Solid-state 17O NMR and computational studies of C -nitrosoarene compounds

Article abstract of DOI:10.1021/ja909656w

We report the first solid-state ¹⁷O NMR determination of the ¹⁷O quadrupole coupling (QC) tensor and chemical shift (CS) tensor for four ¹⁷O-labeled C-nitrosoarene compounds: p-[ ¹⁷O]nitroso-N,N-dimethylaniline

Full text of DOI:10.1021/ja909656w

Published on Web 03/22/2010
Solid-State 17O NMR and Computational Studies of
C-Nitrosoarene Compounds
Gang Wu,*,† Jianfeng Zhu,† Xin Mo,† Ruiyao Wang,† and Victor Terskikh‡
Department of Chemistry, Queen’s UniVersity, 90 Bader Lane, Kingston, Ontario, Canada K7L
3N6, and Steacie Institute for Molecular Sciences, National Research Council Canada, Ottawa,
Ontario, Canada K1A 0R6
Received November 17, 2009; E-mail: gang.wu@chem.queensu.ca
Abstract: We report the first solid-state 17O NMR determination of the 17O quadrupole coupling (QC) tensor
and chemical shift (CS) tensor for four 17O-labeled C-nitrosoarene compounds: p-[17O]nitroso-N,N-
dimethylaniline ([17O]NODMA), SnCl2(CH3)2([17O]NODMA)2, ZnCl2([17O]NODMA)2, and [17O]NODMA·HCl.
The 17O quadrupole coupling constants (CQ) observed in these C-nitrosoarene compounds are on the order
of 10-15 MHz, among the largest values found to date for organic compounds. The 17O CS tensor in
these compounds exhibits remarkable sensitivity toward the nitroso bonding scheme with the chemical
shift anisotropy (δ11 - δ33) ranging from just 350 ppm in [17O]NODMA·HCl to over 2800 ppm in [17O]NODMA.
This latter value is among the largest 17O chemical shift anisotropies reported in the literature. These
extremely anisotropic 17O NMR interactions make C-nitrosoarene compounds excellent test cases that
allow us to assess the detection limit of solid-state 17O NMR. Our results suggest that, at 21.14 T, solid-
state 17O NMR should be applicable to all oxygen-containing organic functional groups. We also show that
density functional theory (DFT) calculations can reproduce reasonably well the experimental 17O QC and
CS tensors for these challenging molecules. By combining quantum chemical calculations with experimental
solid-state 17O NMR results, we are able to determine the 17O QC and CS tensor orientations in the molecular
frame of reference for C-nitrosoarenes. We present a detailed analysis illustrating how magnetic field-
induced mixing between individual molecular orbitals (MOs) contributes to the 17O shielding tensor in
C-nitrosoarene compounds. We also perform a Townes-Dailey analysis for the observed 17O QC tensors
and show that 17O CS and QC tensors are intrinsically related through the π bond order of the NdO bond.
Furthermore, we are able for the first time to examine the parallelism between individual 17O and 15N CS
tensor components in C-nitrosoarenes.
chemistry, and biochemistry.5-7 The C-nitroso compounds have
also found a wide range of pharmaceutical applications.8-10 In
the context of biological systems, C-nitrosoarenes (as well as
HNO) can form stable complexes with heme proteins such as
cytochrome P450, myoglobin and hemoglobin. Therefore,
C-nitrosoarene metal complexes are also of considerable interest.
Three basic binding modes are known for mononuclear C-
nitrosoarene metal complexes;6 see Figure 1. Although the κ1-
N-binding mode is the most common one, examples of κ1-O-
1. Introduction
Oxygen is one of the most common elements found in organic
and biological molecules. Oxygen-containing functional groups
can be found in all biologically important molecules such as
proteins, nucleic acids, carbohydrates, and phospholipids.
Oxygen is often present at the center of action in many
biological structures and processes. For example, oxygen-
containing functional groups are directly involved around the
catalytic center of many enzymes, at the ion binding sites of
metalloenzymes, and at the substrate-enzyme interaction
interfaces. Although the potential of 17O NMR spectroscopy was
recognized many years ago,1,2 the progress has been rather slow
until recently. In the past several years, significant effort has
been devoted to the experimental characterization of funda-
mental 17O quadrupole coupling (QC) and chemical shift (CS)
tensors for oxygen-containing organic functional groups.3,4
C-nitrosoarenes represent an important class of organic
compounds widely used in organic chemistry, coordination
(3) Lemaitre, V.; Smith, M. E.; Watts, A. Solid State Nucl. Magn. Reson.
2004, 26, 215–235.
(4) Wu, G. Prog. Nucl. Magn. Reson. Spectrosc. 2008, 52, 118–169.
(5) Gowenlock, B. G.; Richter-Addo, G. B. Chem. ReV. 2004, 104, 3315–
3340.
(6) Lee, J.; Chen, L.; West, A. H.; Richter-Addo, G. B. Chem. ReV. 2002,
102, 1019–1065.
(7) Cameron, M.; Gowenlock, B. G.; Vasapollo, G. Chem. Soc. ReV. 1990,
19, 355–79.
(8) Rice, W. G.; Schaeffer, C. A.; Harten, B.; Villinger, F.; South, T. L.;
Summers, M. F.; Henderson, L. E.; Bess, J. W.; Arthur, L. O.;
McDougal, J. S.; Orloff, S. L.; Mendeleyev, J.; Kun, E. Nature 1993,
361, 473–475.
† Queen’s University.
‡ Steacie Institute for Molecular Sciences.
(9) King, S. B., C-Nitroso compounds, oximes, N-hydroxyguanidines and
N-hydroxyureas. In Nitric Oxide Donors; Wang, P. G., Cai, T. B.,
Taniguchi, N., Eds.; Wiley-VCH: Weinheim, Germany, 2005; pp 177-
199.
(1) Klemperer, W. G. Angew. Chem., Int. Ed. 1978, 17, 246–254.
(2) Kintzinger, J. P., Oxygen N. M. R. Characteristic Parameters and
Applications. In NMR Basic Principles and Progress; Diehl, P., Fluck,
E., Kosfeld, R., Eds.; Springer-Verlag: Berlin, 1981; Vol. 17, pp 1-
64.
(10) Gooden, D. M.; Chakrapani, H.; Toone, E. J. Curr. Top. Med. Chem.
2005, 5, 687–705.
9
10.1021/ja909656w  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 5143–5155 5143

