Stereochemical features making deoxycholic acid derived tropos biphenylphosphites efficient chiral ligands for rhodium: The asymmetric hydrogenation of dimethylitaconate as a case study

Article abstract of DOI:10.1021/jo701385y

(Chemical Equation Presented) Different deoxycholic acid derived biphenylphosphites, whose tropos nature was ascertained by NMR and CD measurements, were used in the rhodium-catalyzed asymmetric hydrogenation of dimethylitaconate achieving enantiomeric excesses up to 91%. The comparison of these results to those obtained using the corresponding atropoisomeric binaphthyl analogues, together with NMR and CD measurements on the rhodium complexes of some phosphites, allowed us to shed light on the nature of the active catalytic species and on the asymmetric induction process and hence to recognize the most appropriate stereochemical features to reach good levels of enantioselectivity.

Full text of DOI:10.1021/jo701385y

Stereochemical Features Making Deoxycholic Acid Derived tropos
Biphenylphosphites Efficient Chiral Ligands for Rhodium: The
Asymmetric Hydrogenation of Dimethylitaconate as a Case Study
Anna Iuliano,* Debora Losi, and Sarah Facchetti
Dipartimento di Chimica e Chimica Industriale, UniVersita` di Pisa, Via Risorgimento 35, 56126 Pisa, Italy
iuliano@dcci.unipi.it
ReceiVed July 4, 2007
Different deoxycholic acid derived biphenylphosphites, whose tropos nature was ascertained by NMR
and CD measurements, were used in the rhodium-catalyzed asymmetric hydrogenation of dimethylitaconate
achieving enantiomeric excesses up to 91%. The comparison of these results to those obtained using the
corresponding atropoisomeric binaphthyl analogues, together with NMR and CD measurements on the
rhodium complexes of some phosphites, allowed us to shed light on the nature of the active catalytic
species and on the asymmetric induction process and hence to recognize the most appropriate
stereochemical features to reach good levels of enantioselectivity.
Introduction
level of enantioselectivity represented a turning point in the
design of efficient chiral auxiliaries for this reaction. In fact,
monophosphorous ligands, such as phosphites, phosphoramid-
ites, and phosphinites having different structures7 show the
advantage of being easily accessible and quite inexpensive and,
in addition, of possessing comparable or even better enantiose-
lectivity than bidentate phosphorus ligands.8 Thus, the last 5
years have witnessed an explosive growth of newly designed
monophosphorous ligands to be used in the rhodium-catalyzed
asymmetric hydrogenation.9 Further progress in the design of
inexpensive and readily accessible ligands is represented by the
asymmetric activation of tropos species10 in the synthesis of
The rhodium-catalyzed asymmetric hydrogenation of func-
tionalized olefins represents one of the most extensively used
enantioselective reactions not only at the academic level but
also for the industrial preparation of enantiomerically enriched
pharmaceutical compounds.1 High levels of asymmetric induc-
tion were reached using chiral bidentate phosphine ligands2 and,
for about 30 years, the formation of a conformationally defined
metallocycle, guaranteed by the use of a C2 bisphosphorous
ligand,3 was considered mandatory for the stereochemical
outcome of the reaction. This prompted scientists to dedicate
several efforts to the preparation of new and more efficient chiral
ligands having these structural features.2 However, synthesis
of these kinds of ligands is often cumbersome, time-consuming,
and expensive; therefore, the breakthrough of Feringa,4 Pringle,5
and Reetz6 that monodentate phosphorus ligands can success-
fully promote the asymmetric catalytic hydrogenation with high
(4) (a) van den Berg, M.; Minnaard, A. J.; Schudde, E. P.; van Esch, J.;
de Vries, A. H. M.; de Vries, J. G.; Feringa, B. L. J. Am. Chem. Soc. 2000,
122, 11539-11540. (b) van den Berg, M.; Minnaard, A. J.; Haak, R. M.;
Leeman, M.; Shudde, E. P.; Meetsma, A.; Feringa, B. L.; de Vries, A. H.
M.; Maljaars, C. E. P.; Willans, C. E.; Hyett, D.; Boogers, J. A. F.;
Henderickx, H. J. W.; de Vries, J. G. AdV. Synth. Catal. 2003, 345, 308-
323.
(5) Claver, C.; Fernandez, E.; Gillon, A.; Heslop, K.; Hyett, D. J.;
Martorell, A.; Orpen, A. G.; Pringle, P. G. Chem. Commun. 2000, 961-
962.
(6) Reetz, M. T.; Mehler, G. Angew. Chem., Int. Ed. 2000, 39, 3889-
3890.
(7) For a review on monophosphorous ligands, see: Jerphagnon, T.;
Renaud, J.-L.; Bruneau, C. Tetrahedron: Asymmetry 2004, 15, 2101-2111.
(1) Brown, J. M. In ComprehensiVe Asymmetric Catalysis; Jacobsen, E.
N., Pfaltz, A., Yamamoto, H., Eds.; Springer: Berlin, 1999; Vol. 1, Chapter
5.1.
(2) Tang, W.; Zhang, X. Chem. ReV. 2003, 103, 3029-3069.
