Studies on adducts of rhodium(II) tetraacetate and rhodium(II) tetratrifluoroacetate with some amines in CDCl3 solution using 1H, 13C and 15N NMR

Article abstract of DOI:10.1016/j.molstruc.2005.03.035

¹H, ¹³C and ¹⁵N NMR spectroscopy has been applied for investigation of amine adducts with rhodium(II) tetraacetate dimer and rhodium(II) tetratrifluoroacetate dimer in CDCl₃ solution. Subsequent formation of two adducts, 1:1 and 2:1, was proved by NMR and VIS titration experiments, and by NMR measurements at reduced temperatures, from 233 to 273 K. The adduct formation shift, defined as Δδ= δ_adduct-δ_ligand and characterizing complexation reaction, varies from ca. 0 to +1.6 ppm for ¹H, from ca. -10 to +6 ppm for ¹³C and from -4.4 to -39 ppm for ¹⁵N NMR. Formation of N-Rh bond slows the inversiof on the nitrogen atom and generates, in the case of N-methyl-(1-phenylethyl)-amine, a nitrogenous chiral center in the molecule. VIS spectra of amine-dirhodium salt mixture contain two bands in the 532-597 nm spectral range, assigned to 1:1- and 2:1-adducts.

Full text of DOI:10.1016/j.molstruc.2005.03.035

Journal of Molecular Structure 750 (2005) 7–17
www.elsevier.com/locate/molstruc
Studies on adducts of rhodium(II) tetraacetate and rhodium(II)
tetratrifluoroacetate with some amines in CDCl3 solution
1
using H, C and N NMR
13
15
´
´
Jarosław Jazwinski
Institute of Organic Chemistry, Polish Academy of Sciences, 01-224 Warszawa. ul. Kasprzaka 44/52, Poland
Received 21 February 2005; revised 31 March 2005; accepted 31 March 2005
Available online 31 May 2005
Abstract
1H, 13C and 15N NMR spectroscopy has been applied for investigation of amine adducts with rhodium(II) tetraacetate dimer and
rhodium(II) tetratrifluoroacetate dimer in CDCl3 solution. Subsequent formation of two adducts, 1:1 and 2:1, was proved by NMR and VIS
titration experiments, and by NMR measurements at reduced temperatures, from 233 to 273 K. The adduct formation shift, defined as DdZ
dadductKdligand and characterizing complexation reaction, varies from ca. 0 to C1.6 ppm for 1H, from ca. K10 to C6 ppm for 13C and from
K4.4 to K39 ppm for 15N NMR. Formation of N–Rh bond slows the inversiof on the nitrogen atom and generates, in the case of N-methyl-
(1-phenylethyl)-amine, a nitrogenous chiral center in the molecule. VIS spectra of amine-dirhodium salt mixture contain two bands in the
532–597 nm spectral range, assigned to 1:1- and 2:1-adducts.
q 2005 Elsevier B.V. All rights reserved.
Keywords: NMR; 1H NMR; 13C NMR; 15N NMR; Amines; Adduct formation shift; Dirhodium complex; Dirhodium tetraacylate; Titration
1. Introduction
is of great importance. Generally, dirhodium dimers are able
to produce various adducts with organic ligands, namely
1:1- and 2:1-adducts with ligands in axial position (Scheme
1) and more complex structures with rearranged dirhodium
cores [1,24]. Although the structure of an adduct in the
crystalline state is well defined and can be determined by
X-ray technique, the composition of the solution is not
obvious, due to ligand exchange and various equilibria
between species.
Rhodium(II) tetraacylate dimers (Scheme 1) are very
well known [1] and numerous papers have proved their
importance. Rhodium(II) tetraacetate was applied by
organic chemist as the homogenous catalyst; rhodium(II)
tetratrifluoroacetate was used as a reagent in the preparation
of isomerically pure a,b-unsaturated carbonyl compounds,
e.g. some papers published recently [2–9] and references
cited herein. Rhodium(II) tetraacylate dimers have been
applied as auxiliary ligands in determination of absolute
configuration of chiral compounds by electronic spec-
troscopy [10–13]. The Mosher’s acid derivative of
rhodium(II) can act as a chiral recognition reagent in
NMR spectroscopy, useful in determination of optical purity
of chiral compounds [14–23].
NMR spectroscopy appeared as a useful tool for the
investigation of dirhodium complexes with organic ligands
containing selenium, sulphur or phosphorus atom [17–23].
Recent 1H, 13C and 15N NMR studies include adducts with
mesoionic compounds and derivatives of pyridines [15,25].
In order to extend these studies, some amines have been
selected as model ligands for the further examination.
There were two purposes of the work. First, is it possible to
detect all species existing in solution by NMR spec-
troscopy? Secondly, how do the NMR parameters reflect
adduct formation? Rhodium(II) tetraacetate dimer Rh2[ac]4
and rhodium(II) tetratrifluoroacetate dimer Rh2[tfa]4 have
been applied as metal substrate; the set of various amines
used, 1–11, are presented on Scheme 1. Since the solvents
containing the N, O or S heteroatom (i.e. DMSO, DMF,
All applications of dirhodium tetraacylates are based on
initial adducts formation with organic molecule, thus the
knowledge of ligation mode and composition of the solution
E-mail address: jarjazw@icho.edu.pl.
0022-2860/$ - see front matter q 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2005.03.035

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
8
(a)
(b)
(c)
R
R
R
R
R
R
O
O
O
O
R
R
R
O
O
O
O
O
O
R
L
Rh Rh
O
O
O
O
O
O
O
O
O
O
O
O
Rh
O
Rh
O
L
L
L
Rh
O
Rh
O
Rh Rh
O
O
O
Rh
O
Rh
O
L
O
O
O
R
O
O
O
O
O
O
R
R
R
R
R
R
R
R
R
adduct 1:1
R = CH3 Rh2[ac]4
R = CF3 Rh2[tfa]4
adduct 2:1
dimer of 1:1 adduct
(d)
1
2
No
R
R'
R''
---------------------------------------
N(C2H5)3
7
8
9
10
R = H
R = C2H5
N
R
1
2
3
CH3 CH3 CH3
C2H5
C2H5 CH3 CH3
3
NR'R''
CH
H
H
2
1
R
CH3
4
5
Ph
Ph
H
H
H
CH3
N
3
4'
NH2
11
1'
6
Ph
CH3 CH3
3'
2'
Scheme 1. Compounds discussed in the present paper. (a) Rhodium(II) tetraacylate dimers; (b) the axial 1:1- and 2:1-adducts of dirhodium(II) tetraacylate and
organic ligands L; (c) dimeric form of 1:1 adduct [30]; (d) amines used as model ligands.
acetone, acetonitrile, methanol) are able to bind the
dirhodium core and compete with amines for the
complexation site, CDCl3 has been used as the solvent of
choice for all experiments.
of amine solution into the tube. Typically, the solutions
containing 0.5:1, 1:1, 1.5:1, 2:1 and 2.5:1 ligand to rhodium
salt molar ratios were measured.
