaining complete formation about NMR tensors is the
single-crystal NMR experiment. However, large single crystals
suitable for NMR experiments are often difficult to obtain. In
the case of 17O NMR, the necessity of introducing 17O-lables
makes it even more difficult to grow large single crystals
because of the small quantity of labeled samples. Recently, we
have combined multiple-field 17O NMR experiments with
quantum chemical calculations and established the absolute
orientation of the 17O EFG and CS tensors for the carbonyl
oxygen in benzamide.23 Using the result for benzamide as a
starting point, we were able to analyze the experimental 17O
NMR spectra shown in Figure 3 in a straightforward fashion.
The magnitude and orientation of the carbonyl 17O EFG and
CS tensors for compounds 1-3 are presented in Table 1 and
illustrated in Figure 4. The spans of the carbonyl 17O chemical
shift tensors for compounds 1-3 are larger than that in
benzamide, 500 ppm. The larger spans in 1-3 result mainly
from opposite changes in δ11 and δ33. In particular, for
compound 1 (benzanilide), δ11 ) 570 ppm, and δ33 ) -50 ppm,
whereas for benzamide, δ11 ) 500 ppm, and δ33 ) 0 ppm.
As also seen from Table 1, the 17O CS tensors of compounds
1-3 show very similar orientations with respect to the corre-
sponding EFG tensors, that is, R ≈ 0°, â ≈ 90°, and γ ≈ 70°.
molecules,45-53 relatively little data are available concerning 17
O
chemical shift tensors, especially for hydrogen-bonded molecular
systems. We have recently shown that, to reproduce the
experimental 17O NMR results for an amide oxygen, it is
necessary to include a complete intermolecular HB network in
the quantum chemical calculations.23 The carbonyl oxygen in
each of the secondary amide systems studied here is directly
(42) Palmer, M. H. Z. Naturforsch. 1996, 51A, 442.
(43) Sitkoff, D.; Case, D. A. Prog. Nucl. Magn. Reson. Spectrosc. 1998,
32, 165.
(44) (a) Chesnut, D. B.; Phung, C. G. In Nuclear Magnetic Shieldings
and Molecular Structure; Tossell, J. A., Ed.; Kluwer Academic Publish-
ers: Dordrecht, The Netherlands, 1993; p 221. (b) Chesnut, D. B. Annu.
Rep. NMR Spectrosc. 1994, 29, 71.
(45) Kutzelnigg, W.; Fleischer, U.; Schindler, M. In NMR - Basic
Principles and Progress; Springer-Verlag: Berlin, 1990; Vol. 23, p 165.
(46) Chesnut, D. B.; Phung, C. G. J. Chem. Phys. 1989, 91, 6238.
(47) Kaupp, M.; Malkin, V. G.; Malkina, O. L.; Salahub, D. R. J. Am.
Chem. Soc. 1995, 117, 1851.
(48) (a) Schreckenbach, G.; Ziegler, T. J. Phys. Chem. 1995, 99, 606.
(b) Ruiz-Morales, Y.; Schreckenbach, G.; Ziegler, T. J. Phys. Chem. 1996,
100, 3359.
(49) (a) Gauss, J. Chem. Phys. Lett. 1992, 191, 614. (b) Gauss, J.; Stanton,
J. F. J. Chem. Phys. 1996, 104, 2574.
(50) Hansen, P. E.; Abildgaard, J.; Hansen, A. E. Chem. Phys. Lett. 1994,
224, 275.
(51) Barszczewicz, A.; Jaszunski, M.; Jackowski, K. Chem. Phys. Lett.
1993, 203, 404.
(52) Olah, G. A.; Burrichter, A.; Rasul, G.; Gnann, R.; Christe, K. O.;
Prakash, G. K. S. J. Am. Chem. Soc. 1997, 119, 8035.
(53) Kaupp, M.; Malkina, O. L.; Malkin, V. G. J. Chem. Phys. 1997,
106, 9201.

