Zr and Hf-metal-organic frameworks: Efficient and recyclable heterogeneous catalysts for the synthesis of 2-arylbenzoxazole via ring open pathway acylation reaction

Article abstract of DOI:10.1016/j.jcat.2019.04.023

Zirconium- and hafnium-based metal–organic frameworks which constructed by 12-coordinated clusters and 6-coodinated clusters were shown to be highly effective heterogeneous catalysts for the ring opening acylation of benzoxazole to 2-arylbenzoxazole under solvent free conditions. Owning the wide opening spaces structures and inherent formate sites, MOFs based on 6-connected Zr₆/Hf₆ node were able to identify a significantly enhanced yield in Br?nsted acid catalyzed reactions under conventional heating and microwave irradiation. In addition, the detailed mechanism of active sites of the ring opening acylation reaction was confirmed by employing of density functional theory (DFT) calculations.

Full text of DOI:10.1016/j.jcat.2019.04.023

Journal of Catalysis 374 (2019) 110–117
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Zr and Hf-metal-organic frameworks: Efficient and recyclable
heterogeneous catalysts for the synthesis of 2-arylbenzoxazole via ring
open pathway acylation reaction
Linh Ho Thuy Nguyen a,b, Trang Thi Thu Nguyen a, Phuong Hoang Tran b, Yoshiyuki Kawazoe c,
Hung Minh Le a, , Tan Le Hoang Doan a,b,
⇑
⇑
a Center for Innovative Materials and Architectures (INOMAR), Vietnam National University-Ho Chi Minh City (VNU-HCM), Ho Chi Minh City 721337, Viet Nam
b Department of Organic Chemistry, Faculty of Chemistry, University of Science, Vietnam National University-Ho Chi Minh City, Ho Chi Minh City 721337, Viet Nam
c New Industry Creation Hatchery Center, Tohoku University, Sendai 980-8579, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Zirconium- and hafnium-based metal–organic frameworks which constructed by 12-coordinated clusters
and 6-coodinated clusters were shown to be highly effective heterogeneous catalysts for the ring opening
acylation of benzoxazole to 2-arylbenzoxazole under solvent free conditions. Owning the wide opening
spaces structures and inherent formate sites, MOFs based on 6-connected Zr6/Hf6 node were able to iden-
tify a significantly enhanced yield in Brønsted acid catalyzed reactions under conventional heating and
microwave irradiation. In addition, the detailed mechanism of active sites of the ring opening acylation
reaction was confirmed by employing of density functional theory (DFT) calculations.
Ó 2019 Elsevier Inc. All rights reserved.
Received 28 January 2019
Revised 12 April 2019
Accepted 15 April 2019
Keywords:
Acylation
Metal-organic framework
Benzoxazoles
Ring-opening mechanism
Heterogeneous catalysis
1. Introduction
of the catalysts [16], and the use of oxidizing agents or strong base
with halogenated solvents [17]. Consequently, the development
2-Aryl substituted benzoxazoles are an important class of hete-
rocyclic organic molecules in various fields including pharmaceuti-
cal research, natural products chemistry, and agrochemical
industry [1–4]. There are several reported strategies for the synthe-
sis of these substances. Among them, two typical approaches are
the condensation of 2-aminophenols with carboxylic acid deriva-
tives [5] or aldehydes [6–8], and the transition-metal-catalyzed
direct C-H functionalization of benzoxazoles with acylating agents
[9–11]. Yet, they have some drawbacks such as the use of expen-
sive catalysts and organic ligands which cannot be reused and
recovered [11–13], exhibit long reaction times [6,14], volatile
organic solvents [7,12], or requirement of prefunctional substrates
[9,15]. Recently, the ring-opening and ring-closing reaction
between benzoxazoles and benzoyl chlorides [16] or aldehydes
[17] has been emerged as a potential pathway for the construction
of benzoxazoles. Though this method shows noticeable advantage
of high yield and mild reaction conditions, its application remains
challenging due to high loading of catalysts, difficulty in recovery
and exploration of effective systems for acylation of benzoxazole
are highly desired.
Metal-organic frameworks (MOFs), a class of porous materials
constructed from inorganic metal cluster as nodes and organic
linkers, have attracted enormous attention for catalysis [18–21].
