Drain

Subscriber access provided by University of Missouri-Columbia

Article

Unraveling the Issue of Ag Migration in Printable Source/Drain Electrodes Compatible with Versatile Solution-Processed Oxide Semiconductors for Printed Thin-Film Transistor Applications Gyu Ri Hong, Sun Sook Lee, Hye Jin Park, Yejin Jo, Ju Young Kim, Hoi Sung Lee, Yun Chan Kang, Beyong-Hwan Ryu, Aeran Song, Kwun-Bum Chung, Youngmin Choi, and Sunho Jeong ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b00524 • Publication Date (Web): 07 Apr 2017 Downloaded from http://pubs.acs.org on April 9, 2017

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 free 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 accessible to all readers and 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.

ACS Applied Materials & Interfaces 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 34

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

Unraveling the Issue of Ag Migration in Printable Source/Drain Electrodes Compatible with Versatile SolutionProcessed Oxide Semiconductors for Printed Thin-Film Transistor Applications Gyu Ri Hong,a,b Sun Sook Lee,a Hye Jin Park,a Yejin Jo,a Ju Young Kim,a Hoi Sung Lee,a Yun Chan Kang,b Beyong-Hwan Ryu,a Aeran Song,c Kwun-Bum Chung,c,*Youngmin Choi,a,* Sunho Jeong a,*

a

Division of Advanced Materials, Korea Research Institute of Chemical Technology

(KRICT), 141 Kajeongro, Daejeon 305-600, Republic of Korea. b

Department of Materials Science and Engineering, Korea University, Seoul 136-713,

Republic of Korea c

Division of Physics and Semiconductor Science, Dongguk University, Seoul, 100-715,

Korea

KEYWORDS: Ag, migration, print, solution-process; transistor

1

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

ABSTRACT In recent decades, solution-processable, printable oxide thin-film transistors have garnered a tremendous amount of attention given their potential for use in low-cost, large-area electronics. However, printable metallic source/drain electrodes undergo undesirable electrical/thermal migration at an interfacial stack of the oxide semiconductor and metal electrode. In this study, we report oleic-acid-capped Ag nanoparticles that effectively suppress the significant Ag migration and facilitate high field-effect mobilities in oxide transistors. The origin of the role of surface-capped Ag nanoparticles is clarified with comparative studies based on X-ray photoelectron spectroscopy and X-ray absorption spectroscopy.

2


Page 2 of 34

Page 3 of 34

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 In recent decades, a variety of oxide semiconductors have been the subject of tremendous levels of attention from researchers studying various optoelectronic applications owing to their high transparency, relatively high field-effect mobility, excellent film uniformity, environmental stability, and good electrical stability.1,2 In particular, the capability of forming a device-quality channel layer in a thin-film transistor (TFT) architecture via a wet-chemical methodology has been suggested to offer potential toward the practical realization of low-cost, high performance circuits. However, apart from an extensive effort to exploit highperformance, solution-processable oxide semiconductors3,4, easily processable highly conductive electrodes are rarely reported for use as source/drain electrodes. Either aluminum or indium tin oxide has been widely used in vacuum-deposited source/drain electrodes, facilitating an efficient charge injection into oxide channel layers. Printable metal nanoparticles can be regarded as a cost-effective alternative, as metal nanoparticles approximately a few tens of nanometers in size are capable of being transformed into a highly conductive layer through a size-dependent, effective densification reaction at low temperatures below 250 oC.5 As another alternative, precursors were used as printable materials, with organic compartments thermally decomposed by annealing at elevated temperatures, resulting in highly conductive electrodes.6 The amount of material waste is extremely low due to the use of a direct-printing process without the need for the costly photo-mask used in the conventional sequential, time-consuming photolithography process.7 3



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

However, metal nanoparticles for use as source/drain electrodes well matched electrically to oxide semiconductors have not been demonstrated, while metal-nanoparticle-based printed electrodes have been commonly reported in combination with solution-processed organic semiconductors.8,9 Ag nanoparticles can be formulated into printable fluids with the advantages of relative cost-effectiveness over their Au counterparts and chemical/environmental inertness over their Cu counterparts. It is very demanding to suppress the evolution of surface oxide in the nano-sized Cu phase due to the thermodynamic stability of oxide phases.10 It is also quite difficult to reduce chemically other metal species when synthesizing nanoparticles owing to their low reduction chemical potentials. However, in practical electronic applications, despite the various merits of the Ag nano-phase, a sequential stack between a metallic Ag phase and an inorganic functional layer undergoes undesirable thermally induced and bias-dependent migration,11,12 resulting in unpredictable degradation of the device performance. In fact, it has been reported that thin-film transistors employing oxide semiconducting channel layers show inferior device performance when Ag-nanoparticle-driven, highly conductive conductors are utilized as a source/drain electrode.13 The diffusivity of copper, another conductive element, in oxide skeletons is known to be much lower compared to that the Ag element; however, even with vacuum-deposited copper source/drain electrodes, the Cu phase diffuses in wellengineered oxide semiconductors, forming either a conductive metallic phase or an insulating oxide phase in an oxide semiconductor matrix.14,15 Ag-nanoparticle-based electrodes have an economical restriction as compared to their Cu counterparts, and the electrical interconnections constituting most electrode lines in practical circuitry should be formed with low-cost, printed Cu electrodes. However, for source/drain electrodes in contact with oxide 4


Page 4 of 34

Page 5 of 34

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


semiconductors, the realization stable electrical performance would be of paramount importance, and easily processable, chemically inert Ag is likely a more promising candidate to facilitate high-performance, printed oxide transistors. In this study, we designed printable metallic fluids containing Ag nanoparticles surrounded firmly by an organic capping molecule, oleic acid, so as effectively to suppress the thermal/electrical migration inside In-Zn-O (IZO) oxide semiconductors in a type of thinfilm transistor device architecture. Oleic acid has a long hydrocarbon chain and a carboxyl group as a functional end group that can anchor chemically with surficial atoms in metal nanoparticles. The means by which it prevents the migration of the Ag phase and its critical impact on device performance capabilities of TFTs are clarified based on a comparative study with different types of printable Ag conductive fluids in conjunction with in-depth X-ray photoelectron spectroscopy (XPS) and X-ray absorption spectroscopy (XAS) analyses. It is revealed that the combined effect of the amount of Ag phase diffused inside the oxide semiconductors and the chemical state of the Ag phase critically impacts device performance capabilities, providing a new chemical pathway for high-performance oxide transistors with printed Ag electrodes.

