Subscriber access provided by Kaohsiung Medical University
Letter
Fully Aerosol Jet Printed, High-Performance Nano-porous ZnO Ultraviolet Photodetectors Anubha Gupta, Shivaram Arunachalam, Sylvain G. Cloutier, and Ricardo Izquierdo ACS Photonics, Just Accepted Manuscript • DOI: 10.1021/acsphotonics.8b00829 • Publication Date (Web): 12 Sep 2018 Downloaded from http://pubs.acs.org on September 13, 2018
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 16 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 Photonics
Fully Aerosol Jet Printed, High-Performance Nano-porous ZnO Ultraviolet Photodetectors Anubha A Gupta,∗ Shivaram Arunachalam, Sylvain G. Cloutier, and Ricardo Izquierdo∗ ´ Electrical Engineering Department, Ecole de Technologie Sup´erieure, Qu´ebec, Canada E-mail:
[email protected];
[email protected] Abstract
Printed UV photodetectors are of significant interest for missile detection, chemical and biological threat interception, environmental pollution monitoring, optical communications and others. 1–3 ZnO (among other metal oxides such as, Ga2 O3 , SnO2 , In2 O3 , ZnGeO and InGe2 O7 ) shows great potential as a UV photodetector 4 due to its direct-wide band gap of 3.37 eV at room temperature and a corresponding band absorbance of 250-400 nm. 5
We report fully aerosol-jet printed ultraviolet photodetectors based on zinc oxide nano-crystal networks having a porous morphology. The developed photodetectors exhibit selective optical photo response in the ultraviolet spectrum at wavelengths ranging from 250 nm upto 400 nm. These devices show a high ON/OFF ratio of ∼ 106 and short response times. The characteristic time constants for the rise and fall edges were measured to be ∼0.4 s and ∼1.3 s, respectively. In this work, we propose a direct approach for aerosol printing pre-synthesized nanocrystals to overcome the limitations of high post-annealing temperatures and tedious fabrication methods to obtain high aspect ratio nanostructures. The high performance characteristics of these devices are attributed to Schottky barrier modification under the influence of oxygen, which is enhanced by the porosity of the semiconductor material. The random orientation of the crystals aids the formation of air traps in the network, thereby enhancing the surface area-tovolume ratio. Most importantly, the complete processing of these devices is performed below 150 o C, which makes this technology compatible with the processing on a wide range of mechanically flexible, recyclable or inexpensive substrates such as paper and plastics.
Several morphologies of ZnO nano-materials have been proposed to achieve enhanced photosensitivity due to their large surface area-to-volume ratio, 6,7 since the background oxygen concentration has a prominent effect on the behavior of ZnO semiconductors. 8 In particular, ZnO nanoparticle detectors were shown to have reduced dark currents as well as increased responsivity. 9 Single nano-wire detectors have been reported to have similar properties, 7,8,10–12 and the photocurrent can be scaled up by increasing the number of nanowires. 13 Alternative structures such as microtubes, 14 nanorods, 15 nanospheres 16 and tetrapods 17 have also been proposed. However, these structures either require high processing temperatures, 18 tedious fabrication techniques such as electro spinning, 10 vapor deposition techniques which require vacuum processing 8 or precise control of the synthesis and morphology of the nano-materials, 19,20 which impede their use in scaled-up fabrication.
