Integrated Structure and Device Engineering for High Performance and Scalable Quantum Dot Infrared Photodetectors

Abstract

Colloidal quantum dots (CQDs) are emerging as promising materials for the next generation infrared (IR) photodetectors, due to their easy solution processing, low cost manufacturing, size‐tunable optoelectronic properties, and flexibility. Tremendous efforts including material engineering and device structure manipulation have been made to improve the performance of the photodetectors based on CQDs. In recent years, benefiting from the facial integration with materials such as 2D structure, perovskite and silicon, as well as device engineering, the performance of CQD IR photodetectors have been developing rapidly. On the other hand, to prompt the application of CQD IR photodetectors, scalable device structures that are compatible with commercial systems are developed. Herein, recent advances of CQD based IR photodetectors are summarized, especially material integration, device engineering, and scalable device structures.

Full text

2003397 (1 of 20) © 2020 Wiley-VCH GmbH
www.small-journal.com
Review
Integrated Structure and Device Engineering for High
Performance and Scalable Quantum Dot Infrared
Photodetectors
Kaimin Xu, Wenjia Zhou,* and Zhijun Ning*
K. Xu, Dr. W. Zhou, Prof. Z. Ning
School of Physics Science and Technology
ShanghaiTech University
Shanghai 201210, China
E-mail: zhouwj@shanghaitech.edu.cn; ningzhj@shanghaitech.edu.cn
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/smll.202003397.
DOI: 10.1002/smll.202003397
1. Introduction
Photodetectors that can convert light into electronic signal play
an important role in modern society. According to the detection
wavelength, photodetectors can be categorized as X-ray photo-
detectors, ultraviolet (UV) photodetectors, visible (Vis) light
photodetectors, and infrared (IR) photodetectors. X-ray photode-
tectors, usually used in medical field such as Computed Tomo-
graphy (CT) machine, are fabricated from materials that have a
high absorption factor for X-ray. UV photodetectors are mainly
used in military field such as confidential communication, mis-
sile defense. The conventional UV photodetectors are based on
silicon devices, which respond to visible and near-infrared light
at the same time. Visible photodetectors, such as CCD (Charge-
coupled Device) and CMOS (Complementary Metal Oxide Semi-
conductor) imaging sensors, widely used in smart phone and
digital camera, are the most commonly used photodetectors.
IR photodetectors, the subject of this review, play an important
role in military detecting, imaging, communications, bio-sensing
and weather forecast, etc. The mechanism of IR photodetector is
similar with other kinds of photodetectors. The main dierences
Colloidal quantum dots (CQDs) are emerging as promising materials for
the next generation infrared (IR) photodetectors, due to their easy solution
processing, low cost manufacturing, size-tunable optoelectronic properties,
and flexibility. Tremendous eorts including material engineering and device
structure manipulation have been made to improve the performance of the
photodetectors based on CQDs. In recent years, benefiting from the facial
integration with materials such as 2D structure, perovskite and silicon, as
well as device engineering, the performance of CQD IR photodetectors have
been developing rapidly. On the other hand, to prompt the application of
CQD IR photodetectors, scalable device structures that are compatible with
commercial systems are developed. Herein, recent advances of CQD based
IR photodetectors are summarized, especially material integration, device
engineering, and scalable device structures.
of them are usage scenarios and photo-
active materials. To detect long wavelength
light, IR photodetectors have to use narrow
bandgap semiconductors as photo-active
materials. Hence there are limited mate-
rials can be used for infrared detecting and
these materials generally show lower car-
rier mobility and high defect density. As a
result, infrared sensing materials generally
show lower detectivity in comparison to vis-
ible photodetectors.
Commercialized IR photodetectors are
generally based on traditional infrared
semiconductor bulk materials with
narrow bandgap such as Ge, InGaAs, and
HgCdTe. However, they are usually grown
by costly methods under high temperature
and high vacuum, and their performances
are approaching their limitation. On the other hand, due to
the increasing demand for real-time monitoring of human
body, such as heart rate, arterial oxygen saturation and respira-
tory rate, wearable flexible infrared photodetectors are of great
interest,[1] however traditional bulk IR semiconductors are gen-
erally not compatible with flexible devices. It is hence urgent to
reduce the size, weight, power consumption and improve the
performance and flexibility of infrared photodetectors.
Colloidal quantum dots (CQDs) with smaller size than their
exciton Bohr radius, having tunable bandgap, excellent light
harvesting and emitting properties.[2,3] CQDs such as lead chal-
cogenide (PbE, E = S, Se, Te),[4–6] mercury chalcogenide (HgE,
E = S, Se, Te)[7–9] and silver chalcogenide (Ag2E, E = S, Se, Te)[10]
are alternatives to traditional bulk IR semiconductors since
their broad absorption in IR region and decent carriers trans-
port. In addition, CQDs are solution synthesized and solution-
processable, which enable it to be easily integrated with variable
materials and structures to improve performances. In recent
years, the integration of CQDs with other materials is able to
reduce CQDs surface defect density, enhance carrier mobility
and diusion length.[11–13] The exploration of device engineering
is able to enhance carrier separation, transportation and extrac-
tion, significantly enhancing device performance.[14] With the
optimization of material engineering and device structure engi-
neering, more and more CQD based IR photodetectors have
been fabricated (Figure1a–c). This review will summarize the
recent progress of CQD based IR photodetectors.
CQD based IR photodetectors can be classified as
near-infrared (NIR, 0.7–1 µm), short-wavelength infrared
(SWIR, 1–2.7 µm) photodetectors, mid-wavelength infrared
Small 2020, 2003397

2003397 (2 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
(MWIR, 3–5 µm) and long-wavelength infrared (LWIR,
8–14µm) photodetectors. Mercury chalcogenide and silver chal-
cogenide CQDs are mainly used in MWIR and LWIR CQDs
based photodetectors due to their narrower bandgap. Hafizetal.
have reported a worthy review to summarize the recent devel-
opment of MWIR and LWIR photodetectors based on mercury
chalcogenide and Ag2Se CQDs.[15] Hence, in this review, we
focus on the SWIR photodetectors which mainly made by lead
chalcogenide CQDs. First, the fundamental principles of CQDs
and IR photodetectors will be introduced. Second, the recent
progress of integrated materials and device engineering for high
performance CQD based IR photodetectors will be summarized.
Third, the development of scalable CQD based IR photodetec-
tors and their attempts will be listed. At last, current challenges
in this field and possible solutions will be discussed. Progress in
SWIR photo detectors is summarized in
Table1
.
2. Fundamental Principles
2.1. Colloidal Quantum Dots
Commonly used infrared sensitive CQDs include lead chalcoge-
nide, mercury chalcogenide, and silver chalcogenide. Mercury
and silver based CQDs can achieve longer wavelength detection
(Figure2a), but the synthesis of lead based CQDs is more conven-
ient and the quality is high. Most CQDs are synthesized by hot-
injection method and show size monodispersion (Figure2b–d).
The size of CQDs can be tuned by adjusting the amount of
ligands, injection temperature and reaction time. Quantum con-
finement eect endows colloidal quantum dots with size-tunable
bandgap. Figure2c shows the absorption spectra of dierent size
of PbS CQDs, the first exciton peaks are distinct.[16] Solution-pro-
cessable character enable CQD films to be fabricated by readily
available manufacturing techniques, such as spin coating, doctor
blading, spray coating, and inkjet printing (Figure2e).[17–20]
The surface of the initial CQDs have many ligands such as
oleic acid (OA) and oleylamine (OLA), which can stabilize CQD
in solution and prevent aggregation.[21] Carrier mobility of the
initial CQD film is generally low cause of the long alkyl chain of
OA and OLA on CQDs. Therefore, a subsequent ligand exchange
processes is required to convert long-chain ligands to short-chain
ligands to improve the mobility of CQD films.[22] There are two
commonly used strategies for ligands exchange: solid ligand
exchange (SLE) and liquid ligand exchange (LLE).[23–26] Solid
ligand exchanges process is done after film fabrication; as for the
liquid ligand exchange, the exchange process is made before film
fabrication. Through ligand exchange, the energy level structure,
doping type and carrier densities of CQD films can be adjusted,
facilitates the realization of high device performance.[27–29]
2.2. Figure of Merits for Photodetectors
The figure of merits for IR photodetectors are the same with
other kinds of photodetectors, and they have been listed by
many reviews.[30–32] In this review, we only list the five most
important figure of merits for IR photodetectors: responsivity
(R), external quantum eciency (EQE), noise equivalent power
(NEP), detectivity (D*), and bandwidth.
Responsivity (R) is a basic figure of merit for photodetectors,
which equals the photocurrent divided by the incident light power
λ
λ
()
()
=
RI
P
ph
ph
(1)
External quantum eciency (EQE), the ratio of photo-gener-
ated carriers circulating in the device to incident photons, can
be calculated by the equation
λλ
λλλ
)) )
(( (
=≅
EQ
E
1.24
R
hc
qR (2)
Figure 1. Development of CQD based photodetectors. a) Materials integrated with CQDs. b) Device structure engineering of CQD based photodetec-
tors based on conventional device structures. c) CQD based photodetector arrays integrated with commercial ROICs to form IR image sensor. Adapted
with permission.[84] Copyright 2017, Springer Nature.
Small 2020, 2003397

