This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Materials, School of Materials Science and Engineering,
University of Science and Technology Beijing, Beijing
100083, People’s Republic of China. 2 Key Laboratory of New Energy Materials and
Technologies, University of Science and Technology
Beijing, Beijing 100083, People’s Republic of China.
Piezo-modulated interface engineering to enhance the photoresponse of
all-oxide Cu2O/ZnO heterojunction was firstly reported. Under the
illumination of 17.2 mw/cm2, a photoresponse increase of 18.6% has
been obtained when applying a -0.88% compressive strain.
2
Enhanced Photoresponse of Cu2O/ZnO heterojunction with Piezo-modulated Interface Engineering
Pei Lin1, §, Xiaoqin Yan1, §, Xiang Chen1 , Zheng Zhang1, Haoge Yuan1, Peifeng Li1, Yanguang Zhao1, Yue Zhang1,2 () 1 State Key Laboratory for Advanced Metals and Materials, School of Materials Science and Engineering, University of Science and
Technology Beijing, Beijing 100083, People’s Republic of China. 2 Key Laboratory of New Energy Materials and Technologies, University of Science and Technology Beijing, Beijing 100083,
The growth was conducted at 368 K for 4 h. Then, a
100 nm thick tin‐doped indium oxide (ITO) was
sputtered as the top electrode and form Ohmic
contact with ZnO. The morphology of as‐synthesized
materials was characterized with scanning electron
microscope (FESEM, FEI QUANTA 3D). The optical
absorption spectrum of Cu2O, ZnO and Cu2O/ZnO
heterostructure were recorded by UV‐Vis‐NIR
spectrophotometer (VARIAN, Cary 5000) in the
wavelength range of 200‐800 nm. X‐ray diffraction
(XRD, Rigaku DMAX‐RB, Cu Kα) and confocal
Raman microscopy (JY‐HR800) were performed to
4
evaluate the materials’ structural property. The
characterization of sputtered ZnO seed layer was
provided in Fig. S1 in the Electronic Supplementary
Material (ESM). Capacitance‐Voltage measurement
was performed under dark with an AC amplitude of
50 mV and frequency of 10 kHz to determine the
carrier concentration of Cu2O and ZnO. The light
was provided by a solar simulator (Oriel, 91159A)
with adjusted power and was illuminated from the
Cu2O side. All of the electrical characterizations were
recorded at room temperature with an
electrochemical station (Solartron SI 1287).
2 Results and discussion
The schematic of device is sketched in Fig. 1(a).
Figure. 1(b) shows typical cross‐sectional SEM image
of the as‐synthesized Cu2O/ZnO. Much higher
nutrition concentration used in this work resulted in
the densely packed ZnO NRs film where nanorod
merged ultimately with each other [26]. This densely
packed ZnO NRs could facilitate the sputtering of
ITO top electrode and reduce current leakage. Top
view of the ZnO NRs film in the inset of Fig. 1(b)
shows typical hexagonal morphology. It can be
perceived that the synthesized ZnO NRs grows
epitaxially on the Cu2O/FTO substrate with preferred
+c‐orientation due to the existence of pre‐sputtered
ZnO seed layer [25]. The structural properties of
deposited Cu2O were characterized with X‐ray
diffractometer and Raman spectrum, as presented in
Fig. 1(c). The XRD pattern indicates that Cu2O has
cubic structure with (111) preferred orientation,
while the Raman spectrum shows characteristic
phonon frequencies of Cu2O without the CuO
impurity [29‐30]. Carrier concentrations of
as‐prepared Cu2O and ZnO were quantified through
the capacitance‐voltage measurement with
Mott‐Schottky equation (1), as shown in Fig. 1(d):
2 FB20
21 = - - kTV V eC A q N
(1)
Where C represents the capacitance of the space
charge region, V the external applied bias, VFB flat
band potential, k the Boltzmann constant, � the
dielectric constant, �0 the vacuum permittivity and
N the carrier concentration [21]. By fitting the curve
of 1/C2 versus V, carrier concentrations of Cu2O and
ZnO were calculated to be 3.1× 1015 cm‐3 and 1.55×
1017 cm‐3 respectively. Because the NZnO is two orders
higher than NCu2O and �ZnO has similar scale as �Cu2O,
most of the charge depletion region falls in Cu2O
side, which is favorable for the photoresponse of
Cu2O/ZnO heterojunction [20].
5
Figure 1. (a) Schematic of the fabricated device. (b) Cross-sectional SEM image of the as-prepared Cu2O/ZnO heterostructure. (c) XRD pattern and Raman spectrum (inset) of the deposited Cu2O. (d) Mott-Schottky plots of synthesized Cu2O and ZnO.
