Rational Designed Polymer as a Metal-Free Catalyst for Hydroxylation of Benzene to Phenol with Dioxygen

Article abstract of DOI:10.1007/s10562-020-03392-9

Abstract: Direct hydroxylation of benzene to phenol with dioxygen is a green route for synthesis of phenol. The higher bonding energies of C–H in benzene and O–O in dioxygen make metal-free catalysis a challenge. A novel metal-free catalyst of quinone-cyl

Full text of DOI:10.1007/s10562-020-03392-9

Catalysis Letters
https://doi.org/10.1007/s10562-020-03392-9
Rational Designed Polymer as a Metal‑Free Catalyst for Hydroxylation
of Benzene to Phenol with Dioxygen
Weitao Wang1 · Yaoyao Wei1 · Xulu Jiang1 · Zhen‑Hong He1 · Cunshe Zhang2 · Zhao‑Tie Liu1,3
Received: 23 July 2020 / Accepted: 4 September 2020
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Direct hydroxylation of benzene to phenol with dioxygen is a green route for synthesis of phenol. The higher bonding
energies of C–H in benzene and O–O in dioxygen make metal-free catalysis a challenge. A novel metal-free catalyst of
quinone-cylcohexanone-formaldehyde polymer was rationally designed and prepared. The polymer was characterized with
FT-IR, XRD, Raman, SEM, 13C NMR and XPS. Abundant oxygen functional groups of cyclic ketonic, quinone carbonyl
and hydroquinone were found on the surface of polymer. The catalytic performances for hydroxylation of benzene to phenol
with dioxygen over the metal-free catalyst were investigated and a phenol yielding of 11.4% was achieved. Moreover, the
reaction mechanism was proposed.
Graphic Abstract
Keywords Green chemistry · Catalysis · Aerobic oxidation · Liquid-phase oxidation · Partial oxidation · Hydroxylation ·
Benzene · Metal-free catalysis
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-020-03392-9) contains
supplementary material, which is available to authorized users.
2
* Weitao Wang
Shaanxi Key Laboratory of Petroleum Fine Chemicals,
Shaanxi Provincial Research and Design Institute
of Petroleum and Chemical Industry, Xi’an, Shaanxi 710054,
China
wangweitao@sust.edu.cn
* Zhao-Tie Liu
ztliu@snnu.edu.cn
3
School of Chemistry & Chemical Engineering, Shaanxi
Normal University, Xi’an, Shaanxi 710119, China
1
Shaanxi Key Laboratory of Chemical Additives
for Industry, College of Chemistry & Chemical
Engineering, Shaanxi University of Science & Technology,
Xi’an, Shaanxi 710021, China
Vol.:(0123456789)
1 3

W. Wang et al.
the metal-free catalytic reaction for hydroxylation of ben-
zene with dioxygen, giving the phenol yield of 12.5% and
10.1%, respectively. The cyclic ketonic and quinone C= O
over the surface of carbon are proposed as the active sites
for the metal-free carbon-based catalyst [5, 17, 18]. Previ-
ously, we have developed a quinone-amine based polymer
as a metal-free catalyst for hydroxylation of benzene to
phenol with dioxygen [19]. The hydroquinone and quinone
were recognized as the active sites. However, the hydrogen
peroxide and diamine are employed in the quione-amine
polymer preparation, which may involve the safety and
cost concerns. Therefore, these inspired us to design a
safety and easily prepared catalyst with abundant cyclic
ketonic, quinone C = O and hydroquinone groups as a
metal-free catalyst.
1 Introduction
Phenol is an important intermediate in chemical industry
[1] and it is industrially prepared by an isopropyl benzene
process in three steps. However, this synthetic method has
some disadvantages, such as long process fow, by-product
of acetone and waste water containing phenol [2]. The
direct hydroxylation of benzene to phenol with O2 is a
green route for preparation of phenol due to its high atomic
economy, low cost and easy availability of O2. Oxygen is
widely used as an oxidant for hydroxylation of benzene to
phenol. However, direct hydroxylation of benzene to phe-
nol is still a challenge due to the higher bonding energies
of C–H and O–O in benzene and O2 [3–5].
