New catalysts for the asymmetric hydrosilylation of ketones discovered by mass spectrometry screening

Article abstract of DOI:10.1021/jo026365e

A method for determining enantiomeric excess by mass spectrometry was employed to screen a family of chiral phosphite P,N-ligands for activity in the rhodium-catalyzed asymmetric hydrosilylation of ketones. The identification of an effective set of ligand

Full text of DOI:10.1021/jo026365e

New Ca ta lysts for th e Asym m etr ic Hyd r osilyla tion of Keton es
Discover ed by Ma ss Sp ectr om etr y Scr een in g
Sulan Yao, J un-Cai Meng, Gary Siuzdak, and M. G. Finn*
Department of Chemistry, The Scripps Research Institute, 10550 North Torrey Pines Road,
La J olla, California 92037
mgfinn@scripps.edu.
Received August 28, 2002
A method for determining enantiomeric excess by mass spectrometry was employed to screen a
family of chiral phosphite P,N-ligands for activity in the rhodium-catalyzed asymmetric hydro-
silylation of ketones. The identification of an effective set of ligands was followed by preliminary
studies of the reaction scope and mechanism. Asymmetric induction of 84-88% ee for larger-scale
reactions was observed, which is close to the level of the best alternative catalysts previously
discovered. The screening method was shown to be applicable to a variety of substrates without
the need for special optimization.
3-16
The asymmetric reduction of prochiral ketones is a
powerful approach to the synthesis of chiral alcohols.1
Hydrosilylation provides for reaction conditions that are
exceedingly mild, the process being driven by the forma-
tion of a strong Si-O bond. Asymmetric variants of this
reaction have been studied for many years.2 A con-
venient, albeit relatively undemanding, standard for
comparison is acetophenonesthe one substrate common
to the testing of almost all reported catalysts. By this
measure, the state of the art is 94-97% ee, reached by
several systems. Bis(oxazolidine)-pyridine (Pybox) com-
plexes of rhodium give up to 94% ee,4 while three
to Rh- and Ru-catalyzed hydrosilylation.1
Particularly
appropriate for prospects of modular, and therefore
combinatorial, synthesis were chiral phosphite and phos-
phonate ligands reported by the groups of Seebach (such
Scharf, Pastor, and Pfaltz (2, used for
asymmetric Pd-catalyzed allylation). Also intriguing
was the report of dramatic temperature and solvent
as 1),1
7,18
19
20
,3
21
1
9
effects in the asymmetric hydrosilylation reaction,
making methods of efficient reaction optimization of
interest for the practical development of effective condi-
tions for particular substrates.
ferrocenyl-based systems provide levels of enantioselec-
tivity in the range of 95-97% ee.5
-10
We were attracted
to this reaction as a suitable test of our recently developed
mass spectrometry enantiomeric excess determination
(
MSEED) analytical method11,12 in the context of a
combinatorial-style search for new catalysts. An attrac-
tive feature was the prevalent application of P,N-ligands
A set of 21 phosphite P,N-ligands related to 1 and 2,
containing chiral diol and amino alcohol components 3
(
1) Itsuno, S. Org. React. 1998, 52, 395-576.
(
2) Wallbaum, S.; Martens, J . Tetrahedron: Asymmetry 1992, 3,
1
475-1504.
(12) D ´ı az, D. D.; Yao, S.; Finn, M. G. Tetrahedron Lett. 2001, 42,
2617-2619.
(13) Kuwano, R.; Sawamura, M.; Shirai, J .; Takahashi, M.; Ito, Y.
Bull. Chem. Soc. J pn. 2000, 73, 485-496.
(14) Tsuruta, H.; Imamoto, T. Tetrahedron: Asymmetry 1999, 10,
877-882.
(15) Newman, L. M.; Williams, J . M. J .; McCague, R.; Potter, G. A.
Tetrahedron: Asymmetry 1996, 7, 1597-1598.
(16) Moreau, C.; Frost, C. G.; Murrer, B. Tetrahedron Lett. 1999,
40, 5617-5620.
