Nanoisozymes: The Origin behind Pristine CeO2 as Enzyme Mimetics

Article abstract of DOI:10.1002/chem.202001597

It is known that the interplay between molecules and active sites on the topmost surface of a solid catalyst determines its activity in heterogeneous catalysis. The electron density of the active site is believed to affect both adsorption and activation o

Full text of DOI:10.1002/chem.202001597

Accepted Article
Accepted Article
Title: Nanoisozymes: The Origin behind Pristine CeO2 as Enzyme-
mimetics
Authors: Zicong Tan, Yu-Cheng Chen, Jieru Zhang, Jyh-Pin Chou,
Alice Hu, and Yung-Kang Peng
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: Chem. Eur. J. 10.1002/chem.202001597
Link to VoR: https://doi.org/10.1002/chem.202001597
01/2020

10.1002/chem.202001597
Chemistry - A European Journal
Nanoisozymes: The Origin behind Pristine CeO2 as Enzyme-mimetics
Zicong Tan, Yu-Cheng Chen, Jieru Zhang, Jyh-Pin Chou, Alice Hu and Yung-Kang Peng*
Zicong Tan, Jieru Zhang, Dr. Yung-Kang Peng
Department of Chemistry, City University of Hong Kong, Hong Kong, China
E-mail: ykpeng@cityu.edu.hk
Dr. Yung-Kang Peng
City University of Hong Kong Shenzhen Research Institute, Shenzhen 518057, China.
Yu-Cheng Chen, Dr. Jyh-Pin Chou, Dr. Alice Hu
Department of Mechanical Engineering, City University of Hong Kong, Hong Kong, China
Keywords: Nanoisozyme, Ceria, Facet-dependent electron density, Enzyme-like activity
It is well-known that the interplay between molecules and active sites on the topmost
surface of solid catalyst determines its activity in heterogeneous catalysis. The electron
density of active site is believed to affect both adsorption and activation of reactant molecules
at surface. Unfortunately, commercial X-ray photoelectron spectroscopy (XPS) often adopted
for such characterization is not sensitive enough to catalyst’ topmost surface. In our opinion,
most researchers fail to acknowledge this point during their catalytic correlation, leading to
different interpretations found in literatures over the past decades. Recent studies on pristine
Cu2O (Nat. Catal. 2019, 2, 889; Nat. Energy 2019, 4, 957) have clearly suggested that the
electron density of surface Cu is facet dependent and plays a key role in CO2 reduction.
Herein, we show that pristine CeO2 can reach 2506%/1133% increase in phosphatase-
/peroxidase-like activity once the exposed surface is wisely selected. Using nuclear magnetic
resonance (NMR) with surface probe, the electron density of surface Ce (i.e. the active site) is
found facet-dependent and the key factor dictating their enzyme mimicking activities. Most
importantly, the surface area of those CeO2 morphologies was demonstrated to become a
factor only when their surface Ce can activate the adsorbed reactant molecules.
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
1. Introduction
Phosphatase is an essential enzyme widely found in living organisms that removes a
phosphate group (PO43−) from phosphorylated substrates by the hydrolysis of the phosphate
ester bond (Scheme 1a). The phosphate ester bond of many biologically important substrates
has been demonstrated to regulate protein activity, DNA/RNA stability and cellular
signaling[1-3]. For example, protein phosphatases can dephosphorylate an amino acid residue
of its protein substrate, which is crucial to regular intracellular signaling with protein kinases.
Nucleotidases can catalyze the hydrolysis of a nucleotide into a nucleoside and a phosphate,
which is essential to maintain a balanced nucleotides and nucleosides with nucleotide kinases
(cellular homeostasis). Despite Enzyme Commission identifies over one hundred distinct
phosphatase families, all phosphatases basically catalyze the same hydrolysis reaction
(Scheme 1a). Recently, CeO2 has been demonstrated a promising material to mimic
phosphatase via the interaction between surface Ce and substrates’ phosphoryl oxygen (P=O)
for the subsequent ester hydrolysis[4,5]. A surface Ce4+-mediated pathway was firstly
suggested by Tan et al. using para-nitrophenyl phosphate (p-NPP) as a substrate[6]. However,
Baker and colleagues later found the mimicking activity was inhibited once surface Ce3+ is
oxidized to Ce4+[7]. A balanced surface Ce4+/Ce3+ ratio[8] and surface oxygen vacancy[9,10] have
also been proposed recently the decisive factor affecting CeO2-catalyzed dephosphorylation.
Accordingly, the role of surface Ce in this reaction still remains unclear for pristine CeO2 not
to mention for their derivatives (i.e. CeO2-based composites/supports)[6-11]
.
