Catalytic oxidation of aromatic oxygenates by the heterogeneous catalyst Co-ZIF-9

Article abstract of DOI:10.1016/j.apcata.2010.12.026

The catalytic activity of Co-ZIF-9, a zeolitic imidazolate framework of sodalite topology, in the oxidation of small aromatic molecules including phthalan, vanillyl alcohol, guaiacol, syringol, veratryl alcohol, and cinnamyl alcohol by molecular oxygen has been investigated. Co-ZIF-9 selectively oxidized the substrates and thus represents a promising heterogeneous catalyst for the oxidation of small aromatic substrates. Phthalan was converted to three different products, i.e. phthalide, phthalaldehyde, and small amounts of phthalic acid. In addition to phthalan, the alcohol functionality of vanillyl alcohol, veratryl alcohol, and cinnamyl alcohol was also similarly oxidized to yield the corresponding aldehydes in high yields and excellent selectivity. The effect of solvent, temperature and NaOH addition on the Co-ZIF-9 catalytic system has been studied. The presence of NaOH in the solution greatly enhanced the oxidation activity. The structure and stability of the heterogeneous catalyst was probed by UV-vis diffuse reflection spectroscopy, confirming the coordination geometry of cobalt in the catalyst framework, thermal gravimetric analysis, establishing the thermal stability of the structure, and XRD to monitor crystallinity of the catalyst. The heterogeneous nature of the catalyst was established by a hot filtration experiment, ICP analysis of the liquid phase, and by UV-vis absorption spectroscopy on the hot filtered reaction mixture.

Full text of DOI:10.1016/j.apcata.2010.12.026

Applied Catalysis A: General 394 (2011) 79–85
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Catalytic oxidation of aromatic oxygenates by the heterogeneous catalyst
Co-ZIF-9
Joseph Zakzeskia, Agnieszka De˛bczaka,b, Pieter C.A. Bruijnincxa, Bert M. Weckhuysena,∗
a Utrecht University, Inorganic Chemistry and Catalysis Group, Debye Institute for NanoMaterials Science, Sorbonnelaan 16, 3584 CA Utrecht, The Netherlands
b University of Maria Curie-Skłodowska, Department of Chemical Technology, Faculty of Chemistry, Pl. M. Curie-Skłodowskiej 3, 20-031 Lublin, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
The catalytic activity of Co-ZIF-9, a zeolitic imidazolate framework of sodalite topology, in the oxidation
of small aromatic molecules including phthalan, vanillyl alcohol, guaiacol, syringol, veratryl alcohol, and
cinnamyl alcohol by molecular oxygen has been investigated. Co-ZIF-9 selectively oxidized the substrates
and thus represents a promising heterogeneous catalyst for the oxidation of small aromatic substrates.
Phthalan was converted to three different products, i.e. phthalide, phthalaldehyde, and small amounts
of phthalic acid. In addition to phthalan, the alcohol functionality of vanillyl alcohol, veratryl alcohol,
and cinnamyl alcohol was also similarly oxidized to yield the corresponding aldehydes in high yields
and excellent selectivity. The effect of solvent, temperature and NaOH addition on the Co-ZIF-9 catalytic
system has been studied. The presence of NaOH in the solution greatly enhanced the oxidation activity.
The structure and stability of the heterogeneous catalyst was probed by UV–vis diffuse reflection spec-
troscopy, confirming the coordination geometry of cobalt in the catalyst framework, thermal gravimetric
analysis, establishing the thermal stability of the structure, and XRD to monitor crystallinity of the cata-
lyst. The heterogeneous nature of the catalyst was established by a hot filtration experiment, ICP analysis
of the liquid phase, and by UV–vis absorption spectroscopy on the hot filtered reaction mixture.
© 2010 Elsevier B.V. All rights reserved.
Received 16 October 2010
Received in revised form
28 November 2010
Accepted 20 December 2010
Available online 24 December 2010
Keywords:
Zeolitic imidazolate framework
Cobalt
Heterogeneous catalysis
Oxidation
Aromatics
1. Introduction
strated that Co salts in the ionic liquid [EMIM][DEP] serve as
excellent homogeneous catalysts for the oxidation of alcohol-
Catalytic oxidations are an important class of reactions for
the production of fine aromatic chemicals. With the depletion of
fossil fuels as chemical feedstock, the fraction of aromatic chem-
icals obtained from renewable resources, such as biomass, can be
expected to increase in the foreseeable future; thus, the develop-
ment of processes for the catalytic oxidation of biomass-derived
feed is becoming increasingly important [1]. Lignin, one of the
most abundant natural polymers, consists of methoxylated phenyl-
propane structures, and the depolymerization and subsequent
valorization of these units represents a viable method to produce
renewable aromatic compounds on which society depends [1].
