The first fluorous biphase hydrogenation catalyst incorporating a perfluoropolyalkylether: [RhCl(PPh2(C6H 4C(O)OCH2CF(CF3)(OCF2CF(CF 3))nF))3] with n = 4-9

Article abstract of DOI:10.1016/j.jfluchem.2012.09.001

The phosphine oxide (4-diphenylphosphinyl) poly(hexafluoropropylene oxide) methylene benzoate [(C₆H₅)₂P(O)(C ₆H₄C(O)OCH₂CF(CF₃)(OCF ₂CF(CF₃))_nF]

Full text of DOI:10.1016/j.jfluchem.2012.09.001

Journal of Fluorine Chemistry 144 (2012) 24–32
Contents lists available at SciVerse ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
The first fluorous biphase hydrogenation catalyst incorporating a
perfluoropolyalkylether: [RhCl(PPh2(C6H4C(O)OCH2CF(CF3)(OCF2CF(CF3))nF))3]
with n = 4–9
a,
b,1
Chadron M. Friesen a, , Craig D. Montgomery , Sebastian A.J.U. Temple
*
*
a Department of Chemistry, Trinity Western University, Langley, BC, Canada V2Y 1Y1
b Department of Chemistry, Simon Fraser University, 8888 University Drive, Burnaby BC, Canada V5A 1S6
A R T I C L E I N F O
A B S T R A C T
Article history:
The phosphine oxide (4-diphenylphosphinyl) poly(hexafluoropropylene oxide) methylene benzoate
[(C6H5)2P(O)(C6H4C(O)OCH2CF(CF3)(OCF2CF(CF3))nF] (with n = 4–9) 1, was prepared by the reaction of
[4-diphenylphosphinyl] benzoyl chloride with poly(hexafluoropropylene oxide) (pHFPO) methylene
alcohol and triethylamine. Subsequent reduction of 1 with HSiCl3 produced the phosphine 2, (4-
diphenylphosphino) pHFPO methylene benzoate, [(C6H5)2P(C6H4C(O)OCH2CF(CF3)(OCF2CF(CF3))nF]. The
phosphine ligand 2 was incorporated into tris(4-diphenylphosphino) pHFPO methylene benzoate
rhodium chloride 4. Partition coefficients in perfluoromethylcyclohexane (PFMCH):toluene were
determined using 19F NMR spectroscopy for both 2 and 4 as 98:2 and 81:19, respectively. Compound 4
was also shown to function as a fluorous biphase catalyst, catalyzing the hydrogenation of 2-cyclohexen-
1-one in biphasic (1:1 toluene:PFMCH; 1:3:3 toluene:hexanes:Krytox1 K6 [F(CF(CF3)CF2O)5CF2CF3])
and monophasic (1:3:3 toluene:hexanes:PFMCH) solvent systems with average turnover frequencies of
30.7, 17.1 and 20.6 hÀ1, respectively. Rh leaching studies were undertaken to confirm the recycling
ability of the catalyst 4 and the average percentage loss of Rh per cycle in the three solvent systems (1:1
toluene:PFMCH; 1:3:3 toluene:hexanes:Krytox1 K6 [F(CF(CF3)CF2O)5CF2CF3]; 1:3:3 toluene:hexa-
nes:PFMCH) was determined to be 0.35%, 0.17% and 0.30%, respectively.
Received 10 July 2012
Received in revised form 29 August 2012
Accepted 3 September 2012
Available online 10 September 2012
Keywords:
Fluorous Biphase Catalysis (FBC)
Poly(hexafluoropropylene oxide)
Perfluoropolyalkylethers (PFPAEs)
19F-NMR measured partition coefficients
Triarylphosphine based ligands
Hydrogenation catalysis
MALDI-TOF-MS characterization
ß 2012 Published by Elsevier B.V.
1. Introduction
solvents through non-covalent interactions [6–8]. In this paper,
consideration will be given to pHFPO and the extension of its
Perfluoropolyalkylethers (PFPAEs) are polymeric or oligomeric
compounds that consist of repeat units such as –[CF2CF2CF2O]x–,
–[CF2CF2O]x–, –[CF2O]x–or, as in this paper, –[OCF2CF(CF3)]x– and
are commercially available and recognized by such trade names as
Demnum1 (Daikin, Japan), Fomblin1 (Solvay Solexis) and Krytox1
(DuPont, USA) [1]. Examples of industrial applications of these
fluids include use as hard disk lubricants, high temperature
greases, vacuum pump fluids, hydraulic oils, aerospace jet engine
oils, satellite instrumentation bearing greases, antilock braking
system fluids, spark plug and boot lubricants, stone coatings,
cosmetic additives and ski wax substitutes [2–5]. Furthermore the
PFPAE employed herein, polyhexafluoropropylene oxide (pHFPO)
with the repeat unit –[OCF2CF(CF3)]x, has been used to extract
porphyrins, heterocyclic bases and metal complexes into fluorous
applications to include utilization in a fluorous triarylphosphine
suitable for Fluorous Biphase Catalysis (FBC).
While homogeneous catalysis offers many advantages over
heterogeneous catalysis, such as selectivity and efficiency, the
problem of separation and recycling of the catalyst remains a
significantone. The potentialofFBCasa solutiontothisproblem was
´
´
first proposed by Vogt in 1991 [9] and Horvath and Rabai in 1994
[10] and has advanced further since then, finding such applications
as hydroboration [11], hydrosilylation [12,13], hydroformylation
[10,14,15] and hydrogenation [16–19]. Furthermore, FBC has
become an area of considerable interest within green chemistry
[20–23] due to its focus on the recycling of catalysts. While the work
described herein employs fluorous solvents in FBC, it may be noted
that another green chemistry approach that has been investigated
elsewhere is to utilize perfluoro-tagged catalysts in non-fluorous
solvents such as methanol [24] and super-critical CO2 [25].
FBC requires that the catalyst be preferentially soluble in the
* Corresponding authors. Tel.: +1 604 888 7511; fax: +1 604 513 2018.
E-mail addresses: chad.friesen@twu.ca (C.M. Friesen), montgome@twu.ca
(C.D. Montgomery).
fluorous phase of
a biphasic system, which then becomes
monophasic at elevated temperatures (Fig. 1). Subsequent cooling
returns the system to a biphasic state for facile separation.
1
Trinity Western University, Canada.
0022-1139/$ – see front matter ß 2012 Published by Elsevier B.V.
http://dx.doi.org/10.1016/j.jfluchem.2012.09.001

C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
25
An alternate strategy would be to employ a different fluorous
moiety in the ponytail. Due to the potential effect on properties
such as solubility, there has been interest expressed previously in
the incorporation of heteroatoms such as oxygen into the
ponytails, as in the case of a perfluoropolyalkylether (PFPAE)
[39]. More specifically, the work reported herein was motivated by
the thought that more flexible ponytails might lead to more
favorable partition coefficients and that perhaps the desired
flexibility might be obtained by employing PFPAE ponytails rather
than perfluorotelomers. The addition of ether linkages disrupts the
ability of the polymer to crystallize and allows for the polymer to
remain liquid over longer chain lengths and broad temperature
ranges. Specifically, pHFPO may be highly suited to this application
due to its stability towards oxidation and thermal degradation; it is
stable with low volatility and low temperature dependence of
viscosity from À100 to 400 8C [40]. While PFPAEs have been
incorporated into (non-phosphine) catalysts [41] and have been
appended to (non-catalytic) phosphines [30,42,43], never have
PFPAEs been incorporated into phosphines which were then
employed for catalysis. As such this is the first report of
incorporation of a PFPAE into a phosphine ligand to be employed
in a catalyst.
Fig. 1. The thermomorphic solubility of a FBC system.
In order to obtain the desired solubility properties of the
catalyst, both trialkyl- and triaryl-phosphine ligands have been
adapted to include perfluoroalkyl segments or fluorous ‘ponytails’.
