Nanocrystalline M-MOF-74 as heterogeneous catalysts in the oxidation of cyclohexene: Correlation of the activity and redox potential

Article abstract of DOI:10.1002/cctc.201402927

In some aspects, the potential of metal-organic framework (MOF) materials as heterogeneous catalysts has been realized, at least in an academic context. However, one of their most promising catalytic properties, that is, the presence of open metal sites, is far from understood properly. In this work, a series of M-MOF-74 (M=Mn, Co, Ni, Cu, Zn) materials, prepared under sustainable conditions, was tested systematically in the oxidation of cyclohexene, which can proceed by either radical or epoxidation routes. Under the optimized reaction conditions, the radical route is spontaneous to some extent and it is enhanced in the presence of any M-MOF-74 that has a metal with a redox character but not Zn. However, the epoxidation of cyclohexene is also promoted by a redox catalyst in such a way that the conversion correlates qualitatively with the redox potential of the metal. Thus, for the first time, a chemical property of M is correlated with the catalytic activity of the M-MOF-74 family. Redox rocks: A series of nanocrystalline M-MOF-74 (M=Mn, Co, Ni, Cu, Zn) materials prepared at room temperature catalyze the oxidation of cyclohexene with tert-butylhydroperoxide. Radical oxidation is similar for any redox-active M, whereas epoxidation is preferred for M with slight oxidant properties. M-MOF-74 catalysts with the redox pairs M³⁺/M²⁺ were stable and those with M²⁺/M⁺ were not.

Full text of DOI:10.1002/cctc.201402927

DOI: 10.1002/cctc.201402927
Full Papers
Nanocrystalline M–MOF-74 as Heterogeneous Catalysts in
the Oxidation of Cyclohexene: Correlation of the Activity
and Redox Potential
Daniel Ruano, Manuel Dꢀaz-Garcꢀa, Almudena Alfayate, and Manuel Sꢁnchez-Sꢁnchez*[a]
In some aspects, the potential of metal–organic framework
(MOF) materials as heterogeneous catalysts has been realized,
at least in an academic context. However, one of their most
promising catalytic properties, that is, the presence of open
metal sites, is far from understood properly. In this work,
a series of M–MOF-74 (M=Mn, Co, Ni, Cu, Zn) materials, pre-
pared under sustainable conditions, was tested systematically
in the oxidation of cyclohexene, which can proceed by either
radical or epoxidation routes. Under the optimized reaction
conditions, the radical route is spontaneous to some extent
and it is enhanced in the presence of any M–MOF-74 that has
a metal with a redox character but not Zn. However, the epoxi-
dation of cyclohexene is also promoted by a redox catalyst in
such a way that the conversion correlates qualitatively with
the redox potential of the metal. Thus, for the first time,
a chemical property of M is correlated with the catalytic activi-
ty of the M–MOF-74 family.
Introduction
Metal–organic framework (MOF) materials with permanent po-
rosity[1] notably extend the properties, versatility, and, impor-
tantly, potential applications of such microporous materials.[2]
Together with the extreme flexibility[3] and catenation[4] of
some MOF networks, the presence of exposed and unsaturat-
ed metal sites in the framework[5] was probably the most re-
markable aspect of MOFs compared with other microporous
materials. These open metal sites are particularly attractive in
two direct applications. On one hand, the highly demanded
adsorption and separation of gases with high energetic and/or
environmental interest, such as H2,[5b,6] CO2,[7] or light hydrocar-
bons,[7b] found in these unprecedented centers is an incentive
with respect to common adsorbents (zeolites, carbons, etc.),[8]
which normally act as inert storage systems. The highest heats
of adsorption for these gases and some of their most efficient
separations were achieved with MOFs that contain open metal
sites.[9] On the other hand, in a heterogeneous catalysis con-
text, MOF materials not only cover the lack of zeolites and zeo-
types for the incorporation of certain metal ions but also intro-
duce exposed metal centers, which allows their direct interac-
tion with reactants/adsorbates. Although tens of academic
studies have focused on the application of these MOF materi-
als in different catalytic reactions,[10] their generally unsystemat-
ic character led rarely to consolidated knowledge. It is particu-
larly strange if we consider the existence of families of isostruc-
tural MOFs that contain open metal sites that can be prepared
with different metal ions. For instance, M–MOF-74 materials,[11]
also known as CPO-27M,[12] M2(dobdc),[13] or M2(dhtp)[14]
(dobdc=2,5-dioxido-1,4-benzenedicarboxylate, dhtp= 2,5-di-
hydroxyterephthalate), can be prepared with Mg, Mn, Fe, Co,
Ni, Cu, Zn, and Cd divalent ions[11–15] or their mixtures.[16] The
reported catalytic studies of this family of MOFs have normally
investigated a particular M–MOF-74 material[17] rather than the
effect of the nature of the different metal ions.
This work focuses on the catalytic activity of the nanocrystal-
line M–MOF-74 materials prepared at room temperature[18] in
the oxidation of cyclohexene. As this catalytic reaction requires
redox centers, Mn, Co, Ni, and Cu were selected as the redox-
active M of the M–MOF-74 materials over, for instance, Mg and
Cd ions. Zn-MOF-74 was selected as a non-redox-active materi-
al because of the chemical similarity of Zn with the other
chosen metal ions. The spontaneous oxidation of cyclohexene
with tert-butylhydroperoxide was accelerated significantly by
the presence of any redox-active M–MOF-74 material, whereas
it was inhibited by Zn-MOF-74. However, the epoxidation of
cyclohexene occurs necessarily in the presence of redox-active
MOF-based catalysts, the activity of which was correlated to
the redox potential of M. Moreover, these nanocrystalline sam-
ples were more active than their micrometer-sized homo-
logues. The stability of the catalysts and metal leaching to the
reaction media were studied and explained in terms of the
possible reaction mechanism.
