An efficient approach to synthesizing 2,5-bis(: N -methyl-aminomethyl)furan from 5-hydroxymethylfurfural via 2,5-bis(N -methyl-iminomethyl)furan using a two-step reaction in one pot

Article abstract of DOI:10.1039/d1gc01635k

In this study, an amination-oxidation-amination-reduction (AOAR) strategy was proposed for the synthesis of 2,5-bis(N-methyl-aminomethyl)furan (BMAF) from HMF via the intermediate 2,5-bis(N-methyl-iminomethyl)furan (BMIF). Firstly, an efficient synthesis of BMIF from HMF using a one-pot amination-oxidation-amination reaction was developed over α-MnO2 under an air atmosphere. A BMIF yield of 98.3% was obtained under mild reaction conditions. The conversion of HMF to BMIF underwent the fast amination of HMF to 5-(methyl-iminomethyl)furfuryl alcohol (MIFA), the subsequent rate-limiting oxidation of MIFA to 5-(methyl-iminomethyl)furfural (MIFF) and the final fast amination of MIFF to BMIF. The quick amination of MIFF to BMIF drove the oxidation reaction equilibrium toward MIFF from MIFA, which ensured the highly efficient conversion of HMF to BMIF. The investigation of the catalytic mechanism showed better lattice oxygen donating ability and oxygen coordination capacity, which made α-MnO2 retain the structural stability in the reaction. The higher ratio and better mobility of the lattice oxygen endowed α-MnO2 with excellent catalytic performance in the oxidation of MIFA to MIFF by the redox cycling of Mn4+/Mn3+, facilitating the conversion of HMF to BMIF. Eventually, a BMAF yield of 96.1% was achieved by the reduction of BMIF with Ru/C after the AOA reaction, realizing the synthesis of BMAF from HMF using a two-step reaction in one pot. This journal is

Full text of DOI:10.1039/d1gc01635k

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
An efficient approach to synthesizing
2,5-bis(N-methyl-aminomethyl)furan from
5-hydroxymethylfurfural via 2,5-bis(N-methyl-
iminomethyl)furan using a two-step reaction
in one pot†
Cite this: Green Chem., 2021, 23,
5656
a,b
Jie Zhang,a Wenlong Jia,a Yong Sun,
*
a,b,c Shuliang Yang,a,b Xing Tang,
Xianhai Zeng a,b and Lu Lina,b
In this study, an amination–oxidation–amination–reduction (AOAR) strategy was proposed for the syn-
thesis of 2,5-bis(N-methyl-aminomethyl)furan (BMAF) from HMF via the intermediate 2,5-bis(N-methyl-
iminomethyl)furan (BMIF). Firstly, an efficient synthesis of BMIF from HMF using a one-pot amination–
oxidation–amination reaction was developed over α-MnO2 under an air atmosphere. A BMIF yield of
98.3% was obtained under mild reaction conditions. The conversion of HMF to BMIF underwent the fast
amination of HMF to 5-(methyl-iminomethyl)furfuryl alcohol (MIFA), the subsequent rate-limiting oxi-
dation of MIFA to 5-(methyl-iminomethyl)furfural (MIFF) and the final fast amination of MIFF to BMIF. The
quick amination of MIFF to BMIF drove the oxidation reaction equilibrium toward MIFF from MIFA, which
ensured the highly efficient conversion of HMF to BMIF. The investigation of the catalytic mechanism
showed better lattice oxygen donating ability and oxygen coordination capacity, which made α-MnO2
retain the structural stability in the reaction. The higher ratio and better mobility of the lattice oxygen
endowed α-MnO2 with excellent catalytic performance in the oxidation of MIFA to MIFF by the redox
cycling of Mn4+/Mn3+, facilitating the conversion of HMF to BMIF. Eventually, a BMAF yield of 96.1% was
achieved by the reduction of BMIF with Ru/C after the AOA reaction, realizing the synthesis of BMAF from
HMF using a two-step reaction in one pot.
Received 9th May 2021,
Accepted 24th June 2021
DOI: 10.1039/d1gc01635k
rsc.li/greenchem
cerns about energy and the environment, the catalytic conver-
sion of lignocellulosic biomass to renewable versatile diamines
Introduction
Diamines are important intermediates in pharmaceuticals, has been extensively researched.5–7 A prevalent strategy is to
food additives and fine-chemicals. They also act as material convert lignocellulosic biomass to platform chemicals,8–13
monomers for the large-scale production of numerous poly- such as 5-hydroxymethylfurfural (HMF), furfural and levulinic
mers, including polyamides, epoxy polymers and polyureas, acid, in the first step, and then the corresponding diamines
which are increasingly used in health applications and aero- can be produced through a series of further reactions.14–17
space, automotive, agrochemical and building industries.1–4
In recent years, some progress has been made in seeking
Biomass is one such resource that has attracted significant sustainable solutions for the production of furan-diamines
attention due to its high abundance. With great social con- from renewable biomass-derived HMF. Several strategies for
the preparation of furan-diamines from HMF have been
explored, including the stepwise amination pathway, oxi-
dation–amination pathway and amidation–amination pathway,
as shown in Scheme 1. For the stepwise amination pathway,
HMF is converted to 5-aminomethyl-furfuryl alcohol (AMF) by
reductive amination or reduced to 2,5-bis(hydroxymethyl)furan
(BHMF), and then AMF or BHMF is directly aminated to furan-
diamines via the hydrogen-borrowing reaction of a hydroxyl
group with an amine source.18–21 This pathway breaks through
the direct amination of the hydroxyl group, however, it
aXiamen Key Laboratory of Clean and High-valued Utilization for Biomass,
College of Energy, Xiamen University, Xiamen, 361102, P. R. China.
