Propylene glycol oxidation with hydrogen peroxide over Zr-containing metal-organic framework UiO-66

Article abstract of DOI:10.1016/j.cattod.2018.11.063

Zirconium-based metal–organic framework UiO-66 catalyzes oxidation of propylene glycol (PG) using hydrogen peroxide as green oxidant. Hydroxyacetone (HA) is the main oxidation product, while the main side product is acetic acid (AcA). The nature of the solvent drastically affects PG adsorption, oxidant utilization efficiency and product yields. The best catalytic performance (85% selectivity towards HA at ca. 10% PG conversion) was achieved with water–acetonitrile (3/7 (v/v)) mixture as a solvent. Additives of radical chain scavengers produce a rate-inhibiting effect, suggesting radical chain mechanism of the oxidation process. The PG oxidation over UiO-66 proceeds without leaching of the active metal into solution, and the catalysis has a truly heterogeneous nature. The catalyst can be recycled without significant loss of activity and selectivity and retains its structure during at least five reuses.

Full text of DOI:10.1016/j.cattod.2018.11.063

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Propylene glycol oxidation with hydrogen peroxide over Zr-containing
metal-organic framework UiO-66
Viktoriia V. Torbinaa, Nadezhda S. Nedoseykinaa, Irina D. Ivanchikovaa,b, Oxana A. Kholdeevaa,b
,
⁎
Olga V. Vodyankinaa,
a Tomsk State University, 36 Lenin Ave., Tomsk, 634050, Russia
b Boreskov Institute of Catalysis, 5 Lavrentiev Ave., Novosibirsk, 630090, Russia
A R T I C L E I N F O
A B S T R A C T
Keywords:
Zirconium-based metal–organic framework UiO-66 catalyzes oxidation of propylene glycol (PG) using hydrogen
peroxide as green oxidant. Hydroxyacetone (HA) is the main oxidation product, while the main side product is acetic
acid (AcA). The nature of the solvent drastically affects PG adsorption, oxidant utilization efficiency and product
yields. The best catalytic performance (85% selectivity towards HA at ca. 10% PG conversion) was achieved with
water–acetonitrile (3/7 (v/v)) mixture as a solvent. Additives of radical chain scavengers produce a rate-inhibiting
effect, suggesting radical chain mechanism of the oxidation process. The PG oxidation over UiO-66 proceeds without
leaching of the active metal into solution, and the catalysis has a truly heterogeneous nature. The catalyst can be
recycled without significant loss of activity and selectivity and retains its structure during at least five reuses.
Propylene glycol
Selective oxidation
Hydroxyacetone
Alcohol oxidation
Hydrogen peroxide
Metal-organic frameworks
Zirconium
UiO-66
1. Introduction
intermediate product were observed. At the same time, selective pro-
duction of α-hydroxyketones in vapor phase was realized over copper
Nowadays, a growing attention is being paid to the problem of re-
placing carbon sources by chemical synthesis from fossil hydrocarbons to
biorenewable resources [1,2]. One of the products that could be obtained
from mono- and polysaccharides extracted from biomass is propylene
glycol (PG) [3]. A regioselective PG oxidation at secondary hydroxyl group
leads to the formation of hydroxyacetone (HA). HA plays an important
role in organic synthesis and finds application in food, textile, cosmetic
industries as well as in the production of polymers [4]. However, selective
oxidation of only one hydroxyl group in polyols is a challenging task in
view of their close reactivity [5]. Traditionally, HA is manufactured by
noncatalytic methods from sodium or potassium formate or acetate [6] or
via direct oxidation of acetone [7]. However, these multistep procedures
use toxic reagents in stoichiometric amounts and produce HA with low
selectivities. Biosynthetic routes using various substrates are developing
rapidly [8,9] but so far, low conversions and product yields, as well as long
reaction times are characteristic of these methods.
catalysts [21]. Nevertheless, there is a growing demand in the devel-
opment of effective catalytic systems for selective oxidation of alcohols,
especially polyols, in liquid phase using environmentally friendly oxi-
dants and easily recyclable heterogeneous catalyst [22–24].
The synthesis and application of metal–organic frameworks (MOFs)
have attracted a great attention during the last two decades due to the
possibility to obtain materials with a large variety of metal centers and
structures. Very high surface areas and pore volume, crystalline struc-
ture and a large fraction of accessible metal centers make MOFs pro-
mising candidates to be used as heterogeneous catalysts [25–30]. Sev-
eral examples of the selective oxidation of alcohols to carbonyl
compounds over MOFs have been reported- [31–36]. Copper-based
MOFs are the most widely used MOFs for alcohol oxidation due to the
intrinsic activity of this transition metal in oxidation catalysis. Cu-MOFs
revealed catalytic activity in the oxidation of various benzene ring-
containing alcohols in the presence of 2,2,6,6-tetramethylpiperidine-N-
oxyl (TEMPO) and/or organic base [31,32]. The use of Cu-MOF with
NH2-functionalized linkers has led to high activity without addition of
any base [33]. Schiff base as an organic linker has allowed avoiding the
addition of co-catalyst and led to the selective formation of benzyl al-
dehyde under mild temperature and solvent-free conditions [34]. Peng
et al. have compared the activity of Co, Zn, Ni, and Cu-containing
Several groups reported the selective formation of HA in liquid-
phase PG oxidation [10–12], while the majority of publications were
devoted to the production of lactic acid from PG using noble metal-
containing catalysts in alkaline conditions [13–16]. PG oxidation in a
gas phase proceeded with preferential formation of methylglyoxal and
CeC cleavage products [17–20]. Only small amounts of HA as an
⁎ Corresponding author.