Rational designed catalyst with specifc active sites is a
key to explore the catalytic system for direct hydroxylation
of benzene to phenol with dioxygen. Redox transition met-
als as the active center for benzene hydroxylation reaction
are widely reported, mainly the V, Cu, Fe, Pd, polyoxo-
metalate based catalysts [6–14]. With molecular oxygen as
the oxidant, metal-based catalysts usually gave the yield of
10–15% for phenol. Satisfed yield of phenol (> 20%) can
be rarely obtained [15, 16]. However, the metal based cata-
lyst often sufers the leaching problem since the acetic acid
aqueous is often employed as the reaction solvent. The
leached toxic metal may raise the environmental concerns.
Therefore, exploring the metal-free catalyst for hydroxyla-
tion of benzene to phenol attracts the interests. However,
designing and developing of metal-free catalyst is still a
challenge for the titled reaction.
Cyclohexanone and formaldehyde can form cyclohex-
anone-formaldehyde resin with abundant cyclic ketonic
C=O (Scheme 1a). Moreover, the phenolic resin can pro-
vide the hydroquinone (Scheme 1b) and can be oxidized into
quinone C=O group (Scheme 1c). The cyclic ketonic C=O,
hydroquinone and quinone C=O in the resin are expected as
active sites for the direct hydroxylation of benzene to phenol
with dioxygen. Therefore, the quinone and cyclohexanone
were employed to polymerize with formaldehyde to form
quinone-cylcohexanone-formaldehyde resin (QCF) in the
present work. QCF is expected to be a metal-free catalyst for
the hydroxylation of benzene to phenol with dioxygen due
to the abundant existence of cyclic ketonic/quinone C=O
and hydroquinone. The synthesized QCF was characterized
to identify its structure. It is found that QCF can catalyze the
hydroxylation of benzene to phenol with a yield of 11.4%.
Moreover, the reaction mechanism was confrmed to be a
radical involved process.
The surface oxygen species-rich carbon has an activity
for the titled reaction [5]. The modifed activated carbon
[17] and N-doped mesoporous carbon [18] can achieve
Scheme 1 Schematic illustra-
tion for the synthesis of QCF
1 3

Rational Designed Polymer as a Metal-Free Catalyst for Hydroxylation of Benzene to Phenol with…
at room temperature. Scanning electron microscope (SEM)
2 Experimental
for the sample morphology was measured at 20 kV and
15 mA on the Hitachi S-4800. The solid state 13C NMR
was conducted on the Agilent 600 M.
2.1 Materials
All reagents were commercially available and used without
additional purifcation. Benzene was obtained from Shang-
hai Macklin Biochemical Co., Ltd. p-Benzoquinone and
acetic acid were obtained from Sinopharm Chemical Rea-
gent CO., Ltd. Formaldehyde solution and cyclohexanone
were purchased from Shanghai Aladdin Reagent CO., Ltd.
2.4 Benzene Hydroxylation Reaction
Aerobic hydroxylation of benzene to phenol was carried
out in a high-pressure seal reactor. In a typical run, 0.5 mL
(5.62 mmol) of benzene, 0.1 g of QCF, 0.4 g of LiOAc,
and 3.0 mL 70%(V/V) acetic acid aqueous solution were
added into the reactor successively. After the reactor was
charged with 3.0 MPa O2 at room temperature, the reaction
was conducted at 140 °C for 12 h with vigorous stirring.
After the reaction, the catalyst was separated by centrifu-
gation. The solution was diluted to 50 mL in a volumetric
fask with methanol. Concentrations of phenol in the solu-
tion were measured by high performance liquid chromatog-
raphy (HPLC, WAYEE LC 3000–2 Series instrument) using
an Ultimate MBC18 (250 mm× 4.60 mm) column by UV
detector at 254 nm wavelength. The temperature of column
is 40 °C. The mobile phase consisted of a methanol/water
solution 30/70 (V/V) fed at a fow-rate of 1.0 mL/min. The
CO2 and some undefned compounds were detected in the
products (Fig. S2, Fig. S3). The yield of phenol was calcu-
lated with the external standard method according to the
formula:
2.2 Catalyst Preparation
The QCF was synthesized by a one-step solvothermal
method (Fig. S1). Typically, 8.0 mmol p-benzoquinone,
8.0 mmol cyclohexanone, and 1.8 g 37% formaldehyde
aqueous solution were mixed in 20 mL ethyl alcohol, then
the mixture was stirred for 20 min at 65 °C. Subsequently,
200 µL 30% sodium hydroxide solution was added in the
mixture. The mixture was raised to 85 °C and stirred for
10 min. After that, the mixture was transferred into a
Tefon-lined autoclave and heated at 200 °C for 5 h. The
product was separated by fltration, washed with ethanol
and dried at 80 °C. The obtained powder was denoted as
QCF-n, in which n represents the molar ratio of quinone
to ketone.
