Catalytic Transfer Hydrogenation of Furfural to Furfuryl Alcohol under Mild Conditions over Zr-MOFs: Exploring the Role of Metal Node Coordination and Modification

Node Coordination and Modi ﬁ cation

Research Article

Catalytic Transfer Hydrogenation of Furfural to Furfuryl Alcohol

under Mild Conditions over Zr-MOFs: Exploring the Role of Metal

Anil H. Valekar, Minhui Lee, Ji Woong Yoon, Jaesung Kwak, Do-Young Hong, Kyung-Ryul Oh,

Ga-Young Cha, Young-Uk Kwon, Jaehoon Jung,* Jong-San Chang, and Young Kyu Hwang*

Cite This: ACS Catal. 2020, 10, 3720 −3732

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: The catalytic transfer hydrogenation (CTH) reaction is

considered as a potential route for upgrading bio-based carbonyls to

their corresponding alcohols. Herein, a series of Zr-based metal−-

organic frameworks (Zr-MOFs) containing various types of metal

node to ligand coordinations were synthesized and tested for CTH of

furfural (FUR) to furfuryl alcohol (FOL). It was found that metal

node coordination plays a more important role than porosity in Zr-

MOFs. MOF-808 (synthesized using a scaled-up approach to achieve a

higher batch yield), with the lowest metal node to ligand coordination

(

coordination number 6), was found to be the most active catalyst

among the various tested Zr-MOFs. Furthermore, M-MOF-808,

modiﬁed by simple methanol activation (M), outperformed the

pristine MOF-808 in CTH of FUR to FOL even at 30 °C in the

presence of 2-propanol (IPA) as the hydrogen source. The simple modiﬁcation of the metal node in the Zr-MOF changed the acid−

base properties of the MOF-808 surface through the development of coordinatively unsaturated sites (CUS), hydroxyl and methoxy

groups in the framework of the Zr-MOF, which probably help to facilitate the adsorption of FUR and IPA onto the metal node

surfaces of the catalyst. To evaluate the versatility of methanol activation in CTH, further substrates, including other types of

biomass and representative carbonyl compounds over M-MOF-808, were tested. To demonstrate heterogeneous catalysis, the

catalyst was recycled for ﬁve consecutive cycles, with little loss after the ﬁrst cycle, and was fully characterized to observe any changes

in its structure. Mechanistic insights were provided by isotopically labeled 2-propanol-d experiments, indicating FUR reduction

through transfer hydrogenation. Finally, the reaction mechanism for CTH of FUR to FOL was proposed in detail using density

functional theory (DFT) calculations over metal node modiﬁed model systems of a 6-connected Zr-MOF.

KEYWORDS: catalytic transfer hydrogenation, zirconium-based metal−organic frameworks, metal node coordination and modiﬁcation,

methanol activation, density functional theory

−12

. INTRODUCTION

years.

In view of this, several Lewis acidic zeolites (Ti,

Sn, Zr, and SnAl) have been reported for the process. For

example, Bui et al. successfully demonstrated that the

combination of Lewis and Brønsted acid sites in zeolites can

directly convert FUR into GVL with near-quantitative yield

through a cascade reaction involving sequential transfer

hydrogenation and ring opening steps. Some Fe-based

catalysts were also found to be capable of transfer hydro-

genation of FUR to FOL.

Catalytic transfer hydrogenation (CTH) is a potential

alternative approach for the hydrogenation of renewable

lignocellulosic biomass to value-added chemicals,

ing organic molecules as the hydrogen source. Several

precious-metal-based catalysts have been reported for the

CTH of furfural (FUR) to furfuryl alcohol (FOL), i.e., the

−3

employ-

−7

most value added chemical derived from FUR.

These

11,12

However, those catalysts could

catalyst systems nonetheless exhibit some drawbacks: for

instance, precious metals are expensive and have limited

availability. Furthermore, the reaction parameters need to be

optimized carefully in order to obtain the desired products

because both CO and CC groups in FUR can be

hydrogenated and thus produce numerous hydrogenated

Received: November 23, 2019

Revised: February 3, 2020

Published: February 3, 2020

products.

The CTH of FUR over non-precious-metal

catalysts has garnered signiﬁcant attention in recent

2020 American Chemical Society

720

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

not produce a quantitative yield of FOL because they suﬀer

development of Zr-based advanced porous materials as

catalysts for room-temperature biomass conversion reactions.

from deactivation due to the change in pore structure and

formation of oxide species that occur with successive cycles.

Numerous Zr-based catalysts have been explored for CTH

. EXPERIMENTAL SECTION

reactions using various biomass-derived carbonyl compounds,

3−23

2.1. Synthesis of MOF-808. Considering the catalytic

including FUR.

Li et al. prepared a series of Zr-based

application of MOF-808 in the production of biomass-derived

bulk chemicals, the synthesis procedure of MOF-808 was

scaled up. Consequently, the concentration of precursors

amorphous acid−base bifunctional nanohybrids which were

demonstrated to be active in CTH reactions of biomass-

16−18

derived and other carbonyl compounds.

Mesoporous Zr-

(

metal and ligand) in the reaction mixture was increased in

SBA-15 has also been tested for the Meerwein−Ponndorf−

Verley (MPV) reduction of FUR using 2-propanol (IPA) as a

sacriﬁcial alcohol, but a poor yield of FOL was obtained,

order to obtain a higher batch yield. The scaled-up synthesis of

MOF-808 was performed in a 700 mL Teﬂon-lined pressure

reactor. Equivalent amounts of ZrOCl ·8H O (13.5 g, 0.04

mol) and 1,3,5-benzenetricarboxylic acid (9.3 g, 0.04 mol)

were dissolved in a DMF/formic acid mixture (164 mL/160

mL). The reaction mixture was stirred at room temperature

before heating at 135 °C for 48 h and then cooled to room

temperature when the reaction was complete. The solid was

collected by ﬁltration and washed twice with DMF (1 L) to

remove unreacted ligand molecules from the pores of MOF-

primarily due to the etheriﬁcation reaction of FOL with IPA.

Zr-based metal−organic frameworks (Zr-MOFs) have also

been examined as MPV catalysts for the selective hydro-

genation of various saturated and unsaturated carbonyl

1−23

compounds.

We reported CTH of ethyl levulinate to

GVL over a series of functionalized and nonfunctionalized Zr-

MOFs using IPA as the hydrogen source. It was observed

that metal nodes in Zr-MOFs were the active sites for CTH

and that greater accessibility to the nodes by substrate

molecules led to enhanced catalytic activity; however, the

role of metal node to ligand coordination for greater

accessibility to the nodes was not fully considered. Moreover,

a strategy to enhance catalytic activity through metal node

modiﬁcation of Zr-MOFs has not yet been implemented for

the transfer hydrogenation of biomass carbonyls. Most

importantly, a variety of heterogeneous catalysts, such as

808. The resultant powder was washed twice with acetone (1

L) to remove DMF solvent from the pores. This puriﬁcation

procedure took 2 days to complete. Finally, the solid powder

was dried overnight at 100 °C in an oven. Synthesis procedures

of other Zr-MOFs are provided in the experimental section of

the Supporting Information .

2.2. Methanol Activation of MOF-808. A dried powder

of MOF-808 (1 g) was immersed into 100 mL of anhydrous

methanol in a 250 mL round-bottom ﬂask (RBF). The

reaction solution was reﬂuxed to the boiling point of methanol

for 8 h. The powder was ﬁltered and washed with 150 mL of

fresh methanol twice over a 24 h period at room temperature.

The obtained sample was dried in a vacuum oven at 80 °C for

12 h. Methanol-treated MOF-808 powder was labeled as M-

MOF-808.

2.3. Catalytic Test and Product Analysis. The catalytic

activity of Zr-MOFs in the transfer hydrogenation of furfural

was evaluated using various hydrogen donor alcohols. Prior to

use, Zr-MOFs were dried in an oven at 100 °C for 12 h. In a

typical procedure, a known amount of dried catalyst, furfural,

naphthalene as an internal standard, and proton donor/solvent

were placed into a two-necked RBF. The reaction mixture in

the RBF equipped with a rubber septum and reﬂux condenser

was heated at the desired temperature and for the required

time. The catalyst was separated by centrifugation, washed

thoroughly with hot methanol, and dried at 100 °C prior to the

next cycle. The freshly distilled furfural neutralized in Na CO

was used only for recycling tests to minimize the colored and

acidic impurities present in it. The ﬁltrate collected after the

reaction was subjected to quantitative analysis using gas

chromatography (GC, FID detector and HP-5 column), and

identiﬁcation of the products was made by GC-MS (Agilent

3,25,26

14,27,28

16,18,30

metal oxides,

metal hydroxides,

metal com-

5,17,29

plexes,

silica,

metal−organic nanohybrids,

supported metals and metal oxides,

mesoporous

9,31

11,12,32−34

,35,36

24,37,38

zeolites,

and MOFs,

have been reported for the

transfer hydrogenation of biomass-derived carbonyl com-

pounds using alcohol as a hydrogen donor, but none of

these heterogeneous catalysts have been shown to be active at

or near room temperature. In addition, a number of noble-

metal-based catalysts, which are active at room temperature for

the hydrogenation of biomass-derived carbonyls using H gas

9−44

as a hydrogen source, need pressurized reaction systems,

and these demand the use of sophisticated instruments that

may incur heavy infrastructural costs on a large scale.

