Towards the Circular Economy: Converting Aromatic Plastic Waste Back to Arenes over a Ru/Nb2O5 Catalyst

Article abstract of DOI:10.1002/anie.202011063

The upgrading of plastic waste is one of the grand challenges for the 21^st century owing to its disruptive impact on the environment. Here, we show the first example of the upgrading of various aromatic plastic wastes with C?O and/or C?C linkages to arenes (75–85 % yield) via catalytic hydrogenolysis over a Ru/Nb₂O₅ catalyst. This catalyst not only allows the selective conversion of single-component aromatic plastic, and more importantly, enables the simultaneous conversion of a mixture of aromatic plastic to arenes. The excellent performance is attributed to unique features including: (1) the small sized Ru clusters on Nb₂O₅, which prevent the adsorption of aromatic ring and its hydrogenation; (2) the strong oxygen affinity of NbO_x species for C?O bond activation and Br?nsted acid sites for C?C bond activation. This study offers a catalytic path to integrate aromatic plastic waste back into the supply chain of plastic production under the context of circular economy.

Full text of DOI:10.1002/anie.202011063

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Towards the circular economy: converting aromatic plastic waste
back to arenes over Ru/Nb2O5 catalyst
Authors: Yaxuan Jing, Yanqin Wang, Shinya Furukawa, Jie Xia,
Chengyang Sun, Max J. Hülsey, Haifeng Wang, Yong Guo,
Xiaohui Liu, and Ning Yan
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.202011063
Link to VoR: https://doi.org/10.1002/anie.202011063

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
Towards the circular economy: converting aromatic plastic waste
back to arenes over Ru/Nb2O5 catalyst
Yaxuan Jing, [a b], Yanqin Wang,[a]* Shinya Furukawa,[c d] Jie Xia,[a] Chengyang Sun,[a] Max J. Hülsey,[b]
Haifeng Wang, [a] Yong Guo,[a] Xiaohui Liu,[a] and Ning Yan[b]*
Abstract: The upgrading of plastic waste is one of the grand
challenges for the 21st century owing to its disruptive impact on
environment. Here, we show the first example of the upgrading of
various aromatic plastic wastes with C–O and/or C–C linkages to
arenes (75-85% yield) via catalytic hydrogenolysis over Ru/Nb2O5
catalyst. This catalyst not only allows the selective conversion of
single-component aromatic plastic, and more importantly, enables the
simultaneous conversion of a mixture of aromatic plastic to arenes.
The excellent performance is attributed to the unique features
including: (1) the small sized Ru clusters on Nb2O5, which prevent the
adsorption of aromatic ring and its hydrogenation; (2) the strong
oxygen affinity of NbOx species for C–O bond activation and Brønsted
acid sites for C–C bond activation. This study offers a catalytic path to
integrate aromatic plastic waste back into the supply chain of plastic
production under the context of circular economy.
recycling. Unfortunately, this method is associated with
deteriorated properties of plastics which significantly limits its
applicability. [1a,3b,5] For example, only 7% of recycled PET can be
used to recast bottles.[6]
So far, reports on chemical degradation of plastic waste were
mainly focused on pyrolysis, which allows to handle relatively
mixed plastic waste streams, especially polyolefins, PS and
polymethyl methacrylate.[3b,7] Alternative strategies (e.g.,
hydrolysis, alcoholysis, hydrogenation and aminolysis) have been
developed to handle specific aromatic plastics in particular
PET,[3b,8] however, none of these are capable of converting a
mixture of aromatic plastics necessitating sorting as a prerequisite.
Considering the strong desire to minimize plastic disposal into the
environment, and the increasing aspiration to generate profit by
plastic reuse and recycling for the petro-chemical sector,[3b,8f]
developing advanced catalytic approaches for the upcycling of
aromatic plastic waste into value-added chemicals is highly
desirable.
Introduction
Arenes (e.g., p-xylene, m-xylene, ethylbenzene, benzene) –
predominantly derived from fossil fuels – are important raw
materials for the manufacture of aromatic plastics and fibers. In
order to achieve a circular economy, the conversion of aromatic
plastic waste back to those arenes must be realized (Figure 1). In
recent years, the production of arenes from non-petroleum based
feedstock such as biomass and CO2 gained particular attention,
presenting an important field of research.[9] Compared to biomass
and CO2, aromatic plastic waste has prominent advantages,
including simpler molecular structure, much lower oxygen
contents and abundant aromatic functionality (Figure S1). In
particular, its simple structure, contrasting the complexity of
biomass, provides grand opportunities to selectively produce a
narrow stream of arenes, such as in the conversion of PET to p-
xylene and PPO to m-xylene. To date, the selective degradation
of aromatic plastic waste into target arenes is exceedingly rare. A
key challenge of this task is to identify a catalyst that enables the
selective cleavage of C–O and/or C–C bonds while preserving the
aromatic rings. Inspired by C–O/C–C activation chemistry,
cooperative catalysis between two or more different active sites,
especially a combination of metal sites with high hydrogenation
property and acidic supports with the strong ability to activate C–
O/C–C bonds, is required to overcome this key challenge. Up
to now, no catalyst can realize this ambitious goal.
