Hydrogen-Binding-Initiated Activation of O?H Bonds on a Nitrogen-Doped Surface for the Catalytic Oxidation of Biomass Hydroxyl Compounds

Article abstract of DOI:10.1002/anie.202103604

Hydrogen binding of molecules on solid surfaces is an attractive interaction that can be used as the driving force for bond activation, material-directed assembly, protein protection, etc. However, the lack of a quantitative characterization method for hydrogen bonds (HBs) on surfaces seriously limits its application. We measured the standard Gibbs free energy change (ΔG⁰) of on-surface HBs using NMR. The HB-accepting ability of the surface was investigated by comparing ΔG⁰ values employing the model biomass platform 5-hydroxymethylfurfural on a series of Co-N-C-n catalysts with adjustable electron-rich nitrogen-doped contents. Decreasing ΔG⁰ improves the HB-accepting ability of the nitrogen-doped surface and promotes the selectively initiated activation of O?H bonds in the oxidation of 5-hydroxymethylfurfural. As a result, the reaction kinetics is accelerated. In addition to the excellent catalytic performance, the turnover frequency (TOF) for this oxidation is much higher than for reported non-noble-metal catalysts.

Full text of DOI:10.1002/anie.202103604

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: Hydrogen Binding Initiated Activation of O−H Bonds on a
Nitrogen-Doped Surface for Catalytic Oxidation of Biomass
Hydroxyl Compounds
Authors: Xin Liu, Yang Luo, Hong Ma, Shujing Zhang, Penghua Che,
Meiyun Zhang, Jin Gao, and Jie Xu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202103604
Link to VoR: https://doi.org/10.1002/anie.202103604

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
Hydrogen Binding Initiated Activation of O−H Bonds on a Nitrogen-
Doped Surface for Catalytic Oxidation of Biomass Hydroxyl
Compounds
Xin Liu,[a], [b] Yang Luo, Hong Ma,* Shujing Zhang,
[a]
[a]
[a], [b]
Penghua Che, Meiyun Zhang,
[a]
[a], [b]
Jin Gao,[a]
and Jie Xu*[a]
[
a]
Xin Liu, Yang Luo, Hong Ma, Shujing Zhang, Penghua Che, Meiyun Zhang, Jin Gao, Jie Xu
State Key Laboratory of Catalysis
Dalian Institute of Chemical Physics, Chinese Academy of Sciences
Dalian National Laboratory for Clean Energy
457 Zhongshan Road, Dalian 116023 (P. R. China)
E-mail: xujie@dicp.ac.cn
mahong@dicp.ac.cn
[
b]
Xin Liu, Shujing Zhang, Meiyun Zhang
University of Chinese Academy of Sciences
Beijing 100049 (P. R. China)
Supporting information for this article is given via a link at the end of the document.
Abstract: Hydrogen binding of molecules on specific solid surface is
pathways in purpose. For example, the selective activation of the
primary hydroxyl group in diol could be accomplished via forming
O−H···O=S HBs with DMSO as a proton acceptor, which was in
favor of the selective alkoxyl radical transformation, rather than
occurring β-scission due to the unselective activation of both two
hydroxyl groups.[ Han et al. presented that the HBs between
the alcohol hydroxyl groups and ionic liquid [EMIM]OAc could
accelerate the rate-determining step of hydroxyl groups activation,
thus significantly promoting the oxidative esterification of inert
aliphatic alcohols. In the case of heterogeneous catalysis,
researchers also found that the surface-grafted nitrogen-
an attractive interaction that can be employed as driving force for
chemical bond activation, material directed assembly, protein
protection, etc. However, the lack of quantitative characterization
methods for hydrogen bonds (HBs) on surface seriously limits its
application. Herein, we measured the standard Gibbs free energy
5]
[6]
0
change (ΔG ) of surface HBs using NMR technique. HBs accepting
0
ability of surface was investigated in term of comparing ΔG values by
employing model biomass platform 5-hydroxymethylfurfural on a
series of Co-N-C-n catalysts with electron-rich doped-nitrogen
0
contents adjusted. Reducing the ΔG effectively improves HBs
accepting ability of the nitrogen-doped surface, and promotes the
O−H bonds selectively initiated activation in the oxidation of 5-
hydroxymethylfurfural. As a result, the reaction kinetics is accelerated
and the rate constant is significantly increased. In addition to excellent
catalytic performance, the turnover frequency (TOF) value for this
oxidation is extremely higher than the reported non-noble metal
catalysts.
2 3
containing ligands on EPy/Pt/Al O catalyst could form HBs with
specific hydroxyl groups.[7] However, the research of HBs on
heterogeneous surface mainly concentrated on the modification
of catalyst surface, promotion of catalytic performance, and
elimination of negative impacts via pre-converting −OH.[8] There
is lack of experimental methods beneficial for understanding the
regularity of HBs on surface, which still remains in the stage of
exploration.
Nuclear magnetic resonance spectroscopy (NMR) is regarded
as a quite appropriate technology to investigate the hydrogen
binding interaction, since the proton chemical shift (δ) is extremely
sensitive to the change of electron density around the hydrogen
nucleus. It is widely accepted that the HBs between hydroxyl
groups and high electron-rich acceptors can generate the
depletion of electron density near the hydroxyl proton and
deshielded nuclei, so that the hydrogen-bonded protons reveal a
downfield shift in the resonance frequency in different degrees.
