Second-generation tags for fluorous chemistry exemplified with a new fluorous Mitsunobu reagent

Article abstract of DOI:10.1021/ol800750q

(Chemical Equation Presented) A new fluorous DEAD reagent bearing two perfluoro-tert-butyloxy groups with propylene spacers shows excellent promise for use in fluorous Mitsunobu reactions. Pure target products were obtained in good yields after removing fluorous byproducts by FSPE. The new reagent serves as a prototype for a greener second generation of fluorous reagents bearing tags that are not expected to degrade in the environment to compounds that are highly persistent or that bioaccumulate in higher organisms.

Full text of DOI:10.1021/ol800750q

ORGANIC
LETTERS
2008
Vol. 10, No. 12
2453-2456
Second-Generation Tags for Fluorous
Chemistry Exemplified with a New
Fluorous Mitsunobu Reagent†
Qianli Chu, Christopher Henry, and Dennis P. Curran*
Department of Chemistry, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15260
curran@pitt.edu
Received April 2, 2008
ABSTRACT
A new fluorous DEAD reagent bearing two perfluoro-tert-butyloxy groups with propylene spacers shows excellent promise for use in fluorous
Mitsunobu reactions. Pure target products were obtained in good yields after removing fluorous byproducts by FSPE. The new reagent serves
as a prototype for a greener second generation of fluorous reagents bearing tags that are not expected to degrade in the environment to
compounds that are highly persistent or that bioaccumulate in higher organisms.
Fluorous techniques for synthesis and separation are finding
increasing use in small molecule synthesis, large molecule
synthesis, proteomics, and microarraying, among other
applications.1,2 Most techniques are loosely classed as “light”
or “heavy” depending on the number and size of fluorous
tags.3 Heavy fluorous molecules typically have 39 or more
fluorine atoms, and separations usually involve either fluorous
liquids or precipitation. Light fluorous molecules usually have
17 fluorines or fewer and are separated with the aid of a
fluorous stationary phase (often fluorous silica gel). Despite
the diversity of fluorous reaction and separation methods,
almost all fluorous molecules are fashioned from the same
two linear perfluoroalkyl groups, perfluorohexyl (C6F13) and
perfluorooctyl (C8F17), along with a spacer. The combination
of the perfluoroalkyl group and the spacer is called a fluorous
tag or ponytail.4
The initial vision for fluorous chemistry of Horva´th and Ra´bai
was as a green method for large-scale chemical production,5
but much of the subsequent development of the field has been
in the direction of small-scale chemical discovery. The original
vision still holds promise, but a significant problem has
arisen—it is now widely recognized that perfluorocarbons and
associated compounds that have no carbon-hydrogen bonds
are environmentally persistent because of their exceptional
chemical stability.6 Solvents such as FC-72 (perfluorohexanes)
can potentially end up in the atmosphere, where they have
exceptionally long half-lives (tens of thousands of years) and
significant global warming potential.7 Alternative hydrofluo-
roether (HFE) solvents8a are much more readily degraded in
the environment than perfluorocarbons and can even outperform
† Dedicated to Professor Dr. Daniel Bellusˇ on the occasion of his 70th
birthday.
(1) Handbook of Fluorous Chemistry; Gladysz, J. A., Curran, D. P.,
Horva´th, I. T., Eds.; Wiley-VCH: Weinheim, Germany, 2004
.
(2) (a) Zhang, W.; Curran, D. P. Tetrahedron 2006, 62, 11837–11865.
(b) Pearson, W. H.; Berry, D. A.; Stoy, P.; Jung, K. Y.; Sercel, A. D. J.
Org. Chem. 2005, 70, 7114–7122. (c) Brittain, S. M.; Ficarro, S. B.; Broack,
S.; Peters, E. C. Nat. Biotechnol. 2005, 23, 463–468. (d) Ko, K. S.; Jaipuri,
F. A.; Pohl, N. L. J. Am. Chem. Soc. 2005, 127, 13162–13163. (e) Vegas,
A. J.; Bradner, J. E.; Tang, W.; McPherson, O. M.; Greenberg, E. F.;
Koehler, A. N.; Schreiber, S. L. Angew. Chem., Int. Ed. 2007, 46, 7960–
7964
.
(3) (a) Gladysz, J. A.; Correa de Costa, R. In The Handbook of Fluorous
Chemistry; Gladysz, J. A., Curran, D. P., Horva´th, I. T., Eds.; Wiley-VCH:
Weinheim, Germany, 2004; pp 24-40. (b) Curran, D. P. In The Handbook
of Fluorous Chemistry; Gladysz, J. A., Curran, D. P., Horva´th, I. T., Eds.;
Wiley-VCH: Weinheim, Germany, 2004; pp 128-156.
(4) Gladysz, J. A. In The Handbook of Fluorous Chemistry; Gladysz,
J. A., Curran, D. P., Horva´th, I. T., Eds.; Wiley-VCH: Weinheim, Germany,
2004; pp 41-55.
(5) Horva´th, I. T.; Ra´bai, J. Science 1994, 266, 72–75.
(6) (a) Tavener, S. J.; Clark, J. H. J. Fluorine Chem. 2003, 123, 31–36.
(b) Sheldon, R. A. Green Chem. 2005, 7, 267–278.
(8) (a) Costello, M. G.; Flynn, R. M.; Owens, J. G. In Kirk-Othmer
Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York,
2004; pp 877-888. (b) Ryu, I.; Matsubara, H.; Emnet, C.; Gladysz, J. A.;
Takeuchi, S.; Nakamura, Y.; Curran, D. P. Green Reaction Media in Organic
Synthesis; Blackwell: Ames, IA, 2005; pp 59-124. (c) Chu, Q.; Yu, M. S.;
Curran, D. P. Tetrahedron 2007, 63, 9890–9895.
(7) Ellis, D. A.; Cahill, T. M.; Mabury, S. A.; Cousins, I. T.; Mackay,
D. In Organofluorines; Neilson, A. H., Ed.; Springer-Verlag: Berlin, 2002;
Vol. 3, pp 63-83.
10.1021/ol800750q CCC: $40.75
Published on Web 05/17/2008
2008 American Chemical Society