Results and Discussion
[a] D. Ruano, M. Dꢀaz-Garcꢀa, A. Alfayate, Dr. M. Sꢁnchez-Sꢁnchez
Instituto de Catꢁlisis y Petroleoquꢀmica, ICP-CSIC
C/Marie Curie 2, 28049 Madrid (Spain)
MOF-74 materials
The synthesis and characterization of four of the five tested
catalysts in this paper are described elsewhere.[18] The main
characteristic of these samples is their nanocrystalline nature,
E-mail: manuel.sanchez@icp.csic.es
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402927.
ChemCatChem 0000, 00, 0 – 0
1
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Table 1. Crystal size and textural properties of the M–MOF-74 catalysts.
M
Crystal size[a]
[nm]
BET area
External BET area[b]
[m2 gÀ1
]
PSD[c]
[nm]
[m2 gÀ1
]
Zn
Cu
Mn
Co
Ni
16.7
14.0
13.6
5.0
948
1103
791
693
514
235
286
273
416
65
29.5
39.8
24.4
2.3
2.9
–
[a] Estimated by the Scherrer equation.[18] [b] Estimated by the t-plot
method. [c] Maximum pore size distribution of the N2 adsorption branch.
which covers a crystal size range from 16.7 (Zn-MOF-74) to
2.9 nm (Ni-MOF-74) (Table 1). As a result of the small crystal
size, they have strong tendency to be agglomerated/aggregat-
ed. The crystalline units above approximately 10 nm are simply
agglomerated and they are isolable by a mild sonication treat-
ment. Conversely, the crystalline units below 10 nm are do-
mains rather than crystals, and they are aggregated in hardly
isolable particles. In both cases, the agglomeration/aggrega-
tion is relatively ordered, and the particles contain certain in-
tercrystalline mesoporosity.[18] Therefore, these materials are ex-
pected to reduce diffusion problems compared to their homol-
ogous micrometer-sized MOF-74 materials prepared under sol-
vothermal conditions.
Figure 1. Powder XRD patterns of the nanocrystalline M–MOF-74 materials
tested in the oxidation of cyclohexene. The intensity of the patterns of Ni-
and Co-MOF-74 have been multiplied by a factor of 5. At the top, a powder
XRD pattern of a conventional micrometer-sized Zn-MOF-74 is shown for
comparison. The powder XRD patterns of the samples based on Cu, Mn, Co,
and Zn have a significant background caused by a fluorescence phenomen-
on.[18]
and the maximum PSD decreases as the average crystal size
decreases. Ni-MOF-74 is not considered in this trend because
its intercrystalline pore diameter, if any, should be in the micro-
porous range.[18]
The most promising first-period transition metal ion as a cat-
alytic center in oxidation reactions is probably Cu,[19] particular-
ly in Cu-MOF materials.[20] This is why a particular effort has
been made to prepare a nanocrystalline Cu-based MOF-74 ma-
terial at room temperature.[18,21] The preparation of a Cu-based
MOF-74-like material has been described for the first time re-
cently under solvothermal conditions.[17a] Therefore, the room-
temperature synthesis and characterization of Cu-MOF-74 de-
serve some attention. A detailed description of the synthesis
and characterization of Cu-MOF-74 is given in the Supporting
Information, and the powder XRD pattern is presented in
Figure 1. The crystal size of Cu-MOF-74 is quite similar to that
of Zn- and Mn-MOF-74 (Table 1) and significantly higher than
that of Ni- and Co-MOF-74. If we apply the Scherrer equation
with the same criteria described elsewhere,[18] the crystalline
units have an average size of 14.0 nm, which is only exceeded
by that of the Zn sample (16.7 nm). The other characterization
results (Supporting Information) indicate that the Cu-dhtp
sample is nanocrystalline and has a MOF-74-like structure.
In addition to the crystal size estimation by the Scherrer
equation, some textural properties of the series of M–MOF-74
are given in Table 1, which are extracted from the isotherms
shown in the Supporting Information. The textural properties
of our samples were slightly better than those published,[18]
which is probably because the N2 adsorption–desorption iso-
therms were measured immediately after preparation and
washing. Cu-MOF-74 has the best textural properties of the
series, which confirms the successful preparation of this
sample. Nevertheless, the pore size distribution (PSD) is slightly
higher than expected (Table 1).[18] As described previously,[18]
the external/microporous surface area ratio generally increases
Optimizing the reaction conditions: Choice of oxidant
The choice of cyclohexene oxidation as the catalytic test in this
work was based on the high versatility of this reaction. The
versatility covers a large number of solvents,[22] all M of the
series of M–MOF-74 (except Zn, taken as reference) already
tested in the reaction,[17b,23] two different oxidation routes[24]
(Scheme 1), and different oxidizing agents.[25] Acetonitrile was
selected as the solvent because it is one of the most widely
used in cyclohexene oxidation and it is able to dissolve all re-
actants in a unique liquid phase even if 30 wt% H2O2 in aque-
ous solution was used as the oxidant.[24b]
Scheme 1. Products generated from the oxidation of cyclohexene through
either the radical (top) or epoxidation (bottom) routes.
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
2
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Previous studies of MOF-based catalysts in this reac-
tion[17b,23b,26] do not clarify anything about the influence of the
nature of the oxidant. In our opinion, the oxidant deserves to
be studied, not only because of its oxidizing potential but also
because the solvent in which it is stabilized (water in the case
of H2O2 and n-decane for tert-butylhydroperoxide; TBHP) could
play a key role in both the global catalytic activity and the
MOF stability.
route under the tested conditions. Finally, this radical route re-
quires the homolytic rupture of the OÀO bond of the peroxide,
which has been calculated to be more unfavorable for H2O2
than for TBHP in the presence of a Ni-based catalyst.[29]
As a result of the questioned stability of MOF-based cata-
lysts in the reaction media, the comparison between TBHP and
H2O2 should not be limited to cyclohexene conversion and
structural stability should be also considered. The powder XRD
patterns of Ni-MOF-74 before and after the reaction with both
oxidizing agents are shown in Figure 3. After the oxidation of
The yield of the products from the oxidation of cyclohexene
with either H2O2 or TBHP as the oxidant in the presence of the
nanocrystalline Ni-MOF-74 catalyst are plotted as a function of
reaction time in Figure 2. Surprisingly, no conversion of cyclo-
Figure 3. Powder XRD patterns of the Ni-MOF-74 samples before (top) and
after reaction using TBHP (middle) and H2O2 (bottom) as oxidizing agents.
