Does an axial propeller shape on a dirhodium(III,III) core affect equatorial ligand chirality?

Article abstract of DOI:10.1021/om2003078

The dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) framework, which exists solely in a propeller conformation, has been prepared with nonidentical apical aryl ligands. The presence of these nonidentical ligands on the propeller conformation was expec

Full text of DOI:10.1021/om2003078

ARTICLE
pubs.acs.org/Organometallics
Does an Axial Propeller Shape on a Dirhodium(III,III) Core Affect
Equatorial Ligand Chirality?
Michael P. Doyle,*,† Dmitry Shabashov,† Lei Zhou,† Peter Y. Zavalij,† Christopher Welch,‡ and
Zainab Pirzada‡
†Department of Chemistry, University of Maryland, College Park, Maryland 20742, United States
‡Merck Research Laboratories, Merck & Co., P.O. Box 2000, Rahway, New Jersey 07065, United States
S Supporting Information
b
ABSTRACT: The dirhodium(III,III) tetrakis(1-aza-2-cyclo-
octanoate) framework, which exists solely in a propeller conforma-
tion, has been prepared with nonidentical apical aryl ligands. The
presence of these nonidentical ligands on the propeller conforma-
tion was expected to define this structural unit as a chiral element
and the resulting aryl(1)aryl(2)dirhodium(III,III) tetrakis(1-aza-
2-cyclooctanoate) as a chiral compound. Dirhodium(III,III)
tetrakis(1-aza-2-cyclooctanoate) compounds having phenyl and
p-tolyl, 1-(3-phenylphenyl) and 2-methoxy-6-naphthyl, and p-(N-
p-toluenesulfonylpyrolinamidophenyl) apical ligands were pre-
pared and subjected to chromatographic, NMR, and X-ray crystal-
lographic analysis to ascertain if the propeller core provided a
chiral element that could define two enantiomers when dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) was bound
to nonidentical apical aryl ligands. X-ray analysis suggests chirality in those compounds for which structures could be
obtained, but chromatographic analyses were not able to separate enantiomers. Studies of dirhodium(III,III) tetrakis-
(1-aza-2-cyclooctanoate) compounds possessing different apical ligands by NMR spectroscopy using binuclear chiral
lanthanideꢀsilver shift reagents were also performed, but conclusive evidence for chirality in these dirhodium complexes
was not obtained.
’ INTRODUCTION
We have recently reported the preparation and structural
characterization of thermally and chemically stable paddlewheel
dirhodium(III,III) compounds with phenyl substituents in the
axial positions and lactamates as the bridging ligands.1 When the
lactamate ligand is caprolactam, two noninterconvertable con-
formational isomers are formed, one having a biplanar orienta-
tion of the caprolactamate ligands and the other with a propeller
orientation (Figure 1).2 These conformational isomers differ in
the orientation of the bridging caprolactamate ligands and
possess unique spectral, chromatographic, and chemical
properties.2,3 Attempts to force the thermal interconversion of
these configurational isomers in the solid state (up to 240 °C) or
Figure 1. Bisphenyldirhodium(III,III) tetrakis(caprolactamate) core
in solution (refluxing chlorobenzene at 130 °C, 24 h) have not
shown along the rhodiumꢀrhodium bond axis in (a) biplanar and (b)
been successful. Both remain intact when subjected to these
propeller conformations. Red balls are O, blue are N, and green are
conditions. The biplanar and propeller conformers are related to
Rh atoms.
each other by “flipping” two oppositely placed seven-membered
rings with atomic motion basically restricted to the three
RhꢀOꢀC angle, and a less pronounced closing of the RhꢀNꢀC
methylene groups most distant from the amide functional group,
and the source for the restriction of this atomic motion is not
revealed through either examination of molecular models or
DFT calculations.2 In both conformers, a severe bending of the
RhꢀRhꢀPh bond axis from 180°, a major opening of the
angle from the ideal 120° are observed by single-crystal X-ray
analysis.
Received: April 9, 2011
Published: June 13, 2011
r
2011 American Chemical Society
3619
dx.doi.org/10.1021/om2003078 Organometallics 2011, 30, 3619–3627
|

Organometallics
ARTICLE
tetrakis(1-aza-2-cyclooctanoate) was constructed from dirho-
dium(II,II) 1-aza-2-cyclooctanoate (HACO), phenylboronic
acid, and p-tolylboronic acid, catalyzed by copper(II) sulfate
(eq 1). The products consisted of phenyl(p-tolyl)dirhodium(III,III)
tetrakis(1-aza-2-cyclooctanoate) (1) as well as bisphenyldirhodium-
(III,III) tetrakis(1-aza-2-cyclooctanoate) (2) and bis(p-tolyl)-
dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) (3) in a nearly
statistical 48:22:26 ratio, which was expected. Attempts to chroma-
tographically separate the enantiomers of compound 1 using
comprehensive chiral SFC, HPLC, and RP-HPC screening were
unsuccessful, and only one peak for 1 was detected. While this
battery of chromatographic methods is generally successful in at
least partially resolving the enantiomers of most chiral analytes, one
should not draw a conclusion from these negative results that reflect
on the chirality of the sample.
Figure 2. Propeller structures with nonidentical apical ligands differ-
entiated by reflection or inversion.