Therefore, the CTH reaction of biomass-derived carbonyls at

room or near temperature remains a largely uninvestigated

topic in biomass conversion.

Here, we explored the role of metal node coordination and

modiﬁcation in Zr-MOFs for CTH of FUR to FOL. The

ﬁndings demonstrated that the catalytic activity is mainly

dependent on the metal node to ligand coordination of Zr-

MOFs and further elaborated how metal node modiﬁcation in

MOF-808 could lead to signiﬁcant improvements in catalytic

activity at nearly room temperature (30−40 °C). Owing to the

high catalytic activity, the eﬀects of alternative substrates were

extended to other biomass-derived products, as well as

representative carbonyl compounds. FUR reduction through

transfer hydrogenation was conﬁrmed using the deuterium

isotope of IPA in combination with GC-MS analysis. Catalyst

heterogeneity, leaching, and recycling experiments were also

performed. Computational density functional theory (DFT)

calculations were used extensively to study the CTH

mechanism of FUR to FOL in detail. We anticipate that the

excellent catalytic activity of MOF-808 after methanol

activation under mild reaction conditions will enable the

890 N GC and 5973 N MSD). Furfural conversion and the

yield of furfuryl alcohol were determined using a multiple-

point internal standard calibration method. Other carbonyl

substrates and their hydrogenated products were quantitatively

analyzed using a single-point internal standard calibration

method (GC, FID detector and DB-624 column).

3. RESULTS AND DISCUSSION

3.1. Role of Metal Node Coordination in Zr-MOFs for

CTH. The presence of 12-, 10-, 8-, and 6-connected metal

nodes [Zr (μ -O )(μ -OH) ] in Zr-MOFs has been previously

721

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Table 1. Textural Properties and Catalytic Activity of the Zr-MOFs for the CTH of FUR to FOL

TOF (h⁻¹

)

entry

Zr-MOF (CN)

S_B_E_T(m /g)

PD (Å)

AS (mmol/g)

BS (mmol/g)

conversn (%)

Y_F_O_L(%)

S_F_O_L(%)

none

<2.0

2.3

2.1

5.2

16.4

1.3

0.0

0.1

13.5

66.4

56.5

0.0

1.9

82.3

81.7

UiO-66 (12)

DUT-52 (12)

UiO-67 (12)

DUT-67 (8)

MOF-808 (6)

1210

1208

2518

1362

5, 6.1

7.1

12.7

10, 4.9

12.7, 15.9

0.35

0.65

0.60

0.74

0.85

0.05

0.06

0.04

0.10

0.15

0.26

0.0

0.02

2.0

1313

81.3

11.6

S_B_E_Tis the BET surface area. PD is the pore diameter. Acid sites (AS) determined by NH -TPD. Base sites (BS) determined by CO -TPD;

where Y_F_O_Lis the yield of furfuryl alcohol, S_F_O_Lis the selectivity of the furfuryl alcohol, CN is the coordination number of the metal node to the

linker, and turnover frequency (TOF) = (moles of FOL)/(moles of “Zr” from ICP analysis × time) at 2 h of reaction time. Reaction conditions:

FUR, 1 g (10.4 mmol); IPA, 25 g (416 mmol); cat., 0.1 g; temperature, 82 °C (reﬂux); time, 2 h.

reported in the literature. However, very few eﬀorts have

conversion with 81.7% selectivity for FOL was achieved, a

result 43 times higher than that with 12-connected UiO-67.

Furthermore, the turnover frequencies (TOFs) of all the Zr-

MOFs were presented for fair comparison among the catalysts,

on the basis of their Zr content, which was determined using

ICP analysis (Table 1 and Table S1 ). Once again, MOF-808

outperformed all other catalysts reported in Table 1, with the

been undertaken to examine the role played by the metal node

connectivity in Zr-MOFs in catalytic activity.

44−46

Therefore,

to clearly understand the role of porosity in comparison with

that of metal node coordination in CTH of FUR, we

synthesized a series of Zr-MOFs possessing diﬀerent porosities

and metal node connectivities and tested them for transfer

hydrogenation of FUR. The structures of the synthesized Zr-

MOFs were conﬁrmed using powder X-ray diﬀraction (PXRD)

and thermal gravimetric analysis (TGA), and their textural

−

highest TOF of 11.6 h . These results strongly support the

fact that metal node connectivity plays a more important role

than high porosity in Zr-MOFs for CTH of FUR. In 12-

connected Zr-MOFs, Zr sites are fully saturated by carboxylate

ligands and therefore are the least accessible to substrate

molecule adsorption. Decreasing the metal node connectivity

reduces crowding around the metal node and thus enhances

the accessibility of substrate molecules to Zr sites and results in

properties were determined using N -physisorption, as shown

in Figure S1 . PXRD (Figure S1A ) conﬁrmed the formation of

crystalline phases in all Zr-MOFs, which is in accordance with

6−49

the reported structure.

The crystallinity of MOF-808 is

lower than that of the previously reported original structure

due to the scaled-up synthesis approach, which aimed to

achieve a higher batch yield. TGA curves (Figure S1B) of all

synthesized Zr-MOFs represent their thermal stabilities and

demonstrate that Zr-MOF catalysts are suﬃciently stable to

carry out the CTH reaction at moderate reaction temperatures.

Furthermore, all Zr-MOFs are highly porous and have well-

23,37

a higher catalytic activity, consistent with previous reports.

Acid and base sites in several Zr-based catalysts were

reported as active sites, considerably aﬀecting the catalyst

13−18

performance in transfer hydrogenation reactions.

Con-

sequently, NH -TPD and CO -TPD analyses were performed

in order to determine the acid−base properties of the catalyst,

respectively, and the results are shown in Table 1 and Figure

S2 . The reason for the selection of a lower desorption

temperature was the stability of the Zr-MOFs. Although all of

the Zr-MOFs analyzed here were quite stable up to 350 °C,

deﬁned pore diameters, as conﬁrmed by N -physisorption,

shown in Figure S1C,D.

The textural properties and catalytic activities of various Zr-

MOFs used for transfer hydrogenation of FUR are summarized

in Table 1. When CTH of FUR was carried out without a

catalyst at 82 °C (Table 1, entry 1), no FOL was formed, even

after a 2 h reaction time; moreover, less than 2% FUR

conversion was observed. Three diﬀerent Zr-MOFs (UiO-66,

UiO-67, and DUT-52) with the same ligand to metal node

coordination (12) but diﬀerent porous properties (such as

surface area and pore diameter) were subjected to the CTH

reaction (Table 1, entries 2−4). Despite UiO-67 having a

higher surface area and larger pore diameter in comparison

with UiO-66 and DUT-52, the conversion was increased by

only 3%, indicating that improved pore size and surface area

played no signiﬁcant role in enhancing the catalytic activity of

Zr-MOFs. This further implies that 12-connected (12

carboxylate ligands connected to metal site) Zr-MOFs are

less active in the CTH reaction at 82 °C. When 8-connected

DUT-67 was used, FUR conversion increased marginally in

comparison with UiO-67 (Table 1, entry 5), despite the fact

that the latter possesses a higher surface area and pore

diameter than the former. This demonstrates that the catalytic

activity associated with reducing metal node connectivity in

Zr-MOFs (from 12 to 8) involved in the CTH reaction can be

enhanced. Consequently, when 6-connected MOF-808 was

applied to the CTH reaction of FUR, a substantial enhance-

ment in the catalytic activity was reported: 81.3% FUR

triply bridged μ -OH and formate groups from the modulator,

which were linked to the Zr node for charge balance,

disconnected on heating beyond 200 °C, as evidenced from

the TGA patterns (Figure S1b). The origin of the acid−base

sites in Zr-MOFs could be surface hydroxyl groups (Zr−OH,

base sites), μ -OH (acid sites), Zr−OH (acid sites),

framework oxygen Zr−O−C or Zr−O−Zr (base sites), and

coordinatively unsaturated zirconium sites (Zr-CUS; acid

sites) of the metal nodes. The amount and desorption

temperature of NH illustrate the enhancement in concen-

tration and strength of the acid sites in the Zr-MOFs with

decreasing cluster coordination, and this is presumably due to

the diﬀerence in electron densities around the cluster, which

can alter the charge on the zirconium nodes. The NH -TPD

proﬁle of DUT-67 showed two desorption peaks, likely due to

the interaction of ammonia molecules with the metal node and

the sulfur-containing thiophene ligand. All other Zr-MOFs did

not possess any heteroatoms in their ligand ring structure and

showed only one broad desorption peak, a result of the

interaction of NH with the metal node. Both acid and base

sites in the Zr-MOFs increased with decreasing cluster

coordination. The MOF-808 with the highest acid and base

sites, 0.85 and 0.15 mmol/g, respectively, emerged as the most

active catalyst among the tested Zr-MOFs, as shown in Table

722

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Scheme 1. CTH of FUR to FOL over Metal Node Modiﬁed MOF-808

. Therefore, it can be concluded that an enhancement in

decline in surface area and pore size of MOF-808 following

methanol activation was noted from N₂ adsorption −

desorption isotherms recorded at −196 °C (Figure S3C,D).