One possible strategy to valorize lignin is to produce fine
chemicals via the selective oxidation of monomeric lignin-derived
aromatic molecules. Although many metals are used as oxida-
tion catalysts, promising examples include those that use cobalt
as a catalyst. Cobalt is known to readily activate molecular oxy-
gen for substrate oxidation [2–5], which is attractive for both
environmental reasons and associated lower chemical expenses
relative to other stoichiometric oxidants. Recently, we demon-
containing aromatic compounds, achieving very high turnover
frequencies [6,7]. Other examples include the oxidation of aro-
matic cyclic ethers directly with molecular oxygen, which has
been successfully performed with cobalt salts and complexes [2,3].
The anhydrous cobalt(II) chloride catalyzed oxidation of phthalan,
performed in 1,2-dimethoxyethane (DME) under an oxygen atmo-
sphere, resulted in the highest reported catalytic activity with 82%
phthalide yield [3]. Small quantities of phthalic anhydride were
also formed [3]. The principle disadvantage of all these processes
originates with the homogeneous nature of the catalyst, which
necessitates efficient catalyst separation and recovery industrially.
The development of a heterogeneous catalyst system capable of
using molecular oxygen for the oxidation of lignin-derived aromat-
ics to produce fine chemicals would thus be an attractive industrial
alternative to these homogeneous systems as it would allow for
facile separation and recovery of these catalysts.
Recently, a new class of crystalline porous materials known as
metal organic frameworks, more particularly the family of zeolitic
imidazolate frameworks (ZIFs), has been developed for use in vari-
ous applications such as carbon dioxide capture [8,9]. The primary
building unit of ZIFs [M(Im)4], (M = transition metal center; Co2+
,
Zn2+, Fe2+, Cu2+, Im = imidazolate-type linker) are structurally sim-
ilar to the TO4 tetrahedra of aluminosilicate zeolites and allow for
∗
Corresponding author.
E-mail address: B.M.Weckhuysen@uu.nl (B.M. Weckhuysen).
0926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.12.026

80
J. Zakzeski et al. / Applied Catalysis A: General 394 (2011) 79–85
Fig. 1. From left, single crystal X-ray structure of ZIF-9, a single sodalite cage, and the tetrahedral coordination of the benzimidazolate linkers to cobalt [11].
construction of open framework zeolite-like structures. Several of
these materials, especially ZIF-8, -9, and -10, display considerable
chemical and thermal stability [10]. Purple cubic crystals of Co-ZIF-
9 incorporate nets of corner-sharing tetrahedral CoN4 where the
bond between the benzimidazolate anion and Co(II) is one of the
most stable among N-donor ligands [11] (see Fig. 1). Co-ZIF-9 has an
open-framework structure adopting sodalite topology with hexag-
onal symmetry, composed of bridging benzimidazolate (PhIm)
anions and cobalt cations withtheformulaCo(PhIm)2·(DMF)·(H2O).
Tunable pore size along with unique diffusion pathways (e.g.
sodalite cages interconnected by six-membered rings as in the
case of ZIF-7, ZIF-8, ZIF-9, ZIF-90) make ZIFs attractive not only
for membrane and gas storage applications [11–15], but also for
use as heterogeneous catalytic systems [16,17]. Indeed, several
metal-organic-framework (MOF) catalysts have been developed for
alcohol oxidation [18] using air [19] or H2O2 [20]. The examples
of ZIFs used as heterogeneous catalysts are much more limited,
but Corma et al. showed that Co-ZIF-9, for instance, has excellent
catalytic activity for the oxidation of tetralin [21].
Inspired by the high catalytic activity demonstrated by the
homogeneous cobalt complexes and promising leads using Co-
ZIF-9 as a heterogeneous oxidation catalyst, we hypothesized
that catalytic oxidation of small, highly functionalized aromatic
molecules was also attainable using Co-ZIF-9, which could thus
potentially serve as a heterogeneous catalyst to replace the homo-
geneous systems described above. The objective of this work was
therefore to investigate the ability of Co-ZIF-9 to serve as a catalyst
for the oxidation of aromatic molecules containing benzylic alco-
hols, phenols, methoxy groups, and benzylic ethers, which could
be expected from lignin depolymerization feeds. We analyze the
performance of the catalyst in a wide range of solvents, test the
stability of the catalyst, and ascertain the heterogeneous nature of
the system.
2.2. Catalyst characterization
The powder X-ray diffraction pattern of the as-synthesized
Co-ZIF-9 was recorded with a Bruker AXS D8ADVANCE diffrac-
tometer using CoK␣ radiation. BET surface area and properties of
the pore structure were measured by TriStar 3000 V6.04 A. Ther-
mogravimetrical analysis (TGA) was carried out on a Pyris TGA
thermogravimetrical analyzer from Perkin Elmer with the sample
held in a ceramic pan in a continuous flow nitrogen atmosphere
(10 mL min−1). Samples were heated at a constant rate of 5 K min−1
UV–visible diffuse reflectance spectra of powdered catalyst were
recorded between 200 and 700 nm on a Varian Cary 500 Scan spec-
trophotometer equipped with an integration sphere.
