Organic-fluorous phase switches: A fluorous amine scavenger for purification in solution phase parallel synthesis

Article abstract of DOI:10.1021/jo9823442

The synthesis of the fluorous amine scavenger [(C₆F₁₃CH₂CH₂)₃SSiCH ₂CH₂CH₂]₂NH and its successful application in the automated solution phase parallel synthesis of a urea library are described. Ureas were made by robotic synthesis from organic amines and excess isocyanates. The amine scavenger reacts with excess isocyanate, and the fluorous tag serves to solubilize the resulting adduct in the fluorous phase so it can be removed by fluorous-organic extraction. Organic urea products are isolated in high yields and purities after liquid - liquid extraction. Preliminary biological evaluation shows that several of the ureas have ion channel modulation abilities. In contrast to polymer and ionic quenching methods, the fluorous quench works whether the product is soluble or insoluble in the reaction medium, and ionizable functional groups are tolerated in the products.

Full text of DOI:10.1021/jo9823442

J . Org. Chem. 1999, 64, 2835-2842
2835
Or ga n ic-F lu or ou s P h a se Sw itch es: A F lu or ou s Am in e Sca ven ger
for P u r ifica tion in Solu tion P h a se P a r a llel Syn th esis
Bruno Linclau,† Ashvani K. Sing,‡ and Dennis P. Curran*,†
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 and
Department of Cell Biology and Physiology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261
Received November 30, 1998
The synthesis of the fluorous amine scavenger [(C6F13CH2CH2)3SiCH2CH2CH2]2NH and its successful
application in the automated solution phase parallel synthesis of a urea library are described. Ureas
were made by robotic synthesis from organic amines and excess isocyanates. The amine scavenger
reacts with excess isocyanate, and the fluorous tag serves to solubilize the resulting adduct in the
fluorous phase so it can be removed by fluorous-organic extraction. Organic urea products are
isolated in high yields and purities after liquid-liquid extraction. Preliminary biological evaluation
shows that several of the ureas have ion channel modulation abilities. In contrast to polymer and
ionic quenching methods, the fluorous quench works whether the product is soluble or insoluble in
the reaction medium, and ionizable functional groups are tolerated in the products.
Sch em e 1
In tr od u ction
The successful implementation of solution phase chem-
istry in parallel synthesis is hampered by time-consum-
ing workup and purification procedures. This is in
contrast with reactions on solid support, where simple
filtration effects purification in ideal cases.1 However, by
incorporating workup into the overall reaction planning,
one can design reaction components and isolation pro-
cedures that meet parallel synthesis standards in terms
of speed and efficiency. The best strategies allow the
products to be separated by simple phase-separation
methods such as filtration, extraction, and evaporation.2
Techniques that use scavenging (quenching) agents are
highly effective for rapid purification and isolation of the
desired products obtained in a solution phase reaction.3-9
The concept is based on complementary reactivity of the
components of the mixture after the reaction, and is
illustrated in Scheme 1. Combination of reactants A and
B leads to a product P and a byproduct X, which can
either be an excess of one of the reactants or a side
product. Separation of P from X is achieved by a rapid
second reaction of X with a scavenging agent 1 containing
a phase tag. The resulting tagged adduct 2 is removed
by a simple phase separation, leaving behind the pure
product P in organic solution. More broadly speaking,
scavenging is one of several useful “phase switching”
techniques where chemoselective reactions are used to
switch the phase of one (or more) products relative to
another.2
† Department of Chemistry.
‡ Department of Cell Biology and Physiology.
To date, most scavengers have been bound to poly-
mers.3-8 Scavengers attached to a solid support (Q )
insoluble polymer, Scheme 1) cause the transfer of the
captured species from the organic liquid phase to the solid
phase, where it is easily removed by filtration. This
technique combines the attractive features of solution
phase reactions with the convenience and effectiveness
of workup in solid-phase chemistry. Alternatively, a
scavenger can be linked to an ionizable functional
group7-9 (for example Q ) COOH or NR2), and the
trapped species can be removed by pH-adjusted liquid/
liquid extraction or by solid/liquid extraction.
In early 1997, we introduced a number of “fluorous
synthesis” techniques,10 among which was included a
cursory description of fluorous quenching of alkenes with
a tin hydride reagent.11 In a simple view, the “fluorous
phase” constitutes a third liquid phase which is mutually
immiscible with water and most common organic sol-
(1) (a) Thompson, L. A.; Ellman, J . A. Chem. Rev. 1996, 96, 555-
600. (b) Balkenhohl, F.; von dem Bussche-Hu¨nnefeld, C.; Lansky, A.;
Zechel, C. Angew. Chem., Int. Ed. Engl. 1996, 35, 2289-2337.
(2) (a) Curran, D. P. Angew. Chem., Int. Ed. Engl. 1998, 37, 1174-
1196. (b) Curran, D. P. Chemtracts-Org. Chem. 1996, 9, 75-87.
(3) (a) Kaldor, S. W.; Siegel, M. G. Curr. Opin. Chem. Biol. 1997, 1,
101-106. (b) Showalter, H. D. Chemtracts-Org. Chem. 1997, 10, 673-
676.
(4) (a) Kaldor, S. W.; Siegel, M. G.; Fritz, J . E.; Dressmann, B. A.;
Hahn, P. J . Tetrahedron. Lett. 1996, 37, 7193-7196. (b) Booth, R. J .;
Hodges, J . C. J . Am. Chem. Soc. 1997, 119, 4882-4886.
(5) Keating, T. A.; Armstrong, R. W. J . Am. Chem. Soc. 1996, 118,
2574-2583.
(6) (a) Flynn, D. L.; Crich, J . Z.; Devraj, R. V.; Hockerman, S. L.;
Parlow, J . J .; South, M. S.; Woodard, S. J . Am. Chem. Soc. 1997, 119,
4874-4881. (b) Parlow, J . J .; Flynn, D. L. Tetrahedron 1998, 54, 4013-
4031. (c) Flynn, D. L; Devraj, R. V.; Parlow, J . J . Curr. Opin. Drug
Disk. Dev. 1998, 1, 41.
(7) (a) Siegel, M. G.; Hahn, P. J .; Dressman, B. A.; Fritz, J . E.;
Grunwell, J . R.; Kaldor, S. W. Tetrahedron Lett. 1997, 38, 3357-3360.
(b) Suto, M. J .; Gayo-Fung, L. M.; Palanki, M. S. S.; Sullivan, R.
Tetrahedron 1998, 54, 4141-4150.
(8) (a) Parlow, J . J .; Mischke, D. A.; Woodard, S. S. J . Org. Chem.
1997, 62, 5908-5919. (b) Parlow, J . J .; Naing, W.; South, M. S.; Flynn,
D. L. Tetrahedron Lett. 1997, 38, 7959-7962.
(9) Nikam, S. S.; Kornberg, B. E.; Ault-J ustus, S. E.; Rafferty, M.
F. Tetrahedron Lett. 1998, 39, 1121-1124.
(10) Studer, A.; Hadida, S.; Ferritto, R.; Kim, S.-Y.; J eger, P.; Wipf,
P.; Curran, D. P. Science 1997, 275, 823-826.
10.1021/jo9823442 CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/26/1999

2836 J . Org. Chem., Vol. 64, No. 8, 1999
Linclau et al.
