Cu-MOFs as active, selective and reusable catalysts for oxidative C-O bond coupling reactions by direct C-H activation of formamides, aldehydes and ethers

Article abstract of DOI:10.1039/c4cy00032c

MOFs with Cu²⁺ centers linked to four nitrogen atoms from azaheterocyclic compounds, i.e., pyrimidine [Cu(2-pymo)₂] and imidazole [Cu(im)₂], are active, stable and reusable catalysts for oxidative C-O coupling reactions by direct C-H activation of formamides, aldehydes and ethers. The measured catalytic activities are clearly superior to other homogeneous cupric salts, especially for the [Cu(im)₂] MOF. The previously reported activity of the Cu²⁺ centers for cumene oxidation allows the use of the MOF as a bifunctional catalyst for olefin epoxidation with O₂. The overall catalytic process consists of a cascade reaction in which the Cu-MOF first produces cumyl hydroperoxide and then the same Cu²⁺ centers catalyze the oxidative C-O coupling reaction using the generated hydroperoxide as the oxidant. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4cy00032c

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Cu-MOFs as active, selective and reusable
catalysts for oxidative C–O bond coupling
reactions by direct C–H activation of
formamides, aldehydes and ethers†
Cite this: DOI: 10.1039/c4cy00032c
*
*
I. Luz, A. Corma and F. X. Llabrés i Xamena
MOFs with Cu2+ centers linked to four nitrogen atoms from azaheterocyclic compounds, i.e., pyrimidine
[Cu(2-pymo)2] and imidazole [Cu(im)2], are active, stable and reusable catalysts for oxidative C–O
coupling reactions by direct C–H activation of formamides, aldehydes and ethers. The measured catalytic
activities are clearly superior to other homogeneous cupric salts, especially for the [Cu(im)2] MOF. The
previously reported activity of the Cu2+ centers for cumene oxidation allows the use of the MOF as a
bifunctional catalyst for olefin epoxidation with O2. The overall catalytic process consists of a cascade
reaction in which the Cu-MOF first produces cumyl hydroperoxide and then the same Cu2+ centers cat-
alyze the oxidative C–O coupling reaction using the generated hydroperoxide as the oxidant.
Received 10th January 2014,
Accepted 17th March 2014
DOI: 10.1039/c4cy00032c
www.rsc.org/catalysis
chloroformates. The authors concluded that the presence of
a carbonyl group adjacent to the hydroxy group is essential
1. Introduction
Transition metal catalyzed C–C and C–heteroatom coupling
reactions are essential tools for the preparation of fine
chemicals and key intermediates.1,2 These reactions usually
rely on the use of pre-functionalized compounds (e.g., haloge-
nated compounds and Grignard or organolithium reagents),
and the reaction is usually carried out in the presence of an
organic or inorganic base. An interesting possibility that has
attracted great attention is the coupling reaction by direct
activation of C–H bonds. This avoids the use of pre-
functionalized precursors, thus saving additional preparative
steps and increasing the overall atom efficiency of the pro-
cess while reducing the production of waste.
for the success of the oxidative C–O coupling. This was attrib-
uted4 to the tendency of dicarbonyl compounds to form a
bidentate complex with the copper ions that facilitates the
homolytic cleavage of the TBHP O–O bond. The reaction
was proposed to proceed through the formation of a form-
amide radical species, as shown in Scheme 2. Kumar
et al. also demonstrated that both β-keto esters (a′) and
2-carbonyl-substituted phenols (b′) can be coupled with either
aldehydes (b)3 or ethers (c),5 leading respectively to phenol
esters (b–b′) or ethers (c–b′) and unsymmetrical acetals (c–a′),
also implying in both cases a direct activation of either the
C–H bond of the aldehyde (b) or the α-C–H bond of the ether (c)
(see Scheme 1).