A R T I C L E S
Wu et al.
In the context of 17O NMR spectroscopy, C-nitrosoarenes are
known to exhibit extreme 17O chemical shifts, ca. δ(17O) ≈
1250-1550 ppm,14,15 primarily due to the presence of significant
paramagnetic shielding contributions arising from magnetic
field-induced mixing between low-lying excited states and the
ground state. As a result, the 17O chemical shift anisotropy is
also expected to be very large in C-nitrosoarene compounds.
In fact, the 15N chemical shift anisotropies observed in C-
nitrosoarene compounds are among the largest ever reported in
the literature.13,16,17 Although there is no literature report on
the 17O quadrupole coupling constant (CQ) for C-nitrosoarenes,
there are good reasons to believe that CQ(17O) must be rather
large for a nitroso group (R-NdO). For example, an aldehyde
functional group (H-CdO), being isoelectronic with a nitroso
group, typically has CQ(17O) values of about 10-12 MHz.18,19
Potentially, C-nitrosoarene compounds may pose a tremendous
challenge for solid-state 17O NMR studies, because a combina-
tion of very large 17O quadrupole coupling constants and very
large 17O chemical shift anisotropies could make it exceedingly
difficult to obtain high-quality 17O NMR data. Not surprisingly,
solid-state 17O NMR studies of C-nitrosoarenes have never been
reported in the literature. For precisely the same reasons, we
believe that C-nitrosoarenes can be used as a test case for
investigating the detection limit of solid-state 17O NMR spec-
troscopy. We are interested in exploring the detection limit of
solid-state 17O NMR in two aspects. One is the sensitivity limit
with respect to the largest molecular systems (e.g., proteins)
for which solid-state 17O NMR signals can be detected. The
other is concerned with the greatest anisotropic 17O NMR
interactions (e.g., QC and CS tensors) for which high-quality
solid-state 17O NMR signals can be recorded in practice with
the currently available ultrahigh magnetic fields. Recent studies
by Cross and co-workers20 and by Smith and co-workers21 have
to some extent touched upon the sensitivity issue. The present
study attempts to address the second aspect. To date, the largest
CQ(17O) values attainable by solid-state 17O NMR experiments
are on the order of 11 MHz, such as those found in nitrophe-
nols,22 carbohydrates,23 p-nitrobenzaldehyde,24 and sodium
pyruvate.25 These CQ(17O) values are still much less than the
theoretical upper limit, 20.9 MHz, which is the nuclear
quadrupole coupling constant expected for a single electron in
Figure 1. (a) Basic binding modes of monometallic C-nitrosoarene
complexes. Molecular structures of the C-nitrosoarene compounds inves-
tigated in this study: (b) NODMA · HCl; (c) ZnCl2(NODMA)2; (d)
SnCl2Me2(NODMA)2; and (e) NODMA.
binding and η2-NO-binding do exist in the literature. In
particular, κ1-O-binding mode has been seen in several com-
plexes containing Zn(II), Sn(IV), Fe(III), and Mn(III) metal
centers. For example, Richter-Addo and co-workers11 reported
a binding mode switch from κ1-N in a ferrous [(TPP)Fe-
(II)(ArNO)2] (TPP ) meso-tetraphenylporphyrinato) complex
to κ1-O in a ferric [(TPP)Fe(III)(ArNO)2]+ complex. It appears
that the κ1-O-binding mode is always associated with p-
aminosubstituted C-nirosoarenes in which a dipolar quinonoid
resonance structure has a considerable contribution. The biden-
tate η2-NO-binding mode has been observed in complexes
containing Mo(VI), W(VI), Ru(II), Pt(II), and Cu(I) metal
centers.6,12 In addition, many polynuclear complexes such as
[M2](ArNO) and [M4](ArNO) have also been reported in the
literature.
To explore the possibility of using solid-state 17O NMR as a
new spectroscopic probe of the chemical bonding in C-
nitrosoarenes and their metal complexes, we decided to deter-
mine the relevant 17O NMR tensors for the ArNdO functional
group, the first time such a study is undertaken. We chose to
investigate four representative C-nitrosoarenes consisting of a
parent C-nitrosoarene compound, its hydrochloride salt (where
the NdO group is protonated) and two diamagnetic κ1-O-bound
metal complexes; also see Figure 1. There are three major
reasons for selecting these C-nitrosoarene systems, especially
the two κ1-O-bound metal complexes, for this initial investiga-
tion. First, these compounds are diamagnetic and thus suitable
for NMR studies. Second, the κ1-O-binding mode should cause
the largest changes in 17O NMR properties as compared with
those for the parent compound. Third, similar C-nitrosoarene
compounds, in both κ1-N- and κ1-O-bound forms, have been
investigated previously by Oldfield and co-workers13 using solid-
state 15N NMR spectroscopy. Thus, a detailed comparison
between 17O (QC and CS) and 15N (CS) NMR tensors should
offer an excellent opportunity to obtain complementary informa-
tion about the chemical bonding at the NdO functional group.
the pure 2p atomic orbital of an oxygen atom.19 As for the 17
O
chemical shift anisotropy in organic molecules, the most
anisotropic 17O CS tensors so far measured have anisotropy (δ11
- δ33) on the order of 1000-3000 ppm, such as those found in
(14) Orrell, K. G.; Sik, V.; Stephenson, D. Magn. Reson. Chem. 1987, 25,
1007–1011.
(15) Dahn, H.; Pechy, P.; Flogel, R. HelV. Chim. Acta 1994, 77, 306–316.
(16) Lumsden, M. D.; Wu, G.; Wasylishen, R. E.; Curtis, R. D. J. Am.
Chem. Soc. 1993, 115, 2825–2832.
(17) Mason, J.; Larkworthy, L. F.; Moore, E. A. Chem. ReV. 2002, 102,
913–934.
(18) Flygare, W. H.; Lowe, J. T. J. Chem. Phys. 1965, 43, 3645–3653.
(19) Cheng, C. P.; Brown, T. L. J. Am. Chem. Soc. 1979, 101, 2327–2334.
(20) Hu, J.; Chekmenev, E. Y.; Gan, Z.; Gor’kov, P. L.; Saha, S.; Brey,
W. W.; Cross, T. A. J. Am. Chem. Soc. 2005, 127, 11922–11923.
(21) Wong, A.; Beevers, A. J.; Kukol, A.; Dupree, R.; Smith, M. E. Solid
State Nucl. Magn. Reson. 2008, 33, 72–75.
(22) Dong, S.; Yamada, K.; Wu, G. Z. Naturforsch. A 2000, 55, 21–28.
(23) Sefzik, T. H.; Houseknecht, J. B.; Clark, T. M.; Prasad, S.; Lowary,
T. L.; Gan, Z.; Grandinetti, P. J. Chem. Phys. Lett. 2007, 434, 312–
315.
(11) Wang, L.-S.; Chen, L.; Khan, M. A.; Richter-Addo, G. B. Chem.
Commun. 1996, 323–324.
(12) Wiese, S.; Kapoor, P.; Williams, K. D.; Warren, T. H. J. Am. Chem.
Soc. 2009, 131, 18105–18111.
(24) Wu, G.; Mason, P.; Mo, X.; Terskikh, V. J. Phys. Chem. A 2008,
112, 1024–1032.
(13) Salzmann, R.; Wojdelski, M.; McMahon, M.; Havlin, R. H.; Oldfield,
E. J. Am. Chem. Soc. 1998, 120, 1349–1356.
(25) Zhu, J.; Geris, A. J.; Wu, G. Phys. Chem. Chem. Phys. 2009, 11, 6972–
6980.
9
5144 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

Solid-State 17O NMR of C-Nitrosoarenes
A R T I C L E S
Bruker AXS Crystal Structure Analysis Package.32 Neutral atom
scattering factors were taken from Cromer and Waber.33 The crystal
is triclinic space group P1, based on the systematic absences, E
[17O2]picket fence porphyrin,26 metal nitrosyls,27 p-nitroben-
zaldehyde,24 and sodium pyruvate.25 Among these cases,
however, there have been only two instances so far, p-
nitrobenzaldehyde and sodium pyruvate, where very large 17O
quadrupole coupling constant and chemical shift anisotropy are
present simultaneously. As we argued earlier, C-nitrosoarenes
may represent an even more challenging case.
j
statistics, and successful refinement of the structure. The structure
was solved by direct methods. Full-matrix least-squares refinements
2
2
minimizing the function ∑w (Fo - Fc2) were applied to the
compound. All non-hydrogen atoms were refined anisotropically.
The hydrogen atoms were calculated, and their contributions were
included in the structure factor calculations. Convergence to final
R1 ) 0.0151 and wR2 ) 0.0374 for 1822 (I > 2σ(I)) independent
reflections, and R1 ) 0.0153 and wR2 ) 0.0375 for all 1839 (R(int)
) 0.0135) independent reflections, with 218 parameters and 0
restraints, were achieved. The largest residual peak and hole were
found to be 0.373 and -0.242 e/Å3, respectively.
2. Experimental Section
Synthesis. All common chemicals and solvents were purchased
from Sigma-Aldrich (Oakville, Ontario, Canada). Water (70% 17
O
atom) was purchased from isoSolutions (Ottawa, Ontario, Canada).
p-[17O]nitroso-N,N-dimethylaniline ([17O]NODMA)was prepared by
the following procedure. Approximately 300 mg of N,N-dimethy-
laniline was dissolved in 1 mL of concentrated hydrochloric acid
(33% HCl in 17O-enriched water) followed by slow addition of a
solution of 180 mg of sodium nitrite in 0.3 mL of 17O-enriched
water, while maintaining the solution at 5 °C and stirring. After
keeping the orange mixture at 5 °C for 1 h, solid NaOH was added
until the solution became bright green. The excessive 17O-enriched
water was then recovered on a vacuum line. The solid residue was
washed with 3 × 1 mL water. Recrystallization from 3 mL of 50%
ethanol aqueous solution yielded 219 mg of p-[17O]nitroso-N,N-
dimethylaniline (yield, 59%) as a dark-green polycrystalline solid.
p-[17O]nitroso-N,N-dimethylaniline hydrochloride monohydrate was
obtained by recrystallization of [17O]NODMA from 2 M HCl(aq).
ZnCl2(p-[17O]nitroso-N,N-dimethylaniline)2 and SnCl2(CH3)2(p-
[17O]nitroso-N,N-dimethylaniline)2 were prepared from [17O]N-
Quantum Chemical Calculations. Calculations of 17O (QC and
CS), 15N (CS) and 14N (QC) NMR tensors were performed using
density functional theory (DFT) methods as implemented in
Amsterdam Density Functional (ADF)34,35 and Gaussian03 (G03)36
programs. In the ADF calculations, Vosko-Wilk-Nusair (VWN)
exchange-correlation functional37 was used for the local density
approximation (LDA) and Perdew-Burke-Ernzerhof (PBE) ex-
change-correlation functional38 was applied for the generalized
gradient approximation (GGA). Standard Slater-type-orbital (STO)
basis sets with triple-ꢀ quality plus different numbers of polarization
functions (TZP and TZ2P) were used for all atoms except for Sn.
The QZ4P basis set was used for Sn. The relativistic (scalar) effect
was included using either zeroth-order regular approximation
(ZORA)39,40 or Pauli-type41 Hamiltonians. Mayer bond orders for
the NdO group were calculated using the keyword “EXTEND-
EDPOPAN” and the TZ2P basis set. In the G03 calculations, the
hybrid B3LYP exchange functional42,43 was used with standard
basis sets such as 6-311G(d,p), 6-311++G(d,p), 6-311++G(3df,3pd),
and cc-pVTZ. In both ADF and G03 shielding calculations, the
gauge-including atomic orbital (GIAO) approach44,45 was em-
ployed. All quantum mechanical calculations were performed on
Sun Fire E25K servers running Solaris 10, each with 72 dual-core
UltraSPARC-IV+ 1.5 GHz SMP processors and 576 GB of RAM.
Typically six processors were used for each calculation.
ODMA by the literature methods.28,29 Solution H and 13C NMR
1
spectra of these compounds were obtained to confirm the purity of
the synthesized products. The 17O enrichment level in the final
products was estimated to be approximately 55% using solution
17O NMR.
Solid-State 17O NMR. Solid-state 17O NMR spectra were
recorded at 11.74 and 21.14 T, operating at the 17O Larmor
frequencies of 67.78 and 122.02 MHz, respectively. At 11.74 T, a
4 mm MAS probe was used to obtain 17O NMR spectra for
stationary powder samples. At 21.14 T, a 3.2 mm MAS probe was
used in acquiring static spectra. For MAS experiments, a 2.5 mm
MAS probe was used with a sample spinning frequency of 35 kHz.
High power 1H decoupling was used in all cases. Other experimental
details are given in figure captions. Spectral simulations were
performed using WSOLIDS30 and DMFit31 simulation programs.
X-ray Crystallography. A red, block-shaped single crystal of
SnCl2(CH3)2(p-nitroso-N,N-dimethylaniline)2 (0.281 × 0.168 ×
0.166 mm3) was mounted on a glass fiber with grease and cooled
to -93 °C in a stream of nitrogen gas controlled with Cryostream
Controller 700. Data collection was performed on a Bruker SMART
APEX II X-ray diffractometer with graphite-monochromated Mo
KR radiation (λ ) 0.71073 Å), operating at 50 kV and 30 mA over
2θ ranges of 4.58-50.00°. No significant decay was observed
during the data collection. Data were processed on a PC using the
The principal components of the electric field gradient (EFG)
tensor, eqii (ii ) xx, yy, zz; |eqzz| > |eqyy| > |eqxx| and eqzz + eqyy
+
eqxx ) 0), were computed in atomic units (1 au ) 9.717365 ×
1021 V m-2). The principal components of the shielding tensor (σii)
were reported using the usual convention: σiso ) (σ11 + σ22 + σ33)/3
and σ33 > σ22 > σ11. In solid-state NMR experiments for quadrupolar
nuclei, the measurable quantities for a quadrupole coupling (QC)
(32) Bruker AXS Crystal Structure Analysis package: SHELXTL (Version
6.14), XPREP (Version2005/2), SAINT (Version 7.23A), APEX2
(Version 2.0-.2); Bruker AXS Inc.: Madison, Wisconsin, USA.
(33) Cromer, D. T.; Waber, J. T., International Tables for X-ray Crystal-
lography; Kynoch Press: Birmingham, U.K., 1974.
(34) ADF2006, SCM; Theoretical Chemistry, Vrije University; Amsterdam,
the Netherlands; http://www.scm.com.
(35) Te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra,
C.; Van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput.
Chem. 2001, 22, 931–967.
(36) Frisch, M. J.; et al. Gaussian 03, Revision C.02; Gaussian, Inc.:
Wallingford, CT, 2004.
(26) Oldfield, E.; Lee, H. C.; Coretsopoulos, C.; Adebodun, F.; Park, K. D.;
Yang, S. T.; Chung, J.; Phillips, B. J. Am. Chem. Soc. 1991, 113,
8680–8685.
(37) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200–
1211.
(38) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. ReV. Lett. 1996, 77,
3865–3868.
(27) Godbout, N.; Sanders, L. K.; Salzmann, R.; Havlin, R. H.; Wojdelski,
M.; Oldfield, E. J. Am. Chem. Soc. 1999, 121, 3829–3844.
(28) Hu, S.; Thompson, D. M.; Ikekwere, P. O.; Barton, R. J.; Johnson,
K. E.; Robertson, B. E. Inorg. Chem. 1989, 28, 4552–4554.
(29) Matsubayashi, G.-E.; Nakatsu, K. Inorg. Chim. Acta 1982, 64, L163–
L164.
(39) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
99, 4597–4610.
(40) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994,
101, 9783–9792.
(41) Schreckenbach, G.; Ziegler, T. Int. J. Quantum Chem. 1997, 61, 899–
918.
(30) Eichele, K.; Wasylishen, R. E. WSOLIDS1: Solid-State NMR
Spectrum Simulation, Version 1.19.11; Universita¨t Tu¨bingen: Tu¨bin-
gen, Germany2009.
(31) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calve, S.; Alonso,
B.; Durand, J.-O.; Bujoli, B.; Gan, Z.; Hoatson, G. Magn. Reson.
Chem. 2002, 40, 70–76.
(42) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(43) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. ReV. B 1988, 37, 785–789.
(44) Ditchfield, R. Mol. Phys. 1974, 27, 789–807.
(45) Wolinski, K.; Hilton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112,
8251–8260.
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5145