Sch em e 2
Sch em e 3
ducted to separate the organic urea product 5 from the
fluorous urea 7 and the excess fluorous amine 6.
vents.12 Typical organic compounds have little or no
solubility in fluorous solvents such as FC-72 (perfluoro-
hexanes, commercially available from 3M) or other per-
fluorinated liquids. However, organic compounds can
be rendered fluoroussthat is, induced to partition out of
an organic liquid phase and into a fluorous phasesby
attaching a suitable fluorous phase tag. In this regard,
the readily available tris(perfluorohexylethyl)silyl group
[(C6F13CH2CH2)3Si] and related groups prove valuable as
phase tags in a number of fluorous synthesis tech-
niques.13 Such tags bearing 39 fluorines effectively con-
vert relatively small (MW < 200) nonpolar organic
molecules into fluorous ones. However, the fluorous phase
is very nonpolar, and it is currently not clear what is
required to switch relatively polar organic molecules to
the fluorous phase. We decided to address this question
in the context of a fluorous amine quenching scheme.
In this paper, we report a new variant on the scaveng-
ing technique, wherein an amine scavenger is linked to
a fluorous label (Q ) fluorous tag). The separation
procedure is a fluorous organic liquid/liquid extraction.
This method complements and supplements methods
using ionizable or polymeric tags. Unlike filtration-based
methods, fluorous quenching succeeds independent of
whether the final organic products remain in solution or
precipitate out. And since fluorous quenching is not an
ionization method, there is no limit to the presence of
ionizable groups in a library. Fluorous amine quenching
is easily adaptable for automated synthesis as demon-
strated by the robotic parallel synthesis of several small
libraries of aryl ureas. The library members have been
evaluated for their effects on cAMP-stimulated transepi-
thelial Cl-secretion across a human colonic cancer cell
line T84, and a number of active compounds have been
identified.
At the outset of this work, it was not at all clear what
kinds of fluorinated amines would be needed to provide
ureas 7 that partitioned efficiently out of the organic
phase and into the fluorous phase. To address this
question, we initially decided to prepare primary amine
10 containing 39 fluorines. The synthesis of 10 is
straightforward as shown in Scheme 3. Hydrosilylation
of N-allyl trifluoroacetamide14 with the tris(perfluoro-
hexylethyl) silicon hydride [(C6F13CH2CH2)3SiH] 815 leads
to the adduct 9 in 72% yield. No solvent was used in the
hydrosilylation reaction, and the trifluoroacetyl group
was necessary to solubilize the allylamine in the fluorous
silyl hydride. (Attempts to hydrosilylate allylamine were
not very successful.) The trifluoroacetyl group also fa-
cilitates purification of 9 by column chromatography.
Deprotection of the fluorous trifluoroacetamide derivative
9 leads to the desired fluorous amine 10 in 98% yield.
Amine 10 was isolated as a clear oil and was fully
characterized by normal spectroscopic techniques.
Model experiments (not shown) for urea formation
were conducted in chloroform with benzylamine and
excess benzyl isocyanate. However, addition of the fluo-
rous amine 10 followed by fluorous-organic liquid/liquid
extraction with CHCl3 and FC-72 did not provide pure
organic product (dibenzyl urea) from the organic phase.
The organic urea was instead contaminated with the urea
resulting from trapping of benzyl isocyanate with 10 as
evidenced from the 1H NMR spectrum of the organic
product. Recombining the organic and fluorous products,
and repeating the extraction procedure with other sol-
vents such as acetonitrile or methanol, did not result in
effective separation. Similar experiments for the forma-
tion of amides and sulfonamides by quenching the excess
of acid chloride or sulfonyl chloride with fluorous amine
quenching agent 10 gave analogous organic products with
fluorous impurities. In contrast to the trapped derivatives
of 10, the residual unreacted fluorous amine 10 itself was
not detected in the organic phase of these extractions; it
efficiently partitioned into the fluorous phase.
Resu lts a n d Discu ssion
Develop m en t of a F lu or ou s Am in e Qu en ch in g
Rea gen t. Scheme 2 illustrates the general reaction and
purification plan pursued in this work. Ureas 5 were
prepared in a straightforward fashion by mixing a
limiting amount of an amine 3 with an excess of isocy-
anate 4. After the prescribed reaction time, an excess of
fluorous amine quenching reagent 6 was introduced to
convert the unreacted isocyanate 4 to a fluorous urea 7.
Fluorous-organic liquid/liquid extraction was then con-
These results suggested that the products derived from
quenching with 10 have partition coefficients that are
too low for effective separation. This was readily shown
by independently synthesizing some representative trap-
ping products (11-15) and by determining their parti-
tion coefficients (P ) cFluorous phase/cOrganic phase) between
FC-72 and several organic solvents. This was generally
done gravimetrically on larger amounts or by HPLC
analysis on smaller amounts (see Experimental Section).
The results of this series of experiments are shown in
Figure 1.
(11) See ref 18 in cited in the above paper. A full paper describing
the details of quenching with the fluorous tin hydride will be forthcom-
ing. Curran, D. P.; Kim, S.-Y.; Hadida, S., Luo, Z. J . Am. Chem. Soc.,
in press.
(12) Horvath, I. T.; Rabai, J . Science 1994, 266, 72-75. b) Cornils,
B. Angew. Chem., Int. Ed. Engl. 1997, 36, 2057-2059. c) Horvath, I.
T. Acc. Chem. Res. 1998, 31, 641-650.
(13) (a) Studer, A.; J eger, P.; Wipf, P.; Curran, D. P. J . Org. Chem.
1997, 62, 2917-2924. (b) Studer, A.; Curran, D. P. Tetrahedron 1997,
53, 6681-6696.
(14) Abd El Samii, Z. K. M.; Al Ashmawy, M. I.; Mellor, J . M. J .
Chem. Soc. Perkin Trans. 1 1988, 2517-2522.
(15) Boutevin, B.; Guida-Pietrasanta, F.; Ratsimihety, A.; Caporricio,
G.; Gornowicz, G. J . J . Fluorine Chem. 1993, 60, 211-223.

Organic-Fluorous Phase Switches
J . Org. Chem., Vol. 64, No. 8, 1999 2837
F igu r e 2. Partition coefficients of some F78 fluorous deriva-
tives.
F igu r e 1. Partition coefficients of some F39 fluorous deriva-
tives.
Sch em e 4
In contrast to nonfluorinated ureas and sulfonyl ureas,
which are essentially insoluble in FC-72, all of the
compounds have substantial solubility in FC-72. Thus,
it is possible to draw polar functionalities into a fluorous
phase by fluorination. However, none of these compounds
11-15 is highly fluorous as measured by partition
coefficient.16 For example, the fluorous benzylurea 11
partitions preferentially to organic phases such as chlo-
roform (P ) 0.27) and methanol (P ) 0.27) while with
acetonitrile, there is equal product distribution over the
two phases (P ) 1.00). Similar results were found for the
fluorous p-toluenesulfonamide derivative 12: preferential
partitioning into the organic phase was observed with
chloroform (0.73), acetonitrile (0.37), and tetrahydrofuran
(0.05). The partition coefficient in THF is especially low.
