Biosensors and Bioelectronics 92 (2017) 294–304

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

journal homepage: www.elsevier.com/locate/bios

Photoelectrochemical enzymatic biosensors MARK

⁎
Wei-Wei Zhao , Jing-Juan Xu, Hong-Yuan Chen
State Key Laboratory of Analytical Chemistry for Life Science and Collaborative Innovation, Center of Chemistry for Life Science, School of Chemistry and
Chemical Engineering, Nanjing University, Nanjing 210023, Jiangsu, PR China

A R T I C L E I N F O A BS T RAC T

Keywords: Enzymatic biosensors have been valuable bioanalytical devices for analysis of diverse targets in disease
Photoelectrochemical diagnosis, biological and biomedical research, etc. Photoelectrochemical (PEC) bioanalysis is a recently emerged
Bioanalysis method that promptly becoming a subject of new research interests due to its attractive potential for future
Enzyme bioanalysis with high sensitivity and specificity. PEC enzymatic biosensors integrate the inherent sensitivities of
Biosensors
PEC bioanalysis and the selectivity of enzymes and thus share their both advantages. Currently, PEC enzymatic
Review
biosensors have become a hot topic of significant research and the recent impetus has grown rapidly as
demonstrated by increased research papers. Given the pace of advances in this area, this review will make a
thorough discussion and survey on the fundamentals, sensing strategies, applications and the state of the art in
PEC enzymatic biosensors, followed by future prospects based on our own opinions. We hope this work could
provide an accessible introduction to PEC enzymatic biosensors for any scientist.

1. Introduction and interested readers are referred to these well-documented litera-

In 1956, the invention of oxygen electrode by Leland C. Clark, Jr.
has ushered in a new era in biosensor field and also established himself
as the father of the biosensor concept (Clark and Lyons, 1962; Clark,
1956). In Clark oxygen electrode, the glucose was entrapped by a
dialysis membrane, and the decreased concentration of measured
oxygen was proportional to glucose concentration. This work was so
successful that it catalyzed numerous variations using similar design
and many other (oxidase) enzymes. Since this pioneering work, the
biosensor field has grown enormously and, especially, the enzymatic
biosensors have been one of the most frequently applied in the
biomedical field (Turner et al., 1987; Wilson and Hu, 2000).
Currently, enzymatic biosensors have been valuable bioanalytical
devices for qualitative and quantitative analysis of diverse targets in
disease diagnosis, biological and biomedical research, food safety and
environmental monitoring. In a typical enzymatic biosensor, the
enzyme is the most essential component since it offers the selectivity,
and enzyme electrodes thus combine the high specificity of the enzymes
with the particular advantages of electrochemical detection, such as
high sensitivity, low cost and simple instrumentation (Ronkainen et al.,
2010; Thevenot et al., 1999). Historically, the enzyme electrodes can be
divided into three generations, i.e., the first generation of oxygen-
based, the second generation of mediator-based, and the third genera-
tion of direct electrochemistry-based. Conventional electrochemical
enzymatic biosensors have been discussed in several excellent reviews,
tures (Kimmel et al., 2012; Privett et al., 2008, 2010; Zhu et al., 2015).
With the ever-increasing needs for ultrasensitive bioanalysis, the
analytical chemists have always been pursuing advanced bioanalytical
modalities compatible with future requirements (Chen et al., 2016a; Li
and Shi, 2014). Among various techniques, photoelectrochemical
(PEC) bioanalysis is a recently appeared method that promptly
becoming a subject of new research interests due to its attractive
potential for future bioanalysis with high sensitivity and specificity
(Freeman et al., 2013; Gill et al., 2008; Yue et al., 2013). Essentially,
the PEC bioanalysis is the advanced generation of the traditional
electrochemical bioanalysis. So it naturally inherits the merits of
electrochemical bioanalysis, but possesses higher sensitivity due to its
unique PEC set-up consisting of total separated energy forms of light
and electricity as excitation source and detection signal, respectively
(Lisdat et al., 2013; Zhao et al., 2015b). Besides, due to the use of an
electronic readout, the instruments of PEC bioanalysis are also gen-
erally simpler, cheaper, and easier to miniaturize than those of optical
modalities. Specifically, the PEC instrument system generally includes
an excitation source (irradiation light, a monochromator, and a
chopper), a cell, and an electrochemical workstation with a three-
electrode system (working, reference and counter electrodes).
Especially, the working electrode comprises of various semiconductor
materials that connected to a metal contact and then to the external
electronics and eventually to a metal counter electrode. Concurrently,
the semiconductor-based working electrode is also connected to the

⁎
Corresponding author.
E-mail address: zww@nju.edu.cn (W.-W. Zhao).

http://dx.doi.org/10.1016/j.bios.2016.11.009
Received 21 September 2016; Received in revised form 27 October 2016; Accepted 2 November 2016
Available online 07 November 2016
0956-5663/ © 2016 Elsevier B.V. All rights reserved.
W.-W. Zhao et al.
counter electrode via the solution that integrates the circuit. Based on
such a system, the studied biochemical information (e.g. the analyte
concentration) associated with a specific biorecognition event could be
elegantly converted by the semiconductors to the output electrical
signals. About the semiconductors, common organic/inorganic ones,
e.g. tris(bipyridyl) ruthenium complex [Ru(bpy)3]2+(bpy=2,20-bipyr-
idine) and CdS quantum dots (QDs), are frequently used in reported
works. With the development of this technique, inorganic semicon-
ductor nanomaterials and nanostructures such as TiO2/ZnO/Cu2O
nanoparticles (NPs)/nanowires (NWs)/nanorods (NRs)/nanocones
(NCs)/nanosheets (NSs)/nanoclusters (NCls)/nanoporous films, and
organic semiconductor species such as porphyrin and its derivatives,
phthalocyanine and its derivatives, azo dyes, chlorophyll, bacteriorho-
dopsin, and polymers such as phenylenevinylene (PPV), poly(thio-
phene), and their derivatives, as well as diverse semiconductor-based
hybrids such as CdS/TiO2, Au NPs/TiO2, porphyrin/ZnO, CdS/gra-
pheme have also been gradually exploited for exquisite PEC bioana-
lysis. As to the biological recognition elements, in addition to the usual
ones (i.e., DNA, antibody and enzyme), some other bioactive materials
(e.g., anticalins, affibodies and nanobodies) and recognition elements
(e.g., whole cells, phages and molecularly-imprinted polymers) will
further offer unique platforms for future developments of novel PEC
bioanalysis. Due to its obvious advantages, PEC bioanalysis has been
dynamically developing and extensively studied for innovative biomo-
lecular probing, especially in the areas of DNA analysis (Zhao et al.,
2014b), immunoassay as well as enzymatic biosensing. In fact, limita-
tions and hurdles are also associated with this technique, for example,
the common photoactive species are subject to low photo-to-current
conversion efficiencies and susceptibilities to photobleaching. Besides,
the established signaling mechanisms are still highly limited and it
remains a challenge to implement these protocols in real-world
applications. For more information on PEC bioanalysis, readers could
use many recent reviews and the references therein to pursue their
interest in this field (Devadoss et al., 2015; Freeman et al., 2013; Gill
et al., 2008; Lisdat et al., 2013; Tang et al., 2015; Yue et al., 2013;
Zhang et al., 2013; Zhao et al., 2014b, 2015b, 2016a, 2016b; Zhou
et al., 2015a).
PEC enzymatic biosensors, a new subclass of enzymatic biosensors,
integrate the inherent sensitivities of PEC bioanalysis and the selectiv-
ity of enzymes and thus share their both advantages. Ever since the
very beginning of PEC bioanalysis, the study on enzymatic biosensors
has been a focus of significant research due to its relevant importance.
In typical PEC enzymatic biosensors, the PEC enzymatic systems, upon
irradiation, could convert the specific biocatalytic events into electrical
signals via the interactions between the semiconductor species and the
biocatalyzed reaction chain. Currently, along with PEC DNA biosen-
sors, PEC enzymatic biosensors has also become a hot topic in this area
and the recent impetus has grown rapidly for the advanced PEC
enzymatic biosensors, as demonstrated by increased research papers
(Guo et al., 1995; Han et al., 2013; Huang et al., 2005; Lee et al., 2016;
Liu et al., 2007, 2014b; Ren et al., 2009; Stoll et al., 2006; Swainsbury
et al., 2014; Umar et al., 2009; Zayats et al., 2003; Zhang et al., 2015;
Zhou et al., 2005; Zhu et al., 2007). Previously, we have illustrated the
state of the art in the broad field of PEC bioanalysis, within which the
recent progresses in the subfield of enzymatic biosensing were dis-
cussed briefly (Zhao et al., 2015b). Beyond that, no effort has yet been
made for addressing specifically on the PEC enzymatic biosensors. On
the other hand, the summary about the PEC DNA biosensors has been
presented in an in-depth work and has been well-received previously
(Zhao et al., 2014b). Given the pace of advances in PEC enzymatic
biosensors, this review, using the selected typical examples, will make a
thorough discussion and survey on the fundamentals, sensing strate-
gies, applications and the state of the art in PEC enzymatic biosensors,
followed by future prospects in this area based on our own opinions.
This work will serves as a useful source to inform the interested
audience of the developments in PEC enzymatic biosensors.
295
Biosensors and Bioelectronics 92 (2017) 294–304
Fig. 1. General PEC enzymatic biosensor design. The major process involves the enzyme
immobilization on the photoelectrode substrate as the recognition elements for the
subsequent biocatalytic transformation.
2. PEC enzymatic biosensors
PEC enzymatic biosensors principally involves tracking the elec-
trical signal prior to and after the enzymatic transformation. In a
typical configuration, as shown in Fig. 1, the biorecognition elements of
enzymes are initially confined onto the electrode surfaces, and the
subsequent biocatalytic events would then be recorded by the electrical
signals. This section will first introduce the fundamentals of enzymes
and common enzymatic reactions (Section 2.1), and then summarize
the enzyme immobilization (Section 2.2), followed by the discussion of
PEC enzymatic biosensors according to the respective configurations,
i.e. first generation (Section 2.3), second generation (Section 2.4), third
generation (Section 2.5) and other (Section 2.6) type. Incidentally, the
detailed summary about the PEC transducers have been well-docu-
mented and will not be discussed here, interested readers are referred
to comprehensive reviews in this topic (Devadoss et al., 2015; Tang
et al., 2015; Zhang et al., 2013; Zhao et al., 2015a; Zhou et al., 2015a).
2.1. Common enzymes
Enzymes, generally globular proteins composed of linear chains of
amino acids that fold into a 3D structure, are highly efficient macro-
molecular biological catalysts that could accelerate, or catalyze chemi-
cal reactions. Different from most other catalysts, the sequences of the
amino acids specify the unique 3D structures of enzymes and in turn
endow them much high catalytic activities and specificities,
making them selective for one type of substrates molecules. In
enzymes, the active sites, usually a groove or pocket of the enzymes
where substrate molecules bind and undergo a chemical reaction,
consist of one or more binding sites where residues orient the
substrates via forming temporary bonds and the neighboring cataly-
tic sites where residues catalyze specific reactions of substrates and
convert them into products in normal catalytic cycles. The enzyme–
substrate interactions can be characterized by Michaelis–Menten
kinetics. The active sites can catalyze the specific chemical reactions
repeatedly as their residues are not altered at the end of the reaction,
while the remaining enzyme structure serve to maintain the precise
orientation and dynamics of the active sites. Enzyme structures may
also contain allosteric sites where the binding of a small molecule
could induce a conformational change that increases or decreases
activity. Comparing with those enzymes free of additional components
to show full activity, some other enzymes necessitate non-protein
cofactors to be bound for activity. Cofactors can be either inorganic
(e.g., metal ions Mg2+, Cu+, Mn2+, or iron-sulfur clusters), or organic
compounds (e.g., flavin and heme) that could sometimes further
divided into coenzymes and prosthetic groups. Coenzymes, small
organic molecules (e.g., dihydronicotinamide adenine dinucleotide
(NADH), NADPH, biotin, lipoamide, flavin adenine dinucleotide
(FAD), and adenosine triphosphate (ATP)) that can be loosely or
W.-W. Zhao et al. Biosensors and Bioelectronics 92 (2017) 294–304

