Synthesis, characterization, and applicability of neutral polyhydroxy phospholane derivatives and their rhodium(I) complexes for reactions in organic and aqueous media

Article abstract of DOI:10.1021/ja011500a

Two different protocols for the preparation of water-soluble, enantiomerically pure polyhydroxybisphospholanes from acid-labile acetal and tert-butyldimethylsilyl-protected derivatives are reported. These procedures circumvent two of the commonly encounte

Full text of DOI:10.1021/ja011500a

J. Am. Chem. Soc. 2001, 123, 10207-10213
10207
Synthesis, Characterization, and Applicability of Neutral Polyhydroxy
Phospholane Derivatives and Their Rhodium(I) Complexes for
Reactions in Organic and Aqueous Media
T. V. RajanBabu,* Yuan-Yong Yan,* and Seunghoon Shin
Contribution from the Department of Chemistry, The Ohio State UniVersity,
100 W. 18th AVenue, Columbus, Ohio 43210
ReceiVed June 19, 2001
Abstract: Two different protocols for the preparation of water-soluble, enantiomerically pure polyhydroxy-
bisphospholanes from acid-labile acetal and tert-butyldimethylsilyl-protected derivatives are reported. These
procedures circumvent two of the commonly encountered limitations in the synthesis of these potentially
important ligands: (a) formation of phosphonium salts from the highly basic phosphine under acidic conditions,
and (b) the need to start with preformed, fully protected cationic metal complex. Thus, cationic Rh complexes
of these ligands have been prepared in a separate step, and they have been found to be excellent catalysts for
organic and aqueous phase hydrogenation of dehydroamino acids. The viability of catalyst recovery has been
demonstrated in three different systems, including two cases where > 99% ee can be achieved under recycling
conditions.
Introduction
ligands may find applications for metal-catalyzed reduction of
membrane lipids where the catalyst transport into the bilayer
may be precluded by highly charged groups which lead to
accumulation of the catalysts at the water-organic interface.4
In their original work, Selke and co-workers used vicinal
diarylphosphinite-Rh complexes derived from phenyl â-D-
glucopyranoside (1) for asymmetric hydrogenation of dehydro-
amino acids in water.5 They also showed that enhanced
selectivity in the asymmetric hydrogenation of sparingly soluble
substrates could be achieved by the use of surfactants. In an
attempt to circumvent the limited solubility of monosaccharide
derived ligands, we7and others6 have resorted to disaccharides
such as trehalose (2)7a (or ligands with quaternary ammonium
groups derived from D-salicin such as 37b) as ligand precursors
with the expectation that an increased number of hydroxyl
groups would improve solubility in water.8 Indeed, very high
enantioselectivities can be achieved with Rh complexes of these
ligands, especially in the presence of surfactants.6b Besides the
need for surfactants, which have to be removed at the end of
the reaction, two other problems are inherent with the diaryl-
phosphinite system. Our investigations of the partition coef-
ficient for the Rh complexes between organic and water phases
A number of different strategies have been employed to
increase the solubility of phosphine ligands and their complexes
in aqueous media.1 The majority of successful applications of
these reagents in aqueous homogeneous catalysis involve
compounds that carry sulfonate (anionic) or quaternary am-
monium (cationic) groups. Nonionic, chiral ligands with hydro-
philic groups designed to improVe aqueous solubility are
beginning to receive increasing attention,2-8 principally because
biphasic reactions of poorly soluble organic substrates can be
expected to benefit from these ligands. In addition, such neutral
(1) For general reviews on the subject of organic chemistry in aqueous
media, see: (a) Sinou, D. Bull. Soc. Chim. Fr. 1986, 480. (b) Kunz, E. G.
Chemtech 1987, 17, 570. (c) Kalck, P.; Monteil, F. AdV. Organomet. Chem.
1992, 34, 219. (d) Herrmann, W. A.; Kohlpaintner, W. Angew. Chem., Int.
Ed. Engl. 1993, 32, 1524. (e) Aqueous Organometallic Chemistry and
Catalysis; Horva´th, I. T., Joo´, F., Eds.; Kluwer: Dodrecht, 1995. (f) For
thematic issues dedicated to applications of water-soluble organometallic
compounds see: J. Mol. Catal. A: Chem. 1997, 116 (1, 2), 1-309. Catal.
Today, 1998, 42 (4), 371-478. (g) Nagel, U.; Kinzel, E. Chem. Ber. 1986,
119, 1731.
(2) Bo¨rner and co-workers have published extensively on the use of
hydroxyphosphines, most of them for applications in organic solvents. While
our studies were in progress, the Bo¨rner group published the first report
dealing with a water-soluble tetrahydroxyphospholane (see ref 2c). (a) Holz,
J.; Quirmbach, M.; Bo¨rner, A. Synthesis 1997, 983. (b) Selke, R.; Holz, J.;
Riepe, A.; Bo¨rner, A. Chem.sEur. J. 1998, 4, 769 and references therein.
(c) Holz, J.; Quirmbach, M.; Schmidt, U.; Heller, D.; Stu¨rmer, R.; Bo¨rner,
A. J. Org. Chem. 1998, 63, 8031. (d) Holz, J.; Heller, D.; Stu¨rmer, R.;
Bo¨rner, A. Tetrahedron Lett. 1999, 40, 7059. (e) Bo¨rner, A. Eur. J. Inorg.
Chem. 2001, 327.
(3) For other reports of the synthesis of polyhydroxyphospholanes see:
(a) Li, W.; Zhang, Z.; Xiao, D.; Zhang, X. Tetrahedron Lett. 1999, 40,
6701. (b) Li, W.; Zhang, Z.; Xiao, D.; Zhang, X. J. Org. Chem. 2000, 65,
3489. (c) Our investigations reveal that the hydroxyphospholane ligands
described in these publications are most likely phosphonium salts (vide
infra).
(4) (a) Quinn, P. J.; Taylor, C. E. J. Mol. Catal. 1981, 13, 389. (b)
Madden, T. D.; Peel, W. E.; Quinn, P. J.; Chapman, D. J. Biochem. Biophys.
