Open Access
Add to BibSonomy
Abstract
All glass, lenticular lens array, light guide substrates were fabricated in a single mask and etch procedure enabling 1D local light-confinement from an edge-injected LED array. Lenticular structures of sufficient resolution (width & pitch: 100–200 µm) were etched along the length of the thin-slab glass surface, at depths in the range 40–90 µm using an etchant optimized for the alkali-borosilicate composition (Corning Iris Glass). These structures’ aspect ratio (W/H < 3) and pitch effectively controlled the degree of light confinement (local dimming index (LDI) > 80%) along the lenticular corridors. The all-glass and high-transmission nature of the iris glass composition enabled higher brightness operation and yielded negligible color-shift.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Liquid crystal display LCD [1] panels do not emit light, and are therefore dependent on a backlight unit (BLU) positioned behind the LCD panel to provide transmissive light. Light from the BLU illuminates the LCD panel, which then functions as a light valve that selectively allows light to pass through pixels or be blocked, thereby forming a viewable image. An opportunity for improvement in LCD technology is the LCD contrast ratio, which is most noticeable in darker scenes where dark regions look grayish rather than true black. The difficulty in achieving true black is that liquid-crystal (LC) is not an ideal light blocking device, and there will always be light leakage from the backlight panel, even with the LC in full blocking operation.
To overcome this limitation, manufacturers have incorporated active local dimming [2–9] of the image, in which the illumination can be locally dimmed relative to other regions of the display panel, depending on the local brightness of the image being displayed. Such local dimming (two-dimension: 2D) can be relatively easily incorporated when the light source is positioned directly behind the LCD panel, for example a two-dimensional array of LEDs. While 2D local dimming is very difficult to incorporate with an edge-lit BLU, 1D local dimming can be achieved in an edge-lit BLU by introducing microstructures (such as lenticular lens array) on one (or two) surface(s) of the LGP. As shown below in Fig. 1, briefly, injected light is constrained in the LGP by total internal reflection (TIR), and confined in one dimensional corridors by the microstructures. By selectively switching on and off the LEDs at particular locations, 1D local dimming is achieved. The rays are subsequently reflected and refracted by an array of etched or ink printed white spots at the bottom of the LGP. The light emanating from the top surface of the LGP is dispersed by the diffusion sheet that also weakens bright or dark fringes made by the spots. Usually, two cross prism sheets (BEF) are used to collimate and enhance the light transmitted by the diffusion sheet, and one reflection sheet under the LGP reduces the optical loss from the LGP’s bottom surface. A dual brightness enhancing sheet (DBEF) is optionally introduced into the BLU to manage the polarization property of BLU emission light to further enhance the brightness of a LCD panel.
Fig. 1 (Left) Architecture of the conventional edge-lit backlight module in an LCD-LED TV. (Right) Schematic illustration of an LGP structure used to confine edge-lit illumination into one dimensional linear corridors.
1D local dimming allows for various high-end LCD attributes, such as: increased brightness, high dynamic range (contrast), high refresh rates (the 1D dimming prevents us from seeing a tracking of a moving object on the screen), and energy saving. To date, a variety of techniques have been employed enabling the manufacture of the confining micro structures, most commonly in the shape of lenticular lens array [5–9]. While these may assume a variety of different shapes (see Fig. 2 - left side), they all retain the light confinement attribute. This general behavior is illustrated with ray tracing (see Fig. 2 – right). In-coupled light from each LED diverges quickly across an LGP in the absence of lenticular structures since LED output typically exhibits a Lambertian angular distribution. However, when lenticular structures are added to one of the LGP surfaces, as shown in Fig. 2, the light rays will travel in zig-zag paths in both horizontal (LGP plane) and vertical planes along the lenticular direction due to light reflections from the lenticular-featured surfaces. Consequently, most light rays are confined in a narrow zone along the lenticular direction.