While the development of new efficient metal-catalyzed
oxidative coupling reactions based on direct C–H activation is
a significant step towards more eco-friendly processes, it is
evident that the use of heterogeneous catalysts introduces
additional advantages concerning product isolation and cata-
lyst recovery and reuse. In this sense, it has been demon-
strated that metal organic frameworks can replace
homogeneous molecular catalysts for a number of reac-
tions.6–10 In particular, we have shown that copper-MOFs
containing CuN4 centers with either imidazolate or
2-oxypyrimidinolate ligands ([Cu(im)2] and [Cu(2-pymo)2]) are
highly active catalysts for the liquid phase aerobic oxidation
of activated alkanes (viz., tetralin, ethylbenzene and
cumene),11,12 which implies the effective generation of hydro-
peroxides and conversion to the corresponding alcohols and
In this sense, Kumar et al.3 and Barve et al.,4 have recently
shown that various copper salts can catalyze the oxidative
C–O coupling reaction of dialkyl formamides (a) with either
β-keto esters (a′) or 2-carbonyl-substituted phenols (b′) using
tert-butyl hydroperoxide (TBHP) as the oxidant (see
Scheme 1). This synthetic route leading to enols (a–a′) and
phenol carbamates (a–b′) implies direct activation of the C–H
bond of the formamide, and it was proposed as a phosgene-
free alternative to the use of amines and isocyanates or
Instituto de Tecnología Química, Universidad Politécnica de Valencia, Consejo
Superior de Investigaciones Científicas, Avda. de los Naranjos, s/n, 46022
Valencia, Spain. E-mail: fllabres@itq.upv.es, acorma@itq.upv.es;
Fax: +34 963877809
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4cy00032c
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Paper
Catalysis Science & Technology
Scheme 1
Scheme 2
ketones. Herein we show that these Cu-MOFs can also cata-
2. Materials and methods
2.1. Preparation of the Cu-MOFs
lyze the oxidative coupling reactions of the compounds
shown in Scheme 1, which also rely on the ability of the
materials to effectively react with hydroperoxides, as shown
in Scheme 2. It is worth mentioning that during the prepara-
tion of this manuscript, Phan et al.13 have successfully
applied a Cu-MOF as a catalyst for the oxidative coupling of
ethers and ortho-substituted phenolic compounds, following
an analogous procedure as that depicted in Scheme 1 for the
preparation of c–b′ type compounds.
The Cu-MOFs were prepared according to the corresponding
procedures reported in the original references for [Cu(2-
pymo)2],14 [Cu(im)2],15 and [Cu(BDC)].16 [Cu3(BTC)2] is a
Basolite C300 commercial sample purchased from Sigma-
Aldrich. Copper-exchanged zeolites used for comparison were
prepared by conventional cationic exchange of commercial
samples purchased from Zeolyst International, as described
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Catalysis Science & Technology
Paper
in detail in the ESI.† X-ray diffraction (Philips X'Pert, Cu Kα
radiation) was used to confirm the expected crystalline struc-
ture of all the materials. Characterization of the samples by
elemental analysis, TGA, FTIR and N2 adsorption is provided
in the ESI.†
in the reaction medium to assist the oxidative coupling
between the phenol derivative and DMF to form the corre-
sponding carbamate 3. As already reported by Kumar et al.,3
various CuI and CuII salts can catalyze this reaction. Thus as
expected, when the reaction was carried out under the same
reaction conditions and in the presence of copper acetate
(5 mol% Cu), carbamate 3 was quantitatively formed after 3 h
(Table 1, entry 2), in good agreement with the results previ-
ously reported.3,4 Very similar results were obtained when the
Cu-MOF [Cu(2-pymo)2] was used as a catalyst and the copper-
to-substrate molar ratio was kept constant at 5 mol% (Table 1,
entry 6). This demonstrates that this Cu-MOF can success-
fully replace soluble copper salts as the catalyst for this reac-
tion, with the additional advantage that the solid MOF can
be easily separated from the reaction products by simple fil-
tration or centrifugation. Moreover, the solid recovered at the
end of the reaction was found to remain intact, as deter-
mined by comparing the XRD pattern with that of the fresh
catalyst (see Fig. S17 in the ESI†), which allowed reuse of the
material without significant loss of activity for at least 3 cata-
lytic cycles (Table 1, entry 7). The heterogeneity of the cata-
lytic process is supported by the following evidence: i)
maintenance of the crystalline structure of the solid recov-
ered after the catalytic reaction; ii) according to elemental
analysis of the liquid filtrate after the reaction, only traces of
Cu2+ were present (ca. 0.5 mg L−1), showing that no signifi-
cant amount of copper (less than 0.03% of the total amount
of Cu) leached from the solid to the reaction medium during
the reaction; and iii) a hot filtration test was also performed
(see Fig. S19 in the ESI†), showing that once the solid was
removed from the reaction medium, no further conversion
was observed in the filtrate. This experiment rules out the
presence of a significant amount of homogeneous (soluble)
catalytic species coming from the solid MOF.
2.2. Catalytic reactions
Time evolution of products was determined in all cases by
GC analysis of the samples (Varian 3900, capillary column
HP-5) using n-hexadecane as the external standard. Particular
reaction conditions in each case are indicated in the footnote
of the corresponding tables.