(17) Heldmann, D. K.; Seebach, D. Helv. Chim. Acta 1999, 82, 1096-
1110.
(
3) Brunner, H.; Nishiyama, H.; Itoh, K. In Catalytic Asymmetric
Synthesis; Ojima, I., Ed.; VCH: New York, 1993; Chapter 6, pp 303-
3
22.
(4) Nishiyama, H.; Yamaguchi, S.; Kondo, M.; Itoh, K. J . Org. Chem.
1
992, 57, 4306-4309.
(5) Nishibayashi, Y.; Segawa, K.; Takada, H.; Ohe, K.; Uemura, S.
Chem. Commun. 1996, 847-848.
6) Lee, S.; Lim, C. W.; Song, C. E.; Kim, I. O. Tetrahedron:
Asymmetry 1997, 8, 4027-4031.
7) Sudo, A.; Yoshida, H.; Saigo, K. Tetrahedron: Asymmetry 1997,
, 3205-3208.
8) Nishibayashi, Y.; Takei, I.; Uemura, S.; Hidai, M. Organo-
metallics 1998, 17, 3420-3422.
9) Langer, T.; J anssen, J .; Helmchen, G. Tetrahedron: Asymmetry
996, 7, 1599-1602.
10) Hayashi, T.; Hayashi, C.; Uozumi, Y. Tetrahedron: Asymmetry
995, 6, 2503-2506.
11) Guo, J .; Wu, J .; Siuzdak, G.; Finn, M. G. Angew. Chem., Int.
Ed. 1999, 38, 1755-1758.
(
(
8
(
(18) Sakaki, J .; Schweizer, W. B.; Seebach, D. Helv. Chim. Acta
1993, 76, 2654-2665.
(
(19) Haag, D.; Runsink, J .; Scharf, H.-D. Organometallics 1998, 17,
398-409.
1
1
(
(20) Pastor, S. D.; Shum, S. P. Tetrahedron: Asymmetry 1998, 9,
543-546.
(
(21) Pr e´ t oˆ t, R.; Pfaltz, A. Angew. Chem., Int. Ed. 1998, 37, 323-
325.
1
0.1021/jo026365e CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/27/2003
2
540
J . Org. Chem. 2003, 68, 2540-2546

Catalysts for Asymmetric Hydrosilylation of Ketones
SCHEME 1
CHART 1
and 4, was initially assembled by the general method
shown in Scheme 1. Stepwise attachment of diol and
amino alcohol fragments to phosphorus was accomplished
for each ligand in one pot to give phosphites 6. The
structures of the components employed are shown in
Chart 1. Amino alcohol components were either used as
purchased (i.e., most of the dimethylamino-containing
compounds) or capped at nitrogen prior to reaction with
the intermediate chlorophosphinite 5. All ligands were
obtained in high yields and high purity after flash
chromatography using degassed solvent, and were char-
This screen was performed to provide preliminary evalu-
ation of the candidates and to validate the MSEED
method1
1,12
in a catalysis trial. The reactions were
performed on a 50 µmol scale in substrate to provide
samples for ee determination by chiral HPLC as well as
MSEED. Thus, silyl ether 8 was hydrolyzed to the alcohol
9 for evaluation by HPLC, and an aliquot of the silyl
ether was converted to a mixture of diastereomeric esters
11 by the one-pot combination of fluoride deprotection
and carbodiimide-mediated coupling shown. This proce-
dure is necessary to avoid the incorporation of water into
aliquots analyzed by acylation with 10a + 10b. While
the use of an excess of these mass-tagged acids allows a
small amount of water to be tolerated, too much water
poisons the acylation reaction and leads to poor analytical
results. The use of N-acylproline reagents 10a ,b has been
previously described.11
1
31
acterized initially by H and P NMR. The identities and
purities of the candidates selected for further investiga-
1
3
tion (see below) were further substantiated by C NMR,
IR, and combustion analysis.
Complexes of the assembled ligands with rhodium(I)
were made by mixing with [Rh(COD)Cl] and tested for
2
asymmetric catalytic activity toward the hydrosilylation
of 1-naphthyl methyl ketone (7) as shown in Scheme 2.