With caprolactamate as the axial ligand, the biplanar confor-
mer (Figure 1a) is the major isomer, and the propeller conformer
(Figure 1b) is the minor stereoisomer. However, only bisphenyl-
dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) having the
propeller conformation is formed with the higher homologue of
caprolactam, 1-aza-2-cyclooctanone (ACOH), as the bridging
ligand.2 Molecular propellers have been recognized in organic
chemistry as well-defined architectures that can exhibit restricted
conformational equilibria,4 and significant advances in stereo-
chemical applications have resulted from these studies,5 but there
have been no previous reports of a propeller structure having a
dimetallic core. However, although previously reported carbon-
based molecular propellers often have a low energy barrier for
interconversion between the two forms, the propeller conforma-
tion in bisphenyldirhodium(III,III) carboxamidates does not
exhibit an energetically favorable interconversion between
biplanar and propeller conformers even at high temperatures
as a solid or in solution.2 As a consequence, with the ass-
umption that the two propeller conformational isomers of a
bisaryldirhodium(III,III) carboxamidate (Figure 2, where the
two aryl apical ligands are not identical) do not interconvert,
we envisioned that placing two nonidentical structural units
at the axial positions of dirhodium(III,III) 1-aza-2-cyclo-
octanoate would result in a pair of enantiomers and that
having chiral units in those axial positions would produce
diastereoisomers in numbers that corresponded to the num-
ber of chiral centers.
At this point it is useful to consider the symmetry of biplanar
and propeller complexes. As can be seen from Figure 1a, the
biplanar core of dirhodium(III,III) caprolactamate (excluding its
apical phenyl ligands) has a vertical mirror plane, which, along
with a horizontal 2-fold axis, yields an inversion center and,
therefore, 2/m symmetry. In contrast, the propeller core (from
Figure 1b) only has an inversion center. If both apical ligands are
the same (e.g., phenyl), the symmetry of the structure can be as
high as that of the core assuming that there are no packing or
other distortion effects. On the other hand, the presence of
nonidentical apical ligands on a dirhodium(III,III) carboxami-
date core can yield asymmetric or chiral complexes (Figure 2).
The formal exchange of nonidentical apical ligands on the
propeller core yields a structure of different chirality due to the
presence of the inversion center in the core (an operation that is
equivalent to inversion or reflection).
The crystal structure of 1 was solved in the chiral P21 space
group, showing ordered distribution of phenyl and p-tolyl ligands
in apical positions as shown in Figure 3. The mirror image
configuration can be imagined as the inverse image of the
complex depicted in Figure 3 or, which is identical, the config-
uration when phenyl and p-tolyl apical ligands are exchanged.
However, the crystal of 1 does not consist of a single configura-
tion of 1, as it shows merohedral twinning refined to an
approximate 1:1 ratio. In other words, each crystal domain or
mosaic block consists of only one configuration, but different
domains have different configurations of the complex. Additional
evidence of such a distribution of configurations follows from less
computationally suitable results when centrosymmetric space
group P21/c is used, in which both conformers are superimposed;
this yields a 6.3% R-factor for the P21/c space group versus a 4.5%
R-factor in the chiral P21 group.
Recognizing that the subtle structural differences between
phenyl and p-tolyl may not have provided adequate separation of
enantiomers of 1 on the chiral columns that were employed, and
relying on the observed merohedral twinning refined to an
approximate 1:1 ratio with 1 as a suggestion of chirality, two
bulky and more structurally different aryl groups were used to
prepare [1-(3-phenylphenyl)](2-methoxy-6-naphthyl)dirhodium-
(III,III) tetrakis(1-aza-2-cyclooctanoate) (4) from dirhodium-
(II,II) 1-aza-2-cyclooctanoate, m-biphenylboronic acid, and
2-methoxy-6-naphthylboronic acid, catalyzed by copper(II) sul-
fate (eq 2). The chromatographic ratio of the three products was
22 (5):41 (4):17 (6). Compound 4, with its apical 3-phenylphe-
nyl and 2-methoxy-6-naphthyl ligands, was separated from 5 and
6 by either column chromatography or preparative TLC, isolated
in relatively low yield, and characterized spectroscopically. Even
though the homobis-aryldirhodium(III,III) isomers (5 and 6)
have retention volumes significantly different from that for 4
lying between them, there was no separation of 4 into enantio-
mers observed under a broad set of conditions. Again, attempts
to chromatographically separate the enantiomers of compound 4
’ RESULTS AND DISCUSSION
Using the protocol for formation of mixed bisaryldirhodium-
(III,III) caprolactamates,6 phenyl(p-tolyl)dirhodium(III,III)
3620
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
Figure 3. Phenyl(p-tolyl)dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) (1) in two views that are perpendicular to each other, showing the
anisotropic atomic displacement ellipsoids for the non-hydrogen atoms at the 30% probability level. Hydrogen atoms are displayed with an arbitrarily
small radius.
Figure 4. 1-(3-Phenylphenyl)(2-methoxy-6-naphthyl)dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) (4) in two views that are perpendicular to
each other showing the anisotropic atomic displacement ellipsoids for the non-hydrogen atoms at the 30% probability level. Hydrogen atoms are
displayed with an arbitrarily small radius.
using comprehensive chiral SFC, HPLC, and RP-HPC screening
were unsuccessful.
pair of parallel c-planes. The neighboring layers are a reflection of
each other, differing by the chirality of the molecules.
The crystal structure determination reveals the presence of
merohedral twinning in a 3:2 ratio, which in this particular case
can be described as follows: in the Pc group molecules are of
different chirality, but their orientation (orientation of apical
ligands in this case) is the same. The merohedral twinning
introduces an inversion center that relates different domains so
that the orientation of molecules in domains alternates.
Dirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) 4 crystal-
lizes in the monoclinic non-centrosymmetric space group Pc. A
general view of 4 in two orientations that are perpendicular to
each other is shown in Figure 4; both apical ligands are tilted from
the RhꢀRh axis toward the O atoms of the 1-aza-2-cyclooctano-
ate ligands. As is reported for 1 with poor correlation, fitting the
experimental data for structure 4 with centrosymmetric P21/c
symmetry is much worse than the fit with Pc symmetry. Also,
although the structure of 4 in the chiral P21 group yields a better
fit than with the centrosymmetric model, its agreement with
experiment, including reflection absences, is worse than in the
model with Pc symmetry. Thus the model with Pc symmetry was
accepted as the most adequate description of the structure of 4.
Note that Pc is an acentric symmetry group, but due to the
presence of a c-glide plane, it is not chiral. Thus, contrary to the
chiral P21 group, in which all molecules within the crystal (or in
the presence of merohedral twinning within the domain) are of
the same chirality, in the Pc symmetry group the molecules of the
same chirality (in this particular structure) form a layer between a
In addition to merohedral twinning and the ordered alteration
of chirality in the layers, the molecules of different conformations
are disordered in such a way that their cores perfectly coincide,
but their apical ligands, 3-phenylphenyl (PP) and 2-methoxy-6-
naphthyl (MNP), are superimposed onto each other in a 3:1
ratio. The superimposed conformers shown in Figure 5 are of
opposite chirality and can be transformed into each other by
either inversion or reflection. Swapping apical ligands also
changes chirality and is identical to the above transformations.
The crystal structure of 4 is further complicated by having
every other 1-aza-2-cyclooctanoate ligand on the dirhodium core
randomly accepting an alternate conformation (Figure 6a and b),
which is observed as a superposition of two conformations in an
approximate 7:2 ratio. Only the two ligands opposite each other
alter their shape, while two other ligands show no disorder.
Essentially the same disorder was found in bisphenyldirhodium-
(III,III) 1-aza-2-cyclooctanoate complex 2.2 The main conforma-
tion of complex 4 shown in Figure 5a is found in all known
3621
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
Figure 7. Alternate conformation of ligands: (a) signs of torsion angles
in the conformation described by Figure 6a; (b) signs of torsion angles in
the conformation described by Figure 6b. The gray circles in (b) show
the changes in signs of (b) compared to (a); the main difference is in
flipping the upper two carbon atoms that are opposite the
NꢀC(O) bond.
Figure 5. View of two conformations in 4, major (a) and minor (b), that
are superimposed onto each other in a 0.74 to 0.26 ratio. The
conformations (a) and (b) are inverse images of each other.
spectroscopically and chromatographically. The process em-
ployed with L-proline to form 9L is shown in Scheme 1, and the
same procedures were employed with D-proline to form 9D. For
each of 9L and 9D only one compound was chromatographically
separable, and 9L was cleanly distinguished from 9D. In a separate
experiment racemic D,L-proline was employed with the expecta-
tion of formation of 9L, 9D, and the bis-aryldirhodium(III,III)
tetrakis(1-aza-2-cyclooctanoate) with both D- and L-prolinamide
attachments (9DL), whichwasexpectedtobeformedina1(9L):2-
(9DL):1(9D) ratio (eq 3). Chromatographic analysis, however,
exhibited signals for only two distinguishable materials in a 1:1
ratio, presumably because differentiation was restricted only to
the D- and L-prolinamide attachments. These analyses do indicate
that association with the stationary phase of the column is
localized on only one apical ligand without observable influence
on the second apical ligand in the same molecule.
Figure 6. Views along the RhꢀRh bond axis showing two variants of
the propeller configurations of dirhodium(III,III) with 1-aza-2-cyclooc-
tanoate (ACO) ligands: (a) in 1 and essentially the same main
conformation in 4, and (b) alternate conformation in 4.
propeller structures of dirhodium(III,III) 1-aza-2-cyclooctanoate
including complex 1, whose conformation is essentially the same
as for the major conformer of compound 4. As can be seen from
Figure 6a there are two different shapes of the ligands. These two
alternate conformations of ligands are schematically represented
in Figure 7a and b, respectively, where the signs of corresponding
torsion angles are shown. Thus the main conformation of the
complex 4 and the only conformation in 1 could be denoted by
the sequence a,b,ꢀa,ꢀb describing conformation of individual
ligands, where the negative sign denotes inversion. Note that
signs of torsion angles are opposite for ligands trans to each other
due to inversion symmetry of the core.
The absence of chromatographic separation of what we
believed should be diastereoisomers prompted us to investigate
the possible separation of enantiomers through the use of NMR
chiral shift reagents.7 Addition of lanthanide shift reagents to
organic compounds generally results in shifts of resonances to
higher (or lower) frequency with the shift determined primarily
by the distance of the given type of proton (or other NMR active
atom) from the donor group. Ideally, the six-coordinate lantha-
nide complex forms a weak addition complex that is in fast
exchange with the unbound organic substrate on the NMR time
scale. The induced shifts are caused by a large difference in the
magnetic susceptibility tensors for the seven-coordinate com-
plex. Analyses of chiral alkenes and arenes are particularly
difficult.8 However, a chiral lanthanide shift reagent in combina-
tion with silver ion has been shown to cause detectable shifts.9 In
these cases induced shifts are observed from a mixed complex
that forms between the chiral hydrocarbon that is weakly
associated with the silver ion in solution. The dirhodium(III,III)
complexes 10ꢀ12 are substantially shielded from dipolar inter-
actions, but weak association with silver ion should be feasible,
and the chiral lanthanide reagent is expected to induce NMR
shifts in selected ACO proton signals if the dirhodium complex is
a racemic mixture.