Normally, researchers intuitively judge the formation of HBs
directly in accordance with the chemical shift changes, or
compare the relative strength of HBs by the relative size of
downfield shift.[10] For example, the chemical shift of the hydroxyl
Introduction
The hydroxyl group (−OH) belongs to one of the essential
building blocks in various types of chemicals, materials, protein,
etc.[1] In the chemical transformation of most hydroxyl groups, how
to realize the highly efficient activation of O−H bonds becomes a
critial point.[2] Generally, the proton donating ability of hydroxyl
groups contributes to hydrogen bonds (HBs) formed by
themselves, which can usually lower their reactivity.[3]
Interestingly, however, this characteristic also endows HBs with
the ability to facilitate activation of O−H bonds, through forming
strengthened HBs with other proton acceptors such as electric-
rich heteroatoms and halogen ions.[4] The decreased electron
cloud density near the proton nucleus might weaken O−H bonds,
thus enhancing the reactivity of hydroxyl groups. Recently, some
explorative studies have been performed in utilizing HBs for
promoting the conversion of inert hydroxyl substrates and guiding
[9]
proton of benzyl alcohol presented a gradual downfield shift with
the increase of [EMIM]OAc ILs content, which was mainly
-
ascribed to the formation of O−H···OAc HBs between hydroxyl
groups and the acetate anions.[6] Researchers have been
attempting to find out the relationship between the chemical shift
change (Δδ) and HBs energies, and the present studies still
1
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
mainly depend on empirical equation derivations or DFT
calculations.[11] For example, Schaefer[12] proposed a linear
relationship between DFT-calculated HBs energy and the Δδ of
hydrogen-bonded protons in some ortho-substituted phenol
derivatives (Δδ = −0.4 ± 0.2 + EHB). As a commonly accepted
approach, the Benesi-Hildebrand equation was derived, for
improve the reactivity of feedstock (Scheme 1c). For this purpose,
the key point is to form O−H···X HBs on X sties with sufficient
capacity in accepting protons, and find the corresponding criteria.
Based on the theory in our previous research,[ we proposed the
thermodynamic model for the HBs formation process on surface.
Similarly, the formation of O−H···X HBs is a reversible process in
equilibrium with its backward association reaction. The KHB is the
association constant of HBs donors and acceptors. The important
physical parameter ΔG0 for the formation process of HBs, is
expected to determine the criteria of whether the −OH could form
surface HBs. By performing concentration-variable experiments
17]
calculating association constants (K
benzenethiol and nitrogen contained acceptors (1/Δδ
/(K
) + 1/Δδmax).[13] Accordingly, the HBs energies could
be reflected by K and the resulting Gibbs free energy change
ΔG).[14] Besides chemical shift, relaxation time is another NMR
a
)
of HBs between
=
1
a 0
Δδmax[A]
a
(
parameter that is suitable to reflect binding states of protons.
Recently, D'Agostino et al.[15] demonstrated that the relative
surface affinities could be exhibited by adsorbates of primary
alcohols and cyclohexane within liquid-saturated mesoporous
catalysts by measuring the ratio of NMR relaxation time constants
T1:T2. Despite great efforts have been made for HBs in
homogeneous and heterogeneous catalytic systems, it still lacks
experimental methods for measuring the HBs energy on the
catalyst surface.
of proton acceptors, we studied the establishment of
a
0
quantitative measurement method for ΔG . In addition, this study
proposed a linear equation by determining the change of hydroxyl
proton chemical shifts (Δδ) with the variation of the concentration
of proton acceptors (C), 1/Δδ = a/C + b, by which KHB and ΔG0
could be achieved based on the ratio of intercept to slope (KHB
=
0
b/a, ΔG = −RTln(b/a)). The detailed derivation process can be
observed in our previous work[17] and the explanation of this
formula is provided in the Supporting Information.
Our laboratory has been investigating the catalytic conversion
of biomass and its derivatives as well as trying to deepen the
understanding of the role of HBs in hydroxyl groups and purposely
make use of them.[16] Quite recently, we determined the EHB of
HBs between donors and acceptors by conducting 1H NMR
experiments, and put forward for the first time a linear equation as
a function of chemical shifts (δ) (lnδ + σ
δ
= −EHB/RT + A).[17] These
bases make it feasible to determine whether an acceptor has
sufficient capacity of forming new HBs with a donor by measuring
EHB. In the case of the heterogeneous catalyst, if catalyst surface
was endowed with sufficient capacity in accepting protons,
hydroxyl groups might form new O−H···X HBs on such surface.
Thus, O−H bonds might be weakened and activated, aiming to
enhance the reactivity of raw substrates and promote the catalytic
conversions. Herein, an equation was proposed to determine the
0
ΔG and association constant (KHB) for the formation of O−H···X
HBs on surface. Through the comparison of ΔG0 and KHB, the
proton acceptance ability of catalysts in forming HBs on surface
can be directly compared. As a result, Co-N-C-9.6 catalyst with
the most positive ΔG0 = −11.53 kcal/mol among Co-N-C-n
catalysts exhibited unique ability in initiated activation of O−H
bonds and afforded a turnover frequency (TOF) of 477 mol HMF
mol-1 Co h , which was extremely higher than the reported non-
noble metal catalysts and even comparable to noble metal
catalysts in probing the oxidation of 5-hydroxymethylfurfural
-1
Scheme 1 Schematic illustration for the thermodynamic model of the reversible
association and dissociation process for O−H···X HBs on surface (a). Design
for model catalysts as protons acceptors (b). A model oxidation reaction for
HBs-initiated activation of O−H bonds using HMF as a donor (c).
(
HMF) to 2,5-furandicarboxylicacid (FDCA).