(3) Miyashita, A.; Yasuda, A.; Takaya, H.; Toriumi, K.; Ito, T.; Souchi,
T.; Noyori, R. J. Am. Chem. Soc. 1980, 102, 7932-7934.
10.1021/jo701385y CCC: $37.00 © 2007 American Chemical Society
Published on Web 10/03/2007
8472
J. Org. Chem. 2007, 72, 8472-8477

Asymmetric Hydrogenation of Dimethylitaconate
phosphites and phosphoramidites, a synthetic strategy which
avoids the sometimes difficult and time-consuming resolution
step. According to this strategy, a configurationally fluxional
(tropos) phosphorus ligand can be obtained starting from an
achiral biphenol and a configurationally stable amine or alcohol,
which induces a prevalent sense of twist on the biphenyl moiety
in the formation of the phosphite or phosphoramidite species
or when the phosphorus ligand coordinates to the metal center.
Although these kinds of ligands have been successfully used in
various asymmetric metal-catalyzed reactions,11 only a few
examples concerning their use in the rhodium-catalyzed asym-
metric hydrogenation have been reported, limited to the use of
diphosphites12 or combinatorial mixtures of phosphites and
phosphoramidites.13 To the best of our knowledge, only one
example using monophosphite tropos ligands is reported, which
afforded enantiomeric excess up to 75% in the hydrogenation
of dimethylitaconate.14 Our recent finding that the cholestanic
backbone is able to induce a prevalent sense of twist on the
biphenyl moiety of tropos biphenylphosphites linked at the 12-
position of the deoxycholic acid15 has opened the possibility to
obtain a new class of chiral ligands whose activity and
enantioselectivity can be checked in the asymmetric rhodium-
catalyzed hydrogenation. This kind of ligands has already proven
to be able to promote the copper-catalyzed conjugate addition
FIGURE 1. Structures of the phosphite ligands.
(8) Bernsmann, H.; van den Berg, M.; Hoen, R.; Minnaard, A. J.; Mehler,
G.; Reetz, M. T.; de Vries, J. G.; Feringa, B. L. J. Org. Chem. 2005, 70,
943-951 and references therein.
(9) (a) Hu, A.-G.; Fu, Y.; Xie, J.-H.; Zhou, H.; Wang, L.-X.; Zhou, Q.-
L. Angew. Chem., Int. Ed. 2002, 41, 2348-2350. (b) Hoen, R.; Boogers, J.
A. F.; Bernsmann, H.; Minnaard, A. J.; Meetsma, A.; Tiemersma-Wegman,
T. D.; de Vries, A. H. M.; de Vries, J. G.; Feringa, B. L. Angew. Chem.,
Int. Ed. 2005, 44, 4209-4212. (c) Liu, Y.; Ding, K. J. Am. Chem. Soc.
2005, 127, 10488-10489. (d) Hua, Z.; Vassar, V. C.; Ojima, I. Org. Lett.
2003, 5, 3831-3834. (e) Huang, H.; Liu, X.; Chen, S.; Chen, H.; Zheng,
Z. Tetrahedron: Asymmetry 2005, 16, 693-697. (f) Junge, K.; Hagemann,
B.; Enthaler, S.; Spannenberg, A.; Michalik, M.; Oehme, G.; Monsees, A.;
Riermeier, T.; Beller, M. Tetrahedron: Asymmetry 2004, 15, 2621-2631.
(g) Tang, W.-J.; Huang, Y.-Y.; He, Y.-M.; Fan, Q.-H. Tetrahedron:
Asymmetry 2006, 17, 536-543. (h) Hoen, R.; van den Berg, M.; Bernsmann,
H.; Minnaard, A. J.; de Vries, J. G.; Feringa, B. L. Org. Lett. 2004, 6,
1433-1436. (i) Lyubimov, S. E.; Davankov, V. A.; Loim, N. M.; Popova,
L. N.; Petrovskii, P. V.; Valetskii, P. M.; Gavrilov, K. N. J. Organomet.
Chem. 2006, 691, 5980-5983. (j) Huang, H.; Zheng, Z.; Luo, H.; Bai, C.;
Hu, X.; Chen, H. J. Org. Chem. 2004, 69, 2355-2361. (k) Reetz, M. T.;
Ma, J.-A.; Goddard, R. Angew. Chem., Int. Ed. 2005, 44, 412-415. (l)
Panella, L.; Feringa, B. L.; de Vries, J. G.; Minnaard, A. J. Org. Lett. 2005,
7, 4177-4180. (m) Zeng, Q.-H.; Hu, X.-P.; Duan, Z.-C.; Liang, X.-M.;
Zheng, Z. J. Org. Chem. 2006, 71, 393-396.
FIGURE 2. 31P NMR (121 MHz) spectra of phosphite 5 in (a) toluene-
d8 and (b) THF-d8 solutions and of (c) phosphite 6 in THF-d8 solution,
at the decoalescence temperature.
of diethylzinc to enones in enantioselective fashion.16 The results
reported herein concern the use of tropos deoxycholic acid
derived biphenylphosphites 1-6 (Figure 1) in the asymmetric
Rh(I)-catalyzed hydrogenation of the benchmark substrate
dimethylitaconate. The use of phosphites 7 will allow us to
compare the behavior of flexible and atropoisomeric analogues
in order to evaluate the effectiveness of the tropos motif in the
asymmetric induction process17 as far as this kind of ligand is
concerned. A correlation between asymmetric induction and
stereochemical features of the free ligands or their rhodium
complexes, aimed at gaining information about the factors
governing the enantioselectivity of this class of tropos ligands,
is also presented.