In addition to phosphatase, the redox property of CeO2 has been also reported can
mimic various peroxidases by oxidizing corresponding substrates in the presence of H2O2
(Scheme 1b). For example, the reduction of H2O2 catalyzed by CeO2 can subsequently
oxidize substrates such as tetramethylbenzidine (TMB) of horseradish peroxidase (HRP),
nicotinamide adenine dinucleotide (NADH) of NADH peroxidase and glutathione of
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
glutathione peroxidase (GPx)[4,5]. As H2O2 plays a critical role in redox biology and cell
signalling, those peroxidases are known to maintain the level of H2O2 in the cell below the
toxic threshold, protecting biologically important molecules from oxidative damage induced
by excessive H2O2. The mis-regulation of H2O2, in the long term, could lead to various
cardiovascular diseases, cancer and the disorder of neurodegeneration. Qu’s group attributed
the peroxidase-like activity of CeO2 to surface Ce3+ species[12]. However, Yang et al. later
showed that CeO2 morphologies with similar level of Ce3+/Ce4+ ratio and oxygen vacancies
can display different peroxidase-like activities[13]. Despite the sensing of H2O2 by the surface
Ce is believed the key step affecting its mimicking activity[12-15], an in-depth study of the
interplay between them is still lacking nowadays.
In our opinion, different interpretation found in literatures above is due to the heavy
reliance on X-ray photoelectron spectroscopy (XPS) for active site characterization[16,17]. The
substrate molecules of these two enzyme reactions, p-NPP and H2O2, interact only with Ce
atoms on the topmost surface of CeO2 (also for most CeO2-catalyzed reactions[16,17]).
However, the XPS often collects and averages signals > 10 atomic layers from the surface due
to the long escaping depth of photoelectrons[18]. For example, the penetration depth for Ce3d
photoelectrons with 550 eV kinetic energy can be up to 7 nm from CeO2 surface using
commercial XPS[19]. Since the electron density of surface Ce is believed different from their
counterparts in subsurface/bulk, it makes no sense to correlate XPS results with the catalytic
reaction only involves Ce on the topmost surface. Unfortunately, this point has been ignored
by most researchers over the past decades[4-17]
.
Herein, probe-assisted NMR was employed to investigate the electron density of
surface Ce among CeO2 morphologies. With supports from calculation and a wide range of
techniques, the electron density of surface Ce was found facet-dependent and significantly
affects the activation of adsorbed p-NPP/H2O2. We demonstrated that more than 2506% and
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
1133% increase in phosphatase-/peroxidase-like activity can be easily achieved on pristine
CeO2 once its exposed surface is wisely chosen for future engineering.
2. Results and Discussion
2.1. Characterization and Mimicking Testing of CeO2 Morphologies
Octahedron, rod and cube shaped CeO2 crystallites were firstly prepared according to
literatures[10,20]. Figure 1 shows the corresponding TEM and HRTEM images. The
octahedron is about 20 nm in size (Figure 1a) and enclosed by (111) facets as evidenced by
the subsurface lattice spacing of 0.32 nm (Figure 1d). The rod-shaped CeO2 with length up to
200 nm has a uniform diameter around 10 nm (Figure 1b). Even though three lattice spacings
of (111), (110) and (100) can be identified on rod (Figure 1e), we assume herein the majority
of rod surface is terminated by (110) facet as often assigned[10,17,20]. CeO2 cube (Figure 1c)
has a wider size distribution ranging from 15 to 60 nm with surface mainly enclosed by (100)
facet (Figure 1f). Scanning electron microscopy and the corresponding elemental analysis
(Figures S1) suggest high shape homogeneity with cerium (Ce) to oxygen (O) ratio around
0.5 for all morphologies. Their crystallographic nature was confirmed by powder X-ray
diffraction (XRD, Figure S2). Commercial XPS was also attempted to gain insight into the
chemical state of surface Ce (Figure S3). Unfortunately, no apparent difference in XPS Ce3d
spectra can be observed between morphologies.
Phosphatase-like activity of CeO2 morphologies at 25°C were then evaluated using p-
NPP as substrate in Figure 2a-c. Given that p-NPP and its hydrolyzed form, para-nitrophenol
(p-NP), show different absorption at 310 nm and 405 nm, UV−vis spectroscopy was adopted
to evaluate the corresponding activity (Figure 2a). Figure S4 shows the time-dependent
UV−vis spectra of their conversion over CeO2 morphologies at 25 °C. As shown in Figure
2b, octahedron outperforms the other two morphologies with the yield of p-NP (91%)
followed by rod (69%) and cube (3%) by 8 hours. The first order k constant was calculated
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
(hour-1) 0.2706 for octahedron, 0.1927 for rod and 0.0108 for cube (Figure 2c & Figure
S5a). Surface area (Table S1) clearly plays a neglected role that the rod (95 m2/g) possesses
more than twice of the octahedron (42 m2/g) while exhibits lower activity. Moreover, ~70%
more surface area of the octahedron compared to 25 m2/g of the cube cannot explain the
2506% increase in the corresponding k constant. The peroxidase-like property of those CeO2
morphologies was then examined in Figure 2d-f. UV−vis spectroscopy was also adopted here
because the reduction of H2O2 to water by the surface Ce can initiate TMB oxidation to
generate a color change of solution at 652 nm (i.e. the formation of TMB(ox), Figure 2d). In
stark contrast to the phosphatase-mimicking results (Figure 2b), the octahedron is nearly
inactive as peroxidase mimetics while the cube shows very high mimicking activity (Figure
2e). The cube thus exhibits largest k constant (10-4 min-1) of 7.0005 followed by rod (2.1509)
and octahedron (0.6177) (Figure 2f & Figure S5b). Notice that the very active octahedron in
dephosphorylation is now 1133% less active than the cube and their surface area is clear not
the decisive factor (Table S1). Such extreme phosphatase-/peroxidase-like activities obtained
among CeO2 morphologies suggest that the reactivity of surface Ce is closely associated with
their host facets.