.
2.3. Catalyst testing
Reactions were conducted in a 40 mL Parr stainless steel auto-
clave equipped with a thermocouple and a magnetic driver. Stirring
was done with an impellor at 750 rpm. Into 5 g of toluene there
was successively added: 0.13 mmol of powdered Co-ZIF-9 (0.05 g),
1 mmol NaOH (0.04 g, Aldrich, 97%) and 4.04 mmol of selected
substrate consisting of either phthalan (Aldrich, 97%), veratryl alco-
hol (Aldrich, 96%), syringol (Aldrich, 96%), guaiacol (Aldrich, 98%),
vanillyl alcohol (Acros Organics, 99%), or cinnamyl alcohol (Acros
Organics, 98%). Some substrates were contaminated with oxidized
product (i.e. cinnamyl aldehyde in the case of cinnamyl alcohol,
veratraldehyde in the case of veratryl alcohol, etc.). The quantity of
these contaminants was measured and calculations were adjusted
accordingly. The sealed autoclave with the reaction mixture was
first purged three times, charged with oxygen (0.5 MPa, Hoekloos,
99.5%) and then heated to the designated temperature. During the
course of the reaction, samples of 0.35 g were collected every hour
and for that purpose the autoclave was cooled to 343 K and vented.
The autoclave was again purged and charged with 0.5 MPa O2 and
heated. Between reactions, the autoclave was cleaned with water
and ethanol and then rinsed with acetone. The autoclave vessel was
treated overnight with aqueous NaOH in order to remove contam-
inants left on the walls.
2. Experimental
2.1. Catalyst preparation
Quantitative analysis of the products in solution was performed
using a Shimadzu GC-2010A gas chromatography unit equipped
with a WCOT fused silica CP-WAX 57CB column. Product iden-
tification was conducted using a Shimadzu GC/MS-QP2010 Mass
Spectrometer and compared with pure samples when available.
Co-ZIF-9 was prepared as described previously by Park et al.
[11]. Cobalt(II) nitrate hexahydrate (0.210 g, Sigma–Aldrich, 99%)
and benzimidazole (0.060 g, Sigma–Aldrich, 98%) were dissolved
separately in approximately 9 mL DMF. The solutions were then
combined in a capped 20 mL glass scintillation vial. This vial was
placed in a programmable oven, heated to 403 K from room tem-
perature at 5 K min−1 and held at this temperature for 48 h. The
vial was then cooled to room temperature at 0.4 K min−1. The
purple cubic crystals were separated from the mother liquor,
washed thrice with 3 mL DMF, and allowed to dry in air. The
dried catalysts were crushed using a mortar and pestle prior to
reaction.
2.4. Stability test of Co-ZIF-9
UV–visible spectroscopy was performed on the reaction mixture
after 12 h reaction with Co-ZIF-9 using toluene as solvent. A part
of the reaction mixture was hot filtered directly into a glass vial
and subsequently measured using a Varian Cary 50 Conc UV–vis
spectrometer equipped with an immersion probe. The immersion

J. Zakzeski et al. / Applied Catalysis A: General 394 (2011) 79–85
81
60
100
80
60
40
20
0
100
80
60
40
20
0
1
2
50
40
30
20
10
0
Catalyst removed
by hot filtration
3
4
0
60
120 240
time (min)
0
1
2
3
4
5
6
Fig. 2. Oxidation of phthalan by Co-ZIF-9 with molecular O2 with 1.4 mmol NaOH
(solid black) and without NaOH (dashed grey): phthalan (1), phthalide (2), phtha-
laldehyde (3), and phthalic acid (4). Reaction conditions: 7.00 g toluene, 0.07 g
Co-ZIF-9 (0.13 mmol), 0.675 g phthalan (5.59 mmol), 0.5 MPa O2, 363 K.
Time (h)
Fig. 3. Co-ZIF-9 catalyzed veratryl alcohol oxidation: veratryl alcohol conver-
sion versus time. Reaction conditions: 5.00 g toluene, 0.05 g Co-ZIF-9 (0.13 mmol),
1 mmol NaOH (0.04 g), 4.04 mmol veratryl alcohol (0.68 g), 0.5 MPa O2, 423 K.
probe with a 1 mm path length was placed directly in the solu-
tion. The spectrum of toluene was collected as a background. The
hot filtered solution was also submitted to the Mikroanalytisches
Labor Kolbe (Mülheim, Germany) for ICP elemental cobalt analy-
sis. The remaining solution was transferred into a clean autoclave,
pressurized with O2, and heated again.
