Fluorous biphasic catalysis. 2. Synthesis of fluoroponytailed amine ligands along with fluoroponytailed carboxylate synthons, [M(C8F17(CH2)2CO2) 2] (M = Mn2+ or Co2+)

Article abstract of DOI:10.1139/v01-012

Fluorous biphasic catalysis (FBC) is a relatively new concept for homogeneous catalysis where the fluorocarbon soluble catalyst resides in a separate phase from the substrate and products. Therefore, separation of the catalyst and the products occurs by a

Full text of DOI:10.1139/v01-012

888
Fluorous biphasic catalysis. 2. Synthesis of
fluoroponytailed amine ligands along with
fluoroponytailed carboxylate synthons,
[M(C8F17(CH2)2CO2)2] (M = Mn2+ or Co2+):
Demonstration of a perfluoroheptane soluble
precatalyst for alkane and alkene functionalization
in the presence of tert-butyl hydroperoxide and
oxygen gas
Jean-Marc Vincent, Alain Rabion, Vittal K. Yachandra, and Richard H. Fish
Abstract: Fluorous biphasic catalysis (FBC) is a relatively new concept for homogeneous catalysis where the fluorocar-
bon soluble catalyst resides in a separate phase from the substrate and products. Therefore, separation of the catalyst
and the products occurs by a facile decantation process. In this contribution, we present the synthesis of new
Rf-fluoroponytailed synthons, 2-iodo-1-perfluorooctyl-3-propanol (1), 3-perfluorooctyl-1-propanol (2), and 3-
perfluorooctyl-1-iodopropane (3), a variety of new Rf-fluoroponytailed ligands (4–8), with starting amines, 1,4,7-
triazacyclononane, bis-picolylamine, and bis-picolylaminoethylenediamine, as well as new RfMn2+ and RfCo2+
fluoroponytailed carboxylate synthons, [Mn(O2C(CH2)2C8F17)2] (9), and [Co(O2C(CH2)2C8F17)2] (10), where Rf is C8F17.
The only totally perfluoralkane soluble ligand we found was 1,4,7-tris-N-(4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-
heptadecafluoroundecyl)-1,4,7-triazacyclononane (RfTACN, 4), and it was utilized, along with synthons 9 and 10, to
generate in situ RfMn2+–RfTACN and RfCo2+–RfTACN complexes as precatalysts for functionalization of alkanes and
alkenes. We will demonstrate that indeed this novel FBC approach for the separation of the precatalyst from the sub-
strates and (or) products is viable for oxidation of alkanes and alkenes in the presence of the necessary oxidants, tert-
butyl hydroperoxide (TBHP), and O2 gas. We will also show that these oxidation reactions occur via an autoxidation
mechanism under our FBC conditions, while using electron spin resonance (ESR) techniques to ascertain the redox
chemistry occurring with the starting mononuclear RfMn2+–RfTACN complex.
Key words: fluorous solvents, biphasic catalysis, alkane/alkene oxidation. 895
Résumé : La catalyse biphasique en milieu fluoré est un concept récent de catalyse homogène nécessitant l’emploi de
catalyseurs solubles dans les perfluorocarbures. Le catalyseur se trouvant dans une phase différente du substrat et des
produits de réaction, il peut être recyclé par une simple décantation. Nous présentons la synthèse de nouveaux synthons
(1–3), et ligands (4–8), à chaîne perfluorée, ainsi que de nouveaux complexes métalliques à ligands carboxylates fluorés,
[Mn(O2C(CH2)2C8F17)2] (9) et [Co(O2C(CH2)2C8F17)2] (10). Le ligand 1,4,7-tris-N-(4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-
heptadecafluoroundecyl)-1,4,7-triazacyclononane (RfTACN, 4) est le seul ligand à être soluble dans les perfluorocarbu-
res. A partir des composé 9 et 10 et du ligand 4, les complexes RfMn2+–TACN et Co2+–TACN générés in situ servent
de précatalyseurs pour les oxydations d’alcanes et alcènes. Les réactions sont réalisées en milieu biphasique en pré-
sence des oxydants TBHP et O2 et les produits sont facilement séparés du catalyseur par simple décantation. Dans les
Received September 1, 2000. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on July 5, 2001.
The senior author, RHF, wishes to dedicate this fluorous biphasic catalysis contribution to Brian James, a long-time friend and
colleague, whose homogeneous catalysis studies/textbook have been an inspiration to several generations of chemists interested in
this fascinating discipline.
J.-M. Vincent,1 A. Rabion,2 V.K. Yachandra, and R.H. Fish.3 Lawrence Berkeley National Laboratory, University of California,
Berkeley, CA 94720, U.S.A.
1Present address: Laboratoire de Chimie Organique et Organométallique, UMR CNRS 5802, Université Bordeaux 1, 351 Cours de
la Libération, 33405 Talence CEDEX, France.
2Present address: Groupement de recherche de Lacq, BP 34, 64170 Artix, France.
3Corresponding author (fax: (510) 486-7303; e-mail: rhfish@lbl.gov).
Can. J. Chem. 79: 888–895 (2001)
DOI: 10.1139/cjc-79-5/6-888
© 2001 NRC Canada

Vincent et al.
889
conditions utilisées les réactions se déroulent suivant un mécanisme radicalaire d’autoxydation.
Mots clés : solvants fluorés, catalyse biphasique, oxydation alcane/alcène.