Literatures have often relied on commercial XPS as the key surface tool for both
qualitative and quantitative correlation of facet activity. However, Ce3d photoelectrons excited
by Al source (1486.6 eV) with kinetic energy 550 eV its escaping depth can be up to 7 nm
from the surface[19]. Noted that this value (i.e. 7 nm) is even larger than the size of ultra-small
CeO2 catalysts used in many literatures[16,17]. The signal of Ce atoms on the topmost CeO2
surface is thus averaged by subsurface Ce atoms without involving in catalysis, leading to
similar XPS Ce3d pattern between morphologies (Figure S3). Unfortunately, most researchers
fail to acknowledge this point and even try to establish catalytic correlation based on the
deconvoluted XPS. Same method was adopted here to show how unconvincing the obtained
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
correlation could be (Figure S3). The Ce4+/Ce3+ ratio was then calculated 10.03 for
octahedron > 5.07 for rod > 3.94 for cube (Table S1). This order seems can rationalize their
phosphatase-like activity (Figure 2c) since Ce4+ is more acidic (cf. Ce3+) to catalyze the ester
hydrolysis and hence the octahedron with higher Ce4+/Ce3+ ratio exhibits better performance.
However, no literature[6-11] can explain why the Ce4+ portion (90.93%) of octahedron is very
active in dephosphorylation while it becomes inactive for cube with only 10% less of Ce4+
(79.76%) (Figure 2a-c). On the other hand, it is believed that Ce3+ requires lower
overpotential for H2O2 reduction (cf. Ce4+) and hence the cube with the lowest Ce4+/Ce3+ ratio
provides higher peroxidase-like activity. However, literatures[12-15] again failed to argue why
the Ce3+ portion of octahedron (9.07%) is inactive here while it suddenly became active for
cube with doubled Ce3+ portion (20.24%) (Figure 2d-f). Above discussion clearly suggests
that XPS not only has the detection depth issue for catalytic correlation the provided
“discrete” oxidation state (e.g. 3+ and 4+ for Ce) is also not a proper descriptor.
2.2. Surface Structure Analysis and Probe-assisted NMR
From our point of view, the extreme behavior of CeO2 morphologies observed in
Figure 2 should associate with their surface physiochemical properties and hence the
interaction with reactants. Structural analysis in Figure 3a-c shows that the Ce atom on
CeO2(111) surface (Figure 3a) has only one coordination unsaturated site (cf. 8-coordinated
Ce in bulk) while it has two such sites on that of CeO2(110) (Figure 3b) and CeO2(100)
(Figure 3c) surfaces. Since the electron density of surface Ce is inversely proportional to its
coordination number, the 6-coordinated Ce atom on the CeO2(110) indeed exhibits higher
electron density than its 7-coordinated counterpart on the CeO2(111) surface. While for
surfaces with 6-coordinated Ce, the CeO2(100) can accommodate higher electron density than
the CeO2(110) presumably due to the difference in their surface energy[16,17]. According to
above structural analysis, the electron density of surface Ce should change “continuously”
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
with their host surfaces in the order of (100) > (110) > (111) (cf. 3+ or 4+ “discretely”
provided by XPS). This facet-dependent property of surface Ce is believed to affect the
corresponding adsorption/activation energy of reactants, leading to the extreme CeO2
activities observed in Figure 2.
Using 31P NMR with 31P-containing molecules as surface probe has been shown a
promising method for the surface study of metal oxides[21-27]. Trimethylphosphine oxide
(TMPO) was selected here not only for above purpose but also to investigate the interaction
between surface Ce and the P=O of p-NPP by 31P NMR (Figure 3d). The P=O group of
TMPO can specifically bind to surface Ce cation (Lewis acid, LA) through Lewis acid-base
interaction (i.e. P=O→Ce). Since the surface Ce with higher acidity (i.e. lower electron
density) generates stronger bond with TMPO and pushes δ31P to positive ppm, the δ31P can be
used to evaluate the electron density of surface Ce among CeO2 facets[18]. As shown in Figure
3e, the δ31P (ppm) varies with CeO2 morphologies from 50.2 for octahedron, 47.2 for rod to
45.1 for cube, suggesting that the electron density of surface Ce is facet-dependent (100) >
(110) > (111) as predicted in Figure 3a-c. This result can be evidenced by a recent STEM-
EELS study showing that the oxidation state of Ce atoms on (111) facet is Ce4+-like (more
acidic) while they are Ce3+-like (less acidic) on (100) facet[28]
.