(10) For a review on tropos catalysts, see: Mikami, K.; Yamanaka, M.
Chem. ReV. 2003, 103, 3369-3400.
(11) (a) Alexakis, A.; Rosset, S.; Allamand, J.; March, S.; Guillen, F.;
Benhaim, C. Synlett 2001, 1375 -1378. (b) Alexakis, A.; Benhaim, C.;
Rosset, S.; Humam, J. Am. Chem. Soc. 2002, 124, 5262-5263. (c) Alexakis,
A.; Polet, D.; Benhaim, C.; Rosset, S. Tetrahedron: Asymmetry 2004, 15,
2199-2203. (d) Alexakis, A.; Polet, D.; Rosset, S.; March, S. J. Org. Chem.
2004, 69, 5660-5667. (e) Dieguez, M.; Ruiz, A.; Claver, C. Tetrahedron:
Asymmetry 2001, 12, 2895-2900. (f) Polet, D.; Alexakis, A. Tetrahedron
Lett. 2005, 46, 1529-1532. (g) d’Augustin, M.; Palais, L.; Alexakis, A.
Angew. Chem., Int. Ed. 2005, 44, 1376-1378. (h) Alexakis, A.; Croset, K.
Org. Lett. 2002, 4, 4147-4149. (i) Tissot-Croset, K.; Polet, D.; Alexakis,
A. Angew. Chem., Int. Ed. 2004, 43, 2426-2428. (j) Pamies, O.; Dieguez,
M.; Claver, C. J. Am. Chem. Soc. 2005, 127, 3646-3647. (k) Dieguez, M.;
Pamies, O.; Claver, C. J. Org. Chem. 2005, 70, 3363-3368. (l) Scafato,
P.; Cunsolo, G.; Labano, S.; Rosini, C. Tetrahedron 2004, 60, 8801-8806.
(12) Reetz, M. T.; Neugebauer, T. Angew. Chem., Int. Ed. 1999, 38,
179-181.
Results and Discussion
Phosphites 1-6 were prepared by reacting methyl-3-
acetyldeoxycholate with the suitable biphenylchlorophosphite,
(13) (a) Reetz, M. T.; Li, X. Angew. Chem., Int. Ed. 2005, 44, 2959-
2962. (b) Monti, C.; Gennari, C.; Piarulli, U.; de Vries, J. G.; de Vries, A.
H. M.; Lefort, L. Chem.sEur. J. 2005, 11, 6701-6717.
(14) Chen, W.; Xiao, J. Tetrahedron Lett. 2001, 42, 8737-8740.
(15) Iuliano, A.; Facchetti, S.; Uccello-Barretta, G. J. Org. Chem. 2006,
71, 4943-4950.
(16) Facchetti, S.; Losi, D.; Iuliano, A. Tetrahedron: Asymmetry 2006,
17, 2993-3003.
(17) Mikami, K.; Aikawa, K.; Yusa, Y.; Jodry, J. J.; Yamanaka, M.
Synlett 2002, 1561-1578.
J. Org. Chem, Vol. 72, No. 22, 2007 8473

Iuliano et al.
SCHEME 1. Synthesis of tropos Phosphites
which was in turn obtained from the corresponding biphenol
and PCl3 (Scheme 1).15
present, which broadens on lowering the temperature and splits
into two signals at 243 K (Figure 2), a decoalescence temper-
ature corresponding to an interconversion barrier of 11.8 kcal/
mol.19 The relative position of the signals corresponding to the
two diastereoisomeric species points out a P screw sense of the
biphenyl moiety of the prevailing diastereoisomer. The inte-
grated area of the two signals reveals a diastereoisomeric ratio
of 70/30, the highest observed prevalence of one sense of twist
for a substituted biphenylphosphite moiety linked at the 12-
position of the cholestanic backbone. The high prevalence of
the P sense of twist, the lack of dependence of the screw sense
on the solvent, and the higher interconversion barrier are likely
attributable to the presence of the two phenyl substituents at
the 3,3′-positions of the biphenyl moiety, which deeply changes
the stereochemical features of the phosphite.
The ligands 1-6, having different substituents on the biphenyl
moiety, but all having tropos nature and the atropoisomeric
phosphites 7a,b, were assayed in the rhodium-catalyzed enan-
tioselective hydrogenation of dimethylitaconate, not only to
check their effectiveness but also to gain information about the
structural features helping the asymmetric induction. In a typical
procedure, the catalytic system was generated “in situ” by
reacting a solution of the Rh(I) salt and the chiral ligand in 1:2
ratio under nitrogen atmosphere for 10 min; the hydrogenation
reaction was carried out at 7 bar hydrogen pressure for 20 h.
The effect of different parameters on the outcome of the reaction
was checked using 1 as chiral ligand, and the results are reported
in Table 1.
