Content uploaded by Danhao Wang
Author content
All content in this area was uploaded by Danhao Wang on Sep 24, 2021
Content may be subject to copyright.
Articles
https://doi.org/10.1038/s41928-021-00640-7
1School of Microelectronics, University of Science and Technology of China, Hefei, People’s Republic of China. 2Hefei National Laboratory for Physical
Sciences at the Microscale, Department of Chemical Physics, University of Science and Technology of China, Hefei, People’s Republic of China.
3Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI, USA. 4Department of Electronic Materials
Engineering, Research School of Physics and Engineering, The Australian National University, Canberra, Australian Capital Territory, Australia.
✉e-mail: haiding@ustc.edu.cn
The semiconductor p–n junction is a fundamental building
block in numerous electronic components, including recti-
fiers/diodes, photovoltaic cells, light-emitting diodes (LEDs)
and photodetectors1–3. All p–n junction-based devices must, how-
ever, obey the physics of unidirectional current flow, which con-
strains the possible functionalities of devices. As a result, various
modified device designs have been developed over the last few
decades. The bipolar junction transistor (a semiconductor triode),
which can amplify or switch electronic signals, was—for a start—
created4, and it laid the foundation for integrated circuits. The
efficiency of a single p–n junction photovoltaic cell is, similarly,
intrinsically limited. However, by designing multi-junction devices
with series-connected/tandem cells, remarkable power conversion
efficiencies—beyond the fundamental limitation of single-junction
solar cells—have been achieved in a range of material systems
(including silicon, III–Vs and perovskite solar cells)5–8. Recently,
such multi-junction or tandem strategies have also been applied in
dual light-emitting and detecting optoelectronic devices within a
single structure for energy-harvesting displays9.
When combined with the photoelectric effect, p–n junctions can
be used to build solid-state photodetectors. However, the detection
capabilities of such devices are constrained to a certain spectrum
range due to two basic working principles: the incident photon
energy needs to be larger than the bandgap of the semiconductor,
and the resultant photocurrent flows in the same direction (uni-
polar photoresponse) under a certain applied bias voltage1,10–12. To
break the limit of conventional p–n junction-based photodetectors,
wavelength-induced photocurrent polarity switching in solid-state
p–n heterojunctions made of SnS/ZnO (ref. 13) and Sb2Se3/ZnO
(ref. 14) semiconductors has been developed; the capabilities of such
devices can be attributed to the competition between photovoltaic
and photothermoelectric effects in the junction15. These results sug-
gest that bidirectional photocurrent behaviour could be used in
applications such as switchable light imaging and optical commu-
nication, as well as filter-less colour discrimination.
In this Article, we report a light-detection electrochemical cell
based on uniform and vertically aligned p-AlGaN/n-GaN p–n het-
erojunction nanowires. After decorating the nanowires with plati-
num (Pt) nanoparticles, the junctions are exposed to an electrolyte
environment that acts like a third electrode (analogous to the third
electrode in a bipolar junction transistor), and a distinctive photo-
response with reversed polarity from the cell under different light
exposures can be observed at a constant applied bias. Illumination
of the device at two different wavelengths (254 and 365 nm) trig-
gers opposite redox reactions at the nanowire/electrolyte interface,
inducing polarity reversal of the photocurrent. The device exhib-
its a responsivity of up to −175 mA W−1 at 254 nm and 31 mA W−1
at 365 nm, both at 0 V. The resulting photodetector operates under
a combination of physical processes (photoelectric conversion
and carrier transport in a single p–n junction) and chemical pro-
cess (redox reaction on the nanowire surface), and provides a fast
approach to distinguish different spectral bands by simply verifying
the polarity of the photocurrent.
Construction of light-detection electrochemical cell
Our light-detection electrochemical cell is composed of
p-AlGaN/n-GaN nanowires grown on n-type Si(111)
(Supplementary Fig. 1). Each nanowire consists of a 200 nm
Si-doped n-GaN and a 200 nm Mg-doped p-AlGaN segment
along the axial direction (Fig. 1b and Supplementary Fig. 1). The
top p-AlGaN segment with an aluminium composition of ~35%
is specifically designed to absorb light within the deep ultravio-
let band (for example, 254 nm light used for our test), which can
also be absorbed by the underlying n-GaN segment. Lower-energy
photons (such as the 365 nm light used for our test) can only be
absorbed by the bottom n-GaN part since the p-AlGaN segment
Bidirectional photocurrent in p–n heterojunction
nanowires
Danhao Wang1, Xin Liu1, Yang Kang1, Xiaoning Wang2, Yuanpeng Wu3, Shi Fang 1, Huabin Yu1,
Muhammad Hunain Memon 1, Haochen Zhang1, Wei Hu2, Zetian Mi 3, Lan Fu4, Haiding Sun 1 ✉
and Shibing Long 1
Semiconductor p–n junctions provide rectification behaviour and act as building blocks in many electronic devices. However, the
typical junction configuration restricts the potential functionalities of devices. Here we report a light-detection electrochemi-
cal cell that is based on vertically aligned p-AlGaN/n-GaN p–n heterojunction nanowires in an electrolyte environment. After
decorating the nanowires with platinum nanoparticles, the cell exhibits a photoresponse in which the photocurrent polarity is
reversed depending on the wavelength of light. In particular, illumination of the device at two different wavelengths (254 nm
and 365 nm) triggers different redox reactions at the nanowire/electrolyte interface, inducing polarity reversal of the photocur-
rent. The device offers a responsivity of up to −175 mA W−1 at 254 nm and 31 mA W−1 at 365 nm, both at 0 V.
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics 645
Articles Nature electroNics
is transparent to the 365 nm light. Additionally, the n-Si substrate
also plays an important role for efficient carrier transport because
of the relatively small band offset between the n-Si and n-GaN
conduction band edges16,17. Then, the as-grown p-AlGaN/n-GaN
nanowires act as photoelectrodes to construct the light-detection
electrochemical cell.