Due to poor solubility of Rh2[ac]4 in CDCl3, the sample
containing Rh2[ac]4 and 0.7 ml of solvent after addition of
each volume of amine stock solution, was placed for a few
minutes in the ultrasonic bath and then stored for a next few
minutes until insoluble material was deposited on the
bottom of NMR tube. Since ligation of metal salt is a driving
force transporting Rh2[ac]4 to the solution, such a procedure
causes different real constitution of the solution than
calculated one, based on quantity of reagents used. As a
consequence, the real composition of the solution was
2. Experimental
1
All H. 13C and 15N NMR measurements were carried
out on a Bruker DRX-500 spectrometer using the XWIN
NMR acquisition and processing program. The 5 mm triple
broadband inverse probe equipped with z-gradient coil was
1
used for H, 13C and 15N NMR measurements. Variable
1
determined by means of signal integration in a suitable H
temperature experiments were recorded with the BVT 3000
temperature unit. Temperatures were read directly from the
instrument panel; no additional temperature corrections
were done.
NMR spectrum. In contrast, solubility of Rh2[tfa]4 in CDCl3
is higher, and the real constitution of the solution is the same
as the quantity of reagents applied.
The amine solution, of ca. 0.04 M, was used for running
of the ligand reference spectra.
Amines and dirhodium salts are commercially available.
In the case of chiral amines racemic mixtures were used.
Amines were dried over KOH pellets and distilled; then a
weighted amount of amine was dissolved in CDCl3 (99.8%
D atom, stabilized by Ag), using 1 cm3 volumetric flask, in
order to obtaining ca. 0.8 M amine stock solution. All
samples were prepared in situ, in the NMR tubes. A
weighted quantity of dirhodium salt, from 6 to 12 mg, was
placed into the NMR tube, and dissolved Rh2[tfa]4 or
suspended Rh2[ac]4 in 0.7 ml of CDCl3. Then a suitable
volume of amine solution was added to the NMR tube by the
25 ml syringe and the sample was measured. All measure-
ments of a given amine and rhodium salt combination were
performed using one NMR sample, by subsequent addition
Residual solvent peaks were used as secondary chemical
shift standards: d(1H)Z7.26 ppm (CDCl3 residual signal),
d(13C)Z77.2 ppm (CHCl3 signal) or d(13C)Z77.0 (CDCl3
central signal). All foregoing chemical shifts are given with
respect to TMS signals, (0 ppm). 15N NMR spectra were
referred to external CH3NO2 signal (0 ppm).
1H NMR spectra were gained with parameters acqui-
sition time 2.5 s, flip angle 308, relaxation delay 1 ms and
spectral width ca. 13 ppm; from 16 to 64 scans were
acquired, depending on the sample. 32K points were used
for data acquisition and Fourier transformation; thus giving
the spectral digital resolution of 0.2 Hz per point.

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
9
1
13C, H and 15N,1H 2D correlation experiments were
tetraacetate Rh2[ac]4, thus quantity of Rh2[tfa]4 added to the
NMR tube dissolved completely. Consequently, the solution
can contain potentially adducts, uncomplexed ligand and
uncomplexed Rh2[tfa]4. Adducts of Rh2[tfa]4 with primary
or secondary amines appeared unstable in the solution and
decomposed during measurement, within few hours. Thus,
with some exceptions, they were not measured. Generally,
NMR spectra of the amine-Rh2[tfa]4 mixture gained at room
temperature contain, with exceptions of 1 and 8, only one
set of averaged signals each. In the case of 1 and 8, the
signals of both adducts were observed even at room
temperature; only temperature increase up to 320 K caused
signal coalescence. Application of low temperature NMR
technique allowed observing the signals of all species in
solution. Such phenomenon was in contrast to the
experiments with Rh2[ac]4, when temperature decrease
usually did not cause the signals separation (see next
Section).
gained by means of inverse gradient technique. 13C,1H
HSQC (phase sensitive, E/A-TPPI gradient selection with
decoupling during acquisition), 13C,1H HMBC (gradient
selection, magnitude mode, no decoupling during acqui-
sition) and 15N, H HMQC (gradient selection, magnitude
1
mode, no decoupling during acquisition) techniques using
‘invietgs’, ‘inv4gplplrnd’ and ‘inv4gplrnd’ Bruker’s pulse
programs were applied.
Typically, a 512!2048 matrix, zero filled to 1024!
2024 was used for 13C,1H 2D HSQC and HMBC spectra,
with parameters acquisition time of 0.27 s, relaxation delay
1.2 s, and 2–8 scans per experiment. Spectral width of
160 ppm in the F2 (13C) domain and 9 ppm in the F2 (1H)
domain were used giving spectral digital resolutions of ca.
20 Hz per point (13C) and ca. 2.2 Hz per point (1H). A 256!
2048 matrix, zero filled to 1024!2048 was used for gaining
15N,1H HMQC correlation spectra. The parameters acqui-
sition time 0.27 s, relaxation delay 1.5 s, a delay from 50 to
60 ms for evolution of long range nJ(15N–1H) coupling, and
spectral width of 9 ppm in the F2 (1H) domain giving
spectral resolution of ca. 2.2 Hz per point was used. Usually,
the measurements were run twice: a preliminary experiment
with spectral width of 200 ppm in F1 (15N) domain, and the
second spectrum with reduced spectral width, 30 ppm,
giving spectral resolution ca. 1.5 Hz per point. Typically,
from 32 to 64 scans per experiment were acquired.
However, in the case of some samples, no signals were
observed after 64 scans per experiment; in such case ca.
800–1000 scans were taken on and only 64!2048 matrix,
zero filled to 512!2024, was used (overnight experiment).
Despite the poor FID resolution in F1 (15N) domain, the
experiments were sufficient for satisfactory results.
19F NMR spectra were conducted on a Varian Mercury
400 apparatus, using 5 mm broadband probe. The spectra
were gained with parameters acquisition time 1.5 s, flip
angle 308, relaxation delay 4 s and spectral width ca.
50 ppm; typically 16 scans were acquired. 57K(64K) points
were used for data acquisition and Fourier transformation,
respectively, giving the resulting spectral digital resolution
of 0.58 Hz per point. The spectra were referred to 19F peak
of external CFCl3 (0 ppm).
1
The H NMR spectra of 8-Rh2[tfa]4 mixtures shown in
Fig. 1 illustrate the typical titration experiment. At the
beginning of titration (0.6:1), the spectrum contains only
1:1-adduct signals. Adding a next portion of the amine
(1.7:1) leads to the appearance of the second set of signals
belonging to the 2:1-adduct. Finally (2.3:1), only the signals
of the 2:1-adduct and the signals of uncomplexed amine are
observed.