tightly bound to an enzyme, could transport chemical groups from one Table 1
enzyme to another. For example, NAD or NADP+ carries the hydride Common enzymes with specific catalytic mechanisms.
ion (H−), ATP carries the phosphate group, coenzyme A carries the
Enzymes Mechanisms/ Typical reactions
acetyl group, and folic acid carries the formyl, methenyl or methyl functions
groups. Because coenzymes are chemically altered due to the enzyme
action, they could be recognized as a special class of substrates, or Oxidase Catalyze an oxidation- AH2+O2→A+H2O2
reduction reaction
second substrates, which are common to many different enzymes. For
typically with O2 is
instance, several hundreds of enzymes are known to use the coenzyme reduced to H2O or
NADH. When the cofactor that is tightly or even covalently bound it H2O2;
can be termed a prosthetic group. Enzyme activities can be affected by e.g. Glucose oxidase (GOx) e.g. GOx catalyzes the β-D-glucose+O2→D-
inhibitors (e.g. many drugs and poisons) and activators. Inhibitors oxidation of glucose to glucono-δ-lactone+H2O2
H2O2 and D-glucono-δ-
are molecules that bind to enzymes and decrease their enzymatic
lactone;
activities, whereas activators are molecules that bind to enzymes and Peroxidase Catalyze the oxidization AH2 + H2O2→A + 2H2O
increase their enzymatic activities. According to the International of substrate generally
Union of Biochemistry and Molecular Biology, the top-level classifica- using H2O2 as the
oxidizing agent;
tion of enzymes is EC 1 of oxidoreductases (catalyze oxidation/
e.g. Horseradish HRP was found in the 4-chloro-1-naphthol
reduction reactions), EC 2 of transferases (transfer a functional peroxidase (HRP) roots of horseradish; +H2O2→benzo-4-chloro-
group, e.g. a phosphate group), EC 3 of hydrolases (catalyze the hexadienone+2H2O
hydrolysis of various bonds), EC 4 of lyases (cleave various bonds by Phosphatase e.g. Remove a phosphate Ascorbic acid 2-
means other than hydrolysis and oxidation), EC 5 of isomerases Alkaline phosphatase groups from its phosphate→ascorbic acid
(ALP) substrate;
(catalyze isomerization changes within a single molecule), as well as
Dehydrogenase Oxidize a substrate by AH + B→A + BH
EC 6 of ligases (join two molecules with covalent bonds). Such e.g. Glucose transferring a hydrogen D-glucose+ubiquinone→
classification broadly classifies the enzymes based on their functions dehydrogenase (GDH) to an electron acceptor, D-glucono−1,5-lactone

and mechanisms. Besides, the specific name of an enzyme is often usually NAD+ or FAD; +ubiquinol
Acetylcholinesterase Catalyze the hydrolysis Acetylcholine→acetate
derived from its substrate or the chemical reaction it catalyzes, for
(AChE) of acetylcholine and +choline
example, glucose oxidase, formaldehyde dehydrogenase and DNA some other choline
polymerase. Different enzymes that catalyze the same chemical reac- esters;
tion are called isozymes. β-galactosidase (β-Gal) Catalyze the hydrolysis β-D-galactosides+H2O→
Significantly, nucleases, most of which belong to EC 3 of hydro- of β-galactosides into D-galactose + alcohol
monosaccharides via
lases, are a special family of enzymes capable of cleaving the phospho-
breaking the glycosidic
diester bonds between the nucleotide subunits of nucleic acids. bond;
Nucleases can be generally classified into two kinds of ribonuclease Proteases/peptidase Catalyze the hydrolysis Protein+H2O→product A
(RNase) and deoxyribonuclease (DNase). RNase catalyzes the of the peptide bonds +product B
that link amino acids
hydrolytic cleavage of phosphodiester linkages in the RNA backbone
together in a
and degrades the RNA, while DNase degrades DNA into smaller polypeptide chain;
components. According to the action mode, nucleases are usually Kinase Catalyze the transfer of ATP+substrate-
further divided into endonucleases and exonucleases. a phosphate moiety OH→ADP
Endonucleases cleave the phosphodiester bond in the middle (endo) from a high energy and +phosphorylated
phosphate-donating substrate
of a polynucleotide chain, whereas exonucleases cleave nucleotides one
molecule (such as ATP)
at a time from the end (exo) of a polynucleotide chain. Some to their substrate
endonucleases (e.g. deoxyribonuclease I) cut DNA without regard to molecule;
sequence (nonspecifically), while many restriction endonucleases Nucleases Cleave the DNA→DNA fragment A
or restriction enzymes cleave only at very specific nucleotide sequences phosphodiester bonds +fragment B
between the nucleotide
called the restriction sites (specifically). Incidentally, the cleaved subunits of nucleic
fragments can be joined by DNA ligase via catalyzing the formation of acids;
a phosphodiester bond. Differently, DNA polymerases are enzymes DNA ligase Join DNA strands DNA strand A+strand
that synthesize DNA molecules from deoxynucleoside triphosphate together by catalyzing B→DNA
the formation of a
(dNTP, the building blocks of DNA).
phosphodiester bond;
Although most are proteins, a few enzymes are catalytic nucleic acid DNA polymerases Synthesize DNA dNTP+DNAn→
molecules, i.e. ribozymes (RNAzymes) and deoxyribozymes molecules from dNTP diphosphate + DNAn+1
(DNAzymes). Note that these enzymes are named with the word Ribozymes (RNAzymes) Catalyze specific RNA message→cleaved
ending in “-zyme” rather than “-ase”. These catalytic nucleic acid biochemical reactions RNA messages
Deoxyribozymes Catalyze specific Substrate DNA strand
molecules are DNA or RNA molecules that are capable of catalyzing (DNAzymes) biochemical reactions (with cleavage
specific biochemical reactions. Like many protein enzymes, metal site)→DNA fragment A
binding is also critical to the function of many ribozymes RNAzymes, +fragment B
while most DNAzymes suffer from product inhibition and hence exhibit
only single-turnover behavior. Also note that RNAzymes/DNAzymes
should not be confused with RNA/DNA aptamers which are oligonu- directly, which removes the labor-intensive and time-consuming
cleotides that selectively bind a target, but do not catalyze a specific sample pretreatment procedures. Such inherent selectivity could avoid
chemical reaction. the disturbance of concomitant interfering species in the samples. In
In the biosensors fields, enzymes are the oldest and still most Table 1, we summarize some common enzymes with corresponding
commonly used biorecognition elements. On one hand, due to their mechanisms for PEC bioanalysis.
high efficiency, the enzyme could greatly amplify the signal by
generating numerous detectable products for each enzyme molecule. 2.2. Enzyme immobilization
On the other hand, due to their exquisite specificity for particular
substrates, they can detect individual substances in a sample matrix PEC enzymatic biosensors necessitate the spatial contact between