Methods 1980, 2, 19. (c) For biomedical applications of hydroxyphosphines,
see: Katti, K. V.; Gali, H.; Smith, C. J.; Berning, D. E. Acc. Chem. Res.
1999, 32, 9.
(5) (a) Kumar, A.; Oehme, G.; Roque, J. P.; Schwarze, M.; Selke, R.
Angew. Chem., Int. Ed. Engl. 1994, 33, 2197 and references therein. (b)
Selke, R. J. Organomet. Chem. 1989, 370, 241 and 249. (c) Graseert, I.;
Paetzold, E.; Oehme, G. Tetrahedron 1993, 49, 6605. (d) Selke, R.; Ohff,
M.; Riepe, A. Tetrahedron 1996, 52, 15079.
(6) (a) Yonehara, K.; Hashizume, T.; Mori, K.; Ohe, K.; Uemura, S. J.
Org. Chem. 1999, 64, 5593. (b) Yonehara, K.; Hashizume, T.; Mori, K.;
Ohe, K.; Uemura, S. J. Org. Chem. 1999, 64, 9381.
(7) (a) Shin, S.; RajanBabu, T. V. Org. Lett. 1999, 1, 1229. (b) Yan, Y.;
RajanBabu, T. V. J. Org. Chem. 2001, 66, 3277.
(8) For other examples of the use of carbohydrates for the preparation
of water-soluble phosphines see: (a) Mitchell, T. N.; Heesche-Wagner, K.
J. Organomet. Chem. 1992, 436, 43. (b) Sawamura, M.; Kitayama, K.; Ito,
Y. Tetrahedron: Asymmetry 1993, 4, 1829. (c) Beller, M.; Krauter, J. G.
E.; Zapf, A. Angew. Chem., Int. Ed. Engl. 1997, 36, 772. See also: (d)
Ferrara, M. L.; Orabona, I.; Ruffo, F.; Funicello, M.; Panunzi, A.
Organometallics 1998, 17, 3832.
10.1021/ja011500a CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/26/2001

10208 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
RajanBabu et al.
Scheme 1. Synthesis of Mesylate Precursor for 7
Scheme 2. Synthesis of Bishydroxymethylphospholane 8
prototypical Pd-catalyzed allylation11b and allylic amination11c
reactions in organic media. In this paper, we report new
protocols for the direct conversion of these ketal derivatives
and of other silyl-protected compounds (for example, 8) to the
corresponding fully deprotected polyhydroxyphosphines. In the
past, direct acid hydrolysis has been found to be problematic,3a-c,12
and successful syntheses of such ligands bearing electron-rich
phosphorus atoms have relied on either the formation of
the corresponding BH3 adducts9c or precomplexation with
Rh(I).2d,5-7,11b We also record the complete characterization (C,
H analysis and H, 13C, and 13P NMR) of these potentially
1
(by the inductively coupled plasma spectroscopy method)
revealed that, despite the increased number of hydroxyl groups,
the Rh(I)-complexes still possessed considerable solubility in
the organic medium. Ohe, Uemura, and co-workers have noticed
significant deterioration of catalytic activity upon reuse of the
aqueous phase for subsequent reactions.6a One possible explana-
tion of such decrease in activity could be the hydrolytic
degradation of the P-O bond, as we have seen in the case of
related salicin-derived diarylphosphinite complexes (3).7b We
believe that an attractive solution to the dual problems of
hydrolytic instability and hydrophobicity can be addressed
through the use of polyhydroxyphosphines bearing robust P-C
bonds (vis-a´-vis P-O bonds) and a reduced number of aryl
substituents on the phosphorus atom. The latter should help to
decrease the hydrophobicity of the catalyst. With these goals
in mind,9 we chose the C2-symmetric phospholane system,
which forms the backbone of the enormously successful DuPhos
ligands (e.g., 4), discovered by Burk10 and co-workers. Last year
we reported a general route for synthesis of several highly
functionalized phospholane derivatives 5-7 from readily avail-
able D-mannitol.11a Some of these are superb catalysts for
important ligands and their applications in asymmetric hydro-
genation in aqueous and organic media. The reuse of an aqueous
solution of the catalyst (up to four cycles) with no deleterious
effects on the enantioselectivity will also be demonstrated for
the first time.
Results and Discussion
(a) Synthesis of the Ligands. The syntheses of protected
derivatives 53a,11a and 711a from D-mannitol have been reported.13
Minor modifications of the procedures used for the synthesis
of these compounds enabled the preparation of 6 and 814
from the corresponding diols 9a and 10a. Diol 9a was
synthesized from 1:2;5:6-dianhydro-3,4-O-isopropylidene-D-
mannitol15 (Scheme 1) and bis(ButMe2Si)ether, 10b, was
prepared from the tetrahydroxy compound 10a. Phospholane 8
was synthesized from 10b in very good yield as shown in
Scheme 2.
Because base-sensitive OH-protecting groups such as esters8c,d
are incompatible with the highly nucleophilic phosphide anion
used in the synthesis of the phospholanes, synthesis of the
polyhydroxyphospholanes is a nontrivial problem, especially
when the phosphine is electron-rich. Use of acid-sensitive
(9) For other reports of the synthesis and utility of functionalized
phospholanes, see: (a) Hitchcock, P. B.; Lappert, M. F.; Yin, P. J. Chem.
Soc., Chem. Commun. 1992, 1598. (b) See ref 2c. (c) Carmichael, D.;
Doucet, H.; Brown, J. M. J. Chem. Soc., Chem. Commun. 1999, 261. (d)
For the initial report from the Zhang group, see ref 3a.
(10) (a) Burk, M. J.; Feaster, J. E.; Harlow, R. L. Tetrahedron:
Asymmetry 1991, 2, 569. (b) Burk, M. J.; Feaster, J. E.; Nugent, W. A.;
Harlow, R. L. J. Am. Chem. Soc. 1993, 115, 10125. (c) Burk, M. J.; Harper,
T. G. P.; Kalberg, C. S. J. Am. Chem. Soc. 1995, 117, 4423. (d) Burk, M.
J.; Kalberg, C. S.; Pizzano, A. J. Am. Chem. Soc. 1998, 120, 4345. (e)
Burk, M.; Gross, M. F.; Harper, T. G. P.; Kalberg, C. S.; Lee, J. R.;
Martinez, J. P. Pure Appl. Chem. 1996, 68, 37 and references therein.