The 1H and 13C NMR data of three 1-phenylethylamines,
4, 5 and 6 (Table 1) illustrate the influence of ligand
structure on adduct spectral properties. These amines
possess NH2, NH(CH3) and N(CH3)2 group, respectively.
As it was mentioned, the solution containing Rh2[tfa]4 and
primary or secondary amines was unstable. However, time-
efficient running of experiments allowed obtaining diag-
nostic data also for 4 and 5. The 1H NMR spectra of 4, 5 and
6, conducted at 273 K, contained two sets of signals
deriving from the 1:1- and 2:1-adduct. It is worth noting
that the two N(CH3)2 methyl groups in 6 appeared non-
equivalent in adduct, resulting two signals in NMR
spectrum (Fig. 2). Non-equivalency of methyl groups can
be caused either by influence of the neighbouring chiral
centre, or by hindered rotation around carbon–nitrogen
(Ph)(H)(CH3)C–N bond. Absence of such non-equivalent
methyl groups in the case of achiral tertiary amine 1
suggests rather the former reason. An analogous evidence
(i.e. two non-equivalent methyl groups) was observed for
the 3-Rh2[tfa]4 adduct.
Absorption measurements (VIS) were carried out on a
Varian Cary IE spectrometer using CDCl3 as solvent, in
order to retain the same experimental condition as for NMR
experiment.
However, the question arises why non-equivalency of
N(CH3)2 groups is not observed in the case of Rh2[ac]4
complexes, e.g. for Rh2[ac]4 adducts with 3 or 6 (Table 1).
Different behaviour of amine-Rh2[ac]4 complex can be
rationalized by assuming the fast ligand exchange. Gener-
ally, the NMR signals of individual Rh2[tfa]4 adducts can be
observed quite easily, at relatively high temperature. In
contrast, even low temperature experiments are usually
ineffective for amine-Rh2[ac]4 complexes. Thus, the ligand
exchange seems to be faster in the case of Rh2[ac]4 adducts.
3. Results and discussion
All 1H, 13C and 15N NMR data are collected in Table 1.
3.1. Rh2[tfa]4 adducts with amines
The solubility of rhodium(II) tetratrifluoroacetate
Rh2[tfa]4 in CDCl3 is relatively high relative to rhodium(II)

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
10
Table 1
1H, 13C and 15N NMR data (in ppm) of some amines and their adducts with dirhodium(II) tetracetate dimer Rh2[ac]4 and dirhodium(II) tetratrifluoroacetate
dimer Rh2[tfa]4 in CDCl3 solution
Amine
Atom
Free liganda
L: Rh2[ac]4Z1: 1
L: Rh2[ac]4Z2:1
L: Rh2[tfa]4Z1:1
L: Rh2[tfa]4Z2:1
d (ppm)
DdZdadductKdfree ligand (ppm)b
1
303 Kc
273 Kd
CH
2.59 [54.7]
1.01 [18.5]
2.23 [41.1]
K347.7
1.04 (C4.6)
0.35 (C0.1)
0.52 (C3.1)
K17.2
0.77
0.26
0.41
n.m
1.15 (n.o.)
0.39 (C0.4)
0.58 (C4.0)
K15.0
1.07 (C5.1)
0.37 (K0.5)
0.64(C3.8)
K7.9
CH(CH3)2
N(CH3)2
N
2
303 Kc
CH
2.77 [48.4]
1.02 [23.4]
1.31 [32.8]
0.88 [10.6]
1.16
0.84 (C2.5)
0.50 (K3.0)
0.71, 0.47 (K1.5)
0.27 (K1.0)
[4.31] br
0.88
CH3CH
CH2
0.50
0.70, 0.46
n.m.
n.m
CH3CH2
NH2
0.26
n.o.e
N
K339.5
K14.5
303 Kc
n.m
3
263 Kd
CH
2.36 [60.9]
0.99 (C4.3)
0.41 (K1.3)
[1.43, 2.13] (K0.5)
0.12 (C0.9)
0.53 (C2.4)
0.76
1.09 br (C4.2)
0.48 (K1.5)
[1.90, 1.54 m] (K0.6)
0.04 (C0.5)
0.63 (C5.4);
0.53 (C0.3)
K4.4f
1.09 br (C4.2)
0.45 (K1.5)
CH3CH
CH2
0.93 [13.4]
0.30
1.26, 1.53 [26.2]
0.88 [11.2]
[1.39, 2.00]
0.09
[1.98, 1.54 m] (K0.6)
0.06(C0.5)
0.68(C4.8);
CH3CH2
N(CH3)2
2.21 [40.9]
0.42
0.58(C0.3)
N
K351.1
K13.2
303 Kc
n.m.
4
273 Kd,g
CH
4.12 [51.3]
1.39 [25.9]
[147.7]
0.72 (C2.8)
0.48 (K2.7)
– (K2.4)
0.78
0.48
0.68
0.79
0.47
CH3
0.47
(C10)
H20(C20)
H30(C30)
H40(C40)
NH
7.34 m [125.8, 127.9]
7.34 m
0.29 [126.2]
0.14 [128.9]
0.15 (C0.7)
[4.74]
0.29
0.32
0.32
0.17
0.21
[5.06]
n.m.
0.11
0.17
7.23 m [126.8]
1.54
0.11
0.20
[4.59 br]
n.m.
[5.29]
N
K338.8
K15.2
263 Kh,i
n.m
5
263 Kd,g,i,j
1.33* (K0.7)
0.96 (C3.6)
0.44* (K9.6)
0.30 (K1.5)
0.55* (K3.8)
0.36 (C3.5)
n.m
CH
3.64 [60.2]
1.36 [23.8]
2.31 [34.5]
1.36 (K2.6)
0.90* br (C2.7)
0.43 (K9.8)
0.32* (K0.2)
0.48 (K4.8)
0.37*(C3.1)
n.m.
0.95br
0.34
1.41 (K1.5)
1.03*(C2.7)
0.45(K9.9)
0.33*(K1.6)
0.65 (K4.6)
0.47*(C2.9)
n.m.
CH3
NCH3
0.43
(C10)
– [145.4]
n.m.
H20(C20)
H30(C30)
H40(C40)
NH
7.31 m 128.4, 126.5]
7.31 m
0.31; 0.23 [126.8]
0.18 [128.1]
0.15 (C0.3)
[4.87], [4.58]
K18.2k
0.23
0.30 m [126.9]
0.22 m [128.3]
0.20 m (C1.1)
[5.41*], [5.19]
K16.6*, K17.4
273 Kd
0.30 m [127.3]
0.20 m [128.7]
0.17 m (C1.1)
[5.11], [4.89]
n.m.