The cell operates differently from conventional p–n
junction-based photodetectors, mainly due to its electrolyte-assisted
carrier transport characteristic arising from a combination of
physical and chemical processes. Figure 1a,c describes the detailed
working principle of the light-detection electrochemical cell oper-
ating under exposure of different wavelengths of light. The energy
band diagrams along the growth direction and the correspond-
ing photocurrent signals are also schematically illustrated. When
p-AlGaN/n-GaN nanowires are illuminated with 254 nm light,
electron–hole pairs are generated in both p-AlGaN and n-GaN
segments (Fig. 1a). The downward surface band bending facilitates
electrons in the p-AlGaN segment drifting towards the nanowire
surface to drive the proton reduction reaction (hydrogen evolution
reaction (HER))18, while the holes migrate towards the space charge
region in the p–n junction, efficiently tunnelling and recombining
with electrons generated from the n-GaN segment. Simultaneously,
the photogenerated holes in the n-GaN segment migrate through
the external circuit and oxidize the water molecules into oxygen at
the Pt counter electrode, hence exhibiting a negative photocurrent
signal (Supplementary Fig. 2a).
When the nanowires are exposed to 365 nm light, the operation
mechanism is entirely different (Fig. 1c). Since only the n-GaN seg-
ment can absorb 365 nm light, the photogenerated holes in n-GaN
readily drift to the n-GaN/electrolyte interface and undergo water
oxidation reaction (oxygen evolution reaction (OER)), with the
upward surface band bending as the driving force18. Meanwhile,
band bending at the n-GaN/electrolyte interface and built-in elec-
tric field of the p–n junction push electrons drifting towards the
external circuit, recorded as a positive photocurrent through the
electrochemical workstation. These electrons eventually participate
in HER at the counter electrode (Supplementary Fig. 2b).
Essentially, the axial p–n heterojunction nanowire configura-
tion is critical for constructing a light-detection electrochemical
cell. We can benefit from the flexibility in tailoring the bandgap of
each segment in the nanowires by choosing proper semiconductor
materials, as well as the direct contact of all the nanowire segments
with the electrolyte environment for electrochemical reactions.
Such a design makes it possible to distinguish spectral regimes by
monitoring the polarity of the photocurrent arising from differ-
ent redox reactions. In this work, the p-AlGaN/n-GaN p–n het-
erojunction nanowires were grown on silicon by a plasma-assisted
molecular beam epitaxy (MBE) system (Supplementary Note and
Supplementary Fig. 3 provide a detailed discussion of the growth
process and crystalline quality of nanowires).
Theoretical calculations
Unlike conventional p–n junction photodetectors, which rely only
on the photoelectric effect for light detection, the photocurrent of a
nanowire-based light-detection electrochemical cell is determined
by the number of photogenerated carriers that participate in the
redox reactions. In other words, the magnitude of the photocur-
rent is affected by the number of photogenerated carriers as well
as the redox reaction rates. Generally, the rate of HER is mainly
determined by the hydrogen adsorption free energy, ΔGH (ref. 19).
Unfortunately, the adsorption energies of the reactive intermedi-
ates on bare p-AlGaN nanowire surface are not suitable for HER. To
boost the detection efficiency of 254 nm light, a modification of the
ΔGH value of the p-AlGaN segment is required.
In the past few decades, it has been recognized that co-catalysts
play an essential role in boosting the electrochemical reaction rates.
Through co-catalyst decoration, the adsorption energies of reactive
intermediates and carrier separation efficiency can be simultane-
ously improved20. Thus, we conducted theoretical investigations
based on density functional theory (DFT) calculations and found
that Pt nanoparticles could be one of the most efficient proton
reduction co-catalysts to decorate the AlGaN surface. Figure 2a
shows the calculated Gibbs free energy of the adsorption of interme-
diate hydrogen of AlGaN(000
¯
1
)/Pt and AlGaN(000
¯
1
). Figure 2b,c
depicts the corresponding atomic configurations of one H+ adsorp-
tion step. Obviously, the AlGaN(000
¯
1
)/Pt model exhibits a relatively
low barrier for the Volmer step, with an energy barrier of −0.38 eV.
In contrast, the energy barrier is as high as −2.177 eV on a bare
AlGaN(000
¯
1
) surface. This result indicates that by incorporating
Pt nanoparticles on the AlGaN(000
¯
1
) surface, the proton reduction
thermodynamics can be optimized to facilitate the HER process.
To achieve high photodetection performance, charge separa-
tion efficiency should also be considered. Fortunately, in addi-
tion to the reduced proton reduction barrier, Pt nanoparticles can
also improve the charge separation and extraction efficiency21.
Differential charge density analysis was carried out to elucidate the
interaction between Pt nanoparticles and AlGaN alloy. Figure 2d
illustrates that the accumulation of charge (red region) occurs at
the Pt atoms/AlGaN interface while charge reduction (blue region)
is found around the Pt and N atoms, revealing that the electrons
are apparently shared by both Pt and AlGaN alloy, which results in
new hybridized states in the forbidden band22. To further confirm
the newly induced electronic states, the density of states for AlGaN
bulk, AlGaN(000
¯
1
) surface and AlGaN(000
¯
1
)/Pt was calculated
based on the DFT method. First, the near-perfect agreement of the
calculated bandgap (4.252 eV; Supplementary Fig. 4) of AlGaN bulk
with its experimental value (4.2 eV; Supplementary Fig. 3) verifies
a
b
c
External circuit
External circuit
Positive photocurrent
Negative photocurrent
p-AlGaN
254 nm
365 nm
H
2
O
O
2
h
+
e
–
h
+
e
–
H
2
H
+
n-Sin-GaN
Fig. 1 | Design of p-AlGaN/n-GaN nanowires for light-detection
electrochemical cell. a, Operation model of the cell under 254nm
irradiation. The energy band diagrams, light-induced photocarrier excitation,
carrier recombination through tunnelling, carrier diffusion process, proton
reduction at the p-AlGaN surface and corresponding negative photocurrent
signal are shown. b, Schematic of the as-grown p-AlGaN/n-GaN nanowires
on n-Si substrate. c, Operation model of the cell under 365nm irradiation
along with the energy band diagrams, light-induced photocarrier excitation,
carrier diffusion process, water oxidation at the n-GaN surface and
corresponding positive photocurrent signal.