them in some applications.8b,c
with a standard gradient starting from 80% aqueous aceto-
nitrile, increasing to 100% acetonitrile over 30 min. The
retention times of the esters are shown in Figure 1.
At first blush, the persistence problem seems limited to
fluorous solvents because fluorous reaction components
inevitably have a large organic domain that is subject to
degradation upon disposal. The concern is that if this
degradation occurs in an oxidative environment, then the
residual fluorous domain will potentially end up as a
perfluorocarboxylic acid.9 The degraded acids resulting from
the current generation fluorous tags, perfluoroheptanoic acid
(C6F17CO2H) and perfluorononanoic acid (C8F17CO2H),
bracket the very well-known perfluorooctanoic acid
(PFOA).10 This and related compounds (for example, per-
fluorooctane sulfonic acid, PFOS) are persistent in the
environment and bioaccumulate in higher organisms.
These environmental concerns can potentially be addressed
by using smaller perfluoroalkyl groups, and many materials
applications now focus on perfluorobutyl groups.11 The
perfluoro-tert-butyl group ranks very high on Ra´bai’s scale
of fluorophilicity,12 and perfluoro-tert-butanol (1,1,1,3,3,3-
hexafluoro-2-(trifluoromethyl)-2-propanol, (CF3)3COH), has
recently become commercially available at a reasonable
price.13 Ra´bai has fashioned a family of amines bearing one
or several fluorous tert-butyloxy groups and suggested that
they may be useful fluorous tagging reagents (ponytails).14
Figure 1. Structures of model compounds 1t,n and 2t,n and
retention times on fluorous HPLC.a
Pleasingly, the pair of compounds 1t,2t bearing the
perfluoro-t-butyloxy groups had marginally longer retention
times than the controls 1n,2n with perfluoro-n-butyl groups.
Because the pairs of compounds are not isomers (the “t”
series compounds have an extra oxygen atom compared to
the “n” series), it is not safe to conclude that compounds
bearing perfluoro-t-butyl groups will be better retained than
those bearing perfluoro-n-butyl groups. Nonetheless, it is
clear that the two groups are, at a minimum, roughly
comparable in retention behavior. The pair of compounds
2t/2n with the two C4F9 groups were also very well retained
(TR > 25 min), and they are accordingly projected to be
easily separated from organic compounds by standard
fluorous solid phase extraction.
Here we realize the goal of using perfluoro-tert-butyloxy
groups as fluorous tags for synthesis and separation in a
Mitsunobu setting.15 Taken together, Ra´bai’s results14 and
ours suggest that the pairing of a perfluoro-tert-butyloxy
group with a suitable spacer could form the basis of a second
generation of fluorous tags that will be superior for green
chemistry applications to today’s first-generation tags.
To quickly probe the ability of the perfluoro-tert-butyloxy
group to retain compounds on fluorous silica gel, we prepared
benzoate ester 1t and a control bearing a perfluoro-n-butyl
group 1n along with the analogous phthalates 2t and 2n
bearing two t-C4F9O or n-C4F9 groups, respectively (see
Supporting Information). Each sample was injected onto a
FluoroFlash PF-C8 HPLC column (4.6 mm × 150 mm) with
a fluorocarbon bonded phase.16 The samples were analyzed
To evaluate prospects for FSPE separations,2a we selected
the Mitsunobu setting15 because of its importance in synthesis
because Mitsunobu reagents conveniently accommodate two
fluorous groups and, finally, because we have made and
tested several dozen fluorous Mitsunobu reagents over the
past several years.17 This collection of reagents serves as a
convenient calibration for reaction and separation properties
of new reagents.
Since spacer effects can be very important in fluorous
Mitsunobu reactions, we initially targeted a pair of hydrazides
5e,p with ethylene (e) and propylene (p) spacers, respec-
tively. Fluorous hydrazide 5p was synthesized from per-
fluoro-t-butanol in three steps in 66% overall yield as shown
in Scheme 1. Specifically, treatment of the alcohol with KOH
in THF followed by addition of BrCH2CH2CH2OH gave
fluorous alcohol 3p in 79% yield after distillation. Alcohol
3p was reacted with carbonyl diimidazole (CDI) to provide
an intermediate (presumably imidazolide 4p),17b which was
directly reacted with hydrazine monohydrogen chloride and
triethylamine. The fluorous hydrazide 5p was isolated as a
(9) Prevedouros, K.; Cousins, I. T.; Buck, R. C.; Korzeniowski, S. H.
EnViron. Sci. Technol. 2006, 40, 32–44.
(10) Ellis, D. A.; Moody, C. A.; Mabury, S. A. In Organofluorines;
Neilson, A. H., Ed.; Springer-Verlag: Berlin, 2002; Vol. 3, pp 103-120.
(11) For example, 3M reformulated Scotchgard to substitute perfluo-
robutane sulfonic acid (PFBS) for PFOS: http://en.wikipedia.org/wiki/
Scotchgard.
(12) Kiss, L. E.; Kovesdi, I.; Rabai, J. J. Fluorine Chem. 2001, 108,
95–109.
(13) Ryan Scientific (www.ryansci.com) offers perfluoro-tert-butanol
1,1,1,3,3,3-hexafluoro-2-(trifluoromethyl)-2-propanol (or nonafluoro-tert-
butanol, (CF3)3COH, CAS 2378-02-1) at a price of $3420/kg for 1 kg and
$2700/kg for 10 kg.
(14) Szabo, D.; Mohl, J.; Balint, A.-M.; Bodor, A.; Ra´bai, J. J. Fluorine
Chem. 2006, 127, 1496–1504.
(15) For reviews on separation-friendly Mitsunobu reactions, see: (a)
Dembinski, R. Eur. J. Org. Chem. 2004, 2763–2772. (b) Dandapani, S.;
Curran, D. P. Chem.—Eur. J. 2004, 10, 3130–3138. For recent highlights
on the application of fluorous Mitsunobu reactions in total synthesis, see:
(c) Dakas, P.-Y.; Barluenga, S.; Totzke, F.; Zirrgiebel, U.; Winssinger, N
Angew. Chem., Int. Ed. 2007, 46, 6899–6902. (d) Gerard, B.; Cencic, R.;
Pelletier, J.; Porco, J. A, Jr Angew. Chem., Int. Ed. 2007, 46, 7831–7834.
(16) (a) The FHPLC column, FSPE cartridge, and Mitsunobu reagents
are commercially available from Fluorous Technologies, Inc. (b) DPC owns
an equity interest in this company.
(17) (a) Dandapani, S.; Curran, D. P. Tetrahedron 2002, 58, 3855–3864.
(b) Dandapani, S.; Curran, D. P. J. Org. Chem. 2004, 69, 8751–8757. (c)
Curran, D. P.; Bajpai, R.; Sanger, E. AdV. Synth. Catal. 2006, 348, 1621–
1624.
2454
Org. Lett., Vol. 10, No. 12, 2008