Among MOF materials, Zr(IV)- and Hf(IV)-based MOFs have been
emerged as promising heterogeneous catalysts for organic reac-
tions because of their structural efficiency [22–24]. These materials
have highly active sites including Zr4+ or Hf4+ (Lewis acid) [25,26]
and metal-bound hydroxides/water (Brønsted acid) on the Zr(IV)
or Hf(IV)-oxo clusters within the frameworks [27–29]. Due to
strong affinity of the Zr4+ or Hf4+ ions with carboxylate groups of
organic linkers, their structures achieve high chemical stability
leading to the ease of recycling and possibility for microwave-
assisted reactions [23,30]. We have recently studied the efficiently
heterogeneous catalytic activity of Zr- and Hf- MOFs in various
organic transformations including 2-arylbenzoxazole synthesis
[31]. Adopting knowledge from the previous works, herein, we
describe unprecedented ring-opening and ring-closing 2-
arylbenzoxazole synthesis from benzoxazole and benzoyl chlorides
catalyzed by Zr- and Hf-MOFs. In this research, we focused on the
reactivity of MOFs built from 12-connected and 6-connected Zr6/
⇑
Corresponding authors.
E-mail addresses: hung.m.le@hotmail.com (H.M. Le), dlhtan@inomar.edu.vn
(T.L.H. Doan).
https://doi.org/10.1016/j.jcat.2019.04.023
0021-9517/Ó 2019 Elsevier Inc. All rights reserved.

L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
111
Hf6 clusters coordinated with ditopic and tritopic carboxylate link-
2.2. Catalytic performance
ers, respectively (Fig. 1). According to the reaction studies, Zr- and
Hf-MOFs exhibited high catalytic activity in the solvent-free 2-
arylbenzoxazole synthesis under conventional heating and micro-
wave irradiation. It should be noted that the efficiency of VNU-11
(Hf), 6-connected Hf-cluster-based MOF, was significantly higher
in comparison to the other MOFs and catalysts. In addition, theo-
retical calculations were employed to determine role of the metal
cluster in the proposed reaction mechanism.
Following the material preparation, the catalytic activity of the
materials was initially studied through the optimization of the
exemplary reaction between benzoxazole (1) and benzoyl chloride
(2) under catalysis of VNU-11(Hf) (Table 1 and Section S6). Specif-
ically, the effect of solvent on the reaction was investigated by
screening various solvents and under solvent-free conditions. The
investigation indicated that the best result obtained at 170 °C for
3 h using 1 mol% catalyst under solvent-free conditions could
afford the desired compound 3 (2- phenylbenzoxazole) with an
excellent yield of 98% (Table 1, Entry 7). Whereas reasonable yields
could achieve in dichloromethane and toluene (60 vs 59%, Table 1,
Entries 4 and 5), 1,4-dioxane used as reaction solvent reached 96%
yield after 3 h (Table 1, Entry 6). However, the reaction in the polar
solvents gave low yields (less than 25%, Table 1, Entries 1–3).
After screening the optimized conditions for the reaction, the
catalytic behavior of Zr- and Hf-MOFs was compared by adopting
the arylation of benzoxazole as model reaction. As shown in
Table 2, the catalytic activity of MOF-808(Zr) and VNU-11(Hf) in
the reaction was superior to that of UiO-type MOFs, VNU-1(Zr),
and VNU-2(Hf) (Table 2, Entries 2–9). Under the reaction condi-
tions without a solvent, the performance of Hf-MOFs was more
efficient than that of the Zr analogues due to higher oxophilic char-
acter and stronger acidity of the Hf-oxo clusters, as described in
Section S5 [27,29,35]. The higher catalytic activity of the MOFs
with larger pore size revealed that the reaction proceeded not only
on the surface on the surface but also within the pores. Also, the
best catalytic performance was obtained using MOF-808(Zr) or
2. Results and discussion
2.1. Catalysts characterization
All Zr- and Hf-MOFs were synthesized via solvothermal reaction
with slight modifications to the methods reported in the literature
[31–34]. The detailed synthesis and activation procedures were
described in the Supporting Information (SI). In general, microcrys-
tal powder of the MOFs were obtained through the reaction
between ZrOCl2Á8H2O or HfOCl2Á8H2O and organic linkers in N,N-
dimethylformamide (DMF) with monocarboxylic acid added as a
modulator. The reaction was heated at 120 °C for 1 day in an
isothermal oven. As-synthesized sample was washed with DMF
to remove guest molecules and solvent-exchanged. The solvent-
exchanged sample was activated by heating under reduced pres-
sure for 24 h. The activated samples were confirmed by various
techniques including Powder X-ray Diffraction (PXRD) patterns
and N2 sorption isotherms to confirm successful activation and
pure phase of the materials.
Fig. 1. Zr6/Hf6 cluster (a) is connected with organic linkers to form MOFs. UiO-type MOF structures of Zr/Hf-UiO-66 (b) and Zr-, Hf-UiO-67 (b) are built by 12-connected
cluster and ditopic linkers, H2BDC and H2BPDC, respectively. Interpenetrated structure (d) of VNU-1(Zr) and VNU-2(Hf) is constructed from 12-connected cluster and slim,
linear H2CPEB linker. Structures (e) of MOF-808(Zr) and VNU-11(Hf) are based on 6-connected cluster and tritopic linkers H3BTC. Color code: C, black; O, red; Zr or Hf, blue
polyhedra. The yellow and purple balls represent the cages of the structures. (For interpretation of the references to color in this figure legend, the reader is referred to the
web version of this article.)