[Traduit par la Rédaction]
hydrogen peroxide (H2O2) in acetone, were previously
achieved in one homogeneous phase by the in situ prepara-
tion of Mn2+ catalysts, using 2,2′-bipyridine and tris-N-
methyl-TACN ligands, respectively, and further state that
separation of the Mn2+ catalyst from the product will be rel-
atively difficult or impossible (8).
Introduction
Vincent et al.
Recently, Horváth and Rábai (1) published an important
paper on a new homogeneous catalysis concept, fluorous
biphasic catalysis (FBC), that entailed the use of a fluoro-
carbon as one phase containing a modified fluoroponytailed
catalyst, while the substrate and the product were soluble in
a second hydrocarbon phase; this FBC concept was also re-
ported in a Ph.D thesis that was not readily accessible by the
catalysis community (2). More importantly, the Horváth and
Rábai paper has been a virtual catalyst for chemists inter-
ested in new biphasic concepts, while this novel FBC ap-
proach has led to many contributions in a very short time
frame (3). The process of developing a new paradigm for the
separation of the homogeneous catalyst from the substrate
and the product must start with the solvent system. There-
fore, it is well-known that perfluorohydrocarbons have un-
usual properties that entail being extremely hydrophobic and
lacking hydrogen bonding capabilities, which renders them
to be relatively insoluble in their hydrocarbon analogs, and
they are also able to dissolve various gases such as oxygen,
hydrogen, and carbon dioxide, as examples (4). Thus, it is
surprising that it took so long for this FBC approach to be
discovered, since the above mentioned parameters would allow
a number of critical catalytic reactions to be demonstrated
using perfluorohydrocarbons as one phase in a biphasic sol-
vent mode (1–3).
We have been interested in homogeneous oxidation cataly-
sis for many years with the focus on functionalizing alkanes
and alkenes at ambient temperatures (5). Our fascination
with the FBC process provided an opportunity to also dem-
onstrate this novel concept for the functionalization of al-
kanes and alkenes, and in a preliminary account (6), we did
indeed verify this new biphasic approach for alkane and
alkene functionalization, and in fact, a similar approach was
also published by Pozzi et al. (7) very soon after. Therefore,
in this full account, we present the synthesis of Rf-
fluoroponytailed synthons (1–3), a variety of new Rf-
fluoroponytailed ligands (4–8), with starting amines, 1,4,7-
triazacyclononane (TACN), bis-picolylamine (BPA), and bis-
picolylaminoethylenediamine (BisPICEN), as well as new
RfMn2+ and RfCo2+ fluoroponytailed carboxylate synthons,
[Mn(O2C(CH2)2C8F17)2] (9), and [Co(O2C(CH2)2C8F17)2]
(10), where Rf is C8F17, and clearly demonstrate the critical
need to pay close attention to the structures of the ponytailed
ligands as it pertains to their solubility in perfluorocarbons.
To reiterate, we will further demonstrate the utilization of
the FBC paradigm with the functionalization of alkanes and
(or) alkenes, using in situ generated RfMn2+–RfTACN and
RfCo2+–RfTACN as precatalysts that are totally soluble in
the fluorous phase, and in the presence of the necessary oxi-
dants, tert-butyl hydroperoxide (TBHP) and O2 gas (6). In
comparison to the new FBC concept presented in this paper,
it is important to note that the oxidation of alkanes with
TBHP in acetonitrile, and epoxidation of alkenes with
Results and discussion
Synthesis of fluoroponytailed ligands 4–8
One of the key components of the FBC approach includes
the synthesis of fluoroponytailed ligands that are totally sol-
uble in fluorocarbon solvents at ambient temperature. There-
fore, we set out initially to create a fluoroponytail synthon
that we could append to our target ligands, which were
aliphatic and aromatic substituted aliphatic amines. We
chose aliphatic amines as ligands for several reasons that in-
cluded their ease in appending fluoropontytails via
alkylation reactions, and their metal complexes have been
widely used in previous homogeneous oxidation chemistry
of hydrocarbons (8). In our initial experiments to find a suit-
able fluoroponytailed synthon, we evaluated a compound
that had a two-carbon spacer (C8F17CH2CH2I) but quickly
found that in the presence of the aliphatic amine ligand that
loss of HI prevailed rather than alkylation. Therefore, it is
important to recognize that the three-methylene-carbon spacer
in compound 3 was necessary, not only to insulate the amine
from the powerful e-withdrawing effect of the perfluoroalkyl
group, but also to avoid the above mentioned facile elimina-
tion reaction of HI that predominately occurs when a two-
carbon spacer was used during the formation of ligands 4–8.
The first step to the three-carbon spacer synthon,
perfluoroalkyl iodide (3) proceeded by a free radical addi-
tion of a perfluoroalkyl iodide (C8F17I) to allyl alcohol that
was initiated with AIBN to provide the perfluoroalkylated
iodohydrin 1 (this procedure was slightly different than the
one described by Kotora et al., who synthesized 3-
perfluorohexyl-1-propanol (C6F13(CH2)3OH) using copper
powder as a free radical initiator in neat solution followed
by tin hydride reduction. See ref. 9). Compound 1 was then
reduced to the perfluoroalkyl alcohol 2, by using tributyltin
hydride in dry benzotrifluoride. Iodination of 2 to form 3
was achieved using 85% phosphoric acid in the presence of
phosphorous pentoxide and potassium iodide in 85% yield
(Scheme 1).