At present, 380 million tons of plastics are produced around the
world each year, and ~ 80% of which end up discarded.[1] Due to
the decades- or even centries-long degradation kinetics, a
majority of all plastics ever produced on earth are still lying either
in land or in oceans.[1e,2] Many of the widely used plastics, such
as polyethylene terephthalate (PET), polystyrene (PS),
polycarbonate (PC), and polyphenylene oxide (PPO), are
constructed from aromatic monomers by interunit C–O and/or C–
C linkages.[3] Indeed, PET, PS and PC are all listed among the
top ten plastics used globaly.[4] These aromatic plastics are
produced on an annual scale of ~60 million tons, out of which ~50
million tons are disposed into earth’s ecosystem.[3b]
Commercialized route to reuse them mainly relies on mechanical
[a]
Y. Jing, C. Sun, Dr. Y. Guo, Dr. X. Liu, Prof. Dr. Y. Wang
Key Laboratory for Advanced Materials and Joint International
Research Laboratory of Precision Chemistry and Molecular
Engineering, Feringa Nobel Prize Scientist Joint Research Center,
Research Institute of Industrial Catalysis, School of Chemistry and
Molecular Engineering, East China University of Science and
Technology, Shanghai 200237, China
E-mail: wangyanqin@ecust.edu.cn
[b]
M. Hülsey, Prof. Dr. N. Yan
Herein, we report that a multifunctional Ru/Nb2O5 catalyst can
directly upgrade aromatic plastic waste into arenes in unusually
high yields (Figure 1). We demonstrate that Ru/Nb2O5 does not
only catalyze the selective conversion of single-component plastic
waste into bulk chemicals (PET to p-xylene and PPO to m-xylene),
but also enables the conversion of mixed aromatic plastics into
arenes. This catalyst combines two key functions, namely NbOx
species for C–O bond activation and Brønsted acid sites for C–C
bond activation, respectively, making the production of
monocyclic arenes feasible. We confirmed that Ru species on
Department of Chemical & Biomolecular Engineering, National
University of Singapore, 4 Engineering Drive 4, Singapore 117585
(Singapore)
E-mail: ning.yan@nus.edu.sg
Prof. Dr. S. Furukawa
Institute for Catalysis, Hokkaido University, N-21, W-10, Sapporo
001-0021, Japan
Prof. Dr. S. Furukawa
Elements Strategy Initiative for Catalysis and Battery, Kyoto
University, Kyoto Daigaku Katsura, Nishikyo-ku, Kyoto 615-8510,
Japan
Spporting information for this article is given via a link at the end of
the document.
[c]
[d]
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
Nb2O5 present as sub-nano particles with low coordination
number (C.N.= 5-6), preventing the hydrogenation of the benzene
ring thus affording a distinct selectivity to arenes. To the best of
our knowledge, this is the first time to achieve the upgrading of
aromatic plastic waste into arenes through the selective cleavage
of interunit C–O and/or C–C linkages, and the catalytic
performance of Ru/Nb2O5 is unparalleled by other catalysts.
Figure 1. The integration of a C–O and C–C bond cleavage catalyst into the circular plastic economy. Ru/Nb2O5 is proposed as a possible catalyst for the conversion
of various aromatic plastic waste (PET – polyethylene terephthalate, PS – polystyrene, PPO – poly(p-phenylene oxide), PC – polycarbonate) and mixtures thereof
into arenes.
inter-chain interactions. Moreover, water does not react with
hydrogen over metal-solid acid catalysts at the reaction
Results and Discussion
temperature. In contrast, some common polar organic solvents,
such as THF, acetone, ethanol et al, may suffer from
hydrogenation/hydrogenolysis reactions.[11] Among all catalysts
(Ru/Nb2O5, Pd/Nb2O5, Pt/Nb2O5, Ru/TiO2, Ru/ZrO2 and
Ru/HZSM-5) investigated, Ru/Nb2O5 gave an exceptional yield of
monomers (95.2%) and a high selectivity to arenes (87.1%)
(calculated based on the ratio between the mole of product and
the mole of aromatic units in polymer feedstock). Over Pd/Nb2O5
and Pt/Nb2O5, a total monomer yield of ~20% was obtained, much
lower than that over Ru/Nb2O5. Noteworthy, the main products
obtained in the latter two cases were ring saturated products
(compound 6 & 7 in Figure 2). Hence, we suspected the poor
performance of Pd and Pt catalysts was due to the fast
hydrogenation of aromatic rings, resulting in the formation of
chemically more robust C–O linkages. To verify this, we
conducted the conversion of methyl p-toluate over Nb2O5-loaded
Ru, Pt, and Pd catalysts (Figure 2b) and observed arenes as the
PET, PPO, PS and PC are chosen as model compounds since
they comprise all common types of linkages in aromatic plastics,
namely ester linkage, ether linkage, C–C and combinations of C–
O and C–C linkages. Moreover, PET, PS and PC represent the
largest three groups produced among various aromatic
plastics.[1e,4] Initially, PET, PPO, PS and PC from commercial
sources were analyzed by FT-IR and elemental analysis (Figure
S2 and Table S1). The results confirmed the chemical structure
and relatively high purity of these plastics,[10] offering the basis to
calculate product yields in catalytic reactions. The shape and
porous properties of Nb2O5 and Ru/Nb2O5 are provided in Figure
S3.
The conversion of PET into p-xylene in aqueous solution was
employed as a probe reaction to investigate the cleavage of
interunit C–O linkages (Figure 2a). The main reason we use water
as solvent is that water is a polar solvent that binds reasonably
strongly with polar functionalities in PET. This helps to weak the
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
main products over Ru/Nb2O5 (80.1% selectivity). In sharp
contrast, the main product was ring-saturated compound 9 over
Pt/Nb2O5 and Pd/Nb2O5. This result indicates that Pd- and Pt-
based catalysts are more active for the hydrogenation of benzene
rings, and that the cleavage of C–O bonds in the saturated ring is
more challenging. The latter was further confirmed by employing
compound 9 as substrate (Table S2), whose conversion is only
13.6% under the same conditions. Two factors are involved in the
poor reactivity. First, the C–O bond energy in compound 9 is
higher than that in methyl p-toluate because of the
conjugation effect between the benzene ring and the ester group.
