High-performance perovskite memristor based on methyl ammonium lead halides

Article abstract of DOI:10.1039/c6tc00141f

Memristors are devices that can store and process information based on their switchable internal resistance. Although these devices offer better performance than the conventional technology, the use of materials such as complex metal oxides usually requires high-temperature annealing processing or vacuum processing such as sputtering, which complicates the fabrication of the devices and hinders their development for practical use. Here we show a high-performance memristor based on organometal trihalides and electrochemical active metals, which achieved an on-off current ratio of 1.9 × 10⁹. The devices can be solution-processed at low temperature and in air, which may be further developed into printable electronics. We explored the influence of different metal electrodes and device structures on memristor performance and the results indicated the great potential of methyl ammonium lead halide perovskite for information storage and computing. Our work provides new application prospects for these materials and may also contribute to the better understanding of other perovskite-based optoelectronic devices.

Full text of DOI:10.1039/c6tc00141f

View Article Online
View Journal
Journal of
Materials Chemistry C
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: K. Yan, M. Peng, X.
Yu, X. Cai, S. Chen, H. Hu, B. Chen, X. Gao, B. Dong and D. Zou, J. Mater. Chem. C, 2016, DOI:
1
0.1039/C6TC00141F.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/materialsC

Page 1 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
High-Performance Perovskite Memristor Based on Methyl Ammonium Lead Halides
1
1
1
1
1
1
1
1
Kai Yan , Ming Peng , Xiao Yu , Xin Cai , Si Chen , Hsienwei Hu , Buxin Chen , Xue Gao , Bin
1
1, 2, *
Dong , Dechun Zou
1
Beijing National Laboratory for Molecular Sciences, Key Laboratory of Polymer Chemistry and
Physics of Ministry of Education, Center for Soft Matter Science and Engineering, College of
Chemistry and Molecular Engineering, Peking University, Beijing 100871, China. Email:
dczou@pku.edu.cn
2
Beijing Engineering Research Center for Active Matrix Display, Peking University, Beijing
1
00871, China.
Abstract
Memristors are devices that can store and process information based on their switchable internal
resistance. Although these devices offer better performance than the conventional technology, the
use of materials such as complex metal oxides usually requires highꢀtemperature annealing
processing or vacuum processing such as sputtering, which complicates the fabrication of the
devices and hinders their development for practical use. Here we show a highꢀperformance
memristor based on organometal trihalides and electrochemical active metals, which achieved onꢀ
9
off current ratio of 1.9×10 . The devices can be solutionꢀprocessed at low temperature and in air,
which may be further developed into printable electronics. We explored the influence of different
metal electrodes and device structures on memristor performance and results indicated the great
potential of methyl ammonium lead halide perovskite for information storage and computing. Our
work provides new application prospects for these materials and may also contribute to the better
understanding of other perovskiteꢀbased optoelectronic devices.
1

Journal of Materials Chemistry C
Page 2 of 16
View Article Online
DOI: 10.1039/C6TC00141F
Memristors are devices that can store and process information based on their switchable internal
resistance, and with the everꢀincreasing demands for much denser memory and computing devices,
[
1ꢀ3]
they have drawn more and more attention due to the potential to go beyond Moore’s Law
.
Specifically, memristors can be mainly classified into two categories, namely, anionic and cationic
[3]
devices . The former uses electrochemically inert metallic materials, such as Au and Pt, which
[
4ꢀ6]
function only as electrodes , whereas the latter uses electrochemically active materials, such as
[7ꢀ9]
Ag and Cu, which participate in certain chemical reactions and serve as cation sources . The key
to get a robust memristor device is to find a suitable material that can endow the device with highꢀ
speed, nonꢀvolatile and lowꢀenergy switching property. Among all the reported materials, several
perovskite oxides (e.g. BiFeO
3
, BaTiO
3 3
, SrTiO ) have been successfully made into memristor
[10ꢀ13]
devices
, which show promising device performance. However, this class of materials usually
requires highꢀtemperature annealing processing or vacuum processing such as sputtering, which
complicates the fabrication of the device and hinders their development for practical applications.
Thus searching for alternative materials that can be prepared under mild condition has been one of
the main concerns in this cuttingꢀedge field.
Methyl ammonium lead halide perovskite has attracted growing research interest as a potential
[14ꢀ16]
materials for nextꢀgeneration solar cells and lightꢀemitting devices in recent years
. Except for
tunable optical band gap and efficient photoelectric conversion, its high charge mobility could
contribute to other electronic devices. The interface of active metals and perovskite materials is not
preferable in typical perovskite solar cells because their chemical reaction would impair
[17]
photovoltaic performance . Thus the opportunity of integrating the good charge transport and
chemical activity of this solid Lewis acid for other functional electronics is open.
3 3 X
In this work, highꢀperformance cation memristors based on CH NH PbCl I3ꢀX perovskite and
chemically active materials were fabricated. The device achieved onꢀoff current ratio as high as
9
1
.9×10 , which is among the highest values of any kinds of memristors reported thus far. Because
methyl ammonium lead halide perovskite can be prepared using solution process at low temperature
2