0.15
7.24 [126.8]
n.o.e
0.11
[4.58 br]
K10.5k
N
K340.1
6
303 Kc
CH
3.25 [65.9]
1.38 [20.1]
2.20 [43.1]
1.26 (C1.4)
0.28 (K2.5)
0.41 (C0.5)
0.88
0.21
0.30
1.56 (C1.1)
0.29 (K2.2)
0.42 (C5.4)
0.64 (K4.3)
– (K6.9)
1.62 (C0.3)
0.32 (K2.2)
0.51 (C4.8)
O0.69 (K4.7)
– (K6.3)
CH3
N(CH3)2
C10
– [144.0]
– (K4.2)
H20(C20)
H30(C30)
H40(C40)
N
7.30 m [127.4, 128.1]
7.30 m
0.24 [129.7]
0.13 [127.7]
0.15 (C0.6)
K16.1
0.17
0.09
0.09
K9.1
0.31 [130.5]
0.20 [128.3]
0.24 (K0.5)
K19.1
7.23 [126.8]
K342.9
K13.5
7
8
303 Kc
273 Kd
CH2
CH3
N
2.49 [46.3]
1.00 [11.6]
K331.6
0.63 br (C1.7)
0.18 (K2.1)
K19.9
0.38
0.09
n.m.
0.81 (C2.8)
0.22 (K2.7)
K21.6
0.87 (C1.9)
0.20 (K2.7)
K16.0
233 Kd
303 Kd
H1 (C1)
H2 (C2)
H3 (C3)
2.84 [47.8]
1.52 [26.7]
1.72 [20.7]
0.68 (C2.0)
0.26 (K0.6)
0.44 (K0.4)
0.76 (C1.4)
0.20 (K0.4)
0.38 (K0.1)
0.79 (C3.2)
0.61 (C0.1)
0.54 (C0.0)
0.83 (C2.3)
0.54 (C0.3)
0.48 (C0.4)
(continued on next page)

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
11
Table 1 (continued)
Amine
Atom
Free liganda
L: Rh2[ac]4Z1: 1
DdZdadductKdfree ligand (ppm)b
L: Rh2[ac]4Z2:1
L: Rh2[tfa]4Z1:1
K13.7
L: Rh2[tfa]4Z2:1
K8.6
d (ppm)
N
K364.5
223 K
n.o.l
233 Kd
n.o.l
9
H1(C1)
H2 (C2)
H3 (C3)
NH
2.98eq, 2.51ax [46.9]
1.56eq, 1.40ax [26.5]
1.72eq, 1.26ax [24.1]
n.o.
0.81eq, 0.73ax (C2.1)
0.94eq, 0.80ax (C2.3)
[1.98, 1.85] (C1.1)m
[2.03, 1.79] (0.0)m
[4.13]o
n.m.
n.m.
[4.40]n
K24.0
223 Kp
N
K340.5k
K11.1
10
233 K
263 Kp
H1 (C1)
H2 (C2)
H3 (C3)
CH2
2.94; 1.74 [53.9]
1.62, 1.50 [25.7]
1.69, 1.12 [24.3]
2.32 [52.8]
1.06 [12.2]
K326.1
[3.61, 3.27] (K1.8)
[2.14, 1.52m] (K6.3)
[1.91m, 1.65m] (0.0)
1.00 (K6.7)
[3.67, 3.31] (K0.5)
[2.23, 1.67] (K5.4)
[1.98, 1.69m] (0.1)
1.12 (K5.3)
CH3
0.11 (K1.3)
0.12 (K1.2)
N
K39.0
243 Kq
K33.6
243 Kq
11
(q)
(C10)
[146.3]
– (K2.7)
0.69 (C4.2)
0.22 (K0.2)
0.39 (C3.8)
[6.28]
H20(C20)
H30(C30)
H40(C40)
NH2
6.74 [115.1]
7.17 [129.3]
6.77 [118.5]
3.64
0.62
0.19
0.32
[6.17]
K13.3
[7.12]
[6.93]
N
K326.8
K17.0
K25.1
K17.7
Abbreviations: n.m., not measured; n.o, not observed in the spectrum; br, broad signal; m, unresolved multiplet; ax,eq, axial and equatorial hydrogen atom,
respectively.
a
Reference chemical shifts for free (uncomplexed) ligand: 1H chemical shifts (ppm), 13C chemical shifts (squared brackets, ppm) and 15N chemical shifts
(ppm). All reference data were collected at 303 K with exceptions of data for 9 and 10, obtained at 233 K.
b
1H, 13C (in parenthesis) and 15N NMR adduct formation shifts, defined as DdZdadductKdfree ligand (ppm). The suitable chemical shifts (ppm) are given in
square brackets if Dd was not calculated due to lack of the signal assignments.
c
The Dd data for the solutions in equilibrium containing 1:1 and 2:1 mixture of reagents; only one set of the signals at averaged positions is observed (i.e. no
signals of the individual adducts have been detected).
d
The Dd values calculated using the chemical shifts of the individual 1:1- and 2:1-adducts.
The signal too broad to be observed.
e
f
Measurement at room temperature (303 K); the value for the 1.5: 1 equilibrium mixture.
g
The sample is unstable.
h
The signals of 1:1-adduct in the 0.5:1-mixture of reagents have been observed at 263 K. The signals of 2:1-mixture are broad; no signal separation is
observed.
i
Asterisk denotes the major component in the solution (if assigned).
j
No significant differences between the spectra taken at 263 and 303 K.
k
Measurements at room temperature (303 K).
l
No signals after overnight acquisition.
1H chemical shifts taken from 13C–1H 2D correlation spectra. 1H and 13C chemical shifts are similar for both adducts, the signals are not assigned precisely.
m
n
Triplet, 12 Hz.
o
Triplet, 13.5 Hz.
p
The signals are not assigned precisely to the two adducts due to the signal overlapping.
q The 100% 15N-enriched aniline was used. The NH signal appears as a doublet; 1J(15N–1H)Z78 Hz for free aniline, 73 Hz for the Rh2[ac]4 1:1-adduct and
72 Hz for the Rh2[tfa]4 1:1-adduct. The signals at averaged position are observed for the Rh2[ac]4 -aniline mixture; the signals of both adducts are observed for
the Rh2[tfa]4 mixture. However, the signals are broad and not assigned completely.
On the other hand, exchange of the methyl groups via
nitrogen inversion is possible only in the free ligand, not in
the adduct. Consequently, the rate of methyl group
exchange is significantly greater in the case of the
Rh2[ac]4 complex making both methyl groups equivalent.
The most interesting behaviour was noted for the
5-Rh2[tfa]4 complexes, containing a NH(CH3) group.