The first alkylation approach we utilized to introduce the
fluoroponytail, 2, was the aromatic nucleophilic substitution
with the N-pentafluorophenyl derivatives of the TACN and
BPA ligands, leading to ligands 5 and 8 (Scheme 1). We in-
troduced 2, via a nucleophilic substitution of the p-fluoro
substituent of the N-pentafluorophenyl group by using phase
transfer catalysis in a mixture of 50% NaOH–trifluorotoluene,
and in the presence of the phase transfer agent, Aliquat 336.
The other alkylation approach, which directly alkylates
© 2001 NRC Canada

890
Can. J. Chem. Vol. 79, 2001
amine ligands (TACN), (BPA), and (BisPICEN) to provide
4, 6, and 7 with 3, was conducted using K2CO3 in DMSO.
would be very difficult to solubilze in perfluorocarbons.
Therefore, the new RfMn2+ and RfCo2+ complexes
[Mn(C8F17(CH2)2CO2)2] (9) and [Co(C8F17(CH2)2CO2)2]
(10) were synthesized by reaction of Mn(ClO4)2·6H2O or
Co(ClO4)2·6H2O in acetone with
2
equiv of the
triethylammonium salt of the 3-Rf-alkylpropionic acid
C8F17(CH2)2CO2H; elemental analysis of 9 and 10 provided
a metal:carboxylate ligand ratio of 1:2, while 9 was found to
be slightly soluble in perfluorocarbons, exhibited an intense
and broad ESR signal at g = 2 (9 K), and displayed the
broad six-line hyperfine structure (J = 90 G). This could be
attributed either to an oligomeric Mn2+ structure, consistent
with a Mn(OAc)2 trimeric structure, or a mononuclear Mn2+
complex of low solubility in perfluorocarbons at 9K (10). In
addition, it appears that the carboxylate group is chelating,
rather than being in a monodentate coordination mode, in
agreement with the small difference (∆ν = 132 cm–1) ob-
served in the IR spectrum of 9 between the asymmetric and
symmetric C—O stretch (ν = 1578 and 1446 cm–1), respec-
tively. Complex 10 had a similar IR spectrum as 9 and we
suggest that the Co2+ analog has a similar structure as the
Mn2+ complex.
Interestingly, when 1 equiv of 4 was added to a solution
of 9 in perfluoroheptane, the complete solubilization of 9 oc-
curred concomitantly with the appearance of a new UV ab-
sorption band at 320 nm tentatively suggesting that a new
[RfMn(RfTACN)]2+ complex (11) was formed in situ. A sim-
ilar result was also achieved when ligand 4 and complex 10
were added together to provide [RfCo(RfTACN)]2+ (complex
12).
We were totally amazed to find that 4, tris-N-
(4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-heptadecafluoroundecyl)-
1,4,7-triazacyclononane (RfTACN), was the only fluoropony-
tailed ligand found to be soluble in perfluorocarbons, such
as perfluoroheptane (PFH) or perfluoromethylcyclohexane
(PFMC), at room temperature. Fluoroponytailed ligands 5–
8, with a fluorine content for 5–8 of <60% were only soluble
in hot perfluorocarbons (-50°C), while 4 had a fluorine con-
tent of 64.2%. These results vividly demonstrate that polar
groups such as a pentafluorophenyl group or a nonfluorinated
pyridine have to be avoided to obtain good solublity in
perfluorocarbons; these are extremely nonpolar solvents. It
also shows that the solubilzation of ligands, and moreover,
the subsequent metal complexes, could represent an impor-
tant limiting step in the FBC approach for oxidation chemis-
try. Moreover, it is of interest to also note that 4
represented at that time, to our knowledge, the first exam-
ple of an Rf–amine ligand (6) to be found to be soluble in a
perfluorocarbon solvent, subsequently also verified by Pozzi
et al. (7) with their tetra-azamacrocycle systems.
Functionalization of alkanes and alkenes utilizing in situ
generated [RfMn(RfTACN)]2+ and [RfCo(RfTACN)]2+
(complexes 11 and 12) as precatalysts
The FBC oxidation results for several substrates are pre-
sented in Table 1. All the experiments were carried out un-
der biphasic conditions, generating complexes 11 or 12, in
situ in perfluoroheptane (Scheme 2), while the upper phase
was the substrate itself. Importantly, the oxidation products
were detected only in the colorless upper phase after
decantation (GC), while the colored perfluoroheptane lower
phase contains only traces (<5%) of product. Clearly, a fac-
ile and rapid separation of the products from the catalyst was
achieved using this FBC approach.
Table 1 demonstrates that the olefin with allylic hydro-
gens, cyclohexene, provides the highest % yield of oxidation
products (650%/TBHP in 3 h) with TBHP–O2, while alkane
oxidation was much lower as noted with cyclohexane to
cyclohexanol–cyclohexanone (CyOH–CyONE, 12%/TBHP
in 24 h) and toluene to benzyl alcohol–benzaldehyde
(PhCH2OH–PhCHO, 65%/TBHP in 24 h). In the presence of
11, TBHP, and O2 (1 atm (1 atm = 101.325 kPa)), and under
vigorous stirring, cyclohexene was converted to a product ra-
tio of 2-cyclohexen-1-one (CyenONE, -65%), 2-
cyclohexen-1-ol (CyenOH, -35%), and a small amount of
cyclohexene oxide (<2%), while with complex 12 similar
oxidation results were obtained.