3. Results and discussion
3.1. Synthesis of carbamates by direct C–H activation of
formamides
When 2-hydroxyacetophenone (1) was contacted with TBHP
(1.5 eq.) in a large excess of N,N′-dimethylformamide (DMF,
2 mL, ca. 26 eq.), no reaction took place even after 24 h at
80 °C (Table 1, entry 1). The presence of a catalyst is necessary
Table
1 Oxidative coupling of 2-hydroxyacetophenone and N,N′-
dimethylformamidea
Yield of
3b (mol%)
Better results were obtained when the [Cu(im)2] MOF was
used as the catalyst, since in this case the time needed for
full phenol conversion was halved to only 1.5 h (Table 1,
entry 8). Also the crystalline structure of [Cu(im)2] was
maintained during the reaction (see Fig. S18 in the ESI†) and
the material was reusable for at least 3 catalytic cycles
(Table 1, entry 9). We also evaluated the catalytic activity of
other Cu-MOFs containing Cu2 paddle-wheel dimers: viz.
[Cu3(BTC)2] and [Cu(BDC)]. Although these materials were
also found to be active and stable catalysts for this reaction,
they were considerably worse performing than the two other
Cu-MOFs containing CuN4 centers. For instance, a maximum
yield of 71% was attained over [Cu(BDC)] after 24 h of reac-
tion (Table 1, entry 10). It is also worth mentioning that this
reaction seems to be exclusively catalyzed by copper ions,
while other metals give no reaction under similar conditions.
In this sense, Barve et al. reported a preliminary screening of
various substances that are known to catalyze radical reac-
tions, including Fe2O3 or ZnCl2, but none of them was able
to catalyze this C–O coupling reaction.4 Analogously, we
observed that the reaction between 2-hydroxyacetophenone
and DMF did not proceed in the presence of other MOFs
Entry Catalyst
ROOH Time
1
2
3
4
5
6
7
No catalyst
Cu(OAc)2
TBHP 24 h
TBHP 3 h
TBHP 3 h
<2
95
93
98
9/24
93
Cu2+–ZSM-5
Cu2+–USY
TBHP 30 min
Cu2+–USY (3rd cycle) TBHP 30 min/1 h
[Cu(2-pymo)2]
[Cu(2-pymo)2]
(3rd cycle)
TBHP 3 h
TBHP 3 h
92
8
9
[Cu(im)2]
TBHP 30 min/1.5 h
67/97
60/96
71
97
95
[Cu(im)2] (3rd cycle) TBHP 30 min/1.5 h
[Cu(BDC)]
Cu(OAc)2
[Cu(2-pymo)2]
[Cu(im)2]
10
11
12
13
14
TBHP 24 h
CmHP 30 min
CmHP 30 min
CmHP 30 min
CmHP 15 min/30 min/3 h 2/4/75
CmHP 15 min/30 min/3 h 1/3/77
CmHP 15 min/30 min/3 h 7/23/74
99
Cu(OAc)2
(1 mol% Cu)
[Cu(2-pymo)2]
(1 mol% Cu)
[Cu(im)2]
15
16
(1 mol% Cu)
a
Reaction conditions:
1
mmol 2-hydroxyacetophenone (103 μl),
1.5 mmol ROOH, 2 mL of DMF and Cu-catalyst (5 mol% Cu, unless
otherwise indicated), 80 °C. Determined by GC using hexadecane
as the external standard. TBHP
CmHP = cumyl hydroperoxide.
b
= tert-butyl hydroperoxide and
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Paper
Catalysis Science & Technology
containing various metal ions aside from copper, including
Cr-MIL-101 and CPO-27 materials of Mn, Co and Ni.
replaced by cumyl hydroperoxide (CmHP), we observed that
the oxidative C–O coupling reactions over copper acetate,
[Cu(im)2] and [Cu(2-pymo)2] were significantly faster than
that over TBHP, attaining full conversion of 1 to carbamate 3
in only 30 min in all cases (Table 1, entries 11–13). In order
to appreciate the differences between copper acetate and the
two Cu-MOFs, the reaction was repeated using only 1 mol%
of copper with respect to phenol. Under these conditions,
[Cu(im)2] clearly outperformed the other Cu-MOF and copper
acetate at short reaction times (23% yield after 30 min for
[Cu(im)2] vs. 3–4% for copper acetate and [Cu(2-pymo)2],
Table 1, entries 14–16). The corresponding turnover frequen-
cies (TOFs) calculated for the catalysts were 1 h−1, ~5 h−1 and
~28 h−1 for [Cu(2-pymo)2], Cu(OAc)2 and [Cu(im)2], respec-
tively, thus making the latter MOF clearly the most active cat-
alyst for this reaction. Nevertheless, the three materials led to
very similar ca. 75–80% product yields after 3 h. It is worth
mentioning that the reaction stops before full phenol conver-
sion due to the complete consumption of CmHP. Indeed,
when a larger excess of CmHP was used (3 eq.), carbamate 3
was quantitatively formed after 3 h. The reason for this spuri-
ous consumption of the hydroperoxide is the occurrence of a
decomposition side reaction competing with the oxidative
C–O coupling, which can be either a thermally activated pro-
cess or it can also be catalyzed by the same copper centers,
as shown in Scheme 3. Thus the amount of carbamate 3
formed will be determined by the relative reaction rates of
both competing reactions, and this, in turn, will depend on
the relative stability of the hydroperoxide used (either TBHP
or CmHP), the substrates of the C–O coupling reaction, the
reaction temperature and, of course, the local structure of
the copper active centers of the catalyst. This holds true for
the rest of the reactions discussed in this work.
