Syntheses and reactivity of analogues of Grubbs' second generation metathesis catalyst with fluorous phosphines: A new phase-transfer strategy for catalyst activation

Article abstract of DOI:10.1002/adsc.200600461

Reactions of the bis(pyridine) complex (H₂IMes)(Py) ₂(Cl)₂Ru(=CHPh), (H₂IMes = 1,3-dimesityl-4,5- dihydroimidazol-2-ylidene) and the highly fluorophilic phosphines [R _fn(CH₂)_m]

Full text of DOI:10.1002/adsc.200600461

FULL PAPERS
DOI: 10.1002/adsc.200600461
Syntheses and Reactivity of Analogues of Grubbsꢀ Second
Generation Metathesis Catalyst with Fluorous Phosphines:
A New Phase-Transfer Strategy for Catalyst Activation
Rosenildo CorrÞa da Costaa and John A. Gladysza,*
a
Institut für Organische Chemie, Friedrich-Alexander-Universität Erlangen-Nürnberg, Henkestraße 42, 91054 Erlangen,
Germany
Fax : (+49)-9131-85-26865; e-mail: gladysz@chemie.uni-erlangen.de
Received: September 11, 2006
Abstract: Reactions of the bis
(H2IMes)(Py)2(Cl)2Ru(=CHPh),
ACHTREUNG
AHCTREUNG
mesityl-4,5-dihydroimidazol-2-ylidene)
highly fluorophilic phosphines [Rfn
AHCTREUNG
m/n=a, 2/6; b, 2/8; c, 2/10; d, 3/8; Rfn =CF3
G
]
give (H2IMes){[Rfn(CH2)m]3P}(Cl)2Ru
A
N
ACHTREUNG
d, 64–78%), which are analogues of Grubbsꢀ second lyst, 4c, can be recycled by extracting the reaction
generation alkene metathesis catalyst. Complexes mixtures with perfluoro(methylcyclohexane).
4a,b are effective catalysts for conversions of 1,6-
dienes to cyclopentenes under monophasic and bi-
phasic conditions in CH2Cl2 and CH2Cl2/fluorous sol- Keywords: alkene metathesis; fluorous; Grubbsꢀ cat-
vent mixtures. The latter generally exhibit rate accel- alyst; phase transfer; phosphines; recycling
Introduction
tion (k2[alkene]@kÀ1[phosphine]), phosphine dissoci-
C
Fluorous catalysts most commonly feature ponytails
of the formulae Rfn
G
G
and exhibit moderate to marked affinities for fluorous
liquid and solid phases.[1] As such, they can easily be
recovered and reused. However, it seemed to us that
fluorous methodologies also present intriguing oppor-
tunities for catalyst activation. For example, there are
many transition metal-based catalyst precursors from
which a ligand must first dissociate before the catalyt-
ic cycle can be entered. The reverse process can slow
the overall rate. Thus, if the ligand could be efficiently
scavenged, faster reactions should often occur. Most
scavenging strategies involve chemical derivatiza-
tion.[2] However, we wondered whether phase transfer
of a fluorous ligand into a fluorous phase might also
be exploited.
A case in point would be the ruthenium alkene
metathesis catalysts developed by Grubbs.[3] As sum-
marized in Scheme 1 (top), the dissociation of a phos-
phine (k1 or initiation step) can be followed either by
unproductive recoordination (kÀ1 step) or productive
alkene binding (k2 step). Under conditions where the
product of k2 and the alkene concentration is much
Scheme 1. Initiation sequence for a Grubbs-type ruthenium
greater than that of kÀ1 and the phosphine concentra- alkene metathesis catalyst.
Adv. Synth. Catal. 2007, 349, 243 – 254
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
243

Rosenildo CorrÞa da Costa and John A. Gladysz
FULL PAPERS
little or no effect. However, kinetics studies have ses under monophasic and biphasic conditions in
identified many catalysts, including the commercially CH2Cl2 and CH2Cl2/fluorous solvent mixtures; (3)
available first and second generation systems, for substantial rate accelerations under the latter condi-
which kÀ1 ꢀk2, even with very reactive alkenes (e.g., tions, and (4) successful recycling experiments. A por-
ethyl vinyl ether).[3]
tion of this work has been communicated,[12] and addi-
We wondered whether rate enhancements might be tional details are supplied elsewhere.[13]
realized when similar metathesis catalysts bearing flu-
orous phosphines were employed under organic/fluo-
rous liquid/liquid biphase conditions. As illustrated in Results
Scheme 1 (bottom), fluorous phosphines can have
high thermodynamic affinities for fluorous phases, Synthesis and Characterization of New Metathesis
whereas the active catalyst and alkenes would have Catalysts
high thermodynamic affinities for organic phases. For
example, even the non-polar alkene 1-decene shows a As shown in Scheme 2, the bis
ACHTREUNG
marked preference (95:5) to partition into toluene complex (H2IMes)(Py)2(Cl)2Ru
versus perfluoro(methylcyclohexane) (CF3C6F11).[4]
This bias should be even more pronounced for func-
tionalized alkenes and the more polar solvents com-
monly used for alkene metathesis. To the extent that
the less fluorous catalyst precursor partitions into the
fluorous phase, some counteracting rate loss would be
expected.
ACHTREUNG
As depicted in Figure 1, several fluorous versions of
Grubbsꢀ catalysts have been reported.[5–-7] Other cata-
lysts that contain two-or threec-arbon perfluoroalkyl
segments,[8] or (CF3)3CO ligands,[9] are known. How-
ever, the fluorinated moieties in all of these systems
are located in either non-dissociating spectator li-
gands or “boomerang” carbene ligands. To our knowl-
Scheme 2. Syntheses of fluorous Grubbsꢀ catalysts 4a–d.
edge, only one metathesis catalyst with a fluoroalky- fluorous phosphines 1a–d were reacted under homo-
lated phosphine, (p-CF3C6H4)3P, [3b] has ever been re- geneous conditions in the amphiphilic solvent trifluoro-
ported. Accordingly, we set out to prepare and evalu- methylbenzene (CF3C6H5). Work-ups gave the target
ate catalysts containing the highly fluorophilic aliphat- complexes 4a–d as analytically pure pink solids in 64–
ic phosphines [Rfn(CH2)m]3P (1; m/n=a, 2/6; b, 2/8; c, 78% yields. They were characterized by NMR spec-
G
2/10; d, 3/8),[10] which offer a spectrum of phase prop- troscopy (1H, 13C, 31P, 19F) and mass spectrometry, as
erties. Due to the short (CH2)m spacer segments, 1a–d summarized in the Experimental Section. Most data
are much less basic and nucleophilic than tri
ACHTREUNG
phosphines.[11]
In this paper, we report (1) the facile synthesis of a complexes of 1a–d.
family of fluorous phosphine analogues of Grubbsꢀ Complexes 4a–d were air-stable as solids for ex-
second generation catalyst (H2IMes)(Cy3P)(Cl)2Ru- tended periods, as well as overnight in solution. How-
(=CHPh) (2; H2IMes=1,3-dimesityl-4,5-dihydroimi- ever, crystallization attempts did afford an unusual
AHCTREUNG
ACHTREUNG
dazol-2-ylidene); (2) applications in alkene metathe- oxidation product, a ruthenium(III) species described
G
Figure 1. Previously reported fluorous alkene metathesis catalysts.
244
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 243 – 254