This is expected because THF has good solubilizing power
for fluorous compounds. To probe if hydrogen bonding is
a contributing factor in the partitioning, a fluorous
derivative 13 that cannot donate a hydrogen bond was
synthesized. Despite the higher molecular weight of 13
relative to 12, enhanced partitioning to the fluorous
phase is indeed observed, except with chloroform as
solvent. However, the effect is small. Similar results are
found for the fluorous p-toluoylamide derivatives 14 and
15. Thus, it appears that hydrogen bonding is not a major
factor contributing to the low partition coefficients of
these compounds.
Increasing number of fluorine atoms in the fluorous
scavenger should increase the partition coefficient of the
scavenged product toward the fluorous phase. This can
be done either by increasing the length of the existing
fluorous chains or by adding more chains. Past experience
has shown that increasing the length of the chains tends
to lead to waxy compounds with high partition coef-
ficients but low absolute solubilities.13a We opted there-
fore to increase the number of chains, and secondary
amine 16 was designed as a fluorous scavenger. Amine
16 contains 78 fluorine atoms present in six chains.
Amine 16 was again synthesized by hydrosilylation,
as shown in Scheme 4. Although N,N-diallyl trifluoro-
acetamide 2117 was fully soluble in the fluorous silyl
hydride 8, the yield of the hydrosilylation reaction leading
to 22 was much lower (33-37%) compared to the hy-
drosilylation of N-allyltrifluoro acetamide (Scheme 3).
This was due to competing reduction of the double bond,18
leading to 23 as the main product in 50% isolated yield.
The two compounds could easily be separated by column
chromatography. As expected, the desired doubly hy-
drosilylated compound 22 has a higher Rf value than the
singly hydrosilylated side product 23 (solvent ) hexanes:
benzotrifluoride (BTF)19 55:45). Removal of the trifluo-
roacetamide group in 22 with LAH occurred in 98% yield
to give 16 as a clear, colorless oil of moderate viscosity.
Derivatization of 16 with different electrophiles led to
products 17-20 (Figure 2), for which the partition
coefficients were determined in different fluorous/organic
solvent mixtures. The partition coefficients of these
compounds containing 78 fluorines are dramatically
increased relative to the compounds containing 39 fluo-
rines and are sufficiently high for effective extraction.
Even in THF, the partition coefficient lies within a useful
range for liquid-liquid extraction. As a result, we
considered 16 to be a suitable fluorous scavenger.
Not surprisingly, scavenger 16 is not very soluble in
common organic solvents. Apart from fluorous solvents,
it dissolves in benzotrifluoride (BTF). However, since
many ureas are not soluble in BTF, this solvent seemed
unsuitable for the quenching procedure. Furthermore,
BTF tends to homogenize fluorous and organic solvents
so it would have to be removed prior to the extraction.
By conducting a few trial experiments, we found that
biphasic reaction conditions can be used for the scaveng-
ing reaction with THF as the organic solvent and FC-72
as the fluorous solvent. Thus, the method outlined in
(16) Nonpolar compounds of roughly comparable fluorous/organic
ratios exhibit higher partition coefficents. See, for example: (a)
Guillevic, M. A.; Rocaboy, C.; Arif, A. M.; Horvath, I. T.; Gladysz, J . A.
Organometallics 1998, 17, 707-717. (b) Curran, D. P.; Hadida, S. J .
Am. Chem. Soc. 1996, 118, 2531-2532. (c) Curran, D. P.; Luo, Z. Med.
Chem. Res. 1998, 8, 261-265.
(18) Koroleva, G. N., Reikhsfel′d, V. O. Zh. Obshch. Khim. 1967,
37, 2768-2774, Engl. Transl. 2634-2639.
(19) Ogawa, A.; Curran, D. P. J . Org. Chem. 1997, 62, 450-451.
(17) Nugent, W. A.; Feldman, J .; Calabrese, J . C. J . Am. Chem. Soc.
1995, 117, 8992-8998.

2838 J . Org. Chem., Vol. 64, No. 8, 1999
Linclau et al.
Sch em e 5
Ta ble 1. In itia l 3 × 4 Libr a r y
Scheme 5 was developed with automated parallel syn-
thesis experiments in mind. Amine 24 and excess isocy-
anate 25 were allowed to react for 2 h in THF. Fluorous
quenching agent 16 dissolved in FC-72 was then added
to scavenge the excess isocyanate 25. The scavenging
reaction was complete after 5 h at ambient temperature.
Separation of the THF and FC-72 phases followed by two
more fluorous washes of the THF phase led to pure
organic urea 26 after evaporation of the THF (1H NMR).
Robotic Syn th esis w ith F lu or ou s Sca ven gin g. The
usefulness of the scavenging method for automated
parallel synthesis of libraries was demonstrated through
the synthesis of 25 disubstituted ureas. A Hewlett-
Packard automatic synthesizer (HP 7686) was used to
conduct all of the operations involved in the reaction and
isolation; product analyses were conducted manually. The
robot was programmed to perform the following sequence
of handlings:20 (1) preparation of a 0.5 M solution in THF
of the starting materials by transferring the required
amount of reactant into a separate vial and dilution with
THF, (2) combination of the amine and isocyanate in a
1:1.5 molar ratio followed by vortex mixing, (3) dissolu-
tion of the fluorous scavenger in FC 72 and addition of
this solution to the reaction vials followed again by
vortexing (two times, 15 min each), (4) separation of the
organic and fluorous phases, (5) extraction of the organic
phase four more times with FC-72, and finally (6)
evaporation of the organic solvent in the reaction vial.
The resulting solids were then removed from the robot
and dried in their uncapped vials in a vacuum oven at
50 °C for 12 h. All amines and isocyanates were used as
obtained from commercial sources. Some isocyanates
were filtered before use to remove solid impurities
(presumably sym-ureas derived from hydrolysis during
storage). The results of the initial 3 × 4 library experi-
ment are listed in Table 1.
urea
HPLC
entry no.
R1
R2
n
yield (%) puritya (%) %Fb
1
2
3
4
5
6
7
8
9
26a 4-MeO 4-MeO
26b 4-MeO 4-MeO
26c 4-MeO 4-CF3
26d 4-MeO 4-F-3-CF3
26e 4-CF3O 4-MeO
26f 4-CF3O 4-MeO
26g 4-CF3O 4-CF3
26h 4-CF3O 4-F-3-CF3
2
1
1
0
2
1
1
0
2
1
1
0
48
96
95
34
52
46
69
47
70
64
67
38
94.4 (0.6)
97.3 (d)
-
-
97.7 (0.2)
99.0 (1.0)
97.6 (0.4)
98.3 (1.0)
98.5 (0.3)
98.9 (d)
96.9 (1.8)
96.2 (0.8)
98.5 (0.6)
97.5 (2.5)
c
0.6
c
0.2
0.1
0.2
-
-
0.2
0.3
26i 3-Me
26j 3-Me
26k 3-Me
26l 3-Me
4-MeO
4-MeO
4-CF3
10
11
12
4-F-3-CF3
a
Waters Nova-Pack C18 3.9 × 150 mm. Flow rate 1.5 mL/min.
Detection at 210 nm, solvent: acetonitrile 40 to 60% in water. The
numbers in parentheses refer to the amount of symmetrical urea.
Mol %, determined in the 19F NMR spectrum, relative to the
b
product-CF3-group. c Not detectable in the 19F-spectrum. Sum of
d
product and symmetrical urea.
