Functionalization of non-activated C-H bonds of alkanes: An effective and recyclable catalytic system based on fluorinated silver catalysts and solvents

Article abstract of DOI:10.1002/chem.201203566

The complexes F_n-Tp 4Bo,3R fAg(L) (F_n-Tp 4Bo,3R f=a perfluorinated hydrotris(indazolyl) borate ligand; L=acetone or tetrahydrofuran) efficiently catalyze the functionalization of non-activated alkanes such as hexane, 2,3-dimethylbutane, or 2-methylpentane by insertion of CHCO ₂Et units (from N₂CHCO₂Et, ethyl diazoacetate, EDA) into their C-H bonds. The reactions are quantitative (EDA-based), with no byproducts derived from diazo coupling being formed. In the case of hexane, the functionalization of the methyl C-H bonds has been achieved with the highest regioselectivity known to date with this diazo compound. This catalytic system also operates under biphasic conditions by using fluorous solvents such as Fomblin or perfluorophenanthrene. Several cycles of catalyst recovery and reuse have been performed, with identical chemo- and regioselectivities. Copyright

Full text of DOI:10.1002/chem.201203566

FULL PAPER
DOI: 10.1002/chem.201203566
À
Functionalization of Non-activated C H Bonds of Alkanes:
An Effective and Recyclable Catalytic System Based on
Fluorinated Silver Catalysts and Solvents
M. ꢀngeles Fuentes,[a] Bianca K. MuÇoz,[b, c] Kane Jacob,[b, c] Laure Vendier,[b, c]
Ana Caballero,*[a] Michel Etienne,*[b, c] and Pedro J. Pꢁrez*[a]
Abstract:
The
complexes
Fn-
The reactions are quantitative (EDA-
based), with no byproducts derived
from diazo coupling being formed. In
the case of hexane, the functionaliza-
been achieved with the highest regiose-
lectivity known to date with this diazo
compound. This catalytic system also
operates under biphasic conditions by
using fluorous solvents such as Fomblin
or perfluorophenanthrene. Several
cycles of catalyst recovery and reuse
have been performed, with identical
chemo- and regioselectivities.
Tp4Bo,3R Ag(L) (Fn-Tp4Bo,3R =a per-
fluorinated hydrotris(indazolyl) borate
ligand; L=acetone or tetrahydrofuran)
efficiently catalyze the functionaliza-
tion of non-activated alkanes such as
hexane, 2,3-dimethylbutane, or 2-meth-
ylpentane by insertion of CHCO2Et
units (from N2CHCO2Et, ethyl diazo-
À
f
f
À
tion of the methyl C H bonds has
Keywords: alkanes · biphasic catal-
ysis · carbenes · fluorous catalysis ·
silver
ACHTUNGTRENNUNGacetate, EDA) into their C H bonds.
Introduction
The functionalization of non-activated alkanes remains one
of the most challenging goals of modern catalysis,[1] because
their high inertness usually preclude their conversion into
value-added products by simple reactions and under mild
conditions. Yet, their availability as raw materials make
them attractive starting materials in reactions in which new
[2]
À
À
C C or C X bonds are formed. Despite considerable ef-
forts in the last few decades, only a few catalytic systems
based on organometallic activations that operate with
À
simple, non-activated C H bonds have been described;
[a] M. ꢀ. Fuentes, Dr. A. Caballero, Prof. P. J. Pꢁrez
Laboratorio de Catꢂlisis Homogꢁnea
Departamento de Quꢃmica y Ciencia de los Materiales
Unidad Asociada al CSIC, Centro de Investigaciꢄn
en Quꢃmica Sostenible (CIQSO)
À
Scheme 1. Alkane C H bond catalytic-functionalization processes involv-
À
À
ing the formation of M C and/or M H bonds.
among them, the electrophilic activation (Scheme 1a),[3] the
Campus de El Carmen s/n
Universidad de Huelva, 21007-Huelva (Spain)
Fax : (+34)959219942
E-mail: ana.caballero@dqcm.uhu.es
perez@dqcm.uhu.es
[4]
À
À
C H/C D exchange (Scheme 1b), the alkane dehydrogen-
ation (Scheme 1, c),[5] and the alkane borylation
(Scheme 1d).[6] These transformations occur through the in-
teraction of the carbon–hydrogen bonds with the metal
[b] Dr. B. K. MuÇoz, K. Jacob, Dr. L. Vendier, Prof. M. Etienne
CNRS; LCC (Laboratoire de Chimie de Coordination)
205, route de Narbonne, BP44099
À
À
center, and the subsequent formation of M C and/or M H
bonds.[7]
F-31077 Toulouse Cedex 4 (France)
Precisely, the formation of these linkages has been in-
voked as the explanation for the lack of a larger number of
catalytic systems based on this strategy: their high stability
and inertness usually precludes the formation of catalytic
cycles beyond their formation.[1,2] The alkane oxygenation
reaction has also provided neat alkane functionalization al-
though through different pathways.[1]
[c] Dr. B. K. MuÇoz, K. Jacob, Dr. L. Vendier, Prof. M. Etienne
Universitꢁ de Toulouse
UPS, INPT; LCC
F-31077 Toulouse Cedex 4 (France)
E-mail: michel.etienne@lcc-toulouse.fr
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201203566.