112
L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
Table 1
Screening of the reaction conditions.
Entry
Type of solvent
Solvents
Cat. (mol%)
Isolated yield (%)
1
2
3
4
5
6
7
Polar protic
Ethanol
n-Buthanol
DMF
Dichloromethane
Toluene
1,4-Dioxane
–
1
1
1
1
1
1
1
0.5
0.1
0
23
17
2
60
59
96
98
53
31
0
Polar aprotic
Nonpolar
Free-solvent
A mixture of VNU-11(Hf), benzoyl chloride (0.140 g, 1 mmol) and benzoxazole (0.119 g, 1 mmol) in a flask bottom was stirred under solvent condition at 170 °C for 3 h. Yield
of compound 3 was isolated by column chromatography (acetone/petroleum).
Table 2
Comparative study of catalysts used in the synthesis of 2-phenylbenzoxazolea.
Entry
1
Type
Catalyst
_
ABbET (m2 gÀ1
_
)
PSDc (Å)
_
Yieldd (%)
4
TOF (hÀ1
_
)
Free catalyst
MOFs
2
3
4
5
6
7
8
9
Hf-UiO-66
Zr-UiO-66
Hf-UiO-67
Zr-UiO-67
VNU-2(Hf)
VNU-1(Zr)
VNU-11(Hf)
MOF-808(Zr)
800
11.2
11.2
14.5
14.5
20.0
20.0
15.0
15.0
56
48
76
60
85
74
98
89
18.7
16.0
25.3
20.0
28.3
24.7
32.7
29.7
1150
1720
2105
1700
2100
1530
2060
10
11
12
13
14
15
16
Metal salts
ZrO2
TiO2
AlCl3
FeCl3
_
_
_
_
_
_
_
_
_
_
_
_
_
_
27
32
46
40
43
23
20
9.0
10.7
15.3
13.3
14.3
7.7
ZnCl2
HfOCl2Á8H2O
ZrOCl2Á8H2O
6.7
a
b
c
Reaction conditions: benzoxazole (5 mmol) and benzoyl chloride (5 mmol) were stirred under free solvent at 170 °C for 3 h in the presence of catalyst (1 mol%).
Calculated Brunauer–Emmett–Teller surface areas.
Pore size distribution.
Yield of compound 3 was isolated by column chromatography (acetone/petroleum).
d
VNU-11(Hf). This mainly results from the Zr-OH/Zr-OH2 or Hf-OH/
Hf-OH2 facing the pores in the structure, which makes the catalyt-
ically active sites more accessible for the substrates to react. How-
ever, the traditional Lewis acids including metal salts and metal
oxides were not well reactive with the yields obtained from 20%
to 46% (Table 2, Entries 10–16). In order to highlight role of acid
sites in the acylation, the control experiment was performed with-
out the presence of any catalysts. From the result in Table 2, it was
clear that the reaction did not proceed by heating alone (Table 2,
Entry 1).
In order to broaden the results of this study, we carried out the
acylation reaction of various functional groups in the aromatic ring
of acyl chloride, including 4-methoxybenzoyl chloride (4), 4-
fluorobenzoyl chloride, 4-nitrobenzoyl chloride (8). The influence
of substituted aromatic chloride such as electron-donating (p-
OMe) and halo substituents (p-F) substituents on the aromatic ring
was investigated (Table 3). The synthesis of 2-(4-methoxyphenyl)
benzoxazole (5), 2-(4-fluorophenyl)benzoxazole (7), and 2-(4-
nitrophenyl)benzoxazole (Table 3, Entry 2–4) in the presence of
the VNU-11(Hf) catalyst were obtained in 97%, 92%, and 11% yields,
respectively. In the next series, these optimized conditions were
then applied to benzoxazole derivatives, and the reactions were
found to be successful for all benzoxazole derivatives. The use of
5-methylbenzoxazole (10) and 5-nitrobenzoxazole (14) proceeded
more smoothly with benzoyl chloride (2) giving 97 to 86% isolated
yields (Table 3, Entries 8–15). However, the reaction between ben-
zoxazole derivatives and several aromatic acyl chlorides showed
lower efficiency than the others with the expected products 5-m
ethyl-2-(4-methoxyphenyl)benzoxazole (12) (84% yield, Table 3,
Entry 6) and 5-methyl-2-(4-fluorophenyl)benzoxazole (13) (65%
yield, Table 3, Entry 7). Finally, the reaction of (14) with aromatic
acyl chloride led to 5-nitro-2-(4-fluorophenyl)benzoxazole (17)
with a moderate yield of 36% after 7 h (Table 3, Entry 10). This
result could be explained due to the weak nucleophilic character-
istics of aromatic acyl chloride.