E-mail: sunyong@xmu.edu.cn
bFujian Engineering and Research Center of Clean and High-valued Technologies for
Biomass, Xiamen, 361102, P. R. China
cHubei Key Laboratory of Pollutant Analysis & Reuse Technology Hubei Normal
University, Huangshi, 435002 Hubei, P. R. China
†Electronic supplementary information (ESI) available: Experimental details,
catalyst characterization and tests. See DOI: 10.1039/d1gc01635k
5656 | Green Chem., 2021, 23, 5656–5664
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
that enables the ultraselective production of BMIF by integrat-
ing the oxidation and amination of HMF in one pot using air
as an oxidant over MnO2. As is well known to all, MnO2 has
various crystallographic structures, which significantly influ-
ence the catalytic property during oxidation.28,29 At the outset
of our work, the investigation was focused on the performance
of MnO2 catalysts with different crystalline phases (including
α-, β-, and γ-MnO2). Excitingly, these three crystalline MnO2
catalysts exhibited excellent catalytic activity, especially
α-MnO2 gave a better BMIF yield (98.3%) than γ- and β-MnO2
(Table 1, entries 2–4). In addition, several other transition
metal catalysts (such as Fe-, Cu-, Zn-, V-, and Mo-based oxides
or salts) usually used for the oxidation of HMF to DFF were
also examined in this AOA reaction,30–34 but all candidates
failed to give satisfactory results (Table 1, entries 5–11). It is
worth noting that the targeted products, such as MIFA and
BMIF, were not found in Table 1, entries 9–11, however, an
unknown compound with a relative molecular mass of 237
(denoted as MW237) as the main product was detected by
GC-MS measurement (Fig. S1†). In addition, MIFA was also
detected in 0.5 h and 1.5 h when used V2O5 as a catalyst
(Table S1,† entries 1 and 2). Importantly, the nucleophilic reac-
tion between methylamine and HMF in this study was rapidly
completed within 10 min (Table 1, entry 1). Therefore, it is
speculated that the unknown compound MW237 should be a
dimer from MIFA. Specifically, the reaction solution from the
catalysis of MnO2 was light yellow, however, it became brown-
ish red when V2O5 was used as a catalyst (Fig. S2†). The dar-
kened colour of the reaction solution (Fig. S2b†) was probably
caused by the polymerization of MIFA with its derivatives.
The reaction profiles of the three kinds of MnO2 in the AOA
reaction are shown in Fig. S3a.† It can be seen intuitively that
the catalytic performance of these three MnO2 catalysts always
followed the order of α-MnO2 > γ-MnO2 > β-MnO2 in the whole
Scheme 1 Renewable synthesis pathway of HMF to furan-diamines.
requires a homogeneous noble metal catalyst to obtain high
efficiency. In contrast, prolonging the reaction time and
increasing the reaction temperature are necessary for hetero-
geneous non-noble metal catalysts. Since aldehyde groups
facilely react with amines to form furan-amines through reduc-
tive amination, 2,5-diformylfuran (DFF) is an ideal starting
material to produce furan-diamines. Therefore, the oxidation–
amination of HMF is based on the first oxidation of HMF to
DFF, and then the reductive amination of DFF.22–24 Although
this reaction route is promising, some problems including an
efficient and simple access to produce DFF and the polymeriz-
ation of furan-amines with DFF by nucleophilic addition,
resulting in the drop in the synthesis efficiency of furan-
amines, are still needed to be resolved.25 As for the amidation–
amination pathway, furan-amines are generally prepared by
the first amidation of the hydroxymethyl group in HMF with
CH3CN using the Ritter reaction, the subsequent amination of
the aldehyde group and the final hydrolysis of the acylamino
group.26 This reaction system is relatively complex and the syn-
thesis efficiency is still needed to be enhanced.
In this study, a novel efficient protocol was proposed to syn-
thesize 2,5-bis(N-methyl-aminomethyl)furan (BMAF) from
HMF using a two-step reaction in one pot. Firstly, HMF was
converted to 2,5-bis(N-methyl-iminomethyl)furan (BMIF) using
the amination–oxidation–amination (AOA) reaction in one pot,
followed by the reduction of BMIF to BMAF in the same reac-
tion system. This protocol is characterized by no generation of
furan amine-derived polymers that usually occur in the reduc-
tive amination of DFF to furan-amines. More specifically,
α-MnO2 showed excellent performance in this system for acti-
vating oxygen in air to oxidize HMF with high selectivity.
Encouraged by these results, the reaction pathway of HMF to
BMIF and the catalytic mechanism were deeply investigated.
Table 1 Catalytic conversion of HMF to BMIF over different catalystsa
Yieldb (%)
BMIF
Entry
Catalyst
Conv. (%)
MIFAc
1d
2
3
4
5
6
7
8
9
10
11
12e
13e
14e
—
100
100
100
100
100
100
100
100
100
100
100
100
100
100
0
>99
<1
α-MnO2
β-MnO2
γ-MnO2
α-Fe2O3
γ-Fe2O3
CuO
98.3
84.8
93.1
3.7
6.9
5.4
3.9
0
0
0
97.5
42.7
67.3
13.9
5.7
76.9
81.7
41.7
53.1
0
0
0
1.3
56.1
31.6
ZnO
V2O5
Results and discussion
Catalyst screening for the synthesis of 2,5-bis(N-methyl-
VOSO4
MoO3
α-MnO2
β-MnO2
γ-MnO2
iminomethyl)furan (BMIF)
Building upon our previous work, we found that MnO2 exhibi-
ted excellent performance in the activation of air oxygen to
oxidize HMF or 5-[(formate)methyl]furfural (FMF) for the pro-
duction of maleic anhydride.27 Moreover, 2,5-diformylfuran
was found as a crucial oxidative intermediate in the reaction.
Spurred by this, we tried to develop a novel synthesis system
a Reaction conditions: 0.5 mmol HMF (63 mg), 5.0 mL of 30 wt%
methylamine methanol solution, 22 mg of catalyst, 1.0 MPa air, 90 °C,
3 h. b Other products were derivatives produced by MIFA. c MIFA, 5-
(methyl-iminomethyl)furfuryl alcohol. d No catalyst, 10 min. e The
recovered catalyst from the first cycle.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 5656–5664 | 5657

View Article Online
Paper
Green Chemistry
Table 2 Oxidation reactions of HMF or MIFA under various conditions
over α-MnO2
reaction period. The turnover frequencies (TOFs) (Fig. S3b†) of
these catalysts showed that their catalytic activity followed the
above order. More importantly, the reusability of β- and
γ-MnO2 was poor (Table 1, entries 12–14). Thus, α-MnO2 was
selected as the preferred catalyst for the subsequent study.
Meanwhile, the catalytic performance of these three crystalline
MnO2 catalysts clearly demonstrated that the crystal structure
has a great influence on the conversion of HMF to BMIF.
a
Yield (%)
Conv.
(%)
Molar
Entry Conditions
BMIF DFF Othersd balance
1b
40 mmol
Methylamine
—
100
>99
0
—
99
2b
3
60.2
100
23.5
57.3
3.9 32.0
0 40.3
99
98
H2SO4
(pH = 4, 20 mg)
40 mmol
Reaction pathway of HMF to BMIF
methylamine
—
—
pH = 6
pH = 4
O2 (1 MPa)
O2 (1 MPa), pH = 6
O2 (1 MPa), pH = 4
NaHCO3 (20 mg)
In general, the lone-pair electron on the N atom of amine
spontaneously attacks the carbonyl carbocation on aldehyde
via a nucleophilic addition, followed by dehydration to form
an imine. Adding excess amine to the reaction is a common
technique to ensure the stable existence of aldimine.35–37 In
this study, HMF was completely converted into 5-(methyl-imi-
nomethyl)furfuryl alcohol (MIFA) within 10 minutes via the
nucleophilic addition of the aldehyde group with methylamine
even without a catalyst at room temperature, characterized by a
fast reaction (Fig. 1 and Fig. S4†).22 By prolonging the reaction
time, BMIF was gradually detected accompanied by a drop in
the yield of MIFA, as shown in Fig. 1. The product distribution
with the reaction course inferred that the conversion of HMF
to BMIF certainly underwent a stepwise reaction of HMF to
MIFA, followed by MIFA to BMIF. During the reaction, a little
5-(methyl-iminomethyl)furfural (MIFF) (Fig. S5†) was detected,
based on the fact that the fast reaction characteristic of the
nucleophilic addition between the aldehyde group and methyl-
amine leads to the formation of an aldimine in this reaction,
and it is deduced that the oxidation of a hydroxyl group in
HMF to aldehyde should be the rate-limiting step. More impor-
tantly, the decrease in the amount of MIFA was matched with
an increase in the amount of BMIF, inferring the surface reac-
tion control for the oxidation of MIFA to MIFF.
4
37.2
68.1
40.3
71.1
62.5
43.8
80.3
78.8
—
22.7 11.7
27.3 37.6
13.7 25.7
92
95
98
97
96
95
96
62
5c
6
—
—
—
—
—
—
—
7
8
9
10
11
<1
69.3
33.1 26.6
8.7 32.8
<1
77.4
48.7 0e
a Reaction conditions: 0.5 mmol HMF, 5.0 mL of MeOH, 22 mg of
α-MnO2, 1.0 MPa air, 90 °C, 3 h. b The reaction substrate is 0.5 mmol
MIFA. c 8 h. d Total yields of other products (see the details in
Table S2†). e GC not detected.
99% BMIF yield was obtained under the investigated con-
ditions (Table 2, entry 1). It has been well recognized that the
nucleophilic reaction of an amine and aldehyde to form an
aldimine is a reversible reaction. When using MIFA as the
starting material without adding methylamine, the 23.5%
yield of BMIF and 19.7% yield of HMF suggested the reverse of
the reaction equilibrium between HMF and MIFA. More
importantly, the 12.3% yield of MIFF and 3.9% yield of DFF
further confirmed the oxidation of the hydroxyl group in HMF
to the aldehyde group (Scheme 2a, Fig. S6 and S7† and
Table 2, entry 2). During this process, methylamine hydrolyzed
from MIFA provided the amine source for the formation of
BMIF. Therefore, based on the results mentioned above, a reac-
tion pathway for the amination–oxidation–amination of HMF
to BMIF is proposed, as illustrated in Scheme 2b. HMF is first
rapidly aminated to MIFA, and then the hydroxymethyl group
in MIFA is oxidized to an aldehyde group to form MIFF.
Subsequently, MIFF continues to quickly aminate to BMIF in
the presence of methylamine.
To gain insight into the molecular evolution that was
crucial for the conversion of HMF to BMIF, MIFA was used as
the starting material in this study. As expected, more than
The above results showed that α-MnO2 exhibited excellent
performance in the oxidative amination of HMF to BMIF.
Confusingly, by employing α-MnO2 as a catalyst for the oxi-
dation of HMF in a methanol solution with air, the yield of
DFF was less than 23% (Table 2, entry 4). Even when 1 MPa
Fig. 1 Time course of the conversion of HMF to BMIF over α-MnO2.
Reaction conditions: 0.5 mmol HMF (63 mg), 5.0 mL of 30 wt% methyl-
amine methanol solution, 22 mg of α-MnO2, 1.0 MPa air, 90 °C, 3 h.
Scheme 2 Oxidation of MIFA (a) and the reaction pathway of AOA from
HMF to BMIF (b).
5658 | Green Chem., 2021, 23, 5656–5664
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
oxygen was used to replace air or the reaction time was pro-
longed, the yield of DFF was not significantly improved
(Table 2, entries 5 and 8). Generally, a relatively acidic environ-
ment facilitates the enhancement of the electrophilicity of the
carbonyl group, while an alkaline environment favors dehydra-
tion to form an imine. Accordingly, the AOA reaction of HMF
was conducted in an acidic environment by adding a little
H2SO4 (Table 2, entry 3), and disappointingly, the yield of
BMIF dropped significantly, which was probably due to the
instability of the Schiff base (MIFA, MIFF and BMIF) in an
acidic environment. Likewise, in an acidic environment, less
than 15% DFF was obtained from the oxidation of HMF using
air as an oxidant, and almost no DFF was generated under a 1
MPa O2 atmosphere (Table 2, entries 6, 7, 9 and 10), whereas
an alkaline environment clearly favored the oxidation of HMF
to DFF when using NaHCO3 as an alkali source (Table 2, entry
11). In summary, an alkaline environment was made for the
oxidation of a hydroxyl group to an aldehyde group and MIFA,
MIFF and BMIF were formed in this study. More importantly,
the amination of MIFF to BMIF characterized by a fast reaction
also facilitated the oxidation reaction equilibrium toward
MIFF from MIFA, which enhanced the apparent efficiency in
the conversion of HMF to BMIF.
Fig. 2 Characterization analyses of MnO2: XRD patterns of the fresh
α-MnO2, β-MnO2, and γ-MnO2 (a); XPS spectra of Mn 2p (b) and O 1s (c)
of MnO2; O2-TPD (d) and H2-TPR (e) of fresh MnO2; and O2-TPO (f) of
the recovered MnO2.
Catalytic mechanism of MnO2 for the conversion of HMF to
BMIF
To elucidate the structural characteristics of the synthetic
MnO2, XRD, N2 adsorption/desorption isotherm, SEM and
FT-IR analyses were conducted. Initially, the phase structure of
the synthetic MnO2 was characterized by powder XRD. As
shown in Fig. 2a, the XRD patterns were in good agreement
with tetragonal α-MnO2 (JCPDS 44-0141), tetragonal β-MnO2
(JCPDS 24-0735) and orthorhombic γ-MnO2 (JCPDS 14-0644),
respectively, meaning that these three kinds of MnO2 were suc-
cessfully synthesized.38 The average sizes of fresh α-, β-, and
γ-MnO2 calculated using the Scherrer equation on the basis of
the XRD patterns were 22.1, 45.7, and 11.3 nm, respectively. As
shown in the FT-IR pattern (Fig. S8†), the band at 3450 cm−1 is
attributed to the O–H stretching vibration, and the bands
at 1630, 1385, 1120, and 1025 cm−1 are normally attributed
to the O–H bending vibrations combined with Mn atoms.
In the fingerprint area, α-, β-, and γ-MnO2 present different
IR spectra with the stretching vibration originating from the
Mn–O–Mn and Mn–O bonds in [MnO6] octahedra.39–41
Perhaps, the distinct crystalline or morphological character-
istics are mainly attributed to the difference in bonding
among these three MnO2 catalysts, ultimately affecting their
catalytic performance.
presented rod-like nanorods and γ-MnO2 consisted of straight
and radially grown nanorods characterized by urchin-like
structures. As shown in Fig. S12†, these three kinds of MnO2
showed type IV isotherms with a type H1 hysteresis loop, indi-
cating the mesoporous structure of the MnO2 nanorods.42
Specifically, Brunauer–Emmett–Teller (BET) characterization
by N2 adsorption/desorption showed that the specific surface
area and pore volume of α-MnO2 were higher than those of
β-MnO2 and γ-MnO2, corresponding to its slender needle-like
structure (Table 3). A larger BET surface area and pore volume,
together with a relatively smaller crystal size, imply a larger
crystal surface area, which facilitates the shift of oxygen
among the surface lattice, absorbed oxygen on the crystal
surface and substrate.
To identify the surface species of Mn and O, these three
MnO2 catalysts were examined by XPS measurement. As shown
in Fig. 2b and c, the characteristic peaks of Mn 2p3/2 in α-, β-,
and γ-MnO2 were located at around 642.2 eV, indicating that
Mn4+ was predominant in these three crystalline MnO2 cata-
To further figure out the morphology of the catalyst surface,
the synthetic MnO2 was observed by SEM (Fig. S9–S11†). These
three MnO2 catalysts with different crystal phases obviously
presented different morphologies, and most of them were less
than 1.5 micrometers in length and tens of nanometers in
width, with a large aspect ratio, which agreed well with the
XRD result. Obviously, α-, β-, and γ-MnO2 were 1D structural
materials, and α-MnO2 showed a needle-like shape, β-MnO2
Table 3 Particle morphology, surface area, pore volume, and crystal
size of MnO2 catalysts
BET surface
area (m2 g−1
Pore volume
Crystal size
(nm)
Sample
Morphology
)
(cm3 g−1
)
α-MnO2
β-MnO2
γ-MnO2
Needle-like
Rod-like
Urchin-like
22.12
13.06
20.54
0.124
0.075
0.112
22.1
45.7
11.3
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 5656–5664 | 5659

View Article Online
Paper
Green Chemistry
lysts.43 The broad O 1s peaks can be deconvoluted into peaks sumption were observed at 342 and 440 °C. As for γ-MnO2,
at 529.4–529.6 eV, 530.8–531.5 eV, and 533.1–533.7 eV, corres- strong and weak reduction peaks of H2 consumption were
ponding to the lattice oxygen (O2−) (denoted as OL); the observed at 300 and 405 °C. The strong and weak reduction
adsorbed oxygen O22−/O−, hydroxyl species and carbonate peaks represent the reduction of MnO2 to Mn3O4 and the sub-
2−
(CO3
)
(denoted as OA); the adsorbed molecular water sequent reduction of Mn3O4 to MnO, respectively.48,49 Obviously,
(denoted as OW), respectively.44 The surface oxygen species the reducibility of γ-MnO2 was better than that of β-MnO2,
ratios (OL/(OA + OL)) were calculated and are summarized in whereas for α-MnO2, two peaks of H2 consumption were
Table S3,† entries 1–3. Obviously, α-MnO2 (77.3%) had a observed at 294 °C and 307 °C; the former reduction peak may
higher ratio of OL/(OA + OL) than γ-MnO2 (68.9%) and β-MnO2 be attributed to the reduction of MnO2 to Mn2O3 (or Mn3O4),
(63.7%), which was in good accordance with the order of the and the latter was assigned to the final reduction to MnO. The
catalytic activity for the conversion of HMF to BMIF using the reduction temperature of α-MnO2 was obviously lower than
AOA reaction.
those of β-MnO2 and γ-MnO2, especially in the reduction stage
The mobility of oxygen species in these three MnO2 cata- of Mn2O3 or Mn3O4 to MnO. As a result, α-MnO2 possessed the
lysts was assessed by O2-TPD. As shown in Fig. 2d, the O2-TPD most reducibility. In general, a strong Mn–O bond also limits
profiles are divided into three stages: Low-Temperature-Stage the mobility of the lattice oxygen. The average bond lengths of
(LTS) peaks at <300 °C, Medium-Temperature-Stage (MTS) Mn–O in α-MnO2, β-MnO2 and γ-MnO2 are 1.98, 1.88, and
peaks at 300–600 °C and High-Temperature-Stage (HTS) peaks 1.91 Å, respectively,43 meaning that the Mn–O bond strength
at >600 °C, corresponding to the physically adsorbed oxygen decreases in the order of α-MnO2 < γ-MnO2 < β-MnO2, whereas
(LTS-oxygen), strong chemically adsorbed oxygen (MTS-oxygen) the reducibility is opposite. Taken together, the oxygen species
and its bulk lattice oxygen (HTS-oxygen), respectively.45,46 For mobility of the three kinds of MnO2 is consistent with the
all the MnO2 samples, the desorption of LTS-oxygen is similar, order of their catalytic activity, inferring that α-MnO2 has
and three small humps are observed at 85 °C (Fig. 2d inset). excellent lattice oxygen donating ability.
Besides, a weak peak at 515 °C and a strong peak at 740 °C
Furthermore, O2-TPO analysis was employed to investigate
were observed for α-MnO2, while two strong peaks centred at the oxidability of the recovered MnO2 after the reaction. As
530 °C and 760 °C for β-MnO2, and significant peaks at 500 °C shown in Fig. 2f, O2-TPO profiles of the recovered MnO2 are
and 780 °C for γ-MnO2 were observed. As is known, LTS- and also split into three stages, one below 250 °C is assigned to the
MTS-oxygen participates in the suprafacial reaction mecha- adsorption–desorption of O2 at low temperature. The latter two
nism, while HTS-oxygen is relevant for a Mars-van Krevelen stages at 250–600 °C and above 600 °C correspond to the oxi-
redox cycle.47 In the O2-TPD profiles, the intensity of the de- dation of Mn2+ to Mn3+ and the subsequent oxidation process
sorption peaks related to the chemically adsorbed oxygen fol- (Mn3+→Mn4+), respectively.42 The O2-TPO profiles showed that
lowed the tendency: β-MnO2 > γ-MnO2 ≫ α-MnO2. Based on the reduced state of Mn2+ was dominant in the recovered
the catalytic activities of these three MnO2 catalysts, obviously, β-MnO2 and γ-MnO2, however, the reduced state in the recov-
the chemically adsorbed oxygen of MnO2 is hardly involved in ered α-MnO2 was almost in the form of Mn3+, which suggested
the oxidation of MIFA to MIFF, unlike HTS-oxygen. Curiously, the redox cycling of Mn4+/Mn3+ for α-MnO2 as a catalyst, and
although the discrepancy in the HTS-oxygen desorption temp- the redox cycling of Mn4+/Mn3+/Mn2+ for β-MnO2 and γ-MnO2.
erature was not obvious, the HTS-oxygen desorption tempera- Obviously, the oxygen consumption required for the recovered
ture was not consistent with the catalytic activity, probably MnO2 followed the rule of β-MnO2 > γ-MnO2 ≫ α-MnO2,
caused by their different structure. The escape velocity of the meaning that α-MnO2 had a higher valence after the reaction
internal lattice-oxygen probably initially influenced by tunnel and a stronger oxygen coordination capacity compared to
sizes of 1 × 2 (2.3 Å × 4.6 Å) in γ-MnO2, until the collapse of γ-MnO2 and β-MnO2 under an air oxygen atmosphere. To
the crystal. While for β-MnO2 with tunnel sizes of 1 × 1 (2.3 Å × further confirm the valence state of the manganese species
2.3 Å), owing to the limitation of narrower tunnel sizes, the during the reaction, the recovered α-, β-, and γ-MnO2 after the
lattice-oxygen probably mainly desorbed for the crystal plane reaction under a N2 atmosphere (denoted as the used α-, β-,
at the initial desorption stage, resulting in a relatively lower de- and γ-MnO2 in N2) were analysed by XPS. As shown in
sorption temperature than that of γ-MnO2. More importantly,
the lattice oxygen ratio of MnO2 (integrated from the curve at
MTS and HTS, SHTS/SMTS) intuitively followed the order of the
catalytic activity: α-MnO2 > γ-MnO2 > β-MnO2, agreeing with
the results of XPS analysis. Accordingly, it is deduced that the
lattice oxygen ratio of MnO2 is a vital factor for the AOA reac-
tion of HMF to BMIF.
H2-TPR analysis was performed to investigate the relative
reducibility of these three MnO2 catalysts, as shown in Fig. 2e.
The green colour after H2-TPR (see Fig. S13†) indicated that
MnO2 was completely reduced to MnO, the final reduction
state. For β-MnO2, strong and weak reduction peaks of H2 con- α-MnO2 (reaction under a 1.0 MPa N2- or air-atmosphere).
Fig. 3 O 1s (a) and Mn 2p (b) spectra of the fresh and the recovered
5660 | Green Chem., 2021, 23, 5656–5664
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
Fig. S14,† the recovered α-MnO2 did only contain Mn4+ and
Mn3+, while a large amount of Mn2+ was observed in β-MnO2
and γ-MnO2, which further confirmed the above-mentioned
result of O2-TPO analysis.
To evaluate the oxygen coordination capacity of α-MnO2,
the recovered α-MnO2 (reaction under a 1 MPa N2- or air-atmo-
sphere, denoted as the used α-MnO2 in N2 and the used
α-MnO2 in air) was subjected to XPS analysis. The OL/(OA + OL)
ratio of the used α-MnO2 in N2 decreased from 77.3% to
62.9%, but that in air only dropped slightly (Fig. 3a). Similarly,
the proportion of Mn4+/Mn (denoted as RMn4+) for the used
α-MnO2 in N2 was dropped to 79% from 93%, whereas almost
no change was observed for the used α-MnO2 in air (Fig. 3b).
Generally, three crystalline MnO2 catalysts are constructed
from the MnO6 octahedron in different bonding ways, and the
dimensions of the channels (2 × 2) (4.6 Å × 4.6 Å) in α-MnO2
are larger than those in β-MnO2 (1 × 1, 2.3 Å × 2.3 Å) and
γ-MnO2 (2 × 1, 4.6 Å × 2.3 Å); moreover, the molecule diameter
of oxygen is 0.35 nm, which is conducive to the shift of the
surface oxygen into the lattice of α-MnO2.29 These results
demonstrated that the lattice oxygen of α-MnO2 certainly drove
the oxidation of the hydroxyl group to the aldehyde group and
was easily recovered under an oxygen atmosphere.
