Chiral Phosphate in Rhodium-Catalyzed Asymmetric [2+2+2] Cycloaddition: Ligand, Counterion, or Both?

Article abstract of DOI:10.1002/chem.201601188

Investigations based on NMR spectroscopy, mass spectrometry, and DFT calculations shed light on the metallic species generated in the rhodium-catalyzed asymmetric [2+2+2] cycloaddition reaction between diynes and isocyanates with the chiral phosphate TRIP. The catalytic mixture comprising [{Rh(cod)Cl}₂], 1,4-diphenylphosphinobutane (dppb), and Ag(S)-TRIP actually gives rise to two species, both having an effect on the stereoselectivity. One is a rhodium(I) complex in which TRIP is a weakly coordinating counterion, whereas the other is a bimetallic Rh/Ag complex in which TRIP is a strongly coordinating X-type ligand.

Full text of DOI:10.1002/chem.201601188

DOI: 10.1002/chem.201601188
Full Paper
&
Organometallic Chemistry |Hot Paper|
Chiral Phosphate in Rhodium-Catalyzed Asymmetric [2+2+2]
Cycloaddition: Ligand, Counterion, or Both?
Marion Barbazanges,*[a] Elsa Caytan,[b] Denis Lesage,[a] Corinne Aubert,*[a]
Louis Fensterbank,*[a] Vincent Gandon,[c, d] and Cyril Ollivier[a]
Abstract: Investigations based on NMR spectroscopy, mass
spectrometry, and DFT calculations shed light on the metallic
species generated in the rhodium-catalyzed asymmetric
[2+2+2] cycloaddition reaction between diynes and isocya-
nates with the chiral phosphate TRIP. The catalytic mixture
comprising [{Rh(cod)Cl}2], 1,4-diphenylphosphinobutane
(dppb), and Ag(S)-TRIP actually gives rise to two species,
both having an effect on the stereoselectivity. One is a rho-
dium(I) complex in which TRIP is a weakly coordinating
counterion, whereas the other is a bimetallic Rh/Ag complex
in which TRIP is a strongly coordinating X-type ligand.
phate [TRIP]À as the chirality carrier[9] constituted a real break-
through, as high enantioselectivities could be achieved for a va-
riety of transformations.[5,10] However, the exact nature of the
phosphate chirality carrier, that is, X-type ligand or counterion,
is often a source of controversy. If it remains bound to the
metal during the stereochemistry-determining step, then its
mode of action is not very different from a classical L-type
monodentate ligand. If it is well separated from the metal
center (loose ion pair), then the transfer of stereochemical in-
formation would require an anchoring of the counterion on
the ligands or on the substrate itself.[11] In this case, the induc-
tion models should be completely distinct. To date, no report
has brought firm evidence on this issue. The strong depen-
dence of the ee on solvent polarity suggests that a large inter-
ionic distance is deleterious to the chirality transfer but does
not disclose the actual location of the phosphate.[12] X-ray
structure analyses generally reveal short metal–oxygen bonds,
in agreement with strong covalent bonding.[13] However, this
does not mean that such bonds remain covalent or undergo
solvation during catalysis. For instance, although the phos-
phate–gold complexes A and B were proved to be covalent in
CD2Cl2 (Figure 1),[13b,14] the phosphate moiety is believed to be
a “true” counterion in catalysis due to the linear geometry of
the [(R3P)Au(substrate)]+ complex C, which imposes the phos-
phate keeping off the coordination sphere of the metal to
allow substrate activation.[8] This reasoning based on geometri-
cal constraints cannot be so obviously applied with metal ions
of Ir, Pd, Cu, Ru, or Rh.[15] In these cases, DFT studies have been
helpful to support a counterion role of the phosphate on the
Ir complex D,[10c] and on the palladium complex E (Figure 1).[16]
However, to date, no physical evidence on the counterion or
ligand nature of the phosphate has been reported.