2.3. The Origin of Pristine CeO2 as Phosphatase Mimetics
As shown in Figure 4a, CeO2 octahedron with (111) terminal surfaces shows a strong
STEM-EELS signal of phosphorus from the adsorbed p-NPP while only a weak signal can be
observed for that of CeO2 cube terminated by (100) surfaces (see experimental section for
details). This result suggests that the electron density of surface Ce can significantly affect p-
NPP adsorption among CeO2 morphologies. Since the dephosphorylation of p-NPP is
activated by the surface Ce, TMPO was further adopted to investigate the interplay between
P=O group and surface Ce among morphologies by X-ray absorption spectroscopy (XAS)
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
(Figure 4b and Figure S6a). Figure 4c summarizes the corresponding curve-fitting results.
The decrease in the percentage of coordination unsaturated Ce (Ceun) upon TMPO adsorption
suggests the chelating of P=O group to surface Ceun. A significant drop of Ceun percentage can
be observed for both octahedron and rod after TMPO adsorption while not for cube, implying
that the Ce atoms on (111) and (110) surfaces can bind strongly with p-NPP (via P=O→Ce)
and hence activate subsequent ester hydrolysis (Figure 2a-c). This observation can be further
supported by the corresponding local structure analysis (Figure S6b). The averaged Ce
coordination number indeed increases from 6 to around 7.5 for octahedron and rod in the
presence of TMPO, while it remains almost the same for cube (Table S2).
The strong adsorption of p-NPP on CeO2 surface is believed to lower its activation
energy for subsequent ester hydrolysis. To confirm this, we further carried out this reaction at
different temperatures over CeO2 morphologies for kinetic study (Figure S7 and Table S3).
The activation energy (kJ/mol) was obtained 44.24 for octahedron, 50.81 for rod and 83.69
for cube (Figure S7d), indicating that the surface Ce with lower electron density can indeed
activate ester hydrolysis easily and hence accelerate this reaction. Surface area plays a very
minor role here since no enhanced activity can be obtained for cube (25 m2/g) when its testing
amount was doubled (i.e. 8 mg) or even quadrupled (i.e. 16 mg) in this reaction (Figure S8).
The reaction energy profile of p-NPP dephosphorylation on CeO2(111) facet was recently
calculated by Xu’s group.[29]. The desorption of phosphate from CeO2(111) was found with
the highest energy barrier (~115 kJ/mol) and hence the rate-limiting step instead of the P=O
bond activation concluded above. However, this value deviates significantly from the
activation energy of our (111) enclosed octahedron (i.e. 44.24 kJ/mol). The deviation can be
attributed to the solvation effect in aqueous solution which can facilitate the removal of
phosphate from CeO2 surfaces and hence free the active site for next catalytic cycle[30]
.
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Accordingly, we believe the rate-limiting step for this reaction is the activation of P=O bond
(by surface Ce) rather than the desorption of phosphate from CeO2 surface.
Figure 4d summarizes the correlation between the Ce electron density (using δ31P as
descriptor), activation energy (kJ/mol) and the resulting rate constant (hour-1) among CeO2
morphologies for this reaction. In addition to the facilitated p-NPP adsorption, the increase of
Ce acidity from 45.1 ppm of cube, 47.2 ppm of rod to 50.2 ppm of octahedron can lower the
activation energy of p-NPP (green line) and hence increases the corresponding phosphatase-
like activity (brown line). Since rod is active while cube is inactive at 25°C (Figure 2b), we
believe the threshold of Ce electron density (δ31P) for catalyzing this reaction is between 47.2
ppm and 45.1 ppm.
2.4. The Origin of Pristine CeO2 as Peroxidase Mimetics
For the peroxidase-mimicking reaction (Figure 2d-f), the oxidation of TMB (i.e. color
indicator) is catalyzed by the reduction of H2O2 to water in the presence of CeO2. The rate-
limiting step of this reaction is believed the sensing of H2O2 by the surface Ce[4,5]. To confirm
this, steady-state Michaelis–Menten kinetics was adopted to study the affinity of H2O2 and
TMB to CeO2 surfaces (Figure S9). A series of Michaelis–Menten constant (Km) and
maximum initial rates (Vmax) were obtained in Table 1 (see experimental section for details).
In general, a lower Km value suggests a higher affinity of substrate to catalyst and hence
higher enzymatic activity[4,5]. As expected, CeO2 cube shows the lowest Km value towards
H2O2 followed by rod and octahedron, matching well with the order of their H2O2 Vmax
(Table 1) and k constant (Figure 2f) (i.e. cube > rod > octahedron). Given that octahedron
and cube with comparable TMB Km (Table 1) while exhibit extreme activity as peroxidase
mimetics, the sensing of H2O2 by the surface Ce is thus confirmed the key step dictating their
peroxidase-like activity.