!Caution: The flash point of benzimidazole is 419 K and extra
care must be taken to ensure that excess benzimidazole is removed
from the catalyst prior to conducting oxidation reactions to prevent
explosions at higher temperatures.
version in 20 min (see below for further discussion of NaOH in
the reaction). Phthalide was the most abundant product, which
was formed in 89% yield after one hour of reaction, followed by
phthalaldehyde (8% yield). Phthalic acid yield increased slightly
with increasing reaction time, whereas the quantity of the other
products remained approximately constant.
As depicted in Scheme 1, the reaction may proceed mecha-
nistically by the reaction of the phthalan with cobalt-activated
molecular oxygen to form a radical intermediate (A) as previously
observed with homogeneous cobalt catalysts [2,3]. This radical
intermediate can then react with molecular oxygen to form radical
intermediate (B), which either reacts with the Co-ZIF-9 catalyst to
form phthalide (C) or with water to form 1-hydroxyphthalane (D).
The phthalaldehyde was likely produced via ring opening of the
hemiacetal 1-hydroxyphthalane toform speciesE, and thehydroxyl
group is then subsequently oxidized by the Co-ZIF-9. Since species
D was not detected at the conclusion of the reaction, this isomer-
ization and subsequent oxidation likely occurs rapidly. According
to the findings of Bovicelli et al. the oxidation of cyclic hemiacetals
(species D) is faster than the cleavage of the phthalan C(1)–O(2)
bond [22], which is necessary to form species E. In support of
these observations, considerably higher yields of phthalide were
obtained than phthalaldehyde. The formation of the phthalic acid
likely occurs through the formation of radical intermediates (G)
and (H) in analogy to the formation of species (A) and (B). Fur-
ther reaction of radical intermediate H with Co-ZIF-9 results in
phthalic anhydride formation (not detected), which hydrolyzes to
form phthalic acid.
3. Results and discussion
3.1. Substrate oxidation
We begin our analysis of the Co-ZIF-9 system for the oxidation
of aromatic substrates with phthalan, although, as indicated below,
the Co-ZIF-9 system is capable of oxidizing a wide range of sub-
strates. The selective catalytic oxidation of phthalan represents a
potential route to obtain phthalide, which is an attractive industrial
intermediate for pharmaceuticals and fine chemicals. Initially, with
Co-ZIF-9 as catalyst, several solvents including water, acetonitrile,
methanol, ethanol, dimethylformamide and toluene were tested.
The best results were obtained using toluene, and this solvent was
used for all subsequent reactions. Initial experiments were con-
ducted at temperatures 363, 383, 403 and 423 K. Conversion was
found to increase with increasing temperature, and the highest cat-
alyst activity was achieved at 423 K (see below for analysis of the
stability of the catalyst versus temperature). The results of the Co-
ZIF-9 catalyzed oxidation of phthalan both with and without NaOH
using toluene as the solvent are given in Fig. 2. In both cases, phtha-
lan was readily oxidized to produce phthalide, phthalaldehyde and
phthalic acid using molecular O2 and Co-ZIF-9:
3.2. Other substrates
Based on the success of the Co-ZIF-9 catalyst in phthalan oxi-
dation, we extended the system to other aromatic compounds,
including veratryl alcohol, vannilyl alcohol, and cinnamyl alco-
hol. Veratryl alcohol (3,4-dimethoxybenzyl alcohol), synthesized
de novo by the P. chrysosporium fungus during its ligninolytic phase
[23–25], is a representative non-phenolic model compound for
lignin substructures and is widely used as a substrate in catalytic
oxidation reactions [26]. The results are presented in Fig. 3. The Co-
ZIF-9 catalyst selectively oxidized veratryl alcohol at the benzylic
position affording veratraldehyde, which was the only observed
No reaction occurred in the absence of the catalyst. Although
rapid phthalan oxidation occurred using just the Co-ZIF-9 (61% con-
version after 20 min), the presence of NaOH further accelerated the
oxidation rate, which resulted in nearly complete phthalan con-

82
J. Zakzeski et al. / Applied Catalysis A: General 394 (2011) 79–85
Scheme 1. Possible reaction path for the oxidation of phthalan.
product of the reaction:
nolics substituted with simple alkyl groups [27]. In order to test
for this reactivity, syringol (2,6-dimethoxyphenol) and guaiacol (3-
methoxyphenol) were used to study the performance of Co-ZIF-9
catalytic system for the oxidation. Both compounds were reactive
under the applied reaction conditions, however, benzoquinones
were not observed. Within 1 h of the reaction the guaiacol concen-
tration decreased from 3.78 mmol to 0.72 mmol (81%) and syringol
from 4.15 mmol to 0.49 mmol (88%). No monomeric products were
observed by GC, suggesting possible oligomerization of the sub-
strates, and a brown-colored solid formed on the bottom of the
autoclave.
After 4 h veratraldehyde was obtained in 46% yield, which cor-
responded to 1.9 mmol of veratryl alcohol consumed.