2. MATERIALS AND METHODS 2.1. Synthesis of oleic acid-capped Ag nanoparticles. Octylamine (99%), oleic acid (90%), phenylhydrazine (97%) were purchased from Aldrich, and Ag nitrate (99.9%) was purchased from Kojima Chemicals. Toluene (99.5%) was purchased from Duksan. For synthesizing Ag nanoparticles, 9.5 g of Ag nitrate and 25.4 ml of oleic acid were added to a three-neck round5



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

bottomed flask containing 93.2 ml of octylamine. The prepared reacting solution was heated to 80 oC and stirred with a magnetic stirrer under a refluxing condition. When the temperature reached 80 oC, 87.4 g of phenylhydrazine was injected with an injection rate of 5 ml/min. The reaction was continued for 60 min and then cooled to room temperature. The synthesized Ag nanoparticles were washed with toluene by centrifugation. 2.2. Preparation of Ag nanoparticle paste (Ag NP-paste). For preparing Ag nanoparticle pastes, the synthesized Ag nanoparticles were mixed with ethyl cellulose (100cP, 48.0-49.5%, Aldrich) as a binder in α-Terpineol (C10H18O, 90%, Aldrich) as a solvent. The solid loading was 80 wt% and the composition of ethyl cellulose was 1 wt% with respect to Ag nanoparticles. 2.3. Preparation of SIS-incorporated Ag flake paste (Ag F-paste #2). In order for synthesizing amine-functionalized multi-walled carbon nanotubes (NH2-MWCNT), a mixture composed of 1.4 g of MWCNT (97%, length: about 10 µm, Applied Carbon Nano Co. Ltd.), 0.35 g of perylene-3,4,9,10-tetracarboxylic dianhydride (PTDA, 97%, Aldrich), 350 mL of methylene chloride (99.5%, Samchum), 70 mL of triethylamine (99%, Samchun), and 14 mL of ethylenediamine (≥99%, Aldrich) was sonicated for 1 h and stirred vigorously for 24 h. The mixture was centrifuged, and obtained samples were washed with methanol, methylene chloride and methanol in order by centrifugation, and then, dried in a vacuum overnight. poly (acrylic acid)-capped silver nanoparticles were synthesized as follows. Ag nanoparticles were synthesized via chemical reduction of Ag ions in DI water. To prevent the interparticular agglomeration, polyvinylpyrrolidone (PVP, average MW ~10,000) and poly(acrylic acid) (PAA, sodium salt, average Mw ~15,000, 35 wt% in H2O) was incorporated as a surfacecapping molecule, and sodium borohydride was used as a reducing agent. 4.7 g of Ag nitrate 6


Page 6 of 34

Page 7 of 34

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


(99.9%, Kojima Chemicals), 3.8 g of poly(acrylic acid), and 6.0 g of PVP were added into a three-neck, round-bottomed flask containing 100 mL of DI water with pH of 11. The prepared reacting solution was heated to 60 °C and stirred with a magnetic stirrer under a refluxing condition. When the temperature reached 60 °C, 9.7 g of a mixture of sodium borohydride (98.5%, Kojima Chemicals) and DI water (pH 11) was injected. After the reaction for 60 min at 60 °C, the synthesized Ag nanoparticles were selectively separated by a centrifugation method, and the obtained Ag nanoparticles were washed with DI water by centrifugation. In order for decorating the NH2-MWCNT with PAA-capped Ag nanoparticles, 50 g of an aqueous NH2-MWCNT solution with a concentration of 3 mg/ml was mixed with 6.75 g of an aqueous poly(acrylic acid) capped-Ag NP solution with a concentration of 20 wt%. After a subsequent sonication/homogenization mixing process, the mixture was centrifuged to collect the PAA-Ag/NH2-MWCNT hybrid material. Then, the precipitates were dispersed by adding DI-water, followed by the pH adjustment to be 4. After a centrifugation process, the precipitates were re-dispersed with ethyl alcohol and the precipitates obtained by another centrifugation process were mixed with a proper amount of Ag flake (SF120, Ames Advanced Materials Corporation) and 50 g of toluene. After a centrifugation process, the precipitates were mixed manually with a proper amount of polystyrene-polyisoprenepolystyrene (SIS, styrene 22%, 12 poise @ 25 wt% in toluene, Aldrich) and 1,3dichlorobenzene (≥99.0%, Aldrich) in an agate mortar for 3 min. In paste, the ratio of conductive fillers to SIS was 94 wt% and the solid loading was 84 wt%. 2.4. Preparation of oxide semiconductor precursor solutions. For synthesizing the IZO precursor solution, 0.09 g of indium(III) nitrate hydrate (In(NO3)3 · xH2O, 99.999%, Aldrich) was dissolved in 6 g of 2-methoxyethanol (99.8%, anhydrous, Aldrich) at 80 oC for 40 min. 7



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

Then, 0.08 g of zinc acetylacetonate hydrate (Zn(C5H7O2)2 · xH2O, 99.995%, Aldrich) was added and stirred at 80 oC for 40 min. The resulting solution was stirred at room temperature for 24 hr. For synthesizing the In2O3 precursor solution, 0.9 g of acetylacetone (CH3COCH2COCH3, ≥99%, Aldrich) was added in 15 g of 2-Methoxyethanol (99.8%, anhydrous, Aldrich). Then, 0.9 g of indium(III) nitrate hydrate (In(NO3)3 · xH2O, 99.999%, Aldrich) and 0.27 g of 10 M ammonium hydroxide solution was added and stirred at room temperature for 3 hr. For synthesizing the IGZO precursor solution, 0.1 g of H2O and 0.33 g of ethanolamine (NH2CH2CH2OH, ≥99.0%, Aldrich) was added in 6 g of 2-methoxyethanol (99.8%, anhydrous, Aldrich). Then, 0.13 g of zinc acetate dihydrate (Zn(CH3COO)2 · 2H2O, ≥98%, Aldrich) and 0.06 g of gallium(III) nitrate hydrate (Ga(NO3)3 · xH2O, 99.9%, Aldrich) was added and dissolved at room temperature. Subsequently, 0.43 g of indium(III) nitrate hydrate (In(NO3)3 · xH2O, 99.999%, Aldrich) was added and dissolved at 80 oC for 40 min. The resulting solution was stirred at room temperature for 24 hr. For synthesizing the ZTO precursor solution, 0.11 g of ethanolamine (NH2CH2CH2OH, ≥99.0%, Aldrich) was added in 6 g of 2-methoxyethanol (99.8%, anhydrous, Aldrich). Then, 0.06 g of zinc nitrate hydrate (Zn(NO3)2 · xH2O, 99.999%, Aldrich) was added and dissolved at room temperature for 20 min, and 0.07 g of tin(II) chloride dehydrate (SnCl2 · 2H2O, ≥99.99%, Aldrich) was added and dissolved at 80 oC for 40 min. The resulting solution was stirred at room temperature for 24 hr. 2.5. Device fabrication. For fabricating the device with a structure of a top contact and a bottom gate, the heavily-doped Si wafers with 100 nm-thick SiO2 layer were washed subsequently with acetone, isopropyl alcohol, methyl alcohol, and DI-water and then, treated with ozone cleaner. The semiconducting channel layers were formed by spin-coating the 8


Page 8 of 34

Page 9 of 34

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


oxide precursor solutions. The spin-coated IZO and In2O3 layers were annealed for 2 hr at 350 and 250 oC, respectively. The IGZO and ZTO channel layers were annealed at 400 oC for 2 hr. All conductive pastes were printed using a programmable dispenser (Image Master 350PC Smart, Musashi) and a nozzle with an inner diameter of either 100 or 350 µm. The schematic showing a printing process and the picture of printing machine were shown in Figure S1. For the case of Ag F-paste #2, a part of solvent was evaporated at room temperature on purpose for adjusting the rheological properties. As a reference device, the Al electrodes were deposited through a shadow mask. For IZO, In2O3, IGZO and ZTO devices, the channel dimension was 200 µm in length and 1000 µm in width for transistors with Al source/drain electrodes, and 280-450 µm in length and 1000 µm in width for transistors with printed source/drain electrodes. Ag nanoparticle paste-based electrodes were annealed at 200 or 300 oC for 2 hr, and Ag flake paste-based electrodes were dried at 80 oC for 2 hr. 2.6. Characterization. The chemical structures of oxide semiconductors were examined by X-ray photoelectron spectroscopy (XPS, ESCA Probe, Omicron). The surface XPS data were collected using monochromatic AlKα radiation (1486.6 eV) in an ultrahigh vacuum system with a base pressure of ~10-10 Torr. The electronic structure near the conduction bands was measured by near edge X-ray absorption spectroscopy using total electron yield (TEY) mode in BL-10D of the Pohang Accelerator Laboratory (PAL) in Korea. The thermal decomposition behaviors were monitored by thermogravimetric analysis (TGA, Thermo plus EVO II TG 8120, Rigaku). The electrical performance of the produced transistors were analyzed in ambient conditions using an Agilent E5270B source-measure unit. Saturation mobilities were extracted from the slope of (drain current)1/2 versus gate voltage derived from the device transfer plot. 9



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

3. RESULTS AND DISCUSSION 3.1. Basic Properties of Electrodes Printed from Ag Nanoparticle- and Flake-Pastes. Scheme 1 depicts a schematic showing the device structure tested in this study and the experimental procedures associated with the printed oxide thin-film transistors. Devices with a bottom common gate and a top contact geometry were fabricated by spin-coating oxide semiconductor precursor solutions and then directly printing Ag nanoparticle/flake pastes by a dispensing technique. A silver nanoparticle paste was prepared by mixing oleic-acid capped Ag nanoparticles (with a diameter of 20 nm) in terpineol as a solvent with the incorporation of ethyl cellulose as a binder. The composition of the polymeric binder was 1 wt% with regard to Ag nanoparticles. The Ag-nanoparticle-based paste is denoted as Ag NP-paste in this study. The oleic acid-capped Ag nanoparticles were synthesized by reducing chemically Ag ions in a mixture of oleic acid as a capping molecule, octyl amine as a solvent, and phenyl hydrazine as a reducing agent, as reported in our previous study.16 As control experiments to prove the superiority of oleic-acid-capped Ag nanoparticles as a primary metallic ingredient for printable fluids, we prepared two types of Ag-flake-based pastes. One of the Ag flake pastes was purchased from CANS (Model No. ELCOAT P-100) and the other was formulated by mixing Ag flakes, Ag-nanoparticle-decorated multi-walled carbon nanotubes and a thermoplastic triblock copolymer, polystyrene-polyisoprene-polystyrene (SIS), in 1,3dichlorobezene as a solvent.17 The Ag-nanoparticle-decorated multi-walled carbon nanotubes were incorporated to improve the electrical conductivity by reinforcing the electrical connection between the Ag flakes. The composition of the thermoplastic triblock copolymer was 8 wt% in the paste. The commercial Ag flake paste used here and the aforementioned 10


Page 10 of 34

Page 11 of 34

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


SIS-incorporated Ag flake paste are denoted as Ag F-paste #1 and Ag F-paste #2, respectively, in this study. Note that we used a dispensing technique allowing for a printing process of highly viscous fluids, rather than an ink-jet printing process. To date, when forming printed electrode features, the easily processable ink-jet printing process has been commonly adopted;5-10 however, due to limited viscosity (in general, below a few tens of mPa·s), viscous printable metallic fluids are not applicable to the ink-jet printing process, particularly when using pastes with a sufficient amount of polymer as an ingredient. With the ink-jet printing process, various printable fluids cannot be investigated to clarify the critical role of organic moieties incorporated to suppress electrical/thermal Ag migration. In this study, we studied different types of printed metallic electrodes in which the charge transport is dominated by (i) direct conduction between metal particles fused by a sintering process, and (ii) percolation conduction between neighboring flakes in an intact contact after a drying process. With regard to the percolation conduction mechanism, even with the presence of a noticeable amount of a polymeric substance, charge conduction could occur effectively as long as enough conductive moieties to exceed the percolation threshold volume are included;18 thus, the resulting metallic fluid should be viscous to some extent so that it can be ejected through a nozzle by a dispensing technique as opposed to an ink-jet printing process. As shown in Figure S2, the rheological properties of the three types of printable Ag fluids were regulated to be appropriate for a dispensing printing process. In general, for a facile dispensing process, a storage modulus over 103 Pa is a minimal requisite to prevent loaded fluids from dropping according to gravitational force in an “off” state during printing process. However, in order to narrow the linewidth of the printed patterns, we increased the 11



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

storage modulus so that it ranged from 104 to 105 Pa with viscosities of around 106 – 107 cP, allowing for the formation of printed Ag electrodes with a linewidth nearly identical to the diameter of the nozzle (Figure S3). For the Ag NP-paste sample, the conductivities of the printed electrodes varied with the annealing temperature. As-printed electrodes are resistive after drying at 80 oC, and resistivity of approximately 7.6 µΩ·cm, comparable to values observed in well-formed nanoparticle-driven electrodes,19,20 was noted after annealing at temperatures over 200 oC. This is attributable to the well-known fact that charge conduction cannot occur in nanoparticle assemblies, and a thermally triggered sintering process enables the formation of a partially fused microstructure in which neighboring nanoparticles are connected to each other, establishing overall charge conduction pathways. Along with a microstructural transformation from a particle assembly to a film-like structure, the resident organic moieties are decomposed thermally, leaving behind electrodes with small amounts of organic impurities. For both types of Ag flake pastes, the conductive nature evolves simply after a drying process at 80 oC. The resistivities were measured and found to be 110 and 74.6 µΩ·cm for the electrodes printed using Ag F-pastes #1 and #2, respectively. Two-dimensional metallic flakes could be stacked to create face-to-face contact with each other, even in a polymer matrix. In fact, this arrangement is strengthened in a polymer matrix due to the structural reinforcement caused by the attractive interaction between polymeric chains during the solvent drying process. As the composition of the metallic flakes increases with regard to the matrix polymer, the distance between neighboring metallic flakes becomes narrower and eventually close enough to allow for charge tunneling conduction over a specific composition (known as the percolation threshold composition).18 This percolation conduction is favorable for one-dimensional and two-dimensional directional materials, rather than zero-dimensional