The configuration of the 1-aza-2-cyclooctanoate ligands that
are trans to each other on the dirhodium core changes from a to b
and ꢀa to ꢀb, respectively, by flipping two carbon atoms that are
opposite the NꢀC(O) bond (Figure 6). Thus when ligands that
are trans to each other have the alternative configuration of that
shown in Figure 6b, all ligands become identical (with respect to
the inversion), as shown in Figure 7b, and the configuration of
the core can be described as b,b,ꢀb,ꢀb.
To better probe the propeller conformation of dirhodium(III,
III) tetrakis(1-aza-2-cyclooctanoate) as a chiral unit, we prepared
bis-aryldirhodium(III,III) tetrakis(1-aza-2-cyclooctanoate) with
chiral N-p-toluenesulfonylprolinamide attachments that were at
the para positions. Both D- and L-prolinamide derivatives were
prepared, yielding single compounds that were characterized
3622
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
Scheme 1. Synthesis of Arylboronic Acid 8 and Dirhodium Complex 9
(Figure 8). Only line broadening of the 1H NMR signals of both
aromatic and aliphatic hydrogens was observed with the addi-
tions of the chiral lanthanide salts, suggesting the absence of
association between the chiral praseodymium (Figure 8) or
ytterbium (Figure 9) camphorates and 10 or 11. However, upon
addition of Ag(fod), an immediate and continuing separation of
signals at δ 3.2 (hydrogens of CH2 adjacent to the amide N) and
2.3 (hydrogens of CH2 adjacent to the carbonyl group) into two
multiplets occurred. Changes in the chemical shift positions of
these signals to lower field occurred from sequential additions of
silver ion, along with the increasing separation of their multiplets,
whose integration was 1:1. The separation of multiplets originally
at δ 3.2 and 2.3 was approximately the same, but greater with
dirhodium(III,III) compound 11 than with 10. In addition, the
chemical shift positions of the hydrogens on the aromatic rings
moved to lower field upon addition of silver ion, with the change
in chemical shift positions being much greater for the phenyl
substituent than for the para-substituted ring substituent. This is
consistent with the expected π-coordination of silver(I) with
arenes, and the preference for the phenyl ring is expected from
published reports.10
When the order of addition of silver(I) and chiral shift
reagents was reversed, there was a surprising change in the
outcome. Upon addition of Ag(fod), there was an immediate
and continuing separation of signals at δ 3.2 and 2.3 into two
multiplets with 11, but much less so with 10. In contrast to the
opposite mode of addition where the multiplet separations at δ
3.2 and 2.3 were nearly identical, the separation of signals was
Initial experiments were performed on 10 and 11 with chiral
praseodymium or ytterbium camphorates (13) and the silver
diketonate Ag(fod) (14). The dirhodium compound was dis-
solved in CDCl3, and the chiral lanthanide reagent was added in
portions, followed by the silver diketonate, also in portions
3623
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
Figure 8. Proton NMR data of the aliphatic proton region for (a) 10 (5.7 mM in CDCl3) and (b) 11 (5.3 mM in CDCl3) with the PrꢀAg reagent (13a
then 14).
Figure 9. Proton NMR data for the aliphatic proton region of (a) 10 (3.7 mM in CDCl3) and (b) 11 (3.5 mM in CDCl3) with the YbꢀAg reagent (14
then 13b).
greater for the resonances at δ 2.3 than for those at δ 3.2 with the
initial addition of silver(I). However, the integral ratios of the
separated signals were again 1:1. Subsequent addition of the
chiral lanthanide complex led to broadening and further separa-
tion of the signals (Figure 9). As in the previous mode of
addition, the chemical shift positions of the hydrogens on the
aromatic rings moved to lower field upon addition of silver ion,
with the change in chemical shift positions being much greater
for the phenyl substituent than for the para-substituted ring.
How does Ag(fod) induce a separation of signals at δ 3.2 and
2.3? One answer would be the preferential association onto one
face of the aryl rings of either 10 or 11. Preferential association
should be on the more open face of the aryl ring, which would be
anti to the two amide nitrogens on one face of the dirhodium(III,
III) compound, and this association would be expected to
influence the signals at δ 2.3 more than those at δ 3.2, which is
what is observed. The appearance of what is seen as a triplet in the
emerging separation (at 2.4ꢀ2.5 ppm in Figure 9) might suggest
that the signals that are being separated in this case are the “axial”
and “equatorial” hydrogens of the amide ligands (15). In other
words, signal separation may be due to differentiation of proton
signals on each amide, independent of the two aryl substituents.
If this explanation is correct, however, Ag(fod) association with
the symmetrical diphenyldirhodium(III,III) compound 12
should produce the same outcome, but, as seen from Figure 10,
this is not the case. There is no separation of signals when
Ag(fod) is added to 12, and only line broadening occurs when
chiral lanthanide reagent is added.
3624
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
When the symmetrical complex 12 was examined by 13C
NMR spectroscopy, only one set of signals was detected for each
carbon of the bridging lactam ligands. However, two sets of
signals were observed for bridging lactam units of the unsymme-
trical bisaryldirhodium(III,III) compounds. Those sets come
from the 2,2-cis geometrical arrangement of bridging ligands,
and attaching different apical ligands to the dirhodium core
allows them to be differentiated. Consequent to this observation,
we have examined the bisaryldirhodium(III,III) complex by 13C
NMR using a chiral silver reagent. The reagent of choice was a
chiral silver camphorate derivative prepared according to a
published procedure.11 When this compound was added to 10,
collected 13C NMR data showed no splitting due to coordination
An alternative explanation for this NMR signal splitting is
that the onꢀoff rate for silver ion association with each aryl
group is rapid on the NMR time scale, but association with the
phenyl ring moves the associated ligand signals more than
association with the substituted aromatic ring. Complexation of
Ag(fod) with the unsubstituted phenyl group on the dirhodium-
(III,III) core is preferred due to its higher electron density,10
and this is supported by examination of aromatic regions in
spectral data for dirhodium complexes 10 and 11 and compar-
ing them with symmetrical diphenyldirhodium(III,III) com-
pound 12 (Figure 11). Introduction of Ag(fod) to solutions of
these dirhodium complexes induces proton NMR shifts of aryl
group signals. Larger shifts are observed for the proton signals
of the unsubstituted phenyl group, while the electron-poor
aryl group possessing a fluorine atom has relatively small shifts
of proton signals (Figure 11). Thus, the separation of signals at
3.2 and 2.3 ppm arises from the differences in association of
the silver salt with aryl groups on opposite sides of dirhodium
compounds.
Figure 12. View of the bisphenydirhodium(III,III)ꢀsilver salt adduct
(12). Hydrogen atoms are removed for clarity. Selected bond lengths
ꢀ
Figure 10. Proton NMR spectra of the aromatic proton region of 12
(3.8 mM in CDCl3) with the YbꢀAg reagent (14 then 13b).
(Å) and angles (deg): RhꢀRh 2.506; RhꢀC 2.004 and 2.019; AgꢀAg
2.933; AgꢀC 2.544; CꢀRhꢀRh 159.0 and 155.0; CꢀAgꢀAg 121.5.
Figure 11. Proton NMR spectra of the aromatic proton regions for (a) 10 (3.7 mM in CDCl3) and (b) 11 (3.5 mM in CDCl3) with the YbꢀAg reagent
(14 then 13b).
3625
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
with the chiral silver reagent and did not provide evidence for
dirhodium complex chirality. No splitting of signals from the
bridging lactam or aryl groups was detected.
tetrakis(1-aza-2-cyclooctanoate) [Rh2(ACO)4(CH3CN)2]2 (48.0 mg,
0.060 mmol), phenylboronic acid (18.3 mg, 2.50 equiv), p-tolylboro-
nic acid (20.4 mg, 2.50 equiv), and NaHCO3 (50.4 mg, 10.0 equiv)
were dissolved in 6.00 mL of CH2Cl2. Copper(II) sulfate (10.0 mol %)
in 1.50 mL of MeOH was added to the above solution. The result-
ing purple mixture was stirred under air at room temperature for
12 h, during which the solution turned greenish-brown. After filtra-
tion of NaHCO3 and evaporation of the solvent under reduced
pressure, the crude product was purified by column chromatography
on silica gel using CH2Cl2 as eluent to give the mixture of 1, 2, and 3
as a greenish-brown solid in 38% isolated yield (20.0 mg). Column
chromatography could not separate these complexes. HPLC analysis
on a C-18 column (CH3CN/iPrOH, 87:13, 1.0 mL/min, 254 nm
UVꢀvis) shows three peaks, which include two homoaryl com-
pounds (2 and 3) and one heteroaryl (1) compound. Calculation
based on HPLC and 1H NMR analyses show the molar ratio of three
compounds to be PhPh:PhTol:TolTol = 22:48:26. For HPLC
analyses of the mixture, retention times of 2, 1, and 3 are 12.0,
14.6, and 17.8 min, respectively. A single crystal suitable for X-ray
analysis was obtained from the mixture of 1, 2, and 3 in CH2Cl2. 1: 1H
NMR (400 MHz, CD2Cl2) δ 7.47ꢀ7.50 (comp, 2H), 7.33 (d, 2H,
J = 8.0 Hz), 7.09ꢀ7.11 (comp, 3H), 6.94 (d, 2H, J = 8.0 Hz),
3.15ꢀ3.26 (comp, 8H), 2.41 (s, 3H), 2.24ꢀ2.31 (comp, 8H),
1.22ꢀ1.58 (comp, 32H); 13C NMR (150 MHz, CDCl3) δ
181.9, 136.5, 136.0, 132.8, 126.9, 125.8, 123.6, 53.3, 50.1, 33.4, 32.6,
28.0, 26.1, 25.1; HRMS (ESI) for C41H61N4O4Rh2 [M þ H]þ calcd
879.2803, found 879.1859; for 2: HRMS (ESI) for C40H59N4O4Rh2
[M þ H]þ calcd 865.2646, found 865.1860; for 3: HRMS (ESI) for
C42H63N4O4Rh2 [M þ H]þ calcd 893.2959, found 893.2034; IR
Complex formation between the silver salt, Ag(fod), and
phenyl group in bisphenyldirhodium(III,III) tetrakis(1-aza-2-
cyclooctanoate) (2) was confirmed by single-crystal X-ray dif-
fraction (Figure 12). The original diketonate ligand on silver salt
14 was transformed during crystallization into perfluorobutyrate,
what most likely resulted through hydrolysis of the initial ligand
by residual moisture. The structure depicted in Figure 12 is one
of the first intermetallic silverꢀaryl complexes characterized by
X-ray diffraction and is the only one with rhodium. Intermetallic
platinum and gold silverꢀalkyne complexes have been
reported.12 Coordination of the silver ion with a phenyl group
of 2 occurs in η1 fashion and places silver atop the ortho carbon
on the phenyl ligand. The AgꢀCꢀC angles are 87.6° and 95.9°,
respectively. The silver atom is also coordinated with one of the
carboxamidate ligand carbonyl oxygens. The RhꢀRhꢀC angles
of 159.0° and 155.0° did not change significantly compared to
the structure without the silver complex.2 Interestingly, apical
phenyl ligands are not coplanar in this complex, and one of the
phenyl groups has been rotated about 90° around the rhodium
phenyl axis. The RhꢀRh distance has not been affected by
ꢀ
complex formation with silver and is shorter only by 0.02 Å.
’ CONCLUSION
In summary, the chirality of paddlewheel dirhodium(III,III)
complexes having different apical aryl substituents that is sug-
gested from X-ray diffraction studies could not be confirmed by
chromatographic or NMR spectroscopic analyses. Application of
a broad variety of chromatographic and NMR methods did not
distinguish between two enantiomers of a presumed racemic
mixture. Although unlikely by analogy with the absence of
conformational interconversion between biplanar and propeller
conformations of bisphenyldirhodium(III,III) tetrakis(caprol-
actamate),2 the possibility exists that the two propeller enantio-
mers are in rapid equilibrium. The absence of evidence for two
enantiomers for nonsymmetrical apical aryl-substituted dirho-
dium(III,III) tetrakis(1-aza-2-cyclooctanoate) using well-estab-
lished methodologies does suggest the limits of these methods or
the existence of an as yet unknown characteristic of these unique
organometallic compounds.
(neat) 2923, 2854, 1658, 1633, 1579, 1454 cmꢀ1
.
Procedure for the Synthesis of 1-(3-Phenylphenyl)(2-
methoxy-6-naphthyl)dirhodium(III,III) Tetrakis(1-aza-2-cyclo-
octanoate) (4). Rh2(ACO)4(CH3CN)2 (48.0 mg, 0.060 mmol),
3-biphenylboronic acid (24.0 mg, 2.00 equiv), 6-methoxy-2-naphthale-
neboronic acid (24.0 mg, 2.00 equiv), and NaHCO3 (50.4 mg,
10.0 equiv) were dissolved in 6.00 mL of CH2Cl2. Copper(II) sulfate
(10.0 mol %) in 1.50 mL of MeOH was added to the above solution. The
resulting purple mixture was stirred under air at room temperature for
12 h, during which the solution turned yellowish-brown. After filtration
and evaporation of the solvent under reduced pressure, the crude
product was purified by preparative TLC on silica gel using CH2Cl2
as eluent to give the compound mixture of 4, 5, and 6 as a yellowish-
brown solid in 10% yield. HPLC analysis of the reaction mixture on a
C-18 column (CH3CN/iPrOH, 87:13, 1.0 mL/min, 254 nm UVꢀvis)
shows three peaks, which include two homoaryl compounds (5 and 6)
and one heteroaryl compound (4) having retention times of 10.4 (6),
11.9 (4), and 13.3 min (5). The molar ratio of the three peaks is
22:41:17. A single crystal of 4 suitable for X-ray analysis was obtained by
crystallization of 4 in CH2Cl2. 4: 1H NMR (400 MHz, CD2Cl2) δ 7.90
(s, 1H), 7.72ꢀ7.76 (comp, 4H), 7.63ꢀ7.66 (m, 1H), 7.54ꢀ7.57 (comp,
2H), 7.46ꢀ7.50 (comp, 2H), 7.37ꢀ7.42 (comp, 2H), 7.21ꢀ7.26
(comp, 2H), 7.11ꢀ7.14 (m, 1H), 3.98 (s, 3H), 3.25ꢀ3.38 (comp,
8H), 2.35ꢀ2.41 (comp, 8H), 1.27ꢀ1.66 (comp, 32H); 13C NMR (150 MHz,
CD2Cl2) δ 182.1, 157.7, 156.6, 146.7, 141.9, 140.4, 138.7, 135.6, 135.0,
133.5, 132.5, 129.5, 128.4, 128.2, 127.0, 126.6, 125.8, 125.3, 123.2, 122.8,
118.9, 117.1, 105.3, 55.1, 50.2, 33.3, 32.7, 28.0, 26.0, 25.1; HRMS (ESI)
for C51H67N4O5Rh2 [M þ H]þ calcd 1021.3222, found 1021.3059; IR
’ EXPERIMENTAL SECTION
General Procedures. Reactions were performed in oven-dried
(140 °C) or flame-dried glassware under an atmosphere of dry N2.
Dichloromethane (DCM) was passed through a solvent column prior to
use and was not distilled. Methanol was not distilled. Thin layer
chromatography (TLC) was carried out using EM Science silica gel
60 F254 plates, and the products on the developed chromatograms were
visualized by UV lamp (254 nm). Column chromatography was
performed using flash chromatography of the indicated system on silica
gel (230ꢀ400 mesh). Arylboronic acids were purchased from Aldrich
and used as received. 1H NMR and 13C NMR spectra were recorded in
CDCl3 or CD2Cl2 on either Bruker Avance 400 or 600 MHz spectro-
meters. Chemical shifts are reported in ppm with the solvent signals as
reference, and coupling constants (J) are given in hertz. IR spectra were
recorded on a Jasco FTIR 4100 spectrometer. Mass spectra were
obtained with a JEOL AccuTOF-CS spectrometer.
(neat) 2923, 2854, 1648, 1572, 1541, 1454 cmꢀ1
.
Procedure for the Synthesis of p-(N-p-Toluenesulfonyl-
pyrolinamido)phenyl Boronic Acid Pinacol Ester (7a). L-Pro-
line (1.00 g, 8.70 mmol) and p-toluenesulfonyl chloride (1.70 g,
8.90 mmol, 1.02 equiv) were dissolved in 30.0 mL of dry DCM. Triethyla-
mine (3.00 mL, 2.50 equiv) was then added dropwise. The resulting
mixture was stirred at room temperature for 16 h. Dilute hydrochloric
Procedure for the Synthesis of Phenyl(p-tolyl)dirhodium-
(III,III) Tetrakis(1-aza-2-cyclooctanoate) (1). Bisacetonitriledirhodium
3626
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627

Organometallics
ARTICLE
acid (0.50 M) was added to the mixture until the aqueous layer was at
pH = 3. The mixture was washed with brine and extracted with DCM.
The organic layer was dried over anhydrous magnesium sulfate and
decanted. After evaporation of DCM, 1.83 g of N-p-toluenesulfonyl-L-
proline was obtained as white solid in 78% yield and, without further
purification, was dissolved in 10.0 mL of dry DCM. Thionyl chloride
(1.00 mL, 2.00 equiv) was added, and the resulting mixture was stirred
under nitrogen at room temperature for 16 h. The solvent was then
evaporated under reduced pressure to give a yellow, viscous liquid that
was dissolved in 30.0 mL of dry DCM. 4-Aminophenylboronic acid
pinacol ester (0.60 g) was added to the DCM solution, followed by
triethylamine (3.00 mL). The resulting mixture was stirred at 0 °C for
1 h. After evaporation of the solvent under reduced pressure, the crude
product was purified by flash column chromatography on silica gel using
2:1 (v/v) hexanes/ethyl acetate as eluent to give 1.13 g (2.40 mmol) of
7a as a pale yellow solid in 69% overall yield from L-proline. The
D-proline and racemic DL-proline derivatives were prepared using the
same procedure and in the same overall yields. 7: 1H NMR (400 MHz,
CDCl3) δ 8.83 (s, 1H), 7.71ꢀ7.76 (dd, 4H, J = 12.0, 8.0 Hz), 7.56ꢀ7.58
(d, 2H, J = 8.0 Hz), 7.32ꢀ7.34 (d, 2H, J = 8.0 Hz), 4.13ꢀ4.16 (m, 1H),
3.55ꢀ3.60 (m, 1H), 3.18ꢀ3.24 (m, 1H), 2.40 (s, 3H), 2.27ꢀ2.32
(m, 1H), 1.48ꢀ1.82 (comp, 3H), 1.30 (s, 12H); 13C NMR (100 MHz,
CDCl3) δ 169.1, 144.7, 140.1, 135.7, 132.5, 130.1, 127.9, 118.7, 83.7,
63.0, 50.1, 29.5, 24.8, 24.4, 21.6; HRMS (ESI) for C24H32BN2O5S
[M þ H]þ calcd 471.2125, found 471.2039; IR (neat) 3265, 2989, 2972,
(ESI) for C64H86N8NaO10Rh2S2 [M þ Na]þ calcd 1419.3916, found
1419.3378; IR (neat) 3342, 2927, 1677, 1597, 1581, 1510, 1342, 1158 cmꢀ1
.
General Procedure for NMR Analyses with Chiral Shift
Reagents. The bisaryldirhodium(III,III) (1-aza-2-cycloctanoate) was
dissolved in CDCl3, and proton NMR data were obtained. Next, addition of
lanthanide shift reagent or silver(I) salt followed portionwise. Proton NMR
spectra were collected after addition of each shift reagent portion.
’ ASSOCIATED CONTENT
S
Supporting Information. NMR spectra of all new com-
b
pounds. Crystallographic data for compounds 1, 4, and 12. This
material is available free of charge via the Internet at http://pubs.
acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: mdoyle3@umd.edu.
’ ACKNOWLEDGMENT
We are grateful to the National Science Foundation (CHE-
0456911) for their generous support. We also wish to thank J.-H.
Xie for his early efforts to develop a detailed understanding of
bisaryl dirhodium(III,III) carboxamidates.
1681, 1598, 1538, 1396, 1342 cmꢀ1
.
Procedure for the Synthesis of p-(N-p-Toluenesulfonyl-
pyrolinamido)phenyl Boronic Acid (8a). Compound 7 (470 mg,
1.00 mmol) and sodium periodate (642 mg, 3.00 equiv) were stirred in
8.00 mL of a 4:1 mixture of THF and water for 1 h. Then aqueous
hydrochloric acid (1.00 M, 0.70 mL) was added to the suspension. The
resulting yellow mixture was stirred at ambient temperature for 6 h, then
diluted with water and extracted with ethyl acetate. The combined
organic layer was washed with water and brine, dried over anhydrous
sodium sulfate, and decanted. After evaporation of the solvent, 360 mg of
a light yellow solid of 8 was obtained in 93% yield (0.93 mmol). D-Proline
and racemic DL-proline derivatives were prepared using the same procedure
and in the same overall yields. 8: 1H NMR (400 MHz, CDCl3) δ 8.94 (s,
1H), 8.17 (d, 2H, J = 8.0 Hz), 7.70ꢀ7.76 (dd, 4H, J = 16.0, 8.0 Hz), 7.36 (d,
2H, J = 8.0 Hz), 4.19ꢀ4.22 (m, 1H), 3.57ꢀ3.65 (m, 1H), 3.22ꢀ3.28 (m,
1H), 2.42 (s, 3H), 2.25ꢀ2.35 (m, 1H), 1.81ꢀ1.84 (m, 1H), 1.67ꢀ1.80
(comp, 2H); 13C NMR (100 MHz, CDCl3) δ 171.2, 144.7, 141.3, 136.7,
134.9, 132.5, 130.1, 127.9, 119.0, 63.0, 60.4, 24.5, 21.6, 21.0; HRMS (ESI) for
C18H22BN2O5S [M þ 1]þ calcd 389.1342, found 389.1316; IR (neat)
’ REFERENCES
(1) Nichols, J. M.; Wolf, J.; Zavalij, P.; Varughese, B.; Doyle, M. P.