(11) (a) Yan, Y.; RajanBabu, T. V. J. Org. Chem. 2000, 65, 900. (b)
Yan, Y.; RajanBabu, T. V. Org. Lett. 2000, 2, 199. (c) Yan, Y.; RajanBabu,
T. V. Unpublished results.
(12) Direct hydrolysis of acetals with a less basic diphenylphosphino
group is known: (a) Bo¨rner, A.; Holz, J.; Ward, J.; Kagan, H. B. J. Org.
Chem. 1993, 58, 6814. (b) Bo¨rner, A.; Ward, J.; Kortus, K.; Kagan, H. B.
Tetrahedron: Asymmetry 1993, 4, 2219. (c) Tamao, K.; Nakamura, K.;
Shigehiro, Y.; Shiro, M.; Saito, S. Chem. Lett. 1996, 1007. (d) See also ref
8a.
(13) For synthesis of other derivatives see refs 2c,d and 9c.
(14) Originally synthesized in 1999 by S. Shin of our group from 10a.
For 10a, see: Corey, E. J.; Hopkins, P. B. Tetrahedron Lett. 1982, 23,
1979. (b) Marzi, M.; Misiti, D. Tetrahedron Lett. 1989, 30, 6075. (c)
Machinaga, N.; Kibayashi, C. J. Org. Chem. 1992, 57, 5178.
(15) Le Merrer, Y.; Dure´ault, A.; Greck, C.; Micas-Languin, D.; Gravier,
C.; Depezay, J.-C. Heterocycles 1987, 25, 541.

Neutral Polyhydroxy Phospholane DeriVatiVes
J. Am. Chem. Soc., Vol. 123, No. 42, 2001 10209
Table 1. Hydrogenation of Dehydroamino acids Using
protecting groups is often the only alternative, and this procedure
is fraught with problems.16 For example, acid hydrolysis of
DIOP (trans-4,5-bis[(diphenylphosphino)methyl]-2,2-dimethyl-
1,3-dioxolane) derivatives with aqueous mineral acids gave
considerable amounts of quaternary phosphonium salts.17 Bo¨rner
has reported that methanesulfonic acid in aqueous methanol can
be used to liberate hydroxyphosphines from the corresponding
tetrahydropyranyl (THP) derivatives, as long as the phosphine
carried aromatic substituents (in the case cited, diphenylphos-
phino groups).12 In most cases reported to date, including one
example of a hydroxymethyl bisphospholane system,2d the
hydroxyl protecting groups are removed after the formation of
the cationic Rh chelate.5-7 This protocol, while perfectly
acceptable for Rh-catalyzed reactions, imposes serious limita-
tions if other metals are to be considered for catalysis in aqueous
media. Therefore, we thought that it would be highly desirable
to have a direct route to the metal-free polyhydroxyphosphine
ligands, and our initial efforts were directed toward this goal.
While our early attempts to use strong acids in aqueous and
alcoholic media to liberate the hydroxyl groups from 5, 7, and
8 were facing considerable difficulties, Zhang et al. reported3a,b
what appeared to be a simple solution to the problem. These
workers reported that ligand 5, upon treatment with excess
methanesulfonic acid in refluxing aqueous methanol (10 h), gave
the deprotected derivative 12 in 90% yield as an oil! However,
the proton, and especially the phosphorus-31 NMR data (broad
peak at δ +11.9), suggested that the product of this reaction
was most likely a phosphonium salt (15), and not the free
hydroxyphospholane as claimed (vide infra).
[L]Rh+[Cod] BF4
-
substrate
(R in eq 6)
entry
L
solvent
MeOH
MeOH
MeOH
MeOH
1:1 MeOH/H2O
H2O
MeOH
% ee (GC)
-
1
2
3
4
5
6
7
8
3-F-C6H4
12
12
13
13
13
13
14
14
91
>99
97
>99
>99
>99
95
H
-
3-F-C6H4
H
H
H
-
3-F-C6H4
H
MeOH
97
Zhang.3 We also noticed significant differences in the 13C and
1H NMR spectra of the two compounds. For example, the
chemical shifts of the protons in the phosphonium salt, as
expected, are in general at lower fields compared to those of
the neutral ligand (see Experimental Section for details). Neutral
ligands 13 and 14, prepared by a similar route, also show
characteristic 31P NMR chemical shifts (CD3OD) at δ -10.36
and -14.24. The cationic Rh complexes (Rh+[L][COD] BF4
]
-
prepared from these ligands show clean doublets for the JRh-P
(L ) 12, 151.6 Hz; L ) 13, 151.2 Hz; L ) 14, 152.7 Hz),
confirming the C2-symmetric nature of the ligands.
We have since discovered two separate procedures for the
synthesis of the polyhydroxyphospholanes, depending on the
protecting groups on the alcohol. The ketal protecting groups
from 5-7 are removed by treatment with acidic resin (AG 50
WX-8) in anhydrous methanol (eqs 1-3). Best yields are
obtained when the resin is carefully washed with and then
swollen in methanol overnight at room temperature. It is
subsequently washed with methanol three times and dried before
use for hydrolysis. At the end of the reaction, the insoluble
residue is filtered with the aid of Celite. The solvent is removed,
the residue is washed with hexane, and the product is dried. In
each case, the hydroxyphospholanes 12-14 are obtained as a
white solid in 75-80% isolated yield (eqs 1-3). These
potentially important ligands were fully characterized by
elemental analysis (C, H) and 1H, 13C, and 31P NMR. For
example, ligand 12, synthesized according to our protocol, is a
white solid whose 31P NMR spectrum in CD3OD is characterized
by a sharp singlet at δ -3.77. The corresponding phosphonium
salt (15) prepared by treatment of 12 with 3 equiv of methane-
sulfonic acid in CD3OD has a characteristic broad absorption
at δ +15 (eq 4), suspiciously similar to that of the product
obtained under aqueous hydrolysis conditions reported by
An alternate procedure was used for the synthesis of 16. The
use of AG 50 WX-8 for the deprotection of 8 gave poor yields
of the expected product, primarily because of low recovery of
the highly polar tetrahydroxyphosphine from the resin. However,
the desilylation can be accomplished by reaction of 8 with CF3-
CO2H in MeOH (eq 5). After evaporation of the solvent, the
residue (a solid which, by NMR, appears to be a mixture of
phosphonium salts) is dissolved in methanol, and the methanol
solution is passed through a short bed of diethylaminomethyl
polystyrene. Evaporation of the filtrate gives the expected
product in 84% yield. A very sharp peak at δ -12.4 (CD3OD)
is characteristic of the free phosphine. Further confirmation of
1
the structure comes from detailed analysis of the H, 13C, and
HETCOR spectra (see Experimental Section and Supporting
Information for details and spectra).
