Content uploaded by Selorm Yao-Say Solomon Adade
Author content
All content in this area was uploaded by Selorm Yao-Say Solomon Adade on Jan 30, 2022
Content may be subject to copyright.
Received: April Revised: June Accepted: July
DOI: ./-.
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
Advanced applications of chemo-responsive dyes based odor
imaging technology for fast sensing food quality and safety:
Areview
Wencui Kang Hao Lin Hao Jiang Selorm Yao-Say Solomon Adade
Zhaoli Xue Quansheng Chen
School of Food and Biological
Engineering, Jiangsu University,
Zhenjiang, P. R. China
Correspondence
Hao Lin and Quansheng Chen, School of
Food and Biological Engineering, Jiangsu
University,Zhenjiang, Jiangsu , P. R.
China.
Email: linhaolt@.com and
q.s.chen@ujs.edu.cn
Funding information
National Natural Science Foundationof
China, Grant/AwardNumber: ;
Jiangsu Agricultural independent inno-
vation fund, Grant/Award Number:
SCX; Project of Facultyof Agri-
cultural Equipment of Jiangsu University,
Grant/AwardNumber: NZXB;
Key R&D Program of Jiangsu Province,
Grant/AwardNumber: BE
Abstract
Public attention to foodquality and safety has been increased significantly. There-
fore, appropriate analytical tools are needed to analyze and sense the food quality
and safety. Volatile organic compounds (VOCs) are important indicators for the
quality and safety of food products. Odor imaging technology based on chemo-
responsive dyes is one of the most promising methods for analysis of food prod-
ucts. This article reviews the sensing and imaging fundamentals of odor imag-
ing technology based on chemo-responsive dyes. The aim is to give detailed out-
lines about the theory and principles of using odor imaging technology for VOCs
detection, and to focus primarily on its applications in the field of quality and
safety evaluation of food products, as well as its future applicability in mod-
ern food industries and research. The literatures presented in this review clearly
demonstrated that imaging technology based on chemo-responsive dyes has the
exciting effect to inspect such as quality assessment of cereal , wine and vinegar
flavored foods , poultry meat, aquatic products, fruits and vegetables, and tea. It
has the potential for the rapid, reliable, and inline assessment of food safety and
quality by providing odor-image-basedmonitoring tool.
Practical Application: The literatures presented in this review clearly demon-
strated that imaging technology based on chemo-responsive dyes has the exciting
effect to inspect such as quality assessment of cereal , wine and vinegar flavored
foods, poultry meat, aquatic products, fruits and vegetables, and tea.
KEYWORDS
chemo-responsive dyes, detection, food quality and safety, odor imaging technology, sensing,
VOCs
1 INTRODUCTION
Odor is an important indicator of food quality and safety.
During the storage and processing of food, due to some
changes such as its own freshness, infection from microor-
ganisms such as the fungi from external environment, and
changes in composition, gases of different components are
usually volatilized (Moreira et al., ;Picoetal.,).
Because of the slow detection efficiency and higher cost
(Gu, et al., ), traditional detection methods including
artificial sensory evaluation method, chromatographic
chemical analysis method, and so forth, are difficult to
achieve rapid detection of food quality and safety (Feng
et al., ; Sanaeifar et al., ). Odor imaging technology
Compr Rev Food Sci Food Saf. ;–. © Institute of Food Technologists R
1wileyonlinelibrary.com/journal/crf
2A - .. .
globally is currently a new branch in the research field
of artificial olfaction technology (Sung et al., ). This
technology is based on colorimetric sensors composed
of chemo-responsive dyes sensitive to volatile gases to
capture and then express response information in the
form of imaging. Hence, it is a rapid, economical, and a
nondestructive technology for the detection of food qual-
ity and safety (Deshmukh et al., ;Radietal.,).
More prominently, imaging technology based on chemo-
responsive dyes enables olfactory information to convert
intovisual information, making odors intuitively visible
and easily analyzed. Compared with traditional electronic
olfactory technology, the odor sensing that can be used in
imaging technology based on chemo-responsive dyes is
broader and more sustainable, and the detection results
are presented more intuitively and vividly (Ezhilan et al.,
; Q. Liu et al., ; Majchrzak et al., ). Through
theoretical analysis, as well as latest research introduction
and application results, it systematically reveals the basic
principles, sensor production, sensitization processing,
detection system, and the processing and analysis of
sensing signals. Overall, this article expounds the appli-
cation prospects of imaging technology based on chemo-
responsive dyes in order to provide certain reference values
for its promotion in the field of food quality and safety.
Odor imaging technology (also known as colorimetric
sensor technology) was originated in and was first
proposed by Professor Kenneth S. Suslick of the Univer-
sity of Illinois at Urbana-Champaign (Rakow et al., ;
Rakow et al., ). Metalloporphyrin (M-porphyrin) was
employed as a sensor by Suslick’s group for the qualitative
and quantitative detection of volatile organic substances.
As a result, it lays the foundation for the application of odor
imaging technology. It exploits the color change of chem-
ical dyes before and after exposure to the detected gas to
aid the visualization of qualitative and quantitative analy-
sis of gas. While traditional electronic olfaction technolo-
gies rely on weak forces such as physical adsorption or van
der Waals forces, odor imaging technology mainly relies on
strong forces of covalent bonds. Also, the technology has a
good anti-interference ability on water vapor in the envi-
ronment, wherefore to well make up for the shortcomings
of the existing biological and chemical sensor technology.
The colorimetric sensor is mainly composed of some
chemo-responsive dyes with specific recognition capabil-
ities. After these dye molecules interact with the detected
object, the color of the dye molecules changes significantly.
Through computer processing, the red-green-blue (RGB)
data are formed. And the data and chemical index related
to odor are both used to establish recognition pattern, then
regression analysis is performed. With the advancement
of material processing technology and computer data pro-
cessing, imaging detection technology has been proved to
have great application prospects in environmental mon-
itoring, food and beverage quality monitoring, disease
diagnosis, and other fields. The current researches pro-
posed the theoretical analysis of coloring mechanism in
the area of imaging technology based on chemo-responsive
dyes, such as density functional theory (DFT). The anal-
ysis of coloring mechanism has improved the selectivity
of chemo-responsive dyes that are applied in the odor
imaging technology. Some research findings introduced
new types of chemo-response materials used in odor
imaging technology in recent years. Meanwhile, metal-
loporphyrins (M-porphyrins) and boron-dipyrromethene
(BODIPY) could be further modified, so as to improve the
sensitivity and stability of the sensor. Besides, the combi-
nation of colorimetric sensor array and imaging detection
system, as well as the application of various signal pro-
cessing and data analysis methods, brings a wider range of
applications of odor imaging technology based on chemo-
responsive dyes for fast sensing food quality and safety. As
mentioned above, the article has reflected some advances
of odor imaging technology. In recent years, imaging detec-
tion technology has gradually been applied to the odor
detection of some highly volatile foods such as vinegar,
white wine, and mildew of grains (Chen, Liu, & Zhao, ;
Chen, Hui, et al., ; Chen, Li, et al., ;Fengetal.,
; Guan et al., ;Guetal.,; Huo et al., ;
Huangetal.,,; Lin et al., ,;Lin,Duan,
et al., ; Lin, Kang, et al., ;Lin,Yan,etal.,;
Lin, Kang, et al., ; Lin, Wang, et al., ; Khulal et al.,
; Qin et al., ; Qian & Lin, ; Salinas et al., ).
Therefore, in order to objectively evaluate odor qual-
ity, it is incumbent on researchers to find a new way
to design an odor sensor with the same performance as
the human nose. This odor imaging technology based on
chemo-responsive dyes mimics the mammalian olfactory
system by digital imaging to quantify odorants. The study is
aimed at imaging technology based on chemo-responsive
dyes how to recognize the odorant molecular from the
sensing and imaging process and the development and
design is in accordance with Figure .
2CHEMO-RESPONSIVE DYES
In terms of the basic structure of imaging technology,
the chemo-responsive dyes with strong color reaction effi-
ciency are employed, and imaging analysis is performed
based on the combination of chemo-responsive dyes and
volatile gases. The crux of imaging sensor to detect volatile
substances is the interaction between molecules, and for-
mation and fracture of chemical bonds involved in these
interactions. Therefore, the change of the data signal
before and after the reaction between the sensor and the
A - .. . 3
FIGURE 1 The design of imaging technology based on
chemo-responsive dyes
detected object is essentially determined by the physi-
cal and chemical properties of the chemically responsive
dye molecules that make up the sensor. Correspondingly,
chemo-responsive dyes become the key to imaging tech-
nology. The material used in imaging sensors must meet
the following two basic conditions: () The dye should
have at least one interaction center that involves interac-
tions as, π-πmolecular complexion, acid-base interactions,
bond formation, and van der Waals interaction; () a cer-
tain color change can occur after chemo-responsive mate-
rial exposure to substance. The same amount of external
groups or different amounts of the same group are proved
to produce different color changes so that volatile organic
compounds (VOCs) can be qualitatively or quantitatively
analyzed according to the degree of color change (Askim
& Suslick, ).
In order to overcome obstruction in the recognition
ability of sensing materials in connection with odorant
molecular features, it is necessary to study sensing tech-
nologies to recognize the odorant molecular features. Nor-
mally, M-porphyrin, BODIPY, some natural pigments, and
pH indicators could fulfill these requirements, and they
are frequently-used as chemo-responsive dyes in exper-
iments. Below are overviews on the characteristics of
the functional groups and rendering principles of chemo-
responsive dyes (Figure ).
2.1 Metalloporphyrin
M-porphyrins are composed of large π-conjugated
molecules (porphyrin ring) and metal ions (M). They are
a kind of macrocyclic polymer compounds containing
metal ions, of which molecular structure is shown in
Figure a. The metal ion is located in the central part
of the porphyrin ring, and the ring system of porphyrin
is basically on a plane and is simultaneously a highly
conjugated system, thus porphyrins have stable internal
structure and bright color performance. The metal ion
located in the center of the M-porphyrin is connected to
the nitrogen atoms on the four pyrrole rings. On account
of the broad axial ligand of the M-porphyrin, the metal
ion yet have an approach to other groups above or below
the porphyrin plane to take the shape of a metallic bond
(Mamardashvili et al., ). More importantly, the metal
porphyrin (M-porphyrin) as a chemical reaction acceptor
generates color changes when the external group enters
the porphyrin ring to connect with the central metal
ion or other groups. The nitrogen-containing ligand is
axially coordinated with the iron porphyrin, thereby the
electron cloud density both on the iron ion and pyrrole
ring will increase. It reflects as red shift of the band peak
in the spectrum. Hence, M-porphyrin has significant
advantages as a receptor, which can help to identify
molecular size, shape, functional group, and chiral isomer.
Rakow and Suslick () used different M-porphyrin
dyes to detect volatile inorganic and organic compounds.
Table shows different species of M-porphyrins and their
substituted groups and corresponding sensitive reaction
substances.
2.2 Boron-dipyrromethene
BODIPY was a general term for the derivatives of fluo-
robarpyrrole after being replaced by various substituents,
first discovered in . BODIPY as an intermediate prod-
uct in the synthesis of porphyrins retains part of the
chromophore (two pyrrole rings), undoubtedly, so it has
strong color rendering properties. As shown in Figure b,
the substituents are grafted on the core of BODIPY, and
they can be forced to separate from each other with-
out affecting the properties of the chromophore. This
prevents the π-πstacking to form a conformationally
restricted cyclic structure, enabling the fluorobarpyrrole
compounds to have simple single-conjugated chemical
structures (Swamy et al., ). This kind of compound
has absorption mostly in the red to near-infrared regions.
Consequently, its spectral characteristics can be adjusted
to occur blue shift or red shift as a result of minor
changes in structure. The substituted compounds with
high sensitivity and specificity can be easily obtained in the
application of ambient gas detection. The simple chemi-
cal structure makes it easy to be functionalized and nano-
modified. Recent studies have excellent sensing effect
4A - .. .
FIGURE 2 The characteristics of chemo-responsive dyes
achievements. On this basis, it has been widely used in the
fields of fluorescent labeling, sensor, and laser dye produc-
tion (Lin et al., ).