2003397 (3 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
Noise equivalent power (NEP) is the minimum measurable
power, which is defined as the incident light power when the
eective value of the signal current output by the photodetector
is equal to the mean square root current value of the noise.
Itcan be calculated by the equation
λλ
)
)
((
=
NE
P
2
I
R
n (3)
Detectivity (D*) is the most important figure of merit for
photodetectors, as it is a comprehensive parameter allows
Table 1. Progress in CQD based photodetectors.
Year Photoactive
material
Exciton peak
[nm]
Type Ligands Spectral range
[nm]
Responsibility
[A W−1]
Detectivity [Jones] Bandwidth
[Hz]
Ref.
2009 PbS:PCBM:P3HTa) –PD + ROICs OA 1000–1900 0.5 2.3 × 1092.5k [80]
2009 PbS:PCBM 1200 PT OA 400–1400 1.6 @514 nm 2.5 × 1010 @1200 nm – [53]
2012 PbS/grapheneb) 950, 1450 PT EDT 600–1600 1077 × 1013 10 [110]
2012 PbS/graphene 990 PT Pyridine Vis–NIR 107 @895 nm – – [114]
2013 PbS 1450 PD + ROICs MPA Vis-1600 5.73 @1550 nm 1.42 × 1012 @1550 nm [81]
2013 PbS – PD + ROICs – 400–1500 – 1 × 1012 – [82]
2014 PbS:Ag NPs 900 PC MPA 350–800 3.8 × 10−37.1 × 1010 9.4 [106]
2014 PbS:Au NPs 1084 PC EDT 350–1000 9 × 10−41.1 × 1010 1.02 [59]
2014 PbS:CdS CQDs 980 PC EDT Vis–NIR 2.2 × 10−2 @650 nm 2.1 × 1010 @650 nm – [115]
2015 PbS/p-Si 1230 PD TBAI 400–1300 – 1.5 × 1011 @1230 nm – [85]
2015 PbS/MoS21380 PT EDT 400–1500 6 × 1057 × 1014 – [116]
2016 PbS 1040 PD + LED BDT 400–1400 – 1.23 × 1013 – [91]
2016 PbS/graphene 1600 PD + PT EDT 600–1800 4 × 107 V W−11013 1.5k [117]
2016 PbS/SnS2939 PT EDT 300–1000 ≈1 × 1052.4 × 1011 – [118]
2016 PbSe 1180 Vertical PT EDT 808 2 × 104 @808 nm 7 × 1012 @808 nm – [72]
2017 PbS ≈950 PC TBAI/EDT NIR – 1.71 × 1012 [119]
2017 PbS/perovskite 1240 PD MAPbI3400–1600 – 4 × 1012 @1240 nm 60k [105]
2017 PbS/graphene 1050 PT EDT 500–1200 10 2 × 1011 – [111]
2017 PbS/graphene 1670 PT + ROICs EDT 300–2000 1071012 [84]
2017 PbS/Si 1300, 1500 PVFET TBAI 400–1600 104 @1300 nm 1.8 × 1012 100k [86]
2017 PbS/P3HT 1170 Vertical PT EDT Vis–NIR 9 × 104 @808 nm 2 × 1013 @808 nm – [120]
2018 PbS/organic ≈800 PD TBAI 400–1000 6.32 1.12 × 1013 – [121]
2018 HgTe ≈2500 PD EDT SWIR – 3 × 108>10k [122]
2018 PbS/graphene 1550 PT – Vis–NIR 104 @1550 nm 1012 @1550 nm – [77]
2018 PbS/MoS2953 PT TBAI/EDT Vis–NIR 5.4 × 1041 × 1011 – [112]
2018 PbS 1550 Vertical PC TBAI Vis–NIR 5.15 @1550 nm 1.96 × 1010 @1550 nm – [123]
2019 PbS:PVP ≈950 PC BiI3 NIR 1.5 6 × 1011 >3k [124]
2019 HgTe – PD EDT + HCl SWIR-MWIR – 1 × 1010 – [71]
2019 PbS 2100 PD TBAI/EDT 400–2600 0.385@2100 nm 1.5 × 1011 @2100 nm – [125]
2019 PbS ≈1250 PD + ROICs – 1100–1400 5.9 @1250 nm – – [126]
2019 PbS/perovskite 1400 PT SCN−300–1500 1.58 @940 nm 3 × 1012 @940nm – [67]
2019 PbS/WS21800 PT ZnI2+ MPA 800–2200 1442 1 × 1012 – [113]
2019 PbS/graphene ≈1200 PT EDT NIR – >109@1200 nm – [127]
2019 PbS ≈1050 Vertical PT MPA NIR 291@1050 nm 1.84 × 1014 @1050 nm [128]
2020 PbS/n-Si 1473 PD TBAI/EDT 400–1600 0.22 @1490 nm 4.08 × 1011 @1490 nm – [87]
2020 PbS/n-Si 1540 PD EDT 400–1700 0.26 @1540 nm 1.47 × 1011 @1540 nm 29.8k [88]
2020 PbS 1500 PD + LED TBAI/EDT 400–1600 20 6.4 × 1012 – [70]
a)Materials are mixed; b)Materials are integrated but not mixed together; c)Abbreviations: PC, photoconductor; PD, photodiode; PT, phototransistor; PVFET, photovoltage
field-eect transistor; ROICs, read-out circuits; MPA, mercaptopropionic acid; EDT, 1, 2-ethanedithiol; TBAI, tetrabutylammonium iodide; BDT, benzene-1,3-dithiol; PCBM,
phenyl-C61-butyric acid methyl ester; P3HT, poly(3-hexylthiophene); PVP, poly(4-vinylpyridine).
Small 2020, 2003397

2003397 (4 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
comparison between photodetector devices with dierent
configurations and area. D* corresponds to the signal-
noise ratio (SNR) if the detection bandwidth is 1Hz and the
device area is 1 cm2 at an incident power of 1 W, and can be
written as
NEP
λλ
()
()
=
∗
D
A (4)
Response time (τ) determines the cut-o frequency or 3dB
bandwidth for photodetectors, which is defined as the time
taken for the output signal to reach (1-e−1) ≈63% of its peak
steady-state value in response to an incident optical signal. The
cut-o frequency or 3 dB bandwidth for photodetector can be
calculated by the equation
1
2
πτ
=fc (5)
Due to the narrow bandgap of IR photo-active mate-
rials, thermal noise (Johnson noise) significantly aect the
performance of IR photodetectors. Thermal noise can be calcu-
lated by equation
4
2
ikT f
R
=∆ (6)
where k is Boltzmann constant, T is temperature, Δf is meas-
urement bandwidth and R is resistance of material. Decreasing
working temperature of IR photodetectors can suppress
thermal noise and improve performance. Hence the perfor-
mance of IR photodetectors is dierent when varying tempera-
ture. When IR photodetectors are characterized at lower than
room-temperature, the temperature should be provided as a
factor for consideration.
3. Material Engineering
The CQD films are consisted with nanoparticles, and carriers
transporting rely on hooping between individual CQD. Due
to the existence of barrier between CQDs, carrier mobility in
Figure 2. a) Spectral ranges of dierent IR sensitive materials. b) Schematic diagram of hot-injection synthesis. c) Size-dependent optical absorption
spectra of PbS CQDs. Reproduced with permission.[16] Copyright 2011, American Chemical Society. d) TEM image of PbS CQDs synthesized by hot-
injection method. e) Schematic diagrams of four kinds of CQD film fabrication approaches.
Small 2020, 2003397

2003397 (5 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
IR CQD films are generally low. In addition, surface defects of
CQDs could bring carriers recombination. As a result, the diu-
sion lengths of carriers of CQD films are generally only tens of
nanometers, which limits the performance of the photodetec-
tors. Recently, some strategies are explored to mix CQDs with
novel materials to improve carrier transport ability and enhance
device performance. According to the final composites, mate-
rials mixed with CQDs are categorized into metal halide perov-
skites (MHPs), organic semiconductors, and nanoparticles.
3.1. Composite Material with MHPs
In recent years, some works applying MHPs into IR photo-
detectors were reported. Due to their high carrier mobility
and strong absorbance,[33] solution-processible MHPs are poten-
tial to achieve high performance, flexible and low cost IR
photodetectors. In 2014, Yang and coworkers applied solution-
processed lead based perovskite thin films to photodetector, the
response wavelength of which reached 800 nm and its detec-
tivity achieved 1014 Jones.[34] In 2016, Tang group reported a
CH3NH3PbI3 nanowire photodetector that not only obtained
high detectivity in NIR region, but also achieved significantly
polarization sensitivity.[35] To extend response wavelength, Xu
and Cao applied Sn–Pb mixed perovskite in photodetectors,
which can extend the response wavelength to 1000 nm.[36,37]
At the same time, Linet al. found that the sub-bandgap trap
states in organic halide perovskite single crystals can be used
to detect longer wavelength infrared light beyond their absorp-
tion edges.[38] The photodetector based on CH3NH3PbI3 single
crystal have response under 1064 nm radiation. It provides
another strategy to extend response wavelength based on perov-
skite materials. However, most of the perovskite based photo-
detectors, whether thin films or single crystals, are limited to
NIR region. To achieve beyond NIR detection ability, MHPs can
integrate with CQDs, which will be discussed in this section.
The mobility of MHPs is much larger than CQDs, but the sen-
sitivity wavelength is shorter.[39] Therefore, mixing MHPs with
CQDs is explored to combine the advantages of high mobility
of MHPs and ecient IR absorption of CQDs.
Ningetal. developed a strategy to integrate CQDs and MHPs
based on liquid ligand exchange process, producing quantum-
dots-in-perovskite (QDiP) material, which greatly promotes
the development of CQD:MHPs hybrid IR photodetectors
(Figure3a,b).[40] In this method, they applied CH3NH3I (MAI)
and PbI2 in PbS CQDs liquid ligand exchange process, then the
CQDs dispersed in butylamine with a given amount of MAI
and PbI2 for hybrid film fabrication. Through this approach,
MAPbI3 perovskite epitaxially grew on PbS CQDs surface in
situ, and the mass ratio of CQDs in film can be tuned conveni-
ently. Based on this method, PbS CQDs integrated with dif-
ferent kinds of perovskite materials, such as MAPbI3, BA2PbI4,
and CsPbBr3, have been reported.[12,41,42]
In the CQD:MHP composite films, CQDs are responsible
for absorbing IR and generating photo-carriers, and the MHPs
are responsible for transferring photo-carriers. IR CQDs and
MHPs usually form type-I bulk-heterojunction in composite
due to the narrow bandgaps of IR CQDs. The photo-generated
carriers in CQDs need to overcome the energy band barriers to
transfer to MHPs. De Arqueretal. designed a photodiode based
PbS CQD:MAPbI3 composite (Figure 3c–e).[43] PbS CQDs and
MAPbI3 formed type-I heterojunction, and the photo-carriers
were confined in PbS CQDs without external bias. They applied
a certain reverse bias on the photodiodes, then the photo-gener-
ated electrons and holes were both extracted out by field assisted
tunneling (Fowler-Nordheim tunneling) due to large electrical
field in composite film. The photodiode shows D* as high as
4×1012 Jones and exhibited large bandwidth of 60 kHz (Figure3f).
Ning group further developed the QDiP to nanowires (NWs)
structure.[44] The use of DMF as a solvent induced the growth of
NWs and the PbS CQDs acted as seeds that accelerated a uni-
form nanowire growth. Edge states and anisotropy properties of
nanowire material provides an attractive platform for infrared
photodetectors.
Figure 3. a) Schematic of quantum-dots-in-perovskite(QDiP). b) TEM image of QDiP. c) Structure of QDiP based photodiode. Reproduced with per-
mission.[40] Copyright 2015, Springer Nature. d) Absorption spectra of pure CQDs, pure perovskite and QDiP. e) Energy band diagram of QDiP based
photodiode. f) Detectivity of QDiP based photodiode. Reproduced with permission.[105] Copyright 2017, Springer Nature.
Small 2020, 2003397