These observations suggested that the local structure of MOF-

808 had been altered, but without signiﬁcant eﬀects on its

crystalline and porous properties. The elemental composition

and porous properties of pristine MOF-808 and M-MOF-808

are speciﬁed in Table S2.

substrate accessibility and in the acid−base properties of Zr-

MOFs by decreasing cluster coordination are the two main

important factors that signiﬁcantly enhance catalytic perform-

ance. In all cases (entries 1−6 in Table 1), the major side

product acetal, followed by two aldol condensation products,

has been observed in GC-MS analyses. Furthermore,

undetectable and undissolved polymeric products (humin)

could be generated due to the highly reactive aldehyde group

To further explore changes in the local structure of MOF-

808 after methanol activation, in situ FTIR analysis was

performed (Figure 1 and Figures S5 −S7 and Table S3). Two

of FUR, which is susceptible to condensation reactions.

.2. Metal Node Modiﬁcation of MOF-808. The

−

uniqueness of the MOF structure lies in that it can be tuned

more precisely than metal oxides and zeolites to create highly

uniform and controllable reactive sites for speciﬁc catalytic

applications. Consequently, owing to the high catalytic activity

of MOF-808 using metal nodes as active sites, we analyzed the

catalytic activity of metal nodes in modiﬁed MOF-808 during

the transfer hydrogenation of FUR to FOL, as shown in

Scheme 1. As-synthesized MOF-808 was activated simply

using methanol at its boiling point. Hereafter, these activated

MOF-808 samples will be designated as M-MOF-808. M-

MOF-808 was characterized using PXRD, TGA, and N₂-

physisorption (Figure S3 ) to understand the changes that

occur in its structural, thermal, and porous properties,

respectively. No loss of crystallinity was observed in PXRD

data of MOF-808 after methanol activation (Figure S3A ).

TGA of pristine MOF-808 (Figure S3B ) shows three weight

loss regions: the ﬁrst around 100 °C corresponding to

physisorbed water, the second at 225 °C−350 °C belonging

characteristic bands at 2931 and 2866 cm assigned to the C−

51,52

H stretching of formate groups linked to Zr (Zr−OOCH)

present in the pristine MOF-808 (Figure 1A) were replaced by

−

two C−H stretching vibration modes at 2929 and 2827 cm

51−53

assigned to methoxy groups linked to Zr (Zr−OCH ).

conﬁrm the grafting of methoxy groups on Zr sites, isotopic

exchange experiments were carried out. MOF-808 was

activated using deuterated methanol (CD OD); the activated

sample was designated as d-M-MOF-808. In the FTIR spectra

of d-M-MOF-808 (Figure 1A), C−H stretching bands from

Zr−OOCH were completely absent, and two new bands at

−

2203 and 2060 cm were detected, resulting from the C−D

stretching vibration of deuterated methoxy groups linked to Zr

1,53,54

−1

(Zr−OCD ).

The peak at 3086 cm assigned to the

C−H stretching of the aromatic group (BTC) was unchanged

upon methanol activation. Additionally, a new C−O vibration

−

peak was detected in methanol-activated samples at 1155 cm ,

further conﬁrming the formation of Zr−O−C bonds in

activated MOF-808 from the grafting of alcohol moieties to

the metal node (Figure S5A ). Note that the methanol

activation process was not carried out under an inert

atmosphere and MOF-808 was exposed to atmospheric

water before and after the activation (reﬂux) step. Therefore,

the replacement of some of the methoxy groups by OH groups

may explain the presence of methoxy as well as terminal OH

groups observed in the FTIR analysis. This kind of

phenomenon has previously been reported by Yang et al. for

UiO-66 and NU-1000, where methoxy groups generated by

methanol activation were replaced by terminal OH groups

to remnant formate, μ -OH, and trace quantities of

physisorbed dimethylformamide (DMF), and the last at

00−550 °C resulting from the degradation of the framework

benzenetricarboxylate (BTC) ligand. After activation, the

second weight loss region vanished completely, and the sample

was degraded at almost the same temperature as for pristine

MOF-808. The absence of a second weight loss region (225

C−350 °C) indicated the removal of coordinated formate, μ -

OH, and physisorbed DMF; this point was corroborated by H

NMR spectra of MOF-808 and M-MOF-808 digested in HF/

DMSO-d (Figure S4 ). The obtained H NMR data showed

that formate groups and trace amounts of DMF in MOF-808

were almost completely exchanged by methanol. The slight

−1

after water vapor exposure at 393 K. The peak at 3674 cm

53,55,56

in MOF-808 was assigned to bridging μ -OH groups;

723

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Figure 2. FTIR spectra of (A) MOF-808 and (B) M-MOF-808

recorded at −173 °C during CO adsorption.

Figure 1. In situ FTIR recorded at room temperature after evacuation

−

for 2 h at 100 °C (A) in the 2000−3250 cm region and (B) in the

species in pristine MOF-808. At higher CO coverage, three

ν(CO) bands located at 2177, 2153, and 2136 cm appeared

simultaneously (Figure 2A). The presence of H-bonded and

−

300−3900 cm region characterizing (a) MOF-808, (b) M-MOF-

−1

808, and (c) d-M-MOF-808.

physisorbed CO can be conﬁrmed using the peaks at 2153 and

2136 cm , respectively.

−

this disappeared after methanol activation, and four new peaks

were generated in the same region (Figure 1B). The peaks at

We further quantiﬁed the amount of Zr-CUS present in M-

MOF-808 using a plot of the peak area of 2188 cm versus the

671 and 3765 cm⁻were assigned to the μ -OH stretch and

−1

non-hydrogen-bonded terminal OH groups at newly formed

amount of CO adsorbed (Figure S8 ). The chemisorption of

CO reached up to 0.48 mmol/g, conﬁrmed using the linear

correlation between peak area and amount of CO adsorption.

The linearity is tented after 0.48 mmol/g of CO adsorption

due to physisorbed CO molecules. This means that around 0.5

mmol/g of Zr-CUS was generated in MOF-808 after methanol

activation. The quantiﬁed value of Zr-CUS may diﬀer when

MOF-808 is synthesized using diﬀerent reaction conditions.

3.3. CTH of FUR to FOL over Metal Node Modiﬁed

MOF-808. Methanol activation of MOF-808 changes its local

structure through modiﬁcation of the metal node. Con-

sequently, its chemical properties change, and it becomes

highly desirable to correlate the local structure modiﬁcation

with its catalytic activity. Accordingly, the M-MOF-808

catalyst was tested for CTH of FUR to FOL, and the results

obtained are summarized in Table 2. When the reaction was

carried out under reﬂux conditions at 82 °C, M-MOF-808

showed higher FUR conversion and FOL selectivity in

comparison with as-synthesized MOF-808 (entries 1 and 2,

Table 2). Nevertheless, the diﬀerence was not signiﬁcant,

probably due to reaching the equilibrium point. As previously

reported, CTH reaction following the principle of the

Meerwein−Ponndorf−Verley (MPV) reduction is an equili-

brium reaction in which the reverse reaction is termed

54,57

surfaces, respectively.

The remaining two peaks, those at

−

−1

725 and 3639 cm , together with a peak at 5266 cm , were

not present after heating the sample at >150 °C (Figures S5 −

S7 ). Therefore, we assigned these peaks to water molecules

coordinated to Zr sites.

We also performed CO-FTIR analysis of pristine and

activated MOF-808, as shown in Figure 2, to determine the

coordinatively unsaturated Zr sites (Zr-CUS). FTIR experi-

ments were conducted by introducing calibrated doses of CO

after activating the samples at 150 °C for 6 h. In the case of M-

−

MOF-808 (Figure 2B), a ν(CO) band centered at 2188 cm

appeared and progressively broadened with increasing CO

−

dosage, shifting to a lower wavenumber of 2180 cm , which is

assigned to CO coordinated to the Lewis acid site (Zr-

9,60

CUS).