Chem. Eur. J. 2013, 19, 1327 – 1334
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1327

An alternative approach consists of avoiding the direct in-
À
teraction of the C H bonds with the metal center, the func-
tionalization of the alkane being achieved through the inter-
action of an electrophilic ligand with that bond. Thus, the
well-known capabilities of a number of transition-metal
complexes to form metallocarbene species upon reaction
with diazocompounds has been employed in that sense,[8]
À
the carbene unit being transferred and inserted into the C
H bond (Scheme 2a). This methodology was first reported
vents,[18] met with limited success, since the catalytic activity
dropped significantly from the first cycle to the next. Based
on the latter report by Endres and Maas[18] with fluorinated
Rh2L4 complexes, and the interest of fluorous biphasic catal-
ysis (FBC)[19] in terms of green chemistry, we decided to
study the potential of a series of perfluorinated tris(indazo-
lyl)borate complexes of silver in the alkane functionalization
reaction with ethyl diazoacetate under FBC conditions. We
have found that with the appropriate election of the cata-
lysts and of the fluorous phase, quantitative (diazo-based)
conversions of alkanes into the corresponding carboxylate
derivatives can be achieved for several cycles of catalyst sep-
aration and reuse, yet an unprecedented feature in the area
of alkane functionalization by carbene insertion.
Scheme 2. Carbon–hydrogen bond functionalization by carbene insertion
following the diazo-compound strategy.
by Scott and DeCicco[9] nearly 40 years ago, and was later
expanded, but yet not exhaustively exploited.[10] Subsequent
work by the groups of Noels,[11] Callot,[12] Davies[13] and
Che[14] demonstrated the potential of dirhodium tetraacetate
and related Rh2L4 complexes to catalyze the functionaliza-
tion of plain, non-activated alkanes such as n-hexane into
carboxylate derivatives (Scheme 2b).
Our group has developed catalytic systems for the alkane
functionalization by the carbene insertion methodology
based on Group 11 metal complexes,[10c,d] bearing either
tris(pyrazolyl)borate or N-heterocyclic carbene ligands.
Along the years we have improved the catalytic activity of
our catalysts, by 1) increasing the electrophilic nature of the
metal center by employing anchoring ligands with low
donor capabilities and 2) providing steric protection of the
metal center. As a result of such progress, we have recently
described the catalytic functionalization of methane,[15] the
most inert alkane, by using silver complexes containing per-
fluorinated tris(indazolyl)borate ligands as catalysts
[Eq. (1)], ethyl diazoacetate as the carbene source and su-
percritical carbon dioxide as the reaction medium.
Results and Discussion
Synthesis and characterization of the silver complexes Fn-
Tp4Bo,3R AgL: Previous work from our laboratories led to the
f
synthesis of the complexes F21-Tp4Bo,3CF Ag
(acetone) (1)[20,21]
3
ACTHNUTRGNEUNG
and F27-Tp4Bo,3CF CF Ag(THF) (2).[15] Given the well-known
2
3
solubility of complexes containing fluorinated ponytails in
fluorinated solvents, we decided to synthesize three more
complexes with ligands similar to those above, bearing
longer fluorinated alkyl chains with three, four, and six
carbon atoms. As a general protocol, complexes 3–5 were
synthesized under inert atmosphere by the reaction of the
corresponding thallium salts TlFn-Tp4Bo,3R with silver triflate
f
in acetone or tetrahydrofuran as solvent (Scheme 3), in the
same manner as that previously reported for compounds 1
The results described to date by the different
groups[10–14,16] involved in this area demonstrate that this
methodology constitutes an efficient tool for the effective
carbon–hydrogen bond functionalization of plain, non-acti-
vated alkanes. However, to facilitate the use of these cata-
lysts for such a purpose, an effective catalyst separation
methodology should be developed. Very few reports based
on biphasic catalysis with ionic liquids[17] or fluorinated sol-
Scheme 3. Synthesis of the tris(indazolyl)borate silver complexes em-
ployed in this work.
1328
www.chemeurj.org
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1327 – 1334

Recyclable Catalytic Systems
FULL PAPER
and 2. Yields of the isolated product ranged within 72–91%;
the complexes being obtained as white solids of composition
Fn-Tp4Bo,3R Ag
C
f
Table 1 contains selected 19F NMR data for compounds 3–
5. At room temperature, most resonances appear as broad
Table 1. 19F NMR chemical shifts (d) and JFF (Hz) of silver complexes 3–
5 in [D6]acetone.
3a
4
5
AHCTNUTGERG(NNNU CF2)2CF3
Figure 1. ORTEP diagram of F33-Tp4Bo,3
AgACTHNGUTERN(UNG THF) (3b). Hydrogen
F4
F5
À145.47 (m)
À145.32 (m)
À145.14 (m)
À165.58 (dd)
3JF5-F6 =3JF5–F4 =15
À154.13 (brs)
À158.2 (brs)
CaÀ107.79 (m)
Cb,gÀ122.13 (m)
CdÀ123.55 (m)
CeÀ126.96 (m)
À81.86 (t)
atoms have been omitted for clarity. Thermal ellipsoids are drawn at the
30% probability level. Selected bond lengths [ꢆ] and angles [8] Ag1–O1
2.228(5), Ag1–N1 2.378(5), Ag1–N3 2.377(5), Ag1–N5 2.406(5); O1-Ag1-
N3 131.10(19), O1-Ag1-N1 126.3(2), N3-Ag1-N1 83.47(17), O1-Ag1-N5
134.90(19), N3-Ag1-N5 80.69(17), N1-Ag1-N5 81.67(17).
À166.68 (brs)
À166.29 (dd)
3JF5–F6 =3JF5–F4 =17
À154.78 (brs)
À160.67 (brs)
CaÀ107.58 (m)
CbÀ123.32 (brs)
CgÀ126.34 (brs)
F6
F7
CF2
À155.12 (brs)
À161.31 (brs)
CaÀ108.17 (brs)
CbÀ127.17 (d)
4J=9.8
that of TpACHUTGNTER(GNNUN CF ) Ag
ACHTNUTRGENG(UN THF) (2.234(4) ꢆ). A C2F5 end (C21
3
2
À
CF3
À81.13 (m)
4J=9.8
À82.17 (t)
4J=10
À
C22) in the C3F7 chain of the N3 N4 indazolyl group was
highly disordered and was only isotropically refined.