HPLC the content of symmetrical urea, which is formed
through hydrolysis of the starting isocyanate. These
amounts proved to be very low (0.2-2.5%, average 0.9%),
showing that the robot was able to make all the transfers
without introducing significant amounts of water. (In-
deed, we believe that most of the sym-urea derives from
the starting isocyanates, see below). The absence of the
starting isocyanate and the small amounts of sym-
metrical urea demonstrate that the fluorous scavenging
was successful. In a separate control experiment, tert-
butylamine (1.5 equiv) was added after the fluorous
quenching reaction, followed by another 1 h of reaction
time. The organic product, obtained after fluorous organic
extraction and evaporation of the solvent, contained less
than 1% of tert-butyl containing urea product, proving
that all the excess isocyanate was quenched by the
fluorous scavenger.21
In this first experiment, the yields of the expected
unsymmetrical ureas varied from 34 to 96% (average
61%), but the HPLC purities were excellent (94.4-99.0%,
1
average 97.5%). All products gave clean H NMR spectra
(see Supporting Information). We also measured by
(20) See Supporting Information for a detailed description of the
programs used.
(21) A control experiment showed that tert-butylamine does not react
with diarylurea product.

Organic-Fluorous Phase Switches
J . Org. Chem., Vol. 64, No. 8, 1999 2839
Ta ble 2. 3 × 3 Libr a r y w ith Op tim ized Extr a ction
The products were also analyzed for fluorous impurities
by measuring their 19F NMR spectra. The reagents in
this initial series of experiments were chosen such that
half of the ureas contained fluorine atoms, and these
served as internal standards. The signal for the terminal
CF3 of the perfluorohexyl group (δ ) -80 ppm) was
integrated to determine the mol % of fluorous impurity.
This is a sensitive analysis because there are 18 fluorines
per molecule of fluorous contaminant and because all the
CF3 groups of all possible contaminants are expected to
have the same chemical shift (In other words, the assay
provides the sum of all fluorous impurities without
providing their structures.) The amount of fluorous
derivatives was lower than 0.6 mol %,22 proving that the
fluorous-organic extraction is effective in removing fluo-
rous product from the organic phase. Indeed, the fluorous
content is so low that it escapes detection if the spec-
trometer is not tuned to perfection. We presume that the
nonfluorinated ureas have similar trace levels of fluorous
contaminants (<1%), but these are difficult to quantify
without the internal standard.
P r oced u r e
urea
HPLC
entry
no.
R1
R2
yield (%) puritya (%) %Fb
1
2
3
4
5
6
7
8
9
26m
26n
26o
26p
26q
26r
26s
26t
4-Br
4-Br
4-Br
3-Br
3-Br
3-Br
2-Br
2-Br
2-Br
4-CF3
3-CF3
2-CF3
4-CF3
3-CF3
2-CF3
4-CF3
3-CF3
2-CF3
87
80
90
79
93
85
73
96
90
98.0 (0.5)
98.5 (0.4)
98.4 (0.2)
95.7 (2.4)
96.5 (1.7)
96.8 (0.9)
92.3 (4.4)
94.0 (3.6)
93.6 (3.6)
0.5
0.3
0.4
0.3
0.3
0.3
0.2
0.2
0.3
26u
a
Waters Nova-Pack C18 3.9 × 150 mm. Flow rate 1.5 mL/min.
Detection at 210 nm, solvent: acetonitrile 40 to 60% in water. The
numbers in parentheses refer to the amount of symmetrical urea.
Mol %, determined in the 19F NMR spectrum, relative to the
b
product-CF3-group.
Although the product purities are excellent, the iso-
lated yields in these experiments are lower and more
variable than expected for simple urea formation. We
traced this problem to the robotic extractions; the “fluo-
rous waste” from the extractions contained variable but
significant amounts of THF (easily seen by the naked eye)
and the THF in turn contained organic urea. The robot
performs the extraction by drawing the lower (fluorous)
phase up through a needle. The extraction is not con-
trolled by monitoring the meniscus but by the input of a
fixed volume to be withdrawn. In these initial experi-
ments, we had set the volume of the fluorous phase to
be removed to be slightly higher (105%) than the volume
initially added. However, the results suggest that this is
too high. Control experiments showed that when mixing
pure FC-72 and THF, the volume of the THF phase
increases, indicating that more FC-72 dissolves in the
THF phase than vice versa. However, when a concen-
trated solution of fluorous urea (100 mg/0.7 mL) and pure
THF are mixed, the volume of the FC-72 phase increases,
indicating that there is a net flow of THF into the
fluorous phase. These effects are solute dependent since
after fluorous organic extraction, the amount of THF
phase in the fluorous waste vials differed significantly
in individual cases.
In two cases (entries 2, 3), a precipitated product was
clearly visible to the naked eye prior to extraction, and
the yield was very high in these cases. Apparently, the
unintentional removal of the organic phase during ex-
traction did not remove the product because it was not
dissolved. The solid was suspended in the (upper) organic
phase and did not clog the needle when the (lower)
fluorous phase was removed by the robot. That product
precipitation does not have a negative effect on the
scavenging reaction/extraction procedure is advantageous
since it is possibleseven probablesthat some members
of a structurally diverse library will not be soluble in the
reaction medium. Product precipitation cannot easily be
accommodated in polymer quenching or ion exchange
procedures since the desired insoluble product would be
removed during the filtration of the insoluble scavenging
reagents and products.
Ta ble 3. 2 × 2 Libr a r y of Ur ea s Con ta in in g Ba sic
F u n ction a lities
urea
yield (%)
HPLC
entry
no.
puritya (%)
%Fb
1
2
3
4
26v
26w
26x
26y
X ) Br
78
68
82
55
96.5 (2.4)
94.6 (2.4)
96.4 (3.6)
95.0 (2.0)
-
0.4
-
X ) OCF3
Y ) Br
Y ) OCF3
0.5
a
Waters Nova-Pack C18 3.9 × 150 mm. Flow rate 1.5 mL/min.
Detection at 210 nm, solvent: acetonitrile 40 to 60% in water. The
numbers in parentheses refer to the amount of symmetrical urea.
Mol %, determined in the 19F NMR spectrum, relative to the
b
product-CF3-group.
Table 2 shows the results of a second 3 × 3 library. In
this experiment, the settings for the fluorous extraction
procedure were adjusted to call for a lower extraction
volume, so little or no organic phase was concomitantly
removed. The yields were considerably higher with the
revised extraction volume (73-96%, average 86%) while
the excellent level of purity of the products was main-
tained. In this experiment, we can trace the sym-urea
impurity back to the starting isocyanate since the level
of this impurity is a direct function of the starting
isocyanate (compare the yields of sym-urea in entries
1-3, 4-6, and 7-9).
To demonstrate the broader applicability of the fluo-
rous quenching method compared to ionic quenching
methods, a very small library of four ureas containing
basic side chain functionalities was prepared by robot
(Table 3). Ureas 26v-y are not accessible via an acid/
base extraction workup procedure since acidic washing
to remove excess amine reagent would be unselective and
would remove the urea product as well from the organic
phase.
Biologica l Eva lu a tion of Libr a r y Com p on en ts. An
important subgoal of this work was the synthesis of ureas
that would exhibit ion channel modulating effects. Tran-
sepithelial Cl- secretion is an integrated process that
involves the polarized distribution of several membrane
(22) We assume that the major impurity is the fluorous urea, and
the mol % was calculated based on this assumption.

