Structural and Optical Properties of Spray Coated Carbon Hybrid Materials Applied to Transparent and Flexible Electrodes

Transparent and flexible electrodes were fabricated with cost-effective spray coating technique on polyethylene terephthalate foil substrates. Particularly designed paint compositions containedmixtures ofmultiwalled carbon nanotubes and graphene platelets to achieve their desired rheology and electrooptical layers parameters. Electrodes were prepared in standard technological conditions without the need of clean rooms or high temperature processing.The sheet resistance and optical transmittance of fabricated layers were tuned with the number of coatings; then the most suitable relation of these parameters was designated through the figure of merit. Optical measurements were performed in the range of wavelengths from 250 to 2500 nm with a spectrophotometer with the integration sphere. Spectral dependence of total and diffusive optical transmission for thin films with graphene platelet covered by multiwalled carbon nanotubes was designated which allowed determining the relative absorbance. Layer parameters such as thickness, refractive index, energy gap, and effective reflectance coefficient show the correlation of electrooptical properties with the technological conditions. Moreover the structural properties of fabricated layers were examined by means of the X-ray diffraction.


Introduction
Transparent electrodes are used in a variety of applications such as dye-sensitized solar cells [1,2], antennas [3,4], electrodes in displays [5,6], supercapacitors [7,8], or heating elements [9].In most cases transparent electrodes are fabricated with physical and chemical vapour deposition methods (PVD and VCD), which can be used for the deposition of conductive thin metal oxide layers like indium tin oxide (ITO) [10], (AZO) [11], (FTO) [12], (GZO) [13], or (ZnO) [14].Vacuum techniques are strongly linked with many technological issues like complicated and expensive equipment and problems in providing stable quality of vacuum in larger chambers and high installation and exploitation costs.Moreover conductive oxides are brittle; thus they are not suitable for flexible electronics applications which could be resistant for multicycle bending.Therefore there is a need for large-scale and vacuum-free alternative methods like printing (screen printing [15], inkjet printing [16], flexography [17], gravure [18], or stamp printing [19]), dip coating [20], spin coating [21], or spray coating [22].The use of these methods needs the development of particular materials and allows the fabrication of new electronic products.New materials need also a deep understanding of properties in aim to establish guidelines for electronic devices design and development.
The use of novel materials like carbon nanomaterials allows the electrodes to be also flexible which was not achievable for metal oxides such as ITO [23,24].Present paper is related to two kinds of carbon nanomaterials used as functional phases, graphene platelets (GNP) and multiwalled carbon nanotubes (CNT); since they exhibit high mechanical strength [25][26][27] and can be used for transparent coatings [28][29][30], they have great electrical properties [31][32][33][34] and exhibit better performance than graphite based coatings which was described in our preliminary results [28].Transparent electrodes should exhibit possibly the highest optical transmission and the lowest sheet resistance.Unfortunately higher optical transmission usually means higher sheet resistance; thus it is important to find the best match between each other.For this purpose figure of merits (FOM) can be used [35][36][37][38].The method of preparation of transparent conductive layers was spray coating since it is a simple, coseffective, and clean room free technique.Moreover it is easily implementable in production lines and can be used to coat nonflat large area surfaces.