In addition, the conjugated aromatic group has a planar structure,
resulting in a smaller steric hindrance compared to non-planar
cyclohexane rings. Therefore, the activation of C–O bonds
adjacent to saturated rings is not favorable, which highlights the
significance of preserving aromatic rings. In the case of Nb2O5-
supported catalysts, Ru with appropriate hydrogenation ability,
instead of Pd and Pt, is the most selective metal site.
With Ru/ZrO2, Ru/TiO2 and Ru/HZSM-5 as catalysts, total
monomer yields of 72.6, 71.9 and 31.8%, respectively, were
obtained from PET, lower than that over Ru/Nb2O5 (95.2%).
Interestingly, the selectivity of arenes also decreased dramatically
(50.8, 46.7 and 28.9%). Those results imply that Ru species on
these three supports, unlike Ru species on Nb2O5, tend to
hydrogenate the aromatic ring. The conversion of methyl p-toluate
was conducted over these four catalysts (Figure 2b). Indeed, the
unusual selectivity towards arenes was only achieved over
Ru/Nb2O5, while it was lower than 50% with compound 9 as the
main product over the other three catalysts.
Previously we found that NbOx species in Nb-based catalysts
have the unique ability to activate C–O bonds in biomass.[11] It is
therefore not unreasonable to speculate that besides the
uncommon properties of Ru clusters on Nb2O5 to preserve
benzene rings, Nb2O5 itself plays an important role in C–O bond
cleavage. To confirm this, we employed compound 9 as substrate
to compare Ru-based catalysts under relatively harsh conditions
(220 ºC, 4 h, 3 MPa H2) (Figure S5). As expected, the highest
conversion and selectivity to 1,4-dimethylcyclohexane were
achieved over Ru/Nb2O5, in agreement with previous results on
the hydrodeoxygenation of biomass and its derivatives,[11] thus
confirming NbOx species possess a strong ability to activate the
C–O bonds in plastics. Combining all experimental observations,
we attribute the exceptional performance of Ru/Nb2O5 to an
effective synergism of the support-mediated C–O bond activation
and the metal-based selective hydrogenation. The influence of
reaction conditions was investigated. The major findings include:
(1) the plastic conversion increased monotonically as a function
of time while there was little variations of selectivity towards each
product; (2) lower temperature led to low reaction rates, whereas
higher temperature favored the undesirable decarboxylation; (3)
higher pressure favored the hydrogenation of aromatic ring,
whereas lower pressure was not able to complete deoxygenation.
Detailed discussions are provided in the Supporting Information
(Figure S6). Finally, the direct conversion of Coca-Cola bottle was
tested over Ru/Nb2O5 and encouragingly, a total monomer yield
of 90.9% was achieved with an arene selectivity of 78.4% (table
S3), confirming the applicability of the established system to
upgrade real PET waste.
We next tested the conversion of PPO, an ether-linked plastic,
over Ru/Nb2O5 with water and octane as medium. PPO
aggregated into a solid globule after the reaction in water,
whereas desirable m-xylene was successfully obtained in octane.
The dissolution/swelling of PPO in two solvents was compared
through stirring at room temperature and it was observed that
PPO dispersed better in octane than in water (Figure S7).
Therefore, octane was employed as reaction medium to perform
the conversion of various aromatic plastics. PET was effectively
converted to arenes (83.6% yield) in octane (Figure 3) as well. A
high yield (85.0%) of m-xylene was achieved starting from PPO,
demonstrating that, besides the ability to convert ester-linked
plastics, Ru/Nb2O5 also enabled the degradation of ether-linked
plastics.
Our objective is to cleave all types of interunit C–O/C–C
linkages in aromatic plastic waste to produce arenes. Hence, we
conducted the conversion of PS formed solely via C–C linkages
over Ru/Nb2O5. Notably, the conversion of PS afforded a total
monocyclic arenes yield of 75.9 %. It should be noted that the
Figure 2. The conversion of PET over various catalysts. a, The direct
conversion of PET over various catalysts. Reaction conditions: 0.1 g PET, 0.1
g catalyst, 10 g H2O, 200 ºC, 12 h, 0.3 MPa H2; b, The conversion of methyl p-
toluate over various catalysts. Reaction conditions: 0.05 g methyl p-toluate,
0.025 g catalyst, 5 g H2O, 200 ºC, 3 h, 0.3 MPa H2.
cleavage always occurs at the sp2-sp3 bond, namely the Caromatic
C connected with aromatic rings, thus affording high selectivity to
–
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
benzene, which is highly consistent with our previous report in
lignin conversion.[12] A certain amount of indane derivatives and
other arenes were generated by the acid-catalyzed alkylation.
This result proved that Ru/Nb2O5 is effective in breaking the C–C
linkages in aromatic plastic waste while preserving aromatic rings.
Previously, degradation of plastics formed via stable C–C
linkages predominately rely on the non-selective pyrolysis route
to pyrolysis oil and then may be further upgraded into arenes via
zeolite-catalyzed aromatization.[3b] Our newly established system
is distinct from conventional pyrolysis process that enables
>250 °C) and Ru/Nb2O5 is capable of producing a high throughput
in such circumstances.
Catalyst recycling was conducted with mixed plastics as
feedstock and some activity lost (~28%) in the second run (Figure
S9a). From TG analysis (~10% carbon deposition on spent
catalyst, Figure S9b), we speculate that the activity loss may
come from the coverage of active sites by carbon. Indeed, after
calcination in air (450 ºC for 3 h) followed by reduction using H2
(400 ºC for 2 h), the catalyst was recharged for the next run and
the same arene yield was obtained, even after 3 cycles. The
consistent high yields of arenes, and the absence of any change
in XRD patterns for the spent catalyst (Figure S9c), clearly
demonstrate the high durability of Ru/Nb2O5 catalyst.
