Microfluidic multicolor display by juxtapositional color mixing with a pattern of primary color pixels

The JMMD was fabricated by a PDMS mold-replica method and a bonding process (figure 3(a)). First, three aluminum molds (top, middle, and bottom layers) with microchannel patterns were prepared by machining (figure 3(i)). The PDMS base polymer and curing agent (Dow Corning Toray Co., Ltd, Silpot 184) were mixed in a ratio of 10:1 to obtain the mixed PDMS prepolymer solution. The prepolymer solution was poured into the top and bottom layer molds. For the middle layer, calcium carbonate powder (Wako Co., 030-00385) was added to the mixed PDMS prepolymer solution at a ratio of 1:10 and suspended by a planetary centrifugal mixer (THINKY Co., ARE-250) to obtain white-colored PDMS prepolymer solution. The mixed white PDMS prepolymer solution was poured into the middle layer mold. The PDMS poured into the aluminum molds was thoroughly degassed in vacuum (figure 3(ii)).

Zoom In Zoom Out Reset image size

For the lids of the top and bottom layer molds, glass plates (Matsunami Glass Ind., Ltd, S9111) were coated with anti-adhesive reagent (Optool, DAIKIN Ind., Ltd, HD-1100TH) by dipping and heated at 100 °C for 5 min. For the lid of the middle layer mold, a glass plate was immersed in the 2% w/w solution of the 3-(2-aminoethyl)-aminopropyltrimethoxysilane (AATMS, Dow Corning Toray Co., Ltd, OFS-6020) in methanol (Wako Co.) for 1 h and baked in an oven at 100 °C for 10 min to prevent the PDMS prepolymer from curing on the glass surface [16, 17]. The top and bottom molds with PDMS prepolymers were covered with the anti-adhesive-coated glass plates, and the middle mold with white-colored PDMS prepolymers was covered with the AATMS-coated glass plates (figure 3(iii)). After heating at 75 °C for 90 min, the PDMS layers were peeled off from the aluminum molds (figure 3(iv)).

All the PDMS layers were bonded by air plasma bonding (MEIWAFOSIS Co., Ltd, SEDE-P). Diluted inks (Cyan, Magenta, Yellow, Black, BROTHER Ind., Ltd, Br-Ink100-4K-2) were injected into the channels without surface treatment (figures 3(v) and (vi)). Since the middle PDMS layer was white colored, the injected ink in the channels at the bottom PDMS layer were optically hidden from the top view. Thus, the injected inks in the cylindrical chambers in the top layer were only displayed as pixels.

On the JMMD, multicolor was expressed by juxtapositional color mixing of pixel patterns of cyan (C), magenta (M), yellow (Y), or black (K). We compared variously diluted CMYK inks introduced in the JMMD to color references for evaluating the color expression: Cyan, magenta, and yellow inks diluted to 10.0%, 5.0%, 1.1%, 1.0%, 0.9%, 0.5%, and 0.1% with pure water, and black ink diluted to 90.0%, 80.0%, 70.0%, 60.0%, 50.0%, 40.0%, 30.0%, 20.0%, 10.0%, 5.0%, and 1.0% with pure water were tested. The pixels filled with these diluted ink solutions and color reference charts (Japan Paint Manufactures Association, Standard Paint Colors Pocket Type, munsell color value 10B5/10 as cyan, 5RP5/12 as magenta, 5Y8.5/12 as yellow, N1 as black) as references for the primary colors were placed side by side, and photographs were taken with a digital camera (Olympus Co., Tg-5) (figure 4(a), detailed images are in supplementary figure S2). The L* a* b* values (a type of coordinate value in color space [18–20]) for the ink-filled pixels and the color references was measured by the image processing program (ImageJ, Lab stack and Histogram). By using the average values of the measured L* a* b*, the color difference (${{\Delta }}E_{00}^*$) [21, 22] between the color chart and the pixels was calculated for CMYK colors (see supporting information 3 for a detailed derivation of the color difference).