Fig. 2 (Left) Lenticular Structures (side view) – Various forms found suitable for 1D light confinement are shown above, with the height of the surface structures (H), their widths (W), and gaps (G) indicated. (Right) Ray Tracing (top view of LGP) - Light propagation in a LGP with lenticular structures are illustrated using a Lambertian ray source.
LGP fabrication by polymeric hot embossing, injection molding or extrusion are among the more common techniques employed to fabricate them due to the low polymer transition temperatures (e.g. PMMA Tg ~125 ± 20 °C) and compliant viscosities that make such processes attractive. Direct shaping of glass is possible too, but significantly more challenging and much higher cost because of the much higher glass transition temperature and viscosities involved. An alternative approach is to laminate plastic lenticular lens array film to one surface of a glass LGP for an edge lit BLU. While attractive, two technological conundrums arise hampering a fuller market acceptance: (a) white light experiences a color shift after propagating through the LGP panel, presumably due to dispersion related to polymer absorption – requiring thin form factors, and (b) reliability delamination failure modes arise with large variations in environment temperature and humidity due to the large plastic-glass CTE-mismatch. In addition, edge lit enables thinner form factors over direct lit with glass enabling thinner form factors than plastic LGPs.
We describe in this paper an approach to forming micro-structures directly in glass for 1D local dimming by screen print mask and etch procedures, to include the procedures, pattern-structures, performance, and prospects of the technology.
2. Experimental
Figure 3 provides a broad overview of the key individual steps comprising the overall screen print “mask and etch” process.
Fig. 3 Overview of “Mask and “Etch” process individual steps - Iris glass substrate was cut-chamfered, cleaned (40kHz ultrasonic agitation 2% Semiclean KG, Linda Yokohama Oils & Fats Industry, Yokohama, Japan), and baked-stored at 200°C to drive off physisorbed interfacial water, prior to screen printing the etch mask. A stainless steel 360 mesh was typically employed, with a string angle of 22°, and 15 µm emulsion thickness. The printing proceeded along the lenticular “grooves” longitudinal direction, and a post-bake thermal cure was promptly applied by inserting the printed substrate into a 140° oven for at least an hour. The cured substrate was either etched in a bath, or with a spray process described in the text. The residual cured screen mask remaining on the substrate was sometimes removed towards the end of the etch procedures, or with a final clean procedure employing the screen mask solvent and acidic detergent.
The key fabrication steps of the “mask and etch” process is shown in the boxed area of Fig. 3. The “baking” step drives off physisorbed interfacial water to promote efficient chemical coupling of screen ink designed for glass HF-etching. While other methods such as oxygen plasma treatment may well apply, we heated our clean glass substrates in an oven held at 200°C for 16 hours, and stored in a desiccator at ~25% relative humidity, prior to screen-printing, to minimize physisorbed water. The etch mask chemistries we used for HF-etch masks were down-selected from numerous candidates to three: CGSN-XG77, ESTS-3000 from Sun Chemical (Parsippany, NJ-USA), and Kiwomask Z 865/2 VP, from the Kiwo Group (Wieslock, Germany). While these HF etch resistant inks were multi component systems containing an organic polymer, dispersants, emulsifiers, crosslinking agents, pigment, antioxidants, solvents, adhesion promoters and an inorganic material, they were principally acrylate resins, epoxy resins, phenolic resins, and polysiloxanes, formulated with varying levels of glass adhesion promoter. The specific dilution of each ink in its’ solvent, to attain a specific rheology, were: CGSN – used as is, ESTS & Kiwo were diluted 5% & 3% in their respective solvents. Ideal rheological behavior was associated with a clean screen-squeegee release that minimized a screen ripping failure mode, no excessive drying, and multiple uses of the same screen before application of the cleaning cycle.