Analogues of Grubbsꢀ Second Generation Metathesis Catalyst
FULL PAPERS
elsewhere.[15] Complexes 4a–d melted without decom-
position at temperatures ranging from 101 to 1188C.
No other phase transitions were noted by DSC. Exo-
therms were observed 20–658C above the melting
points, accompanied by slight mass losses (2.9–4.5%;
TGA). Complexes 4a,b,d were soluble in CH2Cl2,
benzene, and hexane. However, 4c was much less
soluble, paralleling the effect of the Rfn length in the
free phosphines 1a–c. Interestingly, all complexes
were soluble in the more polar solvent methanol.
The CF3C6F11/toluene partition coefficients of 4a–d
were measured by HPLC as described in the Experi-
mental Section. The data are summarized in Table 1,
Table 1. Partition coefficients (CF3C6F11/toluene, 258C) for
fluorous phosphines and new fluorous metathesis catalysts.
Fluorous
Phosphine
Partition Coef- Fluorous
Partition Co-
efficient
ficient[10]
Catalyst
1a
1b
1c
1d
98.8:1.2
4a
4b
4c
4d
13.2: 86.8
39.6:60.4
77.6:22.4
11.5:88.5
>99.7:<0.3
>99.7:<0.3
98.8:1.2
Figure 2. Rates of formation of 6 (room temperature). Sol-
vent systems: CH2Cl2/C8F16O (2.2 mL/1.1 mL); CH2Cl2/
~
*
*
CF3C6F11 (4.0 mL/2.0 mL);
CH2Cl2/CF3C6F11 (5.0 mL/
&
2.5 mL); ” CH2Cl2/CF3
CH2Cl2 (3.1 mL).
(CF2)5CH2OH (2.2 mL/1.1 mL);
together with values for the corresponding phos-
phines.[4,10] The values for 4a–c followed the expected
trend, with 4a predominantly lipophilic (13.2:86.8)
and 4c significantly fluorophilic (77.6:22.4). The addi-
tional methylene group in each ponytail of 4d as com-
pared to 4b significantly enhances lipophilicity
The samples were rapidly stirred, dispersing the flu-
(11.5:88.5 vs. 39.6:60.4). When toluene is replaced by orous phase. The formation of cyclopentene 6 was
a solvent of greater polarity (e.g., CH2Cl2), the pro- monitored by GC, and yields were calculated with re-
portions of 4a,b,d in the non-fluorous phase should in- spect to the internal standard. The initial rate was sig-
crease.
nificantly enhanced in the presence of CF3C6F11.
After 1 and 2 h the yields of 6 were 23% and 44%
&
*
( ), as opposed to 6% and 16% in CH2Cl2 alone ( ).
The effect was reproducible, with yields of 34% and
Metatheses with 4b: Monophasic and Biphasic
Conditions
*
56% in a similar run ( ); these traces represent the
maximum deviation observed in duplicate experi-
Initial experiments focused on the moderately fluoro- ments in the course of this study. An analogous ex-
philic complex 4b. The ring-closing metathesis of di- periment was conducted using perfluoro(2-butyltetra-
ethyl diallylmalonate (5)[16] was investigated under the hydrofuran) (C8F16O). This solvent is somewhat less
conditions depicted in Figure 2. Two reactions were viscous than CF3C6F11,[17] which we imagined might
conducted under N2 using CH2Cl2 that was 0.048– aid phosphine diffusion across the phase boundary.
0.051M in 5 (default diene concentration for all ex- Interestingly, the rate was much faster, with 58% and
~
periments) and 0.0012–0.0013M in 4b (2.5 mol%), 74% yields of 6 after 1 and 2 h ( ).
and contained an internal standard. In one case,
Analogous experiments with perfluorohexane
CF3C6F11 was added, corresponding to 50% of the (C6F14) were complicated by some solvent volatiliza-
CH2Cl2 volume. As noted above, this should not sig- tion, and are detailed elsewhere.[13] However, the rate
nificantly affect the concentration of 5 in the CH2Cl2 was not as fast as with C8F16O, with 36% and 47%
phase. The concentration of 4b should be reduced by yields of 6 after 1 and 2 h. The fluorous alcohol
less than 20% (i.e., the 39.6% that would be extract- 1H,1H-perfluoroheptanol [CF3(CF2)5CH2OH] was
G
ed with a 1:1 volume ratio divided by two; this would also evaluated. Although we are not aware of any
be further decreased due to the substitution of tolu- quantitative viscosity data for this compound, qualita-
ene by CH2Cl2). Visually, all of the catalyst color re- tively it can be judged to be much higher. As depicted
mained in the CH2Cl2 phase.
in Figure 2, only a very slight initial rate enhancement
Adv. Synth. Catal. 2007, 349, 243 – 254
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
245

Rosenildo CorrÞa da Costa and John A. Gladysz
FULL PAPERS
was observed, with 10% and 18% yields of 6 after 1
and 2 h (”).
For the most effective solvent system in Figure 2
(CH2Cl2/C8F16O), the catalyst loading (2.5%) was
varied. As shown in Figure 3, when the loading was
Figure 4. Rates of formation of 8 (room temperature). Sol-
&
*
vent systems: CH2Cl2/CF3C6F11 (4.0 mL/2.0 mL); CH2Cl2
~
(4.0 mL); CH2Cl2/C8F16O (4.6 mL/2.3 mL).
and catalyst in this study for which no rate accelera-
tion was observed. However, as shown in Figure 5,
C8F16O gave a dramatic effect with N,N-diallyltosyl-
amide (9). The yields of cyclopentene 10[6] were 58%
Figure 3. Rates of formation of 6 (room temperature) in 2:1
^
v/v CH2Cl2/C8F16O with different catalyst loadings:
5.0
~
&
mol% (1.2 mL/0.6 mL); 2.5 mol% (2.2 mL/1.1 mL); 1.0
mol% (5.2 mL/2.6 mL).
^
and 91% after 1 and 2 h ( ), as compared to 1–2%
~
&
and 3–4% with CH2Cl2 alone (two runs, and ).
decreased to 1.0 mol%, metathesis became somewhat
slower, with 32% and 58% yields of 6 after 1 and 2 h
&
( ). However, reaction remained at least as fast as
that in CH2Cl2/CF3C6F11 with 2.5 mol% 4b (Figure 2).
In contrast, when the catalyst loading was increased
^
to 5.0%, little effect was observed ( ). The amount
of fluorous solvent was also varied, keeping the
CH2Cl2 volume and concentrations of 6 (0.048–
0.049M) and 4b (2.5 mol%) constant. Interestingly,
*
the results with 5.0:2.5 v/v ( in Figure 2) and 5.0:1.0
v/v CH2Cl2/CF3C6F11 (not depicted) differed very
little, with 12% vs. 10%, 23% vs. 24%, and 35% vs.
36% yields of 6 after 0.5, 1.0, and 1.5 h.
We sought to confirm that these phenomena were
not unique to a single substrate. Figure 4 summarizes
results obtained with diethyl 2-allyl-2-methallylmalo-
nate (7) under standard conditions in CH2Cl2 (0.051–
0.052M in 7 and 0.0013M in 4b). In the presence of
CF3C6F11, the formation of cyclopentene (8)[18] was
again significantly faster. After 1 and 2 h, the yields
*
of 8 were 47% and 73% ( ), as opposed to 20% and
&
51% in CH2Cl2 alone ( ). In contrast, there was little
Figure 5. Rates of formation of 10 (room temperature). Sol-
or no effect in the presence of C8F16O, which repre-
sents the only reaction involving a fluorous solvent (2.0 mL); CH2Cl2 (2.0 mL; duplicate run).
&
^
vent systems:
CH2Cl2/C8F16O (2.0 mL/1.0 mL); CH2Cl2
~
246
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 243 – 254