An intriguing question is why Ru/Nb2O5 can not only catalyze
the C–O and C–C bond cleavage efficiently, but also gives distinct
selectivity to arenes. Detailed characterizations were conducted
to understand the structure of the catalyst and its structure-activity
correlations. Although no diffraction peaks corresponding to Ru
(Figure S10) were identified in the XRD patterns of all Ru-based
samples, the Ru dispersion results revealed that Ru on Nb2O5 is
substantially higher dispersed (73.4%) than on other supports
such as Ru/ZrO2 (21.1%), Ru/TiO2 (7.7%), and Ru/HZSM-5
(15.8%), suggesting the presence of smaller Ru species on Nb2O5
(Table S6). The H2-TPR profiles (Figure 4a) indicated that the
reduction temperature of Ru on Ru/Nb2O5 (130 °C) was higher
than that of Ru/ZrO2 (106 °C), Ru/TiO2 (105 °C) and Ru/HZSM-5
(101 °C), implying the strong interaction between Ru and Nb2O5.
X-ray absorption near-edge structure (XANES) analysis was
carried out to further clarify the chemical state and coordination
environment of Ru (Figure 4b). Ru species on all catalysts are
close to metallic zero-valent state. XANES spectra of Ru on ZrO2,
TiO2 and HZSM-5 are identical to that of Ru foil, suggesting that
Ru species with high Ru-Ru coordination numbers are dominant
on these supports. In sharp contrast, the XANES spectrum of Ru
on Nb2O5 has evidently different amplitude of the fine structure
oscillations, although the Ru is also zero-valent. This smaller
amplitude indicates that Ru on Nb2O5, unlike Ru species on other
supports and Ru foil, has a different coordination environment.
We conducted the Ru K-edge extended X-ray absorption fine
structure (EXAFS) analysis, and the fitting results for Ru/TiO2 and
Ru/ZSM-5 (Figure S11) demonstrate coordination numbers at
around 9 for Ru in Ru/TiO2 and Ru/ZSM-5 catalysts (Table S7).
On the other hand, the extended scattering paths (measured by
transmission or fluorescence mode) for Ru/Nb2O5 and Ru/ZrO2
could not be analyzed due to the large X-ray absorption by Nb
and Zr and the overlap of major Ru Kα1 emission with Kβ2 of Nb
and Zr, respectively. Based on the fact that Nb2O5-supported Ru
has a higher dispersion, we suspect that Ru on Nb2O5 may tend to
be low-coordinated with small particle sizes.
Figure 3. Results of the conversion of various aromatic plastics over Ru/Nb2O5.
Reaction conditions: 30 mg feedstock, 30 mg Ru/Nb2O5, 4 g octane, 0.5 MPa
H2. a) 280 ºC, 8 h; b) 280 ºC, 16 h; c) 300 ºC, 16 h; d) 320 ºC, 16 h; e) Reaction
conditions: 15 mg PET, 15 mg PC, 15 mg PS, 15 mg PPO, 60 mg Ru/Nb2O5, 4
g octane, 0.5 MPa H2, 320 ºC, 16 h.
the selective cleavage of the Caromatic–C linkages to directly
produce monocyclic arenes.
To evaluate the catalyst performance with increasing feedstock
complexity, PC comprising both ester and C–C linkages was
chosen. The corresponding monocyclic aromatic hydrocarbons
were obtained in 83.3% total yield, indicating the simultaneous,
efficient cleavage of both bond types already hinting at the
possibility to convert mixtures of aromatic plastic waste. It is
noteworthy that the PC used here is commercial polycarbonate
board, once again confirming that the catalytic system is able to
work well to address real PC waste. We further evaluated the
potential of Ru/Nb2O5 for the direct conversion of mixed plastics
containing PET, PS, PC and PPO. Encouragingly, the mixture
underwent the cleavage of interunit C–O and C–C bonds to yield
the arenes expected from single component conversion
experiments with
a
total yield of 78.9%, confirming the
XANES spectra are commonly used to reveal the chemical
states of metals, however, less attention has been paid to explore
the information of how coordination environment and particle size
affect XANES spectra.[13] Here, we successfully estimated the
geometric configuration of Ru species on the Nb2O5 support by
comparing the experimental and density functional theory (DFT)
predicted Ru XANES spectra. All shapes including corner, edge,
face and bulk atoms with various coordination numbers (C.N. =
3~9, 12) of Ru species considered for XANES simulation are
presented in Figure 4e. Arctangent functions were generated to
calibrate the baseline so that the simulation well reproduces the
experimental spectrum of the Ru foil (Figure S12). The XANES
applicability of Ru/Nb2O5 to convert mixtures of plastic waste
without interference of individual plastic component. The
concentration of substrate was investigated (Figure S8) and the
results show a negligible decrease in arenes yield (from 78.9% to
75.3%) occurs when the concentration was increased from 1.5%
to 3%, strengthening the potential of this new system. The mass
ratio of substrate to catalyst was increased from 1/1 to 2/1, the
arenes yield reduced slightly (from 78.9% to 68.8 %), indicating
that unlike biomass conversion,[9] the contact problem between
plastic and catalyst has a relatively little influence on the reaction
since plastics can melt under high temperature (generally
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
simulations by DFT calculations (Figure 4c) allows a full
assignment of the spectra for a series of Ru coordination numbers
(C.N. = 3~9, 12). The calculated spectra show that the peak
intensity decreases while the peak position shifts to higher energy
with the decrease of coordination numbers. This is probably
because the electron density on the coordinatively
Figure 4. Characterization of oxide-supported Ru catalysts. a, H2-TPR profiles of various catalysts; b, XANES spectra of various Ru-based catalysts; c, XANES
simulation by DFT with a series of Ru coordination numbers; d, comparison of XANES spectrum of other Ru catalysts and DFT-calculated spectra; e, Various
coordination numbers of Ru species considered for XANES simulation; f, comparison of XANES spectra of Ru/Nb2O5 and DFT-calculated spectra; acetone DRIFTS
spectra for g, Nb2O5, h, TiO2, i, ZrO2, and j, HZSM-5, at increasing temperatures of 35 °C (black), 60 °C (orange), 90 °C (blue), and 120 °C (green).