Formation of the N–Rh bond prevents inversion on nitrogen
atom, and induces new, a nitrogenous chiral centre in the
adduct. Considering that 5 contains a second chiral atom,
two diastereoisomers, RR(SS) or RS(SR) are formed by
binding the amine. If ‘A’ denotes the first diastereoisomer
and ‘B’ the second one, one can expect five adducts,
potentially existing in the solution: A-[Rh–Rh], B-[Rh–Rh],
A-[Rh–Rh]-A, A-[Rh–Rh]-B and B-[Rh–Rh]-B. Conse-
quently, three different signals of ‘A’ are expected on NMR
spectrum: one signal of the 1:1-adduct and two signals of
two diverse 2:1-adducts. The same is expected for ‘B’.
The signal of the N(CH3) methyl group reflects the
formation of all these adducts signal (Fig. 3). Two doublets
at dZ2.67 and 2.86 ppm appearing at the beginning of
titration were assigned to A and B isomers of 1:1-adduct; the
3
signal multiplicity could be explained by the J(NH-CH3)
coupling of 6.4 Hz. Then, under titration, the doublets of

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
12
(a) 8 : Rh2[tfa]4
(b) 8 : Rh2[ac]4
2.2 : 1
2.3 : 1
233 K
3.25
3.50
3.75
1.7 : 1
1.7 : 1
1.1 : 1
0.6 : 1
1.3 : 1
1.0 : 1
uncomplexed
quinuclidine
uncomplexed
quinuclidine
4.00 3.75 3.50 3.25 3.00 2.75
ppm
4.00 3.75 3.50 3.25 3.00 2.75
ppm
Fig. 1. The 1H NMR titration experiment of quinuclidine 8 and rhodium(II) tetraacylate dimers; the a signal of 8 is shown in the spectra. The concentration of
dirhodium salt was constant; the concentration of 8 increased under titration. (a) NMR titration of Rh2[tfa]4, at 303K. The signal of uncomplexed amine and the
signals of two adducts are visible at room temperature; (b) NMR titration of Rh2[ac]4, at 303 K. Inset shows the spectrum taken at 233 K. The signals of both
adducts are visible at low temperature.
1:1-adduct vanished; instead two triplet-like signals
appeared, at dZ2.78 and 2.96 ppm. Assuming that each
triplet consists of two overlapped doublets, one signal can
be assigned to A molecules in 2:1-adducts, and second one
to B molecules. It should be noted that analogous signals of
the NH hydrogen atoms also could be observed in 1H NMR
spectra. (Table 1). The formation of two 1:1 diastereoi-
somers was also observed for 5 and Rh2[ac]4 (Table 1).
The NMR data lead to the general conclusion that
Rh2[tfa]4 tends to form subsequently the 1:1- and 2:1-
adducts, even with tertiary amines. In contrast, Rh2[ac]4
tends to form only the 1:1 adduct with tertiary amines,
under similar experimental condition (see next section).
This finding is in agreement with VIS experiment (see
below; Fig. 5).
3.2. Rh2[ac]4 adducts with amines.
In contrast to Rh2[tfa]4, the solubility of rhodium(II)
tetratrifluoroacetate Rh2[ac]4 in CDCl3 is relatively low. A
real composition of the sample solution depends on the
equilibrium between solution and insoluble rhodium salt,
thus differs from the ratio of components added (see Section
2). For example, a mixture of one equivalent of Rh2[ac]4 and
0.5 equivalent of amine in CDCl3, yielded usually a solution
containing ca. 1:1 molar ratio of components and insoluble
(b)
(a)
6 : Rh2[tfa]4
303 K
1.5 : 1
35
273 K
1.5 : 1
43.1 ppm
δ (13C)
45
ppm
273 K
1.0 : 1
55
273 K
0.5 : 1
2.55
2.75
δ (1H) ppm
2.95
2.9
2.8
2.7
2.6
ppm
Fig. 2. The 1H NMR spectra of 6-Rh2[tfa]4 mixture. The signals of N(CH3)2 group are shown. (a) The signals of either one or two adducts are seen at 273 K,
depending on the ligand to rhodium salt ratio. The absences of 1:1 adduct signals in 1.5:1 mixture of reagents is worth noting. Only averaged signals are visible
at room temperature, but non-equivalency of two methyl groups are retained; (b) 13C-1H correlation spectrum of 1:1 mixture. Opposite Dd(13C) signs of the two
N(CH3)2 methyl groups are observed. The N(CH3)2 groups of uncomplexed amine resonate at dZ43.1 ppm.

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
13
mixture, a high-frequency shift of the 1H signal was
observed. Next, further a high-frequency shift occurred
until an excess of the ligand, more than two equivalents, was
reached. Finally, the a low-frequency shift (shift back) was
noted. Simultaneously, the colour of the solution changed
from blue to violet and finally to red. Such behaviour can be
explained by stepwise formation of two adducts L$Rh2[ac]4
and L2$Rh2[ac]4, and by equilibria: LCRh2½acꢀ4 #L,Rh2
½acꢀ4 and L,Rh2½acꢀ4 CL#L2Rh2½acꢀ4. The equilibrium
was shifted towards the adduct formation; thus there was no
significant concentration of the unbonded amine in the
solution until all uncomplexed dirhodium salt was con-
sumed. After that, free ligand appeared in the solution
5 : Rh2[tfa]4
2.0 : 1
1.5 : 1
1.0 : 1
1
causing ‘shift back’ of the H signal. Such behaviour was
observed also for the amines 2, 4, and 9, i.e. for all primary
and secondary amines.
The signals of 1, 3, 6 and 7 (tertiary amines) exhibited
high-frequency shift at the beginning of titrations. Then,
only ‘shift back’ phenomena were observed under titration,
and the solution remained blue. On addition of very large
ligand excess, ca. 10–20 equivalents, caused colour
transformation to red. Such a result indicates that the
tertiary amines tend to form mainly 1:1-adducts and the
stability of their 2:1-adducts is low with respect to primary
and secondary amines. Quinuclidine (8) forming both
1:1-and 2:1-adducts with Rh2[ac]4 is an exception.
0.5 : 1
3.0
2.8
ppm
2.6
Fig. 3. The 1H NMR spectra of 5-Rh2[tfa]4 mixture at 273 K. The signals of
NCH3 methyl group are shown. Depending on the ligand to rhodium salt
ratio, the signals of two 1:1-adducts (doublets) or the signals of three 2:1-
adducts (triplet-like signals) are visible. For explanation see text.
1
The case of 5-Rf2[ac]4 adduct is worth noting. H NMR
Rh2[ac]4 at the bottom of NMR tube. Due to the poor
solubility of Rh2[ac]4, the solution contained mainly
adducts and eventually uncomplexed amine; there was no
significant concentration of uncomplexed Rh2[ac]4. Then
under titration insoluble dirhodium salt gradually dissolved.