Moreover, we studied the effect of heating methods such as
conventional heating and microwave irradiation on the reaction
between 5-nitrobenzoxazole (14) and various aromatic acyl chlo-
rides under VNU-11(Hf) catalyst. The reaction was carried out at
170 °C for 7 h under conventional heating or for 40 min under
microwave irradiation. The desired products were isolated by col-
umn chromatography and weighted to calculate the yields. As the
results shown in Fig. 2, microwave irradiation exhibited as a pow-
erful tool to enhance catalytic activity of VNU-11(Hf) in the reac-
tion with high yield and short reaction times. For example, the
reaction under microwave irradiation for 40 min could obtain 5-

L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
113
Table 3
Substrate scope of the VNU-11(Hf) catalyzed synthesis of various functionalization from the appropriate benzoxazoles and aromatic acyl chloride.
Entry
1
R1
H
R2
H
Time (h)
3
Product
Yield (%)
98
TOF (hÀ1
)
32.7
2
H
OMe
6
97
16.2
3
4
5
6
7
H
F
6
6
6
6
6
92
11
97
84
65
15.3
1.83
16.7
14.0
10.8
H
NO2
H
CH3
CH3
CH3
OMe
F
8
NO2
NO2
NO2
H
7
7
7
86
48
36
12.3
6.85
5.14
9
OMe
F
10
nitro-2-phenylbenzoxazole (15) or 5-nitro-2-(4-methoxyphenyl)b
enzoxazole (16), and 5-nitro-2-(4-fluorophenyl)benzoxazole (17),
whereas using conventional heating required 7 h for obtaining
the same yield of these products (Fig. 2). We calculated the maxi-
mum catalyst TOF (Turn-Over Frequency) for the acylation of ben-
zoxazole derivatives with different aromatic acyl chlorides. In the
conventional heating condition, the reaction between (1) and (2)
using VNU-11(Hf) catalyst, 1 mol% and 0.5 mol%, achieved the
maximum values of TOF as 32.7 and 35 hÀ1, respectively. Remark-
ably, TOF increases 8.5 and 10 times by using microwave irradia-
tion (Fig. 2) at 170 and 180 °C, respectively, for 40 min.
To investigate the recyclability of the catalysts, the reaction
between benzoxazole and benzoyl chloride with catalysis of
VNU-11(Hf) under microwave was chosen. After the reaction, the
catalyst was collected via centrifugation, washed with acetone
and ethanol, reactivated under reduced pressure, and analyzed
by X-Ray Diffraction, scanning electron microscopy, transmission
electron microscopy, and infrared spectroscopy (Section S7). PXRD
patterns corresponding to the fresh and used VNU-11(Hf) for five
times, which were observed with no considerable change in struc-
ture despite lower intensity (Fig. 3). The leaching test was also
determined by ICP-OES and the Hafnium content in ethyl acetate
was less than 5 ppb, indicating a little leaching of the metal from
the catalyst during the reaction. The recovered VNU-11(Hf) was
Fig. 2. Effect of conditional heating and microwave radiation method in yields
(column) and TOF (line) of (14) 5-nitrobenzoxazole (14) with aromatic acyl chloride
(2), or (4), and catalyzed by VNU-11(Hf).

114
L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
was released from benzoyl chloride, and hosted by the free formate
residue released earlier. Surprisingly, it did not combine with O,
but indeed formed a bond with C in the formate residue. The
r
O atom of such a formate residue, on the other hand, attempted
to establish bonding to C(2) of benzoyl chloride. Therefore, there
are three elementary steps occurring in the first primary process,
as shown in transition state 1 (TS1). In terms of free energy, TS1
is 22.22 kcal/mol (0.96 eV) above the initial state 1. Upon perform-
ing vibrational analysis for TS1, we observe one imaginary
wavenumber of À97 cmÀ1, which is associated with the bond
breaking process. After this process, we obtained complex 2, which
was 16.21 kcal/mol (0.70 eV) above 1 in terms of free energy. In
complex 2, the benzoyl residue was successfully attached to N with
electron transfer from N to benzoyl chloride as confirmed by Mul-
liken charge analysis. More importantly, the CAN bond in the five
member ring was reduced from
p to r. This is important for a com-
Fig. 3. PXRD analysis of VNU-11(Hf) before (blue) and after (red) the symthesis 2-
phenylbenzoxazole in comparison to the simulated pattern. Inset: The recycling
experiments of the reaction over five cycles (the numbers on the top represent the
yield of reaction. (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)
plete dissociation of CAN, so that C(2) from benzoyl could come in
to substitute the old C(1) site from the five member ring. At this
point, the C(1)AN single bond distance was 1.46 Å. To investigate
the C(1)AN bond dissociation, we performed partial optimization
for the cleavage of the bond. When reaching a distance of 2.79 Å,
the CAN bond underwent a transition state for dissociation (TS2),
and the reaction free energy barrier was 31.21 kcal/mol (1.35 eV).