The optical absorption spectrum of ZnO, Cu2O
and Cu2O/ZnO heterostructure are shown in the Fig.
2(a). Significant absorption increase of ZnO could
only be observed below the wavelength of 378 nm,
while the Cu2O shows much favorable light
absorption in the range of 400 nm‐800 nm than ZnO.
For Cu2O/ZnO heterostructure, light was illuminated
from the Cu2O side and shows similar absorption
characteristic with Cu2O. This result indicates that
vast majority of the illumination is absorbed by Cu2O.
For direct band gap material, the absorption
coefficient α as a function of photon energy hυ could
be expressed as follows [31]:
2 = gA h E (2)
Where A is a constant, Eg is the band gap energy.
Through linear extrapolation of the curve α2 versus
hυ, band gap of ZnO and Cu2O was calculated to be
3.28 eV and 2.24 eV, respectively, as provided in Fig.
2(b) and (c).
6
Figure 2. (a) Optical absorption spectrum of ZnO NR, Cu2O and Cu2O/ZnO heterostructure. (b) and (c) Optical absorption spectrum of
ZnO and Cu2O drawn as α2 versus photon energy hυ.
Figure 3(a) presents the I‐V characteristics of the
device under dark and AM 1.5 illumination. Under
dark, the device shows typical rectifying behavior
with reverse leakage current to be 8× 10‐6 A at ‐1 V,
indicating the formation of well‐defined p‐n junction
at Cu2O/ZnO interface. Under illumination from the
Cu2O side, the current increases dramatically at
forward bias and a strong photoresponse
(Ilight‐Idark)/Idark of about 530% could be obtained at +1
V. Illumination intensity dependent photoresponse at
0 V was performed and the result shows in Fig. 3(b).
It is straightforward to see that the photocurrent
increases linearly with the light intensity, showing
that the device could be used as a self‐powered
photodetector (PD) [17]. Compared with the
photoconductive PDs, this device possesses
relatively short response (0.22 s) and decay (0.32 s)
time due to the absence of deep level traps and
surface molecules absorption, as shown in Fig. 3(c).
Considering the time resolution of testing instrument
and setting parameters, much shorter response time
could be expected. The mechanism is depicted in Fig.
3(d), band bending and space charge region at Cu2O
side provide the critical driving force to separate the
excitons generated under illumination [17]. So
besides contact resistance and carrier mobility, the
most important factor influencing the photocurrent
is the magnitude of depletion region δCu2O in Cu2O.
7
Figure 3. (a) I-V characteristics of Cu2O/ZnO under dark and AM 1.5 illumination. (b) Illumination intensity dependent photocurrent response. (c) Response and decay time. (d) Energy band bending of Cu2O/ZnO heterojunction and the self-powered sensing mechanism at 0 V.
As previously reported, piezoelectric polarization
at the interface could tune band structure and charge
transport properties locally without altering interface
structure [20]. Piezotronic effect on the performance
of this fabricated device was comprehensively
investigated. As shown in Fig. 4(a), under the
illumination density of 17.2 mw/cm2, photoresponse
increases gradually with the increase of compressive
strain. Because of the preferred +c‐orientation growth
of ZnO NRs, a permanent positive piezopotential
would be produced at the Cu2O/ZnO interface when
applying compressive strain, which could
redistribute the charge carriers and band bending, as
presented in Fig. 4(b) [18]. The positive
piezopotential lowers the energy of conduction band
in ZnO and induces a Δφpz, Cu2O potential change in
Cu2O. Like externally applied positive bias, this
potential change produces a sharper and extended
built‐in field in Cu2O, as shown below [20]:
2 2 2
2
2
0 , ,
,
2 Cu O bi Cu O pz Cu O
pz Cu OCu OqN
(3)
Where φbi,Cu2O represents the original built‐in
potential in Cu2O. The enlarged space charge region
in Cu2O means that more excitons are generated and
could be separated apart more effectively, leading to
the enhanced photoresponse current.
Quantitative calculations about the space charge
region change in Cu2O under different strain are
provided in Fig. S2 in the Electronic Supplementary
Material (ESM). It demonstrates that the built‐in
potential in Cu2O increases with the increase of
applied compressive strain. Meanwhile, the
comparable magnitude of Δφpz, Cu2O (~ 0.2 V per 1.0%
strain) with the original φbi,Cu2O (~ 0.65 V) suggests the
possibility of effective interface engineering with
piezopotential [20].
Furthermore, influence of illumination density on
the piezoelectric modulation ability is also studied.