4J=10.5
We have already reported that the reaction of complex 1
with carbon monoxide in dichloromethane formed the cor-
[20]
3
signals, a feature that is enhanced when the fluorinated po-
nytail is enlarged. This effect is similar to that reported for
perfluorobutyric and perfluorooctanoic acids.[22] The spectra
are quite simple since they contain, in all cases, a unique set
of resonances for the three indazolyl rings, as a consequence
of the C3v symmetry in solution. The four different aromatic
responding silver-carbonyl adduct F21-Tp4Bo,CF Ag(CO),
a
non-classical carbonyl complex[24] that was spectroscopically
and structurally characterized. The IR spectrum (dichloro-
methane) showed a strong absorption centered at 2167 cmÀ1
assigned to n(CO), slightly higher than that reported for Tp-
Ag(CO) in the same solvent (n(CO)=2164 cmÀ1).[23b]
AHCTUNGTERG(NNUN CF3)2
À
C F nuclei give rise to four resonances within the range d=
Similar treatment of complexes 2–5 with CO led to the in
À166 to À145 ppm, whereas the fluorinated ponytails exhib-
it two distinct sets of resonances for the CF3 (at ca. d=
À82 ppm) and the CF2 groups. For the latter, the fluorine
nuclei resonate around d=À120 ppm, and each added CF2
group shifts to higher field. Thus, for the longer chain in F51-
situ observation of the corresponding series of carbonyl ad-
ducts Fn-Tp4Bo,3R Ag(CO), for which we focused on their
f
n(CO) values. As shown in Scheme 4, there is no change in
the n(CO) value with the number of CF2 groups sequentially
added to the ligand. Therefore, this group of ligands pro-
vides similar electron density at the metal center, whereas
somewhat different steric requirements could be expected.
Tp4Bo,3
AgACHTUNGTRENNUNG(CH3COCH3) (5), the four CF2 resonances
ACHTUNGTRENNUNG(CF2)5CF3
appear at d=À107.8, À122.1, À122.1 À123.5 and
À126.9 ppm, for CaF2, Cb, Cg, Cd, and Ce, respectively, with
Ca directly bonded to the indazolyl skeleton.
Catalytic functionalization of carbon–hydrogen bonds of al-
kanes under homogeneous conditions: The tris(indazolyl)-
borate silver complexes 1–5 have been employed as catalysts
Crystallization of complex F33-Tp4Bo,3
AgACTHNGUTERN(UNG THF) (3b)
AHCTNUTGERG(NNNU CF2)2CF3
gave single crystals suitable for X-ray diffraction studies. As
shown in Figure 1, the silver center is coordinated to the
three nitrogen donors of the tris(indazolyl)borate ligand in a
k3-N,N’,N’’ fashion, as well as to the O atom of tetrahydro-
furan. The angles around the metal center indicate a distort-
ed tetrahedral geometry. The distances and angles observed
are very similar to those already reported for other silver
complexes containing perfluorinated indazolylborate ligands.
À
The average Ag N distance of 2.387 ꢆ compares well with
that of TpACHTUNGTRENNUNG(CF ) Ag
G
3
2
than the same distance in F21-Tp4Bo,CF Ag(CO) (2.310 ꢆ).
[20]
3
Scheme 4. The IR data (dichloromethane solution) for the series of car-
bonyl adducts showing similar electron density at silver.
À
The Ag O bond length of 2.228(5) ꢆ is also very close to
Chem. Eur. J. 2013, 19, 1327 – 1334
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1329

P. J. Pꢁrez, A. Caballero et al.
in the reaction of n-hexane with ethyl diazoacetate
(N2CHCO2Et, EDA), by using the alkane as the reaction
solvent. Complex 1 had already been tested in this transfor-
mation, leading to the formation of a mixture of three prod-
ucts derived from the insertion of the CHCO2Et group into
ACHTUNGTRENNUNG
Catalytic functionalization of carbon–hydrogen bonds of al-
kanes. Screening of the fluorous phase: Having demonstrat-
ed the catalytic ability of complexes 1–5 for the functional-
À
À
the three different types of hexane C H bonds (Scheme 5
ACHUTNGERNiNUG zaAHCTUNGTRNENtUGN ion of non-activated C H bonds of alkanes by carbene
insertion, we focused on the use of fluorous biphasic condi-
tions. We chose complex 4 as a representative sample of our
catalysts and hexane as a probe, with several fluorinated
substrates as the fluorous phase: 1,1,1-trifluorotoluene, per-
fluorotoluene, Fomblin, and perfluorophenanthrene.
A
series of two cycles (in which the fluorous phase was sepa-
rated and reused) were then performed, with the results
shown in Table 3.
Scheme 5. The functionalization of hexane with ethyl diazoacetate by
using complexes 1–5 as catalysts.
Table 3. Screening of the fluorous phase in the functionalization of
hexane with EDA and complex 4 as the catalyst.
