Bias-controlled Tunable Electronic Transport with Memory

Subscriber access provided by OCCIDENTAL COLL

Functional Nanostructured Materials (including low-D carbon)

Bias-controlled Tunable Electronic Transport with Memory Characteristics in an Individual ZnO Nanowire for Realization of Self-driven UV Photodetector with Symmetrical Two Electrodes Shujuan Wang, Jie Zhao, Tao Tong, Baochang Cheng, Yanhe Xiao, and Shuijin Lei ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.9b00267 • Publication Date (Web): 28 Mar 2019 Downloaded from http://pubs.acs.org on March 28, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Bias-controlled Tunable Electronic Transport with Memory Characteristics in an Individual ZnO Nanowire for Realization of Self-driven UV Photodetector with Symmetrical Two Electrodes Shujuan Wang,† Jie Zhao,‡ Tao Tong,† Baochang Cheng,†,‡,* Yanhe Xiao,‡ and Shuijin Lei,‡ †

Nanoscale Science and Technology Laboratory, Institute for Advanced Study, Nanchang University,

Jiangxi 330031, P. R. China, and ‡ School of Materials Science and Engineering, Nanchang University, Jiangxi 330031, P. R. China ABSTRACT: For ZnO nanostructures, they are exceedingly important building blocks for nanodevices due to wide bandgap and large exciton binding energy. However, their electronic transport characteristics are unstable and unrepeatable with external environment variation. Here, we demonstrate that electron transport of an individual ZnO nanowire-based device with two same electrodes can controllably be modulated by applying relatively large uni-/bi-directional bias. After being modulated, moreover, their electrical properties can well be maintained at relatively low operation bias and room temperature, demonstrating a memory behavior. The presence of surface states related to lattice periodicity breaking and traps associated with oxygen vacancy (Vo) and zinc interstitial (Zni) deep level defects plays a crucial role in tunable electron transport with memory feature. For the single nanowirebased two-terminal device, two back-to-back connected surface barrier diodes with series resistance are formed. The filling and emptying of traps near two end electrodes can remarkably adjust surface barrier width and height. At a relatively low bias, the unmodulated conductance is governed by the electron hopping of bulk traps since the height of emptied traps is higher than that of surface barrier. At a

*

Address correspondence to [email protected]. ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

relatively large bias, whereas, it is dominated by thermion emission due to a dramatic decrease of surface barrier width resulting from the electron injection into traps from negative electrode. Moreover, it will be beneficial for thin surface barrier to penetrate UV light and separate photoexcited electronhole pairs. After being asymmetrically modulated by a unidirectional injection, it can successfully be applied to realize a self-driven UV photodetector based on a photovoltaic effect in the symmetrical twoelectrode structure. Our work provides a new route to tunable electrical properties of nanostructures, which may inspire the development of novel electronics and optoelectronics.

KEYWORDS: ZnO · electronic transport · bias modulation · traps · memory · self-driven photoconductor.

ACS Paragon Plus Environment

2

Page 3 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. INTRODUCTION ZnO is a wide direct bandgap semiconductor of 3.374 eV at 300 K, having high thermal conductivity, electronic mobility, and large exciton binding energy of 60 meV.1,2 Therefore, it has attracted a strongly increasing interest since it can be applied in electronic and optoelectronic devices including field effect transistor,3,4 nanogenerator,5 resistive switching,6,7 gas sensor,8,10 memory,11,12 photodetector,13-20 and piezoelectric diode.21-23 For nanostructured ZnO with very large surface-to-volume ratio and typical ntype properties, however, dangling bonds can induce quantities of acceptor-type surface states due to a breaking of lattice periodicity on its surface, resulting in band bending upward and carrier-depletion layer in the vicinity of surfaces, and correspondingly surface barrier-related diode can be formed.7,13,18 Moreover, high densities of surface states can give rise to Fermi level pinning, and hence its interface barrier is independent of metal work function and semiconductor electron affinity. For ZnO nanostructure-based devices with high surface barrier, therefore, it is difficult to conduct at a low operation bias, showing a high resistance state (HRS)13,18,24,25 In addition, quantities of defects, such as oxygen vacancy (Vo) and zinc interstitial (Zni), exist in ZnO lattice, resulting in the formation of traps with different levels. The filling and emptying of traps will intensively affect the performance associated with electron transport as well, and thus abundant surface states and traps will play a vital role in the physical properties of nanostructures.7,11,12,26 To meet the ever-increasing demand in device function, a detailed investigation of the charge transport parameters is essential to optimize and improve the performance of ZnO nanostructure-based nanodevices. In particular, it is imperative to tailor the properties by designing traps-related surface states of nanostructures4,13,27,28 or modulating surface barrier using external environment such as bias voltage, illumination, and temperature.29-38 Up to now, the modulation of electronic transport characteristics has not been investigated extensively in a single ZnO nanowire (NW)-based nanodevice, and moreover their mechanisms have not been fully understood as well. For the nanostructure semiconductors with very large surface-to-volume ratio, it is hardly for their surface states to eliminate completely, and therefore, how to make use of surface barrier


3


Page 4 of 29

to improve the performance of known nanodevices and to design novel nanodevices is still a great challenge. For pure ZnO NW-based photodetectors, in particular, they show generally long-persistent photoconductivity phenomenon due to the presence of deep level traps. 13 Therefore, it is extremely indispensable to improve response and recovery speed for practical application. In addition, selfpowered photodetectors are all almost based on a complex p-n junction configuration at present,39-42 and hence it is more meaningful to realize self-powered photodetection in a simple symmetrical structure by modulation. For a single one-dimensional (1D) nanostructure-based two-terminal device, it is prone for the origin of physical mechanism to be identified due to configuration simplicity.13,18,24,27 In this work, therefore, the two-terminal device based on a single ZnO NW with two same Ag electrodes was constructed. After being applied different large uni-/bi-directional biases, interestingly, their electron transport can controllably be modulated, which can be ascribed to the decrease of surface barrier near negative electrode end due to a large bias induced electron injection into the traps located in surface depletion region from negative electrode. Moreover, it is favorable for the modulated thin surface barrier to separate photoexcited electron-hole pairs. Therefore, the self-driven UV photodetector with fast response and recovery speed can successfully be realized in the two symmetrical electrode structure after being asymmetrically modulated by a relatively large unidirectional bias. The understanding and realization of bias-governed electrical properties make it inspire for the design and fabrication of novel electronic and photoelectronic nanodevices.