The reaction carried out at room temperature in dichlo-
romethane as solvent, using a 200:1 substrate/catalytic system
ratio, gives the hydrogenated product in good enantiomeric
excess but in low yield (entry 1). Increasing the amount of
catalytic system affords quantitative yield of the hydrogenated
product without affecting the stereochemical outcome of the
reaction (entry 2). Lowering the reaction temperature until -10
°C does not improve the enantiomeric excess (entries 3 and 4).
The use of the Rh(I) triflate affords worse results in terms of
enantioselectivity (entry 5). An improvement of asymmetric
induction is achieved using Rh(cod)2BF4 as catalytic precursor
in dichloroethane as solvent: under these reaction conditions,
a hydrogenation product in 90% ee is obtained (entry 6). The
use of Rh(cod)2SbF6 gives similar results (entry 7). Therefore,
because its higher solubility allows the use of different solvents,
the reaction was performed in THF, the solvent wherein the
sense of twist of the biphenyl moiety of 1 changes.15 It was
performed in THF in order to verify if the opposite enantiomer
The tropos nature of phosphites 1-4 as well as the prevalent
sense of twist of their biphenyl moiety has been already
determined by circular dichroism (CD) and variable-temperature
31P NMR measurements.15,16 A correlation between the relative
position of the 31P NMR signals of the two diastereoisomeric
species and the sense of twist of the biphenyl moiety has been
also found.16 On this basis it is possible to ascertain the prevalent
sense of twist of phosphites 5 and 6 only using variable-
temperature 31P NMR measurements. In fact, the CD spectra
of both these phosphites (Supporting Information) cannot be
compared to the CD spectrum of 1, for which a correlation
between the sign of the Cotton effects and the sense of twist of
the biphenyl moiety has been determined,15 due to the presence
of phenyl or benzoyl groups conjugated to the biphenyl moiety.
It significantly alters the nature of the electronic transitions of
the biphenyl chromophore.18 The sole information we were able
to extract from the CD measurements is the dependence of the
sense of twist of the biphenyl moiety on the solvent only in the
case of 5. In passing from acetonitrile solution to THF solution,
the CD signals of 5 vanish, whereas the CD spectrum of 6
remains unaltered, as far as number sign and intensity of the
Cotton effects are concerned. For this reason, we measured 31
P
NMR spectra both in THF and toluene solution for 5 and only
in THF solution for 6. The variable-temperature NMR profile
of 5 (Supporting Information) confirms the presence of two
diastereoisomeric species, which fast interconvert at room
temperature, and shows a decoalescence temperature of 193 K
both in THF and in toluene solutions (Figure 2), corresponding
to an interconversion barrier for the two diastereoisomeric
species of 9.7 kcal/mol.19
On the basis of the relative position of the 31P signals at the
decoalescence temperature, it is possible to attribute a M
prevalent sense of twist in toluene solution and a P sense of
twist in THF solution to the biphenyl moiety of 5. An 8%
prevalence of the M sense of twist in toluene and P screw sense
in THF is calculated on the basis of the integrated area of the
NMR signals. The variable-temperature 31P NMR measurements
of 6 (Supporting Information) show that a fast equilibrium at
room temperature between two diastereoisomeric species is also
present in this case. At room temperature, only one signal is
(18) Jaffe´, A.; Orchin, M. Theory and Application of UV Spectroscopy;
Wiley: New York, 1962.
(19) (a) Oki, M.; Yamamoto, G. Bull. Chem. Soc. Jpn. 1971, 44, 266-
270. (b) Oki, M. Top. Stereochem. 1983, 14, 1-81.
8474 J. Org. Chem., Vol. 72, No. 22, 2007

Asymmetric Hydrogenation of Dimethylitaconate
TABLE 1. Rhodium-Catalyzed Asymmetric Hydrogenationa of
Dimethylitaconate in the Presence of Ligands 1-7
entry L*
“Rh”
solvent
T (°C) yield (%) ee% (ac)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1
1
1
1
1
1
1
1
2
3
4
5
6
Rh(cod)2BF4
Rh(cod)2BF4
Rh(cod)2BF4
Rh(cod)2BF4
Rh(cod)2OTf CH2Cl2
Rh(cod)2BF4
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
25
25
0
30b
100
81 (S)
80 (S)
75 (S)
82 (S)
52 (S)
90 (S)
91 (S)
racemic
74 (S)
70 (S)
62 (S)
70 (S)
racemic
68 (R)
94 (S)
100
100
100
100
100
100
100
100
100
100
100
100
100
-10
25
25
25
25
25
25
25
25
25
25
25
(CH2)2Cl2
Rh(cod)2SbF6 (CH2)2Cl2
Rh(cod)2SbF6 THF
FIGURE 3. 31P NMR (121 MHz) spectra of 1:2 Rh(cod)2BF4/2 mixture
in CD2Cl2 solution at (a) 298 K, (b) 273 K, (c) 233 K, and (d) 213 K.