Keywords
On the other hand, efforts have been made in solution processing for the fabrication of thin film ZnO phoZinc oxide, UV photodetector, aerosol jet printed, fully 9,21–23 todetectors. However, processing techniques like printed, flexible electronics, printable electronics screen printing, spin coating and dip coating waste a large fraction of the material inks, leading to higher costs in the fabrication process. The use of direct-write Letter additive techniques, such as inkjet printing can signifPrinted electronics is a fast developing field, promisicantly reduce material waste by 94-97% 24 . However, ing low-cost fabrication on large area substrates, with the narrow allowable viscosity range of the inks in inkjet processing limits the amount of usable materials. applications in photovoltaics, RFID tags, digital logic, sensors, etc. In particular, due to the low temperature Moreover, the risk of nozzle clogging reduces process processing, electronics can be printed on unconventional throughput. Recently, direct-writing techniques have substrates such as conformal plastics and flexible PET. attracted interest for the fabrication of ZnO thin films, ACS Paragon Plus Environment 1
ACS Photonics
although a processing temperature of 200o C-400o C is still required to anneal the precursor ink. 25 Aerosol gas stream In this work, we demonstrate a first fully-printed, low Sheath gas temperature processed nano-porous ZnO photodetector (a) on rigid as well as flexible substrates. The entire device is fabricated using Aerosol Jet Printing (AJP), a direct-write fabrication method capable of printing inks with a wide viscosity difference ranging from 1-1000 cP and with improved resilience to nozzle clogging. Recently, printing of ZnO precursor ink formulations have 1-5 mm been explored with AJP. 25,26 However, it has not been Substrate so far employed to print directly from a pre-synthesized nano-crystal ink. In this study, the aerosol jet technique is used to deposit the ZnO nano-crystals from a dispersion ink. As the solvent evaporates, a well(b) connected porous film is formed having high surface roughness. As a result of this porous morphology, the ZnO devices exhibit a high surface-to-volume ratio, and show Ag high ON/OFF ratio and fast speed, comparable in perAg formance to previously published nanostructured ZnO photodetectors. 9,10,13 Another important aspect of this I work is that the complete fabrication process is carried o out below 150 C, making it compatible for processing V - + on large area flexible substrates. Experimental Section Aerosol Setup and Synthesis Procedure Figure (c) 1 (a) describes the aerosol jet printing process for the fabrication of ZnO photodetectors. In this process, a stream of inert gas is blown at high pressure over the surface of the functional ink to be printed. This generates a mist of the material which, after compression by a virtual impactor, is directed towards the deposition head. At the printing head, a second flow of inert gas called sheath gas surrounds the aerosol gas stream and guides it to the substrate, as shown in Figure 1 (a). This helps focusing the printing stream, thereby reducing sprayover. It also prevents the material to come in contact with the nozzle to avoid clogging. The temper100 nm ature of the substrate and the feed tube to the printing head can be controlled to improve ink drying. Printed 1 μm structures can be thermally-annealed or laser-sintered. Figure 1 (b) shows the schematic device structure for Figure 1: (a) Schematic illustration of aerosol jet deposition the printed ZnO photodetectors. A pair of 1 mm2 Ag head demonstrating the aerodynamic focusing of aerosol gas contact pads was printed to facilitate electrical probing stream using sheath gas; (b) Schematic diagram depicting the device architecture of the printed photodetector on PET of the device. Ag electrodes with 30 µm width and substrate; (c) Microscope image of the device and SEM ima gap of 80 µm were printed from the contact pads. A ages of the ZnO films. ZnO porous film was then deposited atop of the metallic contacts to form a metal-semiconductor-metal (MSM) photodetector with Ag-ZnO and ZnO-Ag interfaces. min each followed by drying with nitrogen gas. The Materials and Device Fabrication. All layprinting of both the metal and semiconductor material ers of the device were printed using a commercially was carried out using pneumatic atomizer. available Aerosol jet printing system (AJ300 from OpThe Ag metal electrodes are first printed on the tomec, Inc.) under ambient environment on glass subcleaned substrates using commercial water-based ink strate or PET substrates. Prior to the device fabricaHPS-030AE1 with 55-60 wt% loading from Novacention, the glass substrates were cleaned with soap solutrix without further modification, while the substrate tion/acetone/isopropyl alcohol/deionized water for 10 ACS Paragon Plus Environment
}
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 16
2
Page 3 of 16
3
eV
3.26
200 112 201
110
(F(R)h ν)2
102
103
100
002
Intensity (a.u.)