Synthesis of Mn2+ and Co2+ fluoroponytailed
carboxylate complexes 9 and 10
To insure fluorcarbon solvent solubility of the metal com-
plexes of ligand 4, the only amine ligand we found to have
total fluorocarbon solubility at ambient temperature, we
thought that appending fluoroponytailed groups to the metal
complexes we were going to utilize as synthons would help
in this important aspect, since metal complexes with multi-
ple charges and counter anions, such as ClO−4, NO3−, PF6−,
It is also important to note that styrene, with no allylic hy-
drogens, was not converted to styrene epoxide under these
biphasic conditions. In the absence of O2 or TBHP, only
negligible amounts of CyenOH and CyenONE were detected,
© 2001 NRC Canada

Vincent et al.
891
Scheme 1. Preparation of fluoroponytailed synthon 3 and alkylation to provide ligands 4–8.
Rf =
C8F17
Br
R
R
N
H
RfI
F
OH
, AIBN
75oC, 14 h
75%
KOH
Toluene, 80oC
90%
I
Rf
OH
1
SnH, AIBN
R
R
N
F
C6H5CF3, 80oC, 4 h
F
75%
Rf
F
F
F
OH
2
KI, P2O5
H3PO4 (85%), 120oC, 3h
R
R
N
H
85%
Rf
NaOH 50 %
PTC (aliquat 336)
C6H5CF3, 85 oC, 24 h
I
80%
3
K2CO3
60%
DMSO, 120oC, 22 h
R
R
N
F
F
F
F
O
R
R
N
Rf
5, 8
Rf
4, 6, 7
indicative of an autoxidation reaction, i.e., both were neces-
sary for oxidation to proceed. At the end of the reaction, the
upper phase was removed and new aliquots of cyclohexene
and TBHP were added to provide a 400% yield after 5 h,
showing that the catalyst, after decantation, was only present
in the lower fluorocarbon phase.
cyclohexene, the allylic radical that formed was then trapped
by O2 (k > 1 × 109 M–1 s–1) to provide cyclohexenylperoxy
radicals, which were able (ROO-H = 90 kcal mol–1) to
homolytically remove a hydrogen from benzylic or allylic
–1
C—H bonds (85 kcal mol ), and hence, propagate the radi
-
cal reactions. The secondary cyclohexenyl hydroperoxide
that was formed then decomposes catalytically in the pres-
ence of the RfMn(RfTACN) catalyst to give the alcohol and
the ketone products (12). Scheme 3 defines the initiation,
propagation, termination, and product-forming steps for the
sequence of reactions (eqs. [1]–[8]).
Autoxidation mechanism and fate of the Mn2+ precatalyst
The presented results are in agreement with an autoxidation
mechanism involving alkoxy (RO·) or alkylperoxy (ROO·)
radicals (11), where the reaction was initiated by t-BuO· or
t-BuO2· radicals produced from redox reactions (Haber–
Weiss process with Mn2+/Mn3+). In the case of the substrate
Moreover, the lower yields observed in the case of cyclo-
hexane are consistent with the stronger C—H bond strength
© 2001 NRC Canada

892
Can. J. Chem. Vol. 79, 2001
Table 1. Alkene and alkane functionalization under fluorous biphasic catalysis conditions.a
Precatalyst
Substrate
Oxidant
Products (µmol)
Yield (%)b
t (h)
11
11
11
9
11
11
11
12
12
Cyclohexene
Cyclohexene
Cyclohexene
Cyclohexene
Styrene
TBHP–O2
TBHP
O2
TBHP–O2
TBHP–O2
TBHP–O2
TBHP–O2
TBHP–O2
TBHP–O2
CyenOH (160) CyenONE (300)
CyenOH (<2) CyenONE (<2)
CyenOH (<1) CyenONE (<1)
CyenOH (130) CyenONE (160)
no epoxide
PhCHO (15) PhCH2OH (30)
CyOH (5) CyONE (3.5)
CyenOH (185) CyenONE (360)
CyOH (7) CyONE (5.5)
650
360
3
7
24
12
24
24
24
20
24
Toluene
65
12
750
17
Cyclohexane
Cyclohexene
Cyclohexane
aConditions: [M(Rf(CH2)2CO2)2] (3.5 µmol), Mn or Co, and the RfTACN ligand (3.5 µmol) were dissolved in hot
perfluoroheptane (3 mL) and then the substrate (2 mL) was added. The reaction starts after the addition of TBHP (90%,
72 µmol) under an O2 atmosphere at ambient termperature.
bTotal yield was based on TBHP added.
Scheme 2. Fluorous biphasic oxidation catalysis with cyclohexene
as the substrate and complex 11 formed in situ from 4 and 9.
Scheme 3. Initiation, propagation, termination, and product form-
ing steps with cyclohexene as substrate.
O
OH
+
O2
t-BuO2H
Rf
[Mn2+(Rf(CH2)2CO2)2]
Rf
N
N
N
9
Rf = C8F17
Rf
Perfluoroheptane
O2 (1 atm)
4
298 K
(95 kcal mol–1) (11), which implicates the cyclohexylperoxy
radical, and thus causes a lowering of the rate of the propa-
gation step. The chain termination step presumably comes
mainly from the coupling of two cyclohexenylperoxy radi-
cals to give alcohol, ketone, and O2 (Russell-type mecha-
nism (Scheme 3, eq. [7])) (11).
We found that the [RfMn(RfTACN)] complex, formed dur-
ing the oxidation of 11, was plausibly a RfMn3+Mn4+ dimer.