E-mail address: vodyankina_o@mail.ru (O.V. Vodyankina).
https://doi.org/10.1016/j.cattod.2018.11.063
Received 12 December 2017; Received in revised form 23 November 2018; Accepted 27 November 2018
0920-5861/©2018ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:Torbina,V.V.,CatalysisToday,https://doi.org/10.1016/j.cattod.2018.11.063

V.V. Torbina et al.
CatalysisTodayxxx(xxxx)xxx–xxx
isostructural MOFs in aerobic oxidation of benzyl alcohol [35]. Co(bdc)
(ted)0.5 showed higher benzyl alcohol conversion (81.8%) and se-
lectivity towards benzaldehyde (> 99%) than the other systems. Benzyl
alcohol oxidation over Ir-MOFs in the presence of inorganic base and
iodobenzene afforded selective formation of benzaldehyde at moderate
conversions [36]. Nevertheless, until recently there were no examples
of selective oxidation of polyols over MOFs. Recently, we demonstrated
that the selective oxidation of PG to HA could be realized over Cr-
containing MOFs, MIL-100 and MIL-101, using tert-butylhydroperoxide
as environmentally friendly oxidant [37]. However, use of green oxi-
dants (molecular oxygen or hydrogen peroxide), which produce water
as the only by-product, is more preferable for practical applications
[38]. Moreover, the development of MOFs that can be used as truly
heterogeneous catalysts in aqueous or quasi-aqueous media without
any metal leaching is very attractive.
amounts (3 mmol) of terephthalic acid and ZrO(NO3)2·2H2O were dis-
persed in DMF (60 mL) and then hydrochloric acid (99 mmol) was added.
The mixture was placed into Teflon-lined stainless steel autoclave and
heated at 120 °C for 24 h. The obtained white solid was washed with
DMF and ethanol. Prior to characterization, the sample was degassed
under vacuum at 90 °C during 1 h and at 150 °C during 5 h.
Textural characteristics of the catalyst were determined from ni-
trogen adsorption isotherms (−196 °C; 3Flex instrument, Micromeritics).
Horvath-Kawazoe sphere pore geometry method was used to calculate
pore size distribution. The structure of the UiO-66 material was con-
firmed by XRD (Riguku Miniflex 600 X-ray diffractometer) and FT-IR
spectroscopy (SCIMITAR FTS 2000 spectrometer).
Zirconia was synthesized by thermal decomposition of ZrO(NO3)2∙2H2O
in air at 500 °C for 4 h. According to XRD, the resulting material consists of
a mixture of monoclinic and tetragonal ZrO2 phases (see Supporting in-
formation (SI), Fig. S1). Low-temperature nitrogen adsorption revealed the
formation of a mesoporous material with SBET 47 m2/g, pore volume
0.24 cm3/g, and average pore diameter 17 nm (Fig. S2 in SI).
In 2008, Cavka et al. reported the first synthesis of a porous zirconium
terephthalate UiO-66 [39]. This material attracted a lot of attention due to
unprecedentedly high thermal (up to 540 °C) and solvothermal stability
among MOFs. It consists of Zr-oxo-hydroxo clusters Zr6O4(OH)4 co-
ordinated by terephthalate ligands (Fig. 1). This MOF possesses two types
of pores with windows of 0.6 and 1 nm, the surface area of ca. 1200 m2/g,
and pore volume of 0.7 сm3/g. It has found wide application in adsorption
and gas storage [40,41] as well as an active and selective catalyst for
Lewis-acid catalyzed transformations [42,43]. However, to the best of our
knowledge, UiO-66 was not used as a heterogeneous catalyst for oxidation
reactions, particularly, with hydrogen peroxide as oxidant.
2.3. Catalytic experiments and product analysis
Catalytic oxidation tests were carried out in a thermostatted glass vessel
at 40–70 °C under vigorous stirring (500 rpm). Typically, the reaction was
initiated by the addition of H2O2 to a mixture containing 1 mmol of PG, 1 ml
of solvent and 3.7 mg (0.013 mmol Zr) of catalyst. Samples (1 μL) were
removed periodically through a septum and analyzed by GC (Chromatec
Crystal 5000.1, flame ionization detector, 30 m × 0.22 mm × 0.5 μm ZB-
WAX capillary column, chlorobenzene as an internal standard). At the end
of the reaction, the catalyst was separated and the filtrates were analyzed by
GC-MC (Agilent 7000B system with triple-quadrupole mass-selective de-
tector Agilent 7000 and GC Agilent 7890B, 30 m × 0.25 mm × 0.25 μm ZB-
WAX capillary column), and HPLC (Shimadzu Prominence-I LC-2030C,
Rezex ROA-Organic Acid H + column). Each experiment was reproduced
2–3 times. 1H NMR spectra were collected using Bruker Avance-400 spec-
trometer (400.13 MHz). The concentration of hydrogen peroxide was de-
termined by iodometric titration.
The aim of this work was to explore the potential of UiO-66 for the
selective oxidation of PG with hydrogen peroxide. The effects of reac-
tion conditions (solvent nature, temperature, concentrations of reagents
and atmosphere) on the selectivity and yield of HA and acetic acid
(AcA) have been studied. The nature of catalysis was addressed, and the
catalyst reusability was evaluated.