2840 J . Org. Chem., Vol. 64, No. 8, 1999
Linclau et al.
transporters.23 Several biophysically, pharmacologically,
and molecularly distinct types of K+ and Cl- channels,
as well as ion pumps and ion co-transporters, are thought
to be involved in Cl- secretion. Potential therapeutic uses
of Cl- secretion/absorption modulators include treatment
of respiratory disorders such as the genetic disease cystic
fibrosis (which affects one in 4000 live births worldwide),
asthma, sinusitis, rhinorrhea, and COPD (congestive
obstructive pulmonary disease).24 Defective Cl- secretion/
absorption is also shown to play a major role in gas-
trointestinal disorders such as diarrhea, several cardio-
vascular, neuronal, skeletal, and smooth muscle disorders,
and other diseases such as sickle cell anemia, ischemia,
hypertension, myotonia, and Dent’s disease.
We have evaluated the activity of aryl ureas generated
in this combinatorial library for their effects on cAMP-
stimulated transepithelial Cl- secretion across a human
colonic cancer cell line T84 in Ussing chambers using
published protocols.25 The initial screening was done with
a concentration of 100 µM of aryl ureas, and the ones
that were promising were further evaluated in a dose-
dependent manner to calculate their inhibition constant
(Ki). Several of the compounds were entirely inactive at
100 µM while others, e.g., 26q, 26e and 26g, demon-
strated Ki’s of 1.5, 2.4, and 2.5 µM, respectively, and are
therefore better than the existing Cl- secretion inhibitors.
Studies are in progress to obtain a more complete
evaluation of the biological activity of the diarylureas and
to ascertain their site and mechanism of action and will
be presented in a separate manuscript.
ably reactive electrophiles and it is premature to conclude
that amine 10 will be useful for less reactive electro-
philes.
Limitations notwithstanding, the work clearly suggests
that fluorous quenching procedures have features and
capabilities that make them convenient and attractive
for use in small scale parallel synthesis.
Exp er im en ta l Section
Gen er a l Meth od s. All glassware was dried in an oven at
120 °C prior to use. All experiments were conducted under an
atmosphere of dry nitrogen unless indicated otherwise. Sol-
vents were dried as follows: THF was distilled from sodium/
benzophenone. Dichloromethane was distilled from calcium
hydride; chloroform and benzotrifluoride were distilled from
potassium carbonate. 19F NMR spectra were recorded with
fluorotrichloromethane as an internal standard.
N -[3-[T r is (2-p e r flu o r o h e x y le t h y l)s ily l]p r o p y l]t r i-
flu or oa ceta m id e 9. tris(2-Perfluorohexylethyl)silane (1.0 g,
0.93 mmol) was mixed with N-allyltrifluoroacetamide 8 (286
mg, 1.87 mmol) in a sealed tube under nitrogen gas, and
H2PtCl6 (10.6 µL of a 10% solution in i-PrOH) was added. The
vigorously stirred mixture was heated overnight at 80 °C in
the dark, and then it was diluted with FC-72 and extracted
with dichloromethane. The fluorous phase was evaporated, and
the residue was purified by column chromatography (hexanes/
acetone 95:5 to 9:1), yielding 9 as a colorless oil (762 mg,
72%): IR (cm-1): 3310 (w), 2942 (w), 1713 (m), 1569 (w), 1452
(w), 1226 (s, br), 1082 (m), 739 (m); MS (EI) (m/e): 1223 (M+•
,
7), 1204 (10), 1154 (19), 954 (5), 876 (100), 826 (5), 548 (30).
1H NMR (300 MHz, acetone-d6): 8.47 (1H, bs), 3.37 (2H, q, J
) 6.6 Hz), 2.34 (6H, m), 1.72 (2H, m), 1.09 (6H, m), 0.93 (2H,
m) ppm; 13C NMR (75 MHz, acetone-d6): 157.9 (q, J ) 36.0
Hz), 124.5-107.8 (m), 43.5, 26.3 (t, J ) 23.3 Hz), 23.8, 8.9,
1.7 ppm; 19F NMR (471 MHz, acetone-d6): -76.5 (3F), -81.3
(9F), -116.6 (6F), -122.5 (6F), -123.4 (6F), -123.8 (6F),
-126.7 (6F) ppm.
Con clu sion s
We have developed a fluorous scavenging method for
trapping of electrophiles in a solution phase reaction with
a fluorous secondary amine. The method is based on
liquid-liquid extraction and is complementary to resin
capture and ionization methods, which are based on
filtration. The method succeeds whether the final prod-
ucts remain in solution or precipitate out, and ionizable
functional groups in the product do not interfere with the
separation. The chemical yields and the purities are good
to excellent, and the method is nicely suited for robotic
execution.
Along with these advantages come some potential
limitations. The molecular weight of the quenching agent
is quite high, and it thus has limited utility if a very large
excess of the component to be quenched is used. (How-
ever, the reactions are conducted in solution, so very large
excesses of this reagent should not generally be needed.)
The high molecular weight is caused by the large number
of fluorine atoms, which also leads to limited solubility
in organic solvents. This could cause problems with
effective quenching due to biphasic reactions. In this
work, the robotic vortexing provided a convenient solu-
tion to this problem. At this stage, we do not have a good
gauge on the absolute reactivity of amine quenching
reagent. The isocyanates used in this study are reason-
3-[Tr is(2-p er flu or oh exyleth yl)silyl]p r op yla m in e 10. To
a solution of 10 (1.1 g, 0.9 mmol) in a THF (15 mL)-MeOH
(7.5 mL)-H2O (3.75 mL) mixture was added LiOH-H2O (566
mg, 13.5 mmol). The mixture was stirred at room temperature
for 45 h. After evaporation of the organic solvents, dichlo-
romethane was added, and this was extracted with FC72 (3×).
The combined fluorous phases were dried over MgSO4 and
evaporated, yielding 10 (992 mg, 98%) as a colorless liquid:
IR (cm-1): 2935 (m), 2347 (w), 1385-1001 (s, br); MS (EI) (m/
e): 1127 (M+•, 5), 1108 (5), 858 (5), 780 (100), 433 (3). 1H NMR
(300 MHz, CDCl3): 2.73 (2H, t, J ) 6.9 Hz), 2.04 (6H, m), 1.43
(2H, m), 0.89 (6H, m), 0.70 (2H, m) ppm; 13C NMR (75 MHz,
CDCl3): 121.3-107.5 (m), 45.1, 27.3, 25.5 (t, J ) 25.6 Hz), 8.3,
1.2 ppm; 19F NMR (282 MHz, CDCl3): -81.3 (9F), -116.6 (6F),
-122.5 (6F), -123.4 (6F), -123.8 (6F), -126.7 (6F) ppm.
N,N-Bis[3-[t r is(2-p er flu or oh exylet h yl)silyl]p r op yl]-
tr iflu or oa ceta m id e 22. To a well-stirred mixture of tris(2-
perfluorohexylethyl)silane 7 (4.85 g, 4.5 mmol) and N,N-
diallyltrifluoroacetamide 21 (417 mg, 2.2 mmol) under Ar was
added H2PtCl6 (10% in iPrOH, 124 mL). The mixture was
stirred in a sealed tube for 17 h at 80 °C, shielded from light.