2. EXPERIMENTAL METHODS 2.1 Synthesis of ZnO nanowires. The investigated ZnO NWs were synthesized on an Al2O3 ceramic substrate via a physical vapor deposition process based on thermal evaporation of 99.99% ZnO powders at 1470 ℃ under high pure Ar atmosphere. 2.2 Characterization of structure and morphology. As-synthesized product was measured and


4

Page 5 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analyzed by X-ray diffraction (XRD, RIGAKU D/max-3b), field-emission scanning electron microscopy (FESEM, FEI Quanta 200F), and high-resolution transmission electron microscopy (HRTEM, JEOL JEM-2100) to evaluate its structure and morphology. 2.3 Fabrication and measurement of a single NW-based two-terminal device. A relatively long single ZnO NW was separated and transferred on an Al2O3 insulator substrate under an optical microscope, and then the two ends of the single NW were fixed by silver paste. Afterwards, the two Ag electrodes were annealed at 400 ℃ and hold the temperature for 60 minutes under high pure Ar atmosphere. After the furnace was cooled to room temperature, the devices were taken out and connected the two electrodes with copper wires. A single ZnO NW-based two-terminal device with two symmetrical Ag electrodes was fabricated. Electrical signal measurement was carried out with a synthesized function generator (Stanford Research System Model DS345), low-noise preamplifier (Stanford Research System Model SR560), and a low-noise current preamplifier (Stanford Research System Model SR570). For bidirectional bias modulation, relatively large forward and reverse bias voltages were consecutively applied. For unidirectional bias modulation, however, only a relatively large forward or reverse bias voltage was applied. After being modulated by different uni-/bi-directional bias voltages, a relatively low bidirectional bias voltage was used to measure the variation of electrical properties of devices. For the measurement of photoconduction, the continuous and monochromatic wavelength-light source was provided by fluorescence spectrophotometer (Hitachi F-4600) with a 150 W Xe lamp. For the selfdriven UV detection experiment, the device was first modulated by a unidirectional 10 V bias. A 370 nm UV light with power density of about 170 W/cm2 was used as excitation light resource. The opencircuit voltage (Voc) and short-circuit current (Isc) were measured under continuous irradiation at 0.5 V bidirectional bias voltage. At 0 V bias voltage, the self-driven photocurrent and photovoltage were measured as a function of time with UV light chopping with an interval of about 5 s.

3. RESULTS AND DISCUSSION ACS Paragon Plus Environment

5


3.1 Morphology and structure characterization of ZnO nanowires (a)

(101)

(100)

2000

(b)

1500

1000

(103)

(110)

(102)

500

(200) (112) (201)

(002)

Intensity (a.u.)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

0 30

40

50

60

70

2 (degree)

(c)

(d)

200 nm

Figure 1. Structural characterization of as-synthesized ZnO NWs. (a) XRD pattern, revealing that the product consists of hexagonal wurtzite structure ZnO. (b) FESEM image, showing wire-like morphology. (c) Low-magnification bright-field TEM micrograph. (d) High-resolution TEM image, the inset in (d) corresponds to its FFT pattern.

The XRD pattern of as-prepared product presents clear evidence that all the peaks can be indexed to ZnO with a hexagonal wurtzite structure (JCPDS No.79-0205), as illustrated in Figure 1a. FESEM observation of the sample, as shown in Figure 1b, reveals that the products are mainly composed of nanowires whose length ranges from several micrometers to hundreds of micrometers and diameter is about hundreds of nanometers. TEM measurement is further performed to obtain more detailed wirelike information, as depicted in Figure 1c. The high-resolution TEM lattice fringe image and its corresponding fast Fourier transform (FFT) pattern, as illustrated in Figure 1d, reveal that the ZnO NW grows along [001] direction. 3.2 Bias voltage dependence of electron transport ACS Paragon Plus Environment

6

Page 7 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a)

(b)

60 um

Figure 2. Structure diagram of a single ZnO NW-based two-terminal device with Ag electrodes. (a) Schematic diagram; (b) optical microscopic image.

To investigate the impact of external bias voltage on electronic transport characteristics, the cyclic current-voltage (I-V) curves of a single NW-based two-terminal device, whose schematic diagram and optical microscopic image were presented in Figure 2, were measured by applying different bias voltages from 1 to 10 V at room temperature, as illustrated in Figure 3 to 5. Figure 3a describes the I-V characteristics of a single ZnO NW-based two-terminal device at a relatively low bias voltage of 3 V. As seen from Figure 3, the current is relatively low, and moreover it increases nonlinearly with bias voltage. However, it can well be fitted by Poole-Frenkel (P-F) emission mechanism, and a linear relation of ln(I/V) versus V1/2 can be obtained, suggesting that the height of bulk trap barrier is higher than that of surface barrier at a low measurement bias. Therefore, the conduction mechanism in a single NW is predominated by the bulk trap-related P-F hopping rather than the contact barrier at electrode and NW interface at a relatively low operation bias. The P-F emission is related to electric-field-enhanced thermal emission from a trapped state into a continuum of electronic states, and the current through the NW can be given by:43

I = CV exp(−q ( − (qV /  0 r )1/2 ) / kT )

(1)

Where V is applied voltage, q is unit charge, T is absolute temperature, 0 is free space permittivity, r is dynamic dielectric constant, k is Boltzmann constant, and  is barrier height for electron emission from trap.


7


(b)

(a)

Current (nA)

2

1

1

3

0

-3

-2

-2

1

4

-1

-1

2

Ln(I/V)

2

-3

P-F Fitting -1

0

1

2

3

1.0

1.5

V1/2

Voltage (V)

Figure 3. At 3 V bias voltage, I-V characteristics of a single ZnO NW-based two-terminal device. (a) IV cycle curve; (b) the plot of fitted curve by P-F mechanism for the forward bias part in (a).

(b)

5

0

n 0

-5

4 1 Saturated

5 8

region

2

6 7

-10

-10

3

Saturated region B regreakdo ion w

Current (uA)

Current (uA)

10

Bre regakdow ion n

10

(a)

One cycle

10 cycles -15

-5

0

Voltage (V)

-5

5

gion

(c)

5

(d)

2

Ln(I/V )

-2

Line ar re

0.0

0

Voltage (V) -1

0.5

Current (nA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

2

-3

1

-4

F-N Fitting -0.5

-5 -1

0

1

0

Voltage (V)

1

2

1/V

Figure 4. At 5 V bias voltage, I-V characteristics of a single ZnO NW-based two-terminal device. (a) IV curves with 10 consecutive cycles; (b) one cyclic I-V curve; (c) an enlarged view of I-V curve near zero point in (b); (d) the plot of fitted curve by FN tunneling mechanism for the rise step 1 and 2 of forward-biased voltage in (b).