The intensity of the two top patterns was multiplied by 2 for an easier com-
parison. The background in all patterns is caused by a fluorescence phenom-
enon.[18]
Figure 2. Sum of the yields of the products that result from the oxidation of
cyclohexene in the presence of Ni-MOF-74 with H2O2 30 wt% or TBHP as oxi-
dizing agents.
hexene was detected during 26 h reaction if the oxidant was
30 wt% H2O2. In contrast, the use of TBHP led to an acceptable
conversion of 40% of cyclohexene after a similar reaction time.
Moreover, almost the whole oxidation took place through the
allylic mechanism (95.5%). (Yields of the different products at
the end of the reaction are given in the Supporting Informa-
tion). The preference for the radical route was further con-
firmed by the dramatic decrease of the activity in the presence
of a radical inhibitor (Supporting Information).
cyclohexene with TBHP that gave an acceptable cyclohexene
conversion, the catalyst basically conserved its MOF-74 struc-
ture. The slight difference in the patterns of the catalyst before
and after the reaction can be attributed to the chemicals that
occupy the MOF-74 pores because the XRD pattern of the
former was measured as-reacted without any extra washing or
treatment. Conversely, the structure of the same sample after
the reaction in the presence of H2O2 underwent a complete
transformation although no cyclohexene was oxidized. Al-
though it is not completely clear what causes the transforma-
tion of the Ni-MOF-74 catalyst, it must be related to the pres-
ence of H2O2 and/or H2O under the reaction conditions.
As a result of both the cyclohexene conversion and the
structural stability of Ni-MOF-74, TBHP (5.5m in n-decane) was
selected as the oxidizing agent for further catalytic investiga-
tions in this work.
The ultimate reason for this spectacular difference, which
depends on the nature of the peroxide-based oxidant, is not
completely clear, as some features of the system could disfavor
the catalytic conversion in the presence of H2O2. Firstly, the un-
saturated metal sites must have an asymmetric charge distri-
bution,[27] so it is expected that water rather than acetonitrile,
H2O2, or cyclohexene is coordinated to the catalytically active
sites M, which causes the deactivation of these centers. In ad-
dition, H2O is present in excess of the other very polar chemi-
cal species H2O2 in the 30 wt% oxidant source (H2O/H2O2
molar ratio of ꢀ4.4). Conversely, TBHP, although less polar
than H2O2, has to compete with highly hydrophobic n-decane
instead of water. In this sense, the activity of Ti-MCM-41 cata-
lysts in the epoxidation of alkanes is weakened if anhydrous
TBHP becomes hydrated.[28] As a second possible reason, water
acts as an inhibitor in the oxidation of cyclohexene through
the radical route,[24a,b] which dominates over the epoxidation
Influence of the nature of M on the oxidation of cyclohex-
ene catalyzed by M–MOF-74
As mentioned in the Introduction, the M–MOF-74 family offers
a unique opportunity for systematic comparisons of the cata-
lytic activity of different exposed and unsaturated M centers.
In addition, the catalysts tested here were prepared at room
temperature, which provides two extra important advantages:
(i) one in a sustainable aspect, which contrasts with the MOF
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
3
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
materials prepared under conventional solvothermal condi-
tions, and (ii) the other one in a catalytic aspect as this prepara-
tion method implies the formation of nanocrystalline M–MOF-
74[18] (Table 1), which should minimize diffusion problems. The
improvement in reactants/products diffusion associated with
the “nano” nature of the M–MOF-74 catalysts prepared at
room temperature is evidenced clearly for Cu- and Co-MOF-74
(Supporting Information).
more, the presence of M–MOF-74 with a non-redox-active M,
far from enhancing the conversion, seems to act as an inhibi-
tor, which gives extra relevance to the intrinsic catalytic activity
of redox-active M.
The order of the catalytic activity at any time of the reaction
is: Cu>Co>Mn>Ni>blank>Zn. Notably, a very high cyclo-
hexene conversion is achieved by Cu-MOF-74 (90%, Table 2),
which is practically reached after the first 2 h of reaction
(Figure 4). Similarly, Co- and Mn-based samples also reach
a kind of plateau after relatively short reaction times (around
5 h). Conversely, the blank reaction and the reactions catalyzed
by the Ni- and Zn-MOF materials gave very low conversions in
short reaction times, and their conversions did not reach a pla-
teau after 24 h of reaction. The lower activity of Ni-MOF-74
compared with the other M–MOF-74 materials with a redox-
active M is in good agreement with the only published work in
which two M–MOF-74 materials (M=Co and Ni) were tested in
this reaction.[23b]
Plots of the kinetics of the total yield of the products oxi-
dized from cyclohexene using the five M–MOF-74 catalysts and
in the absence of any catalyst (blank experiment) are shown in
Figure 4. Detailed data for the catalysts at the end of the reac-
A detailed comparison of the cyclohexene conversion with
the physical properties of the MOF-74-based redox catalysts
suggests that the conversion correlates linearly with the sur-
face area of the catalyst (Figure 5) irrespective of the nature of
Figure 4. Sum of the yields of the products that result from the oxidation of
cyclohexene in the presence of M–MOF-74 and without catalyst (blank) with
TBHP as a function of the reaction time.
tion (after ꢀ24 h) are shown in Table 2. All M–MOF-74 that
have a redox-active M gave cyclohexene conversions that were
significantly higher than that of the blank reaction. However,
the conversion in the presence of Zn-MOF-74 was lower than
that of the catalyst-free reaction. This result corroborates that
the redox character of M is essential to make the M–MOF-74
material a viable catalyst for cyclohexene oxidation. Further-
Figure 5. Total yield of products generated by the oxidation of cyclohexene
after ꢀ24 h vs. the BET area of the M–MOF-74 catalysts with a redox-active
M. The black line corresponds to the least-squares adjustment of the four
plotted points.