However, the electron-withdrawing nature of the fluorous pony-
tails can also have the effect of lessening the coordinating ability of
the phosphine ligand and can therefore require the presence of an
‘insulator’ or ‘spacer’ such as alkyl, silyl or –(OCH2)n– groups.
Fig. 2 provides a sampling of various representative fluorinated
phosphine ligands that have been prepared [16,18,24,26–36].
While considerable efforts have been made investigating the
utility of various spacers, little consideration has been given to
alternate fluorous ponytails, other than perfluoroalkyl chains. Such
–(CF2)x– chains (x = 3–10 typically) or telomers as they are also
known, are typically derived from poly(tetrafluoroethylene)
(pTFE). However a problem arises with such ponytails that when
the chains are short they tend to be too soluble in the organic
phase, while longer chains can display an undesired crystallinity
[37], which lowers the absolute solubility in both phases. One
attempt to deal with the problem of crystallinity has been to use
Teflon1 tape as means of catalyst recovery [38].
Herein we report:
a) the synthesis and characterization of
a triarylphosphine
ligand that incorporates the perfluoropolyalkylether pHFPO
with a –C(O)OCH2– spacer;
b) the preparation of a Wilkinson’s catalyst derivative using this
ligand;
c) partition coefficients for the ligand and the hydrogenation
catalyst, along with Rh leaching studies;
d) its catalytic activity in the hydrogenation of 2-cyclohexen-1-
one and the efficient recycling of the catalyst using fluorous
biphase conditions.
The solvent systems used in the partition coefficient studies as
well as the specific reaction conditions employed in the catalytic
studies are similar to previous studies [16]; this was done in order
to establish clear comparisons between perfluoroalkyl and
perfluoropolyalkylether ‘ponytails’ regarding their potential utility
in FBC.
2. Results and discussion
2.1. Ligand synthesis
The synthetic route to the ligand (4-diphenylphosphino) pHFPO
methylene benzoate 2 is shown in Scheme 1.
Starting from commercially available diphenyl [(4-methyl)-
phenyl] phosphine, the carboxylic acid phosphine oxide (diphenyl
[(4-carboxy)phenyl] phosphine oxide) is obtained by oxidation
with KMnO4, in the presence of NaOH [44]. In addition to the
formation of the carboxylic acid in preparation for esterification, it
was also necessary to oxidize the phosphorus centre due to its
nucleophilicity. The compound was obtained in 85–97% yield and
characterized by 1H, 13C and 31P{1H} NMR, as well as FT-IR. IR
bands at ꢀ2700 cmÀ1 (O–H str.), 1706 cmÀ1 (C55O str.) and
1088 cmÀ1 (C–O str.), along with the deshielded 13C nucleus at
166.8 ppm in the 13C NMR confirmed the presence of the
carboxylic acid, while the shift in the 31P{1H} NMR signal to
26.04 ppm (DMSO-d6) confirmed that the phosphine had also been
oxidized.
Subsequently, the carboxylic acid was converted to the acid
chloride, 4-diphenylphosphinyl benzoyl chloride using SOCl2
[45]. The FT-IR of the product no longer exhibited a band at ꢀ2700
due to the –OH stretch, while bands were observed at 1774
Fig. 2. Sampling of previously synthesized fluorous phosphine ligands. (I) Ref. [26],
(II) Refs. [18,34,35], (III) Refs. [32,33], (IV) Refs. [16,27–29], (V) Refs. [30,31], (VI) Ref.
[27], (VII) Ref. [30], (VIII) Ref. [24], (IX) Ref. [36].

26
C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
Scheme 2. Reaction to form the pHFPO ester analogue, 3.
reduction was confirmed by the 31P{1H} in which the signal moved
upfield to À3.75 ppm (in CDCl3).
2.2. Synthesis of the Rh(I) hydrogenation catalyst
The synthesis of the Wilkinson’s-type catalyst followed the
biphasic method whereby a solution of 2 in perfluoromethylcy-
clohexane (PFMCH) was stirred for one day under N2 with a
toluene solution of [RhCl(COD)]2 (Scheme 3).
Characterization of the resulting complex 4, [RhCl(PPh2
(C6H4RF))3] (where RF = C(O)OCH2CF(CF3)(OCF2CF(CF3))nF with
n = 4–9), a red oil, was accomplished primarily through 31P{1H}
NMR. An AB2X pattern consisting of a doublet of doublets at
2
31.7 ppm (1JPRh = 143.1 Hz, JPP = 36.4 Hz, 2P) and a doublet of
Scheme 1. Synthetic route to (4-diphenylphosphino) pHFPO methylene benzoate 2.
and 1739 cmÀ1 corresponding to the C55O stretch of the acid
chloride.
It was hoped that the acid chloride could then be readily
converted to the ester by reaction with pHFPO-CH2OH, but the
feasibility of this reaction was first tested using a non-phosphine
benzyl analogue. Benzoyl chloride was found to react with pHFPO-
CH2OH to give the ester, pHFPO methylene benzoate 3 (Scheme 2).
No band corresponding to an O–H stretch appeared in the IR, but a
C55O stretch at 1745 cmÀ1 was still present. In addition, a
2
characteristic JC–F coupling of 33 Hz was observed in the 13C
NMR due to the methylene carbon, now appearing as a doublet at
58.39 ppm.
Given the success of the model reaction, pHFPO-CH2OH (having
an average molecular weight of 1150 g/mol with a homologue
distribution of n = 4–9 based on GC/MS and 19F NMR analysis) was
then reacted with the acid chloride, 4-diphenylphosphinyl benzoyl
chloride. This resulted in the ester, (4-diphenylphosphinyl) pHFPO
methylene benzoate 1. As in the case of the model reaction, the IR
spectrum was lacking in a band due to an O–H stretch, while the
C55O stretch appeared at 1745 cmÀ1. Again in the 13C NMR
spectrum, the methylene carbon signal appeared as a doublet at
2
59.59 ppm with a JC–F coupling of 31.4 Hz. There was only one
signal in the 31P{1H} NMR spectrum, a singlet at 25.89 ppm (in
CDCl3), consistent with a triarylphosphine oxide. Also MALDI-TOF-
MS corroborated the lithium salts of (C6H5)2P(C6H4C(O)OCH2CF(C-
F3)(OCF2CF(CF3))nF with n = 4–9 having peaks at 1124.9, 1290.9,
1456.8, 1622.8, 1788.8, 1954.8 g/mol.
The phosphine oxide 1 was then reduced to the desired ligand,
(4-diphenylphosphino) pHFPO methylene benzoate 2, a pale
yellow viscous liquid, using HSiCl3 and NEt3 in toluene. The
Scheme 3. Synthesis of the fluorous Rh(I) hydrogenation catalyst 4 and the
dimerization of 4.

C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
27
2
triplets at 47.66 ppm (1JPRh = 187.8 Hz, JPP = 35.2 Hz, 1P) was
observed. However
above where Deelman [34] reports P values higher than those
reported herein, it should be noted that those phosphine ligands
possess 6, 9, 12 or 18 flourous ponytails per phosphine in contrast
to the one ponytail per phosphine in the ligand reported here.
While fluorous phosphine ligand 2 does not display the highest
partition coefficient yet observed, nevertheless when allowing for
the relatively low fluorous content of 2, these results indicate that
phosphine ligands with perfluoropolyalkylether ‘ponytails’ do
indeed have significant potential for improving partition coeffi-
cients in FBC presumably due to their flexible nature.
The partition coefficient for catalyst 4 was 81:19. To observe a
decrease in the partition coefficient as the ligand coordinates to a
metal centre, is not entirely unexpected since the fluorous content
of the complex is less than that of the free ligand. Previous studies
[47] have similarly suggested that in some cases the partitioning
into the fluorous phase decreases upon complexation of the ligand
to form the catalyst.