Analogues of Grubbsꢀ Second Generation Metathesis Catalyst
FULL PAPERS
Metatheses with OtherCatalysts
As shown in Figure 6, the metathesis of 7 was also
studied with the less fluorophilic Rf6 catalyst 4a.
Figure 7. Rates of formation of 12 (room temperature). Sol-
~
&
vent systems:
CH2Cl2/C8F16O (2.4 mL/1.2 mL); CH2Cl2
(2.4 mL).
mains some possibility that the preceding trends
might in part represent solvent effects.[20] Hence, con-
trol experiments were conducted with a related non-
fluorous metathesis catalyst, Grubbsꢀ second genera-
tion complex 2. Identical substrate concentrations and
Figure 6. Rates of formation of 8 (room temperature). Sol-
&
*
vent systems: CH2Cl2/CF3C6F11 (4.0 mL/2.0 mL); CH2Cl2
(4.0 mL).
Under conditions analogous to those in Figure 4, simi- catalyst loadings were employed, and the data are
lar rate enhancements were observed using 2:1 v/v presented in Figure 8 and Figure 9. Note that the re-
&
*
CH2Cl2/CF3C6F11 ( ) as compared to CH2Cl2 ( ). The action times are much shorter than with 4b (Figure 2
yields of 8 were 60% vs. 25% and 90% vs. 60% after and Figure 5), indicating 2 to be a more active cata-
1 and 2 h. In both solvent systems, reactions were lyst.
slightly faster than with 4b (Figure 4). Two experi-
ments were also conducted with N-allyl-N-methallyl- cyclopentene 6 were virtually identical in CH2Cl2 ( )
Figure 8 shows that the initial rates of formation of
&
*
tosylamide (11). As shown in Figure 7, the cyclopen- and CH2Cl2/C8F16O ( ), with yields of 50% vs. 49%
tene 12[6] formed slightly faster in CH2Cl2/C8F16O ( ) and 74% vs. 78% after 15 and 30 min. Moderate dif-
~
~
&
than CH2Cl2 ( ), with conversions of 79% vs. 59% ferences were observed with CH2Cl2 ( ) and CH2Cl2/
after 1 h. CF3C6F11 (”). Yields of 6 were 44% vs. 62% and
The appreciably fluorophilic Rf10 catalyst 4c could 65% vs. 78% after 15 and 30 min. The metathesis of
&
not be studied under identical conditions due to its 9 was also checked in CH2Cl2 ( ) and CH2Cl2/C8F16O
^
much lower solubility in CH2Cl2. Accordingly, solu- ( ). As shown in Figure 9, the yield of 10 was 49%
tions that were 0.049–0.050M in 7 and 0.00046– vs. 53% after 15 min. Hence, we conclude that there
0.00051M in 4c (1.0 mol%) were employed, giving is no appreciable fluorous solvent effect with non-flu-
homogeneous reaction conditions. Reactions of 7 con- orous metathesis catalysts.
ducted with 2:1 v/v CH2Cl2/CF3C6F11 were now slower
than those in CH2Cl2. Yields of 8 were 3% vs. 5%,
12% vs. 18%, and 21% vs. 40% after 1, 2, and 3 h. Catalyst Recycling
The CF3C6F11 phase was distinctly more colored than
the CH2Cl2 phase, suggesting inhibition via phase Apart from the applications above, there is the obvi-
transfer of the catalyst precursor.
ous question whether 4a–d might have utility as recy-
In organic/fluorous biphase systems, small amounts clable metathesis catalysts, as demonstrated earlier by
of the fluorous solvent typically partition into the or- Yao and Curran with B and C in Figure 1.[6,7] Since
ganic solvent, and vice-versa.[19] Therefore, there re- 4a–d can be purified via silica gel chromatography, we
Adv. Synth. Catal. 2007, 349, 243 – 254
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
247

Rosenildo CorrÞa da Costa and John A. Gladysz
FULL PAPERS
Figure 9. Rates of formation of 10 (room temperature). Sol-
&
^
vent systems:
CH2Cl2/C8F16O (2.0 mL/1.0 mL); CH2Cl2
(2.0 mL).
Figure 10. Fluorous/organic liquid/liquid biphase recycling of
metathesis catalyst 4c ([Ru]-PRf10).
Figure 8. Rates of formation of 6 (room temperature). Sol-
&
*
vent systems:
CH2Cl2/C8F16O (5.0 mL/2.5 mL); CH2Cl2
~
(5.0 mL); ” CH2Cl2/CF3C6F11 (5.0 mL/2.5 mL);
(5.0 mL).
CH2Cl2
was 0.050M in the diene 5 and contained a GC stan-
dard. The initial charge of 4c (0.0307 g, 2.6 mol%)
presumed that they could be recovered with fluorous would only be partially soluble in this quantity of
silica gel as employed by Curran. We therefore fo- CH2Cl2, but completely dissolved in the presence of
cused on fluorous/organic liquid/liquid biphase meth- the additives. As shown in Figure 11, the yield of 6
ods, using the most fluorophilic catalyst 4c. Note that was 74% after 10 min, and increased only slightly
although 4c has a benzylidene ligand, the recovered thereafter. After 1 h, the mixture was extracted with
catalyst (4c’) will contain a different, educt-derived al- CF3C6F11 (3”2 mL). Solvent removal from the fluo-
kylidene ligand.
rous phase gave ca. 90% recovery of 4c’ (0.0275 g;
The recycling protocol sketched in Figure 10 was in- 90% if 4c’ and 4c have identical formula weights).
vestigated, using a CH2Cl2 solution (10.7 mL) that The CH2Cl2 phase still carried a residual color.
248
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 243 – 254