Therefore, when TBHP was added to the reaction mixture,
the solution turns brown, and exhibits intense UV–vis ab-
sorptions (ε per atom is -9000 and 1130 M–1 cm–1 at 300
and 500 nm, respectively, after 1 h) characteristic of dinuclear
Mn3+Mn3+, Mn3+Mn4+, or Mn4+Mn4+ complexes (13). The
occurrence of such a putative RfMn3+Mn4+ dimer from start-
ing complex 11 (Fig. 1, top spectrum), under oxidation con-
ditions, was unambiguously confirmed by ESR studies at
9 K. Thus, a strong, distinctive 16-line signal of an antiferro-
magnetically coupled RfMn3+Mn4+ dimer complex at g = 2
was observed in the perfluoroheptane phase, after 1 h of re-
action in the presence of TBHP–O2 and cyclohexene (Fig. 1,
middle spectrum). Indeed, it has recently been shown that
the decomposition of TBHP in the presence of a dinuclear
Mn23+(2-OHsalpn)2 complex resulted in the formation of
t-BuO· radicals and a dinuclear Mn3+(µ-O)2Mn4+ complex (14).
Furthermore, rxn. [1] (Scheme 3) was favored over rxn.
[2] with complex 11 for two important reasons: (i) it has
been previously shown that in nonpolar solvents, eq. [2] was
found to be very slow (15); and (ii) the RfMn3+Mn4+ dimeric
species formed upon the addtition of TBHP to complex 11
was not very efficient in the oxidation of TBHP. Interest-
ingly, after 3 h (Fig. 1, lower spectrum), the 16-line ESR
signal had almost disappeared, suggesting that this
RfMn3+Mn4+ dinuclear mixed valent species might be in-
volved in the catalytic decomposition of the cyclohexenyl
hydroperoxide intermediate to provide the alcohol and ketone
products.
© 2001 NRC Canada

Vincent et al.
893
Fig. 1. ESR experiments at 20 K and 9 K (see Experimental for
details). Top spectrum: [RfMn(RfTACN)]2+ complex (11) at 20 K
in benzotrifluoride; middle: after addition of TBHP–O2 and
cyclohexene (9 K); lower: after 3 h of reaction (9 K).
Experimental
Materials and instrumentation
The fluorocarbon solvents and some starting perfluoro de-
rivatives were purchased from Pierce Chemical Co., while
the alkenes and alkanes were from Aldrich Chemical Co.
The oxidant, tert-butyl hydroperoxide (70% TBHP) was a
commercial product from Aldrich Chemical Co., and all
1
were used without further purification. The H NMR spectra
were obtained on a 500 MHz Brucker NMR spectrometer,
while GC and GC–MS analysis were performed on Hewlett–
Packard (HP) instruments. The UV–vis spectra were re-
corded on an HP diode array instrument with accompanying
software. The ESR experiments were performed at the Cal-
vin Laboratory at LBNL and the following describes the
equipment used. EPR spectra were collected on a Varian E-
109 spectrometer with an E-102 microwave bridge. The EPR
spectra were collected using a home built signal averager
and stored for further analysis on a VAX 4000-300. Samples
were maintained at cryogenic temperatures using an Air
Products Heli-tran liquid helium cryostat. For detecting the
Mn2+ complex the conditions were as follows: 3300 ± 1000 G
scan range, 10 mW microwave power, 20 K temperature,
32 G modulation amplitude, 100 kHz modulation frequency,
2 min/scan, 0.25 s time constant. Spectrometer conditions
were as follows for detecting the 16 line spectrum from the
coupled binuclear Mn3+/Mn4+ complex: 2800 ± 500 G scan
range, 30 mW microwave power, 9 K temperature, 32 G
modulation amplitude, 100 kHz modulation frequency,
0.25 s time constant, 9.26 GHz microwave frequency. The
amplitudes were calculated from the low-field and high-field
peak-to-trough amplitudes for each designated peak.
2-Iodo-1-perfluorooctyl-3-propanol (1)
The RfI (Rf = C8F17, 7 g, 12.8 mmol), allyl alcohol (1 mL,
14.7 mmol), and AIBN (84 mg, 0.51 mmol) were heated at
70–75°C under an inert atmosphere for 14 h. Every 2 h, a
new portion of AIBN was added to the reaction mixture. The
pale yellow solid obtained was recrystallized in refluxing
hexane (40 mL). Compound 1 was obtained in 75% yield,
mp 93–94°C. EI-MS: [M+] 604. 1H NMR (400 MHz, CDCl3,
25°C) δ: 4.45 (m, 1H, -CHI-), 3.79 (m, 2H, -CH2-OH), 2.9
(2m, 2H, Rf-CH2), 2.04 (t, -OH). Elemental anal. calcd. for
C11H6F17IO: C 21.87, H 1.00; found: C 22.06, H 0.97.
Conclusions
Finally, as we have clearly shown in this contribution, as
well as being pointed out recently by Pozzi et al. (16), despite
the apparent simplicity of the FBC concept, the concrete
demonstration of this strategy is still a difficult challenge,
since the solubilization of the fluoroponytailed metal com-
plexes used in perfluorocarbons is one limiting step along
with catalyst stability and the favorable recycling of the cat-
alyst system. Thus, we were able to successfully solubilize
complexes 11 and 12 in perfluoroheptane, using both the
new RfTACN ligand 4, and the new RfMn2+ and RfCo2+
complexes 9 and 10, with fluoroponytailed carboxylate lig-
ands. Moreover, we have been able to perform alkane and
alkene oxidations under FBC conditions in the presence of
TBHP and O2 as oxidants, and hence, separate the products
from the catalyst by a simple decantation process. The oxi-
dation process occurs via an autoxidation mechanism, with
formation of alkenyl or alkyl hydroperoxides as the key inter-
mediates, which then are catalytically decomposed by the
RfMn– or RfCo–RfTACN catalyst, probably at the solvent
interface, to provide the alcohol and ketone products. Future
studies will attempt to further the scope of the FBC ap-
proach to oxidation chemistry.