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics
646
Articles
Nature electroNics
the high accuracy of our simulation methods and models. Secondly,
the density of states for AlGaN(000
¯
1
) (blue area) presents a classic
semiconductor characteristic with a bandgap of approximately 4 eV,
as shown in Fig. 2e. A detailed explanation about the small peak in
the forbidden band can be found in Supplementary Fig. 5.
Lastly, when the Pt nanoparticle is anchored on the AlGaN(000
¯
1
)
surface, new states in the bandgap can be observed (pink area), indi-
cating the formation of new migration channels for carrier transfer
from AlGaN to the Pt nanoparticle, eventually reaching the electro-
lyte21. All these results suggest that the decoration of Pt nanopar-
ticles endows the device with ideal hydrogen adsorption free energy
as well as highly efficient charge carrier extraction during the pho-
toelectrochemical light detection process.
Pt decoration and structural characterization
Based on the theoretical guidance, we modified the as-grown bare
p-AlGaN/n-GaN nanowire surface through Pt nanoparticle deco-
ration using selective photochemical deposition, during which Pt
nanoparticles anchored on the surface sites where electrons had
selectively accumulated23–26. Figure 3a shows a schematic of the pho-
tochemical deposition process. In principle, when the nanowires are
in contact with the electrolyte, downward and upward surface band
bending happen in the p-AlGaN and n-GaN segments, respectively.
Therefore, only the p-AlGaN surface accumulates electrons that are
photogenerated by the incident ultraviolet light18,27. In other words,
Pt nanoparticles are preferentially photodeposited on the p-AlGaN
surface due to the intrinsic selectivity of photochemical deposition.
Compared with other random synthesis methods (for example,
chemical reduction), the photodeposition approach is more favour-
able for photogenerated electrons to drift to the photochemically
deposited active sites and participate in the HER process25,26.
To elucidate the chemical nature of the as-decorated Pt species,
the valence states of Pt were investigated by X-ray photoelectron
spectroscopy (XPS). As shown in Fig. 3b, the Pt4f XPS spectrum
of Pt-decorated p-AlGaN/n-GaN nanowires could be assigned as
a mixture of Pt0 and Pt2+ species, corresponding to metallic Pt and
Pt()O species, respectively26. The decrease in the valence state
with respect to the Pt precursor (H2PtCl6), as shown in Fig. 3b, indi-
cates that PtCl62− ions were effectively photoreduced by the photo-
generated electrons. Scanning electron microscopy (SEM) images
clearly indicate that the Pt nanoparticles were selectively formed on
the p-AlGaN segment (as shown in Fig. 3c), tending to deposit on
the top surface of the AlGaN(000
¯
1
) segment. To further character-
ize the structural properties of our Pt:p-AlGaN/n-GaN nanowires,
we carried out scanning transmission electron microscopy (STEM)
measurements. It can be seen that the Pt nanoparticles are prefer-
entially distributed on the top surface of the p-AlGaN(000
¯
1
) seg-
ment (Fig. 3d).
Previous studies reported that when the co-catalysts are selec-
tively deposited on the corresponding electron-concentrated seg-
ment, the redox reaction could be enhanced because electrons
tend to be trapped on the co-catalysts, inducing more efficient
–4
–2
0
2
Reaction coordinate
AlGaN(0001)
AlGaN(0001)/Pt
–0.380
–2.177
Surface (*) H* 1/2H2
𝛥GH* (eV)
Pt Al Ga
N H
Density of states (a.u.)
AlGaN(0001)
AlGaN(0001)/Pt
–8 –6 –4 –2 0 2 4 6 8
Energy (eV)
c
–
–
–
–
a b
d e
Fig. 2 | DFT calculations. a, Calculated adsorption energies of intermediate hydrogen on the AlGaN(000
¯
1
)/Pt and AlGaN(000
¯
1
) surface.
b,c, Corresponding atomic configurations of the proton adsorption step on the AlGaN(000
¯
1
)/Pt (b) and AlGaN(000
¯
1
) (c) surface. The spheres in sky
blue, green, pink, royal blue and white represent H, Al, Ga, N, and Pt atoms, respectively. d, Charge density difference plot of AlGaN(000
¯
1
)/Pt. The
blue and red regions represent the deletion and accumulation of electrons, respectively. e, Calculated density of states of AlGaN(000
¯
1
) (blue area) and
AlGaN(000
¯
1
)/Pt (pink area).
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics 647
Articles Nature electroNics
charge separation20,25. On the contrary, the photocurrent would be
drastically decreased if co-catalysts were deposited on the wrong
segment26. Therefore, in our case, a boosted photodetection perfor-
mance could be expected, as the Pt nanoparticles are mainly located
at the p-AlGaN surface. Detailed high-resolution high-angle
annular dark-field scanning transmission electron microscopy
(HAADF-STEM) images indicate that the crystalline Pt nanopar-
ticles have a particle size of ~4 nm (Fig. 3e), with two distinct inter-
planar spacings, namely, 2.3 and 2 Å (Supplementary Fig. 6), which
could be assigned to Pt(111) and Pt(100), respectively28,29. More
importantly, the energy-dispersive spectroscopy (EDS) elemental
mapping analysis of the STEM images directly confirms the prefer-
ential distribution of Pt nanoparticles on p-AlGaN/n-GaN nanow-
ires (Fig. 3f).
To provide an insight into the carrier transport properties of
p-AlGaN/n-GaN nanowires before and after Pt decoration, we
carried out electrochemical impedance spectroscopy measure-
ments under 254 nm illumination. It is known that the diameter
of the semicircle in a Nyquist plot is a characteristic of the charge
transfer resistance30. As shown in Supplementary Fig. 7, the radius
of Pt:p-AlGaN/n-GaN nanowires is much smaller than that of
p-AlGaN/n-GaN nanowires, thus providing unambiguous evidence
that Pt decoration can reduce the resistivity of electron transfer22.
Meanwhile, Pt nanoparticles can serve as an effective charge storage
medium for facile charge transfer from the p-AlGaN segment to the
electrolyte21. Furthermore, we investigated the carrier characteris-
tics of as-grown and Pt-decorated nanowires by room-temperature
photoluminescence (PL) spectroscopy. It is widely recognized
that the PL peak intensity effectively reflects the carrier trapping,
migration, transfer and recombination efficiency, and it is directly
correlated with the photorelated electrochemical performance31,32.