Page 3 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
and in air, the proposed method can greatly simplify memristor fabrication and reduce the cost of
largeꢀscale production.
To fabricate perovskite memristor, a device with a vertical structure was designed. As shown in
Figure 1a, it was constructed with a fluorineꢀdoped tin oxide (FTO) glass substrate, on which a
3 3 X
perovskite layer (CH NH PbCl I3ꢀX) was formed by spinꢀcoating with a counter metal electrode
sputtered on it. Devices with various metal electrodes were prepared and the I-V characteristics
were tested. Figure 1b shows the typical I-V characteristics of the FTO/perovskite/Au (MꢀAu) and
FTO/perovskite/Ag (MꢀAg) devices. The device consisting of inert metal Au as the counter
electrode showed only the typical I-V characteristics of a resistor, whereas devices consisting of
electrochemically active metal Ag exhibited the characteristic I-V curve of a typical memristor,
[1ꢀ3]
including the following features : (1) the CV result shows a pinched hysteresis loop, (2) the
pinched point occurs at the origin (0, 0), and (3) the resistance state (highꢀresistance state or lowꢀ
resistance state) of the device remains constant while the applied voltage is changed until the
switching point is reached. Similar phenomena were observed using other metals (Figure 2 and
S1). These results indicate that methyl ammonium lead halideꢀbased memristors tend to be cationic
devices and thus must be incorporated with electrochemically active materials to serve as an
electrode. This also hints us that besides transporting current, the active metal electrode may also
take part in certain electrochemical reactions.
Figure 1 a Schematic of a perovskite memristor device. b I-V characteristics of FTO/perovskite/ Au
(MꢀAu) and FTO/perovskite/Ag (MꢀAg) devices at a sweeping rate of 0.1 V/s under dark
3

Journal of Materials Chemistry C
Page 4 of 16
View Article Online
DOI: 10.1039/C6TC00141F
conditions. The arrows and numbers represent the order of the output current during the I-V test of
MꢀAg, which reveals obvious changes of the device’s resistance state. Fluctuations in the output
current may be a result of both the test system itself (the test electrodes only feature physical
contact with the device) and the complex physical and chemical processes that occur during the test.
As to cation memristor devices, choosing an appropriate electrochemically active electrode material
is critical for good device performance. It is supposed to have a suitable electrochemical activity,
which is enough to serve as cation sources but not over certain level that affects the device
3
stability . For the future largeꢀscale application, the price factor is also a key criterion that needs to
be taken into consideration. We integrated various active metals, including Ag, Cu, Ti, Zn and Al,
into the perovskite memristor as one of the electrode and the devices’ I-V characteristics were
showed in Figure 2 (the logꢀscale version was showed in Figure S2) and the detail parameters were
summarized in Table 1. Although all the devices exhibited memristive properties, significant
differences of their I-V characteristics could be observed. Firstly, using different active metal
electrodes, the device shows different turnꢀon and turnꢀoff voltages. These switching voltages
neither equal nor follow the order of the standard electrode potential of the corresponding metals,
which indicated that a more complicated process might exist. Except for the turnꢀoff voltage of Mꢀ
Zn, other switching voltages are all below 2 V, which can be further reduced by simply decreasing
[18,19]
the film thickness of the perovskite layer
. Another significant difference among the devices is
that only MꢀAl and MꢀZn showed instantaneous switching phenomena between low and high
resistance states, whereas resistance gradually changed in MꢀAg, MꢀCu and MꢀTi. Considering a
[20]
popular physical model published in 2008 , this difference may result from whether the key
parameter w/D can reach 1 or not; here, D represents the thicknesses of the perovskite film and w
represents the length of the nanoꢀscale metal filaments in the semiconductor film. When the nanoꢀ
scale metal filaments form contact between the FTO and the metal electrode (w/D = 1), the
resistance of the device abruptly drops, causing the instantaneous current change seen in the I-V
4