The crude product was purified by gradient chromatography
(BTF/hexanes 45:55 to 50:50) to yield the bissilylated product
22 (1.93 g) and the monosilylated product 23 (1.30 g, 47%),
both as colorless oils. Product 22 was further purified by HPLC
(BTF:hexanes 40:60) to yield 1.67 g (33%).
22: IR (cm-1): 2956 (m), 1700 (s), 1448 (m), 1367 (m), 1251-
1136 (s, br), 1075 (m); MS (EI) (m/e): 2333 (M+•, 10), 2314 (7),
1
2264 (6), 2236 (1), 1986 (100), 1658 (30), 1330 (17); H NMR
(23) Frizzell, R. A.; Field, M., Schultz, S. G. Am. J . Physiol. 1979,
236, F1-F8.
(24) Singh, A. K. Current Res. Ion Channel Modulators 1997, 2,
379-381.
(25) (a) Dharmsathaphorn, K.; Mandel, K. G.; Masui, H.; McRoberts,
J . A. J . Clin. Invest. 1985, 75, 462-471. (b) Mandel, K. G.; Dharm-
sathaphorn, K.; McRoberts, J . A. J . Biol. Chem. 1986, 261, 704-712.
(c) Venglarik, C. J .; Schultz, B. D.; Deroos, A. D. G.; Singh, A. K.;
Bridges, R. J . Biophys. J . 1996, 70, 2696-2703.
(300 MHz, acetone-d6): 3.48 (4H, m), 2.33 (12H, m), 1.78 (4H,
m), 1.09 (12H. m), 0.78 (4H, m) ppm; 13C NMR (125 MHz,
acetone-d6): 156.9 (q, J ) 35.6 Hz), 121.7-109.0 (m), 51.2 and
50.5, 26.0 (t, J ) 23.0 Hz), 23.6 and 21.5, 8.7 and 8.4, 1.6 ppm;
19F NMR (282 MHz, acetone-d6): -68.6 (3F), -80.6 (18F),
-115.4 (12F), -121.5 (12F), -122.4 (12F), -122.8 (12F),
-125.8 (12F) ppm.

Organic-Fluorous Phase Switches
J . Org. Chem., Vol. 64, No. 8, 1999 2841
1
23: IR (cm-1): 2985 (m), 2944 (m), 1690 (s), 1441 (m), 1361
(m), 1240-1143 (s, br), 1070 (m), 900 (m); MS (EI) (m/e): 1265
(M+•, 14), 1246 (12), 1236 (5), 1196 (25), 918 (90), 590 (15),
168 (100); 1H NMR (300 MHz, acetone-d6): 3.50-3.36 (4H, m),
2.34 (6H, m), 1.82-1.58 (4H, m), 1.10 (6H, m), 0.94-0.85 (5H,
m) ppm; 13C NMR (75 MHz, acetone-d6): 156.9 (q, J ) 27.8
Hz), 124.4-105.7 (m), 51.2 and 50.1, 50.6 and 49.3, 26.1 (t, J
) 23.3 Hz), 23.7 and 22.7, 21.6 and 20.9, 11.3 and 11.0, 8.8
and 8.5, 1.7 ppm; 19F NMR (282 MHz, acetone-d6): -68.59 and
-68.64 (3F), -80.71 (9F), -115.5 (6F), -121.4 (6F), -122.4
(6F), -122.8 (6F), -125.8 (6F) ppm.
N,N-Bis[3-[t r is(2-p er flu or oh exylet h yl)silyl]p r op yl]-
a m in e 16. To a solution of N,N-bis[3-[tris(2-perfluorohexyl-
ethyl)silyl]propyl]trifluoroacetamide 22 (1.34 g, 0.57 mmol) in
a mixture of BTF (20 mL) and THF (10 mL) was added LiAlH4
(1.0 M solution in Et2O, 2.0 mL). After stirring at room
temperature for 5 h, the mixture was quenched with saturated
Na2SO4, and the resulting mixture was stirred for 2 h. The
solids were removed by filtration, and the solvent was evapo-
rated. The resulting light yellow oil was dissolved in FC-72
and decolorized with active charcoal. Filtration and evapora-
tion yielded the fluorous amine 16 (1.26 g, 98%) as a colorless
oil: IR (cm-1): 2938 (m), 1442 (m), 1365 (m), 1269-1140 (s,
br), 1077 (m), 1018 (m); MS (EI) (m/e): 2236 (M+• - 1, 5), 2218
(5), 1890 (21), 1140 (100), 792 (15); 1H NMR (300 MHz,
acetone-d6): 2.61 (4H, t, J ) 6.7 Hz), 2.33 (12H, m), 1.58 (4H,
m), 1.05 (12H, m), 0.90 (4H, m) ppm; 13C NMR (125 MHz,
acetone-d6): 120.6-108.3 (m), 53.4, 26.1 (t, J ) 23.4 Hz), 24.6,
9.1, 1.7 ppm; 19F NMR (282 MHz, acetone-d6): -80.7 (18F),
-115.4 (12F), -121.5 (12F), -122.4 (12F), -122.7 (12F),
-125.8 (12F) ppm.
184 (100). H NMR (300 MHz, acetone-d6): 7.72 (2H, d, J )
8.3 Hz), 7.39 (2H, d, J ) 8.1 Hz), 6.41 (1H, t, J ) 6.1 Hz), 2.91
(2H, m), 2.41 (3H, s), 2.31 (6H, m), 1.59 (2H, m), 1.03 (6H, m),
0.85 (2H, m) ppm; 13C NMR (75 MHz, acetone-d6): 143.8,
139.4, 130.5, 127.9, 123.9-105.7 (m), 47.1, 26.1 (t, J ) 23.3
Hz), 24.5, 21.4, 8.8, 1.6 ppm; 19F NMR (282 MHz, acetone-d6):
-80.7 (9F), -115.4 (6F), -121.4 (6F), -122.4 (6F), -122.7 (6F),
-125.8 (6F) ppm.
14: IR (cm-1): 3275 (w), 2940 (w), 2876 (w), 1634 (s), 1549
(m), 1507 (m), 1360 (m), 1237-1144 (s, br), 1072 (m), 904 (m);
MS (EI) (m/e): 1245 (M+•, 3), 1226 (5), 1216 (1), 1204 (1), 898
(25); 1H NMR (300 MHz, acetone-d6): 7.77 (2H, d, J ) 8.2 Hz),
7.73 (1H, m), 7.23 (2H, d, J ) 8.1 Hz), 3.42 (2H, q, J ) 6.6
Hz), 2.35 (3H, s), 2.34 (6H, m), 1.72 (2H, m), 1.08 (6H, m),
0.94 (2H, m) ppm; 13C NMR (75 MHz, acetone-d6): 167.3,
142.1, 133.5, 129.7, 128.1, 124.4-105.7 (m), 43.5, 26.1 (t, J )
23.2 Hz), 24.5, 21.4, 9.0, 1.7 ppm; 19F NMR (282 MHz, acetone-
d6): -80.6 (9F), -115.3 (6F), -121.4 (6F), -122.4 (6F), -122.7
(6F), -125.7 (6F) ppm.
