Solution phase measurement of both weak σ and C-H...X - hydrogen bonding interactions in synthetic anion receptors

Article abstract of DOI:10.1021/ja8035652

A series of tripodal receptors preorganize electron-deficient aromatic rings to bind halides in organic solvents using weak σ anion-to-arene interactions or C-H...X^- hydrogen bonds. ¹H NMR spectroscopy proves to be a powerful techniq

Full text of DOI:10.1021/ja8035652

Published on Web 07/29/2008
Solution Phase Measurement of Both Weak σ and C-H · · ·X- Hydrogen
Bonding Interactions in Synthetic Anion Receptors
Orion B. Berryman,† Aaron C. Sather,† Benjamin P. Hay,‡ Jeffrey S. Meisner,† and
Darren W. Johnson*,†
Department of Chemistry and Materials Science Institute, UniVersity of Oregon, Eugene, Oregon 97403-1253, and
Chemical Sciences DiVision, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830-6119
Received May 21, 2008; E-mail: dwj@uoregon.edu
Electron-deficient arenes offer a variety of interaction motifs
complementing traditional anion binding strategies. We have shown
that at least three distinct binding motifs are possible.1,2 These motifs,
illustrated for Cl- complexes with 1,2,4,5-tetracyanobenzene (TCB)
are (i) the centered, noncovalent anion-π interaction (A), (ii) off-
center or “weak σ” interactions (B and C), and (iii) C-H· · ·X-
Figure 1. MP2/aug-cc-pVDZ optimized geometries for Cl- interactions with
hydrogen bonds (when acidic H’s are available, D) (Figure 1).
1,2,4,5-tetracyanobenzene include an unstable anion-π complex (A), weak σ
Although much recent attention has focused on the electrostatic
complexes (B and C), and an aryl H-bond complex (D).1 Atom colors: carbon
gray, hydrogen white, nitrogen blue, and chloride green.
anion-π interaction,3 there is evidence to suggest that this is not the
predominant binding mode for many highly electron-deficient arenes.2
In prior work, we found that strongly electron-deficient arenes,
including TCB, exhibit stable weak σ and H-bonded geometries B-D,
whereas the anion-π motif, A, was not a stable form in the solid and
gas phases.
The majority of prior computational studies on anion-arene
interactions have been representative of the molecules in the gas
phase. When moving from silico to solution, other factors need to
be considered;4–6 therefore, solution phase association constants
(Ka) must be measured to understand fully the selectivity that will
emerge in binding anions with electron-deficient arenes. In one case
Figure 2. Synthesis of tripodal anion receptors 1-3 (left) and ORTEP
representation of the crystal structure of receptor 1 (right, 50% probability
with hydrogens removed, carbon depicted as gray, hydrogen white, nitrogen
blue, and oxygen red).
Ka’s have been determined for model electron-deficient aromatic
rings from UV-vis titrations with halides.7 NMR spectroscopy can
provide complementary structural information that is not obtainable
1
with UV-vis spectroscopy. A handful of examples use H NMR
Table 1. Averagea Ka (M-1) for Receptors 1, 2, and 4b with
Halides
spectroscopy to characterize anion interactions with electron-
deficient aromatic rings in solution, but in many cases, additional
attractive interactions are present (such as π-stacking, ion pairing,
or hydrogen bonding).8
We present experimental and theoretical results on a series of neutral
tripodal receptors that utilize only electron-deficient arenes to bind
halides in solution (Figure 2). These receptors employ steric gearing
to preorganize electron-deficient arenes, and 1H NMR spectroscopic
titrations and DFT calculations confirm that receptors 1-3 bind anions
in a 1:1 stoichiometry (Table 1 and Figure 3).9 These studies highlight
the first designed receptor to quantitatively measure weak σ contacts
between anions and arenes utilizing only electron-deficient aromatic
rings.
Cl-
Br-
I-
c
c
c
1
2
4
26
53
18
35
11
26
<1d
<1d
<1d
a Average Ka reported from 2 to 3 titration experiments (not including
receptor 4). b All titrations were performed in C6D6 with receptor
concentrations of ∼2 mM; errors are estimated at (10%. A titration of
receptor
3
and NBu4+Br- in C6D6 (with slight heat to increase
solubility) yielded a Ka ) 12 M-1 c Tetra-n-heptylammonium halides
.
were used as the salt sources for each experiment, and titrations were
performed at 27 °C. d ∆δ for control receptor 4 were too small to
determine Ka’s.