Supplementary Fig. 8 clearly shows the reduction in the PL peak
intensity after Pt loading, which confirms the effective carrier sepa-
ration or a suppression of the carrier recombination18,33.
Behaviour of spectrally distinctive photodetection
Based on our band structure design, theoretical calculations and
material characterization, the p-AlGaN/n-GaN nanowire-based
Al Ga N Pt
p-AlGaN
n-GaN
Pt NPs
Pt NPs
p-AlGaN
mm
n-Si(111) 80 75 70 65
Intensity (a.u.)
Binding energy (eV)
Pt2+4f7/2
Pt04f7/2
Pt04f5/2
Pt2+4f5/2
[PtCl6]2–
Pt NPs
UV light
n-GaN
p-AlGaN
a b
c d e
f
Overlap
e–
Fig. 3 | Photodepositon process and structural characterization of nanowires. a, Schematic of the photodeposition method to form Pt nanoparticles (Pt
NPs) on a p-AlGaN/n-GaN nanowire. b, XPS spectra of Pt:p-AlGaN/n-GaN nanowires. High-resolution Pt4f spectra is fitted by the spin–orbit splitting of
3.45eV and the area ratio of 4:3 for Pt4f7/2 and Pt4f5 /2. c, Top-view SEM images of bare p-AlGaN/n-GaN nanowires (top; scale bar, 100nm), Pt-decorated
p-AlGaN/n-GaN nanowires (middle; scale bar, 100nm) and side-view cross-sectional image of Pt:p-AlGaN/n-GaN nanowires (bottom; scale bar,
200nm). d,e, Overview of the STEM image of Pt-decorated nanowires (d; scale bar, 100nm) and atomic-resolution HAADF-STEM image of the top part of
Pt:p-AlGaN/n-GaN nanowires (e; scale bar, 2nm). f, Selected-area STEM image and the corresponding EDS elemental mapping images (scale bar, 30nm).
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics
648
Articles
Nature electroNics
photoelectrodes were employed to construct the light-detection
electrochemical cell (Supplementary Fig. 9) and applied for the
detection test of 254/365 nm light. To present a thorough and com-
prehensive understanding of the wavelength-dependent polarity
switching of the photocurrent, we performed the current–time
response (I–t curves) test of our light-detection electrochemical
cell with multiple light-switching cycles under 0 V. With periodi-
cally switched light, the I–t curves of both p-AlGaN/n-GaN and
Pt:p-AlGaN/n-GaN nanowires demonstrated repeatable on–off
switching behaviour.
Figure 4a shows the photoresponse of p-AlGaN/n-GaN nanow-
ires and Pt:p-AlGaN/n-GaN nanowires under 254 and 365 nm illu-
mination at 0 V. Bare p-AlGaN/n-GaN nanowires exhibit a positive
current density of 18.7 μA cm−2 under 365 nm illumination, while
the current in the solar-blind region becomes much smaller with
a current density of only 5.6 μA cm−2. Importantly, both photore-
sponses are unipolar, showing a unidirectional current transient
response. When it comes to the Pt:p-AlGaN/n-GaN nanowires,
the photocurrent polarity can be reversed once the incident light
changes from 365 to 254 nm. The Pt:p-AlGaN/n-GaN nanowires
0 25 50 75 100 125 150 175
–2
–1
0
1
2
3
4
5
Current (µA)
Time (s)
Pt:p-AlGaN/n-GaN nanowires
0 50 100 150 200 250 300
–50
0
50
Time (s)
p-AlGaN/n-GaN nanowires
365254
Pt:p-AlGaN/n-GaN nanowires
–20
0
20
40
0 100 200 300 400 500 600
–60
–30
0
360 µW cm–2 320 µW cm–2 240 µW cm–2 103 µW cm–2
Time (s)
550 µW cm–2 390 µW cm–2 156 µW cm–2
100 150 200 250 300 350
–80
–70
–60
–50
–40
–30
–20
–10
Light power intensity (µW cm–2)
Photocurrent density
–180
–160
–140
–120
–100
–80
–60
Responsivity (mA W–1)
Photocurrent density (µA cm–2)
Responsivity
Pt:p-AlGaN/n-GaN nanowires
(254 nm; 0 V; 0.5 M H2SO4)
200 300 400 500 600
5
10
15
20
10
20
30
λ = 255–620 nm
Light power intensity (µW cm–2)
Pt:p-AlGaN/n-GaN nanowires
(365 nm; 0 V; 0.5 M H2SO4)
Photocurrent density
Responsivity
Photocurrent density (µA cm–2)
Responsivity (mA W–1)
Current density (µA cm–2)
Photocurrent density (µA cm–2)
Current (µA)
0 25 50 75 100 125 150 175
–1
0
1
2
Time (s)
255 265 285 310 340 365 453 529 620
p-AlGaN/n-GaN nanowires
λ = 255–620 nm
620 µW cm–2
255 265 285 310 340 365 453 529 620
ab
c d
ef
off off
Fig. 4 | Spectrally distinctive photodetection characterization. a, Repeated on–off I–t characteristics of p-AlGaN/n-GaN nanowires-based and
Pt:p-AlGaN/n-GaN nanowires-based light-detection electrochemical cell at 0V under 254nm and 365nm light illumination. b, I–t characteristics of
Pt:p-AlGaN/n-GaN nanowires-based light-detection electrochemical cell at 0V under 254nm and 365nm light illumination with different light intensities.
c,d, Extracted photocurrent density and corresponding photoresponsivity of Pt:p-AlGaN/n-GaN nanowires under 254nm (c) and 365nm (d) irradiation
with different light intensity. e,f, Current signals of p-AlGaN/n-GaN nanowires (e) and Pt:p-AlGaN/n-GaN nanowires (f) under LED illuminations at
different wavelengths (λ).
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics 649
Articles Nature electroNics
show a positive current density of 23 μA cm−2 under 365 nm illu-
mination, whereas a negative current density of −64 μA cm−2 under
254 nm illumination.