Entry
Fluorous
phase
Run
% P1
% P2
% P3
Yield [%][a]
and Table 2). This is now being observed with the other four
compounds 2–5, that induced the complete chemoselectivity
toward the alkane functionalization products; no diethyl fu-
1
2
none
–
45
45
10
>99
1
2
44
44
41
41
15
15
>30[b]
>30[b]
1
2
43
43
41
41
16
16
>99
>99
Table 2. Catalytic functionalization of hexane with EDA and complexes
3
4
1–5 as catalysts.[a]
Fomblin
1
2
46
44
41
42
13
14
>99
>99
Catalyst
F21-Tp4Bo,3CF Ag (1)
% P1
% P2
% P3
Yield [%]
3
36
44
45
45
47
49
45
43
45
42
15
10
12
10
11
>99
>99
>99
>99
>99
F27-Tp4Bo,3CF CF Ag (2)
1
2
46
49
40
38
14
13
>99
>99
2
3
5
F33-Tp4Bo,3
F39-Tp4Bo,3
F51-Tp4Bo,3
Ag (3)
Ag (4)
Ag (5)
ACHTUNGTRENNUNG(CF2)2CF3
ACHTUNGTRENNUNG
[a] [Ag]/ACHTUNTRGNEU[GN EDA]=1:20. Reaction time: 12 h. See the Experimental Section
for details. Yields and regioselectivities were determined by GC.
[b] Other products from arene ring activation were observed.
ACHTUNGTRENNUNG
[a] [Ag]/ACHTUNGTRENNUNG[EDA]=1:100. Reaction time: 12 h. See the Experimental Sec-
tion for details. Yields and regioselectivities were determined by GC.
marate or maleate (derived from the catalytic coupling of
two carbene moieties) were detected by GC at the end of
the reaction even if the EDA was added in one portion, in
contrast with most of the other reported catalytic sys-
tems.[8,10] Even more interestingly, the progressive increase
of the fluorinated chain-length of the ligand caused an in-
crease of the amount of the product derived from the func-
tionalization of the methyl group, reaching up to 47% in the
With trifluorotoluene (Table 3, entry 2) the functionaliza-
tion of the arene ring as well as the targeted hexane func-
tionalization was observed. To avoid this undesired reaction,
perfluorinated toluene was considered (Table 3, entry 3).
Yields and regioselectivity were similar to those in the ho-
mogeneous system (Table 3, entry 1). However, the relative-
ly low boiling point of this solvent originated problems
during the workup. The next fluorous phases were therefore
chosen with high boiling points: Fomblin (Table 3, entry 4),
À
case of 5. This translates to a normalized (per C H bond)
primary/secondary regioselectivity of 1:0.84. Dias and co-
a
fluorinated polyester, and perfluorophenanthrene
workers[16] reported
a
1:1.43 ratio for pentane with
(Table 3, entry 5). In both cases, the reactions with hexane
were quantitative into the functionalization products and
the regioselectivity almost identical to that of the homoge-
neous system. On the basis of this screening, the latter two
fluorous solvents were chosen for a in depth study of the
catalytic activity of complexes 1–5 under biphasic conditions
with a series of linear and branched alkanes and ethyl diazo-
acetate as the carbene source.
TpACHTUNGTRENNUNG(CF ) Ag
ACHTUNGTRENNUNG(THF) as the catalyst, a value close to that of com-
3
2
plex 1 (1:1.33), also bearing a CF3 substituent.
Thus the difference in regioselectivity is a double conse-
quence of electronic and steric effects, an indication for
future catalyst design toward enhanced primary functionali-
zation: the more electrophilic and the more sterically crowd-
À
ed the catalyst, the more selective it is for primary C H
bonds. To the best of our knowledge, this is the highest re-
gioselectivity toward the functionalization of primary sites
in linear alkanes with this methodology using ethyl diazo-
General catalytic functionalization of alkanes under fluorous
conditions: This study has been performed (Scheme 6) with
1330
www.chemeurj.org
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1327 – 1334

Recyclable Catalytic Systems
FULL PAPER
In each case, four to five cycles of catalyst separation and
reuse have been carried out to ascertain the degree of recy-
clability of this catalytic system (Figure 2). Overall, 135 dif-
ferent experiments have been performed, from which the
following trends have been extracted. First, the high catalyt-
ic activity of complexes 1–5 have led to the functionalization
of nearly all available different sites in the alkanes investi-
gated. In addition to the already commented three products
derived from hexane (P1–P3, Scheme 5), two products from
2,3-dimethylbutane (P4 and P5, Scheme 7) and four products
from 2-methylpentane (P6–P9, Scheme 7) have been ob-
tained. As observed previously, only one type of secondary
site of 2-methylpentane (that which is remote from the terti-
ary carbon), could be derivatized.[20]
Scheme 6. Catalysts, fluorous solvents, substrates, and number of cycles
used in this study.
The general procedure consisted of the dissolution of the
catalyst 1–5 (0.005 mmol) in the fluorous phase (4 mL) fol-
lowed by the addition of the alkane (1 mL) and ethyl diazo-
acetate (0.1 mmol, 20 equiv with respect to the catalyst).
three alkanes (hexane, 2,3-dimethylbutane and 2-methylpen-
tane), two fluorous phases (Fomblin and perfluorophenan-
threne), and five catalysts (complexes 1–5).
Figure 2. The catalytic functionalization of hexane, 2,3-dimetylbutane and 2-methylpentane in Fomblin or perfluorophenanthrene: regioselectivity as a
function of catalyst type and catalyst recycling.
Chem. Eur. J. 2013, 19, 1327 – 1334
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1331

P. J. Pꢁrez, A. Caballero et al.
Scheme 7. The catalytic functionalization of 2,3-dimethylbutane (top) and
2-methylpentane (bottom).