As illustrated in Scheme 1b, Co-N-C, atomically dispersed
Results and Discussion
cobalt on N-doped carbon, was selected as
a model
heterogeneous catalyst after the careful consideration. The
reasons are presented as follows. Nitrogen is electron-rich that
will pull the electron toward it and may play the role of proton-
acceptor owing to its basic lone pair of electrons. Cobalt exists in
the form of single atom, which can guarantee that the catalytic
performance is not affected by the distribution of particle size, but
is mainly contributed by surface HBs. As illustrated in Scheme 1c,
we took the important biomass platform HMF as the model
hydroxyl compound, and its oxidation to produce FDCA as a
model reaction in order to investigate the HBs-initiated activation
of O−H bonds. Interestingly, the oxidation of hydroxyl group of
HMF at the initial step is extremely crucial for the whole reaction,
Design for surface HBs initiated-activation of O−H bonds.
At first, a strategy of hydrogen binding initiated-activation of O−H
bonds on the catalyst surface was proposed. Scheme 1a
illustrated the thermodynamic model for the formation process of
O−H···X HBs between −OH and the surface proton acceptor sites
(
X). X is expected to be electron-rich and capable of endowing the
proton acceptance ability to the catalyst by introducing electron-
rich atoms or species on surface. Due to the strong electrophilicity
of X, the O−H bond strength of hydrogen bonded hydroxyl group
is weakened, and it becomes easier to break in the process of
hydroxyl group conversion, so as to realize its activation and
2
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
since its untimely consumption always lead to polymerization,
thus inhibiting the whole process.[18] Therefore, the initiated
activation of O−H bonds by HBs could be probed by comparing
1h), Co-N-C-9.6 exhibited a main peak at 1.4 Å, corresponding to
the Co-N first coordination shell. No obvious Co-Co peak (2.17 Å)
and Co-O peak (2.57 Å) could be observed,[20] suggesting that
0
the KHB and ΔG of the HBs on Co-N-C catalysts surface by
x
there was no metallic Co or CoO species in Co-N-C-9.6, which
employing the above established method, and observing the
enhancement of TOF values, which is dramatically influenced by
the improvement of the −OH reactivity, as well as the producibility
to FDCA. Moreover, a high TOF value and productivity of FDCA
can reflect the improvement of the activation effect of O−H bonds
by HBs initiation.
was in consistence with the AC HAADF-STEM and XRD results.
The quantitative coordination of Co and N in the Co-N-C-9.6 is
configured by EXAFS fitting (Figure S2 and Table S2), that is one
4
Co atom coordinated with four N atoms (CoN single site).
By performing N1s XPS analysis, the states of the doped
nitrogen in Co-N-C-9.6 were explored (Figure 1i). Four
overlapped peaks demonstrated the forms of nitrogen include
pyridinic N/pyridinic N-Co (398.4 eV), pyrrolic N (399.9 eV),
graphitic N (401.0 eV), and pyridine N-oxide (403.0 eV).[21] Co-N-
C-n catalysts have close distributions of the doped nitrogen
species on surface (Figure S3 and Table S3), and achievement
of specific nitrogen species sites alone on Co-N-C-n surface
remains a challenge. Considering the high dispersion of nitrogen
element on the Co-N-C-n surface according to the EDX analysis
and extremely close distributions of doped nitrogen species, the
total nitrogen content was used to represent the apparent
concentration of HBs acceptors during the data fitting in
calculation of the KHB and ΔG0.
Characterization of model catalysts. Co-N-C was designed
and synthesized through annealing process by utilizing g-C N as
3 4
the nitrogen-rich precursor, as the model catalyst with the
electron-rich nitrogen species modified the surface to introduce
proton acceptance ability (Figure S1). The as-synthesized
samples were denoted as Co-N-C-n (n is the molar ratio of
nitrogen to cobalt in the catalyst). A series of characterizations
have been conducted in detail to examine the structure and
element states. On the representative Co-N-C-9.6, no metallic
cobalt or cobalt oxide nanoparticles were detected as confirmed
by TEM (Figures 1a-b) and XRD (Figure 1f). The XRD pattern only
displayed two broad carbon peaks centered at 25 and 43 degrees,
assigned to carbon with a higher degree of graphitization.[19] The
AC HAADF-STEM evidenced that the cobalt was predominantly
in the form of atomically dispersed sites (Figure 1c). Furthermore,
the EDX maps of Co-N-C-9.6 illustrated the highly homogeneous
distribution of Co, N and O on the carbon matrix in the entire
region (Figures 1d-e).
Measurement of the surface HBs energy. Afterwards, the
HBs energy of O−H···X on the nitrogen-doped surface of the
1
model catalyst were determined. The concentration-variation H
NMR measurements were performed in the poor HBs-accepting
solvent CDCl
3
, by taking HMF as a probe molecule with the
-1
concentration fixed at 0.41 mol·L , and varying the
-1
concentrations of Co-N-C-n catalysts from 0 to 3.04 g·L .
Tetramethylsilane (TMS) was employed as the internal standard.
Typically, the suspension of uniform dispersed catalyst was mixed
thoroughly and immediately determined within 4 minutes. All the
1
H NMR spectra displayed well resolved proton peaks, and a few
partially overlapped that do not hinder the observation of the
proton peak position. In the presence of Co-N-C-n, although the
signals were broadened to a certain extent, the hydroxyl proton
peaks could still be clearly distinguished. To guarantee the
accuracy and repeatability of the NMR experiment, each sample
was measured in succession for three times. The low standard
deviation verified that the suspension could maintain a uniform
dispersion state during the whole NMR measurement process,
and the surface HBs formation process attained equilibrium.
Figure 1 TEM images (a and b), AC HAADF-STEM image (c), HAADF STEM
image (d), corresponding EDX maps for the overlapped Co, N, O and C (e),
XRD pattern (f), normalized XANES spectra (g), EXAFS spectra (h), and N1s
XPS spectrum (i) of Co-N-C-9.6.