(b) Hydrogenation Studies: Organic and Aqueous Media.
Contrary to an earlier report,3a isopropylidene derivative 5 was
found to be an excellent ligand for Rh-catalyzed asymmetric
hydrogenation, delivering high ee’s for prototypical substrates
shown in eq 6. The diastereomeric phospholane, 7, as expected,
gave opposite selectivity in the hydrogenation of methyl
acetamidoacrylate.
The cationic Rh complexes of the fully deprotected hydroxy-
phospholanes are excellent catalysts for hydrogenation of
(16) See also footnote 10 in ref 3a.
(17) Deschenaux, R.; Stille, J. K. J. Org. Chem. 1985, 50, 2299.

10210 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
RajanBabu et al.
Table 2. Hydrogenation of Dehydroamino Acids Using
[12]Rh+[Cod] BF4 in MeOH: Effect of Water on the
-
Hydrogenation of Methyl Acetamidoacrylate
entry
MeOH:water
1:0
1:1
conversion
% ee
1
2
3
4
100
100
38
>99
69
21
1:2
1:3 (+ 1 equiv L)
100
67
Table 3. Catalyst Recovery and Reuse in Hydrogenation of
Methyl Acetamidoacrylate Using 1 Mol % of [14]Rh+[Cod] BF4
(1:1 MeOH/water with 1 Mol % of Added Ligand)a
-
entry
run number
conversion
% ee
1
2
3
4
1
2
3
4
100
100
100
100
83
86
87
90
a
See Figure 1 for chromatograms.
be recycled with no loss of actiVity or selectiVity. While a
number of examples of such recovery of catalysts containing
ionic ligands are known, most notably phosphine ligands
carrying ionic (-SO3- or Me3N+) side chains,1,20 ligands 12-
14 are among the first nonionic ligands where this has been
possible with no apparent loss of selectiVity. The results for
methyl acetamidoacrylate reductions are shown in Tables 3-5.
The reaction is typically carried out in 1:1 methanol/water with
1 mol % of the isolated cationic Rh complex as the precatalyst
with one mol % of extra ligand added to suppress precipitation
of the metal. The product is separated at the end after each run
by extraction into ether. The aqueous layer containing the
catalyst is reused in subsequent hydrogenations. The results of
dehydroamino acids (Table 1) and simple enamides (eq 7a,b)
in methanol.18 For these studies and the subsequent hydrogena-
tions in aqueous media, we chose methyl acetamidoacrylate as
the substrate, because this substrate is freely soluble in water,
and its recovery by extractive workup would be a good
indication of how effectively the aqueous phase containing the
catalyst can be recycled in repeated operations (vide infra). Of
special note in Table 1 is the complex prepared from 1,5-
diethylphospholane 13 which gave >99% ee in neat water (entry
6). Such high selectiVity in the absence of added surfactants
has seldom been achieVed in purely aqueous medium, especially
using neutral ligands.19 Phosphine ligands with quaternary
ammonium groups structurally related to chiraphos (17) have
been reported to give up to 94% ee in neat water when the
reaction is done in a slurry.20
The enantioselectivity of the reaction is highly dependent on
the structure of the catalyst and the solvent. For example, Table
2 shows the effect of water on the enantioselectivity of
hydrogenation of acetamidoacrylate using ligand 12. While near
100% ee is observed reproducibly in neat methanol, often a
capricious reaction is observed in MeOH/H2O mixtures, with
enantioselectivity dropping off as the proportion of water
increases (99% ee in neat methanol to 21% in 1:3 methanol/
water). At the end of the reaction in several runs, precipitation
of metallic Rh was noticed, which suggested that nonselective
Rh-mediated processes might be responsible for such behavior.
The problems of reproducibility and precipitation of rhodium
can be alleviated by using an extra equivalent of the ligand (entry
4, Table 2). Using this protocol, it is now possible to recycle
the aqueous solution of the catalyst without loss of selectivity.
(c) Catalyst Recovery and Reuse. An often stated, yet rarely
achieved, goal of developing water-soluble ligands is to find
conditions where the aqueous layer containing the catalyst can
hydrogenation using (COD)Rh+[14] BF4 are shown in Table
-
3. Figure 1 shows the corresponding chromatograms. We have
noticed a small, but discernible, increase in selectivity with each
-
run. The diastereomeric complex (COD)Rh+[12] BF4 also
behaves in a similar fashion, even though the selectivity is lower
(Table 4). Finally, Table 5 shows results with diethylphos-
pholane 13, which is one of only two ligands (the other one
being 16, Vide infra), which gaVe consistently high selectiVities
(>99% ee) in both neat methanol as well as water. This catalyst
can be recycled three times in 1:1 MeOH/H2O, albeit with
significant loss of reactivity during the last run (only 35%
conversion under identical conditions). Most gratifyingly, the
enantioselectivity remains unaltered. It is tempting to speculate
that the loss of reactivity is related to the increased hydropho-
bicity of diethyl derivative 13 (vis-a´-vis dimethyl derivative 12)
and the attendant loss of the catalyst in ether which is used to
extract the product after the second run.