This mixture was homogeneous at room temperature, at var-
iance with other biphasic fluorous systems that need heating
to be homogeneous, and was vigorously stirred until no
EDA was detected by GC. In all cases, no products derived
from carbene coupling, that is, diethyl fumarate or maleate,
were detected; thus the alkane functionalization is com-
pletely chemoselective. Recent theoretical studies with the
related TpBr M (M=Cu, Ag) complexes have explained
3
such selectivity and the lack of carbene coupling.[25] To ach-
ieve catalyst recycling and reuse we have performed two dif-
ferent strategies. The first one required cooling down the
mixture to À208C, inducing the formation of two phases,
and the subsequent separation (through a cannula) of the
organic phase containing the products as well as the excess
of the alkane. However, this procedure revealed that some
catalyst leached out of the fluorous phase. Because of this, a
second, more effective procedure was applied, based on a
trap-to-trap distillation under vacuum of the final reaction
mixture: the excess of the alkane as well as the functional-
ized products were separated from the fluorous phase in this
manner, the latter completely retaining the soluble catalyst.
Figure 2 depicts the composition of the final mixtures upon
EDA consumption for the three alkanes, the five catalysts
and the different cycles. A simple look at the data nicely
demonstrates that regioselectivity is maintained all along
the cycles; four in the case of Fomblin and five in the case
of the perfluorophenanthrene. More interestingly, these
values are identical to those obtained under homogenous
conditions, and did not significantly vary from one fluorous
phase to another (see the Supporting Information for the
complete set of values for each experiment).
Figure 3. Variation of the reaction time required for complete EDA con-
sumption along the number of cycles of catalyst recovery, with perfluoro-
phenanthrene as the fluorous phase
catalysts. Beyond that point, the stability of the catalysts
slightly varied for the second and third cycles, most of them
completing the reaction in less than 2 h. The fourth and fifth
cycles showed differences between complexes 1–5 in terms
of stability: complex 1 required the longest reaction times
for the three alkanes, whereas complex 2 provided very
good yields in relatively short reaction times upon five
cycles (1 h for 2-methylpentane, 2 h for 2,3-dimethylbutane
or 3 h for hexane, after 5 cycles each). Complexes 3–5
showed reaction times within the range described by 1 and
2. Therefore, there is not a simple correlation between the
lengths of the fluorinated Rf chain with catalyst stability
under the reaction conditions.
Since the above study was carried out with 0.1 mmol of
EDA as the limiting reagent, we have scaled-up this trans-
formation by a tenfold, employing 1 mmol of EDA. As rep-
resentative examples we have chosen hexane and 2,3-di-
The variation of the reaction times along the cycles of cat-
alyst recovery deserves some comment (Figure 3). In all
cases, the first cycle was completed in less than 1 h for most
ACHUTNGERNmNUG ethAHCTUNGTRNENyUGN lbutane as the reactants, and perfluorophenanthrene
1332
www.chemeurj.org
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1327 – 1334

Recyclable Catalytic Systems
FULL PAPER
After filtration through celite the solvent was evaporated and the solid
was washed several times with pentane to afford 3a as a white powder
(36 mg, 72%). 19F NMR (282.4 MHz, [D6]acetone): d=À166.68 (brs, 1F,
F5), À161.31 (brs, 1F, F7), À155.12 (brs, 1F, F6), À145.47 (m, 1F, F4),
À127.17 (d, J=10 Hz, 2F, CF2), À108.17 (m, 2F, CF2), À81.14 ppm
(pseudo-t, J=9 Hz, 3F, CF3). DCI-MS: m/z: calculated for
C30H2AgBF33N6: 1190.90; found: 1191.00 [MÀacetone]; elemental analy-
sis calcd (%) for C33H7AgBF33N6O: C 31.73, H 0.56, N 6.73; found: C
31.95, H 1.0, N 6.62. Complex 3b was prepared following the same proce-
dure but using tetrahydrofuran instead of acetone as the reaction solvent.
as the fluorous phase. The catalyst activity and the regiose-
lectivities are again maintained for several cycles (see the
Supporting Information), showing the validity of this system
at a higher scale. This set of experiments was carried out by
using a 1:40 catalyst/EDA ratio in contrast to the previous
experiments that were performed with a 1:20 ratio. There-
fore, these scale-up experiments are even more efficient (a
turnover number (TON) per cycle of 40) than the former
(20 TON per cycle) and confirm the absence of byproducts.
AHCTNUTGERG(NNNU CF2)3CF3
Synthesis of F39-Tp4Bo,3
AgACHTUNGRTNENG(U acetone) (4): Following the same proce-
(CF2)3CF3
dure as for 3a, F39-Tp4Bo,3
Tl (50 mg, 0.035 mmol) and silver triflate
A
(8.5 mg, 0.033 mmol) yielded complex 4 as a white powder (45 mg, 91%).
1H NMR
(300.13 MHz, [D6]acetone): d=2.84 ppm (m, 6H,
Conclusion
CH3C(O)CH3); 19F NMR (282.4 MHz, [D6]acetone): d=À166.29
(pseudo-t, J=17 Hz, 1F, F5), À160.66 (brs, 1F, F7), À154.8 (brs, 1F, F6),
À145.32 (m, 1F, F4), À126.31 (m, 2F, CF2), À123.32 (m, 2F, CF2), À107.58
(m, 2F, CF2), À82.17 ppm (pseudo-t, J=10 Hz, 3F, CF3); 11B NMR
(75.47 MHz, [D6]acetone): d=0.3 ppm (br, BH); elemental analysis calcd
(%) for C36H7AgBF39N6O: C 30.90, H 0.50, N 6.01; found: C 30.87, H
0.38, N 6.34.
We have found that the fluorinated tris(indazolyl)borate
silver complexes Fn-Tp4Bo,3R AgL (1–5) efficiently catalyze
f
the transfer of the CHCO2Et group from ethyl diazoacetate
into the C H bonds of hexane, 2,3-dimethylbutane or 2-
À
methylpentane. The transformation is completely chemose-
lective, with no observation of any product derived from the
homocoupling of the carbene groups. In the case of hexane,
the catalyst bearing the longest n-C6F13 ponytail has yielded
the highest regioselectivity toward the primary sites report-
ed with ethyl diazoacetate.