2. Experimental
2.1. Materials
Recycling experiments were performed under the optimized reac-
tion conditions. After the reaction the catalyst was filtered off, washed
with water or acetone, dried at room temperature overnight, and then
reused. The amount of zirconium in the filtrates was determined by
microwave plasma atomic emission spectroscopy (Agilent 4100 mi-
crowave spectrometer).
Propylene glycol (Vekton, 99%) was distilled under vacuum.
Acetonitrile was dried and stored over activated 4 Å molecular sieves.
Terephthalic acid (Acros Organics, 99%), zirconyl (IV) nitrate hydrate
(Acros Organics, 99.5%), N,N-dimethylformamide (DMF, ECOS-1,
99.8%), hydrochloric acid (Vekton, 38% aqueous solution) were used
without additional purification. The concentration of hydrogen per-
oxide (∼35%) was determined by iodometric titration prior to use.
2.4. Adsorption measurements
2.2. Catalyst preparation and characterization
All adsorption measurements were carried out at room temperature
(25 °C). A concentrated (3.3 M) solution of PG (HA or AcA) was added by
portions (30 μL) to a solution containing 2 ml of solvent, 8 μL of chlor-
obenzene (internal standard) and 20 mg of activated (150 °C, 5 h) UiO-
UiO-66 was synthesized by solvothermal method following the pro-
cedure reported by Cavka et al. [39] with some modifications. Equimolar
Fig. 1. Metal-oxide cluster (a) and representation of 3D fra-
mework structure of UiO-66 (b). Zr atoms are shown in dark
purple, O atoms in light purple and C atoms in orange.
Hydrogen atoms are omitted for clarity. (For interpretation of
the references to colour in this figure legend, the reader is
referred to the web version of this article).
Adapted from [54].
2

V.V. Torbina et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Table 1
Effect of solvent nature on PG oxidation with H2O2 over UiO-66.
Solvent
Conversion, %
H2O2 utilization
efficiency, %a
Selectivity, %
PG
H2O2
HA
AcA
H2O
8
6
6
8
8
60
18
22
27
25
14
34
32
36
39
63
68
75
85
83
14
12
13
13
14
MeCN
2/8(v/v)H2O/MeCN
3/7(v/v)H2O/MeCN
4/6(v/v)H2O/MeCN
Reaction conditions: 1 mmol PG, 1 mmol H2O2, 3,7 mg UiO-66, 1 ml of solvent, 50 °C, 2 h.
C(HA) + 3 • C(AcA)
a
H2O2 utilization efficiency =
× 100
.
C(H
2
O
2
)
C(H O )
2 2
at the end of reaction
loaded
–
66. After the addition of each portion, the mixture was stirred for 20 min
(this time is sufficient to reach the adsorption equilibrium) and then the
MOF was separated by centrifugation (6000 rpm, 15 min), and the con-
centration of organic substance was determined by GC (3–5 measure-
ments) using a gas chromatograph Chromos GC-1000 equipped with a
does not tend to replace hydroxyl and NO3 ions in the zirconium co-
ordination sphere to form the required MOF structure. On the other
hand, the formation of UiO-66 structure is expected to be favorable for
halide compounds such as ZrCl4. However, it was reported that acid
modulators, such as hydrofluoric, hydrochloric, formic and other acids,
can be used to improve the crystallinity of MOFs during the sol-
vothermal reactions [46–48]. The amounts of both acid and water had
impact on the crystallization kinetics of the UiO-66 formation [44]. We
found that the addition of a 33-fold excess of HCl relative to ZrO
(NO3)2·2H2O yielded a highly crystalline porous solid. The BET surface
area (1260 m2/g) was close to that of the UiO-66 material prepared by
Cavka [39] while the pore size distribution curve had two maximums
(Fig. S4 in SI), corresponding to two types of micropores (0.6 and
0.9 nm in diameter). The XRD and FT-IR data (Figs. S5 and S6 in SI)
confirmed the formation of the UiO-66 phase. No unidentified reflec-
tions were observed in the XRD pattern.
flame ionization detector and
a quartz capillary column BPX5
(30 m × 0.25 mm). Parallel, blank experiments (without MOF) were
performed. The amount of adsorbed PG (HA or AcA) was determined as a
difference between the concentrations of PG (HA or AcA) in the solutions
with and without MOF multiplied by the solution volume. Adsorption
constants were estimated using the Langmuir adsorption model.
3. Results and discussion
3.1. Synthesis and characterization of UiO-66
Traditionally UiO-66 is synthesized by the solvothermal method
using terephthalic acid and zirconium chloride [39]. However, the major
drawback of ZrCl4 is its high corrosive and hygroscopic behavior re-
sulting in the formation of a mixture with unknown proportions of ZrCl4
and ZrOCl2. As a result, the synthesis procedure has low reproducibility.
To avoid these drawbacks, Ragon et al. [44] used different zirconium
precursors, such as ZrOCl2·8H2O, (Zr[acac]4), Zr(SO4)2·xH2O,
3.2. Catalytic performance of UiO-66 in PG oxidation
3.2.1. Effect of solvent nature
The results of catalytic experiments of PG oxidation with hydrogen
peroxide over UiO-66 in different solvents are presented in Table 1.