unsaturated Ru becomes higher due to the larger number of
dangling bonds. Not unexpectedly, comparison between
experimental and theoretical spectra (Figure 4d) shows that
XANES spectra of Ru species on ZrO2, TiO2 and HZSM-5 are very
similar to those of Ru with C.N.= 9 or 12, confirming the EXAFS
fitting results. Remarkably, the XANES spectrum of Ru/Nb2O5
resembles those simulated for Ru with C.N. = 5 or 6 very closely,
providing strong evidence that the major Ru species on Nb2O5 are
those with low coordination numbers (C.N. = 5~6) (Figure 4f).
Such low coordination numbers indicate that the contribution of
edge sites is dominant, confirming that the dimension of Ru on
the Nb2O5 support is in the sub-nanometer range.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
It is known that the effective co-adsorption of aromatic rings and
H2 is a prerequisite to catalyze the hydrogenation of the aromatic
ring. It is therefore reasonable to believe that the limited surface
area of small Ru clusters on Nb2O5 restrain the co-adsorption of
H2 and aromatic rings due to the increased barrier for aromatic
ring adsorption on metal clusters compared to flat surfaces,[14]
resulting in a unique selectivity to arenes. Meanwhile, the
adsorption of H2 is less affected by the particle size effect and
therefore Ru sub-nano particles maintain sufficient capability of
H2 dissociation to assist C–O/C–C cleavage.
The oxygen affinity and C–O bond activation capability of the
support are essential factors for the conversion of bonds prevalent
in plastics.[9a,11b] To verify the oxygen affinity of Nb2O5, acetone as
a simple oxygen-containing molecule, was selected to perform
adsorption-desorption DRIFTS experiments on various supports
(Figure 4g-j). For acetone, the peak at 1740 cm-1 is assigned to
the C=O stretching vibration.[15] The red shift on Nb2O5 (1694 cm-
1) is evidently stronger than that on ZrO2 (1704 cm-1), TiO2 (1701
cm-1), and HZSM-5 (1704 cm-1) after purging with N2 for 30 min at
35 °C, indicating that Nb2O5 possesses stronger activation ability
than other supports. When the purging temperature rose to
120 °C, acetone on ZrO2 and TiO2 desorbed completely while
minimum C=O bond signal on HZSM-5 was observed. On Nb2O5,
in contrast, the peak of C=O bond was still clearly visible,
providing evidence that Nb2O5 has distinctively strong adsorption
capability to oxygen-containing groups. With the help of
dissociated H species over Ru clusters, the unique synergistic
activity for the cleavage of C–O bonds within aromatic plastics is
rationalized.
for Ru/Nb2O5 and Nb2O5 (Figure S15) confirm that the primary
adsorption sites reside on Nb2O5 and Ru has little effect on
adsorption of benzene ring. While with the purging temperature
rising from 30 to 90 °C, two clear red-shifts at 1603 cm-1 (Δ=6 cm-
1) and 1454 cm-1 (Δ=5 cm-1) were observed over Nb2O5,
suggesting that C=C vibrations are hindered because of the
adsorption-activation of benzene ring on Nb2O5 (Figure 5a).
Meanwhile, an obvious blue-shift (Δ=6 cm-1) at 3058 cm-1 was
found on Nb2O5 probably because the C–H vibrations became
stronger (Figure 5b). The blue-shift provides additional evidence
for the adsorption-activation of benzene ring on Nb2O5, where the
strong adsorption-activation to benzene ring framework lengthens
the C-H bonds, resulting in the stronger C-H vibrations.
In marked contrast to Nb2O5, no corresponding red- or blue-shift
was observed on HZSM-5 when the purging temperature rise to
90 °C, providing evidence that Nb2O5 has distinctively stronger
adsorption-activation capability to benzene than that of HZSM-5.
Next, DFT calculations were performed to comparatively
investigate the Csp2-Csp3 cleavage of diphenylmethane over
Nb2O5(001) and HZSM-5. The optimization of adsorption sites
confirms that on Nb2O5, the primary adsorption sites of benzene
ring reside on NbOx species, but on HZSM-5, Brønsted acid sites
as adsorption sites have the priority over AlOx species (Figure
S15). Optimized structures of this process and the reaction
energy profile are shown in Figure 5c and 5d. Firstly, the
adsorption energy of diphenylmethane on Nb2O5(001) (-1.17 eV)
is larger than that on HZSM-5 (-0.88 eV), in good agreement with
the DRIFTS results, where Nb2O5 enables stronger adsorption to
benzene ring. It is noteworthy that the carbocation intermediate
formed by the Brønsted acid sites-driven protonation process is
crucial for the Csp2-Csp3 cleavage, given that the direct Caromatic–C
The performance of Ru/Nb2O5 and Ru/ZSM-5 on C–C bond
cleavage was tested using diphenylmethane as
a model
compound (Table S8). Over Ru/Nb2O5, the desirable C–C
cleavage products (benzene and toluene) were generated. Over
Ru/HZSM-5, ring-saturated products are predominant, indicating
that the hydrogenation of benzene ring is favored compared with
the desired C–C bond cleavage. The very low total product yield
(4.5%) at a conversion of 17.6% was caused by the cracking of
substrate and products over Ru/HZSM-5. Therefore, Ru/Nb2O5
has superior performance for the desirable C–C bond cleavage
over Ru/HZSM-5.
