Inkjet-Printed Patterned Quantum Dots Film for High-Efficiency Displays

Due to the effect of solution kinetics and crystallization kinetics of quantum dot (QD) during inkjet printing, the inkjet printing technique is prone to form coffee-ring patterns and in-homogenous films. Pioneering efforts are usually limited to mixed solvent or additive inks, in this paper, we investigated polymer QD films with good uniformity obtained by inkjet printing, in order to control the uniformity of quantum dot films in pixels, the mechanism of action of wettability science on inkjet printing was considered through the action of nitrous oxide (N2O) on the surfaces of substrates and banks. In the absence of N2O treatment, the inkjet-printed QD droplets would pile up and adhere to the bank walls due to capillary forces before the pinning process. With the N2O surface treatment, the contact angle of the QD ink on the substrate was increased from 16° to 62°. The wettability of the QDs between the substrate and the bank was reduced, and the side build-up phenomenon was significantly reduced. By changing the thickness of the QD film from 4.5 μm to 15 μm, the color coordinates of the spectrum were changed from (0.2320, 0.1556) to (0.1480, 0.6933).

by pumping QDs through a short-wavelength electroluminescent chip, without the need for a mass transfer process [7]. Because QD-LEDs use QD films as color-conversion layers, they have lower light loss than white Organic Light Emitting Diode (OLED) [8], [9]. This approach is also considered to be one of the most likely solutions for next-generation display technologies (including micro-LEDs), among others [10].
For patterning, the photolithography-etch process is the most established method. However, there are several main technical difficulties in the preparation of QD color photoresist. QDs are easily quenched when encountering water and oxygen, so the requirements for QD materials themselves are very high [11], [12]. The advantage of lithography is to form patterns with fine pitches, but it does not guarantee low material consumption and the fabrication steps are cumbersome [13]. In the vacuum deposition process, the yield of the material from the source to the deposited film is low, compared to photolithography, Inkjet printing is the best candidate for producing large-scale patterned QD films with little material waste, simple process, low cost, highly automated, and maskless patterning.
With the improvement of the resolution in the display field, while meeting the needs of the pixel size, the previously printed pixel units are often larger than 100 microns, and the current mainstream technology should achieve the pixel size below 50 microns. To ensure that the ultraviolet light emitted by the UV LED does not leak, the printed QD color conversion layer film requires a thickness of a few microns. When an ink droplet with a diameter of tens of micrometers lands on the target point, the final morphology of the ink droplet is related to the surface energy of the substrate [14]. On hydrophilic surfaces, it tends to have good wettability, on hydrophobic surfaces, it tends to have poor wettability [15]. Typical drop sizes available with current inkjet printheads range from 10 to 100 µm in diameter (or 0.5 to 500 pl in volume) due to droplet generation limitations. There have been some previous studies investigating the application of various surface treatments to control surface wettability to form homogeneous films. They often applied chemical coatings and physical engineering to make substrates and banks hydrophilic and hydrophobic, and the film thickness that can be achieved below 1 micron, which cannot meet the needs of thick film, and the problem of film thickness non-uniformity occurs [16], [17]. Therefore, it is urgent to develop a new way to meet the uniformity of thick films.
In this paper, we used inkjet printing to develop green matrix patterns. The substrates and banks used were treated with hydrophobic N 2 O, and a uniform QD color conversion layer of more than ten microns was formed on the sample, and the thickness was controlled by inkjet printing instead of the usual Samples treated with hydrophilic substrates and hydrophobic banks. N 2 O plasma surface treatment is a dry process with short treatment time and high efficiency. Improve the surface energy of the material without changing the intrinsic properties of the material. By treating the substrate and dam with different wettability, the effect of photoresist surface wettability on the shape of the printed film was studied. With optimized surface wettability, the light conversion efficiency (LCE) from UV light to green light can be achieved by adjusting the thickness of the QD layer.

II. EXPERIMENTAL SECTION
The process to prepare the patterned QD film is illustrated in Fig. 1. A bank structure where QD ink is ejected from a piezoelectric nozzle onto a substrate. The ink is confined by the black matrix with an average depth of 15 µm and effectively prevents the diffusion of the QD ink. After printing, it cured by UV light.
Blocking dams are required between pixels, and in order to create a color conversion layer that converts UV light into green light. We compared photoresists with different viscosities, and finally chose 420 cP as the best solution. In order to reduce the optical crosstalk effect in the previous work, a mold with windows and retaining walls was prepared. In addition, the shape and size of the prepared banks were analyzed using an optical microscope (OM; VHX-5000; KEYENCE) and an image analysis program, the bank pattern was made using conventional photolithography, Process (spin coating, pre-bake, UV exposure, post-bake and develop). As shown in Fig. 2 The surface treatment of the substrate affects its surface energy and surface roughness, which in turn changes the wettability of the black matrix. To remove surface oil stains, the substrates were sonicated in an acetone solution for 15 minutes. Next, UV ozone surface treatment was performed on the substrate for 20 minutes. After cleaning, the banks were fabricated using conventional photolithography, and then the samples were then surface treated with N 2 O by PECVD method. For contact angle measurements, SL200K (Kino, USA) was used to evaluate the wettability of the black matrix with glass substrate.
The printer model we used was a Dimatix DMP 2831, and it was paired with a DMC-11601 ink cartridge (Fujifilm, Japan) for fine print. And by dynamically adjusting the jetting waveform, firing frequency, firing voltage, printing height, substrate temperature and ink cartridge temperature, the smooth jetting conditions for QD inks were successfully found. Viscous ink droplets without trailing fluid or satellite droplets were achieved, ensuring reliable and continuous ink printing. Fig. 2(a) shows a photograph of the DMP 2831 inkjet printer, (b) the image of the QD ink droplet ejected from the nozzle, (c) the depth of the embankment measured in a single step, (d) and optical microscope image of the formed bank pattern. After printing the quantum dots with different layers, we carried out the encapsulation process. For the encapsulation process, a glue dispenser (300DS, Musashi Engineering Inc.) was used to apply a low viscosity UV glue to the edge of the glass cover and cured with UV light for 5 minutes. The entire encapsulation process was performed in a nitrogen (N 2 ) atmosphere glove box.