Zoom In Zoom Out Reset image size

As a single color gradation, the JMMD was filled with the optimized CMYK inks at fill ratio of 25% (one fourth of all the pixels in the display was filled), 50% (half of the pixels), 75% (three forth fo the pixels), and 100% (all pixels). As a dual color gradation, the JMMD was filled with the cyan and magenta (C-M), magenta and yellow (M-Y), yellow and cyan (Y-C), cyan and black (C-K), magenta and black (M-K), yellow and black (Y-K) inks. The fill ratios of 75%:25%, 50%:50%, and 25%:75% were examined for all six dual color gradations. The JMMD with the inks was photographed by a digital camera (Olympus Co., Tg-5, Macro mode, white balance: white fluorescent light (4200K)) under ambient light environment in a laboratory. The obtained images were black-balanced with an image processing software with (Adobe, Photoshop). The distribution of HSV of the entire display surface was measured by ImageJ (NIH). The averaged values of the measured HSV under each color variations of the JMMD were shown with computational colors converted from HSV by an illustration software (Adobe, Illustrator). The averaged values of the HSV were used to express the color gamut of the dual color gradations using C-M, M-Y and Y-C on a H-S color circle and those using C-K, Y-K and M-K on a S-V plane.

As a demonstration, color images were shown on the JMMD. From the experimental results of the ink concentration optimization, 1% ink concentration for cyan, magenta and yellow, and 5% ink concentration for black were used for the JMMD. A '+' symbol was designed on the pixel array. According to the designed pixel patterns, the inks and air (as blanks) were injected into the channels to display the image. In the case of filling the channel with ink, the ink solution was dropped on the inlet hole and sucked with a syringe from the outlet hole. In the case of introducing the air into the channel, the ink was removed from the inlet hole before sucking with the syringe. For example, to fill pixels with a pattern of 2 inks-4 blanks-2 inks, ink was dropped on the inlet and sucked from the outlet for 2 pixels, then, ink was removed from the inlet and sucked for 4 pixels, and ink was dropped on the inlet again and sucked from the outlet for 2 pixels. The process of selective ink and air injection was performed manually. A '+' symbol was displayed at various fill ratios of the ink: cyan (CMYK:100, 0, 0, 0), magenta (CMYK:0, 100, 0, 0), yellow (CMYK:0, 0, 100, 0), black (CMYK:0, 0, 0, 100), violet (CMYK:50, 50, 0, 0), orange (CMYK:0, 50, 50, 0), and green (CMYK:50, 0, 50, 0). In the same way, characters were displayed on JMMD with the following conditions: 'O' (CMYK:0, 0, 0, 100), 'N' (CMYK:100, 0, 0, 0), 'O' (CMYK:0, 100, 0, 0), 'E' (CMYK:0, 0, 100, 0), 'L' (CMYK:50, 0, 50, 0), 'A' (CMYK:0, 50, 50, 0), and 'B' (CMYK:50, 50, 0, 0). The JMMD with the color image was photographed by a digital camera (Olympus Co., Tg-5, Macro mode, white balance: white fluorescent light (4200 K)) under ambient light environment in a laboratory. The obtained images were black-balanced with an image processing software (Adobe Photoshop).

Our JMMD had 14 × 14 pixels in a 2D display surface at 8.64 ppi resolution, composed of four different 3D microchannels for each CMYK color in the three-layer PDMS sheet (figure 3(b)). When the JMMD was not filled with ink (initial state) (figure 3(b), top), the white PDMS (background screen) was visible. A close-up view showed the array of cylindrical chambers (pixels) in the top layer and the through holes in the middle layer. From the image filled with optimized CMYK ink (figure 3(b), bottom) showed that each of the cyan and yellow pixels was connected in the x-axis direction, and each of the magenta and black pixels was connected in the y-axis direction in every other pixel. The middle layer worked as a background screen because the ink filled in the microchannels at the bottom layer was optically hidden by the middle layer. Since the JMMD was made of soft material (PDMS), the JMMD was highly flexible. The ink array remained stable even when the JMMD filled with ink was bent, indicating that the JMMD could be applied to as a flexible display (figure 3(b) bottom).