Carbonyl Oxygen Tensors in Amides
J. Am. Chem. Soc., Vol. 122, No. 47, 2000 11607
Similar to the 17O chemical shielding calculations, the dimer
models also produce improved 17O EFG results for compound
1-3. As shown in Table 2, inclusion of the hydrogen-bonding
interaction causes a reduction in the calculated 17O QCCs. This
is consistent with the general observation that hydrogen-bonding
interactions tend to reduce the QCC at the 17O nucleus.23,54-56
At the B3LYP/D95** level of theory, the calculated 17O QCCs
are uniformly smaller than the corresponding experimental
values by approximately 0.5 MHz. Again, the orientation of
the carbonyl 17O EFG tensor showed very little variation
between the monomer and dimer models, despite the clear
sensitivity of the magnitude of individual components to the
models (vide infra).
The Hydrogen-Bonding Effects. In the preceding section,
we demonstrated that the hydrogen-bonding interactions in
compounds 1-3 are important. Because CdO‚‚‚H-N hydrogen
bonds are ubiquitous in biological systems, it is, therefore, of
interest to examine the hydrogen-bonding effects on 17O NMR
tensors in more details. In this section, we construct a simple
model containing an N-methylacetamide molecule and a for-
mamide molecule (see Figure 7). The two molecules are
individually optimized in their geometry and are placed together
so that the two amide planes are coincident. Furthermore, the
O‚‚‚H-N hydrogen bond is assumed to be linear with the
CdO‚‚‚H angle being 120°. Such a model was used previously
by de Dios and Oldfield in the study of the 13C CS tensor as a
function of HB geometry.57 The major conclusion of their study
is that the CdO bond length change upon hydrogen-bonding
plays a major role in determining the 13C CS tensor for the
carbonyl carbon. Because the carbonyl oxygen atom is directly
involved in the hydrogen-bonding, it is expected that both the
lengthening of the CdO bond and the presence of the hydrogen-
bond donor would be important to the 17O NMR tensors. Our
recent quantum chemical calculations on the 17O NMR tensor
of an oxonium ion revealed that the presence of the hydrogen-
bond partner is, in fact, more important than the actual structural
change of the H3O+ ion itself.56 For these reasons, we decided
to investigate the dependence of the carbonyl 17O NMR tensors
on the CdO bond length and on the O‚‚‚N hydrogen-bond
distance separately. In the present work, we report the results
for the hydrogen-bond distance dependence. In the calculations,
the hydrogen-bond distance, r(O‚‚‚N), was varied (2.500,
2.800, 3.200, and 3.500 Å) and the hydrogen-bond angle
( CdO‚‚‚N) and the CdO bond length were kept constant at
120° and 1.230 Å, respectively. Our calculated results for the
carbonyl 13C CS tensor (data not shown) are in agreement with
those reported by de Dios and Oldfield,57 thus confirming the
validity of our model. Table 3 summarizes the results of the
17O EFG and CS tensor calculations for the N-methylacetamide/
formamide model.
Figure 5. Hydrogen-bonded dimers used in the quantum chemical
calculations of compounds 1 (A), 2 (B), and 3 (C).
involved in a single CdO‚‚‚H-N hydrogen bond and, therefore,
represents a good model for testing the performance of quantum
chemical calculations in handling weak hydrogen bonds. The
crystals of compound 1 are monoclinic in space group C2/c.33
The molecules are linked together by a CdO‚‚‚H-N hydrogen
bond to form a chain along the b axis. The O‚‚‚N distance is
3.112 Å. Compound 2 crystallizes in space group Pbca.34 The
hydrogen-bonded molecules are related to each other by a 21
axis with an O‚‚‚N hydrogen-bond distance of 2.93 Å. Com-
pound 3 is isostructural with compound 2 with similar unit cell
dimensions, for which the O‚‚‚N hydrogen-bond distance is
2.935 Å.32
To demonstrate the importance of including intermolecular
HB interactions, we performed two sets of calculations. First,
we calculated the 17O EFG and CS tensors for isolated molecules
(monomers) of compounds 1-3. In the second set of calcula-
tions, we included a neighboring molecule that serves as the
hydrogen-bond donor. The hydrogen-bonded molecular dimers
used in the quantum chemical calculations for compounds 1-3
are shown in Figure 5. The calculated carbonyl 17O CS and
EFG tensors are presented in Table 2. Figure 6 shows the
comparison between the experimental and calculated 17O CS
tensors for compounds 1-3. Clearly, the calculated 17O CS
tensor components for monomers deviate significantly from the
observed values. The largest deviations occur at δ11 components,
ca. 160 ppm. For the hydrogen-bonded dimers, the calculations
show a remarkable improvement. For example, the deviations
between the calculated and observed δ11 components are reduced
by 50%, ca. 80 ppm. As seen from Figure 6, the data from the
dimer models produce a slope of 1.125, which compares to
1.261 from the calculations for monomers. Interestingly, the
two models predict nearly identical orientations for the carbonyl
17O CS tensor. This seems to suggest that only the magnitude
of the individual components of the carbonyl 17O CS tensor is
sensitive to HB interactions. It should be emphasized that,
because the present amide systems exhibit relatively weak HB
interactions, it would be dangerous to generalize this observa-
tion. More studies are clearly required to reach a general
understanding about the effects of intermolecular HB interac-
tions on the orientation of the 17O CS tensor.