101
4 is maintained at 45 o C to facilitate drying. The print(a) x 10 4 ing profiles of this Ag ink at various substrate tempera100 002 3.5 tures have been previously reported. 27 The aerosol mist is then generated by fixing the virtual impactor/carrier 3 gas flow rate 610 sccm/650 sccm. The sheath gas at a 2.5 flow rate of 40 sccm is used to focus the aerosol stream. 34 34.5 35 31.4 31.8 32.2 The electrodes are finally laser sintered at a power out2 101 put of 100 mW with a spot size of 30 µm using built-in 1.5 infrared multi-mode laser. The conductivity of Ag was 1 measured at 106 S.m−1 using four-point probe measure35.5 36.3 37 ment. 0.5 For printing the ZnO, the substrate temperature is 0 elevated to 60 o C. The ZnO ink is prepared by mixing 30 40 50 60 70 ZnO nanopowder (Anachemia Chemicals Inc, > 99%) 2θ (Degrees) (b) in distilled water (0.6 M). The gas flow rates, sheath Ultraviolet Visible gas/virtual impactor/carrier gas are set to 100 sccm/800 100 sccm/850 sccm, respectively. The tube temperature is maintained at 80 o C to reduce the water content of 80 material being printed and the printing stage speed is maintained at 2 mm.s−1 . Both the ink formulations 60 are stirred continuously during the printing process run. The printing is followed by annealing at 150 o C on a 40 heating plate for 30 min to remove the residual solvent in controlled environment (nitrogen). The thickness of 20 3.1 3.3 the metal electrodes and the ZnO layer are measured Energy (eV) at 1.85 µm and 2.9 µm respectively using a Dektak XT 0 B516 profilometer. 200 300 400 500 600 700 Wavelength (nm) Material Characterization. An optical image of the fabricated device is shown in Figure 1 (c). The figFigure 2: Characterization of ZnO film with (a) X-ray ure inset shows the scanning electron microscopy (SEM) diffraction pattern; (b) Diffused reflectance spectra, inset images of the ZnO film, showing a well-connected polyindicates the Tauc’s plot. crystalline porous film morphology. The magnified image shows the hexagonal symmetry of the ZnO nanofilm shows 85 % reflectance in the visible spectrum, crystals with large surface-to-volume ratio. The nanocompared to approximately 5 % in the UV region. The crystals were observed to have a size spreading from material shows low absorbance for photons with wave100 nm up to 1 µm, in contrast with colloidal quantum length above 400 nm making it insensitive to visible dot based films having much smaller nanoparticle sizes. wavelengths. The inset of figure 2, represents the Tauc’s This alternative approach leads to a significant reducplot. From this plot, the optical band gap of the semition in the cost of the nano-particle ink. For instance, conductor is calculated at 3.26 eV by the extrapolation a commercial ZnO nanoparticle ink with particle size of the tangent. This energy corresponds to a wavelength spreading in the range of 7-17 nm, from Sigma Aldrich, of 380 nm in accordance with the literature. 5 is 20000 times more expensive than the cost of this ink. Figure 2 (a) shows the X-ray diffraction pattern for Optoelectronic Characterization. The optoelecthe ZnO film printed on glass substrate as measured ustronic characterization of the devices was performed using Bruker’s D8 Advance diffractometer. The pattern ing a Lake Shore probe station (Figure 3 (a)) in ambient distinctly depicts the characteristic (hkl) peaks at 2θ anenvironment and in vacuum. The electrical parameters gles 31.78, 34.43 and 36.27, corresponding to the lattice of these devices were measured using an Agilent B1517A planes (100), (002) and (101) matching ZnO’s hexagsource meter. Current-voltage (IV) curves, shown in onal wurtzite structure. The high intensity of these Figure 3 (b) on a logarithmic scale, are obtained by peaks affirms the crystallinity of the material and the sweeping the bias voltage from -20 V to +20 V. The random orientation of the crystals as also observed from sweep was performed in the dark and during illuminathe SEM micrograph in Figure 1 (c). tion under a 40 µW/cm2 UV lamp with a dominant wavelength at 252 nm. The ratio between light and The optical absorption of the ZnO film was studied dark current, defined as the ON/OFF ratio, is approxiusing diffused reflectance spectroscopy (DRS), obtained mately ∼ 106 . The photodetector is therefore sensitive by a double-beam spectrometer (Lambda 750) on a solid enough to detect weak UV light intensities. sample as shown in figure 2 (b). The ZnO nano-crystal ACS Paragon Plus Environment
Reflectance %
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 Photonics
ACS Photonics
(c)
(a)
2
Current (A)
−5
x 10
1.5
Bias Voltage = + 20 V ON
ON
ON
ON
1
0.5
0 0
(b)
(d)
10
~106
40 Time (s)
Dark
−11
−13
10
−10
0 Bias (V)
10
−8
τ1 = 0.43 sec
−12
10
80
~106 τ3 = 3.2 sec
−10
20
OFF
60
τ2 = 1.3 sec
10
40 μW/cm2
OFF
Bias Voltage = + 20 V
Light
10
10
−20
OFF
−6
Current (A)
−9
20
10
−7
10 10
OFF
−4
10
Light
−5
Current (A)
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 4 of 16
0
Dark 100
200 Time (s)
300
Figure 3: (a) Lake Shore probe station used for optoelectronic characterization; (b) IV curves under 252 nm wavelength light illumination (40 µW/cm2 ) and dark condition; Time-dependent photo current with 20V bias and 40 µW/cm2 UV excitation (switched by a mechanical shutter) in (c) linear and (d) logarithmic scale. The rising and falling edge are fitted with a single order and double order exponential respectively.