Figure 4 depicts I-V characteristics of a single ZnO NW-based device measured at 5 V bias with a voltage sweeping frequency of 0.1 Hz. It can be seen that currents both show abruptly a steep increase at approximately 4 V at the rise stage of forward and reverse bias voltages accompanied symmetrically with relatively weak counterclockwise hysteresis loops, and furthermore the I-V curves maintain


8

Page 9 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


invariably the same shape after successively sweeping 10 cycles. For the curves with unexpected increase can well be fitted by Fowler-Nordheim (F-N) tunneling mechanism, and a linear decrease relation of ln(|I/V2|) versus |1/V| can be found when the applied bias exceeds the transition point, as illustrated in Figure 4d. It indicates that the energy band bends gradually to become triangular with increasing bias, and the height of traps decreases.11 When the applied electric field surpasses the height of traps, quantities of electrons can be injected into traps from negative electrode, which is analogous with field emission and F-N tunneling occurs.44,45 The curves can repeat well after multiple cycles, revealing that only quite a few shallow traps are filled at 5 V bias, and moreover the filling effect cannot be kept permanently. The current through the NW can be given by:



3





3



4 * q 2E 2  − 2m (q ox )2   C 2(q ox )2  2 J = exp = C E exp 1     3hqE E 16 2h ox    







(2)



Where E is electric, q is unit charge, ϕox is energy barrier that electrons must overcome, m* is electron mass in NWs, h is reduced Planck constant. As seen from Figure 4c, in addition, I-V curve presents a linear region near zero point. For the I-V curve at the rise stage of forward and reverse bias voltages, therefore, it can both be divided into linear, saturated, and breakdown three distinguishable regions. In contrast, the saturation platform region is very obvious at 5 V bias, and furthermore it begins to appear at a very low voltage. In order to further explore the influence of external bias on the electronic transport characteristics of an individual NW-based device, we continued to increase the measurement bias voltage from 6 to 10 V and the results were illustrated in Figure 5. As can be seen that the linear region becomes more evident with increasing bias, and furthermore their resistance decreases. Meanwhile, the transition voltage between saturated and breakdown regions increases with increase in applied bias. In addition, the platform height of saturated region increases as well with increase in cycle times and operation bias voltage. For a single HRS ZnO NW-based device, it is apparent that its electronic transport is intensively ACS Paragon Plus Environment

9


dependent of operation bias. At a relatively low operation bias, its conductance originates from bulk trap-related P-F thermal emission. At a moderate operation bias, it comes from shallow trap fillingrelated F-N tunneling. At a larger operation bias, deeper level traps can be filled, resulting in a gradual disappearance of bulk trap-related P-F thermal emission, and therefore surface barrier is dominant conductance.

0

6

24 .8 gi K on

re ar ne

-100

-100 -6

0

Li

-50

1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th

(b)

100

Current (uA)

0

200 Saturated region

Breakdown reg

ion

50

1st 2nd 3rd 4th 5th 6th 7th 8th

Li 39 ne .5 ar re K gio  n

(a)

Breakdown region

Current (uA)

100

Saturated region

-200 -7

0

Voltage (V)



io n

K

eg

.8 20

8th 9th 10th 11th 12th 13th 14th

1st 2nd 3rd 4th 5th 6th 7th

Current (uA)

-100

(d) 250

rr

0

8th 9th 10th 11th 12th 13th 14th

Li ne a

Breakdown region

100

1st 2nd 3rd 4th 5th 6th 7th

Saturated region

(c)

Breakdown region

Current (uA)

200

7

Voltage (V)

Saturated region

200

-200

150 Increase of cycle times 4

8

5

200 0

-200

1st 2nd 3rd 4th 5th 6th 7th 8th 9th

7

8

400

 n K o .4 regi 7 1 ar e n Li

1st 2nd 3rd 4th 5th 6th 7th 8th 9th

n io eg

wn reg ion

(f)

(e) Current (uA)

400

6

Voltage (V)

Voltage (V)

do

0

Saturated region

rr ea in 300 L

Br eak

-8

Current (uA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

-400 Increase of cycle times

-10

0

10

5

Voltage (V)

10

Voltage (V)

Figure 5. At different bias voltages, continuously cyclic I-V characteristics of a single ZnO NW-based two-terminal device. (a) 6 V; (b) 7 V; (c) 8 V; (d) a magnified view for the wine-red dotted frame in (c); (e) 10 V; (f) a magnified view for sky-blue dotted frame in (e).

3.3 Memory function of electron transport modulated by relatively large bias For a single ZnO NW-based two-terminal device at HRS, its conductance increases with increase in operation bias. To further explore the properties of electron transport after being applied a relatively large bias, I-V characteristics were measured as well at a relatively low operation bias after being ACS Paragon Plus Environment

10

Page 11 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


modulated by relatively large uni-/bi-directional voltages, and the results were illustrated in Figure 6-8. Figure 6 shows I-V characteristics of the device modulated by applying different unidirectional forwardbiased voltages. From Figure 6a, c and e, it can be seen that the I-V curves can also show linear, saturated, and breakdown three regions under being applied relatively large unidirectional forward cyclic bias of of 6, 8, and 10 V, respectively. Similarly, the resistance of linear region decreases with increase in operation bias voltage. As seen from Figure 6b, d and f, more interestingly, I-V curves can show a rectification-like asymmetrical feature measured at a relatively low measurement bias of 1 V after being applied a relatively large unidirectional forward bias. At forward bias, the device is conductive, showing a low resistance state (LRS), and furthermore the I-V curves are virtually linear. On the contrary, the device is nonconductive HRS at reverse bias. Moreover, the asymmetrical I-V characteristics can be maintained for a long time, implying a memory effect. To further recognize the conductive feature at low operation bias after being applied a relatively large bidirectional bias, the device is applied subsequently a fixed bias of -10 V after being applied +10 V fixed bias, as depicted in Figure 7. For the almost nonconductive HRS device, it is only conductive in one-way at 4 V bias after being applied +10 V unidirectional bias, as shown in Figure 7c. In contrast, it changes into two-way conduction at 4 V bias after it is subsequently applied -10 V bias, as represented in Figure 7d, implying that it is both conductive in forward and reverse direction after being applied relatively large bidirectional bias. After being modulated by a large uni-/bi-directional bias, similarly, all the modulated LRS I-V characteristics can be remained at a relatively low operation bias and room temperature, demonstrating a memory property. To clarify the modulation behavior of a HRS ZnO NW under a relatively large bias, the device was measured as well by successively cyclic sweeping with 10 V bidirectional bias, as illustrated in Figure 8. the device was first measured at 5 V bias after they were annealed at about 200 oC, as illustrated in Figure 8b. It can be seen that the device is virtually nonconductive HRS at 5 V bias voltage. Under being applied 10 V bias, interestingly, the first I-V cycle curve presents symmetrically two relatively


11


large hysteresis loops at both of forward and reverse bias voltages, as shown in Figure 8c. Compared with the first cycle, however, it is completely different for the second cycle. A apparent linear curve with LRS appears in relatively low bias region, represented in Figure 8d. Meanwhile, the two symmetrical hysteresis loops appear in higher bias region, and furthermore they become much smaller. The subsequent I-V cycle curves almost keep the same shape as the second cycle curve, manifesting that the conductance has become LRS after being applied a relatively large bias once. (b)

10

Current (nA)