Figure
2
Concentration-variation 1H NMR spectra of HMF with different
concentrations of Co-N-C-9.6 (a), the model of hydrogen bonded HMF with an
acceptor of X (b), and locally amplified 1H NMR spectra of HMF with different
XANES and EXAFS were used to investigate the detailed
electronic structure of Co and the coordination environment of Co
and N in Co-N-C-9.6 (Figures 1g-h). The valence state of the Co
population in Co-N-C-9.6 was +2, as evidenced that Co-N-C-9.6
and CoO had almost the same E value (the first inflection point)
0
of 7720 eV (Table S1) in comparison with the cobalt foil and cobalt
oxides in the reference.[20] As shown in the EXAFS spectra (Figure
concentrations of pyridine (c) (400 MHz, CDCl
1
3
).
3
Typical H NMR spectra of HMF and Co-N-C-9.6 in CDCl are
displayed in Figure 2a. After adding Co-N-C-9.6 into HMF in
CDCl , an obvious change in proton chemical shift value was
observed for the hydroxyl proton (H1). For example, the hydroxyl
3
3
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
proton chemical shift appeared 0.1138 ppm downfield at 3.04 g·L-
3
to lower field with increasing catalyst concentration in CDCl at
1
concentration of Co-N-C-9.6 in CDCl
3
, implying that a formation
different speeds, while the chemical shifts of the rest of protons
(H2-H5) remained unchanged (Figure S5). Increasing the doped
nitrogen content of Co-N-C-n could make the chemical shift move
downfield more obviously. It is worthy of note that the downfield
shift of the hydroxyl proton of HMF was also exhibited with the
addition of different concentrations of N-C catalyst (Figure 3e).
However, slight movements were observed when using nitrogen-
free Co/CB catalyst (Figure 3f), verifying that the nitrogen is
indispensable for endowing proton acceptable ability to the Co-N-
C-n catalysts, while there exists no obvious correlation between
the chemical shift change value and the cobalt contents in the four
Co-N-C-n catalysts.
of O−H···X HBs resulted in the proton deshielded (Figure 2b). The
Δδ values increased with the increase of catalyst concentration,
which is similar to our previously reported studies concerning
homogeneous donor-acceptor systems.[17] In contrast, there is no
change in the rest of proton chemical shifts in C−H of
hydroxymethyl group (H2), the furan ring (H3 and H4), and formyl
group (H5), because carbons in such groups are not sufficiently
electron-rich in the formation of C−H···X HBs, as well as the new
formed O−H···X bonds have negligible influence on them.
Considering that besides HBs, the decrease in HMF
concentration may also generate a change of chemical shift of
hydroxyl protons, verification experiments were performed to
confirm that the change of chemical shift caused by addition of
catalysts into the solution of HMF could be ignored (Table S4).
Therefore, the change in proton chemical shifts is closely
associated with the hydrogen binding state of hydroxyl protons of
HMF.
Then, Co-N-C-n catalysts was compared with pyridine, which
was well recognized as a HBs acceptor, in term of the proton
1
acceptance ability. The concentration-variation H NMR spectra
3
of HMF−pyridine couple were recorded in CDCl at the same
concentration of HMF (Figure 2c). As expected, the downfield shift
of the hydroxyl proton in HMF exhibited after the addition of
pyridine, implying the formation of HBs between HMF and
pyridine. When the concentration of pyridine was 0.0180 M, the
chemical shift change of the hydroxyl proton in HMF was 0.1064
Figure 4 The change of chemical shift (Δδ) of hydroxyl proton in HMF with
different concentrations of catalysts (a). Plots of 1/C versus 1/Δδ with HMF as
donors and catalysts as acceptors (b).
ppm. Additionally, Co-N-C-9.6 catalyst could also lead to a close
chemical shift change (0.1138 ppm) at 3.04 g·L-1 (nitrogen
All the HMF−Co-N-C-n couples satisfy the linear quantitative
equation (1/Δδ = a/C + b). Figure 4b shows five linear plots
between the reciprocal of the proton chemical shift changes
concentration is 0.0220 M). Therefore, it is reasonable to attribute
the change of chemical shift to the contribution of hydrogen
binding. After the data fitting of 1/Δδ and 1/C for HMF−pyridine
couple, a linear equation based on 1/Δδ = a/C + b (a = 0.1615, b
(
1/Δδ) and the reciprocal of the concentrations (1/C, calculated
according to the nitrogen content of the catalysts) for Co-N-C-9.6,
2
0.5205, R2 = 0.994) was obtained (Figure S7), with the
determined KHB and ΔG being 3.22 L/mol and −2.90 kcal/mol,
-8.5, -6.3, -4.1, and N-C with satisfied correlation coefficient R of
=
0
0.990, 0.980, 0.972, 0.997, and 0.985, respectively. Based on the
above discussion, the overall KHB for the formation and
dissociation of surface HBs is the ratio of intercept to slope of the
respectively, far smaller than those of Co-N-C-n catalysts.
0
0
fitting curves (KHB = b/a) and ΔG could be obtained by ΔG =
RTlnKHB. Considering the real surface possesses multiple states
−
of the doped nitrogen and the chemical shift change represents
the overall contribution of the catalyst surface, the average across
the catalyst surface was regarded as the proton acceptor. Both
0
K
1
HB and ΔG indicate the apparent proton accepting ability. Table
0
listed ΔG of the surface HBs between HMF and the catalysts.