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Since the adsorption behavior of H2O2 on CeO2 morphologies cannot be visualized by
STEM-EELS (Figure 4a), density function theory (DFT) calculation was carried out to study
its adsorption configuration/energy on CeO2 surfaces (Figure S10). As shown in Figure 5a, a
H2O2 adsorption energy (Ead) of −0.70 eV can be obtained when the H2O2 molecule uses one
oxygen atom to bind to the 7-coordinated Ce on the (111) surface. A stronger Ead of −1.33 eV
can be obtained on the (110) surface with 6-coordinated Ce due to the formation of Ce-
peroxido species (Figure 5b). Similar Ce-peroxido species can be also found on the (100)
surface with surface 6-coordinated Ce (Figure 5c). However, one oxygen atom of Ce-
peroxido species tends to refill the structural O vacancy and hence provides the most negative
Ead of −1.95 eV among three surfaces. The adsorption of H2O2 on CeO2 surfaces is thus facet-
dependent in the order of (100) > (110) > (111), matching well with the peroxidase-like
activity of their morphologies (Figure 2f).
In additional to the facilitated H2O2 adsorption on cube (100), the Ce atom with higher
electron density on this surface (Figure 3e) is believed to reduce H2O2 easier (i.e. the
activation of O-O bond). Since Raman band associated with peroxide species on CeO2 can be
monitored at around 870 cm-1[31,32], in-situ Raman spectra were thus performed to monitor the
fate of surface adsorbed H2O2 (Figure S11). Figure 5d shows the corresponding Raman
spectra with enlarged range between 800 cm-1 and 950 cm-1. This signal was further
normalized by the intensity of bulk Ce-O signal at 463 cm-1 as a function of time (Figure 5e).
It is clear that the peak ratio of I872/I463 for the cube returns to its origin in 15 mins after H2O2
addition while this ratio almost remains the same for the octahedron over the reaction time.
This result confirms that the O-O bond of H2O2 can be activated (or weakened) upon its
adsorption on cube (100) but not for octahedron (111).
According to above discussion, the high peroxidase-like activity of cube (100) can be
attributed to (1) the facilitated H2O2 adsorption and (2) the promoted H2O2 activation by
10
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
surface electron-rich Ce species. Figure 5f summarizes the correlation between Ce electron
density (δ31P), H2O2 affinity (Km) and the resulting rate constant (min-1) for this reaction. The
increasing Ce electron density (δ31P) from octahedron (111), rod (110) to cube (100) (green
line), together with the enhanced H2O2 affinity (i.e. lowered Km), can lower the energy barrier
for H2O2 reduction and hence synergistically increase peroxidase-like activity (brown line).
2.5. Discussion
Since the first hydrothermal synthesis of CeO2 morphologies[20], researchers have
dedicated to study their facet-dependent catalysis in many important reactions[16,17]. However,
most researchers simply correlate the observed activity of a catalyst using commercial XPS
without acknowledging how representative the result could be. Even though the adsorption
and activation of reactants on catalyst’ surface are the two key steps in heterogeneous
catalysis[18], the interplay between them has been often ignored in literatures[4-17]. This
explains why very limited enhancements (< 200%) in the phosphatase/peroxidase-like activity
of CeO2 was reached although lots of efforts have been put on their surface engineering over
the past decades[11,12,14,15]. Herein, we demonstrate that 2506%/1133% increase in the
corresponding k constant can be easily reached even on pristine CeO2 once its terminal
surface is correctly selected.
3. Conclusion
Taking two CeO2-catalyzed reactions as an example, we demonstrated here that the
adsorption/activation of reactants at catalyst’ surface is highly relevant to the electron density
of active sites even for catalysts with the same composition/crystal structure. As revealed by
TMPO-31P NMR, the electron density of surface Ce among CeO2 shapes is facet-dependent in
the order of (100) > (110) > (111). For phosphatase-mimicking reaction, Ce atoms on the
(111) surface with low electron density can effectively promote the adsorption and hydrolysis
of p-NPP at room temperature. In stark contrast, the high peroxidase-like activity of CeO2
11
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
cube can be attributed to the facilitated adsorption of H2O2 and its reduction by the electron-
rich Ce atoms on the (100) surface. This explains the extreme 2506%/1133% rate constant
obtained for pristine CeO2 as phosphatase-/peroxidase-mimetics. Most importantly, “surface
area” was found only becomes an important factor when surface Ce can activate the adsorbed
molecules in both reactions. This study thus provides some fundamental but critical insights
into facet-dependent properties for the future design of CeO2-based nanozymes.
4. Experimental Section
Synthesis of CeO2 Octahedron, Rod and Cube: For octahedron, 10 mL of 0.25M
hexamethylenetetramine (HMT) was mixed with 50 mL of 0.0185M Ce(NO3)3 solution. The
reaction was carried out at 75 °C (water bath) for 3 h. Both rod and cube were prepared by
hydrothermally heating the solution of Ce(NO3)3 and NaOH at 100°C and 180°C for 24 hours.
0.744g of Ce(NO3)3.6H2O was mixed with 30 mL of 6 M NaOH solution for rod, while
0.6513g of Ce(NO3)3.6H2O was mixed with 30 mL of 3M NaOH solution for cube. All as-
prepared CeO2 samples were washed three times by DI water and dried in oven at 200 °C for
12 h before use.