Materials and Methods
To prepare transparent electrodes with spray coating technique special low-viscosity paints (around 1 mPa⋅s) had to be prepared.As a functional phase two carbon nanomaterials were used, graphene platelets (GNP) and multiwalled carbon nanotubes (CNT), both commercially purchased from Graphene Laboratories Inc. GNP were obtained through the graphene oxide reduction and CNT through the catalytic chemical vapour deposition.Graphene platelets had the platelet size between 2 and 25 m and 10 nm average thickness.Multiwalled carbon nanotubes had the diameter between 1 and 15 nm and length from 10 to 50 m.Carbon nanomaterials were suspended in nitromethane (NM) solvent and homogenized.The homogenization was performed in two steps, firstly for 30 s with higher energy with ultrasonic homogenizer Sonics VCX 750 and secondly for 30 min with lower energy in the ultrasonic bath.The homogenization of paints is crucial for the spray coating process performance and allows achieving higher uniformity of prepared electrodes, which is linked to their lower sheet resistance and higher optical transmission.The final paints compositions are presented in the Table 1.
Transparent electrodes were prepared on a flexible polyethylene terephthalate (PET) transparent substrate.Substrates were cleaned before coating with isopropanol, rinsed with distilled water, and dried in a chamber drier in 80 ∘ C for 30 min.The spraying gun was mounted on a  plotter that allows fully controllable motion of the tool in both axes.The spraying gun was supplied by the oil-free air from a special compressing system, to avoid the oil impurities in the layer that could increase the sheet resistance.All samples were cured after layers deposition in a chamber drier in 120 ∘ C for 30 min to evaporate the solvent.samples prepared from paints CNT0.1, CNT0.099/GNP0.001,CNT0.09/GNP0.01,GNP0.1, and CNT0.05/GNP0.05.We can observe that sample with the highest amount of GNP exhibits the lowest amount of CNT agglomerates.In our previous results we have observed that carbon nanotubes are stretched by the graphene platelets along the surface; thus the layer is more homogeneous [39].Unfortunately in the same time several graphene platelets are not tangential with all their surface to the substrate, which can cause higher roughness and refraction.
The XRD (X-ray diffraction) measurements of samples were performed using Philips X'Pert MPD diffractometer with Cu anode.Diffractograms for samples with multiwalled carbon nanotubes with the addition of graphene platelets on borosilicate glass substrates were measured in the angle range 2 of 3-80 deg.
The optical measurements of layers with carbon nanotubes and graphene platelets were carried out in the range from 200 to 2500 nm by the use of PERKIN ELMER Lambda 19 spectrophotometer for samples both on a flexible polyethylene terephthalate transparent substrate and on borosilicate glass.
The sheet resistance measurements were made with the use of 4-point probe method with Keithley 2001 in 5 different places of each sample and average value was calculated.The thickness of the coatings and the surface roughness was measured with a stylus profilometer Veeco Dektak 150 3D.To achieve correct measurements the fraction of 3 × 0.15 mm of surface of each sample was scanned.
The electrical stability of fabricated samples after several bending cycles was examined with custom made laboratory setup.Sheet resistance of samples was measured in the beginning and after 100, 1,000, 5,000, and 10,000 bending cycles of each sample.The bending radius was 15 mm.

