Fluorous triphasic reactions: Transportative deprotection of fluorous silyl ethers with concomitant purification [11]

Full text of DOI:10.1021/ja011716c

J. Am. Chem. Soc. 2001, 123, 10119-10120
Fluorous Triphasic Reactions: Transportative
10119
Deprotection of Fluorous Silyl Ethers with
Concomitant Purification
Hiroyuki Nakamura,1a Bruno Linclau,1b and Dennis P. Curran*
Department of Chemistry
UniVersity of Pittsburgh
Pittsburgh, PennsylVania 15260
ReceiVed July 16, 2001
The use of an impure starting material in a typical organic
reaction is generally a recipe for producing an impure product.
Accordingly, each organic reaction is usually followed by a
separation. Despite the efficiency advantage, reactions and
separations are rarely coupled in any substantive way.2 Herein,
we describe the first examples of “fluorous triphasic reactions”.
In these processes, a liquid-liquid separation is directly coupled
with a chemical reaction to provide a pure product starting from
an impure precursor. The reaction and separation occur simulta-
neously rather than sequentially, and the reaction drives the
separation.
Fluorous triphasic reactions emanate from fluorous biphasic
reactions3 and other techniques based on fluorous/organic liquid-
liquid extractions.4 They also extend so-called “fluorous synthesis”
techniques in which organic molecules are reversibly tagged with
fluorous tags during a part of the synthesis.5 These techniques
typically rely on solid-liquid extraction because the number of
fluorines needed for liquid-liquid extractions is unduly large.6
The transportation features of three liquid phases (usually two
aqueous, one organic) are well-known7 but are rarely used in
synthesis.
Figure 1. Cartoon of Fluorous Triphasic Transportative Detagging
autopurifying process in which an impure starting material reacts
to generate a pure product.
We conducted a series of preliminary experiments with pure
fluorous silyl ethers to identify suitable fluorous tags, reagents,
and solvents for the triphasic reaction (eq 1 and Table 1). The
F-phase in all cases was FC-72 (perfluorohexanes), and aceto-
nitrile was chosen as the standard solvent for the S-phase. P-phase
solvents and reagents were initially probed with silyl ether 1a of
2-(2-naphthyl)ethanol (Rf ) C8F17). This was added to the S-phase
(acetonitrile), and a reagent was added to the P-phase in the
indicated solvent. After completion, the reactions were worked
up to determine the total yield of product from the amount isolated
from the S- and P-phases. In the ideal experiment, all the product
should be in the P-phase.
The present fluorous triphasic reaction is used to remove a
fluorous tag from a precursor with concomitant separation of the
detagged organic product from organic impurities, as shown in
Figure 1. A simple U-tube holds a lower fluorous phase (also
called the “F-phase”) that serves as a barrier to separate two upper
organic phases. The substrate is added to one organic side (the
“S-phase”), and the product is formed in the other (“the P-phase”).
In the generalized reaction in Figure 1, a fluorous-tagged precursor
contaminated with nontagged (organic) impurities is added to the
S-phase, and a reagent to remove the fluorous tag is added to the
P-phase. The reagent should have negligible solubility in the
F-phase to prevent its transport to the S-phase.
Over the course of the reaction, the tagged substrate migrates
from the S-phase through the F-phase to the P-phase, whereupon
the tag is promptly removed by the reagent. The detagged product
is now stranded in the P-phase, because it lacks the tag that
rendered it soluble in the F-phase. The impurities have no tag in
the first place, so they cannot migrate from the S-phase to the
P-phase. The residual tag has a relatively high fluorine content
and partitions back to the F-phase. The result is a kind of
We first examined different cleavage reagents with 95%
MeOH/H2O as the P-phase solvent. In initial experiments (entries
1-6), only the F-phase was stirred. Reasonable yields and P-phase
selectivities of 2-(2-naphthylethanol) 2 were obtained with AcOH
and CsF (entries 1,2), but the best results were obtained by using
H2SO4 or H2SiF6 in aqueous MeOH. Alcohol 2 was observed
only in the P-phase in high yields (entries 3 and 4). Various
solvents were examined with H2SiF6, and MeOH and DMF were
found to be more effective than MeCN (entries 4-6). The reaction
was accelerated when each phase was stirred during the reaction
process (see the modified U-tube reactor in Figure 2 of the
Supporting Information). The desilylation reactions were com-
pleted in 18-20 h (rather than 2-4 days) with H2SO4 or H2SiF6,
and 2 was obtained only in the P-phase (entries 3 and 4 vs 7 and
8). This apparatus was adopted for all subsequent experiments.