M. In other words, the redox-active ions are equally active, if
accessible. The only slight alteration of the otherwise perfect
linear correlation is found for the Mn and Co samples, which
are quite different to each other (very different crystal size, ex-
ternal/microporous surface area ratio, presence of impurities in
the Mn-based sample,[18] etc.). The other possible analysis of
the deviation from linearity shown in Figure 5 implies that only
the Co-based catalyst is slightly away, probably because its
huge external surface area (ꢀ60% of the whole surface area;
Table 1) reduces the diffusional limitations, which suggests that
even the metal centers on the external surface could be cata-
lytically active.[30]
Table 2. Sum of the yields of the different products generated by the ox-
idation of cyclohexene with TBHP at 708C after ꢀ24 h of reaction cata-
lyzed by M–MOF-74. Detailed yield data for the identified products are
given in the Supporting Information.
Catalyst
Total yield Radical yield Epoxidation yield
TOF
]
[a]
[%]
[%]
[%]
[hÀ1
2 h
24 h
Cu-MOF-74 90.0
Co-MOF-74 71.5
Mn-MOF-74 68.2
35.7
43.1
29.0
38.2
5.0
54.3
28.4
39.2
1.8
0
6.70 0.59
4.98 0.49
4.27 0.46
60.57 0.30
0.05 0.04
Ni-MOF-74
Zn-MOF-74
Blank
40.0
5.0
18.8
As the oxidation of cyclohexene can be catalyzed by at least
two different routes that give different chemical products
(Scheme 1),[31] an analysis of the yields of the products generat-
18.8
0
–
–
[a] Total (radical+epoxidation) TOF after 2 and 24 h of reaction.
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
4
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Therefore, the yield of the epoxidation route a priori emerges
as a promising variable to shed light on the catalytic role of
the open metal sites. As the crystal size and/or textural proper-
ties influence the catalytic activity (Figure 5), the ratio between
the yields of the redox-required epoxidation and the spontane-
ous radical pathways (E/R ratio, Table 3) is analyzed instead of
Table 3. E/R ratio of the different M–MOF-74 materials and reduction po-
tential (E0) for the most probable redox pair of each metal ion. Data for
samples that have positive and negative E0 values are written in green
and black, respectively, to underline the effect of the sign of the E0 in the
E/R ratio.
M
E/R ratio
Reduction potential E0 [V][a]
Cu
Mn
Co
Ni
1.52
1.36
0.66
0.05
0
+0.15
+1.54
+1.92
À0.25
À0.76
Zn
[a] E0 values correspond to the redox pairs Cu2+/Cu+, Mn3+/Mn2+
Co3+/Co2+, Ni2+/Ni0, and Zn2+/Zn0 in aqueous solution.[32]
,
the yield of the products formed by the epoxidation mecha-
nism. Thus, diffusional problems, which are unique to each cat-
alyst, should be cancelled, and then the intrinsic catalytic capa-
bility of the metal centers should emerge. Therefore, it seems
reasonable to try to relate the E/R ratio to the redox capability
of M, that is, to its redox potential (Table 3). Although the
redox potential values used in Table 3 are taken from the tabu-
lated reduction potential in aqueous solution,[32] they are ade-
quate to understand the catalytic behavior of metal oxides[33]
and some other properties of MOFs.[34] In any case, these
values will be discussed in qualitative terms in this work. Facili-
tated by the color code, the first message from the data
shown in Table 3 is that only the catalysts based on M with ox-
idizing power, that is, those with reduction potentials E0 >0,
are able to catalyze the cyclohexene epoxidation to a signifi-
cant extent. Conversely, M–MOF-74 with the most reductant
metal ion, Zn, was unable to catalyze the epoxidation of cyclo-
hexene in a detectable level.
Figure 6. Sum of the yields of the products generated by either the radical
oxidation (top) or epoxidation (bottom) of cyclohexene as a function of re-
action time for M–MOF-74 and for the blank experiment.
ed by these pathways can provide additional information. The
selectivity to either allylic oxidation or epoxidation of cyclohex-
ene for all tested catalysts is shown in Figure 6 top and
bottom, respectively. The complete distribution of the prod-
ucts at the end of the reaction is given in the Supporting Infor-
mation. The cyclohexene oxidation in the absence of a catalyst
(blank) and with the non-redox-active Zn-MOF-74 catalyst
takes place through the radical route. In other words, the allyl-
ic oxidation of cyclohexene is somehow spontaneous under
the tested reaction conditions. Nevertheless, it is evident that
any redox-active M enhances the conversion through that
route in the order Co>Ni~Cu>Mn, whereas Zn-MOF-74 hin-
ders the radical-based oxidation.