Of course it is important to note that the ultimate goal is not
merely high partition coefficients, but the ability of the catalyst to
be recycled and (as discussed below) catalyst 4 did in fact
demonstrate good recycling ability. As for the apparent discrep-
ancy between the lower partition coefficient values of the complex
and its good recycling ability, one should note that the partition
coefficients are determined using toluene and PFMCH but when
the recycling of the catalyst takes place, the product molecule is
also present in the organic phase; this may also alter the affinity of
the catalyst for the organic phase. As a result, it was thought that
rhodium leaching studies would be superior to partition coefficient
determinations, as a method of determining the recycling ability of
the catalyst. These studies are discussed later in the paper.
a doublet appearing at 51.51 ppm (d,
1JPRh = 189.98 Hz) was suggestive of an equilibrium between 4
and a chloro-bridged dimer as shown in Scheme 3, as has been
observed previously [16]. The signal due to the dimer appears in the
room temperature 31P{1H} NMR spectrum in a ratio of approxi-
mately 1:12 with the monomer, but when the system is cooled to
273 K, the dimer signal is no longer present.
An excess of ligand 2 was used in this reaction, with any
unreacted ligand being oxidized to the ligand oxide, 1. It is not
possible to remove this excess ligand oxide from the resulting
product mixture, however it does not interfere with the catalysis.
2.3. Partition coefficients
Partition coefficients, indicating the relative affinity of the
catalyst for the fluorous solvent, are obviously useful predictors in
the effectiveness of a fluorous biphase catalytic system. It has been
argued that a complex must have a fluorine content of >60% in
order to display a preference for the fluorinated solvent [46].
However the results of Hope and co-workers [47] would suggest
that the percentage of fluorous content is not as critical as the
extent to which the fluorous portion of the compound is able to
surround the organic core, an ability that presumably would be
enhanced by using more flexible perfluoropolyalkylether ‘pony-
tails’ as in this study.
Partition coefficients were obtained by 19F NMR spectroscopy at
25 8C in PFMCH/toluene; this solvent system and temperature
were employed elsewhere [48] and was therefore chosen here for
purposes of comparison. The results are shown in Table 1.
The ligand 2 displays a high ratio of 98:2 (P = 49) as it partitions
between the fluorous and organic solvents respectively. This result
compares favorably with those reported previously for ligands
with perfluoroalkyl ‘ponytails’ Rf (Rf = CnF2n+1, n = 6, 8, 10) in the
same solvent system (PFMCH/toluene), whether trialkyl phos-
phines [Rf(CH2)x]3P (x = 2–5), (with values ranging from 98.8:1.2 to
>99.7:<0.3) [26,49] triarylphosphines with alkyl spacers
[Rf(CH2)xC6H4]nPPh3Àn (x = 2, 3; n = 6,8,10) (with values ranging
from 19.5:80.5 to 66.6:33.4) [16,25,27] or triarylphosphines with
silyl spacers [Ph3ÀnP(C6H4-p-SiMe3Àx(CH2CH2Rf)x)n] (x = 1, 2, 3;
n = 4, 6, 8, 10) (with values ranging from 21:79 to 89:<11)
[18,34,35]. Deelman reports higher P values for triarylphosphines
with silyl spacers (n = 4, 6, 8, 10) where the aryl ring substitutions
are in the meta position or 3,5-disubstitution (with values of up to
>238) [34].
2.4. Catalytic hydrogenation and catalyst recycling
Once again for purposes of the comparison of perfluoropolyalk-
ylether ‘ponytails’ with those featuring perfluoroalkyls, the
hydrogenation reaction that was chosen to test the catalytic
ability of 4 was the same as that employed in previous studies
employing perfluoroalkyl moieties [16].
The hydrogenations were done using three solvent systems: 1:1
toluene:PFMCH and 1:3:3 toluene:hexanes:Krytox1 K6
[F(CF(CF3)CF2O)5CF2CF3] that remain biphasic even at high
temperatures and a solvent system (1:3:3 solution of toluene:hex-
anes:PFMCH) that becomes monophasic at 36.5 8C (as illustrated in
Fig. 3). The results are shown in Table 2, including cycle by cycle
TOF and average TOF values.
Comparing the partition coefficients in these systems, it is
noteworthy that ligand 2 does compare favorably with the
triarylphosphines having perfluoroalkyl ‘ponytails’ with either
alkyl or silyl spacers. This is true despite ligand 2 having a lower
fluorous content compared with these other ligands, and only one
fluorous ‘‘ponytail’’ per phosphine rather than more than one, such
as ligands I–VII and IX in Fig. 2. For instance in the results cited
It may be noteworthy that in the biphasic 1:1 toluene:PFMCH
system, the t99% value for the first cycle is 22.22 h while the value
drops significantly in subsequent cycles 2–6. It may be that this is
indicative of an inductive period in which the actual catalyst is
formed from the catalyst precursor.
In this study, catalyst 4 displayed turnover frequencies (TOF99%
;
product/catalyst mole ratio per hour calculated at 99% completion)
Table 1
Partition coefficients for the ligand 2 and the hydrogenation catalyst 4.
Solute
Solvent system
Method
Percent partitioning
fluorous:organic (P)
[4-Poly(HFPO)nCH2OC(O)C6H4P(C6H5)2]3RhCl n = 4–9
(Fluorine in catalyst = 52%)
CF3C6F11
:
19F-NMR 25 8C
19F-NMR 25 8C
19F-NMR 25 8C
81:19 (Æ3)
Average MW ꢁ 4452 g/mol
CH3C6H5
(4.26)
4-Poly(HFPO)nCH2OC(O)C6H4P(C6H5)2 n = 4–9
(Fluorine in ligand = 54%)
CF3C6F11
:
98:2 (Æ1)
Average MW ꢁ 1438 g/mol
CH3C6H5
(49)
4-Poly(HFPO)nCH2OC(O)C6H4P(O)(C6H5)2 n = 4–9
(Fluorine in ligand oxide = 53%)
Average MW ꢁ 1454 g/mol
CF3C6F11
:
96:4 (Æ3)
CH3C6H5
(24)
Note: Percent partitioning was done in triplicate.

28
C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
Fig. 3. Catalytic hydrogenation of cyclohexenone under biphasic (emulsion) and monophasic conditions. (a) Reaction 1 utilizes 1:1 toluene:PFMCH as the solvent system,
Reaction 2 utilizes 1:3:3 toluene:hexanes:
Krytox1K6 [F(CF(CF3)CF2O)5CF2CF3] as the solvent system. (b) 1:3:3 solution of toluene:hexanes:PFMCH.
Table 2
Hydrogenation data summary for monophasic and biphasic conditions.
Complex
Conditions
Cycle
TOF99%
(hÀ1
t99% (h)
AVG. TOF
Æ std. dev.)
)
(
Tris (4-diphenylphosphino pHFPO methylene benzoate) rhodium chloride
Biphasic
1
2
3
4
5
6
1
2
3
1
2
8.15
34.53
25.23
40.17
30.48
23.45
17.82
18.61
14.80
21.83
19.42
22.22
5.30
7.18
4.51
5.94
7.73
10.17
9.73
12.24
8.30
9.33
30.7 Æ 6.8
1:1 toluene: PFMCH
Tris (4-diphenylphosphino pHFPO methylene benzoate) rhodium chloride
Tris (4-diphenylphosphino pHFPO methylene benzoate) rhodium chloride
Biphasic
17.1 Æ 2.0
20.6 Æ 0.7
1:3:3 toluene:hexanes:
Krytox1K6 [F(CF(CF3)CF2O)5CF2CF3]
Monophasic
1:3:3 toluene:hexanes:PFMCH
of 30.7 hÀ1 in toluene:PFMCH, 17.1 hÀ1 in toluene:hexanes:Kry-
tox1K6 and 20.6 hÀ1 in the monophasic toluene:PFMCH:hexanes
system. Interestingly, the TOF values were superior in the biphasic
toluene:PFMCH system compared to the monophasic tolue-
ne:PFMCH:hexanes system; Gladysz and co-workers [16] similarly
reported higher TOF values in the biphasic system compared with a
monophasic one. Nevertheless compound 4 does act as a true
Fluorous Biphase Catalyst (Fig. 3) in toluene:PFMCH:hexanes and
thereby demonstrates ‘‘proof of concept’’ and the successful
application of PFPAEs in phosphine ligands for FBC catalysts.