Next, the effect of the partition coefficient (Kp) values on the
reaction was examined by using silyl groups of differing fluorine
content (entries 9-12). The reaction of 1b (Rf ) C10F21) with a
Kp of 2.7 required 6 days to give 2 in quantitative yield (entry
9). However, the long reaction time resulted in back transport,8
and the final product was distributed in a ratio of 84/16 in the P-
(1) Current addresses: (a) Department of Chemistry, Tohoku University,
Graduate School of Science, Sendai 980-8578, Japan. (b) University of
Southampton, Chemistry Department, Highfield, Southampton SO17 1BJ, U.K.
(2) In contrast, shifting equilibrium by simultaneous removal of small
molecule byproducts such as water is common.
(3) (a) Horva´th, I. T.; Ra´bai, J. Science 1994, 266, 72. (b) Horva´th, I. T.
Acc. Chem. Res. 1998, 31, 641.
(4) (a) Curran, D. P. Angew. Chem., Int. Ed. Engl. 1998, 37, 1175. (b)
Curran, D. P. In Stimulating Concepts in Chemistry; Stoddard, F., Reinhoudt,
D., Shibasaki, M., Eds.; Wiley-VCH: New York, 2000; p 25. (c) Kitazume,
T. J. Fluorine Chem. 2000, 105, 265. (d) Furin, G. G. Russ. Chem. ReV. 2000,
69, 491.
(5) Studer, A.; Hadida, S.; Ferritto, R.; Kim, S. Y.; Jeger, P.; Wipf, P.;
Curran, D. P. Science 1997, 275, 823.
(6) (a) Curran, D. P. Synlett, 2001, in press. (b) Curran, D. P.; Luo, Z. Y.
J. Am. Chem. Soc. 1999, 121, 9069. (c) Luo, Z. Y.; Zhang, Q. S.; Oderaotoshi,
Y.; Curran, D. P. Science 2001, 291, 1766.
(7) (a) Newcomb, M.; Toner, J. L.; Helgeson, R. C.; Cram, J. D. J. Am.
Chem. Soc. 1979, 101, 4941. (b) For triphasic reactions with one liquid and
two solid phases, see: Rebek, J. Tetrahedron 1979, 35, 723.
10.1021/ja011716c CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/20/2001

10120 J. Am. Chem. Soc., Vol. 123, No. 41, 2001
Communications to the Editor
Table 1. Deprotection of the Fluorinated Silyl Ethers Using the
Triphasic Reaction System in Eq 1
to give (S)-4, 6, 8, and 14 in 87-95% yields with high P-phase
selectivities (entries 13-15,19). The Kp values of (S)-3, 5, 7, and
13b were 0.72-1.9. However, the reactions of the aliphatic silyl
ethers 9, 11, and 13a, whose Kp values were 5.0-8.2, needed
longer reaction times (7 days or more), which resulted in
decreased selectivities (entries 16-18).9 The fluorous silyl ether
15 of the natural product mappicine6c underwent desilylation at
40 °C to give free mappicine 16 in 65% yield, although the
reaction did not go to completion in 7 days because of the low
Kp (0.11) of 15 (entry 20). Mappicine (16) is highly insoluble in
FC-72, so back transport was not observed.