The results of the preliminary assay are shown in
Figure 1 for two sets of reactions, differing only in the
J . Org. Chem, Vol. 68, No. 7, 2003 2541

Yao et al.
SCHEME 2
ratio of ligand to rhodium. (R)-p-Tolyl-BINAP (12) and
S)-Pybox (13)3
,4,22,23
were included for comparison. In the
(
first set of reactions, employing a ligand:metal ratio of
.0, all of the reactions showed a fair to good degree of
2
completion as measured by a comparison of the HPLC
peaks for the starting ketone and product 9, corrected
for their relative molar absorbtivities. The values for
enantiomeric excess of the product as measured by HPLC
and MSEED were quite similar in most cases and within
(
10% ee for all reactions showing substantial enantio-
selectivity. With the exception of diminished reactivities
for four ligands (ABG, BBG, AAG, and 13), the results
for a ligand:Rh ratio of 4:1 are comparable to those of
the 2:1 set.
F IGURE 1. Results of hydrosilylation of 1-acetonaphthanone
with a selection of chiral ligands. Key: yellow, percent comple-
tion as measured by HPLC; blue, % ee as measured by the
kinetic resolution/mass spectrometry (MSEED) method; red,
Ligand CEA, derived from (+)-TADDOL and (1S,2R)-
%
ee as measured by chiral HPLC (Chiralcel OD).
(
+)-N-methylephedrine, immediately emerged as supe-
rior among the P,N-systems tested, with results slightly
better than those of p-tolyl-BINAP ligand 12. Also
noteworthy among the results are the following observa-
tions: (a) Epimeric ligands CDA and CEA gave dramati-
cally different results, showing that the configuration of
the R-amino carbon center is important. (b) Asymmetric
induction dropped upon changing the N-methyl substit-
uents of CEA to N-butyl groups (CEB), showing that
some steric congestion apparently exists at or near the
amine center. (c) None of the imine-containing ligands
(
incorporating NR components G-P ) were effective,
2
despite the common use of such ligating moieties in the
chemistry of Rh(I) and other electron-rich transition-
metal centers. Unfortunately, imine-containing ligands
incorporating the effective TADDOL-norephedrine back-
bone (such as CEG) were found to be unstable toward
hydrolysis during purification.
The modified set of ligands shown in Chart 2 was
synthesized and tested as above to gain further insight
into structure-activity relationships. The results for
1-acetonaphthanone and diphenylsilane, determined by
chiral HPLC, are summarized in Table 1. The “matched”
nature of the diol and amino alcohol in ligand CEA is
apparent in the poor performance of the catalyst contain-
ing the “mismatched” N-methylephedrine in CF A (entry
2 vs 3). Omitting either the amine substituent (CH, entry
(
22) Nishiyama, H.; Kondo, M.; Nakamura, T.; Itoh, K. Organo-
metallics 1991, 10, 500-508.
23) As shown in Figure 1, Pybox ligand 13 performs poorly under
(
our conditions, in contrast to its reported effectiveness when used in
a 7:1 ligand:Rh ratio (ref 4). Differences between the reaction conditions
include the much shorter time used here for “aging” of the ligand +
metal mixture (1 h vs 24 h), the lack of silver salt in our system, and
the ligand-to-metal ratio.