296
W.-W. Zhao et al.
the PEC transducer and enzymes, so the enzyme immobilization is an
important step for achieving better biosensor performances in terms of
high enzyme loading, bioactivity, stability and thereby the robustness,
rapid response, and high sensitivity of the sensors. Generally, high
levels of enzyme loading are essential to the enhanced sensitivity and to
maximize the signal-to-noise ratio of biosensors. To date, researchers
have established many methods for efficient enzyme immobilization,
mainly including adsorption, covalent binding, affinity binding, as well
as entrapment within certain matrixes.
Adsorption, in the area of probe immobilization, is the adhesion of
biomolecules to an electrode surface, creating a film of the adsorbate on
the surface of the transducer. The adsorption could be classified into
two main kinds, i.e. the chemisorption (e.g. Au-S bonding) and
physisorption (e.g. electrostatic attraction). The former involves a
natural chemical reaction between the substrate surface and the
adsorbate during which new chemical bonds are generated between
them, while the latter represents a process in which the electronic
structure of the molecule is barely perturbed upon adsorption. In PEC
enzymatic bioanalysis, the nanostructures of various semiconductors
could greatly enhance the active surface available for enzyme adsorp-
tion, and the immobilization is mainly electrostatic and influenced by
the surface chemical composition/wettability, pH, the protein charge
and the solution ionic strength. A typical case is that TiO2 nanos-
tructures of large surface areas have been commonly used to immobi-
lize proteins for bioanalytical applications due to their good biocom-
patibility and chemical/thermal stability. The electrostatic interactions
between negatively charged groups on the TiO2 surface and positively
charged surface lysine and/or arginine residues of the protein would
allow the stable and high levels of enzyme adsorption without the loss
of protein structure or activity. For instance, HRP were immobilized on
the photoactive TiO2 nanotubes (TNs) via the adsorption for the
visible-light-activated PEC detection of H2O2 (Chen et al., 2010). ALP
were immobilized onto the TiO2 NPs film for in situ activation of the
inert TiO2 and thus for PEC enzymatic bioanalysis (Zhao et al., 2013a).
Adsorption of enzymes have also been performed on other nanomater-
ials. For example, cytochrome c (cyt. c) was adsorbed onto the Au NPs/
TiO2 nanoneedle film for PEC detection of H2O2 (Zhu et al., 2009a).
Negatively charged HRP was also adsorbed on the surface of CdSe/
mesoporous silica spheres for PEC detection of H2O2 (Yang et al.,
2011). GOx were adsorbed onto the ZnO inverse opal for PEC detection
of glucose (Xia et al., 2014). Sulfite oxidases were adsorbed on the
quantum-dots-modified indium tin oxide electrode for PEC sulfite
detection (Zeng et al., 2015).
Especially, layer by Layer (LbL) assembly has been applied as a
simple and versatile approach for enzyme immobilization. This assem-
bly technique is based on the electrostatic interactions between
sequentially adsorbed layers of oppositely charged materials with wash
steps in between, providing a simple and facile route to fabricate hybrid
thin film structures with controlled thickness, structure, and composi-
tion. The alternating deposition can be achieved by using many
techniques such as immersion, spin, spray, or fluidics, among which
immersion is the most frequently used for enzyme immobilization. For
example, glutamate dehydrogenase (GDH) and CdS-poly (diallyldi-
methylammonium chloride) (PDDA) were alternating adsorbed onto
the surface of the multiwall carbon nanotubes (CNTs) for PEC
dehydrogenase biosensing (Tang et al., 2008). In another work, multi-
layers of CdS and GOx were fabricated by sequential deposition onto
the Nafion (Pt) substrate for PEC glucose biosensor (Sun et al., 2009).
Based on the negatively charged enzymes, positively charged polyelec-
trolyte poly(allylamine hydrochloride) (PAH) and negative charged
sodium poly(styrene sulfonate) (PSS), Lisdat and Parak et al. alter-
nately assembled GOx/PAH and sarcosine oxidase/PAH multilayers to
fabricate the oxygen-sensitive electrodes for PEC glucose (Tanne et al.,
2011) and sarcosine (Riedel et al., 2013) detection. Similarly, they also
developed cyt c/QDs/PAH and ALP/PAH multilayer systems for PEC
protein direct electrochemistry (Göbel et al., 2011) and aminophenyl
297
Biosensors and Bioelectronics 92 (2017) 294–304
phosphate detection (Khalid et al., 2011), respectively. Recently, they
further used this technique to immobilize three enzymes at a CdS/ZnS
QDs electrode for PEC bioanalysis of guanosine monophosphate (Sabir
et al., 2015). Besides, Wang prepared the GOx/PDDA multilayers on
the CdS QDs/NiO electrode for the PEC glucose detection (Wang et al.,
2014).
Covalent binding, another commonly used anchoring approach,
represents the probe immobilization via the formation of covalent
bonds between the functional groups of the enzymes and electrode
materials. For enzymes, it is generally the covalent bonding between
the -NH2 of specific enzymes and the functional moieties or groups on
the activated electrode surface. Most common ways of doing this are
the Schiff base (linking between the -NH2 and -CHO) and carbodiimide
(linking between the -NH2 and -COOH) reactions. To facilitate the
covalent binding, some versatile reagents including chitosan (with
many -NH2 groups), silanization chemicals (e.g. 3-aminopropyl-
triethoxysilane (APTS)) and cross-linking linkers (e.g. glutaraldehyde
(with two -CHO groups) have been frequently applied in specific
immobilization procedures. For example, AChE was covalently linked
to the CdS QDs (Pardo-Yissar et al., 2003) and Au-TiO2-modified Ti foil
(Zhu et al., 2009c) using glutaraldehyde as the bridging unit for PEC
sensing of the enzyme inhibitors. Many other enzymes, such as GOx
(Tel-Vered et al., 2010), cholesterol oxidase (Tang et al., 2014; Yildiz
et al., 2013), alcohol dehydrogenase (Dilgin and Gokcel, 2015; Jafari
et al., 2014b), alcohol oxidase (Yildiz et al., 2014) and adenosine
triphosphatase (ATPase) (Tang et al., 2014), have also used the
glutaraldehyde linkage chemistry for their immobilization. As afore-
mentioned, N-(3-Dimethylaminopropyl)-N-ethyl-carbodiimide hydro-
chloride (EDC) containing the carbodiimide functionality have been
intensively used to activate carboxylic acids towards amide or ester
formation, while the additive of N-hydroxysuccinimide (NHS) is often
added to increase yields and decrease side reactions. In PEC enzymatic
biosensors, such EDC coupling reactions are directed between the
COOH groups on the electrode surface and the NH2 groups of the
enzymes. For instance, GOx was connected to the various QDs (Golub
et al., 2012; Wang et al., 2012; Zhang et al., 2014) using this technique.
The immobilization of some other enzymes such as GDH (Jafari et al.,
2014a) and cytochrome P450 (Qian et al., 2014) have also been
reported. Comparing with adsorption technique, covalent binding is
obviously more robust and has higher stability. However, covalent
binding necessitate more complicated and severe conditions which
might harm the enzymes and thus their catalytic abilities. In fact, in
comparison between LbL adsorption and glutaraldehyde crosslinking,
it was found that glutaraldehyde linking could influence the sensitivity
and the system reproducibility (Tanne et al., 2011).
Entrapment of enzymes within a 3D matrix (e.g. different hydrogel,
xerogel and silica glass) possesses the advantages of enhanced enzyme
loading/decreased enzyme leaking, preserved bioactivity, as well as
protection from the ambient environment. Sol-gel methods could be
used for easy encapsulation of a range of biological macromolecules.
The entrapment of enzymes in the porous and inorganic glass matrix
would permit the diffusion of solution and small analyte molecules
through the glass to the immobilized enzymes. Besides, the high
enzyme loading per unit area and the optical transparency of the glass
makes this approach especially suitable for PEC signal transduction.
For example, silica sol was prepared by the ion-exchange method, and
then QDs-modified electrode was dipped in the mixed solution contain-
ing silica sol, GOx and polyvinyl alcohol to prepare the enzyme
electrode for PEC glucose detection (Du et al., 2012, 2013). However,
the drawbacks of sol-gel fabrication is that it involves high alcohol
concentration and/or extreme pH conditions as well as elevated
temperatures, which are detrimental to the enzymes and would cause
the enzyme denaturation. This has motivated the development and use
of other biocompatible matrixes, which would minimize the enzyme
denaturation during the entrapment. For example, with excellent
biocompatibility, chitosan is an ideal chemical inert, non-toxic biopo-
W.-W. Zhao et al.
lymer with great adhesion properties. It has been used to enwrap GOx
(Zhao et al., 2012a) and AChE (Huang et al., 2013) to fabricate the
specific PEC enzyme biosensors. Lactic dehydrogenase (LDH) was also
entrapped within the CNT-TiO2 NPs-Nafion matrix (Liu et al., 2015).
Affinity binding utilizes the bioaffinity interactions for enzyme
immobilization. For example, AChE antibodies on BiOI Nanoflakes/
TiO2 NPs electrode was used to bind AChE in situ and thus for PEC
enzymactic analysis (Zhao et al., 2013b). Besides, the formation of
coordination biocomplexes or supramolecular such as biotin/avidin,
nitrilotriacetic acid/Cu2+ biotin, phosphonate/Mg2+ phosphate, nitri-
lotriacetic acid/Ni2+ histidine, and adamantane/β-cyclodextrin systems
may also be used for bioreceptor immobilization.
In all, ideal immobilization should ensure the intimate contact
between the enzymes and electrode surfaces while not block the active
sites of enzymes and not alter their geometries severely. Among
aforementioned various immobilization methods, adsorption is the
least stable one. Although with a rather limited lifetime, adsorption is
quite less harmful to the enzymes and thus prepared biosensor is
sufficient for short-term studies and often exhibits superior short-term
performance. Significantly, the LbL adsorption possesses the additional
advantages of tuned deposition cycles and thus the high surface
enzyme density. As compared to the cases of adsorption, the covalent
bonding offers the most stable immobilization followed by cross-
linking and encapsulation. Obviously, in order to protect the enzymatic
activity, such binding process should be performed under as milder
conditions as possible, e.g., near physiological pHs, low ionic strengths
as well as low temperatures. Incidentally, proper entrapment of
enzymes between membranes or covering the enzymes with a mem-
brane on the electrode surface could minimize the interferences, avoid