The color expression on the JMMD was determined by the coloring of CMYK inks introduced in the microchannel. Thus, the concentration of the CMYK inks were optimized by comparing the color difference (${{\Delta }}E_{00}^*$) between the diluted CMYK inks and the color reference chart (figure 4(a)). From the photos of the cyan optimazation experiment (figure 4(b), cyan), the pixels filled with 10% concentration ink had a darker blue color than the color chart (figure 4(b), cyan, right inset). On the contrary, pixels filled with 0.5% concentration ink had a lighter blue color than the color chart (figure 4(b), cyan, left inset). Therefore, cyan at 10% and 0.5% concentration had a larger color difference value from the color chart (${{\Delta }}E_{00}^*$ = 37.5 at 10%, ${{\Delta }}E_{00}^*$ = 23.0 at 0.5%). The graph of cyan (figure 4(b), cyan) showed that the cyan ink concentration closest to the color chart was determined to be 1% (${{\Delta }}E_{00}^*$ = 4.8). For magenta and yellow (figure 4(b), magenta and yellow), ${{\Delta }}E_{00}^*$ showed the minimum value when the concentration was 1%, meaning that the ink concentration closest to the color chart was 1% (${{\Delta }}E_{00}^*$ = 0.5 at magenta, ${{\Delta }}E_{00}^*$ = 9.0 at yellow). In the case of black (figure 4(b), black), there was no significant difference in the values of ${{\Delta }}E_{00}^*$ from 100% to 5% ink concentration, but the value of ${{\Delta }}E_{00}^*$ was high at 1% concentration. Therefore, black can be used as the primary color with any concentration of ink as long as the ink is not too diluted. From the above, the optimal ink concentrations for CMYK pixels were determined to be 1% for cyan, magenta, and yellow, and 5% for black. These ink concentrations determined by the optimization experiment were used in the subsequent experiments.

Firstly, color graduation expressed on the JMMD was investigated by introducing a single color ink with different fill ratios (spatial density of the colored pixels). The single color gradation expressed on the JMMD (figure 5(a), left) shows that the expressed colors became more vivid as the fill ratio increased. In the case of cyan, the hue (H) and value (V) remained constant (H was around 200, and V was around 80), while the saturation (S) increased from 28 to 83 as the fill ratio increased (figure 5(a), right). The saturation of magenta increased from 23 to 54 and that of yellow from 22 to 55. In the case of black, the hue and saturation remained constant (H was around 290, and S was around 7), and the value decreased from 62 to 29. Therefore, the saturation control of CMY and the value control of black were achieved by changing the spatial pixel fill ratio.

Zoom In Zoom Out Reset image size

Multicolor expression by juxtapositional color mixing on the JMMD was next evaluated. The dual color gradation of cyan, magenta and yellow was created on the JMMD and analyzed with HSV value (figure 5(b)). From the measured HSV values, the combination of two different colors mainly dominated the hue of the colors displayed on the JMMD (figure 5(b), left). In the nine conditions, saturation and value showed little variation (the S value was 57.0 ± 5.1, and the V value was 80.6 ± 4.6 (figure 5(b), center)), while the hue changed in sequence (H varied from 9, 25, 71, 144, 183, 225, 263, 308, and 354 for conditions M:Y = 50:50 (middle of the color chart), M:Y = 25:75, Y:C = 75:25, Y:C = 50:50, Y:C = 25:75, C:M = 75:25, C:M = 50:50. C:M = 25:75, and M:Y = 75:25). Therefore, by using the JMMD, combination of only three types of inks (CMY) can express the wide range of colors in the Hue. The distribution of the colors expressed by the combination of cyan, magenta, and yellow was expressed on the H-S color circle (figure 5(b), right). The plot area (dotted line circle) means the color gamut that was expressed on the JMMD, thus indicating that the JMMD can express 360-degree color tones in stages.

The dual color gradation with black (C-K, M-K, and Y-K) on the JMMD (figure 5(c)) shows that the colors became darker as the fill ratio of black increased. From HSV values, the decreases in S and V was confirmed as the fill ratio of black increased. (in the case of cyan, S decreased from 70 to 35, and V decreased from 72 to 46. In the case of magenta, S decreased from 45 to 20, and V decreased from 83 to 45. In the case of yellow, S decreased from 42 to 15, and V decreased from 69 to 44). The color distribution plots on the S-V plane (figure 5(c) bottom) means that the fill ratio of black can design the S and V of the color displayed on the JMMD in stages. Thus, the HSV of the colors displayed on the JMMD can be designed by controlling the fill ratio of the primary colors (CMYK).