Three key findings that have not been observed previously for anions
interacting with electron-deficient arenes in solution are emphasized:
(i) 1H NMR spectroscopy provides sensitive data for determining both
the magnitude of anion binding (Ka) and the structure (π contacts vs
hydrogen bonding), even for weak binding; (ii) this represents the first
observation of receptors binding anions in solution using only electron-
deficient aromatic rings with either weak σ or C-H· · ·X- hydrogen
bonding interactions;10 and (iii) these data provide the first quantitative
comparison of the relatiVe stabilities for such interactions in solution.
The cavity present in syn conformers of 2,4,6-trisubstituted 1,3,5-
triethylbenzene11 derivatives provides access for monatomic anions
to interact with the electron-deficient dinitroarenes via the π-system
or hydrogen bonding (Figure 2). Receptors 1-3 are structural isomers
composed of three electron-deficient arenes differing only in the
position of their nitro substituents, which allows for an understanding
of the effect of substitution pattern on receptor function. A key feature
of the design strategy is that receptor 2 cannot form hydrogen bonds
to anions (due to the bulky nitro groups being positioned ortho to each
acidic aryl hydrogen), allowing us to study only the interaction between
† University of Oregon.
‡ Oak Ridge National Laboratory.
9
10.1021/ja8035652 CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 10895–10897 10895

C O M M U N I C A T I O N S
Figure 3. 1H NMR spectra from titrations of receptor 1 (4.95 mM) with
NBu4+Br- highlighting the chemical shift changes for this system. The 1H
NMR spectrum of 1 is compared to spectra of 1 in the presence of 7, 31,
51, and 101 equiv of NBu4+Br- in ascending order.
Figure 4. Optimized geometries (B3LYP/DZVP) comparing the aryl C-H
hydrogen bonding mode of 1 (top) with the weak σ binding mode of 2
(bottom). Atom colors: carbon gray, hydrogen white, nitrogen blue, oxygen
red, and bromine rust.
the anion and the π-system. Receptor 4, lacking electron-deficient
arenes, was also prepared as a control.
Further evidence for the identity of these binding modes is provided
by DFT calculations.19 B3LYP/DZVP calculations were performed
on 1:1 complexes between Br- and single arenes from 1 and 2 that
possess the same substitution patterns but with methyl ester substituents
(Figure S2, Supporting Information). Interestingly, optimizations
starting from an idealized anion-π geometry, with the Br- anion
located directly over the arene centroid at a distance of 3.2 Å,2 in both
cases failed to yield the expected anion-π complex. Rather, the model
corresponding to 1 gave a H-bonded form (∆E ) -20.7 kcal/mol)20,21
and that of 2 gave a weak σ complex, with the Br- located above a
carbon atom ortho to the ester substituent (∆E ) -19.7 kcal/mol).
This behavior, consistent with that exhibited by other highly electron-
deficient arenes,1,2 suggests that the anion-π interaction is not a
preferred binding motif for the arene acceptors in 1 and 2.
Receptors 1-4 were synthesized in good yields from CsF-Celite-
assisted esterification of known 1,3,5-tris(bromomethyl)-2,4,6-trieth-
ylbenzene 5 with the corresponding benzoic acids (Figure 2, Supporting
Information).12 Interestingly, the structures determined from colorless
single crystals of 1 and 2 (Supporting Information)13,14 reveal they
crystallize with one electron-deficient ring anti with respect to the other
arenes (Figure 2). Nevertheless, receptors 1-4 exhibit time-averaged
C3V symmetry in solution on the NMR time scale, suggesting that the
up, up, down conformation does not dominate in solution.
A previous study of strongly electron-deficient aromatic rings
illustrated that UV-vis spectroscopy is a suitable method to determine
association constants for weak attractive interactions between halides
and electron-deficient aromatic rings.7,15 We chose to investigate the
1
utility of H NMR spectroscopy to determine association constants
The calculations of the model complexes forecast the results of
calculations on Br- complexes of 1 and 2. Geometries for these
complexes, confirmed to be minima by frequency calculations (Figure
4), are fully consistent with the NMR evidence. In 1, Br- forms
bifurcated H-bonds to each arene, with H · · ·Br- distances of 2.72 Å
(HA, Figure 3) and 3.18 Å (HB, Figure 3). In agreement with the
spectra, the shortest hydrogen bonding interaction yields the largest
chemical shift.22 In 2, where H-bonding to the arene is not possible
due to steric repulsions, Br- binds to the arenes via weak σ interactions,
with distances of 3.31 Å to the nearest carbon atom in each arene.