A R T I C L E S
Wu et al.
Figure 2. Experimental (lower trace) and simulated (upper trace) 17O MAS spectra (left column) and “total” MAS line shapes (right column) of (a)
[17O]NODMA·HCl, (b) ZnCl2([17O]NODMA)2, (c) SnCl2Me2([17O]NODMA)2, and (d) [17O]NODMA. All spectra were recorded at 21.14 T. The sample
spinning frequency was (a) 22 kHz and (b, c, d) 35 kHz. Detailed acquisition parameters were: (a) 3862 transients, 5 s recycle time; (b) 28000 transients,
2 s recycle time; (c) 36000 transients, 2 s recycle time; (d) 36148 transients, 2 s recycle time. The sharp peaks marked by * are due to ZrO2 rotor and a trace
amount of 17O-labeled NaNO2 in (c) and (d), respectively.
Table 1. Experimental 17O QC and CS Tensors Determined for C-Nitrosoarene Compounds
a
ηQ
compound
δiso (ppm)a
|CQ|(MHz)a
δ11 (ppm)b
δ22 (ppm)b
δ33 (ppm)b
R (deg)b
(deg)b
γ (deg)b
[17O]NODMA·HCl
ZnCl2([17O]NODMA)2
SnCl2Me2([17O]NODMA)2
[17O]NODMA
263 ( 5
600 ( 5
717 ( 5
1200 ( 5
10.8 ( 0.5
9.6 ( 0.5
10.5 ( 0.5
15.0 ( 0.5
0.4 ( 0.1
1.0 ( 0.1
0.9 ( 0.1
0.3 ( 0.1
450 ( 10
1260 ( 10
1450 ( 10
2900 ( 10
260 ( 10
480 ( 10
600 ( 10
750 ( 10
100 ( 10
60 ( 10
100 ( 10
100 ( 10
90 ( 10
6 ( 10
0 ( 10
18 ( 10
90 ( 2
88 ( 2
89 ( 2
90 ( 2
50 ( 10
74 ( 10
74 ( 10
83 ( 10
a Obtained from analysis of MAS spectra. b Obtained from analysis of static spectra.
tensor (ꢁii where ii ) xx, yy, zz; |ꢁzz| > |ꢁyy| > |ꢁxx| and ꢁzz + ꢁyy
+
3. Results and Discussion
ꢁxx ) 0) are nuclear quadrupole coupling constant (CQ) and
asymmetry parameter (ηQ). To compare calculated EFG tensors with
experimental QC tensors, following equations were used:
Analysis of 17O MAS spectra. Figure 2 shows the experimental
and simulated 17O MAS spectra for C-nitrosoarene compounds
at 21.14 T. In most cases, strong spinning sidebands were
observed even at a sample spinning frequency of 35 kHz,
indicating the presence of significant 17O chemical shift ani-
sotropy. For example, the spinning sidebands observed for
[17O]NODMA span a range of approximately 3000 ppm (∼350
kHz at 21.14 T). To analyze these “slow” MAS spectra, we
first obtained the “total” MAS line shape for each compound
by adding all the spinning sidebands onto the central band, as
also shown in Figure 2. From each “total” MAS line shape,
three 17O NMR spectral parameters (δiso, |CQ|, ηQ) can be
obtained in a straightforward fashion. Table 1 lists the results
from such an analysis for the four C-nitrosoarenes. It can be
CQ[MHz] ) e2qzzQ/h ) ꢁzz ) 243.96 × Q[barn] × eqzz[au]
(1)
ηQ ) (ꢁxx-ꢁyy)/ꢁzz ) (qxx-qyy)/qzz
(2)
where Q is the nuclear quadrupole moment, e is the elementary
charge, and h is the Planck constant. In this study, we used Q(17O)
) -2.558 and Q(14N) ) 2.044 barn (1 barn ) 10-28 m2).46 To
make direct comparison between the calculated shielding values,
σ, and the observed chemical shifts, δ, we used the established
shielding scales:
(46) Pyykko, P. Mol. Phys. 2008, 106, 1965–1974.
(47) Wasylishen, R. E.; Bryce, D. L. J. Chem. Phys. 2002, 117, 10061–
10066.
δ(ppm) ) σref(ppm)-σ(ppm)
(3)
where σref ) 287.5 ppm47 and 244.6 ppm48 for 17O and 15N nuclei,
(48) Jameson, C. J.; Jameson, A. K.; Oppusunggu, D.; Wille, S.; Burrell,
P. M.; Mason, J. J. Chem. Phys. 1981, 74, 81–88.
respectively.
9
5146 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

Solid-State 17O NMR of C-Nitrosoarenes
A R T I C L E S
seen immediately that the observed δiso values are drastically
different among the four C-nitrosoarene compounds, reflecting
the nature of chemical bonding around the nitroso (NdO) group.
It is quite remarkable to see that protonation of the nitroso group
causes a change in the 17O isotropic chemical shift of ∼1000
ppm! For the two O-bonded C-nitrosoarene metal complexes,
coordination of the nitroso group to a metal center has also
resulted in very large 17O isotropic chemical shift changes,
∼500-600 ppm. The values of CQ(17O) observed for the four
C-nitrosoarenes are between 9.6 and 15 MHz. These values are
among the largest so far experimentally measured by solid-state
17O NMR.4 As also seen from Figure 2, the observed spinning
sidebands can be reproduced reasonably well by spectral
simulations. However, as we have mentioned before,24 we
recommend to examine spinning sideband patterns only after
the 17O CS tensor components are obtained from analyses of
static 17O NMR spectra.
Analysis of 17O Static Spectra. To obtain further information
about the 17O CS tensor components and their relative orienta-
tions with respect to the 17O QC tensor, we performed 17O NMR
experiments for nonspinning (stationary) powder samples at two
magnetic fields, 11.74 and 21.14 T. The static 17O NMR spectra
shown in Figure 3 are of excellent quality, even though in some
cases the spectral widths exceed 300 kHz. Because the 17O
isotropic chemical shift (δiso) and quadrupole parameters (|CQ|
and ηQ) have been determined in the previous section, the
remaining spectral parameters to be determined are two
independent CS tensor components and three Euler angles (R,
, γ) that define the relative orientation between the CS and
QC tensors. In practice, we often use the tensor orientations
obtained from high-level quantum chemical calculations as a
starting point in spectral simulation. In most cases, we found
that these initial Euler angles do not require significant changes.
Under such a circumstance, one just needs to find the values
for two independent CS tensor components that would simul-
taneously fit the static 17O NMR spectra obtained at two
magnetic fields. A more detailed description about spectral
analysis for static 17O NMR spectra can be found in one of our
earlier publications.49 Complete solid-state 17O NMR results
obtained for the four C-nitrosoarene compounds are summarized
in Table 1. As expected, the large δiso value observed for
[17O]NODMA is associated with a remarkably large 17O
chemical shift anisotropy (δ11 - δ33 ) 2800 ppm). At the other
extreme, the 17O CS tensor in [17O]NODMA·HCl exhibits a
rather small chemical shift anisotropy, δ11 - δ33 ) 350 ppm.
We can also see from Table 1 that δ11 and δ22 tensor components
can change significantly, whereas δ33 values remain essentially
invariant. This trend is in line with the previous observations
for carbonyl compounds.24 As mentioned earlier, after we
obtained the 17O CS tensor components and their relative
orientations, we were able to simulate the spinning sideband
patterns in the MAS spectra; see Figure 2. The simulated
spinning sideband patterns match closely with the experimental
spectra, providing further confirmation about the reliability of
the solid-state 17O NMR parameters reported in Table 1.
At this point, it is important to point out that, because
SnCl2Me2(NODMA)2 and ZnCl2(NODMA)2 have ηQ ≈ 1, there
exist two sets of Euler angles that can produce rather similar
(but not identical) 17O NMR spectra. In the present cases, the
Figure 3. Experimental (lower trace) and simulated (upper trace) 17O NMR
spectra for stationary powder samples of (a) [17O]NODMA·HCl, (b)
ZnCl2([17O]NODMA)2, (c) SnCl2Me2([17O]NODMA)2, and (d) [17O]N-
ODMA obtained at 11.74 T (left column) and 21.14 T (right column).
Detailed experimental parameters are given below. For data collected at
11.74 T: (a) 6912 transients, 10 s recycle time; (b) 5198 transients, 10 s
recycle time; (c) 123602 transients, 1 s recycle time; (d) 35384 transients,
2 s recycle time. For data collected at 21.14 T: (a) 6563 transients, 2 s
recycle time; (b) 30578 transients, 2 s recycle time; (c) 35204 transients,
2 s recycle time; (d) 12631 transients, 2 s recycle time.
≈ 90°, and γ1 ) 90° - γ2. These two sets of Euler angles
describe such tensor orientations that, in one case, δ11 is roughly
along the direction of ꢁyy, whereas in the other case, δ11 is
roughly along the direction of ꢁzz. The origin of this ambiguity
is quite easy to understand. When a QC tensor has ηQ ≈ 1, ꢁzz
and ꢁyy components would have very similar magnitude but
opposite signs. Because NMR spectra are independent of the
absolute signs of the QC tensor components, these two sets of
Euler angles in fact describe essentially the same physical
situation, thus producing very similar static and MAS spectra.
This is a general problem often encountered whenever the QC
tensor under study has ηQ ≈ 1. Therefore, caution should be
exercised in deciding which set of the Euler angles is the correct
one (Vide infra).
Crystal Structure of SnCl2Me2(NODMA)2. Before we
present further 17O NMR results and quantum chemical com-
putations, in this section we describe a revised crystal structure
of SnCl2Me2(NODMA)2. The crystal structure of SnCl2-
Me2(NODMA)2 was first reported in 1982 by Matsubayashi and
Nakatsu29 in which the NdO bond length was found to be
two sets of Euler angles are related by R1 ) R2 ≈ 0°,
)
1
2
(49) Yamada, K.; Dong, S.; Wu, G. J. Am. Chem. Soc. 2000, 122, 11602–
11609.
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5147

A R T I C L E S
Wu et al.
Table 2. Calculated 17O (CS and QC), 15N (CS) and 14N (QC) NMR Tensors for C-Nitrosoarene Compounds. Experimental Results from
Solid-State 17Oa and 15Nb NMR and 14Nc NQR Measurements Are Also Listed for Easy Comparison
17O
15N
14N
compound
method
δiso (ppm) δ11 (ppm) δ22 (ppm) δ33 (ppm) CQ (MHz)f
ηQ
δiso (ppm) δ11 (ppm) δ22 (ppm) δ33 (ppm) CQ (MHz)
ηQ
NODMA
ADFd
1552
1523
1250
919
935
717
1001
968
600
272
263
3933
3784
2900
1927
2002
1450
2245
2136
1260
569
742
806
750
700
728
600
709
735
480
177
158
260
-18
-21
100
130
74
100
50
32
60
71
74
100
15.7 0.06
16.5 0.03
15.0 0.3
10.9 0.68
12.2 0.63
10.5 0.9
12.0 0.58
13.5 0.55
9.6 1.0
-11.5 0.31
-12.2 0.38
-10.8 0.4
918
943
802
679
772
596
772
813
583
502
516
429
2111
2153
1692
1382
1638
1174
1653
1744
1165
915
511
554
537
514
556
472
525
569
445
417
437
394
131
122
175
140
122
145
138
125
141
174
166
146
-6.391 0.479
-6.934 0.563
-5.825 0.488
-6.376 0.276
-7.262 0.430
G03/B3LYPe
(exptl)
SnCl2Me2(NODMA)2 ADFd
G03/B3LYPe
(exptl)
s
s
ZnCl2(NODMA)2
ADFd
-6.605 0.308
-7.205 0.444
G03/B3LYPe
(exptl)
s
s
NODMA·HCl
ADFd
-6.512 0.255
-6.910 0.158
G03/B3LYPe
(exptl)
558
450
946
747
270
s
s
a This work. b From ref 13. c From ref 50. d Computed with TZ2P basis set and a relativistic Pauli-type Hamiltonian. e Computed with the
6-311++G(3df,3pd) basis set for nonmetal atoms and the DZVP basis set for metal atoms. f The sign of experimental CQ was assumed to be the same
as the calculated ones.
1.218(4) Å. As Oldfield and co-workers13 noted previously, this
NdO bond length is unusually short and most likely results
from experimental errors. For this reason, we decided to re-
examine the crystal structure of this compound. New crystal-
lographic data and structural details for SnCl2Me2(NODMA)2
are provided in the Supporting Information. Although the unit
cell parameters determined by us are quite similar to those
reported by Matsubayashi and Nakatsu,29 we did observe a
significant disorder phenomenon that had not been mentioned
in the previous study. As shown in Figure S1 (Supporting
Information), the two symmetry-related NODMA molecules can
orient themselves in two different ways related by a two-fold
axis through the metal center. We found that the two orientations
have an occupancy ratio of about 3:1. The NdO bond lengths
are 1.296(4) and 1.283(15) Å for the major (75%) and minor
(25%) orientations, respectively. These NdO bond lengths are
significantly longer than 1.218 Å, but consistent with those
found in other nitroso-metal complexes, e.g., 1.304 Å in
ZnCl2(NODMA)2.28 These new crystallographic results suggest
that it was the unsolved disorder that had led to the erroneous
NdO bond length reported by Matsubayashi and Nakatsu.29 In
a broader context, significant crystallographic disorder is com-
monly observed in κ1-O-bonded C-nitroso complexes; many
reported NdO bond distances in the literature may need to be
re-examined.6 In the next sections, we use this correct crystal
structure to build molecular models of SnCl2Me2(NODMA)2
for quantum chemical calculations.
Calculations of 17O and 14/15N NMR tensors. In this section,
we present results of extensive DFT computations for C-
nitrosoarenes and their metal complexes. Because 15N CS and
14N QC tensors have been previously reported for these
C-nitrosoarene compounds in the literature,13,50 we decided to
examine 17O (QC and CS), 14N (QC) and 15N (CS) NMR tensors
in a unified fashion. This would allow us to fully assess the
quality of quantum chemical computations. We performed two
types of DFT calculations. One is to use the exchange functional
built in the ADF package and the other to employ the hybrid
exchange functional B3LYP as implemented in the G03 suite
of programs. We have performed calculations using a variety
of standard basis sets and different spin-orbit relativistic
methods. For NODMA, ZnCl2(NODMA)2 and NODMA ·HCl,
we used the crystal structures from the literature.28,51,52 For
SnCl2Me2(NODMA)2, we used the revised crystal structure
reported in the previous section. In this case, because the two
ligand orientations due to crystallographic disorder are associated
with slightly different molecular structures, we performed
calculations on both of them. We found that the computed results
are quite similar for the major (75% occupancy) and minor (25%
occupancy) components (within 1% and 8% for the QC and
CS tensor components, respectively). Therefore, we report only
those from the major component. Some representative results
are listed in Table 2; a complete list of results for all basis sets
and methods used in this study can be found in the Supporting
Information. Figure 4 shows comparison between experimental
and computed CS and QC tensor components for the four
C-nitrosoarene compounds. In general, ADF and B3LYP/G03
methods reproduce the experimental NMR data to a similar
degree of accuracy. For the 17O and 14N QC tensor components,
both ADF and B3LYP/G03 methods are able to reproduce the
experimental results within 10%. It is also well-known that even
better agreement can be achieved if “calibrated” Q values are
used.53-57 It is worth mentioning that the value of CQ(14N),
-5.825 MHz, reported for NODMA from a nuclear quadrupole
resonance (NQR) study,50 is also among the largest values
known for 14N nuclei.17,58-60 For the 17O and 15N CS tensors,
although calculations reproduce the general trend, they nonethe-
less overestimate the experimental chemical shift values by
approximately 40%. This is a general phenomenon that results
exclusively from overestimation of the paramagnetic shielding
contributions (in magnitude) in molecular systems by most
computational methods. On the other hand, C-nitrosoarenes
represent a very challenging class of molecules on their own
where substantial electron correlation effects in the NdO group
(51) Romming, C.; Talberg, H. J. Acta Chem. Scand. 1973, 27, 2246–
2248.
(52) Drangfelt, O.; Romming, C. Acta Chem. Scand. 1974, 28, 1101–1105.
(53) Eggenberger, R.; Gerber, S.; Huber, H.; Searles, D.; Welker, M. J.
Mol. Spectrosc. 1992, 151, 474–481.
(54) Ludwig, R.; Weinhold, F.; Farrar, T. C. J. Chem. Phys. 1996, 105,
8223–8230.
(55) Bailey, W. C. Chem. Phys. Lett. 1998, 292, 71–74.
(56) Dong, S.; Ida, R.; Wu, G. J. Phys. Chem. A 2000, 104, 11194–11202.
(57) Wu, G.; Hook, A.; Dong, S.; Yamada, K. J. Phys. Chem. A 2000,
104, 4102–4107.
(58) Lucken, E. A. C. Nuclear Quadrupole Coupling Constants; Academic
Press: New York, 1969.
(59) Edmonds, D. T. Phys. Rep. 1977, 29C, 233–290.
(50) Beepath, N.; Stephenson, D. Hyperfine Interact. 2004, 159, 137–141.
(60) Smith, J. A. S. Chem. Soc. ReV. 1986, 15, 225–260.
9
5148 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