The colour of the solution depends on the molar ratio of
components, and varies from blue (1:1 ratio) via violet
(1.5:1 ratio) to red (2:1 ratio).
spectra of 5-Rf2[ac]4 mixtures, recorded at room tempera-
ture, exhibited only one set of averaged signals. The
spectrum of 5-Rf2[ac]4 as an equimolar mixture, taken at
263 K, contained the signals of two different 1:1-adducts
(1:0.9 ratio). These adducts could be identified as two
1:1-diastereoisomers, like in the case of Rh2[tfa]4
(see previous section). Only broad 1H signals were observed
for a 2:1 Rh2[ac]4-5 mixture despite the low temperature;
thus, the information on individual adducts was lost.
However, the presence of the 2:1-adduct in this mixture
was proofed by VIS data (see below, Fig. 5)
The solution of an amine and Rh2[ac]4 is expected to
contain adducts and uncomplexed ligand to be in equili-
brium, by analogy to Rh2[tfa]4. However, only one set of
averaged signals was observed in NMR spectra conducted at
room temperature, due to the fast ligand exchange (on NMR
time scale). Application of low temperature was supposed to
decrease ligand exchange rate and to allow observing the
signals of individual species. Unfortunately, low tempera-
ture NMR technique appeared, generally, ineffective in the
case of Rh2[ac]4. Temperature reduction resulted either in
signal broadening or made the spectra interpretation
difficult. Quinuclidine (8) appeared to be an exception. At
low temperature (233 K) 1H signals split proving the
presence of two adducts (Fig. 1, right). Low temperature
The NMR data lead to the conclusions that the primary
and secondary amines yield two kinds of adducts with
Rh2[ac]4 in CDCl3, 1:1 and 2:1, similarly to Rh2[tfa]4.
Adducts are formed gradually, depending on the ligand to
Rh2[ac]4 ratio. In contrast to Rh2[tfa]4, tertiary amines tend
to form the 1:1-adducts only with Rh2[ac]4; no significant
concentration of 2:1-adducts was observed, at least in ca.
0.4 mM solutions. The amines 2 and 3 or the amines 4, 5
and 6 can illustrate such feature. The amines possessing
exposed lone pair, like quinuclidine (8), seem to be an
exception.
experiment were successful partly for 5 as well. 1H and 13
C
1
3.3. H and 13C NMR adduct formation shifts
NMR chemical shifts changes remained the sole proof of
complexation in the case of other amines. Generally, two
1
kinds of H NMR signal responses were observed under
The change of a ligand chemical shifts caused by
complexation can be expressed by the adduct formation
titration. As an example of the first manner one can
consider the titration experiment of 8-Rh2[ac]4 (Fig. 1,
right). At the beginning of the titration, i.e. for the 1:1
shift Dd, defined as DdZdadductKduncomplexed
(Table 1). Generally, in the case of Rh2[tfa]4, the signals
ligand

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
14
of both adducts could be observed in NMR spectra at
reduced temperature. The Dd values of Rh2[tfa]4 adducts
given in the Table are related to the individual species. In
contrast, low temperature NMR appeared usually ineffec-
tive in the case of Rh2[ac]4 adducts, due to fast ligand
exchange (with exception of the amine 5 and 8). In
consequence, only the Dd values of equilibrium mixtures
(1:1 and 2:1 ligand to Rh2[ac4] ratio) are given in the Table.
Adduct formation resulted in high-frequency shift of all
1H signals. Positive Dd(1H) values of hydrogen atom
attached to carbon did not exceed 1.6 ppm. However,
Dd(1H) of NH group reached values of ca. 3 ppm. The Dd
magnitude decreased with the hydrogen–nitrogen distance
dZK75.4 ppm was observed for 2:1-mixtures. Considering
the fact that chemical shift offluorine nucleus varies within the
range of hundreds ppm, 19F NMR technique is relatively
insensitive in the case of our model reaction.
3.5. 15N NMR experiments
With a view to the fact that nitrogen atom is directly
involved in adduct formation, 15N chemical shift is
expected to be the most diagnostic NMR parameter
describing the complexation process. The 2D inverse
technique allows obtaining the 15N–1H correlation spectra
using diluted samples containing a few milligrams of
amine. The most accurate and reliable results stem from
experiments showing signals of all species in the
solution, i.e. measurements of Rh2[tfa]4 adducts at low
temperatures.
1
reduction, i.e. the distance from the ligation site. The H
NMR data of Rh2[tfa]4 adducts with 7 and 8 can serve as a
typical example (Table 1). Some irregularities have been
also observed, e.g. Rh2[tfa]4 adducts with 1, 3, 5 and 6. For
these adducts, maximal Dd values from 0.96 to 1.62 ppm
have been noted for the N–CH signal (tertiary hydrogen
atom) whereas smaller values from 0.36 to 0.69 ppm have
been found for NCH3 methyl group, in spite of similar
distance from the nitrogen atom. Very similar Dd values,
from 0.29 to 0.48 ppm, have been observed for farther
N–CH–CH3 methyl group. The 8-Rh2[ac4] adduct is another
example; Dd value for H(3) is greater than for (H2).
Dependence of 13C chemical shifts on the adduct
structure is not so obvious. Dd(13C) values observed appear
in the range from C5.4 to K9.9 ppm. Typically, a few ppm
values, either positive for the a-carbon atoms or negative for
b-carbon atoms have been observed, e.g. in the Rh2[tfa]4
adduct with 7 or the Rh2[ac]4 adducts with 2, 3, 4, 6, 7 or 8.
On the other hand, more complex relationships were also
noted, e.g. 1:1-adducts of Rh2[ac]4 and 5. Dd(13C) values
of K2.6 ppm for the N–CH atom, of K9.8 ppm for the
N–CH–CH3 atom and K4.8 ppm for the NCH3 atom were
found for the one 1:1-adduct, whereas the values of C2.7,
K0.2 and C3.1 ppm were noted for the second one. Large
Dd differences for both diastereoisomers, in particularly of
the CH-CH3 atoms, are of interest. One should point also to
the Rh2[tfa]4 adducts with the amines possessing N(CH3)2
groups, e.g. 6. Two non-equivalent CH3 groups exhibit
positive and negative Dd(13C), respectively, of ca. C5 and
K5 ppm. Thus, sign and magnitude of Dd(13C) depend
rather on the ligand arrangement in the adduct than on the
carbon-nitrogen distance.
Ligation caused significant changes of 15N chemical
shifts. The Dd(15N)Zd(15N)adductKd(15N)free
value
ligand
varied from K4 ppm (3-Rh2[ac]4 adduct), to K39.0 ppm
(10-Rh2[tfa]4 adduct). Typically, Dd values of K10 to
K20 ppm have been observed. There was not straightfor-
ward relationship between amines structure and Dd
magnitudes. If the signals of individual adducts were
detectable, the Dd(15N) magnitudes of 1:1-adduct were
greater than those of the respective 2:1-adducts for a given
amine and dirhodium salt combination (Fig. 4). The typical
difference was equal to ca. 6 ppm. This rule was not fulfilled
for mixtures in equilibria, e.g. mixtures of Rh2[ac]4 with 5 or
6. In the latter case, Dd(15N) was disturbed by the presence
of uncomplexed amine. Such findings exhibit that two
ligand molecules, attached to the two-dirhodium sites, are
not independent. It is worth noting that 15N NMR technique
is able to recognize diastereoismers, e.g. 15N data for two
1:1 5-Rh2[tfa]4 adducts (Table 1).