In summary, from the results of XPS, O2-TPD, H2-TPR, and
O2-TPO analyses, combining the catalytic performance of these
three MnO2 catalysts, it is reasonably concluded that the
lattice oxygen in MnO2 played a key role in the oxidation of
HMF by the redox cycling of Mn4+/Mn3+ for α-MnO2, and the
redox cycling of Mn4+/Mn3+ and Mn3+/Mn2+ for β-MnO2 and
γ-MnO2; moreover, α-MnO2 exhibited the better mobility of
lattice oxygen.
Scheme 3 A plausible mechanistic pathway for the AOA transformation
of HMF into BMIF with α-MnO2 under air (Ovac and Olat represent
oxygen vacancy and lattice oxygen).
group by the lattice oxygen, accompanying the reduction of
Mn4+ to Mn3+. The resultant MIFF is dissociated from the
surface of MnO2, followed by a further amination to obtain
BMIF. Finally, Mn3+ is oxidized to Mn4+ by the adsorbed oxygen
on the surface of the catalyst, resulting in the recovery of MnO2.
Leaching and reusability tests of α-MnO2
In order to investigate the stability of these three MnO2 cata-
lysts in the AOA reaction of HMF to BMIF, the recycling tests of
the recovered MnO2 were conducted. The catalyst was separ-
ated from the reaction mixture by filtration after the reaction,
and then dried at 60 °C for 2 h before reuse in the next cycle.
Excitingly, the catalytic activity of the recovered α-MnO2 (97.5%
yield) still retained a high level (Fig. 4), whereas a significant
decrease for the recovered β-MnO2 (42.7% yield) and γ-MnO2
(67.3% yield) was observed (Table 1, entries 12–14). Then, the
recovered α-MnO2, β-MnO2 and γ-MnO2 after the reactions
(Table 1, entries 12–14) were characterized by XRD (Fig. S15†).
It was found that only α-MnO2 still retained its original crystal-
line structure (Fig. S16a†), and the Mn–O and Mn–O–Mn
bonds in the fingerprint area showed hardly any change
(Fig. S16b†), even after the fifth cycle, while structural collapse
was observed for β-MnO2 and γ-MnO2. These results showed
that the structural stability of α-MnO2 also was a crucial factor
for its excellent performance in the AOA reaction of HMF
to BMIF.
Based on the results described above, the catalytic pathway
for the AOA reaction of HMF over α-MnO2 was proposed as
shown in Scheme 3. Initially, the aldehyde group attached to
HMF spontaneously reacts with methylamine to form MIFA.
Meanwhile, the hydroxyl group in MIFA is adsorbed on the
surface of MnO2 and subsequently oxidized to the aldehyde
To further investigate the stability of α-MnO2, the leaching
test was carried out after the reaction for 1 h. No further con-
version of HMF was observed even prolonging the reaction
time to 5 h (Fig. 1), demonstrating that the active species did
not leach into the reaction solution.
Reduction of BMIF to BMAF
Fig. 4 Recycling tests of α-MnO2 on the conversion of HMF to BMIF.
Reaction conditions: 0.5 mmol HMF (63 mg), 5.0 mL of 30 wt% methyl-
amine methanol solution, 22 mg of fresh or recovered α-MnO2, 1.0 MPa
air, 90 °C, 3 h.
To realize the synthesis of BMAF from HMF using a two-step
reaction in one pot (Scheme 4), several commercial catalysts,
such as RANEY®-Ni, RANEY®-Co, Pt/C, Pd/C and Ru/C, were
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 5656–5664 | 5661

View Article Online
Paper
Green Chemistry
characteristic diffraction peak of Ru was observed for the fresh
Ru/C catalyst, indicating good dispersion of Ru. However, the
recovered Ru/C catalyst showed a few peaks associated with
Ru, which may be caused by the agglomeration of Ru, but the
catalytic performance of the recovered Ru/C catalyst was hardly
affected (Fig. S18†).
Scheme 4 Synthesis of BMAF from HMF using a two-step reaction in
one pot. Reaction conditions: step A, 0.5 mmol HMF (63 mg), 5.0 mL of
30 wt% methylamine methanol solution, 22 mg of α-MnO2, 1.0 MPa air,
90 °C, 3 h; step B, 5.0 mL of liquid products from step A, 1.0 mL of
30 wt% methylamine methanol solution, 45 mg of Ru/C, 2.0 MPa H2,
130 °C, 3 h.
Scope of the AOAR reaction
Finally, several common amines including ammonia, primary
amine and secondary amine were examined in the investigated
reaction to synthesize other furan-diamines, as shown in
Table 5. Almost all amines employed in this reaction obtained
Table 4 Catalytic reduction of BMIF over different catalystsa
a
satisfactory yield of the corresponding imine, except
ammonia and secondary amine (Table 5, entries 1–8). Owing
to the high activity of the primary imine, the imines generated
during the reaction of HMF and ammonia easily underwent
polymerization to form polymers (Fig. S19a†).20,25 Generally,
the reaction of a secondary amine with a carbonyl group gener-
ated the corresponding enamine compound.36 However, when
using HMF as a carbonyl-containing compound, the corres-
ponding enamine was not formed due to no hydrogen at the
2-position of the furan ring (Fig. S19b†). When HMF reacted
with a primary amine, owing to no hydrogen at the secondary
imine group and 2-position of the furan ring in HMF, the
nucleophilic reaction of the aldehyde group with the secondary
imine group hardly took place. Therefore, the secondary furan
Entry
Catalyst
Time (h)
Conv. (%)
Yield (%)
1
2
3
4
RANEY®-Ni
RANEY®-Co
Pt/C (5% Pt)
Pd/C (5% Pd)
Ru/C (5% Ru)
Ru/C
3
3
4
4
3
3
3
100
100
100
100
100
100
100
64.7
88.7
12.3
8.8
96.1
94.7
95.4
5
6b
7c
Ru/C
a Reaction conditions: 5.0 mL of liquid products from the AOA reaction
by α-MnO2, 1.0 mL of 30 wt% methylamine methanol solution, 45 mg
of catalyst, 2.0 MPa H2, 130 °C. b The catalyst was recovered and reused
for the 1st run. c The catalyst was recovered and reused for the 3rd run.