TheTOF99% valuesforcatalyst4reportedinthisstudyat45 8Cand
1 atm (30.7 hÀ1 in toluene:PFMCH; 17.1 hÀ1 in toluene:hexane-
s:Krytox1K6; 20.6 hÀ1 in the monophasic toluene:PFMCH:hexanes
system) are comparable to both fluorous and non-fluorous
Wilkinson’s catalysts for the hydrogenation of cyclohexen-1-one.
That is, when comparing to fluorous catalysts employing perfluor-
opolyalkyl ‘ponytails’ the TOF99% values reported herein in biphasic
conditions are less than those reported by Gladysz and co-workers
[16] (47.5 hÀ1 in biphasic conditions at 45 8C and 1 atm H2).
However the values in this study obtained in monophasicconditions
are in fact higher than those reported elsewhere for catalysts
employing perfluoropolyalkyl analogues (8.25 hÀ1 in monophasic
conditions at 45 8C and 1 atm H2) [16]. When comparing to
non-fluorous catalysts, the TOF value reported by Bezuidenhoudt
and co-workers was somewhat higher (78 hÀ1). However in that
case, a higher temperature (80 8C) and pressure (10 bar) was
employed [50].
As important as the TOF values, is the recycling ability of the
catalyst since that is the critical feature of FBC. The system reported
herein shows potential with respect to the recycling of the catalyst.
Whether in biphasic or monophasic conditions, the catalyst was
recycled and run through multiple hydrogenation cycles while
exhibiting no significant trend of diminishing catalytic ability. This
would seem to suggest that leaching of the catalyst into the organic
phase is not significant here. For example, the highest TOF values
were obtained in the final (third) cycle of the biphasic toluene:K-
rytox1K6:hexanes, and in the fourth cycle of the biphasic
toluene:PFMCH system. When considering the standard deviations

C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
29
Table 3
Determination of Rh leaching into the organic phase.
Conditions
Cycle
1–8a
Rh in organic
% Rh loss
2.8
Avg. % Rh loss
phase (
mg)
per cycle (Æ std. dev.)
Biphasic
11.96
0.35
1:1 toluene:PFMCH
Biphasic
1
2
3
1
2
1.164
0.522
0.52
0.27
0.12
0.12
0.49
0.12
0.17 (Æ 0.09)
0.30 (Æ0.26)
1:3:3 toluene:hexanes:Krytox1K6 [F(CF(CF3)CF2O)5CF2CF3]
Monophasic
2.102
1:3:3 toluene:hexanes:PFMCH
0.5185
a
The organic phases from cycles 1–8 were combined in the 1:1 toluene:PFMCH biphasic system.
for the average TOF values, it should be noted that this study
involves a homologue distribution (which varies from batch to
batch due to the industrial manufacturing process). This may lead
to consistency issues such as microdroplet size (in the case of the
biphase systems). Nevertheless very little apparent diminishing of
catalytic ability was observed with repeated recycling.
To further confirm the recycling ability of the catalyst, leaching
studies were undertaken. After each catalytic cycle, the organic
phase was removed and the solvent evaporated. The residue was
then analyzed for rhodium by ICP-AAS. The results are shown in
Table 3.
These results confirm that very little catalyst is lost into the
organic phase with each catalytic cycle; the average percentage
loss of Rh per cycle ranges from 0.17% to 0.35% in the three systems
employed. Indeed the loss of Rh is significantly lower than that
predicted by the partition coefficients, likely due to the fact that
the conditions for partition coefficient determinations do not
exactly match those of the catalytic cycles. The extent of Rh
leaching found here using PFPAE ponytails (0.17–0.35% per cycle)
is also comparable to that reported by Deelman and co-workers
using perfluoroalkyl ‘ponytails’ (0.1–0.3%) [18,19].
Hydrogen gas (Praxair, 5.0) was used as received. The identity of
molecular compounds was checked by their solution-state NMR
spectra, IR spectra where appropriate as well as mass spectrome-
try. Bruker AVANCE III 400 MHz, Bruker AVANCE III 500 MHz as
well as Bruker AVANCE II 600 MHz ‘‘QNP 600’’ running TopSpin 2.1
were employed as required for obtaining solution-state NMR. For
gas chromatography/mass spectrometry (GC/MS) analyses of the
hydrogenation products an Agilent Technologies 6890N GC was
coupled with an Agilent Technologies 7638B series injector and
Agilent Technologies 5975B inert mass spectrometer detector
(MSD) was employed with electron impact (EI) as the mode of
ionization. All starting materials were commercially available, if
not described below. Matrix assisted laser desorption ionization-
time-of-flight mass spectrometry (MALDI-TOF-MS) experiments
were determined with a Bruker Ultraflex with a positive ionization
method. For sample preparation, samples were 1/50 diluted in
perfluorokerosene. 1
on the target first and dried, then 0.5
spotted on top of the LiCl layer and dried, and finally 0.5
m
L of LiCl/MeOH (10 mg/mL) was deposited
L of sample solution was
L of
m
m
a-
cyano-4-hydroxycinnamic acid (CCA) matrix was applied on top
and dried. IR (attenuated total reflectance, ATR) spectra were
recorded with a Thermo Nicolet 380 FT-IR instrument using ZnSe
or Germanium crystals. All IR (KBr DRIFT) spectra were recorded
with a Mattson Galaxy Series FT-IR 3000. The ICP-AAS analyses for
the Rh leaching studies were performed by Canadian Microana-
lytical Services, Ltd.
3. Conclusions
For the first time, perfluoropolyalkylether moieties have been
employed in a triarylphosphine ligand for Fluorous Biphase
Catalysis. The partition coefficient values compare favorably to
analogous ligands employing perfluoroalkyl ‘ponytails’, likely due
to the flexibility of the perfluoropolyalkylethers. The resulting
Wilkinson-type Rh(I) complex is thus not only able to catalyze the
hydrogenation of 2-cyclohexen-1-one, but it also displays ability to
be recycled. Rhodium leaching studies further support the strong
ability of this catalyst to be recycled. While the partition coefficient
and turnover frequency results reported herein, may not represent
significant improvements on the state of the art, nevertheless they
do suggest potential for such improvements with the introduction
of a new class of Fluorous Biphase catalysts, those employing
perfluoropolyalkylether ‘ponytails’. It is hoped that other triar-
ylphosphine ligands with other PFPEs, with more than one
‘ponytail’ per phosphine and having varying perfluoropolyalkyl
ether lengths, fluorous content, insulator groups, and points of
attachment will allow for the fine-tuning of these catalysts.
4.2. Preparation of diphenyl [4-carboxyphenyl] phosphine oxide
In a typical synthesis KMnO4 (22.161 g, 140 mmol) was added
in
4 portions to a solution of diphenyl [(4-methyl)phenyl]
phosphine (Aldrich, 96%) (10.083 g, 32 mmol), water (130 mL)
and NaOH (2.220 g, 56 mmol). The mixture was allowed to reflux
for 12 h, after which point the hot brown suspension was filtered.
