Appendix B: Sanyo

Executive Managing Director and Representative Director

Manager, LCD Division, Engineering Development
Department, Section 1

Chief Engineer, Engineering Development Department,
Section 2

MEETING SCHEDULE

We were met Sunday evening at Tottori Airport by Professors Hiroshi Kobayashi and Shosaku Tanaka of Tottori University and were joined later that evening at our hotel by Masuharu Takuma, Tadanobu Yamazawa, and Kenichi Narita. The visit to the Sanyo factory began at 1:00 pm the next day. We first saw a video that introduced Sanyo and its product lines. Sanyo is a large electrical company that produces consumer products ranging from household kitchen appliances to advanced electronic products for office automation and home use. The Tottori facility, established in 1966, employs 3000 people and has annual sales of 12 billion yen.

THE FACTORY TOUR

The STN LCD factory that we toured is less than 1 year old. Two years of planning were required prior to construction, and construction took 1 year to complete. The first STN flat- panel display produced on the manufacturing line was assembled in August of 1991, approximately 6 months after the completion of the facility. At another site, Sanyo has an equivalent facility producing STN LCDs. A second factory at the Tottori site is in the planning phase, with display production to begin in 1993. This facility will employ MIM technology to manufacture active matrix LCDs.

The factory we toured contains 11,000 square meters of floor space. The manufacturing line is approximately 300 meters long. The majority of the line is housed in two large clean rooms, which were subdivided into smaller rooms of varying clean room classifications. Two displays are constructed on a single 300 mm x 400mm x 0.7 mm sheet of Nippon Electric Glass boric silicate glass. Two of these sheets are sandwiched to form two complete STN LCDs.

One of the two sheets of glass that form a pair of displays is coated with ITO prior to the first station of processing. It was not clear whether or not this sheet was delivered to the factory with the ITO coating or whether ITO was applied at the Tottori site. All coatings for the complementary sheet, the sheet containing the color filters, are applied at the Tottori site. The colored filters are applied to the glass in the new building. First a polymer gel is spin-coated onto the glass. This is patterned by a UV photo process. The gel is dyed red, green, or blue and then baked. The process is repeated three more times to add the two remaining colors and, last, the black matrix mask. Next, a leveling layer is spin-coated onto the glass with filters. Finally the ITO layer is applied. The sheet is then ready for the processing steps required to pattern the ITO. The dye materials and suppliers are proprietary details of the manufacturing process.

Cassettes hold 30 sheets of glass each as they proceed down the manufacturing line. The line is almost completely automated, and the cassettes are transported automatically without human intervention from station to station throughout most of the process.

Figure Sanyo.1 shows the approximate layout of the two buildings that house the STN LCD factory. The numbers correspond to the locations of various stations or phases of the manufacturing process. The first 11 stations of the manufacturing process are all contained in the new building. Station 1 is for cleaning the sheets of glass. Before the glass sheets arrive at this station, the ITO and/or colored filters have been applied. There are 10 cleaning steps, with a through-put time of 5 seconds per plate. At the second station a photo-resist coating is applied by a roller coater manufactured by DaiNippon Screen. There is an in-line bake of 10 minutes. Three proximity printers expose the address line pattern at station 3. This step takes approximately 5-10 seconds per sheet. Development, etching, and stripping take place at stations 4, 5, and 6. The alignment layer is applied at station 7. Three lines are used to apply, bake, and rub the alignment layer. Six rubbing machines, of a design unique to Sanyo, are at station 8.

The adhesive is applied to the two processed sheets of glass (one with colored filters and one without) at station 9. The adhesives are manufactured by Mitsui Toatsu Chemical. The two sheets are aligned and laminated at the next station (station 10). The spacers are sprayed on with water prior to the lamination. The cell gap is 6 microns, with a tolerance of 3%. The alignment is performed manually by an operator, but the process will be automated at a future time. Finally, the laminated sheets are cured for one hour in UV light at station 11.

The laminated plates are taken to the second building, where they are scribed on one of two manually operated Mitsubishi diamond scribe machines at station 12. In the same area, the separation is performed by a person who breaks the two displays apart by hand. At station 13 each display is filled with liquid crystal material and sealed. Eight dual-chamber vacuum machines are used for the filling operation. The displays are filled through three holes in the adhesive layer, which are sealed and cured after filling. The final station, station 14 in figure Sanyo.1, is for inspection and packaging. The application of polarizers, retardation films, and IC drivers is done at another site 50 km from Tottori. The polarizers and retardation films are produced by Nitto Denko and Sumitomo Chemical.

The factory is operated 24 hours per day. A crew of 60 people is required to operate the facility during a single shift. We were not given data on the frequency of maintenance, cleaning, or breakdown of the factory. This factory produces approximately 150,000 displays per month.

DISCUSSION FOLLOWING TOUR

We were given a set of answers to the questions we had sent to the Sanyo team prior to our arrival. The questions and answers are included with this report. Sanyo believes that its product line in liquid crystal displays will include black-and-white STN devices and a future color MIM active matrix display device. Sanyo sees its greatest challenges as improving the contrast of both types of liquid crystal displays (STN and MIM) and improving the speed of passive STN display devices. Current problems with MIM development include unevenness in the pixel elements and a memory effect that is associated with a threshold shift in the drivers.

MIM technology was chosen for the active matrix display because it is believed to produce better manufacturing yields than TFT devices and because it is believed to be easier to adapt to FLC and PDLC materials.