Among the ligands we prepared, 16 gave the best results in
hydrogenation in 100% water. The enantioselectivity and
recyclability is outstanding with the catalyst derived from this
ligand as shown in Table 6 and Figure 1. In four sequential
runs, ∼99% ee and >90% isolated yield (100% conversion)
were obtained for the hydrogenation of methyl acetamidoacrylate
in neat water. The recycling was repeated for up to seven runs
with a small loss of selectivity. The catalytic efficiency suffers
in runs beyond the fourth cycle. For example, in the seventh
cycle, it is approximately 50%, as judged by the 12 h needed
to complete the reaction instead of the usual 6 h. A surprisingly
high enantioselectivity of 95% was observed even for this run
(Figure 1).
(18) Zhang has also reported similar results in methanol. See ref 3b.
(19) For a notable exception see Bo¨rner et al., ref 2d. Excellent ee’s
using a surfactant (sodium dodecyl sulfate) in an aqueous medium have
been recorded; see for example refs 5a and 6b.
(20) Toth, I.; Hanson, B. E.; Davis, M. E. Tetrahedron: Asymmetry 1990,
1, 913 and references therein. The recovery and reuse of the catalyst has
also been demonstrated with this catalyst.
Summary and Conclusions
We record two different protocols for the preparation of
water-soluble, enantiomerically pure polyhydroxybisphos-

Neutral Polyhydroxy Phospholane DeriVatiVes
J. Am. Chem. Soc., Vol. 123, No. 42, 2001 10211
Figure 1. GCs of hydrogenation products (ee’s in brackets) from successive runs using recycled aqueous solutions of (a) [14]Rh+[COD] BF4
-
(Table 3) and (b) [16]Rh+[NBD] SbF6 (Table 6).
-
Table 4. Catalyst Recovery and Reuse in Hydrogenation of
Methyl Acetamidoacrylate Using 1 Mol % of [12]Rh+[Cod] BF4
(1:1 MeOH/water with 1 Mol % of Added Ligand)
amino acids. The viability of catalyst recovery has been
demonstrated in three different systems, including two cases
where >99% ee can be achieved under recycling conditions.
We are currently exploring applications of these ligands in other
catalytic processes.
-
entry
run number
conversion
% ee
1
2
3
4
1
2
3
4
100
100
100
100
66
80
81
84
Experimental Section
General Methods. All anaerobic reactions were carried out in an
inert atmosphere of nitrogen in a Vacuum Atmospheres drybox or by
using Schlenk techniques. Methylene chloride was distilled from
calcium hydride under nitrogen and stored over molecular sieves.
Tetrahydrofuran (THF) and diethyl ether were distilled under nitrogen
from sodium/benzophenone ketyl. All chemicals were purchased from
Aldrich Chemical Co. unless otherwise noted. Analytical TLC was done
on E. Merck precoated (0.25 mm) silica gel 60 F254 plates. Column
chromatography was conducted by using silica gel 40 (Scientific
Adsorbents Incorporated, Microns Flash). 1H NMR spectra were
recorded in CDCl3 unless otherwise mentioned; coupling values are
reported in hertz. Gas chromatographic analyses were performed using
an HP-ultra-1 cross-linked methyl silicone capillary column (25 m
length × 0.2 mm i.d.).
Table 5. Catalyst Recovery and Reuse in Hydrogenation of
Methyl Acetamidoacrylate Using 1 Mol % of [13]Rh+[Cod] BF4
-
(1:1 MeOH/water with 1 Mol % of Added Ligand)
entry
run number
conversion
% ee
1
2
3
1
2
3
100
100
35
>99
>99
>99
Table 6. Hydrogenation of Methyl Acetamidoacrylate Using 1
Mol % of [16]Rh+[NBD] SbF6 in Water. Catalyst Recovery and
-
Reuse in 100% Water with 1 Mol % of Added Ligand.a
entry
run number
conversion
% ee
The hydrogenation reactions were carried out as follows. In a drybox,
a Fischer-Porter tube was charged with the enamide substrate (0.1
mmol), the appropriate solvent (2 mL), and preformed [COD]Rh+LX-
(1 mol %). After sealing, the tube was removed from the drybox and
placed behind proper shielding. After five vacuum-refilling cycles with
hydrogen, the tube was brought to the appropriate pressure (40 psi) of
H2, and the mixture was vigorously stirred for 3-6 h. After removing
the catalyst on a plug of silica gel, the ee’s of the product were
determined by chiral GC (Chirasil-L-Val on WCOT fused silica 25 m
× 0.25 mm). Baseline separation of the enantiomers is observed, and
the ee’s are reproducible within ( 0.5%.
Hydrogenation in aqueous media was also carried out as described
in the previous paragraph. Water used for these experiments was
degassed by a repeated freeze-thaw procedure. In the recycling
experiments, the Fischer-Porter tube was taken inside the drybox, and
the product was extracted into ether (5 mL each, three times). To the
aqueous layer was added an additional 0.5 mL degassed water to
1
2
3
4
5
1
2
3
4
5
100
100
100
100
100
>99
>99
>99
>99
∼97
a
See Figure 1 for chromatograms.
pholanes from acid-labile protected derivatives. These proce-
dures circumvent two of the commonly encountered limitations
in the synthesis of these potentially important ligands: (a)
formation of phosphonium salts from the highly basic phosphine
under acidic conditions, and (b) the need to start with preformed
fully protected cationic metal complex. Thus, cationic Rh
complexes, (and presumably other metal complexes) can be
prepared in a separate step, and they have been found to be
excellent catalysts for aqueous phase hydrogenation of dehydro-

10212 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
RajanBabu et al.
facilitate the separation of the two layers in the subsequent runs. Fresh
substrate was added to the aqueous solution of the recovered catalyst,
and the hydrogenation was continued at 40 psi of hydrogen. The ether
layer was concentrated and analyzed by gas chromatography as
mentioned earlier.
Syntheses of precursors 53a,11a and 711a from D-mannitol have been
reported.
(t, J ) 5.1), 83.84, 130.11, 133.34 (t, J ) 3.1), 143.35 (d, J ) 4.0). 31
NMR (CD3OD): δ -14.24 (s).