The residual MnO2 was washed with hot water. The filtrate was
acidified with 50% H2SO4 precipitating out the crude white product
that was collected and dissolved in 10% NaOH and extracted with
Et2O to remove residual starting material. The aqueous phase was
again treated with 50% H2SO4 re-precipitating the desired white
product. The product was collected via filtration and dried in a
vacuum oven at 75 8C yielding the product as a white powder
(9.980 g, 31 mmol, 97%). Alternatively: p-tolyldiphenylphosphine
(Aldrich, 96%) (2.142 g, 7.75 mmol), pyridine (15 mL) and H2O
(50 mL) were added to a 250 mL three neck round-bottom flask
equipped with magnetic stir bar, condenser and N2 blanket and
heated to approximately 70 8C. KMnO4 (6.224 g, 39.4 mmol) were
added in 5 small portions at 30 min intervals. Following the last
addition, solution temperature was raised to 90 8C and allowed to
stir for 20 h (reflux). The hot suspension was filtered, and the
residual MnO2 washed with hot water. The filtrate was cooled to
room temperature and acidified in an ice bath using concentrated
HCl to precipitate the crude product. 1H NMR indicated incomplete
oxidation and so was reoxidized in water (60 mL), KOH (0.484 g,
4. Experimental
4.1. General
All non-aqueous reactions were carried out under an atmo-
sphere of nitrogen using oven- or flame-dried glassware, unless
otherwise noted. All catalytic reactions were degassed prior to
admitting hydrogen and were verified to be fully reproducible
through several repetitions. Solvents were dried using standard
methods, and oxygen was removed via freeze-pump-thawing.

30
C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
1
8.62 mmol) and KMnO4 (1.630 g, 10.3 mmol) and allowed to
reflux overnight. The suspension was filtered while hot and the
resulting filtrate was cooled on an ice bath and acidified using
concentrated HCl precipitating out the crude product. The white
solid was filtered using Whatman #40 filter paper and washed
with cold deionized water affording the product (2.018 g,
6.26 mmol, 80%) as white crystals. 1H NMR (600 MHz, DMSO-
(s, C55O), 139.08 (d, JCP(ipso) = 100.0 Hz), 132.40 (m, not clearly
resolvable, overlapping ortho and para signals), 132.03
(d, JCP(ortho) = 10.1 Hz), 132.02 (d. br. JCP(para) = 2.9 Hz), 131.48
(d, JCP(ipso) = 103.8 Hz), 129.66 (d, JCP(meta) = 12.0 Hz), 128.72
2
4
1
3
3
1
2
(d, JCP(meta) = 12.2 Hz), 117.35 (qd, JCF = 288.3, JCF = 31.0,
–OCF2CF(CF3)–), 115.75 (td,
1JCF = 287.8, 2JCF = 31.9,
–OCF2CF(CF3)–), 106.88 (dq,1JCF = 251.6 Hz, 2JCF = 35.6 Hz,
–
d6, 25 8C):
d
= 8.11 (dd, 3J = 8.5 Hz, 4J = 2.3 Hz, 2H), 7.78 (dd,
CH2CF(CF3)O–) 102.37 (ds, 1JCF = 272.8 Hz, 2JCF = 38.0 Hz,
3JHP = 11.4 JHH = 8.5 Hz, 2H), 7.71–7.65 (m, 6H) 7.55 (td,
–CF(CF3)O–), 60.27 (d, JCF = 31.2 Hz, –CH2CF(CF3)O–); 19F NMR
3
2
3J = 8.3 Hz, 4J = 3.0, 4H); 13C NMR (151 MHz, DMSO-d6, 25 8C):
(376.41 MHz, CDCl3, 25 8C):
(–CH2CF(CF3)–), À134.73, À131.04 (–OCF2CF2CF3), À84.26, À84.12,
À82.81 (–CF2–), À81.44(–CF3); IR (ATR) = 3049 (arom. C–H), 1747
(C55O), 1308 (C–F), 1232, 1198, 1118, 985, 806, 730, 696
(st, mono-sub. arom. C–H out-of-plane bend) cmÀ1
d
= À146.22 (–OCF2CF(CF3)–), À135.18
1
d
= 166.61 (s, C55O), 137.32 (d, JCP(ipso) = 99.7 Hz), 133.86 (d,
4JCP(para) = 2.7 Hz), 132.22 (d, 4JCP(para) = 2.4 Hz), 132.03 (d, 1JCP(ip-
n
2
so) = 103.2 Hz), 131.79 (d, JCP(ortho) = 10.0 Hz), 131.48 (d,
2JCP(ortho) = 9.8 Hz), 129.36 (d, JCP(meta) = 11.9 Hz), 128.83 (d,
3
.
3JCP(meta) = 11.8 Hz); 31P{1H} NMR (242.88 MHz, DMSO-d6,
25 8C):
d
26.04 (s, P55O); FT-IR (KBr DRIFT):
n
= 3060 (aromatic
4.5. Preparation of (4-diphenylphosphino) pHFPO methylene
benzoate (2)
C–H stretch); 2885, 2764, 2597, 2475 (COOH split OH stretch);
1706 (C55O stretch); 1255 (P55O), 1156, 1109, 695 (st, mono-sub.
arom. C–H out-of-plane bend) cmÀ1
.
To a stirred solution of toluene (35 mL) and HSiCl3 (0.557 g;
0.415 mL; 4.11 mmol) under N2, (4-diphenylphosphinyl) pHFPO
methylene benzoate (1.196 g; 0.0822 mmol) was added. Triethy-
lamine (0.633 mL; 0.459 g; 4.54 mmol) was added via syringe and
the solution was heated to reflux for 5 h. Following the reflux
period, the excess HSiCl3 as well as approximately half of the
toluene was removed under reduced pressure. Saturated NaHCO3
(3 mL) was added and the mixture was allowed to stir for 5 min,
and filtered through diatomaceous earth (Celite). The RBF and
residual SiO2 were washed with several portions of toluene
(ꢀ10 mL) and Freon I E-fluid [F(CF(CF3)CF2O)CFHCF3] fluorinated
solvent. The fluorous phase was collected and dried over MgSO4
and the solvents were removed in vacuo. The reduced product 2
was collected as a pale yellow viscous liquid (80–84% based on
average molecular weight). 1H NMR (600 MHz, CDCl3, 25 8C):
4.3. Preparation of [4-diphenylphosphinyl] benzoyl chloride
The synthesis was accomplished according to the method of El-
Deek et al. [45]. To a stirred solution of DCM (30 mL) and diphenyl
[4-carboxyphenyl] phosphine oxide (0.642 g, 1.99 mmol), SOCl2
(0.416 g, 3.50 mmol) was added via syringe and refluxed for 5 h.
Following 3 h of reflux, the solution changed from a turbid white
suspension to a clear colorless solution. The solution was allowed
to cool to room temperature and evaporated under nitrogen and
finally the volatiles pumped off under reduced pressure affording
the product as an off-white solid (0.671 g, 1.97 mmol, 99%). 1H
NMR (600 MHz, CDCl3, 25 8C)
d = 7.52 (td, J = 7.8, 2.7 Hz, 4H), 7.61
(t, J = 7.5 Hz, 2H), 7.68 (dd, J = 12.2, 7.5 Hz, 4H) 7.87 (dd,
3
3
4
3JHP = 11.3 Hz, JHH = 8.4 Hz, 2H), 8.23–8.19 (d, 2H); 13C NMR
d
= 4.89 (dq, JH–F = 26.3, JH–F = 13.4, 2H, –OCH2CF(CF3)–), 7.38
1
(151 MHz, CDCl3, 25 8C):
d
= 167.93 (s, C = O), 139.96 (d, JCP(ip-
(m), 7.51 (m), 7.61 (m), 7.70 (m), 7.84 (m), 7.98 (m), 8.14 (m); 13
C
4
4
so) = 99.1 Hz), 136.07 (d, JCP(para) = 2.8 Hz), 132.68 (d, JCP(par-
a) = 8.3 Hz), 132.38 (d, 1JCP(ipso) = 102.2 Hz), 132.20 (d,
NMR (151 MHz, CDCl3) d 164.50 (s, C55O), 146.09 (d, J = 15.6 Hz,
unassigned), 135.95 (d, 1JCP(ipso) = 10.6 Hz), 134.07 (d,
2JCP(ortho) = 10.3 Hz), 132.08 (d, JCP(ortho) = 10.1 Hz), 130.96 (d,
2JCP(ortho) = 20.2 Hz), 133.29 (d, JCP(ortho) = 18.6 Hz), 132.40 (s,
2
2
3JCP(meta) = 12.0 Hz), 128.89 (d, JCP(meta) = 12.5 Hz); 31P{1H} NMR
CP(para)), 132.09 (d, JCP(ipso) = 10.0 Hz), 129.70 (d, JCP(me-
3
1
3
3
(242.88 MHz, CDCl3, 25 8C):
= 3055, 1774 (C55O), 1739 (C55O), 2846, 1593, 1437 (st), 1393 (P-
Phenyl), 1202 (st), 1116 (st), 721 (st), 695 (st, mono-sub. arom. C–H
out-of-plane bend), 539 cmÀ1
d
= 29.954 (s, P55O); IR(KBr Drift):
ta) = 6.1 Hz), 129.55 (s, unassigned), 128.77 (d, JCP(meta) = 7.1 Hz),
127.86 (s, CP(para)), 116.90 (qd, JCF = 289.5, JCF = 31.9,
OCF2CF(CF3)–)
115.42 (td, 1JCF = 289.04, 2JCF = 33.9,
1
2
n
–
–
.