electrode fouling, and stabilizes the sensor response, etc.
2.3. First generation
PEC enzymatic biosensors are generally based on oxidoreductase
enzymes coupled with PEC detection. The biocatalytic conversion of the
targets is associated to an electron transfer process at the electrode
surface and thus generates the photocurrent signal that is proportional
with the analyte concentration, which is quantified using an ampero-
metric set-up. Note that direct redox reactions between the enzymes
and electrodes are very rare because of the deeply buried active site of
enzymes and thus the quite slow and irreversible direct electron
transfer processes. The widely accepted Marcus theory of electron
transfer also demonstrates that electron transfer decays exponentially
with distance (Bard et al., 1980; Bixon and Jortner, 1999). Under such
circumstances, numerous enzyme biosensors have been developed, and
these biosensors could historically be divided into three generations:
oxygen (co-substrate)-based first generation, mediator-based second
generation, and direct electrochemistry-based third generation. PEC
enzymatic biosensors could also be classified correspondingly.
The first generation utilizes the specificity of an enzyme reaction.
Using the general reaction scheme of S (substrate)+S′ (co-substrate)→
P (product)+P′ (by-product), this strategy detects the enzymatic
conversion by monitoring the consumption of a co-substrate S′ (e.g.
the depletion of O2 by oxidase) or the formation of an electro-active
product P (e.g. the generation of H2O2 by oxidoreductase). Following
this idea, several enzymatic reactions could be integrated with the PEC
electrode if a co-substrate or product can be effectively detected at the
illuminated electrodes.
One typical example is the competitive consumption of dissolved
O2, which acts both as the co-substrate and electron acceptor in a
specific PEC enzymatic system. Such O2-sensitivity-based PEC enzy-
matic biosensor was first exemplified with GOx to analyze its substrate
of glucose by Lisdat et al. (Tanne et al., 2011). As shown in Fig. 2, with
O2 as electron acceptor, the electron flow from the electrode to the
CdSe/ZnS QDs and then to the dissolved oxygen would result in the
generation of cathodic photocurrent. When coupled with GOx, the
298
Biosensors and Bioelectronics 92 (2017) 294–304
Fig. 2. Illustration of the electron transfer steps after illumination of the GOx and CdSe/
ZnS QDs coupled system. A represents the dissolved oxygen (reprinted with permission
from Tanne et al. (2011)).
oxygen-consuming enzyme reaction would cause O2 depletion in front
of the QD surface and thus to the photocurrent signal suppression.
Such competitive O2 reduction process upon the O2-sensitive electrode
could obviously be utilized for the evaluation of the GOx activity and
glucose detection on the basis of the influence of the O2 content in
solution upon the photocurrent generation. Using same signaling
strategy but with different oxygen-sensitive QDs, several groups then
constructed similar systems for the glucose detection (Wang et al.,
2014, 2012; Zhang et al., 2014). Obviously, such strategy can be
integrated with different oxidase to analyze the substrate of the
respective enzyme reaction. For example, by using sarcosine oxidase
(SOD), the biocatalytic O2 reduction induced photocurrent suppression
has been successfully used for sarcosine detection (Riedel et al., 2013).
Nevertheless, some other work gave different explanation upon this
GOx catalytic reaction. For example, with a Nafion(Pt)/CdS/GOx
system and in the presence of dissolved O2, glucose is oxidized under
the catalysis of GOx and generating H2O2 in this redox process. Then
the H2O2 diffuses to the CdS QDs and is oxidized by the valence band
hole. During the oxidation of H2O2 at the semiconductor surface, two
electrons are transferred directly to the electrode and thus generate the
anodic photocurrent. The biosensor was applied to detect glucose since
the photocurrent response of the enzyme electrode increased with
glucose concentration (Sun et al., 2009). Relying on the use of H2O2 as
electron donor or acceptor, some other work has also been reported (Tu
et al., 2014; Zhao et al., 2012a). These explanations should be
attributed to the different surface potentials associated with the used
semiconductor photoelectodes. Note that dissolved O2 and H2O2 are
not only the basic electron acceptor or donor species of QDs but also
the reactant or reaction product of the enzyme-catalyzed reaction.
Significantly,O2 are more efficient than H2O2 for PEC (Zhang et al.,
2014) and also electrochemiluminescent (Jiang and Ju, 2007) proper-
ties of QDs.
Another case in point is that AChE can be utilized in this protocol.
AChE can catalyzes the hydrolysis of acetylthiocholine to generate the
acetate and thiocholine, the latter can act as sacrificial electron donor
and be oxidized at the illuminated semiconductor electrode in the PEC
systems. Apparently, the observed photocurrent indicates the concen-
tration of solution-solubilized substrate. However, since acetylthiocho-
line is an artificial substrate, this enzymatic reaction is mainly directed
for the analysis of enzyme activity or the detection of potential
inhibitors. Using CdS QDs electrode, Willner et al. first reported the
use of such strategy for PEC sensing of the enzyme inhibitors (Pardo-
Yissar et al., 2003). As shown in Fig. 3, at sufficient substrate
concentration, the generation of photocurrents was controlled by the
enzyme activities and the biocatalyzed formation of thiocholine that
scavenges the photogenerated valence-band holes. However, the pre-
sence of inhibitors could restrain the enzyme and thus decrease the
photocurrent generation, paving the way for developing PEC enzymatic
biosensor toward the respective inhibitors. Using such strategy, many
PEC bioanalytical systems have been developed for the detection of
W.-W. Zhao et al.
Fig. 3. CdS QDs/AChE hybrid system used for the PEC detection of the enzyme activity
(reprinted with permission from Pardo-Yissar et al. (2003)).
various AChE inhibitors (Gong et al., 2012; Huang et al., 2013; Li et al.,
2015; Liu et al., 2014a; Zhu et al., 2009c). Incidentally, because of their
operating mechanism and applications, these sensors may be termed as
PEC enzyme inhibition/toxin biosensors.
One more example is the PEC oxidation of NADH, which serves as
the coenzyme to several hundreds of enzymes. NADH and correspond-
ing oxidized form (NAD+) are commonly involved with various
dehydrogenase enzymes and in the corresponding electron transfer
chain of biological systems. NADH could be oxidized in a rather wide
potential range around 0 V vs. Ag/AgCl, and the anodic photocurrent
rises corresponding to the enhanced NADH concentration at proper
electrode potentials. Obviously, the involvement of NADH in many
dehydrogenase reactions could be used for PEC bioanalytical applica-
tions. This has been exemplified by using GDH as label for the
development of PEC immunoassay (Zhao et al., 2014a). GDH could
catalyzes the reaction of β-D-glucose to D-glucono-1,5-lactone while
reducing its cofactor NAD+ to NADH, which was then PEC oxidized at
the photoanode and regenerate NAD+. Note that glucose cannot be
oxidized on the electrode at low potential. For PEC enzymatic biosen-
sing, as shown in Fig. 4, Dilgin et al. electropolymerized hematoxylin
(HT) on poly-amidoamine (PAMAM) dendrimers and then immobi-
lized GDH on the film. Dependent on the NAD+/NADH redox, a
concentration-dependent photocurrent can be obtained for glucose
detection (Dilgin and Gokcel, 2015). Recently, they further reported
works based on the same mechanism (Ertek et al., 2016; Ertek and
Dilgin, 2016). Also on the basis of NAD+/NADH redox, Liu et al.
reported the construction of a PEC LDH biosensor toward lactate
detection (Liu et al., 2015). It is worth to mention that direct
electrochemical oxidation of NADH to NAD+ at the bare electrode
surface often involves high overpotentials and suffers from electrode
surface fouling associated with the adsorption and accumulation of
reaction products. To solve this problem, the electrode surface could be
modified with various electron transfer mediators that serve as efficient
electrocatalytic system for oxidation of NADH. Following this idea,
Salimi et al. conducted the electrochemical oxidation of chlorproma-
zine (CPZ) to chlorpromazine-sulfoxide (CPZ-SO) onto the CdS QDs/
graphene modified electrode. And the electrogenerated CPZ-SO could
serve as the electron transfer mediator to assist the efficient PEC
Fig. 4. Schematic representation of the PEC GDH biosensor (reprinted with permission
from Dilgin and Gokcel (2015)).
299
Biosensors and Bioelectronics 92 (2017) 294–304
oxidation of NADH, which was coupled with alcohol dehydrogenase
(ADH) enzymatic system for the successful detection of ethanol (Jafari
et al., 2014b). Similarly, they further electropolymerized Nile blue (NB)
as the electron transfer mediator and coupled to GDH reaction for the
detection of glucose (Jafari et al., 2014a).
Some other enzymes may also be applied for PEC enzymatic
biosensors. For instance, ALP could catalyze the transformation of its
substrate p-aminophenylphosphate (pAPP) to the reaction product of
p-aminophenol (pAP), which could be oxidized at the illuminated QDs
electrode and generates photocurrent signals for sensitive pAPP
detection (Khalid et al., 2011). Recently, adenosine triphosphatase
(ATPase), a metabolically important enzyme that converts between
reversible oxidation and reduction states, was coupled with Co3O4–
TiO2 photoanode for probing the ATP or cholesterol levels. Specifically,
the photogenerated holes could oxidize the water molecules to H2O2 or
O2. And the O2 formed on the electrode could serve as a charge shuttle
between the photo-excited electrode surface and the redox enzyme that
mimics a simple metabolism process, based on which the could be
successfully probed (Tang et al., 2014).
2.4. Second generation
As mentioned before, the redox centers at active sites of most
enzymes are usually deeply embedded within their tertiary protein
structures that near the centroid of the proteins, and the electrons
produced from the enzyme-catalyzed reaction could not be easily and
rapidly transferred to the electrode surface, causing a rather slow rate
of heterogeneous electron transfer and very limited electrical commu-
nication between them. Therefore proper bridging is needed to
facilitate the electron transfer.
The second generation relies on the use of mediators to shuttle
electrons between the active site of the enzyme and the electrode. Such
mediators should ideally be small, soluble, nontoxic and stable (both in
the oxidized and reduced forms, independent of the pH, and unreactive