Solid-State 17O NMR of C-Nitrosoarenes
A R T I C L E S
Figure 4. Comparison between experimental and calculated NMR tensor components for C-nitrosoarene compounds. (a) 17O and 14N QC tensor components
and (b) 17O and 15N CS tensor components.
are expected. Given that the current data set covers nearly the
entire chemical shift ranges for 17O and 15N nuclei, the computed
results shown in Figure 4 are indeed very encouraging. It is
particularly satisfying to see in Figure 4 that all 17O, 14N and
15N NMR tensors behave in a similar fashion, suggesting that
the current computational methods are valid. The observed
systematic discrepancy in the computed CS tensor components
for C-nitrosoarene compounds is somewhat larger than that seen
in carbonyl compounds.24,49,61-66 This is not surprising because
the 17O chemical shift anisotropy observed in a NdO group is
more than twice of that seen in a CdO group. Fortunately, once
this systematic discrepancy is identified, it is possible to make
appropriate corrections so that these computational methods can
yield more precise 17O CS tensors in other oxygen-containing
functional groups.
To further examine the dependence of 17O NMR tensors on
the binding mode of C-nitrosoarene metal complexes, we
performed quantum chemical calculations on the 17O NMR
tensors in Fe(TPP)(Py)(NODMA)27 and Pt(Ph3P)2(PhNO),67
which can be considered as representative models for κ1-N- and
η2-N-binding modes, respectively; see Figure 1. The computa-
tional results suggest that both 17O QC and CS tensors are
remarkably sensitive to the way that the nitroso group is bound
to the metal center. In particular, the value of δiso(17O) changes
drastically: κ1-N (1300 ppm) > κ1-O (1000 ppm) > η2-ON (400
ppm). Similarly, the value of CQ(17O) shows the following trend:
(61) Wu, G.; Yamada, K.; Dong, S.; Grondey, H. J. Am. Chem. Soc. 2000,
122, 4215–4216.
(62) Wu, G.; Dong, S.; Ida, R.; Reen, N. J. Am. Chem. Soc. 2002, 124,
1768–1777.
(63) Gervais, C.; Dupree, R.; Pike, K. J.; Bonhomme, C.; Profeta, M.;
Pickard, C. J.; Mauri, F. J. Phys. Chem. A 2005, 109, 6960–6969.
(64) Chekmenev, E. Y.; Waddell, K. W.; Hu, J.; Gan, Z.; Wittebort, R. J.;
Cross, T. A. J. Am. Chem. Soc. 2006, 128, 9849–9855.
(65) Waddell, K. W.; Chekmenev, E. Y.; Wittebort, R. J. J. Phys. Chem.
B 2006, 110, 22935–22941.
(66) Kwan, I. C. M.; Mo, X.; Wu, G. J. Am. Chem. Soc. 2007, 129, 2398–
2407.
(67) Pizzotti, M.; Porta, F.; Cenini, S.; Demartin, F.; Masciocchi, N. J.
Organomet. Chem. 1987, 330, 265–278.
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5149

A R T I C L E S
Wu et al.
Figure 5. Illustration of the orientations of the 17O QC (top) and CS
(bottom) tensors in the molecular frame of reference determined for (a)
NODMA, SnCl2Me2(NODMA)2, ZnCl2(NODMA)2 and (b) NODMA ·HCl.
In (a) and (b), δ33 is nearly perpendicular (<18°) and exactly perpendicular
to the C-NdO plane, respectively.
η2-ON (14 MHz) > κ1-N (13 MHz) > κ1-O (11 MHz). These
results strongly indicate that 17O NMR tensors can be used as
a sensitive probe of the binding mode in C-nitrosoarene metal
complexes.
Figure 6. Comparison of 17O (a) QC and (b) CS tensor orientations in
several related X ) O (X ) O, N, C) functional groups. Actual values of
the individual tensor components (CS in ppm and QC in MHz) are given
in the parentheses. From left to right: ozone, an oxyheme model, NODMA,
and p-nitrobenzaldehyde. Only one of the resonance structures is shown
for ozone and the oxyheme model to emphasize chemical bonding similarity
for the terminal oxygen.
NMR Tensor Orientations in the Molecular Frame. As
mentioned earlier, analyses of 17O NMR spectra can yield only
the relatiVe orientation between QC and CS tensors. To link
the observed NMR tensors to the molecular frame of reference,
we often rely on high-level quantum chemical calculations. As
discussed in the previous section, a higher degree of accuracy
can often be achieved in the calculation of 17O QC tensors.
Consequently, it is more reliable to use the computed 17O QC
tensor as an internal reference, which then allows us to place
experimentally determined 17O CS tensors into the molecular
frame of reference. For the C-nistrosoarenes under consideration,
we found that ꢁzz is invariantly along the direction of the
nonbonding (electron lone pair) orbital. In the cases of NODMA,
SnCl2Me2(NODMA)2 and ZnCl2(NODMA)2, this corresponds
to the direction lying in the C-NdO plane and perpendicular
to the NdO bond. For NODMA ·HCl, on the other hand, this
direction is within the NdO-H plane making an angle of
approximately 20° with respect to the NdO bond. These 17O
QC tensor orientations are depicted in Figure 5. It should also
be noted that the DFT calculations suggest that CQ > 0 in
NODMA, SnCl2Me2(NODMA)2, and ZnCl2(NODMA)2, but CQ
< 0 in NODMA·HCl. Once the QC tensor is fixed in the
molecular frame, we can use the experimentally determined
relative orientation between the QC and CS tensors (Euler angles
R, , γ) to decide on the CS tensor orientation. As shown in
Figure 5, the 17O CS tensors in NODMA, SnCl2Me2(NODMA)2,
and ZnCl2(NODMA)2 have essentially the same orientation such
that δ11 is almost along the NdO bond (within 7-16°). The
17O CS tensor for NODMA·HCl has a quite different orienta-
tion. In this case, δ11 is nearly along the O-H bond direction
making an angle of 70° with respect to the NdO bond. A
common feature among the 17O CS tensors of the four
C-nitrosoarenes is that δ33 is perpendicular to the C-NdO
plane. These tensor orientations are in good agreement with the
computed results.
model complexes in which they showed that the 17O CS and
QC tensor of the O2 ligand in an oxyheme model have the same
orientations as those in O3. Because an aldehyde group
(H-CdO) is isoelectronic with a nitroso group (-NdO), the
17O QC and CS tensors recently observed for p-nitrobenzalde-
hyde24 were also shown for comparison. Within the series of
functional groups shown in Figure 6, ꢁzz (or CQ) is invariantly
positive with its direction being in-plane and perpendicular to
the NdO or CdO bond. In fact, this is also true for most of
the carbonyl compounds.4 Because ꢁzz is positive, the other two
QC tensor components (ꢁyy and ꢁxx) must have negative values.
Close inspection of the QC tensors in these functional groups
reveals that, on going from O3 to aldehyde, as the magnitude
of ꢁzz decreases, the magnitude of the out-of-plane component
increases and the component along the X ) O bond decreases.
As a consequence, there is an apparent switch of QC tensor
orientation between nitroso and aldehyde groups. This is simply
due to the way that these two components are defined (i.e.,
|ꢁyy| > |ꢁxx|). Aside from this aspect, it is clear that the 17O QC
tensors in these functional groups are closely related. Quite
interestingly, these functional groups all have very similar 17
O
CS tensor orientation in the molecular frame, despite the fact
that individual CS tensor components are drastically different
in magnitude.
It is also interesting to note the 17O QC and CS tensors
observed for NODMA are similar to those predicted for the
terminal oxygen of O3 and for the O2 ligand in oxyheme model
complexes.70 In this context, there exists a long-standing
problem in the literature regarding the 17O NMR properties of
the O2 ligand in oxyheme complexes. Oldfield et al.26 reported
the first solid-state 17O NMR study of [17O2] bound to picket
fence porphyrin (an oxyheme model), hemoglobin and myo-
globin. Their spectral analyses for oxypicket fence porphyrin
To put the aforementioned NMR tensor orientations in
context, we show in Figure 6 what has been known about the
17O QC and CS tensors in several related functional groups.
The 17O QC tensor in ozone (O3) was determined from hyperfine
structures observed in rotational spectra.68,69 Several years ago,
Kaupp et al.70 reported an extensive DFT study of oxyheme
(68) Cohen, E. A.; Pickett, H. M. J. Mol. Struct. 1983, 97, 97–100.
(69) Cohen, E. A.; Hillig II, K. W.; Pickett, H. M. J. Mol. Struct. 1995,
352/353, 273–282.
(70) Kaupp, M.; Rovira, C.; Parrinello, M. J. Phys. Chem. B 2000, 104,
5200–5208.
9
5150 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