OCF2CF(CF3)–), 107.69 (m, –CF2CF2CF3), 106.00, 102.14 (ds,
2
2
1JCF = 271.6 Hz, JCF = 37.40, –CF(CF3)–), 59.59 (d, JCF = 31.4 Hz);
31P{1H} NMR (242.92 MHz, CDCl3, 25 8C):
= À3.75 (s, P:), 28.82 (s,
P55O impurity); 19F{1H} NMR (471 MHz, CDCl3, 25 8C):
= À146.3
4.4. Preparation of (4-diphenylphosphinyl) pHFPO methylene
benzoate (1)
d
d
(–OCF2CF(CF3)–), À134.89 (–CH2CF(CF3)–), À131.25 –OCF2
CF2CF3), À84.38, À84.25, À82.98 (–CF2–), À81.60 (–CF3); IR
To a stirred solution of pHFPO methylene alcohol (4.00 g,
3.48 mmol) and triethylamine (1.795 g, 17.4 mmol), [4-diphenyl-
phosphinyl] benzoyl chloride (3.814 g, 11.2 mmol) and dry THF
(8 mL) were added yielding a peach colored solution. The solution
was allowed to stir for 12 h at room temperature and was then
diluted with 20 mL of Freon I E-fluid [F(CF(CF3)CF2O)CFHCF3]. The
resulting peach-coloured opaque mixture was extracted with
several portions of 70/30 solution of CH3CN/H2O until no more
color was apparent in the organic phase. The resulting peach colored
fluorous phase was collected, dried over MgSO4 and the solvent was
removed under reduced pressure affording 1 as a very viscous
colorless liquid (72%). MALDI-TOF-MS [M + 166n + Li]+ = 1124.9
(n = 4), 1290.9 (n = 5), 1456.8 (n = 6), 1622.8 (n = 7), 1788.8
(n = 8), 1954.8 (n = 9). 31P{1H} NMR (242.92 MHz, CDCl3, 25 8C):
(ATR):
n = 3060 (ar. C–H); 1741 (C55O), 1304, 1224, 1117 (pHFPO
C–F), 983, 803, 749, 696 (monosub. arom. C–H out-of-plane oop.
bend) cmÀ1
.
4.6. Preparation of pHFPO methylene benzoate (3)
To a stirred solution of pHFPO methylene alcohol (1.982 g,
1.72 mmol) and triethylamine (0.347 g, 3.43 mmol), benzoyl
chloride (0.472 g, 3.46 mmol) was added resulting in a thick white
paste. The mixture was extracted with several portions of 50/50
solution of H2O/CH3CN affording 3 as a viscous colorless liquid. 1H
3
NMR (60 MHz, NEAT, 25 8C):
d
= 4.2 (br. d, 2H, JH–F not clearly
resolved, –C(O)O–CH2CF(CF3)–)–, 6.74–7.36 (m, 5H, ArH); 13C NMR
2
d
= 25.89 (s, P55O); IR (ATR):
n
= 1745 (C55O), 1312, 1228, 1114, 981,
(15.089 MHz, NEAT, 25 8C):
d
= 58.39 (d, JC–F = 33 Hz.), {87.88,
806, 749, 710 cmÀ1 1H NMR (600 MHz, CDCl3, 25 8C): 4.91 (dq, 3JH–
F = 24.1 Hz, 4JH–F = 14.0 Hz, 2H, –OCH2CF(CF3)–), 7.5 (m, 4H), 7.59 (t,
3J = 6.5 Hz, 2H), 7.68 (m, 4H), 7.82 (t, 3J = 10.1 Hz, 2H) 8.12
90.07, 92.46, 95.41, 97.96, 100.47, 102.61, 106.98, 109.04, 110.74,
113.05, 115.48, 117.22} (F[CF(CF3)CF2O]–), 126.56, 128.30, 129.60,
132.83, 136.59, 145.21, 147.32, 163.82 (C55O) ppm; 19F NMR
(d, 3J = 8.9 Hz, 2H); 13C NMR (151 MHz, CDCl3, 25 8C)
d
= 163.85
(56.45 MHz, NEAT, 25 8C):
d
= À146.21 (–OCF2CF(CF3)–), –135.14

C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
31
(–CH2CF(CF3)–), –131.49 (–OCF2CF2CF3), À84.29, À83.53 (–CF2–),
À81.77 (–CF3); IR (ATR): = 1745 (C55O), 1312, 1228, 1114, 981,
806, 749, GC/MS (EI 70 eV): m/z = 235
710 cmÀ1
4.10. Catalytic hydrogenation
n
;
4.10.1. Biphasic conditions (1:1 toluene:PFMCH)
(C6H5C(O)OCH2CF(CF3)+), 169 (CF3CF2CF2+), 135 (C6H5C(O)O+),
105 (C6H5CO+), 77 (C6H5+), 69 (CF3+).
These conditions were selected to allow for comparisons with
´
Soos and Gladysz’ catalysts described earlier. The following is
representative and is illustrated in Fig. 3 (Reaction 1). In an inert
atmosphere glovebox, a glass 3-necked 6 mL reactor with J-young
valves equipped with a small magnetic stir bar and hydrogenation
balloon was charged with 2 mL of a degassed PFMCH solution
containing tris-pHFPO-modified triarylphosphine rhodium chlo-
ride (0.019 g, 4.20EÀ03 mmol, 0.5 mol%). To this solution, 2 mL of
4.7. Preparation of tris (4-diphenylphosphino) pHFPO methylene
benzoate rhodium chloride (4)
The synthesis was accomplished according to the procedure of
Gladysz et al. [16]. Degassed (4-diphenylphosphino) pHFPO
methylene benzoate (1.102 g, 0.766 mmol) was dissolved in
degassed PFMCH (10 mL) in a 50 mL 3-neck RBF in a glovebox
(Solution 1). Solution 2, a degassed solution of cyclooctadiene
dry, degassed toluene was added as well as 74
mL of degassed 2-
cyclohexen-1-one. The reactor was sealed and removed from the
glovebox and the contents degassed via freeze-pump-thaw
method. The system was backfilled with hydrogen and set in a
45 8C constant temperature water bath. The biphasic system was
vigorously stirred and the reaction’s progress was analyzed via GC/
MS (3.0 h, 50% conversion to cyclohexanone; 6.1 h, 99% conver-
sion). Upon completion of the reaction, the contents of the reactor
were degassed and returned to the glovebox. The organic phase
was removed via pipette and the vial recharged with a fresh aliquot
rhodium chloride dimer [Rh(COD)(m-Cl)]2 (0.0629 g, 0.091 mmol)
in toluene (7.5 mL) was added to solution 1 in the glovebox and
allowed to stir for 1 day. Following the stirring period, the upper
organic layer was decanted. The lower fluorinated phase was
washed once with 2 mL of dry degassed toluene and the volatiles
from the fluorinated layer were removed in vacuo affording 4 as a
red oil (52% yield by 31P{1H} NMR). 31P{1H} NMR (202.43 MHz,
1
2
CDCl3, 25 8C):
d
= 31.7 (dd; JPRh = 143.1 Hz, JPP = 36.4 Hz, 2P),
of substrate (74 mL) and toluene (2 mL). This was repeated through
47.66 (dt, 1JPRh = 187.8 Hz, 2JPP = 35.2 Hz, 1P), 51.51 (d,
6 cycles for a total of 1096 turnovers. A seventh cycle was run,
raising the temperature in an attempt to render the system
monophasic, however the catalytic activity dropped. An eighth
cycle was then employed to test if the heating had caused the
catalyst to degrade. These last two cycles were not included in the
TOF determinations.