Executing vibrational analysis for TS2, we obtained a negative
wavenumber of À85 cmÀ1. In fact, this free energy barrier is the
highest barrier in the entire reaction scheme, and plays the deci-
sive role in the chemical reaction. Fortunately, this reaction was
actually carried out under the activation of microwave; therefore,
that high reaction barrier could be overcome. One other reason
for the high reaction barrier is that, during this process, not only
the NAC(1) bond was broken. Indeed, we observed the formation
of chloro-formaldehyde. In a previous paper [17], the CAO bond
was weakened under the influence of OHÀ, and subsequently suf-
fered dissociation to achieve ring opening. Our reaction was carried
out under the acidic condition, and our DFT calculations verified
that it was not possible to break the CAO bond. Indeed, the rupture
of CAN bond was the most sensible strategy so that the reaction
would come to the C substitution in a later stage.
reused in the reaction of (1) and (2) for 5 times with slightly fall of
reactivity and insignificant decrease of yield by about 5%. This
should be able to imply the recyclability of VNU-11(Hf) as an
important feature for industrial application.
2.3. Mechanistic considerations
A study mechanism for the ring open path way acylation reac-
tion has been also investigated computationally using benzoxazole
and benzoyl chloride. Initially, we allowed benzoxazole to interact
with the Hf cluster without the presence of benzoyl chloride. It was
confirmed experimentally by 13C NMR that upon the use of ben-
zaldehyde, there must be a C substitution in the five-membered
ring, and it was not possible to undergo CAH activation [36]. More
specifically, there must be a ring opening at first, then C from the
carbonyl group in benzoyl chloride would come in and replace
the C in the five-membered ring. Therefore, while exploring the
reaction mechanism, we strictly followed the pathway involving
C substitution.
At this point, N already connected to benzoyl, while still
remained bonding with the original six-membered phenyl ring.
Moreover, there was a formate residue attached to the aromatic
phenyl ring. The subsequent process concerned the release of
chloro-formaldehyde. In complex 3, the
formaldehyde still linked with N through a weak bond with a dis-
tance of 1.50 Å. In this case, the CAO bond was not a bond, and O
Since N possessed one pair of electrons, it was highly attracted
by one Hf site. In the initial step, the HfAN bond could be estab-
lished easily. Since the reactant came to bind with Hf, the nearby
Hf-formate bond could be released with a low energy barrier of
0.36 eV (8.3 kcal/mol). Such dissociated formate residue will
involve in a later stage of the catalyzing process. According to
our optimization, it is interestingly observed that the resultant pro-
duct in this step was more energetically stable than the initial pro-
duct by an amount of 2.64 kcal/mol. After accomplishing those
benchmarking calculations, we introduced benzoyl chloride to
the system and started to explore the reaction mechanism. This
structure is denoted as complex 1 in our reaction mechanism in
Scheme 1. For the ease of discussion, we denote the C atom in
the five-membered ring of benzoxazole as C(1), while the C of ben-
zoyl chloride is denoted as C(2). Those two C atoms played a signif-
icant role in the overall reaction mechanism. The attachment of N
to Hf was important, because it would help to reduce the C(1)@N
bond. As a result, C(1) became more attractive and subsequently
established a bond to O from benzoyl chloride. On the other hand,
upon the consideration of O from benzyol chloride attaching to Hf,
it would be more difficult for this O atom to form a bond with C(1)
from the unattached benzoxazole. More importantly, the bonding
interaction between N and Hf helps to weaken the CAN bond, thus
leads to its rupture later on.
C
from chloro-
p
had to complete its electron configuration with a linkage with Hf.
To release chloro-formaldehyde, the NAC bond had to be broken,
and we found a transition barrier of 14.81 kcal/mol (0.64 eV) with
an imaginary vibrational wavenumber of À56 cmÀ1 (TS3). Chloro-
formaldehyde was formed in this stage (complex 4), and it is soon
broken into HCl and CO. The presence of HCl can be confirmed
experimentally by the decrease of pH (2.542 before reaction and
1.567 after reaction). For simplicity of calculations, we eliminate
chloro-formadehyde from the entire structure from this point.
After eliminating chloro-formaldehyde, we optimized complex 5,
as shown in Scheme 1. It should be noticed that after the departure
of chloro-formaldehyde, the structure was unstable due to the lack
of bonding on the C(1) site.