In contrast to veratryl alcohol, the reaction of vanillyl alco-
hol resulted in the consumption of 90% of the initial of
substrate (3.99 mmol) after 4 h to form vanillin (4-hydroxy-3-
methoxybenzaldehyde):
Cinnamyl alcohol contains both a C␥–OH and a C–C double
bond functionality and thus serves as an interesting compound to
100
80
60
40
20
0
The results of the oxidation are given in Fig. 4. Note that the
reaction rate tends to increase with time. This activity increase
may be attributed to changes in the long range structure of the
material, as discussed in further detail in Section 3.4 (XRD analy-
sis). No products other than vanillin were observed. Over-oxidation
of vanillin could give vanillic acid or quinones, which require
cleavage of carbon–carbon bond para to the phenolic group, but
these compounds were not detected. Successful oxidations with
Co/O2 complexes affording benzoquinones have been previously
reported, mostly on phenolics lacking para substituents or phe-
0
1
2
3
4
Time (h)
Fig. 4. Co-ZIF-9 catalyzed vanillyl alcohol oxidation: vanillyl alcohol conversion
versus time. Reaction conditions: 5.00 g toluene, 0.05 g Co-ZIF-9 (0.13 mmol),
1 mmol NaOH (0.04 g), 3.99 mmol vanillyl alcohol (0.48 g), 0.5 MPa O2, 423 K.

J. Zakzeski et al. / Applied Catalysis A: General 394 (2011) 79–85
83
40
35
30
25
20
15
10
5
40
35
30
25
20
15
10
5
50
40
30
20
10
0
9
Without NaOH
With NaOH
10
11
4
0
0
0
1
2
3
Time (h)
0
1
2
3
4
Fig. 5. Co-ZIF-9-catalyzed cinnamyl alcohol oxidation: conversion of cinnamyl
alcohol (1) to cinnamaldehyde (2) and epoxide (3). Reaction conditions: 5.00 g
toluene, 0.05 g Co-ZIF-9 (0.13 mmol), 1 mmol NaOH (0.04 g), 3.99 mmol cinnamyl
alcohol (0.53 g), 0.5 MPa O2, 423 K.
Time (h)
Fig. 6. Conversion of veratryl alcohol to veratraldehyde performed in the presence
of NaOH (0.04 g) and absence of NaOH. Reaction conditions: toluene (5.00 g), 0.05 g
Co-ZIF-9 (0.13 mmol), 4.04 mmol veratryl alcohol (0.68 g), 0.5 MPa O2, 423 K.
assess the activity of the catalyst towards these functional groups
[28–32]. The products resulting from its oxidation most often
include cinnamaldehyde, formed after oxidation of the C␥–OH alco-
hol functionality, along with other products derived from reaction
of the C–C double bond, depending on the reaction conditions and
metal catalyst used [28,33]. As shown in Fig. 5, cinnamyl alcohol
was the least reactive with oxygen in the presence of cobalt ZIF-9
compared to the previously discussed non-phenolic and phenolic
aromatic compounds.
present in the axial position of the complex stabilizes the Co–O2
bond [35], while in [EMIM][DEP] solution it accelerated the forma-
tion of the active catalytic complex by aiding in deprotonation and
coordination of the substrate [6], and similar activating effects may
be occurring in the Co-ZIF-9 material depending on the substrate.
In the case of phthalan, where substrate deprotonation is not a rel-
evant mechanistic step, the increase in rate is possibly attributed to
−OH enhanced O2 activation on the Co. In contrast, veratryl, vanillyl,
and cinnamyl alcohols may first require −OH aided deprotonation
and catalyst coordination for rapid oxidation to occur.
3.4. Catalyst characterization and stability under reaction
conditions
In order to confirm the structure and assess the stability of the
synthesized Co-ZIF-9 materials used for the oxidation of the com-
pounds listed above, XRD patterns of the catalyst were recorded.
The results obtained for the fresh catalyst are depicted in Fig. 7.
The recorded pattern corresponds well with the XRD pattern for
the isostructural Zn-ZIF-7 reported by Park et al. [11]. In addi-
tion, the structure and the coordination geometry of cobalt were
further confirmed by UV–vis diffuse reflection spectroscopy (see
Fig. 8). Taken together, the results confirm the successful synthesis
of Co-ZIF-9. The Co-ZIF-9 material exhibited bands in the region
Both functional groups of the molecule included in the propyl
side chain, C␥–OH and the carbon–carbon double bond, were sus-
ceptible to oxidation, but the formation of cinnamaldehyde was
distinctly favored. In addition to cinnamylaldehyde, a small quan-
tity of the epoxide species was detected.
