Oxidative aromatization of olefins with dioxygen catalyzed by palladium trifluoroacetate

Article abstract of DOI:10.1021/jo8016296

(Chemical Equation Presented) Molecular oxygen can replace sacrificial olefins as the hydrogen acceptor in the palladium trifluoroacetate catalyzed dehydrogenation of cyclohexene and related cyclic olefins into aromatics. One of the major drawbacks of the homogeneous system is the tendency of the palladium trifluoroacetate to precipitate as palladium(0) at elevated temperatures. The use of better ligands affords catalysts that can operate at higher temperatures, although they are less reactive than palladium trifluoroacetate.

Full text of DOI:10.1021/jo8016296

Oxidative Aromatization of Olefins with Dioxygen
Catalyzed by Palladium Trifluoroacetate
be possible to replace the sacrificial olefin with O2 in dehydro-
genative reactions. Here, we report our findings about reactions
between palladium trifluoroacetate and olefins in the presence
of O2.
John E. Bercaw,* Nilay Hazari, and Jay A. Labinger*
Arnold and Mabel Beckman Laboratories of Chemical
Synthesis, California Institute of Technology,
Pasadena, California 91125
Whereas precipitation of Pd occurs rapidly upon addition of
cyclohexene to an acetone solution of palladium trifluoroacetate
bercaw@caltech.edu; jal@caltech.edu
8
under inert atmosphere, the same procedure under 1 atm of
1
ReceiVed July 23, 2008
O resulted in a homogeneous orange solution. Analysis by H
2
NMR spectroscopy indicated that after 24 h only benzene had
formed and no cyclohexane was present (eq 2). An additional
broad resonance at δ 4.1 ppm, which grew in over the course
of the reaction, is attributed to water (formed through the
reduction of O2), in exchange with free trifluoroacetic acid
(
formed through protonation of the trifluoroacetate ligand). In
support of this assignment, addition of a small amount of water
to the final solution caused the broad resonance to increase in
intensity and shift slightly in position.
Molecular oxygen can replace sacrificial olefins as the hy-
drogen acceptor in the palladium trifluoroacetate catalyzed
dehydrogenation of cyclohexene and related cyclic olefins
into aromatics. One of the major drawbacks of the homo-
geneous system is the tendency of the palladium trifluoro-
acetate to precipitate as palladium(0) at elevated tempera-
tures. The use of better ligands affords catalysts that can
operate at higher temperatures, although they are less reactive
than palladium trifluoroacetate.
Further confirmation of the role of O2 in the reaction was
obtained by performing reactions with O2 as the limiting reagent.
In these experiments, initially the solution remained clear, and
Two of the major challenges in the functionalization of orga-
1
1
-3
benzene was the only product detected by H NMR spectros-
nic molecules are the activation of C-H bonds
and the
copy. At later stages, once all of the O2 had been consumed, a
black solid precipitated out of solution and formation of
cyclohexane occurred, as a result of the heterogeneous dispro-
utilization of O2 as a terminal oxidant in oxidative transforma-
4
-7
tions.
In both cases the development of practical methods
could have a significant environmental and economic impact
in a number of different areas. An early and intriguing example
of facile C-H bond activation was a 1980 report by Trost and
Metzner, that cyclohexene is catalytically disproportionated into
8
portionation reaction described by Trost. The ratio of direct
conversion of cyclohexene to benzene compared with dispro-
portionation was controlled by the amount of O2 present. For
reactions performed using the same amount of catalyst, solvent,
and cyclohexene under 1 atm of O2 in a J. Young NMR tube,
the ratio of benzene to cyclohexane at the end of the reaction
was approximately 1:1, whereas in a flask with a 10 mL
headspace the ratio was 3.5:1. No disproportionation was
observed so long as excess O2 was present.
1
equiv of benzene and 2 equiv of cyclohexane (eq 1) or
dehydrogenated to benzene using dimethyl fumarate or maleic
acid as sacrificial hydrogen acceptors (giving dimethyl succinate
or succinic acid, respectively) in the presence of an acetone
8
solution of palladium trifluoroacetate. The reactions were
9
proposed to be heterogeneously catalyzed and to involve an
A number of lines of evidence indicate that the dehydroge-
nation with O2 is homogeneously catalyzed. First, in the early
stages of the reaction the solution appears completely homo-
intermediate palladium hydride species. Recently, several reports
have established similarities between the reactions of O2 and
1
0-14
olefins with palladium hydrides,
suggesting that it might
1
5
geneous with no solid particles visible. Over time a fine
precipitate starts to appear, but the formation of benzene is well
(
(
(
(
(
(
(
(
(
1) Crabtree, R. H. J. Chem. Soc., Dalton Trans. 2001, 17, 2437–2450.
2) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507–514.
3) Fekl, U.; Goldberg, K. I. AdV. Inorg. Chem. 2003, 54, 259–320.
4) Stoltz, B. M. Chem. Lett. 2004, 33, 362–367.
5) Sigman, M. S.; Schultz, M. J. Org. Biomol. Chem. 2004, 2, 2551–2554.
6) Stahl, S. S. Angew. Chem., Int. Ed. 2004, 43, 3400–3420.
7) Stahl, S. S. Science 2005, 309, 1824–1826.
1
underway before this begins; monitoring the reaction by H
NMR spectroscopy indicates no initiation period and no
(12) Popp, B. V.; Thorman, J. L.; Stahl, S. S. J. Mol. Catal. A: Chem. 2006,
251, 2–7.
(13) Chowdhury, S.; Rivalta, I.; Russo, N.; Sicilia, E. Chem. Phys. Lett. 2007,
443, 183–189.
8) Trost, B. M.; Metzner, P. J. J. Am. Chem. Soc. 1980, 102, 3572–3577.
9) For confirmation that the reaction is heterogeneous, see: Williams, T. J.;
Caffyn, A. J. M.; Hazari, N.; Oblad, P. F.; Labinger, J. A.; Bercaw, J. E. J. Am.
Chem. Soc. 2008, 130, 2418–2419.
(14) Popp, B. V.; Stahl, S. S. Top. Organomet. Chem. 2007, 22, 149–189.
(15) Of course, this observation does not exclude the possibility of a colloidal
heterogeneous catalyst. The usual “mercury drop” test is uninformative, as
elemental Hg rapidly reduces palladium trifluoroacetate to Pd(0) under reaction
conditions.
(
10) Muzart, J. Chem. Asian J. 2006, 1, 508–515.
(11) Konnick, M. M.; Gandhi, B. A.; Guzei, I. A.; Stahl, S. S. Angew. Chem.,
Int. Ed. 2006, 45, 2904–2907.
8
654 J. Org. Chem. 2008, 73, 8654–8657
10.1021/jo8016296 CCC: $40.75  2008 American Chemical Society
Published on Web 09/27/2008

TABLE 1. Ratio of Benzene to Cyclohexane and Conversion at
Different Temperatures in Reactions between Palladium
Trifluoroacetate and Cyclohexene under 1 atm of O2
Insertion of O2 into the Pd-H bond (in the absence of O2 the
latter would quickly decompose to give Pd metal) then produces
a peroxy intermediate that reacts with trifluoroacetic acid to
regenerate the catalyst. In a similar cycle (not shown) 1,3-
cyclohexadiene would react analogously to give benzene.
Independent experiments using 1,3-cyclohexadiene as a substrate
demonstrate that it is completely oxidized to benzene in 3 h,
under conditions analogous to those described for cyclohexene,
confirming that it is considerably more reactive and would not
accumulate.
a
temp (°C)
ratio C6H6:C6H12
conversion (%)
2
4
6
5
0
0
no cyclohexane
0.96:1
0.82:1
25
37
85
a
Reactions contained 0.016 mmol of palladium trifluoroacetate, 0.32
mmol of cyclohexene, and 0.6 mL of d6-acetone in a J. Young NMR
tube under 1 atm of O2.
Likewise, the absence of detectable hydrogen peroxide could
indicate either that it is a better oxidant than O2 and is rapidly
consumed by a related cycle or that the catalytic system
SCHEME 1
1
7
efficiently decomposes it to O2 and water. Addition of a
stoichiometric (relative to cyclohexene) amount of hydrogen
peroxide to a solution containing cyclohexene and palladium
trifluoroacetate resulted in vigorous bubbling, and after 15 min
little or no peroxide remained (iodide test). After 12 h palladium
black had precipitated out of solution and cyclohexane and
benzene had been formed in an approximately 2:1 ratio
(cyclohexene was also still present), consistent with dispropor-
tionation but no oxidative aromatization having occurred. Clearly
hydrogen peroxide is not a competitive oxidant but instead is
rapidly decomposed under catalytic conditions.