This problem was soon taken care of by an internal re-
arrangement, and an intermediate seven-membered ring was
established, as seen in complex 5. It should be noticed, however,
that C(1) connected to three surrounding O sites. This bonding
behavior was indicative of structural instability. Therefore, the
intermediate complex 5 would soon suffer from bond dissociation;
one of the three OAC bond was broken very easily (to TS4 as a mat-
ter of structural re-arrangement). The total energy barrier for this
Electron-rich C(2) attacked the N site of the initial reactant, and
attempted to establish a new
r bond. Almost at the same time, Cl

L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
115
Scheme 1. Ring opening acylation reaction between benzoxazole (1) and benzoyl chloride (2)-reaction path starting with capping VNU-11(Hf) step.
process was only 0.78 kcal/mol (0.04 eV). More surprisingly, the
free energy barrier was observed to be quite negative (À2.95 kcal/-
mol), while implied the rapid conversion at this step. Moreover, the
C(2)AO bond suffered cleavage at the same time; consequently, C1
was accompanied by two O- to form a formate residue and
released from the seven-membered ring during the process. After
this stage, complex 6 was produced with C(2) connected to N via
a triple bond (the bond length was 1.19 Å with a partial positive
charge on N), while the O atom from the initial reactant connected
to the phenyl ring and one Hf site. In the last process, such an O
atom and C(2) approached each other to make a new bond, and
established a new five-membered ring in the last process. To
achieve TS5, the structure overcame an energy barrier of
4.71 kcal/mol (0.20 eV). For TS5, an imaginary wavenumber of
À127 cmÀ1 was observed. In the final product (complex 7), the
C(2)AN bond distance changed to 1.31 Å, which was more or less
detailed preparation of all MOF catalysts was described in
Section S2.
3.2. General methods
Powder X-ray diffraction patterns were recorded using a D8
Advance diffractometer equipped with
a LYNXEYE detector
(Bragg-Brentano geometry, Cu K radiation k = 1.54056 Å. Fourier
a
transform infrared (FT-IR) spectra were measured on a Bruker
E400 FT-IR spectrometer using potassium bromide pellets. Solu-
tion NMR spectra were acquired on a Bruker Advance-500 MHz
NMR spectrometer. ICP-MS analyses were performed on a Perki-
nElmer NexION 350X. Microwave irradiation was performed on a
CEM Discover BenchMate apparatus, which is offered for micro-
wave synthesis with safe pressure regulation using a 10 mL pres-
surized glass tube with Teflon-coated septum and vertically
focused IR temperature sensor controlling reaction temperature.
descriptive of a
p bond. The final product was found to be
7.25 kcal/mol below TS5.
We then performed a new set of calculations for the reactants,
transition states, and final product with Zr instead of Hf. The
results showed that the new free energy barriers catalyzed by Zr
were higher than those catalyzed by Hf. Especially, to arrive at
TS1, the complex was predicted to overcome a free energy barrier
of 32.23 kcal/mol. The key step, which involves CAN rupture
3.3. Procedure of 2-arylbenzoxazoles derivatives
A mixture of VNU-11(Hf) (0.018 g, 0.01 mmol), aromatic acyl
chloride (0.140 g, 1 mmol) and benzoxazole (0.119 g, 1 mmol)
was heated at 170 °C under conventional heating for a few hours
or microwave irradiation of a CEM Discover apparatus for a few
minutes. After completion of reaction, the catalyst was filtered
from the reaction mixture. The filtrate was diluted with ethyl acet-
ate (50 mL), washed with H2O (3 Â 20 mL), aqueous NaHCO3
(2 Â 20 mL), and dried over Na2SO4. The solvent was removed in
a rotary evaporator. The crude product was purified by flash chro-
matography (90:10 acetone/petroleum ether to give corresponding
product. The purity and identity of the products were confirmed by
GC–MS spectra, which were compared with the spectra in the NIST
library, and by 1H and 13C NMR spectroscopy.
through TS2, occurred with
a remarkably high barrier of
41.65 kcal/mol. As illustrated in Fig. 4, the other reaction barriers
(shown in bracket) were also predicted to be higher than the pre-
vious case.
3. Experimental
3.1. Materials
All chemicals for synthesized MOF were obtained from Sigma-
Aldrich Chemical Company, Merck, and Alfa Aesar. All reagents
and solvents for catalytic studies were obtained from Acros Organ-
ics. Deuterated solvents, CDCl3, acetone d6 and DMSO d6, were pur-
chased from Cambridge Isotope Laboratories (Andover, MA). The
3.4. Theoretical calculations
In this section, density functional theory [37,38] (DFT) calcula-
tions were employed to explore the stepwise reaction mechanism.

116
L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
Fig. 4. Energy profile of ring open acylation between benzoxazole (1) and benzoyl chloride (2) following the reaction path starting with the capping metal-organic framework
step (Scheme 1) B3LYP energies calculated for MOF-808(Zr) and VNU-11(Hf) are reported.