Gen er a l P r oced u r e for P r ep a r a tion of 13 a n d 15. The
N-methyl analogue of 10 was prepared as above, except
N-methyl-N-allyltrifluoroacetamide was used as substrate for
the hydrosilylation (71%), and LiAlH4 in THF was used for
the removal of the trifluoroacetamide group (100%). N-Me 10
was dissolved in BTF (0.06-0.1 M), followed by addition of
triethylamine (4 equiv), a solution of electrophile (1.5 equiv)
in BTF, and a catalytic amount of DMAP. The reaction was
stirred for 6-10 h. The organic phase was washed with 10%
HCl and brine and dried (MgSO4). After evaporation of the
solvent, the obtained crude mixture was purified by column
chromatography (hexanes:acetone 9:1) to yield 13 (84%) and
15 (94%).
Rep r esen ta tive Qu en ch in g Exp er im en t w ith F lu or ou s
Sca ven ger 6. To a solution of benzylamine (14.7 mL, 0.14
mmol) in acetonitrile (0.50 mL) was added a solution of benzyl
isocyanate (33.3 mL 0.27 mmol) in acetonitrile (0.50 mL) at
room temperature. The mixture was stirred for 1 h, followed
by quenching with 16 (380 mg, 0.38 mmol) in an 1.4:1
acetonitrile/benzotrifluoride mixture (1.7 mL). After stirring
for 1 h at room temperature, acetonitrile (2 mL) was added,
and the organic phase was extracted twice with FC-84 (2 × 3
mL). Evaporation of the organic phase gave dibenzylurea,
contaminated with 15% (1H NMR) of the fluorous benzylurea
(59.7 mg, theor 32.3 mg).
13: IR (cm-1): 2937 (w), 1600 (w), 1441 (w), 1346 (m), 1232-
1143 (s, br), 1069 (m). MS (EI) (m/e): 1295 (M+•, 1), 1276 (6),
1
1230 (1), 1140 (1), 1026 (2), 948 (20), 198 (100). H NMR (300
MHz, acetone-d6): 7.70 (2H, d, J ) 8.2 Hz), 7.42 (2H, d, J )
8.1 Hz), 3.01 (2H, t, J ) 7.0 Hz), 2.70 (3H, s), 2.43 (3H, s),
2.35 (6H, m), 1.67 (2H, m), 1.09 (6H, m), 0.88 (2H, m) ppm;
13C NMR (75 MHz, acetone-d6): 144.2, 136.0, 130.6, 128.4,
124.4-105.7 (m), 53.8, 35.1, 26.1 (t, J ) 23.2 Hz), 22.4, 21.4,
8.5, 1.7 ppm; 19F NMR (282 MHz, acetone-d6): -80.5 (9F),
-115.3 (6F), -121.4 (6F), -122.3 (6F), -122.7 (6F), -125.7
(6F) ppm.
Rep r esen ta tive Qu en ch in g Exp er im en t w ith F lu or ou s
Sca ven ger 16. To a solution of p-methoxybenzylamine (17.6
mL, 0.14 mmol) in THF (1 mL) was added a solution of p-tolyl
isocyanate (25.5 mL, 0.20 mmol) in THF (1 mL) at room
temperature. The mixture was stirred for 2 h, followed by
quenching with 16 (300 mg, 0.14 mmol) in FC-72 (2 mL). After
stirring for 5 h at room temperature, the phases were
separated, and the organic phase was extracted three more
times with FC 72 (3 × 2 mL). Evaporation of the organic phase
gave p-methoxybenzyl p-tolyl urea (35.7 mg, 98%).
15: IR (cm-1): 2931 (s), 2883 (m), 1640 (s), 1515 (m), 1441
(m), 1365 (m), 1315-1120 (m, br), 1070 (m), 899 (m); MS (EI)
(m/e): 1259 (M+•, 8), 1240 (6), 912 (12), 50 (100); 1H NMR (300
MHz, acetone-d6): 7.28 (2H, d, J ) 7.8 Hz), 7.21 (2H, d, J )
7.9 Hz), 3.50 and 3.33 (2H, bs), 2.97 (3H, s), 2.35 (9H, m), 1.74
(2H, bs), 1.08 (6H, bs), 0.92 and 0.69 (2H, bs) ppm; 13C NMR
(125 MHz, acetone-d6): 171.9 and 171.4, 139.9 and 139.8, 135.8
and 135.6, 129.6, 128 and 127.7, 121.1-109.3 (m), 54.6 and
50.8, 37.7 and 32.7, 26.0 (t, J ) 22.8 Hz), 22.9 and 21.6, 21.3,
8.6 and 8.4, 1.6 ppm; 19F NMR (282 MHz, acetone-d6): -80.6
(9F), -115.3 (6F), -121.4 (6F), -122.4 (6F), -122.7 (6F),
-125.7 (6F) ppm.
Gen er a l P r oced u r e for Der iva tiza tion of F lu or ou s
Sca ven ger 10 Lea d in g to 11, 12, a n d 14. To a solution of 6
in chloroform/BTF (0.05-0.08 M) was added triethylamine (4
equiv) and a solution of electrophile (1.2-1.5 equiv) in chlo-
roform. After 3 h, the mixture was extracted with water or 1
M HCl. After drying (MgSO4) and evaporation of the solvent,
the crude mixture was purified by column chromatography
(hexanes/acetone 9:1) to obtain 11, 12, and 14 in 55-75% yield.
11: IR (cm-1): 3324 (m, br), 2929 (m), 1633 (s), 1579 (s),
1454 (m), 1360 (m), 1317-1120 (s, br), 1068 (s), 902 (m); MS
(EI) (m/e): 1260 (M+•, 31), 1241 (7), 913 (23), 823 (25), 780
(15), 150 (100); 1H NMR (300 MHz, acetone-d6): 7.24 (5H, m),
5.82 (1H, m), 5.62 (1H, m), 4.34 (2H, d, J ) 6.0 Hz), 3.16 (2H,
m), 2.32 (6H, m), 1.58 (2H, m), 1.07 (6H, m), 0.85 (2H, m) ppm;
13C NMR (75 MHz, acetone-d6): 159.2, 142.2, 129.2, 128.2,
127.6, 123.9-106.2 (m), 44.5, 44.0, 26.1 (t, J ) 23.0 Hz), 25.4,
8.9, 1.7 ppm; 19F NMR (282 MHz, acetone-d6): -80.6 (9F),
-115.4 (6F), -121.4 (6F), -122.4 (6F), -122.7 (6F), -125.7
(6F) ppm.
Gen er a l P r oced u r e for Der iva tiza tion of F lu or ou s
Sca ven ger 16 Lea d in g to 17-20. To a solution of 16 in BTF,
or a BTF-THF mixture (0.02 M), was added electrophile (neat
or dissolved in THF or BTF, 2-5 equiv). Triethylamine (4-6
equiv) and DMAP (cat) were added, for 17 and 18. After 4-6
h, water or 5% NaHCO3 was added, and the mixture was
extracted with FC-72. The phase separation was generally
difficult. The fluorous phase was washed with brine and after
evaporation was purified by column chromatography (BTF/
hexanes 1:1; for 20: hexanes/acetone 9:1) to obtain 17-20 in
58-89% yield.