2003397 (6 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
3.2. Composite Material with Organic Semiconductors
Organic semiconductors possess the advantages of envi-
ronment-friendly and flexibility, leading to multitudinous
applications in optoelectronic devices, especially flexible
devices.[45–47] Devices based on CQDs combined with organic
semiconductors have low dark current due to the small
intrinsic carrier concentration of organic semiconductors.[48]
What is more, the devices based on CQD:organic semicon-
ductor composites have better flexibility than the pure CQD
based devices.
The CQDs and organic semiconductors composites have
been reported since the CQD based IR photodetector appeared.
Sargent group mixed PbS CQDs into poly(2-methoxy-5-(ethyl-
hexyloxy)-1,4-phenylene vinylene) (MEH-PPV) and applied this
composite into IR photodetectors in the following work.[49,50]
The internal quantum eciency of the device was ≈5 × 10−6 to
10−5 charges/photon under 5V bias. CQDs are combined with
other organic materials such as fullerene,[51,52] [6,6]-phenyl-
C61-butyric acid methyl ester (PCBM),[53,54] poly(3-hexylthio-
phene) (P3HT), poly(N-vinylcarbazole) (PVK),[55] multiwalled
carbon nanotubes (MWCNTs),[56] and graphene[57] have been
reported. As mentioned above, the dark current of CQD:organic
semiconductor based photodetectors are extremely low due to
the small intrinsic carrier concentration in organic semicon-
ductors, which is beneficial for high detectivity. The dark cur-
rent of PbS CQD:PCBM based photoconductor was lower than
10−9 A, but the photocurrent approaching 5 × 10−6 A, achieving
high detectivity.[53]
The CQD:organic semiconductor composite based photo-
detectors exhibit excellent flexibility, which have been widely
used in flexible devices.[52,56,58] Gaoet al. fabricated a wearable
and sensitive heart-rate detector based on PbS CQD:MWCNTs
photoconductor, in which the PbS CQDs are anchoring on
MWCNTs though stirring (Figure4a).[56] MWCNTs have
excellent flexibility and carrier mobility and the detectivity of
this photodetector reached 3.25 × 1012 Jones. The heart-rate
detecting based on this device shows comparable performance
with commercial detector (Figure 4b,c). After 10 000 times
bending cycles, the performance is still maintained.
Yoo et al. applied a direct writing technology to fabricate
single PbS CQD:P3HT composite nanowire (r≈ 250 nm and
l≈ 50 µm) on Au–Al electrodes (NWPD) (Figure4d), and it
obtained high detectivity in UV–vis–NIR range.[58] Under
extreme (up to 100%) and repeated stretching (up to 100 cycles),
it kept nearly identical photo-response (Figure4e,f).
3.3. Composite Material with Nanoparticles
Thanks to the rapid development of nanomaterial synthesis, a
large number of metal nanoparticles(NPs) are prepared. Due to
their solution-processability, NPs can mix with CQDs homoge-
neously. NPs in CQD films can prolong carrier life time, which
can form local electrical field to improve carriers transport, and
then increase the gain of the device.
Tang group first introduced Au NPs into PbS CQD based
photoconductors.[59] Au NPs have high conductivity, deep
Figure 4. a) Schematic of the anchoring of PbS CQDs on MWCNT. b) Heart-rate detector based on PbS CQD:MWCNTs photoconductor. c) Perfor-
mance of PbS CQD:MWCNTs photoconductor based heart-rate detector. Adapted with permission.[56] Copyright 2014, AIP Publishing. d) Schematic of
a stretchable NWPD array of single hybrid nanowire arches on Au–Al electrodes embedded in PDMS. e) Photograph of the NWPD array and a series
of optical microscopy images showing stretching of up to 100%. f) The photoresponse of the NWPDs during stretching of up to 100%. Reproduced
with permission.[58] Copyright 2015, Wiley-VCH.
Small 2020, 2003397

2003397 (7 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
work function, and can form ohmic contacts between adja-
cent PbS CQDs, which promote carrier transport. Hence, the
introduction of Au NPs in PbS CQD film could decreases
carrier transit time, benefitting for gain enhancement.
Ag NPs were also doped into PbS CQD films and brought
an obvious enhancement of photodetector.[60] The photocon-
ductor showed lower dark current and higher photocurrent
than the pure PbS CQD photoconductor, which overcame the
tradeo between dark current and photocurrent (Figure5a).
The detectivity of PbS CQD:Ag NPs photoconductor was
about three time as much as pure PbS CQD photoconductor.
They attributed the better performance to the dierent roles
of Ag NPs in dark and light. In dark, the Ag NPs sever as
energy barriers and hole scattering centers (top panel in
Figure 5b), suppressing the dark current; in light, Ag NPs
could trap the electron considering its deeper work function
(bottom panel in Figure 5b), prolonging the trap lifetime
of holes. As described previously, the gain is defined as the
ratio of trap lifetime and transit time, thus the longer trap
lifetime can increase gain.
Konstantatos group also found that doping Ag NPs into PbS
CQD films can reduce dark current, but they attributed it to
the shallow work function of Ag NPs, which were considered
as electron donors, suppressing the p-type conductivity of PbS
CQDs.[61] By controlling the ratios of PbS CQDs and Ag NPs,
net doping of the resulting PbS CQD film can be manipulated.
This approach also enhances passivation of mid-gap trap states.
As a result, detectivity enhancement of 100% was achieved at
the optimum Ag NPs concentration about 1:0.05 (Figure5c).
In addition to metal NPs, ZnO NPs, CdS CQDs, and PbSe
CQDs were integrated with PbS CQD films as well. Huang
group added ZnO NPs into PbS:P3HT:PCBM films and found
the similar function with metal NPs (Figure 5d).[54] The photo-
diode exhibited high gain with ZnO addition because of the
electrons trapping eect of ZnO NPs (Figure 5e,f ). Hu et al.
found that adding 5% PbSe CQDs in PbS CQD film could
induce 50% improvement of carrier mobility of film, and the
photodetectors showed higher response than pure PbS CQD
based devices.[62]
4. Device Structure Engineering
The general photodetector structure can be categorized into
photoconductors (PC), photodiodes (PD), and phototransis-
tors (PT) (Figure6a–c). Apart from improving the proper-
ties of photoactive layer of CQD based photodetectors, device
structure designing is explored to improve the photodetector
performance as well. By introducing extra layers to form junc-
tion between CQD layers or helping carriers transport, the
performance of photodetectors is remarkably increased. In
this section, the progresses of the integration of CQD layer
and other material layer in dierent type of photodetectors are
summarized.
Figure 5. a) Schematic demonstration of PbS CQD:Ag NPs based photoconductor. b) Proposed working mechanism in the dark and under illumi-
nation. Reproduced with permission.[106] Copyright 2014, American Chemical Society. c) Detectvity of PbS CQD:Ag NPs based photoconductor with
dierent amount of Ag NPs. Reproduced with permission.[107] Copyright 2015, Wiley-VCH. d) Device structure of PbS:P3HT:PCBM:ZnO photodiode.
e)Energy diagram of PbS:P3HT:PCBM:ZnO photodiode under reverse bias and illumination. f) EQE of dierent blend film based photodiodes. Repro-
duced with permission.[54] Copyright 2014, Wiley-VCH.
Small 2020, 2003397

2003397 (8 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
4.1. Photoconductors
Conventional photoconductors are planar devices consisting
of two electrodes and a photoactive channel between them
(Figure6a). The advantage of photoconductor structure is the
large gain (G). Gain in photoconductor is attributed to sec-
ondary photocurrent originating from the injection and transit
of carriers from the device electrodes (Figure 6b). It can be
calculated by Equation(7)[63]
;
lt
tt
lt
2tt
2
G
V
l
l
F
l
V
τ
τ
τµ τ
µµ
== ==




 (7)
where τlt is the lifetime of the majority carrier, and τtt is the
transit time of the majority carrier, μ is mobility, l is the dis-
tance between the electrodes, V is bias and F is electric field.
If τlt> τtt, gain can be generated. However, the large τlt of photo-
conductors could lead slow response speed. The large dark
current of photoconductors is one important factor limiting
the detectivity of photoconductors. Several strategies have been
reported to suppress dark current.
Sargent group deposited a MoO3 layer on the traditional
PbS CQD based photoconductor to form a photojunction field-
eect transistor (photo-JFET) (Figure7a).[64] In dark, the PbS
CQD channel is fully depleted, attributing to the work func-
tion dierence between MoO3 (Φ= 5.4eV) top contact and the
n-doped PbS CQD film (Φ= 4.2eV), as well as the low carrier
density of the CQD film (n0≈ ND≈ 1016 cm−3) and small thick-
ness (Figure 7b). Therefore, the dark current of photo-JFET
is suppressed comparing with traditional photoconductor.
Under illumination, the photogenerated holes draw toward
the MoO3 and electrons are filling the channel, giving rise to
much increased current. In the photo-JFET, a desirably small
τtt can be obtained without unacceptably increasing the dark
current.
In 2017, Ren et al. reported a bilayer PbS CQD based
photo conductors, the channel of which were consisted of two
layers consistent with tetrabutylammonium iodide (TBAI)
and 1, 2-ethanedithiol (EDT)-modified PbS CQDs, respec-
tively (Figure 7d).[65] Due to the formation of the depletion
region at the PbS-TBAI/PbS-EDT interface, the dark current
of bilayer photoconductors were obviously lower than single-
layer photoconductors (Figure 7f). The photocurrent of the
bilayer photoconductor was higher than the PbS-TBAI photo-
conductor, which can be explained by the low recombination
rate attributing to the spatial separation of photogenerated
electrons and holes (Figure 7e). The detectivity of bilayer
photo conductor was 1.71 × 1012 Jones, much higher than
single-layer devices.
In 2018, Weiet al. reported a PbS CQD/organic photocon-
ductor, in which a p-type and n-type organic semiconductor
composite film was deposited on the top of CQD films to
form the hybrid structure.[66] Owing to heterojunction between
organic material and CQD films, the channel material is
depleted and dark current can be suppressed. Photocarriers
are spatial separated as above bilayer photoconductor that can
prevent recombination and lead to enhanced photocurrent. The
detectivity of the device was as high as 1.12 × 1013 Jones.
In 2019, Zhang et al. adopted the similar strategy to
improve the performance of CQD photoconductors. They
deposited a MAPbI3 perovskite layer on the PbS CQD based
photoconductor to suppress dark current and enhance
the photocarrier separation to obtain higher photocurrent
(Figure7g,h).[67] The detectivity of device are 4.9 × 1013 Jones
at 365nm and 3.0 × 1011 Jones at 940nm, much higher than
the intrinsic PbS CQD photoconductors. A 10 × 10 PbS
CQD/perovskite based photoconductors array was fabricated,
Figure 6. Schematic images of conventional device architectures and their energy band diagrams. a,b) Photoconductors. c,d) Photodiodes.
e,f)Phototransistors.
Small 2020, 2003397

2003397 (9 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
which constituted a high-performance broadband image
sensor (Figure7i).
Very recently, flexible photodetector device based on CQD/
graphene hybrid photoconductor was reported for wearable fit-
ness monitoring.[68] These prototype wearable devices are able
to monitor vital health signs noninvasively, including heart rate,
arterial oxygen saturation (SpO2), and respiratory rate.
4.2. Photodiodes
Conventional photodiodes are vertical devices with rectification
eect mainly composed of electron transfer layer (ETL), hole
transfer layer (HTL) and photoactive layer (Figure6c). Photo-
generated carriers in photodiodes are separated by the built-
in electric field in device and carrier collections are usually
assisted by revers bias (Figure6d). Photodiodes usually have no
gain, but high gain can be achieved in photodiodes based on
charge tunneling injection induced by traps.
In 2015, Lee et al. introduced traps in the electron
blocking layer instead of in IR sensitive layer, higher gain
was realized.[69] The band diagrams at dark and under illu-
mination of devices are shown in Figure8a,b, the TAPC
(1,1-bis[(di-4-tolylamino) phenyl]cyclohexane) layer serving as
electron blocker. The traps were introduced in TAPC layer
by the Ag atoms penetration during thermal evaporation
process. Similar with Huang’s work,[54] the energy bands of
TAPC are bended due to the hole trapping under IR irra-
diation, and then triggered electron tunneling injection.
The gain of this device was 187 and the detectivity reached
7 × 1013 Jones at 1200 nm, higher than the commercial
InGaAs photodetector (Figure8c).
Figure 7. a) Schematic of the photoJFET. b) In dark, PbS CQDs layer is fully depleted by MoO3 layer. c) Under illumination, photogenerated carriers
are filled with depleted PbS CQDs layer. Reproduced with permission.[108] Copyright 2015, American Chemical Society. d) Schematic of the bilayer photo-
conductor. e) Energy band diagram of the bilayer photoconductor. f) I–V curves of three kinds of photoconductors. Reproduced with permission.[65]
Copyright 2017, Wiley-VCH. g) Schematic of PbS CQD/MAPbI3 photoconductor. h) Schematic energy band diagram and optoelectronic processes
in the PbS/MAPbI3 layers. i) Imaging illustration of PbS CQD/MAPbI3 photoconductor. Reproduced with permission.[67] Copyright 2019, American
Chemical Society.
Small 2020, 2003397