1JPRh = 190.0 Hz); 13C NMR (100 MHz, CDCl3, 25 8C):
d = 60.13 (d,
2JC–F = 30.1 Hz); 128.3; 128.6; 128.8; 129.6; 129.7; 132.0; 132.1;
132.3; 132.4; 132.44; 163.9 (C55O); 1H NMR (400 MHz, CDCl3,
3
4
25 8C):
d = 4.82 (dq, JHF = 25.3 Hz, JHF = 13.3 Hz, 2H), 7.39–
7.44 ppm (m, 4H) 7.49–7.62 ppm (m, 6H), 7.71–7.76 (m, 2H),
8.01–8.04 (m, 2H); 19F{1H} NMR (376 MHz, CDCl3, 25 8C):
d
= À145.09 (m, –OCF2CF(CF3)–), À133.9 (m), À133.5 (m),
À129.8 (m, ÀOCF2CF2CF3), À83.01, À82.87 (–CF2–), À81.53,
À80.70, À80.19 (–CF3); IR(ATR): = 1744 (C55O stretch); 1307,
4.10.2. Biphasic conditions (1:3:3 toluene:hexanes:Krytox1K6
[F(CF(CF3)CF2O)5CF2CF3])
n
The following is representative and is illustrated in Fig.
3
1234, 1197, 1115 (pHFPO C–F stretches), 976 (st), 743, 696
(Reaction 2). In an inert atmosphere glovebox, a glass 3-necked
6 mL reactor with J-young valves equipped with a small magnetic
stir bar and hydrogenation balloon was charged with 2 mL of a
degassed PFMCH solution containing tris-pHFPO-modified triar-
ylphosphine rhodium chloride (0.01936 g, 4.20EÀ03 mmol,
0.5 mol%). The PFMCH was removed in vacuo and 2 mL of
Krytox1K6 [F(CF(CF3)CF2O)5CF2CF3] was added to the reaction
flask. To this solution, 2 mL of dry, degassed hexanes was added,
(monosub. arom. C–H out-of-plane bend) cmÀ1
.
4.8. Partition coefficients
The following is representative of the partition coefficient
protocol. In a glovebox, a 5 mL vial was charged with 2 mL of a
PFMCH solution containing 0.038 g of tris(4-diphenylphosphino)
pHFPO methylene benzoate. To the vial, 2 mL of dry degassed
toluene were added. The vial was sealed with a compression cap
fitting, wrapped with Parafilm, vigorously shaken for 20 minutes
and placed in a constant temperature bath at 25 8C for 48 hours
and allowed to equilibrate. Following the 48 h equilibration period,
a 0.50 mL aliquot was removed from each phase (organic and
fluorous) and transferred to separate Wilmad external coaxial
0.7 mL of dry, degassed toluene, as well as 74
mL of degassed 2-
cyclohexen-1-one. The reactor was sealed and removed from the
glovebox and the contents degassed via freeze–pump-thaw
method. The system was backfilled with hydrogen and set in a
45 8C constant temperature water bath. The biphasic system was
vigorously stirred and the reaction’s progress was analyzed via GC/
MS (4.1 h, 50% conversion to cyclohexanone; 10.7 h, 99% conver-
sion). Upon completion of the reaction, the contents of the reactor
were degassed and returned to the glovebox. The organic phase
was removed via pipette and the vial recharged with a fresh aliquot
NMR tubes. A coaxial stem insert containing a 10% (v/v)
a,a,a-
trifluorotoluene in toluene solution was inserted as the external
standard. The samples were analyzed by 19F NMR (integration of
the signal at À62.39 ppm corresponding to the external standard
of substrate (74
mL) and toluene (2 mL). This was repeated through
a
,
a
,
a
-trifluorotoluene against the multiplet at À146.3 ppm for
3 cycles for a total of 548 turnovers with no apparent catalyst loss.
ligand 3 and at À145.09 ppm for catalyst 4, corresponding to the (–
OCF2CF(CF3)–) of the pHFPO fluorous tail). The procedure was
repeated in triplicate, giving an average partition coefficient.
4.10.3. Monophasic conditions
The monophasic conditions were achieved using a solvent
system comprising a 1:3:3 solution of toluene:hexanes:PFMCH as
illustrated in Fig. 3. In an inert atmosphere glovebox, a glass 3-
necked 6 mL reactor with J-young valves equipped with a small
magnetic stir bar and hydrogenation balloon was charged with
2.14 mL of a PFMCH solution containing tris-pHFPO-modified
triarylphosphine rhodium chloride (0.019 g, 4.20EÀ03 mmol,
0.5 mol%), this was a bright orange solution. To this solution,
2.14 mL of dry, degassed hexanes, 0.7 mL of dry, degassed toluene
4.9. Rh leaching studies
At the conclusion of each catalytic cycle, the organic phase was
removed and the solvent was allowed to evaporate. The rhodium
content was then determined by ICP-AAS. In the case of the 1:3:3
toluene:hexanes:Krytox1K6 [F(CF(CF3)CF2O)5CF2CF3] biphasic
system and the PFMCH:hexanes:toluene monophasic system,
the organic phase was analyzed after each cycle. In the case of
the PFMCH:toluene biphasic system, the organic phases for all
eight cycles were combined and analyzed for the rhodium content.
as well as 74
mL of degassed 2-cyclohexen-1-one were added. The
reactor was sealed and removed from the glovebox and the
contents degassed via freeze–pump-thaw method. The system was

32
C.M. Friesen et al. / Journal of Fluorine Chemistry 144 (2012) 24–32
[14] I.T. Horvath, G. Kiss, R.A. Cook, J.E. Bond, P.A. Stevens, J. Rabai, E.J. Mozeleski,
Journal of the American Chemical Society 120 (1998) 3133–3143.
[15] A. Aghmiz, C. Claver, A.M. Masdeu-Bulto, D. Maillard, D. Sinou, Journal of Molec-
ular Catalysis A: Chemical 208 (2004) 97–101.
[16] T. Soos, B.L. Bennett, D. Rutherford, L.P. Barthel-Rosa, J.A. Gladysz, Organome-
tallics 20 (2001) 3079–3086.
backfilled with hydrogen and set in a 45 8C constant temperature
water bath. The monophasic system was vigorously stirred and the
reaction’s progress was analyzed via GC/MS (3.5 h, 50% conversion
to cyclohexanone; 8.8 h, 99% conversion). Upon completion of the
reaction, the contents of the reactor were degassed and returned to
the glovebox. The organic phase was removed via pipette and the
[17] D. Rutherford, J.J.J. Juliette, C. Rocaboy, I.T. Horvath, J.A. Gladysz, Catalysis Today
42 (1998) 381–388.