Introduction
Enantioselective metal-catalyzed transformations are a corner-
stone of modern synthetic organic chemistry.[1,2] Over the last
decade, the limits of this discipline have been pushed even
further, notably through the design of original ligands[3] or the
use of multiple stereodifferentiation approaches.[4] Recent
years have also witnessed the rise of the asymmetric counter-
ion-directed catalysis (ACDC) approach, whereby the counter-
ion of a catalytically active cationic metal species is employed
as vehicle for the transfer of stereochemical information to the
reaction products.[5] Disclosed in 2000 by Arndtsen and co-
workers for a copper-catalyzed borate-assisted styrene aziridi-
nation,[6] ACDC has been subject to intense interest following
the reports by the groups of Krische, List, and Toste in 2006
and 2007.[7,8] The use of the BINOL-derived hindered phos-
[a] Dr. M. Barbazanges, Dr. D. Lesage, Dr. C. Aubert, Prof. L. Fensterbank,
Dr. C. Ollivier
Sorbonne UniversitØs, UPMC Univ. Paris 06
Institut Parisien de Chimie MolØculaire (UMR CNRS 8232)
Case Courrier 229, 4 place Jussieu, 75252 Paris Cedex 05 (France)
E-mail: marion.barbazanges@upmc.fr
corinne.aubert@upmc.fr
louis.fensterbank@upmc.fr
[b] Dr. E. Caytan
Institut des Sciences Chimiques de Rennes, UniversitØ de Rennes 1
Campus de Beaulieu, 35042 Rennes Cedex (France)
[c] Prof. V. Gandon
ICMMO, Univ. Paris-Sud, CNRS UMR 8182
UniversitØ Paris-Saclay, 91405 Orsay cedex (France)
[d] Prof. V. Gandon
ICSN, CNRS UPR 2301, Univ. Paris-Sud
UniversitØ Paris-Saclay, 91198 Gif-sur-Yvette (France)
Supporting information for this article, containing full NMR and ESI charac-
terizations for complexes [I]+[(S)-TRIP]À, II and III, experimental procedures,
HPLC traces; kinetic studies, and coordinates and energy of the calculated
species, is available on the WWW under http://dx.doi.org/10.1002/
chem.201601188.
We describe herein a study aimed at shedding light on the
issue of the ligand or counterion nature of a chiral phosphate
on catalytically active square planar Rh+ complexes. We re-
cently reported the first case of chiral phosphate-directed
Chem. Eur. J. 2016, 22, 8553 – 8558
8553
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the silver salt. Such a complex could adopt various coordina-
tion modes. Even though the square planar geometry should
be always favored, the phosphate could be an X-type ligand if
cod or dppb acts as a monodentate ligand. On the other hand,
if both cod and dppb allow an h2,h2-Rh complex, then the
phosphate should be
a
counterion as in [I]+[(S)-TRIP]À
(Figure 2) and the complex could be either a tight or a loose
ion pair. Under the optimized conditions, Ag(S)-TRIP is used in
a 1.5-fold excess compared to Rh. If the anion metathesis is
complete, 2.5 mol% of Ag(S)-TRIP should still be present. Since
the silver complex is not catalytically active, the impact of the
presence of this salt on the stereoselectivity is puzzling. Thus,
we decided to study carefully the species contained in this cat-
alytic mixture,[17] focusing on the nature, ligand or counterion,
of the chiral phosphate.
Figure 1. Ligand vs. counterion character of chiral phosphates in transition
metal complexes.
Results and Discussion
asymmetric [2+2+2] cycloaddition (Scheme 1).[10b] We have
shown that diynes 1 and isocyanates 2 lead to enantiomeri-
cally enriched pyridones 3 in up to 91:9 e.r., in the presence
of the phosphate [(S)-TRIP]À (Figure 2) as the sole chirality
source of a catalytic mixture comprising [{Rh(cod)Cl}2] (cod=
1,5-cyclooctadiene), 1,4-diphenylphosphinobutane (dppb), and
Ag(S)-TRIP. A likely component of this mixture would be
[Rh(cod)(dppb){(S)-TRIP}] formed after chloride abstraction by
We first examined the species formed by mixing [{Rh(cod)Cl}2]
(0.5 equiv), dppb (1 equiv) and Ag(S)-TRIP (1 equiv) in (CH2Cl)2
at 808C under argon for 15 min (sealed NMR tube), which cor-
responds to the premix conditions for the cycloaddition reac-
tion shown in Scheme 1. 1D and 2D NMR experiments were in
agreement with the proposed structure [I]+[(S)-TRIP]À
(Figure 2; see the Supporting Information, sections 2.1 to 2.3).