The red shift in peak position reﬂects the

possibilities of formation of multicarbonyl species on Zr-

CUS. At higher CO coverage and only after the saturation of

Zr-CUS, two additional ν(CO) bands were developed at 2151

−1

and 2135 cm . We assigned these to H-bonded CO molecules

with μ -OH and physisorbed CO molecules, respectively. In

pristine MOF-808, virtually no peaks were detected. An

extremely low intensity peak could be visible upon magnifying

the initial spectrum more than 20 times (inset of Figure 2A),

−

situated at 2183 cm . This peak was detected at initial CO

coverage and indicated a negligible amount of vacant Zr

Oppenauer oxidation. To clearly diﬀerentiate the catalytic

activity of these two catalysts, we performed the reaction at 40

724

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Table 2. Comparison of CTH of FUR over MOF-808, M-MOF-808, and Other Zirconium-Based Catalysts

entry

Cat.

MOF-808

M-MOF-808

T (°C)

time (h)

mass ratio (Cat./FUR/IPA)

conversn (%)

Y_F_O_L(%)

S_F_O_L(%)

TOF (h⁻¹)

0.1/1/25

0.1/0.24/7.9

0.1/0.1/7.9

0.1/0.19/7.9

0.1/0.5/12.5

81.3

89.3

2.7

8.9

16.4

37.9

99.2

97.5

66.4

79.1

0.9

0.2

7.1

13.5

34.1

93.0

81.7

88.6

11.6

15.0

0.2

0.01

0.6

2.0

5.4

17.6

3.1

4.4

Zr O (OH) (Mc)

ZrO₂

ZrO(OH)₂

DUT-67

M-DUT-67

M-MOF-808

ZrPn

7.4

79.8

82.3

89.9

93.8

96.8

98.5

94.0

92.3

88.6

100

140

Zr-LS

Zr-FDCA

MOF-808

M-MOF-808

2.7

0.2

2.9

27.5

96.5

25.4

85.5

Turnover frequency (TOF) = (mole of FOL)/(mole of “Zr” from ICP analysis × time) at 2 h reaction time. A batch type pressure reactor was

used. From ref 16 . From ref 63 . From ref 17 .

Figure 3. (A) Hot ﬁltration test: FUR, 1.0 g (10.4 mmol); IPA, 25 g (416 mmol); cat., 0.1 g; temperature, 82 °C (reﬂux). (B) Catalyst recyclability

test: FUR, 0.25 g; IPA, 25 g (416 mmol); cat., 0.1 g; temperature, 82 °C (reﬂux); time, 1 h.

and 30 °C. Another objective was to conﬁrm the eﬃciency of

the catalysts at low temperature, since this is the favored route

for green and sustainable development of biomass-derived

chemicals. Moreover, a large amount of low-grade waste heat

metal node and resulting in higher FUR conversion and FOL

yield.

Considering the high catalytic activity of M-MOF-808 in

CTH of FUR, we compared its activity with that of other Zr-

based catalysts (entries 3−7 and 9−11, Table 2). Using the

fully coordinated methyl acrylate (Mc) Zr cluster

Zr O (OH) (Mc) as a homogeneous catalyst enabled only

(

speciﬁcally below 40 °C) is discharged into the environment

during many industrial processes and cannot generally be

61,62

utilized due to its low-quality energy.

Utilizing this low-

3% conversion, similar to the low activity of 12-connected Zr-

grade waste heat for another application (e.g., in the

production of biomass-derived chemicals) would assist

MOFs. ZrO also showed minor conversion (2.7%), with only

61,62

0.2% FOL yield. ZrO(OH) showed better performance than

industries in improving their thermal energy eﬃciency,

ZrO₂with 8.9% conversion and, importantly, good FOL

selectivity. It has already been demonstrated that hydrous

zirconia can strongly activate an alcohol as a proton donor

through the hydroxyl groups present at the surface of the

which the catalyst must be active at low temperatures. One

could clearly observe the diﬀerence in catalytic activities of

MOF-808 and M-MOF-808 at 40 and 30 °C (entries 12 and

3 in Table 2 and Figure S9 ). After 24 h of reaction time,

MOF-808 converted only 27.5% of FUR, whereas M-MOF-

08 attained 96.5% FUR conversion at 40 °C. At 30 °C, FUR

catalyst and subsequently facilitate hydride ion transfer.

Conversely, fewer hydroxyl groups on calcined ZrO are

responsible for lower catalytic activity. M-MOF-808 out-

conversion was 6.7% and 76.1% with 6.2% and 68.5% FOL

yield for MOF-808 and M-MOF-808, respectively. The

remarkable enhancement in catalytic activity of M-MOF-808

can be attributed to its modiﬁed local structure. The proton

topology is changed in MOF-808 by replacing formate groups

with terminal OH and methoxy groups, and the formation of

Zr-CUS enhances the density of the acid and base sites (Figure

S10 ), likely facilitating the adsorption of FUR and IPA on its

performed all other catalysts given in Table 2 in terms of TOF.

It reaches almost quantitative conversion (99.3%) with 93.3%

FOL yield by increasing the temperature from 82 to 100 °C

(

entry 8). Other catalysts, such as zirconium phosphonate,

lignosulfonate, and carboxylate complexes, were also active for

this reaction and also showed a quantitative yield of FOL

(entries 9−11); however, these catalysts required higher

temperatures and larger amounts of catalyst in order to reach

725

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Table 3. Substrates Tested for CTH of Various Carbonyl Compounds over M-MOF-808

Reaction conditions: substrate, 2.6 mmol; IPA, 416 mmol; cat., 0.1 g (10.6 mol %); temperature, 82 °C (reﬂux).

the optimum yield of FOL. Furthermore, to generalize the

concept of enhancement in the catalytic activity of Zr-MOFs in

the CTH reaction by modiﬁcation of its metal node, we

selected another undercoordinated Zr-MOF: DUT-67 (8-

connected). Upon methanol activation, DUT-67 undergoes a

metal node modiﬁcation similar to that for MOF-808 (Figure

S11 ), and the catalytic activity of M-DUT-67 is also enhanced

considerably (Table 2 entries 6 and 7). Reaction parameters,

such as the proton donor/solvent, reaction time, and

concentration of IPA and FUR, were optimized in order to

reach the higher yield and selectivity of FOL over M-MOF-808

conversion of FUR to FOL, whereas almost no reaction was

observed to occur for the reaction system in which the catalyst

was ﬁltered out after 1 h. Inductively coupled plasma (ICP)

analysis showed that both ﬁltrates contained no “Zr” species,

demonstrating the heterogeneous character of M-MOF-808

catalyst in the CTH of FUR. The recyclability of the M-MOF-

808 catalyst was also studied under optimized reaction

conditions, however, at lower conversion in order to clearly

observe the changes in catalytic activity (Figure 3B). A small

decrease in catalytic activity was observed in the second cycle,

and the conversion dropped from around 75% in ﬁrst cycle to

64% in the second cycle. In subsequent catalytic cycles, the

conversion remained stable at around 64%. FOL selectivity was

maintained at over 97% in all cycles. This indicates that some

of the active sites of the catalyst might have been blocked after

the ﬁrst cycle due to minor amounts of substrate or produced

molecules having strong interactions with the catalyst surface.

To investigate the structural changes in M-MOF-808, PXRD,

TGA, in situ FTIR, SEM, and TEM were used and are

discussed in detail in the results and discussion part of the

Supporting Information (Figures S15 and S16).

3.5. Substrate Scope. Having demonstrated the excellent

catalytic performance of M-MOF-808 in CTH of FUR, the

scope of multiple substrates was tested using other biomass-

derived carbonyl compounds, as well as representative

aldehydes and ketones (Table 3). All examined carbonyl

compounds can be eﬃciently catalyzed by M-MOF-808 to

produce their corresponding alcohols in high yields. Notably, a

longer reaction time is required for CTH of ketones (entries 4

and 7). The M-MOF-808 catalyst can promote the transfer

hydrogenation of the important biomass-derived platform

chemical 5-hydroxymethylfurfural (HMF) into 2,5-bis-

(Figure S12 ). Using IPA as the hydrogen donor allowed us to

reach the maximum yield of FOL (94.1% with 99.1%

selectivity in 2 h of reaction time under reﬂux conditions).

Additionally, considering the good performance of this catalyst

near room temperature (Table 2 and Figure S9), the reaction

conditions were further optimized by changing the IPA and

initial FUR concentrations at 30 °C in order to achieve a

quantitative yield and high selectivity of the FOL. The

maximum FOL yield reached 90% with 94.2% selectivity

after 24 h of reaction time (Figure S13 ). A detailed explanation

regarding reaction optimization is provided in the results and

discussion (Figures S12 and S13 ) of the Supporting

Information . Major byproducts of the reaction were analyzed

using GC-MS (Figure S14 ), from which a plausible reaction

network was proposed and described (Scheme S1 ).

.4. Reusability of the Catalyst. To demonstrate the

heterogeneity of the catalyst, two parallel experiments were

performed using M-MOF-808 catalyst with or without

separating the catalyst from the hot reaction mixture after 1

h (Figure 3A). After stirring for 4 h under identical conditions,

M-MOF-808 had continued to smoothly catalyze the

726

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Figure 4. (A) Optimized geometries for model systems describing singly (top) and doubly (bottom) modiﬁed Zr nodes in MOF-808. A hydroxyl

group and water molecule are used for the modiﬁcation of the Zr node while other Zr centers are still blocked with formate groups to reduce the

complexity of DFT computations. The reaction site is illustrated using a ball-and-stick model (Zr, blue; O, red; C, dark gray; H, white), and the

other parts are simply presented with a stick model for clarity. (B) Proposed CTH reaction path.