Initially, water and a polar organic solvent, acetonitrile, were em-
ployed. In both cases, HA was the main product while predominated
side products were C–C cleavage products, including AcA, formic acid,
acetaldehyde, and formaldehyde. Scheme 1 shows possible routes for
their formation. No other overoxidation products, such as methyl-
glyoxal or pyruvic acid, were found in the reaction mixture (Sheme 1).
In water, at equimolar concentrations of PG and hydrogen peroxide,
conversion of the oxidant was much higher relative to conversion of the
substrate, which indicated that unproductive decomposition of the
oxidant was significant. Selectivity towards HA was 63%. In acetoni-
trile, H2O2 conversion did not exceed 18%, but PG conversion was also
low. At the same time, H2O2 decomposition was significant in the ab-
sence of organic substrate, but it did not occur without catalyst (Fig. 2).
This fact confirmed the ability of UiO-66 to activate hydrogen peroxide.
i
(CH3CO2)xZr(OH)y (x + y = 4), Zr(OH)2CO3ZrO2, and Zr(O-iPr)4 PrOH.
Only the use of ZrOCl2·8H2O in DMF led to pure UiO-66 solid.
When we attempted to employ a more available reactant, ZrO
(NO3)2·2H2O, instead of ZrCl4 or ZrOCl2·8H2O in the standard synthetic
procedure, an amorphous material with low surface area was obtained.
The low-temperature nitrogen adsorption isotherm revealed a hyster-
esis in the region of average relative pressures, showing the presence of
mesopores (Fig. S3 in SI), which could correspond to mesoporous ZrO2.
This can be rationalized if we remember that ability of zirconium to
attract ions and molecules into its coordination sphere decreases in the
following order: OH− > carbonate and α-oxycarboxylates > F-
> HSO4−, NO3−, carboxylate > H2O > alcohols > halide ions (Cl−
,
Br−, I−) [45]. Thus, if we start from ZrO(NO3)2·2H2O, terephthalic acid
Scheme 1. Possible routes of PG transformation over UiO-66.
3

V.V. Torbina et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 2. H2O2 decomposition in MeCN over UiO-66 (○), ZrO2 (●), and without
catalyst (■). Reaction conditions: [H2O2] = 1 M, 14.8 mg of catalyst, 4 ml of
solvent, 50 °C.
Fig. 4. PG oxidation with H2O2 without catalyst (♦), over 1.9 (●), 3.7 (▲), 7.4
(■), and 11.1 mg (▼) UiO-66. Filled symbols – PG, open symbols – HA.
Reaction conditions: 1 mmol PG, 1 mmol H2O2, 1 ml 7/3(v/v)MeCN/H2O,
50 °C.
The dependence of the initial reaction rate on the catalyst amount was
linear (Fig. S7 in SI), indicating that the reaction is first order in cata-
lyst, which suggests the absence of external diffusion limitations. Ad-
ditionally, we performed an experiment using mesoporous zirconia
instead of UiO-66 as catalyst. Previously, Jung and Bell investigated
structure of active sites on the surface of ZrO2 for samples with
monoclinic and tetragonal modification and their mixture and found
that all the zirconia samples have coordinatively unsaturated Zr4+ ca-
tions and possess Lewis acidity [49]. Studies on hydrogen peroxide
decomposition in the absence of organic substrate showed that zirconia
is more active in this reaction than UiO-66 (Fig. 2). Nevertheless, PG
conversion over zirconia did not exceed 1% although oxidant conver-
sion was close to 100%. On the other hand, the structure of UiO-66 [39]
as it is does not imply the presence of coordinatively unsaturated Zr4+
,
Fig. 3. Adsorption of PG on UiO-66 from MeCN solution (●), its fitting with
Langmuir model (○), and adsorption of PG from 7/3(v/v) MeCN/H2O solution
(■).
but namely this MOF enabled PG oxidation to HA. This may indicate the
crucial role of the specific microporous confinement of UiO-66 in this
reaction. We may assume that both enlarged PG concentration in the
MOF pores and the spatial arrangement of the Zr active sites therein are
significant for the observed catalysis. We cannot also exclude that some
defects present in the structure of UiO-66 [50] may contribute to the
activation of H2O2. However, further studies are needed to elucidate
their role.
To clarify the observed effect of solvent on the PG oxidation over
UiO-66, we studied PG adsorption from different solvents on this MOF
(Fig. 3). Indeed, UiO-66 is able to adsorb PG from MeCN and the ad-
sorption constant evaluated using Langmuir model is quite high
(290
60 M−1). This allowed us to suppose that adsorbed PG mole-
cules block zirconium active sites and thereby suppress adsorption and
activation of the oxidant. On the other hand, no adsorption of PG oc-
curred from a mixed solvent H2O/MeCN 3/7(v/v) (Fig. 3).
3.2.3. Effect of H2O2 and PG amount
Although complete conversion of hydrogen peroxide was not at-
tained under the condition of the experiments presented in Table 1, a
two-fold reduction of its amount led to decreasing PG conversion and
HA yield (Table 2). In turn, augmentation of H2O2 concentration did
In order to optimize adsorption of the reactants, different amounts
of water were added to acetonitrile (Table 1). The replacement of
20 vol.% of acetonitrile by water led to a slight increase in H2O2 con-
version and HA yield. Increase in water amount up to 30 vol.% resulted
in enhancement of both PG conversion (to the level of that attained in
water) and H2O2. Moreover, HA selectivity increased markedly (85
versus 63% in water) and carbon mass balance (HA + AcA) became
close to 100%. The oxidant utilization efficiency also enlarged from 14
to 36%. Additional increase in water concentration (40 vol.%) pro-
duced no significant changes of PG conversion and product selectivities.