Following this reasoning, it could seem surprising that the
E/R ratio decreases as the positive reduction potential of M in-
creases (Table 3). However, it must be taken into account that
M is not the oxidizing species but the catalytic center, which
must complete the redox cycle (oxidation and subsequent re-
duction half reactions) to become ready to catalyze the oxida-
tion of another cyclohexene molecule.[35] This means that
a metal ion with a high oxidizing power such as Co will be re-
duced easily from Co3+ to Co2+ but the recovery the oxidation
state Co3+ from Co2+ will be kinetically difficult. However, the
cycle in the case of the Cu2+/Cu+ redox pair will not have such
a high barrier to overcome, so both the oxidation and reduc-
tion half reactions will be kinetically favored in comparison to
that of the Co3+/Co2+ redox pair. The need to complete the
redox cycle is evident from the turnover frequency (TOF)
values given in Table 1, which is 23.5 hÀ1 for Cu-MOF-74 after
0.5 h of reaction. A similar correlation between the catalytic
If we take into account the spontaneity of the allylic oxida-
tion of cyclohexene, it is not easy to extract any information
on the catalytic role of the redox center M. In addition, this
aim is even more complicated if we consider that the M–MOF-
74 materials have quite different crystal sizes and surface
areas, that is, different degrees of accessibility to the metal
centers. However, the presence of an MOF-74 catalyst that con-
tains a redox-active M is compulsory to oxidize cyclohexene
through the epoxidation pathway (Table 2 and Figure 6
bottom). Indeed, the blank and the Zn-MOF-74-catalyzed reac-
tions did not yield any product from the cyclohexene epoxide.
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
5
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
behavior and the reduction potential of the metal ions has
been described for metal oxides in the literature.[33]
Table 4. Metal leaching of the filtered mixtures after reaction measured
by TXRF and nature of the phase in the resultant solid according to
powder XRD (Figure 7).
Catalyst stability and metal leaching
M
Metal leaching
Phase identification
[mgLÀ1
]
The structural and compositional versatility of MOFs allows
them to potentially close the gap between selective homoge-
neous catalysts and more sustainable but less active/selective
heterogeneous catalysts. However, to be strictly considered as
heterogeneous, a catalyst must necessarily obey some guide-
lines, which includes no significant leaching of the active cen-
ters (metal ions in this case) and maintaining an intact struc-
ture. The study of these two points is particularly interesting in
this work because (i) the thermal and chemical stabilities of
MOFs are always under suspicion and (ii) metal leaching and/or
stability of the systematic series of M–MOF-74 catalysts could
provide some important information to help understand their
catalytic behavior.
Cu
Mn
Co
Ni
20
23
6
0.16
7
No crystalline phase
MOF-74
MOF-74
MOF-74
MOF-74
Zn
Ni as predicted from the much higher catalytic activity (and E/
R ratio) of the former. Notably, these two materials consist of
large particles formed by fused nanodomains rather than by
isolable crystals unlike Zn-, Cu-, and Mn-MOF-74. It is even pos-
sible that some disaggregated nanocrystals (<20 nm) of the
latter materials are able to cross the filtration system (the pore
diameter of the filter is 0.45 mm) so that the metal leaching is
overestimated by TXRF analysis. In any case, the detected
metal leaching seems to be the sum of two contributions:
(i) the catalytic activity of M (mainly through the epoxidation
pathway) and (ii) the presence of the disaggregated nanounits
in M–MOF-74.
The most relevant results of the total X-ray fluorescence
(TXRF) analysis of the filtered reaction mixtures as well as the
XRD characterization (Figure 7) of the M–MOF-74 catalysts
The powder XRD patterns of the five M–MOF-74 materials
tested in the cyclohexene oxidation are shown both before
and after reaction in Figure 7. The reacted samples were
simply collected by filtration before the patterns were mea-
sured. Moreover, the quality of the patterns of the samples
measured after reaction could show the effects of: (i) nanocrys-
tallinity, such as before the reaction, (ii) the presence of chemi-
cal species (reactants, solvent, products, etc.) within the pores,
and/or (iii) the scarce amount of the recovered sample, which
makes it difficult to measure a reliable quantitative diffracto-
gram. In any case, they indicate clearly that, except for the Cu-
based sample, all catalysts keep the MOF-74 structure to some
extent. Despite the temptation to attribute the structural trans-
formation of Cu-MOF-74 to its high catalytic conversion of cy-
clohexene, particularly by the epoxidation route, the fact that
other quite active M–MOF-74 catalysts maintain the MOF-74
structure leads us to think that it is related to another feature
of this sample.
Figure 7. Powder XRD patterns of the M–MOF-74 materials before (left) and
after (right) use as catalysts in the oxidation of cyclohexene.
after reaction are summarized in Table 4. Although metal leach-
ing could seem random at first glance, detailed analyses led us
to relate these values to different systematic tendencies in
both the catalytic behavior and the physicochemical properties
of these materials. Generally, metal leaching increases with the
catalytic activity (Figure 4) and with the E/R ratio (Tables 2 and
3) of the M–MOF-74. For instance, high leaching values (above
20 ppm) were found in the reaction mixtures catalyzed by Cu-
and Mn-MOF-74, which gave high catalytic conversions (90
and 68%, respectively; Table 2) and the highest E/R ratios (1.52
and 1.36; Table 3). Similarly, the almost inactive Zn-MOF-74 cat-
alyst releases three times less metal ions to the solution than
the other materials. However, the metal leaching generated by
Co- and Ni-MOF-74 is markedly lower than expected according
to the trend, although Co leaching is much higher than that of
The low chemical stability of Cu-MOF-74 in reaction media
has been made clear in the sole reported catalytic study with
this material.[15c] Cu-MOF-74 was transformed to copper(I) chlo-
ride under conventional conditions, although Cu was present
as CuII in the MOF-74 material and the tested reaction (acyla-
tion of anisole with acetyl chloride) does not require redox
centers. However, the tested reaction in the current work re-
quires redox centers. Unlike the other two most active cata-
lysts of the series, Mn- and Co-MOF-74, the redox cycle of Cu
is presumably between the oxidation states 2+ and 1+. It is
quite probable that the MOF-74 structure undergoes an irre-
versible structural transformation as CuII is reduced to CuI. That
reduction would imply the need to compensate a negative
framework charge in a region in which an exposed and unsa-
turated metal center that has a positive electronic density
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
6
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
acetate monohydrate was used as the Cu source. 2,5-Dihydroxyter-
ephthalic acid (dhtp) and the corresponding divalent metal acetate
(with different degrees of hydration) were dissolved individually in
DMF. A metal acetate/dhtp molar ratio of 2.6 was used. The linker
solution was added dropwise to the metal solution at RT under
stirring. The addition produces the immediate appearance of a pre-
cipitate. The resultant suspension was stirred at 238C for 18 h. The
solid was recovered by centrifugation and then washed with DMF
several times and subsequently with methanol three times. The
solid was kept submerged in methanol until the isotherms were
measured or until it was tested in the catalytic reaction.