Previously, Dd(15N) values covering the range from K53
to K75 ppm have been found for pyridines; a value of ca.
K40 ppm has been noted for mesoionic oxatriazole [25].
Thus, the Dd(15N) values observed for rhodium carboxylate
adducts with the amines discussed here are relatively low.
3.6. VIS experiments
Effective colour change of the Rh2[ac]4 or Rh2[tfa]4
solution under titration by amine from blue to violet and
finally to red suggests the use of electronic absorption
spectroscopy in the visible light (VIS) as a supplemental
tool of examination. The measurements were done using
three amines with similar structure, 1-phenylethylamine 4
(primary amine), N-methyl-1-phenylethylamine 5 (sec-
ondary amine) and N,N-dimethyl-1-phenylethylamine 6
(tertiary amine). Some results of titration VIS experiments
are collected on the Fig. 5.
3.4. 19F NMR experiments
19F NMR measurements of 8-Rh2[tfa]4 mixtures, at 303 K,
were acquired for the comparison purposes. A spectrum of
Rh2[tfa]4 in CDCl3 contained one signal, at dZK75.0 ppm.
The spectrum of amine-Rh2[tfa]4 mixtures contained two
signals, at dZK75.0 ppm (Rh2[tfa]4) and K75.2 ppm (the
1:1-adduct). The 1:1 mixture of components resulted in two
signals in the spectrum, at dZK75.2 ppm (the 1:1-adduct)
and K75.4 ppm (the 2:1-adduct). Finally, only one signal at
Typically, the red shifted band, within the range
575–597 nm appeared at the beginning of titration. Then,
under titration, this band vanished, instead of that blue

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
15
(a)
(c)
(b)
1 : 1
1 : 1 2 : 1
2 : 1
Uncomplexed 7
–355
–350
–345
–340
–335
–330
–325
–353.2
–353.2
–347.6
–347.6
δ (15N)
ppm
–331.6
0.9
0.9
0.9
1.3
1.3
1.1
δ (1H) ppm
1.3
1.1
1.1
Fig. 4. Example of the 15N–1H correlation experiments of 7-Rh2[tfa]4 mixtures, at 263 K. Depending on ligand to rhodium salt ratio, the signals of 1:1 adduct
(a), 1:1- and 2:1-adducts (b) 2:1-adduct and uncomplexed ligand (c) are visible.
shifted band was formed, in 532–547 nm. These signals
were assigned to the 1:1 and 2:1 adduct, respectively.
Isobestic point clearly indicated the subsequent formation of
both 1:1 and 2: 1 adducts. The spectra set of 6-Rh2[ac]4
mixture was an exception; in this case practically one band
was noted, at 597 nm. In fact, slight blue band shift under
titration can be observed, but the equilibrium is clearly
shifted towards the 1:1-adduct. Thus, VIS titration
(b) 0.5
0.8
0.6
(a)
4-Rh2[ac]4
5-Rh2[ac]4
547 nm
582 nm
0.4
0.3
0.2
532 nm
578 nm
1 : 1
1 : 1
1.5 : 1
2 : 1
Abs
Abs
1.5 : 1
0.4
2 : 1
2.5 : 1
0.2
0.0
0.1
0.0
400
500
600
λ (nm)
700
800
400
500
600
700
800
λ (nm)
(c) 0.5
0.4
(d)
0.4
0.3
6-Rh2[ac]4
6-Rh2[tfa]4
597 nm
594 nm
1 : 1
2 : 1
1.5 : 1
2 : 1
0.3
Abs
Abs
535 nm
0.2
0.1
0.0
575 nm
0.2
0.5 : 1
1 : 1
Rh2ac4
x 10
0.1
0.0
1.5 : 1
400
500
600
700
800
400
500
600
λ (nm)
700
800
λ (nm)
Fig. 5. VIS absorption electronic spectra (titration experiment) of Rh2[ac]4, Rh2[tfa]4 and 4–6. The concentration of the dirhodium salt was constant; the amine
concentration was increased; ca. 7 mM solutions of dirhodium salt in CDCl3 were used. The formation of two adducts is clearly visible (a, b and d). The
formation of mainly 1:1-adduct is observed for 6-Rh2[ac]4 (c). Tenfold enlarged spectrum of Rh2[ac]4 in CDCl3 (saturated solution) is shown for comparison
purposes (c).

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
16
experiments are in agreement with NMR findings and
provide information on adduct stoichiometry.
adducts in solution. In contrast, low temperature
technique is less effective in the case of Rh2[ac]4
adducts. Temperature decrease causes often broad-
ening of NMR signals only. However, there are
exceptions from this rule.
3.7. The adduct structure
The structure of the 2:1-adduct in the solution is rather
clear. The most possible structure consists of the dirhodium
(iii) It is known [31] that dynamic effects visible in NMR
spectrum caused by ligand exchange depend on ligand
exchange rate, Dd values and temperature. Since Dd
values (adduct formation shifts) are comparable for
both Rh2[ac]4 and Rh2[tfa]4 adducts (for the same
amine), different behaviour of Rh2[ac]4 and Rh2[tfa]4
adducts with respect to temperature change suggests
slower ligand exchange in the case of Rh2[tfa]4.
(iv) Formation of Rh–N bond slows down structure inver-
sion of nitrogen atom. Hindered inversion causes, in
certain cases, either non-equivalency of two N(CH3)2
methyl groups or formation of nitrogenous chiral centre
in the case of secondary amines (R0R0 0NH).
1
core and two axially bonded ligands. H NMR spectra of
Rh2[ac]4-adducts contained only one CH3 signal of Rh2[ac]4
core revealing equivalency of all methyl groups and
confirming axial ligation mode. The same was observed in
19F NMR spectroscopy for Rh2[tfa]4 adducts. Such ligand
arrangement was found, by X-ray, for the following
adducts: Rh2(tBuCOO)4 with NEt3 [26], Rh2[tfa]4 with
PPh3 [27], and Rh2[ac]4 with NH(C2H5)2 [28] or pyridine
[29]. One can expect that the adduct structures in solution
and in solid state are similar.
Axial ligation mode is expected also for 1:1-adducts.
However, the structure of the whole adduct is not obvious. It
is known that the dirhodium salt tends to form 2:1-adducts.