Lacking direct interaction with aryl hydrogen atoms, this binding motif
is consistent with the small chemical shifts observed in the 1H NMR
spectrum for 2·Br-.
between halides (Cl-, Br-, and I-) and electron-deficient arenes in
benzene, in part for the additional structural information provided by
this technique. It is necessary when performing titration experiments
to obtain data where there is maximal change in the binding isotherm.16
For solubility reasons, it was challenging to find an organic solvent
where subtle interactions could be measured and the anion concentra-
tion could reach a large excess of the receptor concentration. The low
solubility of NBu4+ - and NBu4+Cl- in C6D6 prompted us to perform
I
titration experiments with tetra-n-heptylammonium halide salts
(NHep4+Cl-, NHep4+Br-, and NHep4+I-) at 27 °C. All three
electron-deficient receptors 1-3 turned pale yellow upon addition of
Br- (see Table of Contents graphic, middle, picture exemplifies 90
equiv of tetra-n-butylammonium bromide, NBu4+Br-). Whereas 2
showed relatively little change in the 1H NMR spectrum, significant
changes occurred with receptors 1 and 3 when NHep4+Br- was titrated
into C6D6 solution of receptors (3 was poorly soluble which prevented
acquiring quantitative data). Association constants determined for
Further solution studies, conducted with NHep4+Cl- and NHep4+ -
I
(Supporting Information), yield results that further support conclusions
obtained from the Br- studies. Whereas the 1·Cl- complex is colorless
in C6D6, the 2·Cl- complex presents a pale yellow color in solution.
An orange color change was observed for all receptors 1-3 with
NHep4+I- (Table of Contents graphic, right, demonstrates the color
observed when 84 equiv of NHep4+I- are present at 27 °C).23
Analogous to the Br- titration experiments in C6D6, significantly larger
chemical shift changes were observed for 1 over those of 2, again
consistent with the fact that 1 can form aryl CH · · ·X- H-bonds while
2 cannot. Titrations of 1 and 2 with NHep4+Cl- exhibited the largest
receptors 1 and 2 measured 18-35 M-1 17
while control receptor
,
4sdistinctly lacking electron-deficient arenessexhibited no measurable
binding by NMR and no visible color change (Table 1). These results
lend support to our hypothesis that electron-deficient aromatic rings
are required to bind anions in this neutral system.
Titrations of receptors 1 and 3 with NBu4+Br- displayed changes
in chemical shifts of over 1 ppm,18 and in the case of receptor 1 with
NBu4+Br-, the peak juxtaposition even changed over the course of
the titrations (Figure 3).10 Conversely, 3,5-dinitro-substituted receptor
2 exhibited much smaller chemical shift changes (maximum of 0.038
ppm for NHep4+Br-), but binding constants were determined to be
on the same order of magnitude. The striking differences in ∆δ for
receptors 1 and 3 versus that for receptor 2 indicate that different
binding modes are occurring in solution for these receptors.
1
Ka values (ranging from 26 to 53 M-1, Table 1). For I-, H NMR
titrations at 27 °C reveal association constants ranging from 11 to 26
M-1. As with Br-, control receptor 4 fails to form colored complexes
with Cl- or I- and exhibits no measurable association in C6D6.
Through a series of receptors utilizing only electron-deficient arenes
to bind anions, we have shown that 1H NMR spectroscopy is a practical
means to measure these subtle interactions in solution. DFT calculations
and 1H NMR titrations establish that the nitro group substitution pattern
9
10896 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

C O M M U N I C A T I O N S
(8) For solution examples of anions interacting with electron-deficient arenes,
see: (a) Berryman, O. B.; Hof, F.; Hynes, M. J.; Johnson, D. W. Chem.