III. RESULTS AND DISCUSSION
The photoluminescence (PL) and UV-vis absorption spectra of the QDs are shown in Fig. 3, The emission peak is located at 516 nm and 627 nm, and the full width at half maximum (FWHM) is 31 nm and 28 nm. Fig. 4 shows the contact angle images of water on the glass substrate and the bank. The screen shows that the water contact angle in the black matrix without surface treatment shows hydrophobic properties, and the value is 85°, and the value decreases after treatment. In addition, water exhibits hydrophilic properties on glass substrates, and its hydrophilicity increases after UV ozone treatment. After treatment with N 2 O hydrophobic molecules, the contact angle increased significantly, as previously reported in the literature [18], [19], [20], [21]. However, the water-to-glass contact angle was 16°without surface treatment and 85°on the shore, and after N 2 O treatment, the glass-towater contact angle increased from 16°to 62°, the surface energy between the bottom of the cavity and the sidewall of the black  matrix changes, and the green QD ink can be smoothly deposited in the pixel dots surrounded by the black matrix. Due to the difference in wettability, the QD droplets formed by successive stacking with increasing thickness have poor diffusivity and will not be trapped in the bank but accumulate on the bank wall, a flat QD film cannot be formed. It is found that N 2 O can reduce the wettability of QD inks on glass and banks. And it will increase the printing thickness of the QDs and reduce the leakage of UV light.
The film formation process of the droplets is: (i) impingement of the droplet on the substrate, (ii) spreading, (iii) pinning [22], [23], [24]. Here, the characteristic pattern of deposition is attributed to a form of capillary flow, pinning of the contact lines induces outward convective flow as evaporation proceeds, and the pinning process of droplets is strongly related to the surface energy of the substrate. At the perimeter of the sidewall, the liquid shrinks, but the radius does not decrease because its contact line is fixed. In the previous literature, increasing the wettability of the substrate increases the size of the final formed droplet ejected from the ink port, while decreasing the wettability of the substrate decreases the size of the droplet after falling [18].
In order to confirm the speculation that wettability affects the surface shape and size, a simulation analysis was carried out, in which the size of the substrate surface and black matrix wettability was selected as the main influencing factor of the deposition effect. First, for the substrate without surface treatment, the glass contact angle is 16°and the photoresist contact angle is 60°; for the N 2 O treated substrate, the glass contact angle is 38°and the photoresist contact angle is 89°. Obviously, when treated, the wettability is reduced, and a weakened capillary phenomenon can be observed, as shown in Fig. 5 below. These simulation results show that the reduction of the surface energy of the substrate has a promoting effect on the formation of a flat and uniform QD film.
Controlling the deposition process is critical for making thin films with high uniformity due to size effects, as shown in Fig. 6(c) and (d), after N 2 O treatment of the substrate for 1 minute, the QD films could be uniformly formed under the microscope, The QD films without surface treatment were mainly distributed at the edge of the pixel. Without surface treatment, the printed QD droplets could not form a uniform film, but spread on the substrate and accumulated heavily on the bank walls, as shown in Fig. 6(a). High surface wettability, the three-phase contact lines of the ink droplets were located on the bank, and the shape of the print was determined by the contact angle. However, for 60 s of N 2 O treatment, the pinning process hardly occurred, as shown in Fig. 6(b). Since the bank space was small, to ensure the printing effect, we used a 1 pl nozzle to control the final morphology and thickness of the QD film in the black matrix pixel unit. Fig. 7(a) and (b) show the thickness and luminescence images of the inkjet-printed QD layer without any treatment, the step meter showed that the inkjet-printed QD droplets formed QD films on the bank walls, for UV/ozone treated the sample, the thickness of the QD films became smaller because the wettability of both the substrate and the bank was increased, and more UV light was transmitted out, forming a concave in the middle and a convex image around it. When the wetting was reduced by N 2 O. The thickness of the QD film increased and became smoother, and the ability to absorb ultraviolet light became stronger.