The use of a fluorous phase (Fomblin or perfluorophenan-
threne) has provided a recyclable system that maintains
both the chemo- and regioselectivity observed in the homo-
geneous phase. In addition, the catalysts can be separated
and reused four or five times without loss of both selectivi-
ties, and in a number of cases, with a relatively small in-
crease of the reaction times.
Synthesis of F51-Tp4Bo,3
AgACTHNGUTERN(UNG acetone) (5): Following the same proce-
AHCTNUTGERG(NNNU CF2)5CF3
dure as for 3a, F51-Tp4Bo,3
Tl (110 mg; 0.063 mmol) and silver trif-
AHCTNUTGERG(NNNU CF2)5CF3
AHCTUNGERTGlNNUN ate (14.8 mg; 0.057 mmol) yielded complex 4 as a white powder (79 mg,
82%). 1H NMR (300.13 MHz, [D6]acetone): d=2.85 ppm (m, 6H,
CH3C(O)CH3). 19F NMR (282.4 MHz, [D6]acetone): d=À165.58
(pseudo-t, J=15 Hz, 1F, F5), À159.48 (brs, 1F, F7), À154.13 (brs, 1F, F6),
À145.14 (m, 1F, F4), À126.96 (brs, 2F, CF2), À123.55 (brs, 2F, CF2),
À122.13 (brs, 4F, CF2CF2), À107.80 (brs, 2F, CF2), À81.86 ppm (pseudo-t,
J=10.5 Hz, 3F, CF3); 11B NMR (75.47 MHz, [D6]acetone): d=À0.7 ppm
(br, BH); elemental analysis calcd. (%) for C42H7AgBF51N6O: C 29.69, H
0.42, N 4.95; found: C 29.53, H 0.30, N 4.71.
X-ray analysis of 3b: Data for 3b was collected at 180 K on an Xcalibur
Oxford Diffraction diffractometer equipped with an Oxford Instrument
Cooler Device using a graphite-monochromated MoKa radiation (l=
0.71073 ꢆ). The structure has been solved by direct methods using
SIR92,[26] and refined by means of least-squares procedures on F2 with
the aid of the program SHELXL97[27] included in the software package
WinGX version 1.63.[28] Hydrogen atoms were geometrically placed and
refined by using a riding model. All but a few non-hydrogens atoms were
anisotropically refined, and in the last cycles of refinement a weighting
Scheme was used, in which weights are calculated from the following for-
mula: w=1/[s
ripheral CF2CF3 was too highly disordered to be refined in anisotropic
mode. The disorder was treated by using the “part” option of
SHELXL97 and the atoms C(21), C(22), F(10), F(11), F
Experimental Section
General procedures: All syntheses requiring inert atmosphere (Ar or N2)
were carried out by using Schlenk tube or glovebox techniques. The sol-
vents were dried and distilled using a SBS-Mbraun system or by conven-
tional methods: diethyl ether and tetrahydrofuran (Na/benzophenone),
dichloromethane and pentane (calcium hydride). Acetone was degassed
2
N
ACHUTGTNREN(NUG 12 A), FACHTUNGTRENNUNG(12B),
and stored on molecular sieves under argon. TlFn-Tp4Bo,3R complexes
f
F
A
ACHTUNGTRENNUNG
were prepared according to the method previously reported.[15] A full ac-
count of their syntheses and properties will be reported in due course.
ble for the generation of alerts in the chekcif. Crystal data and refine-
ment: crystal size 0.2ꢇ0.12ꢇ0.03 mm, monoclinic, P2/n, a=18.3960(11),
b=13.0560(5), c=19.0630(11) ꢆ, b=117.290(7)8, V=4068.9(4) ꢆ3, Z=4,
The complexes F21-Tp4Bo,3CF Ag
N
(THF)
3
2
3
were prepared according to the literature.[15,20] The silver triflate, alkanes,
fluorous phases, and ethyl diazoacetate were purchased from Aldrich and
employed without further purification. GC data were collected with a
Varian 3900 instrument. NMR experiments were acquired at 298 K using
ARX250, DPX300, AV300, AV400 Bruker spectrometers and a Varian
Mercury 400 MHz. Elemental analyses were performed in the Analytic
Services of our laboratory. Mass spectrometry measurements were re-
corded on a QTRAP Applied Biosystems Mass Spectrometer.
1calcd =2.062 MgmÀ3, q range=2.94 to 25.688, F
collected/unique 28767/7725 [R
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
25.688=99.8%, semiempirical absorption correction from equivalents
(m=0.689 mmÀ1, max./min. transmission=0.978/0.932), data/restraints/pa-
rameters=7725/9/650, gof on F2 =1.064, final R indices [I>2s(I)] R1=
0.0637, wR2=0.180, R indices (all data) R1=0.098, wR2=0.1956, largest
diff. peak/hole=2.328/À1.895e ꢆÀ3. The drawing of Figure 1 was per-
formed with the program ORTEP32[29] with 30% probability displace-
ment ellipsoids for non-hydrogen atoms.