mmol of phenol
Yield of phenol =
× 100
mmol of initial benzene
2.3 Characterization
XRD (X-ray difraction) patterns were obtained by using
Cu Kα radiation, nickel flter, 40 kV/40 mA, scanning
mode of 2 theta/theta, scanning type of continuous scan-
ning, and a scanning range from 3° to 90° at a scanning
rate of 8°/min (Ultima IV, Rigaku). XPS (X-ray photoelec-
tron spectroscopy) spectra (Axis Ultra, Kratos Analytical
Ltd.) were using 300 W Al Kα radiation with the binding
energies calibrated at 284.6 eV from C1s of the adventi-
tious carbon. Asymmetrical XPS peaks were deconvoluted
by using the curve-ftting approach with CasaXPS soft-
ware, as well as by applying Shirley background subtrac-
tion and Lorentzian–Gaussian functions (30%G). Surface
area of the catalysts was conducted by the adsorption of N2
at −196 °C, the catalyst sample was degassed in vacuum
at 150 °C for 10 h before N2 adsorption (Gemini VII 2.00,
Micromeritics Instrument Corporation). Raman spectra
were recorded on a LabRam-1B microscopic Raman spec-
trometer with 532 nm laser excitation. Fourier transform
infrared spectroscopy (FT-IR) was performed on a VEC-
TOR-22 using the KBr pellet technique with a resolution
of 2 cm−1 from 400 to 4000 cm−1 and consisted of 32 scans
3 Results and Discussion
The functional groups of the QCF were characterized by the
FT-IR (Fig. S4a). All the resins showed a similar IR spectra.
The peaks at 3300–3600 cm−1 were attributed to the vibra-
tion of –OH. The peaks at~2923 cm−1 and~2858 cm−1 were
assigned to the C–H vibration in –CH2– and –CH–. The
peaks at about 1790, 1709, and 1652 cm−1 were contributed
by C=O in aldehyde, ketone, and quinone [20], respectively.
With the molar ratio of quinone to ketone increased, the
intensity of quinone group peak relatively increased, indicat-
ing the content of quinone groups in the QCF increased. The
peak at 1447 cm−1 was assigned to the σC–H of methylene
linker while peaks at 1380 cm−1and 1330 cm−1 were attrib-
uted to the σC–H of aromatic ring [20, 21]. The 1583 cm−1
and 1250 cm−1 can be assigned to the νC=C in aromatic ring
and νC–O in hydroquinone [22], suggesting the existence of
hydroquinone groups in the QCFs. The X-difraction pat-
terns exhibited broad peaks for the QCFs, indicated that
QCFs were amorphous (Fig. S4b). Two peaks can be found
1 3

W. Wang et al.
Table 1 Hydroxylation of benzene to phenol over QCF
in the Raman spectra of the QCFs (Fig. S4c). The broad
peak at ⁓1353 cm−1 was attributed to the D mode related
to the sp3-hybridized carbon. The peak at ⁓1585 cm−1 in
the Raman spectra was attributed to G peak related to the
sp2-hybridized carbon [23]. The ID/IG can be employed to
indicate the relatively content of sp3- and sp2- carbon. It can
be found that the content of sp3-hybridized carbon in QCF
decreased with an increase of the molar ratio of quinone
to ketone. N2 adsorption–desorption analysis of the QCFs
(Fig. S4d) revealed a typical type III isotherm. The BET
surface areas were 8.6, 8.4, 13.6 m2·g−1 for QCF-0.5, QCF-
1, and QCF-2, respectively. This indicated that the molar
ratio of quinone to ketone afected the texture structure of
the resins by the polymerization structures. The solid 13C
NMR was characterized to verifed the structure of the QCF
(Fig. S5). The peak at about 27 ppm revealed the existence
of –CH2– in the linker and cyclohexanone groups. The peaks
at about 48 ppm and 79 ppm corresponded to the –CH– and
–CH2–OH. The peaks at about 122 ppm and 146 ppm were
attributed to the carbon in benzene and phenolic hydroxyl,
respectively. The C=O in quinone and ketone were found at
about 186 ppm and 214 ppm. The solid 13C NMR spectrum
verifed the structure of QCF in Scheme 1. The images of
SEM shows the QCF was in the form of spherical particles
(Fig. S6).