Rh(cod)2BF4
Rh(cod)2BF4
Rh(cod)2BF4
Rh(cod)2BF4
Rh(cod)2BF4
(CH2)2Cl2
(CH2)2Cl2
(CH2)2Cl2
(CH2)2Cl2
(CH2)2Cl2
(CH2)2Cl2
(CH2)2Cl2
of the biphenyl moiety of 1-5 in passing from the free ligand
to the coordinated species or to a different mechanism of
asymmetric induction in passing from tropos to atropoisomeric
ligands. In this last case, the matched pair of tropos phosphites
could be different from that of atropoisomeric analogues. The
asymmetric induction results obtained with the tropos ligands
are surprising if the extent of the prevalence of one sense of
twist for the biphenyl moiety of each phosphite is taken into
account. In fact, phosphite 6, showing the highest prevalence,
is that one affording the racemic product, whereas an 8% of
prevalence in the case of phosphite 5 corresponds to the
achievement of a product in 70% ee. These results clearly show
that the extent of the prevalence of one sense of twist of the
biphenyl moiety in the free ligand is not a stereochemical feature
directly correlated to the extent of asymmetric induction.
Therefore, to shed further light on the nature of the catalytic
species, we performed 31P NMR measurements on the catalytic
precursor obtained starting from phosphite 2, which shows a
low prevalence of one sense of twist of its biphenyl moiety16
but affords good asymmetric induction, and 6, which gives a
racemic product although its biphenyl moiety shows the highest
prevalence of one screw sense. The NMR measurements were
performed on a dichloromethane solution of 1:2 mixture of Rh-
(cod)2BF4/phosphite because hydrogenation reactions were
performed using this ratio and on the basis of the proved
hypothesis that the catalytically active species has two phosphite
ligands coordinated to the metal.21 The 31P NMR spectrum of
the Rh(cod)2BF4/2 mixture (Figure 3) recorded at room tem-
perature shows the presence of a broad doublet signal at 121
ppm with the typical coupling constant Rh-P (J ) 263 Hz),
which sharpens on lowering the temperature until 213 K
(decoalescence temperature for the signals of the free ligand),
but does not undergo any doubling.
7a Rh(cod)2BF4
7b Rh(cod)2BF4
a Reaction conditions: L* (0.01 mmol), [Rh(cod)2BF4] (0.005 mmol),
dimethylitaconate (0.5 mmol), solvent (2.5 mL), H2 (7 bar), rt, 20 h. Yields
and ee values were determined by GC equipped with a chiral capillary
column (Chiraldex G-TA). Absolute configuration of the prevailing
enantiomer was determined by comparison with literature data.20 b A 200:1
ratio substrate/“Rh” was used.
of the hydrogenation product can be obtained, as already
observed in the copper-catalyzed asymmetric conjugate addition
of diethylzinc to enones.16 Unfortunately, under these reaction
conditions, a racemic product is obtained (entry 8), as observed
with other kinds of phosphites.9d The reaction conditions
affording the best results in terms of yield and enantiomeric
excess of the product were used to check the activity and
enantioselectivity of the other phosphites in the asymmetric
hydrogenation of dimethylitaconate (Table 1). All the ligands
give a catalytic system showing comparable activity, with
quantitative yield of the hydrogenation product being obtained
in all cases. The enantioselectivity exhibited by phosphites
having substituents at the 5,5′-positions of the biphenyl moiety
is scarcely dependent on the nature of the substituent, with the
enantiomeric excess values ranging from 70 to 74% (entries 9,
10, and 12) except in the case of 4, which gives only 62% ee
of the hydrogenated product (entry 11). On the contrary,
substitution at the 3,3′-positions of the biphenyl moiety is
detrimental to the asymmetric induction: using phosphite 6, a
racemic product is obtained (entry 13). All phosphites give an
S configured prevailing enantiomer. Phosphites 7a and 7b
bearing an R and S atropoisomeric binaphthyl moiety, respec-
tively, afford enantiomeric hydrogenation products (entries 14
and 15), suggesting that, as observed with other types of
phosphites,7,9d the sense of asymmetric induction depends on
the absolute configuration of the biaryl moiety. In addition, the
higher value of asymmetric induction obtained using 7b suggests
that the (S)-binaphthylphosphite moiety linked at the 12-position
of deoxycholic acid constitutes the matched pair for this reaction.
Given that the M sense of twist corresponds to the R absolute
configuration, this result should indicate that the biphenyl
moieties of phosphites 1-5 are twisted in the “wrong” sense.
However, this does not explain why 6, whose biphenyl moiety
is twisted in a P sense, gives a racemic product. In addition,
the sense of asymmetric induction of 7a is different from that
of 1-5. This can be attributed to a change of the screw sense
Given that the interconversion barrier between the M and P
diastereoisomeric forms of tropos ligands becomes higher once
they are complexed to the metal center,22 the presence of the
doublet in the 31P NMR spectrum that does not undergo
doubling on lowering the temperature is likely due to the
formation of only one species, where the biphenyl moieties of
(20) Berens, U.; Burk, M. J.; Gerlach, A.; Hems, W. Angew. Chem.,
Int. Ed. 2000, 39, 1981-1984.
(21) Reetz, M. T.; Meiswinkel, A.; Mehler, G.; Angermund, K.; Graf,
M.; Thiel, W.; Mynott, R.; Blackmond, D. G. J. Am. Chem. Soc. 2005,
127, 10305-10313. The reaction performed using a 1:1 molar ratio ligand/
“Rh” did not afford hydrogenation product.