8
Figure 4(c) shows the photoresponse under varying
compressive strains and fixed illumination density of
87.8 mw/cm2, which is similar to Fig. 4(a) but with
larger photoresponsivity. Both of the photoresponse
increase linearly with the applied compressive strain,
as shown in Fig. 4(d). However, a 2.2% increasement
per 0.1% strain could be obtained under illumination
density of 17.2 mw/cm2, while only 1.2% for 87.8
mw/cm2. This modification discrepancy arises from
the screening effect of piezopotential [16]. It has been
reported that the free carriers, charged surface or
bulk trap states could largely screen the
piezopotential generated in ZnO [20]. Under
stronger illumination, carrier concentration in ZnO
increases, the screening effect becomes significant
and modulation ability is thus weakened.
Figure 4. (a) Time-resolved photoresponse under different compressive strains and illumination density of 17.2 mw/cm2. (b) Band diagram of Cu2O/ZnO interface with and without the appearance of positive piezopotential, shown in solid red and dotted blue curves respectively (c) Time dependent photoresponse under varying compressive strains and illumination density of 87.8 mw/cm2. (d) Calculated photoresponse enhancement versus compressive strains under illumination of 17.2 mw/cm2 and 87.8 mw/cm2.
Under compressive strain, piezoresistive effect and
change of Cu2O/ZnO interfacial states may also
influence the device performances [18, 20]. In order
to illustrate that piezotronic effect plays the
dominate role in the enhancement of photoresponse,
comparative device structure is designed, as shown
in Fig. 5(a). The fabrication process and synthesis
parameters are similar to Fig. 2 except that
vertically‐aligned ZnO NRs film was firstly
synthesized on FTO substrate with pre‐sputtered
ZnO seed layer, then the Cu2O was electrochemically
deposited on ZnO/FTO. The 100 nm thick Au was
sputtered as the top electrode to form Ohmic contact
with Cu2O. Figure 5(b) and inset present
cross‐sectional and top view of the device, where
densely packed ZnO NRs and typical Cu2O
morphology can be seen. Time‐resolved
photoresponse at 0 V under different compressive
9
strains is described in Fig. 5(c), the light was
illuminated from Cu2O side as Fig. 2. It can be
observed that the photoresponse decrease step by
step with increasing compressive strain.
Contrary to the previously device, permanent
negative piezopotential is produced at ZnO/Cu2O
interface in this circumstance, as showed in Fig. 5(d),
which raises the conduction band of ZnO and the
space charge region in Cu2O shrinks as follows:
2 2 2
2
2
0 , ,
,
2 Cu O bi Cu O pz Cu O
pz Cu OCu OqN
(4)
The reduced depletion region results in the decrease
of generated excitons under light excitation and
weakens the strength of built‐in electric field in Cu2O.
Therefore, the recombination of electron‐hole pairs is
enhanced, which jeopardizes the photoresponse
property.
Figure 5. (a) Schematic of the control device. (b) Typical scanning electron microscopy image of synthesized ZnO/Cu2O. (c) Time dependent photoresponse of the control device under different compressive strains. (d) Schematic band structure of ZnO/Cu2O heterojunction with and without the appearance of negative piezopotential, as shown in solid red and dotted blue respectively.
In addition, the reversibility of piezo‐modulated
photoresponse under different strain was also tested,
as presented in Fig. 6. The results indicate that
photocurrent could return to its original value when
decreasing the compressive strains gradually.
Because the strain applied in our work is less than
1%, which is in the extent of elastic deformation of
ZnO NR. The piezopotential remains in the crystal as
the strain remains, while the potential vanishes with
the release of strain.
10
Figure. 6 Reversibility test of the piezoelectric modulation
with decreasing the applied compressive strain.
From the above, we can conclude that through
proper structure design, an enhanced
photoresponse of Cu2O/ZnO heterojunction could
be achieved. The result also indicates that the
observed enhancement is dominated by the polar
piezotronic effect rather than any other nonpolar
effect, such as piezoresistance effect, defects state
or contact reasons.
3 Conclusion In conclusion, an enhanced photoresponse of
all‐oxide Cu2O/ZnO heterostructure has been
achieved with piezo‐modulated interface
engineering. Under the illumination density of 17.2
mw/cm2, a photoresponse increase of 18.6% is
obtained when applying a ‐0.88% compressive
strain. Our results also demonstrate the
illumination‐density‐dependent photoresponse
enhancement due to the screening effect, a 2.2%
response enhancement per 0.1% strain could be
obtained under the illumination density of 17.2
mw/cm2, while only 1.2% for 87.8 mw/cm2.
Comparative experiment has confirmed that this
enhancement primarily originates from the polar
piezotronic effect. The results reported in this
paper may provide a new alternative route to
improve the performance of all‐oxide
optoelectronics and offer prototypical support for
future devices fabrication.
Acknowledgements
This work was supported by the National Major
Research Program of China (2013CB932602), Major
Project of International Cooperation and Exchanges
(2012DFA50990), the Program of Introducing
Talents of Discipline to Universities, NSFC
(51232001, 51172022, 51372023), the Research Fund
of Co‐construction Program from Beijing Municipal
Commission of Education, the Fundamental
Research Funds for the Central Universities, the
Program for Changjiang Scholars and Innovative
Research Team in University.
Electronic Supplementary Material:
Supplementary material (SEM and TEM
characterization of the sputtered ZnO seed layer,
quantitative calculation of space charge region
change in Cu2O under different strain) is
available in the online version of this article at
http://dx.doi.org/10.1007/s12274‐***‐****‐*
(automatically inserted by the publisher).
References
[1] Mannhart, J.; Schlom, D. G. Oxide interfaces—an
1 State Key Laboratory for Advanced Metals and Materials, School of Materials Science and Engineering, University of Science and
Technology Beijing, Beijing 100083, People’s Republic of China. 2 Key Laboratory of New Energy Materials and Technologies, University of Science and Technology Beijing, Beijing 100083, People’s
Republic of China. § These two authors contributed equally to this work.
Supporting information to DOI 10.1007/s12274-****-****-* (automatically inserted by the publisher)
S1. Characterization of the RF Sputtered ZnO Seed Layer
RF sputtering parameters such as vacuum pressure, sputtering power, gas flow are manipulated elaborately to
optimize the properties of seed layer. The base vacuum pressure was evacuated to 4×10‐4 Pa. The radio
frequency power and Argon flow rate were set to 80 W and 30 SCCM, respectively. The sputtering vacuum was
maintained at 0.8 Pa.
Sputtered seed layer on Si substrate with the same parameters in the paper was prepared for characterization.
Typical SEM images of the top and cross‐sectional view are depicted in Fig. S1 (a) and (b), which shows that
the as‐grown seed layer consists of multiple mesoscopic columnar grains. It is reported c‐axis is the preferred
growth direction of ZnO synthesized with RF sputtering method [1]. Alignment of the c‐axis within these
columnar grains could also give rise to macroscopically observed piezoelectricity of the ZnO polycrystalline
seed layer, as presented in Fig. S1 (c).
Figure S1. (a) and (b) Top and cross-sectional SEM images of the sputtered ZnO seed layer. (c) Schematic and low-magnification TEM image of the mesoscopic ZnO grains.
14
S2. Calculation of the Space Charge Region under different strains
As shown in Fig. 1(d), the carrier concentrations of Cu2O (Np) and ZnO (Nn) were calculated to be 3.1×1015 cm‐3
and 1.55×1017cm‐3, respectively. Due to the large discrepancy between Np and Nn, Cu2O/ZnO heterojunction
could be treated as single‐side abrupt n+p junction approximately and most of the space charge region falls in
the Cu2O side.
Therefore, the width of space charge region W could be simplified as [2]:
2 s bi R
p
V Vw
eN
(s1)
Where Vbi is the built‐in potential, VR is the bias voltage, �s is the dielectric constant. The depletion capacitance
C can be determined by equation:
2 21 bi R
s p
V V
C e N
(s2)
By fitting the curve of 1/C2 versus VR, Vbi could be obtained from the intercept. Then, width of depletion region
can be calculated through equation (s1).
Capacitance‐Voltage measurement of Cu2O/ZnO heterojunction under different strains was carried out to
determine the change of Vbi and depletion width W. The setting test parameters is the same as used in the
paper.
Figure S2. (a-g) Linear fitting of the curve 1/C2 versus VR. (h) Calculated change of built-in bias in Cu2O side under different compressive strain. (i) Calculated depletion width in Cu2O side under different compressive strain
15
Figure S2 (a) shows the 1/C2 versus VR of Cu2O/ZnO heterostructure without the presence of compressive
strain, a built‐in bias of 0.65 V in Cu2O could be obtained, which is close to the reported value [3]. The
calculated Vbi under different compressive strain is present in Fig. S2 (b‐g), it is clear that the built‐in bias in
Cu2O increases with the increase of applied strain. And the change of built‐in bias in Cu2O has a linear
relationship with the strain, as indicated in Fig. S2 (h). The comparable change of Vbi (~ 0.2 V per 1.0% strain calculated from Fig. S2 (h)) with the original (~ 0.65 V) suggests the possibility of effective interface
engineering with piezopotential.
By inserting all the known Vbi, Np and �s (6.2) of Cu2O into equation (s1), the corresponding depletion width
could be calculated, as shown in Fig. S2 (I).
References
[1] Wen, X.; Wu, W.; Ding, Y.; Wang, Z. L. Piezotronic effect in flexible thin-film based devices. Advanced Materials