Commun. 2006, 506. (b) Gil-Ramirez, G.; Escudero, E. C.; Benet-Buchholz,
B.; Ballester, P. Angew. Chem., Int. Ed. 2008, 44, 4114. (c) See ref 7. (d)
Lakshminarayanan, P. S.; Ravikumar, I.; Suresh, E.; Ghosh, P. Inorg. Chem.
2007, 46, 4769. (e) Fairchild, R. M.; Holman, K. T. J. Am. Chem. Soc.
2005, 127, 16364. (f) Maeda, H.; Osuka, A.; Furuta, H. J. Inclusion Phenom.
Macrocycl. Chem. 2004, 49, 33. (g) Gorteau, V.; Bollot, G.; Mareda, J.;
Perez-Velasco, A.; Matile, S. J. Am. Chem. Soc. 2006, 128, 14788. (h)
Schneider, H. J.; Werner, F.; Blatter, T. J. Phys. Org. Chem. 1993, 6, 590.
(i) For a recent solid state example, see: Albrecht, M.; Wessel, C.; Groot,
M.; Rissanen, K.; Luchow, A. J. Am. Chem. Soc. 2008, 130, 4600.
(9) Job plots measured at concentrations lower than the dissociation constants
involved are inherently unreliable; see: Huang, C. Y.; Zhou, R.; Yang,
D. C. H.; Chock, P. B. Biophys. Chem. 2003, 100, 143. Unfortunately, this
system is limited to combined concentrations of ∼20 mM. However, fitting
the isotherms of all 16 titrations to higher order binding stoichiometries
resulted in Ka’s that were highly dependent on starting estimates.
Furthermore, these fits were not reproducible among experiments run in
duplicate or triplicate.
is critical to the binding mode adopted by the receptor. The 2,4- and
3,4-substitution patterns in 1 and 3 engender the interaction with anions
through aryl C-H· · ·X- hydrogen bonding, while the 3,5-substitution
pattern in 2 promotes weak σ interactions. With two NO2 and one
ester substituent, these highly electron-deficient arenes adopt binding
motifs of weak σ and aryl H-bonding instead of the anion-π motif.
The differences in binding modes between isomeric receptors 1 and 2
have allowed us to quantify for the first time distinction between aryl
H-bonds and anion/arene π contacts, which are in conflict when acidic
aryl hydrogens are present. Receptors 1 and 2 exhibit the strongest
interactions with Cl- followed by Br- and I-, and larger association
constants are observed when the halide is restricted to interact solely
through contacts to the π-system (receptor 2). Does this approach hint
at an emerging selectivity for anion binding in solution using electron-
deficient arenes? Receptors that exhibit larger binding constants with
more striking differences will need to be studied to further address
this issue.
(10) 1H NMR spectroscopy has been used previously to demonstrate C-H· · ·X-
hydrogen bonds in solution; see: (a) Li, Y.; Flood, A. H. Angew. Chem.,
Int. Ed. 2008, 47, 2649. (b) Juwarker, H.; Lenhardt, J. M.; Pham, D. M.;
Craig, S. L. Angew. Chem., Int. Ed. 2008, 47, 3740. For other examples of
complementary C-H· · ·X- hydrogen bonding modes in receptors that bind
anions using traditional H-bonds (c-g) or electrostatic attractions (h,i), see:
(c) Lee, C.-H.; Na, H.-K.; Yoon, D.-W.; Won, D.-H.; Cho, W.-S.; Lynch,
V. M.; Shevchuk, S. V.; Sessler, J. L. J. Am. Chem. Soc. 2003, 125, 7301.
(d) Kwon, J. Y.; Jang, Y. J.; Kim, S. K.; Lee, K.-H.; Kim, J. S.; Yoon, J.
J. Org. Chem. 2004, 69, 5155. (e) Chen, Q.-Y.; Chen, C.-F. Tetrahedron
Lett. 2004, 45, 6493. (f) Ghosh, S.; Choudhury, A. R.; Row, T. N. G.;
Maitra, U. Org. Lett. 2005, 7, 1441. (g) Fujimoto, C.; Kusunose, Y.; Maeda,
H. J. Org. Chem. 2006, 71, 2389. (h) Wallace, K. J.; Belcher, W. J.; Turner,
D. R.; Syed, K. F.; Steed, J. W. J. Am. Chem. Soc. 2003, 125, 9699. (i)
Vega, I. E. D.; Gale, P. A.; Light, M. E.; Loeb, S. J. Chem. Commun.
2005, 4913.
Acknowledgment. O.B.B. acknowledges the NSF for an Integra-
tive Graduate Education and Research Traineeship (DGE-0549503).
J.S.M. recognizes the Ronald E. McNair Program for its support.
D.W.J. is a Cottrell Scholar of Research Corporation and gratefully
acknowledges the NSF for a CAREER award. Dr. Lev N. Zakharov
is acknowledged for assistance with the single-crystal X-ray diffraction
experiments. B.P.H. acknowledges support from the Division of
Chemical Sciences, Geosciences, and Biosciences, Office of Basic
Energy Sciences, U.S. Department of Energy (DOE) under contract
number DE-AC05-00OR22725 with Oak Ridge National Laboratory
(managed by UT-Battelle, LLC). DFT calculations were performed
using the Molecular Science Computing Facility (MSCF) in the
William R. Wiley Environmental Molecular Sciences Laboratory, a
national scientific user facility sponsored by the DOE’s Office of
Biological and Environmental Research and located at Pacific North-
west National Laboratory managed for DOE by Battelle.
(11) Hennrich, G.; Anslyn, E. V. Chem.sEur. J. 2002, 8, 2218.
(12) Lee, J. C.; Choi, Y. Synth. Commun. 1998, 28, 2021.
j
(13) Crystal data for 1: C36H30N6O18, M ) 834.66, triclinic, P1, a ) 5.6949(12)
Å, b ) 19.833(4) Å, c ) 33.883(7) Å, R ) 84.106(4)°, ꢀ ) 85.278(5)°, γ
) 81.873(5)°, V ) 3759.6(14) Å3, Z ) 4, µ(Mo KR) ) 0.121 mm-1; R1
) 0.0566 (1321 parameters, 16401 reflections with I > 2σ(I)); R1 ) 0.1090,
wR2 ) 0.1446, GoF ) 1.007 for all 23621 data.
(14) Crystal data for 2: (C36H30N6O18)(C2H6OS)3, M ) 1069.04, monoclinic,
P2(1)/n, a ) 28.272(12) Å, b ) 5.075(2) Å, c ) 37.126(15) Å, ꢀ )
110.669(7)°, V ) 4984(4) Å3, Z ) 4, µ(Mo KR) ) 0.091 mm-1; R1 )
0.0857 (544 parameters, 8731 reflections with I > 2σ(I)); R1 ) 0.1736,
wR2 ) 0.2640, GoF ) 0.946 for all 33339 data.
(15) UV-vis titrations for 1 and 2 with NHep4+I- at 21 °C were performed to
corroborate the 1H NMR titrations herein. Regrettably, the charge transfer
band that grows in throughout the titration appears as a shoulder on the
residual NHep4+I- band that increases throughout the titration (see
Supporting Information).
Supporting Information Available: Full experimental details, includ-
ing synthesis, spectroscopic, titration, and structural details for all
compounds described; crystallographic data of 1 and 2 (CCDC #s
661394-661395); details of DFT calculations and relevant references are
available. This material is available free of charge via the Internet at http://
pubs.acs.org.
(16) Hirose, K. J. Inclusion Phenom. Macrocycl. Chem. 2001, 39, 193.
(17) Titrations of receptors 1 and 2 (∼5 mM) with tetra-n-butylammonium
bromide (NBu4+Br-) were performed at room temperature resulting in
lower association constants (averaging 4-6 M-1). These data indicate that
countercation and/or temperature plays a role in the anion binding ability
of this system. Nevertheless, titrations of receptor 1 at these concentrations
(∼5 mM) better illustrate the dramatic ∆δ for this receptor.
(18) Analogous titrations of 1 (2 mM) at 27 °C with NHep4+ halides also exhibit
striking peak movement (up to 0.632 ppm for NHep4+Cl-, 0.500 ppm for
NHep4+Br-, and 0.390 ppm for NHep4+I-) throughout the experiment
(Table 1 and Supporting Information).
References
(1) Berryman, O. B.; Bryantsev, V. S.; Stay, D. P.; Johnson, D. W.; Hay, B. P.
J. Am. Chem. Soc. 2007, 129, 48.
(2) Hay, B. P.; Bryantsev, V. S. Chem. Commun. 2008, 2417.
(3) Weak, reversible attractions between anions and electron-deficient arenes,
and putative structures for the different interaction types (including each
of the interactions A-D), have been reported. For representative examples,
see: (a) Buncel, E.; Norris, A. R.; Russell, K. E. Q. ReV. Chem. Soc. 1968,
22, 123. (b) Hiraoka, K.; Mizuse, S. J. Phys. Chem. 1987, 91, 5294. (c)
Kebarle, P.; Chowdhury, S. Chem. ReV. 1987, 87, 513. Recent computa-
tional studies have attracted considerable attention and have attempted to
characterize the interaction(s) in greater detail; see: (d) Alkorta, I.; Rozas,
I.; Elguero, J. J. Am. Chem. Soc. 2002, 124, 8593. (e) Mascal, M.;
Armstrong, A.; Bartberger, M. D. J. Am. Chem. Soc. 2002, 124, 6274. (f)
Quinonero, D.; Garau, C.; Rotger, C.; Frontera, A.; Ballester, P.; Costa,
A.; Deya, P. M. Angew. Chem., Int. Ed. 2002, 41, 3389. For additional
recent reviews, see: (g) Gamez, P.; Mooibroek, T. J.; Teat, S. J.; Reedijk,
J. Acc. Chem. Res. 2007, 40, 435. (h) Schottel, B. L.; Chifotides, H. T.;
Dunbar, K. R. Chem. Soc. ReV. 2008, 37, 68.
(19) DFT calculations were performed with the NWCHEM program (a) using
the B3LYP functional (b-e) with the DZVP basis set and DGauss A1
coulomb fitting (f). (Optimized geometries and absolute energies for all
structures are provided as Supporting Information.) (a) Bylaska, W.;
NWChem, A Computational Chemistry Package for Parallel Computers,
version 5.0; Pacific Northwest National Laboratory: Richland, WA, 2006;
for a full author list, see Supporting Information. (b) Becke, A. D. Phys.
ReV. A 1988, 38, 3098. (c) Becke, A. D. In The Challenge of d and f
Electrons: Theory and Computation; Salahub, D. R., Zerner, M. C., Eds.;
ACS Symposium Series, No. 394; American Chemical Society: Washington,
DC, 1989; p 166. (d) Becke, A. D. Int. J. Quantum Chem. Symp. 1989, 23,
599. (e) Perdew, J. P. Phys. ReV. B 1986, 33, 8822. (f) Godbout, N.;
Salahub, D. R.; Andzelm, J.; Wimmer, E. Can. J. Chem. 1992, 70, 560.
(20) Bryantsev, V. S.; Hay, B. P. Org. Lett. 2005, 7, 5031.
(21) Bryantsev, V. S.; Hay, B. P. J. Am. Chem. Soc. 2005, 127, 8282.
(22) An alternate conformation, 6.0 kcal/mol higher in energy, was located for
1 · Br-, where the Br- forms a weak σ complex (3.432 Å) with one arene
and bifurcated aryl H-bonds (2.832, 3.043, 2.909, and 3.021 Å) with the
other two (see Supporting Information). It is possible that the H-bond
complex and/or the weak σ structure is responsible for the colors observed
in solution when receptors 1 and 3 are mixed with Br- or I-.
(4) Bianchi, A., Bowman-James, K., Garcia-Espana, E., Eds. Supramolecular
Chemistry of Anions; Wiley-VCH: New York, 1997.
(5) Sessler, J. L.; Gale, P. A.; Cho, W.-S. Anion Receptor Chemistry; The Royal
Society of Chemistry: Cambridge, UK, 2006.
(6) Sessler, J. L.; Gross, D. E.; Cho, W.-S.; Lynch, V. M.; Schmidtchen, F. P.;
Bates, G. W.; Light, M. E.; Gale, P. A. J. Am. Chem. Soc. 2006, 128,
12281.
(23) As a control, no color is observed when NHep4+Cl-, NHep4+Br- or
NHep4+I- are dissolved in C6D6 and heated to 27 °C.
(7) Rosokha, Y. S.; Lindeman, S. V.; Rosokha, S. V.; Kochi, J. K. Angew.
Chem., Int. Ed. 2004, 43, 4650.
JA8035652
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 10897