CCDC-891320 (3b) contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif
Synthesis of silver complexes
ACHTUNGTRENNUNG(CF2)2CF3
Synthesis of F33-Tp4Bo,3
Ag(L): L=acetone (3a) or THF (3b): In a
ACHTUNGTRENNUNG(CF2)2CF3
glove-box, F33-Tp4Bo,3
Tl (50 mg, 0.039 mmol) and silver triflate
General catalytic experiment under homogeneous conditions: Fn-
Tp4Bo,3R Ag
C
f
(10 mg, 0.039 mmol) were mixed in a Schlenk tube covered with alumi-
num foil and acetone (5 mL) was added. The solution was stirred at
room temperature for 2 h, then the solvent was removed under reduced
pressure and the residue was extracted into dichloromethane (10 mL).
ing alkane (10 mL). Ethyl diazoacetate (52 mL, 0.5 mmol) was added in
one portion. The reaction was followed by GC. After stirring overnight,
no ethyl diazoacetate was detected. The products were identified by
Chem. Eur. J. 2013, 19, 1327 – 1334
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1333

P. J. Pꢁrez, A. Caballero et al.
NMR spectroscopy and GC studies. Quantification was carried out by
GC using calibration curves previously registered with pure products.
The Supporting Information contains GC traces, as well as NMR data of
reaction mixtures showing the lack of diethyl fumarate and maleate.
[9] L. T. Scott, G. J. DeCicco, J. Am. Chem. Soc. 1974, 96, 322–323.
[10] a) H. M. L. Davies, A. R. Dick, Top. Curr. Chem. 2010, 292, 303–
345; b) M. M. Dꢃaz-Requejo, P. J. Pꢁrez, Chem. Rev. 2008, 108,
3379–3394; c) M. M. Dꢃaz-Requejo, T. R. Belderraꢃn, M. C. Nicasio,
P. J. Pꢁrez, Dalton Trans. 2006, 5559–5566; d) H. M. L. Davies,
R. E. J. Beckwith, Chem. Rev. 2003, 103, 2861–2903.
[11] a) A. Demonceau, A. F. Noels, A. Hubert, P. Teyssiꢁ, J. Chem. Soc.
Chem. Commun. 1981, 688–689; b) A. Demonceau, A. F. Noels, A.
Hubert, P. Teyssiꢁ, Bull. Soc. Chim. Belg. 1984, 93, 945–948; c) A.
Demonceau, A. F. Noels, P. Teyssiꢁ, A. Hubert, J. Mol. Catal. 1988,
49, L13–L17; d) A. Demonceau, A. F. Noels, A. Hubert, J. Mol.
Catal. 1989, 57, 149–152.
[12] H. J. Callot, F. Metz, Tetrahedron Lett. 1982, 23, 4321–4324.
[13] a) H. M. L. Davies, T. Hansen, M. R. J. Churchill, J. Am. Chem. Soc.
2000, 122, 3063–3070; b) H. M. L. Davies, T. Hansen, J. Am. Chem.
Soc. 1997, 119, 9075–9076; c) H. M. L. Davies, E. G. Antoulinakis, J.
Organomet. Chem. 2001, 617–618, 47–55.
[14] H.-Y. Thu, G. S.-M. Tong, J.-S. Huang, S. L.-F. Chan, Q.-H. Deng,
C.-M. Che, Angew. Chem. 2008, 120, 9893–9897; Angew. Chem. Int.
Ed. 2008, 47, 9747–9751.
The yields of isolated products from experiments carried out at a higher
scale[20] were obtained within the range 88–93% upon column chroma-
tography using petroleum ether/ethyl acetate as eluent (5:1).
General catalytic experiment in a biphasic system: Fomblin HVAC 140/
13 as a fluorous medium: Fn-Tp4Bo,3R Ag
C
f
dissolved in Fomblin HVAC (4 mL, 1.2 mmol) and the corresponding
alkane (1 mL, 7.6 mmol) and ethyl diazoacetate (10.5 mL, 0.1 mmol) were
added at room temperature, leading to a homogeneous mixture. After a
first catalytic run, the products and starting material were collected by
trap-to-trap vacuum distillation at room temperature. The soluble cata-
lyst and the fluorous phase remained and fresh alkane (1 mL, 7.6 mmol)
and ethyl diazoacetate (10.5 mL, 0.1 mmol) were added for the second
run, this procedure being repeated along the cycles. Yields were deter-
mined as described above.
f
Perfluorophenanthrene as
a
fluorous medium: Fn-Tp4Bo,3R Ag-
ACHTUNGTRENNUNG(CH3COCH3) (0.005 mmol) was dissolved in perfluorophenanthrene
[15] A. Caballero, E. Despagnet-Ayoub, M. M. Dꢃaz-Requejo, A. Dꢃaz-
Rodrꢃguez, M. E. Gonzꢂlez-NfflÇez, R. Mello, B. K. MuÇoz, W.-S.
Ojo, G. Asensio, M. Etienne, P. J. Pꢁrez, Science 2011, 332, 835–838.
[16] a) H. V. R. Dias, R. G. Browning, S. A. Richey, C. J. Lovely, Organo-
metallics 2004, 23, 1200–1202; b) H. V. R. Dias, R. G. Browning,
S. A. Richey, C. J. Lovely, Organometallics 2005, 24, 5784–5784.
[17] P. Rodrꢃguez, E. ꢀlvarez, M. C. Nicasio, P. J. Pꢁrez, Organometallics
2007, 26, 6661–6668.
(4 mL, 3.2 mmol) and the corresponding alkane (1 mL, 7.6 mmol) and
ethyl diazoacetate (10.5 mL, 0.1 mmol) were added at room temperature.
Again, the mixture was homogeneous at room temperature. The proce-
dure from this step was identical to that detailed above with Fomblin.
Scale-up experiments were carried out with EDA (1 mmol) and the silver
catalyst (0.025 mmol) following the above procedure but adding the ethyl
diazoacetate with the aid of a syringe pump at a low rate for 2 h.
[18] A. Endres, G. Maas, J. Organomet. Chem. 2002, 643, 174–180.