Further optimization of the reaction conditions was performed using
the optimal mixture of the two solvents.
Table 2
Effect of H2O2 amount on PG oxidation over UiO-66.
[H2O2], M PG conversion, % Selectivity, %
H2O2 utilization efficiency,
%
HA
AcA
0.5
1
5
92
85
9
41
8
13
36
1.5
1b
1
9
72(80a) 12(20a)
> 34
34
10
10c
85
–
10
95
54
3.2.2. Effect of catalyst loading
The corresponding kinetic curves for PG oxidation with H2O2 ac-
quired using different amounts of UiO-66 are presented in Fig. 4. En-
larging of the catalyst amount increased initial reaction rates, but just
slightly influenced on the substrate conversion and product yields.
Without catalyst, the reaction was much slower and PG conversion did
not exceed 7% after 24 h. HA selectivity in this case was about 70%.
Reaction conditions: 1 mmol PG, 1 mmol H2O2, 3,7 mg UiO-66, 1 ml 3/7(v/v)
H2O/MeCN, 50 °C, 2 h.
a
b
Selectivity in 1 h.
Concentration corresponding to the total amount of H2O2 added (slow
addition during 1.5 h).
c
HA (0.2 mmol) was used as substrate.
4

V.V. Torbina et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 5. Arrhenius plot for PG oxidation with H2O2 over UiO-66.
Fig. 6. PG oxidation over UiO-66 with addition of BHT (●), chloroform (▲),
hydroquinone (■), and without addition of any inhibitor (▼). Filled symbols –
PG, open symbols – HA. Reaction conditions as in Fig. 4, inhibitor 0.03 mmol.
not affect the substrate conversion but it decreased the selectivity to-
wards HA and resulted in increasing amount of C–C cleavage products,
including AcA. It is noteworthy that the kinetic curves of HA and AcA
accumulation had maximum (Table 2, entry 3), indicating that further
oxidation (probably to COx) takes place. When HA was used as sub-
strate instead of PG, the selective formation of AcA was observed
(Table 2). The initial rate of PG oxidation was proportional to the H2O2
amount and independent on PG concentration, demonstrating that the
reaction is first reaction order in oxidant and zero order in substrate
(Fig. S8 in SI). Thus, we can infer that PG does not participate in the
rate-limiting step of the oxidation process.
PG adsorption on UiO-66 from mixtures of MeCN/H2O (Fig. 3) coupled
with zero reaction order in substrate (Fig. S8 in SI), we may suggest that
the first step of the catalytic reaction is homolytic decomposition of
H2O2 on the active zirconium sites. The hydroxyl radicals formed
during H2O2 degradation abstract H atom bound to C2 atom of the
substrate molecule, resulting in the formation of PG secondary radicals,
followed by their further transformations (Scheme 2). Given that me-
soporous ZrO2 is more active in H2O2 decomposition than UiO-66 (see
Fig. 2) but, in contrast to the Zr-MOF, it is not able to perform selective
conversion of PG to HA, we may assume that UiO-66 ensures an optimal
rate of H2O2 decomposition and concentration of PG within the mi-
cropores.
Gradual addition of H2O2 into the reaction mixture did not improve
significantly oxidant utilization efficiency, but it slightly increased PG
conversion and HA yield (see Table 2). We may suppose that the oxi-
dant addition by portions disfavors the formation of C–C cleavage
products.
3.4. Comparison of UiO-66 with other catalysts
3.2.4. Effect of temperature
Comparison of the catalytic performance of UiO-66 with other
catalytic systems reported in the literature for liquid-phase PG oxida-
tion to HA with H2O2 is presented in Table 3. The attainable level of PG
conversion was lower for UiO-66, but selectivity to HA was higher than
in case of PG oxidation over Pd nanoparticles [12] or Ti-containing
polyoxometalate (Ti-POM) [11]. TS-1 enabled higher HA selectivity at
higher PG conversion, but reusability of this catalyst and its stability
under the reaction conditions was not reported [10]. Moreover, milder
reaction conditions (50 °C) and shorter reaction times could be used for
oxidation of PG over UiO-66 (Table 3).
The increase in the reaction temperature from 40 to 60 °C ac-
celerated the reaction but had only little effect on the PG conversion
and product selectivities. Further rising of the temperature led to de-
creasing selectivity towards HA. The initial rates of PG oxidation over
UiO-66 exhibited a typical Arrhenius dependence (Fig. 5). The activa-
tion energy estimated from the corresponding plot equaled to 88 kJ/
mol. Such value is typical of reactions controlled by the chemical in-
teraction rather than diffusion of reactants [51].
3.3. Mechanistic tests
The presence of molecular oxygen can significantly influence on the
oxidation reactions with peroxides over MOF catalysts [52]. Recently,
we demonstrated participation of dioxygen as an oxidant together with
tert-butylhydroperoxide in the PG oxidation over Cr-MIL-100 and Cr-
MIL-101 [37]. Therefore, an experiment under inert atmosphere was
carried out to clarify the effect of O2 on PG oxidation with hydrogen
peroxide over UiO-66. The replacement of air for Ar produced no
changes in the rates of PG consumption and product accumulation (Fig.