17: IR (cm-1): 2933 (m, br), 1637 (s), 1425 (m), 1362 (s),
1321-1120 (s, br), 1070 (s), 903 (m); MS (EI) (m/e): 2355 (M+•
,
12), 2336 (7), 2008 (15), 1680 (7), 1352 (6), 1286 (32), 898 (36),
792 (48), 145 (100); 1H NMR (500 MHz, acetone-d6): 7.25 (2H,
d, J ) 7.9 Hz), 7.19 (2H, d, J ) 7.9 Hz), 3.50 and 3.32 (4H,
bs), 2.33 (15H, m), 1.78 and 1.68 (4H, bs), 1.06 (12H, m), 0.90
and 0.65 (4H, bs) ppm; 13C NMR (125 MHz, acetone-d6): 171.0,
138.7, 135.1, 128.6, 126.5, 120.7-108.0 (m), 51.6 and 47.1, 25.0
(t, J ) 23.1 Hz), 22.4 and 21.1, 20.2, 7.6, 0.6 ppm; 19F NMR
12: mp 47 °C; IR (cm-1): 3282 (m, br), 2941 (m), 2874 (m),
1600 (w), 1441 (m), 1320 (m), 1238-1145 (s, br), 1072 (m); MS
(EI) (m/e): 1262 (M+• - 19, 4), 1126 (1), 1012 (1), 934 (35),

2842 J . Org. Chem., Vol. 64, No. 8, 1999
Linclau et al.
(282 MHz, acetone-d6): -80.6 (18F), -115.4 (6F), -121.4 (6F),
-122.4 (6F), -122.7 (6F), -125.8 (6F) ppm.
partition coefficient was determined as the ratio of the amount
of product that was dissolved in each phase.
18: IR (cm-1): 2943 (w), 1634 (m), 1442 (m), 1360 (m),
1235-1130 (s, br), 1070 (m), 904 (m); MS (FAB) (m/e): 2418
([M + H]+, 24), 2195 (5), 862 (20), 577 (80), 488 (100); 1H NMR
(500 MHz, acetone-d6): 7.67 (4H, m), 7.46 (4H, m), 7.39 (1H,
m), 3.53 and 3.40 (4H, bs), 2.32 (12H, m), 1.81 and 1.73 (4H,
bs), 1.07 (12H, m), 0.92 and 0.69 (4H, bs) ppm; 13C NMR (75
MHz, acetone-d6): 170.8, 141.6, 140.2, 136.9, 129.0, 127.8,
127.3, 126.8, 123.0-105.3 (m), 51.8 and 47.2, 25.2 (t, J ) 23.1
Hz), 22.6 and 21.3, 8.0, 0.7 ppm; 19F NMR (282 MHz, acetone-
d6): -80.6 (18F), -115.3 (12F), -121.4 (12F), -122.4 (12F),
-122.7 (12F), -125.8 (12F) ppm.
(B) By HPLC: a mixture of the fluorous compound (1-10
mg) in FC-72 (1 mL) and organic solvent (1 mL) was vigorously
stirred in a vial. The phases were allowed to settle, and the
amount of product present in each phase was determined by
analytical HPLC (column: NEOS Fluofix, flow rate: 1 mL/min,
solvent: THF, injection volume: 10-50 µL). The partition
coefficient was determined by the ratio of the peak areas. The
injection volume was chosen so that the ratios of the obtained
peak areas did not exceed 15:1. In some cases, the fluorous
phase had to be diluted.
19: mp 42 °C; IR (cm-1): 2947 (w), 2872 (w), 1644 (m), 1593
(w), 1520 (m), 1361 (m), 1251-1145 (s, br), 1067 (m), 897 (m);
MS (EI) (m/e): 2370 (M+•, 12), 2351 (3), 2236 (15), 2218 (15),
2023 (6), 1968 (11), 1890 (62), 1562 (10), 1140 (100), 792 (65);
1H NMR (300 MHz, acetone-d6): 7.46 (1H, s), 7.34 (2H, d, J )
8.4 Hz), 6.99 (2H, d, J ) 8.3 Hz), 3.41 (4H, t, J ) 7.3 Hz), 2.37
(12H, m), 2.23 (3H, s), 1.76 (4H, m), 1.09 (12H, m), 0.87 (4H,
m) ppm; 13C NMR (125 MHz, acetone-d6): 155.8, 139.0, 131.9,
129.5, 120.9, 121.7-109.1 (m), 50.9, 26.1 (J ) 23.1 Hz), 23.3,
20.7, 8.8, 1.7 ppm; 19F NMR (282 MHz, acetone-d6): -80.7
(18F), -115.5 (12F), -121.5 (12F), -122.5 (12F), -122.7 (12F),
-125.8 (12F) ppm.
Ack n ow led gm en t. Bruno Linclau is indebted to the
Belgian American Educational Foundation (BAEF) for
a fellowship. We thank Hewlett-Packard for the loan of
the HP 7686 synthesizer, and Drs. Paul Hoeprich,
William Leber, and Charles Knipe for their helpful
advice. We are thankful to Dr. Kasi V. Somayajula and
Vyacheslow N. Fishman for the mass spectral data. We
acknowledge Prof. Robert J . Bridges for all the helpful
discussions regarding biological evaluation of the com-
binatorial libraries, the technical assistance of Mai-
trayee Sahu in tissue culture, and Matthew Green in
Ussing chamber experiments. This work was supported
by the Drug Chemistry Core (A.K.S.) funded by the
Cystic Fibrosis Foundation Research Development Pro-
gram (RDP) FRIZZE 97R0 and by the National Insti-
tutes of Health (D.P.C.). We thank OxyChem for pro-
viding BTF and Elf Atochem for perfluoroaklyl iodides.
20: IR (cm-1): 3291 (w, br), 2933 (w, br), 1686 (s), 1588 (s),
1545 (s), 1476 (s), 1463 (s), 1443 (s), 1426 (s), 1364-1078 (s,
br), 904 (s); MS (FAB) (m/e): 2435 ([M + H]+, 90), 2363 (5),
1
2238 (10), 2166 (3), 2087 (11); 1140 (23), 368 (100) H NMR
(500 MHz, acetone-d6): 9.07 (1H, s), 7.85 (2H, d, J ) 8.3 Hz),
7.33 (2H, d, J ) 8.2 Hz), 3.30 (4H, bs), 2.40 (3H, s), 2.29 (12H,
m), 1.65 (4H, bs), 1.05 (12H, m), 0.77 (4H, bs) ppm; 13C NMR
(125 MHz, acetone-d6): 152.3, 144.5, 139.1, 129.8, 129.2,
123.1-109.1 (m), 50.31, 25.97 (t, J ) 22.5 Hz), 22.8, 21.4, 8.7,
1.6 ppm; 19F NMR (282 MHz, acetone-d6): -80.6 (18F), -115.4
(12F), -121.4 (12F), -122.4 (12F), -122.7 (12F), -125.8 (12F)
ppm.
Deter m in a tion of th e P a r tition Coefficien ts. (A) Gravi-
metric: a mixture of the fluorous compound (80-150 mg) in
FC-72 (50 mL) and organic solvent (50 mL) was vigorously
shaken in a separating funnel. The phases were allowed to
settle and were separated. The solvent was removed and the
Su p p or tin g In for m a tion Ava ila ble: Contains a descrip-
tion of the programming sequence used for the robotic experi-
ments, HRMS data for the urea products, and copies of the
1H NMR spectra of the fluorous amine 16 and its precursor
22 and the crude urea products from the library experiments.
This material is available free of charge via the Internet at
http://pubs.acs.org.
J O9823442