To show that triphasic reactions can be used to purify products
derived from impure reactants, we examined the purificative
deprotection of silyl ethers doped with impurities (eq 2).
Fluorinated compound 1a was mixed with various amounts of
(rac)-1-(2-naphthyl)ethanol, (rac)-4. This mixture was added to
the S-phase and subjected to the usual triphasic reaction condi-
tions. The corresponding alcohol 2 was obtained free of 1-(2-
naphthyl)ethanol, (rac)-4 (entries 1-4, in Table 2).
product
ratio
yield (P-/S-
time product (%) phase)
sub-
solvent
(R-phase)
strate Kpa reagentb
1
2
3
4
5
6
7
8
9
1a
1a
1a
1a
1a
1a
1a
1a
1b
0.92 AcOH MeOH-H2Oc 6 dh
CsF
MeOH-H2Oc 4 dh
2
2
2
2
2
2
2
2
2
2
2
2
54d
96/4
80e >99/1
92 >99/1
96 >99/1
89 >99/1
H2SO4 MeOH-H2Oc 4 dh
H2SiF6 MeOH
H2SiF6 DMF
H2SiF6 MeCN
2 dh
1 di
1 di
99
91/9
H2SO4 MeOH-H2Oc 18 hi
97 >99/1
92 >99/1
>99
90
96
97
87
H2SiF6 MeOH
H2SiF6 MeOH
0.39 H2SiF6 MeOH
0.12 H2SiF6 MeOH
0.015 H2SiF6 MeOH
H2SiF6 MeOH
H2SiF6 MeOH
0.72 H2SiF6 MeOH
20 hi
6 di
2.7
84/16
86/14
67/33
59/41
98/2
10 1c
11 1d
12 1e
1.5 di
3 di
7 di
13 (S)-3 1.5
2 di
(S)-4
6
8
10
12
14
14
15
16
5
7
9
1.3
1.5 di
3 di
90 >99/1
95
91
95/5
84/16
8.2
5.0
H2SiF6 MeOH
H2SiF6 MeOH
H2SiF6 MeOH
H2SiF6 MeOH
7 di
17 11
18 13a 5.7
19 13b 1.9
7 di
38f 74/26
7 di
>99
93
91/9
96/4
1.5 di 14
20 15
0.11 H2SO4 MeOH-H2Oc 7 di
16
65g >99/1
Purificative deprotections of the chiral silyl ethers ((S)-3 and
(R)-17) were examined in the presence of the enantiomerically
pure mirror image alcohols using the triphasic reaction system.
The deprotection reaction of (S)-3 (1.0 equiv) proceeded in the
presence of (R)-(+)-1-(2-naphthyl)ethanol, (R)-4 (1.0 equiv), and
1-(2-naphthyl)ethanol, 4, was obtained in 83% total yield in a
ratio with 41/59 in the P-/S-phases (entry 5). The ee values of
1-(2-naphthyl)ethanol obtained were >97% (S) and 90% (R) in
the P- and S-phases, respectively. The chiral silyl ether (R)-17
(1.0 equiv) also underwent the purificative deprotection in the
presence of (S)-(-)-2-phenylpropanol (1.0 equiv), (S)-18, and
2-phenylpropyl benzoate was obtained in 76% total yield in a
39/61 ratio in P-/S-phases after protection of 2-phenylpropanol
with benzoyl chloride. The ee values of 2-phenylpropyl benzoate
were 89% (R) and 87% (S) in the P- and S-phases, respectively.
These triphasic reactions use the fluorous tag to allow
transportation of the impure tagged substrate from the S-phase
through the F-phase to the P-phase where it is detagged to provide
the final pure product. The detagging reaction drives the non-
equilibrium transport of the product to the P-phase in this intimate
coupling of reaction and separation. This class of fluorous triphasic
reactions is ideal for use in final detagging of a fluorous substrate
in a fluorous synthesis exercise.5 In addition, enantioselective
fluorous tagging of racemic mixtures10 followed by triphasic
detagging provides a new option for resolution of racemates.