5p. Not surprisingly, hydrazide 6n with eight more fluorines
than the other three compounds shows a significantly longer
retention time (34.7 min).
Scheme 1
.
Synthesis of Spacer Reagents 3e,p and Hydrazides
5e,p
Under these HPLC conditions, compounds with retention
times more than about 12 min should be suitable for FSPE
separation. We expected the Mitsunobu reagent derived from
5p to exhibit superior reaction performance to that derived
from 5e,17b and conveniently, 5p was also better retained
than 5e. Thus, the preferred Mitsunobu reagent is clearly
7p (eq 1), which was produced in quantitative yield by
exposure of 5p to bromine and pyridine in benzotrifluoride
(C6H5CF3, BTF). The crude bright yellow liquid 7p obtained
upon workup exhibited excellent purity by 1H NMR analysis
and was used as is in subsequent reactions.
white solid in 83% yield after standard chromatography.
Fluorous hydrazide 5e was synthesized by a similar route
starting with BrCH2CH2OH and was isolated in 63% overall
yield. The simple, high-yielding preparation of alcohols 3e
and 3p recommends them as general reagents to introduce
t-C4F9O groups with spacers in place.
Hydrazides are the byproducts of Mitsunobu reactions that
need to be separated from target products. To evaluate the
performance of the new fluorous hydrazides, samples of 5e,
5p, and control fluorous hydrazides 5n (bearing (perfluoro-
n-butyl)propyl groups) and 6n (bearing larger (perfluoro-n-
hexyl)propyl groups) were injected onto FluoroFlash HPLC
under the above conditions.
The Mitsunobu reaction between 4-cyanophenol 8 and
p-fluorobenzyl alcohol 9 was chosen as a first test for fluorous
DEAD 7p (eq 2) because its success has been shown to be
predicted on the presence of suitable spacers in the fluorous
Mitsunobu reagent.17b A solution of 7p (0.42 mmol) in THF
(5 mL) was slowly added to a solution of 8 (0.42 mmol), 9
(0.23 mmol), and a fluorous triphenylphosphine (Ph2PC6H4-
p-CH2CH2C8F17, F-TPP, 0.42 mmol) in THF (5 mL). After
8 h, the crude mixture was concentrated and the residue taken
up in ether and washed with aqueous sodium hydroxide (to
remove excess phenol).
The retention times for these injections are listed in Figure
2. The retention times of 5e (13.4 min) and 5p (16.9 min)
The concentrated crude reaction mixture was then loaded
onto a 20 g FluoroFlash cartridge18 with MeOH.16 With the
aid of a new SPE apparatus (see Supporting Information),
the cartridge was washed with MeOH/H2O (80/20) to elute
the organic product. The MeOH/H2O fraction was concen-
trated and dried to give 4-cyanophenyl 4-fluorobenzyl ether
10a in 90% yield and 96% GC purity. A second wash with
100% MeOH eluted 5p and the fluorous triphenylphosphine
oxide (F-TPPO) in 95% recovery. Hydrazide 5p and F-TPPO
were separated by flash chromatography. The reagents were
reactivated for later use for other experiments.
Figure 2. Structures of new (5p,e) and control hydrazides and
retention times on fluorous HPLC.a
are much shorter than that of the phthalates 2t (28.4 min,
Figure 1). This is because of the polar hydrazide functional-
ity. The fluorous hydrazide 5p with propylene spacer showed
a retention time of about 1.5 min longer than the control 5n
(15.4 min). As before, the compounds are not isomers, but
this result reinforces the conclusion that compounds with
O-t-C4F9 groups will be better retained on fluorous silica gel
compared to those with n-C4F9 groups. We expected the
hydrazide with the ethylene spacer 5e to have a shorter
retention time,17b and indeed, it eluted about 3.5 min before
To explore the scope of the new fluorous DEAD reagent,
five different nucleophiles with various pKa’s were coupled
with assorted primary and secondary alcohols using 7p and
(18) This corresponds to about a 3% weight loading of the cartridge.
We attempted a similar experiment with a 5 g cartridge (the size used for
the standard Mitsunobo reagent with 26 fluorines), but significant break-
through (∼20%) of the fluorous products into the organic fraction was
observed. Thus, the presence of fewer fluorines in 7p (18 fluorines) must
be offset by lower loading in the FSPE.
Org. Lett., Vol. 10, No. 12, 2008
2455