The Perdew-Burke-Ernzerhof functional [39–41] was used in con-
junction with the split-valence 6–31 g basis set [42,43] for C, H,
O, N, Cl, while the effective core potential method [44] with the
LANL2dz basis set [45–47] was employed to describe the valence
wave functions of Hf. To account for long-range van der Waals
interaction, we used the well-established D3 empirical correction
terms developed by Grimme [48]. All calculations were executed
using the Gaussian 16 package. First, the core cluster consisting
of six Hf (or Zr) cations, which was the main recipe to constitute
VNU-11(Hf), was optimized. For computational feasibility, all link-
ers connected to the metal cluster were simplified by formate
(COO–) residues as shown in Fig. 4.
Minh City for instrument use. We thank Ha L. Nguyen at UC Berke-
ley and Hung. A. T. Hang at the University of Science for their valu-
able discussions.
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2019.04.023.
References
[1] S. Noël, S. Cadet, E. Gras, C. Hureau, Chem. Soc. Rev. 42 (2013) 7747–7762.
[2] K. Seth, S.K. Garg, R. Kumar, P. Purohit, S.V. Meena, R. Goyal, C.U. Banerjee, K.A.
Chakraborti, ACS Med. Chem. Lett. 5 (2014) 512–516.
[3] A.D. Rodríguez, C. Ramírez, I.I. Rodríguez, E. Gonazález, Org. Lett. 1 (1999) 527–
530.
4. Conclusions
[4] T. Tian, L.-H. Cao, J. Chin. Chem. Soc. 54 (2007) 1369–1373.
[5] X. Wen, J.E. Bakali, R. Deprez-Poulain, B. Deprez, Tetrahedron Lett. 53 (2012)
2440–2443.
[6] Y. Kawashita, N. Nakamichi, H. Kawabata, M. Hayashi, Org. Lett. 5 (2003)
3713–3715.
[7] D. Yang, P. Liu, N. Zhang, W. Wei, M. Yue, J. You, H. Wang, ChemCatChem 6
(2014) 3434–3439.
[8] B. Maleki, M. Baghayeri, S.M. Vahdat, A. Mohammadzadeh, S. Akhoondi, RSC
Adv. 5 (2015) 46545–46551.
[9] F. Zhu, Z.-X. Wang, Org. Lett. 17 (2015) 1601–1604.
[10] P. Feng, G. Ma, T. Zhang, C. Wang, Asian J. Org. Chem. 7 (2018) 788–792.
[11] F. Gao, B.-S. Kim, P.J. Walsh, Chem. Commun. 50 (2014) 10661–10664.
[12] J. Huang, J. Chan, Y. Chen, Y.C.J. Borths, K.D. Baucom, R.D. Larsen, M.M. Faul, J.
Am. Chem. Soc. 132 (2010) 3674–3675.
[13] X.-B. Shen, Y. Zhang, W.-X. Chen, Z.-K. Xiao, T.-T. Hu, L.-X. Shao, Org. Lett. 16
(2014) 1984–1987.
[14] N.-N. Jia, X.-C. Tian, X.-X. Qu, X.-X. Chen, Y.-N. Cao, Y.-X. Yao, F. Gao, X.-L. Zhou,
Sci. Rep. 7 (2017) 43758.
The heterogeneously catalytic capability of Zr- and Hf-MOFs in
the solvents-free acylation of benzoxazole derivative with different
aromatic acyl chloride was investigated under conventional heat-
ing and microwave irradiation. As a result, the MOF catalysis in
the reaction showed the good efficiency with high yield and ease
of recycling. Moreover, the catalytic activity of MOF-808(Zr) and
VNU-11(Hf) were significantly higher in comparison to UiO-type
MOFs, VNU-1(Zr), and VNU-2(Hf) catalysts. Experimental investi-
gation was accompanied by the DFT-modeling of the reaction
mechanism catalyzed by formate, Brønsted and Lewis acid sites.
The mechanism involves two important key steps: the broken of
a C@N double bond under the activation of the formate residue
and Hf site, and the combination of NAC (from benzoyl chloride).
In the final product, it was clearly observed that C from benzoyl
chloride successfully came in and made a replacement to establish
a new five-membered ring. Finally, the reaction was shown to
occur more easily under the influence of Hf in comparison with Zr.
[15] F. Zhu, J.-L. Tao, Z.-X. Wang, Org. Lett. 17 (2015) 4926–4929.
[16] S. Liu, R. Chen, X. Guo, H. Yang, G. Deng, C.-J. Li, Green Chem. 14 (2012) 1577–
1580.
[17] L. Wang, X. Ren, J. Yu, Y. Jiang, J. Cheng, J. Org. Chem. 78 (2013) 12076–12081.
[18] H. Furukawa, K.E. Cordova, M. O’Keeffe, O.M. Yaghi, Science 341 (2013)
1230444.