charge is present. In this sense, it is remarkable that the ther-
mogravimetric (TG) profile of Cu-MOF-74 registered under an
air flow (Figure S4) shows the very low thermal stability of this
sample in comparison with its homologues (i.e., it is ꢀ1508C
less stable than Zn-MOF-74 and almost 3008C less stable than
nanocrystalline Mg-MOF-74). Therefore, in both oxidant media
(reaction and TG analysis under air flow), Cu-MOF-74 has
a much lower stability than its homologues. However, the in-
trinsic stability of Cu-MOF-74 is not much lower, which is dem-
onstrated if TG analysis is performed under an inert flow.[15c]
CoIII (or MnIII) is stabilized easily within the MOF-74 framework,
for instance, by linking an OHÀ group to the open metal site.
These OHÀ groups are generated in the initiation step of the
radical pathway.[36] Therefore, Mn- and Co-MOF-74, but not Cu-
MOF-74, could work as real catalysts that are able to recover
their structure and metal environment after a catalytic cycle.
The catalysts were studied by powder XRD before and after the re-
action to follow any possible structural modification. The PXRD
patterns were recorded by using a Philips X’PERT diffractometer
using CuKa radiation. (l=1.54 ꢃ) in the 2q range of 4–908. The
step size was 0.28, and the accumulation time was 50 s per step
with variable slit. The textural properties of the catalysts were stud-
ied by N2 adsorption–desorption at À1968C. Surface areas were es-
timated by applying the BET method, whereas the t-plot method
was used to estimate both the microporous and external surface
areas. The samples were evacuated at 1008C for at least 16 h
before the isotherms were measured.
Conclusions
A series of nanocrystalline M–MOF-74 (M=Mn, Co, Ni, Cu, or
Zn) materials prepared at room temperature has been tested
systematically in the catalytic oxidation of cyclohexene using
peroxides as oxidizing agents. tert-Butylhydroperoxide (TBHP)
was a more efficient oxidizing agent than H2O2 (30 wt% in
water), which, at least in the presence of the Ni-MOF-74 cata-
lyst, was absolutely inactive. The oxidation of cyclohexene with
TBHP takes place even in the absence of a catalyst. The nano-
crystalline M–MOF-74 materials were much more active than
their micrometer-sized homologues. M–MOF-74 with a redox-
active metal accelerates the spontaneous reaction, whereas
Zn-MOF-74 inhibits it partially. Any accessible redox center can
catalyze the cyclohexene oxidation to a similar extent in such
a way that the textural properties of the material rather than
the nature of M govern the total catalytic activity of M–MOF-
74. However, the route of cyclohexene oxidation (either
through a radical or epoxidation route) depends strongly on
the nature of M. In particular, the proportion of epoxidation
with respect to the total oxidation is notable for oxidant M
and negligible for reductant M. The maximum epoxidation
proportion is reached by M–MOF-74 with a slightly positive re-
duction potential (Cu), as it can complete its catalytic cycle
easily. Metal leaching increases with both the activity of the
catalyst and the ease of their particles to be disintegrated into
their nanocrystals/nanodomains. Amongst the most redox-
active M–MOF-74 (M=Cu, Co, or Mn), the Cu-based material is
unique in that it loses its MOF-74 structure during the reaction,
presumably because its reduced form CuI cannot be accommo-
dated in the M–MOF-74 structure with divalent M.
As a pretreatment step before the catalytic test, M–MOF-74 sam-
ples were placed in a round-bottomed flask submerged in a silicone
bath at 1508C overnight under a constant N2 flow. The oxidation
of cyclohexene was performed at 708C under atmospheric pres-
sure with stirring (400 rpm). The system was refrigerated with
water at 58C (the reactant evaporation was below 3 wt% after
48 h of reaction). The substrate cyclohexene (10 times the weight
of the evacuated catalyst), the solvent acetonitrile (cyclohexene/
acetonitrile molar ratio of 30), and the internal standard toluene
(the same weight as the catalyst) for GC were added to the cata-
lyst. Once this mixture reached 708C (the temperature was con-
trolled with a thermometer in contact with the mixture), the oxidiz-
ing agent (either H2O2 or TBHP) was added, and the moment of
this addition was taken as reaction time zero. Aliquots at different
reaction times (0–25 h) were taken under stirring to follow the ki-
netics. These aliquots were analyzed by GC (Varian 430 GC; capilla-
ry column of 15 m length, 0.25 mm diameter, 1 mm stationary
phase thickness; flame ionization detection; FID) and TXRF after fil-
tration (filter paper with pores of 0.45 mm). Identification of the
noncommercial 2-cyclohexenyl peroxide product was performed as
described elsewhere.[24a] As a result of the high volatility of cyclo-
hexene, its conversion was calculated indirectly through the sum
of the detected products. The mass balance was always below
10%.
Acknowledgements
The authors acknowledge the Spanish Government, MINECO
(MAT2012-31127). A.A. acknowledges CSIC for a PhD JAE-predoc
fellowship. Dr. E. Sastre is acknowledged for his help with the GC
analysis.
The results of this work provide a significant contribution to
the general knowledge of the catalytic behavior of MOF mate-
rials that contain open metal sites.