Entry Catalyst
Solvent
Tempera- Yield (%)
ture (°C)
1
2
3
4
5
6
7
8
QCF-1
QCF-1
QCF-1
QCF-1
QCF-1
QCF-1
QCF-1
QCF-1
QCF-0.5
QCF-2
Acetonitrile
Ethanol
DMF
140
140
140
140
trace
trace
trace
trace
10.4
2.8
6.9
6.1
6.1
3.3
DMSO
Acetic acid (70%) 140
Acetic acid (70%) 120
Acetic acid (70%) 130
Acetic acid (70%) 150
Acetic acid (70%) 140
Acetic acid (70%) 140
9
10
11
QCF-carbona Acetic acid (70%) 140
trace
Catalytic reaction conditions: benzene (0.5 mL), catalyst (100 mg),
solvent (3.0 mL), LiOAc (0.4 g), O2 (3.0 MPa); 12 h
aQCF was calcined at 900 °C under nitrogen atmosphere
The prepared QCF was employed as the catalyst for the
hydroxylation of benzene to phenol. The reaction solvent
was screened (Table 1). It was found that no phenol was
yielded in the solvents of acetonitrile, ethanol, DMF, and
DMSO. It can yield 10.4% phenol in the solvent of acetic
acid aqueous solution. This indicated that the acid proton sol-
vent was important for the QCF catalytic system. The reac-
tion temperature was examined from 120 °C to 150 °C. The
yield of phenol increased frstly, and then decreased with the
increase of temperature, giving a maximum yield at 140 °C.
This was consistent with the reaction temperature range of
120 °C to 150 °C for the reductant-free metal-based cata-
lytic reaction system [19]. As expected, the loading of the
QCF afected the catalytic performance. With the increase
of the QCF loading, the yield of phenol increased frst, and
then declined (Table S1). The yield of phenol increased with
the increase of oxygen pressure and reached the platform
under the pressure of 3.0 MPa (Table S2). The amount of
LiOAc additive afected the reaction. The optimized amount
XPS revealed that three peaks of carbon in QCFs
(Fig. 1a). The peaks of C1s at 284.6, 286.1, and 288.6 eV
were assigned to the aromatic/aliphatic carbon, C–OH, and
C=O [5], respectively. Two kinds of oxygen species were
found on the surface of the QCFs (Fig. 1b), in which the
binding energy of O1s at 531.8 eV was attributed to C=O
while the peak at 533.8 eV was the C–OH [19]. The surface
oxygen content increased from 19.0 to 20.2 at.% with the
increase of molar ratio of quinone to ketone from 0.5 to 1.
But the oxygen content decreased to 17.4 at.% when the
molar ratio increased to 2. From the above characterization,
it can be confrmed that the existence of the groups of qui-
none, phenolic hydroxyl, ketone, and CH2 linker. Therefore,
the QCF structure was deduced as Scheme 1.
Fig. 1 The XPS spectra of a C1s and b O1s in QCFs. c Relationship between the phenol yield and the surface oxygen content of QCF
1 3

Rational Designed Polymer as a Metal-Free Catalyst for Hydroxylation of Benzene to Phenol with…
of LiOAc additve was 0.4 g (Table S3). With the increase of
the reaction time, the yield of phenol increased and reached
11.4% at 24 h (Table S4). The yield of phenol with QCF-1
as catalyst reached catalytic performance level of the metal-
based catalyst (Table S5). The QCF-2 and QCF-0.5 showed
lower yield of phenol than that of QCF-1, implying that the
group content in the QCF afected the reaction activity. The
relationship between the oxygen content (C=O and phenolic
hydroxyl groups) and the yield of phenol was roughly linear
(Fig. 1c), indicating the surface oxygen functional groups
are active sites. This was further confrmed by that there was
no phenol detected when the oxygen groups of the QCF were
removed by calcination of QCF at 900 °C under nitrogen
atmosphere (Table 1, entry 11).
reaction environment [11]. When the radical scavengers of
the t-butanol (•OH radical scavenger) and BHT (butylated
hydroxytoluene, •O2− radical scavenger) were added in the
reaction system, the yield of phenol reduced to 2.9% and
0.4%, respectively, indicating that both hydroxyl radical
−
(•OH) and superoxide radical (•O2 ) were involved in the
reaction. To test whether the oxygen activation was involved
in the initial step in the aerobic oxidation reaction, the reac-
tion was carried out under molecular oxygen-free condition
(Table 2, entry 10). The yield of phenol was not detected,
excluding the involving of surface oxygen species over the
QCF in the initial step of reaction but indicating that the
oxygen molecule activation was involved in the initial step.
