Licai
Zhu
ab,
Jin
Tao
*a,
Panyuan
Li
ab,
Wenchao
Sun
ab,
Jiwei
Li
ab,
KaiLi
Fan
ab,
Jinguang
Lv
a,
Yuxin
Qin
a,
Kaifeng
Zheng
a,
Baixuan
Zhao
a,
Yingze
Zhao
a,
Yupeng
Chen
a,
Yingwen
Tang
c,
Weibiao
Wang
a and
Jingqiu
Liang
*a
aState Key Laboratory of Applied Optics, Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun, Jilin 130033, China. E-mail: taojin@ciomp.ac.cn
bUniversity of Chinese Academy of Sciences, Beijing 100049, China
cCollege of Physics and Information Engineering, Minnan Normal University, Zhangzhou 363000, China
First published on 11th April 2023
This paper presents an easy and intact process based on microfluidics static droplet array (SDA) technology to fabricate quantum dot (QD) arrays for full-color micro-LED displays. A minimal sub-pixel size of 20 μm was achieved, and the fluorescence-converted red and green arrays provide good light uniformity of 98.58% and 98.72%, respectively.
In our previous study, 140 × 50 μm sub-pixels were created using conventional microfluidic technology.27 However, due to their lateral size exceeding 100 μm, they were unsuitable for using in micro-LED pixels. Moreover, because there was no alternative method to individually isolate pixels within an array, pixels of the same color were linked in a linear sequence through a single microchannel. This resulted in a crosstalk effect at the microbridge junction of adjacent pixels of the same color, which affected the performance of the device.28
In this paper, we propose a simple and intact method to fabricate QDCCLs that exploits microfluidic static droplet array (SDA) technology. As a typical representative of droplet microfluidics, SDA technology can generate, transport, immobilize, fuse, and store micro-droplets29,30 ranging from 1 μm to 1 mm,31–35 which is comparable to the pixel size of micro-LEDs. Red and green QD solutions are conducted in separate channels of the same SDA and after removing the excess QD solution and solidification thereof, a dual-color QD array is formed with uniform size and separated sub-pixels. Compared with photolithography patterning, the method proposed herein produces intact QDs with maximal fluorescence conversion efficiency. The output has a neater morphology and higher PPI than is possible with inkjet-printing because the pixel size is determined by the photolithography process of SDA and is more consistent and reliable than transfer printing.
Fig. 1(a)–(c) show the process for preparing dual-color QDCCLs with a SDA microfluidic chip. All microfluidic devices discussed herein were fabricated using standard soft lithography.36 Fig. S1† shows the process for preparing a SDA microfluidic chip. Red and green QD solutions are first injected into the SDA microfluidic chip to completely fill the pattern [Fig. 1(a)] and then transparent sealing fluid is injected [Fig. 1(b)] to create the pixel dots. Finally, in Fig. 1(c), the microfluidic chip's inlet and outlet are sealed to keep out air and water vapor and safeguard the QDs. Fig. 1(d) shows a single SDA element, which consists essentially of a channel, a bridge and a well. Fig. 1(e) shows the preparation of a single QD pixel. Fig. S2† shows a photograph of fluorescence from a dual-color QDCCL and the optical micrographs of the SDA microfluidic chips.
Fig. 1 (a–c) Schematic diagram of a QDCCL prepared by SDA. (d) Single structure of SDA. (e) Preparation of a single QD pixel. |
The experimental images of a dual-color QDCCL with a pixel size of 50 μm and array dimensions of 18 × 20 were created by SDA to demonstrate the specifics of QDCCL preparation, as shown in Fig. 2. The time to complete a QDCCL is 257 s, which can be significantly decreased by raising the pressure of the injection pump or by simultaneously injecting the dual-color QD solutions. The red QD solution is injected into the channels and flows across the bridges and into the wells, as shown in Fig. 2(a)–(d). A portion of the solution entering the wells progressively fills the entire chamber because of the capillary force.37 The process for injecting green QD solution is shown in Fig. 2(e)–(h); the same process works for red QDs. Next, the injection port is used to inject the transparent sealing fluid. Dow Corning medical silicone oil was used as the transparent sealing liquid,38 which totally drained the QD solution in the channels because it is immiscible with the QD solution. For details on transparent sealing fluid, please see the ESI.† Given that the wells are filled with the QD solution and that the bridges are considerably smaller than the channels, the transparent sealing fluid has difficulty entering the wells, but it flows out of the liquid outlet along the channels due to the large hydraulic pressure differential. According to Fig. 2(i)–(l), only the final QD solution remains in the wells, creating a QD-pixel array. See the ESI movie† for further details on the injection procedure.