P
Deprotection of 6: Synthesis of Tetrahydroxyphospholane 13
[(S,S,S,S)]. The title compound was prepared by a route similar to that
of the previous experiment for the synthesis of 14, except starting with
the bisphospholane of 6 instead of the bisphospholane of 7 (yield: 74%).
1H NMR (CD3OD): δ 0.91 (t, J ) 7.3, 6H), 0.93 (t, J ) 7.3, 6H),
1.27 (m, 2H), 1.44 (m, 4H), 1.71 (m, 2H), 1.92 (m, 2H), 2.58 (m, 2H),
2.87 (m, 2H), 3.38 (m, 2H), 4.31 (m, 4H), 7.28 (dd, J ) 3.2 and 5.4,
2H), 7.97 (dd, J ) 3.0 and 5.7, 2H). 13C NMR (CD3OD): δ 14.45 (t,
J ) 4.0), 14.81 (t, J ) 5.2), 21.32, 23.20 (t, J ) 16.1), 43.56 (t, J )
7.8), 46.98, 79.40, 80.23, 129.01, 134.89, 144.38 (t, J ) 6.0). 31P NMR
(CD3OD): δ -10.36 (s). Elemental analysis for C22H36O4P2 Calcd: C,
61.96; H, 8.51. Found: C, 62.22; H, 8.42.
Deprotection of 5: Synthesis of Tetrahydroxyphospholane 12
[(S,S,S,S)]. The title compound was prepared by a route similar to that
of the previous experiment for the synthesis of 14, except starting with
the bisphospholane of 5 instead of the bisphospholane of 7 (yield: 82%).
1H NMR (CD3OD): δ 0.83 (m, 6H), 1.27 (m, 6H), 2.77 (m, 2H), 2.91
(m, 2H), 4.13 (m, 4H), 7.28 (dd, J ) 3.3 and 5.6, 2H), 7.70 (m, 2H).
13C NMR (CD3OD): δ 13.07, 14.32 (t, J ) 17.4), 34.77 (t, J ) 6.7),
35.28, 80.72, 80.83, 129.34, 133.13, 143.54 (t, J ) 4.9). 31P NMR (CD3-
OD): δ -3.77 (s). Elemental analysis for C18H28O4P2 Calcd: C, 58.37;
Synthesis of the Mesylate Precursor 9b. The title compound was
prepared in two steps from 1,2:5,6-dianhydro-3,4-O-isopropylidene-
D-mannitol (Scheme 1). A 1.4 M solution of MeLi in ether (89 mL,
124.6 mmol) was added to a suspension of CuI (11.8 g, 62 mmol) in
ether (25 mL) at -20 °C. After 2 h at -10 °C, the mixture was cooled
to -30 °C, and a solution of the diepoxide (3.86 g, 20.7 mmol) in
ether (10 mL) and THF (10 mL) was added. After stirring at -20 °C
for 3 h, the reaction mixture was poured into cold saturated NH4Cl,
and the product was extracted with ether. The extract was washed with
brine, dried over MgSO4, and evaporated on a rotary evaporator to
leave an oil, which was chromatographed on silica gel with ethyl
acetate/hexane as eluant to get 9a as a colorless oil (4.38 g, 97%, crude).
1H NMR (CDCl3): δ 1.00 (t, J ) 7.4, 6H), 1.35 (s, 6H), 1.82 (m, 2H),
3.52 (m, 4H), 3.66 (dd, J ) 1.9 and 6.0, 2H); 13C NMR: δ 9.24, 26.82,
26.99, 74.28, 82.83, 108.66.
To a mixture of 2.0 g (9.16 mmol) of diol 9a and 30 mg of DMAP
in 10 mL of pyridine and 5 mL of CH2Cl2 was added, slowly, 1.95 mL
(3 equiv) of methanesulfonyl chloride at 0 °C. After stirring for 6 h at
that temperature, the reaction mixture was poured into 100 mL of ice-
cold 2 N HCl, and the product was extracted with CH2Cl2 (3 × 30
mL). The combined extracts were successively washed with 30 mL of
water and saturated NaHCO3 solution and then dried with anhydrous
MgSO4. Removal of the solvent on a rotary evaporator gave the crude
product which was purified by column chromatography on silica gel
(elution with ethyl acetate/hexane, 70/30). Dimesylate 9b was obtained
in 87% yield as an oil. 1H NMR: δ 1.07 (t, J ) 7.4, 6H), 1.42 (s, 6H),
1.87 (m, 4H), 3.10 (s, 6H), 4.19 (dd, J ) 1.5 and 3.9, 2H), 4.71 (m,
2H). 13C NMR: δ 9.02, 23.73, 27.10, 38.86, 78.57, 82.80, 111.07.
Synthesis of Bisphospholane 6. To a solution of 1,2-bisphosphano-
benzene (142.1 mg, 1.0 mmol) in 5 mL of THF was added n-BuLi
(1.4 mL of 1.6 M solution in hexane) via a syringe at -78 °C. The
solution was warmed to room temperature and was stirred for an
additional 1 h. After cooling to 0 °C, a solution of 9b (749.4 mg, 2.0
mmol) in 4 mL of THF was added, and the mixture was stirred for 2
h. Then, 2.2 equiv of n-BuLi (1.4 mL) were added dropwise at 0 °C,
and the orange suspension was stirred overnight. After 14 h, 8 mL of
water was added, and the THF layer was separated. The aqueous phase
was extracted with ether (2 × 15 mL). The combined organic extracts
were dried over sodium sulfate, and the solvent was removed under
reduced pressure. The crude product was purified by flash chromatog-
raphy on silica gel inside the drybox (elution with 5:95 v/v ether/hexane)
to obtain 115 mg (23%) of diphospholane 6 as a white solid. 1H NMR
(CDCl3): δ 0.77 (t, J ) 7.0, 6H), 0.86 (m, 2H), 0.97 (t, J ) 7.3, 6H),
1.32 (m, 4H), 1.47 (s, 12H), 2.00 (m, 2H), 2.20 (m, 2H), 2.62 (m, 2H),
4.40 (dd, J ) 6.6 and 10.4, 2H), 4.47 (dd, J ) 7.4, 10.4, 2H), 7.37 (m,
4H). 13C NMR: δ 13.08 (t, J ) 6.1), 14.57 (t, J ) 2.2), 21.15 (t, J )
14.4), 21.39, 27.25, 27.36, 32.85 (t, J ) 9.7), 32.98 (t, J ) 2.6), 81.37
(t, J ) 6.1), 82.30, 117.05, 129.15, 131.05, (t, J ) 2.7), 141.29 (t, J )
6.0). 31P NMR (CDCl3): δ 35.63. Elemental analysis for C28H44O4P2
Calcd: C, 66.38; H, 8.75. Found: C, 66.35; H, 8.79.