The formation of this species at room temperature in CD2Cl2
was also ascertained by NMR spectroscopy.[18] In particular, the
31P spectrum shows two signals, one accounting for the dppb
ligand [d=24.6 ppm, d, J(103Rh,31P)=144 Hz] and the other for
1
the phosphate (broad s, d=4.8 ppm).[18] The H–103Rh HMQC
spectrum shows a triplet at d=À8466 ppm.[19,20a] This is consis-
tent with a [Rh(P)2(alkene)n]-type complex.[21] ESI mass spec-
trometry experiments carried out on the mixture also con-
firmed this hypothesis, as both cation [I]+ (positive mode, m/z
637, Figure 3) and anion [(S)-TRIP]À (negative mode, m/z 751)
were detected (see the Supporting Information, section 2.4).
The question of the interaction between [I]+ and [(S)-TRIP]À
remained to be answered. In addition to conductivity measure-
ments and IR and 1D NMR spectroscopy,[6,13b] a powerful tool
Scheme 1. Asymmetric counterion-directed Rh-catalyzed [2+2+2] cycloaddi-
tion.[10b]
Figure 3. Simulated and experimental ESI-MS (positive mode) of species
[I]+[(S)-TRIP]À.
Figure 2. Species involved in this study.
Chem. Eur. J. 2016, 22, 8553 – 8558
www.chemeurj.org
8554
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
to evidence ion pairing is diffusion ordered spectroscopy
(DOSY). This technique has been extensively used by the
groups of Pregosin and Macchioni to study cationic organome-
tallic species.[22] To date, application of this method to com-
plexes relevant to ACDC is limited to a single example:[23] for
[Rh(cod)(BINAP)][B(BINOL)2], Brown and co-workers observed
the same diffusion coefficients for [Rh(cod)(BINAP)] and
[B(BINOL)2]. They therefore concluded that this complex is the
tight ion pair [Rh(cod)(BINAP)]+[B(BINOL)2]À in solution, the as-
sembly of the two ionic entities being achieved through non-
covalent interactions. To definitively settle the ligand or coun-
terion character of (S)-TRIP in [I][(S)-TRIP], we carried out three
NMR experiments: i) 1D HMQC 31P–103Rh to detect long-range
couplings between the phosphate and the rhodium nucleus;
ii) 2D NOESY and ROESY to detect any noncovalent interactions
between the phosphate and the ligands; iii) DOSY (Figure 4).
Scheme 2. Evaluation of [I]+[(S)-TRIP]À in enantioselective [2+2+2] cycload-
dition.
species discussed above, was tested in a [2+2+2] cycloaddi-
tion between diyne 1a and isocyanate 2a and led to the de-
sired pyridone (À)-3a in 79% yield (Scheme 2, conditions B).
Surprisingly, a decrease of the enantiomeric ratio from 85:15 in
the ACDC-optimized conditions (Scheme 2, conditions A) to
75:25 was observed. The results of conditions A and B were re-
producible (see the Supporting Information, section 2.5).
The influence of the amount of Ag(S)-TRIP on the rhodium
complexes was then assessed by 31P NMR spectroscopy. When
1 equivalent of Ag(S)-TRIP vs. Rh was used (0.5 equiv
[{Rh(cod)Cl}2], 1 equiv dppb, 1 equiv Ag(S)-TRIP), the major
product was [I]+[(S)-TRIP]À, as discussed above (Figure 5, spec-
trum 1). When adding more Ag(S)-TRIP, the latter does not ac-
cumulate in the reaction mixture but generates two new spe-
cies II and III. Their proportions grow with increasing amounts
of Ag(S)-TRIP (Figure 5, spectra 2–6). Disappearance of [I]+[(S)-
TRIP]À is virtually achieved with 2.2 equivalents of Ag(S)-TRIP
(spectrum 4).
Species II shows four different sets of signals, at d=21.6 [d,
1J(103Rh,31P)=152 Hz], 12.8 [d, 1J(109Ag,31P)=820 Hz; d,
1J(107Ag,31P)=710 Hz], 7.1 [br s, P(O)OAg], and 3.5 ppm [s,
P(O)]. It is a highly dissymmetric complex, bearing one excep-
Figure 4. DOSY experiment (CD2Cl2, 600 MHz, 300 K).