Analogues of Grubbsꢀ Second Generation Metathesis Catalyst
FULL PAPERS
phase. The fluorophilicity of our lead catalyst 4b is
likely somewhat greater than optimum, but could
easily be attenuated by modification of the non-fluo-
rous ligands. The reactivity trends have been con-
firmed in many additional experiments employing
non-standard catalyst concentrations or solvent ratios.
Although only a few of the metatheses in Fig-
ures 2–11 attain yields of>90%, this is not unusual
for ruthenium-based catalysts. Deactivation can occur,
and under preparative conditions it is common to add
multiple charges, for example 2”2.5 mol%. However,
in most cases the starting diene was completely con-
sumed. In order to maintain the integrities of the sol-
vent mixtures, no attempts were made to surmount
possible equilibrium constraints by aspirating the vol-
atile ethylene coproduct. Note that despite the ca.
75% yields in Figure 11, the recovered catalysts
retain their activities. Hence, the non-quantitative
yields in these cases cannot be ascribed to catalyst de-
activation.
In all of the above metatheses with 4a–c, the four-
teen valence-electron methylidene intermediate
Figure 11. Rates of formation of 6 (room temperature)
&
(H2IMes)(Cl)2Ru(=CH2) is generated in the cyclo-
A
^
under the conditions of Figure 10;
first cycle;
second
~
cycle; third cycle.
pentene-forming steps. The same intermediate is ob-
tained when starting with the second generation cata-
lyst 2, and Grubbs has shown that if it combines with
A second cycle was conducted with 4c’ and a the dissociated Cy3P, it is essentially incapable of re-
0.049M CH2Cl2 solution of 5, with the quantity re- entering the catalytic cycle.[3a] The dissociated Cy3P is
duced slightly to preserve a 2.6 mol% loading with also intimately involved in the decomposition of 2 to
the decreased amount of catalyst. As shown in non-alkylidene species.[22] Therefore, another advant-
Figure 11, comparable results were obtained, with a age of the phase transfer activation strategy can be
73% yield of 6 after 10 min. Approximately 85% of longer-lived catalysts.
the charge of 4c’ was recovered. The rate of formation
Detailed kinetic measurements, which would be
of 6 during the third cycle was comparable (66% complicated by the partial partitioning of catalyst pre-
after 10 min), but less 4c’ (57%) could be recovered. cursors 4a–c into the fluorous phases, are beyond the
A parallel sequence conducted in air gave similar re- scope of this initial feasibility study. However, previ-
sults up to this cycle, when the rate slowed somewhat ous investigations involving a variety of ruthenium
(40% and 61% after 10 and 20 min; 60% recovery of catalysts have established that there is no simple rela-
4c’).[21] A fourth cycle gave significantly poorer results. tionship between the rate of phosphine dissociation
Nonetheless, these data show that the new catalyst (k1; Scheme 1) and the k-1/k2 ratio.[3] Triarylphosphine
family can be recycled by fluorous/organic liquid/ ligands with electron-withdrawing substituents give
liquid biphase techniques.
greater k1 values, in accord with their weaker Brønst-
ed basicities and therefore enhanced leaving group
abilities.[3b] Hence, 4a–d would be expected to be
faster initiating catalysts than isosteric analogues with
Discussion
more basic tri
ACHTREUNG
The data presented above are consistent with what
C
ACHTREUNG
sketched for alkene metathesis in Scheme 1, that in
Nonetheless, comparison of Figure 8 and 9 with
theory can be (1) applied to many catalyst precursors Figure 2 and Figure 5 shows that Grubbsꢀ second gen-
from which a ligand must dissociate in a pre-equilibri- eration catalyst 2 is much more active than 4b, despite
um step before the catalyst cycle can be entered, and the much stronger Brønsted basicity of the phosphine
(2) extended to other phases, for example a highly Cy3P compared to 1b.[11] Presumably steric factors,
polar ligand that would preferentially partition into which could affect the relative kÀ1 values, are also at
water. Ideally, the catalyst precursor should have a work. An interesting extension of this study would in-
very high affinity for the reaction phase, and the dis- volve catalysts with fluorous versions of Cy3P, perhaps
sociated ligand a very high affinity for the orthogonal with two Rfn groups attached to the 4-position of each
Adv. Synth. Catal. 2007, 349, 243 – 254
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
249

Rosenildo CorrÞa da Costa and John A. Gladysz
FULL PAPERS
ring. Finally, it should be emphasized that any cata- to equilibrium effects but rather to transport rates.
lyst/substrate combination for which ligand dissocia- One parameter that can affect transport rates is vis-
tion is a rate-determining as opposed to a pre-equilib- cosity, but there are others and the origin of these ef-
rium step would not be a candidate for phase transfer fects will require further study.[30]
activation.
Other approaches to inhibiting the kÀ1 step in
Scheme 1 have been investigated. Grubbs,[23] Nolan,[24] Conclusions
and Blechert[25] have studied various types of copper
species and Lewis or Brønsted acids that are believed In summary, the data in this paper are consistent with
to bind the dissociated phosphine. Rate accelerations a new catalyst activation mechanism featuring phase
have been observed in certain cases, but are some- transfer to fluorous media for the removal of dissoci-
times accompanied by earlier catalyst deactivation. ated ligands that can compete with substrate mole-
Phenol additives can have positive effects, but other cules for binding to a reactive metal center. This strat-
mechanisms are thought to be involved.[26] Highly un- egy is potentially extendible to a wide variety of or-
saturated catalyst precursors that lack any ligand L thogonal phase combinations and ligand phase tags,
have also been developed, but since there are concur- and defines a very promising new frontier for future
rent changes in the ancillary ligands these are not au- catalyst design.
tomatically more active.[9b,27]
A possible enhancement of our methodology would
involve the addition of a fluorous scavenger for the Experimental Section
phosphines 1a–d. This would shift the phase equilibri-
um beyond what can be achieved via partition coeffi- General Remarks
cients alone. In this regard, Vincent has reported
All reactions were conducted under N2 unless noted. Chemi-
the highly fluorous dicopper tetracarboxylate
cals were treated as follows: ether, toluene, hexanes, distil-
Cu
2[O2CCH(CH2CH2CH2Rf8)2]4 (13).[28] Complex 13
G
led from Na/benzophenone; perfluoro(methylcyclohexane)
(CF3C6F11; ABCR), CF3C6H5 (ABCR), distilled from CaH2;
diethyl diallylmalonate (5; Lancaster), tridecane (Aldrich),
perfluoro(2-butyltetrahydrofuran) (C8F16O; Apollo), per-
fluorohexane (C6F14; Fluorochem), 1H,1H-perfluorohepta-
efficiently extracts non-fluorous pyridines into fluo-
rous phases, including C60 derivatives. As detailed
elsewhere, we have studied reactions of the bis-
A
nol (CF3(CF2)5CH2OH; ABCR), CD2Cl2 (Cambridge Iso-
A
rous/organic liquid/liquid biphase mixtures containing
13.[13] However, only rapid catalyst deactivation was
observed.
tope or Aldrich) and other solvents, used as received. Dieth-
yl 2-allyl-2-methallylmalonate (7),[31] N,N-diallyltosylamide
(9)[32] and N-allyl-N-methallyltosylamide (11)[6] were synthe-
sized by literature procedures.
NMR spectra were recorded on 400 MHz spectrometers
at ambient probe temperatures and referenced to residual
internal CHDCl2 (1 H, d=5.32), internal CD2Cl2 (13C, d=
53.23 ppm), internal C6F6 (19F, d=À162.0 ppm) or external
H3PO4 (31P, d=0.00 ppm). IR spectra were measured on an
ASI React-IR spectrometer. GC data were acquired using a
ThermoQuest Trace GC 2000 instrument fitted with a capil-
lary column (OPTIMA 5–0.25 mm; 25 m ” 0.32 mm). HPLC
data were acquired using a Thermoquest instrument pack-
age (pump/autosampler/detector P4000/AS3000/UV6000
LP). DSC and TGA data were recorded with a Mettler-
Toledo DSC821 instrument and treated by standard meth-
ods.[33] Elemental analyses were conducted with a Carlo
Erba EA1110 instrument.
A final point concerns ligand transport across the
phase boundary, which of course must be rapid on the
time scale of the catalytic reaction for a rate enhance-
ment to be possible. This issue underlies all phase-
transfer catalysis and has been extensively studied.[29]
To our knowledge, there are no quantitative data in-
volving heavy fluorous species. However, colored flu-
orous compounds rapidly equilibrate across stirred
phase boundaries, and we presume that phosphines
1a–d behave similarly. Given the extremely high fluo-
rophilicities of 1a–d, we have tentatively attributed
A
N
A Schlenk flask was charged with (H2IMes)(Py)2(Cl)2Ru-
A
(1a;[10] 0.1524 g, 0.142 mmol), and CF3C6H5 (3.0 mL). The
mixture was stirred (1 h). The solvent was removed by oil
pump vacuum to give a brownish pink residue that was
passed through a silica gel plug (4”2 cm) using hexanes and
then hexanes/ether (10:1 v/v). The solvent was removed
from the filtrate by rotary evaporation and oil pump
the rate differences with various fluorous solvents not vacuum to give 4a as
a
pink solid; yield: 0.1482 g
250
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 243 – 254