Notably, there are photocurrent peaks in both I–t response
curves. Such transient photocurrent peaks in the photoelectro-
chemical cell might be attributed to the transient carrier accu-
mulation and recombination near the semiconductor/electrolyte
interface through the trapping of electrons/holes at the surface
states34,35. Alternatively, it might be caused by the pyroelectric effect
because such transient photocurrent peaks were also observed in
ZnO-based pyroelectric photodetectors36. Since our GaN-based
nanowires also possess strong pyroelectricity effect and additionally,
the size and shape of the nanowires can also influence the pyroelec-
tric properties in nitride nanowires37, we suspect that the pyroelec-
tric effect of p-AlGaN/n-GaN nanowires, therefore, may lead to
such transient photocurrent peaks. Nevertheless, the dual-polarity
photoresponse behaviour in our electrochemical cell can be imple-
mented in light-induced photocurrent polarity-switchable devices
for light imaging system, visible-light communication and accurate
indoor positioning applications13,14. Moreover, such a cell can be
further miniaturized by choosing nanowire arrays with high peri-
odicity and uniformity in device fabrication38, which could lead to
next-generation high-pixel, high-resolution imaging/sensors.
To further characterize the photodetection performance, we
define photocurrent density Iphoto as follows39: Iphoto = Ilight – Idark, where
Ilight and Idark are the current density with or without light, respectively,
as extracted from the I–t curves. The light-intensity-dependent
photocurrent Iphoto-254 under different 254 nm light intensity (Popt-
254) varying from 103 to 360 μW cm−2 and Iphoto-365 under different
365 nm light intensity (Popt-365) varying from 156 to 620 μW cm−2
were then evaluated. The relationship between Iphoto-254, Iphoto-365
and irradiance power intensity is presented in Fig. 4b. It can be found
that both Iphoto-254 and Iphoto-365 exhibit a positive correlation trend
with continuously increasing light power density, showing potential
for light intensity quantification. The Iphoto-254 value can reach up
to –63.4 µA cm−2 at Popt-254 = 360 μW cm−2 and the Iphoto-365 value is
19.2 µA cm−2 at Popt-365 = 620 μW cm−2. Figure 4c,d shows the Iphoto
and responsivity (R) values as a function of different Popt values. The
responsivity was calculated as follows39: R = Iphoto/Popt, where Iphoto
is the photocurrent density and Popt is the corresponding incident
light intensity. Along with ramping up of the incident irradiation,
the responsivity of Pt:p-AlGaN/n-GaN nanowires under 254 nm
illumination increases from –79 to –175 mA W−1, while the respon-
sivity increases from 13 to 31 mA W−1 under 365 nm illumination.
We also performed the durability test of our device and the results
are discussed in Supplementary Fig. 10a,b.
To confirm the switching point of photocurrent polarity, we
measured and compared the output current under different light
illuminations. Figures 4e,f presents the measured current signals
of p-AlGaN/n-GaN nanowires and Pt:p-AlGaN/n-GaN nanowires
under LED illuminations with different wavelengths at 0 V in 0.5 M
H2SO4. It shows that the photocurrent of bare p-AlGaN/n-GaN
nanowires exhibits a constant positive polarity (Fig. 4e), while the
photocurrent polarity of Pt-decorated p-AlGaN/n-GaN nanow-
ires is highly dependent on the wavelength of incident light (Fig.
4f). First, we observed negative current signals on light illumina-
tions with wavelengths lower than 265 nm. Thereafter, when the
device was exposed to light with longer wavelengths (285 nm or
higher), the photocurrent polarity was reversed from negative to
positive, which exactly follows the working principle illustrated in
Fig. 1. This phenomenon confirms the success of our device design
for distinguishing spectral bands by simply measuring the polar-
ity of the photocurrent. Besides, the photocurrent corresponding
to visible-light irradiation (incident light wavelength is 453 nm or
higher) can be negligible, which reflects the excellent visible-blind
characteristic of our device.
Moreover, in a classic solid-state photodetector, photoresponse
can be merely tuned by changing the applied voltage or incident
light intensity. For our light-detection electrochemical cells, essen-
tially, their photocurrent can also be easily tuned by the surround-
ing environment, especially the electrolyte, which has a profound
effect on the photoresponse behaviour. In Supplementary Fig. 10c,d,
the value of Iphoto-254 has a linear relationship with increasing H2SO4
concentration, while Iphoto-365 is maintained at a constant level,
which demonstrates that our light-detection electrochemical cell is
also sensitive to the electrolyte concentration. Such additional tun-
ability could further broaden the application of our devices towards
complex optical and imaging sensing applications in a sophisti-
cated environment. In short, all the above features demonstrate that
Pt:p-AlGaN/n-GaN nanowires exhibit great potential for spectrally
selective and sensitive applications.
Conclusions
We have reported bidirectional photoconductivity behaviour in
a light-detection electrochemical cell composed of Pt-decorated
p-AlGaN/n-GaN nanowires. The reversal of the flow of photo-
current is triggered by diverse redox reactions at the nanowire/
electrolyte interface under different light illuminations: under
254 nm illumination, the nanowires exhibit a negative responsivity
of –175 mA W–1, whereas under 365 nm illumination, the nanow-
ires exhibit a positive responsivity of 31 mA W–1, both at 0 V. This
bipolar photoconductivity behaviour illustrates that vertical p–n
junction nanowires can provide spectral-band-distinguishable pho-
todetection. Our device architecture could potentially provide dis-
tinct responses to ultraviolet, visible and infrared illumination by
constructing nanowire p–n heterojunctions based on a combination
of binary or ternary III–V semiconductors with different material
compositions (such as III-nitrides or III-arsenides). It could thus
provide a direct and cost-effective approach to build multi-channel
photo- and biosensors, portable spectrometers, optically controlled
logic circuits and computing, and high-resolution photosensing/
imaging devices.