We used stainless steel 360 mesh screens for high resolution printing exclusively from Sefar (Toledo, OH-USA), with a string angle of 22°, 15 µm emulsion thickness, and longitudinal pattern oriented parallel to and along the screen print direction. The side wall smoothness is also sensitively dependent on the interplay between the visco-elastic screen-ink’s rheology during printing through the emulsion-pattern gaps, and the screen opening. The specific shear-thinning forces that manifest is related to the screen’s squeegee velocity (~50 cm/s) and applied pressure (3 psi above ambient), the screen-substrate gap (2 mm), the emulsion thickness (5 µm – 30 µm), screen thickness, and mesh string angle (0° - 30°). The Fig. 4 below is a scanning electron microscopic (SEM) image (100 x) of one of our preferred screen formats used to produce all-glass lenticular LGPs with high LDI values.
Fig. 4 SEM image (100 x) of screen optimized for producing high LDI “smooth” wall lenticular LGPs – The screen used for this etch-mask application employed a stainless steel 360 mesh with 56 µm opening, string angle of 22°, and a 7-15 µm thick emulsion with the 150 µm wide lenticular pattern.
HF- H2SO4 etchant mixtures were preferred over HF-HCl, and HF-HNO3 mixtures for etching Iris glass based on a higher etch rate, and smoother residual post-etch surface (see Table 1). The 10% HF- 20% H2SO4 mixture was chosen for bath “wet” etching due to fast etch rates (1.4 µm/min) attained with a minimal occurrence of “striations” exhibited with the 10% HF – 30% H2SO4 etchant-solution, regardless of how much mixing or agitation was introduced. The striations manifested as a “wavy” “line-like” residue in the etched glass portions. In contrast, these striations were not observed with spray etching (Chemcut Corporation, State College, PA, USA).
Table 1. - Matrix of Different Etchant Baths explored for optimal IRIS glass etch rate study.
3. Results and discussion
We define a local dimming index (LDI) below to quantitatively assess the degree of light confinement achieved for a given sample pattern (see Fig. 5) using the mask and etch procedures described above.
Fig. 5 Schematic illustrating parameters used to calculate the LDI light confinement, zone width and straightness [7–9] from luminance measurement image data. Unless otherwise stated, we exclusively used local dimming zone widths ~150 mm.
The LDI is calculated using the luminance image of an edge injected sample, as shown in Fig. 5 [5]. Standardized measurement conditions included illumination from a thin edge-lit LED array with a reflector sheet back plate taken from a Dell S2440L 24-Inch screen LED-lit monitor, and imaged 600 mm away in the direction normal to light injection with a Konica Minolta DTS 140 spectro-radiometer imaging camera. Since our samples were 300 mm x 300 mm, we measured at 150 mm away in the direction normal to light injection. For a particular nth corridor, the integrated luminance from the An area, Ln, located 250 mm distal from the LED edge-lit injection site, is calculated, along with the integrated luminance adjacent areas, Ln-1and Ln + 1, according to:
Color point shift measurements were similarly measured with the spectral-radiometer camera in the same location the luminance measurements were made. The integrated CIE-y color point was compared with the injection CIE-y values, and the difference reported as CIE-Δy.
A variety of different lenticular shaped samples were prepared exhibiting different degrees of light confinement (see Fig. 6). While the general degree of light confinement scaled almost linearly with the aspect ratio, the lenticular shapes were significantly different, and yet no discernible impact on the LDI was manifest. Other features however materialized due to the lenticular shapes: most noticeably, with the far-field scattering luminance.
Fig. 6 (Top row) Collection of lenticular etched-pattern samples (side view) prepared by the “mask and etch” procedures described in this paper, whose SEM images are laid out, with increasing aspect ratio and height, from left to right. (Bottom row) The corresponding LDI measurement images (top view) and values are laid out beneath its associated lenticular etched-pattern. Samples with LDI values greater than 80% were deemed acceptable for commercial purposes, sufficiently equivalent in behavior with commercially available plastic sheet.