Analogues of Grubbsꢀ Second Generation Metathesis Catalyst
FULL PAPERS
(0.090 mmol, 71%): mp 113–1188C (capillary). DSC: Ti/Te/
Tp/Tc/Tf 83.8/114.3/123.7/128.0/129.38C (endotherm), 129.4/
134.6/146.8/154.0/159.58C (exotherm). TGA: onset of first
and second mass loss regimes (Te), 127.7 (3.4%) and
171.38C. Anal. calcd. (%) for C52H45Cl2F39N2PRu (1641.8):
C 38.04, H 2.76, N 1.71; found: C 37.85, H 2.83, N 1.67.
1H NMR (400 MHz, CD2Cl2): d=18.91 (s, 1H, Ru=CH),
7.83 (d, J=7.6 Hz, 2H of Ph), 7.51 (t, J=7.7 Hz, 1H of Ph),
7.18 (t, J=7.8 Hz, 2H of Ph), 6.99 (s, 2H of 2Mes), 6.37 (s,
2H of 2Mes), 4.20–3.90 (4H, NCH2CH2N), 2.59 (s, 6H,
2CH3), 2.28 (s, 3H, CH3), 2.23 (s, 6H, 2CH3), 1.94 (s, 3H,
1H NMR (400 MHz, CD2Cl2): d=18.90 (s, 1H, Ru=CH),
7.83 (d, J=7.6 Hz, 2H of Ph), 7.49 (t, J=6.4 Hz, 1H of Ph),
7.17 (t, J=7.8 Hz, 2H of Ph), 6.98 (s, 2H of 2Mes), 6.36 (s,
2H of 2Mes), 4.12–3.90 (2 m, 4H, NCH2CH2N), 2.59 (s, 6H,
2CH3), 2.27 (s, 3H, CH3), 2.22 (s, 6H, 2CH3), 1.93 (s, 3H,
CH3), 1.90–1.50 (br m, 12H,
P
[CH2)2]; 31P{1H} NMR
C
1673
(17)
[P(CH2CH2Rf10)3 +H]+,
568
(32)
[MÀP(CH2CH2Rf10)3]+, 532 (55) [MÀP(CH2CH2Rf10)3ÀCl]+.
CH3), 1.90–1.60 [br m, 12H,
P
(CH2)2]; 13C{1H} NMR
A
(100 MHz, CD2Cl2, partial): d=217.8 [d, J=90.2 Hz,
RuC(N)2], 150.9 (d, J=1.6 Hz), 139.6, 139.3, 138.4, 137.1,
136.9, 134.3, 130.8, 130.6, 129.8, 129.5, 129.2, 128.8, 52.2 (d,
J=3.4 Hz), 51.8 (d, J=1.6 Hz), 25.4 (t, J=22.7 Hz), 20.9,
20.8, 20.2, 18.4, 12.3 (d, J=21.6 Hz); 31P{1H} NMR
(162 MHz, CD2Cl2): d=19.96 (s); 19F NMR (282 MHz,
CD2Cl2, partial): d=À81.64 (t, J=9.3, CF3). MS (FAB, 3-
NBA): m/z (%)=1640 (5) [MÀH]+, 1605 (2.5) [MÀCl]+,
1073 (2.5) [P(CH2CH2Rf6)3 +H]+.
A
G
Complex 3 (0.0790 g, 0.109 mmol),[14] P(CH2CH2CH2Rf8)3
(1d;[10] 0,1700 g, 0.120 mmol), and CF3C6H5 (3.0 mL) were
combined in a procedure analogous to that for 4a. An iden-
tical work-up gave 4d as a pink solid; yield: 0.1695 g
(0.086 mmol, 78%); mp 101–1038C (capillary). DSC: Ti/Te/
Tp/Tc/Tf 102.6/106.6/109.4/111.4/117.88C (endotherm), 153.6/
167.3/173.3/176.8/189.88C (exotherm). TGA: onset of first
and second mass loss regimes (Te), 142.8 (3.9%) and
188.88C. Anal. calcd. (%) for C61H51Cl2F51N2PRu (1983.9):
C 36.93, H 2.59, N 1.41; found: C 36.74, H 2.51, N 1.35.
1H NMR (400 MHz, CD2Cl2): d=18.88 (s, 1H, Ru=CH),
7.88 (d, J=7.8 Hz, 2H of Ph), 7.50 (t, J=7.4 Hz, 1H of Ph),
7.15 (t, J=7.8 Hz, 2H of Ph), 6.98 (s, 2H of 2Mes), 6.34 (s,
2H of 2Mes), 4.12–3.85 (2 m, J=40.9 Hz, 4H, NCH2CH2N),
2.61 (s, 6H, 2CH3), 2.29 (s, 3H, CH3), 2.22 (s, 6H, 2CH3),
1.93 (overlapping s, 3H, CH3), 1.99–1.80, 1.60–1.44 and
A
N
Complex 3 (0.1201 g, 0.165 mmol),[14] P(CH2CH2Rf8)3 (1b;[10]
0.1524 g, 0.142 mmol), and CF3C6H5 (4.0 mL) were com-
bined in a procedure analogous to that for 4a. An identical
workup gave 4b as a pink solid; yield: 0.2050 g (0.106 mmol,
64%); mp 113–1158C (capillary). DSC: Ti/Te/Tp/Tc/Tf, 96.6/
112.4/117.1/120.0/130.38C (endotherm), 159.5/175.7/178.8/
180.0/186.88C (exotherm). TGA: onset of first and second
mass loss regimes (Te), 154.8 (4.5%) and 198.28C. Anal.
calcd (%) for C58H45Cl2F51N2PRu (1941.9): C 35.87, H 2.34,
1.25–1.10 [overlapping m, m, and m, 6H, 6H, 6H, P(CH2)3];
G
13C{1H} NMR (100 MHz, CD2Cl2, partial): d=218.4 [d, J=
90.6 Hz, RuC(N)2], 150.9, 139.6, 139.3, 138.4, 137.0, 136.9,
134.2, 130.7, 130.6, 129.7, 129.5, 128.8, 66.0, 52.2 (d, J=
3.7 Hz), 51.7, 25.4 (t, J=22.9 Hz), 20.9, 20.8, 20.2, 18.4, 12.3
(d, J=23.3 Hz); 31P{1H} NMR (162 MHz, CD2Cl2): d=22.37;
19F NMR (282 MHz, CD2Cl2, partial): d=À81.64 (t, J=
9.9 Hz, CF3). MS (FAB, 3-NBA): m/z (%)=1982 (65)
[MÀH]+, 1947 (25) [MÀCl]+, 1415 (100) [P(CH2CH2Rf8)3 +
H]+, 568 (10) [MÀP(CH2CH2CH2Rf8)3]+, 532 (20)
[MÀP(CH2CH2CH2Rf8)3ÀCl]+.
1
N 1.44; found: C 35.31, H 2.38, N 1.31; H NMR (400 MHz,
CD2Cl2): d=18.90 (s, 1H, Ru=CH), 7.83 (d, J=7.7 Hz, 2H
of Ph), 7.50 (t, J=7.3 Hz, 1H of Ph), 7.18 (t, J=7.7 Hz, 2H
of Ph), 6.99 (s, 2H of 2Mes), 6.40 (s, 2H of 2Mes), 4.20–
4.90 (dm, J=36.6 Hz, 4H, NCH2CH2N), 2.59 (s, 6H, 2CH3),
2.27 (s, 3H, CH3), 2.23 (s, 6H, 2CH3), 1.93 (s, 3H, CH3),
1.90–1.60 [br m, 12H, P
(CH2)2]; 13C{1H} NMR (100 MHz,
T
CD2Cl2, partial): d=218.4 [d, J=90.5 Hz, RuC(N)2], 150.9,
139.6, 139.3, 138.4, 137.0, 136.9, 134.2, 130.7, 130.6, 129.7,
129.4, 128.7, 52.2 (d, J=3.7 Hz), 51.7 (d, J=2.8 Hz), 25.4
(t, J=22.9 Hz), 20.9, 20.8, 20.2, 18.4, 12.2 (d, J=21.1 Hz);
31P{1H} NMR (162 MHz, CD2Cl2): d=20.17 (s); 19F NMR
(282 MHz, CD2Cl2, partial): d=À81.74 (t, J=10.0 Hz, CF3);
MS (FAB, 2-nitrophenyl octyl ether): m/z (%)=1940 (90)
[MÀH]+, 1905 (40) [MÀCl]+, 1373 (100) [P(CH2CH2Rf8)3 +
Experiments in Figure 2
(A): A two-neck flask was charged with 5 (0.0381 g,
0.159 mmol) and tridecane (0.0248 g, 0.135 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(3.1 mL, giving a 0.051M solution). The solution was stirred
and 4b (0.0075 g, 0.0039 mmol, 2.5 mol%) was added
against a stream of nitrogen. Samples were periodically re-
moved by syringe for GC analyses.[34]
H]+,
568
(75)
[MÀP(CH2CH2Rf8)3]+,
532
(85)
[MÀP(CH2CH2Rf8)3ÀCl]+.
(B): A two-neck flask was similarly charged with 5
(0.0574 g, 0.239 mmol) and tridecane (0.0481 g, 0.261 mmol),
and flushed with nitrogen. Freshly distilled CH2Cl2 (5.0 mL,
giving a 0.048M solution) and CF3C6F11 (2.5 mL) were
added. The biphasic mixture was stirred and 4b (0.0116 g,
0.00598 mmol, 2.5 mol%) was added against a stream of ni-
trogen.
(B’): A two-neck flask was similarly charged with 5
(0.0490 g, 0.204 mmol), tridecane (0.0590 g, 0.320 mmol; GC
standard), freshly distilled CH2Cl2 (4.0 mL, giving a 0.051M
solution), CF3C6F11 (2.0 mL), and 4b (0.0097 g, 0.0050 mmol,
2.5 mol%).
A
N
Complex 3 (0.0800 g, 0.110 mmol),[14] P(CH2CH2Rf10)3 (1c;[10]
0.1840 g, 0.110 mmol), and CF3C6H5 (3.0 mL) were com-
bined in a procedure analogous to that for 4a. An identical
work-up gave 4c as
a
(0.081 mmol, 73%); mp 111–1148C (capillary). DSC: Ti/Te/
Tp/Tc/Tf 98.8/114.6/116.8/118.6/126.28C (endotherm), 153.6/
167.3/173.3/176.8/189.88C (exotherm). TGA: onset of first
and second mass loss regimes (Te), 173.4 (2.9%) and
198.08C. Anal. calcd. (%) for C64H45Cl2F63N2PRu (2241.9):
C 34.29, H 2.02, N 1.25; found: C 34.31, H 2.00, N 1.21.
Adv. Synth. Catal. 2007, 349, 243 – 254
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
251