From the fndings of the controlled experiments, it can
be demonstrated that (1) the active sites were the surface
oxygen species, (2) hydroxyl radical and superoxide radical
were involved in the reaction, (3) O2 activation was involved
in the initial step. Based these, the possible reaction route
was proposed for the present hydroxylation of benzene to
phenol (Scheme 2). The O2 was chemisorbed on the surface
oxygen functional groups (ketonic/quinone C=O, and hyd-
The phenol could not be yielded without the QCF cata-
lyst, suggesting that the QCF provided the active sites for
the hydroxylation of benzene to phenol (Table 2). In absence
of the LiOAc additive, the yield of phenol decreased from
10.4% to 7.3%. With LiCl as the additive, trace phenol was
found, indicating that the Li+ ion was not the main efec-
tive content in the additive. This was due to the enthalpy of
the formation of Li+(Ben) and Li+(phenol) was nearly equal
[24, 25] (Table S6), indicating the interaction between Li+
ion and benzene or phenol was equal. In addition, the yield
of phenol varied little when NaOAc additive was added in
the catalytic system instead of LiOAc. Therefore, it can be
deduced that the OAc− was the efective content provided
by the additives. The acetated salts can form the bufer solu-
tion with the acetic acid aqueous solution to give the acidic
−
roquinone) and then was activated to generate •O2 [18].
At the acidic environment, •O2− can combine H+ to form
HO2•. Subsequently, two HO2• radicals easily combine to
form H2O2 and O2. Because rate constant of combination
reaction of HO2• radicals is larger than that of the reaction
for HO2• radicals attacking benzene [19, 26]. Therefore,
H2O2 was generated by two HO2• radicals. •OH was gener-
ated from H2O2 on the surface of the QCF. It is the •OH
radical that attacked the benzene ring to form phenol due
to the reaction rate for •O2− and HO2• attacking benzene
is much lower than that of •OH [26]. From this reaction
route, it can explain why the hydroxyl radical (•OH) and
Table 2 The controlled experiments for hydroxylation of benzene to
phenol
−
superoxide radical (•O2 ) scavengers can greatly decrease
Entry
Conditions
Yield (%)
the yield of phenol. Moreover, this was proved by the yield
of 3.7% phenol with the H2O2 as the oxidant without O2
a
1
No catalyst
−
2
No LiOAc
7.3
3
4
5
6
7
8
9
10
11
Standard condition
10.4
trace
10.7
0.5
0.4
5.9
LiClb
H+
NaOAcb
-
H2O2
O2
BHT (5.7 mmol)
BHT (8.5 mmol)
t-Butanol (5.7 mmol)
t-Butanol (8.5 mmol)
No O2
O
n
OH
O
O2
OH
OH
OH
2.9
ac
QCF
−
n
n
OH
H2O2
3.7d
OH
O
Catalytic reaction conditions: benzene (0.5 mL), catalyst (100 mg),
70%(V/V) acetic acid (3.0 mL), LiOAc (0.4 g), O2 (3.0 MPa), 12 h
- H+
n
O
aNo phenol was detected
H
OH
bEquivalent LiCl or NaOAc was added instead of LiOAc
cThe reaction was carried out under N2 atmosphere
dThe reaction was carried out with H2O2 (30 wt.%, 1.0 mL) as the
oxidant under N2 atmosphere.
Scheme 2 Proposed reaction pathway for hydroxylation of benzene
to phenol over QCF
1 3

W. Wang et al.
4. Huang Y-L, Pellegrinelli C, Wachsman ED (2017) ACS Catal
7:5766–5772
(Table 2, entry 11), but no phenol was yielded without O2
(Table 2, entry 10).