with O2) molecules capable of undergoing rapid and reversible redox
reactions. The most common mediators are organometallic redox
compounds, including quinones, quinoid dyes, organic salts, Ru or
Os complexes, and various ferrocene or ferricyanide derivatives. These
mediators are easily reduced by the enzymatic redox reactions and then
oxidized at relatively low potentials to generate current signals at the
working photoelectrodes. The state of the enzyme redox active center
and the substrate S concentration could thus be detected. For example,
as shown in Fig. 5, Liu et al. incorporated [cobalt(o-phen)3]2+/3+ into
the GOx-CdSe@CdS QDs/TiO2 nanocomposite for PEC glucose biosen-
Fig. 5. Photocurrent generation from GOx/CdSe@CdS QDs/TiO2 nanocomposite (rep-
rinted with permission from Zheng et al. (2011)).
W.-W. Zhao et al.
Fig. 6. The photoanode consists of a [Ru(bpy)3]2+-tethered GOx-modified Au electrode (reprinted with permission from Tel-Vered et al. (2010)).
sing. In such a system, [cobalt(o-phen)3]2+ ions behave as donors for
hole scavenging in the illuminated QDs, while the [cobalt(o-phen)3]3+
ions oxidizes the FADH2 groups in GOx catalyzed glucose oxidation.
Since a high potential was used for the analysis (+0.4 V vs. Ag/AgCl),
the photocurrent generated is attributed to the mediated electron
transfer and is also partly due to the direct glucose oxidation by the
generated holes (Zheng et al., 2011).
Significantly, by plugging into enzymes with light, Willner's group
reported the photonic “wiring” of GOx with electrode for construction
of a novel photobiofuel cells. As shown in Fig. 6, Ru(II)-tris-bipyridine
complex [Ru(bpy)3]2+ was initially tethered to the GOx monolayer.
Upon illumination, photoexcitation of the bipyridine complex
[Ru(bpy)3]2+ could result in the injection of the electrons to the Au
electrode and the formation of [Ru(bpy)3]3+, which would then mediate
the oxidation of the FAD center and activate the biocatalyzed oxidation
of glucose. The photocurrent intensity was found dependent on the
variable concentrations of glucose (Tel-Vered et al., 2010). Using
similar protocol, Yildiz then reported the “wiring” of cholesterol
oxidase and alcohol oxidase for the PEC detection of cholesterol and
alcohol, respectively (Yildiz et al., 2014, 2013).
In comparison with the first generation, the second generation
necessitate the use of mediators to replace the co-substrate of S′ (eg. O2
or NADH) as the electron shuttles. If the reaction rate of the mediator
with enzyme is much higher than that of the co-substrate with enzyme,
the second generation type could eliminate the sensor dependence on
O2/NADH, and thus exhibit better performance than the first genera-
tion counterpart. When selecting a mediator, suitable redox potential
range is that within which the biorecognition event could take place
while the possible sample interferents such as urate/ascorbate could
not be oxidized or reduced.
2.5. Third generation
Unlike most other enzymes, a limited number of enzymes (e.g.
heme-containing enzymes such as HRP) and low-molecule weight
redox proteins (e.g. cytochrome c, Cyt c) have proven capable of direct
electron transfer between their active sites and the electrode. On the
other hand, intensive efforts have also been continuously made to
improve the electrical communication between the redox center of the
enzyme and the electrode surface. Particularly, nanomaterials and
nanostructures have been found to be advantageous for direct electron
transfer since their size and surface properties are often benefit the
oriented interaction with the biomolecule. Besides, establishing “mo-
300
Biosensors and Bioelectronics 92 (2017) 294–304
lecular wire” may also provide electron tunneling for the direct electron
transfer route (Bai et al., 2014).
The third generation thus exploits the direct electrochemistry of
enzymes. It is fascinating because it underlies the more convenient
mediator-free enzyme biosensors. For example, as shown in Fig. 7,
Curri et al. and Vastarella et al. immobilized CdS QDs on gold
electrodes combined with formaldehyde dehydrogenase (FDH) to
conduct the catalytic oxidation of formaldehyde. It has been found
that the CdS QDs in close contact with FDH is an effective photoactive
material to replace the NAD+/NADH role of electron shuttle in the
enzymatic reaction (Curri et al., 2002; Vastarella and Nicastri, 2005).
Later on, on the basis of same mechanism, self-assembly of CdS QDs
and GDH onto multiwall carbon nanotubes were performed for PEC
glutamate detection, and the direct electrochemistry of GDH with the
CdS could induce the cofactor-independent dehydrogenase enzymatic
reaction (Tang et al., 2008). Nevertheless, in such a system the specific
electron pathway needs further investigation since the substrates may
be oxidized directly by the illuminated semiconductors.
HRP, with heme as the active site, could also exhibit direct
electrochemistry at the semiconductors. For instance, upon HRP–
TiO2 nanotubes hybrid system, Li et al. realized the PEC HRP biosensor
towards H2O2 biosensing. When upon visible light illumination, the
photogenerated electron–hole pairs could quickly move to the surface
Fig. 7. Schematic illustration of photoinduced oxidation of formaldehyde using FDH
coupled to CdS as electron mediator instead of NAD+ co-factor (reprinted with
permission from Curri et al. (2002)).
W.-W. Zhao et al.
of HRP/TiO2 and initiate the direct electron transfer of the FeIII/FeII
redox couple at HRP's redox center. Because of specific redox reaction
between HRP and H2O2, the presence of H2O2 could enhance the
electron-transfer reactivity of HRP, which would be enhanced as the
concentration of H2O2 increased (Chen et al., 2010). In yet another
innovative application of this mechanism, Zhu et al. coupled HRP to
CdSe QDs embedded in mesoporous silica spheres, and the PEC
detection of H2O2 is also feasible without the presence of mediators
(Yang et al., 2011).
GOx, especially, has also demonstrated the property of direct
electrochemistry, which could contribute to the glucose detection at
low potentials which are slightly positive from the redox potential of
GOx (Zhu et al., 2015). In PEC enzymatic biosensor, Liu et al.
immobilized the GOx onto the ZnO inverse opal photonic crystals
and applied the system for glucose detection. The direct electrochem-
istry of GOx, which involves a two-proton redox reaction, permitted the
direct electron transfer between the hosted GOx and the electrode (Xia
et al., 2014).
Sulfite oxidase (SO) could catalyze the oxidation of sulfite to sulfate,
coupled with the subsequent reduction of two equivalents of cyt c.
Recently, SO was coupled with CdS QDs and its effective electrochem-
istry was studied. Direct electron transfer from SO to CdS QDs
electrode was verified, and the photocurrent signal increased with the
enhanced addition of sulfite. The dpendence of the catalytic current on
sulfite concentration was also derived (Zeng et al., 2015).
Cyt c, a redox protein existing in both the oxidized and reduced
states, can operate as an electron acceptor or donor, respectively, in
photochemical systems (Pardo-Yissar et al., 2003). It is by far the most
intensively investigated redox protein, and many works have utilized
Cyt c combined with various PEC species to control the cathodic or
anodic photocurrent generation (Stoll et al., 2006), and also to
construct signal chains for bioanalytical applications. For example, as
shown in Fig. 8, the presence of reduced Cyt c could induce the
generation of anodic photocurrent while the oxidized Cyt c induce the
generation of cathodic photocurrent. When coupled with enzymes, the
formed biocatalytic cascades can amplify the photocurrent generation,
which was attributed to the fast biocatalytic regeneration of reduced/
oxidized Cyt c on the CdS NPs interface. Specifically, when coupled
with LDH and its substrate lactate, the anodic photocurrent increased
with the enhanced lactate concentration, and the saturated photocur-
Fig. 8. (A) Amplified anodic photocurrent generation in the presence of reduced Cyt c, LDH and lactate; (B) Amplified cathodic photocurrent generation in the presence of oxidized Cyt
c, NR and nitrate (reprinted with permission from Stoll et al. (2006)).
301
Biosensors and Bioelectronics 92 (2017) 294–304
rent value is approximately nine-fold higher than that of bare reduced
Cyt c. On the contrary, when coupled with nitrate reductase (NR) and
its substrate NO3−, the cathodic photocurrent increased with the
enhanced NO3− concentration, and the saturated photocurrent value
is approximately eightfold higher than that of bare oxidized Cyt c.
Obviously, such biocatalytic cascades could be employed for the
detection of specific enzyme substrate (Katz et al., 2006). A case in
point is that the reaction of superoxide radicals with Cyt c was then
used to detect radical concentrations. In detail, xanthine oxidase
(XOD) was selected to catalyze the conversion of hypoxanthine to uric
acid for the generation of superoxide radicals in solution. The
generated O2− radicals could rapidly interact with the oxidized Cyt c
and generate the reduced Cyt c, which subsequent re-oxidized at the
illuminated QDs electrode. By varying the XOD activity used for the
radical generation, different superoxide concentrations can be followed
(Stoll et al., 2008). Note that the Cyt c was in solution in these above
examples. Later on, Tian et al. immobilized Cyt c onto the Au NPs/TiO2
nanoneedle film, and switched the direction of plasmon-induced
photocurrents. Besides, the immobilized Cyt c could maintain its
inherent enzymatic activity toward H2O2 and provide the enhanced
analytical performance in H2O2 detection (Zhu et al., 2009a, 2009b).
Recently, Liu et al. immobilized the cytochrome P450 2D6 (CYP2D6)
onto the CdTe QDs, and the nanohybrids were then applied for light-
driven drug metabolism. The catalytic cycle of P450-mediated oxyge-
nation reactions occurred by the formation of radical species on QDs
could immediately activate the bound P450 enzyme and then cause the
direct electron transfer from the CdTe QDs to CYP2D6 (Qian et al.,
2014).
Substantially efforts have been exerted to develop the third-
generation biosensors based on the direct electron transfer, and many
reported work have achieve the successful determination of various
targets of interest. Due to the self-contained nature that eliminates the
mediators, the third-generation biosensors are especially ideal for
repeated, real-time, continuous and in vivo monitoring the analytes.
2.6. Others
When several enzymes are co-immobilized upon the same electrode
surface, several strategies for improving biosensor performance can be
developed. For example, several enzymes facilitate the biological
W.-W. Zhao et al. Biosensors and Bioelectronics 92 (2017) 294–304