The first two experiments showed no phosphate–103Rh correla-
tion signal (see the Supporting Information, section 2.2) or
phosphate–ligand interaction (see the Supporting Information,
section 2.6). From DOSY, we derived the following diffusions
coefficients: D[TRIP] =7.5EÀ10 m2 sÀ1 (i.e., a hydrodynamic radius
of rH =7.2 Š); D[I] =8.7EÀ10 m2 sÀ1 (i.e., rH =6.2 Š; see the Sup-
porting Information, section 2.6).[24,25,26] Since the two frag-
ments have different diffusion coefficients, it is clear that they
move independently. On the basis of these three experiments,
it can be concluded that [(S)-TRIP]À can be qualified as a coun-
terion and that complex [I]+[(S)-TRIP]À behaves as a loose ion
pair in solution.
[I]+[(S)-TRIP]À was independently synthesized to evaluate its
catalytic activity and stereoselectivity. This isolated complex,
the NMR data of which fit with those of the in situ-generated
Figure 5. 31P NMR monitoring of the metallic species as a function of Ag(S)-
TRIP equivalents.[18]
Chem. Eur. J. 2016, 22, 8553 – 8558
www.chemeurj.org
8555
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tionally shielded cod double bond [1H NMR: d=2.80 (1H),
2.63 ppm (1H); 13C NMR: d=70.5, 68.0 ppm]. Its 103Rh NMR
signal splits as a doublet (rather than a triplet as in species [I]+
[(S)-TRIP]À) and shifts 993 ppm downfield, from d=À8644 ppm
(species [I]+[(S)-TRIP]À) to d=À7651 ppm.[19,20b] Both represent
the replacement of one of the phosphane by an oxygen
ligand.[21] The latter downfield shift would probably arise from
decoordination of one end of the dppb ligand by excess Ag(S)-
TRIP and concomitant coordination of the phosphate. Indeed,
Figure 6. Evolution of the species in the reaction mixture [31P NMR, (CD2Cl)2,
201 MHz, 343 K, c=0.005m].
31P–109Ag HMQC revealed
a
doublet at d=426 ppm
[1J(31P,109Ag=820 Hz)], consistent with a two-coordinate linear
silver complex.[27] The postulated [Rh(cod)(dppb){(S)-TRI-
P}{Ag(S)-TRIP}] structure II (Figure 2) matches 1D and 2D ex-
periments (see the Supporting Information, sections 3.1 to 3.3).
The 31P–103Rh 1D HMQC experiment confirmed that, in this
case, the phosphate is an X-type ligand (see the Supporting In-
formation, spectrum SI-16).[28] Moreover, the bulky cationic
structure [II-(S)-TRIP]+ was detected by ESI(+) analysis (see the
Supporting Information, section 3.5). Note that only traces of
the latter were detected, which is consistent with the low ionic
character of the Rh-TRIP bound in this complex. This cation is
pretty stable, as increasing the energy does not lead to subse-
quent fragmentation. In particular, cation [I]+ (m/z 637) was
not detected by MS/MS analysis.
The NMR data of species III reveal a symmetric structure
with two phosphorus patterns at d=11.8 [d, J(109Ag,31P)=
820 Hz and d, J(107Ag,31P)=710 Hz], and 7.7 ppm [br s,
P(O)OAg]. The postulated bis-silver structure [(dppb)(Ag(S)-
TRIP)2] for this complex was corroborated by 1D and 2D ex-
periments, as well as ESI(+)-MS which revealed ion peaks for
[III+Ag]+ at m/z 2253.9 and [IIIÀTRIP]+ at m/z 1393.5 (see the
Supporting Information, section 4). The same NMR features
were recorded for a mixture of dppb with 2 equivalents of
Ag(S)-TRIP in CD2Cl2 at RT. Notably, this independently synthe-
sized complex is catalytically inactive.
Scheme 3. 1H EXSY study of the equilibrium between species [I]+[(S)-TRIP]À/
Ag(S)-TRIP and II (300 K).
[(MeO)2PO2]À as model phosphate (Scheme 4). The calculations
were carried out at the M06/SDD(Rh,Ag)6-31+G(d,p)//B3LYP/
LANL2DZ(Rh,Ag)6-31Gd) level using the Gaussian 09 software
package (see the Supporting Information, section 6).
As shown by the quite long CHÀO distances of 2.06 and
2.08 Š, respectively, in the calculated [I][(MeO)2PO2] ([I]
[(MeO)2PO2]-calc), the binding of the phosphate to [I]+ is
achieved through weak hydrogen bonds in the gas phase. One
phenyl moiety and one CH2 unit of dppb serve as hydrogen
bond donors. Notably, another isomer in which the cod ligand
plays the role of hydrogen bond donor was also computed
but it was found to be less stable than [I][(MeO)2PO2]-calc by
3 kcalmolÀ1. The computed DG298 for [I][(MeO)2PO2]-calc+
Ag(MeO)2PO2!II-calc is of À40.4 kcalmolÀ1. This large exergo-
nicity supports the proposed phosphane exchange.
We then monitored the evolution of the species in condi-
tions close to those of the reaction, in the presence of the
diyne and isocyanate reagents [(CD2Cl)2, 708C] with 2 equiva-
lents of Ag(S)-TRIP (10 mol%) vs. Rh, starting from a [I]+[(S)-
TRIP]À/II/III ratio of 1:3.5:1 (Figure 6). A 2:1.4:1 ratio was
reached after 4.5 h. It is worth underlining that during this
study, no reaction intermediate or other catalytic species could
1
be detected by 31P or H NMR spectroscopy
1H exchange spectroscopy (EXSY) showed that species [I]+
[(S)-TRIP]À and II are indeed in equilibrium in the reaction mix-
ture, that is, coordination/decoordination of silver phosphate
occurs with exchange rate constants k1 =0.012Æ0.001 sÀ1 and
We then evaluated these rhodium species in the [2+2+2]
cycloaddition reaction of diyne 1a with isocyanate 2a under
the conditions defined in Figure 7. As shown above, with
5 mol% of Ag(S)-TRIP, the catalytic mixture is mostly composed
of [I]+[(S)-TRIP]À. With 10 mol% of Ag(S)-TRIP, it is enriched in
species II. Both catalytic systems led to full conversion in
about 1 h at 708C. However, the initial rate was higher with
5 mol% Ag(S)-TRIP.
k
À1 =0.049Æ0.01 sÀ1 at 300 K (Scheme 3). A conformational
equilibrium was also revealed by the observation of the NMR
signals of the cod protons which exchange in pairs. Since the
dppb protons do not exchange, this equilibrium can be ex-
plained by the flip of the Rh–dppb–Ag(S)-TRIP side chain,
above and below the rhodium atom. This second equilibrium
occurs with an exchange rate constant of k’=0.034Æ0.005 sÀ1
(see the Supporting Information, section 5).[29]
We then analyzed the enantiomeric excess of the resulting
pyridone 3a as a function of the Ag(S)-TRIP/[Rh] ratio (Table 1).
With a 0.5:1 ratio, a low e.r. of 55:45 was measured (Table 1,
entry 1). A significant increase of the e.r. to 77:23 was observed
with a 0.9:1 ratio (entry 2). An excess of silver compared to
rhodium further raised the e.r. (Table 1, entries 3–5). When
The free energy of the Rh/Ag transmetallation of one phos-
phane moiety was estimated by DFT computations using
Chem. Eur. J. 2016, 22, 8553 – 8558
www.chemeurj.org
8556
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Influence of Ag(S)-TRIP/[Rh] ratio on the e.r.
Entry
Conditions
x [mol%]
Ag(S)-TRIP/[Rh]
Yield [%]
e.r.[a]
1
2
3
4
5
6
7
8
A
A
A
A
A
A
B
2.5
4.5
6.5
7.5
10
15
0
2.5
0.5:1
0.9:1
1.3:1
1.5:1
2:1
3:1
1:1
1.5:1
87
91
91
93
93
76
79
70
55:45
77:23
84:16
84:16
85:15
83:17
75:25
88:12
C
[a] Derived from HPLC (see the Supporting Information, section 3.4).
(5 mol%) without or with addition of Ag(S)-TRIP (2.5 mol%),
which should generate the [I]+[(S)-TRIP]À/II mixture in solution.
Whereas, as noted in Scheme 2, the isolated complex [I]+[(S)-
TRIP]À alone led to a 75:25 e.r. (Table 1, entry 7), pyridone 3a
was obtained in an improved e.r. of 88:12 when additional
Ag(S)-TRIP was used (Table 1, entry 8). This finding supports
the beneficial role of species II in the enantioselectivity of the
cycloaddition. However, as [I]+[(S)-TRIP]À and II are in equilibri-
um, we cannot exclude that II is catalytically inactive and that I
is the only active species. The boost in enantioselectivity in the
presence of II may therefore result from the excess Ag(S)-TRIP
in the reaction mixture.
Scheme 4. Computed Rh/Ag transmetallation (kcalmolÀ1, distances in Š).
Conclusion
For the first time, some physical evidences of the counterionic
character of the TRIP anion in an organometallic species have
been established. By using NMR spectroscopy, mass spectrom-
etry, and DFT calculations, we studied the organometallic spe-
cies involved in phosphate-directed Rh-catalyzed [2+2+2] cy-
cloaddition of diynes to isocyanates. The use of Ag(S)-TRIP, in
slight excess compared to the rhodium complex, gives rise to
various species in equilibrium. The major one, [I]+[(S)-TRIP]À,
leads to the desired pyridone in 75:25 e.r., even in the absence
of excess of silver salt. It is a loose ion pair and the counterion-
ic character of the phosphate in this complex has been ascer-
tained by DOSY and other NMR experiments. With an excess of
silver salt, a second species appears in equilibrium with the
first. It is the bimetallic [Rh(cod)(dppb){(S)-TRIP}{Ag(S)-TRIP}]
species II, the presence of which boosts the reaction enantio-
selectivity. In this species, the 31P–103Rh 1D HMQC experiment
Figure 7. 1H NMR monitoring of the reaction [600 MHz, (CD2Cl)2, 343 K].
a larger excess of Ag(S)-TRIP was introduced (3:1 ratio), species
III accumulated but the e.r. remained similar (83:17; Table 1,
entry 6). These results suggest that species II has a positive
effect on the stereoselectivity. To probe this hypothesis, we
compared the use of the isolated complex [I]+[(S)-TRIP]À
Chem. Eur. J. 2016, 22, 8553 – 8558
www.chemeurj.org
8557
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
H. Amouri, C. Aubert, L. Fensterbank, V. Gandon, M. Malacria, J. Moussa,
C. Ollivier, Chem. Commun. 2013, 49, 7833; c) M. Barbazanges, M. AugØ,
J. Moussa, H. Amouri, C. Aubert, C. Desmarets, L. Fensterbank, V.
Gandon, M. Malacria, C. Ollivier, Chem. Eur. J. 2011, 17, 13789.
[11] D. Zuccaccia, L. Belpassi, F. Tarantelli, A. Macchioni, J. Am. Chem. Soc.
2009, 131, 3170.
showed that the phosphate is coordinated to the metal. Clear-
ly the question as to whether a phosphate is a ligand or
a counterion in the organometallic species involved in ACDC
cannot be answered simply. Our work suggests that the two
options may coexist.
[12] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, 3rd ed.,
Wiley-VCH, Weinheim, 2003.
[13] a) K. Aikawa, M. Kojima, K. Mikami, Angew. Chem. Int. Ed. 2009, 48, 6073;
Angew. Chem. 2009, 121, 6189; b) M. Raducan, M. Moreno, C. Bour,
A. M. Echavarren, Chem. Commun. 2012, 48, 52; c) G. Jiang, R. Halde r, Y.
Fang, B. List, Angew. Chem. Int. Ed. 2011, 50, 9752; Angew. Chem. 2011,
123, 9926.
[14] B. N. Nguyen, L. A. Adrio, E. M. Barreiro, J. B. Brazier, P. Haycock, K. K. Hii,
M. Nachtegaal, M. A. Newton, J. Szlachetko, Organometallics 2012, 31,
2395.
Acknowledgements
This work was supported by the CNRS, UPMC, IUF and Agence
Nationale de la Recherche (ANR-09-BLAN-108 SACCAOR). We
warmly thank A. Bernard, G. Gontard, and Dr. F. Ribot for NMR
studies, E. Kolodziej for HPLC analysis, Dr. H. Amouri, Dr. J.
Moussa, and Dr. M. MØnand for useful discussion. We used the
computing facility of the CRIHAN (project 2006-013).
[15] a) G. Jiang, B. List, Chem. Commun. 2011, 47, 10022; b) K. Ohmatsu, M.