F-TPP. The products were isolated by FSPE as above, and
their structures along with yields and purities are summarized
in Figure 3.
yield, which is comparable to the result with the commercial
reagent 6n (60%).17b The moderate yield is presumably due
to the relatively high pKa of 3-methylphenol.
In summary, a new fluorous DEAD reagent 7p bearing
two perfluoro-t-butyloxy groups with propylene spacers
shows excellent promise for use in fluorous Mitsunobu
reactions. Following reactions of assorted nucleophiles and
alcohols with 7p and a fluorous phosphine, we obtained pure
target products in good yield after removing fluorous
byproducts by FSPE. The FSPEs succeed even though 7p
has eight fewer fluorines than the current commercial reagent
6n, but this lower fluorine content must be offset by lower
loading on FSPE. Reagent 7p serves as a prototype for a
greener second generation of fluorous reagents bearing tags
that are not expected to degrade in the environment to
compounds that are highly persistent or that bioaccumulate
in higher organisms.
Acknowledgment. We thank the Institute of General
Medical Sciences of the National Institutes of Health for
funding of this work. Mr. Simon Nielsen, visiting graduate
student from the Danish University of Pharmaceutical
Sciences, designed the F-SPE apparatus.
Figure 3. Yield and purities of products from Mitsunobu reactions
with 7p.
Supporting Information Available: Complete experi-
mental procedures. This material is available free of charge
via the Internet at http://pubs.acs.org.
In five of the six reactions, including the one with
2-octanol, the yields were higher than 90% and crude product
purities were very good (92-97%). The reaction of 3-me-
thylphenol and p-fluorobenzyl alcohol provided 10c in 64%
OL800750Q
2456
Org. Lett., Vol. 10, No. 12, 2008