Solid-State 17O NMR of C-Nitrosoarenes
A R T I C L E S
yielded very large 17O chemical shift anisotropy but rather small
17O quadrupole coupling constant (CQ < 2 MHz). As pointed
out by Kaupp et al.,70 the reported solid-state 17O NMR
parameters are inconsistent with the solution 17O NMR results
reported by Gerothanassis and co-workers.71,72 To certain extent,
our solid-state 17O NMR results for the R-NdO group
indirectly support the calculations of Kaupp et al.70 in that the
17O quadrupole coupling constant for the terminal oxygen of
the O2 ligand in oxyheme complexes should be quite large, on
the order of 17 MHz. Of course, this value should be considered
as the rigid-lattice value for the O2 ligand in oxyheme systems.
Nonetheless, it appears that simple fast Fe-O2 axial rotation
motion as proposed by Oldfield and co-workers26,27 can explain
some but not all aspects of the experimental observations.
Further research is clearly needed in order to solve this puzzle.
At least, our new results demonstrate convincingly that it is
feasible to obtain high-quality solid-state 17O NMR spectra even
when CQ(17O) is as large as 15 MHz. It would be useful to re-
examine solid-state 17O NMR of some oxyheme models at
ultrahigh magnetic fields.
shielding term is inversely proportional to the energy gap
between the ground state and the excited state.
In the formulation implemented in the ADF software package,
σp is further partitioned into three different parts:74
σp ) σp(gauge) + σp(occ-occ) + σp(occ-vir)
(7)
where σp(gauge), σp(occ-occ) and σp(occ-vir) describe para-
magnetic shielding contributions from the gauge, coupling
between occupied and occupied MOs, and coupling between
occupied and virtual MOs, respectively. A summary of various
contributions to the 17O shielding tensor for C-nitrosoarenes is
provided in the Supporting Information (Table S8). Several
general trends are noted. First, σd is essentially isotropic (very
little dependence on molecular orientation) and exhibits very
little variation among different C-nitrosoarene molecules.
Second, σp(gauge) is generally very small, < 20 ppm. Third,
σp(occ-occ) has slightly larger values/and variations than does
σp(gauge), but generally does not exhibit any clear trend. Fourth,
σp(occ-vir) is the predominant term that accounts for both the
orientation dependence within each molecule and the large
variations among different molecules. Taking NODMA as an
example, σp(occ-vir) has a value of -4093 ppm along the NdO
bond (σ11) but only -107 ppm along the direction normal to
the C-NdO plane (σ33). In comparison, the corresponding
σp(occ-vir) values in NODMA·HCl are -704 and -169 ppm.
Because σp(occ-vir) is the most important source of para-
magnetic shielding contribution, we examine it further in order
to gain better understanding of the origin of the extreme 17O
shielding tensors observed in C-nitrosoarenes. It turns out that,
for C-nitrosoarenes, σp(occ-vir) is determined essentially by just
a few MO couplings out of a large number of possible MO
couplings, making it easier to discuss them in detail. From this
point further, we switch our terminology from “MO coupling”
to “magnetic field-induced mixing” to emphasis the physical
origin of σp(occ-vir). Table 3 lists the major contributors to
σp(occ-vir) along the directions of individual shielding tensor
components, together with the relevant energy gap, ∆E.
In the discussion that follows, we use NODMA as an example
to illustrate how magnetic field-induced mixing between indi-
vidual MOs contributes to the 17O shielding tensor. Our
discussion is generally within the framework of the Ramsey’s
theory (eqs 4-6) and at the same time makes use of the results
given by Jameson and Gutowsky.75 To learn more about the
general relationship between magnetic field-induced MO mixing
and a shielding tensor, the reader may consult an excellent
review written by Widdifield and Schurko.76 Figure 7 shows
the four most important MOs with regard to the origin of σp(occ-
vir). It is interesting to note that the same set of MOs also make
the largest contributions to σp(occ-vir) at the 15N nucleus within
the NdO group. As seen in Figure 7, because a 90° rotation of
the n orbital along the NdO bond would generate a significant
overlap with the π*(NO) orbital, this pair of MOs can be
“mixed” by the presence of a strong magnetic field along the
NdO bond. In addition, because the energy gap between the n
and π*(NO) MOs is very small (1.07 eV in NODMA), magnetic
field-induced mixing between these MOs would thus produce
a significant amount of σp(occ-vir), according to eq 6. In
Analysis of the 17O Shielding Tensors. In this section, we
present a detailed analysis of the 17O shielding tensors in
C-nitrosoarene compounds using a computational approach. This
analysis is performed in two steps. First, we examine shielding
contributions from different origins. Second, we analyze the
major paramagnetic shielding contributions from individual
molecular orbitals (MOs).
According to Ramsey’s theory of nuclear shielding,73 the total
shielding tensor at a nucleus can be divided into diamagnetic
and paramagnetic contributions:
σii ) σdii + σpii
(4)
where the subscript ii indicates the individual principal com-
ponents of the shielding tensor (i ) x, y, z). For i ) x, the two
shielding contributions can be written as:
µ0 e
4π
(y2j + z2j )
2
σdxx
)
〈0|
|0〉
(5)
∑
2
2me
rj3
j
〈0|
L r-3|k〉〈k|
L |0〉 + cc
∑
∑
xj
j
xj
2
µ0 e
4π
j
j
σpxx
)
∑
[
]
2
E0 - Ek
2me
k>0
(6)
where 〈0| is the ground-state electronic wave function, the sum
over k is over all excited electronic states, the sum over j is
over all electrons, Lx is the electron orbital angular momentum
operator, rj is the distance between the jth electron and the
nucleus of interest, cc indicates complex conjugate, E0 and Ek
are the energy values for the ground- and excited-states,
respectively, other symbols such as µ0, e, and me are standard
constants. Qualitatively, the diamagnetic shielding term is
dominated by the core electrons and consequently exhibits little
orientation dependence. On the other hand, the paramagnetic
shielding contribution is responsible for the anisotropic nature
of the shielding tensor. It is clear from eq 6 that the paramagnetic
(71) Gerothanassis, I. P.; Momenteau, M. J. Am. Chem. Soc. 1987, 109,
6944–6947.
(74) Schreckenbach, G.; Ziegler, T. J. Phys. Chem. 1995, 99, 606–611.
(75) Jameson, C. J.; Gutowsky, H. S. J. Chem. Phys. 1964, 40, 1714–
1724.
(72) Gerothanassis, I. P.; Momenteau, M.; Loock, B. J. Am. Chem. Soc.
1989, 111, 7006–7012.
(76) Widdifield, C. M.; Schurko, R. W. Concepts Magn. Reson., Part A
2009, 34, 91–123.
(73) Ramsey, N. F. Phys. ReV. 1950, 78, 699–703.
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5151