Ito, T. Kunieda, T. Ooi, J. Am. Chem. Soc. 2013, 135, 590; c) K. Ohmatsu,
Y. Hara, T. Ooi, Chem. Sci. 2014, 5, 3645.
[16] G. Jindal, R. B. Sunoj, J. Org. Chem. 2014, 79, 7600.
Keywords: chirality · cycloaddition · homogeneous catalysis ·
ligand effects · rhodium
[17] For a related study, see: V. M. Williams, J. R. Kong, B. J. Ko, Y. Mantri, J. S.
Brodbelt, M.-H. Baik, M. J. Krische, J. Am. Chem. Soc. 2009, 131, 16054.
[18] NMR studies were performed in a sealed NMR tube, loaded in a glove
box. As long as no heating is required, the ion pair was preformed and
studied in CD2Cl2, at 300 K, at c=0.02m (see the Supporting Informa-
tion, section 1).
[19] 103Rh NMR were performed in CD2Cl2, 19 MHz, 300 K, X 3.186 MHz, see
L. Carlton in Annual Reports on NMR Spectroscopy, Vol. 63 (Ed.: G. A.
Webb), Elsevier Academic Press, Amsterdam, 2008, pp. 49–178.
[20] a) d=À8466 ppm relative to X 3.186 MHz (i.e., d=À168 ppm relative
to X 3.16 MHz); b) d=À7651 ppm relative to X 3.186 MHz (i.e., d=
654 ppm relative to X 3.16 MHz).
[1] B. Kasprzyk-Hordern, Chem. Soc. Rev. 2010, 39, 4466.
[2] For two recent reviews on enantioselective metal-catalyzed transforma-
tions, see: a) A. H. Cherney, N. T. Kadunce, S. E. Reisman, Chem. Rev.
2015, 115, 9587; b) P. Loxq, E. Manoury, R. Poli, E. Deydier, A. Labande,
Coord. Chem. Rev. 2016, 308, 131.
[3] For recent reviews, see: a) Y.-M. Wang, A. D. Lackner, F. D. Toste, Acc.
Chem. Res. 2014, 47, 889; b) B. M. Trost, M. Rao, Angew. Chem. Int. Ed.
2015, 54, 5026; Angew. Chem. 2015, 127, 5112. For recent examples,
see: c) S. Handa, L. M. Slaughter, Angew. Chem. Int. Ed. 2012, 51, 2912;
Angew. Chem. 2012, 124, 2966; d) M. Barbazanges, L. Fensterbank,
ChemCatChem 2012, 4, 1065; e) K. Ohmatsu, M. Ito, T. Kunieda, T. Ooi,
Nat. Chem. 2013, 5, 473.
[21] a) J. M. Ernsting, S. Gaemers, C. J. Elsevier, Magn. Reson. Chem. 2004, 42,
721; b) A. Fabrello, C. Dinoi, L. Perrin, P. Kalck, L. Maron, M. Urrutigoity,
O. Dechy-Cabaret, Magn. Reson. Chem. 2010, 48, 848.
[22] For selected reviews on DOSY, see: a) P. S. Pregosin, Pure Appl. Chem.
2009, 81, 615; b) A. Macchioni, G. Ciancaleoni, C. Zuccaccia, D. Zuccac-
cia, Chem. Soc. Rev. 2008, 37, 479; c) G. Bellachioma, G. Ciancaleoni, C.
Zuccaccia, D. Zuccaccia, A. Macchioni, Coord. Chem. Rev. 2008, 252,
2224; d) P. S. Pregosin, Prog. Nucl. Magn. Reson. Spectrosc. 2006, 49, 261;
e) A. Macchioni, Chem. Rev. 2005, 105, 2039; g) P. S. Pregosin, P. G. A.
Kumar, I. Fernandez, Chem. Rev. 2005, 105, 2977.
[23] J. A. Raskatov, A. L. Thompson, A. R. Cowley, T. D. W. Claridge, J. M.
Brown, Chem. Sci. 2013, 4, 3140.
[24] Dynamics Center software, (Bruker Biospin GmbH, Rheinstetten, 2015)
was used to calculate the D values.
[4] O. I. Kolodiazhnyi, Tetrahedron 2003, 59, 5953.
[5] For reviews, see: a) J. Lacour, D. Moraleda, Chem. Commun. 2009, 7073;
b) C. Zhong, X. Shi, Eur. J. Org. Chem. 2010, 2999; c) R. J. Phipps, G. L.