3-Perfluorooctyl-1-propanol (2)
Compound 1, RfCH2CHICH2OH (8 g, 13.2 mmol), and
AIBN (52 mg, 0.31 mmol) were partially dissolved in dry
trifluorotoluene (40 mL). The tributyltin hydride (4.2 mL,
15.8 mmol) was then added dropwise and the reaction mix-
ture was heated at 80°C under an inert atmosphere for 4 h
(reaction followed by GC). After removing the solvent under
vacuum, the resulting residue was dissolved in
a
perfluoroheptane–toluene mixture. After decantation, the
lower phase was separated from the upper one, which mainly
contains (Bu)3SnI. By removing the perfluoroheptane, com-
pound 2 was isolated as a white powder in 75% yield.
Compound 2 can be recrystallized from hexane. EI-MS:
1
[M+] 477. H NMR (400 MHz, CDCl3, 25°C) δ: 3.75 (m,
2H, -CH2-OH), 2.20 (m, 2H, Rf-CH2-), 1.88 (m, 2H, -CH2-
© 2001 NRC Canada

894
Can. J. Chem. Vol. 79, 2001
CH2-CH2-), 1.57 (s, -OH). Elemental anal. calcd. for
C11H7F17O: C 27.63, H 1.46; found: C 27.72, H 1.63.
(t, 2H, pyr), 7.46 (d, 2H, pyr), 7.16 (m, 2H, pyr), 3.83 (s,
4H, -N-CH2-pyr), 2.63 (t, 2H, -N-CH2-CH2-), 2.03 (m, 2H,
-CH2-CH2-CH2), 1.76 (m, 2H, -CH2-CH2-Rf). FAB-MS : [M
+ H+] 660.
3-Perfluorooctyl-1-iodopropane (3)
The phosphorous pentaoxide (3.4 g) was added to phos-
phoric acid (85%, 7.1 mL) in a 50 mL round-bottom flask
and then the mixture was cooled to 0°C. The KI (1.39 g,
8.2 mmol) was first added, followed immediately by
RfCH2CH2CH2OH (1.5 g, 3.1 mmol). The mixture was
heated at 120°C for 3.5 h. At ambient temperature, 10 mL of
water was added, and the resulting brown solution was ex-
tracted four times with 25 mL of diethyl ether. The organic
layer was washed twice with 25 mL thiosulfate (0.1 M), then
dried over Na2SO4. After removing the solvent, the iodo de-
rivative was obtained as an oil, which solidified at 4°C.
Compound 3 was used without further purification (yield
85%); however, 3 can be recrystallized from methanol. EI-
Synthesis of ligand 7
The BisPicen (50 mg, 0.206 mmol), K2CO3 (80 mg,
0.580 mmol), and RfCH2CH2CH2I (266 mg, 0.450 mmol)
were dissolved in DMSO (2.5 mL, distilled over CaH2) and
then heated at 90°C for 24 h. The solvent, benzotrifluoride,
was added at room temperature to dissolve a brown oil, and
then the solvent was removed under vaccum. To the sticky
residue was added dichloromethane and then the solution
was filtered, and the solvent removed under vaccum. The
ligand (7) was purified first by alumina column chromatog-
raphy (CH2Cl2–MeOH) and then by recrystallization from
1
hexane to obtain a yellowish powder in 40% yield. H NMR
1
MS: [M+] 588. H NMR (400 MHz, CDCl3, 25°C) δ: 3.26
(400 MHz, CDCl3, 25°C) δ: 8.54 (d, 2H, pyr), 7.66 (t, 2H,
pyr), 7.46 (d, 2H, pyr), 7.16 (m, 2H, pyr), 3.83 (s, 4H, -N-
CH2-pyr), 2.63 (t, 2H, -N-CH2-CH2-), 2.03 (m, 2H, -CH2-
CH2-CH2), 1.76 (m, 2H, -CH2-CH2-Rf). FAB-MS: [M + H+]
1163. Anal. calcd. for C36 H28 N4 F34: C 37.2, H 2.40, N
4.82; found: C 36.51, H 2.33, N 4.38.
(t, 2H, I-CH2-), 2.16 (2m, 4H, -CH2-CH2-Rf). Elemental
anal. calcd. for C11F17H6I: C 22.47, H 1.02; found: C. 22.80,
H 1.26.