Rosenildo CorrÞa da Costa and John A. Gladysz
FULL PAPERS
(C): A two-neck flask was similarly charged with 5
(0.0260 g, 0.108 mmol), tridecane (0.0212 g, 0.115 mmol),
freshly distilled CH2Cl2 (2.2 mL, giving a 0.049M solution),
C8F16O (1.1 mL), and 4b (0.0052 g, 0.0027 mmol, 2.5 mol%).
(D): A two-neck flask was similarly charged with 5
(0.0275 g, 0.114 mmol), tridecane (0.0194 g, 0.105 mmol),
freshly distilled CH2Cl2 (2.2 mL, giving a 0.052M solution),
Experiments in Figure 6
(A): A two-neck flask was charged with 7 (0.0510 g,
0.201 mmol) and hexadecane (0.0459 g, 0.203 mmol) and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(4.0 mL, giving a 0.050M solution). The solution was stirred
and 4a (0.0085 g, 0.0052 mmol, 2.5 mol%) was added against
a stream of nitrogen. Samples were periodically removed by
syringe for GC analysis.[34]
(B): A two-neck flask was similarly charged with 7
(0.0524 g, 0.206 mmol), hexadecane (0.0471 g, 0.208 mmol),
freshly distilled CH2Cl2 (4.0 mL, giving a 0.052M solution),
CF3C6F11 (2.0 mL), and 4a (0.0085 g, 0.00518 mmol, 2.5
mol%).
CF3(CF2)5CH2OH (1.1 mL), and 4b (0.0057 g, 0.0029 mmol,
A
2.5 mol%).
Experiments in Figure 3
One experiment is identical to C in Figure 2 and the others
were conducted analogously.
(A): A two-neck flask was similarly charged with 5
(0.0133 g, 0.055 mmol), tridecane (0.0093 g, 0.050 mmol),
freshly distilled CH2Cl2 (1.2 mL, giving a 0.046M solution),
C8F16O (0.6 mL), and 4b (0.0053 g, 0.0027 mmol, 5.0 mol%).
(B): A two-neck flask was similarly charged with 5
(0.0644 g, 0.268 mmol), tridecane (0.0482 g, 0.261 mmol),
freshly distilled CH2Cl2 (5.2 mL, giving a 0.052M solution),
C8F16O (2.6 mL), and 4b (0.0052 g, 0.0027 mmol, 1.0 mol%).
Experiments in Figure 7
(A): A two-neck flask was charged with 11 (0.0325 g,
0.123 mmol)[6] and hexadecane (0.0301 g, 0.133 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(2.4 mL, giving a 0.051M solution). The solution was stirred
and 4a (0.0050 g, 0.00305 mmol, 2.5 mol%) was added
against a stream of nitrogen. Samples were periodically re-
moved by syringe for GC analysis.[34] Due to a technical
problem, conversions of 11 to 12 are plotted.
Experiments in Figure 4
(A): A two-neck flask was charged with 7 (0.0524 g,
0.206 mmol)[31] and hexadecane (0.0530 g, 0.234 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(4.0 mL, giving a 0.052M solution). The solution was stirred
and 4b (0.0100 g, 0.00516 mmol, 2.5 mol%) was added
against a stream of nitrogen. Samples were periodically re-
moved by syringe for GC analyses.[34]
(B): A two-neck flask was similarly charged with 7
(0.0531 g, 0.209 mmol), hexadecane (0.0530 g, 0.234 mmol),
and freshly distilled CH2Cl2 (4.0 mL, giving a 0.052M solu-
tion), CF3C6F11 (2.0 mL) and 4b (0.0100 g, 0.00516 mmol, 2.5
mol%).
(C): A two-neck flask was similarly charged with 7
(0.0600 g, 0.236 mmol), hexadecane (0.0510 g, 0.225 mmol),
freshly distilled CH2Cl2 (4.6 mL, giving a 0.051M solution),
C8F16O (2.3 mL), and 4b (0.0113 g, 0.00582 mmol, 2.5
mol%).
(B): A two-neck flask was similarly charged with 11
(0.0328 g, 0.124 mmol), hexadecane (0.0316 g, 0.140 mmol),
freshly distilled CH2Cl2 (2.4 mL, giving a 0.052M solution),
C8F16O (1.2 mL), and 4a (0.0050 g, 0.00305 mmol, 2.5
mol%).
Experiments in Figure 8
(A): A two-neck flask was charged with 5 (0.0603 g,
0.251 mmol) and tridecane (0.0424 g, 0.230 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(5.0 mL, giving a 0.050M solution). The solution was stirred
and 2 (0.0052 g, 0.00612 mmol, 2.5 mol%) was added against
a stream of nitrogen. Samples were periodically removed by
syringe for GC analysis.[34]
(B): A two-neck flask was similarly charged with 5
(0.0600 g, 0.250 mmol), tridecane (0.0440 g, 0.239 mmol),
freshly distilled CH2Cl2 (5.0 mL, giving a 0.050M solution),
C8F16O (2.5 mL), and 2 (0.0100 g, 0.00516 mmol, 2.5 mol%).
(C): A two-neck flask was similarly charged with 5
(0.0590 g, 0.246 mmol), tridecane (0.0451 g, 0.245 mmol),
freshly distilled CH2Cl2 (5.0 mL, giving a 0.049M solution),
and 2 (0.0053 g, 0.00624 mmol, 2.5 mol%).
Experiments in Figure 5
(A): A two-neck flask was charged with 9 (0.0260 g,
0.104 mmol)[32] and octadecane (0.0212 g, 0.083 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(2.0 mL, giving a 0.052M solution). The solution was stirred
and 4b (0.0053 g, 0.0027 mmol, 2.5 mol%) was added
against a stream of nitrogen. Samples were periodically re-
moved by syringe for GC analysis.[34]
(A’): A two-neck flask was similarly charged with 9
(0.0260 g, 0.104 mmol), octadecane (0.0210 g, 0.083 mmol),
freshly distilled CH2Cl2 (2.0 mL, giving a 0.052M solution),
and 4b (0.0050 g, 0.0026 mmol, 2.5 mol%).
(D): A two-neck flask was similarly charged with 5
(0.0593 g, 0.247 mmol), tridecane (0.0460 g, 0.250 mmol),
freshly distilled CH2Cl2 (5.0 mL, giving a 0.049M solution),
CF3C6F11 (2.5 mL), and
2 (0.0053 g, 0.00625 mmol, 2.5
mol%) was added against a stream of nitrogen.
Experiments in Figure 9.
(B): A two-neck flask was similarly charged with 9
(0.0260 g, 0.104 mmol), octadecane (0.0212 g, 0.083 mmol),
freshly distilled CH2Cl2 (2.0 mL, giving a 0.052M solution),
C8F16O (1.0 mL), and 4b (0.0052 g, 0.0027 mmol, 2.5 mol%).
(A): A two-neck flask was charged with 9 (0.0260 g,
0.104 mmol) and octadecane (0.0211 g, 0.083 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(2.0 mL, giving a 0.050M solution). The solution was stirred
and 2 (0.0022 g, 0.00259 mmol, 2.5 mol%) was added against
a stream of nitrogen. Samples were periodically removed by
syringe for GC analysis.[34]
252
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 243 – 254