Methods
MBE growth of p-AlGaN/n-GaN nanowires. e p-AlGaN/n-GaN nanowires
were grown on an n-type Si(111) substrate by radio-frequency plasma-assisted
MBE in nitrogen-rich conditions. To remove organic contaminants, the Si substrate
was thoroughly cleaned with acetone and methanol solvent. ereaer, the native
oxide on the substrate was removed by 10% hydrouoric acid before loading into
the MBE chamber. e residual oxide was then desorbed by in situ annealing
of the substrate at ~780 °C before growth initiation. e reection high-energy
electron diraction with a reconstructed pattern further conrms successful clean
of the Si(111) surface. Further, Al, Ga, Si and Mg uxes were controlled using the
respective thermal eusion cells, with an approximate beam equivalent pressure of
2.2 × 10–8, 7.1 × 10–8, 2.5 × 10–9 and 3.6 × 10–9 torr, respectively, whereas the nitrogen
radicals were supplied from a radio-frequency plasma source with a forward plasma
power of 350 W. e nanowires were grown with a substrate temperature at 780 °C
for GaN segment with Si doping and 835 °C for AlGaN segment with Mg doping.
Deposition of co-catalyst Pt nanoparticles. The co-catalyst Pt nanoparticles
were deposited on the p-AlGaN/n-GaN surface by using in situ selective
photodeposition method at room temperature. Normally, the nanowire sample was
immersed into a precursor solution including 100 μl H2PtCl6 solution (20 mg ml–1),
15 ml methanol and 55 ml deionized water without pH adjustment. The nanowires
were then irradiated by a Hg lamp for 15 min. After Pt deposition, the sample was
washed with ethanol and deionized water for more than three times and finally
dried using N2 gas. The as-obtained sample was used for material characterizations
and photodetection measurement.
Photoelectrode preparation. Bare nanowires and Pt-decorated nanowires were
further fabricated as the photoelectrode to evaluate the corresponding detection
ability. To form an ohmic contact, we firstly scraped off the native oxide layer
formed on the backside of the Si substrate by a diamond pen. Then, In–Ga eutectic
alloy (Alfa Aesar) was deposited on the scraped area, which was subsequently
mounted onto a copper sheet using a silver paste (SPI Supplies). Subsequently,
the back and edge of the as-fabricated electrode device was covered by insulating
epoxy to prevent current leakage, except for the nanowire surface. Finally, the
photoelectrodes were cured in air at least 24 h before any measurement.
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics
650
Articles
Nature electroNics
Construction of light-detection electrochemical cell. The as-prepared
p-AlGaN/n-GaN nanowire photoelectrode, Pt counter electrode and saturated
Ag/AgCl reference electrode were constructed in a high-ultraviolet-transmittance
quartz reaction cell (CEL-CPE50) with an electrolyte solution (here we
used 0.500, 0.100, 0.050, 0.010 and 0.001 M H2SO4), thereby forming the
light-detection electrochemical cell. Thereafter, the photodetection performance
of the light-detection electrochemical cell was performed using a CHI660E
electrochemical workstation with a standard three-electrode system, which was
used to record the current signal and applied bias potential at room temperature.
A Hg lamp was used to generate monochromatic lights with wavelengths of 254
and 365 nm to illuminate the photoelectrode. A series of LEDs with different
emission wavelengths were also used for the experiments. The light intensity was
calibrated by an optical power meter (Newport model 2936-R). Amperometric I−t
curves were obtained at a fixed potential of 0 V with a sampling interval of 0.05 s.
In this configuration, the applied voltage is set between the working and reference
electrodes. Electrochemical impedance spectroscopy measurements were carried
out in a frequency range from 0.1 Hz to 500,000.0 Hz with 5 mV amplitude.
Structural and spectroscopic characterization. Detailed morphologies of
the as-prepared nanowires were examined using high-resolution transmission
electron microscopy on JEOL-2100F systems at 200 kV and SEM on Hitachi
SU8220 systems. Sub-ångström-resolution aberration-corrected HAADF-STEM
measurements were acquired on a JEM-ARM200F instrument (University
of Science and Technology of China) at 200 kV. EDS was performed using a
26FEI Talos F200X device at 200 kV. Room-temperature PL measurements were
performed using a 266 nm excitation pulse laser. The PL signal was collected
by using an ultraviolet objective and then measured by an OceanOptics QE Pro
spectrometer. XPS measurements were conducted on a Thermo Scientific K-Alpha
XPS instrument equipped with an Al Kα source (hν = 1,486.6 eV, where h is
Planck’s constant and v is the frequency) at 15 kV. The binding energy scale of all
the measurements was calibrated by referencing C1s to 284.8 eV.
DFT calculation. All the related calculations based on DFT in this paper were
performed by the Vienna ab initio simulation package40,41 with the generalized
gradient approximation for the exchange–correlation potential and with the
projector augmented wave. The energy cutoff for the plane-wave basis was set to
520 eV for all the calculations. The effect of the van der Waals interactions was
considered by the DFT-D3 method proposed by Grimme et al.42. The Brillouin
zones were sampled with 6 × 6 × 6 and 1 × 1 × 1 grids for the primitive cells of bulk
phase and the surface structure according to the Monkhorst–Pack procedure,
respectively. The model of the (000
¯
1
) surface of AlGaN is constructed with six
Al–Ga–N layers, where the bottom three layers were fixed at the corresponding
bulk structure and the dangling bonds at the bottom of the slab were passivated
with hydrogen atoms. It is noteworthy that for real-world ternary nitride alloys,
nitrogen atoms always occupy the anion sites, while the cations (in this work, Al
and Ga) are randomly distributed among the cation sites. It was experimentally
observed that different types of ordering may exist in III-nitride ternary alloys.
Fortunately, the specially designed structure chosen here can well represent the
microscopic structure of a random alloy for the calculation43,44. Thus, an Al/Ga
ratio of 3/8 was employed for the AlGaN alloy, while the rest of the compositions
were covered by quadratic regression. A vacuum layer of 15 Å was inserted between
the periodically repeated slabs along the c axis to avoid interactions among them. A
6 × 6 supercell of the surface was used to investigate the adsorption of H2. The Pt10
cluster was used to decorated the surface. The structure of the Pt10 cluster takes the
form of a pyramid, which is the lowest-energy minimum of the gas-phase Pt10 and
has been used as a catalyst for the hydrogenation of styrene and CO oxidation with
high catalytic activity45,46. The electronic properties of bulk AlGaN were predicted
with the hybrid functional (HSE06). The geometric structures of bulk AlGaN were
relaxed until the atomic forces were less than 0.002 eV Å–1 for each atom and total
energies were converged to 10–5 eV. The optimized lattice constants are a = 6.40 Å
and c = 5.19 Å, which are in agreement with the previous report44.