2003397 (10 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
In 2020, Zhou et al. reported a high gain PbS CQD based
photodiode structure. Instead of adding Ag atoms into electron
blocking layer, they added a certain amount of Ag NPs into ZnO
layer (electron transport layer).[70] The Ag NPs in ZnO layer would
trap electrons under illumination, and induce ZnO layer energy
band bending (Figure8d,e). The holes could inject into CQD layer
by tunneling eect giving rise to the increase of gain. As a result,
the photodetector showed an EQE of 8000%, a detectivity of 6 ×
1012 Jones, and a response time of millisecond level at around
1500nm, which was on par with the best previously-reported PbS
based infrared photodetectors of similar bandgap. (Figure8f).
Photodetector structure engineering is not limited to achieve
high gain. Tang et al. reported a dual-band infrared photo-
detector based on HgTe CQDs.[71] The dual-band photodiode
structure is shown in Figure 8g. Two kinds of dierent size
HgTe CQDs were used as SWIR and MWIR active materials,
respectively. They added Ag2Te NCs between two kinds of
dierent size HgTe CQDs layers to achieve p doping for HgTe
CQDs, and applied Bi2Se3 NCs as n-type layer. A n-p-n config-
uration was constructed (Figure8h). Under negative bias, the
SWIR diode works; under positive bias, the MWIR diode works
(Figure8i). By controlling the bias polarity and magnitude, the
detector can be rapidly switched between SWIR and MWIR at
modulation frequencies up to 100kHz with D* above 1010 Jones
at cryogenic temperature. Furthermore, they applied this dual-
band photodetector in infrared imaging and remote tempera-
ture monitoring, displaying good performance.
4.3. Phototransistors
Conventional phototransistors are three-terminal devices con-
sisting of source, drain, gate, and photoactive channel (Figure6e).
Comparing with photoconductors, the conductivity of the
Figure 8. Schematic energy band diagrams of a gain PbS CQD based photodiode a) in the dark and b) in the IR illumination. c) The function of Ag
layer thickness and gain. Reproduced with permission.[109] Copyright 2015, Wiley-VCH. d,e) Energy band diagrams of the PbS CQD based photodiodes
with or without Ag NPs addition. f) Detectivity of PbS CQD based photodiodes with Ag NPs addition. Reproduced with permission.[70] Copyright 2020,
Springer Nature. g) Device structure of dual-band HgTe CQD based IR photodiode. h) Energy band diagram of n-p-n configuration. i) Spectral response
of the dual-band photodiode under a bias voltage from positive (+500mV) to negative (−300 mV) voltage at 85 K. Reproduced with permission.[71]
Copyright 2019, Springer Nature.
Small 2020, 2003397

2003397 (11 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
channel can be control by the gate voltage and induce lower dark
current and higher photo-response.[72] Because the gain is directly
proportional to the mobility of the charge carrier in channel, the
CQD films are integrated with high carrier mobility materials,
such as graphene and transition metal dichalcogenides (TMD).
The photo-generated carriers in CQD layer will transfer to high
carrier mobility material and to be extracted by source and drain
electrodes under VDS bias (Figure6f).
4.3.1. CQD/Graphene Hybrid Phototransistors
Graphene is an attractive material for optoelectronics due to
the large carrier mobility.[73,74] CQD/graphene hybrid phototran-
sistors have been fabricated, in which graphene plays a role of
channel material while CQDs as infrared absorption layer. As
the gain is inverse proportional to the transit time of photo-
carrier, large mobility can lead to high gain.
In 2012, Konstantatos et al. reported a hybrid PbS CQD/
graphene phototransistor using mechanically graphene
(Figure9a).[75] The combination of PbS CQD and graphene
resulted in ultrahigh gain (108) and ultrahigh detectivity of
7 × 1013 Jones. Built-in field was generated between the PbS
CQDs and graphene. Under light illumination, photo-carriers
are separated and photo-induced holes can inject to the gra-
phene (Figure9b). As graphene has large mobility, the transit
time for holes to across the channel is very fast, shorter than
carrier life-time. The holes can circulate in the device for many
Figure 9. CQD/graphene hybrid phototransistors. a) Schematic of the hybrid PbS CQD/graphene phototransistor. b) Energy band diagram of the
graphene/CQDs interface. c) Spectral responsivity of two phototransistors based on exciton peaks at either 950nm (top panel) or 1450nm (bottom
panel) PbS CQDs. Reproduced with permission.[110] Copyright 2012, Springer Nature. d) Sandwich structure of graphene/PbS CQD/graphene hybrid
phototransistor. e) Fabrication flow of the laser shock imprinting enabled graphene/PbS CQD/graphene seamless hybrid structure. f ) I–t curve of gra-
phene/PbS CQD/graphene hybrid phototransistor after the laser shock process at the gate voltage of 20V. Reproduced with permission.[111] Copyright
2017, American Chemical Society. g) STM analysis of the decorated graphene. h) Transfer curves of the hybrid PbS CQD/graphene photodetector under
various illumination powers. i) The specific detectivity of the hybrid photodetector as a function of the gate voltage for fixed driving voltages of 1 and
0.1V under of 1550nm illumination. Reproduced with permission.[77] Copyright 2018, Wiley-VCH.
Small 2020, 2003397

2003397 (12 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
times and result in high gain. The dark current can be reduced
by applying a gate voltage to close the conductive channel. The
photodetector has a response time of ≈10ms level.
The interfacial resistance between graphene and CQDs is
generally large, which impede carrier transfer from CQDs
to graphene. A sandwich structure of graphene/PbS CQD/
graphene fabricated by laser shock imprinting was explored
to enable close contact between graphene and CQDs
(Figure9d,e).[76] The laser shocking imprint method tuned the
morphology of graphene to perfectly wrap on CQDs, allowing
more eective carrier transfer between them (Figure9f).
Besides, graphene is generally p-type doped due to the
external environment and transfer process. A rather large
gate bias over 20 V is needed to tune the fermi level of gra-
phene, which is much larger than the maximum output voltage
of current ROIC technology. To overcome this problem, N,
S co-decorated graphene fabricated by chemical vapor deposi-
tion (CVD) was also deposited with PbS CQDs to fabricate a
hybrid phototransistor with CMOS compatibility (Figure9g).[77]
P-type doping can be eectively inhibited by N, S co-decorating.
As a result, the required gate voltage can be reduced to less than
3.3 V that is compatible with CMOS technology (Figure9h).
The hybrid phototransistor exhibits ultrahigh responsivity of
>104A W−1 and a detectivity on the order of 1012 Jones with a
low driving voltage of 1V(Figure9i).
Despite CQD/graphene hybrid phototransistor have large
EQE and responsivity, the response time is slow. In order
to overcome the limitations of a phototransistor in terms of
speed, and increase linear dynamic range, one more complex
structure integrating CQDs photodiode with graphene tran-
sistor was fabricated.[78] In this structure, ITO top transparent
electrode is cathode while graphene is the anode as well as the
channel material of 2D transistor. A voltage is added on the
ITO electrode of the photodiode to enlarge the depletion layer
and improve charge transfer from CQD film to graphene. As
a result, when operating as a phototransistor, large gain can
be achieved. The devices operated in the Vis, NIR, and SWIR
ranges and showed an EQE as high as 70–80%, a sub-milli-
second temporal response, a gain-bandwidth product on the
order of 108 and a linear dynamic range in excess of 110 dB.
This highly performing detector also exhibited very high sensi-
tivity with experimentally measured D* of 1 × 1013 Jones.
4.3.2. CQD/TMD Hybrid Phototransistors
Except graphene, other 2D materials such as TMD are used for
hybrid phototransistors as well. Compared to graphene, TMD
2D materials have intrinsic bandgaps leading to low dark cur-
rent. In an early report, Kuferetal. reported a PbS CQD/MoS2
phototransistor structure.[79] the detector has the potential to
reach high sensitivity in the shot noise limit with D* shot-noise
limit reaching up to 7 × 1014 Jones at Vg of −100V.
In the CQD/TMD phototransistors, photo-carriers in CQDs
should be transferred to TMD from CQD films, which is
aected by the band alignment of the CQDs and TMD. Paketal.
fabricated a type-II junction upon the MoS2 (Figure10a). The
junction consists of PbS-TBAI and PbS-EDT layer, and its func-
tion is separating carriers spatially and reduces carrier recom-
bination. The CQD/MoS2 phototransistor with type-II junction
Figure 10. CQD/TMD hybrid phototransistors. a) Device structure of CQD/MoS2 phototransistor. b) Illustration of the band alignments of the MoS2/
PbS-TBAI and MoS2/PbS-TBAI/PbS-EDT structures. c) Temporal response of the photocurrent of two devices under 850nm laser illumination. Repro-
duced with permission.[112] Copyright 2018, American Chemical Society. d,e) Energy band diagrams and responsivity of CQD/WS2 and CQD/MoS2
phototransistors. Reproduced with permission.[113] Copyright 2019, American Chemical Society.
Small 2020, 2003397

2003397 (13 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
showed higher photocurrent (Figure10b,c). The response time
was also decreased to a value of 0.95ms, which is fast among
the 0D/2D hybrid phototransistors.
Konstantatos group reported a WS2 based 0D/2D hybrid
phototransistor, extending the sensitivity from 1.5 to 2 µm.
Compared with MoS2, WS2 has much deeper energy band,
which leads lower barriers for the transfer of photogenerated
electrons. PbS CQD/WS2 exhibits type-II junction benefiting
carriers transfer (Figure 10d). Therefore, the response and
detectivity at 1.8 µm under 1 V bias of PbS CQD/WS2 photo-
transistor was as high as 1400 A W−1 and 1012 Jones, respectively
(Figure10e).
5. Scalable Device Structure of CQD IR
Photodetectors
Generally, it is dicult to apply independent CQD based IR
photodetectors for imaging. Fabricating photodetector arrays or
integrating with other devices can greatly enrich the photodetec-
tor’s application scenarios in imaging. Focal plane arrays (FPAs)
are the most promising IR imaging technique which integrates
the photodetector arrays with readout circuits (ROICs). Several
works about the integration of CQD based IR photodetec-
tors with commercial ROICs have been reported.[80–84] Since
most of commercial ROICs are based on silicon, the integra-
tion of CQD based IR photodetectors with silicon can simplify
the fabrication processes for IR FPAs detectors and reduce
cost.[85–88] Integrating with LED for upconversion is another
attractive application of CQD based IR photodetectors. Upcon-
version devices can convert the invisible IR into visible light
without ROICs and have made significant progress in recent
years.[70,89–92] This section will discuss the progress of scalable
application of CQD based IR photodetectors in details.
5.1. Integration with Commercial ROICs for IR Imaging
In 2009, Tobiasetal. reported an CQD based IR imaging sensor
which consisted of PbS CQD:organic bulk heterojunction photo-
diodes and commercially available a-Si active matrix (AM) thin-
film transistor (TFT) panels (Figure11a).[93] This image sensor
enabled NIR imaging up to 1.8 µm, with rectification ratios of
≈6000 and EQE up to 51%. The sensor could take NIR images of
insects (Figure11b) and the maximum resolution of it reached
to ≈3 lp mm−1 (line pairs per millimeter). However, the normal-
ized detectivity of this device was only 2.3 × 109 Jones due to
the poor performance of PbS CQD based photodiodes. In 2017,
Konstantatos group fabricated a PbS CQDs, graphene and
CMOS integration image sensor (Figure11c).[84] The graphene
was transferred to the top of CMOS and patterned to pixels,
and then the PbS CQDs were spin-coated on the graphene. The
mechanism of this image sensor is similar with the CQD/gra-
phene hybrid phototransistor discussed in Chapter 4.3.1. Photo-
generated carriers in PbS CQDs layer transfer to graphene and
then are extracted out by the bias voltage applied between the
two pixel contacts (Figure 10d). Because of the high mobility
of graphene, this photoconductor structure exhibits ultrahigh
gain of 108 and responsivity above 107 A W−1, which is a great
improvement compared with photodetectors and imaging
systems based on CQDs only. The normalized detectivity was
above 1012 Jones. The image sensor was with high resolution
(388 × 288 pixels) and was used in IR imaging under external
IR irradiation (Figure11e).
Figure 11. CQD based photodetectors integrated with commercial ROICs. a) Schematic of the imager with an a-Si AM backplane and an PbS CQD
based photodiode. b) Infrared shadow cast at 1310nm of a slide showing a monarch butterfly. Reproduced with permission.[80] Copyright 2009,
Springer Nature. c) Computer-rendered impression of the CVD graphene transfer process on a single die containing an image sensor read-out
circuit that consists of 388 × 288 pixels. d) NIR and SWIR light photograph of an apple and pear. Reproduced with permission.[84] Copyright 2017,
Springer Nature.
Small 2020, 2003397