S9 in SI) indicating that hydrogen peroxide was the sole oxidant in the
system.
3.5. Catalyst stability and reusability
High thermal and hydrothermal stability of UiO-66 was documented
[39], but its stability under turnover conditions of a catalytic reaction
in the presence of high amounts of polar organic substances and hy-
drogen peroxide has not yet been addressed. To verify the nature of the
catalysis in PG oxidation over UiO-66, we performed a hot filtration test
[53] during the reaction course. After catalyst separation, the reaction
in the filtrate stopped, which confirmed the truly heterogeneous nature
of the catalysis over UiO-66 (Fig. 7). The amount of zirconium in the
filtrate did not exceed 2 ppm. The retention of MOF structure was
confirmed by the FT-IR and XRD techniques (Figs. S10 and S11 in SI).
An increase in the relative intensity of the IR absorption band in the
range of 3000–3600 cm−1 can be due to an increased amount of guest
water molecules after catalysis in the water-containing medium. The
surface area of UiO-66 after the catalytic experiment slightly decreased
from 1260 to 1101 m2/g. The reason for that could be small amounts of
unwashed adsorbed components of the reaction mixture. The corre-
sponding pore size distribution plots (Fig. S12 in SI) confirm
Additives of conventional radical chain scavengers (2,6-di-tert-
butyl-4-methylphenol (BHT), hydroquinone and chloroform) led to
different rate-limiting effects, depending on the scavenger nature
(Fig. 6). The smallest effect was observed in the case of bulky BHT. It is
not surprising if we take into account that the kinetic diameter of the
BHT molecule exceeds the pore entrance of UiO-66 (0.6 nm) [38].
Nevertheless, based on the shape of the kinetic curves in the presence of
three different inhibitors, we can conclude that PG oxidation with H2O2
over UiO-66 proceeds through radical chain mechanism. The absence of
5

V.V. Torbina et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Scheme 2. Tentative mechanism of PG oxidation with H2O2 over UiO-66.
preservation of the specific porosity of UiO-66 after the catalysis.
Given that the structure of UiO-66 is stable under the turnover
conditions, we may suppose that a possible reason for incomplete
conversions of substrate and H2O2 could be adsorption of the reaction
products of the active Zr(IV) sites and/or blocking of the micropores.
Adsorption experiments revealed no adsorption of the main oxidation
product (HA) from the mixed solvent MeCN/H2O (Fig. S13 in SI). On
the other hand, additives of acetic acid produced a strong rate-in-
hibiting effect on the PG oxidation over UiO-66 (Fig. 8). In turn, ad-
sorption measurements confirmed the ability of UiO-66 to adsorb acetic
acid (Fig. S13 in SI). Therefore, we may conclude that adsorption of
AcA on the MOF surface is, most likely, responsible for the incomplete
conversion of the reactants observed during one catalytic run.
The recycling performance of UiO-66 strongly depends on the ac-
tivation method used between operation cycles (Table 4). While cata-
lyst washing with water led to progressive decrease in PG conversion
and HA yield, catalytic activity and selectivity of UiO-66 remained at
the same level during, at least, 5 reuses if the catalyst was washed with
acetone. Overall, after five operation cycles, ca. 50% of the starting PG
could be converted to HA with ca. 85% selectivity.
Fig. 7. Hot catalyst filtration test at 50 °C. ▼and ■ symbols correspond to
kinetic curves acquired without and with catalyst separation, respectively.
Filled symbols – PG, open symbols – HA.
4. Conclusions
In this work, we first demonstrated that the metal-organic frame-
work UiO-66 can be successfully synthesized using zirconyl nitrate as
the Zr source with the addition of an appropriate amount of HCl and
the resulting material can be employed as a heterogeneous catalyst in
liquid-phase selective oxidation using H2O2 as oxidant. In particular,
we found that this MOF catalyzes selectively oxidation of the chal-
lenging substrate, propylene glycol. The reaction proceeds with high
regioselectivity, leading to the formation of hydroxyacetone as the
main product. After optimization of the reaction conditions, we
managed to obtain HA with 85% selectivity at 10% PG conversion.
Under these conditions, zirconia is highly active in H2O2 decomposi-
tion but it gives PG conversion below 1%. The optimal solvent for PG
oxidation over UiO-66 is a mixture of MeCN and H2O with 30% vol.
water content. PG oxidation with hydrogen peroxide over UiO-66
proceeds via radical chain mechanism, but molecular oxygen does not
participate in the oxidation process. The nature of the catalysis over
UiO-66 is truly heterogeneous, and no metal leaching into solution
occurs. The progressive formation of the by-product acetic acid leads
to deactivation of UiO-66 caused by strong adsorption of the acid on
the catalyst. Importantly, UiO-66 can be recycled with the
Fig. 8. PG oxidation over UiO-66 with (■) and without (▼) addition of
0.2 mmol AcA. Filled symbols – PG, open symbols – HA. Reaction conditions as
in Fig. 4.
maintenance of the catalytic performance and MOF structure during,
at least, five operation cycles, provided the catalyst is washed with
acetone prior to reuse. The excellent recycling performance of UiO-66
partially compensates the relatively low PG conversion acquired
during one operation cycle.
Table 3
PG oxidation with H2O2 over various catalysts.