Keywords: heterogeneous
catalysis
·
metal–organic
frameworks · oxidation · radical reactions · redox chemistry
Experimental Section
[1] a) H. Li, M. Eddaoudi, M. O’Keeffe, O. M. Yaghi, Nature 1999, 402, 276–
279; b) S. S. Y. Chui, S. M. F. Lo, J. P. H. Charmant, A. G. Orpen, I. D. Wil-
liams, Science 1999, 283, 1148–1150; c) H. Li, M. Eddaoudi, T. L. Groy,
O. M. Yaghi, J. Am. Chem. Soc. 1998, 120, 8571–8572.
The five M–MOF-74 materials were prepared at RT according to the
method described in the literature.[18] In the preparation of Cu-
MOF-74, which has not been described expressly at RT, copper(II)
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
7
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[2] a) R. J. Kuppler, D. J. Timmons, Q.-R. Fang, J.-R. Li, T. A. Makal, M. D.
Young, D. Yuan, D. Zhao, W. Zhuang, H.-C. Zhou, Coord. Chem. Rev.
2009, 253, 3042–3066; b) Issue 5, Chem. Soc. Rev. 2009, 38, 1201–1508;
c) Issue 2 Chem. Rev. 2012, 112, 673–1268; d) Special issue on metal–
organic frameworks (Eds.: H.-C. Zhou and S. Kitagawa): Chem. Soc. Rev.
2014, 43, 16.
[3] C. Serre, F. Millange, C. Thouvenot, M. Nogues, G. Marsolier, D. Louer, G.
Ferey, J. Am. Chem. Soc. 2002, 124, 13519–13526.
[4] B. L. Chen, M. Eddaoudi, S. T. Hyde, M. O’Keeffe, O. M. Yaghi, Science
2001, 291, 1021–1023.
[5] a) B. L. Chen, M. Eddaoudi, T. M. Reineke, J. W. Kampf, M. O’Keeffe, O. M.
Yaghi, J. Am. Chem. Soc. 2000, 122, 11559–11560; b) J. L. C. Rowsell,
O. M. Yaghi, Angew. Chem. Int. Ed. 2005, 44, 4670–4679; Angew. Chem.
2005, 117, 4748–4758.
[6] M. Dinca˘, J. R. Long, Angew. Chem. Int. Ed. 2008, 47, 6766–6779; Angew.
Chem. 2008, 120, 6870–6884.
[7] a) D. Britt, H. Furukawa, B. Wang, T. G. Glover, O. M. Yaghi, Proc. Natl.
Acad. Sci. USA 2009, 106, 20637–20640; b) P. D. C. Dietzel, V. Besikiotis,
R. Blom, J. Mater. Chem. 2009, 19, 7362–7370.
[8] J. R. Li, J. Sculley, H. C. Zhou, Chem. Rev. 2012, 112, 869–932.
[9] a) M. P. Suh, H. J. Park, T. K. Prasad, D. W. Lim, Chem. Rev. 2012, 112,
782–835; b) K. Sumida, D. L. Rogow, J. A. Mason, T. M. McDonald, E. D.
Bloch, Z. R. Herm, T. H. Bae, J. R. Long, Chem. Rev. 2012, 112, 724–781;
c) H. Wu, Q. Gong, D. H. Olson, J. Li, Chem. Rev. 2012, 112, 836–868.
[10] a) A. Corma, H. Garcia, F. X. Llabres i Xamena, Chem. Rev. 2010, 110,
4606–4655; b) D. Farrusseng, S. Aguado, C. Pinel, Angew. Chem. Int. Ed.
2009, 48, 7502–7513; Angew. Chem. 2009, 121, 7638–7649; c) A. Dhak-
shinamoorthy, H. Garcia, Chem. Soc. Rev. 2014, 43, 5750–5765.
[11] N. L. Rosi, J. Kim, M. Eddaoudi, B. L. Chen, M. O’Keeffe, O. M. Yaghi, J.
Am. Chem. Soc. 2005, 127, 1504–1518.
[19] a) J. Sun, Q. Kan, Z. Li, G. Yu, H. Liu, X. Yang, Q. Huo, J. Guan, RSC Adv.
2014, 4, 2310–2317; b) H. Su, Z. Li, Q. Huo, J. Guan, Q. Kan, RSC Adv.
2014, 4, 9990–9996.
ˇ
[20] A. Dhakshinamoorthy, M. Opanasenko, J. Cejka, H. Garcia, Catal. Sci.
Technol. 2013, 3, 2509–2540.
[21] D. J. Tranchemontagne, J. R. Hunt, O. M. Yaghi, Tetrahedron 2008, 64,
8553–8557.
[22] A. Corma, P. Esteve, A. Martinez, J. Catal. 1996, 161, 11–19.
[23] a) R. Raja, S. O. Lee, M. Sanchez-Sanchez, G. Sankar, K. D. M. Harris,
B. F. G. Johnson, J. M. Thomas, Top. Catal. 2002, 20, 85–88; b) Y. H. Fu,
D. R. Sun, M. Qin, R. K. Huang, Z. H. Li, RSC Adv. 2012, 2, 3309–3314.
[24] a) A. Alfayate, C. Mꢁrquez-ꢄlvarez, M. Grande-Casas, B. Bernardo-Maes-
tro, M. Sꢁnchez-Sꢁnchez, J. Pꢅrez-Pariente, Catal. Today 2013, 213, 211–
218; b) A. Alfayate, C. Mꢁrquez-ꢄlvarez, M. Grande-Casas, M. Sanchez-
Sꢁnchez, J. Perez-Pariente, Catal. Today 2014, 227, 57–64; c) M. Guidot-
ti, C. Pirovano, N. Ravasio, B. Lꢁzaro, J. M. Fraile, J. A. Mayoral, B. Coq, A.