We conducted the blank experiments and selected PET and PS
as starting materials because PET and PS representing C-O and
C-C linked plastic respectively. The results are present in Table
S4. For PET conversion, terephthalic acid as the only product was
found in catalyst-free and Nb2O5 conditions, indicating that
conventional hydrolysis of PET can occur over Nb2O5 or in pure
water under high reaction temperature. As to the conversion of
PS over Nb2O5 and catalyst-free conditions, no products were
detected. Furthermore, no conversion was observed under
catalyst-free or pure Nb2O5 condition when diphenylmethane was
employed as a model compound to test the C-C cleavage
performances. These results show that the C-C cleavage of PS
only occurs over the multifunctional Ru/Nb2O5 catalyst.
bond cleavage is extremely difficult[12]
quantitatively calculated the reaction
.
Furthermore, we
energies and
energy barriers of protonation process and Csp2-Csp3 cleavage of
+
the protonated intermediate (C6H5(H)C7H7 ), respectively. The
following key information are obtained: (i) On Nb2O5, both steps
show evidently lower kinetic barriers than these on HZSM-5; (ii)
the whole reaction energy change of the C-C cleavage on Nb2O5
(0.43 eV) is lower than that on the HZSM-5 (1.09 eV). From above,
Nb2O5 is superior to HZSM-5, since it facilitates the protonation,
Csp2-Csp3 cleavage and modified the overall reaction energy profile.
The above study confirmed that the multifunctional Ru/Nb2O5
catalyst plays two roles in the selective cleavage of C–O/C–C
bonds in aromatic plastic waste into arenes: 1) small size Ru
species with low coordination numbers (C.N.= 5-6) as
hydrogenation sites limit co-adsorption of H2 and aromatic rings,
affording a high selectivity to arenes; 2) NbOx is responsible for
the activation of C–O bonds and benzene ring while Brønsted acid
sites assist in cleaving C–C bonds. The cooperative effect
between small-sized Ru, NbOx species and Brønsted acid sites
render this catalyst exceptionally suitable for the decomposition
of plastic mixtures into arenes. The cooperative effects of different
components of catalyst are highlighted in Figure 5e. For C-O bond
cleavage, Lewis acid sites (NbOx species) enable the selective
adsorption and activation of C-O bond. Then, with the help of
dissociated H species over Ru, the cleavage of C-O bonds in
aromatic plastics is efficiently achieved. For C-C bond cleavage,
benzene ring is first adsorbed on Lewis acid sites (NbOx species)
of Nb2O5, then the adsorbed benzene ring is protonated by
we conducted DRIFTS analysis of toluene adsorption to
compare the adsorption-activation of benzene ring on Nb2O5 and
HZSM-5 (Figure 5). The peaks at 3450 and 1623 cm-1 are
attributed to hydrogen-bonded OH stretching and HOH bending
vibration of the adsorbed water molecules (Figure S14). For gas
phase toluene, the peaks at 1454-1603 cm-1 and 3026-3085 cm-1
are ascribed to the stretching mode of the C=C bonds and C-H
stretching mode of benzene ring, respectively. The same spectra
Brønsted acid sites on Nb2O5 to proceed the activation of the Csp2
-
Csp3 bond. Following that, the dissociated H species on Ru
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
clusters attack the weakened Csp2-Csp3 bond to break it. The small
Ru clusters play a key role to achieve highly selective arene
production during C-O/C-C cleavage, where the limited surface of
small Ru clusters restrain the co-adsorption of H2 and aromatic
rings.
Figure 5. a, Toluene DRIFTS spectra (1454-1603 cm-1) of Nb2O5 and HZSM-5; b, Toluene DRIFTS spectra (3026-3085 cm-1) of Nb2O5 and HZSM-5; c. Optimized
structures of diphenylmethane adsorption on Nb2O5 and HZSM-5, diphenylmethane adsorption (IS); TS of protonation (TS-1), intermediate state (IM), TS of Csp2-
Csp3 cleavage (TS-2) and final state (FS); d. Energy profiles of the Csp2–Csp3 cleavage processes of diphenylmethane; g. Proposed mechanism of catalytic C-
O/C-C cleavage over Ru/Nb2O5.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
Based on experimental data, we performed a preliminary
mass balance analysis for the new route (Figure 6). Starting from
1 ton aromatic plastic mixture, 60.5 wt% of the feedstock can be
converted into arenes in a single-step with the consumption of
only 3.8 wt% H2, which compares favorably with other emerging
technologies for renewable arene production from woody
biomass and CO2 (Figure S16 and S17). Only 20.7~34.4 wt% of
1 ton raw woody biomass and 9.8~15.8 wt% of 1 ton CO2—
significantly lower than for aromatic plastics valorization—can be
converted into arenes based on open data in the literature (Figure
S16a and S17a). Moreover, the additional use of ethylene for
biomass conversion and the much higher demand of H2 for CO2
hydrogenation inevitably increase the cost of the competing
technologies. As a result, the newly established route adds a
promising option with unequaled efficiency for arene production
from waste resources. The separation of arenes can be achieved
through distillation. For isomers, such as p-xylene and m-xylene,
dependence on fossil fuels for the production of aromatic
chemicals and downstream polymers.
We highlight the following limitations of the new approach
deserving consideration and further development: a) Despite the
superior catalytic performance of Ru/Nb2O5 in cleaving C–C and
C–O bonds in plastics, the solid-solid contact problem between
recalcitrant plastic and solid catalyst, remains a challenge.
Designing advanced solvent system to increase plastic swelling
and solubility is highly desirable; b) H2 gas is used as a reductant.