0
The negative values of ΔG suggested that the hydroxyl proton of
HMF could form HBs on all these catalyst surface. The ΔG values
0
increase in the order of Co-N-C-9.6 < Co-N-C-8.5 < Co-N-C-6.3
<
Co-N-C-4.1, which is consistent with the decrease order of
nitrogen content in Co-N-C-n, indicating that increase in the
1
Table 1. Parameters of hydrogen bonds determined by H NMR
N
K
HB
ΔG0
Catalysts
a
b
(
wt%)
(L/mol)
104.79
(kcal/mol)
1
Figure 3 Locally amplified H NMR spectra of HMF with different concentrations
of Co-N-C-9.6 (a), Co-N-C-8.5 (b), Co-N-C-6.3 (c), Co-N-C-4.1 (d), N-C (e), and
Co/CB (f) (400 MHz, CDCl ).
3
Co-N-C-9.6
Co-N-C-8.5
Co-N-C-6.3
Co-N-C-4.1
N-C
10.1
0.0585
0.0643
0.0742
0.0755
0.1056
6.1303
5.7357
5.7167
4.7843
5.0298
−11.53
−11.13
−10.76
−10.28
−9.57
8.4
7.1
5.9
5.5
89.20
77.04
63.37
47.63
After confirming the reliability of the NMR measurements, we
carefully checked the change of the chemical shifts by increasing
the concentration of Co-N-C-9.6, Co-N-C-8.5, Co-N-C-6.3, and
Co-N-C-4.1, respectively. As shown in Figures 3a-d, for all these
four catalysts, the hydroxyl proton chemical shifts in HMF moved
4
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
nitrogen content for the Co-N-C-n catalysts led to decreasing
energy of the formed surface HBs, as well as the improvement in
the ability of the Co-N-C-n catalysts to construct surface HBs.
It is acknowledged that there are multiple states of the doped
nitrogen on the Co-N-C-n surface, which could be distinguished
by performing N1s XPS analysis (Figures 1i and S3). Based on
our previous work,[16d] the protonation energy of pyridinic N was
4.1, consistent with the increase order of nitrogen contents, and
Co-N-C-9.6 catalyst with the highest nitrogen content and the
strongest ability to form surface HBs (the lowest ΔG0 value)
afforded the best catalytic performance for HMF oxidation to
FDCA (Table 2, entry 4). As a result, Co-N-C-9.6 catalyst afforded
an excellent catalytic performance (> 99% conversion of HMF,
-1
94.3% yield of FDCA) with a high TOF value (477 mol HMF mol
-1
negative value (E
P
= −0.26 eV), while the protonation energy of
Co h ). In the case of nitrogen-free Co/AC and metal-free N-C,
the results are close to that of blank experiment. The distinct
difference in catalytic performance verified the indispensable role
of doped nitrogen in carbon, and revealed a strong correlation
other nitrogen species (pyrrolic N, graphitic N and pyridine N-
oxide) were all positive values, suggesting that the pyridinic N
might play the role of proton acceptor just liking pyridine which
has been evidenced to form HBs with HMF as discussed above.
However, the KHB and −ΔG0 of pyridine-HMF couple were far
smaller than those of Co-N-C-n catalysts. This obvious difference
may come from the strong proton accepting ability of Co-N-C-n
surface. Owing to that the pyridinic N contents on the Co-N-C-n
surfaces determined by N1s XPS analysis were all higher than
0
with the KHB and ΔG (discussed below).
It is worthwhile pointing out that the FDCA productivity, to the
best of our knowledge, achieved the highest value in comparison
with previously report non-noble metal heterogeneous catalysts
(Table 2, entries 12-13). Generally, non-noble metals have the
disadvantages of large amount loading and low efficiency in the
synthesis of FDCA from HMF, such as manganese catalyst,
50% (Table S3), the nitrogen-doped surface possessing stronger
ability to accept protons may be due to the increasing electron
donating ability of doped nitrogen through conjugation with
surface carbon species.[22]
Surface Hydrogen binding initiated activation of O−H
bonds for oxidations. Subsequently, we probed the surface HBs
induced activation of O−H bonds in term of catalytic performances
and TOF values in the oxidation of HMF to FDCA. The oxidation
of HMF was performed over five different heterogeneous
catalysts with well-defined compositions and structures
containing the Co-N-C-n catalysts with varied nitrogen content
requiring the MnO
2
and HMF molar ratio of 0.4 even high to 5.7.[23]
Such a high loading amount of non-noble metal catalyst needed
is mainly due to that HMF is extremely unstable and easy to
polymerize on the catalyst surface in the harsh oxidation system
when it is not converted timely, poisoning the catalyst and
stopping the conversion.[ In contrast, the surface HBs-initiated
activation of O−H bonds efficiently promoted the TOF values, with
the Co-N-C-n catalyst dosage being significantly reduced. The
lowest molar ratio of Co and HMF can reach 0.007. In addition,
the gram scale oxidation of HMF was also carried out and
obtained 82.1% yield of isolated FDCA (Figure S12). The stability
and reusability of Co-N-C-9.6 was investigated by recycling the
used catalyst for successive reuse. There was only a slight
attenuation of activity of Co-N-C-9.6 catalyst after five cycles
(Figure S13), suggesting that the Co-N-C-9.6 catalyst remained
highly stable and reusable.
18]
(
Co-N-C-4.1, -6.3, -8.5, and 9.6), Co NPs supported on active
carbon (Co/AC), and nitrogen-doped carbon (N-C) (Table 2). As
demonstrated above, Co-N-C-n catalysts were predicted to have
the ability in initiated activation of O−H bonds due to their negative
0
ΔG value. As expected, all of them achieved rapid conversions
o
of HMF under 1.0 MPa O
2
at 120 C for 1 h. The yields to FDCA
are in order of Co-N-C-9.6 > Co-N-C-8.5 > Co-N-C-6.3 > Co-N-C-
Table 2. The oxidation of furan hydroxyl compounds to carboxylic acids over various heterogeneous catalysts[a]
Yield of
target acid
%)
Productivity
of target acid
O
2
T
(oC)
t
Conv.