Free energy calculation. In principle, an ideal HER catalyst should hold a
thermal-neutral (close to 0 eV) Gibbs free energy of hydrogen adsorption
(ΔGH* = 0 eV). Under standard conditions, the HER pathway can be described as
follows.
H
++
e
−+↔
H
∗(
ΔG
=
0 eV
)
The Gibbs free energy of the adsorption of intermediate hydrogen (∆GH*) on
the catalyst is a key descriptor for the HER activity of the catalyst and is obtained
as follows.
ΔGH
∗=
ΔEH
∗+
ΔEZPE
−
TΔSH
∗
T is the temperature and ∆SH* is the entropy. Here ∆EZPE can be obtained by
ΔEZPE
=
EH
ZPE
−
1/2EH
2
ZPE,
where
EH
ZPE
is the zero-point energy of atomic hydrogen in the catalyst and
EH
2
ZPE
is
the zero-point energy of H2 in the gas phase. The entropy is given by
S(T)=
3N
∑
i
=
1[
−Rln
(
1−e−hvi
kBT
)
+NAhvi
T
e−
hv
i
/k
B
T
1−e−hvi/kBT
]
,
where R is the universal gas constant, kB is the Boltzmann constant, h is Pl anc k’s
constant, NA is Avogadro’s number, vi represents the frequency and N is the number
of adsorbed atoms47.
Data availability
The data that support the plots within this paper and other findings of this study
are available from the corresponding author upon reasonable request.
Received: 22 December 2020; Accepted: 29 July 2021;
Published online: 23 September 2021
References
1. Zhang, Y. J. et al. Enhanced intrinsic photovoltaic eect in tungsten disulde
nanotubes. Nature 570, 349–353 (2019).
2. Taniyasu, Y., Kasu, M. & Makimoto, T. An aluminium nitride light-emitting
diode with a wavelength of 210 nanometres. Nature 441, 325–328 (2006).
3. Choi, T., Lee, S., Choi, Y. J., Kiryukhin, V. & Cheong, S. W. Switchable
ferroelectric diode and photovoltaic eect in BiFeO3. Science 324,
63–66 (2009).
4. Bardeen, J. & Brattain, W. H. e transistor, a semi-conductor triode. Phys.
Rev. 74, 230 (1948).
5. Xiao, K. et al. All-perovskite tandem solar cells with 24.2% certied eciency
and area over 1 cm2 using surface-anchoring zwitterionic antioxidant.
Nat. Energy 5, 870–880 (2020).
6. Lin, R. X. et al. Monolithic all-perovskite tandem solar cells with 24.8%
eciency exploiting comproportionation to suppress Sn(II) oxidation in
precursor ink. Nat. Energy 4, 864–873 (2019).
7. Guo, F. et al. A generic concept to overcome bandgap limitations for
designing highly ecient multi-junction photovoltaic cells. Nat. Commun. 6,
7730 (2015).
8. Essig, S. et al. Raising the one-sun conversion eciency of III–V/Si solar cells
to 32.8% for two junctions and 35.9% for three junctions. Nat. Energy 2,
17144 (2017).
9. Oh, N. et al. Double-heterojunction nanorod light-responsive LEDs for
display applications. Science 355, 616–619 (2017).
10. Chen, Y. et al. Strain engineering and epitaxial stabilization of halide
perovskites. Nature 577, 209–215 (2020).
11. Gu, L. et al. A biomimetic eye with a hemispherical perovskite nanowire
array retina. Nature 581, 278–282 (2020).
12. García de Arquer, F. P., Armin, A., Meredith, P. & Sargent, E. H.
Solution-processed semiconductors for next-generation photodetectors.
Nat. Rev. Mater. 2, 16100 (2017).
13. Ouyang, B., Zhang, K. & Yang, Y. Photocurrent polarity controlled by light
wavelength in self-powered ZnO nanowires/SnS photodetector system.
iScience 1, 16–23 (2018).
14. Ouyang, B., Zhao, H., Wang, Z. L. & Yang, Y. Dual-polarity response in
self-powered ZnO NWs/Sb2Se3 lm heterojunction photodetector array for
optical communication. Nano Energy 68, 104312 (2020).
15. Ouyang, B., Wang, Y. H., Zhang, R. Y., Olin, H. & Yang, Y. Dual-polarity
output response-based photoelectric devices. Cell Rep. Phys. Sci. 2,
100418 (2021).
16. Wang, Y. J. et al. An In0.42Ga0.58N tunnel junction nanowire photocathode
monolithically integrated on a nonplanar Si wafer. Nano Energy 57,
405–413 (2019).
17. Fan, S. et al. High eciency solar-to-hydrogen conversion on a monolithically
integrated InGaN/GaN/Si adaptive tunnel junction photocathode. Nano Lett.
15, 2721–2726 (2015).
18. Kibria, M. G. et al. Visible light-driven ecient overall water splitting using
p-type metal-nitride nanowire arrays. Nat. Commun. 6, 6797 (2015).
19. Seh, Z. W. et al. Combining theory and experiment in electrocatalysis:
insights into materials design. Science 355, 6321 (2017).
20. Chen, S. S., Takata, T. & Domen, K. Particulate photocatalysts for overall
water splitting. Nat. Rev. Mater. 2, 17050 (2017).
21. He, Y. M. et al. Dependence of interface energetics and kinetics on catalyst
loading in a photoelectrochemical system. Nano Res. 12, 2378–2384 (2019).
22. Zhou, B. et al. Gallium nitride nanowire as a linker of molybdenum suldes
and silicon for photoelectrocatalytic water splitting. Nat. Commun. 9,
3856 (2018).
23. Wenderich, K. & Mul, G. Methods, mechanism, and applications
of photodeposition in photocatalysis: a review. Chem. Rev. 116,
14587–14619 (2016).
24. Li, Z. et al. Surface-polarity-induced spatial charge separation boosts
photocatalytic overall water splitting on GaN nanorod arrays. Angew. Chem.
Int. Ed. 132, 945–952 (2020).
