Flexible electroluminescent device with inkjet-printed carbon nanotube electrodes This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2012 Nanotechnology 23 344003 (http://iopscience.iop.org/0957-4484/23/34/344003) Download details: IP Address: 136.159.235.223 The article was downloaded on 25/04/2013 at 10:02 Please note that terms and conditions apply. View the table of contents for this issue, or go to the journal homepage for more Home Search Collections Journals About Contact us My IOPscience
Transcript
Flexible electroluminescent device with inkjet-printed carbon nanotube electrodes
This article has been downloaded from IOPscience. Please scroll down to see the full text article.
Casali Institute of Applied Chemistry, Institute of Chemistry, The Center for Nanoscience andNanotechnology, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel
Received 2 January 2012Published 10 August 2012Online at stacks.iop.org/Nano/23/344003
AbstractCarbon nanotube (CNTs) inks may provide an effective route for producing flexible electronicdevices by digital printing. In this paper we report on the formulation of highly concentratedaqueous CNT inks and demonstrate the fabrication of flexible electroluminescent (EL) devicesby inkjet printing combined with wet coating. We also report, for the first time, on theformation of flexible EL devices in which all the electrodes are formed by inkjet printing oflow-cost multi-walled carbon nanotubes (MWCNTs). Several flexible EL devices werefabricated by using different materials for the production of back and counter electrodes:ITO/MWCNT and MWCNT/MWCNT. Transparent electrodes were obtained either bycoating a thin layer of the CNTs or by inkjet printing a grid which is composed of empty cellssurrounded by MWCNTs. It was found that the conductivity and transparency of theelectrodes are mainly controlled by the MWCNT film thickness, and that the dominant factorin the luminance intensity is the transparency of the electrode.
(Some figures may appear in colour only in the online journal)
1. Introduction
The unique properties of carbon nanotubes (CNTs), suchas conductivity and good mechanical strength, make theman attractive material for nanoelectronics and optoelectronicsapplications [1]. Unlike Indium tin oxide (ITO), which maycrack with bending [2], CNTs can provide an effective optionfor producing flexible electronic devices.
Recently, CNTs were employed as a transparent electrodefor a flexible electroluminescence (EL) device. Kim et al [3]demonstrated a device in which a conventional transparentITO-coated glass was replaced by a CNT electrode. Thiselectrode was prepared by filtering a dispersion of single-walled carbon nanotubes (SWCNT) and by transferring thecollected CNTs to a PET substrate. Schrage et al [4] reportedon an EL device composed of back and counter electrodesprepared by spray coating of SWCNT dispersions. Bothreports show that the transparency and sheet resistance of thefilms have a significant effect on the brightness of the resultingEL device.
In these reports, due to the selected deposition methods,the resulting EL devices were composed of a continuousluminescent film.
However, there is a current need for EL devices withpatterned electrodes. Such patterning can be achieved byinkjet printing. Inkjet printing of a CNT dispersion hasbeen reported in a small number of publications as analternative method for generating electrically conductivepatterns on substrates, such as plastics and paper [5–16], andfor obtaining thin film transistors [5, 9, 15, 16] and fieldemitter devices [14].
The aim of the present report is to formulate inks whichare composed of aqueous dispersions of multi-walled carbonnanotubes (MWCNT) and their utilization as electrodes forthe fabrication of flexible electroluminescent devices. Inparticular, we report for the first time on the formation of anEL device, in which all the electrodes are formed by inkjetprinting of multi-walled carbon nanotubes. In this device,the counter electrode is inkjet printed and the transparentelectrode is an inkjet-printed grid that is composed of emptycells surrounded by lines of MWCNTs.
Three types of EL devices were prepared in this work,differing in the type of back electrode: one, an ITO/MWCNTdevice (type A) and, two, an all-MWCNT device (types Band C) (figure 1). In device type A the back electrode wasa commercial ITO-coated polyethylene terephthalate (PET)substrate. In device type B, the back electrodes were madeof MWCNTs prepared by coating PET with Ink B. Thecoating was performed by drawdown (K Control Coater, RK Print-Coat Instruments Ltd, UK) with bar coatings thatgive a wet film thickness of 6, 12, 24, 40, 50 and 80 µm.The coatings were performed at 3 m min−1 and the filmswere dried at 90 ◦C. In device type C, the back electrodewas prepared by inkjet printing Ink A on a PET subtract,in a grid pattern composed of cells having dimensions of300 µm× 800 µm, with a linewidth of 100 µm. The printingwas performed with a Microfab JetDrive III inkjet printerwith a 60 µm diameter single nozzle. The applied waveformwas 100 V; rise time 3 µs; echo time 80 µs; dwell time40 µs and fall time 5 µs, while printing at 50 Hz. Themovement of the substrate was controlled by a DMC-21x3XY table (Galil motion control). The substrate temperaturewas set to 60 ◦C with a Peltier heater/cooler and the humiditywithin the printing chamber was 38% RH during the printingexperiments.