2003397 (14 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
In addition, other eorts in the integration of CQD based
photodetectors and ROICs have been reported. Klem and
coworkers fabricated PbS CQD/C60 photodiodes on passive
“fan-out chips,” directly measuring individual photodetector
with structures representative of commercial ROICs without
integrated amplification or signal processing.[82] The average
detectivity of device was 1 × 1012 Jones at room temperature and
3 × 1012 Jones at −20 °C, which was comparable with commer-
cial InGaAs photodetectors.
5.2. Integrated with IGZO TFTs
Amorphous metal oxide semiconductor-based thin-film
transistors (TFTs) have attracted much attention due to
their high mobility, high transparency and low o current
level. Indium-gallium-zinc-oxide (IGZO) TFTs have been
widely used in large scale active matrix backplanes as a pixel
driver for high-resolution liquid crystal displays and organic
light-emitting diodes display. Hwanget al. suggested a PbS
CQD/IGZO hybrid transistor and developed a prototype IR
image sensor.[94] In this structure, PbS CQDs are light absorp-
tion layer and IGZO is the channel material. Photocarriers
in PbS CQDs can easily transferred to IGZO channel, which
has large mobility and high gain can be generated. The
hybrid phototransistor exhibited a responsivity of 106 A W−1
and a detectivity of 1013 Jones at 1000 nm. By connecting a
load resistor to the unit phototransistor, pronounced output
voltage signals with a high output voltage photo-gain of ≈4.9V
at close to 99% were shown.
Kim group then fabricated a skin-like, large area, flex-
ible, pixelated and full color CQD/IGZO hybrid photo-
transistor.[95] By direct photo-patterning approach, the
phototransistor pixels were fabricated to form a high-reso-
lution image sensor that is capable of realizing position-
dependent full-color photodetection. The device structure is
CMOS-compatible, which can response to variable stimuli
simultaneously and selectively, showing a facile route for
bio-imaging applications.
5.3. Integrated with Silicon
The fabrication of junction between CQD and silicon could
allow the direct combination of CQDs with ROIC structure
without intermediate layers.[96–99] However, the energy band
misalignment and the lattice mismatch for IR material and
silicon imped the performance improvement of devices. The
tunable bandgaps and solution phase synthesis of CQDs are
potential to address these problems.
In 2015, Masala et al. designed a PbS CQD/p-Si hetero-
junction photodiode. They introduced CH3I to passivates
the silicon and CQD terminations.[100] Further, an inter-
face dipole was introduced to the interface, which pro-
duced a favorable band alignment between the two solids.
The photo diode exhibited a high normalized detectivity of
1.5×1011 Jones at 1230nm under 7V reverse bias.[85] Xuetal.
reported an inverted PbS CQD/Si heterojunction photodiode
(Figure12a).[88] Inverted structure with p-type PbS CQD/n-
type Si heterojunction has better energy band alignment
than n-type PbS CQD/p-type Si structure, hence the inverted
photodiode showing broader detecting wavelength range,
high response and ultra-low working bias. It obtained a high
normalized detectivity of 1.47 × 1011 Jones at 1540nm without
reverse bias (Figure 12b), and the bandwidth of it reached
29.8kHz.
In order to further passivate the silicon surface and tune
the energy band alignment, Xiao etal. inserted solution pro-
cessed ZnO nanoparticles between silicon and PbS CQD film
(Figure 12c).[87] Via applying ZnO nanoparticles, the surface
dangling bond of silicon was remarkably suppressed, which
reduced carrier recombination. What is more, the valence band
Figure 12. Device structures and performance of dierent kinds of PbS CQD/silicon photodetectors. a,b) Inverted PbS CQD/silicon
photodiode. Reproduced with permission.[88] Copyright 2020, American Chemical Society. c,d) ZnO passivated PbS CQD/silicon photodiode.
Reproduced with permission.[87] Copyright 2020, AIP Publishing. e,f) PbS CQD based PVFET. Adapted with permission.[86] Copyright 2017,
Springer Nature.
Small 2020, 2003397

2003397 (15 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
structure aligns well between CQDs and silicon, enabling eec-
tive carrier extraction under low bias. As a result, the photo-
detector shows a detectivity as high as 4.08 × 1011 Jones under
−0.25V bias and the longest wavelength is up to 1600 nm
(Figure12d).
Sargent group designed a photovoltage field-eect transistor
(PVFET) as shown in Figure12e, which used silicon for charge
transport and PbS CQDs for IR absorber.[86] A lightly doping
p-type silicon channel was epitaxially grown on an heavy doped
n-type silicon substrate that acted as a gate. The source and
drain were deposited on the p-type silicon channel with alu-
minium forming ohmic contact. In the PVFET, gain is adjusted
by tuning the doping of the channel and the eect of the gate
allows high gain at low dark current. Gain based on photo-
voltage and transconductance is distinct from photoconductive
type. It enables simultaneously high signal amplification and
rapid response (Figure12f). In this work, the Si/CQD PVFET
showed high gain (>104) even in the infrared (wavelengths
of >1500 nm), high speed (100 kHz) and contained dark cur-
rent (10−1–101 A cm−2). The detectvity was measured as 1.8 ×
1012 Jones.
5.4. Integrated with LED for Upconversion
Flip-chip hybridization bonding process to bond IR photo-
detector with ROIC one to one is required for traditional IR
image cameras, which increases the fabrication costs and limits
device’s area and resolution. An alternative method is to use an
upconversion device to convert infrared light into visible light
and then image it with a commercial silicon-based CCD or
human eyes. In this case, the upconversion devices do not need
discrete pixels, especially the expensive indium pillar welding
process is avoided.
Photodetector-light-emitting-diode upconversion is one
promising method widely used in the infrared imaging
field. The device consists of a photodetector layer and a light
emitting diode (LED) layer both of which are stacked and
connected in series. The infrared photons are converted into
electrical signals in photodetector layer, and then the electrical
signals inject into LED layer and induce visible light. The
most important figure of merits for upconversion device is
photon-to-photon conversion eciency (ηcon) that can be cal-
culated by Equation(7)
η
λλ
λλ
λ
)
)
(
(
=
∫d
con
photons
IR IR
I
Rhc
P
hc
(7)
So group is the pioneer in CQD based photodetector-
light-emitting-diode upconversion field. In 2011, they reported
an upconversion device by integrating an PbSe CQD material
with a green organic LED, which can convert infrared photons
(1300nm) to green photons (520 nm) with a photon-to-photon
eciency of 1.3%.[101] In this device, a ZnO blocking layer is key
to keep the device o in the absence of IR excitation. This is the
first demonstration of upconversion device integrating CQDs
photodetectors with LEDs (Figure13a).
Based on this structure, they further developed an imaging
system by combining the upconversion device with a commer-
cially-available visible CCD.[102] In the former structure, the
light emits from the bottom and large ratio of emitting visible
light will be reabsorbed by the CQD film. In order to avoid the
reabsorption, top emitting diode used a semitransparent elec-
trode. An IR pass visible mirror layer was inserted to increase
the intensity of emitting light. The reflective top emitting
device showed a p-p eciency of 4.1% while the transparent
device without mirror layer showed a p-p eciency of 1.1%.
With the aids of a 1200nm IR flash light, photo images can
be clearly taken in dark environment using this upconversion-
CCD imaging system (Figure13d).
Figure 13. a
–c) Device structures of dierent kinds of upconversion devices. Reproduced with permission.[101] Copyright 2011, American Chemical
Society. d) Upconversion-CCD imaging system. Reproduced with permission.[90] Copyright 2014, Springer Nature. e) Bio-imaging application
of upconversion device. Reproduced with permission.[70] Copyright 2020, Springer Nature. (a) Reproduced with permission.[89] Copyright 2011,
American Chemical Society. (b) Reproduced with permission.[92] Copyright 2019, AIP Publishing
Small 2020, 2003397

2003397 (16 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
Zhou et al. recently reported a two terminal upconver-
sion photodetector based on CQDs that was fabricated by
solution process, which would save the fabrication cost
and induce low working voltage (Figure 13c).[70] PbS CQDs
(300–1600 nm) were used as IR absorption layer of the photo-
detector and luminance CdSe/ZnS CQDs were adopted as
light emitting layer. Due to the high gain of the IR detector
part (The mechanism of IR photodetector is discussed in
Chapter 4.2.), the p-p eciency of the upconversion device
achieved 6.5% while the turn on voltage was ultralow of
−2.5V. In order to show its application, flexible devices were
fabricated and showed considerable performance. Bio-imaging
application was demonstrated by using this upconversion
detector and cancerous issue could be distinguished under IR
illumination (Figure13e).
Sun and coworkers reported a solution processed CQD
upconversion device using a Schottky junction structure
(Figure13b).[92] PbS CQDs was deposited on the top of a red
QLED. Under dark condition, electron injection is prevented
by the PbS layer due to large barrier between metal electrode
and PbS CQDs. Under IR illumination, photo-generated
electrons can be transferred to ZnO layers and then trans-
ferred to the emitting zone, where the electron and holes
recombined and emitted visible light. The upconversion
devices showed a p-p eciency of 3.35% and a luminance
on–o ratio of ≈8 × 103.
Although the successfully demonstration of upconver-
sion device, the p-p eciency was still low, which limits
their practical application. In order to address this issue, So
and coworkers further demonstrated an infrared-to-visible
upconversion light-emitting phototransistor (LEPT) with high
gain.[103] Unlike upconversion devices that combines a photo-
diode and LED which is a two-terminal device, the LEPT is
a three-terminal device, in which the photodetector part is a
vertical phototransistor where a PbS CQD photodiode is the
photogate of the transistor and the current passing through
channel material drives the OLED (Figure14a). The conduc-
tivity of the channel material can be controlled by the photo-
voltage generated by the CQD photodiode: under zero bias
and without light illumination, carrier injection is inhibited
due to large carrier barrier; under IR light illumination, the
photovoltage causes electron accumulating, which further
causes energy band bending, finally carrier injection to the
channel material happens and the conductivity of channel
material is increased (Figure 14b,c). This work mechanism
makes the LEPT can have a large detectivity of 1.2 × 1013 Jones
Figure 14. LEPT upconversion device. a) Device structure of LEPT. b,c) Physical cross-sectional views of the vertical phototransistor in LEPT in dark and
under illumination. d) Luminance transfer curve with and without infrared illumination. The p-to-p external quantum eciency (open stars) reaches
over 1000% at VDS= 10V. e) The p-to-e eciency and e-to-p eciency curves of the LEPT. Reproduced with permission.[91] Copyright 2016, Springer
Nature.
Small 2020, 2003397

2003397 (17 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
and a large p-p eciency of over 1000% at a small IR intensity
(Figure14d,e).
6. Challenges and Perspectives
Although CQD based IR photodetectors have made impres-
sive progress in the past few years and some figures of merits
of them exceeds the commercial IR photodetectors, there are
still some important problems should be investigated further.
Some important challenges:
1. The detecting wavelength of CQD based photodetectors are
concentrated in <2 µm range. Only few of them can extend to
over 2 µm, especially for device can achieve MWIR (>3 µm)
detecting. In addition, the detectivity beyond 2 µm is still low.
Objects emit blackbody radiation at wavelength in MWIR
range, and the photodetectors which are able to detect MWIR
can be applied in night vision. In contrast, generally photo-
detectors with a shorter detection wavelength than MWIR
need external infrared radiation to achieve infrared imaging.
The short detection wavelength of CQD based photodetectors
limit their application scenarios.
2. Although high EQE and detectivity are partly achieved, the
response time of CQD based photodetectors are slow. There
are great demands for high speed devices in science and tech-
nology application, such as information communication and
time of flight imaging. For photodetector based on traditional
semiconductor, ns or faster response time can be achieved.
However, the response time of CQD based photodetector is
only millisecond or microsecond level, which is due to low
mobility and large trap state of CQDs material compared to
traditional single crystal materials.
3. Large area with ultrahigh pixel number is critical to realize
high resolution imaging, however, there are lack of report for
large area devices, and the performance of the device that is
compatible with large scale fabrication is much lower than the
best performance up to now. It is highly desirable to further
improve the performance of devices with scalable structures.
Overall, the performance of CQD IR photodetectors remains
to be significantly enhanced to meet application requirement.
The improvement of carrier mobility of the film plays a key
role to enhance device performance, e.g. responsivity, gain, and
response speed. In addition, the construction of high quality
interface between active materials are critical to ensure eec-
tive carriers extraction and decrease of carrier recombination.
To address the challenges above, more attention need to be paid
to the following aspects.
1. The most frequently used CQDs in IR photodetectors are
lead chalcogenide attributing to high performance. However,
the synthesis of the first exciton peak beyond 2 µm is dicult
due to their relatively larger bandgap (The bandgaps of PbS,
PbSe, and PbTe bulk materials are 0.41, 0.28, and 0.31 eV
respectively.[104]). Mercury chalcogenide CQDs are suitable
for MWIR and LWIR photodetectors, but the investigation
about them is limited by their hypertoxic mercury precursor.
More works need to be conducted to establish a safe synthesis
method for high quality mercury chalcogenide CQDs, and
explore them for MWIR sensitive photodetectors.
2. Develop CQD film with lower defect density and high car-
rier mobility. The carrier mobility of CQD films fabricated by
liquid ligand exchange are higher than solid ligand exchange
process, but it is limited in small size CQD nowadays. De-
velop eective strategies to increase photoactive layer carrier
mobility can enhance response speed of photodetectors and
allow the use of thicker photoactive layer.
3. Fabricate super-lattice CQD films in IR photodetectors. In
ordinary CQD films, the stack of CQDs is disorder, hence
carrier in films have to hoop between CQDs to be extracted.
However, for super-lattice CQD films, the carrier can be trans-
ported not only by hooping, but also by band-like transport,
which leads high carrier mobility. The development of CQD
films with super-lattice structure is a promising strategy to
improve carrier mobility of the film.
7. Summary
This review summarizes the recent progress of CQD infrared
photodetectors, and highlight the progress of using CQDs in
hybrid materials and structures, device engineering and scal-
able device structure fabrication. The performance of CQD
IR photodetectors is growing quickly, and it show promise for
next-generation infrared photodetectors. The integration of
CQDs with other materials to form hybrid materials and struc-
tures reduce defect density and increase carrier transport. By
structure design and control, the photodetector can take advan-
tages of each material to enhance performances. Device engi-
neering based on electrical field manipulation improve carriers
transport, and significantly improve device performance. The
integration of CQD with silicon could enable the convenient
fabrication of large scale devices. The development of upconver-
sion photodetector oers an alternative way for IR imaging that
does not need discrete pixels. In the end, a discussion about
the challenges and possible solutions are presented to shadow
some light for future development of CQD IR photodetectors.
Acknowledgements
The authors gratefully acknowledge financial support from the National
Key Research and Development Program of China (under Grant No.
2016YFA0204000), ShanghaiTech start-up funding, National Natural
Science Foundation of China (Grant Nos. 61935016), and Shanghai key
research program (Grant No. 16JC1402100).
Conflict of Interest
The authors declare no conflict of interest.
Keywords
colloidal quantum dots, hybrid structure, infrared imaging,
photodetectors
Received: June 2, 2020
Revised: July 30, 2020
Published online:
Small 2020, 2003397

2003397 (18 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
[1] X.Pang, L. Zhao, W.Han, X. Xin, Z. Lin, Nat. Nanotechnol. 2013,
8, 426.
[2] X.Dai, Z.Zhang, Y.Jin, Y.Niu, H.Cao, X.Liang, L.Chen, J.Wang,
X.Peng, Nature 2014, 515, 96.
[3] R. Wang, X. Wu, K. Xu, W. Zhou, Y. Shang, H. Tang, H. Chen,
Z.Ning, Adv. Mater. 2018, 30, 1704882.
[4] M. A.Hines, G. D.Scholes, Adv. Mater. 2003, 15, 1844.
[5] A. Sashchiuk, L. Langof, R. Chaim, E. Lifshitz, J. Cryst. Growth
2002, 240, 431.
[6] J. E. Murphy, M. C. Beard, A. G. Norman, S. P. Ahrenkiel,
J. C. Johnson, P. R. Yu, O. I. Micic, R. J. Ellingson, A. J. Nozik,
J. Am. Chem. Soc. 2006, 128, 3241.
[7] S.Keuleyan, E.Lhuillier, P.Guyot-Sionnest, J. Am. Chem. Soc. 2011,
133, 16422.
[8] E.Lhuillier, M.Scarafagio, P.Hease, B.Nadal, H.Aubin, X. Z.Xu,
N. Lequeux, G. Patriarche, S. Ithurria, B. Dubertret, Nano Lett.
2016, 16, 1282.
[9] K. S. Jeong, Z. Deng, S. Keuleyan, H. Liu, P. Guyot-Sionnest,
J. Phys. Chem. Lett. 2014, 5, 1139.
[10] A. Sahu, L. Qi, M. S. Kang, D.Deng, D. J. Norris, J. Am. Chem.
Soc. 2011, 133, 6509.
[11] J. Xu, O. Voznyy, M. Liu, A. R. Kirmani, G. Walters, R. Munir,
M.Abdelsamie, A. H.Proppe, A.Sarkar, F. P.Garcia de Arquer, M.Wei,
B. Sun, M. Liu, O. Ouellette, R. Quintero-Bermudez, J. Li, J. Fan,
L. Quan, P. Todorovic, H. Tan, S.Hoogland, S. O. Kelley, M. Stefik,
A.Amassian, E. H.Sargent, Nat. Nanotechnol. 2018, 13, 456.
[12] M.Liu, Y.Chen, C. S. Tan, R.Quintero-Bermudez, A. H. Proppe,
R.Munir, H.Tan, O.Voznyy, B.Scheel, G.Walters, A. P. T.Kam,
B. Sun, M. J. Choi, S. Hoogland, A. Amassian, S. O. Kelley,
F. P.Garcia de Arquer, E. H.Sargent, Nature 2019, 570, 96.
[13] B.Sun, M. Vafaie, L.Levina, M.Wei, Y.Dong, Y. Gao, H. T.Kung,
M. Biondi, A. H. Proppe, B. Chen, M. J. Choi, L. K. Sagar,
O. Voznyy, S. O. Kelley, F. Laquai, Z. H. Lu, S. Hoogland,
F. P.Garcia de Arquer, E. H.Sargent, Nano Lett. 2020, 20, 3694.
[14] C. R.Kagan, E. Lifshitz, E. H.Sargent, D. V.Talapin, Science 2016,
353, 885.
[15] S. B. Hafiz, M. Scimeca, A. Sahu, D. K. Ko, Nano Convergence
2019, 6, 7.
[16] I. Moreels, Y. Justo, B. De Geyter, K. Haustraete, J. C. Martins,
Z.Hens, ACS Nano 2011, 5, 2004.
[17] F. P. G.De Arquer, A.Armin, P.Meredith, E. H.Sargent, Nat. Rev.
Mater. 2017, 2, 16100.
[18] M.Li, W.Zhang, G.Shao, H.Kan, Z.Song, S.Xu, H.Yu, S.Jiang,
J.Luo, H.Liu, Thin Solid Films 2016, 618, 271.
[19] H.Aqoma, S.-Y.Jang, Energy Environ. Sci. 2018, 11, 1603.
[20] I. J. Kramer, J. C. Minor, G. Morenobautista, L. R. Rollny,
P.Kanjanaboos, D.Kopilovic, S. M.Thon, G. H.Carey, K. W.Chou,
D.Zhitomirsky, Adv. Mater. 2015, 27, 116.
[21] J.Park, J.Joo, S. G.Kwon, Y.Jang, T.Hyeon, Angew. Chem., Int. Ed.
Engl. 2007, 46, 4630.
[22] R. Wang, Y. Shang, P. Kanjanaboos, W. Zhou, Z. Ning,
E. H.Sargent, Energy Environ. Sci. 2016, 9, 1130.
[23] J. M. Luther, M. Law, Q. Song, C. L. Perkins, M. C. Beard,
A. J.Nozik, ACS Nano 2008, 2, 271.
[24] J. W. Jo, J.Choi, F. P. García de Arquer, A. Seifitokaldani, B. Sun,
Y. Kim, H. Ahn, J. Fan, R.Quintero-Bermudez, J.Kim, M.-J.Choi,
S.-W. Baek, A. H. Proppe, G. Walters, D.-H. Nam, S. Kelley,
S.Hoogland, O.Voznyy, E. H.Sargent, Nano Lett. 2018, 18, 4417.
[25] J. Xu, O. Voznyy, M. Liu, A. R. Kirmani, G. Walters, R. Munir,
M.Abdelsamie, A. H.Proppe, A.Sarkar, F. P.García de Arquer, M.Wei,
B. Sun, M. Liu, O. Ouellette, R. Quintero-Bermudez, J. Li, J. Fan,
L. Quan, P. Todorovic, H. Tan, S.Hoogland, S. O. Kelley, M. Stefik,
A.Amassian, E. H.Sargent, Nat. Nanotechnol. 2018, 13, 456.
[26] X. Y.Jiang, H. S.Li, Y. Q.Shang, F.Wang, H.Chen, K. M.Xu, M.Yin,
H. F.Liu, W. J.Zhou, Z. J.Ning, Chem. Commun. 2019, 55, 9483.
[27] P. R. Brown, D. Kim, R. R. Lunt, N. Zhao, M. G. Bawendi,
J. C.Grossman, V.Bulovic, ACS Nano 2014, 8, 5863.
[28] Z.Ning, O.Voznyy, J. Pan, S.Hoogland, V.Adinolfi, J. Xu, M. Li,
A. R.Kirmani, J.Sun, J. C.Minor, Nat. Mater. 2014, 13, 822.
[29] M. J.Choi, F. P. Garcia de Arquer, A. H.Proppe, A. Seifitokaldani,
J.Choi, J. Kim, S. W.Baek, M.Liu, B.Sun, M. Biondi, B. Scheel,
G. Walters, D. H. Nam, J. W. Jo, O. Ouellette, O. Voznyy,
S.Hoogland, S. O.Kelley, Y. S.Jung, E. H.Sargent, Nat. Commun.
2020, 11, 103.
[30] R.Saran, R. J.Curry, Nat. Photonics 2016, 10, 81.
[31] F. P. García de Arquer, A. Armin, P. Meredith, E. H. Sargent,
Nat. Rev. Mater. 2017, 2, 16100.
[32] J.Hu, Y.Shi, Z.Zhang, R.Zhi, S.Yang, B.Zou, Chin. Phys. B 2019,
28, 020701.
[33] Y. Wang, Y. Zhang, P. Zhang, W.Zhang, Phys. Chem. Chem. Phys.
2015, 17, 11516.
[34] L.Dou, Y. M. Yang, J.You, Z.Hong, W. H.Chang, G. Li, Y.Yang,
Nat. Commun. 2014, 5, 5404.
[35] L.Gao, K. Zeng, J. Guo, C.Ge, J.Du, Y.Zhao, C. Chen, H. Deng,
Y.He, H.Song, G.Niu, J.Tang, Nano Lett. 2016, 16, 7446.
[36] X. Xu, C.-C. Chueh, P. Jing, Z. Yang, X. Shi, T. Zhao, L. Y. Lin,
A. K. Y.Jen, Adv. Funct. Mater. 2017, 27, 1701053.
[37] F.Cao, J.Chen, D.Yu, S.Wang, X.Xu, J.Liu, Z. Han, B. Huang,
Y.Gu, K. L.Choy, H.Zeng, Adv. Mater. 2020, 32, 1905362.
[38] Q.Lin, A.Armin, P. L.Burn, P.Meredith, Laser Photonics Rev. 2016,
10, 1047.
[39] Y.Liu, Z.Yang, S.Liu, Adv. Sci. 2018, 5, 1700471.
[40] Z. Ning, X. Gong, R. Comin, G. Walters, F. Fan, O. Voznyy,
E. Yassitepe, A. Buin, S. Hoogland, E. H. Sargent, Nature 2015,
523, 324.
[41] M.Sytnykt, S. Yakunin, W. Schofberger, R. T. Lechner, M. Burian,
L.Ludescher, N. A.Killilea, A. YousefiAmin, D.Kriegner, J.Stangl,
H.Groiss, W.Heiss, ACS Nano 2017, 11, 1246.
[42] Z. Y. Yang, O. Voznyy, G. Walters, J. Z. Fan, M. Liu, S. Kinge,
S.Hoogland, E. H.Sargent, ACS Photonics 2017, 4, 830.
[43] F. P. G. De Arquer, X. Gong, R. P. Sabatini, M. Liu, G. Kim,
B. R. Sutherland, O. Voznyy, J. Xu, Y. Pang, S. Hoogland,
Nat. Commun. 2017, 8, 14757.
[44] R. L.Wang, F.Wang, W. J.Zhou, J. Z.Fan, P. G.deArquer, K. M.Xu,
E. H.Sargent, Z. J.Ning, Infrared Phys. Technol. 2019, 98, 16.
[45] J.Gao, J.Wang, Q.An, X.Ma, Z.Hu, C.Xu, X. L.Zhang, F.Zhang,
Sci. China: Chem. 2020, 63, 83.
[46] H. Uoyama, K. Goushi, K. Shizu, H.Nomura, C. Adachi, Nature
2012, 492, 234.
[47] K.Baeg, M.Binda, D.Natali, M.Caironi, Y.Noh, Adv. Mater. 2013,
25, 4267.
[48] X. Gong, M. Tong, Y. Xia, W. Cai, J. S. Moon, Y. Cao, G. Yu,
C.-L.Shieh, B.Nilsson, A. J.Heeger, Science 2009, 325, 1665.
[49] L.Bakueva, S.Musikhin, M. A.Hines, T. W. F.Chang, M.Tzolov,
G. D.Scholes, E. H.Sargent, Appl. Phys. Lett. 2003, 82, 2895.
[50] S. A. McDonald, P. W. Cyr, L. Levina, E. H. Sargent, Appl. Phys.
Lett. 2004, 85, 2089.
[51] R. Saran, M. N. Nordin, R. J.Curry, Adv. Funct. Mater. 2013, 23,
4149.
[52] R.Saran, V.Stolojan, R. J.Curry, Sci. Rep. 2014, 4, 5041.
[53] K. Szendrei, F.Cordella, M. V. Kovalenko, M. Boberl, G.Hesser,
M.Yarema, D.Jarzab, O. V.Mikhnenko, A. Gocalinska, M. Saba,
F.Quochi, A. Mura, G. Bongiovanni, P. W. M.Blom, W. G.Heiss,
M. A.Loi, Adv. Mater. 2009, 21, 683.
[54] R. Dong, C. Bi, Q. F. Dong, F. W. Guo, Y. B. Yuan, Y. J. Fang,
Z. G.Xiao, J. S.Huang, Adv. Opt. Mater. 2014, 2, 549.
[55] M. Sulaman, S. Yang, A. Bukhtiar, C. Fu, T. Song, H. Wang,
Y.Wang, H.Bo, Y.Tang, B.Zou, RSC Adv. 2016, 6, 44514.
[56] L.Gao, D. D. Dong, J. G.He, K. K.Qiao, F. R.Cao, M.Li, H.Liu,
Y. B.Cheng, J.Tang, H. S.Song, Appl. Phys. Lett. 2014, 105, 4.
Small 2020, 2003397

2003397 (19 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
[57] B. S.Kim, D. C.Neo, B.Hou, J. B.Park, Y.Cho, N.Zhang, J.Hong,
S. Pak, S. Lee, J. I. Sohn, H. E. Assender, A. A. Watt, S. Cha,
J. M.Kim, ACS Appl. Mater. Interfaces 2016, 8, 13902.
[58] J.Yoo, S.Jeong, S.Kim, J. H.Je, Adv. Mater. 2015, 27, 1712.
[59] J.He, M. Luo, L. Hu, Y. Zhou, S. Jiang, H. Song, R.Ye, J.Chen,
L.Gao, J.Tang, J. Alloys Compd. 2014, 596, 73.
[60] J.He, K.Qiao, L.Gao, H. Song, L.Hu, S.Jiang, J. Zhong, J.Tang,
ACS Photonics 2014, 1, 936.
[61] F. P. G. De Arquer, T. Lasanta, M. Bernechea, G. Konstantatos,
Small 2015, 11, 2636.
[62] L. Hu, S. Huang, R. Patterson, J. E.Halpert, J. Mater. Chem. C
2019, 7, 4497.
[63] M.Leng, Y.Yang, K.Zeng, Z.Chen, Z.Tan, S.Li, J.Li, B.Xu, D.Li,
M. P. Hautzinger, Y. Fu, T.Zhai, L. Xu, G. Niu, S. Jin, J. Tang,
Adv. Funct. Mater. 2018, 28, 1704446.
[64] V. Adinolfi, I. J. Kramer, A. Labelle, B. Sutherland, S. Hoogland,
E.Sargent, ACS Nano 2015, 9, 356.
[65] Z. Ren, J. Sun, H. Li, P.Mao, Y. Wei, X. Zhong, J.Hu, S. Yang,
J.Wang, Adv. Mater. 2017, 29, 1702055.
[66] Y. Wei, Z.Ren, A. Zhang, P.Mao, H. Li, X. Zhong, W.Li, S.Yang,
J.Wang, Adv. Funct. Mater. 2018, 28, 1706690.
[67] J. Y.Zhang, J. L.Xu, T.Chen, X.Gao, S. D.Wang, ACS Appl. Mater.
Interfaces 2019, 11, 44430.
[68] E. O. Polat, G. Mercier, I. Nikitskiy, E. Puma, T. Galan, S. Gupta,
M.Montagut, J. J.Piqueras, M.Bouwens, T.Durduran, G.Konstantatos,
S.Goossens, F.Koppens, Sci. Adv. 2019, 5, eaaw7846.
[69] J. W.Lee, D.Kim, F.So, Adv. Funct. Mater. 2015, 25, 1233.
[70] W. Zhou, Y.Shang, F. P.García de Arquer, K.Xu, R.Wang, S.Luo,
X.Xiao, X. Zhou, R.Huang, E. H.Sargent, Z. Ning, Nat. Electron.
2020, 3, 251.
[71] X.Tang, M. M.Ackerman, M. Chen, P.Guyot-Sionnest, Nat. Pho-
tonics 2019, 13, 277.
[72] H. Zhang, Y. Zhang, X. Song, Y. Yu, M. Cao, Y. Che, J. Wang,
J.Yang, H.Dai, G.Zhang, J.Yao, Nanotechnology 2016, 27, 425204.
[73] M. Long, P.Wang, H.Fang, W. Hu, Adv. Funct. Mater. 2019, 29,
1803807.
[74] N.Huo, G.Konstantatos, Adv. Mater. 2018, 30, 1801164.
[75] G. Konstantatos, M. Badioli, L. Gaudreau, J. Osmond,
M. Bernechea, F. P. G. De Arquer, F. Gatti, F. H. L. Koppens,
Nat. Nanotechnol. 2012, 7, 363.
[76] Q.Nian, L.Gao, Y.Hu, B. Deng, J. Tang, G. J.Cheng, ACS Appl.
Mater. Interfaces 2017, 9, 44715.
[77] L.Zheng, W.Zhou, Z.Ning, G. Wang, X.Cheng, W.Hu, W. Zhou,
Z. Liu, S. Yang, K.Xu, M. Luo, Y. Yu, Adv. Opt. Mater. 2018, 6,
1800985.
[78] I. Nikitskiy, S. Goossens, D. Kufer, T. Lasanta, G. Navickaite,
F. H. L.Koppens, G.Konstantatos, Nat. Commun. 2016, 7, 11954.
[79] D.Kufer, I.Nikitskiy, T. Lasanta, G. Navickaite, F. H. L.Koppens,
G.Konstantatos, Adv. Mater. 2015, 27, 176.
[80] T.Rauch, M.Böberl, S. F.Tedde, J.Fürst, M. V.Kovalenko, G.Hesser,
U.Lemmer, W.Heiss, O.Hayden, Nat. Photonics 2009, 3, 332.
[81] E. Heves, C. Ozturk, V. Ozguz, Y.Gurbuz, IEEE Electron Device
Lett. 2013, 34, 662.
[82] P. D. LeVan, E. Klem, J. Lewis, C. Gregory, G. Cunningham,
D. Temple, A. D’Souza, E. Robinson, P. S. Wijewarnasuriya,
N. Dhar, A. K. Sood, P. S. Wijewarnasuriya, A. I. D’Souza, Proc.
SPIE 2013, 8868, 886806.
[83] E. J. D. Klem, C. Gregory, D.Temple, J.Lewis, Proc. SPIE 2015,
9451, 945104.
[84] S. Goossens, G. Navickaite, C. Monasterio, S. Gupta,
J. J.Piqueras, R.Pérez, G.Burwell, I.Nikitskiy, T.Lasanta, T.Galán,
E.Puma, A.Centeno, A. Pesquera, A. Zurutuza, G. Konstantatos,
F.Koppens, Nat. Photonics 2017, 11, 366.
[85] S. Masala, V. Adinolfi, J. P. Sun, S. Del Gobbo, O. Voznyy,
I. J.Kramer, I. G.Hill, E. H.Sargent, Adv. Mater. 2015, 27, 7445.
[86] V.Adinolfi, E. H.Sargent, Nature 2017, 542, 324.
[87] X.Xiao, K.Xu, M.Yin, Y.Qiu, W. Zhou, L.Zheng, X. Cheng, Y.Yu,
Z.Ning, Appl. Phys. Lett. 2020, 116, 101102.
[88] K. Xu, X. Xiao, W. Zhou, X. Jiang, Q. Wei, H. Chen, Z. Deng,
J.Huang, B.Chen, Z. Ning, ACS Appl. Mater. Interfaces 2020, 12,
15414.
[89] D. Y. Kim, K. R. Choudhury, J. W. Lee, D. W.Song, G.Sarasqueta,
F.So, Nano Lett. 2011, 11, 2109.
[90] D. Y. Kim, T. H.Lai, J. W. Lee, J. R.Manders, F.So, Sci. Rep. 2014,
4, 5946.
[91] H. Yu, D.Kim, J. Lee, S. Baek, J.Lee, R. Singh, F. So, Nat. Pho-
tonics 2016, 10, 129.
[92] N. Zhang, H. Tang, K. Shi, W. Wang, W. Deng, B. Xu, K. Wang,
X. W.Sun, Appl. Phys. Lett. 2019, 115, 221103.
[93] T.Rauch, M.Boberl, S. F.Tedde, J. D.Furst, M. V.Kovalenko, G.Hesser,
U.Lemmer, W.Heiss, O.Hayden, Nat. Photonics 2009, 3, 332.
[94] D. K. Hwang, Y. T. Lee, H. S. Lee, Y. J. Lee, S. H. Shokouh,
J.-h. Kyhm, J. Lee, H. H. Kim, T.-H.Yoo, S. H.Nam, D. I. Son,
B.-K.Ju, M.-C.Park, J. D.Song, W. K.Choi, S.Im, NPG Asia Mater.
2016, 8, 233.
[95] J. Kim, S. M. Kwon, Y. K.Kang, Y. H.Kim, M. J. Lee, K. Han,
A.Facchetti, M. G.Kim, S. K.Park, Sci. Adv. 2019, 5, eaax8801.
[96] A. G. Kontos, A. Kaltzoglou, E. Siranidi, D. Palles,
G. K.Angeli, M. K.Arfanis, V.Psycharis, Y. S.Raptis, E. I.Kamitsos,
P. N.Trikalitis, C. C.Stoumpos, M. G.Kanatzidis, P.Falaras, Inorg.
Chem. 2017, 56, 84.
[97] T. C. Jellicoe, J. M. Richter, H. F. Glass, M. Tabachnyk,
R. Brady, S. E. Dutton, A. Rao, R. H. Friend, D. Credgington,
N. C.Greenham, M. L.Bohm, J. Am. Chem. Soc. 2016, 138, 2941.
[98] K. Y. Lu, Y. J. Wang, Z. K. Liu, L. Han, G. Z. Shi, H. H. Fang,
J.Chen, X. C. Ye, S.Chen, F.Yang, A. G.Shulga, T. Wu, M. F.Gu,
S. J.Zhou, J.Fan, M. A.Loi, W. L.Ma, Adv. Mater. 2018, 30, 9.
[99] A. D. Antu, Z. F. Jiang, S. M. Premathilka, Y. T. Tang, J. J. Hu,
A.Roy, L. F.Sun, Chem. Mater. 2018, 30, 3697.
[100] S.Masala, V.Adinolfi, J.Sun, S. D.Gobbo, O.Voznyy, I. J.Kramer,
I. G.Hill, E. H.Sargent, Adv. Mater. 2015, 27, 7445.
[101] D. Kim, K. R. Choudhury, J. W. Lee, D. W. Song, G. Sarasqueta,
F.So, Nano Lett. 2011, 11, 2109.
[102] D.Kim, T.Lai, J. W.Lee, J. R.Manders, F.So, Sci. Rep. 2015, 4, 5946.
[103] H. Yu, D. Kim, J. Lee, S. Baek, J. W. Lee, R. K. Singh, F. So,
Nat. Photonics 2016, 10, 129.
[104] G. E. Tudury, M. V. Marquezini, L. G. Ferreira, L. C. Barbosa,
C. L.Cesar, Phys. Rev. B 2000, 62, 7357.
[105] F. P.Garcia de Arquer, X.Gong, R. P.Sabatini, M.Liu, G. H.Kim,
B. R.Sutherland, O.Voznyy, J.Xu, Y.Pang, S.Hoogland, D.Sinton,
E.Sargent, Nat. Commun. 2017, 8, 14757.
[106] J. G. He, K. K. Qiao, L. Gao, H. S. Song, L. Hu, S. L. Jiang,
J.Zhong, J.Tang, ACS Photonics 2014, 1, 936.
[107] F. Pelayo Garcia de Arquer, T. Lasanta, M. Bernechea,
G.Konstantatos, Small 2015, 11, 2636.
[108] J. Zhang, X. Liu, P. Jiang, H. Chen, Y. Wang, J. Ma, R. Zhang,
F.Yang, M.Wang, J.Zhang, G.Tu, Nano Energy 2019, 66, 104142.
[109] J. W.Lee, D. Y.Kim, F.So, Adv. Funct. Mater. 2015, 25, 1233.
[110] G. Konstantatos, M. Badioli, L. Gaudreau, J. Osmond,
M. Bernechea, F. P. Garcia de Arquer, F. Gatti, F. H. Koppens,
Nat. Nanotechnol. 2012, 7, 363.
[111] Q. Nian, L. Gao, Y. W. Hu, B. W. Deng, J. Tang, G. J. Cheng,
ACS Appl. Mater. Interfaces 2017, 9, 44715.
[112] S.Pak, Y.Cho, J.Hong, J.Lee, S.Lee, B.Hou, G. H.An, Y. W.Lee,
J. E. Jang, H. Im, S. M. Morris, J. I. Sohn, S. Cha, J. M. Kim,
ACS Appl. Mater. Interfaces 2018, 10, 38264.
[113] O.Özdemir, I.Ramiro, S.Gupta, G.Konstantatos, ACS Photonics
2019, 6, 2381.
[114] Z. H.Sun, Z. K.Liu, J. H.Li, G. A.Tai, S. P.Lau, F.Yan, Adv. Mater.
2012, 24, 5878.
Small 2020, 2003397