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics 651
Articles Nature electroNics
43. Bernardini, F. & Fiorentini, V. Nonlinear macroscopic polarization in III-V
nitride alloys. Phys. Rev. B 64, 085207 (2001).
44. Liu, K. K. et al. Wurtzite BAlN and BGaN alloys for heterointerface
polarization engineering. Appl. Phys. Lett. 111, 222106 (2017).
45. Yin, C. R. et al. Alumina-supported sub-nanometer Pt10 clusters:
amorphization and role of the support material in a highly active CO
oxidation catalyst. J. Mater. Chem. A 5, 4923–4931 (2017).
46. Imaoka, T. et al. Platinum clusters with precise numbers of atoms for
preparative-scale catalysis. Nat. Commun. 8, 688 (2017).
47. Zhu, Y. A., Chen, D., Zhou, X. G. & Yuan, W. K. DFT studies of dry
reforming of methane on Ni catalyst. Catal. Today 148, 260–267 (2009).
Acknowledgements
This work was funded by the National Natural Science Foundation of China (grant
nos. 61905236 and 51961145110), the Fundamental Research Funds for the Central
Universities (grant no. WK2100230020), USTC Research Funds of the Double First-Class
Initiative (grant no. YD3480002002) and USTC National Synchrotron Radiation
Laboratory (grant no. KY2100000099), and was partially carried out at the USTC Center
for Micro and Nanoscale Research and Fabrication. We thank W. Wu from USTC for the
support of DFT calculation.
Author contributions
H.S. developed the idea and designed the experiments. D.W., X.L., Y.K., Y.W., S.F.,
H.Y., M.H.M., H.Z. and Z.M. performed the MBE growth and characterizations, XPS
measurement and photodetection experiments, as well as collected and analysed the data.
D.W., X.L., Y.K. and S.F. performed Pt nanoparticle decoration and material investigation.
X.W. and W.H. conducted and discussed the theoretical calculations. D.W. and H.S
performed the aberration-corrected STEM characterization. D.W., L.F., S.L and H.S.
co-wrote the paper. All the authors discussed the results and commented on the manuscript.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41928-021-00640-7.
Correspondence and requests for materials should be addressed to Haiding Sun.
Peer review information Nature Electronics thanks Ya Yang and Daoyou Guo for their
contribution to the peer review of this work.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2021
25. Takata, T. et al. Photocatalytic water splitting with a quantum eciency of
almost unity. Nature 581, 411–414 (2020).
26. Li, R. et al. Spatial separation of photogenerated electrons and holes among
{010} and {110} crystal facets of BiVO4. Nat. Commun. 4, 1432 (2013).
27. Kibria, M. G. et al. Tuning the surface Fermi level on p-type gallium nitride
nanowires for ecient overall water splitting. Nat. Commun. 5, 3825 (2014).
28. Cao, L. et al. Atomically dispersed iron hydroxide anchored on Pt for
preferential oxidation of CO in H2. Nature 565, 631–635 (2019).
29. Wang, Y. J. et al. A single-junction cathodic approach for stable unassisted
solar water splitting. Joule 3, 2444–2456 (2019).
30. Varadhan, P. et al. Surface passivation of GaN nanowires for enhanced
photoelectrochemical water-splitting. Nano Lett. 17, 1520–1528 (2017).
31. Chowdhury, F. A., Trudeau, M. L., Guo, H. & Mi, Z. A photochemical diode
articial photosynthesis system for unassisted high eciency overall pure
water splitting. Nat. Commun. 9, 1707 (2018).
32. Cao, Y. et al. Atomic-level insight into optimizing the hydrogen evolution
pathway over a Co1-N4 single-site photocatalyst. Angew. Chem. Int. Ed. 56,
12191–12196 (2017).
33. Liu, W. et al. Single-site active cobalt-based photocatalyst with a long carrier
lifetime for spontaneous overall water splitting. Angew. Chem. Int. Ed. 129,
9440–9445 (2017).
34. Klahr, B., Gimenez, S., Fabregat-Santiago, F., Bisquert, J. & Hamann, T. W.
Photoelectrochemical and impedance spectroscopic investigation of water
oxidation with ‘Co–Pi’-coated hematite electrodes. J. Am. Chem. Soc. 134,
16693–16700 (2012).
35. Dotan, H., Sivula, K., Grätzel, M., Rothschild, A. & Warren, S. C. Probing the
photoelectrochemical properties of hematite (α-Fe2O3) electrodes using
hydrogen peroxide as a hole scavenger. Energy Environ. Sci. 4, 958–964 (2011).
36. Ma, N., Zhang, K. & Yang, Y. Photovoltaic–pyroelectric coupled eect
induced electricity for self-powered photodetector system. Adv. Mater. 29,
1703694 (2017).
37. Jiang, H. P., Su, Y. J., Zhu, J., Lu, H. M. & Meng, X. K. Piezoelectric and
pyroelectric properties of intrinsic GaN nanowires and nanotubes: size and
shape eects. Nano Energy 45, 359–367 (2018).
38. Ra, Y. H. et al. An electrically pumped surface-emitting semiconductor green
laser. Sci. Adv. 6, eaav7523 (2020).
39. Xie, Z. J. et al. Ultrathin 2D nonlayered tellurium nanosheets: facile
liquid-phase exfoliation, characterization, and photoresponse with high
performance and enhanced stability. Adv. Funct. Mater. 28, 1705833 (2018).
40. Kresse, G. & Furthmuller, J. Eciency of ab-initio total energy calculations
for metals and semiconductors using a plane-wave basis set. Comput. Mater.
Sci. 6, 15–50 (1996).
41. Kresse, G. & Furthmuller, J. Ecient iterative schemes for ab initio
total-energy calculations using a plane-wave basis set. Phys. Rev. B 54,
11169–11186 (1996).
42. Grimme, S. Semiempirical GGA-type density functional constructed with a
long-range dispersion correction. J. Comput. Chem. 27, 1787–1799 (2006).
NATURE ELECTRONICS | VOL 4 | SEPTEMBER 2021 | 645–652 | www.nature.com/natureelectronics
652
A preview of this full-text is provided by Springer Nature.
Content available from Nature Electronics
This content is subject to copyright. Terms and conditions apply.