Fig. 9. Detection of GMP using an NADH-dependent electrochemical assay (reprinted

with permission from Sabir et al. (2015)).

recognition by sequentially converting the product of a series of

enzymatic reactions into a final electroactive form. A type case is the
recently reported PEC bioanalysis of guanosine monophosphate (GMP)
by Parak et al., with the enzymatic reaction cascade as demonstrated in
Fig. 10. Proposed PEC assembly consisting of ALP catalyzed generation of AA for in Situ
Fig. 9. Upon a CdS@ZnS QDs electrode, three enzymes were combined
activating the TiO2 NPs (reprinted with permission from Zhao et al. (2013a)).
in a coupled reaction assay. In the first reaction, guanylate kinase
(GMPK) catalyzed the adenosine triphosphate (ATP)-dependent con-
4-chloro-1-naphthol (4-CN) by H2O2 was performed upon a TiO2/CdS
version of GMP to guanosine diphosphate (GDP), which warranted
electrode to produce the insoluble and insulating product benzo-4-
substrate specificity. In the second reaction, the catalytic products ADP
chlorohexadienone on the photoelectrode surface for PEC detection of
and GDP from the first reaction were co-substrates for the pyruvate
H2O2 (Zhao et al., 2011). Upon a CdS/GOx system, a PEC biosensor
kinase (PK) catalyzed conversion of phosphoenolpyruvate (PEP) to
was realized for probing enzyme activities without external irradiation.
pyruvate, which was the co-substrate of the third reaction for LDH that
Hemin/G-quadruplex-stimulated chemiluminescence resonance en-
produced lactate via oxidation of NADH to NAD+. In this sequential
ergy transfer (CRET) was directed for the generation of photocurrent
enzyme reactions, a signal chain was established from the GMPK
signals (Golub et al., 2012). Based on the GOx catalyzed H2O2-
reaction in the first step through the PK and LDH dependent reactions
mediated enlargement of Au NPs on a nanoporous TiO2 substrate,
and finally to the NADH oxidation at the QD electrode, where the last
localized surface plasmon resonance-based photoelectrochemistry and
step is a redox reaction that detected by photocurrent measurements.
nanoparticle size effect were also coupled for the PEC enzymatic
Obviously, the consumption of NADH is proportional to the GMPK
detection of glucose (Zhao et al., 2012b). Using the ALP catalytic
activity, thereby indirectly probing the amount of GMP (Sabir et al.,
process to couple with the unique surface chemistry of nanocrystalline
2015). One obvious advantage of such a set-up is that it allows a much
TiO2, as shown in Fig. 10, a strategy for PEC enzymatic analysis has
wider range of possible analytes for enzymatic biosensors. Besides,
also been developed for enzyme activity or inhibitor detection (Zhao
aiming for signal amplification, the appropriate application of multiple
et al., 2013a). Based on enzymatic product-mediated in situ formation
enzymes in series may efficiently regenerate the co-substrate of the first
of CdS QDs, PEC detection of AChE activity and inhibition has been
enzyme and thus achieve real signal amplification. In addition, aiming
reported recently (Hou et al., 2016). Proteases have also been involved
for improving the selectivity, the proper application of multiple
in PEC bioanalysis. For example, PEC sensing of caspase-3 (one of the
enzymes in parallel may decrease the local concentration of electro-
most frequently activated cysteine proteases) activity was developed to
chemical interfering substance and hence accomplish the enhanced
evaluate the drug resistance of chronic myeloid leukemia (CML) cells
biosensor selectivity.
during nilotinib treatment. After cleavage by active caspase-3, the
Some other extension types of the PEC enzymatic bioanalysis have
residual peptides were combined with ALP label for enzymatic
also been reported. Note that we term them as “enzymatic bioanalysis”
catalyzed generation of ascorbic acid (AA) as electron donor for
rather than “enzymatic biosensor” since they could not be defined as
photocurrent generation. The activity of caspase-3, which depends on
“biosensor” by the IUPAC definition on account of their specific
the amount of residual peptide on the electrode, was thus inversely
configurations. For example, PEC bioanalysis of tyrosinase (TR)
proportional to the amount of AA (Zhou et al., 2014, 2015b).
activity was realized by using CdS QDs as signal reporter units.
Tyrosinase, a Cu(II)-containing monooxygenase, is one important
Oxidation of the capping layer by TR/O2 could yield the respective L-
metalloproteinases. Based on the use of tyrosinase, PEC biosensor
DOPA and dopaquinone products, the reduction of this mixture of
was also developed for thrombin detection (Chen et al., 2016b).
products with citric acid would yield the L-DOPA derivative. And then
Addressing protein kinase activity and inhibition, various PEC bioana-
the association of the L-DOPA derivative-functionalized CdS QDs to a
lysis have also been reported (Yan et al., 2016; Yin et al., 2015; Zhou
boronic acid monolayer-modified electrode could enable the PEC
et al., 2015c).
transduction of TR activity by the generation of photocurrents in the
With the development of enzymatic bioanalysis, various nucleic
presence of triethanolamine as a sacrificial electron donor (Yildiz et al.,
acid-based or nucleic acid-processing enzymes have been increasingly
2008). Harnessing enzyme reaction to produce an insoluble product on
involved with PEC biomolecular detection. For example, using DNA
the transducer, a physical effect such as steric hindrance or an
repair enzymes, label-Free and selective PEC detection of chemical
insulating effect could be generated and thus inhibit the interfacial
DNA methylation damage has been reported (Wu et al., 2013). Using
electron communication between the PEC species and the electron
DNase I, PEC biosensing was realized for the direct detection of
acceptor/donor. Using this strategy, the HRP-accelerated oxidation of
microRNAs (Cao et al., 2014). Using DNAzyme-catalyzed precipitation

302
W.-W. Zhao et al. Biosensors and Bioelectronics 92 (2017) 294–304

Table 2 reduced cost, easy operation, miniaturization, and especially real-time

Comparison of different PEC enzymatic biosensor types. operation without prior separation or pretreatment. When combined
with PEC biomolecular detection, the PEC enzymatic biosensors have
Sensor Main characteristics Typical enzymes Some applications
types further remarkable performances such as lower detection limits and
higher sensitivity. In spite of its short development time, this newly
First
genera-
• Utilizes the • Oxidase,
specificity of an GOx;
e.g. • To analyze the
substrate of the
emerged methodology has been continuously expanding with increas-
tion enzyme reaction; • Phosphatase, e.g. respective enzyme
ing papers published in the past decade. In this work, we have
• Detects the ALP; reaction (e.g. the specifically reviewed this dynamically developing field of PEC enzy-
enzymatic
conversion by
• Dehydrogenase,
e.g. GDH;
depletion of O by
oxidase);
2
matic biosensor. As indicated, the remarkable and rapid advances in
this field demonstrated the intensified scientific awareness of its
monitoring the
consumption of a
• AChE • Toenzymeanalyze the
inhibition
attractive potentials and also a bright research future. The inherent
sensitivity, simplicity, cost benefits and long-term application needs
co-substrate S′ or by use of the
the formation of generated product will continue to push forward its further advancements.
an electro-active as electron donor/ Future developments can expand their application scope for clinical
product P; acceptor; and industrial purposes. Advanced management of enzyme immobili-
Second
genera-
• Uses mediators to
shuttle electrons
• Most of common •
enzymes that
To detect the
substrate by the
zation can improve the surface coverage, orientation, accessibility,
tion between the with their redox use of mediators activity and stability of the enzyme molecules. Various nucleic acid-
active site of the centers deeply as electron based enzymes will afford exciting opportunities for specific and
enzyme and the embedded within transporter; sensitive detection of many other biochemical species. New progresses
electrode; their tertiary
•
in nanotechnology and material science as well as in custom engineer-
Mediators are protein
ing of biomolecules can contribute to the new detection principles with
easily reduced by structures;
the enzymatic signal amplification. The synergy effect could not only promote the
reactions and biosensor performances such as better sensitivity, specificity, reprodu-
then oxidized at cibility and stability, but also permit the indirect detection of more
relatively low
targets of low concentrations. Further miniaturization and integration
potentials at the
photoelectrodes;
in autonomously operated and portable instruments and (paper-based)
Third
genera-
• Relies on the
direct redox
• Heme-containing
enzymes e.g. such
• Tosubstrate
detect the
with a
microfluidic formats could simplify and extend its applicability to
bedside patient and home testing. The development of PEC enzymatic
tion reactions as HRP; mediator-free
• Low-molecule
biosensor arrays with screen-printed functionalized electrodes that
between the way;
enzymes and
electrodes;
weight redox
proteins, e.g. Cyt
• To exploit the
direct
capable of automated multi-analysis would also lead to remarkable
advantages compared to current single-channel system in terms of cost,
• Improved
electrical contact
c; electrochemistry
of specific
speediness as well as simplicity.

between the proteins;

redox center of
the enzyme and
the electrode
surface;
Others • Variable
protocols with
– • Tomultiple
different
purposes and
advantages;
amplification, ultrasensitive PEC monitoring of polynucleotide kinase
activity was reported (Zhuang et al., 2015). PEC bioanalysis of M.SssI
methyltransfease activity and inhibitor has also been proposed recently
(Shen et al., 2015; Yang et al., 2015). M.HhaI MTase can use
cysteamine as the substrate to displace the hydroxyl group of hydro-
xymethylcytosine (5-hmC). Using such enzyme, PEC detection of 5-
hmC was also realized based on in situ electron donor producing
strategy (Yang et al., 2016).
In addition, the photovoltaic effect of semiconductors that generate
hydroxyl radicals, superoxide anions, ozone and electron pairs at the
irradiated semiconductor–liquid interfaces has also been demonstrated
for contributing to the electron transfer reactivity and the enzyme
catalytic activity of various enzymes, such as HRP, hemoglobin,
microperoxidase and even GOx (Zhu et al., 2009d). Table 2 has
compared the different PEC enzymatic biosensor types.
3. Conclusion and perspectives
Enzymatic biosensors possess several distinct advantages such as
Acknowledgments
We thank the National Natural Science Foundation of China (Grant
Nos. 21327902, 21135003, 21305063, and 21675080) and the Natural
couple the Science Funds of Jiangsu Province (Grant BK20130553) for support.
This work was also supported by a Project Funded by the Priority
enzymatic
reactions that Academic Program Development of Jiangsu Higher Education
allow a much Institutions.
wider range of
possible analytes References
or signal
amplification;
Bai, Y.-F., Xu, T.-B., Luong, J.H.T., Cui, H.-F., 2014. Anal. Chem. 86, 4910–4918.
Bard, A.J., Faulkner, L.R., Leddy, J., Zoski, C.G., 1980. Electrochemical Methods:
Fundamentals and Applications. Wiley, New York.
Bixon, M., Jortner, J. (1999) Electron Transfer—from Isolated Molecules to
Biomolecules, in Advances in Chemical Physics: Electron Transfer - from Isolated
Molecules to Biomolecules. Part 1, Volume 106 (eds I. Prigogine and S. A. Rice),
John Wiley & Sons, Inc., Hoboken, NJ, USA. http://dx.doi.org/10.1002/
9780470141656.ch3
Cao, H., Liu, S., Tu, W., Bao, J., Dai, Z., 2014. Chem. Commun. 50, 13315–13318.
Chen, D., Zhang, H., Li, X., Li, J., 2010. Anal. Chem. 82, 2253–2261.
Chen, H., Gao, Q., Li, J., Lin, J.-M., 2016a. J. Photochem. Photobiol. C 27, 54–71.
Chen, J., Liu, Y., Zhao, G.-C., 2016b. Sensors 16, 135.
Clark, L.C., Lyons, C., 1962. Ann. N.Y. Acad. Sci. 102, 29–45.
Clark, L.C.J., 1956. Trans. Am. Soc. Artif. Intern. Organs 2, 41–48.
Curri, M.L., Agostiano, A., Leo, G., Mallardi, A., Cosma, P., Della Monica, M.,, 2002.
Mater. Sci. Eng. C 22, 449–452.
Devadoss, A., Sudhagar, P., Terashima, C., Nakata, K., Fujishima, A., 2015. J.
Photochem. Photobiol. C 24, 43–63.
Dilgin, D.G., Gokcel, H.I., 2015. Anal. Methods 7, 990–999.
Du, J., Yu, X., Di, J., 2012. Biosens. Bioelectron. 37, 88–93.
Du, J., Yu, X., Wu, Y., Di, J., 2013. Mater. Sci. Eng. C 33, 2031–2036.
Ertek, B., Dilgin, Y., 2016. Bioelectrochemistry 112, 138–144.
Ertek, B., Akgul, C., Dilgin, Y., 2016. RSC Adv. 6, 20058–20066.
Freeman, R., Girsh, J., Willner, I., 2013. ACS Appl. Mat. Interfaces 5, 2815–2834.
Gill, R., Zayats, M., Willner, I., 2008. Angew. Chem. Int. Ed. 47, 7602–7625.
Göbel, G., Schubert, K., Schubart, I.W., Khalid, W., Parak, W.J., Lisdat, F., 2011.
Electrochim. Acta 56, 6397–6400.
Golub, E., Niazov, A., Freeman, R., Zatsepin, M., Willner, I., 2012. J. Phys. Chem. C 116,