In exploring the scope of the reaction, we found that 1,2-
dihydronaphthalene behaves very similarly to cyclohexene:
reaction under 1 atm of O2 resulted in rapid conversion to
naphthalene (eq 3), whereas the same reaction performed under
argon gave disproportionation to tetralin and naphthalene.
Substituted cyclohexenes exhibit reduced reactivity: none of
evidence for autocatalytic kinetics. Most significantly, the
heterogeneously catalyzed disproportionation of cyclohexene
into benzene and cyclohexane is inhibited by O2. When the
precipitate from a reaction under inert atmosphere was isolated
and added to a fresh solution of cyclohexene in acetone under
O2, no conversion of cyclohexene at all was observed, whereas
under argon the same precipitate was highly active for dispro-
portionation. Hence if the oxidative dehydrogenation is hetero-
geneously catalyzed, there would have to be two different solid
catalysts present in the precipitate: one that converts cyclohexene
directly to benzene using O2 as the oxidant and a second that
disproportionates cyclohexene to benzene and cyclohexane and
is suppressed by O2. It seems far more likely that the former
reaction is homogeneous.
1
-methylcyclohexene, 4-methylcyclohexene, or 3-chlorocyclo-
hexene underwent any reaction at room temperature, but at 60
C (where black solid precipitated) products consistent with both
°
direct oxidation and disproportionation were observed.
For 1-methylcyclohexene and 4-methylcyclohexene the ratios
of toluene to methylcyclohexane were 1.38 and 1.32 to 1,
respectively, with conversions of around 80% (after 24 h). With
3
-chlorocyclohexene the ratio of chlorobenzene to chlorocy-
clohexane was around 1.3 to 1, with competing dechlorination
giving some benzene, cyclohexane, and cyclohexene. These
results demonstrate that a direct oxidation pathway is feasible
for substituted cyclohexenes, but because the reactions must
be run at higher temperatures the heterogeneous disproportion-
ation pathway becomes a significant side reaction.
Consistent with this interpretation, when the reaction was
carried out at higher temperatures in order to increase the rate
of conversion, black solid precipitated out of solution almost
immediately upon heating, and analysis of the reaction mixtures
1
after 24 h by H NMR spectroscopy showed that both benzene
and cyclohexane had formed (Table 1). It appears that although
at room temperature O2 is able to keep the palladium catalyst
almost completely in solution, as the temperature is raised,
precipitation sets in and the heterogeneously catalyzed dispro-
portionation becomes a viable side reaction even in the presence
of O2.
Linear olefins do not exhibit analogous chemistry. No reaction
was observed between palladium trifluoroacetate and 20 equiv
of 1-hexene under 1 atm of O2 at room temperature. When the
reaction mixture was heated a black precipitate crashed out of
solution; when the mixture was maintained at 80 °C for 1 week,
slow isomerization to a mixture of cis- and trans-2-hexene took
place. This process is heterogeneously catalyzed, as the black
solid could be isolated by filtration, washed with acetone, and
then reused successfully to catalyze 1-hexene isomerization, as
well as disproportionation of cyclohexene to benzene plus
cyclohexane. As with the latter, isomerization was inhibited by
O2: it proceeded significantly faster under argon (reaction took
A plausible mechanism for the oxidative dehydrogenation is
shown in Scheme 1, including C-H activation similar to that
8
proposed by Trost for the heterogeneous case and oxidation
related to that proposed by Muzart and Pete for the dehydro-
16
genation of cyclohexanones using O2 as the oxidant. Insertion
of the highly electrophilic palladium into the allylic C-H bond
of cyclohexene, followed by elimination of trifluoroacetic acid,
gives an intermediate that undergoes ꢀ-hydride elimination to
give 1,3-cyclohexadiene and a palladium hydride species.
(
16) Muzart, J.; Pete, J. P. J. Mol. Catal. 1982, 15, 373–376.
(17) Steinhoff, B. A.; Stahl, S. S. J. Am. Chem. Soc. 2006, 128, 4348–4355.
J. Org. Chem. Vol. 73, No. 21, 2008 8655

around 3 days). Presumably, O2 inhibits both reactions by
covering the catalyst surface and blocking olefin coordination,
which is consistent with the O2 inhibition being less significant
at higher temperatures. Surprisingly, neither cis- nor trans-2-
hexene is isomerized by palladium trifluoroacetate at 80 °C.
We believe this is because the sterically more hindered olefins
cannot easily coordinate to the metal, which is presumably the
first step in isomerization.
into aromatics. This represents a significant improvement as O2
is cheap and abundant and produces only water as a byproduct.
It appears that catalysis using O2 as the oxidant occurs
homogeneously, whereas when a sacrificial olefin is used, the
reaction is heterogeneous. One of the major drawbacks of the
homogeneous system is the tendency of the palladium trifluo-
roacetate to precipitate as palladium(0) at elevated temperatures.
The use of better ligands affords catalysts that can operate at
higher temperatures; ongoing studies are aimed at finding ligands
that stabilize the palladium complexes at higher temperatures
without sacrificing activity. Sacrificial olefins are requisite co-
reagents in many instances of low-temperature, homogeneously
We also examined the behavior of 1,2,3,4-tetrahydronaph-
thalene (tetralin), which contains benzylic but no allylic C-H
bonds. There was no reaction of palladium trifluoroacetate with
1
2 equiv of tetralin under 1 atm of O2 at room temperature, but
1
8
heating to 80 °C led to the formation of 0.5 equiv of naphthalene
per palladium (eq 4), along with formation of black precipitate;
the same result was obtained when the reaction was performed
under argon. These results suggest that O2 plays no role in the
reaction and is consistent with the reaction being stoichiometric
in palladium (2 equiv of palladium are required to convert
tetralin to naphthalene). However, when maleic acid (1.5 equiv
relative to tetralin) was added to the reaction mixture (under
argon), the ratio of tetralin to naphthalene at the end of the
reaction was 4:1 and a considerable amount of succinic acid
had been generated. This suggests that hydride transfer from a
palladium species to maleic acid is occurring and in this case
the reaction is catalytic in palladium. The reaction with maleic
acid appears to be heterogeneous and only occurs at elevated
temperatures, presumably because it is more difficult to activate
the C-H bonds in tetralin. The results shown above suggest
that O2 is unable to keep the catalyst in solution at elevated
temperatures, and this may be why O2 is unable to work as a
sacrificial hydride acceptor in this reaction.
catalyzed alkane dehydrogenation; our ultimate goal is to be
able to substitute O2 for a much wider range of substrates.
Experimental Section
Palladium trifluoroacetate (purchased from Sigma Aldrich) and
other organic compounds used in this work were reagent-grade
commercial samples used without further purification. The hydroxy-
II
2+
bridged palladium dimer, [(diimine)Pd (µ
2
-OH)]
2
(diimine )
t
ArNdC(Me)-C(Me)dNAr; Ar ) 3,5- Bu
2
C
H
6 3
), was synthesized
9
1
using a literature procedure. H NMR spectra were recorded at
ambient temperature using a Varian Mercury 300 MHz spec-
trometer. GC analyses were performed on an HP model 6890N
chromatograph equipped with a 10 m × 0.10 mm × 0.40 µm
DB-1 column. GC/MS analyses were performed on an HP model
6890N chromatograph equipped with a 30 m × 25 mm × 0.40
µm HP5-1 column and equipped with an HP 5973 mass selective
EI detector.
Standard Reaction Protocol. The following procedure was used
for a standard reaction catalyzed by palladium trifluoroacetate.
Palladium trifluoroacetate (5 mg, 0.016 mmol) was dissolved in
6
d -acetone (0.6 mL) in a J. Young NMR tube. Cyclohexene (30
µL, 0.32 mmol, 20 equiv) was then added, and the solution
immediately placed in liquid nitrogen. If an organic additive was
used, it was added at the same stage as cyclohexene. The reaction
mixture was degassed using three consecutive freeze-
One approach to finding a palladium catalyst that is less likely
to aggregate as Pd(0) at elevated temperatures is to use better
ligands than trifluoroacetate. We have recently shown that
2
pump-thaw cycles and placed under 1 atm of O using a Schlenk
line (a similar procedure was used to perform control reactions
under argon). Reactions were heated at the appropriate temper-
1
II
2+
ature and monitored using H NMR spectroscopy. At the
complexes of the type [(diimine)Pd (µ2-OH)]2 (diimine )
completion of the reactions the ratio of products was determined
t
ArNdC(Me)-C(Me)dNAr; Ar ) 3,5- Bu2C6H3 or 2,4,6-
1
using both H NMR spectroscopy and gas chromatography (GC).