A R T I C L E S
Wu et al.
Table 3. Calculated Paramagnetic Shielding Contributions (in ppm) Arising from Magnetic Field-Induced Mixing between Occupied and
Virtual MOs, σp (occ-vir), along Three Principal 17O Shielding Tensor Components for C-Nitrosoarene Compoundsa
compound
MO mixing
σ11
σ22
σ33
∆E (eV)b
NODMA
σ(NO) f π*(NO)
n f π*(NO)
16
-249 (30%)
-99 (12%)
0
-163 (20%)
-17 (2%)
0
0
0
0
-1
1
-1
0
-43
8.95
1.07
4.14
9.13
1.63
4.69
8.90
1.50
4.32
10.5
2.94
6.11
-3649 (90%)
-327 (8%)
-1
-1868 (83%)
-105 (5%)
6
-1795 (78%)
-78 (3%)
-16
-275 (40%)
-60 (4%)
n f π*(NC)
SnCl2Me2(NODMA)2
ZnCl2(NODMA)2
NODMA·HCl
σ(NO) f π*(NO)
n f π*(NO)
n f π*(NC)
σ(NO) f π*(NO)
n f π*(NO)
n f π*(NC)
σ(NO) f π*(NO)
n f π*(NO)
n f π*(NC)
-176 (22%)
8 (1%)
-1
0
0
0
0
-72 (26%)
-77 (28%)
-18 (6%)
a Percentage contributions shown in parentheses are defined as σp(occ-vir)/σp(total) for each shielding tensor component. b Averaged values if several
nearly degenerate MOs exist.
individual shielding tensor components rather than the trace of
the shielding tensor that should be correlated to the electronic
transitions.
For the Zn and Sn complexes of NODMA, the O f M
bonding is essentially σ-donation from the ligand molecule. A
natural bond orbital (NBO) analysis suggests that there is very
little back-donation from the metal orbitals to the π* MO of
the NdO group. Therefore, it is not surprising that the two metal
complexes exhibit rather similar 17O NMR parameters.
Townes-Dailey Analysis of the 17O QC Tensors. In this
section, we describe results from a Townes-Dailey analysis78
of the observed 17O QC tensors for C-nitrosoarenes. For
NODMA, SnCl2Me2(NODMA)2, and ZnCl2(NODMA)2, we
used the expressions derived by Brown and co-workers79–81 for
treating N-O-containing compounds:
1
1
ꢁzz ) (2 - /2Pπ - /2Pσ)ꢁ0
(8)
(9)
1
ꢁyy ) (-1 - /2Pπ + Pσ)ꢁ0
1
Figure 7. Selected occupied and virtual MOs that have the largest
paramagnetic shielding contributions and their associated energy levels for
NODMA.
ꢁxx ) (-1 + Pπ - /2Pσ)ꢁ0
(10)
where ꢁzz and ꢁyy are the 17O QC tensor components that are
perpendicular and parallel to the NdO bond, respectively; ꢁxx
is the 17O QC tensor component normal to the C-NdO plane;
ꢁ0 ) +20.9 MHz, corresponding to the 17O quadrupole coupling
constant from a single electron in a pure 2p atomic orbital; Pπ
and Pσ are the orbital populations in the π and σ bonds of the
NdO group, respectively.
For NODMA ·HCl, because the oxygen atom of interest is
dicoordinated (NdO-H), a slightly different model has to be
used. If we assume that the oxygen atom in question has sp2
hybrid orbitals within the plane forming two σ bonds and a
lone pair and a pure 2p orbital normal to the NdO-H plane
contributing to a π bond, we can obtain the following expres-
sions for the 17O QC tensor components:58,82,83
comparison, the MO mixing between n and π*(NC) has the
same symmetry, but a much larger ∆E (3.80 eV) and, thus,
much smaller σp(occ-vir). Another factor is that π*(NC) is not
localized as much as π*(NO) around the oxygen atom. However,
as seen in Figure 7 and Table 3, the n f π* type of MO mixing
accounts for 98% of the total paramagnetic shielding along the
direction of σ11. In contrast, the σ f π* mixing is responsible
for σp(occ-vir) along the direction that is in-plane but perpen-
dicular to the NdO bond (the direction of σ22). As seen from
Table 3, the contribution from n f π* mixing decreases to 88%
in SnCl2Me2(NODMA)2, to 78% in ZnCl2(NODMA)2, and 44%
in NODMA ·HCl. This decrease is primarily due to the increase
in energy gap between the two MOs. As expected from eq 6,
σp depends linearly on 1/∆E for both n f π* and σ f π*
mixing, as illustrated in Figure S2 in the Supporting Inforamtion.
The similar slopes observed for the two types of mixing suggest
that the degrees of MO overlaps are similar for n f π* and σ
f π*. An approximately linear relationship between 17O
chemical shifts and lowest-energy electronic transitions mea-
sured in optical spectra was first observed in the early years of
17O NMR spectroscopy.77 However, caution should be exercised
if one uses this line of reasoning to interpret chemical shifts,
i.e., to link the observed isotropic chemical shift to a particular
electronic transition. In principle, as we showed here, it is the
ꢁzz ) [-1/2Pπ + /2(2 - s)Pσ + /2(s - 1)PLP]ꢁ0
1
1
(11)
(12)
1
1
ꢁyy ) [Pπ - /2(s + 1)Pσ + /2(s - 1)PLP]ꢁ0
ꢁxx ) [-1/2Pπ + /2(2s - 1)Pσ + (1 - s)PLP]ꢁ0 (13)
1
where ꢁzz and ꢁxx are the 17O QC tensor components that are in
the NdO-H plane with ꢁxx nearly bisecting the NdO-H angle;
ꢁyy is the 17O QC tensor component normal to the NdO-H
9
5152 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

Solid-State 17O NMR of C-Nitrosoarenes
A R T I C L E S
Table 4. Experimental N-O Bond Lengths, NBO Atomic Charges, Oxygen Orbital Populations from a Townes-Dailey Analysis of the
Experimental 17O QC Tensors, and Calculated Mayer Bond Orders for C-Nitrosoarene Compounds
Townes-Dailey analysis
ADF calculation
compound
rNO/Å
QO(e)
Pπ
Pσ
π bond ordera
Mayer bond order
NODMA
1.247
1.296
1.304
1.374
-0.3938
-0.4980
-0.5719
-0.4783
1.362
1.202
1.356
1.388
1.248
0.709
0.400
0.340
0.183
1.64
1.47
1.40
1.04
SnCl2Me2(NODMA)2
ZnCl2(NODMA)2
NODMA·HCl
1.639
1.694
1.835
a The π bond order ) (2 - Pπ)/0.9 as suggested by Brown and co-workers.79–81
plane; Pπ is the orbital populations in the NdO π bond; Pσ is
the averaged orbital population in the N-O and O-H σ bonds;
PLP is the orbital population for the electron lone pair within
the NdO-H plane (PLP ) 2); s is the s-character in the sp2
hybrid orbitals. In this study, we used s ) 0.406 for
NODMA·HCl, assuming that the NdO-H angle is 115°.
Applying eqs 8-13 to the experimental 17O QC tensor
components, we obtained orbital populations for C-nitrosoare-
nes; see Table 4. In general, on going from NODMA,
SnCl2Me2(NODMA)2, ZnCl2(NoDMA)2 to NODMA ·HCl, we
observe an increase in Pπ, thus a decrease in the π bond order.
This is in agreement with the observed trend in the NdO bond
length. Furthermore, the π bond order can be calculated from
Pπ using (2 - Pπ)/0.9 where the factor of 0.9 arises from a
consideration of the electronegativity difference between N and
O atoms, as described by Brown and co-workers.79–81 Table 4
also lists the calculated Mayer bond order for the four C-
nitrosoarenes. Quite interestingly, the Mayer bond orders are
in good correlation with the π bond orders. In fact, if we assume
that the σ bond order is approximately 1 in the NdO bond, the
total bond order (a simple sum of the π and σ bond orders) is
in excellent agreement with the calculated Mayer bond order.
For completeness, we also performed a Townes-Dailey
analysis of the 14N QC tensor experimentally determined for
NODMA.50 Assuming ꢁ0 ) -11 MHz for a single electron in
a pure 2p atomic orbital of the neutral nitrogen atom,84,85 we
obtained Pπ ) 1.099 and Pσ ) 1.272 for the NdO bond in
NODMA. If we use exactly the same electronegativity argument
as mentioned earlier, a π bond order of 0.819 can be calculated
using (2 - Pπ)/1.1. This value is in satisfactory agreement with
that derived from the 17O QC tensor as shown in Table 4. An
NBO analysis indicates that the oxygen atomic charge increases
on going from NODMA to its Sn and Zn complexes, consistent
with the picture that the κ1-O-binding mode is associated with
increased contributions from a dipolar quinonoid resonance
structure. The NBO analysis also confirms that the oxygen atom
in NODMA·HCl is involved in a quite different orbital
hybridization compared with those in the other three C-
nitrosoarene compounds investigated in this study.
14N chemical shifts and nuclear quadrupole coupling constants
in nitroso compounds. This is because both quantities depend
on the imbalance of orbital populations around the atom of
interest, as first pointed out by Saika and Slichter87 and later
elaborated by Karplus and Das.88 There are also several other
studies in the literature reporting on similar correlations.89-91
However, almost all the previous studies have focused only on
isotropic chemical shifts. A careful examination of the complete
CS and QC tensors under question will definitely provide new
insights into the origin of such correlations. Because the
imbalance of orbital populations is intrinsically related to the
bond order, it is expected that both 17O chemical shift and
nuclear quadrupole coupling constant are also correlated to the
bond order.1,92 Indeed, in a previous solid-state 15N NMR study
of C-nitrosoarenes, Oldfield and co-workers13 observed a linear
relationship between δ11 of the 15N CS tensor and the Mayer
bond order of the NdO group. In this study, we employ a
slightly different approach to examine a similar correlation
between 17O CS tensor components and the NdO bond order.
In particular, we have obtained the values of π bond order for
the C-nitrosoarenes from a Townes-Dailey analysis of the 17
O
QC tensors, as discussed in the previous section. These π bond
orders can be considered as “experimental values”. As seen from
Figure 8, when the 17O CS tensor components are plotted against
the π bond order for C-nitrosoarenes, a clear trend is observed.
That is, δ11 and δ22 are strongly correlated with the π bond
order, but δ33 is essentially invariant. As a result, δiso(17O) also
increases with the π bond order, as first noted by Kidd more
than 40 years ago.93 Recently, we reported a quite general
correlation between 17O CS tensor components and CQ for
carbonyl compounds.4,24 While direct comparison of CQ is valid
within a class of closely related molecules, the underlying
assumption is that these molecules have the same 17O QC tensor
orientation in the molecular frame of the CdO moiety. When
the 17O QC tensor orientation is different or the sign of CQ is
changed, which is the case for the four C-nitrosoarenes studied
here, it will not be very useful if one examines only the
magnitude of CQ. Therefore, it is much more important and
fundamentally correct to examine the nature of chemical bonding
byanalyzingthefull17OQCtensor(e.g.,usingtheTownes-Dailey
model). The trend observed in Figure 8 suggests that the 17O
CS and QC tensors in C-nitrosoarenes are intrinsically related
through the π bond order of the NdO bond. Moreover, the
Relationship between 17O QC and CS Tensors. More than
three decades ago, Mason86 demonstrated a correlation between
(77) Figgis, B. N.; Kidd, R. G.; Nyholm, R. S. Proc. R. Soc. London, A
1962, 269, 469–480.
(78) Townes, C. H.; Dailey, B. P. J. Chem. Phys. 1949, 17, 782–796.
(79) Cheng, C. P.; Brown, T. L. J. Am. Chem. Soc. 1980, 102, 6418–6421.
(80) Hiyama, Y.; Brown, T. L. J. Phys. Chem. 1981, 85, 1698–1700.
(81) Woyciesjes, P. M.; Janes, N.; Ganapathy, S.; Hiyama, Y.; Brown, T. L.;
Oldfield, E. Magn. Reson. Chem. 1985, 23, 315–321.
(87) Saika, A.; Slichter, C. P. J. Chem. Phys. 1954, 22, 26–28.
(88) Karplus, M.; Das, T. P. J. Chem. Phys. 1961, 34, 1683–1692.
(89) Bancroft, G. M.; Clark, H. C.; Kidd, R. G.; Rake, A. T.; Spinney,
H. G. Inorg. Chem. 1973, 12, 728–731.
(82) Brosnan, S. G. P.; Edmonds, D. T.; Poplett, I. J. F. J. Magn. Reson.
1981, 45, 451–60.
(90) Jones, C. H. W.; Sharma, R. D. J. Chem. Phys. 1987, 86, 5294–5299.
(91) Chung, S. C.; Chan, J. C. C.; Auyeung, S. C. F.; Xu, X. J. Phys.
Chem. 1993, 97, 12685–12690.
(83) Janes, N.; Oldfield, E. J. Am. Chem. Soc. 1986, 108, 5743–5753.
(84) Schirmacher, A.; Winter, H. Phys. ReV. A 1993, 47, 4891–4907.
(85) Tokman, M.; Sundholm, D.; Pyykko, P.; Olsen, J. Chem. Phys. Lett.
1997, 265, 60–64.
(92) Gerothanassis, I. P. Oxygen-17 NMR. In Encyclopedia of Nuclear
Magnetic Resonance; Grant, D. M., Harris, R. K., Eds. Wiley: New
York, 1996; Vol. 5, pp 3430-3440.
(86) Mason, J. J. Chem. Soc., Faraday Trans. 2 1976, 72, 2064–2068.
(93) Kidd, R. G. Can. J. Chem. 1967, 45, 605–608.
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5153