Page 5 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
characteristics tests. By contrast, if the nanoꢀscale metal filaments do not connect the two electrodes
during the I-V characteristics test, no sudden change of the output current is observed.
Despite their obvious differences, the devices consisting of different metal electrodes showed
−
2
similar orders of magnitude (10 A) of the corresponding current at a certain voltage when they
were switched to the lowꢀresistance state (onꢀstate) in the I-V characteristics test. This result
supports the assumption that the memristive property originates from the formation of metal
filaments in the perovskite film because the conductivities of the metals are of the same orders of
magnitude. By comparing the overall I-V characteristics performance of the MꢀAg, MꢀCu, MꢀTi,
MꢀZn and MꢀAl devices, especially their instantaneous switching ability which is practically useful,
and also the price factor, we selected Al as the active metal electrode for subsequent experiments.
Figure 2 I-V characteristics of MꢀAg, MꢀCu, MꢀTi, MꢀZn and MꢀAl. The inset shows that the highꢀ
−6
resistance state current of MꢀAl is approximately10 A. MꢀZn and MꢀAl show instantaneous
switching phenomena which are marked by arrows.
Table 1 Summarized parameters from Figure 2 and Figure S2.
5

Journal of Materials Chemistry C
Page 6 of 16
View Article Online
DOI: 10.1039/C6TC00141F
Device
Turn-on Voltage
Turn-off Voltage
High-Risistance Current
(A)
Low-Risistance Current at 1V
(A)
Instantaneous
Price*
Type
MꢀAg
MꢀCu
MꢀTi
(V)
(V)
Switching
(RMB / Kg)
3390
ꢀ4
ꢀ3
ꢀ3
0.35
0.35
0.65
1.80
1.10
ꢀ1.40
ꢀ0.45
ꢀ1.20
ꢀ2.55
ꢀ1.65
10 ~10
2.72×10
3.05×10
2.46×10
2.85×10
3.81×10
No
No
ꢀ7
ꢀ6
ꢀ3
ꢀ3
ꢀ3
ꢀ3
10 ~10
39.5
ꢀ
10
ꢀ8
10 ~10
No
47.5
ꢀ6
ꢀ3
MꢀZn
MꢀAl
10 ~10
Yes
Yes
14.2
ꢀ7
ꢀ6
10 ~10
10.4
*
The price data are from www.smm.cn.
Because the lowꢀresistance state current is in the same orders of magnitude according to the metal
filament model, the highꢀresistance state current must be maintained at the lowest possible value to
ꢌꢍꢎꢏꢐꢑꢒꢓꢔꢓꢕꢖꢗꢘꢒꢃꢓꢕꢖꢕꢒ
obtain a high onꢀoff current ratio ( ꢀꢁ − ꢀꢂꢂꢃꢄꢅꢆꢆꢇꢁꢈꢃꢆꢉꢈꢊꢋ = ꢌ
). The highꢀ
ꢙꢔꢚꢙꢐꢑꢒꢓꢔꢓꢕꢖꢗꢘꢒꢃꢓꢕꢖꢕꢒ
−6
resistance state current of MꢀAl was already as low as 10 A (Inset of Figure 2), indicating that the
perovskite film functioned as an insulating layer in the device. Thus, to reduce the highꢀresistance
state current further, adding additional insulating layers into the device may be a feasible method.
We used a semiconducting TiO
leakage current in perovskite solar cells, as an interlayer between FTO and the perovskite film
cTiO ꢀMꢀAl). Figure 3a showed the I-V characteristics result of the device which indicated that
2
compact layer, which is a commonly used function layer to reduce
(
2
−
11
−10
the additional interlayer could reduce the highꢀresistance state current to 10 –10 A (the logꢀ
scale version was showed in Figure S3). The lowꢀresistance state current remained at
−2
approximately 10 A, thereby increasing the onꢀoff current ratio by 4–5 orders of magnitude, that
8
9
9
is, 10 –10 (the maximum value was 1.9×10 at 0.14 V), which, to the best of our knowledge, is
[
21,22]
among the highest values reported thus far
. Higher onꢀoff current ratio leads to faster reading,
[23]
fewer errors and less power consumption . With onꢀoff current ratio as high as billions, the asꢀ
fabricated perovskite memristor shows great potential for future practical use. Moreover, by adding
an extra compact layer, a pinholeꢀfree and improved crystallized perovskite film (Figure 3b and 3c),
which could reduce the possibility of shortꢀcircuit, was obtained. To further investigate the TiO
2
compact layer’s effects, a device with the structure of FTO/TiO compact layer/Al (cTiO ꢀAl) was
2
2
fabricated and its I-V characteristics results are showed in Figure S4. The device also showed
6

Page 7 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
[3]
memristive property as expected, because TiO is a traditional memristive material . The highꢀ
2
−
11
−10
resistance state current was about 10 –10 A, which equaled that of cTiO
because TiO compact layer that highꢀresistance state current of cTiO ꢀMꢀAl could be this small.
The switching voltages of cTiO ꢀAl were higher than those of MꢀAl, and in the negative voltage
2
ꢀMꢀAl. Thus it is
2
2
2
part, a very large reset voltage (ꢀ5 V) was needed to totally turn off the device, which is clearly
disadvantageous for device performance. In contrast, the turnꢀon and turnꢀoff voltage of cTiO ꢀMꢀ
2
Al were both smaller than 2 V and almost same as those of MꢀAl, which showed the perovskite
2
layer’s role in maintaining a small switching voltage. Thus, by integrating TiO compact layer into
MꢀAl, we could decrease the highꢀresistance state current without affecting the switching voltages.
Further study is needed for revealing the specific physical and chemical process happened in the
interface of TiO and perovskite and their effects on device performances.
2
The device was kept and tested in an inert gasꢀfree atmosphere to determine its stability, several
tests methods were used, including consecutive I-V characteristics test, longꢀinterval I-V
characteristics test and pulseꢀvoltage test. The results (Figure S5) revealed the perovskite
memristor’s good stability, which represents another vital index for practical use. A more stable
device performance can be expected by applying a suitable packaging design to the device.
Performance comparisons between our device and other typical memristors, including a recently
3 3 X
published paper that based on CH NH PbCl I3ꢀX hybrid perovskite (the same material we used
here) and chemically inert metal Au, were summarized in Table 2. Our device showed competitive
device performance with other oxideꢀbased memristors and compared with their highꢀtemperature
annealing process or vacuum processing such as sputtering, the solution process of
3 3 X
CH NH PbCl I3ꢀX showed distinct advantages in simplifying device fabrication and reducing
preparation cost. Specially, compared with the paper mentioned above, our device showed
comparable small switching voltages but far bigger onꢀoff current ratio which is advantageous for
[
23]
achieving faster reading, fewer errors and less power consumption . The first reason for the
difference was the using of electrochemically metals rather than inert metals in our devices, which
7