(

hydroxymethyl)furan (BHMF) with a high selectivity of

8.0% (Table 3, entry 2). When a methyl group is introduced

sequential metal hydride formation and subsequent carbonyl

16,37

group reduction occurred.

We applied the same method

in FUR at position 5 (5-MF), a longer reaction time (6 h) is

essential to achieve a similar conversion to FUR (95%). This is

due to the electron-donating nature of methyl groups

increasing the electron density on the C atom of the carbonyl

group. Consequently, the situation did not favor hydride ion

attack on carbonyl carbon. Moreover, steric hindrance can also

contribute to the slowing of the reaction rate. In contrast, an

electron-withdrawing group attached to the aromatic ring at

the para position improved the rate of the CTH reaction

with M-MOF-808 to understand the mode of the reaction

mentioned above. Here, 2-propanol-d₈was used with 2-

methyl-2-propanol in a molar ratio of 1:3. The absence of β-H

in 2-methyl-2-propanol prevents it from playing the role of a

hydrogen donor. However, it is capable of exchanging H atoms

from its OH groups to the OD groups of 2-propanol-d , as well

as with the active H/D from the catalyst surface. Further, due

to the excess of 2-methyl-2-propanol used, it can replace

almost all of the OD of 2-propanol-d into OH, leaving 2-

(entries 5 and 6, Table 3). To illustrate the competency of M-

propanol-d as the major D donor to FUR (Scheme S2). FOL

MOF-808 in transfer hydrogenation reactions near room

temperature, a substrate scope study was undertaken using the

same carbonyls utilized in Table 3 at 30 °C (Table S4). All of

the carbonyl compounds were successfully converted into their

respective alcohols with high selectivity, though lower substrate

to catalyst ratios and longer reaction times were required

products formed by the hydrogenation of FUR with 2-

propanol-d and 2-propanol-d were analyzed by GC-MS, and

the diﬀerence of 1 amu (from 98 to 99) in their molecular ion

peaks (Figure S17 ) suggests that hydrogen from IPA was

directly transferred to the carbonyl group of FUR without

being adsorbed on the catalyst surface. Mechanistic insights

therefore suggest that basic sites of the catalyst (Zr−OH/Zr−

(

Table S4 ). These results illustrate the great potential

applications of M-MOF-808 for CTH reactions in both

organic synthesis and biomass transformation at low reaction

temperatures.

OCH ) absorb the hydrogen atoms from IPA, which assist the

formation of 2-propoxide on Zr sites and that subsequently

facilitates the transformation of β-H from alcohol via an eight-

membered transition state.

.6. Mechanistic Insights Using Isotopically Labeled

-Propanol. Isotopically labeled 2-propanol-d₈GC-MS

3.7. Reaction Mechanism Based on Computational

Study. The metal node modiﬁcation of Zr-based MOF

accomplished by the removal of formate groups is an essential

procedure in achieving the high catalytic activity for CTH of

FUR to FOL. The poor reactivity of 12-connected Zr-MOFs

methods have been used in previous studies with diﬀerent

heterogeneous catalyst systems to determine whether the

reaction proceeds through direct hydrogen transfer from 2-

propanol to the carbonyl groups of the substrate or whether

727

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

Figure 5. Reaction energy diagrams for CTH paths of FUR to FOL (see Figure 4B) with singly (blue) and doubly (black) modiﬁed Zr node model

systems. The Gibbs free energy (ΔG) throughout the reaction processes is presented in kcal/mol.

can thus be easily understood through modiﬁcation available

only at the defective sites of the Zr node due to their lack of

formate groups. Nevertheless, the dependence of CTH on the

coordination and modiﬁcation of Zr nodes and its detailed

reaction mechanism has not yet been described for Zr-based

MOFs. To assess this, we carried out a computational study

based on DFT (see the Supporting Information for

computation details). On the basis of the three-dimensional

network of MOF-808, the feasible geometries of Zr nodes due

to modiﬁcation, i.e., singly (top) and doubly (bottom)

modiﬁed Zr nodes (s and d Zr nodes) of MOF-808, are

depicted in Figure 4, where modiﬁed Zr centers are occupied

with one −OH group and one water molecule. In the proposed

model systems, the −OMe group introduced by methanol

activation (see Scheme 1) was substituted with an −OH group

in the modiﬁed Zr center, whereas the other Zr centers

remained blocked with formate groups, in order to reduce the

complexity of DFT computations. Note that replacing the

node model systems and indicate that our approach does not

signiﬁcantly inﬂuence the reaction energy proﬁle. The catalytic

cycle after binding the ﬁrst IPA molecule to a Zr node, i.e., the

formation of A, is schematically presented in Figure 4B, where

colored hydrogens indicate transferred hydrogen species

during the reaction. The ﬁrst and second IPA molecules

provide hydrogens as a hydride (H1, B → C) and as a proton

(H2, D → E) to FUR. The ﬁrst hydrogen transfers from IPA to

FUR (B → C) occurs with an activation energy (E ) of 11.83

and 11.47 kcal/mol for the s and d Zr nodes, respectively,

following the replacement of the remaining water molecules

with FUR (A → B). The hydride transfer from alkoxy carbon

to carbonyl carbon is the rate-determining step, and the

corresponding E values are almost identical for both the s and

d Zr nodes (Figure 5), where the transition state structure can

be deﬁned with an atomic conﬁ guration of an eight-membered

ring involving two Zr centers (Figure S18). The CTH reaction

of aldehyde to alcohol has usually been thought to follow the

Meerwein−Ponndorf−Verley (MPV) mechanism, which re-

−

OMe groups with −OH groups might occur due to

experimental conditions as reported earlier in this paper

quires a six-membered-ring transition state for hydride

21−23

(section 2.2) and in a previous study. The computationally

transfer.

The binding of FUR and IPA molecules to the

suggested CTH mechanism, where two IPA molecules are

sequentially required as a hydrogen donor, is presented in

Figure 4B, and the corresponding reaction energy proﬁles

same Zr metal center is required for the MPV reaction process.

Neither the local minimum nor the transition state structures

corresponding to the traditional MPV mechanism on the s Zr

node system were found because of the signiﬁcant steric

hindrance of the neighboring formate groups. Conversely, the

d Zr node systems may lead to the binding of FUR and IPA to

the same Zr metal center, which has no bond with formate

groups (see Figures S20 and S21 for the reaction energy

diagram and optimized geometries, respectively, for the MPV

process in the d Zr node). However, the MPV process in d Zr

nodes requires a much higher energy for hydrogen transfer (B

→ C) in comparison with the proposed mechanism in the

presented computational results, which has the eight-

membered-ring transition state involving the closest two Zr

centers (Figure 5). A similar reaction mechanism deviated

from the MPV process and was also reported in a previous

study regarding the CTH reaction of cyclohexanone with IPA

(

ΔG) obtained with DFT calculations are provided in Figure 5

see Figure S18 for optimized geometries). The initiation step

of CTH, i.e., the binding process of the ﬁrst IPA molecule onto

a surface-modiﬁed Zr node, proceeds through the substitution

of an −OH group with a deprotonated IPA (S → A). Whereas

the reaction energy for S → A is estimated to be endothermic

by +8.49 kcal/mol for the s Zr node of MOF-808 (blue), the

corresponding reaction energy for the d Zr node (black) is

reduced to −11.26 kcal/mol, as shown in Figure 5. Therefore,

the d Zr node can easily initiate the CTH reaction with a

highly exothermic binding process involving the ﬁrst IPA

molecule in comparison with the s Zr node, for which the

initiation step is thermodynamically unfavorable with endo-

thermic energy. In addition, the corresponding reaction

energies for the s and d Zr nodes are +8.46 and −9.63 kcal/

mol, respectively, when the −OMe group is not substituted

with an −OH group for simplicity (Figure S19). These values

are comparable with those obtained using the simpliﬁed Zr

on a zirconia surface.

Thereafter, the acetone molecule should be substituted with

a second IPA as a hydrogen donor (C → D). This process is

expected to be facilitated due to hydrogen bonding between a

728

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

https://pubs.acs.org/doi/10.1021/acscatal.9b05085.