Catalyst
Solvent
Molar ratio
T, °C
time, h
PG conversion, %
HA selectivity, %
Refs.
PG/Metal
PG/H2O2
TS-1
H2O
150
12.5
7
2.5
90
70
95
50
8
32
35
35
10
94
77
60
85
[10]
Ti-POM
Pd-black
UiO-66
MeCN
H2O
0.25
0.025
1
5
[11]
10
2.5
[12]
H2O/MeCN
75
This work
6

V.V. Torbina et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Table 4
[18] D.Yu. Ebert, A.S. Savel’eva, N.V. Dorofeeva, O.V. Vodyankina, Kinet. Catal. 58
(2017) 701–706.
Recycling of UiO-66 in PG oxidation.
[19] J. Shen, W. Shan, Y. Zhang, J. Du, H. Xu, K. Fan, W. Shen, Y. Tang, J. Catal. 237
(2006) 94–101.
Washing solvent
Cycle
PG conversion, %
HA yield, %
[20] J. Shen, W. Shan, Y. Zhang, J. Du, H. Xu, K. Fan, W. Shen, Y. Tang, Chem. Commun.
(2004) 2880–2881.
water
I
10
6
8.5
5.0
3.2
8.5
8.7
8.7
8.5
8.6
[21] S. Sato, R. Takahashi, T. Sodesawa, H. Fukuda, T. Sekine, E. Tsukuda, Catal.
Commun. 6 (2005) 607–610.
II
III
4
[22] H.-B. Ji, K. Ebitani, T. Mizugaki, K. Kaneda, Catal. Commun. 3 (2002) 511–517.
[23] V. Cortés Corberán, M.E. González-Pérez, S. Martínez-González, A. Gómez-Avilés,
Appl. Catal. A 474 (2014) 211–223.
acetone
I
10
10
10
10
10
II
III
IV
V
[24] A. Villa, N. Dimitratos, C.E. Chan-Thaw, C. Hammond, L. Prati, G.J. Hutchings, Acc.
Chem. Res. 48 (2015) 1403–1412.
[25] J.L.C. Rowsell, O.M. Yaghi, Microporous Mesoporous Mater. 73 (2004) 3–14.
[26] A. Corma, H. García, F.X. Llabrés i Xamena, Chem. Rev. 110 (2010) 4606–4655.
[27] P. Silva, S.M.F. Vilela, J.P.C. Tomé, F.A. Almeida Paz, Chem. Soc. Rev. 44 (2015)
6774–6803.
Reaction conditions: 1 mmol PG, 1 mmol H2O2 (slow addition during 1.5 h),
3.7 mg UiO-66, 1 ml 3/7(v/v)H2O/MeCN, 50 °C, 2.5 h.
[28] O.A. Kholdeeva, Catal. Today 278 (2016) 22–29.
[29] A. Dhakshinamoorthy, M. Alvaro, H. Garсia, Catal. Sci. Technol. 1 (2011) 856–867.
[30] A. Dhakshinamoorthy, M. Opansenko, J. Čejka, H. Garcia, Catal. Sci. Technol. 3
(2013) 2509–2540.
Acknowledgements
[31] A. Dhakshinamoorthy, M. Alvaro, H. Garсia, ACS Catal. 1 (2011) 48–53.
[32] B.R. Kim, J.S. Oh, J. Kim, Ch.Y. Lee, Bull. Korean Chem. Soc. 36 (2015) 2799–2800.
[33] A. Taher, D.W. Kim, I.-M. Lee, RSC Adv. 7 (2017) 17806–17812.
[34] S. Aryanejad, Gh. Bagherzade, A. Farrokhi, Inorg. Chem. Commun. 81 (2017)
37–42.
The authors thank A.V. Livanova for XRD measurements and Dr. I.Y.
Skobelev for his help with adsorption studies. The research was sup-
ported by “The Tomsk State University competitiveness improvement
programme” (8.2.03.2018) and the Russian Ministry of Education and
Science (projectno. 4.4590.2017/6.7).
[35] L. Peng, Sh. Wu, X. Yang, J. Hu, X. Fu, Q. Huo, J. Guan, RSC Adv. 6 (2016) 72433.
[36] S. Abednatanzi, P.G. Derakhshandeh, A. Abbasi, P. Van Der Voort, K. Leus,
ChemCatChem 8 (2016) 3672–3679.
[37] V.V. Torbina, I.D. Ivanchikova, O.A. Kholdeeva, I. Yu, O.V. Skobelev, Vodyankina,
Catal. Today 278 (2016) 97–103.
Appendix A. Supplementary data
[38] M.G. Clerici, O.A. Kholdeeva (Eds.), Liquid Phase Oxidation Via Heterogeneous
Catalysis: Organic Synthesis and Industrial Applications, Wiley, New Jersey, 2013.
[39] J.H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S. Bordiga,
K.P. Lillerud, J. Am. Chem. Soc. 130 (2008) 13850–138051.
[40] A.D. Wiersum, E. Soubeyrand-Lenoir, Q. Yang, B. Moulin, V. Guillerm, M. Ben
Yahia, S. Bourrelly, A. Vimont, S. Miller, Ch. Vagner, M. Daturi, G. Clet, Ch. Serre,
G. Maurin, Ph.L. Liewellyn, Chem.–An Asian J. 6 (2011) 3270–3280.
[41] S. Chavan, J.G. Vitillo, D. Gianolio, O. Zavorotynska, B. Civalleri, S. Jakobsen,
M.H. Nilsen, L. Valenzano, C. Lamberty, K.P. Lillerud, S. Borgia, Phys. Chem. Chem.