(%)
Entry
Catalyst
Ref.
(
MPa)
(h)
-
1
(
(h )
16.1
23.7
25.5
30.2
n.d.
n.d.
n.d.
20.2
20.3
23.0
27.5
2.2
1[b]
2[b]
3[b]
4[b]
5[b]
6[b]
7[b]
8[c]
Co-N-C-4.1
Co-N-C-6.3
Co-N-C-8.5
Co-N-C-9.6
Co/AC
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.3
0.5
1.0
120
120
120
120
120
120
120
120
120
70
1
1
1
1
1
1
1
1
1
4
6
3
5
> 99
> 99
> 99
> 99
23.9
26.5
21.9
> 99
> 99
100
59.2
81.2
88.4
94.3
0.3
This
work
N-C
0.2
Blank
0.4
Co-N-C-9.6
Co-N-C-9.6
Au-Ce
96.5
96.7
92
9[d] [e]
0[b]
1[b]
2[b]
3[b]
[24]
[16e]
[25]
[26]
1
1
1
1
Au/HY
60
> 99
98.3
> 99
> 99
95.3
95.2
CoO
x
-MC
80
Co-Mn-0.25
120
0.1
[
a]
3 2
Reaction condotions: 0.5 mmol furan hydroxyl compounds, 0.024 mmol Co in Co-N-C-n catalysts, 2.0 mmol NaHCO , 5 mL of H
O. [b] The substrate: HMF, the
target acid: FDCA. [c] The substrate: HMFCA, the target acid: FDCA. [d] The substrate: furfuryl alcohol, the target acid: 2-furancarboxylic acid. [e] 1.0 mmol NaHCO
3
.
5
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
In this paper, we reported the initiated activation of O−H bonds
with hydrogen binding on nitrogen-doped surface as driving force
for catalytic oxidation. The formation trend of surface hydrogen
bonds is reflected by hydrogen binding standard Gibbs free
0
energy change (ΔG ) based on model 5-hydroxymethylfurfural on
a series of Co-N-C-n catalysts with electron-rich doped nitrogen
contents adjusted. We developed a method to determine the ΔG0
as well as the association constant (KHB) on catalyst surface via
NMR experiments by a linear equation as a function of the
hydroxyl proton chemical shift changes (1/Δδ = a/C + b, KHB = b/a,
ΔG0 = −RTln(b/a)). The proposed method is of significance in
demonstrating the high ability of Co-N-C-9.6 catalyst surface to
0
form O−H···X HBs with ΔG value of −11.53 kcal/mol, which is
much more negative than that of pyridine. Remarkably, in probing
the oxidation of HMF to FDCA, the obtained parameters of k and
0
TOF describing the catalytic activity, together with ΔG presenting
the proton acceptable ability, are linearly correlated with the
doped nitrogen contents with high correlation coefficients,
suggesting that hydrogen binding interaction could drive the
initiated activation of O−H bonds. The stronger the HBs formation
ability is, the better the capability of HBs initiated activation of O−H
bonds on the catalyst is. Excellent catalytic performance (> 99%
conversion of HMF and 94.3% yield of FDCA in 1 h) and high TOF
value (477 mol HMF mol Co h ) were achieved by Co-N-C-9.6
catalyst with high metal utilization and high stability. This study
provides an experimental method for the determination of solid
surface HBs energy, and provides the guidance for promoting the
activation O−H bonds in biomass-derived hydroxyl compounds by
the HBs interactions as well as the design of oxidation catalysts.
Figure 5 The time courses of HMF conversion over Co-N-C-n catalysts (a). The
linear fitting of ln([HMF]
0
/[HMF]
t
) against the reaction time of the HMF oxidation
-1
-1
over Co-N-C-n catalysts (b). Linear relationships between the apparent rate
constants k (c, left) and TOF values (c, middle) of HMF oxidation over Co-N-C-
n catalysts, together with the −ΔG0 for HMF and Co-N-C-n catalysts (c, right)
with nitrogen contents of catalysts (c). Reaction condition: for the time courses
and fitting of k, 1.0 mmol HMF, 0.024 mmol Co in Co-N-C-n catalysts, 4.0 mmol
NaHCO
HMF, 0.0035 mmol Co in Co-N-C-n catalysts, 2.0 mmol NaHCO
20 oC, 1.0 MPa O
, 10 min.
3
, 10 mL of H
2
O, 100 oC, 0.5 MPa O
2
; for calculating of TOF, 0.5 mmol
3
, 10 mL of H O,
2
1
2
Acknowledgements
As illustrated in Figure 5 and Table 1, the obtained parameters
This work was supported by National Natural Science Foundation
of China (21790331, 21872138, and 21690084) and the Strategic
Priority Research Program of Chinese Academy of Sciences
(Grant No: XDA21030400 and XDB17020300).
were linearly corrected with doped nitrogen contents of the
catalysts. The apparent rate constants (k) and TOF are the most
commonly used quantities describing the catalytic activity, while
0
ΔG represents the proton acceptable ability on catalyst surface.
0
Interestingly, the k, TOF, and −ΔG were linearly correlated with
the Co-N-C-n catalysts in term of the nitrogen contents with high
R of 0.981, 0.966, and 0.985, respectively. All the trends in k,
Keywords: surface hydrogen binding • O−H bond • initiated
activation • selective oxidation • biomass hydroxyl compound
2
TOF, and −ΔG0 are in the same order as that of the doped
nitrogen contents (Co-N-C-9.6 > Co-N-C-8.5 > Co-N-C-6.3 > Co-
N-C-4.1). These indicate that the higher nitrogen content in the
catalysts could contribute to higher proton accepting capacity on
[
1]
a) T. Li, C. Chen, A. H. Brozena, J. Y. Zhu, L. Xu, C. Driemeier, J. Dai,
O. J. Rojas, A. Isogai, L. Wagberg, L. Hu, Nature 2021, 590, 47-56; b) J.
Iglesias, I. Martinez-Salazar, P. Maireles-Torres, D. M. Alonso, R.