In all the devices, the electroluminescent (EL) layer wasprepared by silk printing an EL paste composed of ZnS and adielectric paste of BaTiO3 (MOBIChem, Israel) on top of thetransparent electrode. The counter electrodes were preparedby inkjet printing Ink A (in several layers) on top of the ELlayer.
Figure 1. Three types of devices prepared: (1) ITO/MWCNT(device type A) and (2) all-MWCNT (device types B and C).
2.3. Device characterization and instrumentation
The light transmittance of the MWCNT electrodes was mea-sured at 600 nm, using a Cary 100 UV–vis spectrophotometer(Varian, USA). The sheet resistance was measured by aCascadeTM Microtech four-point probe (Beaverton, USA)and the resistivity was measured by a two-point probe coupledto an Extech milliohm meter (model 380562, Waltham, USA).SEM observations were performed using the Extra HighResolution SEM Magellan 400 and the Environmental SEMQuanta 200. The luminance of each device was measuredby using the Tricor photometer (model 820, Tricor SystemsInc., USA) with the Eyeppearance 4.2 software and byimage analysis of optical microscope images (model SQF-II,Catic Fujian, China) with Matlab software. The luminancewas measured under an applied voltage of 200 V at afrequency of 1 kHz, using the AC power supply Compactix Series (California Instruments, USA). The flexibility ofthe electrodes and devices was evaluated by measuring theresistivity and device performance changes after bendingthem several times to an angle of 180◦. The bendingexperiments were performed using a Single Column TestingSystems for Low-force Testing (Instron, model 3345, USA).
3. Results
3.1. MWCNT inks
Two highly concentrated MWCNT aqueous inks wereprepared using two different types of MWCNTs. Ink Acontained short MWCNTs with a viscosity of 1.89 cP andsurface tension of 30.0±0.4 mN m−1, which made it suitablefor inkjet printing. Ink B contained long MWCNTs andwas suitable for wet coatings, with a viscosity of 3.24 cPand surface tension of 24.6 ± 0.2 mN m−1. The resistivitiescalculated for lines (with known dimensions) made of thedried ink was 2.03 ± 0.01 × 10−3 � m and 1.90 ± 0.2 ×10−3 � m for Ink A and Ink B, respectively.
2
Nanotechnology 23 (2012) 344003 S Azoubel et al
Figure 2. Counter electrode for an EL device prepared inkjet printing MWCNT dispersion. The vertical lines: 12 printed layers; horizontallines: 2–10 printed layers. (A) The front side of the counter electrode and (B) the back side of the device after the voltage is applied.
3.2. Device type A
This EL device is composed of ITO as the back electrode andinkjet-printed MWCNTs as the counter electrodes. Duringthe first printing experiments, lines with various number oflayers (2–12 layers) were printed (figure 2) in order to checkthe performance of the device. It should be noted that, inorder to achieve luminescence, a satisfactory percolation ofthe MWCNTs should be accomplished. As reported earlier,the number of inkjet-printed layers affects the electrical sheetresistance of the CNT pattern (6, 12 and 13). The results showthat, in order to achieve the required MWCNT percolation thatwill provide the EL effect, each line needs to be printed in atleast six layers.
When a liquid droplet containing solid particles isdeposited in a substrate and the liquid is dried, the particles arepushed to the margin of the droplet and form a ring known asthe ‘coffee ring effect’, as reported by Deegan et al [17]. Thiseffect is also observed when printing the MWCNT ink overthe EL layer. As shown in figure 3(A), each ink droplet formsa ring, where the MWCNTs are found mainly in the margin.When printing several layers, the formed rings interpolate,forming a continuous MWCNT dense line (figures 3(B) and(C)). The dense ring margin and percolation of the nanotubesare shown in figure 3(D).
After the preliminary printing experiments, a comparisonof the intensities of the EL of devices with different numbersof printed layers (from 6 to 14) was made, measuring theluminance of each line after applying a voltage of 200 V(figure 4). In the line of six printed layers the luminanceintensity was 133.2±4.8 cd m−2, increasing with the increasein the number of printed layers, reaching 188.8± 4.2 cd m−2
for 10 printed layers. Above that, at 12 and 14 printedlayers, the luminance decreased to 171.7 ± 3.8 and 166.0 ±3.1 cd m−2, respectively.
3.3. Device type B
This flexible EL device is composed of inkjet-printedMWCNTs as the counter electrodes and MWCNT-deposited
film (by drawdown of ink B) replacing the transparentITO back electrode. The performance of the devices ismainly determined by the resistance and transmittance of theelectrodes. Films with a low sheet resistance (R) and a highlight transmittance (T) should yield the highest luminance. Inan attempt to find the best performing devices, several filmswere prepared with various wet film thicknesses (6–80 µm).It was found that, as the wet thickness of the film increases,the sheet resistance decreases, but as could be expected, thisis accompanied by a decrease in transmittance (figure 4).The thinnest film (wet thickness of 6 µm) has R = 16.3 ±2.9 k�/� and T = 66.3±0.4%. For the 24 µm wet thicknessthe sheet resistance decreased to 2.8 ± 0.4 k�/� and thetransmittance decreased to 27.5± 5.2%. For the thickest film(wet coating of 80 µm), the values dropped to R = 0.7 ±0.2 k�/� and T = 0.2 ± 0.1%, meaning that the film is nottransparent at all.
Several flexible electroluminescent devices were formedby using MWCNT films with different thicknesses asthe transparent electrode, and an identical inkjet-printedMWCNT counter electrode. The performance of the deviceswas compared by luminance measurements, after applyinga voltage of 200 V. Figure 5(A) shows optical microscopeimages of all the devices and figure 5(B) shows theimage analysis of these images indicating the light emissionintensity. It can also be seen that the first device (figures 5(A-1) and (B-1)), whose electrode has the highest sheet resistance(16.33 ± 2.9 k�/�) and the highest transmittance (66.3 ±0.4%), shows the highest light emission. It might be expectedthat increasing the MWCNT film thickness would lead to adecrease in electrode sheet resistance and, therefore, improvethe device’s performance. However, the increase of theMWCNT concentration in the film decreases the electrode’stransparency, and it seems that, within the studied rangesof resistance, the most dominant factor for such devicesis transparency and not resistance. The overall dependenceof the luminescence intensity on electrode transmittance ispresented in figure 6. Table 1 summarizes the effect of theelectrodes’ characteristics on the light emission intensity ofthe devices.
3
Nanotechnology 23 (2012) 344003 S Azoubel et al
Figure 3. SEM observation of the ring formed by the MWCNT-printed ink (A) and the interpolation of the rings forming a continuous line((B) and (C)). The MWCNT dense percolation is achieved in the ring margin (D).
Figure 4. The effect of wet coating thickness on transmittance (at600 nm) and sheet resistance of the films.
The flexibility was evaluated by bending the MWCNTfilms and the EL devices for 20 times. It was found that theMWCNT electrodes’ resistivities were not affected when bentto an angle of 180◦ and the overall performance of the devicesremains unchanged after 20 bendings to 180◦ (figure 7, seethe video abstract that accompanies the online version of thisarticle).
3.4. Device type C
In this flexible EL device all of the electrodes were obtainedby inkjet printing of Ink A, with the short MWCNTs. Thetransparent electrode was obtained by printing a grid ofCNT lines, in which the spaces between the lines providethe transparency (the linewidth was about 100 µm and thedimensions of the formed empty cells were 300µm×800µm;figure 8(A)). The counter electrode was a single inkjet-printedline composed of 14 layers of MWCNT ink.
The luminance measured for this device (above theMWCNT line of the counter electrode) was 35.2 ± cd m−2
when applying a voltage of 200 V, indicating that the printedgrid can indeed function as a transparent and flexible electrode(figure 8(B)).
4. Conclusions
It was demonstrated that it is possible to form flexible ELdevices in which all the electrodes are obtained by inkjetprinting MWCNTs. In order to obtain a working device byinkjet printing, contacts between the CNTs are required, andwhen using a MWCNT dispersion of 1 wt%, each line shouldbe printed at least six times.
It is expected that increasing the CNT concentration inthe ink would enable a decrease in the number of printedlayers. This may lead to an increase in the viscosity of the
4
Nanotechnology 23 (2012) 344003 S Azoubel et al
Figure 5. Light emission of devices composed by MWCNT transparent electrodes prepared by different wet film thickness (A) denotes themicroscope image; (B) denotes the intensity determined by image analysis of the photos. The film thicknesses are given in table 1.
Table 1. MWCNT electrode characteristics and device luminance intensity.
Figure 6. The effect of the MWCNT electrode transmittance on thedevice luminance.
ink (with the current dispersing agents), beyond that whichis required for inkjet printing. Therefore, further work shouldbe focused on the preparation of CNT dispersions at higher
concentrations, without significantly increasing the ratio ofdispersant to CNT. At present, the dispersant content is 50%of the MWCNTs, which is about 30% of the dry printed film.This, obviously, has a negative effect on conductivity. It wasalso found that the transparent electrode can be obtained eitherby coating a thin layer of the MWCNTs or by printing a gridcomposed of empty cells surrounded by lines of inkjet-printedMWCNTs. Obviously, the transparency and conductivity ofthe films have critical effects on the luminance output, andboth should be maximal. However, increased conductivity isobtained by greater concentration of CNTs in the film, whichhas a negative impact on the transparency.
Acknowledgments
This research is supported by the Singapore NationalResearch Foundation under its CREATE program: Nanoma-terials for Energy and Water Management, and the IsraelMinistry of Trade and Industry under its NES Magnet
5
Nanotechnology 23 (2012) 344003 S Azoubel et al
Figure 7. Flexibility test showing the light emission during bending from 0◦ to 180◦.
Figure 8. MWCNT inkjet-printed grid in PET (A) to be used as the flexible transparent electrode in device type C. (B) Shows the back sideof the device. The luminance of the counter electrode (a single printed line) is achieved after applying voltage to the device.
program. We would also like to thank CPC HI Technologiesand Professor Banin for their assistance in measuring theluminescence.
References
[1] Avouris P and Chen J 2006 Nanotube electronics andoptoelectronics Mater. Today 9 46–54
[2] Kumar A and Zhou C W 2010 The race to replace tin-dopedindium oxide: which material will win? ACS Nano 4 11–4
[3] Kim M J, Shin D W, Kim J Y, Park S H, Han I T andYoo J B 2009 The production of a flexibleelectroluminescent device on polyethylene terephthalatefilms using transparent conducting carbon nanotubeelectrode Carbon 47 3461–5
[4] Schrage C and Kaskel S 2009 Flexible and transparentSWCNT electrodes for alternating currentelectroluminescence devices ACS Appl. Mater. Interfaces1 1640–4
[5] Beecher P et al 2007 Ink-jet printing of carbon nanotube thinfilm transistors J. Appl. Phys. 102 043710
[6] Fan Z J, Wei T, Luo G H and Wei F 2005 Fabrication andcharacterization of multi-walled carbon nanotubes-basedink J. Mater. Sci. 40 5075–7
[7] Kordas K, Mustonen T, Toth G, Jantunen H, Lajunen M,Soldano C, Talapatra S, Kar S, Vajtai R andAjayan P M 2006 Inkjet printing of electrically conductivepatterns of carbon nanotubes Small 2 1021–5
[8] Lok B K, Ng Y M, Liang Y N and Hu X 2010 Inkjet printingof multi-walled carbon nanotube/polymer composite thin
film for interconnection J. Nanosci. Nanotechnol.10 4711–5
[9] Okimoto H, Takenobu T, Yanagi K, Miyata Y, Kataura H,Asano T and Iwasa Y 2009 Ink-jet printing of asingle-walled carbon nanotube thin film transistor Japan. J.Appl. Phys. 48 06FF03
[10] Panhuis M I H, Heurtematte A, Small W R andPaunov V N 2007 Inkjet printed water sensitive transparentfilms from natural gum–carbon nanotube composites SoftMatter 3 840–3
[11] Shigematsu S, Ishida Y, Nakashima N and Asano T 2008Electrostatic inkjet printing of carbon nanotube for coldcathode application Japan. J. Appl. Phys. 47 5109–12
[12] Small W R and Panhuis M I H 2007 Inkjet printing oftransparent, electrically conducting single-waitedcarbon-nanotube composites Small 3 1500–3
[13] Song J W, Kim J, Yoon Y H, Choi B S, Kim J H andHan C S 2008 Inkjet printing of single-walled carbonnanotubes and electrical characterization of the line patternNanotechnology 19 095702
[14] Song J W et al 2009 The production of transparent carbonnanotube field emitters using inkjet printing Physica E41 1513–6
[15] Takenobu T, Miura N, Lu S Y, Okimoto H, Asano T,Shiraishi M and Iwasa Y 2009 Ink-jet printing of carbonnanotube thin-film transistors on flexible plastic substratesAppl. Phys. Express 2 025005
[16] Han X, Janzen D C, Vaillancourt J and Lu X 2007 Printablehigh-speed thin-film transistor on flexible substrate usingcarbon nanotube solution Micro Nano Lett. 2 96–8
[17] Deegan R D, Bakajin O, Dupont T F, Huber G, Nagel S R andWitten T A 1997 Capillary flow as the cause of ring stainsfrom dried liquid drops Nature 389 827–9