Results
Profilometer measurements and sheet resistance where done for 10 samples of each layer material.Profilometer measurements showed that the thickness of prepared samples was quite similar (between 520 nm and 640 nm), but the slight difference between the roughness of the sample with the highest amount of graphene platelets and the other samples was observed.It is caused by the fact that graphene platelets used in the experiment had relatively big diameters, between 2 m and 25 m, and they usually are not coplanar with their surface to the substrate, what we have described in our previous research [39].The samples had also similar sheet resistance which is presented in Table 2.
X-ray diffraction measurements allow identifying the structural properties of composite of various graphene platelets contents.The XRD diffractograms of all composites are presented in Figure 6.They reveal two peaks for angle 2: at about 2 ∘ and 26.5 ∘ .The first one is related to (001) graphite oxide, observed also by other authors in graphite oxide material [40].The second is related to the (002) graphite peak.
In Figure 7, one can see the XRD diffractograms of all samples but with the zoom of areas where characteristic peaks occur.Peak at about 2 ∘ does not exist in samples CNT0.1 and CNT0.09/GNP0.01(Figure 7(a)).The analysis showed differences between the layer of carbon nanotubes with a  small amount of graphene platelets (CNT0.099/GNP0.001)and a coating with a higher amount of graphene platelets (CNT0.05/GNP0.05).The diffraction pattern of a samples CNT0.099/GNP0.001,CNT0.05/GNP0.05,and GNP0.1 contains (001) graphite oxide peak at about 2 ∘ (Figure 7(a)) [41].This peak may be related to carbon nanotubes as well as to graphene platelets.Additionally the distinct (002) graphite peak for angle 2 about 26.5 ∘ is present in (Figure 7(b)) [42].The intensity of this peak strongly depends on GNP content in technological mixture.For this reason for layers without graphene platelets (CNT0.1)and with very small GNP content (CNT0.099/GNP0.001)the peak 26.5 ∘ does not exist.
Using the Scherrer equation [42] the mean size of grains was calculated (perpendicular to the film surface).From FWHM (full-width at half-maximum) values for (002) graphite diffraction peaks, the obtained D values are in the range from 23 to 25 nm.
The diffraction patterns of base powders reveal no graphite oxide peak at small values of 2-theta (Figure 8).Due to this fact we assume that graphite oxide (in CNT0.099/GNP0.001,CNT0.05/GNP0.05,and GNP0.1)) is formed during samples preparation, for example, during sonication with nitromethane (NM) solvent and CNT. Figure 9 shows the spectral dependence of total transmittance for flexible PET and how it changes after applying the composite layers with diverse compositions.In visible range the total transmittance decreases from about 85% to 60% for the most transparent composition, which was the layer with the lowest amount graphene platelets Both CNT and pure GNP and all other hybrid combinations exhibited lower transmission.In the near-infrared between wavelength of 1600 and 2500 nm a significant amount of absorption bands from the substrate was observed.It does not allow a reliable assessment of the optical properties of the material in this spectral range.To provide the reliable measurements in the near IR, samples were also deposited on a glass substrate.The results of total transmission on glass substrates are presented on Figure 10. Figure 10 shows similar decrease of transmittance after applying the carbon coatings in the visible wavelengths.Again the composition with the slight adding of graphene platelets exhibited better performance.Figure 10 shows also that the transmission of hybrid materials in the visible spectrum behaves like a logarithmic curve opposite to pure CNT and GNP where it is similar to the curves with right angles.Glass substrate allowed also the observation in the near IR region.It could be seen that the transmission for all hybrid materials is increasing with the wavelengths opposite to pure CNT ang GNP, where after achieving the peak at around 1600 nm is slowly decreasing.
Figure 11 presents the results related to the electrical stability of samples after several bending cycles.All of examined materials exhibited excellent stability regardless the CNT/ GNP ratio (sheet resistance changes less than 3.4% in relation).Our results show even a small drop of sheet resistance after 10,000 bending cycles (about 1.8% as regards the value before bending), which is probably caused by the selfarrangement of the carbon nanostructures in the fabricated layers.

Conclusions
Carbon nanomaterials consisted of functional phases, graphene platelets (GNP) and multiwalled carbon nanotubes (CNT), which revealed optical properties for applications as transparent coatings of high optical transmission and the low sheet resistance.The analysed layers have a complex structure and a variation of the content of graphene platelets in the structure of layers brings change in the optical properties of material.Spectral dependence of total transmittance in the middle of visible region (i.e., 550 nm) and sheet resistance for composite layers deposited onto flexible PET were shown in Table 3.It could be seen that the best performance was observed for layers made with CNT0.099/GNP0.001composition.X-ray diffractometry indicates the presence of (001) graphite oxide peak (at 2 = 2 ∘ ) in CNT0.099/GNP0.001,CNT0.05/GNP0.05,and GNP 0.1, strongly associated with the technological process.In graphene platelets reach samples (CNT0.09/GNP0.01,CNT0.05/GNP0.05,and GNP0.1) a characteristic (002) graphite peak at 2 = 26.5 ∘ is observed.The highest intensity is achieved for GNP0.1 sample.This result is in good correlation with optical transmission; sample GNP0.1 has the highest absorption in whole range of wavelength.
Results obtained in this research may be useful for understanding the nature of graphene platelets, carbon nanotubes, and their hybrid compositions in functional coatings in optoelectronic

Figure 6 :
Figure 6: Diffraction patterns of all composites deposited on borosilicate glass.

Figure 9 :Figure 10 :
Figure 9: Spectral dependence of the total transmittance for flexible PET substrate and PET with diverse CNT and GNP layers.

Figure 11 :
Figure 11: Cyclic bending of examined layers made on PET flexible substrate.

Table 2 :
Sheet resistance, thickness, and roughness of prepared samples.

Table 3 :
Comparison of hybrid and pure CNT and GNP electrodes on PET.