2.3 Natural pigments
Natural pigments with low toxicity are extracted from
natural products such as spinach (Spinacia oleracea), red
radish (Raphanus sativus L.), winter jasmine (Jasminum
nudiflorum), and black rice (Oryza sativa L. indica). Antho-
cyanins are often found in these natural products (Huang
et al., ), and the molecular structure of which is
shown in Figure c. Anthocyanin can bind with bases
such as amines, due to the presence of carbonyl and
hydroxyl groups (Huang et al., ). Because natural pig-
ments with color are in different ranges of visible light, it
offers a promising prospect for the application of chemo-
responsive dyes.
2.4 pH indicator
pH indicators are generally weak acids or weak alkali. With
the change of pH in the environment, the indicator gains
protons so as to convert to acid type or loses protons to base
type (Gabor et al., ; Stone & Vaughn, ). Due to the
different structures of acidic and alkaline indicators, they
appear in different colors. Take methyl orange as an exam-
ple, it is an alkaline indicator with the function of two-
color indicator. There are dissociation balance and color
change as shown in Figure d. Obviously, when the H+
concentration increases, the reaction proceeds to the right
and methyl orange mainly exists as a quinoid (acid color
type), showing red. The reaction proceeds to the left as the
H+concentration decreases. In this case, methyl orange
mainly exists in the azo (alkaline color type) and appears
yellow (Chen, Li, et al., ). The pH indicator not only
senses the acid-alkali changes in the environment but also
causes changes in color with change in the polarity of sub-
stances (Motellier & Toulhoat, ).
In summary, the chemo-responsive dye used may be M-
porphyrin, BODIPY, natural pigments, and pH indicator,
which shows difference in color for different analytes (Shi,
et al., ). M-porphyrin and BODIPY compounds are
widely considered as the most commonly used materials
with better performance. Besides, experts and scholars in
the field of chemistry or research on environmental safety
are developing some novel chemo-responsive materials. A
new rhodamine-formaldehyde probe based on the fluores-
cence on–off strategy for hydrogen sulfide detection was
developed which captures HS in live cells. (Hou et al.,
A - .. . 5
TABLE 1 Metalloporphyrins with different substituted groups and corresponding sensitive substances
Species Substituted group Sensitive substance Reference
Co(TPP) Phenyl Methanol (Brunink et al., )
Ethanol (Rakow & Suslick, et al., ;Chen,Liu,
et al., )
Diethylamine (Di Natale et al., )
Nitrogen (Alimelli et al., )
Quinine (Paolesse et al., )
White wine (Verrelli et al., )
CopNOTPP P-nitrophenyl Thiophene (Brunink et al., )
Diacetyle (Di Natale et al., )
CopOCH-TPP P-methoxypheny Methanol (Di Natale et al., )
Zn(TPP) Phenyl Ammonia (Vaughan et al., )
Triethylamine (Vaughan et al., ; Huang, Li, et al., )
Ethyl acetate (Zhang & Suslick, et al., )
Dimethylformamide (Rakow & Suslick, )
Tri-n-butylphosphine (Rakow & Suslick, )
Hexanal (Lin et al., )
MnTPP Phenyl Hexylamine (Rakow & Suslick, )
HS (Zhang & Suslick, et al., )
NaCl, (Paolesse et al. )
Acetic acid (Chen, Liu, et al., )
-Methylbutanal (Lin et al., )
MnTPPCl P-chlorobenzene Acid acetic (Di Natale et al., )
-Methylbutyraldehyde (Lin et al., )
SnTPP Phenyl Tetrahydrofuran (Rakow & Suslick, )
Oxygen (Paolesse et al., )
SnTPP(Cl) P-chlorobenzene N-octylamine (Rakow & Suslick, )
Butanol (Di Natale et al., )
Fe(TPP) Phenyl Acetonitrile (Rakow & Ssulick, )
HCl (Paolesse et al., )
SO (Verrelli et al., )
FeCl(TPP)F P-chlorofluorobenzene Acetone, pyridine,
-crown-
(Lin et al., )
Cu(TPP) Phenyl CHCl (Rakow & Suslick, )
Propanol (Suslick et al., )
HS (Verrelli et al., )
Ag(TPP) Phenyl P(OCH) (Rakow et al., )
Propionaldehyde (B. A. Suslick et al., )
Ru(TPP) Phenyl CHCl (Rakow & Suslick, )
Propane (B. A. Suslick et al., )
Abbreviation: TTP, tetraphenylporphyrin.
). Hao et al. () used a naphthalimide-based azo
dye as a colorimetric and ratiometric probe for the detec-
tion of CN_synthesized by incorporating a salicylalde-
hyde moiety. The synthesis and discovery of these develop-
ing chemo-responsive dyes undoubtedly provide a broad
application basis for odor imaging technology, which is
conducive to the wide application of odor imaging tech-
nology based on chemo-responsive dyes in the fields of
food quality and safety, environmental monitoring, and
biopharmaceuticals, and so forth.
6A - .. .
3 ANALYSIS OF COLORING
MECHANISM
Chemo-responsive dyes are the core for building colori-
metric sensors used in odor imaging technology. As one
of the common chemo-responsive dyes, M-porphyrins are
in a π-shaped conjugated system composed of four pyr-
role rings and have excellent color rendering properties.
On the one hand, the diversified porphyrins provide a
broad choice for chemo-responsive materials and thereby
are beneficial to the construction of specific colorimet-
ric sensors, but inevitably, they also bring certain difficul-
ties in the selection of chemo-responsive materials. Conse-
quently, the analysis of coloring mechanism is of extreme
significance to research directions in the area of imaging
technology based on chemo-responsive dyes.
3.1 Dynamic analysis of ultraviolet
spectroscopy
Chemo-responsive materials mostly have large π-conju-
gated structures and rigid planar structures. They pos-
sess chromophore groups that can interact with a vari-
ety of substances through intermolecular interactions such
as axial coordination, hydrogen bonding, and electrostatic
interaction. In consequence, changes in the configure
ration of porphyrins can be detected by absorption spec-
troscopy, fluorescence spectroscopy, and so forth. Also,
chemo-responsive materials can bind to the volatile gas
through molecular coordination, which leads to a change
in the energy level of electronic transition. This change
reflects as a color change called the color reaction. The spa-
tial structure of the material molecule and the activity of
the central metal ion have a great influence on the result
of the coordination reaction. Figure a shows the ultravio-
let (UV) spectrum of the color reaction of tetraphenylpor-
phyrin (TPP) with volatile ethanol. It is noted that ultra-
violet and visible (UV-Vis) spectrophotometry spectrum of
TPP shows a band at nm and named Soret band, which
is caused by the electron transition from the highly occu-
pied orbital 𝛼1𝜇(π) to the lowest empty orbital 𝑒𝑔(π*) (X. Li
et al., ). Figure b,cshows the absorption spectra and
difference spectra of the TPP exposure to different concen-
trations of ethanol. It can be seen from the figure that as
the concentration of ethanol increases, the absorbance at
the Soret band peak of TPP also increases gradually.
In the coordination reaction, the coordination number
and equilibrium constant are vital parameters for study-
ing the interaction between molecules, which can be cal-
culated using the Benesi-Hilde brand equation, as shown
in Equation ().
lg 𝐾 = lg 𝐴−𝐴1
𝐴2 − 𝐴 −𝑛lg[𝐶]()
in the equation, A1 is the blank absorbance of the por-
phyrin without ethanol; A2 is the absorbance of the solu-
tion when the porphyrin is completely protonated; Ais the
absorbance of the porphyrin solution at any point in the
titration process; and C is the concentration of the added
ethanol. According to the change of absorbance at a cer-
tain wavelength, the slope of the curve obtained by plot-
ting lg((A-A1)/(A2-A)) versus lg[C] can be used to obtain
the coordination number n, and the equilibrium constant
lgKcan be obtained from the intercept (Kuntz et al., ;
Wong & Suslick, ).
3.2 DFT analysis
The color changes after the chemo-responsive dyes expo-
sure to volatile gas involve molecular interactions, and
the interactions have some relationships with the electron
cloud distribution among nuclei. Therefore, the reaction
mechanism of the chemo-responsive dyes and the volatile
characteristic gas are theoretically calculated through
DFT. Hohenberg and Kohn proposed Hohenberg-Kohn
theorems and published them in the authoritative journal
Physical Review (Brandenburg & Grimme, ). The the-
orem states that the electron density at the relatively lowest
energy ρis the ground state density of the system, and the
corresponding energy is also the ground state energy of the
system. Hence, the common functional E[ρ] will be mini-
mized to get the energy and electron density of the ground
state. For a multi-electron system, its energy E(ρ)ismainly
composed of three parts, as shown in Equation ().
𝐸[𝜌] = 𝑇𝑒[𝜌] + 𝑉𝑒−𝑒[𝜌] + 𝑉ext [𝜌] ()
in the equation, Te[ρ] is the kinetic energy of electrons,
𝑉ext[𝜌] is the interaction energy between nuclei and
electrons, and Ve-e[ρ] is the interaction energy between
electrons and electrons. Among these three energies,
only𝑉ext[𝜌] is known, and Ve-e[ρ]andTe[ρ]canonlybe
approximated by functional.
Kohn and Sham describe the electron density mainly
through a series of independent molecular orbitals, as
showninEquation()
𝜌(𝑟)=
occ
∑
𝑖
∣Φ
𝑖(𝑟)∣2()
A - . .. 7
FIGURE 3 (a) The ultraviolet (UV) spectrum of the color reaction of tetraphenylporphyrin (TPP) with volatile ethanol. (b and c) The
absorption spectra and difference spectra of the TPP exposure to different concentrations of ethanol
Equation () is the summation of the occupied orbital.
Therefore, the effect of classical Coulomb and kinetic
energy can be described as Equations ()and()
𝑇𝑠[𝜌]=−
1
2
occ
∑
1
<Φ
1∣∇
2∣Φ
1>()
𝑉𝐻[𝜌]=1
2∫𝜌(𝑟1)(
𝑟2)
∣𝑟
1−𝑟
2∣𝑑𝑟1𝑟2()
Therefore, the functional of the energy system can be
transformed into Equation ():
𝐸[𝜌]=𝑇
𝑠[𝜌]+𝑉
𝐻[𝜌]+𝑉
ext [𝜌]+𝐸
XC [𝜌]()
𝐸XC[𝜌] in the equation is a commutative-correlation (XC)
functional, and its expression is shown in Equation ().
𝐸XC [𝜌]=(𝑇[𝜌]−𝑇
𝑠[𝜌]) +(𝑉
𝑒−𝑒 [𝜌]−𝑉
𝐻[𝜌])()
8A - .. .
FIGURE 4 Density functional theory (DFT) analysis after the chemo-responsive dyes exposure to volatile organic compounds (VOCs)
Taking variation operation on the corresponding orbit,
Equation () can be obtained.
[−1
2∇2+∫𝜌𝑟′
∣𝑟−𝑟
′∣𝑑𝑟′+𝑉
𝑒𝑥𝑡 (𝑟)+𝑉
𝑋𝐶 (𝑟)]Φ(𝑅)
=𝜀
𝑖Φ𝑖(𝑟)()
In Equation (), we have
𝑉𝑋𝐶 (𝑟)=𝜕𝐸𝑋𝐶 [𝜌]
𝜕𝜌 (𝑟).()
The research is developed in the binding energy, orbital
energy level gap between highest occupied molecular
orbital (HOMO) and lowest unoccupied molecular orbital
(LUMO), molecular plane angle, distance and charge
changes and dipole moment (Figure ) to investigate the
mechanism of the combination of chemo-responsive mate-
rials and volatile gas molecules. The popularization of
computers has led to considerable development in quan-
tum chemical calculations. (Gu et al., ,).
.. Binding energy
The energy released during the reaction between por-
phyrin and ethanol can be used to analyze the binding abil-
ity of porphyrin compounds and ethanol. That is, the bind-
ing energy is the difference between the total energy before
the reaction with ethanol and the energy of the product
in the system after the reaction. It is shown in Figure
that the conversion equation between the energy unit Har-
tri (a.u) and the commonly used energy unit kcal/mol is
a.u =. kcal/mol.
.. Orbital energy level gap
As the frontier molecular orbital theory proposed by Fukui
Kenichi states, orbital energy level determines the elec-
trons transfer and other important chemical properties of
molecules. In view of this, there are HOMO and LUMO
in the molecule. The electron is distributed on HOMO
with the highest electron energy, while no electron is dis-
tributed on LUMO and the electron energy is lowest. In
the chemical reaction, the electrons on the HOMO in the
high-energy state suffer the least bound with the highest
activity. And they are liable to transfer to the LUMO in
the low-energy state, thus causing the transfer of electrons
as presented in Figure . To sum up, HOMO and LUMO
orbitals are the key to the reaction among molecules in a
system. There are studies showing that the less the energy
required to excite electrons, the more favorable it is for elec-
trons to transition from HOMO to LUMO. It indicated that
the smaller energy gap is more likely to initiate chemical
or physical reaction.
A - . .. 9
.. Molecular plane angle
In the combination of the chemo-responsive materials and
the volatile gas molecule, the change of the substituent
on the heterocycle of the materials will lead to change
in the distance and the angle between the two molecular
planes, as shown in Figure . This change is often closely
related to the combination between the material and the
molecule to be tested. Therefore, the sensitivity of the
chemo-responsive material to the tested molecule can be
revealed in the distance and angle between the two.
.. Distance and charge changes among
atoms
The essence of a chemical reaction is the sharing of elec-
trons between the reactants, which can be characterized by
the change in charge before and after the chemical reac-
tion. The strength of the intermolecular interaction can be
reflected in the distance between them to a certain extent
(Gu et al., ). Chemical calculation softwares (such as
Mercury) are used to calculate the distances between the
-F of BODIPY and the -C, -C, and -C of hexanal so that
the interaction strength of BODIPY and hexanal molecu-
lar can be explored. The breaking and formation of chemi-
cal bonds in the process of chemical reaction is essentially
the transfer of electrons among reactants. Reasonably, the
transfer can be characterized by the charge change before
and after the compound reaction.
.. Dipole moment
The dipole moment refers to the result of distance between
center of positive charge (r) and negative charge (q) multi-
plied by the electric quantity given as 𝜇=𝑟×𝑞, D (debye).
The dipole moment can be employed to judge the polar-
ity of the molecule. If the dipole moment is equal to ,
it is a nonpolar molecule; otherwise, the molecule is a
polar molecule with dipole moment not equal to . The
greater the dipole moment, the stronger the polarity of the
molecule. The comparison between porphyrin and vari-
ous M-porphyrins before and after the action of ethanol is
shown in Figure .
The above discussion indicated that the coloring mecha-
nism of the specific combination of chemo-responsive dyes
molecules and VOCs molecules detected by odor imag-
ing technology mainly involves spatial structure, reaction
energy,orbital energy level difference, and dipole moment.
Akman et al. () found that the molecular spatial struc-
ture of chemo-responsive dyes has a greater impact on
its binding to VOCs, and the binding capacity between
the detected substance and M-porphyrin would change
with the optimization of the molecular spatial structure of
chemo-responsive materials.
4MODIFICATION AND
IMPROVEMENT OF
CHEMO-RESPONSIVE DYES
Conventional chemo-responsive dyes have a good detec-
tion effect on constant gas, but poor detection performance
to trace gases. It primarily comes from the fact that chemo-
responsive materials with small particle size do not have
an effective and sufficient contact with the tested gas.
M-porphyrins and BODIPY compounds are heterocyclic
compounds with a π-πconjugated structure formed by
connecting pyrrole molecules, and the H atom of the NH
situated in center of the molecule can be replaced by metal
ions to form M-porphyrin. With a good color sensitivity,
easy modification and stable properties of the π-conjugated
system, the chemo-responsive material is usually under
the treatment of nano-dispersion to increase the bind-
ing force of the colorimetric medium substrate and gas
molecule ultimately to improve its sensitivity. Nanomate-
rials are a kind of typical mesoscopic systems, and macro-
scopic objects reduced to the nanoscale to cause obvious
changes in their optical, mechanical, and chemical prop-
erties (Hu et al., ;H.Zhouetal.,;X.B.Zhang
et al., ). Due to the large surface area of nanomateri-
als, nanoparticles have become a good choice for immo-
bilizing and supporting biochemical molecules. Currently,
M-porphyrins and BODIPY are used to fabricate colori-
metric sensors. In terms of the application, they can be
further modified to improve the sensitivity and stability
of the sensor. Based on the specific surface area effect,
small size effect, quantum effect, and interface effect of
polymer nanospheres (Liao et al., ; Scholl et al., ;
Sousa et al., ), it can be polymerized with chemo-
responsive materials to form nanomaterials with higher
sensitivity and chemical activity, which can interact sta-
bly with the detected VOCs to enhance the color rendering
effect (Charoensuk et al., ). The common nanomodi-
fication methods of chemo-responsive dyes are discussed
below.
4.1 Nano-self-assembly
Thanks to the intermolecular forces being noncova-
lent bonds, the nanoparticles self-assembly of chemo-
responsive dyes can better ensure that the electronic struc-
ture of nano-scale materials is not destroyed. The method
10 A - .. .
FIGURE 5 The flow chart of nano-self-assembly porphyrin
is widely used because it is simple and easy to con-
trol the shape of nano-porphyrins. Self-assembly refers
to a technology in which molecules spontaneously com-
bine through a certain assembly method so that these
molecules can have a certain ordered structure and a cer-
tain regular geometric appearance. It is mainly based on
hydrogen bonds, coordination bonds, π-πstacking, and
the role of noncovalent bonds such as van der Waals
forces (Beer et al., ;D.Lietal.,;Palmer&
Stupp, ; Shimizu et al., ). Currently, the com-
monly used self-assembly methods are systematized into
solid-phase self-assembly method and liquid-phase self-
assembly method. The solid-phase self-assembly method
is mainly a physical vapor deposition method, and
the liquid-phase self-assembly method mainly includes
reprecipitation method, surfactant-assisted method, and
ion self-assembly method. Among them, the surfactant-
assisted method is an improvement of the reprecipitation
method, which is simple to operate and can be widely
used in chemo-responsive materials of nanoparticles self-
assembly (Guan et al., ). Figure shows the flow chart
of nanoself-assembly porphyrin.
Chem-responsive dyes will generate rod-like structures
or sheet-like structures after nano-self-assembly. If the
face-to-face H-type polymer is formed by intermolecular
π-πinteraction, the average width of self-assembled nano-
particles is about – nm, and the length is about
.–. µm. The following Figure a,b and Figure c,d
are, respectively, the scanning electron microscopy (SEM)
and transmission electron microscopy (TEM) images of
self-assembled two porphyrin materials (TPP-Zn, TPP-
Mn). It can be seen from the figures that self-assembled
MnTPP has a square piece with a certain thickness, and
the side length of the square is about nm. It can be
A - . .. 11
FIGURE 6 The synthesis diagram and the transmission electron microscopy (TEM) images of poly (styrene-co-acrylic acid) (PSA)
microspheres
further proved that in the self-assembly process of
N-ZnTPP, in addition to the π-πbetween porphyrin
molecules interaction, hydrogen bonding is also included.
4.2 Nano polymerization
Poly (styrene-co-acrylic acid) (PSA) is a kind of self-
assembly polymer nanosphere with surface aggregation
function. The PSA is prepared through soap-free emulsion
copolymerization, and styrene (St) and acrylic acid (AA)
are used as monomers while ammonium persulfate (APS)
is used as a nonbuffer initiator in the medium. Styrene
and acrylic acids are used as raw materials, and then
chemo-responsive materials are added. Heat is applied
to increase the temperature of the mixed system. Due to
the increase in temperature, the solubility of the chemo-
responsive dyes increased greatly (Yu et al., ). On the
other hand, the activity of the acrylic chains on the sur-
face of the nanospheres is enhanced, resulting in more
free volume. This can promote the interaction between the
chemo-responsive dyes and the nanospheres in the solu-
tion. If they are in full contact with the surface of the
nanospheres, hydrogen bond will be formed between the
material molecules and the carboxyl groups of acrylic acid
to increase the amount of material adsorbed on the outer
hydration layer. A concentration gradient will then be cre-
ated between the surface and the inside of the nanosphere.
Due to the strong hydrophobic interaction between the
material and the styrene segment, the chemo-responsive
dyes diffuse from the surface of the nanosphere to the
inside, reducing the concentration in the water phase
(Ruenraroengsak & Tetley, ). To maintain balance,
more material in the suspended particles will dissolve until
all the molecules are absorbed by the nanospheres. The
synthesis diagram is shown in Figure . In the above-
mentioned reaction conditions, this experiment has been
studied on the particle size of nanosphere (the amount
of acrylic acid), the mass ratio of nanospheres to chemo-
responsive materials, and the selection of emulsifiers to
optimize the synthesized nano-chemo-responsive materi-
als so as to enhance the performance of nano-colorimetric
sensors for gas detection (Lin, Kang et al., )
The prepared PSA nanosphere sample was diluted
approximately – times with deionized water and son-
icated for min, and then dropped onto a Cu net, and
irradiated under an infrared lamp. After reaching a dry
state, TEM was used to observe the shape, size, and dis-
persion of the latex nanoparticles prepared. From a micro
perspective, Figure gives the TEM images of PSA micro-
spheres. It can be seen that the PSA nanospheres are rela-
tively smooth, uniform, and spherical. It can be seen that
in the fabrication of nano-colorimetric sensors, the par-
ticle size of polymerized nanospheres, and the choice of
surfactants can have an impact on the quality of the sen-
sor. In macro perspective, it is obvious that the application
of polymeric nanospheres in imaging technology based
on chemo-responsive dyes has played a significant role
in enhancing the color development of chemo-responsive
dyes.
4.3 Nanoporous modification
Mesoporous silicas nanospheres materials are porous due
to the characteristics of large specific surface and pore
12 A - .. .
FIGURE 7 The preparation process of nano porous modified dye and the scanning electron microscopy (SEM), transmission electron
microscopy (TEM), and energy dispersive X-ray spectroscopy (EDS) image of porous silica nanospheres (PSN)
volume of porous silica nanospheres (PSN), as well as tun-
able pore size (Baeckmann et al., ; Dumanoğulları
et al., ;Y.Wangetal.,). Herein, we described
an approach to fabricate a colorimetric sensor using some
chemo-responsive dyes based on PSN for VOCs detection.
The screened chemo-responsive dyes were mixed with
PSN and dispersed in N, N-dimethylacetamide (DMAC)
and ethanol, and then stirred at ◦Cforhr(S.Xu
et al., ). The specific process is shown in Figure .As
shown from the figure, experimentally synthesized PSN
is white powder with the appearance of ordered pores.
Figure shows the SEM, TEM, and energy dispersive X-
ray spectroscopy (EDS) image. The white powder is com-
posed of spherical nanoparticles, the average diameter of
which is estimated to be nm. The corresponding EDS
(the small image in Figure a) finds out the presence of
Si and O elements, indicating that the spherical nanopar-
ticles are composed of silica. At higher magnification
(Figure b), it can be seen that the nanospheres are porous.
Figure c–e shows TEM images of different surface-
modified PSN. In terms of particle size, pore size and sur-
face morphology, compared with PSN, the PSN-CH, PSN-
COOH, and PSN-NHhave no significant changes. How-
ever, the shell of PSN-CHis smoother than that of other
samples (Figure d). Since the surface chemistry of the par-
ticles has an important influence on the Zeta potential, the
particles grafted with different silane coupling agents have
a variety of surface chemistry.
It is thus clear that the nanomaterial modification
method makes it smooth for VOCs molecules to diffuse
in the modified chemo-responsive dye pores in order
to improve the effective contact area and binding effi-
ciency of the chemo-responsive dyes and the gas. The
above discussed nano-modification methods of chemo-
responsive dyes includes nano-self-assembly, nano poly-
merization, and nanoporous modification, and they would
have good performance results combined with different
chemo-responsive dyes. Among them, nanoporous mod-
ification such as PSN would help improve the diffusion of
volatile gases in the pores of the sensing substrate, thus
form a pre-enrichment effect to improve the sensitivity
(Duan et al., ). The specific surface area of nano poly-
merization such as P (St-co-AA) could greatly increase
the specific surface area of the chemo-responsive dyes
molecules combined with the outside, which has strong
sensitization potential (Y. Liu et al., ).
A - . .. 13
5ODOR IMAGING SYSTEM
Through the imaging detection technology, odor infor-
mation is converted into visual information to make
odors ‘visible.’ Compared with traditional electronic nose
technology, olfactory imaging technology is more intuitive
and vivid. Sensor arrays are the main working units of
the imaging technology based on chemo-responsive dyes;
therefore, the fabrication of sensor arrays is a key step
in the design of imaging detection system. The imaging
detection system is composed of gas-collecting chamber,
vacuum pump, light source, image acquisition device,
and reaction chamber. Typically, the choice of light source
and image acquisition device and the design of reaction
chamber have significant effects on the performance of
the imaging system based on chemo-responsive dyes.
5.1 Fabrication of sensor arrays
The fabrication of the colorimetric sensor array is the
core technology of the odor imaging system. For a sen-
sor, its basic requirements include high sensitivity, anti-
interference stability (insensitive to noise), linearity, high
reliability, repeatability, safety, interchangeability, high
precision, high response rate, wide measuring range, and
wide working temperature range. In addition, there should
be special requirements for specific types of sensors. At
present, the substrates used in the production of imag-
ing sensors mainly include hydrophobic materials such as
reverse phase of silica gel plates and polytetrafluoroethy-
lene. A chemical developer can be made into a single visual
sensor, the solid support materials for colorimetric sensor
array (Long et al., ;K.S.Suslicketal.,; Zaragozá
et al., ). Sensors made up of different color develop-
ers have different sensitivity characteristics. Multiple sen-
sors made up of different color developers together are
arranged to form a visual gas sensor array. The combi-
nation of the sensors has improved the detection accu-
racy, and greatly expanded the application range of imag-
ing (Jiang et al., ). Therefore, imaging sensors gener-
ally appear in the form of arrays. The manufacturing pro-
cess of the imaging sensor array is mainly divided into four
steps, namely:
∙Selection of gas chemo-responsive materials: Hydropho-
bic porphyrins, BODIPY, and pH indicators that can
have color changes to gases are selected as gas chemo-
responsive materials.
∙Selection of solvent: According to the materials selected
in step (), the corresponding solvent is selected, which
can be used to dissolve the gas chemo-responsive mate-
rials at the concentration of .–. mol/L.
∙Selection and production of sensor substrate materi-
als: (i) White polytetrafluoroethylene material ( mm ≥
thickness ≥ mm) are selected and processed into rect-
angles or squares; (ii) each square is neatly engraved into
a square with a side length of .∼. mm (or a circle
with a diameter of .∼. mm) and a depth of .∼.
mm, which are treated as marks; (iii) each mark is cov-
ered with thick hydrophobic white stable agent with a
layer of .∼. mm.
∙Production of array: .∼ µl of chemo-responsive
material solution is taken through a micro-sampling
device then fixed on the stabilizer at the mark of the sub-
strate plate, and then dry the solvent to obtain CSA.
5.2 Composition of imaging system
Generally, the imaging detection system is mainly com-
posed of light source, image acquisition method, and
reaction chamber. The schematic diagram of the system
designed by Chen, Liu, and Zhao (),Huangetal.
(), and Lin et al. () is presented in Figure .Scan-
ners and cameras are the main methods of image acquisi-
tion at present. Specifically, scanners have the advantage of
lower price and the lamp tube can move with the stepping
motor to provide uniform illumination. However, com-
pared with cameras, scanners are bulk, slow in scanning
speed, and generally not supported to carry out secondary
development. The camera has the advantages of being dex-
terous, capable of collecting images online, and being able
to be used for secondary development. Therefore, the cam-
era is selected as the image acquisition method to capture
images. The image in () in Figure is the physical image
of the CCD and CCD camera. The light collection area of
the three CCD image sensors of the CCD camera is rela-
tively large, making the CCD camera have a higher signal-
to-noise ratio, better sensitivity, and a wide dynamic range
than a CCD camera (Lin et al., ).
The uniformity of the light source will directly affect the
quality of the image, thereby affecting the performance of
the device. Uniformity, brightness, efficiency, service life,
and spectral characteristics, and so forth. are the major fac-
tors in the selection of the camera light source. The com-
monly used light sources are halogen lamps (fiber light
sources), high frequency fluorescent lamps, and LED light
sources (Ferroudj et al., ;X.Zhouetal.,). For
example, Zhang et al. () carried out a study using
Hamamatsu L- pulse xenon lamp for the NH con-
centration detection. Compared with other light sources,
LEDs not only have high brightness and long service life
but also they can be designed into a complex structure to
achieve different angles of light source illumination. Gen-
erally, there are strip diffuse reflection LED light sources
14 A - .. .
FIGURE 8 The schematic diagram of the imaging detection system
and diffuse reflection integrating sphere light source being
used, as shown in () in Figure . Overall, concerning the
latter, a circle of LED light sources on the bottom plane can
be evenly reflected on the image so that the brightness of
the image is uniform
The stability, repeatability, signal response time, and
degree of the sensor signal will be greatly affected by the
structure of the reaction chamber of the odor imaging sys-
tem. During the reaction process, it is necessary to ensure
that the gas flow can be uniformly contacted with each
chemo-responsive material. Therefore, the optimization of
the reaction chamber is of great significance to improve
the detection capability of imaging sensors. It can be seen
from () in Figure (a–e) when the baffle is too close to the
inlet, the airflow velocity on both sides of the sensor array
becomes smaller. After continuous adjustment of the cur-
vature and position of the baffle, the optimal results of the
curvature and position of the baffle are shown in () in Fig-
ure . It was not conducive to improve the reaction between
CSA and VOCs because the airflow was extremely concen-
trated in the middle of the chamber. The application of par-
ticle tracking that shows the particle distributions also sup-
ports the results on simulation of the velocity fields. The
simulation of particle tracking matching to different kinds
with baffles is shown in () of Figure . All of the reaction
chambers with a small or large curvature baffle and the
one without baffle cannot meet the requirements because
their particle cannot cover the CSA and the distribution are
not as uniform as the reaction chamber with an optimized
baffle.
5.3 Signals processing and data analysis
The image signals processing module is the core com-
ponent of the imaging system, which is responsible for
extracting and analyzing the color-difference information.
And characteristic difference matrix is generally used to
represent the difference in information before and after the
reaction of each color area of the visual sensor array. On the
premise of the requirement, the design and application of
image processing algorithm is ought to occupy small cal-
culation amount and shorter calculation time. The current
image processing process is composed of the positioning
of the graphic center point, the selection of the charac-
teristic region, and the establishment of the color space
model. The specific implementation steps include filter
noise reduction, binarization, morphological processing,
figure centering, characteristic region selection, and so
forth. in order to segment the color regions of the sensor
array and extract the corresponding color information, and
so forth, which is shown in Figure .
A - . .. 15
FIGURE 9 The image processing flowchart of colorimetric sensor
.. Characterization of images of
colorimetric sensors
In the detection of gas with olfaction visualization technol-
ogy, the characteristic values of target image are generally
extracted from the color regions under the RGB, HSV, and
Lab color space. Applying different color spaces in detect-
ing different objects would be beneficial to the characteris-
tic extraction for pattern recognition analysis. Among the
three spaces, the RGB color model (red-green-blue three
primary color model) is the most commonly used (James
et al., ). In the HSV model, three-dimensional values
of hue, saturation, and brightness indicate color. The lab
color space uses three elements to instruct colors. The coor-
dinate L is the representative to the brightness of the color,
but the a and b channels represent the chromaticity (Leon
et al., ). The a channel represents the degree of color
changes between green and red, but the b channel reflects
the different degrees of color between blue and yellow. In
summary, the characteristic variable matrix of the sensor
color area is calculated based on the RGB, HSV, and Lab
color models, as shown in Figure . Besides, for the con-
venience of data visualization and comparison, the values
of each channel obtained in all spaces are normalized to
the interval of , .
.. Selection of region of interest
The color-difference variables are obtained by subtracting
the two images before and after the reaction between the
visualization sensor and the gas. The precise position of
each sensor before and after the reaction is firstly deter-
mined so that the color region in the image can be sub-
tracted in one-to-one correspondence. The region of inter-
est (ROI) can be selected as the mark of unique position of
each sensor, which further affects the accuracy of extract-
ing color characteristic differences. In view of this, the
method of first moment, minimum bounding rectangle,
and ellipse fitting were employed to find the coordinates
of the ROI, as shown in Figure .
With the first moment method of obtaining the cen-
ter of gravity of the image, the center of gravity of the
color-developing region of the visualization sensor can be
regarded as its heart (Phelan, ). Supposing that the
image can be represented by the function f(x, y), because
the points on the image are two-dimensionally discrete, the
p+qmoment of the image f(x, y) can be defined as Equa-
tion in Figure .
The basic idea of the least squares ellipse fitting algo-
rithm takes advantage of the mathematical meaning of
ellipse, which can be expressed as Equation in Figure .
The optimal parameter can be solved by Equation , that
is, the optimal solution of the linear equation system is
also the optimal parameter of the ellipse, and the center is
calculated. Using minimum bounding rectangle method,
the contour of the target image needs to be scanned first
to obtain the outer regular rectangle (Jo & Jung, ).
Then, the minimum bounding rectangle is obtained by
rotating and translating the main or auxiliary axis of the
regular rectangle, and the center point coordinates o(x,y)
of the minimum bounding rectangle are calculated accord-
ing to the four vertex coordinates p—p, as shown in
Figure .
16 A - . ..
FIGURE 10 The flowchart of signals processing and data analysis
A - . .. 17
.. Image processing after ROI selection
In the actual production of the sensor array, the chemo-
responsive materials need to be formulated into a certain
concentration solution. When the solution is spotted on
the solid substrate with a capillary, the diffusion of the
solution is a process, which results in different color layers
surrounding the center of spotting. In order to accurately
extract the color difference before and after the reaction
between sensors and measured objects, it is necessary to
perform image processing after ROI selection.
Generally, after selecting the ROI of the color-
developing area of the sensor, the center point of the
sensor color-developing area is the center of a circle, of
which the radius is a certain pixel length (it needs to be
less than the smallest inscribed circle of the color area).
The circular area is artificially set as an ROI, and named
as a characteristic area, as shown in Figure . The flood
filling method is a kind of region filling method. With the
former method, it can fill areas with very similar colors or
brightness so as to achieve the region segmentation. The
purpose of applying the flood filling method is to separate
the central color layer and the peripheral color layer
of the sensor, and the characteristic extraction regions
(color layers) of the color change of sensor as shown in
Figure . After obtaining different regions of interest,
the gray values of each channel of the ROI before and
after the VOCs exposure to chemo-responsive materials
are extracted and subtracted to characterize the change of
channels. And the response to VOCs (detected analyte)
could be characterized with a N-dimensional vector (N
dyes x color component difference (RGB)).
.. Data analysis and modeling
After the imaging processing, the gray values of the three
components of R, G, and B of ROI image before and after
the colorimetric sensor exposure to detected VOCs are
extracted. The feature matrix of gray values is enlarged by
normalization, and the normalization equation is named
as Equations –. Here, taking the R component as an
example, Δ𝑅𝑖 and Δ𝑅𝑖′are the characteristic variables
before and after the normalization of the R component of
the 𝑖-th chemo-responsive dye, and Δ𝑅𝑚𝑎𝑥 and Δ𝑅𝑚𝑖𝑛
are the maximum and minimum original characteristic
variables of the R component of all chemo-responsive dyes.
The normalized matrix is used to generate gray-scale dif-
ference images of the three components of R, G, and B.
Finally, the gray-scale images are superimposed to obtain
the difference feature image before and after colorimetric
sensors exposure to the gas.
To identify different unknown odor, it is necessary to
use the pattern recognition method to train and predict
the known samples and construct a discriminant model.
In general, some chemo-metric methods, including prin-
cipal component analysis (PCA), linear discriminant anal-
ysis (LDA), K-nearest neighbor (KNN), and partial least
squares (PLS) were used to minimize the prediction vari-
able complexity (Domínguez-Aragón et al., ). PCA is a
statistical method that has received considerable attention
(Lin et al., ). By orthogonal transformation, many vari-
ables are converted into a set of few variables called princi-
pal components (PCs) (Qin, et al., ). As a pattern recog-
nition method, LDA is carried out based on minimizing the
within-class distance and maximizing the between-class
distance (Skrbic & Durisic-Mladenovic, ). KNN is a
kind of method which stores all the sample data of the cali-
bration set in the computer and the distance between each
unknown sample. Then, each training sample is calculated
one by one to find the nearest k based on the assumption
that samples of the same kind are close to each other in
the simulation space. PLS is a supervised method used to
build the linear and multivariate calibration model, which
has the advantages of avoiding collinearity and interaction
problems in the detection indexes data (Lin, Duan, et al.,
;Zhu et al., ).
6APPLICATION IN SENSING FOOD
QUALITY AND SAFETY
Imaging technology based on colorimetric sensor fabri-
cated by chemo-responsive dyes is novel e-nose for the
detection and classification of food materials based on
their VOCs, it has gained multitudinous attention due to
its intuitive detection results, excellent detection perfor-
mance, and wide-spectrum detection range. In fact, imag-
ing technology based on chemo-responsive dyes is playing
an increasingly important role in rapid nondestructive test-
ing of food quality and safety. Scholars at home and abroad
have conducted several studies on the application of this
field. Imaging technology based on chemo-responsive dyes
has been proposed to evaluate the states of some com-
modities or agro-products such as detection of cereal (Lin,
Kang, et al., ; Lin, Wang, et al., ), wine and vine-
gar flavored foods (Guan et al., ,; Lin et al., ;
Lin, Jiang, et al., ), poultry meat (Alimelli et al., ;
Huang et al., ), aquatic products, fruits and vegetables,
and tea.
18 A - . ..
FIGURE 11 (a) The interactive sensor to distinguish the storage period of rice; (b) diagram of three-dimensional principal component
scores from volatile organic compounds (VOCs) of rice with different storage time; (c) identification rates from K-nearest neighbor (KNN)
model; (d) identification rates from linear discriminant analysis (LDA) model
6.1 Application in grain quality
assessment
The quality of agricultural products (agro-products)
declines during storage and transportation. This decline
in quality increases with increasing storage time which
reflects in the increase of lipids, sulfides, and furan
compounds in grains. The characteristic volatile gases of
the agro-products with mold infection are captured to
detect and analyze the condition of the grains, eventually
providing guarantee for the quality and safety of the
grains. VOCs can have a color-developing effect with
chemo-responsive materials, thereby enabling the ‘visual-
ization’ of odor information. Based on this, volatile odors
would be present through visual images. The intuitive
detection results and excellent detection performance of
the odor imaging technology have gained great attention
and research in the field of grain quality detection.
While grain is constantly aging, the important quali-
ties such as color and smell will decrease. Guan et al.
() realized correct identification of % in the storage
time of rice samples through the odor imaging technology.
Lin et al. () used imaging method to assess and iden-
tify VOCs of rice with different storage. Figure a shows
a characteristic diagram of rice. There appears different
degrees of color change and forms a characteristic map cor-
responding to each rice storage. Each colorimetric unit of
the image is converted into three colorful RGB difference
images, and a total of variables are obtained ( color sen-
sitive materials × color components).
Figure b is a three-dimensional scatter plot with prin-
cipal component scores of five types of samples as the
A - . .. 19
FIGURE 12 (a) The initial image, final image and characteristic image of green tea, oolong tea, and black tea; (b) the classification
results of green tea, oolong tea, and black tea
input. The cumulative contribution rate reaches .%
(>%). From the Figure c, we can conclude that the
KNN model achieves the recognition rate of the calibra-
tion set and prediction set for %. As can be seen in Fig-
ure d, the recognition rates of the calibration set and pre-
diction set are the highest % and %, respectively. It can
be obtained that rice with different storage periods can be
qualitatively distinguished by imaging technology based
on chemo-responsive dyes.
6.2 Application in classification of tea
types
Due to the difference of processing techniques, tea can
be divided into many different types, of which green tea,
oolong tea, and black tea are the three most common types.
The rapid and safe classification of tea types is an urgent
need for the agriculture and food industry to provide a reli-
able tool to tea traders and consumers, thereby limiting
fraudulent labeling (Ye, ). L. Li et al. ()evaluated
green tea quality by combining olfactory visualization sys-
tems (odor imaging technology) and hyperspectral imag-
ing with a classification accuracy of %. In addition, the
sensory evaluation of black tea aroma quality based on an
olfactory visual sensor system was proposed.
Providing three color components of R, G, and B as
the characteristic value by each color reagent, there are
a total of characteristic variables ( color reagent ×
color components) according to the study by Chen, Liu,
Zhao, Ouyang, et al. (). Figure a shows the initial
image, final image and characteristic image of green tea,
oolong tea, and black tea. Utilizing the PCA and LDA pat-
tern recognition methods, classification results were com-
pared and analyzed. As can be clearly seen in Figure b,
compared with the boundary between black tea and green
tea, the boundary between oolong tea and green tea is
more obvious and can be better distinguished. As shown in
20 A - . ..
the figure, a good classification effect has been achieved,
and the interactive verification of different fermented tea
reached a % correct rate.
6.3 Application in liquid foods quality
identification
Odor is an important evaluation basis for authenticity
identification and quality classification of liquid foods
such as vinegar and wine. Chemo-responsive materials are
employed to interact with the volatile gas of the detected
object and the color of the dye molecule changes signifi-
cantly. After computer processing, a specific digital signal
is formed, and finally the corresponding pattern recogni-
tion method is used to perform qualitative and quantitative
analysis (Chen, Liu, Zhao, Ouyang, et al., ; Chen, Liu,
&Zhao,; Chen, Hui, et al., ; Chen, Li, et al., ;
Guan et al., ,)
In order to adequately monitor acetic fermentation, it is
necessary to consider the changes in the content of charac-
teristic VOCs such as alcohols, esters, and acids to grasp the
main condition, and also pay attention to the detection and
characterization of other volatile gases to obtain the overall
information. Guan et al. () screened out nine kinds of
porphyrins and three kinds of pH indicators to fabricate a
sensor array with four rows and three columns ( ×). As
shown in Figure a, the characteristic image of the CSA
before min has been changing, indicating that the reac-
tion has not yet reached equilibrium. When the reaction
time reaches min, the color of the difference graph tends
to be stable. Based on gas chromatography-mass spectrom-
etry (GC-MS) to detect the alcoholic strength during vine-
gar culture fermentation, Figure b shows the changes
in ethanol content during acetic fermentation. Figure c
shows the correlation between the measured value of alco-
hol and the predicted value of the BP-ANN model. It can
be seen from the figure that the odor information of dif-
ferent fermentation days collected by the imaging method
is highly correlated with the ethanol content measured by
GC-MS.
In the process of rice wine fermentation, it is absolutely
necessary to monitor some vital physical and chemical
indicators, such as the dynamic changes of alcohol and
aromatic substances (Ouyang, Zhao, Chen, et al., ;
Ouyang, Zhao, & Chen, ). The odor imaging system is
not susceptible to changes in humidity easily, thussuitable
for analyzing and detecting the characteristic odor of liq-
uid food. According to the characteristics of the rice wine
odor, nine porphyrins and six pH indicators are preferably
screened as gas-sensitive materials to form a sensor array
with good selectivity and sensitivity. After min of rice
wine samples exposure to CSA, the characteristic images of
samples with different storage time are obtained, as shown
in Figure a. Different degrees of color changes occurred
after each chemo-responsive material reacting with rice
wine of different storage time. Red-green-blue color com-
ponents of each color reagent were extracted as the char-
acteristic value; the rice wine sample got characteristic
variables ( color reagents× color components). These
variables were processed by LDA to distinguish rice wine
of different storage time. Figure b indicates the scores of
LD and LD. And the two discriminant functions embrace
.% of the whole information. As displayed in the figure,
the four types of rice wine with different storage time can
be completely distinguished, and there will be no overlap
between different categories.
6.4 Application in poultry and red meat
quality and safety detection
The demand for poultry as one of the fast-growing foods is
on the increase globally. With the continuous increase in
demand for chicken, more and more attention from con-
sumers and researchers has been gradually drawn to the
freshness of chicken, especially focusing on the content of
TVB-N, an important reference index for chicken fresh-
ness. Odor imaging technology presents a colorimetric
dye-based array for naked-eye detection of chicken meat
spoilage, which demonstrated that sensors response was
correlated with meat spoilage progress (Magnaghi et al.,
). A fusion strategy based on colorimetric sensors and
NIRs was applied to rapidly identify Pseudomonas spp.in
chicken, and eventually BP-ANN achieved .% classifi-
cation rate in the predication set (Y. Xu et al., ).
New electronic noses with low-cost imaging technology
based on chemo-responsive dyes have also been explored
and applied to rapid detection in this field. Chen, Hui,
et al. () fabricated a novel and low-cost colorimet-
ric sensors array, which has a high potential in evalu-
ating chicken freshness with the classification results of
both % in the calibration and prediction sets by the
orthogonal linear discriminant analysis (OLDA) and adap-
tive boosting (AdaBoost) algorithm. The national stan-
dard GB/T .- was adopted as the determination
method. In terms of imaging method, the ×sensor
array is made and a scan is performed to obtain the images
before and after the reaction between the CSA and chicken
samples (Khulal et al., ). It concluded that the low-
est RMSEP can be obtained when the principal component
number is and φis .. In the prediction set, when Rp
equaled to . and RMSEP equaled to .mg/g,
the best BP-ANN model is obtained. Therefore, imaging
technology based on chemo-responsive dyes can quantita-
tively detect the freshness of chicken.
A - . .. 21
FIGURE 13 (a) Difference images of colorimetric sensor arrays exposure to vinegar substrate to different time arranged from to
min; (b) variation of alcoholic strength through acetic acid fermentation; (c) Correlation between the measured value and the BP-ANN model
prediction of ethanol
Pork is one of the most important red meat on the table
because of its fine and soft fiber, less connective tissue,
and more intramuscular fat in muscle tissue. Therefore,
predicting pork meat quality traits is considered a key fac-
tor that protects consumers’ health and safety. The control
of the origin of red meat is a major concern for the con-
sumers, producers, and distributors (Haddi et al., ).
The purpose of the Begoña’s study (de la Roza-Delgado
et al., ) was to assess the suitability of near infrared
spectroscopy (NIRs) technology to establish a physico-
chemical characterization of the Asturcelta meat. H. Li
et al. () investigated the feasibility of rapid evaluation
of pork freshness using a portable e-nose based on an odor
imaging technology by printing chemically responsive
dyes. The optimum discrimination rates were % and
.% for the training and prediction sets, respectively.
22 A - . ..
FIGURE 14 (a) Feature images of rice
wine from different storage ages; (b) linear
discriminant analysis (LDA) result for
differentiating rice wine from different
storage ages by olfactory imaging sensor
Valdez et al. () developed ForceSpun polydiacetylene
nanofibers as colorimetric sensor and demonstrated good
detection of meat spoilage ability based on released amine
vapors during food degradation. These indicated that
this technology has a high potential for real-time use in
monitoring red meat quality in meat processing industries.
6.5 Application in aquatic products
quality evaluation
Aquatic products such as freshwater fish have a relatively
high content of unsaturated fatty acids and are easily oxi-
dized. At the same time, the protein in the fish body
can easily decompose to produce bad odors, resulting in
flavor changes and even endangering consumers’ health.
During the spoilage process of fish, the volatile compo-
nents mainly include hydrocarbons, alcohols, aldehydes,
ketones, esters, phenols, sulfur, and nitrogen compounds.
A visual sensor array is made of porphyrin compounds and
a pH indicator, which can detect changes in fish freshness.
Huang et al. () showed that the sensor array measure-
ments of fish were classified by radial basis function (RBF)
algorithm into three groups, corresponding to day (group
), day , , , and (group ), and day and (group ). The
total classification accuracy was .%. This suggests that
the system is capable of inspecting quality of fish and per-
haps other foods containing high protein. A new optoelec-
tronic nose based on eight sensing materials containing pH
indicators, Lewis acids, and an oxidation-reduction indica-
tor for the shelf-life assessment of fresh sea bream in cold
storage has been developed (Zaragoza et al., ). Olafs-
dottir et al. () presented a multisensor technique for
fish quality detection based on visible (Vis) spectroscopy,
image analysis, and electronic noses. Dowlati et al. ()
summarized the application of machine-vision techniques
in fish quality assessment (Wu et al., ).