Establish hydrogen transfer systems using renewable alcohols
and acids as hydrogen sources, is a way to reduce process
complexity and cost; c) Non-noble-metal based catalysts with
comparable performance should be explored to improve the
overall economic competitiveness. Developments along these
lines are going on in our labs.
adsorption is a viable approach[21]
.
Acknowledgements
This work was supported financially by the NSFC of China (No.
21832002, 21872050, 21808063), the Science and Technology
Commission of Shanghai Municipality (2018SHZDZX03), the
Programme of Introducing Talents of Discipline to Universities
(B16017) in China, the Fundamental Research Funds for the
Central Universities (222201718003), and the NUS Flagship
Green Energy Programme (R-279-000-553-646, and R-279-000-
553-731). The XAS analysis was performed with the approval of
Japan Synchrotron Radiation Research Institute (JASRI: No.
2019B1620). Computation time was provided by the
supercomputer systems in Institute for Chemical Research, Kyoto
University. Y. J. is grateful for financial support from the China
Scholarship Council (CSC grant number 201906740056).
Figure 6. Mass balance of aromatic plastics into arenes. The values are
estimated based on the reaction stoichiometries in this study and assuming the
conversion of 1 ton feedstocks.
Keywords: plastic • arenes • C-O/C-C cleavage • Ru/Nb2O5 •
Conclusion
circular economy
In summary, we have developed the direct upgrading of aromatic
plastic waste mixtures back to arenes in high yield via the precise
cleavage of interunit C–O and C–C linkages over a Ru/Nb2O5
catalyst. Ru/Nb2O5 is able to selectively break all common types
of linkages in aromatic plastics, including ester, ether and even
C–C linkages to obtain monocyclic arenes. Ru species on Nb2O5,
unlike those on other supports, have ultra-small particle sizes,
preventing the hydrogenation of aromatic rings. Along with NbOx
species for C–O bond activation and Brønsted acid sites for C–C
bond cleavage, desired reactivity towards aromatic plastic
depolymerization is achieved.
Despite the long-standing research efforts for chemical
upgrading of plastic waste, a majority of previously developed
routes is only applicable for specific plastic waste streams.
Pyrolysis is an exception, since it enables the transformation of
relatively mixed plastic waste streams, but still is unable to work
well with the plastic contaminations including PVC, PU and
PET.[3b] Thus, a combination of different technologies may be
applied in the future to provide a satisfactory solution of plastic
waste valorization. The here reported catalytic system converting
aromatic plastic waste into value-added arenes adds new options
in plastic upcycling offering possible solutions to two key
challenges, namely, the accumulation of highly persistent
aromatic plastics in the environment and the excessive
[1]
a) Nat. Catal. 2019, 2, 945-946; b) V. Tournier, C. Topham, A. Gilles, B.
David, C. Folgoas, E. Moya-Leclair, E. Kamionka, M.-L. Desrousseaux,
H. Texier, S. Gavalda, Nature 2020, 580, 216-219; c) J. M. Garcia, M. L.
Robertson, Science 2017, 358, 870-872; d) J. R. Jambeck, R. Geyer, C.
Wilcox, T. R. Siegler, M. Perryman, A. Andrady, R. Narayan, K. L. Law,
Science 2015, 347, 768-771; e) R. Geyer, J. R. Jambeck, K. L. Law, Sci.
Adv. 2017, 3, e1700782; f) B. F. S. Nunes, M. C. Oliveira, A. C.
Fernandes, Green Chem. 2020; g) T. Uekert, M. F. Kuehnel, D. W.
Wakerley, E. Reisner, Energy Environ. Sci. 2018, 11, 2853-2857; h) A.
Rahimi, J. M. García, Nat. Rev. Chem. 2017, 1, 1-11.
[2]
[3]
H. P. Austin, M. D. Allen, B. S. Donohoe, N. A. Rorrer, F. L. Kearns, R. L.
Silveira, B. C. Pollard, G. Dominick, R. Duman, K. El Omari, Proc. Natl.
Acad. Sci. U.S.A. 2018, 115, E4350-E4357.
a) C. J. Rhodes, Sci. Prog. 2018, 101, 207-260; b) I. Vollmer, M. J. Jenks,
M. C. Roelands, R. J. White, T. van Harmelen, P. de Wild, G. P. van der
Laan, F. Meirer, J. T. Keurentjes, B. M. Weckhuysen, Angew. Chem. Int.
Ed. 2020. doi:10.1002/anie.201915651.
[4]
a) World Plastics Production 1950–2015 (Plastics Europe); b) Plastics–
the Facts 2019: An analysis of European plastics production, demand
and waste data (Plastics Europe).
[5]
[6]
[7]
N. A. Rorrer, S. Nicholson, A. Carpenter, M. J. Biddy, N. J. Grundl, G. T.
Beckham, Joule 2019, 3, 1006-1027.
T. Uekert, H. Kasap, E. Reisner, J. Am. Chem. Soc. 2019, 141, 15201-
15210.
a) A. B. Raheem, Z. Z. Noor, A. Hassan, M. K. A. Hamid, S. A. Samsudin,
A. H. Sabeen, J. Clean. Prod. 2019; b) R. Aguado, M. n. Olazar, B.
Gaisán, R. Prieto, J. Bilbao, Chem. Eng. J. 2003, 92, 91-99; c) M. Artetxe,
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
G. Lopez, M. Amutio, G. Elordi, M. Olazar, J. Bilbao, Ind. Eng. Chem.
Res. 2010, 49, 2064-2069.
[8]
a) S. Westhues, J. Idel, J. Klankermayer, Sci. Adv. 2018, 4, eaat9669; b)
H. Tang, N. Li, G. Li, A. Wang, Y. Cong, G. Xu, X. Wang, T. Zhang, Green
Chem. 2019, 21, 2709-2719; c) H. Tang, Y. Hu, G. Li, A. Wang, G. Xu, C.