54. 5K



20

After being applying 6 V unidirectional bias

0 -1

0

1

Voltage (V)

(d)

30


20

10

Current (uA)

100

(c)

32. 4K

Current (uA)

200



1st cycle 2nd cycle 3rd cycle 4th cycle 5th cycle 6th cycle 7th cycle 8th cycle 9th cycle 10th cycle

0

0 0

8

-1

0

(f)

20

10

Current (nA)

30


K

200

40

(e)

1st cycle 2nd cycle 3rd cycle 4th cycle 5th cycle 6th cycle 7th cycle 8th cycle 9th cycle 10th cycle

29

400

1

Voltage (V)

Voltage (V)

Current (nA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

0

0 0

10

Voltage (V)

-1

0

1

Voltage (V)

Figure 6. Under different unidirectional forward bias voltages and after being modulated by unidirectional biases, I-V characteristics of a single ZnO NW-based two-terminal device. The left figures (a), (c), and (e) correspond to successively cyclic I-V curves under being applied a unidirectional forward sweeping bias of 6, 8, and 10 V, respectively. The right figures (b), (d) and (f) are their corresponding I-V curves at 1 V bias after being applied a unidirectional forward sweeping bias of 6, 8, and 10 V, respectively, showing asymmetrical rectification-like feature after being modulated unidirectionally.


12

Page 13 of 29

Unidirectional conduction

Bidirectional 50 conduction

Reverse pulse

0

0

(b) Initial I-V cyclic curve

Current (uA) Current (uA)

HRS

Voltage (V)

20

Forward pulse

0.01

(a) 40

0.00

-50

-0.01 0

50

-4

100

Time (s)

0

4

Voltage (V)

(d)

(c) 20

0 -4

Current (uA)

20

Current (uA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


After only applying +10 V unidirectional fixed bias

0

4

Applying +10 and -10 V fixed biases successively 0

-4

Voltage (V)

0

4

Voltage (V)

Figure 7. Modulated by consecutively applying +10 V and -10 V fixed bias, electrical characteristics of a single ZnO NW-based two-terminal device. (a) The current response curve with loading different bias voltages; (b) At 4 V bias, initial I-V curve corresponding to dotted purple frame in (a), revealing an HRS; (c) after only being applied +10 V unidirectional fixed bias, I-V curve at 4 V bias corresponding to pink dotted frame in (a), showing a rectification-like asymmetrical conduction; (d) after being successively applied +10 and -10 V bidirectional fixed bias, I-V curve at 4 V bias corresponding to green dotted frame in (a), indicative of a nearly symmetrical conduction with LRS.

From the above experiments, it can be drawn a conclusion that the electron transport of a single ZnO HRS NW-based device is strongly dependent on the magnitude and direction of loaded bias voltage, showing a tunable memory feature. After being modulated by a relatively large unidirectional bias, the devices are conductive in the same direction as large bias. After being modulated by a relatively large bidirectional bias, the devices are both conductive in forward and reverse direction. Moreover, the modulated LRS electron transport performance can well be maintained at a relatively low operation bias and room temperature, demonstrating a memory behavior.


13


(a)

40

6 successive cycles measured at 10 V bias

20

(b) 1.0

Initial I-V cyclic curve measured only at 5 V

Current (uA)

Current (uA)

60

0.5

0

Initial I-V at 5 V 1st cycle 2nd cycle 3rd cycle 4th cycle 5th cycle 6th cycle

-20 -40 -10

0

0.0

-0.5

10

-5

0

Voltage (V)

20

60

(c) First cyclic sweeping I-V curve

40 20

0 Ln (I/V2)

0

-20

-4 0

(d) Second cyclic sweeping I-V curve

0

-20

-2

-40 -10

5

Voltage (V)

Current (uA)

40

Current (uA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

-40 1/V

2

0

Voltage (V)

10

-10

0

10

Voltage (V)

Figure 8. Under 10 V bidirectional bias, successively cyclic I-V characteristics of a single ZnO NWbased two-terminal HRS device. (a) 6 successively cyclic I-V curves at 10 V bias voltage; (b) original IV cycle curve measured at 5 V bias voltage, revealing a HRS; (c) the first I-V cyclic curve under a bias voltage of 10 V, revealing a huge transition of resistance at the voltage higher than about ±6 V, and the inset corresponds to the plot of fitted curve by F-N tunneling for the rise stage of forward bias; (d) the second I-V cyclic curve under a bias voltage of 10 V, showing that LRS has been formed in a low voltage region.

3.4 Application of modulated electron transport in the realization of self-driven UV photodetection with symmetrical two electrodes For the modulated electrical properties, to confirm whether it can improve the performance of known devices and realize novel electronic and optoelectronic devices, the classical photoelectric conversionrelated detection was selected as a verification experiment. For a single NW-based device with the structure of two back-to-back connected diodes, regardless of whether it is applied a forward or reverse bias voltage, one of them is invariably polarized in forward and the second one is in reverse direction. Therefore, it is very arduous to conduct at a relatively low bias.13,27 Under illumination, its conduction


14

Page 15 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


should originate from the reverse tunneling or avalanche effect of reverse-biased surface barrier. Due to the presence of quantities of traps associated with Vo and Zni, pure ZnO 1D nanostructures can generally show a long-persistent response and decay phenomenon, and moreover the dark current is very large at a relatively large measurement bias.13 Therefore, it is extremely disadvantageous for photodetection performance. However, the above researches reveal the width of surface barrier, connected with negative electrode, can become very thin after being applied a relatively large bias due to the injection of electrons into traps, while that connected with positive electrode almost remains unchanged. The symmetrical two back-to-back diodes with similar barriers will change into an asymmetrical configuration with different barriers. For the thin surface barrier with filled traps, it is extremely advantageous to separate free electron-hole pairs excited by external action.31 As seen from Figure 6b, d, and f, the devices are conductive in the same direction as large modulated bias, and moreover the conductive current shows a linear relation with bias. If we can utilize the modulated surface barrier, the disadvantage could be solved. Figure 9 illustrates the photoresponse measurement results at 1 V fixed bias after being applied +10 V unidirectional fixed bias. After being applied +10 V fixed bias, as seen from Figure 9a, the device can both represent a photoresponse to the UV light close to ZnO bandgap energy at both of forward and reverse bias voltages of 1 V. Comparatively, the device can show a stronger photoresponse to shorter wavelength UV light at +1 V than that at -1 V. From Figure 9b and c, it can further be found that the device can show fast response and recovery time lower than 1s at +1 V forward bias, and conversely it shows a long-persistent photoconductivity at -1 V reverse bias. At +1 V forward bias, however, the dark current is relatively large, and the on/off ratio is only about 5, and hence the sensitivity is very underdeveloped although its response and recovery speed is relatively rapid. To improve the sensitivity of photodetector modulated by a large unidirectional bias, it is necessary to further decrease the measurement voltage. If we can utilize the rapid separation of photoexcited electron-hole pairs in modulated thin surface barrier connected with negative electrode at zero bias, the self-driven UV


15


photoconductor based photovoltaic effect would be realized for symmetrical two-electrode device with asymmetrically modulated surface barriers. 800

Current (nA)

(a)

+1 V -1 V

150

600

400

200

200

400

600

800

Wavelength (nm)

Current (nA)

800

+1 V off

off

(b) off

off

600

400

200

on

on

on

on

0

50

100

Time (s)

0

Current (nA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

on

on

(c)

on

on

-5

off -10

-1 V 0

off

off

off

50

Time (s)

Figure 9. After being modulated by a unidirectional fixed bias of +10 V, the photoresponse properties of a single ZnO NW-based two-terminal device measured at 1 V fixed bias. (a) At +1 (red line) and -1 V (blue line), respectively, the photoresponse vs wavelength; (b) and (c) at +1 and -1 V, respectively, the on/off photoresponse to 370 nm UV light with a power density of 170 W/cm2, revealing that the longpersistent photoconductivity disappears at +1 V and contrarily it still exists at -1 V.


16

Page 17 of 29

(a)

(b)

Voc

Current (nA)

0

Current (uA)

0

-5

-100

In dark UV light

-10

-0.5

0.0

In dark UV light

Isc 0.5

-200

0.00

0.05

Voltage (V)

(d) on

Voltage (mV)

on

100

0.10

Voltage (V)

(c) Current (nA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


50

off

50

0 0

20

40

Time (s)

60

off

0 0

20

40

60

Time (s)

Figure 10. Under illumination of UV light, the photoresponse of self-driven photodetector based on a single ZnO nanowire after being modulated by a unidirectional -10 V fixed bias. (a) At 0.5 V bias, I-V characteristics in dark and UV illumination, revealing a more manifest photoconduction effect in the same direction as applied bias; (b) I-V characteristics near zero point in dark and UV illumination, namely, the magnified view of green dotted frame in (a), implying the presence of notable photovoltaic effect; (c) and (d) photoexcited current and voltage as a function of time with light chopping with an interval of 5 s at 0 V bias, showing fast response and recovery speed.

After being applied a unidirectional -10 V fixed bias, the self-driven UV photoresponse of a single ZnO nanowire-based two-terminal detector was measured, as shown in Figure 10. As seen from Figure 10a, the dark current is relatively large in the same direction as applied large modulated bias although the photocurrent is very large, indicating that the ratio of on to off current is relatively low, and accordingly the sensitivity is exceedingly poor. As seen from the Figure 10b of a magnified view near zero point, the device can show Voc of 90 mV and Isc of 150 nA. Moreover, the potential is positive at the end connected with the applied large negative bias. For the rectification-like asymmetrical curve, therefore, it shows an opposite photovoltaic response law to the classic diode rectification curve. Figure


17


Page 18 of 29

10c and d further show the photogenerated current and voltage as a function of time with light chopping at zero bias. It can be seen the photoexcited current of about 130 nA and voltage of about 80 mV can both show fast rise and decay with time lower than 0.2 s, and moreover their stability and repeatability are both excellent as well. The above results not only verify firmly that the optical and electrical properties of ZnO nanostructures are strongly dependent on trap filling, but also demonstrate that some high-performance devices can be realized by modulating electron transport using a bias, such as self-driven ZnO UV photodetector with fast response and decay speed by a relatively large unidirectional fixed bias modulation. 3.5 Mechanism of bias-modulated electron transport with memory behavior Ag work function of 4.26 eV is lower than ZnO electron affinity of 4.35 eV. In general, it would be expected to form Ohmic contact when Ag is in contact with n-type ZnO NWs. For ZnO NWs with very large surface-to-volume ratio and typical n-type properties, however, dangling bonds due to a breaking of lattice periodicity on the surface can trigger the presence of quantities of acceptor-type surface states, triggering a band bending upward and an existence of carrier-depletion layer in the vicinity of surfaces, where majority carriers (electrons) are almost completely depleted, and correspondingly surface barriers are formed. Due to a great difference in electronegativity between O and Zn, in addition, quantities of defects such as Vo and Zni exist in ZnO lattice, bringing out in the formation of traps with different levels. For nanostructure ZnO, high densities of surface states will lead to Fermi level pinning, and therefore its surface barrier is independent of Ag work function and ZnO electron affinity. Due to the presence of high densities of surface states, majority carriers (electrons) are depleted close to surface. To maintain electrical neutrality, moreover, quantities of positive charges exist at the boundary of depletion region near the interior. In addition, enormous number of traps are underfilled by electrons in the interior of NW. The electrons trapped by deep levels cannot move freely. Especially for the traps located in surface space charge region, they are emptied completely. A wide depletion region is formed


18

Page 19 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


as well, and therefore it is generally arduous for ZnO based-nanodevices to conduct at a relatively low operation bias. If the height of band bending (Vd) is lower than that of traps (ϕ), its conduction is mainly governed by the hopping of trapped electrons, as illustrated diagrammatically in Figure 11a. If Vd is larger than ϕ, the device is almost nonconductive. (a)

(b)

(c)

(d)

Figure 11. Schematic sketches of bias dependent electron transport and photoresponse of a single ZnO NW-based two-terminal device.  represents trap barrier height. Ec, EV, Ei and Ef correspond to conduction band, valence band, intermediate, and Fermi levels, respectively. Vd corresponds to band bending height, V1 and V2 represent barrier reduction height under UV illumination. (a) Before being applied a relatively large bias, internal electron traps are almost completely emptied, the surface Femi level are pinned, the surface barrier is wide, and the conduction is dominated by the hopping of trapped electrons. (b) Under a relatively large unidirectional bias, electrons are injected into traps from negative electrode, resulting in the filling of traps and the decrease of barrier height and width near negative electrode. (c) After being successively applied relatively large bidirectional bias, most traps are full up, the width of barriers at two ends becomes very thin, and the conduction is dominated by the thermion emission of surface barrier. (d) Under UV illumination, the light can completely penetrate depletion layer to reach the interior of NW for thin surface barrier with overfilled traps, while it can completely be


19


Page 20 of 29

absorbed in a part of depletion layer near surface for the thick surface barrier with underfilled traps.

With increasing bias voltage, subsequently, the energy band tilts gradually in depletion region since the externally loaded bias almost completely drops on the surface barrier with large resistance, and meanwhile the height of trap barrier decreases.25,45 When the height of tilting exceeds that of trap barriers near depletion region, electrons can be injected into traps from negative electrode, resulting in a filling of traps near the negative electrode. At the bias voltage higher than a certain value, the abrupt jumping of device current provides a substantial support for this view of trap filling. If the applied bias is not high enough such as 5 V bias, only some shallow level traps, which can be excited thermally at room temperature, are filled, and therefore I-V curves can remain initial shape very well after multiple cycles. If the applied bias voltage is higher than 5 V, deeper level traps will be filled, and moreover the filled electrons cannot spontaneously be excited thermally at room temperature. Therefore, I-V curves cannot restore to their initial states after being applied a relatively large bias once, showing a memory behavior. It also indicates that the height and width of the surface barrier connected with negative decrease drastically after bias induced-electron filling in traps. Especially for its width, it becomes very thin, the conduction mechanism changes from the internal traps-controlled electron hopping into thin surface barrier-controlled thermion emission, and correspondingly I-V curves can appear linear, saturated, and breakdown three evident regions.16,46 With the extend of injection time and the increase of applied bias, the resistance of linear region decreases, and the saturation platform upshifts, indicative of a further decrease of barrier height resulting from the filling of deeper level traps. If the devices are successively applied large forward and reverse bidirectional bias voltages, the electron traps are all filled up at two ends, and hence the device is both conductive at relatively low forward- and reversebiased voltages, as shown schematically in Figure 11c. If only a relatively large unidirectional forwardor reverse-biased voltage is applied, the device only shows a rectification-like one-way electrical conduction with the same direction as the applied large modulated bias. The corresponding schematic ACS Paragon Plus Environment

20

Page 21 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


diagram is illustrated in Figure 11b. In breakdown region, the conductance arises mainly from the large bias-induced avalanche effect of reverse-biased surface barrier.51 Therefore, the current can rapidly increase, and meanwhile shows a small hysteresis loop at high bias breakdown region. Since the accurate control of electrical conduction of a single ZnO NW can be realized by the appropriate filling of traps near two end electrodes at a relatively large bias, and moreover the low operation bias and temperature cannot induce the filling and emptying of deep level traps. At relatively low operation bias and room temperature, therefore, the modulated electrical conduction can show a memory feature after being applied relatively large uni-/bi-directional bias. After only being applied a relatively large unidirectional bias, the width of surface barrier at the end connected with negative electrode becomes very thin due to the overfilling of traps, while the width at another end is still thick. For a short wavelength UV light, in addition, its absorption coefficient is relatively large, that is, its penetration depth is very shallow. Moreover, the shorter the wavelength is, the shallower the depth is.57 For the thin surface barrier with overfilled traps, the incident light can penetrate the thin depletion layer to reach the interior of NW under UV illumination. The absorption region also includes a part of NW interior besides the entire depletion region. Due to the overfilling of traps, the disturbance of traps to electronic transport can be ignored. Therefore, it is very easy for photogenerated electron-hole pairs to be separated by surface built-in electric filed.16 For the wide surface barrier with emptied traps, on the contrary, the incident light can completely be absorbed by a part of depletion region near surface under UV illumination. Due to the exhaustion of traps in depletion layer, the disturbance of traps to the electronic transport is very strong. For photoexcited electron-hole pairs, therefore, their separated efficiency is relatively low, and further free carriers can be captured by emptied traps. Under illuminating entire device with UV light at zero bias, the photoexcited current and voltage of self-driven detector should originate from the photovoltaic effect of thin surface barrier with overfilled traps. Although the I-V curve of two back-to-back diodes with a unidirectional injection is the same shape as a classic single diode, therefore, their direction of photoexcited current and voltage is just


21


Page 22 of 29

opposite. For the two-terminal devices with a unidirectional electron injection, moreover, they can show fast response and recovery speed, resulting in an immense improvement of long-persistent photoconductivity. Due to the presence of emptied traps, in addition, it is more inclined for the shorter wavelength UV light to recombine. After being injected by a large unidirectional forward bias, accordingly, the device can show a much wider photoresponse to shorter wavelength UV light at a low forward bias in comparison with a low reverse bias. The corresponding schematic diagram is illustrated in Figure 11d.

4. CONCLUSION For a single ZnO NW-based two-terminal device, in summary, two back-to-back connected surface barrier diodes with series resistance can be formed due to the presence of abundant surface states. In surface space charge region, the filling and emptying of traps can modulate the surface barrier. At a relatively low operation bias, generally, the devices have a relatively inferior conductivity due to the empty of traps, and the corresponding conductivity mainly arises from the electron hopping of traps. At a relatively large bias voltage, however, electrons can be injected into traps from negative electrode. The filling of traps gives rise to a dramatic decrease of surface barrier at the end connected with the negative electrode. The electrical conduction changes into a thin surface barrier-controlled thermion emission, and correspondingly the I-V curves can show linear, saturation, breakdown three distinguishable regions. After only being applied a large unidirectional bias, additionally, a rectification-like feature can be formed at a low operation bias due to the asymmetrical decrease of two back-to-back diode barriers. The filled and emptied traps can be maintained for a relatively long time at low operation bias and room temperature, demonstrating a tunable memory feature. For the modulated thin surface barrier, it is extremely beneficial to penetrate UV light and separate photoexcited electron-hole pairs. At zero bias, therefore, the asymmetrically modulated two-terminal devices with two same electrodes can show outstanding self-driven UV photodetection performance with fast response and recovery speed. For single ZnO NW, the recognition and accurate control of electronic transport not only clarify the origin


22

Page 23 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of their conductance, but also put forward a strategy to the design of novel electron and photoelectron device based on ZnO nanostructures. AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Fax: +86-791-8396-9329.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT

This work was financially supported by the National Natural Science Foundation of China (51571107, 51462023),

and

the

Natural

Science

Foundation

of Jiangxi

Province

(20152ACB20010,

20181BAB216008). REFERENCES (1) Dasgupta, N. P.; Sun, J. W.; Liu, C.; Brittman, S.; Andrews, S. C.; Hanwei Gao, H. W.; Yan, R. X.; Yang, P. D. 25th Anniversary Article: Semiconductor Nanowires–synthesis, Characterization, and Applications. Adv. Mater. 2014, 26, 2137-2184. (2) Vishnukumar, P.; Vivekanandhan, S.; Misra, M.; Mohanty, A. K. Recent Advances and Emerging Opportunities in Phytochemical Synthesis of ZnO Nanostructures. Mater. Sci. Semi. Proc. 2018, 80, 143-161. (3) Shokouh, S. H. H.; Pezeshki, A.; Raza, A.; Raza, S.; Lee, H. S.; Min, S.; Im, S. High-Gain Subnanowatt Power Consumption Hybrid Complementary Logic Inverter with WSe2 Nanosheet and ZnO Nanowire Transistors on Glass. Adv. Mater. 2015, 27, 150-156. (4) Tao, L.; Cinquanta, E.; Chiappe, D.; Grazianetti, C.; Fanciulli, M.; Akinwande, D. Silicene FieldEffect Transistors Operating at Room Temperature. Nat. Nanotechnol. 2015, 10, 227. ACS Paragon Plus Environment

23


Page 24 of 29

(5) Wang, Z. L.; Song, J. H. Piezoelectric Nanogenerators Based on Zinc Oxide Nanowire Arrays. Science 2006, 312, 242-246. (6) Anoop, G.; Panwar, V.; Kim, T. Y.; Jo, J. Y. Resistive Switching in ZnO Nanorods/Graphene Oxide Hybrid Multilayer Structures. Adv. Electron. Mater. 2017, 3, 1600418. (7) Cheng, B. C.; Zhao, J.; Xiao, L.; Cai, Q. S.; Guo, R.; Xiao, Y. H.; Lei, S. J. PMMA InterlayerModulated Memory Effects by Space Charge Polarization in Resistive Switching Based on CuSCNnanopyramids/ZnO-nanorods p-n heterojunction. Sci. Rep. 2015, 5, 17859. (8) Horsfall, L. A.; Pugh, D. C.; Blackman, C. S.; Parkin, I. P. An Array of WO3 and CTO Heterojunction Semiconducting Metal Oxide Gas Sensors Used as a Tool for Explosive Detection. J. Mater. Chem. A 2017, 5, 2172-2179. (9) Lupan, O.; Postica, V.; Grottrup, J.; Mishra, A. K.; de Leeuw, N. H.; Carreira, J. F. C.; Cretu, V. Hybridization of Zinc Oxide Tetrapods for Selective Gas Sensing Applications. ACS Appl. Mater. Interfaces 2017, 9, 4084-4099. (10) Poloju, M.; Jayababu, N.; Manikandan, E.; Reddy, M. R. Enhancement of the Isopropanol Gas Sensing Performance of SnO2/ZnO Core/shell Nanocomposites. J. Mater. Chem. C 2017, 5, 2662-2668. (11) Zhao, J.; Cheng, B. C.; Xiao, Y. H.; Guo, R.; Lei, S. J. Ultrahigh Performance Negative ThermalResistance Switching Based on Individual ZnO:K,Cl Micro/nanowires for Multibit Nonvolatile Resistance Random Access Memory Dual-written/erased Repeatedly by Temperature or Bias. J. Mater. Chem. C 2015, 3, 12220-12229. (12) Zhao, J.; Cheng, B. C.; Xiao, Y. H.; Guo, R.; Lei, S. J. Gate-Free Controlled Multibit Memories Based on Individual ZnO:In Micro/Nanowire Back-to-Back Diodes. Adv. Electron. Mater. 2016, 2, 1500395. (13) Cheng, B. C.; Xu, J.; Ouyang, Z. Y.; Xie, C. C.; Su, X. H.; Xiao, Y. H.; Lei, S. J. Individual ZnO Nanowires for Photodetectors with Wide Response Range From Solar-blind Ultraviolet to Near-infrared Modulated by Bias Voltage and Illumination Intensity. Opt. Express 2013, 21, 29719.


24

Page 25 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(14) Mikhelashvili, V.; Padmanabhan, R.; Eisenstein, G. Simplified Parameter Extraction Method for Single and Back-to-back Schottky Diodes Fabricated on Silicon-on-insulator Substrates. J. Appl. Phys. 2017, 122, 034503. (15) Luo, B.; Zhao, J.; Cheng, B. C.; Chen, G. W.; Ouyang, T. F.; Pan, Y. X.; Li, B. H.; Xiao, Y. H.; Lei, S. J. A Surface State-controlled, High-performance, Self-powered Photovoltaic Detector Based on an Individual SnS Nanorod with a Symmetrical Electrode Structure. J. Mater. Chem. C 2018, 10, 1039. (16) Jin, Z. W.; Zhou, Q.; Chen, Y. H.; Mao, P.; Li, H.; Liu, H. B.; Wang, J. Z.; Li, Y. L. Graphdiyne: ZnO Nanocomposites for High-Performance UV Photodetectors. Adv. Mater. 2016, 28, 3697-3702. (17) Shen, Y. W.; Yan, X. Q.; Si, H. N.; Lin, P.; Liu, Y. C.; Sun, Y. H.; Zhang, Y. Improved Photoresponse Performance of Self-powered ZnO/spiro-MeOTAD Heterojunction Ultraviolet Photodetector by Piezo-phototronic Effect. ACS Appl. Mater. Interface 2016, 8, 6137-6143. (18) Cheng, B. C.; Xu, J.; Ouyang, Z. Y.; Su, X. H.; Xiao, Y. H.; Lei, S. J. Individual Ohmic Contacted ZnO/Zn2SnO4 Radial Heterostructured Nanowires as Photodetectors with a Broad-spectralResponse: Injection of Electrons into/from Interface States. J. Mater. Chem. C 2014, 2, 1808-1814. (19) Zhao, B.; Wang, F.; Chen, H. Y.; Zheng, L. X.; Su, L. X.; Zhao, D. X.; Fang, X. S. An Ultrahigh Responsivity (9.7 mA W-1) Self-Powered Solar-Blind Photodetector Based on Individual ZnO-Ga2O3 Heterostructures. Adv. Funct. Mater. 2017, 27, 1700264. (20) Yu, P. P.; Hu, K.; Chen, H. Y.; Zheng, L. X.; Fang, X. S. Novel p-p Heterojunctions SelfPowered Broadband Photodetectors with Ultrafast Speed and High Responsivity. Adv. Funct. Mater. 2017, 27, 1703166. (21) Kang, J. H.; Ebaid, M.; Jeong, D. K.; Lee, J. K.; Ryu, S. W. Efficient Energy Harvesting of a GaN p-n Junction Piezoelectric Generator Through Suppressed Internal Field Screening. J. Mater. Chem. C 2016, 4, 3337-3341. (22) Luo, L.; Zhang, Y.; Li, L. J. Piezotronic PIN Diode for Microwave and Piezophototronic Devices. Semicond. Sci. Tech. 2017, 32, 044002.


25


Page 26 of 29

(23) Xue, F.; Chen, L. B.; Chen, J.; Liu, J. B.; Wang, L. F.; Chen, M. X.; Pang, Y. K.; Yang, X. N.; Gao, G. Y.; Zhai, J. Y.; Wang, Z. L. p-Type MoS2 and n-Type ZnO Diode and Its Performance Enhancement by the Piezophototronic Effect. Adv. Mater. 2016, 28, 3391-3398. (24) Zheng, J. P.; Cheng, B. C.; Wu, F. Z.; Su, X. H.; Xiao, Y. H.; Guo, R.; Lei, S. J. Modulation of Surface Trap Induced Resistive Switching by Electrode Annealing in Individual PbS Micro/NanowireBased Devices for Resistance Random Access Memory. ACS Appl. Mater. Interfaces 2014, 6, 2081220818. (25) Park, J. Y.; Cho, J.; Jun, S. C. Review of Contact-resistance Analysis in Nano-material. J. Mech. Sci. Technol. 2018, 32, 539-547. (26) Cheng, B. C.; Wu, G. Z.; Ouyang, Z. Y.; Su, X. H.; Xiao, Y. H.; Lei, S. J. Effects of Interface States on Photoexcited Carriers in ZnO/Zn2SnO4 Type-II Radial Heterostructure Nanowires. ACS Appl. Mater. Interfaces 2014, 6, 4057-4062. (27) Tong, T.; Wang, S. J.; Zhao, J.; Cheng, B. C.; Xiao, Y. H.; Lei, S. J. Erasable Memory Properties of Spectral Selectivity Modulated by Temperature and Bias in an Individual CdS Nanobelt-based Photodetector. Nanoscale Horiz. 2019, 4, 138-147. (28) Gazquez, G. C.; Lei, S. D.; George, A.; Gullapalli, H.; Boukamp, B. A.; Ajayan, P. M.; Ten Elshof, J. E. Low-Cost, Large-Area, Facile, and Rapid Fabrication of Aligned ZnO Nanowire Device Arrays. ACS Appl. Mater. Interface 2016, 8,13466-13471. (29) Jeong, H.; Oh, H. M.; Bang, S.; Jeong, H. J.; An, S. J.; Han, G.; Lee, Y. H. Metal-insulatorsemiconductor Diode Consisting of Two-dimensional Nanomaterials. Nano Lett. 2016, 16, 1858-1862. (30) Jain, S. K.; Aggarwal, N.; Krishna, S.; Kumar, R.; Husale, S.; Gupta, V.; Gupta, G. GaN-UV Photodetector Integrated with Asymmetric Metal Semiconductor Metal Structure for Enhanced Responsivity. J. Mater. Sci. 2018, 29, 8958-8963. (31) Zhao, J.; Cao, M. H.; Cheng, B. C.; Wu, G. H.; Guo, H. J.; Ai, Y. L.; Su, X. H.; Xiao, Y. H.; Lei, S. J. Carbon-encapsulated CdSe Quantum Dot Inorganic Hybrid Nanobelts for High Performance


26

Page 27 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Photoelectronic Devices Based on the Efficient Separation and Transfer of Photoinduced Holes. J. Mater. Chem. C, 2015, 3, 2471-2478. (32) Liu, Y. Y.; Stradins, P.; Wei, S. H. Van der Waals Metal-semiconductor Junction: Weak Fermi Level Pinning Enables Effective Tuning of Schottky Barrier. Sci. Adv. 2016, 2, 1600069. (33) Selman, A. M.; Hassan, Z. Fabrication and Characterization of Metal-semiconductor-metal Ultraviolet Photodetector Based on Rutile TiO2 Nanorod. Mater. Res. Bull. 2016, 73, 29-37. (34) Das, M.; Datta, J.; Sil, S.; Dey, A.; Jana, R.; Halder, S.; Ray, P. P. Equivalent Circuit Analysis of Al/rGO-TiO2 Metal-semiconductor Interface Via Impedance Spectroscopy: Graphene Induced Improvement in Carrier Mobility and Lifetime. Mat. Sci. Semicond. Proc. 2018, 82, 104-111. (35) Yu, H.; Schaekers, M.; Barla, K.; Horiguchi, N.; Collaert, N.; Thean, A. V. Y.; De Meyer, K. Contact Resistivities of Metal-insulator-semiconductor Contacts and Metal-semiconductor Contacts. Appl. Phys. Lett. 2016, 108, 171602. (36) Shi, H. P.; Cheng, B. C.; Cai, Q. S.; Su, X. H; Xiao, Y. H.; Lei, S. J. Surface State Controlled Ultrahigh Selectivity and Sensitivity for UV Photodetectors Based on Individual SnO2 Nanowires. J. Mater. Chem. C. 2016, 4, 8399-8406. (37) Shi, H. P.; Zheng, J. P.; Cheng, B. C.; Zhao, J.; Su, X. H.; Xiao, Y. H.; Lei, S. J. Rewritable Nonvolatile Stress Information Memory by Bulk Trap-induced Giant Piezoresistance Effect in Individual PbS Micro/nanowires. J. Mater. Chem. C 2017, 5, 229-237. (38) Chiquito, A. J.; Amorim, C. A.; Berengue, O. M.; Araujo, L. S.; Bernardo, E. P.; Leite, E. R. Back-to-back Schottky Diodes: The Generalization of the Diode Theory in Analysis and Extraction of Electrical Parameters of Nanodevices. J. Phys. 2012, 24, 225303. (39) Ning, Y.; Zhang, Z. M.; Teng, F.; Fang, X. S. Novel Transparent and Self-Powered UV Photodetector Based on Crossed ZnO Nanofiber Array Homojunction. Small, 2018, 14, 1703754. (40) Fu, Q. M.; He, D. C.; Yao, Z. C.; Peng, J. L.; Zhao, H. Y.; Tao, H.; Chen, Z.; Tu, Y. F.; Tian, Y.; Zhou, D.; Zheng, G.; Ma, Z. B. Self-powered Ultraviolet Photodetector Based on ZnO Nanorod Arrays


27


Page 28 of 29

Decorated with Sea Anemone-like CuO Nanostructures. Mater. Lett. 2018, 222, 74-77. (41) Fan, M. M., Liu, K. W.; Chen, X.; Zhang, Z. Z.; Li, B. H.; Shen, D. Z. A Self-powered SolarBlind Ultraviolet Photodetector Based on a Ag/ZnMgO/ZnO Structure with Fast Response Speed. RSC Adv. 2017, 7, 13092-13096. (42) Duan, L.; He, F. N.; Tian, Y.; Sun, B.; Fan, J. B.; Yu, X. C.; Ni, L.; Zhang, Y.; Chen, Y. N.; Zhang, W. X. Fabrication of Self-powered Fast-response Ultraviolet Photodetectors Based on Graphene/ZnO:Al Nanorod-array-film Structure with Stable Schottky Barrier. ACS Appl. Mater. Interfaces 2017, 9, 8161-8168. (43) Vietmeyer, F.; Tchelidze, T.; Tsou, V.; Janko, B.; Kuno, M. Electric Field-induced Emission Enhancement and Modulation in Individual CdSe Nanowires. ACS Nano 2012, 6, 9133-9140. (44) Gomer, R. Field Emission and Field Ionization; American Institute of Physics: New York. 1992, pp 1-126. (45) Sze, S. M.; Ng, K. K. Physics of Semiconductor Devices, 3rd ed.; John Wiley & Sons: Hoboken, New Jersey, 2007, pp 79-196. (46) Chiquito, A. J.; Amorim, C. A.; Berengue, O. M.; Araujo, L. S.; Bernardo, E. P.; Leite, E. R. Back-to-back Schottky Diodes: The Generalization of the Diode Theory in Analysis and Extraction of Electrical Parameters of Nanodevices. J. Phys. 2012, 24, 225303.


28

Page 29 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents graphic


29

Bias-controlled Tunable Electronic Transport with Memory

Recommend Documents