Mariscal, M. L. Granados, Chem. Soc. Rev. 2020, 49, 5704-5771; c) R.
Gerardy, D. P. Debecker, J. Estager, P. Luis, J. C. M. Monbaliu, Chem.
Rev. 2020, 120, 7219-7347; d) C. H. Liu, M. R. Niazi, D. F. Perepichka,
Angew. Chem. Int. Ed. 2019, 58, 17312-17321; Angew. Chem. 2019, 131,
0
catalyst surface (lower ΔG ), thus enhancing the abilities in
initiated activation of O−H bonds and promoting the oxidation of
HMF to FDCA. Regarding the oxidation of other furan hydroxyl
compounds such as 5-(methoxymethyl)-2-furoicacid (HMFCA)
and furfuryl alcohol, the high TOF values and yields of
corresponding acids may also be attributed to such hydrogen
binding initiated activation on the Co-N-C-9.6 catalyst (Table 2,
entries 8-9). All the results indicated that the ΔG0 values from
NMR measurements are in complete consistence with the
experimental results in the hydrogen binding initiated activation of
O−H bonds and oxidation of biomass-derived hydroxyl
compounds.
1
7473-17482; e) W. W. Park, K. M. Lee, B. S. Lee, Y. J. Kim, S. H. Joo,
S. K. Kwak, T. H. Yoo, O. H. Kwon, Angew. Chem. Int. Ed. 2020, 59,
089-7096; Angew. Chem. 2020, 132, 7155-7162.
7
[2]
a) S. Chen, R. Wojcieszak, F. Dumeignil, E. Marceau, S. Royer, Chem.
Rev. 2018, 118, 11023-11117; b) C. S. Lancefield, B. Flker, R. C. Cioc,
K. Stanciakova, R. E. Bulo, M. Lutz, M. Crockatt, P. C. A. Bruijnincx,
Angew. Chem. Int. Ed. 2020, 59, 23480-23484; Angew. Chem. 2020, 132,
23686-23690; c) Y. X. Lu, C. L. Dong, Y. C. Huang, Y. Q. Zou, Z. J. Liu,
Y. B. Liu, Y. Y. Li, N. H. He, J. Q. Shi, S. Y. Wang, Angew. Chem. Int.
Ed. 2020, 59, 19215-19221; Angew. Chem. 2020, 132, 19377-19383.
a) H. V. D. Nguyen, R. De Vries, S. D. Stoyanov, ACS Sustainable Chem.
Eng. 2020, 8, 14166-14178; b) H. S. Huang, J. Zhong, X. P. Tan, X. G.
Guo, B. F. Yuan, Y. H. Lin, J. S. Francisco, X. C. Zeng, J. Am. Chem.
Soc. 2020, 142, 10314-10318; c) Y. Yu, T. Tyrikos-Ergas, Y. T. Zhu, G.
Fittolani, V. Bordoni, A. Singhal, R. J. Fair, A. Grafmuller, P. H.
[
3]
Conclusion
6
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
Seeberger, M. Delbianco, Angew. Chem. Int. Ed. 2019, 58, 13127-
[25] X. Y. Liu, M. Zhang, Z. H. Li, ACS Sustainable Chem. Eng. 2020, 8, 4801-
4808.
13132; Angew. Chem. 2019, 131, 13261-13266.
[
4]
a) Z. Q. Yang, J. H. Si, Z. X. Cui, J. H. Ye, X. F. Wang, Q. T. Wang, K.
[26] K. T. V. Rao, J. L. Rogers, S. Souzanchi, L. Dessbesell, M. B. Ray, C. B.
Xu, ChemSusChem 2018, 11, 3323-3334.
P. Peng, W. Z. Chen, S. C. Chen, Carbohydr. Polym. 2017, 174, 750-
759; b) X. Li, H. C. Li, T. T. You, X. M. Chen, S. Ramaswamy, Y. Y. Wu,
F. Xu, ACS Sustainable Chem. Eng. 2019, 7, 19176-19184; c) H. Wang,
Y. F. Zhao, F. T. Zhang, Y. Y. Wu, R. P. Li, J. F. Xiang, Z. P. Wang, B.
X. Han, Z. M. Liu, Angew. Chem. Int. Ed. 2020, 59, 11850-11855; Angew.
Chem. 2020, 132, 11948-11953; d) V. V. Korolkov, M. Baldoni, K.
Watanabe, T. Taniguchi, E. Besley, P. H. Beton, Nat. Chem. 2017, 9,
1191-1197.
[
[
5]
6]
Y. C. Zhu, Z. Y. Zhang, R. Jin, J. Z. Liu, G. Q. Liu, B. Han, N. Jiao, Angew.
Chem. Int. Ed. 2020, 59, 19851-19856; Angew. Chem. 2020, 132,
20023-20028.
M. Y. Liu, Z. R. Zhang, H. Z. Liu, Z. B. Xie, Q. Q. Mei, B. X. Han, Sci.
Adv. 2018, 4, eaas9319.
[
[
7]
8]
M. Waki, S. Muratsugu, M. Tada, Chem. Commun. 2013, 49, 7283-7285.
a) C. D. Liang, S. Dai, J. Am. Chem. Soc. 2006, 128, 5316-5317; b) M.
Hara, Energy Environ. Sci. 2010, 3, 601-607.
[
[
9]
a) S. Q. Xiang, R. L. Narayanan, S. Becker, M. Zweckstetter, Angew.