5. Wang W, Shi L, Li N, Ma Y (2017) RSC Adv 7:12738–12744
6. Long Z, Chen G, Liu S, Huang F, Sun L, Qin Z, Wang Q, Zhou
Y, Wang J (2018) Chem Eng J 334:873–881
7. Shang S, Chen B, Wang L, Dai W, Zhang Y, Gao S (2015) RSC
Adv 5:31965–31971
4 Conclusion
8. Okamura J, Nishiyama S, Tsuruya S, Masai M (1998) J Mol Catal
A: Chem 135:133–142
In summary, we have developed a novel quinone-cylcohex-
anone-formaldehyde polymer as a metal-free catalyst for
hydroxylation of benzene to phenol with dioxygen. The
polymer was characterized to possess the oxygen functional
groups of quinone, phenolic hydroxyl, and ketone. The oxy-
gen groups were the active sites for the reaction. Under the
optimized reaction condition, 11.4% phenol can be obtained.
This catalyst confrmed the idea that rationally designed cat-
alyst with the oxygen groups of quinone, phenolic hydroxyl,
and ketone can be employed to catalyze the hydroxylation
reaction. This provides a promising candidate for designing
the metal-free catalyst for the hydroxylation of benzene to
phenol.
9. Okemoto A, Ueyama K, Taniya K, Ichihashi Y, Nishiyama S
(2017) Catal Commun 100:29–32
10. Shang S, Hua Y, Li J, Bo C, Ying L, Shuang G (2014) ChemPlu-
sChem 79:680–683
11. Qin Q, Liu Y, Shan W, Hou W, Wang K, Ling X, Zhou Y, Wang
J (2017) Ind Eng Chem Res 56:12289–12296
12. Long Z, Zhou Y, Chen G, Ge W, Wang J (2014) Sci Rep
4:3651–3651
13. Liu Y, Murata K, Inaba M (2006) J Mol Catal A: Chem
256:247–255
14. Wang W, Yao M, Ma Y, Zhang J (2014) Prog Chem 26:1665–1672
15. Li Y, Li B, Chen T, Zhou Z, Wang J, Huang J (2015) Chinese J
Catal 36:1086–1092
16. Wang W, Li N, Tang H, Ma Y, Yang X (2018) Mol Catal
453:113–120
17. Zhu J, Li G, Wang Q, Zhou Y, Wang J (2019) Ind Eng Chem Res
58:20226–20235
Acknowledgements The authors gratefully acknowledge the fnancial
supports from the National Natural Science Foundation of China (No.
21403136, 21978160), Natural Science Basic Research Plan in Shaanxi
Province of China (No. 2019JM-080, 2019JLM-16, 2019ZDLGY06-
04), and Foundation of Chemical Industry Shaanxi Key Laboratory of
Petroleum Fine Chemicals.
18. Shan W, Li S, Cai X, Zhu J, Zhou Y, Wang J (2019) Chem-
CatChem 11:1076–1085
19. Wang W, Tang H, Jiang X, Huang F-E, Ma Y (2019) Chem Com-
mun 55:7772–7775
20. Shiraishi Y, Takii T, Hagi T, Mori S, Kofuji Y, Kitagawa Y, Tan-
aka S, Ichikawa S, Hirai T (2019) Nat Mater 18:985
21. Yang G, Jiang J, Zhang Y (2015) Prog Org Coat 78:55–58
22. Poljanšek I, Krajnc M (2005) Acta Chim Slov 52:238–244
23. Ou J, Wang J, Liu S, Mu B, Ren J, Wang H, Yang S (2010) Lang-
muir 26:15830–15836
Compliance with Ethical Standards
Conflict of interest There is no confict of interest for each contribut-
ing author.
24. Feller D, Dixon DA, Nicholas JB (2000) J Phys Chem A
104:11414–11419
25. Zierkiewicz W, Michalska D, Černý J, Hobza P (2006) Mol Phys
104(13–14):2317–2325
References
26. Vysotskaya NA (1973) Russ Chem Rev 42:851–856
1. Zhou Y, Ma Z, Tang J, Yan N, Du Y, Xi X, Wang K, Zhang W,
Wen H, Wang J (2018) Nat Commun 9:2931
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional afliations.
2. Jiang T, Wang W, Han B (2013) New J Chem 37:1654–1664
3. Lücke B, Narayana KV, Martin A, Jähnisch K (2004) Adv Synth
Catal 8:1407–1424
1 3