Many other applications of triphasic reactions can be envisioned.
For example, the use of a fluorous reagent or catalyst in the
F-phase will allow the transportative coupling of two organic
reactants. Finally, the ability to couple triphasic reactions in
sequence may ultimately allow for real-time, multistep synthesis/
separation processes.
a The KP values of substrates were measured between FC-72 and
MeOH. b The amount of reagents used is as follows: HCl (2 equiv) in
entry 1, AcOH (35 equiv) in entry 2, CsF (3 equiv) in entry 3, H2SO4
(1 equiv) in entries 4 and 8, H2SiF6 (2 equiv) in entries 5-7 and 9-20.
c The ratio of MeOH/H2O is 20/1. d Compound 1a was recovered in
36% and 8% yields from the S- and F-phases, respectively. e Compound
1a was recovered in ∼10% yield from the F-phase. f The reaction was
not complete in 7 days, and 11 was recovered in 55% yield from the
F-phase. g The reaction was carried out at 40 °C. The reaction was not
complete in 7 days, and 15 was recovered in 35% yield from the
S-phase. h Stirring of F-phase only. i All three phases stirred; see
apparatus in Supporting Information.
Table 2. Purificative Deprotection from the Mixture of Fluorinated
and Nonfluorinated Compounds Using the Triphasic Reaction
System in Eq 2
ratio of
tagged
substrate
nonfluorinated
yield
(%)
product
(P-/S-phases)
contaminant (equiv)a
time
1 d
1
2
3
4
5
6
1a
1-(2-naphthyl)ethanol
(0.2)
2, 87
99/1
1a
1-(2-naphthyl)ethanol
(0.4)
1-(2-naphthyl)ethanol
(0.6)
1-(2-naphthyl)ethanol
(1.0)
1.5 d 2, 96
3 d 2, 97
2.5 d 2, 96
96/4
95/5
94/6
1a
1a
(S)-3
(R)-(+)-1-(2-naphthyl)-
ethanol (1.0)
2 d
4, 83b 41/59 (>97%ee/
90%ee)c
(R)-17 (S)-(-)-2-phenylpropanol 2 d
18, 76b,d 39/61 (89%ee/
87%ee)c,d
(1.0)
a The equivalents were based on the tagged substrates. b The yields
were based on the total amount of both enantiomers in the reaction.
c The enantiomeric purity of the corresponding alcohols obtained in
each phase is shown in parentheses. The ee values were determined
by optical rotation. d The yields and ee values were determined after
conversion of 2-phenylpropanol into the corresponding benzoate.
Acknowledgment. We thank the National Institutes of Health for
funding this work. We are grateful for the experimental contributions of
Dr. Shigeru Wada, who initially developed the fluorous silyl groups.
Supporting Information Available: Detailed experimental proce-
dures for all the data in the Tables along with transport rates, partition
coefficients, and a photograph of the apparatus (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
and S-phases. Reactions of 1c,d, which contain fewer fluorines
(C6F13, C4F9) with lower Kp values (0.39, 0.12), also required
longer reaction times; this caused an increase in the back transport
of product 2 (entries 10,11). The control substrate 1e bearing no
fluorines (iC3H7) gave especially poor results (entry 12). These
results suggested that the optimal Kp for triphasic deprotection
reactions of small organic molecules in U-tubes is roughly 1.
To assay generality, fluorinated silyl ethers derived from
various alcohols were examined. The aromatic silyl ethers (S)-3,
5, 7, and 13b underwent the triphasic deprotection very smoothly
JA011716C
(8) See Supporting Information for control experiments measuring transport
rates of nonfluorous compounds.
(9) The reaction of 11 was not completed even after 2 days under the
ordinary monophasic conditions, whereas the reaction of 1 was completed in
30 min under the same conditions.
(10) Hungerhoff, B.; Sonnenschein, H.; Theil, F. Angew. Chem., Int. Ed.
2001, 40, 2492.