The unoccupied site of a 1:1-dirhodium complex usually
bonds either a solvent or a water molecule or just an oxygen
atom of the second dirhodium unit [1]. In the crystalline
state, the dimeric 2:2 structure has been found in the case of
Rh2[ac]4 and PhNC (Scheme 1). As similar dimeric
structure was observed for the adduct of Rh2[ac]4 with
triphenylphosphine [30].
(v) The adduct formation shift DdZdadductKdligand (ppm)
is a useful NMR parameter characterizing the binding
process. Dd(1H) is positive and does not exceed the
value of 2 ppm. The magnitude of Dd(13C) does not
exceed a few ppm; the sign of Dd(13C) is either positive
or negative. The Dd(15N) values of the nitrogen atom,
involved in binding, occur within the range from K4 to
K39 ppm. Generally, the magnitude of Dd(15N) of 1:
1-adducts is a few ppm lower than that of 2:1-adducts.
(vi) Typically, VIS spectra of the rodium tetracarboxylate-
amine mixture taken on in CDCl3 contain two bands in
the 532–597 nm spectral range. Red bands can be
assigned to 1:1-adducts, whereas blue shifted bands
refer to 2:1 adducts. VIS titration experiments exhibit-
ing isobestic point confirm consecutive formation of
two adducts.
Our NMR and VIS data does not allow to state what is the
real structure of the 1:1-adduct in the solution, monomeric or
dimeric. Precisely speaking, in the present paper the 1:1-
adduct means dirhodium core with one site occupied by one
molecule of amine, while the 2:1-adduct means one
dirhodium core containing two molecules of amine.
1
The H NMR data of the Rh2[tfa]4 with piperidine 9
adduct provide some information on this ligand’s arrange-
ment. The NH signal of the amine shows triplet-like
structure, with ‘large’ coupling constant of ca. 13 Hz. Such
multiplicity can be explained by vicinal coupling of axial
NH hydrogen with the two axial a-hydrogens in the ring.
Consequently, the dirhodium unit is attached to the ring via
equatorial N–Rh bond.
Acknowledgements
This work was supported by the State Committee for
Scientific Research under Grant No. 4 T09A 060 25.
3.8. Conclusions
References
The results of NMR and VIS measurements can be
summarized as follows:
[1] F.A. Cotton, R.A. Walton, Multiple bond between metal atom, second
ed., Clarendon Pres, Oxford, 1993 (Chapter 7, pp. 431).
[2] S. Muthusamy, C. Gunanathan, M. Nethaji, Synlett 4 (2004) 639.
[3] D. Lee, R. Otte, J. Org. Chem. 69 (2004) 3569.
(i) In titration experiments, amines and dirhodium salts,
Rh2[ac]4 and Rh2[tfa]4, form 1:1- and-2:1 adducts
subsequently. Tertiary amines with Rh2[ac]4 tend to
form only 1:1-adducts under similar experimental
conditions, i.e. in 0.2O0.4 mM solutions. Quinucli-
dine (8) containing exposed lone electron pair and
giving 2:1-adduct with Rh2[ac]4, is an exception.
[4] B. Ronan, E. Bacque, J. Barriere, Tetrahedron 60 (2004) 3819.
[5] J. Yadar, B. Reddy, P. Reddy, Adv. Synth. Catal. 346 (2004) 53.
[6] E. Cuevas-Yanez, J. Muchowski, R. Cruz-Almanza, Tetrahedron 60
(2004) 1505.
[7] N. Karsche, S. Jaschak, D. Dhavale, J. Org. Chem. 68 (2003) 4531.
[8] N. Karsche, S. Jaschak, D. Dhavale, J. Org. Chem. 66 (2003) 6323.
[9] A. Padwa, C. Straub, J. Org. Chem. 68 (2003) 227.
´
[10] J. Frelek, A. Klimek, P. Ruskowska, Curr. Org. Chem. 7 (2003) 1081.
´
[11] J. Frelek, J. Jagodzinski, H. Meyer-Figge, W.S. Sheldrick,
(ii) NMR measurements conducted at low temperatures
allow observing the signals of both Rh2[tfa]4-amine
E. Wieteska, W. Szczepek, Chirality 13 (2001) 313.

´
´
J. Jazwinski / Journal of Molecular Structure 750 (2005) 7–17
17
[12] J. Frelek, Tetrahedron: Asymmetry 10 (1999) 2809.
[13] J. Frelek, W. Szczepek, Tetrahedron-Asymmetry 10 (1999) 1507.
[14] S. Rockitt, H. Duddeck, K. Kiec-Konowicz, Eur. J. Org. Chem. 20
(2000) 3489.
[22] D. Magiera, J. Omelanczuk, K. Dziuba, M. Pietrusiewicz,
H. Duddeck, Organometallics 22 (2003) 2464.
[23] D. Magiera, A. Szmigielska, M. Pietrusiewicz, H. Duddeck, Chirality
16 (2004) 57.
´
´
[15] J. Jazwinski, Z. Rozwadowski, D. Magiera, H. Duddeck, Magn.
[24] J. Telser, R.S. Drago, Inorg. Chem. 23 (1984) 2599.
´
´
Reson. Chem. 41 (2003) 315.
[25] J. Jazwinski, H. Duddeck, Magn. Reson. Chem. 41 (2003) 921.
[26] M. Handa, M. Watanabe, D. Yoshioka, S. Kawabata, R. Nukada,
M. Mikuriya, H. Azuma, K. Kasuga, Bull. Chem. Soc. Jpn. 72 (1999)
2681.
[16] S. Hameed, R. Ahmad, H. Duddeck, Magn. Reson. Chem. 36 (1998)
S47.
[17] S. Hameed, R. Ahmad, H. Duddeck, Heteroatom Chem. 9 (1998)
471.
[27] F.A. Cotton, T.R. Felthouse, S. Klein, Inorg. Chem. 20 (1981) 3037.
[28] Y.-B. Koh, G.G. Christoph, Inorg. Chem. 18 (1979) 112.
[29] N. Mehmet, D.A. Tocher, Inorg. Chim. Acta 188 (1991) 71.
[30] G.G. Christoph, D.J. Kountz, Cambridge Data Base, [ACA, Ser. 2 10
(1982) 23].
[18] C. Meyer, H. Duddeck, Magn. Reson. Chem. 38 (2000) 29.
[19] S. Rockitt, H. Duddeck, J. Omelanczuk, Chirality 13 (2001) 214.
[20] S. Malik, H. Duddeck, J. Omelanczuk, M. Choudhary, Chirality 14
(2002) 407.
¨
[31] J. Sandstrom, Dynamic NMR spectroscopy, Academic Press, London,
[21] D. Magiera, S. Moeller, Z. Drzazga, Z. Pakulski, M. Pietrusiewicz,
H. Duddeck, Chirality 15 (2003) 391.
1982.