Journal of Materials Chemistry C
Page 8 of 16
View Article Online
DOI: 10.1039/C6TC00141F
endowed our devices with cation memristive property. And compared with anion memristors using
[
3]
inert metals, cation memristors usually yield bigger lowꢀresistance current , which could lead to
higher onꢀoff current ratio. The second reason was the using of TiO compact layer, resulting in a
2
ꢀ11
ꢀ4
much smaller highꢀresistance current (10 A) to that of the reported work (>10 A), which could
further increase the onꢀoff current ratio.
Figure 3 a I-V characteristics test results of cTiO ꢀMꢀAl. The inset shows that the highꢀresistance
2
−
10
−12
state current of the device is reduced to less than 1×10 A (3×10 A at 0.14 V). b & c SEM
image and XRD spectra of the FTO/CH
3
NH
3
PbI3ꢀXCl
X
perovskite layer and the FTO/TiO
2
compact
layer/CH NH PbI3ꢀXCl perovskite layer.
3 3 X
Table 2 A brief comparison of device performances
On/Off Voltage
V)
OnꢀOff Current
Ratio
Instantaneous
Switching
Preparation
Method
Comparison with
our device
Material
Ref.
(
9
CH
3
3
NH
3
3
PbCl
X
I
3ꢀX
+1.10/ꢀ1.65
+0.90/ꢀ0.63
+3.0/ꢀ1
10
Yes
No
Spinꢀcoating
Spinꢀcoating
Sputtering
This study
Small onꢀoff current ratio;
No instantaneous switching;
Small onꢀoff current ratio
CH
NH
PbCl
X
I
3ꢀX
< 10
[24]
[25]
[26]
2
Vacuum condition;
High switching voltage
SiO
2
10
Yes
Yes
Vacuum condition;
455℃ annealing
Small onꢀoff current ratio
Hydrothermal
growth
TiO
2
+0.25V/ꢀ2V
<10
8

Page 9 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
Vacuum condition;
00℃ annealing
2
SrZrO
3
ꢀ2.0/+0.8
ꢀ1.1/+2.6
10
Yes
Yes
PLD
[27]
[28]
7
Vacuum condition;
11~13 V pretreatment;
Small onꢀoff current ratio
aꢀSrTiO
3
*
<100
Sputtering
*
“a” is the abbreviation for amorphous.
As mentioned above, a possible mechanism by which memristive property is conferred to the
perovskite materials is formation/deformation of nanoꢀscale metallic filaments in the perovskite
[7ꢀ9,18,19,22]
film
. The intuitive evidence here is that only during combination with electrochemically
active metals could perovskite exhibit memristive property, thereby indicating that the metal
electrode functioned as a metal ion source. To verify this speculation, an oppositeꢀdirection I-V
characteristics test was conducted, that is, negative voltage scanning was applied prior to positive
scanning. When negative voltage scanning was applied firstly to the device instead of positive
scanning, no metal cations could be drawn to the FTO side to be reduced to form metal filaments
due to the internal electric field direction; thus, the device remains in the highꢀresistance state
(Figure 4a). The device could not be switched to the lowꢀresistance state until the positive scanning
voltage was applied. Further more, we measured the device’s onꢀstate resistance with different
temperatures. The result (Figure 4b) showed the onꢀstate resistance increase with the increase of
temperature and have simple metallic dependence on it. The linear equation of resistanceꢀ
temperature derived from the test was ꢛꢜ = 37.827 × (1 + 0.00402 × ꢈ). The derived metallic
temperature coefficient (0.00402/ ℃) is very close to the experimental value of aluminum
(0.00429/℃), suggesting the existence of aluminum conducting channel inside the perovskite
memristor. Although these methods’ results are strong evidence for the metal filament model,
they’re all indirect ways to support this model; the most direct way is inꢀsitu observation of the
filament, which is difficult to achieve because of its nanoꢀscale size. However, it is this property
that confers memristor based storage or computing devices with their ultraꢀhigh density
[3,29,30]
feature
.
9

Journal of Materials Chemistry C
Page 10 of 16
View Article Online
DOI: 10.1039/C6TC00141F
st
nd
Figure 4 a 1 (left) and 2 (right) loop of the oppositeꢀdirection I-V characteristics test of MꢀAl.
The device remained in the highꢀresistance state when the negative voltage was first applied to it
(
marked as 1 and 2), indicating that the moving ions are positive. b Onꢀstate resistance of cTiO
2
ꢀMꢀ
Al under different temperatures.
Based on the I-V characteristics test of MꢀAl described above, the lowꢀresistance state current was
.02 A at 0.5 V, indicating that the resistance was approximately 25 ꢁ. Assuming that (1) the
0
resistance was contributed entirely by the aluminum filaments, (2) the perovskite was
approximately 400 nm thick (Figure S6), which was also the length of the metal filaments, (3) the
filaments were regular cylinders (for simplifying the calculation), and (4) only one metal filament
that connected the two electrodes was formed during the I-V characteristics test, then, by using the
resistance formula ꢛ = ꢝꢞ⁄ꢟ, the diameter of the filament can be determined to be approximately
1
0 nm. Although the actual diameter was likely much smaller considering the errors attributable to
the above assumptions, the derived data are sufficient to show the great potential of the perovskite
1
0

Page 11 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
memristor in producing ultraꢀhigh density storage or computing devices. A feasible method to
validate the fourth assumption is to examine the relationship between the onꢀstate resistance and the
area of the active metal electrode. The result (Figure S7) showed that the onꢀstate resistance of the
devices with varying electrode areas was nearly similar, indicating that the total crossꢀsectional area
of the metal filaments was constant. Thus, only one metal filament most likely existed.
Futher more, we noticed a lot of reported works indicated that the hysteresis phenomena observed
in the I-V characteristics test of perovskite solar cells might be related to the mobile ions in the
[31ꢀ33]
perovskite layer
. As we showed here, the mobile ion was also the origin of the perovskite’s
memristive property; thereby, it may assist in determining the relationship between the hysteresis
[17,34]
and the memristive property
. This brand new way of revealing the confounding hysteresis
phenomena in perovskite solar cells is worth of further investigation.
In summary, our work demonstrated a highꢀperformance memristor based on the methyl ammonium
lead halide perovskite material. By choosing an appropriate electrochemically active metal
2
electrode and also integrating with TiO compact layer, we got a memristor with onꢀoff current ratio
9
as high as 1.9×10 , which is among the highest values of any kinds of memristors reported thus far.
Such high onꢀoff current ratio may lead to faster reading, fewer errors and less power
[23]
consumption . Further more, the solution processing of methyl ammonium lead halide perovskite
will greatly reduce the largeꢀscale production cost of perovskite memristors and enable the
production of printable devices. Except for the potential practical application of the proposed
3 3 X 3ꢀX
memristor in information storage or computing, the memristive property of CH NH PbCl I
displayed by our device may contribute to the better understanding of other perovskite based
optoelectronic devices and open new opportunities for this material.
Experimental Section
Materials: All of the basic materials used in this work were purchased from SigmaꢀAldrich and
applied directly without further purification unless otherwise stated.
1
1

Journal of Materials Chemistry C
Page 12 of 16
View Article Online
DOI: 10.1039/C6TC00141F
Synthesis of CH NH I: CH NH I was synthesized by reacting 30 mL of CH NH and 12.5 mL of HI
3
3
3
3
3
2
in a 250 mL roundꢀbottom flask at 0 °C for 2 h with stirring. The precipitate was placed in a rotary
evaporator at 50 °C to remove the solvents. The product was dissolved in ethanol and recrystallized
using diethyl ether. Finally, the CH
Synthesis of TiO compact-layer solution: Solutions A and B were respectively prepared by adding
69 μL of titanium isopropoxide into 2.53 mL of ethanol and adding 35ꢃꢠL of 2 M HCl solution
3 3
NH I was dried at 60 °C in a vacuum for 24 h.
2
3
into 2.53 mL of ethanol. Solution B was then added dropwise into solution A with stirring; the
resulting mixture was filtered with a PFTE hydrophobic 0.45ꢃꢠm filter prior to use.
Memristor fabrication: The FTO substrate was sequentially washed with deionized water, acetone,
and ethanol followed by oxygen plasma treatment for 15 min. The perovskite layer was then formed
by spinꢀcoating with 40 wt% precursor solution prepared by dissolving CH
molar ratio of 3:1) in DMF at 2000 rmp for 30 s and annealing at 100 °C for 1 h. Finally, the
electrochemically active metal electrode was formed via magnetron sputtering using a mask to form
a circular pattern with radius of 1 mm. To prepare cTiO ꢀMꢀAl, an additonal TiO compact layer
3 3 2
NH I and PbCl mixture
(
2
2
was formed by spinꢀcoating with the compactꢀlayer solution at 2000 rmp for 45 s and then sintering
at 500 °C for 30 min.
Characterization: All of the CV tests were conducted using a PGSTAT302N instrument (Autolab
Corp., Switzerland) under dark conditions at a sweep rate of 0.1 V/s.
Acknowledgements This study is jointly supported by Ministry of Science and Technology of
China (No. 2011CB933300), National Natural Science Foundation of China (No. 91333107, No.
5
1573004) and the fund from Shenzhen City (No. CXZZ20120618162051603).
[
[
[
1]
L. Chua, Semicond. Sci. Technol., 2014, 29, 104001.
2]
3]
T. Prodromakis, C. Toumazou, L. Chua, Nat. Mater., 2012, 11, 478.
J. J. Yang, D. B. Strukov, D. R. Stewart, Nat. Nanotechnol., 2013, 8, 13.
1
2

Page 13 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
[
4]
H. Shima, F. Takano, H. Akinaga, Y. Tamai, I. H. Inoue, H. Takagi, Appl. Phys. Lett., 2007,
1, 012901.
9
[
[
[
5]
6]
7]
H.ꢀH. Huang, W.ꢀC. Shih, C.ꢀH. Lai, Appl. Phys. Lett., 2010, 96, 193505.
I. H. Inoue, S. Yasuda, H. Akinaga, H. Takagi, Phys. Rev. B, 2008, 77, 035105.
K. Tsunoda, Y. Fukuzumi, J. R. Jameson, Z. Wang, P. B. Griffin, Y. Nishi, Appl. Phys.
Lett., 2007, 90, 113501.
[
[
[
[
8]
9]
A. Nayak, T. Ohno, T. Tsuruoka, K. Terabe, T. Hasegawa, J. K. Gimzewski, M. Aono, Adv.
Funct. Mater., 2012, 22, 3606.
S. Peng, F. Zhuge, X. Chen, X. Zhu, B. Hu, L. Pan, B. Chen, R.ꢀW. Li, Appl. Phys. Lett.,
2
012, 100, 072101.
10] N. Maso, H. Beltran, M. Prades, E. Cordoncillo, A. R. West, Phys. Chem. Chem. Phys.,
014, 16, 19408.
11] M. Prades, N. Maso, H. Beltran, E. Cordoncillo, A. R. West, J. Mater. Chem., 2010, 20,
335.
2
5
[
[
12] S. I. Shablaev, A. I. Grachev, Phys. Solid State, 2014, 56, 682.
13] X. B. Yan, K. Li, J. Yin, Y. D. Xia, H. X. Guo, L. Chen, Z. G. Liu, Electrochem. Solid State
Lett., 2010, 13, H87.
[
[
[
14] M. A. Green, K. Emery, Y. Hishikawa, W. Warta, E. D. Dunlop, Progress in Photovoltaics:
Research and Applications, 2015, 23, 1.
15] W. S. Yang, J. H. Noh, N. J. Jeon, Y. C. Kim, S. Ryu, J. Seo, S. I. Seok, Science, 2015, 348,
1234.
16] T. ZhiꢀKuang, R. S. Moghaddam, M. L. Lai, P. Docampo, R. Higler, F. Deschler, M. Price,
A. Sadhanala, L. M. Pazos, D. Credgington, F. Hanusch, T. Bein, H. J. Snaith, R. H. Friend,
Nat. Nanotechnol., 2014, 9, 687.
[
17] Z. Xiao, Y. Yuan, Y. Shao, Q. Wang, Q. Dong, C. Bi, P. Sharma, A. Gruverman, J. Huang,
Nat. Mater., 2015, 14, 193.
1
3

Journal of Materials Chemistry C
Page 14 of 16
View Article Online
DOI: 10.1039/C6TC00141F
[
18] B. J. Choi, D. S. Jeong, S. K. Kim, C. Rohde, S. Choi, J. H. Oh, H. J. Kim, C. S. Hwang, K.
Szot, R. Waser, B. Reichenberg, S. Tiedke, J. Appl. Phys., 2005, 98, 033715.
19] C. Schindler, G. Staikov, R. Waser, Appl. Phys. Lett., 2009, 94, 072109.
20] D. B. Strukov, G. S. Snider, D. R. Stewart, R. S. Williams, Nature, 2008, 453, 80.
21] S. G. Hahm, S. Choi, S.ꢀH. Hong, T. J. Lee, S. Park, D. M. Kim, W.ꢀS. Kwon, K. Kim, O.
Kim, M. Ree, Adv. Funct. Mater., 2008, 18, 3276.
[
[
[
[
[
22] N. Banno, T. Sakamoto, N. Iguchi, H. Sunamura, K. Terabe, T. Hasegawa, M. Aono, IEEE
Trans. Electron Devices, 2008, 55, 3283.
23] A. Chen, J. Hutchby, V. Zhirnov, G. Bourianoff, Emerging Nanoelectronic Devices, John
Wiley & Sons Ltd, Chichester, United Kindom 2014.
[
[
24] E. J. Yoo, M. Lyu, J.ꢀH. Yun, C. J. Kang, Y. J. Choi, L. Wang, Adv. Mater., 2015, 27, 6170.
25] S. Gao, C. Chen, Z. Zhai, H. Y. Liu, Y. S. Lin, S. Z. Li, S. H. Lu, G. Y. Wang, C. Song, F.
Zeng, F. Pan, Appl. Phys. Lett., 2014, 105, 063504.
[
[
26] W. Wu, N. Zhang, B. Shan, K. Feng, Mater. Chem. Phys., 2015, 156, 76.
27] M. X. Li, J. Miao, S. Z. Wu, Q. L. Liu, Y. Jiang, H. Yang, L. J. Qiao, J. Alloys Compd.,
2013, 548, 1.
[
[
28] H. Nili, S. Walia, A. E. Kandjani, R. Ramanathan, P. Gutruf, T. Ahmed, S. Balendhran, V.
Bansal, D. B. Strukov, O. Kavehei, M. Bhaskaran, S. Sriram, Adv. Funct. Mater., 2015, 25,
3172.
29] B. Govoreanu, G. S. Kar, Y. Chen, V. Paraschiv, S. Kubicek, A. Fantini, I. P. Radu, L.
Goux, S. Clima, R. Degraeve, N. Jossart, O. Richard, T. Vandeweyer, K. Seo, P. Hendrickx,
G. Pourtois, H. Bender, L. Altimime, D. J. Wouters, J. A. Kittl, M. Jurczak. 10times10nm 2
Hf/HfO x crossbar resistive RAM with excellent performance, reliability and low-energy
operation. IEEE International Electron Devices Meeting (IEDM), 2011, 31.6.1.
[
30] A. Kawahara, R. Azuma, Y. Ikeda, K. Kawai, Y. Katoh, K. Tanabe, T. Nakamura, Y.
Sumimoto, N. Yamada, N. Nakai, S. Sakamoto, Y. Hayakawa, K. Tsuji, S. Yoneda, A.
1
4

Page 15 of 16
Journal of Materials Chemistry C
View Article Online
DOI: 10.1039/C6TC00141F
Himeno, K. Origasa, K. Shimakawa, T. Takagi, T. Mikawa, K. Aono. An 8Mb multi-layered
cross-point ReRAM macro with 443MB/s write throughput. IEEE International Solid-State
Circuits Conference (ISSCC), 2012, 432.
[
[
[
31] H.ꢀW. Chen, N. Sakai, M. Ikegami, T. Miyasaka, J. Phys. Chem. Lett., 2015, 6, 164.
32] J. S. Manser, P. V. Kamat, Nat. Photonics, 2014, 8, 737.
33] E. L. Unger, E. T. Hoke, C. D. Bailie, W. H. Nguyen, A. R. Bowring, T. Heumueller, M. G.
Christoforo, M. D. McGehee, Energy Environ. Sci., 2014, 7, 3690.
[
34] Y. Zhao, C. Liang, H. Zhang, D. Li, D. Tian, G. Li, X. Jing, W. Zhang, W. Xiao, Q. Liu, F.
Zhang, Z. He, Energy Environ. Sci., 2015, 8, 1256.
1
5

Journal of Materials Chemistry C
Page 16 of 16
View Article Online
DOI: 10.1039/C6TC00141F
High-performance memristors were fabricated based on methyl ammonium lead
9
halides perovskite material and achieved on-off ratio of 1.9×10 .
+
ꢀ
Metal(electrodeꢀ
Perovskite(layerꢀ
.ꢀ
low.resistance(state(ꢀ
0ꢀ
FTO(substrateꢀ
high.resistance(state(ꢀ
0ꢀ
Voltage(ꢀ