(22) Suarez, A.; Pizzano, A.; Fernandez, I.; Chiara, N. Tetrahedron:
Asymmetry 2001, 12, 633-642.
J. Org. Chem, Vol. 72, No. 22, 2007 8475

Iuliano et al.
FIGURE 5. 31P NMR (121 MHz) spectra of 1:2 Rh(cod)2BF4/6 mixture
in CD2Cl2 solution at (a) 308 K, (b) 298 K, and (c) 233 K.
ligands possess M twisted biphenyl moieties suggests that the
different sense of asymmetric induction shown by tropos
phosphites 1-5 with respect to atropoisomeric ligands 7a,b is
only due to a change in the asymmetric induction mechanism.
The 31P NMR spectrum of the Rh(cod)2BF4/6 mixture at room
temperature (Figure 5) shows the presence of two doublet signals
at 131 ppm (J ) 242 Hz) and at 134 ppm (J ) 258 Hz) and a
singlet at 152 ppm. On lowering the temperature, the singlet
broadens and, at 233 K, doubles in two signals at 152 and 155
ppm, whereas the two doublets remain unchanged.
FIGURE 4. CD spectra of phosphite 2 in acetonitrile solution and
1:2 Rh(cod)2BF4/2 mixture in CH2Cl2 solution.
both the coordinated ligands are twisted in the same highly
prevalent screw sense. The sense of twist of the biphenyl
moieties of the two ligands cannot be different because, in that
event, two different signals would be present in the spectrum,
each having a double doublet multiplicity, since the phosphorus
atoms of the two ligands would be heterotopic, giving rise to
different signals, with multiplicity due to the P-P and P-Rh
coupling. In addition, no signals are present in the spectral region
where the resonances of the free ligand are found, suggesting
that the ligand is totally complexed to the metal center. Given
that the NMR data show the presence in solution of only one
complexed species, it is possible to use CD spectroscopy in
order to determine the sense of twist of the biphenyl moiety of
the complexed species and hence to gain information about the
reason why the sense of asymmetric induction changes in
passing from atropoisomeric to tropos ligands. The CD spectrum
of the 1:2 Rh(cod)BF4/2 mixture, normalized to ligand concen-
tration, shown in Figure 4 in comparison to that of 2, is very
similar to the CD spectrum of the free ligand as far as number
and sign of the Cotton effects are concerned; the blue shift of
the biphenyl CD bands in the complex can be attributed both
to the complexation of phosphorus to the Rh center and to the
different solvents where the measurements were carried out. The
most evident difference between the two CD spectra is the
intensity of the Cotton effects that is remarkably higher in the
spectrum of the complex than in the spectrum of the free ligand.
The similarity of the two spectra suggests that the biphenyl
moiety has in the complexed species the same sense of twist as
in the free ligand, whereas the higher intensity of the Cotton
effects in the spectrum of the complex is attributable to the
formation of only one species, as indicated by the NMR
measurements, wherein the biphenyl moieties assume a totally
prevalent M screw sense. Two different interpretations10,17 for
the formation of only one complexed species are possible, both
related to the tropos nature of 2. According to the first one, the
M diastereoisomer is the species coordinating preferentially to
the metal center, and this causes shift of the M-P equilibrium
of the free ligand toward the M form, so that only the complex
containing the M diastereoisomeric ligand is formed. According
to the second one, both diastereoisomeric complexed species
in equilibrium between each other are formed, but the equilib-
rium is rapidly shifted toward the most stable form. In any case,
the formation of only one species wherein both coordinated
The signal at 152 ppm is attributable to the resonance of the
free ligand, whereas the two doublets are likely the signals of
two different rhodium complexes. On the basis of the integrated
areas of the signals, a 45:55 ratio between free ligand and
complexed species and a 70:30 ratio between the two rhodium
complexes were calculated. Therefore, an incomplete complex-
ation of the ligand to the rhodium center can be inferred,
probably due to the steric hindrance exerted by the two phenyl
ring at the 3,3′-positions of the biphenyl moiety on the
phosphorus atom, which prevents complexation to some extent.
The amount of complexed ligand originates two species where
the phosphorus atoms are homotopic, as the multiplicity of the
NMR signals suggests, and hence the two biphenyl units of each
complex have the same sense of twist. This means that the two
species derive from the complexation of the two M and P
diastereoisomers of phosphite 6; this hypothesis seems to be
confirmed by the NMR measurements at 308 K (Figure 5),
where the two doublets broaden and approach each other, a
typical behavior for species which are in a slow equilibrium
that become faster when the temperature is increased. The
formation of two diastereoisomeric species after complexation
of 6 is probably attributable to the higher interconversion barrier
between the M and P forms of free 6, which implies a higher
interconversion barrier also for the complexed ligands, prevent-
ing shift of the equilibrium between M and P diastereoisomers
toward the most stable species; this is possible in the case of 2
by virtue of its lower interconversion barrier. These NMR data
suggest a possible rationalization for the different enantiose-
lectivity shown by 2 and 6. In fact, the reaction between 2 and
Rh(cod)2BF4 gives rise to the quantitative formation of a
diastereomerically pure complex, whereas 6 is only partly bound
to the rhodium center generating two diastereoisomeric com-
plexes in a ratio similar to that observed for the two diastere-
oisomers of the free ligand. It is conceivable that in the case of
this kind of ligands, unlike as observed in other cases,13 none
of the possible complexes has higher activity and enantiose-
8476 J. Org. Chem., Vol. 72, No. 22, 2007

Asymmetric Hydrogenation of Dimethylitaconate
120.9, 129.6, 133.0, 133.4, 134.4, 136.4, 136.6, 139.6, 143.3, 164.2,
199.9; IR (KBr, cm-1) 3111, 1694, 1638, 1588, 1566, 1500,; 1383,
1283, 1144, 1111, 922, 833, 716, 616. Anal. Calcd for C26H18O4:
C, 79.17; H, 4.60; O, 16.23. Found: C, 79.22; H, 4.58.