The illumination is modulated using a shutter assemThis first decay is followed by a second decay of three bly controlled by a frequency generator from Stanford orders of magnitude which occurs this time on a 40 s Research. Figure 3 (c) shows the time photoresponse period. In order to properly fit this behavior, a secof the device on a linear scale when illuminated at 40 ond order exponential fit is necessary, with a first time µW/cm2 with a light source modulated at 0.05 Hz. In constant of 1.3 s and a second slower time constant of order to perform a detailed observation of the response 3.2 s. This is an indication that there are two separate speed, the pulse duration was increased and the phophysical mechanisms controlling the photo decay of this tocurrent response was plotted on a log-linear scale in device, which merits further investigation. Figure 3(d). The plot on Figure 3(d) clearly shows that Figure 4 plots the dependence of the photocurrent on a steady state dark current on the order of a few pA is the illumination intensity in (a) ambient and (b) vacachieved within a short time after the illumination. uum environment. The light intensity was measured To further analyze the dynamic behavior of those deby a standard silicon photodiode(S1226-18BQ) bought vices, we perform a curve fitting on the rising and the from Hamamatsu Corp. In air, the current changes by falling edges of the time photoresponse curve. In Figure four orders of magnitude as the light intensity is in3 (d), we can observe on the rising edge that, as the ilcreased from 2 · 10−4 mW/cm2 to 4 · 10−3 mW/cm2 . lumination is turned on, the photo current increases six The photocurrent increases with irradiance following a orders of magnitude in 10 s. This behavior can be acpower law corresponding to Iph ∝ P 3.5 , while in vacuum curately fitted with a single order exponential growth, the response is linear with a power law fit of Iph ∝ P 1.09 . characterized by a time constant of 0.43 s. On the falling In photoresistors and photodiodes, the photocurrent is edge, when the illumination is turned off, the photocurexpected to increase linearly with the number of incomrent first decreases by three orders of magnitude in 15 s. ing photons. Therefore, an additional mechanism is reACS Paragon Plus Environment 4
Page 5 of 16
3
10
Ipc ∝ P3.5
−6
10
−8
10
1
10
−1
−10
10
−12
10
10
−14
−4
−3
10
−2
10 10 Optical Intensity (mW/cm2)
10
Ipc ∝ P1.09
−6 −8
10
2
10
101
10−10
100
−12
−1
10
−14
10 −4
10
(d)
103
Vacuum
10
e
−1
−3
−2
10 10 2 Optical Intensity (mW/cm )
metal
Ec
Ag
hv Eg
h+
(b) −4 10
10
n-semiconductor
−3
10 10
metal
Responsivity (A/W)
Photocurrent (A)
(c)
Air
Responsivity (A/W)
(a) 10−4
Photocurrent (A)
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 Photonics
Ag
-
Ev Dark
Ag
Oxygen adsorption
ZnO
UV Photodesorption
Ag
ZnO
−2
10
Figure 4: Photo current versus light intensity at 20V bias for (a) ambient air and (b) vacuum condition. A power law fit is made for both conditions. Also plotted are the responsivities, showing light sensitivity as a function of light intensity; Device energy band diagram of (c) MSM photodiode under UV illumination with external bias applied (d) barrier modification due to the adsorption and photo-desorption of oxygen under UV illumination.
quired to explain the superlinear response.
cles by capturing electrons from the n-type ZnO: O2 (g) + e− → O− 2 (ad)
Oxygen Adsorption Mechanism. The difference in characteristics of the device between operation in air and vacuum can be attributed to oxygen adsorption mechanism on the surface of the ZnO crystals. As illustrated in the energy level diagram of the MSM photodiode structure under UV illumination shown in Figure 4 (c), Schottky barriers are formed on the AgZnO interfaces due to difference in work functions. Absorbed photons generate excitons which are split into holes and electrons by the built-in electric field. The Schottky barriers significantly increase the speed of the device, because of the built-in electric field at the junctions 9,21,28,29 . If the barrier has a fixed height, the photocurrent scales linearly with the amount of absorbed photons.
(1)
The oxygen adsorption increases the Schottky barrier at the Ag-ZnO interface, thereby reducing the dark current. Upon illumination, the absorbed UV photons generate excitons and the photo-generated holes release the adsorbed oxygen by the following reaction: h+ + O− 2 (ad) → O2 (g)
(2)
The oxygen desorption will lead to the decrease of the Schottky barrier and results in a larger flow of charge carriers by diffusion through the junction. The current due to the illumination therefore increases in a superlinear fashion. 9 This modification of the Schottky barrier therefore results in a large ratio between dark and illuminated current upon adsorption of oxygen. Moreover, from this relationship,between Schottky barrier height and the amount of adsorbed oxygen, we expect
Figure 4(d) explains how the Schottky barrier could be modified by oxygen adsorption. In the dark, oxygen molecules are adsorbed on the surface of the nanopartiACS Paragon Plus Environment 5
ACS Photonics
(a)
(d)
Light
−5
nm
10
−7
0 μm
y:
.00
μm
x:
.0 10
~106
−9
10
Dark
−11
10
m 0μ
−13
10
40 μW/cm2
−20
(e)
−4
10
Porous ZnO film Non−porous ZnO film
ON
~106
−8
10
−12
10
0
400 Time [s]
Bias voltage = +20V
1
0.6 0.4
OFF
OFF
200
20
0.8
~103
10
−5
10
1.2
ON
−10
0 Bias[V]
1.4
Bias Voltage = + 20 V
−6
10
−10
x 10
Current (A)
(b)
10
10
Current [A]
0.12 μm
Current [A]
600
0.2 0 0
800
(c)
(f)
100
200 7100 Time [s]
10
Number of Devices
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 16
7200
7300
40 Devices
8 6 4 2 0 0
2
4
6 8 ON/OFF Ratio
10
12 5 x 10
Figure 5: (a) AFM image of non-porous ZnO film, detailing a smooth, crystalline surface; (b) Comparison of time-dependent photocurrent in photodetector with porous and non-porous ZnO film under switching 252 nm wavelength light source; (c) Optical image of fully aerosol-jet printed array of ZnO photodiodes on a mechanically flexible PET substrate, with < 150o C processing; (d) IV curve demonstrating same order of ON/OFF ratio for a device printed on flexible substrate as on rigid substrate; (e) Repeatability of time-dependent photo current tested for over 2 hours duration with 20V bias and 40 µW/cm2 UV excitation (switched by a mechanical shutter) on PET substrate, plotted in linear scale; (f) Histogram of ON/OFF ratio measured for 40 devices on PET substrate as shown in Figure 5(c).
ACS Paragon Plus Environment 6
Page 7 of 16 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 Photonics
Table 1: Performance comparison of published ON/OFF Incident Rise Fall Ratio Power Time Time [µW/cm2 ] [s] [s] 6 Gupta et al. (This work) ∼ 10 40 0.43 1.3 / 3.2 Cheng et al. 7 4 · 105 7600 N/A 0.05 Paper
ZnO UV photodetectors. Processing Morphology Temp. [o C] 150 Nano-porous N/A Nano-wire
Jin et al. 9
∼ 106
830