Chem. Int. Edit. 2013, 52, 3525-3528; b) M. L. de la Paz, J. Jimenez-
Barbero, C. Vicent, Chem. Commun. 1998, 465-466; c) T. D. W. Claridge,
High-Resolution NMR Techniques in Organic Chemistry, Elservier Inc,
2009, pp. 322-323.
10] a) N. Berg, S. Bergwinkl, P. Nuernberger, D. Horinek, R. M. Gschwind,
J. Am. Chem. Soc. 2021, 143, 724-735; b) H. Y. Wang, J. J. Cui, Y. L.
Zhao, Z. Y. Li, J. J. Wang, Green. Chem. 2021, 23, 405-411.
[
[
[
11] P. Thordarson, Chem. Soc. Rev. 2011, 40, 1305-1323.
12] T. Schaefer, J. Phys. Chem. 1975, 79, 1888-1890.
13] a) R. Mathur, N. C. Li, E. D. Becker, R. B. Bradley, J. Phys. Chem. 1963,
6
7, 2190-2194; b) M. W. Hanna, A. L. Ashbaugh, J. Phys. Chem. 1964,
8, 811-816.
6
[
[
[
14] B. W. Tresca, R. J. Hansen, C. V. Chau, B. P. Hay, L. N. Zakharov, M.
M. Haley, D. W. Johnson, J. Am. Chem. Soc. 2015, 137, 14959-14967.
15] N. Robinson, C. Robertson, L. F. Gladden, S. J. Jenkins, C. D'Agostino,
ChemPhysChem 2018, 19, 2472-2479.
16] a) J. Z. Chen, F. Lu, X. Q. Si, X. Nie, J. S. Chen, R. Lu, J. Xu,
ChemSusChem 2016, 9, 3353-3360; b) P. H. Che, F. Lu, X. Nie, Y. Z.
Huang, Y. L. Yang, F. Wang, J. Xu, Chem. Commun. 2015, 51, 1077-
1080; c) Y. Luo, H. Ma, Y. X. Sun, P. H. Che, X. Nie, T. L. Wang, J. Xu,
J. Phys. Chem. A 2018, 122, 843-848; d) Y. X. Sun, H. Ma, X. Q. Jia, J.
P. Ma, Y. Luo, J. Gao, J. Xu, ChemCatChem 2016, 8, 2907-2911; e) J.
Y. Cai, H. Ma, J. J. Zhang, Q. Song, Z. T. Du, Y. Z. Huang, J. Xu, Chem.
-
Eur. J. 2013, 19, 14215-14223; f) Q. Song, F. Wang, J. Y. Cai, Y. H.
Wang, J. J. Zhang, W. Q. Yu, J. Xu, Energy Environ. Sci. 2013, 6, 994-
007; g) R. Lu, F. Lu, J. Z. Chen, W. Q. Yu, Q. Q. Huang, J. J. Zhang, J.
Xu, Angew. Chem. Int. Ed. 2016, 55, 249-253; Angew. Chem. 2016, 128,
57-261.
1
2
[17] Y. Luo, H. Ma, S. J. Zhang, D. Y. Zheng, P. H. Che, X. Liu, M. Y. Zhang,
J. Gao, J. Xu, J. Am. Chem. Soc. 2020, 142, 6085-6092.
[
18] M. Ventura, M. Aresta, A. Dibenedetto, ChemSusChem 2016, 9, 1096-
1100.
[
19] J. J. Shi, Y. Y. Wang, W. C. Du, Z. Y. Hou, Carbon 2016, 99, 330-337.
20] W. G. Liu, L. L. Zhang, W. S. Yan, X. Y. Liu, X. F. Yang, S. Miao, W. T.
Wang, A. Q. Wang, T. Zhang, Chem. Sci. 2016, 7, 5758-5764.
[
[21] R. V. Jagadeesh, H. Junge, M. M. Pohl, J. Radnik, A. Bruckner, M. Beller,
J. Am. Chem. Soc. 2013, 135, 10776-10782.
[
22] a) K. Yuan, T. Hu, Y. Z. Xu, R. Graf, L. Shi, M. Forster, T. Pichler, T.
Riedl, Y. W. Chen, U. Scherf, Mater. Chem. Front. 2017, 1, 278-285; b)
R. Arrigo, M. Havecker, S. Wrabetz, R. Blume, M. Lerch, J. McGregor, E.
P. J. Parrott, J. A. Zeitler, L. F. Gladden, A. Knop-Gericke, R. Schlogl, D.
S. Su, J. Am. Chem. Soc. 2010, 132, 9616-9630.
[
[
23] a) E. Hayashi, Y. Yamaguchi, K. Kamata, N. Tsunoda, Y. Kumagai, F.
Oba, M. Hara, J. Am. Chem. Soc. 2019, 141, 890-900; b) E. Hayashi, T.
Komanoya, K. Kamata, M. Hara, ChemSusChem 2017, 10, 654-658; c)
F. Nocito, M. Ventura, M. Aresta, A. Dibenedetto, ACS Omega 2018, 3,
18724-18729.
24] S. Albonetti, A. Lolli, V. Morandi, A. Migliori, C. Lucarelli, F. Cavani, Appl.
Catal., B 2015, 163, 520-530.
7
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202103604
RESEARCH ARTICLE
0
The ΔG of hydrogen binding on surface was obtained by linear fitting of the chemical shifts change of surface hydrogen bonded protons
and the variation of acceptor concentrations via NMR techniques, which reflected to the abilities of catalyst surface to form HBs with
hydroxyl groups. And the surface hydrogen binding was successfully employed to drive the initiated activation of O−H bonds in probing
the oxidation of HMF with Co-N-C catalysts.
8
This article is protected by copyright. All rights reserved.