Research Article

newly inserted IPA and the neighboring oxygen of deproto-

nated FOL formed by hydride transfer to FUR. The

exothermic reaction energy of −11.50 kcal/mol for the s Zr

node is twice as large as that for the d Zr node (−5.08 kcal/

mol), which is related to the strength of the hydrogen bond:

the hydrogen bond lengths for s and d Zr nodes are 1.44 and

The substrate scope was extended to a number of biomass-

derived and other representative carbonyl compounds over M-

MOF-808. Mechanistic insights were provided by isotopically

labeled 2-propanol-d experiments, showing FUR reduction

through the transfer hydrogenation path. Finally, the reaction

mechanism for CTH of FUR to FOL was investigated in detail

using DFT computations, which suggests that a highly

modiﬁed Zr node feasible in a Zr-MOF with lower

connectivity can provide higher catalytic activity. The work

presented here should encourage the development of new Zr-

based advanced porous materials as catalysts to tackle the

challenges associated with room- or low-temperature biomass

conversion reactions.

.54 Å, respectively (Figure S18). The neighboring functional

groups formed with activation processes, i.e. the −OH group

and water molecule, occupy larger spaces in comparison with a

formate group and thus weaken the hydrogen bonding of IPA.

The formation of FOL (D → E) can easily be achieved

through proton transfer from IPA to deprotonated FOL along

the hydrogen bonding between them with small (3.23 kcal/

mol) and negligible (∼0 kcal/mol) E values for both s and d

Zr nodes, respectively, in which the transition state has a six-

ASSOCIATED CONTENT

sı Supporting Information

The Supporting Information is available free of charge at

■

membered-ring conﬁguration also involving two Zr centers

Ulsan 44776, Republic of Korea; orcid.org/0000-0001-

Figure S18). The computationally obtained negative E of a

negligible extent (−0.6 kcal/mol) for a d Zr node indicates a

barrierless reaction process. In contrast to the substitution of

acetone with IPA, the desorption of FOL (+10.60 kcal/mol)

for the s Zr node is twice as unfavorable as that for the d Zr

node (+4.97 kcal/mol), which is also related to hydrogen

bonding between FOL and deprotonated IPA: the hydrogen

bond lengths for s and d Zr nodes are 1.37 and 1.43 Å,

respectively (Figure S18 ). Our DFT computations suggest that

a highly modiﬁed Zr node in a Zr-MOF with lower

connectivity can provide higher catalytic activity due to the

following features: (i) the d Zr node can easily initiate a

catalytic cycle with exothermic insertion of the ﬁrst IPA

molecule and (ii) the produced FOL can be more easily

released from the d Zr node in comparison with the s Zr node

due to slightly weakened hydrogen bonding, although both s

and d Zr nodes have an almost identical activation energy at

the rate-determining step.

Experimental section including synthesis and character-

izations of Zr-MOFs, computational methods and

simulations, additional catalyst characterizations (XRD,

TGA, surface area analysis, digested H NMR, SEM,

TEM, EA, in situ FTIR, NH and CO TPD, mass

spectra of major products), plausible reaction networks,

and additional catalytic tests (PDF )

AUTHOR INFORMATION

Corresponding Authors

■

Jaehoon Jung − Department of Chemistry, University of Ulsan,

550-139X ; Email: jjung2015@ulsan.ac.kr

Young Kyu Hwang − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea; Department of Advanced Materials and Chemical

Engineering, University of Science and Technology (UST),

Daejeon 34113, Korea; orcid.org/0000-0002-3940-3013;

Email: ykhwang@krict.re.kr

. CONCLUSIONS

To summarize, a series of Zr-MOFs exhibiting diﬀerent metal

node connectivities (12, 8, and 6) were synthesized and tested

for CTH of FUR to FOL in order to understand the role of

metal node coordination. It was found that metal node to

linker connectivity plays a dominant role over high porosity of

the Zr-MOFs. MOF-808, with the lowest metal node

connectivity (6), enhances the accessibility of reactant

molecules to active sites (metal node), resulting in the highest

catalytic activity among the other tested Zr-MOFs with a

higher connectivity (8 and 12). Further, the metal node of

MOF-808 was modiﬁed by methanol activation and extensively

studied with various characterization techniques to evaluate

any changes to its local structure. Signiﬁcant enhancement in

the catalytic activity of MOF-808 for CTH of FUR to FOL was

reported at low reaction temperatures (30−40 °C) after metal

node modiﬁcation. The modiﬁcation enhanced the acid−base

sites of MOF-808 by developing Zr-CUS and surface hydroxyl

groups, which presumably facilitate the adsorption of FUR and

Authors

Anil H. Valekar − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea

Minhui Lee − Department of Chemistry, University of Ulsan,

Ulsan 44776, Republic of Korea

Ji Woong Yoon − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea

Jaesung Kwak − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea

Do-Young Hong − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea; Department of Advanced Materials and Chemical

Engineering, University of Science and Technology (UST),

Daejeon 34113, Korea

-propanol onto the Zr node surface.

The catalyst was recycled ﬁve times with little observed drop

in activity and without a change in FOL selectivity after the

ﬁrst cycle. The structural stability of the catalyst was

maintained after ﬁve cycles, as conﬁrmed by PXRD, SEM,

and TEM analyses; however, strong adsorption of organic

moieties from reactant or product molecules might have

blocked some active sites of the catalyst surface after the ﬁrst

cycle. This was recognized using TGA and in situ FTIR study.

Kyung-Ryul Oh − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea; Department of Chemistry, Sungkyunkwan University,

Suwon 16419, Korea

Ga-Young Cha − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea; Department of Advanced Materials and Chemical

729

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

(11) Li, J.; Liu, J. L.; Zhou, H. J.; Fu, Y. Catalytic Transfer

Research Article

Engineering, University of Science and Technology (UST),

Daejeon 34113, Korea

Young-Uk Kwon − Department of Chemistry, Sungkyunkwan

University, Suwon 16419, Korea; School of Materials Science

Engineering, Tianjin Polytechnic University, Tianjin 300387,

People’s Republic of China; orcid.org/0000-0002-8065-

Hydrogenation of Furfural to Furfuryl Alcohol over Nitrogen-Doped

Carbon-Supported Iron Catalysts. ChemSusChem 2016, 9, 1339−

347.

12) Wang, F.; Zhang, Z. Catalytic Transfer Hydrogenation of

Furfural into Furfuryl Alcohol over Magnetic γ -Fe O @HAP Catalyst.

(13) Chia, M.; Dumesic, J. A. Liquid-Phase Catalytic Transfer

ACS Sustainable Chem. Eng. 2017, 5 , 942 −947.

37X

Jong-San Chang − Research Center for Nanocatalysts, Korea

Research Institute of Chemical Technology, Daejeon 34114,

Korea; Department of Chemistry, Sungkyunkwan University,

Suwon 16419, Korea; orcid.org/0000-0003-3640-8190

Hydrogenation and Cyclization of Levulinic Acid and its Esters to γ -

Valerolactone over Metal Oxide Catalysts. Chem. Commun. 2011, 47,

(

2233.

14) Tang, X.; Chen, H.; Hu, L.; Hao, W.; Sun, Y.; Zeng, X.; Lin, L.;

Liu, S. Conversion of Biomass to γ -Valerolactone by Catalytic

Transfer Hydrogenation of Ethyl Levulinate over Metal Hydroxides.

Appl. Catal., B 2014, 147, 827−834.

Complete contact information is available at:

https://pubs.acs.org/10.1021/acscatal.9b05085

(15) Song, J.; Zhou, B.; Zhou, H.; Wu, L.; Meng, Q.; Liu, Z.; Han, B.

Notes

Porous Zirconium −Phytic Acid Hybrid: a Highly Efficient Catalyst

The authors declare no competing ﬁnancial interest.

for Meerwein −Ponndorf −Verley Reductions. Angew. Chem., Int. Ed.

16) Li, H.; He, J.; Riisager, A.; Saravanamurugan, S.; Song, B.; Yang,

015, 54, 9399−9403.

ACKNOWLEDGMENTS

■

S. Acid −Base Bifunctional Zirconium N-Alkyltriphosphate Nano-

hybrid for Hydrogen Transfer of Biomass-Derived Carboxides. ACS

Catal. 2016, 6, 7722−7727.

This work was supported by the Next Generation Carbon

Upcycling Project (2017M1A2A2043143 and

017M1A2A20431414) by the National Research Foundation

(17) Li, H.; Liu, X.; Yang, T.; Zhao, W.; Saravanamurugan, S.; Yang,

of Korea. Dr. D.-Y. Hong and Y. K. Hwang kindly acknowledge

support from the Institutional Research program of KRICT

S. Porous Zirconium −Furandicarboxylate Microspheres for Efficient

Redox Conversion of Biofuranics. ChemSusChem 2017, 10, 1761−

1770.

(18) Li, H.; Fang, Z.; He, J.; Yang, S. Orderly Layered Zr-

Benzylphosphonate Nanohybrids for Efficient Acid −Base-Mediated

Bifunctional/Cascade Catalysis. ChemSusChem 2017, 10, 681−686.

(19) Iglesias, J.; Melero, J.; Morales, G.; Moreno, J.; Segura, Y.;

(

SI2011-10). The authors thank Dr. U.−H. Lee and Dr. D. W.

Hwang for helpful discussions.