lectivity, so that the formation of more complexed species in
lower amount results in worsening of the stereochemical
outcome of the reaction.
Preparation of the Phosphites: Representative Procedure.
A warm solution (60 °C) of biphenol (2.05 mmol) in dry toluene
or THF (25 mL) was slowly added to a solution of PCl3 (2.05 mmol)
and Et3N (4.1 mmol) in dry toluene or THF (5 mL). After 2 h of
stirring, the reaction mixture was filtered under argon atmosphere.
The solution was dropwise added to a solution of DMAP (2.23
mmol) and Et3N (7.3 mmol) in dry toluene or THF (25 mL) at
-60 °C over 2 h, then methyl-3-acetyldeoxycholate (2.1 mmol)
was added, and the mixture was allowed to warm to room
temperature and stirred over 20 h. The solids were removed by
filtration, and the filtrate was evaporated to dryness under reduced
pressure. The crude product was purified by column chromatog-
raphy (SiO2, CH2Cl2/acetone, 97:3), affording the pure phosphite.
Methyl 3R-Acetyloxy-12R-(5,5′-dibenzoylbiphenyl-2,2′-diyl)-
Conclusions
The use of tropos phosphites 1-6 in the rhodium-catalyzed
asymmetric hydrogenation of dimethylitaconate has proven that
the tropos motif is a stereochemical feature affording good
deoxycholic acid based phosphite ligands. In fact, the hydro-
genation product is obtained with enantiomeric excesses up to
91%, a comparable value to the asymmetric induction obtained
using one of the atropoisomeric binaphthylphosphite analogues.
The comparison between tropos and atropoisomeric deoxycholic
acid based phosphites has revealed that the sense of asymmetric
induction depends on the absolute configuration of the biaryl
moiety, and it is different for tropos and atropoisomeric ligands.
The stereochemical outcome of the hydrogenation reactions
demonstrates that the extent of the prevalence of one sense of
twist of the biphenyl moiety in free ligands is not correlated to
the extent of asymmetric induction. Variable-temperature 31P
NMR measurements on Rh complexes of some phosphites have
allowed us to conclude that the extent of asymmetric induction
depends on the formation of only one complexed species, which
is in turn guaranteed by a low interconversion M-P barrier for
the biphenyl moiety joined to a little steric hindrance around
the phosphorus atom. CD measurements on the Rh-phosphite
complex have revealed that the biphenyl moiety of the com-
plexed species has the same sense of twist as in the free ligand.
This result allows us to conclude that the change of the sense
of asymmetric induction from atropoisomeric to tropos ligands
is attributable to a change in the enantioselection mechanism,
and as a consequence, this suggests that different matched and
mismatched pairs are found for the two classes of phosphites.23
In other words, the prevalence of the M sense of twist of the
biphenyl moiety, guaranteed by a suitable substitution, is a
stereochemical feature making these tropos phosphites good
asymmetric rhodium ligands.
phosphite-5â-cholan-24-oate 5: 0.6 g (0.69 mmol, 39%); mp 97-
1
100 °C; [R]22 ) 25.3 (c ) 1.02, CH2Cl2); H NMR (200 MHz,
D
benzene-d6, δ) 0.49 (s, 3H), 0.91 (s, 3H), 1.08 (d, J ) 6.0 Hz, 3H),
1.20-2.00 (m, 25H), 1.70 (s, 3H), 2.22-2.31 (m, 1H), 3.36 (s,
3H), 4.26 (m, 1H), 4.63 (m, 1H), 6.91-7.14 (m, 10H), 7.29 (d, J
) 8.4 Hz, 1H), 7.39 (d, J ) 8.4 Hz, 1H), 7.58 (d, J ) 7.8 Hz,
2H),7.85 (m, 2H); 13C NMR (50 MHz, benzene-d6, δ) 12.2, 17.8,
20.8, 22.9, 23.5, 26.0, 26.8, 27.0, 27.6, 28.8, 30.9, 31.1, 32.4, 33.6,
2
34.2, 35.0, 35.8, 41.7, 46.6, 46.7, 47.8, 50.9, 73.7, 79.4 (d, J )