Besides the higher reaction rate attained as compared to
TBHP, the use of CmHP as the oxidant for the C–O coupling
reaction is interesting since we have previously shown that
CmHP can be generated in situ by liquid phase aerobic oxida-
tion of cumene over Cu-MOF catalysts.12 Thus, taking the
idea of the Sumitomo process for the production of propene
oxide,20,21 we can envisage a one-pot two-step process in
which cumene was oxidized to CmHP using O2 as the oxi-
dant. Then in a second step, CmHP was used to perform the
oxidative C–O coupling of DMF and 2-hydroxyacetophenone.
Eventually, the cumyl alcohol generated in the second step
could be reconverted into cumene through dehydration and
hydrogenation. The whole process is depicted in Scheme 4.
Note that the Cu-MOF catalyzes both cumene oxidation in air
and the oxidative coupling reaction using as the oxidant the
CmHP generated in the first step. Thus, in this sense, the
Cu-MOF acts as a bifunctional catalyst even though the active
centers are the same Cu2+ ions for both reactions.
To the best of our knowledge, no reports exist on other
Cu-containing heterogeneous catalysts for this coupling reac-
tion. Therefore, in order to put the activity of the Cu-MOFs
into perspective, we extended our study to Cu-exchanged zeo-
lites of medium (MFI-type) and large pore (FAU-type) zeolites.
The results are reported in Table 1 (entries 3 and 4). As can
be seen, both zeolites are active catalysts, as expected due to
the typically high degree of coordinative unsaturation of the
Cu2+ counterions in zeolites and their ability to coordinate
with electron donating adsorbed molecules.17–19 The larger
pore opening of Cu2+–USY (7.4 Å) with respect to Cu2+–ZSM-5
(5.5 Å) allows easier diffusion of substrates and products and
is probably responsible for the higher catalytic activity
observed. With respect to the Cu-MOFs, we found that, while
Cu2+–ZSM-5 has a catalytic activity similar to [Cu(2-pymo)2],
the performance of Cu2+–USY is superior to that of [Cu(im)2].
However, Cu2+–USY loses most of its catalytic activity after
one catalytic cycle due to leaching of Cu2+ counterions to the
liquid medium. Fig. 1 shows a comparison of the time–
conversion plots obtained for fresh and reused Cu2+–USY and
[Cu(im)2]. Therefore, in spite of the high initial activity of
Cu2+–USY, the lack of long term stability and limited reus-
ability make the Cu-zeolite a poor catalyst for this oxidative
coupling reaction in comparison with the Cu-MOFs.
As reported by Kumar et al., the selection of TBHP as the
oxidant for this oxidative C–O coupling reaction was found to
be essential; no product formation was observed by these
authors when H2O2, urea-hydrogen peroxide (UHP), di-tert-butyl
peroxide (DTBP), meta-chloroperoxybenzoic acid (mCPBA)
or NaOCl was used as the oxidant.3,5 A similar screening
of various organic and inorganic oxidants led Barve et al.
to the same conclusion.4 Nevertheless, we wanted to
extend the range of applicable oxidants for this reaction to
other organic hydroperoxides. Thus, when TBHP was
Following the process shown in Scheme 4, we contacted
1 mL of cumene (ca. 7 mmol) with [Cu(2-pymo)2] (5 mol% Cu)
at 80 °C under an O2 atmosphere for 2 h to produce CmHP in
ca. 20% yield, whereupon 1 mmol of 2-hydroxyacetophenone
and 1 mL of DMF were added at the same reaction temperature.
Fig. 1 Kinetic data of the oxidative C–O coupling reaction between
2-hydroxyacetophenone and DMF over fresh and reused [Cu(im)2] and
Cu2+–USY.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014