2
542 J . Org. Chem., Vol. 68, No. 7, 2003

Catalysts for Asymmetric Hydrosilylation of Ketones
CHART 2
TABLE 1. Resu lts for Hyd r osilyla tion of
-Aceton a p h th a n on e (7) a t Room Tem p er a tu r e u n d er
Sta n d a r d Con d ition s (0.5% [Rh (cod )Cl]2, 2% Liga n d ,
P h 2SiH2, Tolu en e), Deter m in ed by Ch ir a l HP LC
4
) or chiral substitution on the amino alcohol unit (CGA,
1
entry 5) gave poor enantioselectivity. Both the phenyl and
methyl substituents on the ephedrine backbone are
necessary, as shown by entries 6 and 7 (CEA vs CIA and
CJ A). The poor result with the N,N-dibutyl derivative
CEB (Figure 1) prompted us to explore the steric
environment about the amine center with ligands CEC
and CED, incorporating piperidine and pyrrolidine rings,
respectively. Both of these ligands also gave very poor
catalysts (entries 8 and 9), as did the replacement of the
backbone methyl group with a phenyl moiety in ligand
CBA (entry 10). These results suggest that the steric
environment about the 1-position of the ephedrine com-
ponent is hindered, and/or that there is a strong de-
pendence of catalyst performance on the conformation
about the central ephedrine C-C bond. An attempt to
restrict that conformation with ligand CKA did not
provide good asymmetric induction, although catalytic
activity was high (entry 11). Last, replacement of the
TADDOL phenyl groups with naphthyls (DEA, entry 12)
produced an effective catalyst, although slightly less
enantioselective than the “parent” CEA structure.
%
ee
%
% ee
%
entry ligand (config) conv entry ligand (config) conv
1
2
3
4
5
6
none
CEA
CF A
CH
CGA
CIA
0
59
95
92
97
76
93
7
8
9
10
11
12
CJ A
47 (R)
2 (R)
4 (R)
8 (S)
18 (R)
66 (R)
94
62
29
38
95
91
78 (R)
0.9
14 (S)
30 (S)
31 (S)
CEC
CED
CBA
CKA
DEA
based DIOS (desorption/ionization on silicon) mass spec-
24-27
trometry method.
In this variation, the hydrosilyl-
ation and subsequent mass-tagged derivatization reac-
tions were performed as before (Scheme 2), but the
resulting samples were analyzed by depositing 0.1-0.5
µL aliquots on a porous silicon plate (100 samples per
plate) followed by DIOS-MS analysis, instead of injection
into an electrospray ionization mass spectrometer. The
data collection rate is thereby reduced from 5 min to
24
approximately 10 s per sample. The results are shown
The performance of a subset of the above catalysts with
a selection of seven ketones (7, 15-20) was measured
using the MSEED technique implemented in the chip-
in Table 2. Note that runs giving higher values of % ee
were checked by chiral HPLC, showing the DIOS-
MSEED method to be a reliable and convenient screening
method for asymmetric induction.
The results show that ligands CEA and CJ A afford
the most consistently selective reductions, with enantio-
selectivities in the 70-94% ee range for five of the seven
(24) Shen, Z.; Yao, S.; Crowell, J . E.; Siuzdak, G.; Finn, M. G. Isr.
J . Chem. 2002, 41, 313-316.
(
(
25) Wei, J .; Buriak, J .; Siuzdak, G. Nature 1999, 399, 243-246.
26) Shen, Z.; Thomas, J . J .; Averbuj, C.; Broo, K. M.; Engelhard,
M.; Crowell, J . E.; Finn, M. G.; Siuzdak, G. Anal. Chem. 2001, 73, 612-
19.
27) Thomas, J . J .; Shen, Z.; Crowell, J . E.; Finn, M. G.; Siuzdak,
G. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 4932-4937.
6
(
J . Org. Chem, Vol. 68, No. 7, 2003 2543

Yao et al.
TABLE 2. Scr een in g of Asym m etr ic Hyd r osilyla tion of a Selection of Keton es (Colu m n s) Usin g th e In d ica ted Liga n d s
(
Row s) in a 2:1 Liga n d :Rh Mola r Ra tio u n d er Sta n d a r d Con d ition sa
7
15
16
17
18
19
20
CDA
CEA
DEA
CED
CJ A
CBG
13
80 (81.9)
29
46
58.7 (54.6)
21
67 (62.3)
33
-4
16
36
19
36
59
34
58
51
26
-17
15
6
13
11
3
29
98 (89.1)
99 (92.5)
18
48
17
-30
31
-20
84 (80.4)
102 (93.9)
23
75 (78.2)
60.4 (62.5)
45
99 (93.6)
48
43
39
35
63.0 (64.9)
80.9 (70.0)
60 (63.9)
41
1
20
48
26
1
1
2
3
60 (53.4)
50
16
3
a
The values given are % ee measured by the DIOS-MSEED method; all reactions proceeded to significant levels of completion (>50%),
as determined by relative mass spectral peak intensities. Values in parentheses were obtained by separate analysis on chiral HPLC
Chiralcel OD); negative numbers denote that the S-enantiomer is present in excess. The most selective reactions for each ketone are
(
marked with values in bold type.
SCHEME 3
substrates. Ethyl ketone 16 is accommodated best by
ligand CJ A, but the maximum enantioselectivity (ca.
5
9%) is only fair. Isopropyl ketone 18 is resistant to
asymmetric hydrosilylation by this set of complexes.
Interestingly, ethyl pyruvate (19) is quite efficiently
handled by Rh complexes of CEA or DEA. Ketones 17
and 20 were subjected to reduction on a 1 mmol scale
with ligand CEA to obtain the corresponding (R)-alcohols
in 86.1% ee (54% isolated yield) and 84.0% ee (67%
isolated yield), respectively. Ketone 19 was reduced on a
1
mmol scale using ligand DEA in 88.0% ee (50% yield).
The values obtained in these preparative runs differ from
the screening results due to necessary changes in con-
centration and starting temperature, and presumably can
be further optimized.28
Several additional parameters were checked for a
subset of effective ligands. The addition of silver salts to
rhodium and ruthenium halide catalysts has been re-
1
species (115.6 ppm, d, J P,Rh ) 254 Hz), a minor signal
1
(
125.3 ppm, d, J P,Rh ) 232 Hz, approximately 1/8 the
signal intensity of the major resonance), and a trace
signal for the uncomplexed ligand (140.1 ppm, s). The
analogous 2.1:1 (ligand/Rh) mixture showed the “major”
complex as the only Rh-phosphine species, along with
the free ligand signal at equal intensity. These data
demonstrate that a complex of 1:1 Rh:CEA stoichiom-
etry, presumably a monomer, is formed even in the
ported to give improved activity and enantioselectiv-
ity.1
6,22
However, when screened by normal (i.e., using
electrospray ionization MS) MSEED with 2:1 L/Rh
catalysts involving ligands CBG, CBN, CCE, CEA, CH,
1
presence of excess ligand. The H NMR spectra re-
vealed bound cyclooctadiene to be present, so the pre-
catalyst can be formulated as [Rh(cod)(CEA)]Cl, 20a , or
Rh(cod)(CEA)Cl], 20b. The difference between them is
the occupancy of a coordination site by the pendant amine
or chloride, respectively. Monophosphite rhodium com-
plexes have been previously shown to be good catalysts
of asymmetric hydrogenation reactions.30
and 12, the addition of AgBF
with respect to rhodium gave rise to no improvement in
asymmetric induction, while AgPF and AgOTf slowed
4
in 1-4-fold molar excess
[
6
the reaction noticeably without improving enantioselec-
tivity. The use of other silanes (with analysis of ee by
chiral HPLC) showed Ph
to be the best in the reduction of 7.
2 2 2
SiH and (1-naphthyl)(Ph)SiH
2
9
The sensitivity of the process to the presence, size, and
stereochemistry of the amino substituent of CEA (de-
scribed above) suggests that coordination of N to Rh may
be important in the catalytic cycle. This was supported
by the observation that the addition of triflic acid to a
3
:1 ligand/Rh catalyst mixture in either toluene or THF
poisoned the enantioselectivity of the reaction while
retaining catalytic activity (Scheme 3).