Surface topology of the all-glass lenticular light guide plates (LGP) plays a large role in the overall system efficiency owing to the far-field scattering behavior of optical micron-sized elements [8]. From an efficiency standpoint, the goal would be to map photons directly from one stack to another, with little angular deviation, to reduce stray light loss inefficiencies. The precise surface topology of all-glass lenticular light guide plates (LGP) created by screenprint & etching methods was found to depend strongly on the degree of adhesion of the etch-mask to the substrate. The lenticular topography was found to be adjustable, spanning a range from varying degrees of sinusoidal morphologies to varying degrees of “flat top” morphologies. While interfacial physisorbed water significantly influenced the degree of adhesion during the etching (spray or bath) event, with all ink-chemistries we used, the complexity of the influencing etch-parameters and rich morphologies obtained will be more fully described in subsequent reports. We simply illustrate this phenomena with Fig. 7 below illustrating etch results of similarly screenprinted patterns differing only with the degree of ink adhesion.
Fig. 7 Tunable Topography - From Sinusoidal to “Flat Top” Morphology - (Left column): Lenticular LGP sample was prepared by etching Iris glass sample WS01760 using the an etch mask with no adhesion promotor. (Right Column): Lenticular LGP sample WS01340 was prepared in similar fashion differing only in the use of strong adhesion promotion.
We found that while the lenticular topography here was “adjustable”, spanning a range of morphologies, the light confinement index was not significantly impacted by differing “G”, “W”, and “H” values, if the W/H aspect ratio < 3.
The LDI metric appeared compromised when longitudinal side-wall ‘Scalloping’ having sharp asperities were found. Light confinement using lenticular structures were significantly dependent on the smoothness of the longitudinal lenticular “wall” structures. LDI << 80% metrics (for typical 150 mm zone width) were found to prevail in the presence of side wall “scalloping”, or “waviness”. This is due to confinement photons impinging on local asperities that disrupt the total internal reflection (TIR) condition, manifesting as a loss and driving LDI metrics below 80% confinement. One appreciates this distinction with the information illustrated in Fig. 8 below.
Fig. 8 Light Confinement is Compromised with Longitudinal Side-Wall ‘Scalloping’ with Sharp Asperities – SEM images of two samples with different side wall smoothness were prepared from equivalent regions in the etched lenticular array LGPs, and shown above in the top row. The top left SEM image (100-X) illustrates the relatively smoother-walled sample WS01760 prepared with no adhesion promoter (ESTS-3000), contrasted with the top right SEM image (150-X) of sample WS01072 prepared with an additional adhesion promotion, prior to application of the CGSN etch-mask. Below each SEM image is the related luminance image used to measure the light confinement parameter: LDI. The LDI value for the “smooth” walled sample on the left was ~85% while the LDI value for the “kinked-tube” shaped lenticulars on the right were well below 65%.
The high degree of light confinement for the smooth-wall lenticular manifests as low scattering near the LED injection end at the bottom of luminance image, propagating along the longitudinal lenticular length, and brightening the distal edge, opposite the injection end. This propagation and brightening of the distal edge is not observed with the “kinked tube” lenticular luminance image on the right, and indicative of poor light confinement, low LDI values, and apparent with the large scattering losses piling up near the LED light injection edge, and appearing as a large saturated brightness.
A collection of additional photometric measurements is collated below in the Table 2 summarizing all measurements taken and calculated.
Table 2. – Summary of key technical photometric attributes achieved by “mask and etch” process applied to Iris glass LGP substrates.
Conclusions: All-glass, lenticular lens array, light guide substrates were fabricated in a single mask and etch procedure, enabling local 1D light-confinement from an edge-injected LED array. Lenticular structures of sufficient resolution (100 - 200 µm) were etched along the length of the thin-slab Iris glass surface, at depths in the range 40 - 90 µm using an etchant optimized for the alkali boro-silicate Iris glass composition. These structures’ aspect ratio (W/H < 3) and pitch effectively controlled the degree of light confinement (> 80%) along the lenticular corridors. The all-glass and high-transmission nature of the Iris glass composition yielded high negligible color-shift over 300 mm distance. Control of the lenticular topology was highly dependent on the screen mask-glass interfacial adhesion, their respective chemical composition’s interaction with the acid etchant solution.