REFERENCES

■

(

1) Mariscal, R.; Maireles-Torres, P.; Ojeda, M.; Sadaba, I.; Lopez

Granados, M. Furfural: a renewable and versatile platform molecule

Paniagua, M.; Cambra, A.; Hernan

Catalyst: Activity in Meerwein Ponndorf Verley Reduction. Catalysts

015, 5, 1911−1927.

20) Montes, V.; Min

dez, B. Zr-SBA-15 Lewis Acid

for the synthesis of chemicals and fuels. Energy Environ. Sci. 2016, 9,

(

144−1189.

ambres, J. F.; Khalilov, A. N.; Boutonnet, M.;

2) Assary, R. S.; Curtiss, L. A.; Dumesic, J. A. Exploring Meerwein −

Marinas, J. M.; Urbano, F. J.; Maharramov, A. M.; Marinas, A.

Chemoselective Hydrogenation of Furfural to Furfuryl Alcohol on

ZrO ₂ Systems Synthesized Through the Microemulsion Method.

Catal. Today 2018, 306, 89−95.

Ponndorf −Verley Reduction Chemistry for Biomass Catalysis Using a

First-Principles Approach. ACS Catal. 2013 , 3 , 2694 −2704.

3) Gilkey, M. J.; Xu, B. Heterogeneous Catalytic Transfer

Hydrogenation as an Effective Pathway in Biomass Upgrading. ACS

Catal. 2016, 6, 1420−1436.

(21) Vermoortele, F.; Bueken, B.; Le Bars, G.; Van De Voorde, B.;

Vandichel, M.; Houthoofd, K.; Vimont, A.; Daturi, M.; Waroquier,

M.; Van Speybroeck, V.; Kirschhock, C.; De Vos, D. E. Synthesis

Modulation as a Tool To Increase the Catalytic Activity of Metal −

Organic Frameworks: The Unique Case of UiO-66(Zr). J. Am. Chem.

Soc. 2013, 135, 11465−11468.

(4) Neeli, C. K. P.; Chung, Y.-M.; Ahn, W.-S. Catalytic Transfer

Hydrogenation of Furfural to Furfuryl Alcohol by using Ultrasmall Rh

Nanoparticles Embedded on Diamine-Functionalized KIT-6. Chem-

CatChem 2017, 9, 4570−4579.

(5) Panagiotopoulou, P.; Vlachos, D. G. Liquid Phase Catalytic

(22) Plessers, E.; De Vos, D. E.; Roeffaers, M. B. J. Chemoselective

Transfer Hydrogenation of Furfural over a Ru/C Catalyst. Appl.

Catal., A 2014, 480, 17−24.

Reduction of α ,β -Unsaturated Carbonyl Compounds with UiO-66

Materials. J. Catal. 2016, 340, 136−143.

(6) Panagiotopoulou, P.; Martin, N.; Vlachos, D. G. Effect of

(23) Plessers, E.; Fu, G.; Tan, C.; De Vos, D.; Roeffaers, M. Zr-

Hydrogen Donor on Liquid Phase Catalytic Transfer Hydrogenation

Based MOF-808 as Meerwein −Ponndorf −Verley Reduction Catalyst

for Challenging Carbonyl Compounds. Catalysts 2016, 6, 104.

of Furfural over a Ru/RuO /C Catalyst. J. Mol. Catal. A: Chem. 2014,

(

92, 223−228.

(24) Valekar, A. H.; Cho, K.-H.; Chitale, S. K.; Hong, D.-Y.; Cha,

7) Gilkey, M. J.; Panagiotopoulou, P.; Mironenko, A. V.; Jenness, G.

G.-Y.; Lee, U.-H.; Hwang, D. W.; Serre, C.; Chang, J.-S.; Hwang, Y. K.

Catalytic Transfer Hydrogenation of Ethyl Levulinate to γ -

Valerolactone over Zirconium-Based Metal −Organic Frameworks.

Green Chem. 2016, 18, 4542−4552.

R.; Vlachos, D. G.; Xu, B. Mechanistic Insights into Metal Lewis Acid-

Mediated Catalytic Transfer Hydrogenation of Furfural to 2-

Methylfuran. ACS Catal. 2015 , 5 , 3988 −3994.

(8) Koehle, M.; Lobo, R. F. Lewis Acidic Zeolite Beta Catalyst for

(25) He, J.; Li, H.; Riisager, A.; Yang, S. Catalytic Transfer

the Meerwein −Ponndorf −Verley Reduction of Furfural. Catal. Sci.

Hydrogenation of Furfural to Furfuryl Alcohol with Recyclable Al −

Technol. 2016, 6, 3018−3026.

Zr@Fe Mixed Oxides. ChemCatChem 2018, 10, 430 −438.

(9) Antunes, M. M.; Lima, S.; Neves, P.; Magalhaes, A. L.; Fazio, E.;

(26) He, J.; Schill, L.; Yang, S.; Riisager, A. Catalytic Transfer

Fernandes, A.; Neri, F.; Silva, C. M.; Rocha, S. M.; Ribeiro, M. F.;

Pillinger, M.; Urakawa, A.; Valente, A. A. One-pot conversion of

furfural to useful bio-products in the presence of a Sn,Al-containing

zeolite beta catalyst prepared via post-synthesis routes. J. Catal. 2015,

Hydrogenation of Bio-Based Furfural with NiO Nanoparticles. ACS

Sustainable Chem. Eng. 2018, 6, 17220−17229.

(27) Zhang, J.; Dong, K.; Luo, W.; Guan, H. Selective Transfer

Hydrogenation of Furfural into Furfuryl Alcohol on Zr-Containing

Catalysts Using Lower Alcohols as Hydrogen Donors. ACS Omega

2018, 3, 6206−6216.

10) Bui, L.; Luo, H.; Gunther, W. R.; Roman-Leshkov, Y. Domino

29, 522−537.

Reaction Catalyzed by Zeolites with Brønsted and Lewis Acid Sites

for the Production of g-Valerolactone from Furfural. Angew. Chem.,

Int. Ed. 2013, 52, 8022−8025.

(28) Hao, W.; Li, W.; Tang, X.; Zeng, X.; Sun, Y.; Liu, S.; Lin, L.

Catalytic transfer hydrogenation of biomassderived 5-hydroxymethyl

730

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

furfural to the building block 2,5-bishydroxymethyl furan. Green

Chem. 2016, 18, 1080−1088.

Hafnium-Based Metal −Organic Frameworks. ACS Catal. 2017, 7,

997−1014.

29) Zhou, S.; Dai, F.; Xiang, Z.; Song, T.; Liu, D.; Lu, F.; Qi, H.

(47) Salomon, W.; Roch-Marchal, C.; Mialane, P.; Rouschmeyer, P.;

Serre, C.; Haouas, M.; Taulelle, F.; Yang, S.; Ruhlmann, L.; Dolbecq,

A. Immobilization of Polyoxometalates in the Zr-Based Metal Organic

Framework UiO-67. Chem. Commun. 2015, 51, 2972−2975.

(48) Bon, V.; Senkovska, I.; Weiss, M. S.; Kaskel, S. Tailoring of

Network Dimensionality and Porosity Adjustment in Zr- and Hf-

Based MOFs. CrystEngComm 2013, 15, 9572.

Zirconium −Lignosulfonate Polyphenolic Polymer for Highly Efficient

Hydrogen Transfer of Biomass-Derived Oxygenates under Mild

Conditions. Appl. Catal., B 2019, 248, 31−43.

(30) Wang, R.; Wang, J.; Zi, H.; Wang, H.; Xia, Y.; Liu, X.

Conversion of ethyl levulinate to γ -valerolactone catalyzed by the new

Zr-containing organic −inorganic hybrid catalysts. J. Chin. Chem. Soc.

(

018, 65, 1398−1406.

(49) Bon, V.; Senkovska, I.; Baburin, I. A.; Kaskel, S. Zr- and Hf-

Based Metal −Organic Frameworks: Tracking Down the Poly-

morphism. Cryst. Growth Des. 2013, 13, 1231−1237.

31) Lopez-Asensio, R.; Gomez, C. P. J.; Sancho, C. G.; Moreno-

Tost, R.; Cecilia, J. A.; Maireles-Torres, P. Influence of Structure-

modifying Agents in the Synthesis of Zr-doped SBA-15 Silica and

Their Use as Catalysts in the Furfural Hydrogenation to Obtain High

Value-added Products through the Meerwein-Ponndorf-Verley

Reduction. Int. J. Mol. Sci. 2019, 20, 828.

(50) Furukawa, H.; Gandara, F.; Zhang, Y.-B.; Jiang, J.; Queen, W.

L.; Hudson, M. R.; Yaghi, O. M. Water Adsorption in Porous Metal −

Organic Frameworks and Related Materials. J. Am. Chem. Soc. 2014,

136, 4369−4381.

(32) Gong, W.; Chen, C.; Fan, R.; Zhang, H.; Wang, G.; Zhao, H.

(51) Jung, K. D.; Bell, A. T. Role of Hydrogen Spillover in Methanol

Transfer-hydrogenation of furfural and levulinic acid over supported

Synthesis over Cu/ZrO . J. Catal. 2000, 193 , 207 −223.

copper catalyst. Fuel 2018, 231, 165−171.

(52) Rhodes, M.; Pokrovski, K.; Bell, A. The Effects of Zirconia

Morphology on Methanol Synthesis from CO and H over Cu/ZrO

(33) Villaverde, M. M.; Garetto, T. F.; Marchi, A. J. Liquid-phase

transfer hydrogenation of furfural to furfuryl alcohol on Cu −Mg −Al

catalysts Part II. Transient-Response Infrared Studies. J. Catal. 2005,

catalysts. Catal. Commun. 2015, 58, 6−10.

233, 210−220.

34) Xu, S.; Yu, D.; Ye, T.; Tiana, P. Catalytic transfer

(53) Yang, D.; Bernales, V.; Islamoglu, T.; Farha, O. K.; Hupp, J. T.;

Cramer, C. J.; Gagliardi, L.; Gates, B. C. Tuning the Surface

Chemistry of Metal Organic Framework Nodes: Proton Topology of

hydrogenation of levulinic acid to γ -valerolactone over a bifunctional

tin catalyst. RSC Adv. 2017, 7, 1026 −1031.

35) Rani, P.; Srivastava, R. Integration of a metal −organic

the Metal-Oxide-Like Zr Nodes of UiO-66 and NU-1000. J. Am.

framework with zeolite: a highly sustainable composite catalyst for

the synthesis of g-valerolactone and coumarins. Sustainable Energy

Fuels 2018, 2, 1287−1298.

Chem. Soc. 2016, 138, 15189−15196.

(54) Ouyang, F.; Kondo, J. N.; Maruya, K.; Domen, K. IR Study on

H/D Isotope Exchange Reactions of Formate and Methoxy Species

with D on ZrO . J. Chem. Soc., Faraday Trans. 1997, 93, 169−174.

36) Zhu, Y.; Chuah, G.-K.; Jaenicke, S. Selective Meerwein −

Ponndorf −Verley reduction of α ,β -unsaturated aldehydes over Zr-

zeolite beta. J. Catal. 2006, 241, 25−33.

(55) Planas, N.; Mondloch, J. E.; Tussupbayev, S.; Borycz, J.;

Gagliardi, L.; Hupp, J. T.; Farha, O. K.; Cramer, C. J. Defining the

(37) Rojas-Buzo, S.; Garcia-Garcia, P.; Corma, A. Catalytic Transfer

Proton Topology of the Zr -Based Metal −Organic Framework NU-

Hydrogenation of Biomass-Derived Carbonyls over Hafnium-Based

1000. J. Phys. Chem. Lett. 2014, 5, 3716−3723.

Metal −Organic Frameworks. ChemSusChem 2018, 11, 432−438.

(56) Deria, P.; Bury, W.; Hod, I.; Kung, C. W.; Karagiaridi, O.;

Hupp, J. T.; Farha, O. K. MOF Functionalization via Solvent-Assisted

Ligand Incorporation: Phosphonates vs Carboxylates. Inorg. Chem.

2015, 54, 2185−2192.

(38) Kurisingal, J. F.; Rachuri, Y.; Palakkal, A. S.; Pillai, R. S.; Gu, Y.;

Choe, Y.; Park, D.-W. Water-Tolerant DUT-Series Metal −Organic

Frameworks: A Theoretical −Experimental Study for the Chemical

Fixation of CO ₂ and Catalytic Transfer Hydrogenation of Ethyl

Levulinate to γ -Valerolactone. ACS Appl. Mater. Interfaces 2019, 11,

(57) Kouva, S.; Honkala, K.; Lefferts, L.; Kanervo, J. Review:

Monoclinic Zirconia, its Surface Sites and Their Interaction with

Carbon Monoxide. Catal. Sci. Technol. 2015, 5, 3473−3490.

(58) Vimont, A.; Goupil, J.-M.; Lavalley, J.-C.; Daturi, M.; Surble, S.;

Serre, C.; Millange, F.; Ferey, G.; Audebrand, N. Investigation of Acid

Sites in a Zeotypic Giant Pores Chromium(III) Carboxylate. J. Am.

Chem. Soc. 2006, 128, 3218−3227.

(59) Mautschke, H.-H.; Drache, F.; Senkovska, I.; Kaskel, S.;

Xamena, F. X. L. Catalytic Properties of Pristine and Defect

Engineered Zr-MOF-808 Metal Organic Frameworks. Catal. Sci.

Technol. 2018, 8, 3610−3616.

(60) Driscoll, D. M.; Troya, D.; Usov, P. M.; Maynes, A. J.; Morris,

A. J.; Morris, J. R. Characterization of Undercoordinated Zr Defect

Sites in UiO-66 with Vibrational Spectroscopy of Adsorbed CO. J.

Phys. Chem. C 2018, 122, 14582−14589.

(61) Fang, H.; Xia, J.; Zhu, K.; Su, Y.; Jiang, Y. Industrial Waste Heat

Utilization for Low Temperature District Heating. Energy Policy 2013,

62, 236−246.

(62) Ammar, Y.; Joyce, S.; Norman, R.; Wang, Y.; Roskilly, A. P.

Low Grade Thermal Energy Sources and Uses from the Process

Industry in the UK. Appl. Energy 2012, 89, 3−20.

39) Chen, X.; Zhang, L.; Zhang, B.; Guo, X.; Mu, X. Highly

1458−41471.

selective hydrogenation of furfural to furfuryl alcohol over Pt

nanoparticles supported on g-C N nanosheets catalysts in water.

Sci. Rep. 2016, 6, 28558.

40) Yuan, Q.; Zhang, D.; Haandel, L. V.; Ye, F.; Xue, T.; Hensen, E.

(

J. M.; Guan, Y. Selective liquid phase hydrogenation of furfural to

furfuryl alcohol by Ru/Zr-MOFs. J. Mol. Catal. A: Chem. 2015, 406,

41) Wang, G.; Yao, R.; Xin, H.; Guan, Y.; Wu, P.; Li, X. At room

8−64.

temperature in water: efficient hydrogenation of furfural to furfuryl

alcohol with a Pt/SiC −C catalyst. RSC Adv. 2018, 8, 37243−37253.

(42) Xiao, C.; Goh, T.-W.; Qi, Z.; Goes, S.; Brashler, K.; Perez, C.;

Huang, W. Conversion of Levulinic Acid to γ -Valerolactone over Few-

Layer Graphene-Supported Ruthenium Catalysts. ACS Catal. 2016, 6,

43) Tan, J.; Cui, J.; Zhu, Y.; Cui, X.; Shi, Y.; Yan, W.; Zhao, Y.

93−599.

Complete Aqueous Hydrogenation of 5-Hydroxymethylfurfural at

Room Temperature over Bimetallic RuPd/Graphene Catalyst. ACS

Sustainable Chem. Eng. 2019, 7, 10670−10678.

(63) Zhou, S.; Dai, F.; Xiang, Z.; Song, T.; Liu, D.; Lu, F.; Qi, H.

Zirconium −Lignosulfonate Polyphenolic Polymer for Highly Efficient

Hydrogen Transfer of Biomass-Derived Oxygenates under Mild

Conditions. Appl. Catal., B 2019, 248, 31−43.

(64) Tang, X.; Hu, L.; Sun, Y.; Zhao, G.; Hao, W.; Lin, L.

Conversion of Biomass-Derived Ethyl Levulinate into γ -Valerolactone

via Hydrogen Transfer from Supercritical Ethanol over a ZrO ₂

Catalyst. RSC Adv. 2013, 3, 10277.

(44) Bai, Y.; Dou, Y.; Xie, L.-H.; Rutledge, W.; Li, J.-R.; Zhou, H.-C.

Zr-Based Metal −Organic Frameworks: Design, Synthesis, Structure,

and Applications. Chem. Soc. Rev. 2016, 45, 2327−2367.

(45) Liu, Y.; Klet, R. C.; Hupp, J. T.; Farha, O. Probing the

Correlations Between the Defects in Metal −Organic Frameworks and

Their Catalytic Activity by an Epoxide Ring-Opening Reaction. Chem.

Commun. 2016, 52, 7806−7809.

(46) Rimoldi, M.; Howarth, A. J.; DeStefano, M. R.; Lin, L.;

(65) Gonell, F.; Boronat, M.; Corma, A. Structure −reactivity

Goswami, S.; Li, P.; Hupp, J. T.; Farha, O. K. Catalytic Zirconium/

relationship in isolated Zr sites present in Zr-zeolite and ZrO for

731

ACS Catal. 2020, 10, 3720−3732

ACS Catalysis

Research Article

the Meerwein −Ponndorf −Verley reaction. Catal. Sci. Technol. 2017,

, 2865−2873.

732