Analogues of Grubbsꢀ Second Generation Metathesis Catalyst
FULL PAPERS
(B): A two-neck flask was similarly charged with 9
(0.0260 g, 0.104 mmol), octadecane (0.0213 g, 0.084 mmol),
freshly distilled CH2Cl2 (2.0 mL, giving a 0.050M solution),
C8F16O (1.0 mL), and 2 (0.0022 g, 0.0026 mmol, 2.5 mol%).
[4] J. A. Gladysz, C. Emnet, J. Rµbai, in: Handbook of
Fluorous Chemistry, (Eds.: J. A. Gladysz, D. P. Curran,
I. T. Horvµth), Wiley-VCH, Weinheim, 2004, Ch. 6.
[5] A. Fürstner, L. Ackermann, B. Gabor, R. Goddard,
C. W. Lehmann, R. Mynott, F. Stelzer, O. R. Thiel,
Chem. Eur. J. 2001, 7, 3236.
Experiments in Figure 10 and Figure 11
[6] Q. Yao, Y. Zhang, J. Am. Chem. Soc. 2004, 126, 74.
[7] M. Matsugi, D. P. Curran, J. Org. Chem. 2005, 70, 1636.
[8] a) J. O. Krause, O. Nuyken, K. Wurst, M. R. Buchmeis-
er, Chem. Eur. J. 2004, 10, 777; b) L. Yang, M. Mayr, K.
Wurst, M. R. Buchmeiser, Chem. Eur. J. 2004, 10, 5761;
c) T. S. Halbach, S. Mix, D. Fischer, S. Maechling, J. O.
Krause, C. Sievers, S. Blechert, O. Nuyken, M. R.
Buchmeiser, J. Org. Chem. 2005, 70, 4687.
[9] a) J. H. Oskam, H. H. Fox, K. B. Yap, D. H. McCon-
ville, R. OꢀDell, B. J. Lichtenstein, R. R. Schrock, J. Or-
ganomet. Chem. 1993, 459, 185; b) M. S. Sanford, L. M.
Henling, M. W. Day, R. H. Grubbs, Angew. Chem.
2000, 112, 3593; Angew. Chem. Int. Ed. 2000, 39, 3451.
[10] L. J. Alvey, D. Rutherford, J. J. J. Juliette, J. A. Gladysz,
J. Org. Chem. 1998, 63, 6302.
(A): A two-neck flask was charged with 5 (0.1295 g,
0.539 mmol) and tridecane (0.1000 g, 0.542 mmol), and
flushed with nitrogen. Freshly distilled CH2Cl2 was added
(10.7 mL, giving a 0.050M solution). The solution was stir-
red and 4c (0.0305 g, 0.0136 mmol, 2.6 mol%) was added
against a stream of nitrogen. Samples were periodically
taken by syringe for GC analyses. After 1 h, the sample was
extracted with degassed CF3C6F11 (3”2.0 mL) under nitro-
gen. The fluorous phases were combined in a two-neck
flask. The solvent was removed by oil pump vacuum to give
4c’ as a pinkish residue; yield: 0.0275 g (0.0123 mmol, 90%).
(B): The two-neck flask from cycle A was similarly charg-
ed with
5 (0.1200 g, 0.499 mmol), tridecane (0.0900 g,
0.488 mmol), and CH2Cl2 (10.0 mL, giving a 0.049M solu-
tion). An identical reaction and workup gave 4c’ as a pinkish
residue; yield: 0.0258 g (0.0115 mmol, 85%).
[11] H. Jiao, S. Le Stang, T. Soꢂs, R. Meier, K. Kowski, P.
Rademacher, L. Jafarpour, J.-B. Hamard, S. P. Nolan,
J. A. Gladysz, J. Am. Chem. Soc. 2002, 124, 1516.
[12] R. C. da Costa, J. A. Gladysz, Chem. Commun. 2006,
2619.
(C) The two-neck flask from cycle B was similarly charg-
ed with
5 (0.1117 g, 0.465 mmol), tridecane (0.0871 g,
0.472 mmol), and CH2Cl2 (10.0 mL, giving a 0.047M solu-
tion). An identical reaction and workup gave 4c’ as a pinkish
residue; yield: 0.0172 g (0.00767 mmol, 57%).
[13] R. C. da Costa, Doctoral Dissertation, Universität Er-
langen-Nürnberg, 2006.
[14] M. S. Sanford, J. A. Love, R. H. Grubbs, Organometal-
Partition Coefficients
lics 2001, 20, 5314.
The following is representative, and other data are given
elsewhere.[13] A 10-mL vial was charged with 4a (0.0106 g,
6.46”10À3 mmol), CF3C6F11 (2.000 mL) and toluene
(2.000 mL), fitted with a mininert valve, and vigorously
shaken (2 min). After 2 h (248C), a 0.500 mL aliquot of each
phase was removed. The solvents were evaporated and the
residues dried by oil pump vacuum (2 h). Each residue was
taken up in MeOH (1.000 mL) and analyzed by HPLC
(average of 5 injections, 200”4 mm Nucleosil 100–5 column,
UV/visible detector). The relative peak intensities were
13.2:86.8
[15] R. C. da Costa, F. Hampel, J. A. Gladysz, Polyhedron
2007, 26, in press; DOI 10.1016/j.poly.2006.08.015.
[16] a) W. A. Nugent, J. Feldman, J. C. Calabrese, J. Am.
Chem. Soc. 1995, 117, 8992; b) M. Scholl, S. Ding,
C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953.
[17] S. W. Green, D. S. L. Slinn, R. N. F. Simpson, A. J.
Woytek, in: Organofluorine Chemistry Principles and
Commercial Applications, (Eds.: R. E. Banks, B. E.
Smart, J. C. Tatlow), Plenum Press, New York, 1994,
Chapter 4 and Table 9.
[18] T. A. Kirkland, R. H. Grubbs, J. Org. Chem. 1997, 62,
7310.
[19] J. A. Gladysz, C. Emnet, in: Handbook of Fluorous
Chemistry, (Eds.: J. A. Gladysz, D. P. Curran, I. T.
Horvµth), Wiley-VCH, Weinheim, 2004, Ch. 3.
[20] For some interesting fluorous solvent effects with het-
erogenized non-fluorous ruthenium hydrogenation and
oxidation catalysts, see: a) S. L. Vinson, M. R. Gagnꢃ,
Chem. Commun. 2001, 1130; b) E. Burri, S. M. Leeder,
K. Severin, M. R. Gagnꢃ, Adv. Syn. Catal. 2006, 348,
1640.
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft (DFG, GL
300/3–3), and Johnson Matthey and Materia (precious metal
loans and catalyst donations) for support.
References
[21] An analogous set of recycling experiments with 7 in air
gave poorer results, as detailed elsewhere.[13]
[1] Handbook of Fluorous Chemistry, (Eds.: J. A. Gladysz,
D. P. Curran, I. T. Horvµth), Wiley-VCH, Weinheim,
2004.
[22] S. H. Hong, M. W. Day, R. H. Grubbs, J. Am. Chem.
Soc. 2004, 126, 7414.
[23] a) J. P. Morgan, R. H. Grubbs, Org. Lett. 2000, 2, 3153;
b) D. M. Lynn, B. Mohr, R. H. Grubbs, L. M. Henling,
M. W. Day, J. Am. Chem. Soc. 2000, 122, 6601; c) M. S.
Sanford, L. M. Henling, R. H. Grubbs, Organometallics
1998, 17, 5384; d) E. L. Dias, S. T. Nguyen, R. H.
Grubbs, J. Am. Chem. Soc. 1997, 119, 3887.
[2] L. A. Thompson, Curr. Opinion Chem. Biol. 2000, 4,
324.
[3] a) M. S. Sanford, J. A. Love, R. H. Grubbs, J. Am.
Chem. Soc. 2001, 123, 6543; b) J. A. Love, M. S. San-
ford, M. W. Day, R. H. Grubbs, J. Am. Chem. Soc.
2003, 125, 10103.
Adv. Synth. Catal. 2007, 349, 243 – 254
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
253