Tris-N-(4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-
heptadecafluoroundecyl)-1,4,7-triazacyclononane (4)
The TACN (88.4 mg, 0.73 mmol), K2CO3 (423 mg,
3.1 mmol), and RfCH2CH2CH2I (1.41 g, 2.41 mmol) were
dissolved in DMSO (10 mL, distilled over CaH2) and heated
at 90°C for 24 h. Then perfluoroheptane (20 mL) was added
to the reaction mixture and the brown fluorous lower phase
was separated and then filtered. After removing the solvent,
compound 4 was obtained as a brown oil. After crystalliza-
tion from hot hexane, compound 4 was isolated as a yellow-
ish powder in 50% yield. FAB-MS: [M + H+] 1510. 1H
NMR (400 MHz, CDCl3, 25°C) δ: 2.71 (s, 12H, -N-CH2-
CH2-N), 2.55 (t, 6H, -N-CH2-CH2-), 2.17 (m, 6H, -CH2-
CH2-CH2), 1.58 (m, 6H, -CH2-CH2-Rf). Elemental anal.
calcd. for C39H30F51N3: C 31.03, F 64.20, H 1.98, N 2.78;
found: C 30.74, F 64.31, H 2.02, N 2.70.
Synthesis of tris-N-pentafluorophenyl-1,4,7-
triazacyclononane and ligand 8
To a stirring solution of TACN (3.7 mmol) in anhydrous
toluene (10 mL) in a glovebox was added powdered KOH
(11 mmol) slowly, and then the resulting suspension was
treated with C6F5CH2Br (11 mmol) dropwise. The reaction
mixture was heated to 80°C and stirred at this temperature
for 8 h. After being cooled to room temperature, it was then
treated with an additional portion of powdered KOH and
then C6F5CH2Br. The mixture was stirred and heated to
80°C for another 8 h. After filtration, the toluene is removed
under vacuum leading to a yellow solid, which gave a satis-
factory NMR spectrum. Tris-N-pentafluorophenyl-1,4,7-
triazacyclononane can be further purified by recrystallization
from MeOH; mp 75°C. EI-MS: [M+] 669. 1H NMR
(400 MHz, CDCl3, 25%C) δ: 2.70 (s, 12H, -CH2-TACN),
3.71 (s, 6H, -CH2-Ph). 13C NMR (MHz, CDCl3, 25°C) δ:
48.31, 54.69, 135.99, 138.50, 144.63, 146.62. 19F NMR
(MHz, CDCl3–CFCl3, 25°C) δ: –142.82 (2F), –156.11 (1F),
–162.91 (2F). Elemental anal. calcd. for C28H18F15N3: C
48.4, F 42.5, H 2.7, N 6.3; found: C 48.3, F 42.4, H 2.8, N
6.2.
The PTC catalyst, Aliquat 336 (49 µmol), trifluorotoluene
(5 mL), tris-N-pentafluorophenyl-1,4,7-triazacyclononane
(1.49 mmol), RfCH2CH2CH2OH (4.92 mmol), and NaOH
50% (5 mL) were stirred at 85°C for 22 h. After removing
the solvent in vacuo, the solid was taken off with a mixture
of trifluorotoluene, dichloromethane, and water. After stir-
ring, the cloudy organic phase was isolated and the solvent
stripped in vacuo. After washing the resulting orange solid
with dichloromethane, compound 8 was obtained, after fil-
tration, as a white powder (80%). Compound 8 can be fur-
ther purified by alumina gel chromatography (CH2Cl2–MeOH,
0–2%). m p 98°C. FAB-MS: [M+H+] 2044. 1H NMR
(400 MHz, CDCl3, 25°C) δ: 4.27 (t, 6H, -O-CH2-), 3.68 (s,
6H, N-CH2-C6F4), 2.71 (s, 12H, CH2-TACN), 3.37 (m, 6H,
O-CH2-CH2), 2.10 (m, 6H, -CH2-CH2-Rf). 13C NMR (MHz,
Synthesis of the N-(pentafluorotoluene)-bis-picolylamine
and ligand 5
This procedure for the preparation of 5 was similar to that
used to prepare ligand 8. At the end of the reaction, the tolu-
ene was removed under vacuum, and a brownish solid was
obtained. This solid was triturated with pentane affording a
beige powder. Ligand 5 was obtained in 85% yield and pro-
vided a satisfactory 1H NMR spectrum. 1H NMR (400 MHz,
CDCl3, 25°C) δ: 8.52 (d, 2H, aromatic H), 7.65 (t, 2H, aro-
matic H), 7.51 (d, 2H, aromatic H), 7.15 (m, 2H, aromatic
H), 3.90 (s, 2H, -CH2-C6F5), 3.85 (s, 4H, -CH2-C5H4N).
Synthesis of ligand 6
BPA (59 mg, 0.296 mmol), K2CO3 (57 mg,),
RfCH2CH2CH2I (191 mg, 0.325 mmol) in DMSO (0.5 mL,
distilled over CaH2) were heated at 90°C for 16 h. The reac-
tion mixture was filtered and the solvent removed in vacuo.
Ligand 6 was purified by alumina column chromatography
(CH2Cl2–MeOH) and was recovered as a beige oil, which
solidifies at 4°C providing a yield of 60%. Ligand 6 was
found to be soluble in hot, but not in cold perfluoroheptane.
1H NMR (400 MHz, CDCl3, 25°C) δ: 8.54 (d, 2H, pyr), 7.66
© 2001 NRC Canada

Vincent et al.
895
CDCl3, 25°C) δ: 48.31, 54.69, 135.99, 138.50, 144.63,
146.62. 19F NMR (MHz, CDCl3–CFCl3, 25°C) δ: –81.29
(3F), –114.92 (2F), –122.23 (2F), –122.43 (4F), –123.23
(2F), –124.00 (2F), –126.63 (2F), –144.16 (2F), –158.19
(2F). Anal. calcd. for C60H36N3F63O3: C 35.3, H 1.8, N 2.1,
F 58.6; found: C 35.6, H 1.9, N 2.1, F 59.6.