[19] a) I. T. Horvꢂth, J. Rꢂbai, Science 1994, 266, 72–75; b) J. A. Gladysz,
D. P. Curran, I. T. Horvꢂth, Handbook of Fluorous Chemistry,
Wiley-VCH, Weinheim, 2004; c) I. T. Horvath, Acc. Chem. Res.
1998, 31, 641–650; d) A. P. Dobbs, M. R. Kimberley, J. Fluorine
Chem. 2002, 118, 3–17.
Acknowledgements
We thank the MEC (Proyecto CTQ2011-28942-CO2-01), the ERA-
Chemistry Programme (2nd call “Chemical activation of carbon dioxide
and methane”, contract number 1736154) and the Institut de Chimie of
the CNRS for financial support. M.A.F. also thanks MEC for a research
fellowship. We thank Dr Emmanuelle Despagnet-Ayoub for helpful dis-
cussions.
[20] E. Despagnet-Ayoub, K. Jacob, L. Vendier, M. Etienne, E. ꢀlvarez,
A. Caballero, M. M. Dꢃaz-Requejo, P. J. Pꢁrez, Organometallics
2008, 27, 4779–4787.
[21] The formulation of Fn-Tp4Bo,3R , in which Fn is the total number of F
f
atoms and Rf the nature of the fluorinated alkyl chain at position 3,
is adopted to keep the merit and spirit of the nomenclature pro-
posed by Trofimenko for the parent, fully hydrogenated tris(indazo-
lyl)borate ligand Tp4Bo and the corresponding Tp4Bo,3R derivatives.
See a) S. Trofimenko, Scorpionates, The coordination Chemistry of
Polypyrazolylborate Ligands; Imperial College Press, London, 1999;
b) C. Pettinari, Scorpionates II. Chelating Borate Ligands, Imperial
College Press, London, 2008.
À
[1] a) Alkane C H Activation by Single-Site Metal Catalysis (Ed.: P. J.
Pꢁrez), Springer, Dordrecht, 2012; b) Activation and functionaliza-
tion of C H bonds (Eds.: K. I. Goldberg, A. S. Goldman), ACS
Symp. Ser. 885, Am. Chem. Soc., Washington DC, 2004; c) Hand-
book of C H transformations, Vol. 1 (Ed.: G. Dyker), Wiley-VCH,
Weinheim, 2005; d) A. E. Shilov, G. B. Shul’pin, Activation and Cata-
lytic Reactions of saturated hydrocarbons in the presence of metal
complexes. Kluwer, Dordrecht, 2000; e) see also special issues of the
following: Chem. Rev. 2010, 110, issue 2 and Acc. Chem. Res. 2012,
45, issue 6.
À
À
[22] D. A. Ellis, K. A. Denkenberger, T. E. Burrow, S. A. Mabury, J.
Phys. Chem. A 2004, 108, 10099–10106.
[23] a) H. V. R. Dias, Z. Wang, W. Jin, Inorg. Chem. 1997, 36, 6205–
6215; b) H. V. R. Dias, W. Jin, J. Am. Chem. Soc. 1995, 117, 11381–
11382.
[2] a) R. G. Bergman, Nature 2007, 446, 391–393; b) J. A. Labinger,
J. E. Bercaw, Nature 2002, 417, 507–513; c) B. A. Arndtsen, R. G.
Bergman, T. A. Mobley, T. H. Peterson, Acc. Chem. Res. 1995, 28,
154–162.
[24] P. K. Hurlburt, J. J. Rack, J. S. Luck, S. F. Dec, J. D. Webb, O. P. An-
derson, S. H. Strauss, J. Am. Chem. Soc. 1994, 116, 10003–10014.
[25] A. A. C. Braga, A. Caballero, J. Urbano, M. M. Dꢃaz-Requejo, P. J.
Pꢁrez, F. Maseras, ChemCatChem 2011, 3, 1646–1652.
[3] A. E. Shilov, G. B. Shulꢈpin, Chem. Rev. 1997, 97, 2879–2932.
[4] J. T. Golden, R. A. Andersen, R. G. Bergman, J. Am. Chem. Soc.
2001, 123, 5837–5838.
[5] a) C. M. Jensen, Chem. Commun. 1999, 2443–2449; b) A. S. Gold-
man, A. H. Roy, Z. Huang, R. Ahuja, W. Schinski, M. Brookhart,
Science 2006, 312, 257–261.
[6] I. A. I. Mkhalid, J. H. Barnard, T. B. Marder, J. M. Murphy, J. F.
Hartwig, Chem. Rev. 2010, 110, 890–931.
[7] J. F. Hartwig, Organotransition Metal Chemistry: From Bonding to
Catalysis, University Science books, 2009.
[26] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, J. Appl.
Crystallogr. 1993, 26, 343–350.
[27] SHELX97 [Includes SHELXS97, SHELXL97, CIFTAB]—Programs
for Crystal Structure Analysis (Release 97–2). G. M. Sheldrick, Insti-
tüt für Anorganische Chemie der Universität, Tammanstrasse 4,
3400 Gçttingen, Germany, 1998.
[28] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[29] L. J. Farrugia, J. Appl. Crystallogr. 1997, 30, 565–565.
[8] a) M. P. Doyle, M. A. McKervey, T. Ye, Modern Catalytic Methods
for Organic Synthesis with Diazo Compounds, Wiley, New York,
1998; b) M. P. Doyle, R. Duffy, M. Ratnikov, L. Zhou, Chem. Rev.
2010, 110, 704–724.
Received: October 6, 2012
Revised: November 6, 2012
Published online: November 29, 2012
1334
www.chemeurj.org
ꢅ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1327 – 1334