[19] C.-D. Wu, M. Zhao, Adv. Mater. 29 (2017), 1605446(0-21).
[20] J. Liu, L. Chen, H. Cui, J. Zhang, L. Zhang, C.-Y. Su, Chem. Soc. Rev. 43 (2014)
6011–6061.
Acknowledgements
[21] A. Dhakshinamoorthy, Z. Li, H. Garcia, Chem. Soc. Rev. 47 (2018) 8134–8172.
[22] Y. Bai, Y. Dou, L.-H. Xie, W. Rutledge, J.-R. Li, H.-C. Zhou, Chem. Soc. Rev. 45
(2016) 2327–2367.
[23] M. Rimoldi, A.J. Howarth, M.R. DeStefano, L. Lin, S. Goswami, P. Li, J.T. Hupp, O.-
K. Farha, ACS Cat. 7 (2017) 997–1014.
This work was funded by Vietnam National University, Ho Chi
Minh city (no. VNU-C2018-50-2 and TX-2019-50-01). We grate-
fully acknowledge the support of Center for Innovative Materials
& Architectures (INOMAR), Vietnam National University, Ho Chi

L.H.T. Nguyen et al. / Journal of Catalysis 374 (2019) 110–117
117
[24] Y. Liu, R.C. Klet, J.T. Hupp, O. Farha, Chem. Commun. 52 (2016) 7806–7809.
[25] Z. Hu, D. Zhao, CrystEngComm 19 (2017) 4066–4081.
[26] F. Vermoortele, B. Bueken, G. Le Bars, B. Van de Voorde, M. Vandichel, K.
Houthoofd, A. Vimont, M. Daturi, M. Waroquier, V. Van Speybroeck, C.
Kirschhock, D.E. De Vos, J. Am. Chem. Soc. 135 (2013) 11465–11468.
[27] R.C. Klet, Y. Liu, T.C. Wang, J.T. Hupp, O.K. Farha, J. Mat. Chem. 4 (2016) 1479–
1485.
[35] Z. Hu, Y. Peng, Y. Gao, Y. Qian, S. Ying, D. Yuan, S. Horike, N. Ogiwara, R.
Babarao, Y. Wang, N. Yan, D. Zhao, Chem. Mater. 28 (2016) 2659–2667.
[36] S. Liu, R. Chen, X.I. Guo, H. Yang, G. Deng, C.-J. Li, Green Chem. 14 (2012) 1577–
1580.
[37] P. Hohenberg, W. Kohn, Phys. Rev. 136 (1964) B864–B871.
[38] W. Kohn, L.J. Sham, Phys. Rev. 140 (1965) A1133–A1138.
[39] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865–3868.
[40] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 78 (1997) 1396.
[41] A.D. Becke, J. Chem. Phys. 140 (2014), 18A301(1–18).
[42] R. Ditchfield, W.J. Hehre, J.A. Pople, J. Chem. Phys. 54 (1971) 724–728.
[43] M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, M.S. Gordon, D.J. DeFrees, J.A.
Pople, J. Chem. Phys. 77 (1982) 3654–3665.
[44] A. Alkauskas, A. Baratoff, C. Bruder, J. Phys. Chem. A 108 (2004) 6863–6868.
[45] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270–283.
[46] W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284–298.
[47] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299–310.
[48] S. Grimme, J. Antony, S. Ehrlich, H.A. Krieg, J. Chem. Phys. 132 (2010), 154104
(1–19).
[28] J. Jiang, O.M. Yaghi, Brønsted acidity in metal-organic frameworks, Chem. Rev.
115 (2015) 6966–6997.
[29] M.H. Beyzavi, R.C. Klet, S. Tussupbayev, J. Borycz, N.A. Vermeulen, C.J. Cramer,
J.F. Stoddart, J.T. Hupp, O.K. Farha, J. Am. Chem. Soc. 136 (2014) 15861–15864.
[30] T.L.H. Doan, T.Q. Dao, H.N. Tran, T.N. Le, Dalton Trans. 45 (2016) 7875–7880.
[31] L.H.T. Nguyen, T.T. Nguyen, H.L. Nguyen, T.L.H. Doan, P.H. Tran, Catal. Sci.
Technol. 7 (2017) 4346–4350.
[32] A. Schaate, P. Roy, A. Godt, J. Lippke, F. Waltz, M. Wiebcke, P. Behrens, Eur. J.
Chem. 17 (2011) 6643–6651.
[33] T.L.H. Doan, H.L. Nguyen, H.Q. Pham, N.-N. Pham-Tran, T.N. Le, K.E. Cordova,
Asian J. Chem. 10 (2015) 2660–2668.
[34] J. Jiang, F. Gándara, Y.-B. Zhang, K. Na, O.M. Yaghi, W.G. Klemperer, J. Am.
Chem. Soc. 136 (2014) 12844–12847.