Yu, X. Wang, T. Zhang, N. Li, Green Chem. 2019, 21, 3789-3795; d) E.
Feghali, T. Cantat, ChemSusChem 2015, 8, 980-984; e) Y. Xie, X. Jiao,
K. Zheng, Q. Chen, X. Li, Y. Li, W. Shao, J. Xu, J. Zhu, Y. Pan, Y. Sun,
Angew. Chem. Int. Ed. 2020 doi:10.1002/anie.201915766; f) X. Zhang,
M. Fevre, G. O. Jones, R. M. Waymouth, Chem. Rev. 2018, 118, 839-
885.
[9]
a) Y. Jing, Y. Guo, Q. Xia, X. Liu, Y. Wang, Chem 2019, 5, 2520-2546; b)
W. Schutyser, T. Renders, S. Van den Bosch, S. F. Koelewijn, G. T.
Beckham, B. F. Sels, Chem. Soc. Rev. 2018, 47, 852-908; c) X. Wu, X.
Fan, S. Xie, J. Lin, J. Cheng, Q. Zhang, L. Chen, Y. Wang, Nat. Catal.
2018, 1, 772-780; d) Z. R. Zhang, J. L. Song, B. X. Han, Chem. Rev.
2017, 117, 6834-6880; e) W. Zhou, K. Cheng, J. Kang, C. Zhou, V.
Subramanian, Q. Zhang, Y. Wang, Chem. Soc. Rev. 2019, 48, 3193-3228;
f) S. Saeidi, N. A. S. Amin, M. R. Rahimpour, J. CO2 Util. 2014, 5, 66-81;
g) A. E. Settle, L. Berstis, N. A. Rorrer, Y. Roman-Leshkóv, G. T. Beckham,
R. M. Richards, D. R. Vardon, Green Chem. 2017, 19, 3468-3492.
[10] a) M. Liang, Y.-J. Jhuang, C.-F. Zhang, W.-J. Tsai, H.-C. Feng, Eur. Polym.
J. 2009, 45, 2348-2357; b) S. Vijayakumar, P. Rajakumar, ILCPA 2012,
4, 58-65; c) J.-c. Wang, H. Wang, Sep. Purif. Techno. 2017, 187, 415-
425.
[11] a) Y. Shao, Q. Xia, L. Dong, X. Liu, X. Han, S. F. Parker, Y. Cheng, L. L.
Daemen, A. J. Ramirez-Cuesta, S. Yang, Y. Wang, Nat. Commun. 2017,
8, 16104; b) Q. Xia, Z. Chen, Y. Shao, X. Gong, H. Wang, X. Liu, S. F.
Parker, X. Han, S. Yang, Y. Wang, Nat. Commun. 2016, 7, 11162; c) Q.
N. Xia, Q. Cuan, X. H. Liu, X. Q. Gong, G. Z. Lu, Y. Q. Wang, Angew.
Chem. Int. Ed. 2014, 53, 9755-9760.
[12] L. Dong, L. F. Lin, X. Han, X. Q. Si, X. H. Liu, Y. Guo, F. Lu, S. Rudic, S.
F. Parker, S. H. Yang, Y. Q. Wang, Chem 2019, 5, 1521-1536.
[13] J. Nilsson, P.-A. Carlsson, H. Grönbeck, M. Skoglundh, Top. Catal. 2017,
60, 283-288.
[14] a) U. Sanyal, Y. Song, N. Singh, J. L. Fulton, J. Herranz, A. Jentys, O. Y.
Gutiérrez, J. A. Lercher, ChemCatChem 2019, 11, 575-582; b) N. Singh,
U. Sanyal, G. Ruehl, K. A. Stoerzinger, O. Y. Gutiérrez, D. M. Camaioni,
J. L. Fulton, J. A. Lercher, C. T. Campbell, J. Catal. 2020, 382, 372-384.
[15] Y. Jing, Y. Xin, Y. Guo, X. Liu, Y. Wang, Chin. J. Catal. 2019, 40, 1168-
1177.
[16] a) D. Zhang, A. Duan, Z. Zhao, C. Xu, J. Catal. 2010, 274, 273-286; b) P.
Michaud, J. Lemberton, G. Pérot, Appl. Catal. A 1998, 169, 343-353.
[17] M. Segall, P. J. Lindan, M. a. Probert, C. J. Pickard, P. J. Hasnip, S. Clark,
M. Payne, J. Phys. Condens. Mat. 2002, 14, 2717.
[18] B. Hammer, L. B. Hansen, J. K. Nørskov, Phys. Rev. B 1999, 59, 7413.
[19] D. Vanderbilt, Phys. Rev. B 1990, 41, 7892.
[20] H. Monkshort, J. Pack, Phys. Rev. B 1976, 13, 5188-5192.
[21] Y. Yang, P. Bai, X. Guo, Ind. Eng. Chem. Res. 2017, 56, 14725–14753
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202011063
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Yaxuan Jing, [a b] Yanqin Wang,[a]* Shinya Furukawa,[c d], Jie Xia, [a]
Chengyang Sun,[a] Max J. Hülsey,[b] Haifeng Wang,[a] Yong Guo,[a]
Xiaohui Liu,[a] and Ning Yan[b]*
Towards the circular economy: converting aromatic plastic
waste back to arenes over Ru/Nb2O5 catalyst
While plastic waste upcycling has attracted increasing worldwide concern, catalytic routes capable to convert a mixture of plastics
selectively and effectively into useful chemicals remain exceedingly rare. Here, we report a salient example of the catalytic upgrading
of various aromatic plastic waste to simple arenes in high yield via the selective cleavage of interunit C–O and C–C linkages.
10
This article is protected by copyright. All rights reserved.