2003397 (20 of 20)
www.advancedsciencenews.com
© 2020 Wiley-VCH GmbH
www.small-journal.com
[115] J. He, Y.Yu, L.Zhang, J. Yi, S. Liu, H. Kan, G. Zhang, S.Jiang,
J. Mater. Sci.: Mater. Electron. 2014, 25, 2516.
[116] D.Kufer, I.Nikitskiy, T. Lasanta, G. Navickaite, F. H. L.Koppens,
G.Konstantatos, Adv. Mater. 2015, 27, 176.
[117] I. Nikitskiy, S. Goossens, D. Kufer, T. Lasanta, G. Navickaite,
F. H.Koppens, G.Konstantatos, Nat. Commun. 2016, 7, 11954.
[118] L.Gao, C. Chen, K.Zeng, C.Ge, D.Yang, H.Song, J. Tang, Light:
Sci. Appl. 2016, 5, 16126.
[119] Z. W.Ren, J. K.Sun, H.Li, P.Mao, Y. Z.Wei, X. H.Zhong, J. S.Hu,
S. Y.Yang, J. Z.Wang, Adv. Mater. 2017, 29, 7.
[120] H. Zhang, Y. Zhang, X.Song, Y. Yu, M. Cao, Y. Che, Z. Zhang,
H.Dai, J.Yang, G.Zhang, J.Yao, ACS Photonics 2017, 4, 584.
[121] Y. Z. Wei, Z. W. Ren, A. D. Zhang, P. Mao, H. Li, X. H. Zhong,
W. W.Li, S. Y.Yang, J. Z.Wang, Adv. Funct. Mater. 2018, 28, 8.
[122] A.Jagtap, B.Martinez, N.Goubet, A.Chu, C.Livache, C.Gréboval,
J. Ramade, D. Amelot, P. Trousset, A. Triboulin, S. Ithurria,
M. G.Silly, B.Dubertret, E.Lhuillier, ACS Photonics 2018, 5, 4569.
[123] Y. Tang, F. Wu, F. Chen, Y. Zhou, P. Wang, M. Long, W. Zhou,
Z. Ning, J. He, F. Gong, Z.Zhu, S. Qin, W.Hu, Small 2018, 14,
1803158.
[124] A. YousefiAmin, N. A. Killilea, M. Sytnyk, P.Maisch, K. C. Tam,
H. J. Egelhaaf, S. Langner, T. Stubhan, C. J. Brabec, T. Rejek,
M. Halik, K. Poulsen, J. Niehaus, A. Kock, W. Heiss, ACS Nano
2019, 13, 2389.
[125] C.Dong, S. Liu, N. Barange, J.Lee, T.Pardue, X.Yi, S. Yin, F.So,
ACS Appl. Mater. Interfaces 2019, 11, 44451.
[126] Q. Xu, L. Meng, T. Zeng, K. Sinha, C. Dick, X.Wang, Opt. Lett.
2019, 44, 463.
[127] M. J. Grotevent, C. U. Hail, S.Yakunin, D. N.Dirin, K. Thodkar,
G. Borin Barin, P. Guyot-Sionnest, M. Calame, D. Poulikakos,
M. V.Kovalenko, I.Shorubalko, Adv. Opt. Mater. 2019, 7, 1900019.
[128] M. Sulaman, Y. Song, S. Yang, M. Li, M. I. Saleem,
P. V. Chandraseakar, Y. Jiang, Y. Tang, B. Zou, Nanotechnology
2019, 31, 105203.
Wenjia Zhou received his Bachelor’s degree in Material Chemistry from Sichuan University in
China in 2009. He earned his Ph.D. in Physics from the Institute of Physics, the Chinese Academy
of Sciences in 2015, where he worked on fabricating perovskite oxide films and studying their
photoelectricity properties. In 2015, he joined the School of Physical Science and Technology at
ShanghaiTech University as a postdoctoral fellow. Now he is a research assistant professor in
ShanghaiTech University, working on solar cells, photodetectors, and other optoelectronic devices
from solution-processed semiconductors such as colloidal quantum dots and perovskites.
Zhijun Ning received his Ph.D. degree from Department of Applied Chemistry, East China
University of Science and Technology. From 2009 to 2011, he was a Postdoctoral Scholar at
Royal Institute of Technology, Sweden. From 2011 to 2014, he was a Postdoctoral Scholar in the
Department of Electrical and Computer Engineering, University of Toronto. Since 2015, he holds
a faculty position at School of Physical Science and Technology, ShanghaiTech University. He
was selected as highly cited researcher by Web of Science at 2018 and 2019. His current research
interest focuses on solution processed optoelectronic materials and devices, especially quantum
dot and perovskite.
Kaimin Xu is a Ph.D. candidate under the guidance of Prof. Ning in the School of Physical Science
and Technology at ShanghaiTech University. He received his Bachelor’s degree in School of
Chemistry and Chemical Engineering from Nanjing University in 2016. His current research inter-
ests are colloidal quantum dots based infrared photodetectors and novel photoelectric material
exploration.
Small 2020, 2003397