Phosphatase-mimic Activity: This reaction was carried out at 25 ºC by mixing 4 mg CeO2 and
10 mL of p-NPP solution (0.2 mg/mL). At each time point, 0.5 mL of the reaction solution
was sampled and diluted to 1 mL with ethanol. CeO2 in the collected solution was further
removed by centrifugation. The supernatants were then evaluated by UV-Vis adsorption
spectra (UV-Vis). The concentration of p-NPP (reactant) and p-NP (product) can be
determined by the absorbance at 310 nm and 405 nm. For kinetic study, this reaction was
further carried out at 0 ºC (ice bath), 50 ºC and 75 ºC (oil bath) during reaction.
Peroxidase-mimic Activity and Michaelis–Menten Kinetic Study: 1 mg of CeO2 morphologies
was firstly dispersed in 1 mL of acetate buffer (50 mM). The reaction can be initiated by
mixing 25 µL of this solution with 0.1 mL of TMB solution (5 mM), 2.8 mL of acetate buffer
12
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
and 0.1 mL of H2O2 (1.5 M). The final concentration of H2O2 is thus 0.5 M. The reaction was
carried out at 25°C and monitored at 652 nm (TMBox) every 12 seconds. For Michaelis–
Menten kinetic study, different concentration of H2O2 and TMB were adopted to evaluate the
corresponding affinity to CeO2 morphologies. For Michaelis–Menten kinetic study, the Km
(Michaelis constant) and the maximal reaction velocity (Vmax) can be obtained by the equation
below:
×[C]
V0= V
(1)
max
K +[C]
m
Where [C] is the concentration of substrate (H2O2 or TMB) and V0 is the corresponding initial
rate calculated from the first 3 min of the reaction.
XPS, EPR and CV Measurements: Ce3d XPS spectra were recorded by a Thermo XPS
equipped Al as X-ray source. The X-band CW electron paramagnetic resonance (EPR)
spectrometer (Bruker EMX) was used for collecting the EPR spectra of CeO2 morphologies.
Cyclic voltammetry (CV) measurement of CeO2 morphologies were carried out by CHI650E
Electrochemistry Workstation using Ag/AgCl (3M KCl) as reference electrode and a Pt wire
as counter electrode.
STEM-EELS Experiment for p-NPP Adsorption: To visualize the adsorption over CeO2
surface, scanning transmission electron microscopy-electron energy loss spectroscopy
(STEM-EELS) mappings of phosphorus and cerium were carried out by JEOL 2100F under
STEM mode. Before conducting the mapping, CeO2 cube and octahedron were pretreated in
p-NPP solution. Briefly, 1 mg of ceria was dispersed in 1 mL of DI water. After that, 0.1 mL
of ceria suspension was mixed with 0.9 mL of p-NPP ethanol solution (0.2 mg/mL). The
mixture was kept in ice bath (0℃) for 5 min to avoid the dephosphorylation of the p-NPP.
Finally, the pretreated samples were analyzed by STEM-EELS mapping.
13
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
In-situ Raman for H2O2 Activation: Raman measurements were carried out by BaySpec's
Agility™ Raman spectrometer using an excitation laser wavelength of 785 nm. The H2O2
solution (6mM) was prepared using acetate buffer (pH=4.0, 50 mM) and 30% H2O2. 3 mg of
CeO2 (cube/octahedron) was loaded on a clean glass slide in 2 mm*2 mm area. Raman
measurements were carried out by BaySpec's Agility™ Raman spectrometer using an
excitation laser wavelength of 785 nm. Before adding H2O2 solution, the Raman spectrum
obtained from raw ceria was set as 0 min. After recoding the spectrum of raw ceria, 10 µL of
H2O2 solution (6mM) was added gently on ceria. The spectra were recorded at 1, 4, 10, 15 and
30 min after the addition of H2O2 solution.
Density functional theory (DFT) calculation: Geometry optimizations, total energies and
charge analysis calculations were performed using the Vienna ab initio simulation package,
with projector augmented wave pseudopotential and the PBE-type exchange-correlation
functional. See supporting information (SI) for details.
X-ray absorption spectroscopy (XAS): Ce L3-edge XAS studies on CeO2 morphologies were
carried out at the National Synchrotron Radiation Research Centre in Taiwan.
31P NMR measurement: The surface of CeO2 sample (100 mg) was firstly cleaned at 373 K
under vacuum (10–1 Pa) before trimethylphosphine oxide (TMPO) adsorption. This sample
was then mixed with CH2Cl2 solution containing 0.1M TMPO for probe adsorption. The
TMPO-adsorbed CeO2 sample was dried under vacuum before 31P NMR measurement at
room temperature (Bruker Avance III 400WB). The 31P chemical shifts were reported
relative to 85% aqueous solution of H3PO4. The quantitative analysis of adsorbed TMPO
molecules was then calculated according to the 31P calibration line established in our previous
literatures[21-25]
.
Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.
14
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Conflict of Interest
The authors declare no competing financial interest.
Acknowledgements
We thank the Hong Kong Research Grants Council (CityU 21301719), National
Natural Science Foundation of China (21902138), City University of Hong Kong (projects
9610415, 9667181 and 9048147). The authors would also like to thank Dr. Tai-Sing Wu and
Prof. Yun-Liang Soo for XAS measurements at National Synchrotron Radiation Research
Centre (Taiwan).
Z.T., J.Z. and Y.K.P. carried out all experiments from particle synthesis,
characterization, catalytic testing to data analysis and wrote the manuscript. All DFT
calculations were performed by Y.C.C., J.P.C. and A. H.. Y.K.P. supervised this project.
Received: ((will be filled in by the editorial staff))
Revised: ((will be filled in by the editorial staff))
Published online: ((will be filled in by the editorial staff))
References
[1] J. Svara, N. Weferling, T. Hofmann, Ullmanns Encyclopedia of Industrial Chemistry
2006, DOI 10.1002/14356007.a19_545.pub2.
[2] M. B. Abou-Donia, Archives of Environmental Health: An International Journal
2003, 58, 484.
[3] M. C. Alavanja, J. A. Hoppin, F. Kamel, Annual Review of Public Health 2004, 25, 155.
[4] C. Xu, X. Qu, NPG Asia Materials 2014, 6, DOI 10.1038/am.2013.88.
[5] H. Wang, K. Wan, X. Shi, Advanced Materials 2018, 31, 1805368.
[6] F. Tan, Y. Zhang, J. Wang, J. Wei, Y. Cai, X. Qian, Journal of Mass Spectrometry
2008, 43, 628.
15
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
[7] M. H. Kuchma, C. B. Komanski, J. Colon, A. Teblum, A. E. Masunov, B. Alvarado, S.
Babu, S. Seal, J. Summy, C. H. Baker, Nanomedicine: Nanotechnology, Biology and
Medicine 2010, 6, 738.
[8] A. Vernekar, T. Das, G. Mugesh, Angewandte Chemie International Editions 2015, 128,
1434.
[9] Z.-G. Wang, W.-Z. Bi, S.-C. Ma, N. Lv, J.-L. Zhang, D.-H. Sun, J.-Z. Ni, Particle &
Particle Systems Characterization 2015, 32, 652.
[10] M. J. Manto, P. Xie, C. Wang, ACS Catalysis 2017, 7, 1931.
[11] Yao, T.; Tian, Z.; Zhang, Y.; Qu, Y., ACS Applied Materials & Interfaces 2019, 11 (1),
195-201. DOI 10.1021/acsami.8b17086.
[12] Tian, Z.; Li, J.; Zhang, Z.; Gao, W.; Zhou, X.; Qu, Y., Biomaterials 2015, 59, 116-124.
DOI https://doi.org/10.1016/j.biomaterials.2015.04.039.
[13] Yang, Y.; Mao, Z.; Huang, W.; Liu, L.; Li, J.; Li, J.; Wu, Q., Scientific Reports 2016, 6
(1), 35344. DOI 10.1038/srep35344.
[14] Naganuma, T., Nano Research 2017, 10 (1), 199-217. DOI 10.1007/s12274-016-1278-4.
[15] Baldim, V.; Bedioui, F.; Mignet, N.; Margaill, I.; Berret, J. F., Nanoscale 2018, 10 (15),
6971-6980. DOI 10.1039/C8NR00325D.
[16] Montini, T.; Melchionna, M.; Monai, M.; Fornasiero, P., Chemical Reviews 2016, 116
(10), 5987-6041. DOI 10.1021/acs.chemrev.5b00603.
[17] Trovarelli, A.; Llorca, J., ACS Catalysis 2017, 7 (7), 4716-4735. DOI
10.1021/acscatal.7b01246.
[18] Peng, Y.-K.; Tsang, S. C. E., Nano Today 2018, 18, 15-34. DOI
https://doi.org/10.1016/j.nantod.2017.12.011.
[19] Kosto, Y.; Zanut, A.; Franchi, S.; Yakovlev, Y.; Khalakhan, I.; Matolín, V.; Prince, K.
C.; Valenti, G.; Paolucci, F.; Tsud, N., Applied Surface Science 2019, 488, 351-359. DOI
https://doi.org/10.1016/j.apsusc.2019.05.205.
16
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
[20] Mai, H.-X.; Sun, L.-D.; Zhang, Y.-W.; Si, R.; Feng, W.; Zhang, H.-P.; Liu, H.-C.; Yan,
C.-H., The Journal of Physical Chemistry B 2005, 109 (51), 24380-24385. DOI
10.1021/jp055584b.
[21] Peng, Y.-K.; Ye, L.; Qu, J.; Zhang, L.; Fu, Y.; Teixeira, I. F.; McPherson, I. J.; He, H.;
Tsang, S. C. E., Journal of the American Chemical Society 2016, 138 (7), 2225-2234.
DOI 10.1021/jacs.5b12080.
[22] Peng, Y.-K.; Fu, Y.; Zhang, L.; Teixeira, I. F.; Ye, L.; He, H.; Tsang, S. C. E.,
ChemCatChem 2017, 9 (1), 155-160. DOI 10.1002/cctc.201601341.
[23] Peng, Y.-K.; Hu, Y.; Chou, H.-L.; Fu, Y.; Teixeira, I. F.; Zhang, L.; He, H.; Tsang, S. C.
E., Nature communications 2017, 8 (1), 1-13.
[24] Peng, Y.-K.; Chou, H.-L.; Tsang, S. C. E., Chemical science 2018, 9 (9), 2493-2500.