6.6 Applications in detection of other
fields
Imaging technology based on chemo-responsive dyes, as
a rapid detection method, has a qualitative and quantita-
tive analysis of various food volatile gases, so as to pro-
vide guarantee for food quality and safety monitoring.
It has been widely used in many fields, in addition to
the detailed examples described above, likewise, in other
fields, as shown in Table . Imaging technology based on
chemo-responsive dyes has embodied some effectiveness
of application in the detection of food. It is not only capa-
ble of acquiring imaging information from natural food but
it is also possible to explore the odor changes of processed
foods to monitor their quality, so as to meet the challenges
caused by food sample diversity and volatile component
complexity.
A - . .. 23
TABLE 2 Applications of odor imagingtechnology in other foods detection
Product Detection result Reference
Pork SVMR prediction model with the determination coefficient of . (H. Li et al., )
Milk Detection of streptomycin from simply-treated milk at nmol/L (Z. Liu et al., )
Honey Detection of streptomycin from honey at nmol/L (Z. Liu et al., )
Herbal medicine Discrimination (with % accuracy) of herbal distillates (Hemmateenejad, et al., )
Beer Error rate of categorization identification <% (Zhang, Bailey, et al., )
Pork sausage Detection accuracy rate of storage time of .% by PLS model (Salinas et al., )
Fruit pickle Analysis accuracy of toxic materials in adulterated of . (Bordbar et al., )
Coffee beans In quintuplicate runs of commercial coffees and controls, no confusions
or errors in classification by (hierarchical cluster analysis) HCA in trials
(B. A. Suslick et al., )
Abbreviations: PLS, partial least squares.
TABLE 3 A list of nondestructive methods combined with colorimetric sensors
Method Probe Detection Reference
Fluorescent sensor Manganese tetraphenylporphyrin Total polar material (Gu et al., )
Fluorescent sensor TFP-Graphene oxide Heparin (Cai et al., )
Fluorescent sensor Quantum dot Trinitrotoluene (Kui et al., )
Fluorescent sensor Polymer Bacteria (Liron Silbert et al., )
Colorimetric and fluorescent sensor Nanofibrous membrane(polyimide) HCL gas (Lv et al., )
Surface plasmon resonance Ag coated Au nanoparticles Copper (Lou et al., )
Diffuse reflectance spectroscopy Nix pro color Soil organic carbon (Swagata et al., )
Optoelectronic nose pH indicators, Lewis acids Sea bream shelf life (Zaragozá et al., )
Near infrared Chemo-responsive dye VOCs (Kutsanedzie et al., )
Hyperspectral imaging Colorimetric materials Rice storage time (Lin, Wang, et al., )
UV–visible spectroscopy Ag nanoparticles Copper (Colombo et al., )
Visible spectra Peptide coated SiO n-Butyl phenol (Takatoshi et al., )
Visual sensor RNA probe Bacteria (Sivakumar et al., )
Abbreviations: VOCs, volatile organic compounds.
7 ADVANTAGES AND
DISADVANTAGES OF IMAGING
TECHNOLOGY BASED ON
CHEMO-RESPONSIVE DYES
The applications of imaging technology based on chemo-
responsive dyes using colorimetric sensor as probes in the
evaluation of food safety and quality have many advan-
tages (Huang et al., ).
∙It has good selectivity, high sensitivity, low cost, simple,
and fast response.
∙Friendly to environment, reproducibility, and it is offi-
cially approved.
∙It has good correlation with human sensory for specific
applications in food safety and quality supervision.
∙This nondestructive technique is easy to build, provides
real-time detection and on-line monitoring of volatiles,
and requires only very short analysis time.
Although computer vision has the aforementioned
advantages, there are still some unovercomed limitations.
∙The equipment of imaging technology based on chemo-
responsive dyes are not petite to real-time tracking of
food quality.
∙The sensor arrays based on chemo-responsive dyes are
currently nonrecyclable.
8FUTURE TREND AND CONCLUSION
The foregoing contents suggest that imaging detection
paves the way for detecting and differentiating com-
plex components. In odor imaging technology based on
chemo-responsive dyes, chemo-responsive dyes are gener-
ally utilized as probes to interact with given analytes. This
interaction with odorants could be quantified by digital
imaging. To solve the associated problem, related studies
24 A - . ..
have employed a cellphone camera to capture imaging data
as imaging-collecting tool of imaging technology based on
chemo-responsive dyes. In current studies, Zaragoza et al.
() used a cellphone camera as the imaging tool to col-
lect signals. Askim and Suslick () developed a portable
hand-held device system for imaging detection. Based on
this, a small, less expensive hand-held device for use of
imaging technology based on chemo-responsive dyes was
developed to ensure food quality and safety.
Currently, applications of NIRs and hyperspectral imag-
ing have been explored for food quality detection and
monitoring. Therefore, it is possible to combine spec-
troscopy methods and imaging technology based on
chemo-responsive dyes to capture more information.
Table lists other methods coupled with imaging technol-
ogy in detection of various compounds. There is a need
to capture the odor information in samples by imaging
method and the subsequent use of spectroscopy methods
for the detection of the response signal via data processing
to present a way forward. This might solve the problem of
fewer variables available for screening in imaging meth-
ods. The model established by imaging technology based
on chemo-responsive dyes is more stable and accurate in
the near future and can be used for quantitative analysis of
food detection.
Odor imaging technology based on chemo-responsive
dyes has also been applied in many other fields such as
chemistry (R. Wang et al., ) and environment science
(Song et al., ;Lietal.,). Odor imaging technol-
ogy provides a facile, efficient, and sensitive approach for
the rapid detection and identification of chemical sub-
strates. In summary, the technique is simple, cheap, and
more time-efficient when utilized with miniature devices
such as NIRs. Thus, the industrialization of this technol-
ogy would economize the investment of equipment funds
when applied in food science and other fields. Moreover,
it is friendly to the environment with reproducibility. It
is also competent for detecting flammable, explosive, and
toxic compounds, which indicated the suitability of the
proposed approach in environmental quality. However,
further studies are still needed to facilitate the applica-
tions of the advanced evaluation technologies in various
industries.
This work presents a comprehensive insight into imag-
ing technology, including the sensing principle, imaging
principle and application for food quality and safety. The
implementation of nanotechnology in the sensor field for
rapid sensing gives impetus to the development of imaging
technology based on chemo-responsive dyes. The response
signal of the technology is timely and easy to be char-
acterized by color imaging method. Imaging technology
based on chemo-responsive dyes has proved itself as a lead-
ing detection technique with no or less sample prepara-
tion, nondestructive and rapid nature. In order to break
through the lack of intelligence and dexterity of imag-
ing equipment, it has emerged to replace the camera as
an imaging device for data collection. To make imaging
technique more analytically useful, the trend in future
will be oriented toward imaging technology based on
chemo-responsive dyes combined with spectroscopy meth-
ods with higher developments for the application in the
food industry.
We would like to thank all our nondestructive research
team of Jiangsu University and their deep gratitude for
their miscellaneous support in the current work. This work
has been financially supported by the National Natural
Science Foundation of China (), Self-innovation
Fund Project of Agricultural Science and Technology in
Jiangsu Province (CX(), SCX), the Project of
Faculty of Agricultural Equipment of Jiangsu University
(NZXB), and the Key R&D Program of Jiangsu
Province (BE).
AUTHOR CONTRIBUTIONS
Conceptualization, data curation, formal analysis,
methodology, validation, visualization, and writing-
original draft: Wencui Kang. Funding acquisition and
supervision: Hao Lin. Data curation and software: Hao
Jiang. Writing-review and editing: Selorm Yao-Say
Solomon Adade. Investigation and resources: Zhaoli Xue.
Funding acquisition and supervision: Quansheng Chen.
CONFLICT OF INTEREST
The authors declare no conflict of interest
ORCID
Hao Lin https://orcid.org/---
Quansheng Chen https://orcid.org/---
REFERENCES
Akman, F., Issaoui, N., & Kazachenko, A. S. (). Intermolecu-
lar hydrogen bond interactions in the thiourea/water complexes
(Thio-(HO)n) (n =, ..., ): X-ray, DFT, NBO, AIM, and RDG
analyses. Journal of Molecular Modeling,26,.
Alimelli, A., Pennazza, G., Santonico, M., Paolesse, R., Filippini, D.,
D’Amico, A., Lundstrom, I., & Di Natale, C. (). Fish fresh-
ness detection by a computer screen photoassisted based gas sen-
sor array. Analytica Chimica Acta,582, –.
Askim, J. R., & Suslick, K. S. (). Hand-held reader for colorimetric
sensor arrays. Analytical Chemistry,87,–.
Askim, J. R., Mahmoudi, M., & Suslick, K. S. (). Optical sensor
arrays for chemical sensing: The optoelectronic nose. Chemical
Society Reviews,42, –.
Baeckmann, V. C., Guillet, N. R., Renfer, D., Kahig, H., & Kleite,
F. (). A toolbox for the synthesis of multifunctionalized
mesoporous silica nanoparticles for biomedical applications. ACS
Omega,3,–.
A - . .. 25
Beer, P. D., Cormode, D. P., & Davis, J. J. (). Zinc
metalloporphyrin-functionalised nanoparticle anion sensors.
Chemical Communications,14,–.
Bordbar, M. M., Tashkhourian, J., & Hemmateenejad, B. (). Qual-
itative and quantitative analysis of toxic materials in adulterated
fruit pickle samples by a colorimetric sensor array. Sensors and
Actuators B: Chemical,257,–.
Brandenburg, J. G., & Grimme, S. (). Dispersion corrected
hartree-fock and density functional theory for organic crystal
structure prediction. Topics in Current Chemistry,345,–.
Brunink, J. A. J.,Di Natale, C., Bungaro, F., Davide, F. A. M., D’Amico,
A., Paolesse, R., Boschi, T., Faccio, M., & Feerri, G. (). The
application of metalloporphyrins as coating material for quartz
microbalance-based chemical sensors. Analytica Chimica Acta,
325,–.
Charoensuk, J., Thonglao, J., Wichaiyo, B., Mukdasai, K., Santa-
ladchaiyakit, Y., Srijaranai, S., & Mukdasai, S. (). A sim-
ple and sensitive colorimetric sensor for cadmium (II) detec-
tion based on self-assembled trimethyl tetradecyl ammonium bro-
mide and murexide on colloidal silica. Microchemical Journal,160,
.
Chen, Q., Hui, Z., Zhao, J., & Ouyang, Q. (). Evaluation of
chicken freshness using a low-cost colorimetric sensor array with
AdaBoost–OLDA classification algorithm. LWT - Food Science and
Technology,57, –.
Chen, Q., Li, H., Ouyang, Q., & Zhao, J. (). Identification of
spoilage bacteria using a simple colorimetric sensor array. Sensors
and Actuators B: Chemical,205,–.
Chen, Q., Liu, A., Zhao, J., Ouyang, Q., Sun, Z., & Huang, L. ().
Monitoring vinegar acetic fermentation using a colorimetric sen-
sor array. Sensors and Actuators B: Chemical,183, –.
Chen, Q., Liu, A., & Zhao, J. (). Classification of tea category
using a portable electronic nose based on an odor imaging sen-
sor array. Journal of Pharmaceutical and Biomedical Analysis,84,
–.
Colombo, V., Montoro, C., Maspero, A., Palmisano, G., Mascioc-
chi, N., Galli, S., Barea, E., & Navarro, J. A. (). Tuning
the adsorption properties of isoreticular pyrazolate-based metal-
organic frameworks through ligand modification. Journal of the
American Chemical Society,134, –.
de la Roza-Delgado, B., Soldado, A., de Faria Oliveira, A. F. G.,
Martínez-Fernández, A., & Argamentería, A. (). Assessing the
value of a portable near infrared spectroscopy sensor for predict-
ing pork meat quality traits of “asturcelta autochthonous swine
breed”. Food Analytical Methods,7(), –.
Deshmukh, S., Bandyopadhyay, R., Bhattacharyya, N., Pandey, R.A.,
& Jana, A. (). Application of electronic nose for industrial
odors and gaseous emissions measurement and monitoring—An
overview. Tal ant a,144,–.
Di Natale, C., Paolesse, R., Macagnano, A., Mantini, A., Goletti,
C., & D’Amico, A. (). Characterization and design of
porphyrins-based broad selectivity chemical sensors for electronic
nose applications. Sensors and Actuators B: Chemical,52,–
.
Domínguez-Aragón, A., Olmedo-Martínez, J. A., & Zaragoza-
Contreras, E. A. () Colorimetric sensor based on a poly(ortho-
phenylenediamine-co-aniline) copolymer for the monitoring of
tilapia (Orechromis niloticus) freshness. Sensors and Actuators B:
Chemical,259,–.
Dowlati, M., de la Guardia, M., Dowlati, M., & Mohtasebi, S. S. ().
Application of machine-vision techniques to fish-quality assess-
ment. TrAC Trends in Analytical Chemistry,40,–.
Duan, Y., Lin, H., He, P., & Chen, Q. (). Detection of volatile
marker in the wheat infected with Aspergillus flavus by porous
silica nanospheres doped Bodipy dyes. Sensors and Actuators B:
Chemical,330,.
Dumanoğulları, F. M., Tutel, Y., Küçüköz, B., Sevinç, G., Karatay, A.,
Yılmaz, H., Hayvali, M., & Elmali, A. (). Investigation of ultra-
fast energy transfer mechanism in BODIPY–Porphyrin dyad sys-
tem. Journal of Photochemistry and Photobiology A: Chemistry,373,
–.
Ezhilan, M., Nesakumar, N., Babu, K.J. Srinandan, C. S., & Rayap-
pan, J. B. B. (). An electronic nose for royal delicious apple
quality assessment—A tri-layer approach. Food Research Interna-
tional,109, –.
Feng,T.,Zhuang,H.,Ye,R.,Jin,Z.,Xu,X.,&Xie,Z.().Analysis
of volatile compounds of Mesona Blumes gum/rice extrudates via
GC–MS and electronic nose. Sensors and Actuators B: Chemical,
160,–.
Ferroudj, N., Talbot, D., Michel, A., Davidson, A., & Abramson, S.
(). Increasing the efficiency of magnetic heterogeneous Fen-
ton catalysts with a simple halogen visible lamp. Journal of Photo-
chemistry and Photobiology A: Chemistry,338,–.
Gabor, G., Chadha, S., & Walt, D. R. (). Sensitivity enhancement
of fluorescent pH indicators using pH-dependent energy transfer.
Analytica Chimica Acta,313,–.
Gu, H., Huang, X., Chen, Q., Sun, Y., Tan, C., Zhao, J., & Lv, R. ().
Prediction of the property of colorimetric sensor array based on
density functional theory. Sensors and Materials,31,–.
Gu, H., Huang, X., Yao, L., Teye, E., & Wen, Y. (). Study on
the binding ability of cobalt-porphyrin with small volatile organic
compounds based on density functional theory. Analytical Meth-
ods,6, .
Guan, B., Xue, Z., Chen, Q., Lin, H., & Zhao, J. (). Preparation
of zinc porphyrin nanoparticles and application in monitoring the
ethanol content during the solid-state fermentation of Zhenjiang
Aromatic vinegar. Microchemical Journal,153,.
Guan, B., Zhao, J., Jin, H., & Lin, H. (). Determination of rice stor-
age time with colorimetric sensor array. Food Analytical Methods,
10, –.
Guan, B., Zhao, J., Cai, M., Lin, H., Yao, L., & Sun, L. (). Analy-
sis of volatile organic compounds from Chinese vinegar substrate
during solid-state fermentation using a colorimetric sensor array.
Analytical Methods,6, –.
Guan, B., Zhao, J., Lin, H., & Zou, X. (). Characterization of
volatile organic compounds of vinegars with novel electronic
nose system combined with multivariate analysis. Food Analytical
Methods,7,–.
Haddi, Z., El Barbri, N., Tahri, K., Bougrini, M., El Bari, N., Llobet, E.,
& Bouchikhi, B. (). Instrumental assessment of red meat ori-
gins and their storage time using electronic sensing systems. Ana-
lytical Methods,7(), –.
Hao, Y., Chen, W., Wang, L., Zhou, B., Zang, Q., Chen, S., & Liu, Y.
N. (). A naphthalimide-based azo colorimetric and ratiometric
probe: Synthesis and its application in rapid detection of cyanide
anions. Analytical Methods,6,.
Hu, Y., Xue, Z., He, H., Ai, R., Liu, X., & Lu, X. ().
Photoelectrochemical sensing for hydroquinone based on
26 A - . ..
porphyrin-functionalized Au nanoparticles on graphene. Biosen-
sors & Bioelectronics,47,–.
Huang, X., Xin, J., & Zhao, J. (). A novel technique for rapid eval-
uation of fish freshness using colorimetric sensor array. Journal of
Food Engineering,105,–.
Huang,X.,Zou,X.,Shi,J.,Guo,Y.,Zhao,J.,Zhang,J.,&Lin,H.().
Determination of pork spoilage by colorimetric gas sensor array
based on natural pigments. Food Chemistry,145,–.
Huang, X., Zou, X., Zhao, J., Shi, J., Li, Z., & Shen, T. (). Mon-
itoring the biogenic amines in Chinese traditional salted pork in
jelly (Yao-meat) by colorimetric sensor array based on nine natu-
ral pigments. International Journal of Food Science Technology,50,
–.
Huang, X., Zou, X., Zhao, J., Shi, J., Li, Z., & Zhao, J. (). Colori-
metric sensor arrays based on chemo-responsive dyes for food odor
visualization. Trends in Food Science & Technology,81, –.
Huo, D., Wu, Y., Yang, M., Fa, H., Luo, X., & Hou, C. (). Discrim-
ination of Chinese green tea according to varieties and grade lev-
els using artificial nose and tongue based on colorimetric sensor
arrays. Food Chemistry,145,–.
James, T., Schillaci, C., & Lipani, A. (). Convolutional neural net-
works for water segmentation using sentinel- red, green, blue
(RGB) composites and derived spectral indices. International Jour-
nal of Remote Sensing,42 (), –.
Jiang, H., Xu, W., & Chen, Q. (). Evaluating aroma quality of
black tea by an olfactory visualization system: Selection of feature
sensor using particle swarm optimization. Food Research Interna-
tional,126, .
Jo, B., & Jung, S. (). Quadrant-based minimum bounding
rectangle-tree indexing method for similarity queries over big spa-
tial data in HBase. Sensors,18(), .
Khulal, U., Zhao, J., Hu, W., & Chen, Q. (). Comparison of dif-
ferent chemometric methods in quantifying total volatile basic-
nitrogen (TVB-N) content in chicken meat using a fabricated col-
orimetric sensor array. RSC Advances,6,–.
Kuntz, I. D., Gasparro, F. P., Johnston, M. D., & Taylor, R. P. ().
Molecular interactions and the Benesi-Hildebrand equation. Jour-
nal of the American Chemical Society,90, –.
Kutsanedzie, F. Y. H., Hao, L., Yan, S., Ouyang, Q., & Chen, Q. ().
Near infrared chemo-responsive dye intermediaries spectra-based
in-situ quantification of volatile organic compounds. Sensors and
Actuators B: Chemical,254,–.
Leon, K., Mery, D., Pedreschi, F., & Leon, J. (). Color measure-
ment in La*b* units from RGB digital images. Food Research Inter-
national,39, –.
Li, R., Bu, T., Zhao, Y., Sun, X., Wang, Q., Tian, Y., Bai, F., &
Wang, L. (). Polydopamine coated zirconium metal-organic
frameworks-based immunochromatographic assay for highly sen-
sitive detection of deoxynivalenol. Analytica Chimica Acta,1131,
-.
Liao, S., Wu, S., & Chao, C. () Preparation of black poly (methyl
methacrylate-ethyleneglycol dimethacrylate) microspheres using
metallic complex dyes. Materials Sciences and Applications,11,
–.
Li, D., Huang, J., & Kaner, R. B. (). Polyaniline nanofibers: A
unique polymer nanostructure for versatile applications. Accounts
of Chemical Research,42,–.
Li, H., Chen, Q., Zhao, J., & Ouyang, Q. (). Non-destructive eval-
uation of pork freshness using a portable electronic nose (E-nose)
based on a colorimetric sensor array. Analytical Methods,6(),
–.
Li, H., Kutsanedzie, F., Zhao, J., & Chen, Q. (). Quantifying total
viable count in pork meat using combined hyperspectral imag-
ing and artificial olfaction techniques. Food Analytical Methods,
9,–.
Li, L., Xie, S., Ning, J., Chen, Q., & Zhang,Z. (). Evaluating green
tea quality based on multisensor data fusion combining hyper-
spectral imaging and olfactory visualization systems. Journal of the
Science of Food and Agriculture,99(), –.
Li, R., Bu, T., Zhao, Y., Sun, X., Wang, Q., Tian, Y., Bai, F., &
Wang, L. (). Polydopamine coated zirconium metal-organic
frameworks-based immunochromatographic assay for highly sen-
sitive detection of deoxynivalenol. Analytica Chimica Acta,1131,
–.
Li, X., Xu, W., Wang, X., Jia, H., Zhao, B., Li, B., & Ozaki, Y. ().
Ultraviolet-visible and surface-enhanced Raman scattering spec-
troscopy studies on self-assembled films of ruthenium phthalocya-
nine on organic monolayer-modified silver substrates. Thin Solid
Films,457,–.
Lin, H., Duan, Y., Man, Z., Zareef, M., Wang, Z., & Chen, Q. ().
Quantitation of volatile aldehydes using chemoselective response
dyes combined with multivariable data analysis. Food Chemistry,
353,.
Lin, H., Duan, Y., & Yan, S. (). Quantitative analysis of volatile
organic compound using novel chemoselective response dye based
on Vis-NIRS coupled Si-PLS. Microchemical Journal,145, –
.
Lin, H., Jiang, H., He, P., Haruna, S. A., Chen, Q., Xue, Z., Chan, C.,
& Ali, S. (). Non-destructive detection of heavy metals in veg-
etable oil based on nano-chemoselective response dye combined
with near-infrared spectroscopy. Sensors and Actuators B: Chemi-
cal,335,.
Lin, H., Kang, W., Jin, H., & Man, Z., & Chen, Q. (). Discrimina-
tion of Chinese Baijiu grades based on colorimetric sensor arrays.
Food Science and Biotechnology,39,–.
Lin, H., Kang, W., Kutsanedzie, F. Y. H., & Chen, Q. (). A
novel nanoscaled chemo dye–based sensor for the identification
of volatile organic compounds during the mildewing process of
stored wheat. Food Analytical Methods,12,–.
Lin, H., Man, Z., Guan, B., Chen, Q., Jin, H., & Xue, Z. ().
In situ quantification of volatile ethanol in complex components
based on colorimetric sensor array. Analytical Methods,40,–
.
Lin, H., Man, Z. X., Kang, W. C., Guan, B. B., Chen, Q. S., & Xue,
Z. L. (). A novel colorimetric sensor array based on boron-
dipyrromethene dyes for monitoring the storage time of rice. Food
Chemistry,268, –.
Lin, H., Wang, Z., Ahmad, W., Man, Z., & Duan, Y. (). Identifica-
tion of rice storage time based on colorimetric sensor array com-
bined hyperspectral imaging technology. Journal of Stored Prod-
ucts Research,85,.
Lin, H., Yan, S., Song, B., Wang, Z., & Sun, L. (). Discrimina-
tion of aged rice using colorimetric sensor array combined with
volatile organic compounds. Journal of Food Process Engineering,
42,e.
Liu, Q., Zhao, N., Zhou, D., Sun, Y., Sun, K., Pan, L., & Tu, K. ().
Discrimination and growth tracking of fungi contamination in
peaches using electronic nose. Food Chemistry,262,–.
A - . .. 27
Liu, Y., Goebl, J., & Yin, Y. (). Templated synthesis of nanostruc-
tured materials. Chemical Society Reviews,42, –.
Liu, Z., Zhang, Y., Xie, Y., Sun, Y., Bi, K., Cui, Z., Zhao, L., & Fan,
W. (). An aptamer-based colorimetric sensor for streptomycin
and its application in food inspection. Chemical Research in Chi-
nese Universities,33,–.