Contract no. DE-AC03–76SF00098. We thank Dr. István T.
Horváth and Dr. Jozsef Rábai for many helpful discussions
concerning our FBC program.
References
1. I. T. Horváth and J. Rábai. Science (Washington, D.C.), 266,
72 (1994), and refs. therein.
[Mn(C8F17(CH2)2CO2)2(H2O)2] (9)
2. M. Vogt. Ph.D Thesis. Technical University of Aachen, 1991.
3. (a) I.T. Horváth. Acc. Chem. Res. 31, 641 (1998) and refs.
therein; (b) I.T. Horváth. In Applied homogeneous catalysis
with organometallic compounds. Edited by B. Cornils,
W.A. Hermann). VCH, 1996. p. 601; (c) E. de Wolf, G. van
Koten, and B.-J. Deelman. Chem. Soc. Rev. 28, 37 (1999);
(d) R.H. Fish. Chem. Eur. J. 5, 1677 (1999), and refs. therein.
4. M. Hudlicky and A.E. Pavlath. Chemistry of organic com-
pounds II. Ellis Horwood, Chichester, U.K. 1992.
5. K. Neimann, R. Neumann, A. Rabion, R.M. Buchanan, and
R.H. Fish. Inorg. Chem. 38, 3575 (1999), and refs. therein.
6. J.-M. Vincent, A. Rabion, V.K. Yachandra, and R.H. Fish.
Angew. Chem. Int. Ed. Engl. 36, 2346 (1997).
7. G. Pozzi, M. Cavazzini, S. Quici, and S. Fontana. Tetrahedron
Lett. 38, 7605 (1997).
8. (a) S. Ménage, M.-N. Collomb-Dunand-Sauthier, C. Lambeaux,
and M. Fontecave. J. Chem. Soc. Chem. Commun. 1885 (1994);
(b). D. De Vos and T. Bein. J. Chem. Soc. Chem. Commun.
917 (1996).
9. M. Kotora, M. Hajek, B. Ameduri, and B. Boutevin. J. Fluo-
rine Chem. 68, 85 (1994).
10. P. Mathur and G. Dismukes. J. Am. Chem. Soc. 105, 7093
(1983).
11. R.A. Sheldon and J. K. Kochi. Metal-catalyzed oxidations of
organic compounds. Academic Press, New York. 1981.
12. R.H. Fish, M.S. Konings, K.J. Oberhausen, R.H. Fong, W.M.
Yu, G. Christou, J.B. Vincent, D.K. Coggin, and R.M. Buchanan.
Inorg. Chem. 30, 3002 (1991).
13. Y.-M. Frappart, A. Boussac, R. Albach, E. Anxolabéhère-
Malhart, M. Delroisse, J.-B. Verlhac, G. Blondin, J.-J. Girerd,
J. Guilhem, M. Cesario, A.W. Rutherford, and D. Lexa. J. Am.
Chem. Soc. 118, 2669 (1996).
The CF3(CF2)7CH2CH2CO2H (1.35 g, 2.80 mmol) was
dissolved in acetone (15 mL) and to this solution was added
triethylamine (380 mL, 2.80 mmol). This resulting solution
was added dropwise to Mn(ClO4)2·6H2O (500 mg,
1.37 mmol) dissolved in acetone (30 mL). The sticky precip-
itate which formed was vigorously stirred for 2 h. After fil-
tration, 9 was obtained as a white powder in 75% yield.
Elemental anal. calcd. for C22F34H12O6Mn: C 24.61, F
60.19, H 1.12, Mn 5.12; found: C 25.37, F 59.77, H 0.94,
Mn 5.50.
[Co(C8F17(CH2)2CO2)2(H2O)2] (10)
The same procedure, as conducted with complex 9, was
used for the synthesis of 10, except with Co(ClO4)2·6H2O. A
pink precipitate formed immediately, and after filtration, a
pink powder was obtained in 100% yield. Elemental anal.
calcd. for C22H12F34CoO6: C 24.53, H 1.11, F 59.98, Co
5.46; found: C 25.05, H 1.27, F 60.15, Co 5.35.
ESR–FBC experiment
The ESR spectrum for in situ generated 11 was conducted
at 20 K in benzotrifluoride (top spectrum, Fig. 1). Using
standard oxidation conditions for cyclohexene as outlined in
Table 1, aliquots of the perfluoroheptane layer were removed
and immediately frozen at 100 K in an ESR tube (middle
and lower spectra, Fig. 1). ESR spectra were recorded at 9 K
using microwave power at 10 mW, modulation amplitude
32 G, a time constant of 0.5 s, at a frequency of 9.18 GHz,
and gain of 3.2 × 10–3.
14. M.T. Caudle, P. Riggs-Gelasco, A.K. Gelasco, J.E. Penner-
Hahn, and V.L. Pecoraro. Inorg. Chem. 35, 3577 (1996).
15. R.A. Sheldon and J.K. Kochi. Adv. Catal. 25, 272 (1976).
16. G. Pozzi, S. Banfi, A. Manfredi, F. Montanari, and S. Quici.
Tetrahedron, 52, 36, 11 879 (1996).
Acknowledgment
The studies at LBNL were generously supported by a gift
grant (R.H.F.) and a postdoctoral fellowship to J.-M. V. from
Elf Aquitaine Inc., and the Department of Energy under
© 2001 NRC Canada