Figure 8 displays the dependence of both the 17O CS tensor
components and the isotropic chemical shift on the hydrogen-
bond distance, r(O‚‚‚N). All three tensor components changed
linearly with r(O‚‚‚N). Although δ11 and δ22 increased with
r(O‚‚‚N), δ33 exhibited an opposite trend. As a result, the
isotropic chemical shift increased linearly with the hydrogen-
bond distance. This is in qualitative agreement with previous
findings from solution 17O NMR studies that the stronger the
HB interaction, the more shielded the carbonyl oxygen environ-
(54) (a) Butler, L. G.; Cheng, C. P.; Brown, T. L. J. Am. Chem. Soc.
1981, 103, 2738. (b) Butler, L. G.; Brown, T. L. J. Am. Chem. Soc. 1981,
103, 6541.
(55) Gready, J. E. J. Phys. Chem. 1984, 88, 3497.
(56) Wu, G.; Hook, A.; Dong, S.; Yamada, K. J. Phys. Chem. A 2000,
104, 4102.
(57) de Dios, A. C.; Oldfield, E. J. Am. Chem. Soc. 1994, 116, 11485.

11608 J. Am. Chem. Soc., Vol. 122, No. 47, 2000
Yamada et al.
Table 2. Calculated (B3LYP/D95**) Carbonyl 17O CS and Quadrupole Coupling Tensors for Compounds 1-3a
deg
system
model
δiso
δ11
δ22
δ33
øzz
øyy
øxx
R
â
γ
benzanilide (1)
monomer
dimer
monomer
dimer
monomer
dimer
403
372
356
326
394
363
740
667
673
598
732
659
508
477
467
434
499
464
-39
-27
-71
-55
-50
-33
9.661
9.366
9.147
8.769
9.399
9.005
-6.634
-6.799
-5.611
-5.611
-5.529
-5.808
-3.087
-2.567
-3.536
-3.536
-3.870
-3.197
-2.4
2.0
2.0
-0.7
5.4
89.3
90.4
89.9
89.9
90.4
90.5
78.1
76.6
85.4
84.0
75.6
74.4
N-methylbenzamide (2)
acetanilide (3)
5.9
a All CS tensor components are in parts per million (ppm) and quadrupole coupling tensor components are in MHz.
Table 3. Calculated (B3LYP/D95**) Carbonyl 17O CS and
Quadrupole Coupling Tensors for the N-methylacetamide/
Formamide Complexa
O‚‚‚N
distance (Å) δiso δ11 δ22
δ33
øzz
øyy
øxx
2.500
2.800
3.200
3.500
336 589 416
2
1.487 -1.068 -0.419
340 607 435 -21 1.491 -1.028 -0.463
345 627 451 -43 1.512 -1.002 -0.509
348 639 459 -53 1.530 -0.994 -0.536
a All CS tensor components are in parts per million (ppm) and
quadrupole tensor components are in MHz.
Figure 8. Calculated 17O CS tensor components versus r(O‚‚‚N) for
the N-methylacetamide/formamide model.
Figure 6. Comparison between experimental and calculated 17O CS
tensors for (A) monomers and (B) dimers.
The discrepancy may be due to the semiempirical method and
a small basis set STO-6G used in the study of Kuroki et al. It
is also seen from Figure 8 that tensor components are ap-
proximately four times more sensitive to the hydrogen-bond
distance than is the isotropic chemical shift.
Figure 9 shows the dependence of the carbonyl 17O QCC
and asymmetry parameter on r(O‚‚‚N). Clearly, the carbonyl
17O QCC increases with r(O‚‚‚N). This is consistent with the
experimental observations for compounds 1-3. For example,
compound 1 exhibits the weakest hydrogen-bond strength, with
r(O‚‚‚N) ) 3.112 Å, and the largest 17O QCC, ø ) 8.97 MHz.