The color was patterned by inkjet printing, and the relationship between the color conversion efficiency and the thickness of the QD color conversion layer was discussed, the QD color conversion layer was inkjet printed on the bank patterned substrate with controllable thickness with different numbers of inkjet printing droplets after 60 s N 2 O treatment. Fig. 8 shows the color coordinates of different thicknesses of inkjet-printed green QD color conversion layers. As the number of printed drops increased from 15 to 60 drops, the UV LED intensity decreased significantly and the green light intensity increased significantly.   To evaluate the conversion properties, patterned QDs were printed on a 700-micron-thick piece of glass. After soft-baking and UV curing the entire layer, the QD layer on the glass was excited by tightly applying UV backlight LEDs to the backside of the glass. UV backlight LED, square, fabricated using the same GaN epilayer. The UV light will continuously penetrate the glass and the QD layer so that the unabsorbed UV light was collected by the light energy meter along with the converted red or green light. Based on the obtained data, the optical output power of the unabsorbed blue light (PB1) and the converted red  was used as a reference by irradiating the bare glass. Therefore, the absorption rate, conversion ratio and power conversion efficiency (PCE) [25] can be defined as follows: Power conversion Efficiency = P R1 P B0 or P G1 P B0 The switching properties of the green QD layer are shown in Fig. 10. Three samples were measured and compared, with green QD layer thicknesses of 15, 30, 45, and 60 drops, respectively. As shown in Fig. 10(a), in the case of the backlight LED injecting 100 mA current, the absorption rate increases with the increase of green QDs, at high thickness values, the absorption ratio rises more slowly, and as the thickness increases, the maximum absorption ratio will remain unchanged, when the mean free path of the QD film was shorter, its extinction coefficient was larger, and the ability to absorb UV light was stronger. The conversion rate decreased first, then increased and then decreased with the increase of the thickness of the QDs. As the printing progresses, the absorption capacity of the QDs for blue light increases, and the conversion rate tends to decrease. When the printed ink droplet exceeds a certain thickness, due to the overlap between the absorption and emission spectra of the QDs, this may lead to reabsorption, resulting in high conversion losses, resulting in a decrease in the intensity of the converted light.
To better understand the influencing factors of Color Conversion Efficiency (CCE), we defined CCE as the ratio of the green light emitted by the QDs to the intensity of the original UV LED (without QDs) [26]. When the current was 100 mA and 15 drops were printed, PG1 was 7.8 mW; PB0 was 97.6 mW; When the current was 100 mA and 30 drops were printed, PG1 was 9.2 mW; PB0 was 97.6 mW; When the current was 100 mA and 45 drops were printed, PG1 was 14.4 mW; PB0 was 97.6 mW; When the current was 100 mA and 60 drops were printed, PG1 was 10.3 mW; PB0 was 97.6 mW; The calculated CCEs of QD color conversion layers with 15 drops, 30 drops, 45 drops and 60 drops were 8.0%, 9.4%, 14.7% and 10.5%, respectively. When the film thickness of the printed QD layer is too thin, the residual UV light cannot be suppressed, so it cannot efficiently convert from UV light to green light. We also found that when the number of printed QD layers exceeds a certain value, the CCE actually decreases. These results may be due to the spectral overlap of the materials themselves. The light from the QDs generated by the color conversion layer competes with the reabsorption by the QDs in the QD film. In order to maintain the advantages of QDs, fluorescence resonance energy transfer (FRET) should be considered, reducing nonradiative recombination and avoiding self-absorption between printed QDs. Fig. 10(c) depicts the converted green light (λ = 516 nm) along with the unabsorbed UV light (λ = 365 nm) emitted from the green QD layer. For a 60-layer thick green QD layer, the unabsorbed transmitted UV light was still stronger than the converted green light. Therefore, in the actual full-color demonstration, it is essential to apply an extra layer of green CF to block the unabsorbed UV light.

IV. CONCLUSION
In this study, in order to make the QD pixels exhibit uniform morphology, we surface-treated the substrate and the dykes. Through the measurement of hydrophobicity and hydrophilicity, the results showed that the contact angle between glass and water was 16°without any treatment, and the contact angle was reduced to 9°after UV treatment for 10 minutes. However, after 60 seconds of N 2 O treatment, the contact angle between glass and water increased to 62°. Furthermore, the QD films were clearly attached to the bank walls before surface treatment, but disappeared after N 2 O treatment. Each pixel exhibits uniform photoluminescence, and the thickness of the thin film within the bank can be evaluated by the number of layers of printed QDs. By changing the thickness of the QD film from 4.5 µm to 15 µm, the CIE chromaticity coordinates of the spectrum changed from (0.2320, 0.1556) to (0.1768, 0.5110), (0.1959, 0.5339), (0.1438, 0.6899) for UV LED and (0.1480, 0.6933), changing from near ultraviolet to near pure green. In addition, printing 45 drops of QDs, we found the largest PCE, which provides some references for the recent mass production of micro-LED full-color micro displays.