Phys. 14 (2012) 1614–1626.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.cattod.2018.11.063.
References
[1] P. Gallezot, Chem. Soc. Rev. 41 (2012) 1538–1558.
[2] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827–1870.
[3] F. Cavani, S. Albonetti, F. Basile, A. Gandini, Chemicals and Fuels from Bio-Based
Building Blocks, Wiley, New Jersey, 2016.
[42] U.S.F. Arrozi, H.W. Wijaya, A. Patah, Y. Permana, Appl. Catal. A Gen. 506 (2015)
77–84.
[4] N.H. Mohamad, R. Awang, W.M.Z.W. Yunus, Am. J. Appl. Sci. 8 (2011) 1135–1139.
[5] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C.D.P. Angew, Chem. Int. Ed. 46
(2007) 4434–4440.
[43] J. Hajek, M. Vanduchel, B. Van de Voorde, B. Bueken, D. De Vas, M. Waroquier,
V. Van Speybroeck, J. Catal. 331 (2015) 1–12.
[44] F. Ragon, P. Horcajada, H. Chevreau, Y.K. Hwang, U. Lee, S.R. Miller, Th. Devic, J.-
S. Chang, Ch. Serre, Inorg. Chem. 53 (2014) 2491–2500.
[6] P.A. Levene, A. Walti, Org. Synth. 2 (1943) 5.
[7] Pastureau, Bull. Soc. Chem. 5 (4) (1909) 227.
[45] W.B. Blumenthal, The Chemical Behavior of Zirconium, Van Nostrand, N.Y, 1958.
[46] J. Ren, H.W. Langmi, B.C. North, M. Mathe, D. Bessarabov, Int. J. Hydrogen Energy
39 (2014) 890–895.
[8] J. Ko, I. Kim, S. Yoo, B. Min, K. Kim, C. Park, J. Bacteriol. 187 (2005) 5782–5789.
[9] R. Yao, Q. Liu, H. Hu, T.K. Wood, X. Zhang, Appl. Microbiol. Biotechnol. 99 (2015)
7945–7952.
[47] M.J. Katz, Z.J. Brown, Y.J. Colon, P.W. Siu, K.A. Scheidt, R.Q. Snurr, J.T. Hupp,
O.K. Farha, Chem. Commun. 49 (2013) 9449–9451.
[10] R.A. Sheldon, J. Dakka, Erdöel Erdgas Kohle 109 (1993) 520–522.
[11] O.A. Kholdeeva, B.G. Donoeva, T.A. Trubitsina, G. Al-Kadamany, U. Kortz, Eur. J.
Inorg. Chem. 2009 (2009) 5134–5141.
[48] Y. Han, M. Liu, K. Li, Y. Zuo, Y. Wei, Sh. Xu, G. Zhang, Ch. Song, Z. Zhang, X. Guo,
CrystEngComm 17 (2015) 6434–6440.
[12] R.S. Disselcamp, B.D. Harris, T.R. Hart, Catal. Commun. 9 (2008) 2250–2252.
[13] N. Dimitratos, J.A. Lopez-Sanchez, S. Meenakshisundararam, J.M. Anthonykutty,
G. Brett, A.F. Carley, S.H. Taylor, D.W. Khight, G.J. Hutchings, Green Chem. 11
(2009) 1209–1216.
[49] K.T. Jung, A.T. Bell, J. Mol. Catal. A: Chem. 163 (2000) 27–42.
[50] St. Dissegna, K. Epp, W.R. Heinz, G. Kieslich, R.F. Fisher, Adv. Mater. 1704501
(2018) 1–23.
[51] P. Atkins, J. De Paula, Elements of Physical Chemistry, 6th ed., University Press,
Oxford, 2013.
[14] Yu. Ryabenkova, P.J. Miedziak, N.F. Dummer, St.H. Taylor, N. Dimitratos,
D.J. Willock, D. Bethell, D.W. Knight, G.J. Hutchings, Top. Catal. 55 (2012)
1283–1288.
[52] N.V. Maksimchuk, K.A. Kovalenko, V.P. Fedin, O.A. Kholdeeva, Chem. Commun. 48
(2012) 6812–6814.
[15] E. Redina, A. Greish, R. Nivikov, A. Strelkova, O. Kirichenko, O. Tkachenko,
G. Kapustin, I. Sinev, L. Kustov, Appl. Catal. A Gen. 491 (2015) 170–183.
[16] Y. Feng, W. Xue, H. Yin, M. Meng, A. Wang, Sh. Liu, RSC Adv. 5 (2015)
106918–106929.
[53] R.A. Sheldon, M. Wallau, I.W.C.E. Arends, U. Schuchardt, Acc. Chem. Res. 31
(1998) 485–493.
[54] M.J. Cliffe, W. Wan, X. Zou, Ph.A. Chater, A.K. Kleppe, M.G. Tucker, H. Wilhelm,
N.P. Funnell, F.-X. Coudert, A.L. Goodwin, Nat. Commun. 5 (2014) 4176.
[17] M. Ai, A. Motohashi, S. Abe, Appl. Catal. A Gen. 246 (2003) 97–102.
7