Finally, the color images and characters with juxtapositional color mixing were displayed on the JMMD as demonstrations (figure 6). Based on the illustrated design of the color images of the '+' in various CMYK ratios (figure 6(a), top), the pixel patterns were generated with converted CMYK proportions (figure 6(a), middle). The '+' symbols displayed on the JMMD with the generated pixel patterns (figure 6(a) bottom) shows that the same pixel pattern as the designed pixel pattern was displayed on the JMMD for each color. The '+' signs that were 50% cyan and 50% magenta, 50% magenta and 50% yellow, and 50% yellow and 50% cyan, were visually observed as violet, orange, and green, respectively, by juxtapositional color mixing. Therefore, the JMMD displayed images with a wide range of color variations. Examples of alphabetic characters, 'ONOELAB.' were also demonstrated on the JMMD to confirm the capability of the displayed image variations on the JMMD (figure 6(b)). Therefore, in principle, arbitrary color images can be displayed on JMMD by designing the pixel pattern of the desired color image.

Zoom In Zoom Out Reset image size

We proposed the microfluidic multicolor display using only four primary colors (JMMD) to demonstrate multicolor expression using juxtapositional color mixing on the JMMD. There have been several reports on microfluidic displays that express multicolor [13, 14], but the expressed colors on the reported displays limited to the colors of the inks used in the displays. Thus, the previous microfluidic displays required to prepare many different inks to display multicolor images. In addition, switching between those multiple inks introduced into the microchannels or mixing of the different colors in the pixels was essential for the previous microfluidic displays, causing the complexity of display systems. On the other hand, the proposed JMMD achieved the expression of color gradations by simply controlling the fill ratio of the pixels, not the ink variations. This indicates that the fluid control system required for the JMMD could be very simple, and that the JMMD overcomes one of the factors that hindered the practical application of microfluidic displays.

The JMMD is the first microfluidic display that has multi-color expression function by juxtapositional color mixing. Juxtapositional color mixing is a widespread color display method in commercial optical displays (e.g. PC, smartphone) [23, 24]. A typical light-emitting display consists of light-emitting elements of red, green, and blue, the three primary colors of light. However, while additive color mixing is used for the light-emitting displays, subtractive color mixing [25] must be used for the reflective displays because the reflective displays do not emit light themselves [26–28]. Therefore, cyan, magenta, and yellow, the three primary colors of color, are suitable as a display element for reflective displays.

Generally, display resolution dominates the color quality in a multicolor expression method based on the principle of juxtapositional color mixing. Juxtapositional color mixing with CMY is used in the printing technology of commercial inkjet printers. Inkjet printers print in color by juxtaposing minute circular patterns of cyan, magenta, and yellow (on the 1 μm order) [29, 30]. Therefore, the JMMD shows multicolor in a similar way in principle to current printing technology. The JMMD proposed in this paper has 2.5 mm diameter pixels and 8.46 ppi resolution. Because typical commercial display resolutions are around 100–300 ppi, smoother juxtapositional color mixing could be achieved on the JMMD by downsizing the pixel size of JMMD to around 100 μm. In addition, the color quality is also affected by ink misalignment. For example, in figure 6, ink had been trapped in some of the pixels that should be blanks. In the previous study, the trapped ink was suppressed by controlling the pixel shapes [13]. Therefore, downsizing the pixel size of the JMMD and optimizing the pixel shapes would improve the color quality of the JMMD.

As an alternative approach to increasing the resolution, large area displays such as multi-purpose sign boards used in outdoors, wearing objects including cloths, and exterior/interior of vehicles or architects are possible applications of the JMMD without high pixel resolutions. For the human eye, optical resolution is determined by a minimum distance where two points can be separated and perceived: If two points that are closer together than that minimum distance is not recognized separately [31]. For example, when you look at photographic mosaic art [32, 33] from close up, each photo image (a single pixel ) can be recognized. But when you look at the art from a certain distance, photo images appears to be one illustration. This displaying method stems from the fact that the principle of juxtapositional color mixing utilizes the limits of the resolution of the human eye. Thus, large area displays with the proposed pixel design of the JMMD could open a new way to innovative and low-energy-consumption applications that have not been realized with light-emitting displays.