303
W.-W. Zhao et al.
13827–13834.
Gong, J., Wang, X., Li, X., Wang, K., 2012. Biosens. Bioelectron. 38, 43–49.
Guo, L.-H., Mukamel, S., McLendon, G., 1995. J. Am. Chem. Soc. 117, 546–547.
Han, D.-M., Ma, Z.-Y., Zhao, W.-W., Xu, J.-J., Chen, H.-Y., 2013. Electrochem. Commun.
35, 38–41.
Hou, T., Zhang, L., Sun, X., Li, F., 2016. Biosens. Bioelectron. 75, 359–364.
Huang, Q., Chen, H., Xu, L., Lu, D., Tang, L., Jin, L., Xu, Z., Zhang, W., 2013. Biosens.
Bioelectron. 45, 292–299.
Huang, Y., Zhang, W., Xiao, H., Li, G., 2005. Biosens. Bioelectron. 21, 817–821.
Jafari, F., Salimi, A., Navaee, A., 2014a. Electroanalysis 26, 1782–1793.
Jafari, F., Salimi, A., Navaee, A., 2014b. Electroanalysis 26, 530–540.
Jiang, H., Ju, H., 2007. Chem. Commun., 404–406.
Katz, E., Zayats, M., Willner, I., Lisdat, F., 2006. Chem. Commun., 1395–1397.
Khalid, W., Göbel, G., Hühn, D., Montenegro, J.-M., Rivera-Gil, P., Lisdat, F., Parak,
W.J., 2011. J. Nanobiotechnol. 9, 1–10.
Kimmel, D.W., LeBlanc, G., Meschievitz, M.E., Cliffel, D.E., 2012. Anal. Chem. 84,
685–707.
Lee, C.-Y., Park, H.S., Fontecilla-Camps, J.C., Reisner, E., 2016. Angew. Chem. Int. Ed.
55, 5971–5974.
Li, C., Shi, G., 2014. J. Photochem. Photobiol. C 19, 20–34.
Li, X., Zheng, Z., Liu, X., Zhao, S., Liu, S., 2015. Biosens. Bioelectron. 64, 1–5.
Lisdat, F., Schäfer, D., Kapp, A., 2013. Anal. Bioanal. Chem. 405, 3739–3752.
Liu, Q., Lu, X., Li, J., Yao, X., Li, J., 2007. Biosens. Bioelectron. 22, 3203–3209.
Liu, Q., Cai, J., Huan, J., Dong, X., Wang, C., Qiu, B., Wang, K., 2014a. Analyst 139,
1121–1126.
Liu, X., Yan, R., Zhu, J., Huo, X., Wang, X., 2015. Electrochim. Acta 173, 260–267.
Liu, Y.-F., Chen, J.-X., Xu, M.-Q., Zhao, G.-C., 2014b. Int. J. Electrochem. Sci. 9,
4014–4023.
Pardo-Yissar, V., Katz, E., Wasserman, J., Willner, I., 2003. J. Am. Chem. Soc. 125,
622–623.
Privett, B.J., Shin, J.H., Schoenfisch, M.H., 2008. Anal. Chem. 80, 4499–4517.
Privett, B.J., Shin, J.H., Schoenfisch, M.H., 2010. Anal. Chem. 82, 4723–4741.
Qian, J., Zhu, W., Mi, L., Xu, X., Yu, J., Cui, D., Xue, Y., Liu, S., 2014. J. Electroanal.
Chem. 733, 27–32.
Ren, X., Chen, D., Meng, X., Tang, F., Hou, X., Han, D., Zhang, L., 2009. J. Colloid
Interface Sci. 334, 183–187.
Riedel, M., Göbel, G., Abdelmonem, A.M., Parak, W.J., Lisdat, F., 2013. ChemPhysChem
14, 2338–2342.
Ronkainen, N.J., Halsall, H.B., Heineman, W.R., 2010. Chem. Soc. Rev. 39, 1747–1763.
Sabir, N., Khan, N., Völkner, J., Widdascheck, F., del Pino, P., Witte, G., Riedel, M.,
Lisdat, F., Konrad, M., Parak, W.J., 2015. Small 11, 5844–5850.
Shen, Q., Han, L., Fan, G., Abdel-Halim, E.S., Jiang, L., Zhu, J.-J., 2015. Biosens.
Bioelectron. 64, 449–455.
Stoll, C., Kudera, S., Parak, W.J., Lisdat, F., 2006. Small 2, 741–743.
Stoll, C., Gehring, C., Schubert, K., Zanella, M., Parak, W.J., Lisdat, F., 2008. Biosens.
Bioelectron. 24, 260–265.
Sun, J., Zhu, Y., Yang, X., Li, C., 2009. Particuology 7, 347–352.
Swainsbury, D.J.K., Friebe, V.M., Frese, R.N., Jones, M.R., 2014. Biosens. Bioelectron.
58, 172–178.
Tang, J., Li, J., Da, P., Wang, Y., Zheng, G., 2015. Chem.–Eur. J. 21, 11288–11299.
Tang, J., Wang, Y., Wang, Y., Li, J., Kong, B., Jiang, M., Zheng, G., 2014. J. Mater. Chem.
A 2, 15752–15757.
Tang, L., Zhu, Y., Yang, X., Sun, J., Li, C., 2008. Biosens. Bioelectron. 24, 319–323.
Tanne, J., Schäfer, D., Khalid, W., Parak, W.J., Lisdat, F., 2011. Anal. Chem. 83,
7778–7785.
Tel-Vered, R., Yildiz, H.B., Yan, Y.-M., Willner, I., 2010. Small 6, 1593–1597.
Thevenot, D.R., Tóth, K., Durst, R.A., Wilson, G.S., 1999. Electrochemical biosensors:
Recommended definitions and classification. Pure Appl. Chem., 2333–2348.
Tu, L., Liu, G., Zhang, W., Qin, J., Yue, Z., 2014. IEEE J. Sel. Top. Quantum Electron. 20,
175–183.
Turner, A.P.F., Karube, I., Wilson, G., 1987. Nature 329, (402-402).
Umar, A., Rahman, M.M., Al-Hajry, A., Hahn, Y.B., 2009. Talanta 78, 284–289.
304
Biosensors and Bioelectronics 92 (2017) 294–304
Vastarella, W., Nicastri, R., 2005. Talanta 66, 627–633.
Wang, G.-L., Liu, K.-L., Dong, Y.-M., Wu, X.-M., Li, Z.-J., Zhang, C., 2014. Biosens.
Bioelectron. 62, 66–72.
Wang, W., Bao, L., Lei, J., Tu, W., Ju, H., 2012. Anal. Chim. Acta 744, 33–38.
Wilson, G.S., Hu, Y., 2000. Chem. Rev. 100, 2693–2704.
Wu, Y., Zhang, B., Guo, L.-H., 2013. Anal. Chem. 85, 6908–6914.
Xia, L., Song, J., Xu, R., Liu, D., Dong, B., Xu, L., Song, H., 2014. Biosens. Bioelectron.
59, 350–357.
Yan, Z., Wang, Z., Miao, Z., Liu, Y., 2016. Anal. Chem. 88, 922–929.
Yang, X., Wang, P., Zhu, Y., Li, C., 2011. J. Solid State Electrochem. 15, 731–736.
Yang, Z., Wang, F., Wang, M., Yin, H., Ai, S., 2015. Biosens. Bioelectron. 66, 109–114.
Yang, Z., Shi, Y., Liao, W., Yin, H., Ai, S., 2016. Sens. Actuators B Chem. 223, 621–625.
Yildiz, H.B., Freeman, R., Gill, R., Willner, I., 2008. Anal. Chem. 80, 2811–2816.
Yildiz, H.B., Talaz, O., Kamaci, M., Caliskan, A., Caliskan, S., 2013. J. Macromol. Sci. Part
A Pure Appl. Chem. 50, 1182–1193.
Yildiz, H.B., Kamaci, M., Karaman, M., Toppare, L., Sayin, S., 2014. J. Macromol. Sci.
Part A Pure Appl. Chem. 51, 385–393.
Yin, H., Sun, B., Dong, L., Li, B., Zhou, Y., Ai, S., 2015. Biosens. Bioelectron. 64,
462–468.
Yue, Z., Lisdat, F., Parak, W.J., Hickey, S.G., Tu, L., Sabir, N., Dorfs, D., Bigall, N.C.,
2013. ACS Appl. Mat. Interfaces 5, 2800–2814.
Zayats, M., Kharitonov, A.B., Pogorelova, S.P., Lioubashevski, O., Katz, E., Willner, I.,
2003. J. Am. Chem. Soc. 125, 16006–16014.
Zeng, T., Leimkühler, S., Koetz, J., Wollenberger, U., 2015. ACS Appl. Mat. Interfaces 7,
21487–21494.
Zhang, J., Tu, L., Zhao, S., Liu, G., Wang, Y., Wang, Y., Yue, Z., 2015. Biosens.
Bioelectron. 67, 296–302.
Zhang, X., Guo, Y., Liu, M., Zhang, S., 2013. RSC Adv. 3, 2846–2857.
Zhang, X., Liu, M., Liu, H., Zhang, S., 2014. Biosens. Bioelectron. 56, 307–312.
Zhao, C., Zhang, Z., Zhao, Y., Yu, J., 2012a. Chin. J. Chem. 30, 1851–1856.
Zhao, W.-W., Xu, J.-J., Chen, H.-Y., 2014b. Chem. Rev. 114, 7421–7441.
Zhao, W.-W., Xu, J.-J., Chen, H.-Y., 2015b. Chem. Soc. Rev. 44, 729–741.
Zhao, W.-W., Xu, J.-J., Chen, H.-Y., 2016a. TrAC Trends Anal. Chem. 82, 307–315.
Zhao, W.-W., Xu, J.-J., Chen, H.-Y., 2016b. Analyst 141, 4262–4271.
Zhao, W.-W., Yu, P.-P., Xu, J.-J., Chen, H.-Y., 2011. Electrochem. Commun. 13,
495–497.
Zhao, W.-W., Tian, C.-Y., Xu, J.-J., Chen, H.-Y., 2012b. Chem. Commun. 48, 895–897.
Zhao, W.-W., Ma, Z.-Y., Xu, J.-J., Chen, H.-Y., 2013a. Anal. Chem. 85, 8503–8506.
Zhao, W.-W., Wang, J., Zhu, Y.-C., Xu, J.-J., Chen, H.-Y., 2015a. Anal. Chem. 87,
9520–9531.
Zhao, W.-W., Shan, S., Ma, Z.-Y., Wan, L.-N., Xu, J.-J., Chen, H.-Y., 2013b. Anal. Chem.
85, 11686–11690.
Zhao, W.-W., Liu, Z., Shan, S., Zhang, W.-W., Wang, J., Ma, Z.-Y., Xu, J.-J., Chen, H.-Y.,
2014a. Sci. Rep. 4, 4426.
Zheng, M., Cui, Y., Li, X., Liu, S., Tang, Z., 2011. J. Electroanal. Chem. 656, 167–173.
Zhou, H., Liu, J., Zhang, S., 2015a. TrAC Trends Anal. Chem. 67, 56–73.
Zhou, H., Gan, X., Wang, J., Zhu, X., Li, G., 2005. Anal. Chem. 77, 6102–6104.
Zhou, S., Wang, Y., Zhao, M., Jiang, L.-P., Zhu, J.-J., 2015b. ChemPhysChem 16,
2826–2835.
Zhou, S., Kong, Y., Shen, Q., Ren, X., Zhang, J.-R., Zhu, J.-J., 2014. Anal. Chem. 86,
11680–11689.
Zhou, Y., Wang, M., Yang, Z., Yin, H., Ai, S., 2015c. Sens. Actuators B Chem. 206,
728–734.
Zhu, A., Luo, Y., Tian, Y., 2009a. Anal. Chem. 81, 7243–7247.
Zhu, A., Luo, Y., Tian, Y., 2009b. Chem. Commun., 6448–6450.
Zhu, C., Yang, G., Li, H., Du, D., Lin, Y., 2015. Anal. Chem. 87, 230–249.
Zhu, W., An, Y.-R., Luo, X.-M., Wang, F., Zheng, J.-H., Tang, L.-L., Wang, Q.-J., Zhang,
Z.-H., Zhang, W., Jin, L.-T., 2009c. Chem. Commun., 2682–2684.
Zhu, X., Yuri, I., Gan, X., Suzuki, I., Li, G., 2007. Biosens. Bioelectron. 22, 1600–1604.
Zhu, Y., Cao, H., Tang, L., Yang, X., Li, C., 2009d. Electrochim. Acta 54, 2823–2827.
Zhuang, J., Lai, W., Xu, M., Zhou, Q., Tang, D., 2015. ACS Appl. Mat. Interfaces 7,
8330–8338.