Galarneau, Green Chem. 2009, 11, 1421.
[25] A. Corma, P. Esteve, A. Martinez, S. Valencia, J. Catal. 1995, 152, 18–24.
[26] a) D. Jiang, T. Mallat, D. M. Meier, A. Urakawa, A. Baiker, J. Catal. 2010,
270, 26–33; b) K. Leus, I. Muylaert, M. Vandichel, G. B. Marin, M. Waro-
quier, V. Van Speybroeck, P. Van der Voort, Chem. Commun. 2010, 46,
5085–5087; c) K. Leus, M. Vandichel, Y.-Y. Liu, I. Muylaert, J. Musschoot,
S. Pyl, H. Vrielinck, F. Callens, G. B. Marin, C. Detavernier, P. V. Wiper, Y. Z.
Khimyak, M. Waroquier, V. Van Speybroeck, P. Van Der Voort, J. Catal.
2012, 285, 196–207.
[27] P. Canepa, Y. J. Chabal, T. Thonhauser, Phys. Rev. B 2013, 87, 5094407.
[28] A. Corma, M. Domine, J. A. Gaona, J. L. Jordꢁ, M. T. Navarro, F. Rey, J.
Pꢅrez-Pariente, J. Tsuji, B. McCulloch, L. T. Nemeth, Chem. Commun.
1998, 2211–2212.
[29] X. Solans-Monfort, J. L. Fierro, L. Hermosilla, C. Sieiro, M. Sodupe, R.
[12] P. D. C. Dietzel, Y. Morita, R. Blom, H. Fjellvag, Angew. Chem. Int. Ed.
2005, 44, 6354–6358; Angew. Chem. 2005, 117, 6512–6516.
[13] S. R. Caskey, A. G. Wong-Foy, A. A. J. Matzger, J. Am. Chem. Soc. 2008,
130, 10870.
[14] W. Zhou, H. Wu, T. Yildirim, J. Am. Chem. Soc. 2008, 130, 15268–15269.
[15] a) S. Bhattacharjee, J. S. Choi, S. T. Yang, S. B. Choi, J. Kim, W. S. Ahn, J.
Nanosci. Nanotechnol. 2010, 10, 135–141; b) M. Dꢀaz-Garcꢀa, M.
Sꢁnchez-Sꢁnchez, Microporous Mesoporous Mater. 2014, 190, 248–254;
c) R. Sanz, F. Martinez, G. Orcajo, L. Wojtas, D. Briones, Dalton Trans.
2013, 42, 2392–2398.
[16] a) L. J. Wang, H. Deng, H. Furukawa, F. Gandara, K. E. Cordova, D. Peri,
O. M. Yaghi, Inorg. Chem. 2014, 53, 5881–5883; b) J. A. Botas, G. Calleja,
M. Sanchez-Sanchez, M. G. Orcajo, Int. J. Hydrogen Energy 2011, 36,
10834–10844.
[17] a) G. Calleja, R. Sanz, G. Orcajo, D. Briones, P. Leo, F. Martꢀnez, Catal.
Today 2014, 227, 130–137; b) J.-S. Lee, S. B. Halligudi, N.-H. Jang, D.-W.
Hwang, J.-S. Chang, Y.-K. Hwang, Bull. Korean Chem. Soc. 2010, 31,
1489–1495; c) H.-Y. Cho, D.-A. Yang, J. Kim, S.-Y. Jeong, W.-S. Ahn, Catal.
Today 2012, 185, 35–40.
Mas-Balleste, Dalton Trans. 2011, 40, 6868–6876.
ˇ
[30] M. Polozˇij, M. Rubesˇ, J. Cejka, P. Nachtigall, ChemCatChem 2014, 6,
2821–2824.
[31] J. M. Fraile, J. I. Garcia, J. A. Mayoral, E. Vispe, J. Catal. 2001, 204, 146–
156.
[32] D. R. Lide, CRC Handbook of Chemistry and Physics, 87th ed., 2006–
2007.
[33] K. Klier, J. Catal. 1967, 8, 14–21.
[34] R. Das, P. Pachfule, R. Banerjee, P. Poddar, Nanoscale 2012, 4, 591–599.
[35] a) F. Corꢆ, M. Alfredsson, C. M. Barker, R. G. Bell, M. D. Foster, I. Saa-
doune, A. Simperler, C. R. A. Catlow, J. Solid State Chem. 2003, 176, 496–
529; b) I. Saadoune, F. Cora, M. Alfredsson, C. R. A. Catlow, J. Phys. Chem.
B 2003, 107, 3012–3018.
[36] M. Contel, C. Izuel, M. Laguna, P. R. Villuendas, P. J. Alonso, R. H. Fish,
Chem. Eur. J. 2003, 9, 4168–4178.
Received: November 19, 2014
Revised: December 11, 2014
Published online on && &&, 0000
[18] M. Dꢀaz-Garcꢀa, ꢄ. Mayoral, I. Dꢀaz, M. Sꢁnchez-Sꢁnchez, Cryst. Growth
Des. 2014, 14, 2479–2487.
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
8
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
Redox rocks: A series of nanocrystalline
M–MOF-74 (M=Mn, Co, Ni, Cu, Zn) ma-
terials prepared at room temperature
catalyze the oxidation of cyclohexene
with tert-butylhydroperoxide. Radical
oxidation is similar for any redox-active
M, whereas epoxidation is preferred for
M with slight oxidant properties.
D. Ruano, M. Dꢀaz-Garcꢀa, A. Alfayate,
M. Sꢁnchez-Sꢁnchez*
&& – &&
Nanocrystalline M–MOF-74 as
Heterogeneous Catalysts in the
Oxidation of Cyclohexene: Correlation
of the Activity and Redox Potential
M–MOF-74 catalysts with the redox
pairs M3+/M2+ were stable and those
with M2+/M+ were not.
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
9
ꢂ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