[18] B. Richter, B. Deelman, G. Van Koten, Journal of Molecular Catalysis A: Chemical
145 (1999) 317–321.
[19] B. Richter, A.L. Spek, G. van Koten, B. Deelman, Journal of the American Chemical
Society 122 (2000) 3945–3951.
vial recharged with a fresh aliquot of substrate (74
(2 mL). This was repeated through 2 cycles for a total of 366
turnovers with no apparent catalyst loss.
mL) and toluene
[20] J. Fawcett, E.G. Hope, A.M. Stuart, A.J. West, Green Chemistry 7 (2005) 316–320.
[21] X. Hao, A. Yoshida, J. Nishikido, Green Chemistry 6 (2004) 566–569.
[22] E.G. Hope, A.M. Stuart, A.J. West, Green Chemistry 6 (2004) 345–350.
[23] A. Yoshida, X. Hao, J. Nishikido, Green Chemistry 5 (2003) 554–557.
[24] D.J. Birdsall, E.G. Hope, A.M. Stuart, W. Chen, Y. Hu, J. Xiao, Tetrahedron Letters 42
(2001) 8551–8553.
[25] W. Chen, L. Xu, Y. Hu, A.M. Banet Osuna, J. Xiao, Tetrahedron 58 (2002) 3889–
3899.
[26] L.J. Alvey, D. Rutherford, J.J.J. Juliette, J.A. Gladysz, Journal of Organic Chemistry 63
(1998) 6302–6308.
4.10.4. Calculation of turnover frequencies
For each cycle, a 2nd order polynomial fit of the data was used
to calculate the time taken to reach 99% completion. The TOF99%
value was then calculated from that time. The R2 correlation for the
2nd order polynomial fit was 0.986 or greater in all cases. The
average TOF value reported is the average TOF99%, averaged over all
cycles (with the exception of the 1:1 toluene:PFMCH biphasic
system where the first cycle was not included in the average
calculation, as it clearly represented an induction period).
[27] Q. Zhang, Z. Luo, D.P. Curran, Journal of Organic Chemistry 65 (2000) 8866–8873.
[28] S. Kainz, A. Brinkmann, W. Leitner, A. Pfaltz, Journal of the American Chemical
Society 121 (1999) 6421–6429.
[29] W. Chen, L. Xu, J. Xiao, Organic Letters 2 (2000) 2675–2677.
[30] D. Sinou, G. Pozzi, E.G. Hope, A.M. Stuart, Tetrahedron Letters 40 (1999) 849–
852.
Acknowledgements
[31] D. Sinou, D. Maillard, G. Pozzi, European Journal of Organic Chemistry (2002)
269–275.
[32] P. Bhattacharyya, B. Croxtall, J. Fawcett, J. Fawcett, D. Gudmunsen, E.G. Hope,
R.D.W. Kemmitt, D.R. Paige, D.R. Russell, A.M. Stuart, D.R.W. Wood, Journal of
Fluorine Chemistry 101 (2000) 247–255.
[33] B. Betzemeier, P. Knochel, Angewandte Chemie-International Edition in English
36 (1997) 2623–2624.
[34] E. de Wolf, E. Riccomagno, J.J.M. de Pater, B. Deelman, G. van Koten, Journal of
Combinatorial Chemistry 6 (2004) 363–374.
[35] B. Richter, E. de Wolf, G. van Koten, B. Deelman, Journal of Organic Chemistry 65
(2000) 3885–3893.
[36] B.M. Berven, G.A. Koutsantonis, Synthesis (2008) 2626–2630.
[37] E.G. Hope, A.M. Stuart, Journal of Fluorine Chemistry 100 (1999) 75–83.
[38] L.V. Dinh, J.A. Gladysz, Angewandte Chemie-International Edition 44 (2005)
4095–4097.
The authors wish to acknowledge Dr. Peter Wilson and Dr. Daniel
Leznoff for helpful discussions. Also we thank Dr. Jon Howell of E. I.
du Pont de Nemours Experimental Station, Wilmington, DE for
supplying the starting material, pHFPO-CH2OH and to Dr. Joseph S.
Thrasher of The University of Alabama, Tuscaloosa for the MALDI-
TOF-MS analyses. CDM acknowledges the Institute of Chemistry at
the Hebrew University of Jerusalem for a Visiting Professorship.
Finally we are grateful to Trinity Western University, Simon Fraser
University and the Natural Sciences and Engineering Research
Council of Canada for financial support.
[39] J.A. Gladysz, Handbook of Fluorous Chemistry (2004) 41–55.
[40] J. Scheirs, Modern Fluoropolymers, High Performance Polymers for Diverse
Applications, Wiley & Sons, New York, 1997, p. 435.
[41] W. Keim, M. Vogt, P. Wasserscheid, B. Driessen-Holscher, Journal of Molecular
Catalysis A: Chemical 139 (1999) 171–175.
[42] C. Tamborski, C.E. Snyder Jr., J.B. Christian, U.S. Patent 4,454,349 (1984).
[43] J.L. Howell, K.A. Hay, U.S. Patent (2006).
[44] Q. Lin, S. Unal, A.R. Fornof, R.S. Armentrout, T.E. Long, Polymer 47 (2006)
4085–4093.
[45] M. El-Deek, M.A. Hassan, S. El-Hamshary, Journal of the Indian Chemical Society
58 (1981) 197–199.
[46] C.R. Mathison, D.J. Cole-Hamilton, Catalysis by Metal Complexes 30 (2006)
145–181.
[47] D.J. Adams, D.J. Cole-Hamilton, E.G. Hope, P.J. Pogorzelec, A.M. Stuart, Journal of
Organometallic Chemistry 689 (2004) 1413–1417.
[48] C.S. Consorti, F. Hampel, J.A. Gladysz, Inorganica Chimica Acta 359 (2006)
4874–4884.
References
[1] C.M. Friesen, Ph.D. Dissertation, University of Alabama, Tuscaloosa, 2000.
[2] T.M. O’Connor, M.S. Jhon, C.L. Bauer, B.G. Min, D.Y. Yoon, T.E. Karis, Tribology
Letters 1 (1995) 219–223.
[3] G. Caporiccio, L. Flabbi, G. Marchionni, G.T. Viola, Journal of Synthetic Lubrication
6 (1989) 133–149.
[4] G. Marchionni, G. Ajroldi, G. Pezzin, European Polymer Journal 24 (1988) 1211–1216.
[5] F. Piacenti, M. Camaiti, Journal of Fluorine Chemistry 68 (1994) 227–235.
[6] K.L. O’Neal, S.G. Weber, The Journal of Physical Chemistry B 113 (2009) 7449–7456.
[7] K.L. O’Neal, S.G. Weber, The Journal of Physical Chemistry B 113 (2009) 149–158.
[8] R. Correa da Costa, T. Buffeteau, A. Del Guerzo, N.D. McClenaghan, J. Vincent,
Chemical Communications (Cambridge, U.K.) 47 (2011) 8250–8252.
[9] M. Vogt, Ph.D. Dissertation, Rheinisch-Westfalische Technische Hochschule
Aachen, 1991.
[10] I.T. Horvath, J. Rabai, Science 266 (1994) 72–75.
[11] J.J.J. Juliette, I.T. Horvath, J.A. Gladysz, Angewandte Chemie-International Edition
in English 36 (1997) 1610–1612.
[12] L.V. Dinh, J.A. Gladysz, Tetrahedron Letters 40 (1999) 8995–8998.
[13] E. de Wolf, E.A. Speets, B. Deelman, G. van Koten, Organometallics 20 (2001)
3686–3690.
[49] L.J. Alvey, R. Meier, T. Soos, P. Bernatis, J.A. Gladysz, European Journal of Organic
Chemistry (2000) 1975–1983.
[50] J.H. van Tonder, C. Marais, D.J. Cole-Hamilton, B.C.B. Bezuidenhoudt, Synthesis
(2010) 421–424.