Long, J., Xu, J., Yang, Y., Wen, J., & Jia, C. (). A colorimetric
array of metalloporphyrin derivatives for the detection of volatile
organic compounds. Materials Science and Engineering B,176,
–.
Lv, Y. Y., Wu, J., & Xu, Z.K. (). Colorimetric and fluorescent
sensor constructing from the nanofibrous membrane of porphyri-
nated polyimide for the detection of hydrogen chloride gas. Sen-
sors and Actuators B: Chemical,148,–.
Magnaghi, L. R., Alberti, G., Capone, F., Zanoni, C., Mannucci, B.,
Quadrelli, P., & Biesuz, R. (). Development of a dye–based
device to assess the poultry meat spoilage. Part II: Array on act.
Journal of Agricultural and Food Chemistry,68(), –.
Majchrzak, T., Wojnowski, W., Dymerski, T., Gebicki, J., & Namies-
nik, J. (). Electronic noses in classification and quality control
of edible oils: A review. Food Chemistry,246,–.
Mamardashvili, G. M., Lazovskiy, D. A., Khodov, I. A, Efimov, A.
E., & Mamardashvili, N. Z. (). New polyporphyrin arrays with
controlled fluorescence obtained by diaxial Sn(IV)-Porphyrin Phe-
nolates Chelation with Cu+Cation. Polymers,13(), .
Moreira, N., Lopes, P., Ferreira, H., Cabral, M., & Pinho, P.G.D. ().
Influence of packaging and aging on the red wine volatile compo-
sition and sensory attributes. Food Packaging Shelf,8,–.
Motellier, S., & Toulhoat, P. (). Modified acid-base behaviour of
resin-bound pH indicators. Analytica Chimica Acta,271, –.
Olafsdottir, G., Nesvadba, P., Di Natale, C., Careche, M., Oehlen-
schläger, J., Tryggvadóttir, S. a. V., Schubring, R., Kroeger, M.,
Heia, K., Esaiassen, M., Macagnano, A., & Jørgensen, B. M. ().
Multisensor for fish quality determination. Trends in Food Science
& Technology,15(), –.
Ouyang, Q., Zhao, J., Chen, Q., & Lin, H. (). Classification of
rice wine according to different marked ages using a novel arti-
ficial olfactory technique based on colorimetric sensor array. Food
Chemistry,138,–.
Ouyang, Q., Zhao, J., & Chen, Q. (). Classification of rice
wine according to different marked ages using a portable multi-
electrode electronic tongue coupled with multivariate analysis.
Food Research International,51,–.
Palmer, L. C., & Stupp, S. I. (). Molecular self-assembly into one-
dimensional nanostructures. Accounts of Chemical Research,41,
–.
Paolesse, R., Natale, C. D., Burgio, M., Martinelli, E., Mazzone, E.,
Palleschi, G., & D’ Amico, A. (). Porphyrin-based array of
cross-selective electrodes for anhlysis of liquid samples. Sensors
and Actuators B: Chemical,95, –.
Phelan, J. P. (). First moment. Mindfulness,2,–.
Pico, J., Bernal, J., & Gómez, M. (). Wheat bread aroma com-
pounds in crumb and crust: A review. Food Research International,
75, –.
Qian, S., & Lin, H. (). Colorimetric sensor array for detection
and identification of organophosphorus and carbamate pesticides.
Analytical Chemistry,87, –.
Qin, H., Huo, D., Zhang, L., Yang, L., Zhang, S., Yang, M., Shen, C.,
& Hou, C. (). Colorimetric artificial nose for identification of
Chinese liquor with different geographic origins. Food Research
International,45, –.
Radi Ciptohadijoyo, S., Litananda, W. S., Rivai, M., & Purnomo, M.
H. (). Electronic nose based on partition column integrated
with gas sensor for fruit identification and classification. Comput-
ers and Electronics in Agriculture,121,–.
Rakow, N. A., Sen, A., Janzen, M. C., Ponder, J. B., & Suslick, K. S.
(). Molecular recognition and discrimination of amines with
a colorimetric array. Angewandte Chemie,117, –.
Rakow, N. A., & Suslick, K. S. (). A colorimetric sensor array for
odour visualization. Nature,406(), –.
Ruenraroengsak, P., & Tetley, T. D. (). Differential bioreactivity of
neutral, cationic and anionic polystyrene nanoparticles with cells
from the human alveolar compartment: Robust response of alve-
olar type epithelial cells. Particle and Fibre Toxicology,12,.
Salinas, Y., Ros-Lis, J. V., Vivancos, J.-L., Martínez-Máñez, R., Mar-
cos, M. D., Aucejo, S., Herranz, N., Lorente, I., & Garcia, E.
(). A novel colorimetric sensor array for monitoring freshpork
sausages spoilage. Food Control,35, –.
Sanaeifar, A., Mohtasebi, S. S., Ghasemi-Varnamkhasti, M., &
Ahmadi, H. (). Application of MOS based electronic nose for
the prediction of banana quality properties. Measurement,82,–
.
Scholl, J. A., Garcia-Etxarri, A., Koh, A. L., & Dionne, J. A.
(). Observation of quantum tunneling between two plasmonic
nanoparticles. Nano Letters,13,–.
Shimizu, T., Masuda, M., & Minamikawa, H. (). Supramolecular
nanotube architectures based on amphiphilic molecules. Journal
of Chemical Reviews,105, –.
Shi, Z., Han, X., Hu, W., Bai, H., Peng, B., Ji, L., Fan, Q., Li, L.,
& Huang, W. (). Bioapplications of small molecule Aza-
BODIPY: From rational structural design to in vivo investigations.
Chemical Society Reviews,49,–.
Skrbic, B., & Durisic-Mladenovic, N. (). Principal component
analysis for soil contamination with organochlorine compounds.
Chemosphere,68, –.
Song, G., Sun, R., Du, J., Chen, M., & Tian, Y. (). A highly selec-
tive, colorimetric, and environment-sensitive optical potassium
ion sensor. Chemical Communications,53(), –.
Sousa, A. A., Hassan, S. A., Knittel, L. L., Balbo, A., Aronova, M.
A., Brown, P. H., Schuck, P., & Leapman, R. D. (). Biointerac-
tions of ultrasmall glutathione-coated gold nanoparticles: Effect of
small size variations. Nanoscale,8, –.
Stone, R. D., & Vaughn, J. L. (). The effect of pH indicators chlor-
phenol red and bromocresol purple on the in vitro production of
lepidopteran nuclear polyhedrosis viruses. Journal of Invertebrate
Pathology,48,–.
Su, Z., Shaw, W. L., Miao, Y.-R., You, S., Dlott, D. D., & Suslick, K. S.
(). Shock wave chemistry in a metal-organic framework. Jour-
nal of the American Chemical Society,139, –.
Sung, J., Kim, B. K., Kim, B. S., & Kim, Y. (). Mass spectrometry-
based electric nose system for assessing rice quality during storage
at different temperatures. Journal of Stored Products Research,59,
–.
Suslick, B. A., Feng, L., & Suslick, K. S. (). Discrimination of com-
plex mixtures by a colorimetric sensor array: Coffee aromas. Ana-
lytical Chemistry,82,–.
Suslick, K. S., Neal, A., & Rakow, A. S. (). Colorimetric sensor
arrays for molecular recognition. Tetra hed ron ,60, –.
28 A - . ..
Swamy, P. C. A., Mukherjee, S., & Thilagar, P. (). Multichannel-
emissive V-shaped boryl-BODIPY dyads: Synthesis, structure, and
remarkably diverse response toward fluoride. Inorganic Chem-
istry,53,–.
Valdez, M., Gupta, S. K., Lozano, K., & Mao, Y. (). Force-
Spun polydiacetylene nanofibers as colorimetric sensor for food
spoilage detection. Sensors and Actuators B: Chemical,297,
.
Wang, R., Ruan, G., Sun, Y., Zhao, D., Yu, H., Zhang, C.-W., Li,
L., & Liu, J. (). A full-wavelength coverage colorimetric sen-
sor depending on polymer-carbon nanodots from blue to red for
visual detection of nitrite via smartphone. Dyes and Pigments,191,
.
Wang, Y., Yang, Q., Zhao, M. J., & Su, B. (). Silica-nanochannel-
based interferometric sensor for selective detection of polar and
aromatic volatile organic compounds. Analytical Chemistry,90,
–.
Wong, C. P., & Horrocks, W. D. W. (). New metalloporphyrins.
Thorium and yttrium complexes of tetraphenylporphin. Tetr ahe -
dron Letters,16,–.
Wu, L., Pu, H., & Sun, D. W. (). Novel techniques for evaluat-
ing freshness quality attributes of fish: A review of recent develop-
ments. Trends in Food Science & Technology,83,–.
Xu, S., Hartvickson, S., & Zhao, J. X. (). Increasing surface area of
silica nanoparticles with a rough surface. ACS Applied Materials &
Interfaces,3,–.
Xu, Y., Kutsanedzie, F. Y. H., Sun, H., Wang, M., Chen, Q., Guo, Z.,
& Wu, J. (). Rapid pseudomonas species identification from
chicken by integrating colorimetric sensors with near-infrared
spectroscopy. Food Analytical Methods,11(), –.
Ye, N. S. (). A minireview of analytical methods for the geograph-
ical origin analysis of teas (Camellia sinensis). Critical Reviews in
Food Science and Nutrition,52,–.
Yu,D.G.,An,J.H.,Bae,J.Y.,Ahn,S.D.,Kang,S.Y.,&Suh,K.S.
(). Negatively charged uitrafine black particles of P (MMA-
co-EGDMA) by dispersion polymerization for electrophoretic dis-
plays. Macromolecules,38,–.
Zaragoza, P., Fuentes, A., Ruiz-Rico, M., Vivancos, J. L., Fernandez-
Segovia, I., Ros-Lis, J. V., Barat, J. M., & Martinez-Manez, R. ().
Development of a colorimetric sensor array for squid spoilage
assessment. Food Chemistry,175, -.
Zaragoza, P., Fuentes, A., Fernandez-Segovia, I., Vivancos, J. L., Rizo,
A., Ros-Lis, J. V., Barat, J. M., & Martinez-Manez, R. (). Evalua-
tion of sea bream (Sparus aurata) shelf life using an optoelectronic
nose. Food Chemistry,138(–), –.
Zhu, J., Agyekum, A. A, Kutsanedzie, F. Y. H., Li, H., Chen, Q.,
Ouyang, Q., & Jiang, H. (). Qualitative and quantitative anal-
ysis of chlorpyrifos residues in tea by surface-enhanced Raman
spectroscopy (SERS) combined with chemometric models. LWT
- Food Science and Technology,97, –.
Zhang, X. B., Guo, C. C., Li, Z. Z., Shen, G. L., & Yu, R. Q. ().
An optical fiber chemical sensor for mercury lons based on a por-
phyrin dimer. Analytical Chemistry,74, –.
Zhou, H., Baldini, L., Hong, J. Wilson, A. J., & Hamilton, A. D. ().
Pattern recognition of proteins based on an array of functionalized
porphyrins. Journal of the American Chemical Society,128,–
.
Zhou, X., Yang, H., Wang, C., Mao, X., Wang, Y., Yang, Y., & Liu,
G. (). Visible light induced photocatalytic degradation of rho-
damine B on one-dimensional lron oxide particles. JournalofPhys-
ical Chemistry,114, –.
Zhang, C., Suslick, K. S. (). A colorimetric sensor array for organ-
ics in water. Journal of the American Chemical Society,127,–
.
Zhang, K., Huang, S., Lu, H., Li, F., & Shao, L. (). Experimental
study of the light source characteristics for the NH concentration
detection. Optik,209, .
How to cite this article: Kang W, Lin H, Jiang H,
Adade SY-SS, Xue Z, Chen Q. Advanced
applications of chemo-responsive dyes based odor
imaging technology for fast sensing food quality
and safety: A review. Compr Rev Food Sci Food Saf.
;–. https://doi.org/./-.
A preview of this full-text is provided by Wiley.
Content available from Comprehensive Reviews in Food Science and Food Safety
This content is subject to copyright. Terms and conditions apply.