Hamilton, F. D. Toste, Nat. Chem. 2012, 4, 603; d) M. Mahlau, B. List, Isr.
J. Chem. 2012, 52, 630; e) M. Mahlau, B. List, Angew. Chem. Int. Ed. 2013,
52, 518; Angew. Chem. 2013, 125, 540; f) E. P. vila, G. W. Amarante,
ChemCatChem 2012, 4, 1713; g) A. Parra, S. Reboredo, A. M. Martín Cas-
tro, J. Alemµn, Org. Biomol. Chem. 2012, 10, 5001; h) K. Brak, E. N. Jacob-
sen, Angew. Chem. Int. Ed. 2013, 52, 534; Angew. Chem. 2013, 125, 558;
i) M. Raynal, P. Ballester, A. Vidal-Ferran, P. W. N. M. van Leeuwen, Chem.
Soc. Rev. 2014, 43, 1660; j) D. Parmar, E. Sugiono, S. Raja, M. Rueping,
Chem. Rev. 2014, 114, 9047; k) M. Jia, M. Bandini, ACS Catal. 2015, 5,
1638.
[6] a) D. B. Llewellyn, D. Adamson, B. A. Arndtsen, Org. Lett. 2000, 2, 4165;
b) D. B. Llewellyn, B. A. Arndtsen, Tetrahedron: Asymmetry 2005, 16,
1789.
[7] a) V. Komanduri, M. J. Krische, J. Am. Chem. Soc. 2006, 128, 16448;
b) J. R. Zbieg, E. Yamaguchi, E. L. McInturff, M. J. Krische, Science 2012,
336, 324.
[25] Hydrodynamic radius rH was obtained from the Stokes–Einstein relation
kT
D=6phr , with a viscosity h=0.410EÀ3 kgmÀ1 sÀ1 for CH2Cl2 at 300 K,
H
see: Chemical Properties Handbook; McGraw-Hill, New York, 1999.
[26] Ag(S)-TRIP DOSY led to the following diffusion coefficient: D[Ag(S)ÀTRIP]
=
7.1EÀ10 m2 sÀ1 (i.e., rH =7.5 Š; see the Supporting Information, sec-
tion 2.7).
[27] a) A. Cingolani, F. Marchetti, C. Pettinari, R. Pettinari, B. W. Skelton, A. H.
White, Inorg. Chem. 2002, 41, 1151; b) S. M. Socol, R. A. Verdake, Inorg.
Chem. 1984, 23, 3487; c) S. M. Socol, R. A. Jacobson, J. G. Verkade, Inorg.
Chem. 1984, 23, 88.
[8] a) S. Mukherjee, B. List, J. Am. Chem. Soc. 2007, 129, 11336; b) G. L. Ham-
ilton, E. J. Kang, M. Mba, F. D. Toste, Science 2007, 317, 496; c) R. L. La-
Londe, Z. J. Wang, M. Mba, A. D. Lackner, F. D. Toste, Angew. Chem. Int.
Ed. 2010, 49, 598; Angew. Chem. 2010, 122, 608.
[9] a) D. Uraguchi, M. Terada, J. Am. Chem. Soc. 2004, 126, 5356; b) T. Akiya-
ma, J. Itoh, K. Yokota, K. Fuchibe, Angew. Chem. Int. Ed. 2004, 43, 1566;
Angew. Chem. 2004, 116, 1592.
1
[28] Due to complex H NMR spectrum, DOSY of I/II/III mixture could not be
used (see the Supporting Information, section 3.6).
[29] EXSYCalc software package, Mestrelab Research, Santiago de Compos-
tela, 2007 was used to extract exchange constants from EXSY experi-
ments.
[10] For the contribution of our laboratory in the field of anion-directed
transition metal catalysis, see: a) M. AugØ, A. Feraldi-Xypolia, M. Barba-
zanges, C. Aubert, L. Fensterbank, V. Gandon, C. Ollivier, Org. Lett. 2015,
17, 3754; b) M. AugØ, M. Barbazanges, A. T. Tran, A. Simonneau, P. Elley,
Received: March 12, 2016
Published online on May 11, 2016
Chem. Eur. J. 2016, 22, 8553 – 8558
www.chemeurj.org
8558
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim