Engineering Sensor Arrays Using Aggregation-Induced Emission Luminogens for Pathogen Identification

Article abstract of DOI:10.1002/adfm.201805986

Lacking rapid and reliable pathogen diagnostic platforms, inadequate or delayed antimicrobial therapy could be made, which greatly threatens human life and accelerates the emergence of antibiotic-resistant pathogens. In this contribution, a series of simple and reliable sensor arrays based on tetraphenylethylene (TPE) derivatives are successfully developed for detection and discrimination of pathogens. Each sensor array consists of three TPE-based aggregation-induced emission luminogens (AIEgens) that bear cationic ammonium group and different hydrophobic substitutions, providing tunable logP (n-octanol/water partition coefficient) values to enable the different multivalent interactions with pathogens. On the basis of the distinctive fluorescence response produced by the diverse interaction of AIEgens with pathogens, these sensor arrays can identify different kinds of pathogens, even normal and drug-resistant bacteria, with nearly 100% accuracy. Furthermore, blends of pathogens can also be identified accurately. The sensor arrays exhibit rapid response (about 0.5 h), high-throughput, and easy-to-operate without washing steps.

Full text of DOI:10.1002/adfm.201805986

FULL PAPER
Biosensors
www.afm-journal.de
Engineering Sensor Arrays Using Aggregation-Induced
Emission Luminogens for Pathogen Identification
Chengcheng Zhou, Wenhan Xu, Pengbo Zhang, Meijuan Jiang, Yuncong Chen,
Ryan T. K. Kwok, Michelle M. S. Lee, Guogang Shan, Ruilian Qi, Xin Zhou,
Jacky W. Y. Lam, Shu Wang,* and Ben Zhong Tang*
date, the most widely adopted methods for
Lacking rapid and reliable pathogen diagnostic platforms, inadequate or
delayed antimicrobial therapy could be made, which greatly threatens
human life and accelerates the emergence of antibiotic-resistant pathogens.
In this contribution, a series of simple and reliable sensor arrays based on
tetraphenylethylene (TPE) derivatives are successfully developed for detection
and discrimination of pathogens. Each sensor array consists of three TPE-
based aggregation-induced emission luminogens (AIEgens) that bear cationic
ammonium group and different hydrophobic substitutions, providing tunable
logP (n-octanol/water partition coefficient) values to enable the different
multivalent interactions with pathogens. On the basis of the distinctive
fluorescence response produced by the diverse interaction of AIEgens with
pathogens, these sensor arrays can identify different kinds of pathogens, even
normal and drug-resistant bacteria, with nearly 100% accuracy. Furthermore,
blends of pathogens can also be identified accurately. The sensor arrays
exhibit rapid response (about 0.5 h), high-throughput, and easy-to-operate
without washing steps.
identifying microorganisms include plating
and culturing, observing the morphological
structure, and examining gene and immu-
nological characteristics.[3] However, the
application scopes of these techniques are
limited by their inherent problems. For
example, the plating and culturing method
is time-consuming and usually takes sev-
eral days.[4] Analyzing the morphology
by microscopy hardly discriminate the
pathogens with similar size and shape.[5]
Examining gene and immunological char-
acteristics of pathogens requires technologi-
cally advanced systems such as polymerase
chain reaction (PCR), gene microarray, and
target-specific immunoassays,[3a] which are
complex and require sophisticated instru-
mentations. Moreover, due to multiple
operating steps, the occurrence of false-pos-
itive results is unavoidable.[6] Furthermore,
for some advanced technologies applied in
hospitals and other authorized organizations, such as automated
biochemical instrumentation and matrix-assisted laser desorption
ionization time-of-flight mass spectrometry (MALDI-TOF-MS),
the accuracy rate of microorganism identification is only 90–95%
and several hours of operation are needed.[7] Without access
to timely and reliable pathogen information, point-of-care
treatment decisions were made by the prescription of a subop-
timal antibiotic.[8] Under selective pressure of the suboptimal
1. Introduction
Microorganism identification is very important because
numerous species are greatly associated with human diseases
and death.[1] Currently, more than 300 million severe infection
cases and over 5 million death stem from pathogens every year
in the world.[2] To ensure effective treatments, rapid and reliable
diagnosis of pathogen infection is the first critical step.[1b] Up to
Dr. C. Zhou, Dr. M. Jiang, Dr. Y. Chen, Dr. R. T. K. Kwok, Dr. G. Shan,
Dr. J. W. Y. Lam, Prof. B. Z. Tang
HKUST Shenzhen Research Institute
No. 9 Yuexing 1st RD, South Area, Hi-tech Park, Nanshan,
Shenzhen 518057, China
Dr. P. Zhang, Dr. R. Qi, X. Zhou, Prof. S. Wang
Key Laboratory of Organic Solids
Beijing National Laboratory for Molecular Sciences (BNLMS)
Institute of Chemistry
Chinese Academy of Sciences
E-mail: tangbenz@ust.hk
Beijing 100190, P. R. China
E-mail: wangshu@iccas.ac.cn
Dr. C. Zhou, W. Xu, Dr. M. Jiang, Dr. Y. Chen, Dr. R. T. K. Kwok,
M. M. S. Lee, Dr. G. Shan, Dr. J. W. Y. Lam, Prof. B. Z. Tang
Department of Chemistry
Hong Kong Branch of Chinese National Engineering Research
Center for Tissue Restoration and Reconstruction
Department of Chemical and Biological Engineering
The Hong Kong University of Science and Technology
Clear Water Bay, Kowloon, Hong Kong 999077, China
Prof. B. Z. Tang
State Key Laboratory of Luminescent Materials and Devices
Center for Aggregation-Induced Emission
(Guangzhou International Campus)
South China University of Technology
Guangzhou 510640, China
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/adfm.201805986.
DOI: 10.1002/adfm.201805986
©
Adv. Funct. Mater. 2018, 1805986
1805986 (1 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Scheme 1. Structure of TPE-ARs and schematic illustration of a sensor array composed of three TPE-ARs to achieve pathogen identification. P1–P7
represent seven kinds of pathogens.
prescription, in which large quantities of antimicrobial agents
are present or inadequate doses of drugs are taken, many strains
undergo mutations and consequently acquire antibiotic resist-
ance.[9] It has been predicted that, antibiotic-resistant infections
could kill up to 10 million people worldwide by 2050.[10] Thus, a
fast and reliable method to discern various pathogens is urgently
required.
based on the different surface electric potentials of bac-
teria, the sensor arrays for bacteria detection have been built
using a series of AIEgens with various electric charges.[7c,20]
However, the AIEgen-based sensor array showed only about
93.75% detection accuracy for bacteria samples by analyzing
the collective fluorescence signals of bacteria with the statis-
tical methods.[7c] Therefore, it still remains a great challenge
to rationally design the AIEgen sensor array for enlarging the
difference between fluorescence response of various patho-
gens and thus achieve high detection accuracy. Meanwhile, it
is also significant to make the sensor arrays simple as well as
improving the detection accuracy.[21]
In this work, we introduced multivalent interactions between
AIEgens and pathogens to augment the diversity of the fluores-
cence response of pathogens, taking advantage of hydrophobic
groups that are prevalent on microbial exteriors in addition to
the negatively charged residues. For this purpose, a series of
AIE-active tetraphenylethylene (TPE) derivatives, TPE-ARs, were
designed and synthesized (Scheme 1). Structurally, they bear
one positively charged ammonium group and different hydro-
phobic groups with finely controlled hydrophobicity of calcu-
lated logP (ClogP) values from 3.426 to 6.071 (logP is n-octanol/
water partition coefficient, whose values were estimated using
ChemBioDraw 14.0), which were engineered to tune the electro-
static and hydrophobic interactions between AIEgens and patho-
gens. Additionally, the alkoxy chain was introduced to increase
the water-solubility and flexibility of TPE-ARs. Using these AIE-
gens, we successfully created multiple competent sensor arrays
for rapid and reliable detection and discrimination of different
pathogens, even between normal and drug-resistant pathogens,
with the aid of linear discriminant analysis (LDA).
As an alternative, fluorescent probe is a promising tool for
identifying pathogens, owing to its rich advantages, such as
fast response, superior sensitivity, simplicity, and so on.[11]
Some biochemical sensors based on fluorescent responses
have been developed for pathogens identification.[3b,5,12]
However, conventional fluorescent molecules generally
suffer from the notorious aggregation-caused quenching
(ACQ) effect, where their emissions are often weakened or
quenched at higher concentration or in aggregated state.[13]
This has confined their working concentration to a very low
level, and thus limits the labeling degrees of probes to ana-
lytes, resulting in compromised sensitivity.[14] Moreover,
the ACQ effect of conventional fluorophores often forces
them to work in a fluorescence “turn-off” mode. Inevitably,
light emissions of fluorophores are usually susceptible to
some external factors such as water and air, further compro-
mising sensitivity and accuracy of identification.[14] To over-
come these issues, tailor-made quenchers were introduced
to weaken the emission of fluorophore, and then the system
with weak emission was used to detect bioanalysts in turn-on
fashion.[12b,e,15] Although effective, this approach complicates
the sensor designs.
Diametrically opposed to the conventional ACQ fluoro-
phores, aggregation-induced emission luminogens (AIEgens)
are nonemissive or weakly emissive when molecularly dis-
solved but highly emissive when aggregated.[16] This feature
endows AIEgens with ability to work in a light-up/turn-on
fashion.[17] Moreover, AIE-based probes have low background
and thus without need of repeated washing procedures.[7c]
These merits of AIEgens well meet the requirement of an
ideal fluorescence sensor,[12e,18] which will greatly enhance
the sensitivity and reliability of detection.[19] Very recently,
2. Results and Discussion
2.1. Synthesis and Characterization of TPE-ARs
Seven TPE-ARs (TPE-AMe, TPE-AEt, TPE-APrA, TPE-ABu,
TPE-ACH, TPE-ABn, and TPE-AHex) were prepared by facile
©
Adv. Funct. Mater. 2018, 1805986
1805986 (2 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 1. a) Normalized absorption spectra of TPE-ARs (20 × 10−6 m) in DMSO solution and normalized PL spectra of TPE-ARs (200 × 10−6 m) in
organic solvent/H2O mixture with 96% water fraction, λex: 340 nm. Organic solvent/H2O mixture: MeOH/H2O mixture for TPE-AMe, TPE-APrA and
TPE-ABn, ACN/H2O mixture for TPE-AEt, TPE-ABu, and TPE-ACH, and DMSO/H2O mixture for TPE-AHex. b) PL spectra of TPE-AHex (200 × 10−6 m)
in DMSO/H2O mixture with different water fraction (fw), λex: 340 nm. c) Plots of PL intensity of TPE-ARs versus the dye concentration in PBS solution.
synthetic routes in reasonable yields (Scheme S1, Supporting
Information), and their chemical structures were fully charac-
terized by 1H NMR, 13C NMR, and high-resolution mass spectra
(HRMS) (Figures S1–S7, Supporting Information). Their photo-
physical properties were studied and summarized in Table S1
(Supporting Information). As shown in Figure 1a, one main
absorption peak was observed at 313 or 314 nm for these AIE-
gens in dimethyl sulfoxide (DMSO) solution. Then, their AIE
properties were studied by changing the water content in the
mixtures of water-miscible organic solvent and water. Taking
TPE-AHex as an example (Figure 1b), TPE-AHex emits weakly
in DMSO/H2O mixtures with water fractions from 0 to 90 vol%.
Further increasing the water content to 96%, a strong emission
peak at 476 nm was observed, and the difference can be easily
distinguished by naked eyes (inset in Figure 1b). Similarly,
at higher water fractions, the other six TPE-ARs also showed
obviously enhanced fluorescence emission (Figures S8 and S9,
Supporting Information), exhibiting a typical AIE phenom-
enon due to the formation of aggregates, as confirmed by
dynamic light scattering (DLS) results (Figure S10, Supporting
Information). The maximum emissions of TPE-ARs aggre-
gates all locate at around 470 nm (Figure 1a). Thus, function-
ally decorating TPE in our strategy had little effect on the
absorption and emission profiles, which are expected to greatly
facilitate the pathogen detection simply based on the quantita-
tive analysis of fluorescence intensity. To optimize the working
concentration of TPE-ARs, their critical aggregation con-
centrations (CACs) in phosphate buffered saline (PBS) were
determined by following the fluorescence changes of TPE-
ARs upon aggregation. From the plots of emission intensity
of TPE-ARs against their concentrations (Figure 1c), the CAC
values of TPE-AMe, TPE-AEt, TPE-APrA, TPE-ABu, TPE-ACH,
TPE-ABn, and TPE-AHex in PBS solution can be estimated to
be 42.5, 79.2, 47.4, 44.4, 80.2, 76.5, and 65.3 × 10−6 m, respec-
tively, which are in a good agreement with those from the DLS
results (Figure S11, Supporting Information). Above CACs,
they present diverse aggregate morphologies (Figure S12, Sup-
porting Information). To achieve high detection sensitivity,
20 × 10−6 m of AIEgen PBS solutions with weak fluorescence
background were chosen for the following pathogen detection
experiments.
2.2. Diverse Fluorescence Response of TPE-ARs with Pathogens
2.2.1. Pathogen Staining and Imaging with TPE-ARs
Seven microorganisms were used as targeted pathogens for dem-
onstration. Among them, S. aureus, penicillin-resistant S. aureus
(abbreviated as S. aureusR) and E. faecalis are Gram-positive bac-
teria, E. coli, ampicillin-resistant E. coli (abbreviated as E. coliR)
and P. aeruginosa are Gram-negative bacteria, and C. albicans
is a fungus. The addition of these pathogens to seven TPE-ARs
solutions gave rise to obvious changes of fluorescence intensity
but only little change on the maximum emission wavelength,
which greatly facilitates pathogen detection. Taking TPE-APrA
for example (Figure 2a), after incubating with Gram-positive
S. aureusR, Gram-negative E. coli and fungus C. albicans, the fluo-
rescence intensity of TPE-APrA was turned on with different
extents, following the order of C. albicans > S. aureusR > E. coli. This
reflects the different binding degrees of TPE-APrA to these three
pathogens. Meanwhile, the zeta potentials of pathogens did not
change distinctly after adding TPE-ARs (Figure 2b), suggesting
that these AIEgens inserted into the pathogen membranes or
entered the pathogens. This was further confirmed by their con-
focal microscopy images (Figure 2c). From the confocal images
of three AIEgens TPE-APrA, TPE-ACH, and TPE-AHex with
seven pathogens, it can be found that these AIEgens not only can
efficiently stain the pathogens but also show diverse response
signals. Diverse fluorescence responses presented when one
strain was incubated with different AIEgens and different strains
incubated with the same AIEgen, which is a prerequisite for
creating the sensor array to identify pathogens.
2.2.2. Grouping of TPE-ARs Based on Fluorescence Response
To test the ability of seven TPE-ARs to identify pathogens,
we used the microplate reader, an easy-to-perform and high-
throughput technique, to record the fluorescence intensities
of TPE-ARs at 470 nm with the excitation of 340 nm after
adding each pathogen. The fluorescence of TPE-ARs alone in
PBS was measured as control. The relative fluorescence inten-
sities of TPE-ARs before and after incubation with pathogens,
©
Adv. Funct. Mater. 2018, 1805986
1805986 (3 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 2. a) Fluorescence spectra of 20 × 10−6 m TPE-APrA in PBS and microbes suspensions. b) Zeta potential results of seven pathogens in the
absence and presence of 20 × 10−6 m TPE-ARs. c) CLSM images of seven pathogens after incubation with 20 × 10−6 m AIEgens (TPE-APrA, TPE-ACH,
and TPE-AHex) for 15 min. λex = 405 nm and λem = 430–500 nm.
(I − I0)/I0, were used to characterize the fluorescence response
for each AIEgen against the seven pathogens. As shown in
Figure 3, the seven TPE-ARs with ClogP values from 3.426 to
6.071 showed diverse fluorescence response signals against the
seven pathogens, which should be attributed to the different
multivalent interactions between TPE-ARs and pathogens. This
explicitly demonstrates the possibility of these AIEgens to iden-
tify pathogens.
According to the fluorescence response patterns, we divided
the seven TPE-ARs into three groups with the variation of ClogP
values (Figure 3). The color depth of blue circle was used to
describe the relative fluorescence intensity. With the increase of
the blue depth, the relative fluorescence intensity increases. For
group A, consisting of TPE-AMe, TPE-AEt, and TPE-APrA with
3 < ClogP < 5, the relative fluorescence intensity of TPE-ARs
significantly increases after adding Gram-positive bacteria or
fungi, as compared to the Gram-negative bacteria. This was in
a good agreement with the observation under the fluorescence
microscope as shown in Figure 2c. With ClogP value ranging
from 5 to 6, group B, including TPE-ABu, TPE-ACH, and TPE-
ABn, shows the similar change in the fluorescence intensity
between the tested pathogens. Group B was further divided
into group B1 (TPE-ACH with larger fluorescence change) and
group B2 (TPE-ABu and TPE-ABn with smaller fluorescence
change). TPE-AHex with ClogP > 6 is classified as Group C,
which is diametrically opposed to the AIEgens in group A. The
relative fluorescence intensity of TPE-ARs in this group shows
more increase after adding Gram-negative bacteria than Gram-
positive bacteria and fungi. On the basis of these fluorescence
responses, we can deduce that with increasing ClogP value of
TPE-ARs, the affinity of TPE-ARs toward Gram-positive bacteria
and fungi was gradually weakened and changed into the higher
affinity to Gram-negative bacteria, suggesting that the hydro-
phobic interaction plays more important role in Gram-negative
bacteria compared to Gram-positive bacteria and fungi. Mean-
while, it should be noted that though the fluorescence responses
of TPE-ARs to various pathogens are similar in the same group,
the extents are still different, implying the variance of the weak
interactions between TPE-ARs and pathogens.
2.2.3. Self-Assembly Behavior of TPE-ARs Enriching
the Response Difference
It was interesting to observe the emission intensity of TPE-
ABu and TPE-AHex was attenuated after incubated with
©
Adv. Funct. Mater. 2018, 1805986
1805986 (4 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 3. Fluorescence response of seven TPE-ARs after the addition of microbes (CTPE-AR = 20 × 10−6 m) (Left). Each value was the average of six
independent measurements, error bar shows the standard deviation of these measurements. λex: 340 nm, λem: 470 nm. I0 and I are the fluorescence
intensity of TPE-ARs in the absence and presence of microbes. The grouping criterion of seven TPE-ARs based on the fluorescence intensity change
(Right). The color depth of blue circle stands for the relative fluorescence intensity, that is, with the increase of the blue depth, the change of
fluorescence intensity increases.
Gram-positive bacteria and fungi while increased or almost
unchanged with Gram-negative bacteria (Figures 3 and 4a).
This observation is different from the fluorescence turn-on
response of the other TPE-ARs to the pathogens. To understand
this interesting phenomenon, we rechecked the fluorescence
intensity of seven TPE-ARs below their CAC. It was found that
TPE-ABu and TPE-AHex show moderate emission in contrast
to the other five AIEgens (Figure 4b). This means that below
CAC, TPE-ABu and TPE-AHex tend to form large and loose
premicellar aggregates. As demonstrated by the DLS results
in Figure 4c, the size distribution of TPE-ABu and TPE-AHex
below their CAC (≈1 µm) is much larger than that beyond CAC
(≈20 nm for TPE-ABu and ≈100 nm for TPE-AHex). This is
very similar to the reported aggregation behavior of oligomeric
surfactants, where they first generate large network-like aggre-
gates below the CAC and transform into small micelles with
the increase of their concentration.[22] Based on DLS results
and transmission electron microscopy (TEM) images, TPE-
ABu and TPE-AHex first form large ribbons and sheets below
their CAC (Figure 4d,e), respectively. Beyond CAC, their aggre-
gates changed to spherical aggregates of ≈20 nm for TPE-ABu
(Figure S12d, Supporting Information) and smaller sheets of
about 100 nm for TPE-AHex (Figure S12g, Supporting Infor-
mation). The formation of these premicellar aggregations could
be attributed that TPE-ABu and TPE-AHex bear the longer
hydrophobic chains relative to other TPE-ARs, which provide
the stronger hydrophobic interaction. Given that the strength
of interaction between pathgens and TPE-ARs was weaker than
that between TPE-ARs themselves in the premicellar aggre-
gates, the emission intensity of TPE-ARs could be attenuated
after incubating with pathogens according to the restriction of
intramolecular motions (RIM) mechanism of AIE.[16] Gener-
ally, Gram-positive bacteria and fungi have a relative loose and
poriferous cell wall,[23] thus the weaker interaction between
pathogens and TPE-ARs cannot effectively restrict the intramo-
lecular motion of TPE-ARs, leading to the attenuated fluores-
cence compared to that of the original premicellar aggregate.
In contrast, as for the Gram-negative bacteria, whose cell wall
consists of an outer lipid membrane and a thin cross-linked
peptidoglycan network,[23b] the strong hydrophobic interaction
between the lipid membrane of bacteria and TPE-ARs greatly
restricts the rotation of TPE groups, leading to the increase of
fluorescence. This diverse self-assembly behavior of TPE-ARs
also enriches the interactions of TPE-ARs with pathogens,
contributing to the subtle difference in fluorescence response
toward different pathogens.
2.3. Pathogen Identification Using Fabricated Sensor Arrays
2.3.1. TPE-ARs Sensor Arrays for Identifying Pathogen
To make a balance between the diverse response and simplicity
of the sensor array, we chose one AIEgen from group A, B, and
C (Figure 3) to build up a sensor array. Seven TPE-ARs give
17 combinations of sensor arrays, each consisting of three
AIEgens, as shown in Table S2 (Supporting Information). It
has been verified that the further decrease of AIEgen number
(n < 3) will reduce the identification accuracy. To test the
capability of the built sensor array, the fluorescence response
patterns of pathogens produced by the sensor array were ana-
lyzed using the linear discriminant analysis (LDA), a powerful
statistical method extensively used in pattern recognition.[24]
Taking the sensor array of TPE-APrA, TPE-ACH, and TPE-
AHex (the combination of AB1C) as an example, the fluores-
cence pattern of pathogens (Figure 5a) could be transformed to
2D canonical score plot by LDA (Figure 5b). The seven patho-
gens were well-clustered into seven groups and discriminated
completely from each other. Interestingly, the distribution
of seven pathogens in the 2D canonical score plot obviously
associated with their categories, where Gram-negative bac-
teria were placed at the left, and the Gram-positive bacteria
were at the right. The 100% accuracy of discrimination was
achieved and proved by “leave-one-out” cross-validation in LDA,
demonstrating that our AIEgen sensor array is highly effective
for pathogen identification. Significantly, E. coliR and S. aureusR
©
Adv. Funct. Mater. 2018, 1805986
1805986 (5 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 4. a) Fluorescence spectra of 20 × 10−6 m TPE-AHex in PBS and microbes suspensions. b) Plots of PL intensity of TPE-ARs versus the dye
concentration in PBS solution. c) Size distribution of TPE-ABu and TPE-AHex aggregates in PBS solution at the concentration of 20 and 200 × 10−6 m.
d,e) Cryo-TEM images of 20 × 10−6 m TPE-ABu and TPE-AHex in PBS solution, respectively.
Figure 5. a) Fluorescence response patterns of seven microbes stained by 20 × 10−6 m TPE-APrA, TPE-ACH, and TPE-AHex (combination AB1C)
■
transformed from Figure 3. b) Canonical score plot for the fluorescence response patterns determined by LDA ( is the centroid of each group and
ellipses depict the 95% confidence limits for each pathogen).
©
Adv. Funct. Mater. 2018, 1805986
1805986 (6 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
can also be easily discerned from normal E. coli and S. aureus
samples by our sensor array, which is very important for effec-
tive treatment.
After testing the identification ability of 17 sensor arrays
(Figure S13 and Table S2, Supporting Information), 14 out
of 17 sensor arrays can achieve 100% classification accuracy
for the seven pathogens as well as nearly 100% identification
accuracy as proved by cross-validation.
Further to verify the ability of our sensor arrays to predict
unknown bacteria, we randomly selected 14 microorganism
samples from the seven targeted pathogens. Again, the sensor
array consisting of TPE-APrA, TPE-ACH, and TPE-AHex was
selected as a representative. Fluorescence response patterns
of 14 microorganism samples generated from the sensor
array were transformed to the canonical scores by the discri-
minant functions established from the training samples used
in Figure 5. The Mahalanobis distances of a detected sample
to the respective centroids of seven groups were calculated in
a 3D space (canonical factors 1–3), as shown in Figure S14
(Supporting Information). The shortest Mahalanobis distance
value decides the arrangement of pathogen sample. In this
way, the 14 unknown samples were completely identified with
a detection accuracy of 100%, demonstrating the high reliability
of our sensor array (Table S3, Supporting Information).
arrays. In contrast, the introduction of phenyl ring decreases
the fluorescence response, leading to the compromised sensi-
tivity and accuracy, as exampled by the sensor arrays TPE-AMe,
TPE-ABn, and TPE-AHex; TPE-AEt, TPE-ACH, and TPE-ABn;
and TPE-ACH, TPE-ABn, and TPE-AHex (Table S2, Supporting
Information). This may be attributed that the additional phenyl
ring is a hydrophobic group with large steric hindrance, which
could hinder the interaction of TPE core with pathogens.
Collectively, the positive charge, hydrophobic substitutions,
and the resulted aggregative behavior contribute to the multiva-
lent interactions of TPE-ARs with the pathogens, which greatly
augment the diversity in the fluorescence response patterns,
contributing to the high detection accuracy.
2.3.3. TPE-ARs Sensor Arrays for Identifying Blends of Pathogens
Practically, the clinical diagnosis often faces the mixed pathogen
samples. Thus, the sensor array TPE-APrA, TPE-ACH, and
TPE-AHex was also a representative to identify the pathogen
mixtures. Eight representative blended samples including
four mixtures of two species of microbes and four mixtures of
three species of microbes were used as targeted mixed patho-
gens. Similarly, LDA was used to transform the fluorescence
response patterns of pathogen mixtures to 2D canonical score
plot, where eight blended pathogens were well-clustered into
eight groups with a 100% discrimination accurate (Figure 6).
Based on the established discriminant function, eight randomly
selected blended samples were also completely identified with a
detection accuracy of 100% (Table S4, Supporting Information).
2.3.2. Fabrication Rule of Sensor Arrays
Based on the above results, we found that TPE-ARs bearing
the shorter hydrophobic chain or weaker hydrophobility with
3 < ClogP < 5 show more selectively binding with Gram-
positive bacteria and fungi, whereas those having the longer
hydrophobic chain or stronger hydrophobility with ClogP > 6
show higher affinity for Gram-negative bacteria. When TPE-
ARs present the moderate hydrophobility with 5 < ClogP < 6,
they possess the similar affinity for three kinds of pathogens.
The combination of three groups of AIEgens with different
hydrophobilities succeeded in generating the competent sensor
3. Conclusion
In summary, we have successfully constructed 14 competent
sensor arrays with TPE-based AIEgens for fast and reliable
pathogen identification. Each sensor array is composed of
three TPE-ARs on the basis of the criterion of balancing the
Figure 6. a) Fluorescence response patterns of eight microbe mixtures stained by 20 × 10−6 m TPE-APrA, TPE-ACH, and TPE-AHex (combination
AB1C). Each value was the average of six independent measurements, error bar shows the standard deviation of these measurements. λex: 340 nm,
λ
em: 470 nm. I0 and I are the fluorescence intensity of TPE-ARs in the absence and presence of microbes. b) Canonical score plot for the fluorescence
■
response patterns determined by LDA ( is the centroid of each group and ellipses depict the 95% confidence limits for each pathogen).
©
Adv. Funct. Mater. 2018, 1805986
1805986 (7 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Synthesis of (2-(4-(2-(2-Bromoethoxy)ethoxy)phenyl)ethene-1,1,2-triyl)-
tribenzene (2): Hydroxylated TPE (1) (1.74 g, 5 mmol), dibromoethyl
ether (1.39 g, 6 mmol), and K2CO3 (1.38 g, 10 mmol) were added in
a two-necked round-bottom flask. 30 mL acetone was added under
N2 protection. The reaction mixture was heated to reflux overnight.
After solvent evaporation, the crude product was purified by silica gel
column chromatography using n-hexane/DCM (20: 1) as eluent to give
diversity of fluorescence response and simplicity of the sensor
array. TPE-ARs bear one cationic ammonium group but
different hydrophobic groups with finely controlled hydro-
phobicity, which were engineered to tune the electrostatic and
hydrophobic interactions between AIEgens and pathogens.
Meanwhile, TPE-ARs also show diverse aggregate behaviors,
further enriching the multivalent interactions of TPE-ARs and
pathogens. Thanks to the diverse interactions of TPE-ARs with
the pathogens, each sensor array can give a unique fluores-
cence response pattern for different pathogens. By recognizing
the fluorescence pattern of pathogens with assistance of LDA,
the seven different pathogens, even normal and drug-resistant
bacteria, are identified effectively with nearly 100% accuracy.
Also, our sensor arrays are highly suitable for the complicated
situations including multiple pathogens. Moreover, the iden-
tification procedure is fast (about 0.5 h), very simple without
washing steps and high throughput, which exhibits a great
potential to offer timely and reliable pathogen information for
clinical decisions and monitoring trends of infectious disease.
Currently, we are attempting to identify clinical samples using
our sensor assays.
1
a colorless oil with 65% yield. H NMR (400 MHz, CDCl3, δ): 7.13–7.06
(m, 9H), 7.02 (dtd, J = 9.5, 4.6, 2.1 Hz, 6H), 6.95–6.90 (m, 2H),
6.67–6.62 (m, 2H), 4.06 (dd, J = 5.6, 3.8 Hz, 2H), 3.86–3.81 (m, 4H),
3.49–3.46 (m, 2H).
Synthesis of TPE-AR (3): The mixture of (2-(4-(2-(2-bromoethoxy)-
ethoxy)phenyl)ethene-1,1,2-triyl)tribenzene (2) (2 mmol) and the
corresponsing amine (10 mmol) was stirred and heated to reflux in
ethanol (20 mL) for 24 h. The solvent was removed under vacuum, and
the crude product was repeatedly dissolved by a little methanol, and then
excess THF was added to precipitate a white powder, which was dried to
obtain the desired compound. The corresponding characterization is as
follows:
N,N,N-Trimethyl-2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)ethoxy)ethan-
1-aminium bromide (TPE-AMe): Yield: 82%. 1H NMR (400 MHz,
DMSO-d6, δ): 7.12 (m, 9H), 6.96 (m, 6H), 6.86 (d, J = 8.8 Hz, 2H), 6.70
(d, J = 8.8 Hz, 2H), 4.06–4.00 (m, 2H), 3.88 (m, 2H), 3.78–3.73 (m, 2H),
3.57 – 3.52 (m, 2H), 3.09 (s, 9H). 13C NMR (100 MHz, DMSO-d6, δ):
156.96, 143.61, 143.53, 143.49, 140.25, 139.91, 135.69, 132.07, 130.79,
130.76, 128.03, 127.93, 126.65, 126.57, 113.84, 68.81, 66.71, 64.54,
+
64.24, 60.25, 53.29. MALDI-TOF HRMS: calcd. for C33H36NO2 [M-Br]+:
478.2741, found: 478.2782.
4. Experimental Section
N-Ethyl-N,N-dimethyl-2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)ethoxy)-
ethan-1-aminium bromide (TPE-AEt): Yield: 84%. 1H NMR (400 MHz,
DMSO-d6, δ): 7.17–7.06 (m, 9H), 6.96 (m, 6H), 6.86 (dd, J = 8.8,
3.3 Hz, 2H), 6.70 (dd, J = 8.8, 3.3 Hz, 2H), 4.07–4.01 (m, 2H), 3.88
(s, 2H), 3.79–3.72 (m, 2H), 3.52 (m, 2H), 3.07 (d, J = 3.2 Hz, 2H), 3.03
(d, J = 3.2 Hz, 6H), 1.22 (m, 3H). 13C NMR (100 MHz, DMSO-d6, δ):
156.86, 143.49, 143.41143.37, 140.12, 139.74, 135.52, 131.94, 130.67,
130.64, 127.88, 127.78, 126.49, 126.42, 126.38, 113.71, 68.72, 66.58,
63.93, 63.90, 61.78, 59.69, 50.14, 7.98, 7.95. MALDI-TOF HRMS: calcd.
for C34H38NO2+ [M-Br]+: 492.2897, found: 492.2881.
Materials and Instrumentations: All the materials and solvents were
purchased from the commercial sources (J&K, TCI, and Sigma-Aldrich
Company) and used as received. Seven kinds of microorganisms were
chosen including three Gram-positive bacteria ((S. aureus (ATCC6538),
PenrS. aureus (CGMCC1.879), and E. faecalis (JCM5803)), three Gram-
negative bacteria (E. coli (ATCC25922), AmprE. coli (TOP10), and
P. aeruginosa (JCM5962)), and one fungi (C. albicans (ATCC10231)),
which were purchased from Beijing Bio-Med Technology Development
Co., Ltd. and China General Microbiological Culture Collection Center.
1 × PBS (pH 7.4) was used throughout the work. NMR spectra were
recorded with a Bruker ARX 400 NMR spectrometer and HRMS were
measured in positive mode on a Finnegan MAT TSQ 7000 Mass
Spectrometer. UV–vis absorption spectra were recorded on a Milton
Ray Spectronic 3000 array spectrophotometer. Fluorescence emission
spectra were measured with a Perkin Elmer LS 55 spectromete. The
size distribution and zeta potential were measured on Nano ZS
(ZEN3600). Laser confocal scanning microscope images were collected
on a confocal laser scanning microscopy (FV1200-IX83, Olympus,
Japan). Fluorescence intensities of TPE-ARs before and after adding
pathogens were recorded on a microplate reader (Varioskan Flash)
using black 96-well plates. The morphology of TPE-ARs aggregates
was observed by TEM (JEM 2010) and scanning electron microscopy
(SEM, S-4300).
Synthesis of TPE-ARs: Synthesis of 4-(1,2,2-Triphenylvinyl)phenol (1):
Benzophenone (1.82 g, 10 mmol), 4-hydrobenzophenone (1.98 g,
10 mol), and zinc dust (2.60 g, 40 mmol) were placed into a two-necked
round-bottom flask. The flask was evacuated under vacuum and refilled
with nitrogen three times. Under a nitrogen atmosphere, 70 mL dried
tetrahydrofuran (THF) was added, and then 2.2 mL TiCl4 (20 mmol) was
slowly added under stirring in dry-ice acetone bath. The reaction mixture
was heated to reflux overnight under N2 protection. After cooling to
room temperature, 50 mL dilute hydrochloric acid (1 m) was added to
the mixture, and then the mixture was extracted with dichloromethane
(DCM). The combined organic phase was dried over anhydrous sodium
sulfate and filtered. After the solvent evaporated, the crude product was
purified by silica gel column chromatography using n-hexane/ethyl acetate
(40: 1) as eluent. The white powder was obtained with the yield of 52%.
1H NMR (400 MHz, CDCl3, δ): 7.16–7.09 (m, 8H), 7.07–7.01 (m, 9H),
6.91 (d, J = 8.0, 1H), 6.58 (d, J = 8.0, 1H), 4.78 (s, 1H).
3-Hydroxy-N,N-dimethyl-N-(2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)-
ethoxy)ethyl)propan-1-aminium bromide (TPE-APrA). Yield: 86%.
1H NMR (400 MHz, DMSO-d6, δ): 7.18–7.06 (m, 9H), 7.00–6.92
(m, 6H), 6.86 (d, J = 8.4 Hz, 2H), 6.70 (d, J = 8.4 Hz, 2H), 4.78
(t, J = 5.0 Hz, 1H), 4.06–4.01 (m, 2H), 3.88 (s, 2H), 3.75 (m, 2H),
3.56–3.51 (m, 2H), 3.44 (d, J = 5.0 Hz, 2H), 3.07 (s, 2H), 3.05 (s, 6H),
1.85–1.78 (m, 2H). 13C NMR (100 MHz, DMSO-d6, δ): 157.36, 144.04,
143.95, 140.73, 140.46, 136.25, 132.53, 131.20, 131.17, 128.54, 128.44,
127.21, 127.13, 127.10, 114.30, 69.39, 67.23, 64.58, 63.13, 62.99, 58.27,
+
51.65, 25.85, 25.79. MALDI-TOF HRMS: calcd. for C35H40NO3 [M-Br]+:
522.3003, found: 522.3058.
N,N-Dimethyl-N-(2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)ethoxy)ethyl)-
butan-1-aminium bromide (TPE-ABu). Yield: 89%. 1H NMR (400 MHz,
DMSO-d6, δ): 7.12 (m, 9H), 7.00–6.92 (m, 6H), 6.86 (d, J = 8.7 Hz, 2H),
6.69 (d, J = 8.7 Hz, 2H), 4.03 (m, 2H), 3.89–3.85 (m, 2H), 3.75 (m, 2H),
3.56–3.51 (m, 2H), 3.28 (d, J = 8.8 Hz, 2H), 3.04 (s, 6H), 1.64 (m, 2H),
1.35–1.25 (m, 2H), 0.84 (s, 3H). 13C NMR (100 MHz, DMSO-d6, δ):
156.86, 143.48, 143.40, 143.37, 140.11, 139.73, 135.50, 131.92, 130.66,
130.63, 127.86, 127.77, 126.48, 126.39, 126.37, 113.67, 68.73, 66.63,
63.93, 63.87, 62.52, 62.20, 50.78, 50.66, 23.80, 19.22, 19.15. 13.58, 13.49.
+
MALDI-TOF HRMS: calcd. for C36H42NO2 [M-Br]+: 520.3210, found:
520.3244.
N,N-Dimethyl-N-(2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)ethoxy)ethyl)-
cyclohexanaminium bromide (TPE-ACH). Yield: 84% yield. 1H NMR
(400 MHz, DMSO-d6, δ): 7.11 (m, 9H), 7.01–6.91 (m, 6H), 6.88–6.83
(m, 2H), 6.72–6.66 (m, 2H), 4.04 (m, 2H), 3.87 (d, J = 6.1 Hz, 2H),
3.80–3.73 (m, 2H), 3.56 (m, 3H), 2.98 (s, 6H), 2.08 (m, 2H), 1.85–1.77
(m, 2H), 1.47–1.18 (m, 6H). 13C NMR (100 MHz, DMSO-d6, δ): 157.08,
143.72, 143.64, 143.60, 140.37, 140.04, 135.81, 132.18, 130.88, 130.85,
128.13, 128.06, 126.79, 126.69, 113.90, 72.45, 69.07, 66.93, 64.22, 64.16,
©
Adv. Funct. Mater. 2018, 1805986
1805986 (8 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
61.38, 48.52, 25.68, 25.64, 25.10, 25.04, 24.55. MALDI-TOF HRMS:
calcd. for C38H44NO2+ [M-Br]+: 546.3367, found: 546.3391.
Microorganism Imaging: 500 µL bacteria (OD600 = 1.0) or fungi
(OD600 = 2.0) was transferred to a 1.5 mL centrifuge tube, and harvested
by centrifuging (7100 rpm, 2 min). After removing the supernatant, 50 µL
of 20 × 10−6 m TPE-ARs PBS solutions were added into the centrifuge
tube, and incubated at 37 °C for 15 min. Then 3 µL microorganism
solution was added to glass slide and then covered by a coverslip for
imaging. The image was collected using CLSM with λex = 405 nm and
λem = 430–500 nm.
N-Benzyl-N,N-dimethyl-2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)ethoxy)-
ethan-1-aminium bromide (TPE-ABn). Yield: 65%. 1H NMR (400 MHz,
methanol-d4, δ): 7.62–7.52 (m, 9H), 7.07 (m, 7H), 7.00–6.96 (m, 4H),
6.89 (d, J = 8.8 Hz, 2H), 6.67 (d, J = 8.8 Hz, 2H), 4.64 (s, 2H), 4.15 (dd,
J = 4.5, 2.3 Hz, 2H), 4.09–4.04 (m, 2H), 3.93–3.86 (m, 2H), 3.65–3.58
(m, 2H), 3.11 (s, 6H). 13C NMR (100 MHz, DMSO-d6, δ): 156.88, 143.50,
143.42, 143.39, 140.14, 139.77, 135.55, 133.18, 133.15, 132.84, 131.94,
130.69, 130.65, 130.36, 130.33, 128.98, 128.95, 128.91, 128.05, 127.99,
127.89, 127.79, 126.51, 126.42, 113.74, 68.81, 67.46, 67.28, 67.13, 66.62,
64.00, 63.92, 63.14, 62.65, 56.06, 51.78, 49.83, 48.13, 18.57. MALDI-TOF
HRMS: calcd. for C39H40NO2+ [M-Br]+: 554.3054, found: 554.3025.
N,N-Dimethyl-N-(2-(2-(4-(1,2,2-triphenylvinyl)phenoxy)ethoxy)ethyl)-
hexan-1-aminium bromide (TPE-AHex). Yield: 89%. 1H NMR (400 MHz,
DMSO-d6, δ): 7.12 (ddt, J = 13.4, 8.9, 6.5 Hz, 9H), 6.99–6.91 (m, 6H),
6.88–6.83 (m, 2H), 6.72–6.67 (m, 2H), 4.03 (m, 2H), 3.89–3.86 (m, 2H),
3.75 (m, 2H), 3.54 (m, 2H), 3.33–3.27 (m, 2H), 3.05 (s, 6H), 1.63
(m, 2H), 1.27 (m, 4H), 0.94 (t, J = 7.3 Hz, 2H), 0.86 (t, J = 7.3 Hz, 3H).
13C NMR (100 MHz, DMSO-d6, δ): 156.86, 143.49, 143.42, 143.39,
140.11, 139.74, 135.50, 131.93, 130.67, 130.64, 127.87, 127.79, 127.76,
126.49, 126.40, 113.66, 68.76, 66.64, 64.06, 63.95, 62.18, 50.79, 30.74,
30.66, 25.38, 21.91, 21.84, 21.77, 13.87, 13.82. MALDI-TOF HRMS:
calcd. for C38H46NO2+ [M-Br]+: 548.3523, found: 548.3589.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
C.Z. and W.X. contributed equally to this work. The authors thank the
staff scientists Dr. Bo Guan, Yang Liu, and Jiling Yue at the Center for
Analysis and Testing, ICCAS for their support in cryo-TEM and cryo-SEM
measurement. The authors also greatly appreciate the financial
support from the National Natural Science Foundation of China
(21788102), the Research Grants Council of Hong Kong (16301614,
16308016, 16305015, C6009-17G, C2014-15G, N_HKUST604/14,
and A-HKUST605/16), the Innovation and Technology Commission
(ITC-CNERC14SC01 and ITS/254117), and the Technology Plan of
Shenzhen (JCYJ20160229205601482).
Preparation of TPE-ARs Solution: The stock solution of 1.0 or 5.0 × 10−3
m
TPE-ARs in DMSO was prepared, and diluted to the corresponding
concentration with PBS for experiments.
Preparation of Bacteria and Fungi Solution: A single colony of bacteria
or fungi on solid agar plate was transferred to 10 mL corresponding
liquid culture medium and grown at 37 °C for 6–8 h for bacteria and
30 °C for ≈10 h for fungi (culture medium: LB for E. coli, Ampr E. coli,
and P. aeruginosa, NB for S. aureus and PenrS. aureus, TSB for E. faecalis,
and YPD for C. albicans). Microorganisms were harvested by centrifuging
at 7100 rpm for 2 min. The remaining microorganisms were resuspended
with PBS, and diluted to an optical density of 1.0 for bacteria and 2.0 for
fungi at 600 nm (OD600 = 1.0 for bacteria and OD600 = 2.0 for fungi).
Experimental Procedure for Pathogen Identification: 40 µL bacteria
(OD600 = 1.0) or fungi (OD600 = 2.0) was transferred to the black
96-well plate, and then 60 µL TPE-AR PBS solution was added with
the final volume of 100 µL and the final TPE-AR concentration of
20 × 10−6 m. After shaking for 30 s, the mixtures were incubated at
37 °C for 15 min. Six repeated experiments were performed. All the
fluorescence intensities were recorded on a microplate reader with
λex = 340 nm and λem = 470 nm. TPE-ARs solution without microbes
was also treated under the same conditions as control. The relative
fluorescence intensities of TPE-ARs before and after incubation with
microbes, that is, (I − I0)/I0 were calculated by software Microsoft
Excel, where I is the fluorescence intensity of TPE-ARs after adding
the microbes, and I0 is the fluorescence intensity of TPE-ARs without
microbes. The obtained relative fluorescence intensities (I − I0)/I0 were
introduced in mathematical analysis to receive bacteria identification
results by the software IBM SPSS Statistics 22. The whole process
only needs about 0.5 h.
Dynamic Light Scattering: The size distribution of TPE-ARs
aggregates in organic solvent/water mixtures with water fraction of
96% was measured at 25 °C at a scattering angle of 173° with Nano ZS
(ZEN3600) equipped with a thermostated chamber and a 4 mW He–Ne
laser (λ = 632.8 nm).
ζ-Potential Measurements: Seven pathogens were incubated by seven
TPE-ARs at 37 °C for 15 min, respectively. The microorganisms were
harvested by centrifuging at 7100 rpm for 2 min and resuspended in
H2O for zeta potential measurements. The pathogens without TPE-ARs
were also treated under the same conditions as control.
Cryo-TEM Measurements: TPE-ARs solutions were placed on freshly
carbon-coated holey TEM grids. The excess solutions were removed with
filter paper, and then the TEM grids were placed into liquid nitrogen to
make the samples embedded in a thin layer of vitreous ice. The frozen
samples were observed at 120 kV in low-dose mode by TEM (JEM 2010).
Conflict of Interest
The authors declare no conflict of interest.