J. Am. Chem. Soc. 2007, 129, 3504–3505.
(2) Xie, J.-H.; Zhou, L.; Zavalij, P.; Doyle, M. P.; Sun, Y.-X.; Liu, Y.;
Sun, H. Chem. Commun. 2009, 3005–3007.
(3) Xie, J.-H.; Doyle, M. P. J. Mex. Chem. Soc. 2009, 53, 142–145.
(4) For reviews of molecular propellers see: (a) Mislow, K. Acc.
Chem. Res. 1976, 9, 26–33. (b) Meurer, K. P.; V€ogtle, F. Top. Curr. Chem.
1985, 127, 3–75.(c) Willem, R.; Gielen, M.; Hooggzand, C.; Pepermans,
H. Advances in Dynamic Stereochemistry; Gielen, M., Ed.; Academic
Press: New York, 1985; Vol. 1, p 207. (d) Eliel, E. L.; Wilen, S. H.;
Mander, L. N. Stereochemistry of Organic Compounds; Wiley: New York,
1994; p 1156. (e) Kottas, G. S.; Clarke, L. I.; Horinek, D.; Michl, J. Chem.
Rev. 2005, 105, 1281–1376.
(5) (a) Katoono, R.; Kawai, H.; Fujiwara, K.; Suzuki, T. J. Am. Chem.
Soc. 2009, 131, 16896–16904. (b) Wang, B.; Krꢀal, P. Phys. Rev. Lett.
2007, 98, 266102. (c) Alajarin, M.; Lꢀopez-Leonardo, C.; Bernꢀa, J. Org.
Lett. 2007, 9, 4631–4634. (d) Axe, P.; Bull, S. D.; Davidson, M. G.;
Gilfillin, C. J.; Jones, M. D.; Robinson, D. E. J. E.; Tutner, L. E.; Mithcell,
W. L. Org. Lett. 2007, 9, 223–226.
3378, 3338, 2974, 2954, 1686, 1595, 1532, 1327 cmꢀ1
.
Procedure for the Synthesis of Bis[p-(N-p-toluenesulfo-
nylpyrolinamidophenyl)]dirhodium(III) Tetrakis(1-aza-2-cy-
clooctanoate) (9). Rh2(ACO)4(CH3CN)2 (48.0 mg, 0.060 mmol),
compound 8a (116 mg, 5.00 equiv), and NaHCO3 (50.4 mg, 10.0 equiv)
were dissolved in 6.00 mL of CH2Cl2, and copper(II) sulfate (10.0 mol %)
in 1.50 mL of MeOH was added. The resulting purple mixture was
stirred under air at room temperature for 12 h, during which the solution
turned yellowish-brown. After filtration and evaporation of the solvent
under reduced pressure, the crude product was purified by column
chromatography on silica gel using 1:2 hexanes/ethyl acetate as eluent to
give compound 9 as yellowish-brown solid in 40% yield (34.0 mg,
0.024 mmol). D-Proline and racemic DL-proline derivatives were pre-
pared using the same procedure. 9: 1H NMR (400 MHz, CDCl3) δ 8.72
(s, 2H), 7.73ꢀ7.78 (comp, 4H), 7.34ꢀ7.48 (comp, 8H), 6.85ꢀ6.87
(comp, 4H), 4.14ꢀ4.22 (comp, 2H), 3.78 (s, 4H), 3.58ꢀ3.63
(comp, 2H), 3.19ꢀ3.25 (comp, 4H), 2.44 (s, 6H), 2.26ꢀ2.34 (comp,
4H), 1.34ꢀ1.81(comp, 14H); 13C NMR (150 MHz, CDCl3) δ182.0, 168.5,
156.5, 144.4, 136.3, 132.6, 130.5, 129.9, 128.8, 127.8, 121.6, 119.9, 117.7,
114.0, 62.8, 55.4, 50.0, 33.4, 32.6, 29.6, 28.0, 26.0, 25.0, 24.3, 21.5; HRMS
(6) Xie, J.-H.; Zhou, L.; Lubek, C.; Doyle, M. P. Dalton Trans.
2009, 2871–2877.
(7) Parker, D. Chem. Rev. 1991, 91, 1441–1457.
(8) Behnam, B. A.; Hall, D. M.; Modarai, B. Tetrahedron Lett. 1979,
20, 2619–2620.
(9) (a) Wenzel, T. J.; Bettes, T. C.; Sadlowski, J. E.; Sievers, R. E.
J. Am. Chem. Soc. 1980, 102, 5903–5904. (b) Wenzel, T. J.; Sievers, R. E.
J. Am. Chem. Soc. 1982, 104, 382–388. (c) Offermann, W.; Mannschreck,
A. Tetrahedron Lett. 1981, 22, 3227–3230. (d) Audit, M.; Demerseman,
P.; Goasdoue, N.; Platzer, N. Org. Magn. Reson. 1983, 21, 698–705.
(10) Hartley, F. R. Chem. Rev. 1973, 73, 163–190.
(11) Wenzel, T. J.; Sievers, R. E. Anal. Chem. 1981, 53, 393–399.
(12) (a) Ara, I.; Forniꢀes, J.; Gomez, J.; Lalinde, E.; Moreno, M. T.
Organometallics 2000, 19, 3137–3144. (b) Espinet, P.; Forniꢀes, J.;
ꢀ
Martinez, F.; Sotes, M.; Lalinde, E.; Moreno, M. T.; Ruiz, A.; Welch,
A. J. J. Organomet. Chem. 1991, 403, 253–267. (c) Vicente, J.; Chicote,
M.-T.; Alvarez-Falcꢀon, M. M.; Jones, P. G. Organometallics 2005, 24, -
4666–4675.
3627
dx.doi.org/10.1021/om2003078 |Organometallics 2011, 30, 3619–3627