[25] Peng, Y.-K.; Keeling, B.; Li, Y.; Zheng, J.; Chen, T.; Chou, H.-L.; Puchtler, T. J.;
Taylor, R. A.; Tsang, S. C. E., Chemical communications 2019, 55 (30), 4415-4418.
[26]Tan, Z.; Wu, T.-S.; Soo, Y.-L.; Peng, Y.-K., Applied Catalysis B: Environmental 2020,
264, 118508.
[27] Tan, Z.; Li, G.; Chou, H.-L.; Li, Y.; Yi, X.; Mahadi, A. H.; Zheng, A.; Tsang, S. C. E.;
Peng, Y.-K., ACS Catalysis 2020, 10 (7), 4003-4011.
[28] Goris, B.; Turner, S.; Bals, S.; Van Tendeloo, G., ACS Nano 2014, 8 (10), 10878-10884.
DOI 10.1021/nn5047053.
[29] Zhao, C.; Xu, Y., Catalysis Today 2018, 312, 141-148.
[30]Śmiechowski, M., Journal of Molecular Structure 2009, 924-926, 170-174.
[31] Wang, Y.-J.; Dong, H.; Lyu, G.-M.; Zhang, H.-Y.; Ke, J.; Kang, L.-Q.; Teng, J.-L.; Sun,
L.-D.; Si, R.; Zhang, J., Nanoscale 2015, 7 (33), 13981-13990.
[32] Issa Hamoud, H.; Azambre, B.; Finqueneisel, G., Journal of Chemical Technology &
Biotechnology 2016, 91 (9), 2462-2473.
17
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Scheme 1. Schematic illustration of (a) dephosphorylation of substrates by phosphatases in
the presence of H2O and (b) oxidation of substrates by peroxidases in the presence of H2O2
(TMB: tetramethylbenzidin; NADH: nicotinamide adenine dinucleotide; GSH: glutathione).
Figure 1. Transmission electron microscope (TEM) images of CeO2 (a) octahedron, (b) rod, (c)
cube and (d-f) the corresponding high-resolution TEM (HRTEM) images.
18
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Figure 2. (a-c) Phosphatase-mimicking reaction and (d-f) peroxidase-mimicking reaction
over CeO2 morphologies. The corresponding (b, e) enzyme-like activity as a function of time
and (c, f) rate constant k at room temperature.
19
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Figure 3. (a-c) Surface crystallographic analysis of CeO2 surfaces showing the difference in
electron density between saturated Ce in bulk (8 coordination) and surface unsaturated Ce.
Noted that the electron density of 7-coordinated Ce on CeO2(111) was set at 0.0000 for
comparison. (d) Schematic illustration of the characterization of surface Ce by TMPO-31P
NMR and (e) the corresponding spectra of CeO2 morphologies.
Figure 4. (a) STEM-EELS Ce/O/P mapping of CeO2 octahedron and cube after p-NPP
adsorption. (b) Schematic illustration of the characterization of surface Ce by TMPO-XAS.
(c) The change of coordination unsaturated/saturated Ce w/o TMPO adsorption among CeO2
morphologies. (d) The correlation of Ce electron density with the corresponding k constant
(hour-1) and activation energy (kJ/mol) of CeO2 morphologies.
20
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Figure 5. The calculated adsorption configuration and energy of H2O2 on CeO2 (a) (111), (b)
(110), (c) (100) surfaces (also see Figure S10). (d) The time-dependent in-situ Raman results
of CeO2 octahedron and cube in the presence of H2O2 and (e) the corresponding change of
peak ratio I872/I463. (d) The correlation of Ce electron density with the corresponding k
constant (hour-1) and H2O2 Km (M) of CeO2 morphologies.
Table 1. The summary of the Michaelis–Menten constant (Km) and the maximum reaction rate
(Vmax) of H2O2 and TMB among CeO2 morphologies.
H2O2
TMB
Km (M)
2.2614
1.1328
0.1236
Vmax (µM/s)
0.0135
Km (mM)
0.1019
0.2439
0.1021
Vmax (µM/s)
0.0078
Octahedron
Rod
0.0198
0.0157
0.0284
0.0460
Cube
21
This article is protected by copyright. All rights reserved.

10.1002/chem.202001597
Chemistry - A European Journal
Table of Contents
The electron density of surface Ce is revealed facet-dependent among CeO2
morphologies in the order of cube (100) > rod (110) > octahedron (111), which significantly
affects their enzyme mimicking activities. It is concluded for pristine CeO2 that the Ce atoms
on octahedron (111) with lower electron density can facilitate its phosphatase-like activity,
while its counterpart on cube (100) is the key for high peroxidase-like activity.
Nanoisozyme, Ceria, Facet-dependent electron density, Enzyme-mimetic activity
Zicong Tan, Yu-Cheng Chen, Jieru Zhang, Jyh-Pin Chou, Alice Hu, and Yung-Kang Peng*
Nanoisozymes: The Origin behind Pristine CeO2 as Enzyme-mimetics
22
This article is protected by copyright. All rights reserved.