Keywords
AIEgens, mathematical analysis, multivalent interaction, pathogen
detection, simple and accurate
Received: August 27, 2018
Revised: October 15, 2018
Published online:
[1] a) M. Urdea, L. A. Penny, S. S. Olmsted, M. Y. Giovanni, P. Kaspar,
A. Shepherd, P. Wilson, C. A. Dahl, S. Buchsbaum, G. Moeller,
D. C. H. Burgess, Nature 2006, 444, 73; b) X. Didelot, R. Bowden,
D. J. Wilson, T. E. Peto, D. W. Crook, Nat. Rev. Genet. 2012, 13, 601.
[2] “Common Infectious Diseases Worldwide,” Infoplease. © 2000–
2017 Sandbox Networks, Inc., publishing as Infoplease, https://
www.infoplease.com/science-health/health/common-infectious-
diseases-worldwide/ (accessed: July 2018).
[3] a) O. Lazcka, F. J. Del Campo, F. X. Munoz, Biosens. Bioelectron.
2007, 22, 1205; b) A. Duarte, A. Chworos, S. F. Flagan, G. Hanrahan,
G. C. Bazan, J. Am. Chem. Soc. 2010, 132, 12562.
[4] a) B. S. Reisner, G. L. Woods, J. Clin. Microbiol. 1999, 37,
2024; b) R. P. Peters, M. A. van Agtmael, A. M. Simoons-Smit,
S. A. Danner, C. M. Vandenbroucke-Grauls, P. H. Savelkoul, J. Clin.
Microbiol. 2006, 44, 4186.
[5] H. Bai, H. Chen, R. Hu, M. Li, F. Lv, L. Liu, S. Wang, ACS Appl.
Mater. Interfaces 2016, 8, 31550.
[6] A. Borst, A. T. A. Box, A. C. Fluit, Eur. J. Clin. Microbiol. Infect. Dis.
2004, 23, 289.
©
Adv. Funct. Mater. 2018, 1805986
1805986 (9 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
[7] a) M. Quesada, M. Giménez, S. Molinos, G. Fernández,
M. Sánchez, R. Rivelo, A. Ramírez, G. Banqué, V. Ausina,
Clin. Microbiol. Infect. 2010, 16, 137; b) M.-P. Romero-Gómez,
R. Gómez-Gil, J. R. Paño-Pardo, J. Mingorance, J. Infect. 2012,
65, 513; c) W. Chen, Q. Li, W. Zheng, F. Hu, G. Zhang, Z. Wang,
D. Zhang, X. Jiang, Angew. Chem. 2014, 126, 13954.
[15] a) H. Jiang, P. Taranekar, J. R. Reynolds, K. S. Schanze, Angew.
Chem., Int. Ed. 2009, 48, 4300; b) N. Cheeveewattanagul,
E. Morales-Narváez, A. R. H. Hassan, J. F. Bergua, W. Surareungchai,
M. Somasundrum, A. Merkoçi, Adv. Funct. Mater. 2017, 27, 1702741.
[16] J. Mei, N. L. C. Leung, R. T. K. Kwok, J. W. Y. Lam, B. Z. Tang, Chem.
Rev. 2015, 115, 11718.
[8] M. H. Kollef, Crit. Care 2001, 5, 189.
[17] a) Y. Yuan, S. Xu, X. Cheng, X. Cai, B. Liu, Angew. Chem., Int. Ed.
2016, 55, 6457; b) G. Feng, Y. Yuan, H. Fang, R. Zhang, B. Xing,
G. Zhang, D. Zhang, B. Liu, Chem. Commun. 2015, 51, 12490.
[18] J. Heo, S. Z. Hua, Sensors 2009, 9, 4483.
[19] a) J. Liang, B. Z. Tang, B. Liu, Chem. Soc. Rev. 2015, 44, 2798;
b) R. T. K. Kwok, C. W. T. Leung, J. W. Y. Lam, B. Z. Tang, Chem. Soc.
Rev. 2015, 44, 4228.
[20] G. Liu, S. Tian, C. Li, G. Xing, L. Zhou, ACS Appl. Mater. Interfaces
2017, 9, 28331.
[21] S. M. Yoo, S. Y. Lee, Trends Biotechnol. 2016, 34, 7.
[22] a) Y. Hou, M. Cao, M. Deng, Y. Wang, Langmuir 2008, 24, 10572;
b) Y. Fan, Y. Hou, J. Xiang, D. Yu, C. Wu, M. Tian, Y. Han, Y. Wang,
Langmuir 2011, 27, 10570.
[23] a) W. L. Chaffin, J. L. López-Ribot, M. Casanova, D. Gozalbo,
J. P. Martínez, Microbiol. Mol. Biol. Rev. 1998, 62, 130;
b) Y. Wang, T. S. Corbitt, S. D. Jett, Y. Tang, K. S. Schanze, E. Y. Chi,
D. G. Whitten, Langmuir 2012, 28, 65.
[24] a) P. Anzenbacher Jr., P. Lubal, P. Bucˇek, M. A. Palacios,
M. E. Kozelkova, Chem. Soc. Rev. 2010, 39, 3954; b) J. R. Askim,
M. Mahmoudi, K. S. Suslick, Chem. Soc. Rev. 2013, 42, 8649.
[9] a) J. Davies, Science 1994, 264, 375; b) J. L. Martínez, Science
2008, 321, 365; c) J. L. Martínez, F. Baquero, D. I. Andersson,
Nat. Rev. Microbiol. 2007, 5, 958; d) M. B. Edmond, S. E. Wallace,
D. K. McClish, M. A. Pfaller, R. N. Jones, R. P. Wenzel, Clin. Infect.
Dis. 1999, 29, 239.
[10] F. Walsh, http://www.bbc.com/news/health-30416844 (accessed:
December 2014).
[11] A. B. Chinen, C. M. Guan, J. R. Ferrer, S. N. Barnaby, T. J. Merkel,
C. A. Mirkin, Chem. Rev. 2015, 115, 10530.
[12] a) C. Zhu, L. Liu, Q. Yang, F. Lv, S. Wang, Chem. Rev. 2012, 112,
4687; b) R. L. Phillips, O. R. Miranda, C. C. You, V. M. Rotello,
U. H. Bunz, Angew. Chem., Int. Ed. 2008, 47, 2590; c) H. Yuan,
Z. Liu, L. Liu, F. Lv, Y. Wang, S. Wang, Adv. Mater. 2014, 26,
4333; d) S. Xu, Q. Wang, Q. Zhang, L. Zhang, L. Zuo, J.-D. Jiang,
H.-Y. Hu, Chem. Commun. 2017, 53, 11177; e) H. Pei, J. Li, M. Lv,
J. Wang, J. Gao, J. Lu, Y. Li, Q. Huang, J. Hu, C. Fan, J. Am. Chem.
Soc. 2012, 134, 13843.
[13] J. B. Birks, Photophysics of Aromatic Molecules, Wiley, New York 1970.
[14] D. Ding, K. Li, B. Liu, B. Z. Tang, Acc. Chem. Res. 2013, 46, 2441.
©
Adv. Funct. Mater. 2018, 1805986
1805986 (10 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim