Transparent Electrodes with Nanotubes and Graphene for Printed Optoelectronic Applications

We report here on printed electroluminescent structures containing transparent electrodesmade of carbon nanotubes and graphene nanoplatelets. Screen-printing and spray-coating techniques were employed. Electrodes and structures were examined towards optical parameters using spectrophotometer and irradiation meter. Electromechanical properties of transparent electrodes are exterminated with cyclical bending test. Accelerated aging process was conducted according to EN 62137 standard for reliability tests of electronics. We observed significant negative influence of mechanical bending on sheet resistivity of ITO, while resistivity of nanotube and graphene based electrodes remained stable. Aging process has also negative influence on ITO based structures resulting in delamination of printed layers, while those based on carbon nanomaterials remained intact. We observe negligible changes in irradiation for structures with carbon nanotube electrodes after accelerated aging process. Such materials demonstrate a high application potential in general purpose electroluminescent devices.

We present fabrication of TFEL structures with the use of elaborated transparent electrodes with oligo walled carbon nanotubes (OWCNTs) and graphene nanoplatelets (GNPs).Structures were subjected to series of experiments: transmittance measurements of electrodes and bending tests for evaluation of resistivity stability, irradiance comparison of structures with different colors of luminophores (orange, blue, and blue-green) and different types of electrodes (ITO, OWCNT, and GNP-CNT), and accelerated aging of electroluminescent structures demonstrating influence on mechanical properties of TCLs.

Materials and Preparation
TFEL structures were prepared on diverse transparent electrodes deposited on PET foil.ITO on PET was acquired commercially.Composite layers with CNTs, GNPs, and GNP + CNT were coated on PET foil using spray-coating method.Luminophore, dielectric, and silver electrode layers were screen-printed on top of transparent electrodes.
2.1.Materials.First type of nanomaterial investigated in research was CCVD grown oligowalled carbon nanotubes, commercially available from Cheap Tubes Inc. (USA).Observed Raman spectra (Figure 1(a)) are typical for multiwalled carbon nanotubes with small diameters.Additional scanning electron microscopy (SEM) observations (Figure 1(b)) allowed confirming assumptions that we have carbon nanotubes of small diameter with length around 10-20 m, organized in bundles.
Second type of nanomaterial investigated in research was graphene nanoplatelets from the same supplier.Again, Raman analysis indicates that material consists of few layer graphene structures (Figure 2(a)) which was confirmed by scanning electron microscopy observations (Figure 2(b)).Mean diameter of graphene nanoplatelets was in range of 5 to 15 m.According to supplier GNPs were obtained by chemical method, employing oxidization of graphite with the mixture of H 2 SO 4 , NaNO 3 , and KMnO 4 , known as Hummer's and Offerman method.Obtained graphene oxide was further exfoliated and reduced by hydrazine.
For elaborated structures, luminophore powders from GTP (USA) were acquired: GG45 High Brite-green, GG25 Super Brite-blue-green, and GG13 High Brite-orange, with mean particle size in range of 20-30 m.For dielectric compositions barium titanate BaTiO 3 nanopowder from IAM (Germany) was used.Performed SEM observations and granulometric analysis (Figure 3) showed that the average grain size is 137 nm, though there were also larger grains observed.For metallic electrode commercial silver lacquer L-121 from ITME (Poland) was used.
Final compositions consist of functional filler (CNTs, GNPs, luminophores or dielectric powders) and polymer vehicle.To prepare polymer vehicle poly(methyl methacrylate) (PMMA) M W 350 000 from Sigma-Aldrich was dissolved in diethylene glycol butyl ether acetate with magnetic mixer for 48 hours at 40 ∘ C. Obtained this way 8 wt.% solution of PMMA in solvent was used as vehicle for luminophore and dielectric powders.Spray-coating inks contain 0.02 wt.% solution of PMMA in organic solvent, to achieve desired low viscosity.

Preparation.
To obtain final compositions, carbon nanomaterials and luminophore/dielectric powders were dispersed in corresponding polymer vehicles.
Spray-coating CNT and GNP inks were prepared by mixing carbon fillers and polymer vehicle in an ultrasonic bath for 30 min in 25 ∘ C. Nanotube ink contained 0.31 wt.% OWCNT in vehicle, and GNP ink contained 0.35 wt.% platelets with additional 0.1 wt.% OWCNT in vehicle.
Luminophore powders (60 wt.%) were mixed with the 8 wt.% PMMA vehicle with pestle and mortar followed by homogenization on three-roll mill to break remaining agglomerates.The dielectric paste consisted of 76.3 wt.% BaTiO 3 and additional 0.8 wt.% of MALIALM SC-0505K surfactant from NOF Corp. (Japan).

2.3.
Printing.TFEL structure consists of PET substrate with TCL and three main layers printed on top: luminophore, dielectric, and last metallic electrode.Schematic illustration of such structure is presented in Figure 4.
Transparent electrodes were made using spray-coating method.Airbrush gun with 350 m nozzle diameter was supplied with compressed air (pressure 0,5 MPa) and spraying was done with 100 mm distance between the nozzle and the PET substrate.Screen-printing was made using polyester screen with density 67 T for luminophore, dielectric, and silver layers.Double dielectric layer was printed to ensure that no electric breakdowns in capacitor structure will take place.Spray-coated and screen-printed layers were cured in a chamber dryer in 125 ∘ C for 30 min.

Experimental
Elaborated TFEL structures are supplied with alternating voltage from HV generator, indicated with peak-to-peak values.Main interest in the conducted experiments was focused on comparison of properties of composite carbon nanomaterials electrodes with ITO.Irradiance measurements from all types of structures were conducted with the same supply parameters, measured with Thorlabs PM100D power meter equipped with S121C sensor.Additionally optical properties measurements of all TCLs were taken in 400-900 nm spectrum on Lambda 40 spectrophotometer from Perkin-Elmer.Sheet resistivity measurements were done with Keithley 2636A dual-channel source measure unit and fourprobe method.Mechanical bending tests were performed on specially adapted laboratory stand, with 350 cycles per minute and bending to 0,14 mm −1 curvature.Bent sample was attached from one side to a static fixture and on the other side to a movable grip which caused a warp around symmetry axis.Accelerated aging process was conducted according to EN 62137 standard for reliability tests of electronics in Heraeus HT 7012 S2 thermal shock chamber.The thermal cycle was set to 125 ∘ C for upper and −40 ∘ C for lower temperatures, respectively, and cycled equally in 1-hour period with 5-second temperature shift.Tests were conducted without additional humidification or addition of aggressive atmospheres.

Results and Discussion
Printed electroluminescent structures were fully functional and were emitting a color light according to luminophore type.Structures with TCLs filled only with GNPs were not functional, which is related to high sheet resistivity of that TCL (around 1,5 MΩ/sq).This is significantly higher value than reported form previous experiments [42][43][44].Reported  experiment with GNP and CNT layers deposited without polymer binder or fired at elevated temperature.From that perspective it is possible to use such layers in encapsulated systems, such as capacitors or rigid solar cells, but they do not meet requirements for elastic electronic applications.Although GNP layers exhibited extremely high resistivity, small addition of OWCNTs (0.1 wt.%) allowed lowering resistivity by more than order of magnitude, resulting in functional structures.Picture of working electroluminescent structures is presented in Figure 5.
Thickness of obtained TCL layers was controlled during coating process and remained on the level of 500 nm.Comparable thickness of all layers allows direct comparison of their properties.Final TFEL structure with screen-printed luminophore, dielectric, and silver layers was 35 m thick.

Irradiance Measurements. Measurements of irradiance
flux from all types of structures revealed significant dissimilarities (Figure 6).Emission form luminophores varies depending on different colors.While highest obtained irradiance for blue and blue-green is around 100 W/cm 2 , for orange luminophore it stays below 30 W/cm 2 .Moreover, elaborated elements with nanotube based TCLs have at most 60% of irradiance compared with ITO based structures.This is related to lower optical transmittance of composite layers, what we confirmed in the next experiment.

Transmittance Measurements. Transparent electrodes
were measured with spectrophotometer to draw optical transmittance spectrum.As we can see in Figure 7, ITO has higher values or transmittance, but on uneven level, while CNT based electrodes have linear relation, with minor decay in lower wavelengths.Interestingly graphene based layers  have wavelength independent value of transmittance, which is also known for monolayer graphene [45] and not for any other transparent electrodes.

Electromechanical
Properties.Periodical mechanical stress test was performed on all transparent electrodes.Direct comparison of results is presented in Figure 8. Mechanical stress has relatively low influence on resistance of printed composite electrodes, while ITO samples did not survive the test.Sheet resistivity of CNT and GNP based electrodes remained almost unchanged after 75 000 bending cycles, while sheet resistivity of ITO layer increased more than 10 times just after 20 bends, and shortly after that noted higher resistivity than GNP + CNT layers High mechanical endurance of composite electrodes emerges from carbon nanomaterials, strengthening structure of the layers.Applied mechanical stress does not have any effect on structure of CNTs or GNPs, and additionally very high length-to-diameter ratio of nanotubes and high area of graphene platelets prevent losing electrical contact between separate particles.Sputtered ITO with domain-based polycrystalline structure is less resilient to this type of stresses and is easily destroyed.Detailed observations of microstructure revealed cracked ITO structure, while there were no visible defects in layers with nanomaterials (Figure 9).4.4.Accelerated Aging.We have not observed influence of the accelerated aging process on the irradiance, but we observed degradation of mechanical properties in ITO based structures.After 200 thermal cycles a delamination of luminophore layer from ITO was observed, for all kinds of used luminophores.Adhesion loss resulting in delamination is negative phenomena in case of elastic electronic structures that we have demonstrated in Figure 10.Structures with carbon based layers (OWCNT, GNP + OWCNT) did not exhibit such problems.

Conclusions
Electroluminescent structures with carbon nanotubes and graphene nanoplatelets based transparent electrodes were obtained using screen-printing and spray-coating techniques.Although elaborated structures with nanomaterials have lower luminance than typically ITO based structures, they exhibit much higher resilience to mechanical factors and have stable parameters after aging process.Such composite layers sustain heavy mechanical stress or environmental factors and in that field overcome significantly ITO electrodes.While electrical and optical properties are still superior for ITO layers, optimization processes are on the way.Selection of nanotubes with higher length-to-diameter aspect ratio and thinner graphene platelets will improve optical properties of TCLs.Addition of dispersing agents for elaborated composite layers is goal to improve electrical properties.The results presented in this paper demonstrate that printed TCLs with carbon nanomaterials might be good substitute for ITO layers in elastic optoelectronic application.Presented results are very promising, but we need to improve electrical and optical properties of these new electronic materials.

Figure 5 :Figure 6 :
Figure 5: Printed electroluminescent structures with three types of TCL (from top): ITO, CNTs, and GNPs + CNTs.Colors (from left): orange, green, and blue-green.Inaccurate color mapping and intensities are related to camera technical restrictions.

Figure 7 :Figure 8 :
Figure 7: The relationship between optical transmittance value of TCLs and wavelength.

Figure 9 :
Figure 9: SEM micrographs of ITO and OWCNT layers after cyclical bending tests.