4. Summary
All-glass, lenticular lens array, light guide substrates were fabricated in a single mask and etch procedure, enabling local 1D light-confinement operation from an edge-injected LED array. Lenticular structures of sufficient resolution (width & pitch: 100 - 200 µm) were etched along the length of the thin-slab Iris glass surface, at depths in the range 40 - 90 µm using an etchant (10% HF – 20-30% H2SO4) optimized for the alkali boro-silicate Iris glass composition. These structures’ aspect ratio (W/H < 3) and pitch effectively controlled the degree of light confinement (> 80%) along the lenticular corridors. The all-glass and high-transmission nature of the Iris glass composition yielded negligible color-shift over long distance enabling higher brightness operation. Control of the lenticular topology was sensitively dependent on the precise screen mask-glass interfacial adhesion, their respective chemical composition’s interaction with the acid etchant solution, and use of bath-etch versus spray-etch processes. Significant versatility was additionally found with key spray-etch parameters: nozzle geometry, back pressure, temperature, and dwell time.
Acknowledgments
Authors gratefully acknowledge screen printing assistance from David L Baker and Kevin Maslin (MTE), SEM images from Amanda Hall (CS), color shift measurements from Andy Sullivan (AGI), and input from Jim West, Steve Rosenblum, and Nat Venkataraman. We are also thankful to Todd St Clair, Robert Lee, Aleksandra Boskovic and Adam Ellison for their support.
References
1. D. Yi, “Display Light Guide Plate (LGP) Report – 2013,” IHS Electronics & Media market, technology and supply chain analysis, (2013).
2. P. de Greef and H. Groot-Hulze, “39.1 Adaptive dimming and boosting backlight for LCD-TV systems,” Society for Information Display Symposium Digest of Technical Papers: 38(1), 1332–1335 (2007). [CrossRef]
3. S. Cha, T. Choi, H. Lee, and S. Sull, “An Optimized Backlight Local Dimming Algorithm for Edge-Lit LED Backlight LCDs,” J. Disp. Technol. 11(4), 378–385 (2015). [CrossRef]
4. X. Mi, K. Allen, D. Kuksenkov, J. Tokar, A. Sullivan, and S. Rosenblum, “Patterned holey glass LGP based ultra-thin 2D local dimming backlight” Society for Information Display Symposium Digest of Technical Papers 14(1), 145–148 (2018).
5. S. Jung, M. Kim, D. Kim, and J. Lee, “Local Dimming Design and Optimization for Edge-Type LED Backlight Unit,” Society for Information Display Symposium Digest of Technical Papers, pp.1430–1432 (2011). [CrossRef]
6. Y. C. Kim, H. D. Im, M. G. Lee, and H. Y. Choi, “Directivity Enhanced BLU for Edge-type Local Dimming”, Society for Information Display Symposium Digest of Technical Papers, pp.662- 664 (2011). [CrossRef]
7. K. Käläntär, “A directional backlight with narrow angular luminance distribution for widening the viewing angle for an LCD with a front-surface light-scattering film,” J. Soc. Inf. Disp. 20(3), 133 (2012). [CrossRef]
8. T. Teng,and L. Tseng, “Effect of a light guide plate with lenticular-arrayed surface on optical output for backlight and illumination”, Proc. SPIE 8835, LED-based Illumination Systems, 88350H (30 September 2013).
9. M. Isshiki, Y. Arai, M. Inoue, T. Sasaki, and K. Käläntär, “Monolithic Glass Light-Guide Plate with Built-in Prism Structure for 1D Dimming Large-Area LCD,” Society for Information Display Symposium Digest of Technical Papers, pp.149- 152 (2018). [CrossRef]