Examination of the 31P NMR spectrum of the complex
2
formed from a mixture of [Rh(cod)Cl] and CEA (ligand:
Rh ) 1.2:1) in benzene-d showed a single dominant
6
The question of nitrogen coordination in 20 is of
interest because known P,N-ligands almost invariably
employ sp -hybridized nitrogen centers, expected to be
well matched with the “soft” nature of the Rh(I) center,
(
28) Due to heating of the solution upon addition of silane, for the
2
larger-scale reactions the mixture of ketone, catalyst, and solvent was
cooled to -10 °C before addition of silane. The reaction mixtures were
maintained at 0 °C for 1 h and then allowed to warm to room
temperature.
in contrast to the tertiary amine moiety of CEA and
1
15
31
related ligands. H- N HMQC NMR spectroscopy was
(
29) Hydrosilylation of 7 in the presence of 1% Rh and 2% CEA
toluene, room temperature) gave the following values of enantiomeric
excess (all with the R-isomer in excess): Ph SiH , 77.7%; PhSiH , 1.9%;
PhMeSiH , 5.8%; (1-naphthyl)PhSiH , 77.9%; Et SiH , 34.4%; poly-
methylhydrosiloxane), no reaction.
(
2
2
3
(30) Reetz, M. T.; Mehler, G. Angew. Chem., Int. Ed. 2000, 39, 3889-
3890.
(31) Martin, G. E.; Hadden, C. E. J . Nat. Prod. 2000, 63, 543-585.
2
2
2
2
(
2
544 J . Org. Chem., Vol. 68, No. 7, 2003

Catalysts for Asymmetric Hydrosilylation of Ketones
1
15
2 6 6
F IGURE 2. H- N HMQC NMR spectra of CEA (left) and a 1.05:1 (ligand/Rh) mixture of CEA and [Rh(cod)Cl] (right) in C D
solvent.
used to extract 15N chemical shifts by magnetization
transfer from protons coupled to the nitrogen center of
interest. The spectra, shown in Figure 2, revealed a scant
the precatalyst, and it would be difficult to detect 15N-
03 33,34
1
Rh coupling (expected 8-20 Hz
) under these condi-
tions. Amine coordination to rhodium is therefore not
proven, but we regard it as likely during the catalytic
cycle.
2
.3 ppm shift in the 15N resonance upon complexation of
the ligand to rhodium, suggesting that the nitrogen
center is not coordinated to the metal. However, one of
We have described here the implementation of mass
spectrometry-based screening methods for the discovery
of catalysts for the asymmetric hydrosilylation of a
variety of ketones. The MSEED technique, and especially
the variant incorporating DIOS mass spectrometry, is a
useful tool, particularly when the testing of different
substrates would otherwise require the cumbersome
optimization of chromatographic techniques for enantio-
meric excess determination in each case.
The family of chiral phosphite P,N described here are
easy to make and moderately effective in providing an
environment for asymmetric induction. High sensitivity
to the nature of the silane is also demonstrated. The
chiral ligand that emerges as the best candidate, CEA,
bears diol, amino alcohol, and N-alkyl components that
each are shown to be important to overall performance.
The further optimization of this type of ligand and the
application of our methods to other ligands and reactions
are subjects of current interest in our laboratory.
the NMe
2
methyl groups to which the 15N resonance is
coupled shifted upfield by 0.33 ppm, consistent with its
occupancy of a position in which it experiences shielding,
possibly close to the metal center or an aromatic sub-
stituent. Ephedrine-derived P,N-ligand complexes of
Rh(I) have been previously described in which the
nitrogen center is found to be either bound or free,
depending on the ligand-to-metal ratio and the avail-
ability of coordination sites.32
The addition of 10 equiv of 7 did not completely
dislodge cyclooctadiene, but did give rise to an additional
Rh-monophosphite complex, as evidenced by the ap-
3
1
pearance of a new signal in the P NMR spectrum at
1
1
02.4 ppm (d, J P,Rh ) 240 Hz) in addition to the doublet
at 115.6 ppm. A mixture of CE A, rhodium, and di-
phenylsilane (1.2:1.0:12) again showed two major peaks,
the normal 115.6 ppm doublet and a new resonance at
1
1
[
28.4 ppm ( J P,Rh ) 206 Hz). These data suggest that the
+
Rh(CEA)] center is sensitive to weak donor ligands,
making it difficult to assign the structure of the active
complex. Examination of a reaction mixture (containing
Exp er im en ta l Section
excess Ph
2
SiH
2
and 7 relative to Rh-CEA) after hydro-
Gen er a l Meth od s. Ligand syntheses and catalytic reac-
tions were performed under nitrogen using an inert atmo-
silylation was complete showed COD to be liberated from
the metal and one dominant species by P NMR and two
species (major and minor) by H- N HMQC NMR. The
chemical shifts of these signals are very close to those of
3
1
sphere box or standard Schlenk techniques. CH
2 2
Cl was dried
1
15
over CaH , distilled, and stored over 4 Å molecular sieves.