17.1 Hz), 122.5, 128.1, 129.9, 131.4, 131.9, 132.6, 135.2, 135.3,
137.8, 153.3, 169.5, 173.5, 193.9, 198.1; 31P NMR (121 MHz,
benzene-d6, δ) 161.6; IR (KBr, cm-1) 2940, 2861, 2353, 1734, 1655,
1598, 1484, 1445, 1362, 1318, 1252, 1108, 1028, 963, 884, 805,
696. Anal. Calcd for C51H55O9P: C, 72.67; H, 6.58; O, 17.08; P,
3.67. Found: C, 74.20; H, 6.49; P, 3.75.
Methyl 3R-Acetyloxy-12R-(3,3′-diphenylbiphenyl-2,2′-diyl)-
phosphite-5â-cholan-24-oate 6: 0.53 g (0.65 mmol, 32%); mp 69-
71 °C; [R]21D ) -24.20 (c ) 1.08, CH2Cl2); 1H NMR (300 MHz,
benzene-d6, δ) 0.28 (s, 3H), 0.58 (s, 3H), 0.60 (d, J ) 6.6 Hz, 3H),
1.00-1.72 (m, 25H), 1.78 (s, 3H), 1.90-2.10 (m, 1H), 3.35 (s,
3H), 4.65 (m, 1H), 4.85 (m, 1H), 7.30-7.58 (m, 16H); 13C NMR
(75 MHz, benzene-d6, δ) 12.1, 17.6, 17.7, 21.0, 23.0, 23.4, 25.9,
26.4, 27.2, 27.6, 30.9, 31.0, 32.6, 33.5, 34.0, 34.9, 35.1, 35.9, 42.0,
2
45.8, 46.1, 46.2, 47.1, 50.7, 74.2, 76.5 (d, J ) 17.5 Hz), 124.9,
125.0, 129.6, 129.8, 129.9, 130.5, 130.6, 132.8, 133.4, 135.0, 135.1,
Experimental Section
138.4, 138.5, 146.7 (d, 2J ) 5.5 Hz), 147.1 (d, 2J ) 6.0 Hz), 165.9,
173.8; 31P NMR (121 MHz, benzene-d6, δ) 150.3; IR (KBr, cm-1
)
General experimental details can be found in the Supporting
Information.
2933, 2855, 1727, 1450, 1416, 1377, 1361, 1244, 1183, 1077, 1022,
888, 855, 761, 700. Anal. Calcd for C51H59O7P: C, 75.16; H, 7.30;
O, 13.74; P, 3.80. Found: C, 75.20; H, 7.29; P, 3.82.
5,5′-Dibenzoyl-2,2′-dihydroxybiphenyl. Dry DMF (4 mL) was
dropwise added to AlCl3 (23 g, 0172 mol). The mixture was warmed
to 35 °C, then a solution of biphenol (2 g, 11 mmol) in benzoyl
chloride (3.5 mL, 30 mmol) was added. The reaction mixture was
warmed to 85 °C and stirred at that temperature for 5 h, then, after
cooling to room temperature, a 10% HCl solution was added and
the aqueous phase was extracted with diethyl ether (3 × 15 mL).
The collected organic phases were dried on anhydrous Na2SO4,
then concentrated under reduced pressure. The crude product was
purified by flash chromatography (SiO2, CH2Cl2/acetone 90:10)
affording 1.4 g of pure product (3.5 mmol, 33%) as a solid: mp
230 °C; 1H NMR (200 MHz, CDCl3, δ) 5.00 (s, 2H), 6.63 (d, J )
8.4 Hz, 2H), 6.99-7-36 (m, 14H); 13C NMR (50 MHz, CDCl3, δ)
General Procedure for the Rh-Catalyzed Asymmetric Hy-
drogenation. A solution of the Rh(I) salt (0.005 mmol) and the
phosphite (0.01 mmol) in dry solvent (2.5 mL) was introduced in
a stainless steel autoclave and stirred for 10 min. Dimethylitaconate
(0.5 mmol) was then added; the autoclave was closed and filled
with hydrogen (7 bar). After 20 h of stirring, the autoclave was
opened and the reaction mixture, after filtration over a pad of silica
gel, was analyzed by chiral GC for conversion and enantiomeric
excess determination.
Acknowledgment. The work was supported by the Univer-
sity of Pisa, MIUR (Project “High performance separation
systems based on chemo- and stereoselective molecular recogni-
tion” grant 2005037725).
(23) In the examples reported in the literature concerning the use of
binaphthyl-based phosphites and phosphoramidites (see ref 9) in the
hydrogenation of dimethylitaconate, the (R)-binaphthyl moiety induces the
formation of an S product; by contrast, the use of R atropos biphenyl-
based phosphites (ref 9d) gives an R product. These results suggested a
different induction mechanism for the two kinds of ligands. Our results,
obtained using tropos and atropos ligands having the same configurationally
stable structural motif (cholestanic backbone), strongly point out the different
asymmetric induction mechanism for tropos biphenyl-based phosphites and
atropos binaphthyl-based phosphites.
Supporting Information Available: General experimental
1
details, copy of H NMR spectra of 5, 6, and 5,5′-dibenzoyl-2,2′-
dihydroxybiphenyl, CD spectra of 5 and 6, and variable-temperature
31P NMR spectra. This material is available free of charge via the
Internet at http://pubs.acs.org
JO701385Y
J. Org. Chem, Vol. 72, No. 22, 2007 8477