3.3. NaOH
As indicated above, the presence of −OH is important for the
rapid oxidation of the substrates. The results of the oxidation of
phthalan, discussed above, and veratryl alcohol, given in Fig. 6,
both in the presence and in the absence of NaOH, exemplify this
importance. In the absence of NaOH, phthalan and veratryl alco-
hol oxidation occurred, but at a significantly lower rates than in
the presence of NaOH. In the case of veratryl alcohol and the other
substrates mentioned above, the yield of product exceeded the stoi-
chiometric quantity of NaOH present in the system, which indicates
that the −OH may serve to activate the catalyst or it is regenerated
as a result of the oxidation. Similar results were demonstrated for
several homogeneous cobalt oxidation catalysts. In particular, the
reaction of veratryl alcohol with molecular oxygen in the presence
of Co(salen) complexes was thoroughly enhanced by the presence
of NaOH either in aqueous medium or ionic liquids and differ-
ent reaction conditions were revealed to be crucial for increasing
the extent of veratryl alcohol oxidation [6,34]. The role of sodium
hydroxide was quite different depending on the reaction medium
used. In aqueous solutions, −OH acted as a donor ligand which when
Spent
Fresh
5
10 15 20 25 30 35 40 45 50
2(theta) (deg)
Fig. 7. X-ray diffraction patterns of the Co-ZIF-9 catalyst before and after reaction.

84
J. Zakzeski et al. / Applied Catalysis A: General 394 (2011) 79–85
0.6
1.0
0.8
0.6
0.4
0.2
0.0
7
6
5
4
3
2
1
0
0.5
0.4
0.3
0.2
0.1
0.0
400
500
600
700
Wavelength (nm)
0
1
2
3
400
500
600
700
Recycle #
Wavelength (nm)
Fig. 10. Veratrylaldehyde yield versus number of recycle steps. Reaction conditions:
0.05 g Co-ZIF-9 (0.13 mmol), 0.04 mmol NaOH (1 mmol), 0.5 MPa O2, 423 K, t = 1 h.
Solutions: toluene (5.00 g), 4.04 mmol veratryl alcohol (0.68 g). Catalyst isolated by
centrifugation and rinsed with toluene between reactions.
Fig. 8. UV–vis diffuse reflectance spectrum of Co-ZIF-9 sample and UV–vis of reac-
tion solution after 12 h at 423 K following hot filtration (inset).
between 450 and 650 nm with two distinct maxima at 540 and
590 nm along with a shoulder at 567 nm. In their study of Co-
containing aluminophosphate-based molecular sieves (Co-APO),
Lourenco et al. observed three bands with maxima near 542, 583
and 623 nm in the UV–vis diffuse reflection spectrum, which is
characteristic of cobalt in tetrahedral coordination and taken as
an indication of isomorphous substitution [36]. In the case of Co-
ZIF-9, bonding between the corner sharing nets formed between
the imidazole, which is among the most stable for N-donor ligands,
and Co(II) causes the shift in absorbance energy in UV–vis spectrum
relative to the oxygen-sharing Co-APO structures.
Fig. 9 gives the results of thermal gravimetric analysis of the Co-
ZIF-9 catalyst. Four weight loss steps were observed between 320
and 880 K. The first weight loss of 14.6%, the onset of which occurred
at approximately 400 K, corresponds to the removal of a portion of
the DMF and H2O contained in the framework Co-ZIF-9 (1 DMF
and 1 H2O; calcd. 19.9% and 4.9%, respectively). Most likely, DMF
is leaving the restricted porous system slower than H2O based on
its lower volatility and size relative to H2O. The second weight loss
of 22.3% between 560 and 680 K along with two small weight-loss
steps of 3% and 3.5% (680–880 K) result because of decomposition
of the ZIF framework. It can be concluded that Co-ZIF-9 is stable
under reaction conditions, but temperatures exceeding 560 K have
a deleterious effect on the framework resulting in its decomposi-
tion.
In addition to the XRD of the fresh catalyst, Fig. 6 includes the
XRD of the recovered spent catalyst, recorded after use in the vera-
tryl alcohol oxidation reaction. Comparison of the fresh and spent
catalyst indicates that the Co-ZIF-9 catalyst maintains its crys-
tallinity and overall structure as evident by similarities between the
fresh andspentsamples; however, the long range orderof themate-
rial was altered as evident by the disappearance of peaks between
10 and 15◦ and at 21◦. The decrease in the number of peaks is indica-
tive of increased order in the catalyst. In order to test if DMF or H2O
was expelled from the pore system under reaction conditions led
to this structural change, we heated the Co-ZIF-9 catalyst solely in
the presence of toluene. No differences were observed between this
sample and the fresh catalyst, which indicates that the differences
in the spent catalyst do not occur from loss of spectator molecules
but rather as a result of the substrate oxidation.
In order to further investigate the homogeneous or heteroge-
neous nature of the reactions, we performed UV–vis absorption
spectroscopy on the hot filtered reaction mixture. The spectrum
for this slightly yellow solution (see Fig. 8 inset depicted above)
does not exhibit absorption bands in the region between 500 and
700 nm, which, if present, would confirm the presence of cobalt
ions dissolved in the reaction mixture. When the hot-filtered solu-
tion was returned to a clean autoclave and reheated, no additional
catalytic oxidation activity was observed (see Fig. 3 above). Addi-
tionally, ICP elemental Co analysis of the solution revealed Co
concentrations of less than 1 ppm, which is the detection limit of
the method used. Taken together, these results support the het-
erogeneous nature of the catalyst, and the increase in activity
with reaction time observed with many of the substrates can be
attributed to a solid–solid restructuring of the framework, indi-
cated by XRD, rather than the slow dissolution of the cobalt ions.
The reactions likely take place on coordinatively unsaturated cobalt
defects that are possibly formed during the restructuring on the
external surface of the ZIF-9 framework. Framework cobalt is coor-
dinatively saturated, and although this saturation confers strength
and stability to the Co-ZIF-9 material, it also prevents sites for oxy-
gen or substrate coordination and activation. In addition, the small
pore size of the ZIF-9 relative to the size of the substrates severely
limits substrate diffusion into the ZIF-9, further supporting the
occurrence of the reaction on the exterior surface.
100
14.6%
(1.20 mg)
90
80
22.3%
(1.83 mg)
70
3.0%
(0.25 mg)
60
3.5%
(0.27 mg)
50
400
500
600
700
800
The recyclability of the catalyst was tested by repeatedly con-
ducting the oxidation of veratryl alcohol, isolating and rinsing the
catalyst with toluene, and continuing the oxidation with fresh solu-
tions of veratryl alcohol. The results, given as a percent of the initial
Temperature (K)
Fig. 9. Thermal gravimetrical analysis of Co-ZIF-9.

J. Zakzeski et al. / Applied Catalysis A: General 394 (2011) 79–85
85
veratryl alcohol conversion, are given in Fig. 10. Despite a discol-
oration of the catalyst possibly attributable to the formation of
coke-like species, the Co-ZIF-9 catalyst showed no apparent deac-
tivation even after the 3rd recycle.
After 12 h of reaction in the presence of Co-ZIF-9 (with added
NaOH), the colorless toluene turned slightly yellow. This coloration
can be attributed to toluene itself being slightly reactive towards
the catalytic system under reaction conditions, and two products
of its oxidation, benzyl alcohol and benzaldehyde, were found in
trace quantities (< 0.1%) after 4 h. The formation of the aldehyde
likely occurs from oxidation of the benzyl alcohol [37]. No benzoic
acid was detected.
[2] A.M. Romano, M. Ricci, J. Mol. Catal. A: Chem. 120 (1997) 71–74.
[3] P. Li, H. Alper, J. Mol. Catal. 72 (1992) 143–152.
[4] K. Kervinen, M. Allmendinger, M. Leskelä, T. Repo, B. Rieger, Phys. Chem. Chem.
Phys. 5 (2003) 4450–4454.
[5] K. Kervinen, H. Korpi, J.G. Mesu, F. Soulimani, T. Repo, B. Rieger, M. Leskelä, B.M.
Weckhuysen, Eur. J. Inorg. Chem. 13 (2005) 2591–2599.
[6] J. Zakzeski, A.L. Jongerius, B.M. Weckhuysen, Green Chem. 12 (2010)
1225–1236.
[7] J. Zakzeski, P.C.A. Bruijnincx, B.M. Weckhuysen, Green Chem. (2011),
doi:10.1039/C0GC00437E.
[8] A. Phan, C. Doonan, F.J. Uribe-Romo, C.B. Knobler, M. O’Keeffe, O.M. Yaghi, Acc.
Chem. Res. 43 (2009) 58–67.
[9] R. Banerjee, A. Phan, B. Wang, C. Knobler, H. Furukawa, M. O’Keeffe, O.M. Yaghi,
Science 319 (2008) 939–943.
[10] D. Farrusseng, S. Aguado, C. Pinel, Angew. Chem. Int. Ed. 48 (2009)
7502–7513.
[11] K.S. Park, Z. Ni, A.P. Cote, J.Y. Choi, R. Huang, F.J. Uribe-Romo, H.K. Chae, M.
O’Keeffe, O.M. Yaghi, Proc. Natl. Acad. Sci. U.S.A. 103 (2006) 10186–10191.
[12] Y. Li, F. Liang, H. Bux, W. Yang, J. Caro, J. Membr. Sci. 354 (2010) 48–54.
[13] B. Wang, A.P. Cote, H. Furukawa, M. O’Keeffe, O.M. Yaghi, Nature 453 (2008)
207–211.
[14] A. Phan, C.J. Doonan, F.J. Uribe-Romo, C.B. Knobler, M. O’Keeffe, O.M. Yaghi, Acc.
Chem. Res. 43 (2010) 58–67.
[15] W. Morris, C.J. Doonan, H. Furukawa, R. Banerjee, O.M. Yaghi, J. Am. Chem. Soc.
130 (2008) 12626–12627.
[16] H.-L. Jiang, B. Liu, T. Akita, M. Haruta, H. Sakurai, Q. Xu, J. Am. Chem. Soc. 131
(2009) 11302–11303.
[17] J.Y. Lee, O.K. Farha, J. Roberts, K.A. Scheidt, S.B.T. Nguyen, J.T. Hupp, Chem. Soc.
Rev. 38 (2009) 1450–1459.
[18] A. Corma, H. Garcia, F.X. Llabrés, I. Xamena, Chem. Rev. 110 (2010)
4606–4655.
[19] F.X. Llabrés i Xamena, A. Abad, A. Corma, H. Garcia, J. Catal. 250 (2007)
294–298.
[20] C.N. Kato, M. Hasegawa, T. Sato, A. Yoshizawa, T. Inoue, W. Mori, J. Catal. 230
(2005) 226–236.
[21] F.X. Llabrés i Xamena, O. Casanova, R. Galiasso Tailleur, H. Garcia, A. Corma, J.
Catal. 255 (2008) 220–227.
[22] P. Bovicelli, A. Sanetti, R. Bernini, Tetrahedron 530 (1997) 9755–9760.
[23] R. Bourbonnais, M.G. Paice, Biochem. J. 255 (1988) 445–450.
[24] K.B. Lee, M.B. Gu, S.-H. Moon, Eng. Life Sci. 6 (2001) 238–245.
[25] P.J. Harvey, H.E. Schoemaker, J.M. Palmer, FEBS 195 (1986) 243–246.
[26] K. Kervinen, P. Lathinen, T. Repo, M. Svahn, M. Leskelä, Catal. Today 75 (2002)
183–188.
[27] J.J. Bozell, B.R. Hames, D.R. Dimmel, J. Org. Chem. 60 (1995) 2398–2404.
[28] N. Dimitratos, A. Villa, D. Wang, F. Porta, D. Su, L. Prati, J. Catal. 244 (2006)
113–121.
4. Conclusions
Co-ZIF-9 can be effectively used as a heterogeneous catalyst for
the oxidation of several small aromatic molecules in the presence of
molecular oxygen. After 20 min under reaction conditions, nearly
complete phthalan conversion was obtained to yield three prod-
ucts, consisting of phthalide, phthalaldehyde, and small amounts
of phthalic acid. In addition to phthalan, the alcohol functional-
ity of vanillyl alcohol, veratryl alcohol, and cinnamyl alcohol was
also similarly oxidized to yield the corresponding aldehydes in high
yields and excellent selectivity. UV–vis absorption spectroscopy
and ICP analysis performed on the hot filtered reaction mixture
confirmed the heterogeneous nature of the Co-ZIF-9 material. The
catalyst was stable under the reaction conditions employed, as sup-
ported by TGA analysis; however, temperatures in excess of 560 K
lead to the decomposition of the catalyst framework. Although the
principle framework structure remained intact during reactions,
long-range reordering of the catalyst occurred as determined by
XRD analysis. The reaction rates were significantly enhanced in the
presence of NaOH, which was not ultimately consumed during the
course of the reaction. Co-ZIF-9 materials thus represent an attrac-
tive alternative to homogeneous catalysts for similar oxidations.
[29] M. Musawir, P.N. Davey, G. Kelly, I.V. Kozhevnikov, Chem. Commun. (2003)
1414–1415.
[30] T.L. Stuchinskaya, I.V. Kozhevnikov, Catal. Commun 4 (2003) 417–422.
[31] A. Abad, C. Amela, A. Corma, H. Garcia, Chem. Commun. (2006)
3178–3180.
[32] A. Abad, P. Concepcion, A. Corma, H. Garcia, Angew. Chem. Int. Ed. 44 (2005)
4066–4069.
[33] J.-D. Grunwaldt, C. Keresszegi, T. Mallat, A. Baiker, J. Catal. 213 (2003) 291–295.
[34] V.O. Sippola, A.O.I. Krause, J. Mol. Catal. A: Chem. 194 (2003) 89–97.
[35] K. Kervinen, M. Allmendiger, M. Leskelä, T. Repo, B. Rieger, Phys. Chem. Chem.
Phys. 5 (2003) 4450–4454.
Acknowledgements
JZ gratefully thanks the National Science Foundation Interna-
tional Research Fellowship Program for support of this research
under Award No. 0856754. AD thanks the Erasmus Programme for
financial support.
References
[36] J.P. Lourenco, M.F. Ribeiro, F.R. Ribeiro, Zeolites 18 (1997) 398–407.
[37] K.R. Sheddon, A. Stark, Green Chem. 4 (2002) 119–123.
[1] J. Zakzeski, P.C.A. Bruijnincx, A.L. Jongerius, B.M. Weckhuysen, Chem. Rev. 110
(2010) 3552–3599.