Approximately 75% of the displays Sanyo produces are sold as component technology. Sanyo uses the remaining 25% in its own consumer product lines. They have not considered building displays for the aerospace industry because they see this as a low-volume, highly specialized market. Moreover, they have very little contact with this industry. Their primary market is the high-volume consumer product marketplace.

GENERAL COMMENTS

The Sanyo factory was very impressive. The staff engineers we met with were very well informed about developments in the industry. The required machine tool equipment was clearly available. Some of the factory equipment was standard, and some was specially designed for this particular factory. It is clear that Japan has a deep infrastructure of machine tool manufacturers and component technology suppliers. The engineering staff at Tottori Sanyo clearly spans a large age range; thus a wide range of experience and expertise in research, development, and manufacturing has been tapped to develop this maturing industry. The human resources required for this industry are in place and growing in Japan.

GENERAL QUESTIONS BY THE TEAM

Liquid Crystal Displays

What size panels have been fabricated, in color and in monochrome?

16.7" (1120 x 780 dots) for monochrome and 8.5" (VGA color) for simple multiplexed STN.

How can optical efficiency of panels be improved?

Larger twist angle and optimized liquid can improve it.

What method do you expect will improve off-axis contrast ratio?

Retarder film will improve it by making a z-axis retardance.

What interconnect technology do you use today? How do you plan to approach interconnects in 1995? In 2000?

Now we use zebra connectors, heat-pressing type conductive films and anisotropic conductive adhesive films. We don't know about the technology in the future.

What are prospects for fast-response LCDs (<100 ms) at temperatures <0 degrees centigrade, < -20 degrees centigrade?

What, if any, R&D is being done on new LC modes to improve optical performance (viewing angle, transmission, efficiency, contrast ratio, response time)?

AMLCD+SmC system or AMLCD+PDLC is being studied.

What are the prospects for low-voltage polymer-dispersed LC(PDLC) with high contrast ratio?

Will PDLCs ever be acceptable as direct-view multiplexed LCDs? If so, when and at what level of multiplexing?

Are PDLCs useful for projection systems? Will they replace twisted nematic LCs? If so, when?

What are the prospects for reducing the number of masking steps? Is a 2- or 3- mask process commercially viable?

Are integral drivers commercially feasible? If so, when? What is cost comparison to external drivers?

Now we use fine-pitch and narrow-width TAB ICs.

Ex.: (A) 0.14 (pitch) x 160 (outs) x 11 (widths)

What techniques can be used for driver attachment for high-resolution LCDs (>10 pixel/m)?

Now the level of 5-6 pixels/mm is obtained by anisotropic adhesion techniques.

What are the limitations of MIMs in resolution and display size?

We don't know. Now we are improving the I-V characteristics rapidly.

STN LCDs

What are the size and resolution limits for STN LCD?

What are the prospects for 50 mms response times with >15:1 CR?

What techniques can be used to improve viewing angle?

It can be improved by giving a retardance to z-axis of compensation film.

What are the prospects for wide temp range (-40 to 70 degrees centigrade) with usable CR(>5:1)?

As one of measures, some scanning lines are driven from both sides of the LCD panel. This can improve contrast, too.

What techniques are under development to reduce costs?

To improve the quality of manufacturing facilities and enhance the yield is the first thing to do.

When will STN be replaced by AMLCD, if at all?

We don't think STN market will be replaced by AMLCD.

What are prospects for color STN LCD with >15:1 CR? What response time is feasible?

Now we can offer 350 ms (T(subscript r) + T(subscript f)).

What breakthroughs are needed to advance the performance of STN LCD?

Ferroelectric LCDs

The Nikkei Sangyo newspaper of October 1 reported that Hitachi and Takeda Pharmacy had developed a 3.68-million pixel, 3.3" screen with response <200 ms/picture.

What are prospects for increasing temperature range to 0-50 degrees centigrade? To -40-70 degrees centigrade?

We don't know the recent technological trend about FE LCDs.

Backlights

What backlight technology do you see as most suitable for AMLCD laptops? Portable TV on the wall?

What novel research in backlight technology do you see as most promising? For example, RF fluorescent, electroluminescent, color sequential cold cathode, etc.

What are the parameters, brightness, lumens per watt, power consumption, spectral distribution, size, thickness, lifetime today? In the year 1995? In 2000? What are the limiting issues? What R&D is conducted to address these issues?

LC Materials

What is the most critical LC material problem from the perspective of AM design and processing?

High-level resistivity and electrochemical stability are the most important.

What level of interest is there in non-STN or non-TN materials? For example, VAN, PDLC, SmC, ferroelectric?

What materials cause the most limitations in your display products?

We consider that defects in the alignment layer limit the yield.

What are the technical tradeoffs of the major approaches to color filters for LD, and what are the new developments?

We are now trying to get the pigment-dispersed photo resist for easy mass production.

Alignment Layers

What qualities do you believe the alignment layer should have to reduce the effects of ions in the display (like image sticking)?

Is there a need for low-resistivity alignment layers because of the ion problem? If so, how are plate-to-plate shorts avoided?

What is the best pre-tilt angle for STN devices? How can this high pre-tilt be achieved?

It can be achieved by designing special molecular structures of alignment polymer.

What characteristics of alignment layers are desirable for new devices such as VAN and SmC? What research will lead to these characteristics?

ITO Layers

What resistivity will be required for future passive matrix displays? For video rate STN and SmC devices?

What can be done to minimize the plate-to-plate shorting?

It can be minimized by decreasing particles in every facility.

ATTENDEES