SDAs can create 1 to 100 μm droplets with good morphology and uniformity, as evinced by visible photography. Fig. 3 shows monochromatic and dual-color QDCCLs with pixel sizes of 90, 50 and 20 μm prepared by SDA technology; the number of monochromatic and dual-color pixel arrays is 10 × 10 and 18 × 20, respectively. The horizontal pitch is 150, 90 and 40 μm for pixel sizes of 90, 50 and 20 μm, respectively. The horizontal pitch can be made smaller by increasing the lithography precision and decreasing the channel size. In this experiment, the bridge's width is also investigated. There are four intended bridge sizes: 3, 5, 7 and 9 μm. Comparing the results with the different bridge sizes shows that the pixel arrays of the three sizes all exhibit good results when the bridge's width is 5 μm. The QD pixels produced are destroyed when the bridge width is 7 μm, because the transparent sealing fluid enters the wells and destroys the pixels. However, it is difficult for the QD solution to enter the wells to form pixels when the bridge size is 3 μm.
Fig. 3 (a–c) Green, (d–f) red and (g–i) dual-color QDCCLs with pixel sizes of 90/50/20 μm prepared by SDA, respectively. |
CdSe/ZnS QDs were used in the experiment because of their excellent stability, long fluorescence lifetime, and narrow emission spectrum.11,39Fig. 4(a) and (b) show the absorption spectrum and UV excitation spectrum of the CdSe/ZnS QDs thin film. The emission peaks of red and green QDs are 623 and 529 nm, respectively, with FWHMs of 24 and 21 nm, and quantum yields of 72% and 75%, respectively, when excited by 365 nm UV light.
In addition, the luminous uniformity of pixels is a critical metric for assessing display performance.25,40,41 The average fluorescence intensity distribution of a monochromatic QDCCL with a 10 × 10 pixel layout of 50 μm is shown in Fig. 4(c) and (d), and for 20 μm in Fig. 4(e) and (f), as measured by ImageJ. According to the electronic industry standard “Measure method of light-emitting diode (LED) display screen” of the People's Republic of China, the light uniformity of red and green QDs with a pixel diameter of 50 μm is calculated to be 94.1% and 96.4%, respectively, while the uniformity of red and green QDs with a pixel diameter of 20 μm is calculated to be 98.58% and 98.72%, respectively. This indicates the high fluorescence homogeneity of the QDCCL produced via SDA technology, which conforms to the national standard. The uniformity of 20 μm pixels is improved by 4.54% and 2.35%, respectively, compared to 50 μm pixels, indicating that SDA microfluidic technology produces greater fluorescence intensity uniformity for smaller-sized QD pixels. We believe this is due to the microfluidic chip manufacturing process and the reduced impact of inherent non-uniformity of QD solution on smaller-sized QD pixel arrays. Fig. 4(g) shows the fluorescence intensity of two adjacent red and green QD pixels along the dashed line in the 20 μm dual-color QDCCL under 365 nm UV excitation.42 The green QDs emit a higher intensity than the red QDs when exposed to the same amount of light, and there is minimal optical crosstalk between neighbouring pixels.
As shown in Fig. 4(h), we also investigate the relationship between injection time and the number of filled monochromatic QDCCL pixels for three different pixel sizes and array numbers of 100 when the injection velocity is 10 μL min−1 and the SDA pattern depth is 20 μm. Here, when the QD solution fills a single well in excess of 50%, the well can be regarded as a complete pixel because the capillary force inevitably guides the solution to fill the well completely,31 so this part of the time is not included in this study. As soon as the first pixel emerges, we begin. The injection times of 20, 50 and 90 μm pixel arrays are 60, 80, and 110 s, respectively. The difference is mainly caused by the different volumes of patterns that need to be filled with solution. In addition, the QD solution rapidly fills the SDA at the outset, then slows down, and finally accelerates again. The reason for this behavior is that the QD solution first fills the entire channel; the anterior segment of the channel has a larger hydraulic pressure than the posterior segment, so the QD solution fills the anterior segment of the SDA pattern preferentially. In the anterior segment, the solution gradually flows from bridges to wells, and the filling area of the wells soon reaches 50%, forming partially filled wells, which explains why pixels rapidly form in the early stage. With continuing injection, the solution fills the partially filled wells to make them complete, but during this time, fewer new pixels are generated than before. Once the anterior part of the well array is filled, the speed of the posterior segment pixel generation will resume its rapidity.
We probed blue OLED backlight excitation on the dual-color QDCCL with a pixel size of 50 μm to analyse the performance of the full-color display of the QDCCL by SDA.43Fig. 5(a) shows the excitation spectrum. The comprehensive intensity ratio of blue, green, and red is 1.00:0.22:0.15, so the intensity of the blue light is significantly greater than that of the red and green light, according to the spectral data. In this case, white balances the CIE chromaticity coordinates of the samples mentioned above. The colorspace coverage reaches 119.7% of the National Television System Commission (NTSC) standard, as shown in Fig. 5(b).
Fig. 5 (a) Emission spectrum of blue OLED backlight with a 50 μm dual-color QDCCL. The inset is the same spectrum but with magnified red and green scales. (b) Color gamut diagram. |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2na00765g |
This journal is © The Royal Society of Chemistry 2023 |