12


Page 12 of 34

Page 13 of 34

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


nanoparticles, owing to the low percolation threshold compositions of the former types. Such an underlying mechanism facilitates low-temperature processability for Ag-flake-based, percolation conduction-dominated electrodes, but the resistivity is much higher than direct conduction-dominated, sintered electrodes owing to the presence of polymer constituent materials. The current-voltage characteristics are shown in Figure S4 for all of the printed electrodes. 3.2. Electrical Performance of Devices Employing Source/Drain Electrodes Printed From Ag Nanoparticle- and Flake-Pastes. According to the TGA analysis of the three types of printable Ag pastes (Figure 1a), it was observed that Ag F-paste #1 includes organic components up to a level of 14.3 wt%, while Ag NP-paste is composed of capping molecules and binders with a composition of 7.6 wt%. Ag F-paste #2 was not measurable, as it was not ground into a powder after drying due to the inherent nature of the thermoplastic polymer included in it. It can be deduced that the composition of the organic component is limited to a value of 8 wt%, the level incorporated when formulating the paste. When those printed electrodes are employed as a source/drain electrode in a TFT device structure, it is speculated that Ag F-paste-based electrodes will exhibit superior device performance levels owing to the suppressed Ag migration characteristic. The organic residues surrounding the Ag phase are capable of interrupting the movement of mobile Ag species. The Ag NP-paste-based electrodes lack organic components due to partial thermal decomposition at an elevated temperature above at least 200 oC. As observed in the TGA analysis, the amount of organic moiety is reduced from 7.6 to 1.0 wt% after annealing at 200 oC. Interestingly, unlike such an expectation, transistors with source/drain electrodes printed using Ag NP-paste (after annealing at 200 oC) exhibited the best device performance (Figures 1b-f). The channel layer 13



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 34

was prepared by spin-coating the IZO precursor solution, followed by annealing at 350 oC in air. Al electrodes were thermally evaporated as a reference sample. The detailed device performance parameters are summarized in Table 1. The field-effect mobility, a representative parameter in transistors, were measured and found to be 7.9, 0.05 and 5.3 cm2/V·s for transistors employing source/drain electrodes printed using Ag NP-paste, Ag F-paste #1 and Ag F-paste #2, respectively, while the reference device exhibited field-effect mobility of 9.0 cm2/V·s. The output characteristics of IZO transistors employing Ag NP-paste based (200 oCannealed) and Ag F-paste #2 based source/drain electrodes are shown in Figure S54, exhibiting appropriate ohmic contact behaviors. For transistors with Ag NP-paste-based electrodes annealed at 300 oC, the conductive characteristic evolved in the channel layer. The inferior electrical performance of transistors employing electrodes printed from Ag F-paste #1 is a representative feature observable in oxide transistors with source/drain electrodes prepared from commercially available Ag pastes. 3.3. Spectroscopy Analyses for Understanding the Different Device Performance. In order to clarify the origin of the distinctively different device performances depending on the type of electrode used and the annealing temperature, we analyzed the X-ray photoelectron spectroscopy

(XPS)

spectra

of

IZO

channel

layers

prepared

from

different

electrode/semiconductor stacks. The electrodes were printed on top of IZO layers coated onto Si wafers and then annealed at 200 and 300 oC for Ag NP-paste-based electrodes and dried at 80 oC for Ag F-paste-based electrodes. Subsequently, the printed electrodes were delaminated using a solvent from the semiconducting layers. Figure 2a shows the compositional variation of Ag inside the IZO semiconductors. Notably, it appears clearly that for the Ag F-paste #2 sample, Ag migration into the underlying oxide layer is prohibited effectively with a 14


Page 15 of 34

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


composition of 0.64 at%; this is within the measurable minimal range for an XPS analysis. The composition of Ag was measured and found to be 27.7 at% for the Ag F-paste #1 based electrode. Taking into consideration the amount of the incorporated organic component for both Ag F pastes, this implies the effectiveness of the triblock copolymer in suppressing the migration of the Ag phase in semiconducting layers. The thermoplastic triblock copolymer is composed of two segments, one with a high glass transition temperature (Tg) and one with a low Tg. The SIS used in this study possesses a rubbery isoprene segment with a Tg below -60 o

C and a rigid styrene segment with a Tg above 80 oC. At room temperature and at an elevated

temperature of 80 oC (for the drying process), the rubbery isoprene phase is likely to exhibit a liquid-like nature, by which it wets the surface of the Ag flakes in conformal contact while the rigid styrene segment maintains a film structure. Both the isoprene and the styrene segment do not have any functional groups to capture surficial atoms in Ag flakes. Thus, it is believed that the polymeric conformation effectively surrounding the Ag flakes would restrict the undesirable migration of the Ag phase. However, a polymeric substance can be partially degradable in SIS-incorporated electrodes due to local joule heating under AC operation conditions. An interesting feature in the XPS-based compositional analysis is that the composition of the Ag element in the IZO semiconductor layer is limited to a value of 3.6 at% for the Ag NP-paste-based electrode annealed at 200 oC. Compared with Ag F-paste #1, the amount of the organic component incorporated in the fluid is much lower and the organic residues thermally decompose after annealing at 200 oC, leaving behind a residue-less, highly conductive electrode. Rather, the thermally activated migration of Ag should be accelerated by annealing at an elevated temperature. Figure 2b shows the XPS C 1s spectrum of the as15



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

synthesized, oleic-acid-capped Ag nanoparticles. As shown in Figure S6, the N 1s spectrum is mostly absent, indicative of the absence of octyl amine, which is used as a solvent, along the surface of the Ag nanoparticles. In the C 1s spectrum, the features at binding energies of 284.5 and 287.3 eV are attributed to the alkyl chain (C-C) and the carboxylate (-COO-) moiety, respectively.20 The measured atomic fractions of both sub-peaks were 94.3 and 5.7 at%, corresponding well to the chemical formula of oleic acid. It is well known that a carboxyl group could give rise to a firm chemical binding to transition-metal species through monodentate and bidentate bonding characteristics.21 By virtue of this type of chemical passivation, for even Cu nanoparticles it has been demonstrated that the formation of a thermodynamically stable surface oxide is significantly suppressed, allowing for the chemical synthesis of oxide-free Cu nanoparticles.20,21 According to an examination of the C 1 spectrum (Figure 2c) of Ag nanoparticles annealed at 200 oC, the chemical contribution of carboxylate is more apparent. The sub-peaks positioned at 284.5, 286.4 and 288.5 eV result from alkyl and from carbons bonded with oxygens and carboxylate, respectively. As shown in Figure 2d, the atomic fraction of carboxylate was varied from 5.7 to 10.6 at% for the assynthesized Ag nanoparticles annealed at 200 oC. This indicates that thermal decomposition prevails on hydrocarbon chains rather than on carboxylates during the thermal annealing process in air, preserving the means of capturing surficial Ag atoms. However, the Ag phase migrated significantly inside of the IZO semiconductors after annealing at 300 oC (Figure 2a) owing to the complete decomposition of the organic components and the input of more thermal energy. Note that in practical electrical circuits, the resistance of patterned conductive features that provide interconnections between each active component plays a critical role in 16


Page 16 of 34

Page 17 of 34

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


determining the overall performance of the resulting array devices. At a given geometry of electrode, the resistance is primarily governed by the value of the electrical conductivity. Thus, a nanoparticle-based electrode which undergoes a thermally driven sintering process and thermal decomposition of the organic residue would be more favorable in practical circuit devices. However, to date, a characteristic chemical methodology which effectively prevents the migration of the metallic phase inside oxide semiconductors has not been demonstrated in highly conductive, sintered metallic electrodes. In this study, by introducing firmly passivated Ag nanoparticles, the migration of the Ag element was successfully suppressed in electrodes, exhibiting high conductivity in the resulting electrodes and excellent electrical properties in the resulting thin-film transistor devices. In order to analyze the chemical bonding characteristics of the Ag element diffused inside the IZO semiconductor, the detailed chemical states of Ag were investigated by the deconvolution of the Ag 3d state, as shown in Figure 3. The sub-peaks at binding energies of 368.3 and 367.7 eV are associated with Ag0 in the metallic phase and Ag+ in the oxide phase, respectively.22 As shown in Figure 3a, for the IZO layer with the Ag NP-paste-based electrode annealed at 200 oC, the Ag phase which migrated inside the IZO channel layer preserves the metallic state with the partial formation of an Ag2O phase. Recently, it was demonstrated that the intentional incorporation of conductive moieties (such as carbon nanotubes and graphenes) inside oxide semiconductor layers would improve the field-effect mobility by creating locally ballistic charge conduction pathways, while the value of the sub-threshold swing is degraded accordingly.23-25 It is presumed that the deterioration in the device performance by the high work function of the Ag phase with regard to n-type oxide semiconductors is offset to some degree in the conductive metallic phase-incorporated oxide channel layer, which results in a 17



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

slightly lower field-effect mobility and a moderate value of the subthreshold swing compared to a reference device with thermally evaporated Al electrodes. The metallic state inside the IZO semiconductor layer is also preserved for the Ag NP-paste based electrode/IZO semiconductor stack after annealing at 300 oC, with the partial presence of the Ag2O phase (Figure 3b). The atomic fractions of metallic silver phase were 30.9 and 35.3 % for electrodes annealed at 200 and 300 oC, respectively. For the IZO layer with the printed electrode annealed at 300 oC, the presence of an excessive amount of Ag phase would make the channel layer conductive and cause it to lose its characteristic semiconducting nature. Additional critical feature observed in the device structure of the heavily doped Si/100-nmthick SiO2/Ag NP-paste-based electrode annealed at 300 oC is that the SiO2 layer does not maintain its superior insulating property, showing a leakage current of 0.01 A/cm2 (a compliance level in the measurement) at an electric field of 0.3 MV/cm (Figure S7). This is caused by the significant thermal/electrical migration of the Ag phase into a robust gate dielectric layer, SiO2, confirming the critical issue of metallic phase migration in electronics. In industries, in order to suppress the migration from vacuum-deposited Cu source/drain electrodes, an additional robust oxide layer is inserted beneath the Cu layer.14,15 In this study, we demonstrate that an appropriate chemical strategy can resolve such a critical issue by surrounding the surfaces of the metallic species with chemically anchoring molecules without the involvement of additional layers. The limited value of the field-effect mobility for TFTs with the Ag F-paste #1 based source/drain electrodes is attributable to the excessive formation of Ag2O phase, as shown in Figure 3c. The measured atomic fraction of the metallic silver phase was only 8.4 %. It is well known that the significant degradation in the device performance for oxide TFTs with 18


Page 18 of 34

Page 19 of 34

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


vacuum-deposited Cu source/drain electrodes results from the formation of locally distributed, electrically resistive copper oxide phases inside the oxide semiconductors.14,15 In addition, an inferior charge injection by an energy band mismatch between the metallic Ag phase and the oxide semiconductor would be more problematic compared to Ag NP-paste based (200 oC annealed) electrodes, as an excessive amount of Ag element is diffused inside the semiconductors, creating a long-range interfacial boundary with an energy barrier. Moreover, the characteristic feature of electrodes with a percolation conduction mechanism is that the organic component remains in the finally processed printed structures without undergoing thermal decomposition. Thus, the organic residues present along with conductive metallic fillers could act as an additional barrier against an efficient charge injection. This limitation of percolation conduction-based electrodes can be resolved to some extent by incorporating a thermoplastic triblock copolymer that can suppress the migration of the metallic Ag phase into the oxide semiconductors. As shown in Figure 1f and Table 1, moderate field-effect mobility of 5.3 cm2/V·s was achievable for TFTs employing electrodes printed using Ag Fpaste #2. Given that the migration of the Ag phase is almost completely prohibited, the oxide semiconductors could show inherent properties in terms of the threshold voltage and subthreshold swing, similar to TFTs with thermally evaporated Al electrodes. However, the organic residues are still present in printed electrodes, restricting the charge injection behavior to some degree, limiting the value of the field-effect mobility. The effectiveness of oleic-acid-based surface passivation was also reported for Cu nanoparticle-based electrodes coupled with solution-processed zinc-tin oxide (ZTO) channel layers.26 It was suggested that Cu electrodes (printed from a mixture of oleic-acid-capped Cu nanoparticles, ethyl cellulose as a binder and 3-aminopropyltriethoxysilane (APTS) as an 19



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

adhesion promoter) could be used as source/drain electrodes when annealed at 250 oC under an inert atmosphere, exhibiting field-effect mobility with a value half (2.6 cm2/V·s) that (5.4 cm2/V·s) of ZTO transistors employing sputtered ITO electrodes. This was enabled by the lower diffusivity of the Cu element as well as the presence of a siloxane layer transformed chemically from APTS and a surficial oxide layer formed after a pre-baking process at 200 oC in air. Cu-based electrodes have attracted much attention given their economic processing methods. However, despite the cost-effectiveness of the raw material (compared with silver), the costly provision of an inert atmosphere endows an additional production cost during the process of synthesizing Cu nanoparticles and annealing printed Cu electrodes, overshadowing the critical advantages of using Cu nanoparticles rather than chemically inert Ag nanoparticles. In combination with a recently developed flash annealing process that can be conducted in air on a timescale of ~103, this issue can be resolved, but an in-depth study has not been conducted for Cu nanoparticle-based, flash-annealed electrodes that can suppress the significant migration into oxide semiconductors. The evolution of the electronic structure in the region of the conduction band, related to the incorporation of Ag, was examined by means of X-ray absorption spectroscopy (XAS) in beamline 2A of the Pohang Acceleration Laboratory (PAL). In order to investigate the change of the electronic structure as a function of the degree of Ag incorporation, the XAS spectra near the O K1 edge were compared to those of IZO films, as shown in Figure 4. Normalization of the XAS spectra was carefully performed by subtracting the X-ray beam background and subsequently scaling the pre- and post-edge levels, which can be used to compare the qualitative changes of the conduction band.27 The O K1 edge XAS spectra of the IZO film samples are directly related to the O p-projected states of the conduction band, 20


Page 20 of 34

Page 21 of 34

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


which consists of unoccupied hybridized orbitals for In 5sp, Zn 4sp, and O 2p from 530 eV to 550 eV.28 The XAS spectra of the Ag NP-paste based electrode annealed at 200 oC and the Ag F-paste #2 based electrode are slightly modified by the subtle incorporation of Ag chemical states, by virtue of the presence of Ag2O with the hybridization of Ag 5sp + O 2p.29 Notable findings were recorded in the spectra of the Ag NP-paste based electrode annealed at 300 oC and the Ag F-paste #1 based electrode. The electronic structure related to the hybridization of Ag and O is strongly enhanced by the increase in the absolute amount of the incorporated Ag2O phase. Similar to the XPS results, the electronic structure of Ag2O was mainly represented in the electrode prepared from Ag F-paste #1. 3.4. Versatility of Source/Drain Electrodes Printed from the Pastes Comprising OleicAcid Capped Ag Nanoparticles. To demonstrate the versatility of oleic-acid-capped Ag nanoparticles, we investigated the device performance capabilities of transistors employing various printed electrode/semiconductor stacks by synthesizing indium oxide (In2O3), indium gallium zinc oxide (IGZO), and zinc tin oxide (ZTO) precursor solutions. Both IGZO and ZTO precursor solutions were synthesized via conventional metal-salt-based sol-gel chemistry,30 and the In2O3 precursor solution was synthesized via combustion chemistry accompanied with local internal heat generation to lower the annealing temperature.31 Both the IGZO and ZTO channel layers were annealed at 400 oC, and the In2O3 channel layer was heat-treated at 250 oC. For all devices, the Ag NP-paste-based electrodes were annealed at 200 oC. As shown in Figure S8 and Table S1, for all of the different types of compositional oxide semiconductors, there is no noticeable difference in the field-effect mobility of the transistors with vacuum-deposited Al and printed Ag source/drain electrodes (0.9 and 1.0 cm2/V·s for the In2O3 TFTs with the Al and Ag electrodes, respectively; 2.1 and 2.6 cm2/V·s 21



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

for the IGZO TFTs with the Al and Ag electrodes, respectively; 5.2 and 5.5 cm2/V·s for the ZTO TFTs with the Al and Ag electrodes, respectively). This indicates clearly that the oleicacid-capped surface passivation of Ag nanoparticles is valid universally for compositional oxides that are composed of different constituent elements and prepared via different synthetic pathways.

4. CONCLUSION In summary, we have demonstrated that oleic-acid-capped Ag nanoparticles can be readily formulated into printable pastes to form printed Ag source/drain electrodes compatible with versatile oxide semiconductors, allowing for high-performance, oxide thin-film transistors by suppressing the undesirable migration of Ag inside adjacent oxide semiconductors. It was revealed that by virtue of the capability of capturing surficial Ag atoms along Ag nanoparticles, the migration of Ag was effectively prevented and a subtle amount of Ag residing inside the IZO oxide semiconductors partially preserves the inherent metallic phase. Based on comparative studies using printed electrodes prepared from commercially available and thermoplastic triblock copolymer-incorporated Ag flake pastes, it was suggested that the prevention of Ag migration, the suppressed formation of the Ag2O phase, and the reduced incorporation of organic components inside the electrodes all play critical roles in facilitating the realization of high-performance, oxide thin-film transistors employing printed Ag source/drain electrodes.

22


Page 22 of 34

Page 23 of 34

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 1. Summary of device performance parameters for IZO transistors employing thermally evaporated Al and various printed Ag source/drain electrodes Semicon ductor

Electrode Al 1

IZO

Deposition method evaporation

Ag NP-paste

Ag F-Paste #1 Ag F-Paste #2 1

oleic acid-capped Ag nanoparticle

2

threshold voltage

3

subthreshold swing

printing

Annealing temp. (oC) 200 300 80 80

Mobility (cm2V-1s-1) 9.0 7.9 0.05 5.3

23


2

Vth (V) 11.6 24.6 22.6 11.4

3

S.S (V/dec) 1.24 1.3 2.24 1.1


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

Scheme 1. Schematics showing the dispensing printing process of various Ag nanoparticle/flake pastes on top of a semiconducting layer. Three types of pastes were tested in this study. The inset shows an optical microscope image of the channel area with printed source/drain electrodes.

24


Page 24 of 34

110 10

b

Ag nanoparticle Ag nanoparticle Paste commercial Ag flake paste

100

10

Drain Current (A)

a Weight Loss (%)

90

10

10

- evaporated Al S/D electrodes - mobility: 9.0 cm2V-1s-1 -3

c

- Ag NP-paste (200 oC annealed) - mobility: 7.9 cm2V-1s-1 10

-4

10

-6

10

-8

-5

-7

-9

-10

10

80 200

300

400

500

-11

10

o

Temperature ( C)

10

10

10

20

30

-2

e

-4

10

-4

10

-6

10

-8

-10

0

10

20

30

40

Gate Voltage (V)

10

-10

0

10

20

30

40

- Ag F-paste #2 - mobility: 5.3 cm2V-1s-1 10

-4

10

-6

10

-8

-10

10

-12

-12

-6

-10

Gate Voltage (V)

f

10

-10

10

40

- Ag F-paste #1 - mobility: 0.05 cm2V-1s-1

Drain Current (A)

10

0

Gate Voltage (V)

- Ag NP-paste (300 oC annealed) - mobility: not semicoducting

d

-12

-10

Drain Current (A)

100

Drain 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


Drain Current (A)

Page 25 of 34

0

10

20

30

40

Gate Voltage (V)

10

-10

0

10

20

30

40

Gate Voltage (V)

Figure 1. (a) TGA results for the Ag nanoparticle, Ag NP-paste and Ag F-paste #1 samples. Both pastes were analyzed with dried powders obtained after a drying process. Transfer characteristics of IZO transistors employing (b) thermally evaporated Al, (c) Ag NP-paste based, 200 oC-annealed, (d) Ag NP-paste based, 300 oC-annealed, (e) Ag F-paste #1 based and (f) Ag F-paste #2 based source/drain electrodes. Both Ag flake paste-based electrodes were dried at 80 oC after the completion of the printing process. Drain bias of 40 V was applied to all transistors.

25



b

Ag NP-paste

40 Ag F-paste #1

30

C-C

Intensity (a.u.)

a Composition (at%)

20

carboxylate

10 Ag F-paste #2

282

0 300

c

d

C-C

C with O carboxylate

unknown 282

285

285

288

291

Binding Energy (eV)

oC

288

12

Atomic Fraction (%)

200

oC

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 26 of 34

10 8 6 4 2

0 as-synthesized 200 oC-annealed

291

Binding Energy (eV) Figure 2. (a) Compositional variations of Ag element inside IZO semiconductors stacked with Ag NP-paste based (200 or 300 oC-annealed), Ag F-paste #1 based and Ag F-paste #2 based electrode layers. Both Ag flake paste-based electrodes were dried at 80 oC after the completion of the printing process. Prior to the XPS analysis, the upper electrode layers were eliminated by a washing process with solvents; XPS C 1s spectra for (b) as-synthesized Ag nanoparticles and (c) 200 oC-annealed Ag nanoparticles; (d) atomic fractions of carboxylate in the C 1s spectra for the as-synthesized and 200 oC-annealed Ag nanoparticles.

26


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


Normarlized Intensity (Arb.Units)

Page 27 of 34

o

o

a

b

Ag NP-paste (200 C) Ag NP-paste (300 C) Ag F-paste #1

Ag2O

c

Ag

370

368

366 370

368

366 370

368

366

Binding Energy (eV) Figure 3. XPS Ag 3d5/2 spectra of IZO semiconductors stacked with (a) Ag NP-paste based, 200 oC-annealed, (b) Ag NP-paste based, 300 oC-annealed and (c) Ag F-paste #1 based electrode layers. The Ag flake paste-based electrode was dried at 80 oC after the completion of the printing process. Prior to the XPS analysis, the upper electrode layers were eliminated by a washing process with solvents.

27


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

T.E.Y Intensity (Arb.Units)


In 5s + O 2p Ag 5s + O 2p Zn 4s + O 2p

Page 28 of 34

In 5sp + O 2p Zn 4sp + O 2p Ag 5sp + O 2p

Ag F-paste #2 Ag F-paste #1 o

Ag NP-paste(300 C) o

Ag NP-paste(200 C) IZO

530

535

540

545

550

555

Photon Energy (eV) Figure 4. O K1 edge XAS spectra IZO semiconductors stacked with electrode layers based on Ag NP-paste (200 oC-annealed), Ag NP-paste (300 oC-annealed), Ag F-paste #1, and Ag F-paste #2. Both Ag flake paste-based electrodes were dried at 80 oC after the completion of the printing process. Prior to the XAS analysis, the upper electrode layers were eliminated by a washing process with solvents.

28


Page 29 of 34

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


ASSOCIATED CONTENT Supporting Information. Experimental methods, a summary of device performance parameters for various solution-processed oxide transistors with evaporated Al and printed Ag source/drain electrodes, the rheological properties of all pastes, optical microscope images of printed patterns, current-voltage measurement results for printed electrodes, XPS N 1s spectrum of as-synthesized Ag nanoparticles, current density vs. electric field for MIM devices, and transfer characteristics for various solution-processed oxide transistors with evaporated Al and printed Ag source/drain electrodes.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected] (K.-B. Chung) *E-mail: [email protected] (Y. Choi) *E-mail: [email protected] (S. Jeong) Author Contributions S. Jeong and Y. Choi designed the experiment, supervised all phases of the project, and edited the manuscript. G. R. Hong synthesized the precursor solutions, fabricated the devices, and measured the electrical performance of the devices. H. J. Park, Y. Jo, J. Y. Kim and H. S. Lee formulated the printable conductive pastes. S. S. Lee, B.-H. Ryu and Y. C. Kang analyzed the XPS spectra. A. Song and K. B. Chung analyzed the XAS spectra. G. R. Hong and S. Jeong wrote the manuscript. All authors proofed the manuscript. 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

ACKNOWLEDGMENT This research was supported by the Global Research Laboratory Program of the National Research Foundation (NRF) funded by Ministry of Science, Information and Communication Technologies and Future Planning (NRF-2015K1A1A2029679), and partially supported by the Nano·Material Technology Development Program through the National Research Foundation of Korea funded by the Ministry of Science, Information and Communication Technologies and Future Planning (NRF-2015M3A7B4050306).

30


Page 30 of 34

Page 31 of 34

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


REFERENCES [1] Yu, X.; Marks, T. J.; Facchetti, A. Metal Oxides for Optoelectronic Applications. Nat.

Mater. 2016, 15, 383-396 [2] Huang, W.; Zeng, L.; Yu, X.; Guo, P.; Wang, B.; Ma, Q.; Chang, R. P. H.; Yu, J.; Bedzyk, M. J.; Marks, T. J.; Facchetti, A. Metal Oxide Transistors via Polyethylenimine Doping of the Channel Layer: Interplay of Doping, Microstructure, and Charge Transport. Adv.

Funct. Mater. 2016, 26, 6179-6187 [3] Jeong, S.; Moon, J. Low-Temperature, Solution-Processed Metal Oxide Thin Film Transistors. J. Mater. Chem. 2012, 22, 1243-1250 [4] Kim, Y.-H.; Heo, J.-S.; Kim, T.-H.; Park, S.; Yoon, M.-H.; Kim, J.; Oh, M. S.; Yi, G.-R.; Noh, Y.-Y.; Park, S. K. Flexible Metal-Oxide Devices Made by Room-Temperature Photochemical Activation of Sol-Gel Films. Nature 2012, 489, 128-132 [5] Kamyshny, A.; Magdassi, S. Conductive Nanomaterials for Printed Electronics. Small 2014, 10, 3515–3535 [6] Black, K.; Singh, J.; Mehta, D.; Sung, S.; Sutcliffe, C. J.; Chalker, P. R. Silver Ink Formulations for Sinterfree Printing of Conductive Films. Sci. Rep. 2016, 6, 20814 [7] Kim, S. H.; Hong, K.; Xie, W.; Lee, K. H.; Zhang, S.; Lodge, T. P.; Frisbie, C. D. Electrolyte-Gated Transistors for Organic and Printed Electronics. Adv. Mater. 2013, 25, 1822–1846 [8] Sekitani, T.; Noguchi, Y.; Zschieschang, U.; Klauk, H.; Someya, T. Organic Transistors Manufactured Using Inkjet Technology with Subfemtoliter Accuracy. Proc. Natl. Acad.

Sci. 2008, 105, 4976–4980 [9] Kim, D.; Jeong, S.; Shin, H.; Xia, Y,; Moon, J. Heterogeneous Interfacial Properties of Ink-Jet-Printed Silver Nanoparticulate Electrode and Organic Semiconductor. Adv. Mater. 2008, 20, 3084–3089 [10] Jeong, S.; Woo, K.; Kim, D.; Lim, S.; Kim, J. S.; Xia, Y,; Moon, J. Controlling the Thickness of the Surface Oxide Layer on Cu Nanoparticles for the Fabrication of Conductive Structures by Ink-Jet Printing. Adv. Funct. Mater. 2008, 18, 679–686 [11] Zuo, R.; Li, L.; Gui, Z. Influence of Silver Migration on Dielectric Properties and Reliability of Relaxor Based MLCCs. Ceram. Int. 2000, 26, 673-676 [12] Ando, E.; Miyazaki, M. Moisture Degradation Mechanism of Silver-Based Low31



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 32 of 34

Emissivity Coatings. Thin Solid Films 1999, 351, 308-312 [13] Secor, E. B.; Smith, J.; Marks, T. J.; Hersam, M. C. High-Performance Inkjet-Printed Indium-Gallium-Zinc-Oxide Transistors Enabled by Embedded, Chemically Stable Graphene Electrodes. ACS Appl. Mater. Interfaces 2016, 8, 17428−17434 [14] Lee, C.-K.; Park, S. Y.; Jung, H. Y.; Lee, C.-K.; Son, B.-G.; Kim, H. J. Lee, Y.-J.; Joo, Y.-C.; Jeong, J. K. High Performance Zn–Sn–O Thin Film Transistors with Cu Source/Drain Electrode. Phys. Status Solidi RRL 2013, 7, 196-198 [15] Kim, W.-S.; Moon, Y.-K.; Lee, S.; Kang, B.-W.; Kwon, T.-S.; Kim, K.-T.; Park, J.-W. Copper Source/Drain Electrode Contact Resistance Effects in Amorphous Indium– Gallium–Zinc-Oxide Thin Film Transistors. Phys. Status Solidi RRL 2009, 3, 239-241 [16] Jo, Y.; Oh, S.-J.; Lee, S. S.; Seo, Y.-H.; Ryu, B.-H.; Moon, J.; Choi, Y.; Jeong, S. Extremely Flexible, Printable Ag Conductive Features on PET and Paper Substrates via Continuous Millisecond Photonic Sintering in a Large Area. J. Mater. Chem. C 2014, 2, 9746-9753 [17] Jo, Y.; Kim, J. Y.; Kim, S.-Y.; Seo, Y.-H.; Jang, K.-S.; Lee, S. Y.; Jung, S.; Ryu, B.-H,; Kim, H.-S.; Park, J.-U.; Choi. Y.; Jeong, S. 3D-printable, highly conductive hybrid composites employing chemically-reinforced,

complex dimensional fillers

and

thermoplastic triblock copolymers. Nanoscale 2016, DOI: 10.1039/C6NR09610G. [18] Park, M.; Park, J.; Jeong, U. Design of Conductive Composite Elastomers for Stretchable Electronics. Nano Today 2014, 9, 244-260 [19] Jeong, S.; Song, H. C.; Lee, W. W.; Choi, Y.; Lee, S. S.; Ryu, B.-H. Combined Role of Well-Dispersed Aqueous Ag Ink and the Molecular Adhesive Layer in Inkjet Printing the Narrow and Highly Conductive Ag Features on a Glass Substrate. J. Phys. Chem. C 2010, 114, 22277–22283 [20] Oh, S.-J.; Jo, Y.; Lee, E. J.; Lee, S. S.; Kang, Y. H.; Jeon, H.-J.; Cho, S. Y.; Park, J.-S.; Seo, Y.-H.; Ryu, B.-H.; Choi, Y.; Jeong, S. Ambient Atmosphere-Processable, Printable Cu Electrodes for Flexible Device Applications: Structural Welding on a Millisecond Timescale of Surface Oxide-Free Cu Nanoparticles. Nanoscale 2015, 7, 3997-4004 [21] Jeong, S.; Lee, S. H.; Jo, Y.; Lee, S. S.; Seo, Y.-H.; Ahn, B. W.; Kim, G.; Jang, G.-E.; Park, J.-U.; Ryu, B.-H.; Choi, Y. Air-Stable, Surface-Oxide Free Cu Nanoparticles for Highly Conductive Cu Ink and Their Application to Printed Graphene Transistors. J. 32


Page 33 of 34

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


Mater. Chem. C 2013, 1, 2704-2710 [22] Weaver, J. F.; Hoflund, G. B. Surface Characterization Study of the Thermal Decomposition of AgO. J. Phys. Chem. 1994, 98, 8519-8524 [23] Son, G.-C.; Chee, S.-S.; Jun J.-H.; Son, M.; Lee, S. S.; Choi, Y.; Jeong, S.; Ham, M.-H. Chemically Functionalized, Well-Dispersed Carbon Nanotubes in Lithium-Doped Zinc Oxide for Low-Cost, High-Performance Thin-Film Transistors. Small 2016, 12, 18591865 [24] Dai, M.-K.; Lian, J.-T.; Lin, T.-Y.; Chen, Y.-F. High-Performance Transparent and Flexible Inorganic Thin Film Transistors: a Facile Integration of Graphene Nanosheets and Amorphous InGaZnO. J. Mater. Chem. C 2013, 1, 5064-5071 [25] Liu, X.; Wang, C.; Cai, B.; Xiao, X.; Guo, S.; Fan, Z.; Li, J.; Duan, X.; Liao, L. Rational Design of Amorphous Indium Zinc Oxide/Carbon Nanotube Hybrid Film for Unique Performance Transistors. Nano Lett. 2012, 12, 3596-3601 [26] Han, Y. H.; Won, J.-Y.; Yoo, H.-S; Kim, J.-H.; Choi, R.; Jeong, J. K. High Performance Metal Oxide Field-Effect Transistors with a Reverse Offset Printed Cu Source/Drain Electrode. ACS Appl. Mater. Interfaces 2016, 8, 1156−1163 [27] Chung, K. B.; Long, J. P.; Seo, H.; Lucovsky, G.; Nordlund, D. Thermal Evolution and Electrical Correlation of Defect States in Hf-Based High-K Dielectrics on N-Type Ge (100): Local Atomic Bonding Symmetry. J. Appl. Phys. 2009, 106, 074102 [28] Park, H.-W.; Song, A.; Kwon, S.; Ahn, B. D.; Chung, K.-B. Improvement of Device Performance and Instability of Tungsten-Doped InZnO Thin-Film Transistor with Respect to Doping Concentration. Appl. Phys. Exp. 2016, 9, 111101 [29] Bukhtiyarov, V. I.; Hävecker, M.; Kaichev, V. V.; Knop-Gericke, A.; Mayer, R. W.; Schlögl, R. Atomic Oxygen Species on Silver: Photoelectron Spectroscopy and X-Ray Absorption Studies. Phys. Rev. B 2003, 67, 235422 [30] Kim, S. J.; Kim, A.; Jo, Y.; Yoon, J.-Y.; Lee, S. S.; Choi, Y.; Won, J.; Nahm, S.; Jang, K.-S.; Kim, Y. H.; Jeong, S. Polymeric Mold Soft-Patterned Metal Oxide Field-Effect Transistors: Critical Factors Determining Device Performance. J. Mater. Chem. C 2014, 2, 8486-8491 [31] Kang, Y. H.; Jeong, S.; Ko, J. M.; Lee, J.-Y.; Choi, Y.; Lee, C.; Cho, S. Y. TwoComponent Solution Processing of Oxide Semiconductors for Thin-Film Transistors via 33



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

Self-Combustion Reaction. J. Mater. Chem. C 2014, 2, 4247-4256 Table of Contents (TOC)

34


Page 34 of 34

Drain

Recommend Documents