Rosenildo CorrÞa da Costa and John A. Gladysz
FULL PAPERS
[24] J. Huang, H.-J. Schanz, E. D. Stevens, S. P. Nolan, Orga-
nometallics 1999, 18, 5375.
organic solvent ratios. This has been interpreted as a
concentration effect with respect to radical trapping,
and represents another way in which fluorous biphase
conditions can be manipulated to favor or disfavor a
reaction, see: D. P. Curran, S. Hadida, S.-Y. Kim, Z.
Luo, J. Am. Chem. Soc. 1999, 121, 6607.
[25] M. Rivard, S. Blechert, Eur. J. Org. Chem. 2003, 2225.
[26] G. S. Forman, A. E. McConnell, R. P. Tooze, W. J. van
Rensburg, W. H. Meyer, M. M. Kirk, C. L. Dwyer,
D. W. Serfontein, Organometallics 2005, 24, 4528.
[27] a) P. E. Romero, W. E. Piers, R. McDonald, Angew.
Chem. 2004, 116, 6287; Angew. Chem. Int. Ed. 2004, 43,
6161; b) P. E. Romero, W. E. Piers, J. Am. Chem. Soc.
2005, 127, 5032.
[31] W. J. Doran, H. A. Shonle, J. Am. Chem. Soc. 1937, 59,
1625.
[32] S. Varray, R. Lazaro, J. Martinez, F. Lamaty, Organo-
metallics 2003, 22, 2426.
[33] H. K. Cammenga, M. Epple, Angew. Chem. 1995, 107,
1284; Angew. Chem. Int. Ed. Engl. 1995, 34, 1171.
[34] The identities of 6,[16] 8,[18] 10,[6] and 12[6] were con-
firmed by preparative experiments, and comparison of
their spectroscopic properties to literature data.
[28] M. El Bakkari, N. McClenaghan, J.-M. Vincent, J. Am.
Chem. Soc. 2002, 124, 12942.
[29] Phase-Transfer Catalysis, (Eds.: C. M. Starks, C. L.
Liotta, M. Halpern), Chapman & Hall, New York,
1994.
[30] The product ratios in certain stoichiometric reactions
of fluorous tin hydrides are a function of the fluorous/
254
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 243 – 254