examined for the reduction of BMIF after the filtration of diimine is stable as an intermediate or product in the investi-
α-MnO2 (Table 4, entries 1–5). Owing to the reversible reaction gated reaction. So, the protocol to synthesize furan diamines
of an aldehyde group and amine to generate an imine, the proposed in this study is suitable for secondary furan
addition of methylamine is conducive to the stability of diamines.
imines. Satisfyingly, Ru/C performed well in the reduction of
Emphasizedly, it is difficult to efficiently synthesize the
BMIF to BMAF with a yield of 96.1% (Fig. S17†).
furan primary diamines from DFF by direct reductive amin-
To examine the stability of the catalyst, Ru/C was recovered ation. Recently, Qi et al.23 developed a novel method to syn-
after the reaction for the next run. Encouragingly, no obvious thesize 2,5-bis(aminomethyl)furan from DFF mediated by sec-
decrease in the catalytic performance for the conversion of ondary diimines with the assistance of trans-imination with
BMIF to BMAF was observed after the third cycle (Table 4, ammonia. Combining the protocol of an efficient synthesis of
entries 6 and 7). In the XRD profiles of Ru/C, no obvious BMIF from HMF using a one-pot amination–oxidation–amin-
Table 5 Preparation of different furan diimines and diaminesa
Entry Substrate
Step A-diimine
YieldA (%) Step B-diamine
H2 (MPa) Time (h) Conv. (%) YieldBc (%)
1
Ammonia
0
—
—
—
5
—
—
—
87
97
96
2
—
0
—
—
3b
92
99
99
0.3
0.5
0.5
100
100
100
4
4
5
4
6
99
99
85
0.5
0.5
0.2
4
4
6
100
100
82
96
98
53
7
8b
a Reaction conditions: step A: 0.5 mmol HMF, 5.0 mL of 30 wt% substrate methanol solution, 22 mg of α-MnO2, 1.0 MPa air, 90 °C, 3 h. The con-
versions were 100%. Step B: 5.0 mL of liquid products from step A, 31 mg of Ru/C, 110 °C. In all reactions, hexadecane was utilized as the
internal standard, and the conversions and yields were determined by GC analysis. b Step A: 5.0 mL of 10 wt% substrate methanol solution.
c Furan ring and benzene ring can be reduced.
5662 | Green Chem., 2021, 23, 5656–5664
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
ation reaction in our study with Qi’s study, an alternative strat-
egy for the synthesis of primary furan diamines from HMF by
trans-imination is promising (Fig. S19c†).
T. H. Riermeier, S. Seebald, T. Tacke and H. Trauthwein,
Appl. Catal., A, 2005, 280, 17–46.
7 V. Froidevaux, C. Negrell, S. Caillol, J. Pascault and
B. Boutevin, Chem. Rev., 2016, 116, 14181–14224.
8 A. Mittal, H. M. Pilath and D. K. Johnson, Energy Fuels,
2020, 34, 3284–3293.
9 W. Hao, W. Li, X. Tang, X. Zeng, Y. Sun, S. Liu and L. Lin,
Green Chem., 2016, 18, 1080–1088.
10 T. Wang, M. W. Nolte and B. H. Shanks, Green Chem., 2014,
16, 548–572.
11 X. Kong, Y. Zhu, Z. Fang, J. A. Kozinski, I. S. Butler, L. Xu,
H. Song and X. Wei, Green Chem., 2018, 20, 3657–3682.
12 J. Du, J. Zhang, Y. Sun, W. Jia, Z. Si, H. Gao, X. Tang,
X. Zeng, T. Lei, S. Liu and L. Lin, J. Catal., 2018, 368, 69–78.
13 A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and
C. A. M. Afonso, Green Chem., 2011, 13, 754–793.
14 I. Delidovich, P. J. C. Hausoul, L. Deng, R. Pfützenreuter,
M. Rose and R. Palkovits, Chem. Rev., 2016, 116, 1540–1599.
15 L. T. Mika, E. Cséfalvay and Á. Németh, Chem. Rev., 2018,
118, 505–613.
Conclusions
In conclusion, an amination–oxidation–amination–reduction
strategy was proposed to synthesize BMAF from HMF via the
intermediate BMIF using a two-step reaction in one-pot. The
crystal structure of MnO2 had a great effect on the catalytic per-
formance for the oxidation of MIFA to MIFF. Meanwhile
α-MnO2 retained high stability and activity during the reaction.
The higher lattice oxygen ratio and better mobility endowed
α-MnO2 with excellent catalytic performance in the oxidation
of MIFA to MIFF by the redox cycling of Mn4+/Mn3+, facilitating
the conversion of HMF to BMIF, while the redox cycling of
Mn4+/Mn3+ and Mn3+/Mn2+ was required for β-MnO2 and
γ-MnO2 to finish the oxidation of MIFA to MIFF. Eventually,
BMAF was achieved with a yield of 96.1% by the reduction of
BMIF over Ru/C under a H2-atmosphere after the filtration of
α-MnO2, realizing the synthesis of BMAF from HMF using a
two-step reaction in one pot. More importantly, the protocol of
an efficient synthesis of BMIF from HMF using a one-pot
amination–oxidation–amination reaction provides an alterna-
tive strategy for the synthesis of primary furan diamines from
HMF by trans-imination.
16 Y. Xu, X. Jia, J. Ma, J. Gao, F. Xia, X. Li and J. Xu, Green
Chem., 2018, 20, 2697–2701.
17 J. He, L. Chen, S. Liu, K. Song, S. Yang and A. Riisager,
Green Chem., 2020, 22, 6714–6747.
18 D. Pingen, J. B. Schwaderer, J. Walter, J. Wen, G. Murray,
D. Vogt and S. Mecking, ChemCatChem, 2018, 10, 3027–
3033.
19 K. Zhou, H. Liu, H. Shu, S. Xiao, D. Guo, Y. Liu, Z. Wei and
X. Li, ChemCatChem, 2019, 11, 2649–2656.
20 T. Komanoya, T. Kinemura, Y. Kita, K. Kamata and
M. Hara, J. Am. Chem. Soc., 2017, 139, 11493–11499.
21 H. Yuan, B. T. Kusema, Z. Yan, S. Streiff and F. Shi, RSC
Adv., 2019, 9, 38877–38881.
Conflicts of interest
There are no conflicts to declare.
22 J. J. Roylance and K. Choi, Green Chem., 2016, 18, 5412–
5417.
Acknowledgements
23 H. Qi, F. Liu, L. Zhang, L. Li, Y. Su, J. Yang, R. Hao,
A. Wang and T. Zhang, Green Chem., 2020, 22, 6897–6901.
24 Q. Girka, N. Hausser, B. Estrine, N. Hoffmann, J. Le Bras,
S. Marinković and J. Muzart, Green Chem., 2017, 19, 4074–
4079.
This work was supported by the National Natural Science
Foundation of China (No. 21776234, 21978246) and the
research funds from Hubei Key Laboratory of Pollutant
Analysis & Reuse Technology (Hubei Normal University).
25 N. Le, A. Byun, Y. Han, K. Lee and H. Kim, Green
Sustainable Chem., 2015, 05, 115–127.
Notes and references
26 X. Wang, W. Chen, Z. Li, X. Zeng, X. Tang, Y. Sun, T. Lei
and L. Lin, J. Energy Chem., 2018, 27, 209–214.
27 W. Jia, Z. Si, Y. Feng, X. Zhang, X. Zhao, Y. Sun, X. Tang,
X. Zeng and L. Lin, ACS Sustainable Chem. Eng., 2020, 8,
7901–7908.
1 J. M. García, F. C. García, F. Serna and J. L. de la Peña,
Prog. Polym. Sci., 2010, 35, 623–686.
2 H. Zeng and Z. Guan, J. Am. Chem. Soc., 2011, 133, 1159–
1161.
3 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 28 E. Hayashi, Y. Yamaguchi, K. Kamata, N. Tsunoda,
2411–2502.
Y. Kumagai, F. Oba and M. Hara, J. Am. Chem. Soc., 2019,
141, 890–900.
29 J. Nie and H. Liu, J. Catal., 2014, 316, 57–66.
4 C. Moreau, M. N. Belgacemb and A. Gandini, Top. Catal.,
2004, 27, 11–30.
5 K. Murugesan, T. Senthamarai, V. G. Chandrashekhar, 30 F. Bertinchamps, M. Treinen, P. Eloy, A. M. Dos Santos,
K. Natte, P. C. J. Kamer, M. Beller and R. V. Jagadeesh,
Chem. Soc. Rev., 2020, 49, 6273–6328.
M. M. Mestdagh and E. M. Gaigneaux, Appl. Catal., B, 2007,
70, 360–369.
6 B. Chen, U. Dingerdissen, J. G. E. Krauter, H. G. J. Lansink 31 J. Zhao, A. Jayakumar and J. Lee, ACS Sustainable Chem.
Rotgerink, K. Möbus, D. J. Ostgard, P. Panster,
Eng., 2018, 6, 2976–2982.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 5656–5664 | 5663

View Article Online
Paper
Green Chemistry
32 X. Jia, J. Ma, M. Wang, Z. Du, F. Lu, F. Wang and J. Xu, 41 N. Sui, Y. Duan, X. Jiao and D. Chen, J. Phys. Chem. C,
Appl. Catal., A, 2014, 482, 231–236. 2009, 113, 8560–8565.
33 P. Hu, Y. Chen, X. Yan, W. Lang and Y. Guo, Ind. Eng. 42 F. Gao, X. Tang, H. Yi, C. Chu, N. Li, J. Li and S. Zhao,
Chem. Res., 2019, 58, 4065–4073. Chem. Eng. J., 2017, 322, 525–537.
34 W. Jia, J. Du, H. Liu, Y. Feng, Y. Sun, X. Tang, X. Zeng and 43 S. Liang, F. Teng, G. Bulgan, R. Zong and Y. Zhu, J. Phys.
L. Lin, J. Chem. Technol. Biotechnol., 2019, 94, 3832–3838. Chem. C, 2008, 112, 5307–5315.
35 T. Xiang, X. Liu, P. Yi, M. Guo, Y. Chen, C. Wesdemiotis, 44 M. Kang, E. D. Park, J. M. Kim and J. E. Yie, Appl. Catal., A,
J. Xu and Y. Pang, Polym. Int., 2013, 62, 1517–1523. 2007, 327, 261–269.
36 S. Gomez, J. A. Peters and T. Maschmeyer, Adv. Synth. 45 X. Liao, J. Hou, Y. Wang, H. Zhang, Y. Sun, X. Li, S. Tang,
Catal., 2002, 344, 1037–1057.
37 W. Chen, Y. Sun, J. Du, Z. Si, X. Tang, X. Zeng, L. Lin, S. Liu
K. Kato, M. Yamauchi and Z. Jiang, Green Chem., 2019, 21,
4194–4203.
and T. Lei, J. Chem. Technol. Biotechnol., 2018, 93, 3028– 46 S. Rong, P. Zhang, F. Liu and Y. Yang, ACS Catal., 2018, 8,
3034. 3435–3446.
38 B. Zhao, R. Ran, X. Wu and D. Weng, Appl. Catal., A, 2016, 47 M. R. Morales, B. P. Barbero and L. E. Cadús, Fuel, 2008,
514, 24–34. 87, 1177–1186.
39 C. Yuan, L. Hou, L. Yang, D. Li, L. Shen, F. Zhang and 48 J. Carnö, M. Ferrandon, E. Björnbom and S. Järås, Appl.
X. Zhang, J. Mater. Chem., 2011, 21, 16035–16041. Catal., A, 1997, 155, 265–281.
40 Y. Xie, Y. Yu, X. Gong, Y. Guo, Y. Guo, Y. Wang and G. Lu, 49 R. Xu, X. Wang, D. Wang, K. Zhou and Y. Li, J. Catal., 2006,
CrystEngComm, 2015, 17, 3005–3014. 237, 426–430.
5664 | Green Chem., 2021, 23, 5656–5664
This journal is © The Royal Society of Chemistry 2021