2
(33) Rentsch, G. H.; Kozminski, W.; von Philipsborn, W.; Asaro, F.;
Pellize, G. Magn. Reson. Chem. 1997, 35, 904-907.
(34) Appleton, T. G.; Hall, J . R.; Ralph, S. F. Inorg. Chem. 1988,
27, 4435-4437.
(
32) Kuznetsov, V. F.; Facey, G. A.; Yap, G. P. A.; Alper, H.
Organometallics 1999, 18, 4706-4711.
J . Org. Chem, Vol. 68, No. 7, 2003 2545

Yao et al.
Tetrahydrofuran (THF), benzene, and toluene were dried and
and then was chilled to -30 °C for 20 min. The catalytic
reaction was initiated by the addition of 25 µL (0.08 mmol) of
silane solution. The reaction mixture was agitated with a
rotary shaker in the drybox at room temperature overnight
and then stopped by dilution with 900 µL of benzene.
MSEED An a lysis (Sch em e 2). A 25 µL aliquot (containing
e1.0 µmol of product) of the diluted hydrosilylation reaction
mixture was removed to a new vial and treated sequentially
with 25 µL (e5 µmol) of a CsF suspension (2.25 g, 15.0 mmol,
in 15.0 mL of THF), 120 µL (24 µmol) of a stock solution
containing an equimolar mixture of mass-tagged acids 10a and
distilled from sodium-benzophenone. Silica gel (230-400
1
31
13
mesh) was used for flash chromatography. H, P, and
C
NMR spectra were recorded at 200, 80.9, and 50.3 MHz,
1
15
respectively; two-dimensional H- N HMQC correlation NMR
1
spectroscopy was performed at 600 MHz for H. Mass spec-
trometry ee determination via electrospray ionization and
DIOS-MS measurements were obtained as previously de-
2
5-27
scribed.
Chiral HPLC analysis was performed using a
Chiralpak OD column (hexanes/2-propanol eluent mixtures,
flow rate e0.5 mL/min). Elemental analyses were performed
at Midwest Microlabs (Indianapolis, IN). All chiral diol and
amino alcohol components were obtained from commercial
sources and were used as received. The method for assembly
of ligand components was adapted from those reported for
10b (15.0 mmol each in 150 mL of CH
stock solution containing DMAP (0.12 mmol) and DIC (12.0
mmol) in 60 mL of CH Cl . The resulting reaction was kept at
2 2
Cl ), and 100 µL of a
2
2
room temperature for about 15 h. The solvent was then allowed
to evaporate, the residue was resuspended in 500 µL of
methanol, and the solid material was allowed to settle. A 0.5
µL aliquot of the supernatant was used directly for DIOS-MS
analysis, whereas an aliquot of 8 µL was diluted with methanol
to 1.0 mL for electrospray ionization MS analysis.
1
7,18,21
structures 1 and 2.
R ep r esen t a t ive P r oced u r e for Liga n d Syn t h eses
Liga n d CEA). To a flame-dried Schlenk tube containing PCl
2.0 mmol, 174 uL) in CH Cl (2.0 mL) was added Et N (580
(
3
(
2
2
3
µL, 4.1 mmol) at -40 °C under nitrogen. After the mixture
was stirred for 5 min, a solution of (-)-TADDOL (2.0 mmol,
P r oced u r e for 1-m m ol-Sca le Ca ta lytic Rea ction s. To
a flame-dried Schlenk flask containing [Rh(COD)Cl] (0.01
2
9
3.3 mg) in CH
2
Cl
2
(8 mL) was added by syringe over 10 min.