H, 7.62. Found: C, 56.69; H, 7.46.
31P NMR of the Cationic Rh Complexes. [COD]Rh+[7] BF4
:
:
31P
-
NMR (CD3OD) 65.03 (d, JRh-P ) 152.7 Hz). [COD]Rh+[5] BF4
31P
-
NMR (CD3OD) 77.66 (d, JRh-P ) 151.8 Hz). [NBD]Rh+[8] SbF6
:
-
31P NMR (CD3OD) 63.0 ppm (d, JRh-P ) 154.4 Hz). [COD]Rh+[12]
31P NMR (CD3OD) 77.66 (d, JRh-P ) 151.6 Hz). [COD]Rh+-
-
BF4
:
[13] BF4
:
31P NMR (CD3OD) 70.99 (d, JRh-P ) 151.2 Hz). [COD]-
-
Rh+[14] BF4
:
31P NMR (CD3OD) 65.02 (d, JRh-P ) 152.7 Hz).
-
-
:
31P NMR (CD3OD) 66.0 (d, JRh-P ) 155.4 Hz).
[NBD]Rh+[16] SbF6
1,6-Di-O-t-Butyldimethylsilyl-3,4-dideoxy-D-mannitol (10b). To
a solution of tetrol 10a (0.500 g, 3.33 mmol) and imidazole (0.567 g,
8.32 mmol) in 3 mL of N,N-dimethylformamide was added, at 0 °C, a
solution of tert-butyldimethylsilyl chloride (1.104 g, 7.32 mmol) in 3
mL of DMF. The mixture was stirred at 0 °C for 2.5 h. The reaction
mixture was subsequently diluted with 100 mL of ether and was washed
with brine (50 mL × 2) and water (50 mL). The organic layer was
dried (MgSO4), and the solvent was removed to give 0.890 g (71%) of
white crystalline solid, mp 79-81 °C (lit. 78 °C).
(2S,5S)-1,6-Di(tert-butyldimethylsilyloxy)-2,5-di(methanesulfon-
yloxy)hexane (11). To a stirred solution of diol 10b (258 mg, 0.681
mmol) and triethylamine (380 uL, 2.725 mmol) in 3 mL of CH2Cl2 at
-20 °C was added dropwise methanesulfonyl chloride (158.2 µL, 2.044
mmol). After complete addition, the mixture was stirred at -20 °C for
30 min longer and was diluted with 20 mL of ether. The solution was
washed with water (3 mL × 2) and dried (MgSO4), and the solvent
was removed. The residue was purified on a silica gel column (ethyl
acetate/hexane ) 1:4) to give a colorless oil, which became a crystalline
solid (322.0 mg, 88%) upon keeping in a refrigerator overnight. Mp
74-76 °C. 1H NMR (400 MHz, CDCl3): δ 0.00 (s, 6H, Me-Si × 2),
0.01 (s, 6H, Me-Si × 2), 0.82 (s, 18H, t-Bu × 2), 1.71 (m, 4H, C3/
C4-H), 3.00 (s, 6H, CH3 in Mesyl), 3.66 (d, J ) 5.2 Hz, 4H, C1/
C6-H), 4.66 (t, br, J ) 5.4 Hz, 2H, C2/C5-H). 13C NMR (100 MHz,
CDCl3): δ -5.0, 18.8, 26.3, 26.8, 39.0, 65.5, 83.6.
Preparation of Bisphospholane 8. To a solution of 1,2-bisphos-
phinobenzene (34.0 mg, 0.24 mmol) in 2 mL of THF was added n-BuLi
(0.21 mL of 2.5 M solution in hexane, 2.2 equiv) via syringe at -30
°C, and the mixture was stirred at that temperature for 1.5 h. To the
resulting orange solution was added, dropwise, dimesylate 11 (256.1
mg, 0.48 mmol) in 2 mL of THF at -30 °C, and the mixture was
slowly warmed to room temperature over 1 h. After cooling the mixture
to -30 °C, an additional 2.2 equiv of n-BuLi was added, and the
reaction mixture was allowed to stir overnight. After 16 h, the reaction
mixture was filtered through a Celite pad, and the pad was rinsed with
ether. The crude product was purified by flash chromatography on silica
gel inside the drybox (elution with 5:95 v/v ether/hexane) to give 98.3
Deprotection of 7: Synthesis of Tetrahydroxyphospholane 14
[(R,S,S,R)]. In a drybox, to a solution of 90 mg (0.20 mmol) of
bisphospholane 7 in 3 mL of methanol was added 40 mg of resin (AG
50 WX-8). The resin had been previously swollen in methanol overnight
at room temperature, filtered, washed three times with methanol, and
dried before use in this experiment. The reaction mixture was stirred
at room temperature overnight. The resin was filtered through Celite.
The solvent was removed under high vacuum, and the residue was
washed three times with hexane and dried under high vacuum to get
1
59 mg (80%) of 14 [(R,S,S,R)] as a white solid. H NMR (CD3OD):
1
0.89 (m, 6H), 1.33 (m, 6H), 2.06 (m, 2H), 2.46 (m, 2H), 3.48 (m, 4H),
7.42 (dd, J ) 3.3 and 5.7, 2H), 7.62 (m, 2H). 13C NMR (CD3OD):
15.73 (t, J ) 3.1), 18.18 (t, J ) 17.0), 34.24, 34.73 (t, J ) 6.3), 83.67
mg (50%) of pure bisphospholane 8 as colorless oil. H NMR (500
MHz, CDCl3): δ 0.13 (s, 6H), -0.10 (s, 6H), 0.04 (s, 6H), 0.06 (s,
6H), 0.77 (s, 18H), 0.90 (s, 18H), 1.64 (m, 2H, C3-H), 1.72 (m, 2H,

Neutral Polyhydroxy Phospholane DeriVatiVes
J. Am. Chem. Soc., Vol. 123, No. 42, 2001 10213
C4-H), 2.10 (m, 2H, C4′-H), 2.25 (m, 2H, C3′-H), 2.50 (m, 2H,
C5-H), 2.82 (m, 2H, C2-H), 2.86 (t, J ) 10.3, 2H, C6-H), 3.63 (m,
2H, C1-H), 3.68 (dd, J ) 4.2, 10.3 Hz, 2H, C6′-H), 3.82 (m, 2H,
C1′-H), 7.23 (m, 2H), 7.39 (m, 2H). 13C NMR (125 MHz, CDCl3):
δ 4.94, -4.88, -4.81, -4.79, 18.66, 18.81, 26.33, 26.39, 30.1, 30.5,
41.4 (JP-C ) 7.1), 42.1, 65.7 (JP-C ) 4.8), 67.2 (JP-C ) 27.2), 128.6,
131.9, 141.8. 31P NMR (200 MHz, CDCl3): δ -13.4 ppm (s). 31P NMR
(200 MHz, CD3OD) for cationic rhodium complex [Rh(8)(NBD)]BF4:
δ 63.0 ppm (d, JRh-P ) 154.4).
for 30 min at room temperature. [Rh (16)(NBD)]SbF6 31P NMR (200
MHz, CD3OD) : δ 66.0 ppm (d, JRh-P ) 155.4).
Typical Hydrogenation Procedure using [Rh(16)(NBD)]+ SbF6
.
-
Inside a vacuum atmosphere drybox, the Rh complex (1.6 mg, 2.0 µmol)
and free ligand 16 (0.7 mg, 2.0 µmol) in 2 mL of neat water (degassed
as described earlier) were placed in a Fisher-Porter bottle, and methyl
2-acetamidoacrylate (28.6 mg, 0.200 mmol) was added. The bottle was
sealed and then taken outside the drybox. The bottle was charged with
50 psi of H2 and then was evacuated under house vacuum. This cycle
was repeated five times before it was finally charged with 50 psi of
H2. With efficient stirring, the mixture was hydrogenated at room
temperature for 5-7 h. After this period, the bottle was brought inside
the drybox. The aqueous layer (2 mL) was extracted with 4 mL of
Et2O with vigorous magnetic stirring. This extraction was repeated 3-4
times, and then the ethereal layer was dried (MgSO4) and evaporated
to give methyl N-acetylalanate (isolation: ∼90%). The GLC analysis
100 °C (isothermal run at 100 °C) of the product was done on a chiralsil
L-Val column.
Hydrolysis of the tert-Butyldimethlsilyl Ethers. Tetrahydroxy-
phospholane 16. To a mixture of 8 (64.2 mg, 0.078 mmol) and 2 mL
of MeOH was added CF3CO2H (44 mg, 0.38 mmol, 5.0 equiv), and
the mixture was stirred at room temperature for 10 h. The progress of
deprotection can be checked by TLC (THF as eluent, Rf ) 0.45). The
reaction was stirred for 10 h. After evaporation of solvent to dryness,
the residue was redissolved in MeOH and was filtered through a 2 cm
pad of polystyrene-CH2NEt2 resin to remove residual acid. The solvent
was removed, and the resulting white solid was purified by flash
chromatography using THF as eluent to give 16 as a white solid.
Unprotected phospholane and partially deprotected phospholane were
combined, and the mixture was treated with CF3CO2H in MeOH to
give complete deprotection. The combined yield after chromatography
Recycling and Reuse of the Catalyst ([Rh(16)(NBD)]+ SbF6-).
To the aqueous layer above was added 28.6 mg of methyl 2-acetami-
doacrylate (28.6 mg, 0.200 mmol). The same procedure was repeated
for hydrogenation. After the fourth cycle, more water was added because
of loss of water during extraction.
1
was 24.1 mg (84%). H NMR (500 MHz, CD3OD): δ 1.55 (m, 2H,
C4-H), 1.64 (m, 2H, C3-H), 2.08 (m, 2H, C3′-H), 2.30 (m, 2H,
C4′-H), 2.41 (m, 2H, C2-H), 2.80-2.92 (m, 4H, C5-H, C1-H),
3.49 (dd, J ) 5.4 and 11.1, 2H, C1′-H), 3.55-3.72 (m, 4H, C6-H,
C6′-H), 7.31 (m, 2H), 7.46 (m, 2H). 13C NMR (125 MHz, CD3OD):
δ 29.6 (C4), 30.5 (C3), 42.1 (C5), 42.4 (JP-C ) 6.8 Hz, C2), 64.0 (JP-C
Acknowledgment. This work was supported by the EPA/
NSF Program for Technology for a Sustainable Environment
(R826120-01-0) and the NSF (Grants CHE-9706766 and CHE-
0079948).
) 3.9 Hz, C1), 65.7 (JP-C ) 22 Hz, C6), 128.8, 132.1, 141.6 (JP-C
)
4.8 Hz). 31P NMR (202 MHz, CD3OD): δ -12.4 ppm (s).
Supporting Information Available: NMR spectra (1H, 13C,
and 31P) of 8, 12-14, 16 (PDF). This material is available free
of charge via Internet at http://pubs.acs.org.
Procedure for Formation of Rh Complexes. To a solution of
tetrahydroxybisphospholane 16 (12.4 mg, 33.5 µmol) in 1 mL of
MeOH-d4 was added [Rh(NBD)2]+ SbF6- (15.9 mg, 30.4 µmol) [NBD
) norbornadiene], and the resulting bright orange solution was stirred
JA011500A