A R T I C L E S
Wu et al.
Figure 8. Relationship between 17O CS tensor components and π bond
order for C-nitrosoarenes.
different slopes observed for δ11 and δ22 reflect the influence
from ∆E. Clearly, the approach illustrated in Figure 8 can be
generalized and applied to other functional groups.
Parallelism between 17O and 15N CS Tensors. Previous
solution NMR studies have strongly suggested that there exists
a parallelism between 17O and 15N chemical shifts for nitroso
compounds (X-NdO).15,94 Now combining our solid-state 17
O
NMR data with the 15N CS tensors reported by Oldfield and
co-workers13 for C-nitrosoarenes, we have an opportunity for
the first time to be able to examine the parallelism between 17
O
Figure 9. (a) Correlation between 17O and 15N CS tensor components
observed for the four C-nitrosoarene compounds investigated in this study.
(b) Correlation between isotropic 17O and 15N chemical shifts determined
from both solid-state NMR (solid circles) and solution-state NMR (open
circles) experiments for nitroso compounds (X-NdO where X ) R, Ar,
SR, Halide, OR, NR2). Solution NMR data were obtained from refs 15 and
94. All 14/15N chemical shifts were referenced to that of liquid NH3, δ ) 0
ppm.
and 15N CS tensor components rather than just the isotropic
chemical shifts. Figure 9a illustrates the correlation between
17O and 15N CS tensor components for C-nitrosoarenes. Clearly,
the previously found correlation between isotropic 17O and
14/15N chemical shifts results exclusively from correlations
between only two of the CS tensor components, δ11 and δ22.
For completeness, Figure 9b also shows the correlation observed
for isotropic chemical shifts including both solid-state and
solution NMR data. These data can be fitted by a straight line
of slope 1.8. It is often argued in the literature that this slope of
1.8 is in excellent agreement with the expected ratio of 〈1/r3〉2p
between O and N atoms, i.e., 〈1/r3〉2p(O)/〈1/r3〉2p(N) ) 1.75.95
This type of rationalization stems from the use of the
Karplus-Pople equation to describe the paramagnetic shielding
contribution:96
when chemical shifts for different elements are compared as
demonstrated by Jameson and Gutowsky,75 several other factors
will need to be considered for comparing CS tensor components
between two nuclei within the same functional group (e.g., 15
N
and 17O in the NdO group). As seen from Figure 9, when the
11 and δ22 components are examined, the observed slopes, 2.62
δ
and 3.55, are larger than 1.75. This suggests that the other two
factors (∆E and ∑QAB) in eq 14 must be important. For the
C-nitrosoarenes under investigation, our calculations show that
the same set of MOs are responsible for both 15N and 17O
paramagnetic shielding contributions. Thus the ∆E term will
be identical when applying eq 14 to either 15N or 17O nuclei.
Under such a circumstance, the observed discrepancy in slopes
(2.62 versus 1.75 for δ11 and 3.55 versus 1.75 for δ22) must
reflect the subtle difference in ∑QAB. The fact that both the
observed slopes are greater than 1.75 suggests that changes in
∑QAB are larger for oxygen than for nitrogen. Because the
paramagnetic shielding contribution along the direction of δ33
is negligible, neither δ33(17O) nor δ33(15N) shows much variation
among different compounds.
2
µ0 e
4π
1
∆E
3
σp ) -
1/r 2p
Q
(14)
〈
〉
∑
AB
2
2me
where ∆E is the average electronic excitation energy and ∑QAB
is a summation of the charge density-bond order matrix
elements. According to eq 14, σp depends on 〈1/r3〉2p for nitrogen
and oxygen. However, although it is clear that 〈1/r3〉2p (as well
as 〈1/r3〉d) is indeed a valid measure of the chemical shift range
(94) Andersson, L.-O.; Mason, J. J. Chem. Soc., Dalton Trans. 1974, 202–
205.
(95) Barnes, R. G.; Smith, W. V. Phys. ReV. 1954, 93, 95–98.
(96) Karplus, M.; Pople, J. A. J. Chem. Phys. 1963, 38, 2803–2807.
9
5154 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

Solid-State 17O NMR of C-Nitrosoarenes
A R T I C L E S
4. Conclusion
tensors are remarkably sensitive to the binding mode of
C-nitrosoarene/metal complexes. The new results obtained in
this study on C-nitrosorenes and their metal complexes have
laid a solid foundation for future solid-state 17O NMR studies
of C-nitrosoarenes bound to globin proteins. Further research
in this direction is underway in our group.
In this study, we have used solid-state 17O NMR spectroscopy
to determine the 17O QC and CS tensors in four representative
C-nitrosoarene compounds. This is the first time that such
fundamental 17O NMR tensors have been measured for this
important class of organic compounds. The observed 17O QC
and CS tensors exhibit remarkable sensitivity toward the
chemical bonding scheme at the nitroso group (-NdO). The
17O quadrupole coupling constant and chemical shift anisotropy
observed in C-nitrosoarenes are among the largest values yet
measured for oxygen-containing organic functional groups by
solid-state 17O NMR. Our success in acquiring high-quality
solid-state 17O NMR spectra for this challenging class of
compounds provides strong evidence that all oxygen-containing
functional groups should be accessible by solid-state 17O NMR
spectroscopy at ultrahigh magnetic fields (e.g., 21 T). In addition
to reporting new experimental 17O NMR results, we have also
performed extensive DFT calculations. The computed 17O NMR
tensors are in reasonable agreement with the observed values.
We have presented a detailed analysis of various MO contribu-
tions to the shielding tensor at the oxygen nucleus. We have
analyzed the 17O QC tensors using the Townes-Dailey model
and found that the 17O CS and QC tensors in C-nitrosoarenes
are intrinsically related through the π bond order of the NdO
group. We have also examined the strong parallelism between
17O and 15N CS tensor components (δ11 and δ22). This paral-
lelism arises from the fact that the shielding at both oxygen
and nitrogen nuclei of the nitroso (NdO) moiety is dominated
by the magnetic field-induced mixing of the n f π* and σ f
π* nature. Our computational results also suggest that 17O NMR
Acknowledgment. This work was supported by the Natural
Sciences and Engineering Research Council (NSERC) of Canada.
Quantum chemical calculations were performed at the High
Performance Computing Virtual Laboratory (HPCVL) at Queen’s
University. We thank Dr. Hartmut Schmider for assistance in
performing ADF calculations. Access to the 900 MHz NMR
spectrometer was provided by the National Ultrahigh Field NMR
Facility for Solids (Ottawa, Canada), a national research facility
funded by the Canada Foundation for Innovation, the Ontario
Innovation Trust, Recherche Que´bec, the National Research Council
Canada, and Bruker BioSpin and managed by the University of
Ottawa (www.nmr900.ca). NSERC is acknowledged for a Major
Resources Support grant.
Supporting Information Available: Complete citation of
reference 36; detailed crystallographic and structural data for
SnCl2Me2(NODMA)2 (atomic coordinates, bond lengths, bond
angles, torsion angles, and hydrogen coordinates) and crystal-
lographic information file in CIF format; complete list of
computational results. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA909656W
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5155