It is also interesting to note that because two CdO‚‚‚H-N
hydrogen bonds are present in benzamide (r(O‚‚‚N) ) 2.898
and 2.932 Å58) the carbonyl 17O QCC is rather small, ø ) 8.40
MHz.23 Figure 9a suggests that for the model used in this study,
the dependence of carbonyl 17O QCC on r(O‚‚‚N) is not linear.
In fact, there is no theoretical reason for this dependence to be
linear. Therefore, the previous observation of a linear relation-
ship between 17O QCC and r(O‚‚‚N) may be either accidental
or due to the small range of 17O QCC values.22a
Figure 7. Model consisting of N-methylacetamide and formamide
molecules linked by a CdO‚‚‚H-N hydrogen bond.
ment is. In a previous theoretical work, Kuroki et al. investigated
the HB effect on 17O NMR tensors in a model containing
N-acetyl-N′-methylglycine and two formamide molecules.22b
They showed that all three CS tensor elements increased with
the HB length. Our results are in general agreement with those
of Kuroki et al. except for the opposite trend found for δ33.
Another interesting finding is that the asymmetry parameter
also exhibits a correlation with the hydrogen-bond distance. As
(58) Gao, Q.; Jeffrey, G. A.; Ruble, J. R. Acta Crystallgr. 1991, B47,
742.

Carbonyl Oxygen Tensors in Amides
J. Am. Chem. Soc., Vol. 122, No. 47, 2000 11609
noticed that the asymmetry parameter for the 17O EFG tensor
of benzamide is even larger, η ) 0.37.23 From the data shown
in Table 3, it is clear that the observed dependence for η arises
from the fact that as r(O‚‚‚N) is changed, qzz and qxx change in
opposite directions, whereas qyy remains approximately constant.
This interesting observation in amides is consistent with the
early 17O NQR results reported by Cheng and Brown.59
Conclusions
We have presented a systematic investigation of 17O CS and
EFG tensors for secondary amide functional groups. The
experimental and theoretical results have indicated that both 17
O
CS and EFG tensors of an amide functional group are sensitive
to the CdO‚‚‚H-N hydrogen-bonding environment. In par-
ticular, δ11 and δ22 components increase with hydrogen-bond
distance, whereas δ33 exhibits an opposite dependence. For the
carbonyl 17O quadrupolar coupling tensor, both 17O QCC and
the asymmetry parameter depend on the hydrogen-bond dis-
tance. We have also shown that it is important to obtain high-
quality solid-state 17O NMR spectra in order to extract reliable
information about 17O NMR tensors. At the B3LYP/D95**
level, the quantum chemical calculations have reproduced the
experimental 17O QCC values within 0.5 MHz, whereas the
correlation between the calculated and experimental 17O CS
tensor components exhibits a slope of 1.125 (R2 ) 0.9952). The
benchmark experimental values reported in this study may be
useful in testing future quantum chemical calculations for
CdO‚‚‚H-N hydrogen-bonded systems. The implication of the
present study is that solid-state 17O NMR is potentially useful
as an additional nuclear probe in the study of proteins.
Acknowledgment. This research was supported by the
Natural Sciences and Engineering Research Council (NSERC)
of Canada. We wish to thank Dr. Klaus Eichele for providing
the WSOLIDS program and for stimulating discussions. We
are also grateful to Dr. Minhhuy Ho, Dr. Mike Lumsden, and
Prof. Rod Wasylishen for helpful discussions and to an
anonymous reviewer for very helpful comments.
Figure 9. Calculated 17O QCC (A) and asymmetry parameter (B)
versus r(O‚‚‚N) for the N-methylacetamide/formamide model.
seen from Figure 9b, the asymmetry parameter decreases with
the hydrogen-bond length. This prediction is confirmed by the
experimental data for compounds 1-3. Specifically, compound
1, which has a relatively long r(O‚‚‚N), exhibits a carbonyl 17
O
EFG tensor with η ) 0.15. The asymmetry parameters from
compounds 2 and 3 are 0.30 and 0.20, respectively. We also
JA0008315
(59) Cheng, C. P.; Brown, T. L. J. Am. Chem. Soc. 1979, 101, 2327.