Me3C6H2) can catalyze the conversion of cyclohexene to
2
In reactions under 1 atm of O at room temperature, solutions
9
benzene under 1 atm of O2 at 60 °C. This reaction appears
were initially orange (regardless of the substrate) and appeared
homogeneous before a fine black precipitate formed over time
(hours to days). When these reactions were heated, a black
precipitate formed much faster (hours) and the solution gradually
became colorless. The rate at which this process occurred
increased as the temperature was raised. In reactions under argon,
solutions were initially orange before a black precipitate formed
and the solution became colorless in less than 15 min. In many
cases the black precipitate was isolated through filtration and
then washed with acetone and hexane before being dried under
reduced pressures.
to be homogeneous, indicating that the ligand stabilizes the
intermediates at higher temperatures. Consistent with the
proposal that O2 and olefins may operate interchangeably as
hydride acceptors, when we perform the reaction between
II
2+
cyclohexene and 5 mol % of [(diimine)Pd (µ2-OH)]2 (diimine
)
t
ArNdC(Me)-C(Me)dNAr; Ar ) 3,5- Bu2C6H3) in the
presence of maleic acid, we see conversion to benzene (ap-
proximately 71% after 4 days at 60 °C), in addition to succinic
acid along with a small amount of cyclohexane. Presumably,
cyclohexene can compete with maleic acid to act as the hydride
acceptor. The greater stability of this N-ligated system also
allows us to use O2 as the oxidant for the conversion of tetralin
to naphthalene at 110 °C, although with low conversion (4%
conversion, 2 turnovers per Pd). Unfortunately, as indicated by
the higher required reaction temperatures, these diimine com-
plexes appear significantly less reactive than palladium trifluo-
roacetate.
The following procedure was used for the reaction between
II
2+
[
(diimine)Pd (µ
2
-OH)]
2
(diimine ) ArNdC(Me)-C(Me)dNAr;
t
Ar ) 3,5- Bu C H ), cyclohexene, and maleic acid. [(Diimine)-
2
6
3
II
2+
Pd (µ
mixture of d
0.1 mL) in a J. Young NMR tube. Cyclohexene (20 µL, 0.197
mmol, 22 equiv per Pd) and maleic acid (69 mg, 0.591 mmol,
equiv to cyclohexene) were then added, and the solution
2
-OH)]
2
(3 mg, 0.002 mmol) was dissolved in a 6:1
4
3
-dichloroethane (0.6 mL) and d -trifluoroethanol
(
3
We have demonstrated that O2 can replace sacrificial olefins
as the hydride acceptor in the palladium trifluoroacetate
catalyzed dehydrogenation of cyclohexene and related species
(
18) For example, see: Jensen, C. M. Chem. Commun. 1999, 24, 2443–2449.
Liu, F. C.; Pak, E. B.; Singh, B.; Jensen, C. M.; Goldman, A. S. J. Am. Chem.
Soc. 1999, 121, 4086–4087.
8
656 J. Org. Chem. Vol. 73, No. 21, 2008

immediately placed in liquid nitrogen. The reaction mixture was
end of the reaction 4% of the tetralin had been converted to
naphthalene, which corresponds to approximately two turnovers
per palladium center.
degassed using three consecutive freeze-pump-thaw cycles and
2
placed under 1 atm of O using a Schlenk line. The reaction
1
was heated at 60 °C and monitored using H NMR spectroscopy.
At the completion of the reactions the ratio of products was
Acknowledgment. We thank Professor Shannon Stahl for
1
determined using both H NMR spectroscopy and GC. A similar
useful discussions. This work was supported by BP through the
II
procedure was utilized for a reaction between [(diimine)Pd (µ
2
-
2
MC program.
2
+
OH)]
4
2
(5 mg, 0.004 mmol) and tetralin (50 µL, 0.368 mmol,
5 equiv per Pd), which was heated for 8 days at 110 °C. At the
JO8016296
J. Org. Chem. Vol. 73, No. 21, 2008 8657