The reaction mixture was stirred at -40 to -10 °C for 1 h
mmol, 4.93 mg) in 1 mL of anhydrous toluene was added a
solution of ligand (0.03 mmol) in anhydrous toluene (1 mL)
under nitrogen. The mixture was stirred at room temperature
for 90 min, after which ketone (2.0 mmol) in 1 mL of anhydrous
toluene was added dropwise. After being stirred at room
temperature for 10 min, the mixture was cooled to -10 °C,
and then at room temperature for 2 h. To the resulting solution
was added Et
3
N (1.5 mL, 10 mmol), followed by a solution of
Cl (4 mL) at -40 °C.
(1R,2S)-(-)-N-methylephedrine in CH
2
2
The reaction mixture was allowed to warm to room temper-
ature slowly and was then stirred for 2 days. Filtration,
evaporation of solvents, and flash chromatography purification
2 2
followed by addition of a solution of Ph SiH (2.4 mmol, 445
using 25-40% EtOAc in hexanes (degassed by N
2
purging)
µL) in 2 mL of anhydrous toluene over 15 min. The stirred
reaction solution was maintained at 0 °C for 1 h and then was
allowed to stir at room temperature for 9 h. A solution of 0.6
mL of MeOH containing 1% TsOH was added dropwise, and
the reaction mixture was stirred until it appeared clear.
Evaporation of solvents and flash chromatography (20-25%
EtOAc in hexanes) afforded the product.
afforded the ligand CEA as a white power (925 mg, 69%).
1
Mp: 165-167 °C. H NMR: δ 0.42 (s, 3H), 1.09 (d, J ) 7.1
Hz, 3H), 1.10 (s, 3H), 2.26 (s, 6H), 2.74-2.80 (m, 1H), 4.99-
5
2
9
8
1
1
.09 (m, 2H), 5.58 (dd, J ) 4.8, 9.4 Hz, 1H), 7.02-7.62 (m,
3
1
1
3
5H). P NMR (CDCl
3
): δ 140.31 (s). C NMR (CDCl ): δ
3
.41, 26.80, 27.96, 42.07, 66.20, 78.07, 81.83, 83.23, 83.42,
6.55, 113.69, 127.77, 127.81, 127.83, 128.02, 128.10, 128.16,
28.38, 128.4, 128.54, 128.70, 128.77, 129.01, 129.78, 130.19,
Ack n ow led gm en t. We are grateful to The Skaggs
Institute for Chemical Biology and the National Insti-
tutes of Health (Grant RR-15066) for support of this
work; S.Y. is a Skaggs postdoctoral fellow. We also
thank Prof. Erich Uffelman for introducing us to the
30.23, 142.98, 143.51, 146.94, 147.51. Anal. Calcd for C42
44
H -
NO
5
P: C, 74.87; H, 6.58; N, 2.08. Found: C, 74.85; H, 6.60;
N, 2.11. Characterization data for other ligands appears in the
Supporting Information.
P r oced u r e for Su r vey-Sca le Ca t a lyt ic R ea ct ion s
1
15
H- N HMQC NMR technique, and Dr. Zhouxin Shen
(
Sch em e 2). In the drybox a 2 mL vial was charged with 25
µL (0.0025 mmol) of a [Rh(COD)Cl] stock solution (1.479 g,
.0 mmol, in 30 mL of benzene) and 25 µL (0.01 mmol) of a
for assistance with the MSEED-DIOS experiments.
2
3
Su p p or tin g In for m a tion Ava ila ble: Characterization
data for the ligands and details of experiments involving
additives. This information is available free of charge via the
Internet at http://pubs.acs.org.pubs.acs.org.
ligand stock solution (0.40 mmol in 1.0 mL of benzene). The
mixture was stirred for 1 h at room temperature, followed by
the addition of 25 µL (0.05 mmol) of ketone (substrate) stock
solution (7.0 mmol in 3.5 mL of benzene). The resulting
solution was allowed to stand at room temperature for 10 min
J O026365E
2
546 J . Org. Chem., Vol. 68, No. 7, 2003